diff --git "a/pretrain-checkpoint-10388/trainer_state.json" "b/pretrain-checkpoint-10388/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/pretrain-checkpoint-10388/trainer_state.json"
@@ -0,0 +1,135110 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9997593666530307,
+  "eval_steps": 500,
+  "global_step": 10388,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "lm_loss": 9.0891,
+      "step": 0,
+      "vm_loss": 0.2726
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 9.8504,
+      "step": 0,
+      "vm_loss": 0.2817
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 9.959,
+      "step": 0,
+      "vm_loss": 0.3381
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 10.3082,
+      "step": 0,
+      "vm_loss": 0.3483
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 10.4086,
+      "step": 0,
+      "vm_loss": 0.3033
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 10.3127,
+      "step": 0,
+      "vm_loss": 0.3025
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 9.9909,
+      "step": 0,
+      "vm_loss": 0.2222
+    },
+    {
+      "epoch": 0,
+      "lm_loss": 10.4709,
+      "step": 0,
+      "vm_loss": 0.2834
+    },
+    {
+      "epoch": 0.0001925066775753784,
+      "grad_norm": 67.56311716151845,
+      "learning_rate": 6.41025641025641e-08,
+      "loss": 10.4004,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003850133551507568,
+      "grad_norm": 67.54056470404237,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 10.3209,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005775200327261352,
+      "grad_norm": 68.11791668225818,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 10.3976,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007700267103015136,
+      "grad_norm": 67.99840906336297,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 10.458,
+      "step": 4
+    },
+    {
+      "epoch": 0.000962533387876892,
+      "grad_norm": 67.95113424782106,
+      "learning_rate": 3.205128205128205e-07,
+      "loss": 10.4427,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011550400654522703,
+      "grad_norm": 67.5200481108944,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 10.3632,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013475467430276488,
+      "grad_norm": 67.25707550005988,
+      "learning_rate": 4.4871794871794876e-07,
+      "loss": 10.439,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "grad_norm": 67.13059075776133,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 10.3051,
+      "step": 8
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 9.8455,
+      "step": 8,
+      "vm_loss": 0.2447
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 10.1427,
+      "step": 8,
+      "vm_loss": 0.2629
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 9.5623,
+      "step": 8,
+      "vm_loss": 0.2549
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 9.9834,
+      "step": 8,
+      "vm_loss": 0.3022
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 9.5318,
+      "step": 8,
+      "vm_loss": 0.2696
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 10.0009,
+      "step": 8,
+      "vm_loss": 0.2668
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 10.0723,
+      "step": 8,
+      "vm_loss": 0.2553
+    },
+    {
+      "epoch": 0.0015400534206030272,
+      "lm_loss": 10.1124,
+      "step": 8,
+      "vm_loss": 0.3022
+    },
+    {
+      "epoch": 0.0017325600981784055,
+      "grad_norm": 66.30373484529393,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 10.3016,
+      "step": 9
+    },
+    {
+      "epoch": 0.001925066775753784,
+      "grad_norm": 62.07838358943419,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 10.2373,
+      "step": 10
+    },
+    {
+      "epoch": 0.002117573453329162,
+      "grad_norm": 61.37073291647445,
+      "learning_rate": 7.051282051282052e-07,
+      "loss": 10.1884,
+      "step": 11
+    },
+    {
+      "epoch": 0.0023100801309045406,
+      "grad_norm": 61.400977655344526,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 10.0603,
+      "step": 12
+    },
+    {
+      "epoch": 0.002502586808479919,
+      "grad_norm": 50.182015447034196,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 9.7238,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026950934860552975,
+      "grad_norm": 49.97607872876963,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 9.7353,
+      "step": 14
+    },
+    {
+      "epoch": 0.002887600163630676,
+      "grad_norm": 48.99447824537502,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 9.6302,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "grad_norm": 48.10312448024197,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 9.6281,
+      "step": 16
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.0049,
+      "step": 16,
+      "vm_loss": 0.273
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.1357,
+      "step": 16,
+      "vm_loss": 0.2863
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 8.6543,
+      "step": 16,
+      "vm_loss": 0.2357
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.2545,
+      "step": 16,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.4352,
+      "step": 16,
+      "vm_loss": 0.2756
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.181,
+      "step": 16,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 9.7709,
+      "step": 16,
+      "vm_loss": 0.3593
+    },
+    {
+      "epoch": 0.0030801068412060545,
+      "lm_loss": 8.8362,
+      "step": 16,
+      "vm_loss": 0.2298
+    },
+    {
+      "epoch": 0.003272613518781433,
+      "grad_norm": 45.12580958599678,
+      "learning_rate": 1.0897435897435899e-06,
+      "loss": 9.3547,
+      "step": 17
+    },
+    {
+      "epoch": 0.003465120196356811,
+      "grad_norm": 52.454030929150434,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 8.9004,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036576268739321894,
+      "grad_norm": 49.37823434619467,
+      "learning_rate": 1.217948717948718e-06,
+      "loss": 8.885,
+      "step": 19
+    },
+    {
+      "epoch": 0.003850133551507568,
+      "grad_norm": 44.49325642467285,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 8.7918,
+      "step": 20
+    },
+    {
+      "epoch": 0.004042640229082946,
+      "grad_norm": 39.66645486846694,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 8.7024,
+      "step": 21
+    },
+    {
+      "epoch": 0.004235146906658324,
+      "grad_norm": 38.43222276018352,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 8.6347,
+      "step": 22
+    },
+    {
+      "epoch": 0.004427653584233703,
+      "grad_norm": 37.16065706647283,
+      "learning_rate": 1.4743589743589745e-06,
+      "loss": 8.4031,
+      "step": 23
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "grad_norm": 40.34442755793192,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 8.0557,
+      "step": 24
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.2879,
+      "step": 24,
+      "vm_loss": 0.2595
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.6324,
+      "step": 24,
+      "vm_loss": 0.2813
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.568,
+      "step": 24,
+      "vm_loss": 0.2289
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.6291,
+      "step": 24,
+      "vm_loss": 0.2515
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.6707,
+      "step": 24,
+      "vm_loss": 0.2657
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.5318,
+      "step": 24,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.5985,
+      "step": 24,
+      "vm_loss": 0.2884
+    },
+    {
+      "epoch": 0.004620160261809081,
+      "lm_loss": 7.6362,
+      "step": 24,
+      "vm_loss": 0.3132
+    },
+    {
+      "epoch": 0.00481266693938446,
+      "grad_norm": 39.60870239983037,
+      "learning_rate": 1.602564102564103e-06,
+      "loss": 7.8232,
+      "step": 25
+    },
+    {
+      "epoch": 0.005005173616959838,
+      "grad_norm": 39.4982593276097,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 7.6741,
+      "step": 26
+    },
+    {
+      "epoch": 0.005197680294535217,
+      "grad_norm": 34.99435336916321,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 7.4862,
+      "step": 27
+    },
+    {
+      "epoch": 0.005390186972110595,
+      "grad_norm": 30.470687063139007,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 7.416,
+      "step": 28
+    },
+    {
+      "epoch": 0.005582693649685973,
+      "grad_norm": 34.83143074077658,
+      "learning_rate": 1.8589743589743592e-06,
+      "loss": 7.4327,
+      "step": 29
+    },
+    {
+      "epoch": 0.005775200327261352,
+      "grad_norm": 36.79708928645268,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 7.2392,
+      "step": 30
+    },
+    {
+      "epoch": 0.00596770700483673,
+      "grad_norm": 35.409654561711974,
+      "learning_rate": 1.987179487179487e-06,
+      "loss": 7.0838,
+      "step": 31
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "grad_norm": 28.8551998968067,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 6.8512,
+      "step": 32
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 5.9191,
+      "step": 32,
+      "vm_loss": 0.2617
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.5199,
+      "step": 32,
+      "vm_loss": 0.2488
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.9769,
+      "step": 32,
+      "vm_loss": 0.305
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.4268,
+      "step": 32,
+      "vm_loss": 0.3273
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.5986,
+      "step": 32,
+      "vm_loss": 0.2744
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.6533,
+      "step": 32,
+      "vm_loss": 0.2437
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.4629,
+      "step": 32,
+      "vm_loss": 0.2509
+    },
+    {
+      "epoch": 0.006160213682412109,
+      "lm_loss": 6.1676,
+      "step": 32,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 0.006352720359987487,
+      "grad_norm": 31.308951991177533,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 6.7775,
+      "step": 33
+    },
+    {
+      "epoch": 0.006545227037562866,
+      "grad_norm": 39.603915694655704,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 6.6871,
+      "step": 34
+    },
+    {
+      "epoch": 0.006737733715138244,
+      "grad_norm": 40.0007504045018,
+      "learning_rate": 2.243589743589744e-06,
+      "loss": 6.5464,
+      "step": 35
+    },
+    {
+      "epoch": 0.006930240392713622,
+      "grad_norm": 33.64736860797609,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 6.3299,
+      "step": 36
+    },
+    {
+      "epoch": 0.007122747070289001,
+      "grad_norm": 27.527054295587234,
+      "learning_rate": 2.371794871794872e-06,
+      "loss": 6.1941,
+      "step": 37
+    },
+    {
+      "epoch": 0.007315253747864379,
+      "grad_norm": 26.868103732745308,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 6.0506,
+      "step": 38
+    },
+    {
+      "epoch": 0.007507760425439758,
+      "grad_norm": 34.14811674624274,
+      "learning_rate": 2.5e-06,
+      "loss": 5.9916,
+      "step": 39
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "grad_norm": 30.15744905014044,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 5.8761,
+      "step": 40
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.2691,
+      "step": 40,
+      "vm_loss": 0.2279
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.8057,
+      "step": 40,
+      "vm_loss": 0.2381
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.9513,
+      "step": 40,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.3976,
+      "step": 40,
+      "vm_loss": 0.2235
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.6982,
+      "step": 40,
+      "vm_loss": 0.2372
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 6.0874,
+      "step": 40,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.495,
+      "step": 40,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.007700267103015136,
+      "lm_loss": 5.3463,
+      "step": 40,
+      "vm_loss": 0.2759
+    },
+    {
+      "epoch": 0.007892773780590515,
+      "grad_norm": 24.27723895971218,
+      "learning_rate": 2.6282051282051286e-06,
+      "loss": 5.8348,
+      "step": 41
+    },
+    {
+      "epoch": 0.008085280458165893,
+      "grad_norm": 24.93716131037821,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 5.8076,
+      "step": 42
+    },
+    {
+      "epoch": 0.00827778713574127,
+      "grad_norm": 25.96480278691456,
+      "learning_rate": 2.756410256410257e-06,
+      "loss": 5.5663,
+      "step": 43
+    },
+    {
+      "epoch": 0.008470293813316649,
+      "grad_norm": 22.172212733444184,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 5.4675,
+      "step": 44
+    },
+    {
+      "epoch": 0.008662800490892028,
+      "grad_norm": 20.513481238836366,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 5.4229,
+      "step": 45
+    },
+    {
+      "epoch": 0.008855307168467406,
+      "grad_norm": 21.502472356346956,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 5.3116,
+      "step": 46
+    },
+    {
+      "epoch": 0.009047813846042784,
+      "grad_norm": 19.880574155344647,
+      "learning_rate": 3.012820512820513e-06,
+      "loss": 5.3022,
+      "step": 47
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "grad_norm": 17.837237753145605,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 5.1557,
+      "step": 48
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 5.2803,
+      "step": 48,
+      "vm_loss": 0.2251
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.4979,
+      "step": 48,
+      "vm_loss": 0.2392
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.9999,
+      "step": 48,
+      "vm_loss": 0.2557
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.6382,
+      "step": 48,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.3124,
+      "step": 48,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.6165,
+      "step": 48,
+      "vm_loss": 0.2116
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 4.7613,
+      "step": 48,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 0.009240320523618163,
+      "lm_loss": 5.1469,
+      "step": 48,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.009432827201193542,
+      "grad_norm": 19.125369584951166,
+      "learning_rate": 3.141025641025641e-06,
+      "loss": 5.1127,
+      "step": 49
+    },
+    {
+      "epoch": 0.00962533387876892,
+      "grad_norm": 18.88008556622611,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 4.994,
+      "step": 50
+    },
+    {
+      "epoch": 0.009817840556344298,
+      "grad_norm": 17.793665649291906,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 4.8572,
+      "step": 51
+    },
+    {
+      "epoch": 0.010010347233919676,
+      "grad_norm": 15.95988062195999,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 4.8681,
+      "step": 52
+    },
+    {
+      "epoch": 0.010202853911495054,
+      "grad_norm": 14.079401638833641,
+      "learning_rate": 3.397435897435898e-06,
+      "loss": 4.7364,
+      "step": 53
+    },
+    {
+      "epoch": 0.010395360589070434,
+      "grad_norm": 15.252274114094016,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 4.6981,
+      "step": 54
+    },
+    {
+      "epoch": 0.010587867266645812,
+      "grad_norm": 14.969515784687696,
+      "learning_rate": 3.5256410256410263e-06,
+      "loss": 4.5873,
+      "step": 55
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "grad_norm": 14.594612860831406,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 4.6019,
+      "step": 56
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.0126,
+      "step": 56,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.126,
+      "step": 56,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.2023,
+      "step": 56,
+      "vm_loss": 0.2398
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.4627,
+      "step": 56,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.3084,
+      "step": 56,
+      "vm_loss": 0.2611
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.1678,
+      "step": 56,
+      "vm_loss": 0.2361
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.646,
+      "step": 56,
+      "vm_loss": 0.2391
+    },
+    {
+      "epoch": 0.01078037394422119,
+      "lm_loss": 4.7998,
+      "step": 56,
+      "vm_loss": 0.3156
+    },
+    {
+      "epoch": 0.010972880621796568,
+      "grad_norm": 12.555763755713865,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 4.5263,
+      "step": 57
+    },
+    {
+      "epoch": 0.011165387299371946,
+      "grad_norm": 13.506694461540164,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 4.3911,
+      "step": 58
+    },
+    {
+      "epoch": 0.011357893976947326,
+      "grad_norm": 13.69467907982829,
+      "learning_rate": 3.782051282051282e-06,
+      "loss": 4.4337,
+      "step": 59
+    },
+    {
+      "epoch": 0.011550400654522704,
+      "grad_norm": 11.933853505007217,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 4.3236,
+      "step": 60
+    },
+    {
+      "epoch": 0.011742907332098082,
+      "grad_norm": 11.05845132436078,
+      "learning_rate": 3.910256410256411e-06,
+      "loss": 4.336,
+      "step": 61
+    },
+    {
+      "epoch": 0.01193541400967346,
+      "grad_norm": 12.193715903046733,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 4.3291,
+      "step": 62
+    },
+    {
+      "epoch": 0.01212792068724884,
+      "grad_norm": 11.092158638717908,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 4.1942,
+      "step": 63
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "grad_norm": 10.453586342361072,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 4.2205,
+      "step": 64
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 4.2783,
+      "step": 64,
+      "vm_loss": 0.2464
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 4.4476,
+      "step": 64,
+      "vm_loss": 0.2418
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 4.248,
+      "step": 64,
+      "vm_loss": 0.2461
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 3.9782,
+      "step": 64,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 3.9591,
+      "step": 64,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 4.1235,
+      "step": 64,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 3.9965,
+      "step": 64,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.012320427364824218,
+      "lm_loss": 3.8855,
+      "step": 64,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 0.012512934042399596,
+      "grad_norm": 10.571151275524029,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 4.1813,
+      "step": 65
+    },
+    {
+      "epoch": 0.012705440719974974,
+      "grad_norm": 10.332021130036583,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 4.147,
+      "step": 66
+    },
+    {
+      "epoch": 0.012897947397550352,
+      "grad_norm": 9.078977356392668,
+      "learning_rate": 4.294871794871795e-06,
+      "loss": 4.1122,
+      "step": 67
+    },
+    {
+      "epoch": 0.013090454075125732,
+      "grad_norm": 9.249462822436183,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 4.0811,
+      "step": 68
+    },
+    {
+      "epoch": 0.01328296075270111,
+      "grad_norm": 10.233531959350929,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 4.0307,
+      "step": 69
+    },
+    {
+      "epoch": 0.013475467430276488,
+      "grad_norm": 9.487386336050799,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 3.9154,
+      "step": 70
+    },
+    {
+      "epoch": 0.013667974107851866,
+      "grad_norm": 9.328143981087013,
+      "learning_rate": 4.551282051282052e-06,
+      "loss": 3.8843,
+      "step": 71
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "grad_norm": 8.526946206961572,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 3.894,
+      "step": 72
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.9402,
+      "step": 72,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.6234,
+      "step": 72,
+      "vm_loss": 0.2238
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.8825,
+      "step": 72,
+      "vm_loss": 0.2857
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 4.1888,
+      "step": 72,
+      "vm_loss": 0.2127
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.8483,
+      "step": 72,
+      "vm_loss": 0.2016
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.585,
+      "step": 72,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.6869,
+      "step": 72,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.013860480785427244,
+      "lm_loss": 3.7044,
+      "step": 72,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 0.014052987463002624,
+      "grad_norm": 8.976446161591616,
+      "learning_rate": 4.6794871794871795e-06,
+      "loss": 3.9352,
+      "step": 73
+    },
+    {
+      "epoch": 0.014245494140578002,
+      "grad_norm": 8.57017614217942,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 3.8516,
+      "step": 74
+    },
+    {
+      "epoch": 0.01443800081815338,
+      "grad_norm": 8.143374976367387,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 3.846,
+      "step": 75
+    },
+    {
+      "epoch": 0.014630507495728758,
+      "grad_norm": 8.268330580017869,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 3.7582,
+      "step": 76
+    },
+    {
+      "epoch": 0.014823014173304137,
+      "grad_norm": 7.974463043405866,
+      "learning_rate": 4.935897435897436e-06,
+      "loss": 3.8047,
+      "step": 77
+    },
+    {
+      "epoch": 0.015015520850879515,
+      "grad_norm": 7.243983118995253,
+      "learning_rate": 5e-06,
+      "loss": 3.7461,
+      "step": 78
+    },
+    {
+      "epoch": 0.015208027528454893,
+      "grad_norm": 6.784533296038897,
+      "learning_rate": 5.064102564102565e-06,
+      "loss": 3.7125,
+      "step": 79
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "grad_norm": 7.5594748816071675,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 3.6808,
+      "step": 80
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.1976,
+      "step": 80,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.5419,
+      "step": 80,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.2571,
+      "step": 80,
+      "vm_loss": 0.2571
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.3721,
+      "step": 80,
+      "vm_loss": 0.2068
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.3478,
+      "step": 80,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 2.9339,
+      "step": 80,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.8199,
+      "step": 80,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.015400534206030271,
+      "lm_loss": 3.6178,
+      "step": 80,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.01559304088360565,
+      "grad_norm": 6.465158031015808,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 3.6744,
+      "step": 81
+    },
+    {
+      "epoch": 0.01578554756118103,
+      "grad_norm": 29.311170712167385,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 3.6273,
+      "step": 82
+    },
+    {
+      "epoch": 0.015978054238756405,
+      "grad_norm": 7.604136704214489,
+      "learning_rate": 5.320512820512821e-06,
+      "loss": 3.6615,
+      "step": 83
+    },
+    {
+      "epoch": 0.016170560916331785,
+      "grad_norm": 7.221914551372844,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 3.6641,
+      "step": 84
+    },
+    {
+      "epoch": 0.016363067593907165,
+      "grad_norm": 6.126998492930213,
+      "learning_rate": 5.448717948717949e-06,
+      "loss": 3.5409,
+      "step": 85
+    },
+    {
+      "epoch": 0.01655557427148254,
+      "grad_norm": 6.313025278980163,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 3.5991,
+      "step": 86
+    },
+    {
+      "epoch": 0.01674808094905792,
+      "grad_norm": 5.966017358027714,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 3.5317,
+      "step": 87
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "grad_norm": 5.6035146911464855,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 3.5677,
+      "step": 88
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.6411,
+      "step": 88,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.3762,
+      "step": 88,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.2421,
+      "step": 88,
+      "vm_loss": 0.3299
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.1934,
+      "step": 88,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 2.9709,
+      "step": 88,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.2983,
+      "step": 88,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.6268,
+      "step": 88,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 0.016940587626633297,
+      "lm_loss": 3.1539,
+      "step": 88,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 0.017133094304208677,
+      "grad_norm": 7.174914137397918,
+      "learning_rate": 5.705128205128206e-06,
+      "loss": 3.525,
+      "step": 89
+    },
+    {
+      "epoch": 0.017325600981784057,
+      "grad_norm": 6.813223499274678,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 3.4842,
+      "step": 90
+    },
+    {
+      "epoch": 0.017518107659359433,
+      "grad_norm": 6.679518010719231,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 3.4315,
+      "step": 91
+    },
+    {
+      "epoch": 0.017710614336934813,
+      "grad_norm": 6.572097255682014,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 3.4732,
+      "step": 92
+    },
+    {
+      "epoch": 0.01790312101451019,
+      "grad_norm": 6.183997770654562,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 3.5838,
+      "step": 93
+    },
+    {
+      "epoch": 0.01809562769208557,
+      "grad_norm": 14.567870123394162,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 3.4775,
+      "step": 94
+    },
+    {
+      "epoch": 0.01828813436966095,
+      "grad_norm": 10.831077977064988,
+      "learning_rate": 6.08974358974359e-06,
+      "loss": 3.4321,
+      "step": 95
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "grad_norm": 6.479427660592102,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 3.3915,
+      "step": 96
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.0723,
+      "step": 96,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.2662,
+      "step": 96,
+      "vm_loss": 0.185
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.3414,
+      "step": 96,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.2811,
+      "step": 96,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.3468,
+      "step": 96,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.3057,
+      "step": 96,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.3988,
+      "step": 96,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.018480641047236325,
+      "lm_loss": 3.2662,
+      "step": 96,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 0.018673147724811705,
+      "grad_norm": 5.374474299425232,
+      "learning_rate": 6.217948717948718e-06,
+      "loss": 3.4264,
+      "step": 97
+    },
+    {
+      "epoch": 0.018865654402387085,
+      "grad_norm": 5.717364333879507,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 3.4044,
+      "step": 98
+    },
+    {
+      "epoch": 0.01905816107996246,
+      "grad_norm": 6.112177669084539,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 3.4472,
+      "step": 99
+    },
+    {
+      "epoch": 0.01925066775753784,
+      "grad_norm": 4.467719887841041,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 3.4059,
+      "step": 100
+    },
+    {
+      "epoch": 0.019443174435113217,
+      "grad_norm": 8.25017683762081,
+      "learning_rate": 6.474358974358975e-06,
+      "loss": 3.4254,
+      "step": 101
+    },
+    {
+      "epoch": 0.019635681112688597,
+      "grad_norm": 5.735595100137572,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 3.3524,
+      "step": 102
+    },
+    {
+      "epoch": 0.019828187790263976,
+      "grad_norm": 6.23470405055773,
+      "learning_rate": 6.602564102564103e-06,
+      "loss": 3.3708,
+      "step": 103
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "grad_norm": 6.098411804215811,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 3.358,
+      "step": 104
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.278,
+      "step": 104,
+      "vm_loss": 0.259
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.0134,
+      "step": 104,
+      "vm_loss": 0.2244
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.0034,
+      "step": 104,
+      "vm_loss": 0.2757
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 2.8226,
+      "step": 104,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.3339,
+      "step": 104,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.5293,
+      "step": 104,
+      "vm_loss": 0.2031
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.0764,
+      "step": 104,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.020020694467839353,
+      "lm_loss": 3.2529,
+      "step": 104,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 0.020213201145414732,
+      "grad_norm": 6.136288532192748,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 3.3935,
+      "step": 105
+    },
+    {
+      "epoch": 0.02040570782299011,
+      "grad_norm": 6.163567089968353,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 3.3032,
+      "step": 106
+    },
+    {
+      "epoch": 0.02059821450056549,
+      "grad_norm": 7.695161522115114,
+      "learning_rate": 6.858974358974359e-06,
+      "loss": 3.391,
+      "step": 107
+    },
+    {
+      "epoch": 0.020790721178140868,
+      "grad_norm": 8.284810986761684,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 3.3819,
+      "step": 108
+    },
+    {
+      "epoch": 0.020983227855716245,
+      "grad_norm": 6.829673110076974,
+      "learning_rate": 6.9871794871794876e-06,
+      "loss": 3.3247,
+      "step": 109
+    },
+    {
+      "epoch": 0.021175734533291624,
+      "grad_norm": 6.10533286566713,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 3.3494,
+      "step": 110
+    },
+    {
+      "epoch": 0.021368241210867,
+      "grad_norm": 10.155164642203296,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 3.3479,
+      "step": 111
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "grad_norm": 9.492590327271007,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 3.3817,
+      "step": 112
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 2.8807,
+      "step": 112,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.1213,
+      "step": 112,
+      "vm_loss": 0.0985
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.221,
+      "step": 112,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.2356,
+      "step": 112,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.0598,
+      "step": 112,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 2.9602,
+      "step": 112,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.2267,
+      "step": 112,
+      "vm_loss": 0.2017
+    },
+    {
+      "epoch": 0.02156074788844238,
+      "lm_loss": 3.4341,
+      "step": 112,
+      "vm_loss": 0.2666
+    },
+    {
+      "epoch": 0.02175325456601776,
+      "grad_norm": 7.027929270068158,
+      "learning_rate": 7.243589743589744e-06,
+      "loss": 3.3258,
+      "step": 113
+    },
+    {
+      "epoch": 0.021945761243593136,
+      "grad_norm": 8.43797699584041,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 3.3145,
+      "step": 114
+    },
+    {
+      "epoch": 0.022138267921168516,
+      "grad_norm": 8.73386040218349,
+      "learning_rate": 7.371794871794873e-06,
+      "loss": 3.2792,
+      "step": 115
+    },
+    {
+      "epoch": 0.022330774598743892,
+      "grad_norm": 9.842525194493287,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 3.3344,
+      "step": 116
+    },
+    {
+      "epoch": 0.022523281276319272,
+      "grad_norm": 6.569715725981174,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 3.3751,
+      "step": 117
+    },
+    {
+      "epoch": 0.022715787953894652,
+      "grad_norm": 7.709524952064772,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 3.3565,
+      "step": 118
+    },
+    {
+      "epoch": 0.022908294631470028,
+      "grad_norm": 4.243711290650009,
+      "learning_rate": 7.6282051282051286e-06,
+      "loss": 3.2749,
+      "step": 119
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "grad_norm": 13.538180070203053,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 3.2191,
+      "step": 120
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 3.0914,
+      "step": 120,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 3.1108,
+      "step": 120,
+      "vm_loss": 0.2037
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 3.3028,
+      "step": 120,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 3.2354,
+      "step": 120,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 2.8069,
+      "step": 120,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 2.93,
+      "step": 120,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 3.1315,
+      "step": 120,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.023100801309045408,
+      "lm_loss": 2.965,
+      "step": 120,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 0.023293307986620784,
+      "grad_norm": 7.665741220768339,
+      "learning_rate": 7.756410256410258e-06,
+      "loss": 3.212,
+      "step": 121
+    },
+    {
+      "epoch": 0.023485814664196164,
+      "grad_norm": 10.226398256908233,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 3.2717,
+      "step": 122
+    },
+    {
+      "epoch": 0.023678321341771544,
+      "grad_norm": 5.948491598057429,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 3.2855,
+      "step": 123
+    },
+    {
+      "epoch": 0.02387082801934692,
+      "grad_norm": 14.647611390497222,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 3.2483,
+      "step": 124
+    },
+    {
+      "epoch": 0.0240633346969223,
+      "grad_norm": 8.256742315660889,
+      "learning_rate": 8.012820512820515e-06,
+      "loss": 3.2698,
+      "step": 125
+    },
+    {
+      "epoch": 0.02425584137449768,
+      "grad_norm": 7.138236374002898,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 3.2711,
+      "step": 126
+    },
+    {
+      "epoch": 0.024448348052073056,
+      "grad_norm": 9.256013416446951,
+      "learning_rate": 8.141025641025641e-06,
+      "loss": 3.3322,
+      "step": 127
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "grad_norm": 7.537511602392607,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 3.2525,
+      "step": 128
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.1969,
+      "step": 128,
+      "vm_loss": 0.2097
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.5112,
+      "step": 128,
+      "vm_loss": 0.2458
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.2529,
+      "step": 128,
+      "vm_loss": 0.2193
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.2538,
+      "step": 128,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.1832,
+      "step": 128,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 2.9437,
+      "step": 128,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.3408,
+      "step": 128,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.024640854729648436,
+      "lm_loss": 3.0984,
+      "step": 128,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.024833361407223812,
+      "grad_norm": 8.250978419450655,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 3.3227,
+      "step": 129
+    },
+    {
+      "epoch": 0.02502586808479919,
+      "grad_norm": 7.3628457173539,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 3.2849,
+      "step": 130
+    },
+    {
+      "epoch": 0.02521837476237457,
+      "grad_norm": 8.454176258940308,
+      "learning_rate": 8.397435897435898e-06,
+      "loss": 3.1913,
+      "step": 131
+    },
+    {
+      "epoch": 0.025410881439949948,
+      "grad_norm": 8.376173215840465,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 3.2459,
+      "step": 132
+    },
+    {
+      "epoch": 0.025603388117525328,
+      "grad_norm": 6.87582665691323,
+      "learning_rate": 8.525641025641026e-06,
+      "loss": 3.2125,
+      "step": 133
+    },
+    {
+      "epoch": 0.025795894795100704,
+      "grad_norm": 8.896240523232297,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 3.1578,
+      "step": 134
+    },
+    {
+      "epoch": 0.025988401472676084,
+      "grad_norm": 6.6724173301052865,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 3.2215,
+      "step": 135
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "grad_norm": 6.6790152323660354,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 3.1951,
+      "step": 136
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.963,
+      "step": 136,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.7938,
+      "step": 136,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.9,
+      "step": 136,
+      "vm_loss": 0.2829
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 3.0702,
+      "step": 136,
+      "vm_loss": 0.2815
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.7033,
+      "step": 136,
+      "vm_loss": 0.2858
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.5698,
+      "step": 136,
+      "vm_loss": 0.2013
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 2.808,
+      "step": 136,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 0.026180908150251463,
+      "lm_loss": 3.1382,
+      "step": 136,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.02637341482782684,
+      "grad_norm": 7.2813674097077685,
+      "learning_rate": 8.782051282051283e-06,
+      "loss": 3.2166,
+      "step": 137
+    },
+    {
+      "epoch": 0.02656592150540222,
+      "grad_norm": 8.149438744977191,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 3.1668,
+      "step": 138
+    },
+    {
+      "epoch": 0.026758428182977596,
+      "grad_norm": 8.287910029793027,
+      "learning_rate": 8.910256410256411e-06,
+      "loss": 3.1465,
+      "step": 139
+    },
+    {
+      "epoch": 0.026950934860552975,
+      "grad_norm": 6.719044446835403,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 3.2167,
+      "step": 140
+    },
+    {
+      "epoch": 0.027143441538128355,
+      "grad_norm": 7.011100244291259,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 3.2148,
+      "step": 141
+    },
+    {
+      "epoch": 0.02733594821570373,
+      "grad_norm": 6.1920104301004235,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 3.1763,
+      "step": 142
+    },
+    {
+      "epoch": 0.02752845489327911,
+      "grad_norm": 6.755555458151317,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 3.1924,
+      "step": 143
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "grad_norm": 5.608491649533835,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 3.1493,
+      "step": 144
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 3.0104,
+      "step": 144,
+      "vm_loss": 0.2395
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 2.9678,
+      "step": 144,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 2.9748,
+      "step": 144,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 2.6869,
+      "step": 144,
+      "vm_loss": 0.2255
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 2.6758,
+      "step": 144,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 2.8615,
+      "step": 144,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 3.2455,
+      "step": 144,
+      "vm_loss": 0.2135
+    },
+    {
+      "epoch": 0.027720961570854488,
+      "lm_loss": 3.4464,
+      "step": 144,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.027913468248429867,
+      "grad_norm": 6.808337906129715,
+      "learning_rate": 9.294871794871796e-06,
+      "loss": 3.2025,
+      "step": 145
+    },
+    {
+      "epoch": 0.028105974926005247,
+      "grad_norm": 5.167818737857269,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 3.1271,
+      "step": 146
+    },
+    {
+      "epoch": 0.028298481603580623,
+      "grad_norm": 4.972735180235373,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 3.1229,
+      "step": 147
+    },
+    {
+      "epoch": 0.028490988281156003,
+      "grad_norm": 5.268973889273245,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 3.0918,
+      "step": 148
+    },
+    {
+      "epoch": 0.02868349495873138,
+      "grad_norm": 6.433975035834553,
+      "learning_rate": 9.551282051282053e-06,
+      "loss": 3.088,
+      "step": 149
+    },
+    {
+      "epoch": 0.02887600163630676,
+      "grad_norm": 4.771609768692185,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 3.0797,
+      "step": 150
+    },
+    {
+      "epoch": 0.02906850831388214,
+      "grad_norm": 5.673825586895475,
+      "learning_rate": 9.67948717948718e-06,
+      "loss": 3.132,
+      "step": 151
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "grad_norm": 4.916995450862572,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 3.0724,
+      "step": 152
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.8541,
+      "step": 152,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.814,
+      "step": 152,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 3.0889,
+      "step": 152,
+      "vm_loss": 0.2279
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.9812,
+      "step": 152,
+      "vm_loss": 0.2412
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.6343,
+      "step": 152,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 3.0641,
+      "step": 152,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.9087,
+      "step": 152,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.029261014991457515,
+      "lm_loss": 2.9947,
+      "step": 152,
+      "vm_loss": 0.21
+    },
+    {
+      "epoch": 0.029453521669032895,
+      "grad_norm": 5.345451285169419,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 3.1293,
+      "step": 153
+    },
+    {
+      "epoch": 0.029646028346608275,
+      "grad_norm": 5.255283821767704,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 3.0883,
+      "step": 154
+    },
+    {
+      "epoch": 0.02983853502418365,
+      "grad_norm": 7.053242397925785,
+      "learning_rate": 9.935897435897437e-06,
+      "loss": 3.1023,
+      "step": 155
+    },
+    {
+      "epoch": 0.03003104170175903,
+      "grad_norm": 5.450663155255625,
+      "learning_rate": 1e-05,
+      "loss": 3.0896,
+      "step": 156
+    },
+    {
+      "epoch": 0.030223548379334407,
+      "grad_norm": 6.9974723489596995,
+      "learning_rate": 1.0064102564102565e-05,
+      "loss": 3.0843,
+      "step": 157
+    },
+    {
+      "epoch": 0.030416055056909787,
+      "grad_norm": 6.575972837059719,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 3.053,
+      "step": 158
+    },
+    {
+      "epoch": 0.030608561734485167,
+      "grad_norm": 7.613242070548436,
+      "learning_rate": 1.0192307692307692e-05,
+      "loss": 3.0256,
+      "step": 159
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "grad_norm": 5.877015920330703,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 3.0322,
+      "step": 160
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.9073,
+      "step": 160,
+      "vm_loss": 0.2618
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.6029,
+      "step": 160,
+      "vm_loss": 0.2138
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 3.0235,
+      "step": 160,
+      "vm_loss": 0.2892
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.9285,
+      "step": 160,
+      "vm_loss": 0.2265
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.9395,
+      "step": 160,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.9529,
+      "step": 160,
+      "vm_loss": 0.2529
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.7037,
+      "step": 160,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.030801068412060543,
+      "lm_loss": 2.9467,
+      "step": 160,
+      "vm_loss": 0.2386
+    },
+    {
+      "epoch": 0.030993575089635923,
+      "grad_norm": 7.043293903695604,
+      "learning_rate": 1.0320512820512822e-05,
+      "loss": 3.1206,
+      "step": 161
+    },
+    {
+      "epoch": 0.0311860817672113,
+      "grad_norm": 5.026521806205481,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 3.0801,
+      "step": 162
+    },
+    {
+      "epoch": 0.03137858844478668,
+      "grad_norm": 9.250564457388757,
+      "learning_rate": 1.044871794871795e-05,
+      "loss": 3.0663,
+      "step": 163
+    },
+    {
+      "epoch": 0.03157109512236206,
+      "grad_norm": 6.517869865889641,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 3.057,
+      "step": 164
+    },
+    {
+      "epoch": 0.03176360179993744,
+      "grad_norm": 10.127518582941192,
+      "learning_rate": 1.0576923076923078e-05,
+      "loss": 3.0116,
+      "step": 165
+    },
+    {
+      "epoch": 0.03195610847751281,
+      "grad_norm": 7.467979551777786,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 3.0232,
+      "step": 166
+    },
+    {
+      "epoch": 0.03214861515508819,
+      "grad_norm": 11.873610868910118,
+      "learning_rate": 1.0705128205128205e-05,
+      "loss": 3.0264,
+      "step": 167
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "grad_norm": 7.355551676738349,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 3.0486,
+      "step": 168
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.8384,
+      "step": 168,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.8109,
+      "step": 168,
+      "vm_loss": 0.2679
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 3.158,
+      "step": 168,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.9281,
+      "step": 168,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.754,
+      "step": 168,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.8037,
+      "step": 168,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.6258,
+      "step": 168,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.03234112183266357,
+      "lm_loss": 2.6665,
+      "step": 168,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 0.03253362851023895,
+      "grad_norm": 10.013137269000024,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 3.0366,
+      "step": 169
+    },
+    {
+      "epoch": 0.03272613518781433,
+      "grad_norm": 10.246365030507082,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 3.0152,
+      "step": 170
+    },
+    {
+      "epoch": 0.0329186418653897,
+      "grad_norm": 12.843857885035476,
+      "learning_rate": 1.0961538461538464e-05,
+      "loss": 3.0374,
+      "step": 171
+    },
+    {
+      "epoch": 0.03311114854296508,
+      "grad_norm": 10.140554754776014,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 3.0704,
+      "step": 172
+    },
+    {
+      "epoch": 0.03330365522054046,
+      "grad_norm": 7.867327631267269,
+      "learning_rate": 1.1089743589743592e-05,
+      "loss": 2.9356,
+      "step": 173
+    },
+    {
+      "epoch": 0.03349616189811584,
+      "grad_norm": 8.407609223787436,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 2.9535,
+      "step": 174
+    },
+    {
+      "epoch": 0.03368866857569122,
+      "grad_norm": 8.469515518588421,
+      "learning_rate": 1.1217948717948719e-05,
+      "loss": 3.0489,
+      "step": 175
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "grad_norm": 6.419257757286239,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 2.9658,
+      "step": 176
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 3.1021,
+      "step": 176,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.7029,
+      "step": 176,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 3.1034,
+      "step": 176,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.6592,
+      "step": 176,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.5599,
+      "step": 176,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.65,
+      "step": 176,
+      "vm_loss": 0.2721
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.8335,
+      "step": 176,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.033881175253266595,
+      "lm_loss": 2.7471,
+      "step": 176,
+      "vm_loss": 0.2585
+    },
+    {
+      "epoch": 0.034073681930841974,
+      "grad_norm": 9.656691837353117,
+      "learning_rate": 1.1346153846153847e-05,
+      "loss": 2.977,
+      "step": 177
+    },
+    {
+      "epoch": 0.034266188608417354,
+      "grad_norm": 7.657196372897803,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 2.953,
+      "step": 178
+    },
+    {
+      "epoch": 0.034458695285992734,
+      "grad_norm": 7.68150357555819,
+      "learning_rate": 1.1474358974358974e-05,
+      "loss": 2.9969,
+      "step": 179
+    },
+    {
+      "epoch": 0.034651201963568114,
+      "grad_norm": 6.58729845599872,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 3.0113,
+      "step": 180
+    },
+    {
+      "epoch": 0.03484370864114349,
+      "grad_norm": 6.285673454310952,
+      "learning_rate": 1.1602564102564104e-05,
+      "loss": 2.9444,
+      "step": 181
+    },
+    {
+      "epoch": 0.035036215318718866,
+      "grad_norm": 5.761664043336924,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 2.9881,
+      "step": 182
+    },
+    {
+      "epoch": 0.035228721996294246,
+      "grad_norm": 5.410695641675287,
+      "learning_rate": 1.1730769230769232e-05,
+      "loss": 2.9334,
+      "step": 183
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "grad_norm": 6.023237121264645,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 2.9166,
+      "step": 184
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 2.975,
+      "step": 184,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 3.5106,
+      "step": 184,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 3.418,
+      "step": 184,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 2.9029,
+      "step": 184,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 2.9491,
+      "step": 184,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 2.8136,
+      "step": 184,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 2.8904,
+      "step": 184,
+      "vm_loss": 0.2051
+    },
+    {
+      "epoch": 0.035421228673869626,
+      "lm_loss": 3.2085,
+      "step": 184,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.035613735351445006,
+      "grad_norm": 27.35729649401158,
+      "learning_rate": 1.185897435897436e-05,
+      "loss": 3.2101,
+      "step": 185
+    },
+    {
+      "epoch": 0.03580624202902038,
+      "grad_norm": 12.004355265836677,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 3.2179,
+      "step": 186
+    },
+    {
+      "epoch": 0.03599874870659576,
+      "grad_norm": 13.72068537647716,
+      "learning_rate": 1.1987179487179487e-05,
+      "loss": 3.1634,
+      "step": 187
+    },
+    {
+      "epoch": 0.03619125538417114,
+      "grad_norm": 8.089655402274973,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 3.0601,
+      "step": 188
+    },
+    {
+      "epoch": 0.03638376206174652,
+      "grad_norm": 48.303874189153646,
+      "learning_rate": 1.2115384615384615e-05,
+      "loss": 3.2995,
+      "step": 189
+    },
+    {
+      "epoch": 0.0365762687393219,
+      "grad_norm": 17.441258568769378,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 3.293,
+      "step": 190
+    },
+    {
+      "epoch": 0.03676877541689727,
+      "grad_norm": 12.222015674075696,
+      "learning_rate": 1.2243589743589746e-05,
+      "loss": 3.3524,
+      "step": 191
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "grad_norm": 13.784523262562121,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 3.3192,
+      "step": 192
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 2.9604,
+      "step": 192,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 2.8938,
+      "step": 192,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 3.1123,
+      "step": 192,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 2.9682,
+      "step": 192,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 3.089,
+      "step": 192,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 3.1525,
+      "step": 192,
+      "vm_loss": 0.1352
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 3.0075,
+      "step": 192,
+      "vm_loss": 0.2288
+    },
+    {
+      "epoch": 0.03696128209447265,
+      "lm_loss": 3.1707,
+      "step": 192,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.03715378877204803,
+      "grad_norm": 12.289969252302749,
+      "learning_rate": 1.2371794871794874e-05,
+      "loss": 3.291,
+      "step": 193
+    },
+    {
+      "epoch": 0.03734629544962341,
+      "grad_norm": 10.159277758140643,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 3.136,
+      "step": 194
+    },
+    {
+      "epoch": 0.03753880212719879,
+      "grad_norm": 11.740631375133924,
+      "learning_rate": 1.25e-05,
+      "loss": 3.1739,
+      "step": 195
+    },
+    {
+      "epoch": 0.03773130880477417,
+      "grad_norm": 12.173446170654557,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 3.1562,
+      "step": 196
+    },
+    {
+      "epoch": 0.03792381548234954,
+      "grad_norm": 8.518876013836504,
+      "learning_rate": 1.2628205128205129e-05,
+      "loss": 3.1074,
+      "step": 197
+    },
+    {
+      "epoch": 0.03811632215992492,
+      "grad_norm": 11.30323844302314,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 3.1066,
+      "step": 198
+    },
+    {
+      "epoch": 0.0383088288375003,
+      "grad_norm": 7.691372851932579,
+      "learning_rate": 1.2756410256410257e-05,
+      "loss": 3.048,
+      "step": 199
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "grad_norm": 8.468513935727255,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 3.0237,
+      "step": 200
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.6696,
+      "step": 200,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.7005,
+      "step": 200,
+      "vm_loss": 0.2201
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.8287,
+      "step": 200,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.8452,
+      "step": 200,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.9198,
+      "step": 200,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.913,
+      "step": 200,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.7118,
+      "step": 200,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 0.03850133551507568,
+      "lm_loss": 2.9899,
+      "step": 200,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.03869384219265106,
+      "grad_norm": 7.116072705989573,
+      "learning_rate": 1.2884615384615386e-05,
+      "loss": 3.0521,
+      "step": 201
+    },
+    {
+      "epoch": 0.038886348870226434,
+      "grad_norm": 5.746238686341502,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 2.9985,
+      "step": 202
+    },
+    {
+      "epoch": 0.039078855547801813,
+      "grad_norm": 6.989671432523846,
+      "learning_rate": 1.3012820512820514e-05,
+      "loss": 2.9781,
+      "step": 203
+    },
+    {
+      "epoch": 0.03927136222537719,
+      "grad_norm": 5.9360335543901925,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 2.933,
+      "step": 204
+    },
+    {
+      "epoch": 0.03946386890295257,
+      "grad_norm": 4.85956682392983,
+      "learning_rate": 1.3141025641025642e-05,
+      "loss": 2.9653,
+      "step": 205
+    },
+    {
+      "epoch": 0.03965637558052795,
+      "grad_norm": 6.550014255490412,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 2.9506,
+      "step": 206
+    },
+    {
+      "epoch": 0.039848882258103326,
+      "grad_norm": 6.704605878999232,
+      "learning_rate": 1.3269230769230769e-05,
+      "loss": 2.9385,
+      "step": 207
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "grad_norm": 4.346420782145899,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 2.9289,
+      "step": 208
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.5593,
+      "step": 208,
+      "vm_loss": 0.2811
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.6893,
+      "step": 208,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.7023,
+      "step": 208,
+      "vm_loss": 0.2148
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.8206,
+      "step": 208,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.3611,
+      "step": 208,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.7323,
+      "step": 208,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.5944,
+      "step": 208,
+      "vm_loss": 0.2861
+    },
+    {
+      "epoch": 0.040041388935678705,
+      "lm_loss": 2.584,
+      "step": 208,
+      "vm_loss": 0.2026
+    },
+    {
+      "epoch": 0.040233895613254085,
+      "grad_norm": 5.774392950229877,
+      "learning_rate": 1.3397435897435897e-05,
+      "loss": 2.9859,
+      "step": 209
+    },
+    {
+      "epoch": 0.040426402290829465,
+      "grad_norm": 5.093636173741241,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 2.9473,
+      "step": 210
+    },
+    {
+      "epoch": 0.040618908968404845,
+      "grad_norm": 4.300007709417803,
+      "learning_rate": 1.3525641025641028e-05,
+      "loss": 2.92,
+      "step": 211
+    },
+    {
+      "epoch": 0.04081141564598022,
+      "grad_norm": 5.809945210465388,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 2.9335,
+      "step": 212
+    },
+    {
+      "epoch": 0.0410039223235556,
+      "grad_norm": 5.207221432710424,
+      "learning_rate": 1.3653846153846156e-05,
+      "loss": 2.9261,
+      "step": 213
+    },
+    {
+      "epoch": 0.04119642900113098,
+      "grad_norm": 4.896002202116585,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 2.9121,
+      "step": 214
+    },
+    {
+      "epoch": 0.04138893567870636,
+      "grad_norm": 4.725368576157323,
+      "learning_rate": 1.3782051282051283e-05,
+      "loss": 2.8892,
+      "step": 215
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "grad_norm": 5.002780050146259,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 2.8562,
+      "step": 216
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.8098,
+      "step": 216,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.8638,
+      "step": 216,
+      "vm_loss": 0.2558
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.4597,
+      "step": 216,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.6925,
+      "step": 216,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.6028,
+      "step": 216,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.6288,
+      "step": 216,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 3.1163,
+      "step": 216,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 0.041581442356281736,
+      "lm_loss": 2.6315,
+      "step": 216,
+      "vm_loss": 0.2213
+    },
+    {
+      "epoch": 0.04177394903385711,
+      "grad_norm": 6.479679840921408,
+      "learning_rate": 1.3910256410256411e-05,
+      "loss": 2.9393,
+      "step": 217
+    },
+    {
+      "epoch": 0.04196645571143249,
+      "grad_norm": 4.572757285960612,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 2.8804,
+      "step": 218
+    },
+    {
+      "epoch": 0.04215896238900787,
+      "grad_norm": 5.074904352294016,
+      "learning_rate": 1.403846153846154e-05,
+      "loss": 2.8821,
+      "step": 219
+    },
+    {
+      "epoch": 0.04235146906658325,
+      "grad_norm": 4.806943875282919,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 2.9012,
+      "step": 220
+    },
+    {
+      "epoch": 0.04254397574415863,
+      "grad_norm": 5.470116200121896,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 2.8102,
+      "step": 221
+    },
+    {
+      "epoch": 0.042736482421734,
+      "grad_norm": 4.44921275322167,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 2.8962,
+      "step": 222
+    },
+    {
+      "epoch": 0.04292898909930938,
+      "grad_norm": 5.662064155307462,
+      "learning_rate": 1.4294871794871796e-05,
+      "loss": 2.8708,
+      "step": 223
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "grad_norm": 5.353328706835362,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 2.846,
+      "step": 224
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.4981,
+      "step": 224,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.7507,
+      "step": 224,
+      "vm_loss": 0.1174
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.8412,
+      "step": 224,
+      "vm_loss": 0.267
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.6686,
+      "step": 224,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.3326,
+      "step": 224,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.2581,
+      "step": 224,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.6294,
+      "step": 224,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.04312149577688476,
+      "lm_loss": 2.9014,
+      "step": 224,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 0.04331400245446014,
+      "grad_norm": 4.3566890598549755,
+      "learning_rate": 1.4423076923076924e-05,
+      "loss": 2.8706,
+      "step": 225
+    },
+    {
+      "epoch": 0.04350650913203552,
+      "grad_norm": 5.77724861491947,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 2.8869,
+      "step": 226
+    },
+    {
+      "epoch": 0.04369901580961089,
+      "grad_norm": 4.689389375319442,
+      "learning_rate": 1.4551282051282051e-05,
+      "loss": 2.8259,
+      "step": 227
+    },
+    {
+      "epoch": 0.04389152248718627,
+      "grad_norm": 5.768321825370758,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 2.8594,
+      "step": 228
+    },
+    {
+      "epoch": 0.04408402916476165,
+      "grad_norm": 5.055908443189705,
+      "learning_rate": 1.467948717948718e-05,
+      "loss": 2.8894,
+      "step": 229
+    },
+    {
+      "epoch": 0.04427653584233703,
+      "grad_norm": 4.829683354618881,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 2.8516,
+      "step": 230
+    },
+    {
+      "epoch": 0.04446904251991241,
+      "grad_norm": 5.512977719670376,
+      "learning_rate": 1.480769230769231e-05,
+      "loss": 2.8592,
+      "step": 231
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "grad_norm": 5.113545819702339,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 2.841,
+      "step": 232
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.4145,
+      "step": 232,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.4598,
+      "step": 232,
+      "vm_loss": 0.2467
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.6577,
+      "step": 232,
+      "vm_loss": 0.2049
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.526,
+      "step": 232,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.6536,
+      "step": 232,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.3985,
+      "step": 232,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.5585,
+      "step": 232,
+      "vm_loss": 0.2144
+    },
+    {
+      "epoch": 0.044661549197487785,
+      "lm_loss": 2.8372,
+      "step": 232,
+      "vm_loss": 0.2464
+    },
+    {
+      "epoch": 0.044854055875063165,
+      "grad_norm": 6.417788955615832,
+      "learning_rate": 1.4935897435897438e-05,
+      "loss": 2.8161,
+      "step": 233
+    },
+    {
+      "epoch": 0.045046562552638544,
+      "grad_norm": 3.71111057112546,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 2.8409,
+      "step": 234
+    },
+    {
+      "epoch": 0.045239069230213924,
+      "grad_norm": 5.681294642281556,
+      "learning_rate": 1.5064102564102565e-05,
+      "loss": 2.8796,
+      "step": 235
+    },
+    {
+      "epoch": 0.045431575907789304,
+      "grad_norm": 4.2874115736934595,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 2.846,
+      "step": 236
+    },
+    {
+      "epoch": 0.04562408258536468,
+      "grad_norm": 5.0998004345925905,
+      "learning_rate": 1.5192307692307693e-05,
+      "loss": 2.83,
+      "step": 237
+    },
+    {
+      "epoch": 0.045816589262940056,
+      "grad_norm": 4.857255113532344,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 2.7729,
+      "step": 238
+    },
+    {
+      "epoch": 0.046009095940515436,
+      "grad_norm": 5.731345644511636,
+      "learning_rate": 1.5320512820512823e-05,
+      "loss": 2.8073,
+      "step": 239
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "grad_norm": 4.074735776209843,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 2.8033,
+      "step": 240
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.5881,
+      "step": 240,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.42,
+      "step": 240,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.8157,
+      "step": 240,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.755,
+      "step": 240,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.5736,
+      "step": 240,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.7088,
+      "step": 240,
+      "vm_loss": 0.225
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.4963,
+      "step": 240,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.046201602618090816,
+      "lm_loss": 2.8049,
+      "step": 240,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.046394109295666196,
+      "grad_norm": 5.448771509383658,
+      "learning_rate": 1.544871794871795e-05,
+      "loss": 2.8039,
+      "step": 241
+    },
+    {
+      "epoch": 0.04658661597324157,
+      "grad_norm": 5.181558569222629,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 2.8054,
+      "step": 242
+    },
+    {
+      "epoch": 0.04677912265081695,
+      "grad_norm": 4.837739297340354,
+      "learning_rate": 1.557692307692308e-05,
+      "loss": 2.8442,
+      "step": 243
+    },
+    {
+      "epoch": 0.04697162932839233,
+      "grad_norm": 5.31087875515781,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 2.8542,
+      "step": 244
+    },
+    {
+      "epoch": 0.04716413600596771,
+      "grad_norm": 4.764992258784042,
+      "learning_rate": 1.5705128205128205e-05,
+      "loss": 2.7845,
+      "step": 245
+    },
+    {
+      "epoch": 0.04735664268354309,
+      "grad_norm": 6.284306004867487,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 2.7959,
+      "step": 246
+    },
+    {
+      "epoch": 0.04754914936111846,
+      "grad_norm": 5.321702902470818,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 2.8045,
+      "step": 247
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "grad_norm": 4.79852329988178,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 2.8272,
+      "step": 248
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.1254,
+      "step": 248,
+      "vm_loss": 0.1234
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.7617,
+      "step": 248,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.5905,
+      "step": 248,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.71,
+      "step": 248,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.4503,
+      "step": 248,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.6204,
+      "step": 248,
+      "vm_loss": 0.2362
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.5351,
+      "step": 248,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 0.04774165603869384,
+      "lm_loss": 2.4125,
+      "step": 248,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.04793416271626922,
+      "grad_norm": 6.4595083159010835,
+      "learning_rate": 1.5961538461538465e-05,
+      "loss": 2.8033,
+      "step": 249
+    },
+    {
+      "epoch": 0.0481266693938446,
+      "grad_norm": 4.308638650480195,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 2.8103,
+      "step": 250
+    },
+    {
+      "epoch": 0.04831917607141998,
+      "grad_norm": 6.4101497723491425,
+      "learning_rate": 1.6089743589743593e-05,
+      "loss": 2.7673,
+      "step": 251
+    },
+    {
+      "epoch": 0.04851168274899536,
+      "grad_norm": 4.532903144942222,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 2.8179,
+      "step": 252
+    },
+    {
+      "epoch": 0.04870418942657073,
+      "grad_norm": 4.748829857520044,
+      "learning_rate": 1.6217948717948718e-05,
+      "loss": 2.7881,
+      "step": 253
+    },
+    {
+      "epoch": 0.04889669610414611,
+      "grad_norm": 4.948904119589788,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 2.7535,
+      "step": 254
+    },
+    {
+      "epoch": 0.04908920278172149,
+      "grad_norm": 4.6513772973707574,
+      "learning_rate": 1.6346153846153847e-05,
+      "loss": 2.7733,
+      "step": 255
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "grad_norm": 4.294851496778557,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 2.8135,
+      "step": 256
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.7238,
+      "step": 256,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.5891,
+      "step": 256,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.8327,
+      "step": 256,
+      "vm_loss": 0.1018
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.7131,
+      "step": 256,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.6686,
+      "step": 256,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.6499,
+      "step": 256,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.7558,
+      "step": 256,
+      "vm_loss": 0.1277
+    },
+    {
+      "epoch": 0.04928170945929687,
+      "lm_loss": 2.4651,
+      "step": 256,
+      "vm_loss": 0.1893
+    },
+    {
+      "epoch": 0.04947421613687225,
+      "grad_norm": 5.67444934915225,
+      "learning_rate": 1.6474358974358975e-05,
+      "loss": 2.75,
+      "step": 257
+    },
+    {
+      "epoch": 0.049666722814447624,
+      "grad_norm": 4.179304943641643,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 2.7486,
+      "step": 258
+    },
+    {
+      "epoch": 0.049859229492023004,
+      "grad_norm": 5.4761858408932,
+      "learning_rate": 1.6602564102564103e-05,
+      "loss": 2.7808,
+      "step": 259
+    },
+    {
+      "epoch": 0.05005173616959838,
+      "grad_norm": 4.61371812438267,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.7898,
+      "step": 260
+    },
+    {
+      "epoch": 0.05024424284717376,
+      "grad_norm": 5.207440224754311,
+      "learning_rate": 1.673076923076923e-05,
+      "loss": 2.7671,
+      "step": 261
+    },
+    {
+      "epoch": 0.05043674952474914,
+      "grad_norm": 4.854998652086252,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 2.7799,
+      "step": 262
+    },
+    {
+      "epoch": 0.050629256202324516,
+      "grad_norm": 5.043735751444399,
+      "learning_rate": 1.685897435897436e-05,
+      "loss": 2.7374,
+      "step": 263
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "grad_norm": 5.142440309794773,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 2.7828,
+      "step": 264
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.6345,
+      "step": 264,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.6301,
+      "step": 264,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.5312,
+      "step": 264,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.6304,
+      "step": 264,
+      "vm_loss": 0.201
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.7943,
+      "step": 264,
+      "vm_loss": 0.2887
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.1806,
+      "step": 264,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.6636,
+      "step": 264,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.050821762879899895,
+      "lm_loss": 2.4347,
+      "step": 264,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 0.051014269557475275,
+      "grad_norm": 5.432414537043364,
+      "learning_rate": 1.698717948717949e-05,
+      "loss": 2.7929,
+      "step": 265
+    },
+    {
+      "epoch": 0.051206776235050655,
+      "grad_norm": 5.011886065042031,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 2.752,
+      "step": 266
+    },
+    {
+      "epoch": 0.051399282912626035,
+      "grad_norm": 5.690458115360194,
+      "learning_rate": 1.7115384615384617e-05,
+      "loss": 2.7801,
+      "step": 267
+    },
+    {
+      "epoch": 0.05159178959020141,
+      "grad_norm": 4.9953507143641485,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 2.7851,
+      "step": 268
+    },
+    {
+      "epoch": 0.05178429626777679,
+      "grad_norm": 6.136330745638589,
+      "learning_rate": 1.7243589743589745e-05,
+      "loss": 2.8012,
+      "step": 269
+    },
+    {
+      "epoch": 0.05197680294535217,
+      "grad_norm": 5.398065985219351,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 2.758,
+      "step": 270
+    },
+    {
+      "epoch": 0.05216930962292755,
+      "grad_norm": 5.627593326165706,
+      "learning_rate": 1.7371794871794873e-05,
+      "loss": 2.7479,
+      "step": 271
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "grad_norm": 5.4363601291925345,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 2.7913,
+      "step": 272
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.6982,
+      "step": 272,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.7023,
+      "step": 272,
+      "vm_loss": 0.2113
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.5525,
+      "step": 272,
+      "vm_loss": 0.2256
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.607,
+      "step": 272,
+      "vm_loss": 0.2278
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.7356,
+      "step": 272,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.7488,
+      "step": 272,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.5289,
+      "step": 272,
+      "vm_loss": 0.2322
+    },
+    {
+      "epoch": 0.05236181630050293,
+      "lm_loss": 2.4803,
+      "step": 272,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 0.0525543229780783,
+      "grad_norm": 6.71871940737774,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 2.7886,
+      "step": 273
+    },
+    {
+      "epoch": 0.05274682965565368,
+      "grad_norm": 5.810392290014264,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 2.8243,
+      "step": 274
+    },
+    {
+      "epoch": 0.05293933633322906,
+      "grad_norm": 4.220671687484487,
+      "learning_rate": 1.762820512820513e-05,
+      "loss": 2.7267,
+      "step": 275
+    },
+    {
+      "epoch": 0.05313184301080444,
+      "grad_norm": 3.997163303641086,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 2.7728,
+      "step": 276
+    },
+    {
+      "epoch": 0.05332434968837982,
+      "grad_norm": 5.611693679867312,
+      "learning_rate": 1.775641025641026e-05,
+      "loss": 2.7983,
+      "step": 277
+    },
+    {
+      "epoch": 0.05351685636595519,
+      "grad_norm": 5.971487917375231,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 2.7688,
+      "step": 278
+    },
+    {
+      "epoch": 0.05370936304353057,
+      "grad_norm": 5.502538116528449,
+      "learning_rate": 1.7884615384615387e-05,
+      "loss": 2.7181,
+      "step": 279
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "grad_norm": 5.063807787749939,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 2.688,
+      "step": 280
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.3802,
+      "step": 280,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.3224,
+      "step": 280,
+      "vm_loss": 0.2148
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.7132,
+      "step": 280,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.5248,
+      "step": 280,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.3786,
+      "step": 280,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.5967,
+      "step": 280,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.6854,
+      "step": 280,
+      "vm_loss": 0.2494
+    },
+    {
+      "epoch": 0.05390186972110595,
+      "lm_loss": 2.6236,
+      "step": 280,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 0.05409437639868133,
+      "grad_norm": 5.557173810961538,
+      "learning_rate": 1.8012820512820515e-05,
+      "loss": 2.7521,
+      "step": 281
+    },
+    {
+      "epoch": 0.05428688307625671,
+      "grad_norm": 6.099894343529975,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 2.7306,
+      "step": 282
+    },
+    {
+      "epoch": 0.05447938975383208,
+      "grad_norm": 5.750374294384423,
+      "learning_rate": 1.8141025641025644e-05,
+      "loss": 2.6932,
+      "step": 283
+    },
+    {
+      "epoch": 0.05467189643140746,
+      "grad_norm": 4.596638318229386,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 2.7395,
+      "step": 284
+    },
+    {
+      "epoch": 0.05486440310898284,
+      "grad_norm": 6.061063918190246,
+      "learning_rate": 1.826923076923077e-05,
+      "loss": 2.779,
+      "step": 285
+    },
+    {
+      "epoch": 0.05505690978655822,
+      "grad_norm": 6.843683391493255,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 2.7758,
+      "step": 286
+    },
+    {
+      "epoch": 0.0552494164641336,
+      "grad_norm": 5.7417151251373,
+      "learning_rate": 1.8397435897435897e-05,
+      "loss": 2.6776,
+      "step": 287
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "grad_norm": 5.95014629777928,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 2.7491,
+      "step": 288
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.7807,
+      "step": 288,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.5052,
+      "step": 288,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.3852,
+      "step": 288,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.5051,
+      "step": 288,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.5671,
+      "step": 288,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.7327,
+      "step": 288,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.9177,
+      "step": 288,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 0.055441923141708975,
+      "lm_loss": 2.2264,
+      "step": 288,
+      "vm_loss": 0.2377
+    },
+    {
+      "epoch": 0.055634429819284355,
+      "grad_norm": 7.481916905302612,
+      "learning_rate": 1.852564102564103e-05,
+      "loss": 2.7311,
+      "step": 289
+    },
+    {
+      "epoch": 0.055826936496859735,
+      "grad_norm": 10.666369396641484,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 2.711,
+      "step": 290
+    },
+    {
+      "epoch": 0.056019443174435114,
+      "grad_norm": 6.643483793492336,
+      "learning_rate": 1.8653846153846157e-05,
+      "loss": 2.6804,
+      "step": 291
+    },
+    {
+      "epoch": 0.056211949852010494,
+      "grad_norm": 7.834274382181083,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 2.7641,
+      "step": 292
+    },
+    {
+      "epoch": 0.05640445652958587,
+      "grad_norm": 5.481083990593822,
+      "learning_rate": 1.8782051282051282e-05,
+      "loss": 2.6757,
+      "step": 293
+    },
+    {
+      "epoch": 0.05659696320716125,
+      "grad_norm": 13.494098611415133,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 2.7491,
+      "step": 294
+    },
+    {
+      "epoch": 0.056789469884736626,
+      "grad_norm": 7.929890808822521,
+      "learning_rate": 1.891025641025641e-05,
+      "loss": 2.7485,
+      "step": 295
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "grad_norm": 8.078997650865817,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 2.7422,
+      "step": 296
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.6819,
+      "step": 296,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.5034,
+      "step": 296,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.4533,
+      "step": 296,
+      "vm_loss": 0.2169
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.6871,
+      "step": 296,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.4849,
+      "step": 296,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.5656,
+      "step": 296,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.5809,
+      "step": 296,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.056981976562312006,
+      "lm_loss": 2.4418,
+      "step": 296,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.057174483239887386,
+      "grad_norm": 6.614455299732487,
+      "learning_rate": 1.903846153846154e-05,
+      "loss": 2.7212,
+      "step": 297
+    },
+    {
+      "epoch": 0.05736698991746276,
+      "grad_norm": 5.590732664103167,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 2.7189,
+      "step": 298
+    },
+    {
+      "epoch": 0.05755949659503814,
+      "grad_norm": 17.867175223198412,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 2.7574,
+      "step": 299
+    },
+    {
+      "epoch": 0.05775200327261352,
+      "grad_norm": 6.270512488879316,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 2.7043,
+      "step": 300
+    },
+    {
+      "epoch": 0.0579445099501889,
+      "grad_norm": 7.82771296805973,
+      "learning_rate": 1.9294871794871796e-05,
+      "loss": 2.7745,
+      "step": 301
+    },
+    {
+      "epoch": 0.05813701662776428,
+      "grad_norm": 6.860681826757564,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 2.7372,
+      "step": 302
+    },
+    {
+      "epoch": 0.05832952330533965,
+      "grad_norm": 7.9165276479442,
+      "learning_rate": 1.9423076923076924e-05,
+      "loss": 2.7511,
+      "step": 303
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "grad_norm": 6.403564952036589,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 2.725,
+      "step": 304
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.7356,
+      "step": 304,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.6297,
+      "step": 304,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.334,
+      "step": 304,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.5127,
+      "step": 304,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.5081,
+      "step": 304,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.4344,
+      "step": 304,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.8054,
+      "step": 304,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.05852202998291503,
+      "lm_loss": 2.5351,
+      "step": 304,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 0.05871453666049041,
+      "grad_norm": 6.838258056479387,
+      "learning_rate": 1.9551282051282052e-05,
+      "loss": 2.7151,
+      "step": 305
+    },
+    {
+      "epoch": 0.05890704333806579,
+      "grad_norm": 5.823482406268452,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 2.7605,
+      "step": 306
+    },
+    {
+      "epoch": 0.05909955001564117,
+      "grad_norm": 7.102550818386516,
+      "learning_rate": 1.967948717948718e-05,
+      "loss": 2.7435,
+      "step": 307
+    },
+    {
+      "epoch": 0.05929205669321655,
+      "grad_norm": 5.5919661140690335,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 2.7346,
+      "step": 308
+    },
+    {
+      "epoch": 0.05948456337079192,
+      "grad_norm": 6.437294533210534,
+      "learning_rate": 1.980769230769231e-05,
+      "loss": 2.6936,
+      "step": 309
+    },
+    {
+      "epoch": 0.0596770700483673,
+      "grad_norm": 7.530450249681175,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 2.7374,
+      "step": 310
+    },
+    {
+      "epoch": 0.05986957672594268,
+      "grad_norm": 6.1723717590724005,
+      "learning_rate": 1.9935897435897437e-05,
+      "loss": 2.7802,
+      "step": 311
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "grad_norm": 6.221761453623501,
+      "learning_rate": 2e-05,
+      "loss": 2.7176,
+      "step": 312
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.438,
+      "step": 312,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.4646,
+      "step": 312,
+      "vm_loss": 0.2016
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.6592,
+      "step": 312,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.5624,
+      "step": 312,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.7499,
+      "step": 312,
+      "vm_loss": 0.2068
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.4759,
+      "step": 312,
+      "vm_loss": 0.2478
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.4851,
+      "step": 312,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.06006208340351806,
+      "lm_loss": 2.534,
+      "step": 312,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.06025459008109344,
+      "grad_norm": 6.556227124834291,
+      "learning_rate": 1.9999999513936033e-05,
+      "loss": 2.726,
+      "step": 313
+    },
+    {
+      "epoch": 0.060447096758668814,
+      "grad_norm": 5.959865100550867,
+      "learning_rate": 1.9999998055744173e-05,
+      "loss": 2.7372,
+      "step": 314
+    },
+    {
+      "epoch": 0.060639603436244194,
+      "grad_norm": 5.833405481445536,
+      "learning_rate": 1.9999995625424566e-05,
+      "loss": 2.7118,
+      "step": 315
+    },
+    {
+      "epoch": 0.060832110113819574,
+      "grad_norm": 5.755580657232161,
+      "learning_rate": 1.9999992222977445e-05,
+      "loss": 2.7132,
+      "step": 316
+    },
+    {
+      "epoch": 0.06102461679139495,
+      "grad_norm": 5.755834599779296,
+      "learning_rate": 1.9999987848403148e-05,
+      "loss": 2.7577,
+      "step": 317
+    },
+    {
+      "epoch": 0.06121712346897033,
+      "grad_norm": 5.406238335625812,
+      "learning_rate": 1.999998250170209e-05,
+      "loss": 2.6694,
+      "step": 318
+    },
+    {
+      "epoch": 0.061409630146545706,
+      "grad_norm": 5.44359809849898,
+      "learning_rate": 1.9999976182874795e-05,
+      "loss": 2.702,
+      "step": 319
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "grad_norm": 4.808249430953446,
+      "learning_rate": 1.9999968891921883e-05,
+      "loss": 2.725,
+      "step": 320
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.6041,
+      "step": 320,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.7065,
+      "step": 320,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.1316,
+      "step": 320,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.5332,
+      "step": 320,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.6907,
+      "step": 320,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.5532,
+      "step": 320,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.7553,
+      "step": 320,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 0.061602136824121086,
+      "lm_loss": 2.5232,
+      "step": 320,
+      "vm_loss": 0.2404
+    },
+    {
+      "epoch": 0.061794643501696465,
+      "grad_norm": 6.107397910778231,
+      "learning_rate": 1.9999960628844055e-05,
+      "loss": 2.6995,
+      "step": 321
+    },
+    {
+      "epoch": 0.061987150179271845,
+      "grad_norm": 6.137674852555086,
+      "learning_rate": 1.9999951393642117e-05,
+      "loss": 2.6803,
+      "step": 322
+    },
+    {
+      "epoch": 0.062179656856847225,
+      "grad_norm": 4.68510259297647,
+      "learning_rate": 1.9999941186316965e-05,
+      "loss": 2.7011,
+      "step": 323
+    },
+    {
+      "epoch": 0.0623721635344226,
+      "grad_norm": 5.37521618938404,
+      "learning_rate": 1.9999930006869593e-05,
+      "loss": 2.7482,
+      "step": 324
+    },
+    {
+      "epoch": 0.06256467021199798,
+      "grad_norm": 5.77930171018597,
+      "learning_rate": 1.999991785530109e-05,
+      "loss": 2.703,
+      "step": 325
+    },
+    {
+      "epoch": 0.06275717688957336,
+      "grad_norm": 4.852920462008338,
+      "learning_rate": 1.9999904731612635e-05,
+      "loss": 2.7423,
+      "step": 326
+    },
+    {
+      "epoch": 0.06294968356714874,
+      "grad_norm": 6.152316334811615,
+      "learning_rate": 1.9999890635805506e-05,
+      "loss": 2.7177,
+      "step": 327
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "grad_norm": 5.0602729932716795,
+      "learning_rate": 1.999987556788107e-05,
+      "loss": 2.6476,
+      "step": 328
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.6998,
+      "step": 328,
+      "vm_loss": 0.2544
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.394,
+      "step": 328,
+      "vm_loss": 0.2385
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.6689,
+      "step": 328,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.8713,
+      "step": 328,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.5413,
+      "step": 328,
+      "vm_loss": 0.1094
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.5915,
+      "step": 328,
+      "vm_loss": 0.2122
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.1835,
+      "step": 328,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 0.06314219024472412,
+      "lm_loss": 2.3432,
+      "step": 328,
+      "vm_loss": 0.2225
+    },
+    {
+      "epoch": 0.0633346969222995,
+      "grad_norm": 5.53464512741921,
+      "learning_rate": 1.999985952784079e-05,
+      "loss": 2.7022,
+      "step": 329
+    },
+    {
+      "epoch": 0.06352720359987488,
+      "grad_norm": 5.071347296118864,
+      "learning_rate": 1.999984251568623e-05,
+      "loss": 2.6764,
+      "step": 330
+    },
+    {
+      "epoch": 0.06371971027745024,
+      "grad_norm": 4.8525379853183015,
+      "learning_rate": 1.9999824531419042e-05,
+      "loss": 2.6447,
+      "step": 331
+    },
+    {
+      "epoch": 0.06391221695502562,
+      "grad_norm": 4.974582527869404,
+      "learning_rate": 1.9999805575040976e-05,
+      "loss": 2.6804,
+      "step": 332
+    },
+    {
+      "epoch": 0.064104723632601,
+      "grad_norm": 4.445088280205642,
+      "learning_rate": 1.9999785646553872e-05,
+      "loss": 2.6619,
+      "step": 333
+    },
+    {
+      "epoch": 0.06429723031017638,
+      "grad_norm": 4.9470845625267765,
+      "learning_rate": 1.999976474595967e-05,
+      "loss": 2.7377,
+      "step": 334
+    },
+    {
+      "epoch": 0.06448973698775176,
+      "grad_norm": 5.472564416952947,
+      "learning_rate": 1.99997428732604e-05,
+      "loss": 2.6685,
+      "step": 335
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "grad_norm": 4.325664357046897,
+      "learning_rate": 1.9999720028458185e-05,
+      "loss": 2.6585,
+      "step": 336
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.4,
+      "step": 336,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.4846,
+      "step": 336,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.7019,
+      "step": 336,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.6328,
+      "step": 336,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.6214,
+      "step": 336,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.4843,
+      "step": 336,
+      "vm_loss": 0.1179
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.3489,
+      "step": 336,
+      "vm_loss": 0.2547
+    },
+    {
+      "epoch": 0.06468224366532714,
+      "lm_loss": 2.7837,
+      "step": 336,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.06487475034290252,
+      "grad_norm": 5.665835095279862,
+      "learning_rate": 1.9999696211555254e-05,
+      "loss": 2.6667,
+      "step": 337
+    },
+    {
+      "epoch": 0.0650672570204779,
+      "grad_norm": 5.401863221189648,
+      "learning_rate": 1.9999671422553914e-05,
+      "loss": 2.6879,
+      "step": 338
+    },
+    {
+      "epoch": 0.06525976369805328,
+      "grad_norm": 5.134478680834371,
+      "learning_rate": 1.999964566145658e-05,
+      "loss": 2.6951,
+      "step": 339
+    },
+    {
+      "epoch": 0.06545227037562866,
+      "grad_norm": 5.10687350027081,
+      "learning_rate": 1.9999618928265758e-05,
+      "loss": 2.6853,
+      "step": 340
+    },
+    {
+      "epoch": 0.06564477705320404,
+      "grad_norm": 5.918541336370904,
+      "learning_rate": 1.9999591222984042e-05,
+      "loss": 2.7043,
+      "step": 341
+    },
+    {
+      "epoch": 0.0658372837307794,
+      "grad_norm": 5.401240241834477,
+      "learning_rate": 1.9999562545614125e-05,
+      "loss": 2.7002,
+      "step": 342
+    },
+    {
+      "epoch": 0.06602979040835479,
+      "grad_norm": 5.55776949635833,
+      "learning_rate": 1.9999532896158796e-05,
+      "loss": 2.6814,
+      "step": 343
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "grad_norm": 4.940930329109183,
+      "learning_rate": 1.9999502274620942e-05,
+      "loss": 2.6935,
+      "step": 344
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.3664,
+      "step": 344,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.6444,
+      "step": 344,
+      "vm_loss": 0.2206
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.3721,
+      "step": 344,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.4093,
+      "step": 344,
+      "vm_loss": 0.1907
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.4839,
+      "step": 344,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.3184,
+      "step": 344,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.3154,
+      "step": 344,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 0.06622229708593017,
+      "lm_loss": 2.5662,
+      "step": 344,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 0.06641480376350554,
+      "grad_norm": 5.795138431781234,
+      "learning_rate": 1.999947068100353e-05,
+      "loss": 2.6436,
+      "step": 345
+    },
+    {
+      "epoch": 0.06660731044108092,
+      "grad_norm": 4.1998792847036475,
+      "learning_rate": 1.9999438115309645e-05,
+      "loss": 2.6276,
+      "step": 346
+    },
+    {
+      "epoch": 0.0667998171186563,
+      "grad_norm": 5.81983355879425,
+      "learning_rate": 1.999940457754244e-05,
+      "loss": 2.6668,
+      "step": 347
+    },
+    {
+      "epoch": 0.06699232379623168,
+      "grad_norm": 4.675919662854624,
+      "learning_rate": 1.999937006770518e-05,
+      "loss": 2.6722,
+      "step": 348
+    },
+    {
+      "epoch": 0.06718483047380706,
+      "grad_norm": 5.551831943600783,
+      "learning_rate": 1.9999334585801223e-05,
+      "loss": 2.6038,
+      "step": 349
+    },
+    {
+      "epoch": 0.06737733715138244,
+      "grad_norm": 4.6632356114396005,
+      "learning_rate": 1.9999298131834015e-05,
+      "loss": 2.6595,
+      "step": 350
+    },
+    {
+      "epoch": 0.06756984382895782,
+      "grad_norm": 6.692896453659439,
+      "learning_rate": 1.9999260705807097e-05,
+      "loss": 2.6376,
+      "step": 351
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "grad_norm": 5.343279283106529,
+      "learning_rate": 1.9999222307724112e-05,
+      "loss": 2.6047,
+      "step": 352
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.488,
+      "step": 352,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.4696,
+      "step": 352,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.3815,
+      "step": 352,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.3602,
+      "step": 352,
+      "vm_loss": 0.2399
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.5061,
+      "step": 352,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.4995,
+      "step": 352,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.6202,
+      "step": 352,
+      "vm_loss": 0.2258
+    },
+    {
+      "epoch": 0.06776235050653319,
+      "lm_loss": 2.2618,
+      "step": 352,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 0.06795485718410857,
+      "grad_norm": 4.913554407756347,
+      "learning_rate": 1.9999182937588793e-05,
+      "loss": 2.6339,
+      "step": 353
+    },
+    {
+      "epoch": 0.06814736386168395,
+      "grad_norm": 5.01951707989754,
+      "learning_rate": 1.9999142595404966e-05,
+      "loss": 2.6714,
+      "step": 354
+    },
+    {
+      "epoch": 0.06833987053925933,
+      "grad_norm": 5.629729224985602,
+      "learning_rate": 1.999910128117655e-05,
+      "loss": 2.6332,
+      "step": 355
+    },
+    {
+      "epoch": 0.06853237721683471,
+      "grad_norm": 6.026098965730194,
+      "learning_rate": 1.9999058994907564e-05,
+      "loss": 2.6521,
+      "step": 356
+    },
+    {
+      "epoch": 0.06872488389441009,
+      "grad_norm": 4.482636991401426,
+      "learning_rate": 1.9999015736602118e-05,
+      "loss": 2.6708,
+      "step": 357
+    },
+    {
+      "epoch": 0.06891739057198547,
+      "grad_norm": 4.728802487308409,
+      "learning_rate": 1.999897150626442e-05,
+      "loss": 2.6755,
+      "step": 358
+    },
+    {
+      "epoch": 0.06910989724956085,
+      "grad_norm": 4.421993026950499,
+      "learning_rate": 1.9998926303898766e-05,
+      "loss": 2.6556,
+      "step": 359
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "grad_norm": 4.3441646914180865,
+      "learning_rate": 1.999888012950955e-05,
+      "loss": 2.6744,
+      "step": 360
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.4258,
+      "step": 360,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.5005,
+      "step": 360,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.6775,
+      "step": 360,
+      "vm_loss": 0.2432
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.6832,
+      "step": 360,
+      "vm_loss": 0.1881
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.34,
+      "step": 360,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.4566,
+      "step": 360,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.5994,
+      "step": 360,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 0.06930240392713623,
+      "lm_loss": 2.3162,
+      "step": 360,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 0.06949491060471161,
+      "grad_norm": 5.2506038938294814,
+      "learning_rate": 1.9998832983101265e-05,
+      "loss": 2.6272,
+      "step": 361
+    },
+    {
+      "epoch": 0.06968741728228697,
+      "grad_norm": 5.284536186850513,
+      "learning_rate": 1.9998784864678492e-05,
+      "loss": 2.6672,
+      "step": 362
+    },
+    {
+      "epoch": 0.06987992395986235,
+      "grad_norm": 4.8286689622159455,
+      "learning_rate": 1.9998735774245906e-05,
+      "loss": 2.6614,
+      "step": 363
+    },
+    {
+      "epoch": 0.07007243063743773,
+      "grad_norm": 4.687684021072952,
+      "learning_rate": 1.9998685711808282e-05,
+      "loss": 2.6284,
+      "step": 364
+    },
+    {
+      "epoch": 0.07026493731501311,
+      "grad_norm": 4.1144995267156546,
+      "learning_rate": 1.9998634677370488e-05,
+      "loss": 2.7031,
+      "step": 365
+    },
+    {
+      "epoch": 0.07045744399258849,
+      "grad_norm": 4.292179436658936,
+      "learning_rate": 1.999858267093748e-05,
+      "loss": 2.6015,
+      "step": 366
+    },
+    {
+      "epoch": 0.07064995067016387,
+      "grad_norm": 4.274711940139827,
+      "learning_rate": 1.9998529692514318e-05,
+      "loss": 2.6082,
+      "step": 367
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "grad_norm": 4.6390138924938995,
+      "learning_rate": 1.9998475742106155e-05,
+      "loss": 2.6544,
+      "step": 368
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.0815,
+      "step": 368,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.2327,
+      "step": 368,
+      "vm_loss": 0.2462
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.3392,
+      "step": 368,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.7784,
+      "step": 368,
+      "vm_loss": 0.2694
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.6468,
+      "step": 368,
+      "vm_loss": 0.2823
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.2761,
+      "step": 368,
+      "vm_loss": 0.3097
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.2707,
+      "step": 368,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.07084245734773925,
+      "lm_loss": 2.7296,
+      "step": 368,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 0.07103496402531463,
+      "grad_norm": 4.603914347689331,
+      "learning_rate": 1.9998420819718228e-05,
+      "loss": 2.6714,
+      "step": 369
+    },
+    {
+      "epoch": 0.07122747070289001,
+      "grad_norm": 4.380669375370557,
+      "learning_rate": 1.999836492535588e-05,
+      "loss": 2.626,
+      "step": 370
+    },
+    {
+      "epoch": 0.07141997738046539,
+      "grad_norm": 5.403224556955136,
+      "learning_rate": 1.999830805902455e-05,
+      "loss": 2.6733,
+      "step": 371
+    },
+    {
+      "epoch": 0.07161248405804076,
+      "grad_norm": 4.636809687245444,
+      "learning_rate": 1.9998250220729757e-05,
+      "loss": 2.6269,
+      "step": 372
+    },
+    {
+      "epoch": 0.07180499073561614,
+      "grad_norm": 5.143458583937728,
+      "learning_rate": 1.999819141047713e-05,
+      "loss": 2.6357,
+      "step": 373
+    },
+    {
+      "epoch": 0.07199749741319152,
+      "grad_norm": 4.1139032835810845,
+      "learning_rate": 1.9998131628272383e-05,
+      "loss": 2.6459,
+      "step": 374
+    },
+    {
+      "epoch": 0.0721900040907669,
+      "grad_norm": 4.693406696889142,
+      "learning_rate": 1.9998070874121324e-05,
+      "loss": 2.6304,
+      "step": 375
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "grad_norm": 4.028661359851269,
+      "learning_rate": 1.999800914802987e-05,
+      "loss": 2.6701,
+      "step": 376
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.6267,
+      "step": 376,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.0724,
+      "step": 376,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.661,
+      "step": 376,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.3668,
+      "step": 376,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.7229,
+      "step": 376,
+      "vm_loss": 0.2276
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.6082,
+      "step": 376,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.5906,
+      "step": 376,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.07238251076834228,
+      "lm_loss": 2.4841,
+      "step": 376,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 0.07257501744591766,
+      "grad_norm": 4.189241594722147,
+      "learning_rate": 1.999794645000401e-05,
+      "loss": 2.6249,
+      "step": 377
+    },
+    {
+      "epoch": 0.07276752412349304,
+      "grad_norm": 4.2884548035205325,
+      "learning_rate": 1.9997882780049847e-05,
+      "loss": 2.6262,
+      "step": 378
+    },
+    {
+      "epoch": 0.07296003080106842,
+      "grad_norm": 4.259221190360824,
+      "learning_rate": 1.9997818138173564e-05,
+      "loss": 2.6289,
+      "step": 379
+    },
+    {
+      "epoch": 0.0731525374786438,
+      "grad_norm": 4.170838251896876,
+      "learning_rate": 1.999775252438145e-05,
+      "loss": 2.6766,
+      "step": 380
+    },
+    {
+      "epoch": 0.07334504415621917,
+      "grad_norm": 4.458199664716821,
+      "learning_rate": 1.999768593867988e-05,
+      "loss": 2.6342,
+      "step": 381
+    },
+    {
+      "epoch": 0.07353755083379454,
+      "grad_norm": 5.031529911214012,
+      "learning_rate": 1.9997618381075335e-05,
+      "loss": 2.6229,
+      "step": 382
+    },
+    {
+      "epoch": 0.07373005751136992,
+      "grad_norm": 4.122483247945086,
+      "learning_rate": 1.9997549851574372e-05,
+      "loss": 2.6343,
+      "step": 383
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "grad_norm": 4.238488014723586,
+      "learning_rate": 1.999748035018366e-05,
+      "loss": 2.6133,
+      "step": 384
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.6191,
+      "step": 384,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.5066,
+      "step": 384,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.6732,
+      "step": 384,
+      "vm_loss": 0.2354
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.4402,
+      "step": 384,
+      "vm_loss": 0.2476
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.368,
+      "step": 384,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.3874,
+      "step": 384,
+      "vm_loss": 0.2107
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.2578,
+      "step": 384,
+      "vm_loss": 0.1044
+    },
+    {
+      "epoch": 0.0739225641889453,
+      "lm_loss": 2.3363,
+      "step": 384,
+      "vm_loss": 0.1159
+    },
+    {
+      "epoch": 0.07411507086652068,
+      "grad_norm": 6.204975124102081,
+      "learning_rate": 1.999740987690995e-05,
+      "loss": 2.5971,
+      "step": 385
+    },
+    {
+      "epoch": 0.07430757754409606,
+      "grad_norm": 4.54092037575377,
+      "learning_rate": 1.9997338431760096e-05,
+      "loss": 2.6696,
+      "step": 386
+    },
+    {
+      "epoch": 0.07450008422167144,
+      "grad_norm": 5.842899742795652,
+      "learning_rate": 1.9997266014741043e-05,
+      "loss": 2.5951,
+      "step": 387
+    },
+    {
+      "epoch": 0.07469259089924682,
+      "grad_norm": 4.654436316884126,
+      "learning_rate": 1.9997192625859835e-05,
+      "loss": 2.6004,
+      "step": 388
+    },
+    {
+      "epoch": 0.0748850975768222,
+      "grad_norm": 4.966095790053132,
+      "learning_rate": 1.9997118265123597e-05,
+      "loss": 2.6338,
+      "step": 389
+    },
+    {
+      "epoch": 0.07507760425439758,
+      "grad_norm": 6.255048217829622,
+      "learning_rate": 1.9997042932539568e-05,
+      "loss": 2.6397,
+      "step": 390
+    },
+    {
+      "epoch": 0.07527011093197296,
+      "grad_norm": 5.110274557779918,
+      "learning_rate": 1.9996966628115065e-05,
+      "loss": 2.6717,
+      "step": 391
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "grad_norm": 5.020056293418851,
+      "learning_rate": 1.9996889351857505e-05,
+      "loss": 2.5814,
+      "step": 392
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.3172,
+      "step": 392,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.2725,
+      "step": 392,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.664,
+      "step": 392,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.4934,
+      "step": 392,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.6374,
+      "step": 392,
+      "vm_loss": 0.1812
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.2315,
+      "step": 392,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.5262,
+      "step": 392,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 0.07546261760954834,
+      "lm_loss": 2.2028,
+      "step": 392,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.0756551242871237,
+      "grad_norm": 6.239743206846836,
+      "learning_rate": 1.9996811103774402e-05,
+      "loss": 2.6338,
+      "step": 393
+    },
+    {
+      "epoch": 0.07584763096469908,
+      "grad_norm": 4.919758152236745,
+      "learning_rate": 1.9996731883873367e-05,
+      "loss": 2.6203,
+      "step": 394
+    },
+    {
+      "epoch": 0.07604013764227446,
+      "grad_norm": 5.911200849202589,
+      "learning_rate": 1.9996651692162094e-05,
+      "loss": 2.5953,
+      "step": 395
+    },
+    {
+      "epoch": 0.07623264431984984,
+      "grad_norm": 4.5688952866080745,
+      "learning_rate": 1.9996570528648386e-05,
+      "loss": 2.6177,
+      "step": 396
+    },
+    {
+      "epoch": 0.07642515099742522,
+      "grad_norm": 6.653621247777492,
+      "learning_rate": 1.9996488393340123e-05,
+      "loss": 2.5625,
+      "step": 397
+    },
+    {
+      "epoch": 0.0766176576750006,
+      "grad_norm": 5.109342299682795,
+      "learning_rate": 1.99964052862453e-05,
+      "loss": 2.6379,
+      "step": 398
+    },
+    {
+      "epoch": 0.07681016435257598,
+      "grad_norm": 5.1039939309649425,
+      "learning_rate": 1.9996321207371993e-05,
+      "loss": 2.6631,
+      "step": 399
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "grad_norm": 5.827058092260074,
+      "learning_rate": 1.999623615672837e-05,
+      "loss": 2.5796,
+      "step": 400
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.2977,
+      "step": 400,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.6482,
+      "step": 400,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.35,
+      "step": 400,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.3909,
+      "step": 400,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.5429,
+      "step": 400,
+      "vm_loss": 0.2058
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.4245,
+      "step": 400,
+      "vm_loss": 0.2627
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.2211,
+      "step": 400,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.07700267103015136,
+      "lm_loss": 2.4429,
+      "step": 400,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 0.07719517770772674,
+      "grad_norm": 3.9323240756023714,
+      "learning_rate": 1.9996150134322705e-05,
+      "loss": 2.5995,
+      "step": 401
+    },
+    {
+      "epoch": 0.07738768438530212,
+      "grad_norm": 5.374848687755518,
+      "learning_rate": 1.999606314016336e-05,
+      "loss": 2.5589,
+      "step": 402
+    },
+    {
+      "epoch": 0.07758019106287749,
+      "grad_norm": 5.859427418048798,
+      "learning_rate": 1.9995975174258792e-05,
+      "loss": 2.6392,
+      "step": 403
+    },
+    {
+      "epoch": 0.07777269774045287,
+      "grad_norm": 4.26420722417678,
+      "learning_rate": 1.9995886236617547e-05,
+      "loss": 2.6561,
+      "step": 404
+    },
+    {
+      "epoch": 0.07796520441802825,
+      "grad_norm": 4.72753101872864,
+      "learning_rate": 1.9995796327248277e-05,
+      "loss": 2.6288,
+      "step": 405
+    },
+    {
+      "epoch": 0.07815771109560363,
+      "grad_norm": 3.8915702259247116,
+      "learning_rate": 1.9995705446159724e-05,
+      "loss": 2.5802,
+      "step": 406
+    },
+    {
+      "epoch": 0.078350217773179,
+      "grad_norm": 4.800953125330857,
+      "learning_rate": 1.9995613593360716e-05,
+      "loss": 2.6144,
+      "step": 407
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "grad_norm": 4.942506031210902,
+      "learning_rate": 1.9995520768860185e-05,
+      "loss": 2.6788,
+      "step": 408
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.5765,
+      "step": 408,
+      "vm_loss": 0.2768
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.4396,
+      "step": 408,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.465,
+      "step": 408,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.4295,
+      "step": 408,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.4915,
+      "step": 408,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.3964,
+      "step": 408,
+      "vm_loss": 0.0995
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.2371,
+      "step": 408,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.07854272445075439,
+      "lm_loss": 2.342,
+      "step": 408,
+      "vm_loss": 0.3171
+    },
+    {
+      "epoch": 0.07873523112832977,
+      "grad_norm": 3.9198298213529337,
+      "learning_rate": 1.9995426972667157e-05,
+      "loss": 2.6314,
+      "step": 409
+    },
+    {
+      "epoch": 0.07892773780590515,
+      "grad_norm": 4.353713650508078,
+      "learning_rate": 1.9995332204790746e-05,
+      "loss": 2.5824,
+      "step": 410
+    },
+    {
+      "epoch": 0.07912024448348053,
+      "grad_norm": 5.289327370372657,
+      "learning_rate": 1.9995236465240168e-05,
+      "loss": 2.6272,
+      "step": 411
+    },
+    {
+      "epoch": 0.0793127511610559,
+      "grad_norm": 4.479585755033477,
+      "learning_rate": 1.999513975402473e-05,
+      "loss": 2.5951,
+      "step": 412
+    },
+    {
+      "epoch": 0.07950525783863127,
+      "grad_norm": 5.383623263190479,
+      "learning_rate": 1.9995042071153833e-05,
+      "loss": 2.645,
+      "step": 413
+    },
+    {
+      "epoch": 0.07969776451620665,
+      "grad_norm": 4.110562910059662,
+      "learning_rate": 1.999494341663697e-05,
+      "loss": 2.6094,
+      "step": 414
+    },
+    {
+      "epoch": 0.07989027119378203,
+      "grad_norm": 4.603499928423537,
+      "learning_rate": 1.9994843790483735e-05,
+      "loss": 2.6397,
+      "step": 415
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "grad_norm": 4.732775210366881,
+      "learning_rate": 1.9994743192703815e-05,
+      "loss": 2.6386,
+      "step": 416
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.3904,
+      "step": 416,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.3909,
+      "step": 416,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.7642,
+      "step": 416,
+      "vm_loss": 0.2751
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.3344,
+      "step": 416,
+      "vm_loss": 0.2419
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.3295,
+      "step": 416,
+      "vm_loss": 0.2019
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.4101,
+      "step": 416,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.4027,
+      "step": 416,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 0.08008277787135741,
+      "lm_loss": 2.3477,
+      "step": 416,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 0.08027528454893279,
+      "grad_norm": 4.017623192112295,
+      "learning_rate": 1.9994641623306984e-05,
+      "loss": 2.617,
+      "step": 417
+    },
+    {
+      "epoch": 0.08046779122650817,
+      "grad_norm": 3.8059314028452467,
+      "learning_rate": 1.9994539082303118e-05,
+      "loss": 2.6075,
+      "step": 418
+    },
+    {
+      "epoch": 0.08066029790408355,
+      "grad_norm": 4.329771878505528,
+      "learning_rate": 1.999443556970219e-05,
+      "loss": 2.6457,
+      "step": 419
+    },
+    {
+      "epoch": 0.08085280458165893,
+      "grad_norm": 4.471405981684093,
+      "learning_rate": 1.9994331085514255e-05,
+      "loss": 2.6095,
+      "step": 420
+    },
+    {
+      "epoch": 0.08104531125923431,
+      "grad_norm": 3.8388686285349407,
+      "learning_rate": 1.9994225629749475e-05,
+      "loss": 2.5931,
+      "step": 421
+    },
+    {
+      "epoch": 0.08123781793680969,
+      "grad_norm": 5.032789118058817,
+      "learning_rate": 1.99941192024181e-05,
+      "loss": 2.6351,
+      "step": 422
+    },
+    {
+      "epoch": 0.08143032461438506,
+      "grad_norm": 4.192509045251331,
+      "learning_rate": 1.9994011803530472e-05,
+      "loss": 2.6282,
+      "step": 423
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "grad_norm": 4.812130088917483,
+      "learning_rate": 1.999390343309704e-05,
+      "loss": 2.6337,
+      "step": 424
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.2247,
+      "step": 424,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.073,
+      "step": 424,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.4632,
+      "step": 424,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.4725,
+      "step": 424,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.7145,
+      "step": 424,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 1.8568,
+      "step": 424,
+      "vm_loss": 0.2437
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.5772,
+      "step": 424,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.08162283129196043,
+      "lm_loss": 2.4504,
+      "step": 424,
+      "vm_loss": 0.2325
+    },
+    {
+      "epoch": 0.08181533796953581,
+      "grad_norm": 4.972570940854288,
+      "learning_rate": 1.9993794091128336e-05,
+      "loss": 2.5861,
+      "step": 425
+    },
+    {
+      "epoch": 0.0820078446471112,
+      "grad_norm": 4.260208269673439,
+      "learning_rate": 1.9993683777634987e-05,
+      "loss": 2.6374,
+      "step": 426
+    },
+    {
+      "epoch": 0.08220035132468657,
+      "grad_norm": 5.126930289067559,
+      "learning_rate": 1.9993572492627716e-05,
+      "loss": 2.6622,
+      "step": 427
+    },
+    {
+      "epoch": 0.08239285800226195,
+      "grad_norm": 3.959883748126858,
+      "learning_rate": 1.9993460236117347e-05,
+      "loss": 2.5808,
+      "step": 428
+    },
+    {
+      "epoch": 0.08258536467983733,
+      "grad_norm": 4.136908618927918,
+      "learning_rate": 1.9993347008114788e-05,
+      "loss": 2.6109,
+      "step": 429
+    },
+    {
+      "epoch": 0.08277787135741271,
+      "grad_norm": 4.5427020652420085,
+      "learning_rate": 1.999323280863105e-05,
+      "loss": 2.6326,
+      "step": 430
+    },
+    {
+      "epoch": 0.0829703780349881,
+      "grad_norm": 4.829765682853637,
+      "learning_rate": 1.999311763767723e-05,
+      "loss": 2.5867,
+      "step": 431
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "grad_norm": 3.944981469899085,
+      "learning_rate": 1.9993001495264527e-05,
+      "loss": 2.5931,
+      "step": 432
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.3912,
+      "step": 432,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 1.8917,
+      "step": 432,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.0879,
+      "step": 432,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.5934,
+      "step": 432,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.4406,
+      "step": 432,
+      "vm_loss": 0.1189
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.4875,
+      "step": 432,
+      "vm_loss": 0.2012
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.5105,
+      "step": 432,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 0.08316288471256347,
+      "lm_loss": 2.7124,
+      "step": 432,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.08335539139013884,
+      "grad_norm": 5.376991364774965,
+      "learning_rate": 1.999288438140423e-05,
+      "loss": 2.5925,
+      "step": 433
+    },
+    {
+      "epoch": 0.08354789806771422,
+      "grad_norm": 4.012543783778699,
+      "learning_rate": 1.9992766296107726e-05,
+      "loss": 2.5671,
+      "step": 434
+    },
+    {
+      "epoch": 0.0837404047452896,
+      "grad_norm": 4.407267023707095,
+      "learning_rate": 1.999264723938649e-05,
+      "loss": 2.5952,
+      "step": 435
+    },
+    {
+      "epoch": 0.08393291142286498,
+      "grad_norm": 5.033969418794113,
+      "learning_rate": 1.9992527211252103e-05,
+      "loss": 2.585,
+      "step": 436
+    },
+    {
+      "epoch": 0.08412541810044036,
+      "grad_norm": 4.843202599271736,
+      "learning_rate": 1.9992406211716226e-05,
+      "loss": 2.6513,
+      "step": 437
+    },
+    {
+      "epoch": 0.08431792477801574,
+      "grad_norm": 4.809826868336139,
+      "learning_rate": 1.9992284240790628e-05,
+      "loss": 2.642,
+      "step": 438
+    },
+    {
+      "epoch": 0.08451043145559112,
+      "grad_norm": 5.064952269495695,
+      "learning_rate": 1.999216129848716e-05,
+      "loss": 2.5747,
+      "step": 439
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "grad_norm": 5.072105633259909,
+      "learning_rate": 1.999203738481778e-05,
+      "loss": 2.6159,
+      "step": 440
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.5864,
+      "step": 440,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.3824,
+      "step": 440,
+      "vm_loss": 0.2236
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.6471,
+      "step": 440,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.3837,
+      "step": 440,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.3542,
+      "step": 440,
+      "vm_loss": 0.3
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.7226,
+      "step": 440,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.4981,
+      "step": 440,
+      "vm_loss": 0.2234
+    },
+    {
+      "epoch": 0.0847029381331665,
+      "lm_loss": 2.3625,
+      "step": 440,
+      "vm_loss": 0.2238
+    },
+    {
+      "epoch": 0.08489544481074188,
+      "grad_norm": 4.378055239698308,
+      "learning_rate": 1.9991912499794532e-05,
+      "loss": 2.6005,
+      "step": 441
+    },
+    {
+      "epoch": 0.08508795148831726,
+      "grad_norm": 4.915084013954159,
+      "learning_rate": 1.9991786643429553e-05,
+      "loss": 2.5852,
+      "step": 442
+    },
+    {
+      "epoch": 0.08528045816589262,
+      "grad_norm": 5.52644319746219,
+      "learning_rate": 1.9991659815735075e-05,
+      "loss": 2.6014,
+      "step": 443
+    },
+    {
+      "epoch": 0.085472964843468,
+      "grad_norm": 5.299377509082517,
+      "learning_rate": 1.999153201672344e-05,
+      "loss": 2.5953,
+      "step": 444
+    },
+    {
+      "epoch": 0.08566547152104338,
+      "grad_norm": 5.304927636380555,
+      "learning_rate": 1.9991403246407056e-05,
+      "loss": 2.5809,
+      "step": 445
+    },
+    {
+      "epoch": 0.08585797819861876,
+      "grad_norm": 4.75314150552903,
+      "learning_rate": 1.9991273504798456e-05,
+      "loss": 2.5666,
+      "step": 446
+    },
+    {
+      "epoch": 0.08605048487619414,
+      "grad_norm": 4.380128205715259,
+      "learning_rate": 1.9991142791910243e-05,
+      "loss": 2.5593,
+      "step": 447
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "grad_norm": 4.318705685106545,
+      "learning_rate": 1.9991011107755126e-05,
+      "loss": 2.5956,
+      "step": 448
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.2428,
+      "step": 448,
+      "vm_loss": 0.228
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.4717,
+      "step": 448,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.4566,
+      "step": 448,
+      "vm_loss": 0.2012
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.3733,
+      "step": 448,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.3158,
+      "step": 448,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.4755,
+      "step": 448,
+      "vm_loss": 0.2357
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.1083,
+      "step": 448,
+      "vm_loss": 0.2557
+    },
+    {
+      "epoch": 0.08624299155376952,
+      "lm_loss": 2.395,
+      "step": 448,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 0.0864354982313449,
+      "grad_norm": 5.100189768112497,
+      "learning_rate": 1.999087845234591e-05,
+      "loss": 2.5839,
+      "step": 449
+    },
+    {
+      "epoch": 0.08662800490892028,
+      "grad_norm": 4.566942736823465,
+      "learning_rate": 1.9990744825695485e-05,
+      "loss": 2.6003,
+      "step": 450
+    },
+    {
+      "epoch": 0.08682051158649566,
+      "grad_norm": 4.67341629268445,
+      "learning_rate": 1.9990610227816845e-05,
+      "loss": 2.5668,
+      "step": 451
+    },
+    {
+      "epoch": 0.08701301826407104,
+      "grad_norm": 4.137873231298762,
+      "learning_rate": 1.9990474658723073e-05,
+      "loss": 2.611,
+      "step": 452
+    },
+    {
+      "epoch": 0.08720552494164642,
+      "grad_norm": 3.8051621215802642,
+      "learning_rate": 1.9990338118427347e-05,
+      "loss": 2.5889,
+      "step": 453
+    },
+    {
+      "epoch": 0.08739803161922179,
+      "grad_norm": 5.225626633998041,
+      "learning_rate": 1.9990200606942943e-05,
+      "loss": 2.6107,
+      "step": 454
+    },
+    {
+      "epoch": 0.08759053829679717,
+      "grad_norm": 4.0165052400732755,
+      "learning_rate": 1.9990062124283233e-05,
+      "loss": 2.591,
+      "step": 455
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "grad_norm": 4.661007634311046,
+      "learning_rate": 1.998992267046167e-05,
+      "loss": 2.6033,
+      "step": 456
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.5336,
+      "step": 456,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.3767,
+      "step": 456,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.164,
+      "step": 456,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.3462,
+      "step": 456,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.3675,
+      "step": 456,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.3259,
+      "step": 456,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.4258,
+      "step": 456,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 0.08778304497437255,
+      "lm_loss": 2.4685,
+      "step": 456,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 0.08797555165194793,
+      "grad_norm": 4.616366067523772,
+      "learning_rate": 1.9989782245491816e-05,
+      "loss": 2.5818,
+      "step": 457
+    },
+    {
+      "epoch": 0.0881680583295233,
+      "grad_norm": 4.6502671095176975,
+      "learning_rate": 1.9989640849387318e-05,
+      "loss": 2.566,
+      "step": 458
+    },
+    {
+      "epoch": 0.08836056500709868,
+      "grad_norm": 4.566747237580736,
+      "learning_rate": 1.9989498482161928e-05,
+      "loss": 2.5371,
+      "step": 459
+    },
+    {
+      "epoch": 0.08855307168467406,
+      "grad_norm": 4.5108812834080885,
+      "learning_rate": 1.9989355143829483e-05,
+      "loss": 2.5752,
+      "step": 460
+    },
+    {
+      "epoch": 0.08874557836224944,
+      "grad_norm": 4.33856611291057,
+      "learning_rate": 1.9989210834403915e-05,
+      "loss": 2.6607,
+      "step": 461
+    },
+    {
+      "epoch": 0.08893808503982482,
+      "grad_norm": 4.284488850525381,
+      "learning_rate": 1.9989065553899255e-05,
+      "loss": 2.5752,
+      "step": 462
+    },
+    {
+      "epoch": 0.0891305917174002,
+      "grad_norm": 4.597566654401899,
+      "learning_rate": 1.9988919302329626e-05,
+      "loss": 2.5872,
+      "step": 463
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "grad_norm": 4.123517766480121,
+      "learning_rate": 1.9988772079709244e-05,
+      "loss": 2.6068,
+      "step": 464
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.423,
+      "step": 464,
+      "vm_loss": 0.2284
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.4247,
+      "step": 464,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.526,
+      "step": 464,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.3393,
+      "step": 464,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.6121,
+      "step": 464,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.2647,
+      "step": 464,
+      "vm_loss": 0.2049
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.296,
+      "step": 464,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.08932309839497557,
+      "lm_loss": 2.4049,
+      "step": 464,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.08951560507255095,
+      "grad_norm": 4.992122515913401,
+      "learning_rate": 1.9988623886052425e-05,
+      "loss": 2.6111,
+      "step": 465
+    },
+    {
+      "epoch": 0.08970811175012633,
+      "grad_norm": 4.106694664525713,
+      "learning_rate": 1.998847472137357e-05,
+      "loss": 2.5963,
+      "step": 466
+    },
+    {
+      "epoch": 0.08990061842770171,
+      "grad_norm": 4.874698464927889,
+      "learning_rate": 1.998832458568718e-05,
+      "loss": 2.6133,
+      "step": 467
+    },
+    {
+      "epoch": 0.09009312510527709,
+      "grad_norm": 4.217876395217213,
+      "learning_rate": 1.9988173479007857e-05,
+      "loss": 2.5935,
+      "step": 468
+    },
+    {
+      "epoch": 0.09028563178285247,
+      "grad_norm": 3.6072137765551355,
+      "learning_rate": 1.9988021401350285e-05,
+      "loss": 2.5561,
+      "step": 469
+    },
+    {
+      "epoch": 0.09047813846042785,
+      "grad_norm": 4.334501750094067,
+      "learning_rate": 1.9987868352729247e-05,
+      "loss": 2.5523,
+      "step": 470
+    },
+    {
+      "epoch": 0.09067064513800323,
+      "grad_norm": 4.749684049791752,
+      "learning_rate": 1.9987714333159624e-05,
+      "loss": 2.5596,
+      "step": 471
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "grad_norm": 3.8240568308481566,
+      "learning_rate": 1.9987559342656387e-05,
+      "loss": 2.5928,
+      "step": 472
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.3947,
+      "step": 472,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.6004,
+      "step": 472,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.5982,
+      "step": 472,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.4629,
+      "step": 472,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.4183,
+      "step": 472,
+      "vm_loss": 0.2244
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.3785,
+      "step": 472,
+      "vm_loss": 0.2424
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.1391,
+      "step": 472,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 0.09086315181557861,
+      "lm_loss": 2.6103,
+      "step": 472,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 0.09105565849315399,
+      "grad_norm": 4.663153349680087,
+      "learning_rate": 1.9987403381234604e-05,
+      "loss": 2.6152,
+      "step": 473
+    },
+    {
+      "epoch": 0.09124816517072935,
+      "grad_norm": 4.162147228936526,
+      "learning_rate": 1.9987246448909437e-05,
+      "loss": 2.5616,
+      "step": 474
+    },
+    {
+      "epoch": 0.09144067184830473,
+      "grad_norm": 4.424916784288473,
+      "learning_rate": 1.998708854569614e-05,
+      "loss": 2.5772,
+      "step": 475
+    },
+    {
+      "epoch": 0.09163317852588011,
+      "grad_norm": 3.634107321885927,
+      "learning_rate": 1.998692967161006e-05,
+      "loss": 2.5544,
+      "step": 476
+    },
+    {
+      "epoch": 0.09182568520345549,
+      "grad_norm": 3.8471275085344097,
+      "learning_rate": 1.998676982666665e-05,
+      "loss": 2.5619,
+      "step": 477
+    },
+    {
+      "epoch": 0.09201819188103087,
+      "grad_norm": 4.2965501168743065,
+      "learning_rate": 1.9986609010881447e-05,
+      "loss": 2.5775,
+      "step": 478
+    },
+    {
+      "epoch": 0.09221069855860625,
+      "grad_norm": 3.74049091618792,
+      "learning_rate": 1.998644722427008e-05,
+      "loss": 2.5879,
+      "step": 479
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "grad_norm": 4.04950593906965,
+      "learning_rate": 1.998628446684828e-05,
+      "loss": 2.5497,
+      "step": 480
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.3726,
+      "step": 480,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.658,
+      "step": 480,
+      "vm_loss": 0.1125
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.2042,
+      "step": 480,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.2737,
+      "step": 480,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.56,
+      "step": 480,
+      "vm_loss": 0.2221
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.6275,
+      "step": 480,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.4803,
+      "step": 480,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.09240320523618163,
+      "lm_loss": 2.4612,
+      "step": 480,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 0.09259571191375701,
+      "grad_norm": 4.168779019699929,
+      "learning_rate": 1.9986120738631866e-05,
+      "loss": 2.579,
+      "step": 481
+    },
+    {
+      "epoch": 0.09278821859133239,
+      "grad_norm": 4.520248602112067,
+      "learning_rate": 1.9985956039636755e-05,
+      "loss": 2.6008,
+      "step": 482
+    },
+    {
+      "epoch": 0.09298072526890777,
+      "grad_norm": 4.76251337913576,
+      "learning_rate": 1.9985790369878963e-05,
+      "loss": 2.5414,
+      "step": 483
+    },
+    {
+      "epoch": 0.09317323194648314,
+      "grad_norm": 4.833530895653689,
+      "learning_rate": 1.998562372937459e-05,
+      "loss": 2.5452,
+      "step": 484
+    },
+    {
+      "epoch": 0.09336573862405852,
+      "grad_norm": 4.472736876071227,
+      "learning_rate": 1.9985456118139836e-05,
+      "loss": 2.5462,
+      "step": 485
+    },
+    {
+      "epoch": 0.0935582453016339,
+      "grad_norm": 4.168361170527035,
+      "learning_rate": 1.9985287536190996e-05,
+      "loss": 2.5638,
+      "step": 486
+    },
+    {
+      "epoch": 0.09375075197920928,
+      "grad_norm": 4.587043052267748,
+      "learning_rate": 1.998511798354446e-05,
+      "loss": 2.5742,
+      "step": 487
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "grad_norm": 3.8902232970242565,
+      "learning_rate": 1.9984947460216708e-05,
+      "loss": 2.6027,
+      "step": 488
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.2761,
+      "step": 488,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.3176,
+      "step": 488,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.4059,
+      "step": 488,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.4569,
+      "step": 488,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.1078,
+      "step": 488,
+      "vm_loss": 0.2286
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.3137,
+      "step": 488,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.3958,
+      "step": 488,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 0.09394325865678466,
+      "lm_loss": 2.6177,
+      "step": 488,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 0.09413576533436004,
+      "grad_norm": 3.789531341506008,
+      "learning_rate": 1.998477596622432e-05,
+      "loss": 2.5858,
+      "step": 489
+    },
+    {
+      "epoch": 0.09432827201193542,
+      "grad_norm": 4.721478304358618,
+      "learning_rate": 1.9984603501583963e-05,
+      "loss": 2.5388,
+      "step": 490
+    },
+    {
+      "epoch": 0.0945207786895108,
+      "grad_norm": 3.4771208257789743,
+      "learning_rate": 1.9984430066312407e-05,
+      "loss": 2.6125,
+      "step": 491
+    },
+    {
+      "epoch": 0.09471328536708618,
+      "grad_norm": 4.2134738863418315,
+      "learning_rate": 1.9984255660426508e-05,
+      "loss": 2.6002,
+      "step": 492
+    },
+    {
+      "epoch": 0.09490579204466155,
+      "grad_norm": 4.312397139640785,
+      "learning_rate": 1.9984080283943224e-05,
+      "loss": 2.5406,
+      "step": 493
+    },
+    {
+      "epoch": 0.09509829872223692,
+      "grad_norm": 4.394364013461965,
+      "learning_rate": 1.9983903936879604e-05,
+      "loss": 2.5462,
+      "step": 494
+    },
+    {
+      "epoch": 0.0952908053998123,
+      "grad_norm": 4.800268761098634,
+      "learning_rate": 1.9983726619252792e-05,
+      "loss": 2.5699,
+      "step": 495
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "grad_norm": 4.240062053830481,
+      "learning_rate": 1.998354833108002e-05,
+      "loss": 2.5816,
+      "step": 496
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.4354,
+      "step": 496,
+      "vm_loss": 0.2584
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.3032,
+      "step": 496,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.131,
+      "step": 496,
+      "vm_loss": 0.1004
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.4127,
+      "step": 496,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.6089,
+      "step": 496,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.3863,
+      "step": 496,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.5855,
+      "step": 496,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.09548331207738768,
+      "lm_loss": 2.3004,
+      "step": 496,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.09567581875496306,
+      "grad_norm": 4.447937835853206,
+      "learning_rate": 1.998336907237862e-05,
+      "loss": 2.5998,
+      "step": 497
+    },
+    {
+      "epoch": 0.09586832543253844,
+      "grad_norm": 4.824443198360054,
+      "learning_rate": 1.9983188843166028e-05,
+      "loss": 2.5347,
+      "step": 498
+    },
+    {
+      "epoch": 0.09606083211011382,
+      "grad_norm": 4.820549742018637,
+      "learning_rate": 1.9983007643459757e-05,
+      "loss": 2.5223,
+      "step": 499
+    },
+    {
+      "epoch": 0.0962533387876892,
+      "grad_norm": 5.713518136250507,
+      "learning_rate": 1.998282547327742e-05,
+      "loss": 2.572,
+      "step": 500
+    },
+    {
+      "epoch": 0.09644584546526458,
+      "grad_norm": 4.98156917712925,
+      "learning_rate": 1.998264233263673e-05,
+      "loss": 2.5726,
+      "step": 501
+    },
+    {
+      "epoch": 0.09663835214283996,
+      "grad_norm": 4.978553087117611,
+      "learning_rate": 1.9982458221555493e-05,
+      "loss": 2.5697,
+      "step": 502
+    },
+    {
+      "epoch": 0.09683085882041534,
+      "grad_norm": 4.663907646726513,
+      "learning_rate": 1.9982273140051602e-05,
+      "loss": 2.5794,
+      "step": 503
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "grad_norm": 4.095342225561139,
+      "learning_rate": 1.998208708814305e-05,
+      "loss": 2.6095,
+      "step": 504
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.2248,
+      "step": 504,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.2301,
+      "step": 504,
+      "vm_loss": 0.2263
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.56,
+      "step": 504,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.5923,
+      "step": 504,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.6183,
+      "step": 504,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.3742,
+      "step": 504,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.4218,
+      "step": 504,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 0.09702336549799072,
+      "lm_loss": 2.2415,
+      "step": 504,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.09721587217556608,
+      "grad_norm": 4.483308439993457,
+      "learning_rate": 1.9981900065847927e-05,
+      "loss": 2.5545,
+      "step": 505
+    },
+    {
+      "epoch": 0.09740837885314146,
+      "grad_norm": 4.593686805486022,
+      "learning_rate": 1.998171207318441e-05,
+      "loss": 2.5599,
+      "step": 506
+    },
+    {
+      "epoch": 0.09760088553071684,
+      "grad_norm": 4.312691222179744,
+      "learning_rate": 1.9981523110170778e-05,
+      "loss": 2.5523,
+      "step": 507
+    },
+    {
+      "epoch": 0.09779339220829222,
+      "grad_norm": 4.216721216018249,
+      "learning_rate": 1.9981333176825397e-05,
+      "loss": 2.5527,
+      "step": 508
+    },
+    {
+      "epoch": 0.0979858988858676,
+      "grad_norm": 4.602239657520409,
+      "learning_rate": 1.9981142273166735e-05,
+      "loss": 2.6171,
+      "step": 509
+    },
+    {
+      "epoch": 0.09817840556344298,
+      "grad_norm": 4.195913233258438,
+      "learning_rate": 1.9980950399213344e-05,
+      "loss": 2.5911,
+      "step": 510
+    },
+    {
+      "epoch": 0.09837091224101836,
+      "grad_norm": 4.753370484619016,
+      "learning_rate": 1.9980757554983884e-05,
+      "loss": 2.59,
+      "step": 511
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "grad_norm": 4.2637981024242935,
+      "learning_rate": 1.9980563740497096e-05,
+      "loss": 2.5816,
+      "step": 512
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 1.8557,
+      "step": 512,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.3957,
+      "step": 512,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.4068,
+      "step": 512,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.3137,
+      "step": 512,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.5145,
+      "step": 512,
+      "vm_loss": 0.1226
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.3018,
+      "step": 512,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.4041,
+      "step": 512,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.09856341891859374,
+      "lm_loss": 2.5427,
+      "step": 512,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.09875592559616912,
+      "grad_norm": 3.9858606369345804,
+      "learning_rate": 1.9980368955771824e-05,
+      "loss": 2.5272,
+      "step": 513
+    },
+    {
+      "epoch": 0.0989484322737445,
+      "grad_norm": 4.2679744558612525,
+      "learning_rate": 1.9980173200827004e-05,
+      "loss": 2.5449,
+      "step": 514
+    },
+    {
+      "epoch": 0.09914093895131987,
+      "grad_norm": 3.980811997729134,
+      "learning_rate": 1.9979976475681665e-05,
+      "loss": 2.5657,
+      "step": 515
+    },
+    {
+      "epoch": 0.09933344562889525,
+      "grad_norm": 4.642889480523902,
+      "learning_rate": 1.9979778780354935e-05,
+      "loss": 2.5096,
+      "step": 516
+    },
+    {
+      "epoch": 0.09952595230647063,
+      "grad_norm": 3.3432377538770037,
+      "learning_rate": 1.9979580114866022e-05,
+      "loss": 2.572,
+      "step": 517
+    },
+    {
+      "epoch": 0.09971845898404601,
+      "grad_norm": 4.3653302984121325,
+      "learning_rate": 1.9979380479234253e-05,
+      "loss": 2.5336,
+      "step": 518
+    },
+    {
+      "epoch": 0.09991096566162139,
+      "grad_norm": 4.272995148347355,
+      "learning_rate": 1.9979179873479023e-05,
+      "loss": 2.6058,
+      "step": 519
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "grad_norm": 4.221414867985027,
+      "learning_rate": 1.997897829761984e-05,
+      "loss": 2.5624,
+      "step": 520
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.5836,
+      "step": 520,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.1159,
+      "step": 520,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.4777,
+      "step": 520,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.4033,
+      "step": 520,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.1587,
+      "step": 520,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.5878,
+      "step": 520,
+      "vm_loss": 0.2251
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.5319,
+      "step": 520,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 0.10010347233919677,
+      "lm_loss": 2.5733,
+      "step": 520,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.10029597901677215,
+      "grad_norm": 4.6442495897955665,
+      "learning_rate": 1.99787757516763e-05,
+      "loss": 2.5459,
+      "step": 521
+    },
+    {
+      "epoch": 0.10048848569434753,
+      "grad_norm": 5.069089669449857,
+      "learning_rate": 1.9978572235668085e-05,
+      "loss": 2.544,
+      "step": 522
+    },
+    {
+      "epoch": 0.1006809923719229,
+      "grad_norm": 4.463885809538296,
+      "learning_rate": 1.997836774961499e-05,
+      "loss": 2.5587,
+      "step": 523
+    },
+    {
+      "epoch": 0.10087349904949829,
+      "grad_norm": 4.034778740896075,
+      "learning_rate": 1.997816229353689e-05,
+      "loss": 2.593,
+      "step": 524
+    },
+    {
+      "epoch": 0.10106600572707365,
+      "grad_norm": 4.187143655625298,
+      "learning_rate": 1.9977955867453754e-05,
+      "loss": 2.5303,
+      "step": 525
+    },
+    {
+      "epoch": 0.10125851240464903,
+      "grad_norm": 4.6057469976851335,
+      "learning_rate": 1.9977748471385654e-05,
+      "loss": 2.5759,
+      "step": 526
+    },
+    {
+      "epoch": 0.10145101908222441,
+      "grad_norm": 3.9742265378043165,
+      "learning_rate": 1.997754010535275e-05,
+      "loss": 2.5935,
+      "step": 527
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "grad_norm": 5.629343433645512,
+      "learning_rate": 1.9977330769375298e-05,
+      "loss": 2.594,
+      "step": 528
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.3092,
+      "step": 528,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.235,
+      "step": 528,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.1688,
+      "step": 528,
+      "vm_loss": 0.1011
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.789,
+      "step": 528,
+      "vm_loss": 0.1998
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.2814,
+      "step": 528,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.2993,
+      "step": 528,
+      "vm_loss": 0.2645
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.4847,
+      "step": 528,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.10164352575979979,
+      "lm_loss": 2.4784,
+      "step": 528,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.10183603243737517,
+      "grad_norm": 4.2954753814477185,
+      "learning_rate": 1.9977120463473647e-05,
+      "loss": 2.5763,
+      "step": 529
+    },
+    {
+      "epoch": 0.10202853911495055,
+      "grad_norm": 4.583368002012516,
+      "learning_rate": 1.9976909187668247e-05,
+      "loss": 2.5448,
+      "step": 530
+    },
+    {
+      "epoch": 0.10222104579252593,
+      "grad_norm": 4.543164436570776,
+      "learning_rate": 1.9976696941979625e-05,
+      "loss": 2.5663,
+      "step": 531
+    },
+    {
+      "epoch": 0.10241355247010131,
+      "grad_norm": 4.258242832977958,
+      "learning_rate": 1.9976483726428423e-05,
+      "loss": 2.5309,
+      "step": 532
+    },
+    {
+      "epoch": 0.10260605914767669,
+      "grad_norm": 4.936872883128827,
+      "learning_rate": 1.9976269541035365e-05,
+      "loss": 2.5791,
+      "step": 533
+    },
+    {
+      "epoch": 0.10279856582525207,
+      "grad_norm": 4.294882683273078,
+      "learning_rate": 1.997605438582128e-05,
+      "loss": 2.5756,
+      "step": 534
+    },
+    {
+      "epoch": 0.10299107250282744,
+      "grad_norm": 4.47189972091334,
+      "learning_rate": 1.9975838260807072e-05,
+      "loss": 2.5751,
+      "step": 535
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "grad_norm": 5.534645060445187,
+      "learning_rate": 1.9975621166013758e-05,
+      "loss": 2.5431,
+      "step": 536
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.4895,
+      "step": 536,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.4698,
+      "step": 536,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.4426,
+      "step": 536,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.5139,
+      "step": 536,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.5653,
+      "step": 536,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.3175,
+      "step": 536,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.585,
+      "step": 536,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.10318357918040282,
+      "lm_loss": 2.4231,
+      "step": 536,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 0.1033760858579782,
+      "grad_norm": 4.742741672338863,
+      "learning_rate": 1.997540310146244e-05,
+      "loss": 2.6073,
+      "step": 537
+    },
+    {
+      "epoch": 0.10356859253555357,
+      "grad_norm": 4.7129243689926925,
+      "learning_rate": 1.997518406717432e-05,
+      "loss": 2.5483,
+      "step": 538
+    },
+    {
+      "epoch": 0.10376109921312895,
+      "grad_norm": 5.111649367214757,
+      "learning_rate": 1.997496406317069e-05,
+      "loss": 2.5007,
+      "step": 539
+    },
+    {
+      "epoch": 0.10395360589070433,
+      "grad_norm": 5.272156736996118,
+      "learning_rate": 1.9974743089472935e-05,
+      "loss": 2.5842,
+      "step": 540
+    },
+    {
+      "epoch": 0.10414611256827971,
+      "grad_norm": 4.3925016569700155,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 2.5717,
+      "step": 541
+    },
+    {
+      "epoch": 0.1043386192458551,
+      "grad_norm": 5.31383252091099,
+      "learning_rate": 1.9974298233081072e-05,
+      "loss": 2.5776,
+      "step": 542
+    },
+    {
+      "epoch": 0.10453112592343047,
+      "grad_norm": 4.395161100779029,
+      "learning_rate": 1.997407435043021e-05,
+      "loss": 2.5521,
+      "step": 543
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "grad_norm": 5.2056347769229525,
+      "learning_rate": 1.9973849498171717e-05,
+      "loss": 2.5304,
+      "step": 544
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.303,
+      "step": 544,
+      "vm_loss": 0.1359
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.3496,
+      "step": 544,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.3177,
+      "step": 544,
+      "vm_loss": 0.2901
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 1.9506,
+      "step": 544,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.4513,
+      "step": 544,
+      "vm_loss": 0.2269
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.3186,
+      "step": 544,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.3219,
+      "step": 544,
+      "vm_loss": 0.2259
+    },
+    {
+      "epoch": 0.10472363260100585,
+      "lm_loss": 2.381,
+      "step": 544,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 0.10491613927858122,
+      "grad_norm": 4.452461920680797,
+      "learning_rate": 1.9973623676327447e-05,
+      "loss": 2.5622,
+      "step": 545
+    },
+    {
+      "epoch": 0.1051086459561566,
+      "grad_norm": 4.538593463208113,
+      "learning_rate": 1.997339688491936e-05,
+      "loss": 2.6382,
+      "step": 546
+    },
+    {
+      "epoch": 0.10530115263373198,
+      "grad_norm": 5.829048104888673,
+      "learning_rate": 1.9973169123969495e-05,
+      "loss": 2.5312,
+      "step": 547
+    },
+    {
+      "epoch": 0.10549365931130736,
+      "grad_norm": 4.0765426592671,
+      "learning_rate": 1.9972940393499997e-05,
+      "loss": 2.5475,
+      "step": 548
+    },
+    {
+      "epoch": 0.10568616598888274,
+      "grad_norm": 4.888477649510067,
+      "learning_rate": 1.9972710693533105e-05,
+      "loss": 2.5142,
+      "step": 549
+    },
+    {
+      "epoch": 0.10587867266645812,
+      "grad_norm": 4.364645119237847,
+      "learning_rate": 1.9972480024091143e-05,
+      "loss": 2.5429,
+      "step": 550
+    },
+    {
+      "epoch": 0.1060711793440335,
+      "grad_norm": 4.554702935944056,
+      "learning_rate": 1.9972248385196537e-05,
+      "loss": 2.5523,
+      "step": 551
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "grad_norm": 4.562261967753434,
+      "learning_rate": 1.9972015776871804e-05,
+      "loss": 2.5612,
+      "step": 552
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.2495,
+      "step": 552,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.1304,
+      "step": 552,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.4615,
+      "step": 552,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.2341,
+      "step": 552,
+      "vm_loss": 0.2755
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.2722,
+      "step": 552,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.5001,
+      "step": 552,
+      "vm_loss": 0.233
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 2.1186,
+      "step": 552,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.10626368602160888,
+      "lm_loss": 1.8815,
+      "step": 552,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.10645619269918426,
+      "grad_norm": 4.029308585651912,
+      "learning_rate": 1.9971782199139565e-05,
+      "loss": 2.5253,
+      "step": 553
+    },
+    {
+      "epoch": 0.10664869937675964,
+      "grad_norm": 4.617136440972574,
+      "learning_rate": 1.9971547652022512e-05,
+      "loss": 2.5554,
+      "step": 554
+    },
+    {
+      "epoch": 0.106841206054335,
+      "grad_norm": 4.451605206367319,
+      "learning_rate": 1.9971312135543458e-05,
+      "loss": 2.5417,
+      "step": 555
+    },
+    {
+      "epoch": 0.10703371273191038,
+      "grad_norm": 3.8717236897340293,
+      "learning_rate": 1.9971075649725293e-05,
+      "loss": 2.5716,
+      "step": 556
+    },
+    {
+      "epoch": 0.10722621940948576,
+      "grad_norm": 4.904277182359794,
+      "learning_rate": 1.997083819459101e-05,
+      "loss": 2.546,
+      "step": 557
+    },
+    {
+      "epoch": 0.10741872608706114,
+      "grad_norm": 4.588915419700571,
+      "learning_rate": 1.9970599770163686e-05,
+      "loss": 2.5618,
+      "step": 558
+    },
+    {
+      "epoch": 0.10761123276463652,
+      "grad_norm": 4.54865545281532,
+      "learning_rate": 1.9970360376466506e-05,
+      "loss": 2.5528,
+      "step": 559
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "grad_norm": 5.237608747956109,
+      "learning_rate": 1.9970120013522737e-05,
+      "loss": 2.5461,
+      "step": 560
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.2993,
+      "step": 560,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.0733,
+      "step": 560,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.0348,
+      "step": 560,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.3537,
+      "step": 560,
+      "vm_loss": 0.2592
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.2235,
+      "step": 560,
+      "vm_loss": 0.1084
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.2353,
+      "step": 560,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.2617,
+      "step": 560,
+      "vm_loss": 0.2525
+    },
+    {
+      "epoch": 0.1078037394422119,
+      "lm_loss": 2.2282,
+      "step": 560,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.10799624611978728,
+      "grad_norm": 4.155256976357799,
+      "learning_rate": 1.996987868135575e-05,
+      "loss": 2.5055,
+      "step": 561
+    },
+    {
+      "epoch": 0.10818875279736266,
+      "grad_norm": 5.164610891613552,
+      "learning_rate": 1.9969636379989e-05,
+      "loss": 2.5875,
+      "step": 562
+    },
+    {
+      "epoch": 0.10838125947493804,
+      "grad_norm": 5.18307245733418,
+      "learning_rate": 1.9969393109446052e-05,
+      "loss": 2.5472,
+      "step": 563
+    },
+    {
+      "epoch": 0.10857376615251342,
+      "grad_norm": 5.362741885855717,
+      "learning_rate": 1.996914886975054e-05,
+      "loss": 2.5797,
+      "step": 564
+    },
+    {
+      "epoch": 0.1087662728300888,
+      "grad_norm": 4.295717857849304,
+      "learning_rate": 1.996890366092622e-05,
+      "loss": 2.4908,
+      "step": 565
+    },
+    {
+      "epoch": 0.10895877950766417,
+      "grad_norm": 4.491151245485607,
+      "learning_rate": 1.9968657482996923e-05,
+      "loss": 2.5713,
+      "step": 566
+    },
+    {
+      "epoch": 0.10915128618523955,
+      "grad_norm": 3.5870286838918948,
+      "learning_rate": 1.996841033598658e-05,
+      "loss": 2.5951,
+      "step": 567
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "grad_norm": 5.530468277715157,
+      "learning_rate": 1.9968162219919222e-05,
+      "loss": 2.5357,
+      "step": 568
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.2972,
+      "step": 568,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.4349,
+      "step": 568,
+      "vm_loss": 0.2109
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.6446,
+      "step": 568,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.1997,
+      "step": 568,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.2931,
+      "step": 568,
+      "vm_loss": 0.225
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.4268,
+      "step": 568,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.4335,
+      "step": 568,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.10934379286281493,
+      "lm_loss": 2.517,
+      "step": 568,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 0.1095362995403903,
+      "grad_norm": 4.306208266188474,
+      "learning_rate": 1.996791313481897e-05,
+      "loss": 2.5886,
+      "step": 569
+    },
+    {
+      "epoch": 0.10972880621796569,
+      "grad_norm": 4.84155708381945,
+      "learning_rate": 1.9967663080710026e-05,
+      "loss": 2.5134,
+      "step": 570
+    },
+    {
+      "epoch": 0.10992131289554107,
+      "grad_norm": 4.450894152483306,
+      "learning_rate": 1.9967412057616713e-05,
+      "loss": 2.507,
+      "step": 571
+    },
+    {
+      "epoch": 0.11011381957311644,
+      "grad_norm": 5.182548491387447,
+      "learning_rate": 1.9967160065563425e-05,
+      "loss": 2.5399,
+      "step": 572
+    },
+    {
+      "epoch": 0.11030632625069182,
+      "grad_norm": 4.405104545059833,
+      "learning_rate": 1.9966907104574663e-05,
+      "loss": 2.562,
+      "step": 573
+    },
+    {
+      "epoch": 0.1104988329282672,
+      "grad_norm": 5.0551007934908485,
+      "learning_rate": 1.9966653174675014e-05,
+      "loss": 2.5621,
+      "step": 574
+    },
+    {
+      "epoch": 0.11069133960584258,
+      "grad_norm": 4.499112987514951,
+      "learning_rate": 1.996639827588917e-05,
+      "loss": 2.5634,
+      "step": 575
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "grad_norm": 4.4879177165491315,
+      "learning_rate": 1.99661424082419e-05,
+      "loss": 2.4971,
+      "step": 576
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.4575,
+      "step": 576,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.0256,
+      "step": 576,
+      "vm_loss": 0.0832
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.5426,
+      "step": 576,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.3797,
+      "step": 576,
+      "vm_loss": 0.1893
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.5603,
+      "step": 576,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 1.9635,
+      "step": 576,
+      "vm_loss": 0.1369
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.4542,
+      "step": 576,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.11088384628341795,
+      "lm_loss": 2.2695,
+      "step": 576,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 0.11107635296099333,
+      "grad_norm": 4.4604461446502555,
+      "learning_rate": 1.996588557175809e-05,
+      "loss": 2.5108,
+      "step": 577
+    },
+    {
+      "epoch": 0.11126885963856871,
+      "grad_norm": 3.6797171128629524,
+      "learning_rate": 1.9965627766462697e-05,
+      "loss": 2.5279,
+      "step": 578
+    },
+    {
+      "epoch": 0.11146136631614409,
+      "grad_norm": 3.9133489619595583,
+      "learning_rate": 1.996536899238079e-05,
+      "loss": 2.5544,
+      "step": 579
+    },
+    {
+      "epoch": 0.11165387299371947,
+      "grad_norm": 4.376642195522719,
+      "learning_rate": 1.9965109249537523e-05,
+      "loss": 2.5474,
+      "step": 580
+    },
+    {
+      "epoch": 0.11184637967129485,
+      "grad_norm": 3.8807954406788885,
+      "learning_rate": 1.9964848537958143e-05,
+      "loss": 2.5441,
+      "step": 581
+    },
+    {
+      "epoch": 0.11203888634887023,
+      "grad_norm": 4.443795091187734,
+      "learning_rate": 1.9964586857668002e-05,
+      "loss": 2.4777,
+      "step": 582
+    },
+    {
+      "epoch": 0.11223139302644561,
+      "grad_norm": 4.1714757307448815,
+      "learning_rate": 1.996432420869253e-05,
+      "loss": 2.5991,
+      "step": 583
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "grad_norm": 4.0765940747119185,
+      "learning_rate": 1.996406059105726e-05,
+      "loss": 2.537,
+      "step": 584
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.4565,
+      "step": 584,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.2496,
+      "step": 584,
+      "vm_loss": 0.2172
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.2324,
+      "step": 584,
+      "vm_loss": 0.2361
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.4206,
+      "step": 584,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.0797,
+      "step": 584,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.4471,
+      "step": 584,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 1.8408,
+      "step": 584,
+      "vm_loss": 0.2423
+    },
+    {
+      "epoch": 0.11242389970402099,
+      "lm_loss": 2.0841,
+      "step": 584,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 0.11261640638159637,
+      "grad_norm": 4.56036467020508,
+      "learning_rate": 1.996379600478783e-05,
+      "loss": 2.4848,
+      "step": 585
+    },
+    {
+      "epoch": 0.11280891305917173,
+      "grad_norm": 4.387155945380377,
+      "learning_rate": 1.9963530449909954e-05,
+      "loss": 2.5938,
+      "step": 586
+    },
+    {
+      "epoch": 0.11300141973674711,
+      "grad_norm": 3.4866136531740146,
+      "learning_rate": 1.9963263926449446e-05,
+      "loss": 2.5449,
+      "step": 587
+    },
+    {
+      "epoch": 0.1131939264143225,
+      "grad_norm": 5.749970050845133,
+      "learning_rate": 1.9962996434432215e-05,
+      "loss": 2.5848,
+      "step": 588
+    },
+    {
+      "epoch": 0.11338643309189787,
+      "grad_norm": 4.3566581971219,
+      "learning_rate": 1.996272797388427e-05,
+      "loss": 2.558,
+      "step": 589
+    },
+    {
+      "epoch": 0.11357893976947325,
+      "grad_norm": 4.755457770589791,
+      "learning_rate": 1.9962458544831703e-05,
+      "loss": 2.5339,
+      "step": 590
+    },
+    {
+      "epoch": 0.11377144644704863,
+      "grad_norm": 4.487539284555335,
+      "learning_rate": 1.996218814730071e-05,
+      "loss": 2.571,
+      "step": 591
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "grad_norm": 4.858345835334981,
+      "learning_rate": 1.9961916781317576e-05,
+      "loss": 2.5097,
+      "step": 592
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.1724,
+      "step": 592,
+      "vm_loss": 0.1132
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.1021,
+      "step": 592,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.1455,
+      "step": 592,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.603,
+      "step": 592,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.3863,
+      "step": 592,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.4888,
+      "step": 592,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.3053,
+      "step": 592,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 0.11396395312462401,
+      "lm_loss": 2.1713,
+      "step": 592,
+      "vm_loss": 0.249
+    },
+    {
+      "epoch": 0.11415645980219939,
+      "grad_norm": 3.987979843075841,
+      "learning_rate": 1.996164444690868e-05,
+      "loss": 2.531,
+      "step": 593
+    },
+    {
+      "epoch": 0.11434896647977477,
+      "grad_norm": 4.176283039869517,
+      "learning_rate": 1.9961371144100497e-05,
+      "loss": 2.5312,
+      "step": 594
+    },
+    {
+      "epoch": 0.11454147315735015,
+      "grad_norm": 5.447252848070803,
+      "learning_rate": 1.9961096872919596e-05,
+      "loss": 2.5583,
+      "step": 595
+    },
+    {
+      "epoch": 0.11473397983492552,
+      "grad_norm": 5.334194807974677,
+      "learning_rate": 1.9960821633392635e-05,
+      "loss": 2.4812,
+      "step": 596
+    },
+    {
+      "epoch": 0.1149264865125009,
+      "grad_norm": 6.811477152548263,
+      "learning_rate": 1.996054542554638e-05,
+      "loss": 2.5601,
+      "step": 597
+    },
+    {
+      "epoch": 0.11511899319007628,
+      "grad_norm": 4.334357344693945,
+      "learning_rate": 1.9960268249407674e-05,
+      "loss": 2.5511,
+      "step": 598
+    },
+    {
+      "epoch": 0.11531149986765166,
+      "grad_norm": 5.336792549265995,
+      "learning_rate": 1.9959990105003467e-05,
+      "loss": 2.5667,
+      "step": 599
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "grad_norm": 5.893389448387921,
+      "learning_rate": 1.9959710992360798e-05,
+      "loss": 2.5293,
+      "step": 600
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.5349,
+      "step": 600,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.5449,
+      "step": 600,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.2599,
+      "step": 600,
+      "vm_loss": 0.256
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.4555,
+      "step": 600,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.2912,
+      "step": 600,
+      "vm_loss": 0.2175
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.3837,
+      "step": 600,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.3388,
+      "step": 600,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.11550400654522704,
+      "lm_loss": 2.1978,
+      "step": 600,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 0.11569651322280242,
+      "grad_norm": 5.53391451210367,
+      "learning_rate": 1.9959430911506797e-05,
+      "loss": 2.5925,
+      "step": 601
+    },
+    {
+      "epoch": 0.1158890199003778,
+      "grad_norm": 4.656257247769798,
+      "learning_rate": 1.995914986246869e-05,
+      "loss": 2.5165,
+      "step": 602
+    },
+    {
+      "epoch": 0.11608152657795318,
+      "grad_norm": 6.197808096987862,
+      "learning_rate": 1.9958867845273805e-05,
+      "loss": 2.5937,
+      "step": 603
+    },
+    {
+      "epoch": 0.11627403325552856,
+      "grad_norm": 5.017573613725394,
+      "learning_rate": 1.9958584859949558e-05,
+      "loss": 2.5917,
+      "step": 604
+    },
+    {
+      "epoch": 0.11646653993310394,
+      "grad_norm": 4.739787351976849,
+      "learning_rate": 1.995830090652345e-05,
+      "loss": 2.5321,
+      "step": 605
+    },
+    {
+      "epoch": 0.1166590466106793,
+      "grad_norm": 4.252525284448323,
+      "learning_rate": 1.995801598502309e-05,
+      "loss": 2.5554,
+      "step": 606
+    },
+    {
+      "epoch": 0.11685155328825468,
+      "grad_norm": 4.356549863905389,
+      "learning_rate": 1.9957730095476177e-05,
+      "loss": 2.5117,
+      "step": 607
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "grad_norm": 5.045659854669538,
+      "learning_rate": 1.9957443237910502e-05,
+      "loss": 2.552,
+      "step": 608
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.526,
+      "step": 608,
+      "vm_loss": 0.2532
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.2053,
+      "step": 608,
+      "vm_loss": 0.2431
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.4523,
+      "step": 608,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.7507,
+      "step": 608,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.4816,
+      "step": 608,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.2803,
+      "step": 608,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.3579,
+      "step": 608,
+      "vm_loss": 0.1043
+    },
+    {
+      "epoch": 0.11704405996583006,
+      "lm_loss": 2.283,
+      "step": 608,
+      "vm_loss": 0.2349
+    },
+    {
+      "epoch": 0.11723656664340544,
+      "grad_norm": 4.073958335085406,
+      "learning_rate": 1.995715541235395e-05,
+      "loss": 2.5839,
+      "step": 609
+    },
+    {
+      "epoch": 0.11742907332098082,
+      "grad_norm": 4.043061377223294,
+      "learning_rate": 1.9956866618834508e-05,
+      "loss": 2.5649,
+      "step": 610
+    },
+    {
+      "epoch": 0.1176215799985562,
+      "grad_norm": 4.438789195059056,
+      "learning_rate": 1.995657685738024e-05,
+      "loss": 2.4747,
+      "step": 611
+    },
+    {
+      "epoch": 0.11781408667613158,
+      "grad_norm": 4.190315304510294,
+      "learning_rate": 1.995628612801932e-05,
+      "loss": 2.5526,
+      "step": 612
+    },
+    {
+      "epoch": 0.11800659335370696,
+      "grad_norm": 4.046238216487343,
+      "learning_rate": 1.9955994430780012e-05,
+      "loss": 2.4863,
+      "step": 613
+    },
+    {
+      "epoch": 0.11819910003128234,
+      "grad_norm": 5.132396838070412,
+      "learning_rate": 1.9955701765690668e-05,
+      "loss": 2.5498,
+      "step": 614
+    },
+    {
+      "epoch": 0.11839160670885772,
+      "grad_norm": 3.9392603672246485,
+      "learning_rate": 1.9955408132779747e-05,
+      "loss": 2.5735,
+      "step": 615
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "grad_norm": 5.46153809828257,
+      "learning_rate": 1.9955113532075784e-05,
+      "loss": 2.5266,
+      "step": 616
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.1994,
+      "step": 616,
+      "vm_loss": 0.187
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.1457,
+      "step": 616,
+      "vm_loss": 0.184
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.1226,
+      "step": 616,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.1768,
+      "step": 616,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.6253,
+      "step": 616,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.3991,
+      "step": 616,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.3839,
+      "step": 616,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.1185841133864331,
+      "lm_loss": 2.5259,
+      "step": 616,
+      "vm_loss": 0.1553
+    },
+    {
+      "epoch": 0.11877662006400846,
+      "grad_norm": 4.083237404668489,
+      "learning_rate": 1.9954817963607426e-05,
+      "loss": 2.5464,
+      "step": 617
+    },
+    {
+      "epoch": 0.11896912674158384,
+      "grad_norm": 5.474657002245904,
+      "learning_rate": 1.99545214274034e-05,
+      "loss": 2.5376,
+      "step": 618
+    },
+    {
+      "epoch": 0.11916163341915922,
+      "grad_norm": 4.602705189412951,
+      "learning_rate": 1.9954223923492537e-05,
+      "loss": 2.5094,
+      "step": 619
+    },
+    {
+      "epoch": 0.1193541400967346,
+      "grad_norm": 5.151232392435146,
+      "learning_rate": 1.9953925451903757e-05,
+      "loss": 2.5739,
+      "step": 620
+    },
+    {
+      "epoch": 0.11954664677430998,
+      "grad_norm": 4.846012794073122,
+      "learning_rate": 1.9953626012666073e-05,
+      "loss": 2.5261,
+      "step": 621
+    },
+    {
+      "epoch": 0.11973915345188536,
+      "grad_norm": 4.762067948465738,
+      "learning_rate": 1.99533256058086e-05,
+      "loss": 2.5525,
+      "step": 622
+    },
+    {
+      "epoch": 0.11993166012946074,
+      "grad_norm": 4.450505712860721,
+      "learning_rate": 1.9953024231360537e-05,
+      "loss": 2.4879,
+      "step": 623
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "grad_norm": 3.8589132116601914,
+      "learning_rate": 1.995272188935118e-05,
+      "loss": 2.5234,
+      "step": 624
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.5136,
+      "step": 624,
+      "vm_loss": 0.1301
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.4452,
+      "step": 624,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.1829,
+      "step": 624,
+      "vm_loss": 0.2665
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.1713,
+      "step": 624,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.4625,
+      "step": 624,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.3332,
+      "step": 624,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.2857,
+      "step": 624,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.12012416680703612,
+      "lm_loss": 2.7094,
+      "step": 624,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 0.1203166734846115,
+      "grad_norm": 5.487828235063577,
+      "learning_rate": 1.9952418579809927e-05,
+      "loss": 2.5577,
+      "step": 625
+    },
+    {
+      "epoch": 0.12050918016218688,
+      "grad_norm": 4.119771714153317,
+      "learning_rate": 1.9952114302766255e-05,
+      "loss": 2.5191,
+      "step": 626
+    },
+    {
+      "epoch": 0.12070168683976225,
+      "grad_norm": 4.457015090525097,
+      "learning_rate": 1.995180905824975e-05,
+      "loss": 2.492,
+      "step": 627
+    },
+    {
+      "epoch": 0.12089419351733763,
+      "grad_norm": 3.936169323856248,
+      "learning_rate": 1.995150284629008e-05,
+      "loss": 2.5349,
+      "step": 628
+    },
+    {
+      "epoch": 0.12108670019491301,
+      "grad_norm": 4.887220589290658,
+      "learning_rate": 1.9951195666917025e-05,
+      "loss": 2.5377,
+      "step": 629
+    },
+    {
+      "epoch": 0.12127920687248839,
+      "grad_norm": 5.273023501704847,
+      "learning_rate": 1.9950887520160433e-05,
+      "loss": 2.5603,
+      "step": 630
+    },
+    {
+      "epoch": 0.12147171355006377,
+      "grad_norm": 4.766747051272953,
+      "learning_rate": 1.9950578406050263e-05,
+      "loss": 2.5186,
+      "step": 631
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "grad_norm": 5.756125396693531,
+      "learning_rate": 1.995026832461657e-05,
+      "loss": 2.4871,
+      "step": 632
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.3456,
+      "step": 632,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.5095,
+      "step": 632,
+      "vm_loss": 0.2327
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.155,
+      "step": 632,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.4375,
+      "step": 632,
+      "vm_loss": 0.1977
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.4639,
+      "step": 632,
+      "vm_loss": 0.2188
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.0473,
+      "step": 632,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.4015,
+      "step": 632,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 0.12166422022763915,
+      "lm_loss": 2.4911,
+      "step": 632,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.12185672690521453,
+      "grad_norm": 6.055192025831539,
+      "learning_rate": 1.9949957275889496e-05,
+      "loss": 2.5765,
+      "step": 633
+    },
+    {
+      "epoch": 0.1220492335827899,
+      "grad_norm": 5.737076872236404,
+      "learning_rate": 1.9949645259899272e-05,
+      "loss": 2.4851,
+      "step": 634
+    },
+    {
+      "epoch": 0.12224174026036529,
+      "grad_norm": 4.772299580492336,
+      "learning_rate": 1.994933227667624e-05,
+      "loss": 2.4688,
+      "step": 635
+    },
+    {
+      "epoch": 0.12243424693794067,
+      "grad_norm": 5.643288397928099,
+      "learning_rate": 1.994901832625082e-05,
+      "loss": 2.5595,
+      "step": 636
+    },
+    {
+      "epoch": 0.12262675361551603,
+      "grad_norm": 4.039098871752698,
+      "learning_rate": 1.9948703408653535e-05,
+      "loss": 2.57,
+      "step": 637
+    },
+    {
+      "epoch": 0.12281926029309141,
+      "grad_norm": 4.9711410623891465,
+      "learning_rate": 1.9948387523915e-05,
+      "loss": 2.5426,
+      "step": 638
+    },
+    {
+      "epoch": 0.12301176697066679,
+      "grad_norm": 4.245752283373815,
+      "learning_rate": 1.9948070672065916e-05,
+      "loss": 2.5143,
+      "step": 639
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "grad_norm": 4.509170010482479,
+      "learning_rate": 1.994775285313709e-05,
+      "loss": 2.5062,
+      "step": 640
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.6645,
+      "step": 640,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.3692,
+      "step": 640,
+      "vm_loss": 0.2272
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.458,
+      "step": 640,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.3372,
+      "step": 640,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.3324,
+      "step": 640,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.4398,
+      "step": 640,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.1126,
+      "step": 640,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.12320427364824217,
+      "lm_loss": 2.4645,
+      "step": 640,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 0.12339678032581755,
+      "grad_norm": 4.553496428023007,
+      "learning_rate": 1.994743406715942e-05,
+      "loss": 2.5524,
+      "step": 641
+    },
+    {
+      "epoch": 0.12358928700339293,
+      "grad_norm": 3.8853493766983913,
+      "learning_rate": 1.9947114314163892e-05,
+      "loss": 2.5323,
+      "step": 642
+    },
+    {
+      "epoch": 0.12378179368096831,
+      "grad_norm": 5.3555711966836155,
+      "learning_rate": 1.9946793594181592e-05,
+      "loss": 2.5207,
+      "step": 643
+    },
+    {
+      "epoch": 0.12397430035854369,
+      "grad_norm": 3.9093572886523913,
+      "learning_rate": 1.99464719072437e-05,
+      "loss": 2.5258,
+      "step": 644
+    },
+    {
+      "epoch": 0.12416680703611907,
+      "grad_norm": 4.019594269327376,
+      "learning_rate": 1.9946149253381483e-05,
+      "loss": 2.5043,
+      "step": 645
+    },
+    {
+      "epoch": 0.12435931371369445,
+      "grad_norm": 4.493105766305279,
+      "learning_rate": 1.9945825632626313e-05,
+      "loss": 2.5364,
+      "step": 646
+    },
+    {
+      "epoch": 0.12455182039126982,
+      "grad_norm": 3.403117324921623,
+      "learning_rate": 1.9945501045009647e-05,
+      "loss": 2.524,
+      "step": 647
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "grad_norm": 5.256055958991321,
+      "learning_rate": 1.9945175490563036e-05,
+      "loss": 2.5402,
+      "step": 648
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.5369,
+      "step": 648,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.4691,
+      "step": 648,
+      "vm_loss": 0.1109
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.5041,
+      "step": 648,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.4555,
+      "step": 648,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.5028,
+      "step": 648,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.3323,
+      "step": 648,
+      "vm_loss": 0.1231
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 1.974,
+      "step": 648,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 0.1247443270688452,
+      "lm_loss": 2.3771,
+      "step": 648,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 0.12493683374642058,
+      "grad_norm": 3.51367171662114,
+      "learning_rate": 1.994484896931814e-05,
+      "loss": 2.5417,
+      "step": 649
+    },
+    {
+      "epoch": 0.12512934042399596,
+      "grad_norm": 4.647815520160439,
+      "learning_rate": 1.994452148130668e-05,
+      "loss": 2.5321,
+      "step": 650
+    },
+    {
+      "epoch": 0.12532184710157135,
+      "grad_norm": 3.7809143432498757,
+      "learning_rate": 1.9944193026560513e-05,
+      "loss": 2.566,
+      "step": 651
+    },
+    {
+      "epoch": 0.12551435377914671,
+      "grad_norm": 3.6761052173672835,
+      "learning_rate": 1.9943863605111558e-05,
+      "loss": 2.5652,
+      "step": 652
+    },
+    {
+      "epoch": 0.12570686045672208,
+      "grad_norm": 4.63139988481928,
+      "learning_rate": 1.9943533216991843e-05,
+      "loss": 2.4885,
+      "step": 653
+    },
+    {
+      "epoch": 0.12589936713429747,
+      "grad_norm": 3.7503339358328827,
+      "learning_rate": 1.994320186223348e-05,
+      "loss": 2.5568,
+      "step": 654
+    },
+    {
+      "epoch": 0.12609187381187284,
+      "grad_norm": 4.730960373444081,
+      "learning_rate": 1.994286954086869e-05,
+      "loss": 2.524,
+      "step": 655
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "grad_norm": 3.86790112087384,
+      "learning_rate": 1.994253625292977e-05,
+      "loss": 2.5173,
+      "step": 656
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.4725,
+      "step": 656,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 1.934,
+      "step": 656,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.4389,
+      "step": 656,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.1853,
+      "step": 656,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.5635,
+      "step": 656,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.521,
+      "step": 656,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.2869,
+      "step": 656,
+      "vm_loss": 0.1133
+    },
+    {
+      "epoch": 0.12628438048944823,
+      "lm_loss": 2.3508,
+      "step": 656,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.1264768871670236,
+      "grad_norm": 4.635303142095531,
+      "learning_rate": 1.9942201998449133e-05,
+      "loss": 2.5302,
+      "step": 657
+    },
+    {
+      "epoch": 0.126669393844599,
+      "grad_norm": 3.9988896990150247,
+      "learning_rate": 1.9941866777459253e-05,
+      "loss": 2.5095,
+      "step": 658
+    },
+    {
+      "epoch": 0.12686190052217436,
+      "grad_norm": 3.9154733464983833,
+      "learning_rate": 1.9941530589992733e-05,
+      "loss": 2.5567,
+      "step": 659
+    },
+    {
+      "epoch": 0.12705440719974975,
+      "grad_norm": 4.562643978524786,
+      "learning_rate": 1.994119343608225e-05,
+      "loss": 2.5156,
+      "step": 660
+    },
+    {
+      "epoch": 0.12724691387732512,
+      "grad_norm": 3.4115903425836267,
+      "learning_rate": 1.9940855315760582e-05,
+      "loss": 2.5357,
+      "step": 661
+    },
+    {
+      "epoch": 0.12743942055490048,
+      "grad_norm": 4.691181338330435,
+      "learning_rate": 1.9940516229060596e-05,
+      "loss": 2.5134,
+      "step": 662
+    },
+    {
+      "epoch": 0.12763192723247588,
+      "grad_norm": 3.367208174430154,
+      "learning_rate": 1.9940176176015253e-05,
+      "loss": 2.5319,
+      "step": 663
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "grad_norm": 4.191450775924186,
+      "learning_rate": 1.9939835156657616e-05,
+      "loss": 2.5166,
+      "step": 664
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.2624,
+      "step": 664,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.5022,
+      "step": 664,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.3328,
+      "step": 664,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.4798,
+      "step": 664,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.1554,
+      "step": 664,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.0532,
+      "step": 664,
+      "vm_loss": 0.2217
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.3149,
+      "step": 664,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 0.12782443391005124,
+      "lm_loss": 2.1765,
+      "step": 664,
+      "vm_loss": 0.227
+    },
+    {
+      "epoch": 0.12801694058762664,
+      "grad_norm": 3.4969968141940955,
+      "learning_rate": 1.9939493171020836e-05,
+      "loss": 2.5157,
+      "step": 665
+    },
+    {
+      "epoch": 0.128209447265202,
+      "grad_norm": 3.8711315461127036,
+      "learning_rate": 1.9939150219138154e-05,
+      "loss": 2.5384,
+      "step": 666
+    },
+    {
+      "epoch": 0.1284019539427774,
+      "grad_norm": 3.774071886217187,
+      "learning_rate": 1.993880630104291e-05,
+      "loss": 2.5066,
+      "step": 667
+    },
+    {
+      "epoch": 0.12859446062035276,
+      "grad_norm": 4.038928846906131,
+      "learning_rate": 1.9938461416768544e-05,
+      "loss": 2.5013,
+      "step": 668
+    },
+    {
+      "epoch": 0.12878696729792816,
+      "grad_norm": 3.802648692275223,
+      "learning_rate": 1.9938115566348576e-05,
+      "loss": 2.5594,
+      "step": 669
+    },
+    {
+      "epoch": 0.12897947397550352,
+      "grad_norm": 3.453875121639923,
+      "learning_rate": 1.9937768749816628e-05,
+      "loss": 2.5198,
+      "step": 670
+    },
+    {
+      "epoch": 0.12917198065307892,
+      "grad_norm": 4.2472939694026905,
+      "learning_rate": 1.9937420967206415e-05,
+      "loss": 2.5353,
+      "step": 671
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "grad_norm": 3.9958109356791236,
+      "learning_rate": 1.9937072218551752e-05,
+      "loss": 2.5283,
+      "step": 672
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.2537,
+      "step": 672,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.4968,
+      "step": 672,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.4089,
+      "step": 672,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.5114,
+      "step": 672,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.2481,
+      "step": 672,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.1567,
+      "step": 672,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.313,
+      "step": 672,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.12936448733065428,
+      "lm_loss": 2.4374,
+      "step": 672,
+      "vm_loss": 0.1151
+    },
+    {
+      "epoch": 0.12955699400822965,
+      "grad_norm": 3.9293362103244447,
+      "learning_rate": 1.9936722503886533e-05,
+      "loss": 2.5089,
+      "step": 673
+    },
+    {
+      "epoch": 0.12974950068580504,
+      "grad_norm": 4.632319353205451,
+      "learning_rate": 1.9936371823244764e-05,
+      "loss": 2.5573,
+      "step": 674
+    },
+    {
+      "epoch": 0.1299420073633804,
+      "grad_norm": 3.7231800984411265,
+      "learning_rate": 1.9936020176660524e-05,
+      "loss": 2.5283,
+      "step": 675
+    },
+    {
+      "epoch": 0.1301345140409558,
+      "grad_norm": 4.076787346661957,
+      "learning_rate": 1.9935667564168008e-05,
+      "loss": 2.5272,
+      "step": 676
+    },
+    {
+      "epoch": 0.13032702071853117,
+      "grad_norm": 3.5182137769058786,
+      "learning_rate": 1.993531398580149e-05,
+      "loss": 2.5304,
+      "step": 677
+    },
+    {
+      "epoch": 0.13051952739610656,
+      "grad_norm": 3.9492581536972673,
+      "learning_rate": 1.993495944159534e-05,
+      "loss": 2.5538,
+      "step": 678
+    },
+    {
+      "epoch": 0.13071203407368193,
+      "grad_norm": 4.455982751635128,
+      "learning_rate": 1.9934603931584032e-05,
+      "loss": 2.5007,
+      "step": 679
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "grad_norm": 3.0802307570055922,
+      "learning_rate": 1.9934247455802115e-05,
+      "loss": 2.554,
+      "step": 680
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.243,
+      "step": 680,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.6919,
+      "step": 680,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.2703,
+      "step": 680,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.0587,
+      "step": 680,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.4997,
+      "step": 680,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.2966,
+      "step": 680,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.1264,
+      "step": 680,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.13090454075125732,
+      "lm_loss": 2.1005,
+      "step": 680,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.13109704742883269,
+      "grad_norm": 3.678063288166201,
+      "learning_rate": 1.9933890014284257e-05,
+      "loss": 2.5178,
+      "step": 681
+    },
+    {
+      "epoch": 0.13128955410640808,
+      "grad_norm": 3.871396444452642,
+      "learning_rate": 1.9933531607065193e-05,
+      "loss": 2.4804,
+      "step": 682
+    },
+    {
+      "epoch": 0.13148206078398345,
+      "grad_norm": 4.022100269229644,
+      "learning_rate": 1.993317223417977e-05,
+      "loss": 2.5173,
+      "step": 683
+    },
+    {
+      "epoch": 0.1316745674615588,
+      "grad_norm": 4.756744774856078,
+      "learning_rate": 1.993281189566292e-05,
+      "loss": 2.5432,
+      "step": 684
+    },
+    {
+      "epoch": 0.1318670741391342,
+      "grad_norm": 3.878310002826389,
+      "learning_rate": 1.9932450591549683e-05,
+      "loss": 2.5448,
+      "step": 685
+    },
+    {
+      "epoch": 0.13205958081670957,
+      "grad_norm": 4.18048753467834,
+      "learning_rate": 1.9932088321875175e-05,
+      "loss": 2.4943,
+      "step": 686
+    },
+    {
+      "epoch": 0.13225208749428496,
+      "grad_norm": 4.060923585685389,
+      "learning_rate": 1.9931725086674607e-05,
+      "loss": 2.5957,
+      "step": 687
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "grad_norm": 3.4645542345182805,
+      "learning_rate": 1.99313608859833e-05,
+      "loss": 2.5397,
+      "step": 688
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.5851,
+      "step": 688,
+      "vm_loss": 0.222
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.2303,
+      "step": 688,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.4098,
+      "step": 688,
+      "vm_loss": 0.2349
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.3245,
+      "step": 688,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.5087,
+      "step": 688,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.4198,
+      "step": 688,
+      "vm_loss": 0.2447
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.1676,
+      "step": 688,
+      "vm_loss": 0.2177
+    },
+    {
+      "epoch": 0.13244459417186033,
+      "lm_loss": 2.4587,
+      "step": 688,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 0.13263710084943572,
+      "grad_norm": 3.959500347218049,
+      "learning_rate": 1.9930995719836658e-05,
+      "loss": 2.5491,
+      "step": 689
+    },
+    {
+      "epoch": 0.1328296075270111,
+      "grad_norm": 4.6051983094794835,
+      "learning_rate": 1.9930629588270177e-05,
+      "loss": 2.5777,
+      "step": 690
+    },
+    {
+      "epoch": 0.13302211420458648,
+      "grad_norm": 3.6599835646768666,
+      "learning_rate": 1.993026249131945e-05,
+      "loss": 2.4919,
+      "step": 691
+    },
+    {
+      "epoch": 0.13321462088216185,
+      "grad_norm": 4.050632043887986,
+      "learning_rate": 1.992989442902016e-05,
+      "loss": 2.5217,
+      "step": 692
+    },
+    {
+      "epoch": 0.13340712755973722,
+      "grad_norm": 4.671505037451218,
+      "learning_rate": 1.9929525401408095e-05,
+      "loss": 2.4872,
+      "step": 693
+    },
+    {
+      "epoch": 0.1335996342373126,
+      "grad_norm": 3.7477270661721054,
+      "learning_rate": 1.9929155408519124e-05,
+      "loss": 2.4912,
+      "step": 694
+    },
+    {
+      "epoch": 0.13379214091488797,
+      "grad_norm": 4.540438705544522,
+      "learning_rate": 1.9928784450389215e-05,
+      "loss": 2.532,
+      "step": 695
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "grad_norm": 3.5958084799314105,
+      "learning_rate": 1.9928412527054434e-05,
+      "loss": 2.5505,
+      "step": 696
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.1628,
+      "step": 696,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.0239,
+      "step": 696,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.5022,
+      "step": 696,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.1749,
+      "step": 696,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.4784,
+      "step": 696,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.2219,
+      "step": 696,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.0597,
+      "step": 696,
+      "vm_loss": 0.2234
+    },
+    {
+      "epoch": 0.13398464759246337,
+      "lm_loss": 2.3268,
+      "step": 696,
+      "vm_loss": 0.2184
+    },
+    {
+      "epoch": 0.13417715427003873,
+      "grad_norm": 5.293221432692755,
+      "learning_rate": 1.9928039638550932e-05,
+      "loss": 2.4637,
+      "step": 697
+    },
+    {
+      "epoch": 0.13436966094761413,
+      "grad_norm": 4.181332486198363,
+      "learning_rate": 1.992766578491496e-05,
+      "loss": 2.5523,
+      "step": 698
+    },
+    {
+      "epoch": 0.1345621676251895,
+      "grad_norm": 4.37083363256827,
+      "learning_rate": 1.992729096618286e-05,
+      "loss": 2.4945,
+      "step": 699
+    },
+    {
+      "epoch": 0.1347546743027649,
+      "grad_norm": 4.131147972223744,
+      "learning_rate": 1.9926915182391074e-05,
+      "loss": 2.5517,
+      "step": 700
+    },
+    {
+      "epoch": 0.13494718098034025,
+      "grad_norm": 3.748901096479565,
+      "learning_rate": 1.992653843357613e-05,
+      "loss": 2.4505,
+      "step": 701
+    },
+    {
+      "epoch": 0.13513968765791565,
+      "grad_norm": 3.5877001786459894,
+      "learning_rate": 1.9926160719774656e-05,
+      "loss": 2.4923,
+      "step": 702
+    },
+    {
+      "epoch": 0.135332194335491,
+      "grad_norm": 3.8824528589617797,
+      "learning_rate": 1.992578204102336e-05,
+      "loss": 2.5295,
+      "step": 703
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "grad_norm": 3.8136063237230116,
+      "learning_rate": 1.9925402397359066e-05,
+      "loss": 2.5711,
+      "step": 704
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.2671,
+      "step": 704,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.3235,
+      "step": 704,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.3424,
+      "step": 704,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 1.6294,
+      "step": 704,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.1326,
+      "step": 704,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.3362,
+      "step": 704,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.5109,
+      "step": 704,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 0.13552470101306638,
+      "lm_loss": 2.3083,
+      "step": 704,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.13571720769064177,
+      "grad_norm": 4.0117866110071345,
+      "learning_rate": 1.9925021788818674e-05,
+      "loss": 2.5291,
+      "step": 705
+    },
+    {
+      "epoch": 0.13590971436821714,
+      "grad_norm": 3.7080704525190544,
+      "learning_rate": 1.992464021543919e-05,
+      "loss": 2.4973,
+      "step": 706
+    },
+    {
+      "epoch": 0.13610222104579253,
+      "grad_norm": 4.106054799135807,
+      "learning_rate": 1.99242576772577e-05,
+      "loss": 2.4842,
+      "step": 707
+    },
+    {
+      "epoch": 0.1362947277233679,
+      "grad_norm": 4.402486136739704,
+      "learning_rate": 1.9923874174311394e-05,
+      "loss": 2.4998,
+      "step": 708
+    },
+    {
+      "epoch": 0.1364872344009433,
+      "grad_norm": 4.21005751032192,
+      "learning_rate": 1.992348970663756e-05,
+      "loss": 2.5805,
+      "step": 709
+    },
+    {
+      "epoch": 0.13667974107851866,
+      "grad_norm": 4.087937474001111,
+      "learning_rate": 1.9923104274273564e-05,
+      "loss": 2.5068,
+      "step": 710
+    },
+    {
+      "epoch": 0.13687224775609405,
+      "grad_norm": 3.9605057651128153,
+      "learning_rate": 1.992271787725688e-05,
+      "loss": 2.5333,
+      "step": 711
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "grad_norm": 4.393816375918899,
+      "learning_rate": 1.992233051562507e-05,
+      "loss": 2.4906,
+      "step": 712
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.3329,
+      "step": 712,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.4234,
+      "step": 712,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.1843,
+      "step": 712,
+      "vm_loss": 0.2479
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.2165,
+      "step": 712,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 1.7938,
+      "step": 712,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.5259,
+      "step": 712,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.188,
+      "step": 712,
+      "vm_loss": 0.2361
+    },
+    {
+      "epoch": 0.13706475443366942,
+      "lm_loss": 2.3782,
+      "step": 712,
+      "vm_loss": 0.2383
+    },
+    {
+      "epoch": 0.13725726111124478,
+      "grad_norm": 4.04710857306534,
+      "learning_rate": 1.9921942189415793e-05,
+      "loss": 2.5318,
+      "step": 713
+    },
+    {
+      "epoch": 0.13744976778882018,
+      "grad_norm": 4.4872288353309235,
+      "learning_rate": 1.9921552898666793e-05,
+      "loss": 2.4813,
+      "step": 714
+    },
+    {
+      "epoch": 0.13764227446639554,
+      "grad_norm": 4.016406026527296,
+      "learning_rate": 1.9921162643415918e-05,
+      "loss": 2.4594,
+      "step": 715
+    },
+    {
+      "epoch": 0.13783478114397094,
+      "grad_norm": 4.044491673117378,
+      "learning_rate": 1.9920771423701108e-05,
+      "loss": 2.5857,
+      "step": 716
+    },
+    {
+      "epoch": 0.1380272878215463,
+      "grad_norm": 4.237752379784449,
+      "learning_rate": 1.9920379239560388e-05,
+      "loss": 2.5607,
+      "step": 717
+    },
+    {
+      "epoch": 0.1382197944991217,
+      "grad_norm": 5.218320403597712,
+      "learning_rate": 1.9919986091031885e-05,
+      "loss": 2.5589,
+      "step": 718
+    },
+    {
+      "epoch": 0.13841230117669706,
+      "grad_norm": 4.870607108724517,
+      "learning_rate": 1.9919591978153824e-05,
+      "loss": 2.5335,
+      "step": 719
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "grad_norm": 3.871828669170681,
+      "learning_rate": 1.9919196900964517e-05,
+      "loss": 2.5532,
+      "step": 720
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.47,
+      "step": 720,
+      "vm_loss": 0.1103
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.5248,
+      "step": 720,
+      "vm_loss": 0.209
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.1865,
+      "step": 720,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.6737,
+      "step": 720,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.0873,
+      "step": 720,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.2774,
+      "step": 720,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.57,
+      "step": 720,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 0.13860480785427245,
+      "lm_loss": 2.0699,
+      "step": 720,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.13879731453184782,
+      "grad_norm": 4.075557972706211,
+      "learning_rate": 1.9918800859502366e-05,
+      "loss": 2.5329,
+      "step": 721
+    },
+    {
+      "epoch": 0.13898982120942321,
+      "grad_norm": 4.425157589383468,
+      "learning_rate": 1.991840385380587e-05,
+      "loss": 2.4517,
+      "step": 722
+    },
+    {
+      "epoch": 0.13918232788699858,
+      "grad_norm": 4.908035310675754,
+      "learning_rate": 1.991800588391363e-05,
+      "loss": 2.5122,
+      "step": 723
+    },
+    {
+      "epoch": 0.13937483456457395,
+      "grad_norm": 3.5165633378115744,
+      "learning_rate": 1.9917606949864326e-05,
+      "loss": 2.5141,
+      "step": 724
+    },
+    {
+      "epoch": 0.13956734124214934,
+      "grad_norm": 4.046152078981162,
+      "learning_rate": 1.9917207051696744e-05,
+      "loss": 2.5085,
+      "step": 725
+    },
+    {
+      "epoch": 0.1397598479197247,
+      "grad_norm": 4.221543390010374,
+      "learning_rate": 1.991680618944976e-05,
+      "loss": 2.4703,
+      "step": 726
+    },
+    {
+      "epoch": 0.1399523545973001,
+      "grad_norm": 4.1866542337112005,
+      "learning_rate": 1.991640436316234e-05,
+      "loss": 2.5176,
+      "step": 727
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "grad_norm": 4.428208052748198,
+      "learning_rate": 1.991600157287355e-05,
+      "loss": 2.564,
+      "step": 728
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.3465,
+      "step": 728,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.2343,
+      "step": 728,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.6499,
+      "step": 728,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.4633,
+      "step": 728,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.5692,
+      "step": 728,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.3546,
+      "step": 728,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.4916,
+      "step": 728,
+      "vm_loss": 0.2559
+    },
+    {
+      "epoch": 0.14014486127487547,
+      "lm_loss": 2.5001,
+      "step": 728,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.14033736795245086,
+      "grad_norm": 4.2609027439682015,
+      "learning_rate": 1.9915597818622542e-05,
+      "loss": 2.5709,
+      "step": 729
+    },
+    {
+      "epoch": 0.14052987463002622,
+      "grad_norm": 4.128131763963246,
+      "learning_rate": 1.991519310044857e-05,
+      "loss": 2.4997,
+      "step": 730
+    },
+    {
+      "epoch": 0.14072238130760162,
+      "grad_norm": 3.6429004555596403,
+      "learning_rate": 1.9914787418390975e-05,
+      "loss": 2.5422,
+      "step": 731
+    },
+    {
+      "epoch": 0.14091488798517698,
+      "grad_norm": 3.9901359976747113,
+      "learning_rate": 1.9914380772489195e-05,
+      "loss": 2.5108,
+      "step": 732
+    },
+    {
+      "epoch": 0.14110739466275238,
+      "grad_norm": 4.431513703584899,
+      "learning_rate": 1.9913973162782764e-05,
+      "loss": 2.4951,
+      "step": 733
+    },
+    {
+      "epoch": 0.14129990134032774,
+      "grad_norm": 4.172522236655245,
+      "learning_rate": 1.99135645893113e-05,
+      "loss": 2.5202,
+      "step": 734
+    },
+    {
+      "epoch": 0.1414924080179031,
+      "grad_norm": 3.8441806336495117,
+      "learning_rate": 1.9913155052114533e-05,
+      "loss": 2.5012,
+      "step": 735
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "grad_norm": 4.520989804769229,
+      "learning_rate": 1.9912744551232268e-05,
+      "loss": 2.5034,
+      "step": 736
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.3319,
+      "step": 736,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.4547,
+      "step": 736,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.3391,
+      "step": 736,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.8188,
+      "step": 736,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.169,
+      "step": 736,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.3449,
+      "step": 736,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.4636,
+      "step": 736,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 0.1416849146954785,
+      "lm_loss": 2.175,
+      "step": 736,
+      "vm_loss": 0.1559
+    },
+    {
+      "epoch": 0.14187742137305387,
+      "grad_norm": 3.933989005893063,
+      "learning_rate": 1.991233308670441e-05,
+      "loss": 2.5414,
+      "step": 737
+    },
+    {
+      "epoch": 0.14206992805062926,
+      "grad_norm": 4.136289663795889,
+      "learning_rate": 1.991192065857096e-05,
+      "loss": 2.4705,
+      "step": 738
+    },
+    {
+      "epoch": 0.14226243472820463,
+      "grad_norm": 4.288852677681718,
+      "learning_rate": 1.991150726687201e-05,
+      "loss": 2.5102,
+      "step": 739
+    },
+    {
+      "epoch": 0.14245494140578002,
+      "grad_norm": 3.2252252856749997,
+      "learning_rate": 1.9911092911647752e-05,
+      "loss": 2.4659,
+      "step": 740
+    },
+    {
+      "epoch": 0.1426474480833554,
+      "grad_norm": 4.352948548409321,
+      "learning_rate": 1.9910677592938458e-05,
+      "loss": 2.5335,
+      "step": 741
+    },
+    {
+      "epoch": 0.14283995476093078,
+      "grad_norm": 3.665563505299811,
+      "learning_rate": 1.9910261310784515e-05,
+      "loss": 2.5252,
+      "step": 742
+    },
+    {
+      "epoch": 0.14303246143850615,
+      "grad_norm": 3.46504171500494,
+      "learning_rate": 1.9909844065226376e-05,
+      "loss": 2.5128,
+      "step": 743
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "grad_norm": 4.502805014824155,
+      "learning_rate": 1.9909425856304615e-05,
+      "loss": 2.5114,
+      "step": 744
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.3742,
+      "step": 744,
+      "vm_loss": 0.1837
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.3313,
+      "step": 744,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 1.9662,
+      "step": 744,
+      "vm_loss": 0.2747
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.2348,
+      "step": 744,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.1076,
+      "step": 744,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.706,
+      "step": 744,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.3584,
+      "step": 744,
+      "vm_loss": 0.259
+    },
+    {
+      "epoch": 0.1432249681160815,
+      "lm_loss": 2.3714,
+      "step": 744,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 0.1434174747936569,
+      "grad_norm": 3.600535945519794,
+      "learning_rate": 1.990900668405988e-05,
+      "loss": 2.5253,
+      "step": 745
+    },
+    {
+      "epoch": 0.14360998147123227,
+      "grad_norm": 4.018715609335587,
+      "learning_rate": 1.990858654853292e-05,
+      "loss": 2.4983,
+      "step": 746
+    },
+    {
+      "epoch": 0.14380248814880767,
+      "grad_norm": 3.573250442551011,
+      "learning_rate": 1.990816544976458e-05,
+      "loss": 2.5331,
+      "step": 747
+    },
+    {
+      "epoch": 0.14399499482638303,
+      "grad_norm": 3.653256381564751,
+      "learning_rate": 1.99077433877958e-05,
+      "loss": 2.5117,
+      "step": 748
+    },
+    {
+      "epoch": 0.14418750150395843,
+      "grad_norm": 4.3701896250522045,
+      "learning_rate": 1.9907320362667605e-05,
+      "loss": 2.4485,
+      "step": 749
+    },
+    {
+      "epoch": 0.1443800081815338,
+      "grad_norm": 3.436477164554334,
+      "learning_rate": 1.9906896374421116e-05,
+      "loss": 2.5282,
+      "step": 750
+    },
+    {
+      "epoch": 0.14457251485910919,
+      "grad_norm": 4.534762046031325,
+      "learning_rate": 1.9906471423097552e-05,
+      "loss": 2.4481,
+      "step": 751
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "grad_norm": 3.5823737485713867,
+      "learning_rate": 1.990604550873823e-05,
+      "loss": 2.5235,
+      "step": 752
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.2159,
+      "step": 752,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.0608,
+      "step": 752,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.3342,
+      "step": 752,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.5536,
+      "step": 752,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.667,
+      "step": 752,
+      "vm_loss": 0.1057
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.2093,
+      "step": 752,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.3884,
+      "step": 752,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 0.14476502153668455,
+      "lm_loss": 2.3651,
+      "step": 752,
+      "vm_loss": 0.2497
+    },
+    {
+      "epoch": 0.14495752821425995,
+      "grad_norm": 3.362839374780301,
+      "learning_rate": 1.9905618631384546e-05,
+      "loss": 2.4953,
+      "step": 753
+    },
+    {
+      "epoch": 0.1451500348918353,
+      "grad_norm": 3.890888572852814,
+      "learning_rate": 1.9905190791077998e-05,
+      "loss": 2.4955,
+      "step": 754
+    },
+    {
+      "epoch": 0.14534254156941068,
+      "grad_norm": 3.8663659137447395,
+      "learning_rate": 1.9904761987860187e-05,
+      "loss": 2.4857,
+      "step": 755
+    },
+    {
+      "epoch": 0.14553504824698607,
+      "grad_norm": 3.471695821517422,
+      "learning_rate": 1.9904332221772788e-05,
+      "loss": 2.498,
+      "step": 756
+    },
+    {
+      "epoch": 0.14572755492456144,
+      "grad_norm": 3.6899373830137754,
+      "learning_rate": 1.9903901492857586e-05,
+      "loss": 2.494,
+      "step": 757
+    },
+    {
+      "epoch": 0.14592006160213683,
+      "grad_norm": 4.180522440144561,
+      "learning_rate": 1.990346980115645e-05,
+      "loss": 2.4886,
+      "step": 758
+    },
+    {
+      "epoch": 0.1461125682797122,
+      "grad_norm": 3.836127376099166,
+      "learning_rate": 1.9903037146711348e-05,
+      "loss": 2.5123,
+      "step": 759
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "grad_norm": 4.280059401564708,
+      "learning_rate": 1.9902603529564338e-05,
+      "loss": 2.5147,
+      "step": 760
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.5928,
+      "step": 760,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.3228,
+      "step": 760,
+      "vm_loss": 0.2248
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.1436,
+      "step": 760,
+      "vm_loss": 0.2843
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.4144,
+      "step": 760,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.3729,
+      "step": 760,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.4157,
+      "step": 760,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.2469,
+      "step": 760,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 0.1463050749572876,
+      "lm_loss": 2.2256,
+      "step": 760,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 0.14649758163486296,
+      "grad_norm": 3.428443985869128,
+      "learning_rate": 1.9902168949757572e-05,
+      "loss": 2.5365,
+      "step": 761
+    },
+    {
+      "epoch": 0.14669008831243835,
+      "grad_norm": 3.8166352812621875,
+      "learning_rate": 1.9901733407333303e-05,
+      "loss": 2.4953,
+      "step": 762
+    },
+    {
+      "epoch": 0.14688259499001372,
+      "grad_norm": 3.942498205711259,
+      "learning_rate": 1.990129690233386e-05,
+      "loss": 2.5387,
+      "step": 763
+    },
+    {
+      "epoch": 0.14707510166758908,
+      "grad_norm": 3.317992072568667,
+      "learning_rate": 1.9900859434801686e-05,
+      "loss": 2.5173,
+      "step": 764
+    },
+    {
+      "epoch": 0.14726760834516447,
+      "grad_norm": 3.89094857705775,
+      "learning_rate": 1.9900421004779306e-05,
+      "loss": 2.4946,
+      "step": 765
+    },
+    {
+      "epoch": 0.14746011502273984,
+      "grad_norm": 3.786031588662704,
+      "learning_rate": 1.9899981612309344e-05,
+      "loss": 2.5156,
+      "step": 766
+    },
+    {
+      "epoch": 0.14765262170031523,
+      "grad_norm": 3.8286918262451297,
+      "learning_rate": 1.9899541257434506e-05,
+      "loss": 2.5095,
+      "step": 767
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "grad_norm": 3.931170649796231,
+      "learning_rate": 1.989909994019761e-05,
+      "loss": 2.5604,
+      "step": 768
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.3366,
+      "step": 768,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.3119,
+      "step": 768,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.2553,
+      "step": 768,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.4439,
+      "step": 768,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.3012,
+      "step": 768,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.2725,
+      "step": 768,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.5608,
+      "step": 768,
+      "vm_loss": 0.2307
+    },
+    {
+      "epoch": 0.1478451283778906,
+      "lm_loss": 2.3226,
+      "step": 768,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.148037635055466,
+      "grad_norm": 3.7565547676940634,
+      "learning_rate": 1.9898657660641554e-05,
+      "loss": 2.5324,
+      "step": 769
+    },
+    {
+      "epoch": 0.14823014173304136,
+      "grad_norm": 3.7773119159102504,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 2.5483,
+      "step": 770
+    },
+    {
+      "epoch": 0.14842264841061675,
+      "grad_norm": 3.572617679441565,
+      "learning_rate": 1.989777021474403e-05,
+      "loss": 2.4803,
+      "step": 771
+    },
+    {
+      "epoch": 0.14861515508819212,
+      "grad_norm": 4.244362952287954,
+      "learning_rate": 1.9897325048488833e-05,
+      "loss": 2.5031,
+      "step": 772
+    },
+    {
+      "epoch": 0.1488076617657675,
+      "grad_norm": 3.6258598385306393,
+      "learning_rate": 1.989687892008702e-05,
+      "loss": 2.5049,
+      "step": 773
+    },
+    {
+      "epoch": 0.14900016844334288,
+      "grad_norm": 3.565874285155107,
+      "learning_rate": 1.989643182958196e-05,
+      "loss": 2.4807,
+      "step": 774
+    },
+    {
+      "epoch": 0.14919267512091824,
+      "grad_norm": 3.9979185276266462,
+      "learning_rate": 1.9895983777017113e-05,
+      "loss": 2.516,
+      "step": 775
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "grad_norm": 3.6620937247709993,
+      "learning_rate": 1.9895534762436032e-05,
+      "loss": 2.5312,
+      "step": 776
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 1.929,
+      "step": 776,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.2116,
+      "step": 776,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.2605,
+      "step": 776,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.3101,
+      "step": 776,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.3977,
+      "step": 776,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.4413,
+      "step": 776,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.2528,
+      "step": 776,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 0.14938518179849364,
+      "lm_loss": 2.34,
+      "step": 776,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.149577688476069,
+      "grad_norm": 3.6826109674237744,
+      "learning_rate": 1.9895084785882378e-05,
+      "loss": 2.4569,
+      "step": 777
+    },
+    {
+      "epoch": 0.1497701951536444,
+      "grad_norm": 3.1463332149594,
+      "learning_rate": 1.9894633847399883e-05,
+      "loss": 2.4866,
+      "step": 778
+    },
+    {
+      "epoch": 0.14996270183121976,
+      "grad_norm": 3.4325686250731393,
+      "learning_rate": 1.9894181947032388e-05,
+      "loss": 2.532,
+      "step": 779
+    },
+    {
+      "epoch": 0.15015520850879516,
+      "grad_norm": 3.498797083938133,
+      "learning_rate": 1.9893729084823826e-05,
+      "loss": 2.5261,
+      "step": 780
+    },
+    {
+      "epoch": 0.15034771518637052,
+      "grad_norm": 3.9926524983205844,
+      "learning_rate": 1.9893275260818222e-05,
+      "loss": 2.466,
+      "step": 781
+    },
+    {
+      "epoch": 0.15054022186394592,
+      "grad_norm": 3.6638849402622244,
+      "learning_rate": 1.9892820475059687e-05,
+      "loss": 2.5113,
+      "step": 782
+    },
+    {
+      "epoch": 0.15073272854152128,
+      "grad_norm": 4.326531067224903,
+      "learning_rate": 1.9892364727592436e-05,
+      "loss": 2.4637,
+      "step": 783
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "grad_norm": 3.983710277535217,
+      "learning_rate": 1.9891908018460775e-05,
+      "loss": 2.4953,
+      "step": 784
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.4628,
+      "step": 784,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.5962,
+      "step": 784,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.4448,
+      "step": 784,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.3503,
+      "step": 784,
+      "vm_loss": 0.1932
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.532,
+      "step": 784,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.4043,
+      "step": 784,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.3167,
+      "step": 784,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 0.15092523521909668,
+      "lm_loss": 2.4208,
+      "step": 784,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.15111774189667204,
+      "grad_norm": 4.609559942458938,
+      "learning_rate": 1.98914503477091e-05,
+      "loss": 2.5614,
+      "step": 785
+    },
+    {
+      "epoch": 0.1513102485742474,
+      "grad_norm": 4.460768949829644,
+      "learning_rate": 1.98909917153819e-05,
+      "loss": 2.5284,
+      "step": 786
+    },
+    {
+      "epoch": 0.1515027552518228,
+      "grad_norm": 3.9804700459200415,
+      "learning_rate": 1.9890532121523766e-05,
+      "loss": 2.4754,
+      "step": 787
+    },
+    {
+      "epoch": 0.15169526192939817,
+      "grad_norm": 4.070777584649809,
+      "learning_rate": 1.989007156617937e-05,
+      "loss": 2.5261,
+      "step": 788
+    },
+    {
+      "epoch": 0.15188776860697356,
+      "grad_norm": 4.6263915929846675,
+      "learning_rate": 1.9889610049393487e-05,
+      "loss": 2.5019,
+      "step": 789
+    },
+    {
+      "epoch": 0.15208027528454893,
+      "grad_norm": 3.674440215497087,
+      "learning_rate": 1.9889147571210984e-05,
+      "loss": 2.518,
+      "step": 790
+    },
+    {
+      "epoch": 0.15227278196212432,
+      "grad_norm": 3.9061139696453693,
+      "learning_rate": 1.9888684131676816e-05,
+      "loss": 2.5138,
+      "step": 791
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "grad_norm": 4.578408555264086,
+      "learning_rate": 1.9888219730836038e-05,
+      "loss": 2.5151,
+      "step": 792
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 1.688,
+      "step": 792,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.3761,
+      "step": 792,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.4514,
+      "step": 792,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.3851,
+      "step": 792,
+      "vm_loss": 0.2171
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.3004,
+      "step": 792,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.3727,
+      "step": 792,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.2116,
+      "step": 792,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.1524652886396997,
+      "lm_loss": 2.0514,
+      "step": 792,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 0.15265779531727508,
+      "grad_norm": 3.887310030307114,
+      "learning_rate": 1.9887754368733795e-05,
+      "loss": 2.4626,
+      "step": 793
+    },
+    {
+      "epoch": 0.15285030199485045,
+      "grad_norm": 3.8136046688558816,
+      "learning_rate": 1.9887288045415327e-05,
+      "loss": 2.577,
+      "step": 794
+    },
+    {
+      "epoch": 0.1530428086724258,
+      "grad_norm": 3.2654560002836774,
+      "learning_rate": 1.9886820760925962e-05,
+      "loss": 2.4736,
+      "step": 795
+    },
+    {
+      "epoch": 0.1532353153500012,
+      "grad_norm": 3.779358761458401,
+      "learning_rate": 1.9886352515311134e-05,
+      "loss": 2.5142,
+      "step": 796
+    },
+    {
+      "epoch": 0.15342782202757657,
+      "grad_norm": 3.8790729178900443,
+      "learning_rate": 1.9885883308616355e-05,
+      "loss": 2.5019,
+      "step": 797
+    },
+    {
+      "epoch": 0.15362032870515197,
+      "grad_norm": 3.891944942258998,
+      "learning_rate": 1.9885413140887244e-05,
+      "loss": 2.4944,
+      "step": 798
+    },
+    {
+      "epoch": 0.15381283538272733,
+      "grad_norm": 3.5925555580484185,
+      "learning_rate": 1.9884942012169503e-05,
+      "loss": 2.483,
+      "step": 799
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "grad_norm": 4.474016329028718,
+      "learning_rate": 1.9884469922508934e-05,
+      "loss": 2.5211,
+      "step": 800
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.3964,
+      "step": 800,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.3323,
+      "step": 800,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.4357,
+      "step": 800,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.4901,
+      "step": 800,
+      "vm_loss": 0.2229
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.2118,
+      "step": 800,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.109,
+      "step": 800,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.2988,
+      "step": 800,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.15400534206030272,
+      "lm_loss": 2.4196,
+      "step": 800,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 0.1541978487378781,
+      "grad_norm": 3.8196326951673463,
+      "learning_rate": 1.9883996871951427e-05,
+      "loss": 2.518,
+      "step": 801
+    },
+    {
+      "epoch": 0.15439035541545348,
+      "grad_norm": 3.7698113125530632,
+      "learning_rate": 1.988352286054297e-05,
+      "loss": 2.4896,
+      "step": 802
+    },
+    {
+      "epoch": 0.15458286209302885,
+      "grad_norm": 3.7304889831830463,
+      "learning_rate": 1.9883047888329647e-05,
+      "loss": 2.4967,
+      "step": 803
+    },
+    {
+      "epoch": 0.15477536877060424,
+      "grad_norm": 4.414808272225282,
+      "learning_rate": 1.988257195535763e-05,
+      "loss": 2.5113,
+      "step": 804
+    },
+    {
+      "epoch": 0.1549678754481796,
+      "grad_norm": 3.725610598629842,
+      "learning_rate": 1.9882095061673177e-05,
+      "loss": 2.5329,
+      "step": 805
+    },
+    {
+      "epoch": 0.15516038212575498,
+      "grad_norm": 4.534571389116044,
+      "learning_rate": 1.988161720732266e-05,
+      "loss": 2.5074,
+      "step": 806
+    },
+    {
+      "epoch": 0.15535288880333037,
+      "grad_norm": 4.220101350182375,
+      "learning_rate": 1.9881138392352528e-05,
+      "loss": 2.4932,
+      "step": 807
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "grad_norm": 3.643618220848386,
+      "learning_rate": 1.9880658616809324e-05,
+      "loss": 2.5262,
+      "step": 808
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.4591,
+      "step": 808,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.474,
+      "step": 808,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.4972,
+      "step": 808,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.2781,
+      "step": 808,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.3956,
+      "step": 808,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.4666,
+      "step": 808,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.5471,
+      "step": 808,
+      "vm_loss": 0.2226
+    },
+    {
+      "epoch": 0.15554539548090573,
+      "lm_loss": 2.1856,
+      "step": 808,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 0.15573790215848113,
+      "grad_norm": 3.7097444091461673,
+      "learning_rate": 1.9880177880739695e-05,
+      "loss": 2.5145,
+      "step": 809
+    },
+    {
+      "epoch": 0.1559304088360565,
+      "grad_norm": 4.292306449995038,
+      "learning_rate": 1.987969618419037e-05,
+      "loss": 2.509,
+      "step": 810
+    },
+    {
+      "epoch": 0.1561229155136319,
+      "grad_norm": 3.6939094755972994,
+      "learning_rate": 1.987921352720818e-05,
+      "loss": 2.5302,
+      "step": 811
+    },
+    {
+      "epoch": 0.15631542219120725,
+      "grad_norm": 4.052198836030496,
+      "learning_rate": 1.9878729909840043e-05,
+      "loss": 2.5107,
+      "step": 812
+    },
+    {
+      "epoch": 0.15650792886878265,
+      "grad_norm": 4.37266745542868,
+      "learning_rate": 1.987824533213297e-05,
+      "loss": 2.475,
+      "step": 813
+    },
+    {
+      "epoch": 0.156700435546358,
+      "grad_norm": 3.4514887647106898,
+      "learning_rate": 1.9877759794134072e-05,
+      "loss": 2.5024,
+      "step": 814
+    },
+    {
+      "epoch": 0.15689294222393338,
+      "grad_norm": 3.8212435643259095,
+      "learning_rate": 1.987727329589055e-05,
+      "loss": 2.4874,
+      "step": 815
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "grad_norm": 3.7271970226065485,
+      "learning_rate": 1.9876785837449696e-05,
+      "loss": 2.5019,
+      "step": 816
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.4799,
+      "step": 816,
+      "vm_loss": 0.2399
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.0976,
+      "step": 816,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.0679,
+      "step": 816,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.2839,
+      "step": 816,
+      "vm_loss": 0.2008
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.2638,
+      "step": 816,
+      "vm_loss": 0.2365
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.3225,
+      "step": 816,
+      "vm_loss": 0.2095
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.187,
+      "step": 816,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 0.15708544890150877,
+      "lm_loss": 2.432,
+      "step": 816,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 0.15727795557908414,
+      "grad_norm": 4.094966795142,
+      "learning_rate": 1.9876297418858893e-05,
+      "loss": 2.4807,
+      "step": 817
+    },
+    {
+      "epoch": 0.15747046225665953,
+      "grad_norm": 4.094307033081914,
+      "learning_rate": 1.987580804016563e-05,
+      "loss": 2.5165,
+      "step": 818
+    },
+    {
+      "epoch": 0.1576629689342349,
+      "grad_norm": 4.442790577289076,
+      "learning_rate": 1.987531770141748e-05,
+      "loss": 2.5127,
+      "step": 819
+    },
+    {
+      "epoch": 0.1578554756118103,
+      "grad_norm": 4.139757998924265,
+      "learning_rate": 1.9874826402662105e-05,
+      "loss": 2.4873,
+      "step": 820
+    },
+    {
+      "epoch": 0.15804798228938566,
+      "grad_norm": 4.173747352310305,
+      "learning_rate": 1.9874334143947266e-05,
+      "loss": 2.4998,
+      "step": 821
+    },
+    {
+      "epoch": 0.15824048896696105,
+      "grad_norm": 4.088918468567153,
+      "learning_rate": 1.9873840925320823e-05,
+      "loss": 2.5387,
+      "step": 822
+    },
+    {
+      "epoch": 0.15843299564453642,
+      "grad_norm": 3.7010636854152015,
+      "learning_rate": 1.9873346746830712e-05,
+      "loss": 2.4738,
+      "step": 823
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "grad_norm": 4.680368108743709,
+      "learning_rate": 1.9872851608524987e-05,
+      "loss": 2.4965,
+      "step": 824
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.2447,
+      "step": 824,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.6471,
+      "step": 824,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.3997,
+      "step": 824,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.4949,
+      "step": 824,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.4591,
+      "step": 824,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.5899,
+      "step": 824,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.3551,
+      "step": 824,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.1586255023221118,
+      "lm_loss": 2.114,
+      "step": 824,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 0.15881800899968718,
+      "grad_norm": 4.161591338676701,
+      "learning_rate": 1.9872355510451768e-05,
+      "loss": 2.5179,
+      "step": 825
+    },
+    {
+      "epoch": 0.15901051567726254,
+      "grad_norm": 3.9734022392053303,
+      "learning_rate": 1.9871858452659293e-05,
+      "loss": 2.4518,
+      "step": 826
+    },
+    {
+      "epoch": 0.15920302235483794,
+      "grad_norm": 4.186996813735879,
+      "learning_rate": 1.987136043519588e-05,
+      "loss": 2.5024,
+      "step": 827
+    },
+    {
+      "epoch": 0.1593955290324133,
+      "grad_norm": 4.203275787319111,
+      "learning_rate": 1.987086145810994e-05,
+      "loss": 2.4777,
+      "step": 828
+    },
+    {
+      "epoch": 0.1595880357099887,
+      "grad_norm": 3.7430402733378823,
+      "learning_rate": 1.9870361521449978e-05,
+      "loss": 2.5474,
+      "step": 829
+    },
+    {
+      "epoch": 0.15978054238756406,
+      "grad_norm": 4.129396091441689,
+      "learning_rate": 1.9869860625264602e-05,
+      "loss": 2.4963,
+      "step": 830
+    },
+    {
+      "epoch": 0.15997304906513946,
+      "grad_norm": 3.7953898576288325,
+      "learning_rate": 1.98693587696025e-05,
+      "loss": 2.5042,
+      "step": 831
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "grad_norm": 3.7482538962335794,
+      "learning_rate": 1.9868855954512456e-05,
+      "loss": 2.4534,
+      "step": 832
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.4681,
+      "step": 832,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.0654,
+      "step": 832,
+      "vm_loss": 0.126
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.3052,
+      "step": 832,
+      "vm_loss": 0.2036
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.3923,
+      "step": 832,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.1613,
+      "step": 832,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.2553,
+      "step": 832,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.3776,
+      "step": 832,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 0.16016555574271482,
+      "lm_loss": 2.1413,
+      "step": 832,
+      "vm_loss": 0.1243
+    },
+    {
+      "epoch": 0.16035806242029021,
+      "grad_norm": 3.6482939857870846,
+      "learning_rate": 1.9868352180043358e-05,
+      "loss": 2.4955,
+      "step": 833
+    },
+    {
+      "epoch": 0.16055056909786558,
+      "grad_norm": 3.278285390521105,
+      "learning_rate": 1.9867847446244172e-05,
+      "loss": 2.4619,
+      "step": 834
+    },
+    {
+      "epoch": 0.16074307577544097,
+      "grad_norm": 3.9383036614192806,
+      "learning_rate": 1.9867341753163967e-05,
+      "loss": 2.4281,
+      "step": 835
+    },
+    {
+      "epoch": 0.16093558245301634,
+      "grad_norm": 4.372790171334646,
+      "learning_rate": 1.9866835100851903e-05,
+      "loss": 2.5259,
+      "step": 836
+    },
+    {
+      "epoch": 0.1611280891305917,
+      "grad_norm": 3.5797450957194745,
+      "learning_rate": 1.9866327489357236e-05,
+      "loss": 2.4525,
+      "step": 837
+    },
+    {
+      "epoch": 0.1613205958081671,
+      "grad_norm": 3.5250988605569873,
+      "learning_rate": 1.986581891872931e-05,
+      "loss": 2.4837,
+      "step": 838
+    },
+    {
+      "epoch": 0.16151310248574247,
+      "grad_norm": 3.7970270639036268,
+      "learning_rate": 1.9865309389017566e-05,
+      "loss": 2.5168,
+      "step": 839
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "grad_norm": 3.4834087852061826,
+      "learning_rate": 1.986479890027153e-05,
+      "loss": 2.4664,
+      "step": 840
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.4628,
+      "step": 840,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.1535,
+      "step": 840,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.108,
+      "step": 840,
+      "vm_loss": 0.24
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.3511,
+      "step": 840,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.4522,
+      "step": 840,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.4413,
+      "step": 840,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 1.9965,
+      "step": 840,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.16170560916331786,
+      "lm_loss": 2.656,
+      "step": 840,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.16189811584089323,
+      "grad_norm": 3.8289089820304807,
+      "learning_rate": 1.986428745254084e-05,
+      "loss": 2.4937,
+      "step": 841
+    },
+    {
+      "epoch": 0.16209062251846862,
+      "grad_norm": 3.3458407030892325,
+      "learning_rate": 1.9863775045875206e-05,
+      "loss": 2.4868,
+      "step": 842
+    },
+    {
+      "epoch": 0.16228312919604398,
+      "grad_norm": 3.646234913416789,
+      "learning_rate": 1.986326168032444e-05,
+      "loss": 2.4751,
+      "step": 843
+    },
+    {
+      "epoch": 0.16247563587361938,
+      "grad_norm": 3.432423840366768,
+      "learning_rate": 1.986274735593845e-05,
+      "loss": 2.5625,
+      "step": 844
+    },
+    {
+      "epoch": 0.16266814255119474,
+      "grad_norm": 3.2465491995918563,
+      "learning_rate": 1.986223207276724e-05,
+      "loss": 2.528,
+      "step": 845
+    },
+    {
+      "epoch": 0.1628606492287701,
+      "grad_norm": 3.518014121969006,
+      "learning_rate": 1.9861715830860894e-05,
+      "loss": 2.5338,
+      "step": 846
+    },
+    {
+      "epoch": 0.1630531559063455,
+      "grad_norm": 3.5837082374926092,
+      "learning_rate": 1.9861198630269606e-05,
+      "loss": 2.5171,
+      "step": 847
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "grad_norm": 4.13172619845174,
+      "learning_rate": 1.9860680471043644e-05,
+      "loss": 2.4927,
+      "step": 848
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.2073,
+      "step": 848,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.3082,
+      "step": 848,
+      "vm_loss": 0.2643
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.1762,
+      "step": 848,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.231,
+      "step": 848,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 1.9084,
+      "step": 848,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.5673,
+      "step": 848,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 2.4502,
+      "step": 848,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.16324566258392087,
+      "lm_loss": 1.9246,
+      "step": 848,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 0.16343816926149626,
+      "grad_norm": 3.5100326464907714,
+      "learning_rate": 1.986016135323339e-05,
+      "loss": 2.527,
+      "step": 849
+    },
+    {
+      "epoch": 0.16363067593907163,
+      "grad_norm": 3.7155701556789156,
+      "learning_rate": 1.98596412768893e-05,
+      "loss": 2.4747,
+      "step": 850
+    },
+    {
+      "epoch": 0.16382318261664702,
+      "grad_norm": 3.8818291179174995,
+      "learning_rate": 1.985912024206194e-05,
+      "loss": 2.5051,
+      "step": 851
+    },
+    {
+      "epoch": 0.1640156892942224,
+      "grad_norm": 3.4959880929983753,
+      "learning_rate": 1.9858598248801953e-05,
+      "loss": 2.4791,
+      "step": 852
+    },
+    {
+      "epoch": 0.16420819597179778,
+      "grad_norm": 4.430584916408281,
+      "learning_rate": 1.9858075297160094e-05,
+      "loss": 2.5104,
+      "step": 853
+    },
+    {
+      "epoch": 0.16440070264937315,
+      "grad_norm": 3.6332412505498626,
+      "learning_rate": 1.985755138718719e-05,
+      "loss": 2.4906,
+      "step": 854
+    },
+    {
+      "epoch": 0.16459320932694854,
+      "grad_norm": 4.646771970769478,
+      "learning_rate": 1.9857026518934177e-05,
+      "loss": 2.5212,
+      "step": 855
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "grad_norm": 3.6662405647563308,
+      "learning_rate": 1.985650069245208e-05,
+      "loss": 2.4986,
+      "step": 856
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.5841,
+      "step": 856,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.4106,
+      "step": 856,
+      "vm_loss": 0.1804
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.4381,
+      "step": 856,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.3658,
+      "step": 856,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.1363,
+      "step": 856,
+      "vm_loss": 0.2235
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.0291,
+      "step": 856,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.3136,
+      "step": 856,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.1647857160045239,
+      "lm_loss": 2.2066,
+      "step": 856,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.16497822268209927,
+      "grad_norm": 3.6236749224200433,
+      "learning_rate": 1.985597390779201e-05,
+      "loss": 2.4565,
+      "step": 857
+    },
+    {
+      "epoch": 0.16517072935967467,
+      "grad_norm": 4.343106849979768,
+      "learning_rate": 1.9855446165005185e-05,
+      "loss": 2.4459,
+      "step": 858
+    },
+    {
+      "epoch": 0.16536323603725003,
+      "grad_norm": 4.063166972094065,
+      "learning_rate": 1.9854917464142903e-05,
+      "loss": 2.5496,
+      "step": 859
+    },
+    {
+      "epoch": 0.16555574271482543,
+      "grad_norm": 5.602233108872377,
+      "learning_rate": 1.9854387805256566e-05,
+      "loss": 2.4624,
+      "step": 860
+    },
+    {
+      "epoch": 0.1657482493924008,
+      "grad_norm": 3.4826940006454774,
+      "learning_rate": 1.9853857188397656e-05,
+      "loss": 2.4879,
+      "step": 861
+    },
+    {
+      "epoch": 0.1659407560699762,
+      "grad_norm": 4.0200536560380025,
+      "learning_rate": 1.985332561361776e-05,
+      "loss": 2.5539,
+      "step": 862
+    },
+    {
+      "epoch": 0.16613326274755155,
+      "grad_norm": 4.109384360418663,
+      "learning_rate": 1.9852793080968555e-05,
+      "loss": 2.4912,
+      "step": 863
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "grad_norm": 3.400800481928985,
+      "learning_rate": 1.9852259590501807e-05,
+      "loss": 2.4795,
+      "step": 864
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.1267,
+      "step": 864,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.4347,
+      "step": 864,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.3805,
+      "step": 864,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.279,
+      "step": 864,
+      "vm_loss": 0.2296
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.5198,
+      "step": 864,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.2566,
+      "step": 864,
+      "vm_loss": 0.2571
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.3167,
+      "step": 864,
+      "vm_loss": 0.2091
+    },
+    {
+      "epoch": 0.16632576942512695,
+      "lm_loss": 2.4634,
+      "step": 864,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 0.1665182761027023,
+      "grad_norm": 5.066341540872207,
+      "learning_rate": 1.9851725142269382e-05,
+      "loss": 2.5316,
+      "step": 865
+    },
+    {
+      "epoch": 0.16671078278027768,
+      "grad_norm": 3.7992754143895437,
+      "learning_rate": 1.9851189736323228e-05,
+      "loss": 2.5325,
+      "step": 866
+    },
+    {
+      "epoch": 0.16690328945785307,
+      "grad_norm": 5.075925160303168,
+      "learning_rate": 1.9850653372715403e-05,
+      "loss": 2.4576,
+      "step": 867
+    },
+    {
+      "epoch": 0.16709579613542844,
+      "grad_norm": 3.7371478522469217,
+      "learning_rate": 1.9850116051498043e-05,
+      "loss": 2.554,
+      "step": 868
+    },
+    {
+      "epoch": 0.16728830281300383,
+      "grad_norm": 4.527764418179241,
+      "learning_rate": 1.984957777272338e-05,
+      "loss": 2.4809,
+      "step": 869
+    },
+    {
+      "epoch": 0.1674808094905792,
+      "grad_norm": 4.445340125418732,
+      "learning_rate": 1.9849038536443746e-05,
+      "loss": 2.4831,
+      "step": 870
+    },
+    {
+      "epoch": 0.1676733161681546,
+      "grad_norm": 3.7644555071649606,
+      "learning_rate": 1.9848498342711563e-05,
+      "loss": 2.4676,
+      "step": 871
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "grad_norm": 4.221017632089052,
+      "learning_rate": 1.984795719157934e-05,
+      "loss": 2.4757,
+      "step": 872
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.0428,
+      "step": 872,
+      "vm_loss": 0.0955
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.2914,
+      "step": 872,
+      "vm_loss": 0.1962
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.3406,
+      "step": 872,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.4781,
+      "step": 872,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.5142,
+      "step": 872,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.6147,
+      "step": 872,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.3595,
+      "step": 872,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 0.16786582284572996,
+      "lm_loss": 2.5091,
+      "step": 872,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 0.16805832952330535,
+      "grad_norm": 3.6357447416847624,
+      "learning_rate": 1.9847415083099685e-05,
+      "loss": 2.491,
+      "step": 873
+    },
+    {
+      "epoch": 0.16825083620088072,
+      "grad_norm": 4.0084138737054555,
+      "learning_rate": 1.98468720173253e-05,
+      "loss": 2.5196,
+      "step": 874
+    },
+    {
+      "epoch": 0.1684433428784561,
+      "grad_norm": 3.5142195425760865,
+      "learning_rate": 1.9846327994308975e-05,
+      "loss": 2.5002,
+      "step": 875
+    },
+    {
+      "epoch": 0.16863584955603148,
+      "grad_norm": 4.427925772410468,
+      "learning_rate": 1.98457830141036e-05,
+      "loss": 2.4858,
+      "step": 876
+    },
+    {
+      "epoch": 0.16882835623360684,
+      "grad_norm": 4.542372084774133,
+      "learning_rate": 1.984523707676215e-05,
+      "loss": 2.5025,
+      "step": 877
+    },
+    {
+      "epoch": 0.16902086291118223,
+      "grad_norm": 3.8693440162781862,
+      "learning_rate": 1.98446901823377e-05,
+      "loss": 2.4649,
+      "step": 878
+    },
+    {
+      "epoch": 0.1692133695887576,
+      "grad_norm": 4.796691383754355,
+      "learning_rate": 1.984414233088341e-05,
+      "loss": 2.5548,
+      "step": 879
+    },
+    {
+      "epoch": 0.169405876266333,
+      "grad_norm": 4.551327231107463,
+      "learning_rate": 1.9843593522452546e-05,
+      "loss": 2.5246,
+      "step": 880
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.3588,
+      "step": 880,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.2537,
+      "step": 880,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.1825,
+      "step": 880,
+      "vm_loss": 0.1124
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.0459,
+      "step": 880,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.2926,
+      "step": 880,
+      "vm_loss": 0.2376
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.4112,
+      "step": 880,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.5111,
+      "step": 880,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.169405876266333,
+      "lm_loss": 2.6628,
+      "step": 880,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 0.16959838294390836,
+      "grad_norm": 4.122078849569165,
+      "learning_rate": 1.9843043757098456e-05,
+      "loss": 2.53,
+      "step": 881
+    },
+    {
+      "epoch": 0.16979088962148375,
+      "grad_norm": 4.036775529873049,
+      "learning_rate": 1.984249303487458e-05,
+      "loss": 2.5269,
+      "step": 882
+    },
+    {
+      "epoch": 0.16998339629905912,
+      "grad_norm": 4.021417786112161,
+      "learning_rate": 1.984194135583446e-05,
+      "loss": 2.4871,
+      "step": 883
+    },
+    {
+      "epoch": 0.1701759029766345,
+      "grad_norm": 3.3089788906978823,
+      "learning_rate": 1.9841388720031727e-05,
+      "loss": 2.4875,
+      "step": 884
+    },
+    {
+      "epoch": 0.17036840965420988,
+      "grad_norm": 3.795220669792721,
+      "learning_rate": 1.98408351275201e-05,
+      "loss": 2.4582,
+      "step": 885
+    },
+    {
+      "epoch": 0.17056091633178525,
+      "grad_norm": 3.6286940192394126,
+      "learning_rate": 1.9840280578353397e-05,
+      "loss": 2.4753,
+      "step": 886
+    },
+    {
+      "epoch": 0.17075342300936064,
+      "grad_norm": 4.533779684106499,
+      "learning_rate": 1.983972507258553e-05,
+      "loss": 2.48,
+      "step": 887
+    },
+    {
+      "epoch": 0.170945929686936,
+      "grad_norm": 3.487651041614942,
+      "learning_rate": 1.9839168610270498e-05,
+      "loss": 2.4603,
+      "step": 888
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.3489,
+      "step": 888,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.3151,
+      "step": 888,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.0978,
+      "step": 888,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.1745,
+      "step": 888,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.1319,
+      "step": 888,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.4738,
+      "step": 888,
+      "vm_loss": 0.2573
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.2832,
+      "step": 888,
+      "vm_loss": 0.2312
+    },
+    {
+      "epoch": 0.170945929686936,
+      "lm_loss": 2.6107,
+      "step": 888,
+      "vm_loss": 0.1171
+    },
+    {
+      "epoch": 0.1711384363645114,
+      "grad_norm": 3.3852869513560475,
+      "learning_rate": 1.98386111914624e-05,
+      "loss": 2.4894,
+      "step": 889
+    },
+    {
+      "epoch": 0.17133094304208676,
+      "grad_norm": 3.7437518199127693,
+      "learning_rate": 1.9838052816215416e-05,
+      "loss": 2.5299,
+      "step": 890
+    },
+    {
+      "epoch": 0.17152344971966216,
+      "grad_norm": 3.5930595410013817,
+      "learning_rate": 1.9837493484583838e-05,
+      "loss": 2.4558,
+      "step": 891
+    },
+    {
+      "epoch": 0.17171595639723752,
+      "grad_norm": 3.2708157426537423,
+      "learning_rate": 1.9836933196622033e-05,
+      "loss": 2.4517,
+      "step": 892
+    },
+    {
+      "epoch": 0.17190846307481292,
+      "grad_norm": 3.535767008734297,
+      "learning_rate": 1.983637195238447e-05,
+      "loss": 2.4494,
+      "step": 893
+    },
+    {
+      "epoch": 0.17210096975238828,
+      "grad_norm": 3.8074124989101965,
+      "learning_rate": 1.9835809751925708e-05,
+      "loss": 2.4525,
+      "step": 894
+    },
+    {
+      "epoch": 0.17229347642996368,
+      "grad_norm": 3.3519842148845527,
+      "learning_rate": 1.9835246595300405e-05,
+      "loss": 2.5064,
+      "step": 895
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "grad_norm": 3.541432511442601,
+      "learning_rate": 1.9834682482563302e-05,
+      "loss": 2.5128,
+      "step": 896
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.3096,
+      "step": 896,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.5469,
+      "step": 896,
+      "vm_loss": 0.283
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.3726,
+      "step": 896,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.3474,
+      "step": 896,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.3731,
+      "step": 896,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.3883,
+      "step": 896,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.4883,
+      "step": 896,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.17248598310753904,
+      "lm_loss": 2.5499,
+      "step": 896,
+      "vm_loss": 0.1793
+    },
+    {
+      "epoch": 0.1726784897851144,
+      "grad_norm": 3.37154568668484,
+      "learning_rate": 1.983411741376924e-05,
+      "loss": 2.528,
+      "step": 897
+    },
+    {
+      "epoch": 0.1728709964626898,
+      "grad_norm": 3.6126627739882924,
+      "learning_rate": 1.9833551388973147e-05,
+      "loss": 2.565,
+      "step": 898
+    },
+    {
+      "epoch": 0.17306350314026517,
+      "grad_norm": 4.034874715261827,
+      "learning_rate": 1.9832984408230054e-05,
+      "loss": 2.5266,
+      "step": 899
+    },
+    {
+      "epoch": 0.17325600981784056,
+      "grad_norm": 3.5657829981459397,
+      "learning_rate": 1.9832416471595074e-05,
+      "loss": 2.4913,
+      "step": 900
+    },
+    {
+      "epoch": 0.17344851649541593,
+      "grad_norm": 3.9712059504488115,
+      "learning_rate": 1.983184757912342e-05,
+      "loss": 2.4832,
+      "step": 901
+    },
+    {
+      "epoch": 0.17364102317299132,
+      "grad_norm": 3.1856571783874323,
+      "learning_rate": 1.9831277730870396e-05,
+      "loss": 2.4019,
+      "step": 902
+    },
+    {
+      "epoch": 0.1738335298505667,
+      "grad_norm": 3.7649483355730653,
+      "learning_rate": 1.9830706926891396e-05,
+      "loss": 2.4939,
+      "step": 903
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "grad_norm": 3.529662089759606,
+      "learning_rate": 1.9830135167241916e-05,
+      "loss": 2.5236,
+      "step": 904
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.1607,
+      "step": 904,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.3304,
+      "step": 904,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.3995,
+      "step": 904,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 1.8449,
+      "step": 904,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.53,
+      "step": 904,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.3159,
+      "step": 904,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.3825,
+      "step": 904,
+      "vm_loss": 0.2446
+    },
+    {
+      "epoch": 0.17402603652814208,
+      "lm_loss": 2.4334,
+      "step": 904,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.17421854320571745,
+      "grad_norm": 3.519274166730685,
+      "learning_rate": 1.982956245197753e-05,
+      "loss": 2.4378,
+      "step": 905
+    },
+    {
+      "epoch": 0.17441104988329284,
+      "grad_norm": 4.200817241002008,
+      "learning_rate": 1.9828988781153916e-05,
+      "loss": 2.5199,
+      "step": 906
+    },
+    {
+      "epoch": 0.1746035565608682,
+      "grad_norm": 4.039331012621496,
+      "learning_rate": 1.9828414154826845e-05,
+      "loss": 2.5342,
+      "step": 907
+    },
+    {
+      "epoch": 0.17479606323844357,
+      "grad_norm": 3.8594096740692976,
+      "learning_rate": 1.9827838573052176e-05,
+      "loss": 2.5359,
+      "step": 908
+    },
+    {
+      "epoch": 0.17498856991601897,
+      "grad_norm": 4.299297151095028,
+      "learning_rate": 1.982726203588586e-05,
+      "loss": 2.4471,
+      "step": 909
+    },
+    {
+      "epoch": 0.17518107659359433,
+      "grad_norm": 3.7133107730609445,
+      "learning_rate": 1.982668454338395e-05,
+      "loss": 2.5135,
+      "step": 910
+    },
+    {
+      "epoch": 0.17537358327116973,
+      "grad_norm": 4.271758266916981,
+      "learning_rate": 1.9826106095602582e-05,
+      "loss": 2.4855,
+      "step": 911
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "grad_norm": 4.168423781562003,
+      "learning_rate": 1.982552669259799e-05,
+      "loss": 2.5005,
+      "step": 912
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.1874,
+      "step": 912,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.3565,
+      "step": 912,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.3739,
+      "step": 912,
+      "vm_loss": 0.1962
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.1453,
+      "step": 912,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.4343,
+      "step": 912,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.3087,
+      "step": 912,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.2499,
+      "step": 912,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 0.1755660899487451,
+      "lm_loss": 2.2876,
+      "step": 912,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.17575859662632048,
+      "grad_norm": 3.3981216667121847,
+      "learning_rate": 1.9824946334426497e-05,
+      "loss": 2.4972,
+      "step": 913
+    },
+    {
+      "epoch": 0.17595110330389585,
+      "grad_norm": 4.79842294820674,
+      "learning_rate": 1.982436502114452e-05,
+      "loss": 2.4711,
+      "step": 914
+    },
+    {
+      "epoch": 0.17614360998147124,
+      "grad_norm": 3.76316178362478,
+      "learning_rate": 1.9823782752808574e-05,
+      "loss": 2.4849,
+      "step": 915
+    },
+    {
+      "epoch": 0.1763361166590466,
+      "grad_norm": 4.552832711876154,
+      "learning_rate": 1.9823199529475265e-05,
+      "loss": 2.4903,
+      "step": 916
+    },
+    {
+      "epoch": 0.17652862333662198,
+      "grad_norm": 3.493438864885672,
+      "learning_rate": 1.9822615351201283e-05,
+      "loss": 2.5204,
+      "step": 917
+    },
+    {
+      "epoch": 0.17672113001419737,
+      "grad_norm": 3.2804991008916864,
+      "learning_rate": 1.982203021804342e-05,
+      "loss": 2.48,
+      "step": 918
+    },
+    {
+      "epoch": 0.17691363669177274,
+      "grad_norm": 3.172639943448026,
+      "learning_rate": 1.982144413005856e-05,
+      "loss": 2.4697,
+      "step": 919
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "grad_norm": 3.302360975300849,
+      "learning_rate": 1.982085708730368e-05,
+      "loss": 2.4715,
+      "step": 920
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.3409,
+      "step": 920,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.429,
+      "step": 920,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.0471,
+      "step": 920,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 1.8902,
+      "step": 920,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.2868,
+      "step": 920,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.4169,
+      "step": 920,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.5888,
+      "step": 920,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.17710614336934813,
+      "lm_loss": 2.3245,
+      "step": 920,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 0.1772986500469235,
+      "grad_norm": 3.523413540578112,
+      "learning_rate": 1.9820269089835843e-05,
+      "loss": 2.4697,
+      "step": 921
+    },
+    {
+      "epoch": 0.1774911567244989,
+      "grad_norm": 4.113064333429806,
+      "learning_rate": 1.9819680137712214e-05,
+      "loss": 2.5192,
+      "step": 922
+    },
+    {
+      "epoch": 0.17768366340207425,
+      "grad_norm": 3.140013825303239,
+      "learning_rate": 1.9819090230990046e-05,
+      "loss": 2.4435,
+      "step": 923
+    },
+    {
+      "epoch": 0.17787617007964965,
+      "grad_norm": 4.368978262255207,
+      "learning_rate": 1.9818499369726684e-05,
+      "loss": 2.5135,
+      "step": 924
+    },
+    {
+      "epoch": 0.17806867675722501,
+      "grad_norm": 4.438369945176267,
+      "learning_rate": 1.9817907553979566e-05,
+      "loss": 2.5502,
+      "step": 925
+    },
+    {
+      "epoch": 0.1782611834348004,
+      "grad_norm": 3.7076680064270757,
+      "learning_rate": 1.9817314783806226e-05,
+      "loss": 2.4497,
+      "step": 926
+    },
+    {
+      "epoch": 0.17845369011237577,
+      "grad_norm": 3.9828981444572293,
+      "learning_rate": 1.9816721059264294e-05,
+      "loss": 2.5006,
+      "step": 927
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "grad_norm": 4.0852706255795646,
+      "learning_rate": 1.9816126380411478e-05,
+      "loss": 2.4854,
+      "step": 928
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.4162,
+      "step": 928,
+      "vm_loss": 0.1811
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.3802,
+      "step": 928,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.3401,
+      "step": 928,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.3518,
+      "step": 928,
+      "vm_loss": 0.0999
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 1.9935,
+      "step": 928,
+      "vm_loss": 0.2272
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.1281,
+      "step": 928,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.1832,
+      "step": 928,
+      "vm_loss": 0.2207
+    },
+    {
+      "epoch": 0.17864619678995114,
+      "lm_loss": 2.1482,
+      "step": 928,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 0.17883870346752653,
+      "grad_norm": 4.105545881370151,
+      "learning_rate": 1.9815530747305594e-05,
+      "loss": 2.491,
+      "step": 929
+    },
+    {
+      "epoch": 0.1790312101451019,
+      "grad_norm": 4.25919193880954,
+      "learning_rate": 1.9814934160004543e-05,
+      "loss": 2.5171,
+      "step": 930
+    },
+    {
+      "epoch": 0.1792237168226773,
+      "grad_norm": 3.7556766444343714,
+      "learning_rate": 1.9814336618566324e-05,
+      "loss": 2.4845,
+      "step": 931
+    },
+    {
+      "epoch": 0.17941622350025266,
+      "grad_norm": 3.4500733338305527,
+      "learning_rate": 1.9813738123049022e-05,
+      "loss": 2.5274,
+      "step": 932
+    },
+    {
+      "epoch": 0.17960873017782805,
+      "grad_norm": 3.7655259561953063,
+      "learning_rate": 1.981313867351082e-05,
+      "loss": 2.5393,
+      "step": 933
+    },
+    {
+      "epoch": 0.17980123685540342,
+      "grad_norm": 4.081020466463539,
+      "learning_rate": 1.9812538270009995e-05,
+      "loss": 2.4465,
+      "step": 934
+    },
+    {
+      "epoch": 0.1799937435329788,
+      "grad_norm": 3.3199140286584705,
+      "learning_rate": 1.981193691260491e-05,
+      "loss": 2.4707,
+      "step": 935
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "grad_norm": 3.781768576733766,
+      "learning_rate": 1.9811334601354023e-05,
+      "loss": 2.4506,
+      "step": 936
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.6181,
+      "step": 936,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 1.8245,
+      "step": 936,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.3441,
+      "step": 936,
+      "vm_loss": 0.2084
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.433,
+      "step": 936,
+      "vm_loss": 0.2126
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 1.9038,
+      "step": 936,
+      "vm_loss": 0.225
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.3011,
+      "step": 936,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.2468,
+      "step": 936,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.18018625021055418,
+      "lm_loss": 2.2464,
+      "step": 936,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 0.18037875688812954,
+      "grad_norm": 3.7818729181062607,
+      "learning_rate": 1.981073133631589e-05,
+      "loss": 2.4634,
+      "step": 937
+    },
+    {
+      "epoch": 0.18057126356570494,
+      "grad_norm": 4.113065116976769,
+      "learning_rate": 1.9810127117549158e-05,
+      "loss": 2.474,
+      "step": 938
+    },
+    {
+      "epoch": 0.1807637702432803,
+      "grad_norm": 4.041987198703892,
+      "learning_rate": 1.980952194511256e-05,
+      "loss": 2.4222,
+      "step": 939
+    },
+    {
+      "epoch": 0.1809562769208557,
+      "grad_norm": 3.821931670421027,
+      "learning_rate": 1.9808915819064926e-05,
+      "loss": 2.4771,
+      "step": 940
+    },
+    {
+      "epoch": 0.18114878359843106,
+      "grad_norm": 3.6492934028917734,
+      "learning_rate": 1.9808308739465184e-05,
+      "loss": 2.5089,
+      "step": 941
+    },
+    {
+      "epoch": 0.18134129027600646,
+      "grad_norm": 4.275026258897681,
+      "learning_rate": 1.980770070637235e-05,
+      "loss": 2.4623,
+      "step": 942
+    },
+    {
+      "epoch": 0.18153379695358182,
+      "grad_norm": 3.9938233056880157,
+      "learning_rate": 1.9807091719845527e-05,
+      "loss": 2.4876,
+      "step": 943
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "grad_norm": 3.6050250421607037,
+      "learning_rate": 1.9806481779943922e-05,
+      "loss": 2.5289,
+      "step": 944
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 1.8762,
+      "step": 944,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.4668,
+      "step": 944,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.252,
+      "step": 944,
+      "vm_loss": 0.2038
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.4318,
+      "step": 944,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.4921,
+      "step": 944,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.2405,
+      "step": 944,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 1.9509,
+      "step": 944,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.18172630363115722,
+      "lm_loss": 2.14,
+      "step": 944,
+      "vm_loss": 0.1146
+    },
+    {
+      "epoch": 0.18191881030873258,
+      "grad_norm": 4.6365375048485,
+      "learning_rate": 1.9805870886726826e-05,
+      "loss": 2.4836,
+      "step": 945
+    },
+    {
+      "epoch": 0.18211131698630798,
+      "grad_norm": 3.7434358690966443,
+      "learning_rate": 1.9805259040253624e-05,
+      "loss": 2.4693,
+      "step": 946
+    },
+    {
+      "epoch": 0.18230382366388334,
+      "grad_norm": 4.356568670475241,
+      "learning_rate": 1.9804646240583797e-05,
+      "loss": 2.5226,
+      "step": 947
+    },
+    {
+      "epoch": 0.1824963303414587,
+      "grad_norm": 3.989996873737595,
+      "learning_rate": 1.980403248777692e-05,
+      "loss": 2.538,
+      "step": 948
+    },
+    {
+      "epoch": 0.1826888370190341,
+      "grad_norm": 3.731461961837993,
+      "learning_rate": 1.9803417781892655e-05,
+      "loss": 2.4724,
+      "step": 949
+    },
+    {
+      "epoch": 0.18288134369660947,
+      "grad_norm": 3.75401515076171,
+      "learning_rate": 1.9802802122990758e-05,
+      "loss": 2.4643,
+      "step": 950
+    },
+    {
+      "epoch": 0.18307385037418486,
+      "grad_norm": 3.6429819788861044,
+      "learning_rate": 1.980218551113108e-05,
+      "loss": 2.4784,
+      "step": 951
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "grad_norm": 4.326937072731741,
+      "learning_rate": 1.9801567946373564e-05,
+      "loss": 2.4308,
+      "step": 952
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.3545,
+      "step": 952,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.5824,
+      "step": 952,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 1.7718,
+      "step": 952,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.1943,
+      "step": 952,
+      "vm_loss": 0.1012
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.5159,
+      "step": 952,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.125,
+      "step": 952,
+      "vm_loss": 0.2222
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.3212,
+      "step": 952,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 0.18326635705176023,
+      "lm_loss": 2.2838,
+      "step": 952,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.18345886372933562,
+      "grad_norm": 3.7770294529989483,
+      "learning_rate": 1.9800949428778245e-05,
+      "loss": 2.4882,
+      "step": 953
+    },
+    {
+      "epoch": 0.18365137040691099,
+      "grad_norm": 3.2466566302180664,
+      "learning_rate": 1.9800329958405254e-05,
+      "loss": 2.4741,
+      "step": 954
+    },
+    {
+      "epoch": 0.18384387708448638,
+      "grad_norm": 3.910691484176159,
+      "learning_rate": 1.9799709535314805e-05,
+      "loss": 2.4963,
+      "step": 955
+    },
+    {
+      "epoch": 0.18403638376206174,
+      "grad_norm": 3.8286337347381654,
+      "learning_rate": 1.9799088159567213e-05,
+      "loss": 2.5175,
+      "step": 956
+    },
+    {
+      "epoch": 0.18422889043963714,
+      "grad_norm": 3.5316038579186113,
+      "learning_rate": 1.9798465831222885e-05,
+      "loss": 2.499,
+      "step": 957
+    },
+    {
+      "epoch": 0.1844213971172125,
+      "grad_norm": 4.1670051970138475,
+      "learning_rate": 1.979784255034232e-05,
+      "loss": 2.4936,
+      "step": 958
+    },
+    {
+      "epoch": 0.18461390379478787,
+      "grad_norm": 3.385429672109255,
+      "learning_rate": 1.9797218316986112e-05,
+      "loss": 2.4689,
+      "step": 959
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "grad_norm": 3.3078136609460094,
+      "learning_rate": 1.9796593131214935e-05,
+      "loss": 2.4422,
+      "step": 960
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.2358,
+      "step": 960,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.2605,
+      "step": 960,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.423,
+      "step": 960,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.3081,
+      "step": 960,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 1.4558,
+      "step": 960,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.3081,
+      "step": 960,
+      "vm_loss": 0.2512
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.286,
+      "step": 960,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.18480641047236326,
+      "lm_loss": 2.7513,
+      "step": 960,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 0.18499891714993863,
+      "grad_norm": 3.3534005649504213,
+      "learning_rate": 1.9795966993089574e-05,
+      "loss": 2.4932,
+      "step": 961
+    },
+    {
+      "epoch": 0.18519142382751402,
+      "grad_norm": 3.5507171279996403,
+      "learning_rate": 1.9795339902670893e-05,
+      "loss": 2.4604,
+      "step": 962
+    },
+    {
+      "epoch": 0.1853839305050894,
+      "grad_norm": 3.5148603601270105,
+      "learning_rate": 1.9794711860019852e-05,
+      "loss": 2.4512,
+      "step": 963
+    },
+    {
+      "epoch": 0.18557643718266478,
+      "grad_norm": 3.4062079659895774,
+      "learning_rate": 1.979408286519751e-05,
+      "loss": 2.5454,
+      "step": 964
+    },
+    {
+      "epoch": 0.18576894386024015,
+      "grad_norm": 3.997495292221776,
+      "learning_rate": 1.979345291826501e-05,
+      "loss": 2.4384,
+      "step": 965
+    },
+    {
+      "epoch": 0.18596145053781554,
+      "grad_norm": 3.4630392062229323,
+      "learning_rate": 1.9792822019283594e-05,
+      "loss": 2.4567,
+      "step": 966
+    },
+    {
+      "epoch": 0.1861539572153909,
+      "grad_norm": 4.048113515227834,
+      "learning_rate": 1.979219016831459e-05,
+      "loss": 2.4567,
+      "step": 967
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "grad_norm": 3.4428917692998025,
+      "learning_rate": 1.979155736541942e-05,
+      "loss": 2.4915,
+      "step": 968
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.6,
+      "step": 968,
+      "vm_loss": 0.2593
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.2453,
+      "step": 968,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.4075,
+      "step": 968,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.2265,
+      "step": 968,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.3331,
+      "step": 968,
+      "vm_loss": 0.2269
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.0818,
+      "step": 968,
+      "vm_loss": 0.2031
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.2525,
+      "step": 968,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 0.18634646389296627,
+      "lm_loss": 2.566,
+      "step": 968,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 0.18653897057054167,
+      "grad_norm": 3.9272927603774876,
+      "learning_rate": 1.9790923610659604e-05,
+      "loss": 2.4994,
+      "step": 969
+    },
+    {
+      "epoch": 0.18673147724811703,
+      "grad_norm": 3.6329926998473305,
+      "learning_rate": 1.9790288904096754e-05,
+      "loss": 2.4511,
+      "step": 970
+    },
+    {
+      "epoch": 0.18692398392569243,
+      "grad_norm": 4.653108912322447,
+      "learning_rate": 1.9789653245792565e-05,
+      "loss": 2.4973,
+      "step": 971
+    },
+    {
+      "epoch": 0.1871164906032678,
+      "grad_norm": 4.087354006557861,
+      "learning_rate": 1.9789016635808836e-05,
+      "loss": 2.5106,
+      "step": 972
+    },
+    {
+      "epoch": 0.1873089972808432,
+      "grad_norm": 3.4705490011992333,
+      "learning_rate": 1.978837907420745e-05,
+      "loss": 2.4917,
+      "step": 973
+    },
+    {
+      "epoch": 0.18750150395841855,
+      "grad_norm": 3.780693667091048,
+      "learning_rate": 1.978774056105039e-05,
+      "loss": 2.5198,
+      "step": 974
+    },
+    {
+      "epoch": 0.18769401063599395,
+      "grad_norm": 3.306947713109733,
+      "learning_rate": 1.9787101096399727e-05,
+      "loss": 2.4992,
+      "step": 975
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "grad_norm": 3.9278538748370773,
+      "learning_rate": 1.978646068031762e-05,
+      "loss": 2.4832,
+      "step": 976
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.3012,
+      "step": 976,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.45,
+      "step": 976,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.3898,
+      "step": 976,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.3723,
+      "step": 976,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.063,
+      "step": 976,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.4101,
+      "step": 976,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.1547,
+      "step": 976,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.1878865173135693,
+      "lm_loss": 2.0522,
+      "step": 976,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 0.1880790239911447,
+      "grad_norm": 4.575225456367262,
+      "learning_rate": 1.978581931286633e-05,
+      "loss": 2.4422,
+      "step": 977
+    },
+    {
+      "epoch": 0.18827153066872007,
+      "grad_norm": 3.8872161486108014,
+      "learning_rate": 1.9785176994108207e-05,
+      "loss": 2.4964,
+      "step": 978
+    },
+    {
+      "epoch": 0.18846403734629544,
+      "grad_norm": 4.137412348549307,
+      "learning_rate": 1.9784533724105692e-05,
+      "loss": 2.5189,
+      "step": 979
+    },
+    {
+      "epoch": 0.18865654402387083,
+      "grad_norm": 4.200131678711624,
+      "learning_rate": 1.9783889502921315e-05,
+      "loss": 2.4432,
+      "step": 980
+    },
+    {
+      "epoch": 0.1888490507014462,
+      "grad_norm": 3.2020051860280714,
+      "learning_rate": 1.9783244330617713e-05,
+      "loss": 2.5203,
+      "step": 981
+    },
+    {
+      "epoch": 0.1890415573790216,
+      "grad_norm": 4.378339167178262,
+      "learning_rate": 1.9782598207257592e-05,
+      "loss": 2.514,
+      "step": 982
+    },
+    {
+      "epoch": 0.18923406405659696,
+      "grad_norm": 3.8179046337961053,
+      "learning_rate": 1.978195113290377e-05,
+      "loss": 2.4961,
+      "step": 983
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "grad_norm": 3.795245357283514,
+      "learning_rate": 1.978130310761915e-05,
+      "loss": 2.4645,
+      "step": 984
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.3467,
+      "step": 984,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.3725,
+      "step": 984,
+      "vm_loss": 0.2331
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.5058,
+      "step": 984,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.4235,
+      "step": 984,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.4651,
+      "step": 984,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.2615,
+      "step": 984,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.19,
+      "step": 984,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.18942657073417235,
+      "lm_loss": 2.3705,
+      "step": 984,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 0.18961907741174772,
+      "grad_norm": 3.7093620016369746,
+      "learning_rate": 1.978065413146673e-05,
+      "loss": 2.4914,
+      "step": 985
+    },
+    {
+      "epoch": 0.1898115840893231,
+      "grad_norm": 3.7049314660779817,
+      "learning_rate": 1.9780004204509594e-05,
+      "loss": 2.4619,
+      "step": 986
+    },
+    {
+      "epoch": 0.19000409076689848,
+      "grad_norm": 3.8158072288365994,
+      "learning_rate": 1.977935332681093e-05,
+      "loss": 2.4841,
+      "step": 987
+    },
+    {
+      "epoch": 0.19019659744447384,
+      "grad_norm": 3.6220529851068846,
+      "learning_rate": 1.9778701498434007e-05,
+      "loss": 2.5257,
+      "step": 988
+    },
+    {
+      "epoch": 0.19038910412204924,
+      "grad_norm": 4.103434716649326,
+      "learning_rate": 1.977804871944219e-05,
+      "loss": 2.5211,
+      "step": 989
+    },
+    {
+      "epoch": 0.1905816107996246,
+      "grad_norm": 4.3867567449904685,
+      "learning_rate": 1.9777394989898947e-05,
+      "loss": 2.4672,
+      "step": 990
+    },
+    {
+      "epoch": 0.1907741174772,
+      "grad_norm": 3.935406093562076,
+      "learning_rate": 1.9776740309867813e-05,
+      "loss": 2.4706,
+      "step": 991
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "grad_norm": 3.5199978074956046,
+      "learning_rate": 1.9776084679412443e-05,
+      "loss": 2.5139,
+      "step": 992
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.2626,
+      "step": 992,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.2582,
+      "step": 992,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.5946,
+      "step": 992,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.187,
+      "step": 992,
+      "vm_loss": 0.1913
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.219,
+      "step": 992,
+      "vm_loss": 0.2317
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.1164,
+      "step": 992,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.4844,
+      "step": 992,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 0.19096662415477536,
+      "lm_loss": 2.3042,
+      "step": 992,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.19115913083235075,
+      "grad_norm": 4.032690702016478,
+      "learning_rate": 1.977542809859657e-05,
+      "loss": 2.487,
+      "step": 993
+    },
+    {
+      "epoch": 0.19135163750992612,
+      "grad_norm": 3.7662215343923013,
+      "learning_rate": 1.977477056748402e-05,
+      "loss": 2.458,
+      "step": 994
+    },
+    {
+      "epoch": 0.1915441441875015,
+      "grad_norm": 3.4108699826658255,
+      "learning_rate": 1.9774112086138718e-05,
+      "loss": 2.4901,
+      "step": 995
+    },
+    {
+      "epoch": 0.19173665086507688,
+      "grad_norm": 3.262581332430276,
+      "learning_rate": 1.9773452654624675e-05,
+      "loss": 2.4222,
+      "step": 996
+    },
+    {
+      "epoch": 0.19192915754265227,
+      "grad_norm": 3.4279722162029973,
+      "learning_rate": 1.9772792273005993e-05,
+      "loss": 2.4488,
+      "step": 997
+    },
+    {
+      "epoch": 0.19212166422022764,
+      "grad_norm": 3.490304265099083,
+      "learning_rate": 1.9772130941346866e-05,
+      "loss": 2.5138,
+      "step": 998
+    },
+    {
+      "epoch": 0.192314170897803,
+      "grad_norm": 3.4614595889352406,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 2.4717,
+      "step": 999
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "grad_norm": 3.9858027067320045,
+      "learning_rate": 1.9770805428164558e-05,
+      "loss": 2.5017,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 1.9985,
+      "step": 1000,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.2382,
+      "step": 1000,
+      "vm_loss": 0.305
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.1586,
+      "step": 1000,
+      "vm_loss": 0.2122
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.445,
+      "step": 1000,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 1.9637,
+      "step": 1000,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.442,
+      "step": 1000,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.3797,
+      "step": 1000,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 0.1925066775753784,
+      "lm_loss": 2.2713,
+      "step": 1000,
+      "vm_loss": 0.2927
+    },
+    {
+      "epoch": 0.19269918425295376,
+      "grad_norm": 3.957687521447021,
+      "learning_rate": 1.9770141246770226e-05,
+      "loss": 2.4373,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19289169093052916,
+      "grad_norm": 3.4806180075557793,
+      "learning_rate": 1.976947611559317e-05,
+      "loss": 2.5069,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19308419760810452,
+      "grad_norm": 3.5292738481587933,
+      "learning_rate": 1.9768810034698044e-05,
+      "loss": 2.4729,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19327670428567992,
+      "grad_norm": 3.4395583234702505,
+      "learning_rate": 1.9768143004149607e-05,
+      "loss": 2.4949,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19346921096325528,
+      "grad_norm": 3.6927098490518886,
+      "learning_rate": 1.9767475024012698e-05,
+      "loss": 2.5255,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19366171764083068,
+      "grad_norm": 3.461277809334239,
+      "learning_rate": 1.9766806094352252e-05,
+      "loss": 2.5147,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19385422431840604,
+      "grad_norm": 3.4270695858874936,
+      "learning_rate": 1.9766136215233297e-05,
+      "loss": 2.4699,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "grad_norm": 4.211807705414343,
+      "learning_rate": 1.9765465386720963e-05,
+      "loss": 2.465,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.1625,
+      "step": 1008,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.4017,
+      "step": 1008,
+      "vm_loss": 0.2301
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.5536,
+      "step": 1008,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.126,
+      "step": 1008,
+      "vm_loss": 0.3028
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.1502,
+      "step": 1008,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.2478,
+      "step": 1008,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.1107,
+      "step": 1008,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 0.19404673099598144,
+      "lm_loss": 2.1554,
+      "step": 1008,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.1942392376735568,
+      "grad_norm": 3.159723117226434,
+      "learning_rate": 1.9764793608880454e-05,
+      "loss": 2.4801,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19443174435113217,
+      "grad_norm": 3.540448326004961,
+      "learning_rate": 1.9764120881777077e-05,
+      "loss": 2.4163,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19462425102870756,
+      "grad_norm": 3.776630767621123,
+      "learning_rate": 1.976344720547623e-05,
+      "loss": 2.4728,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19481675770628293,
+      "grad_norm": 2.9084201315534233,
+      "learning_rate": 1.976277258004341e-05,
+      "loss": 2.4444,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19500926438385832,
+      "grad_norm": 3.2219226578541766,
+      "learning_rate": 1.9762097005544187e-05,
+      "loss": 2.4562,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1952017710614337,
+      "grad_norm": 3.5620050413053703,
+      "learning_rate": 1.976142048204424e-05,
+      "loss": 2.4781,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19539427773900908,
+      "grad_norm": 3.1703341459018115,
+      "learning_rate": 1.976074300960934e-05,
+      "loss": 2.4632,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "grad_norm": 3.616403122568653,
+      "learning_rate": 1.9760064588305347e-05,
+      "loss": 2.4929,
+      "step": 1016
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.2893,
+      "step": 1016,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.3458,
+      "step": 1016,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.4711,
+      "step": 1016,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.4778,
+      "step": 1016,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.7431,
+      "step": 1016,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.2508,
+      "step": 1016,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.3955,
+      "step": 1016,
+      "vm_loss": 0.19
+    },
+    {
+      "epoch": 0.19558678441658445,
+      "lm_loss": 2.3489,
+      "step": 1016,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.19577929109415984,
+      "grad_norm": 3.2476158500938044,
+      "learning_rate": 1.9759385218198203e-05,
+      "loss": 2.4891,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1959717977717352,
+      "grad_norm": 4.018487468468996,
+      "learning_rate": 1.9758704899353954e-05,
+      "loss": 2.4978,
+      "step": 1018
+    },
+    {
+      "epoch": 0.19616430444931057,
+      "grad_norm": 3.491169193715776,
+      "learning_rate": 1.9758023631838744e-05,
+      "loss": 2.4992,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19635681112688597,
+      "grad_norm": 4.7145408455772335,
+      "learning_rate": 1.9757341415718794e-05,
+      "loss": 2.5322,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19654931780446133,
+      "grad_norm": 3.162191584676517,
+      "learning_rate": 1.9756658251060425e-05,
+      "loss": 2.5072,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19674182448203673,
+      "grad_norm": 3.818865511666175,
+      "learning_rate": 1.9755974137930053e-05,
+      "loss": 2.466,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1969343311596121,
+      "grad_norm": 3.5139150620052835,
+      "learning_rate": 1.9755289076394176e-05,
+      "loss": 2.4518,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "grad_norm": 3.7638917320882177,
+      "learning_rate": 1.9754603066519395e-05,
+      "loss": 2.4568,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.2327,
+      "step": 1024,
+      "vm_loss": 0.2077
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 1.857,
+      "step": 1024,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.0651,
+      "step": 1024,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.5391,
+      "step": 1024,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.4881,
+      "step": 1024,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.3546,
+      "step": 1024,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.4412,
+      "step": 1024,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 0.19712683783718749,
+      "lm_loss": 2.289,
+      "step": 1024,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.19731934451476285,
+      "grad_norm": 3.386152807398567,
+      "learning_rate": 1.97539161083724e-05,
+      "loss": 2.5099,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19751185119233824,
+      "grad_norm": 3.9435749377278726,
+      "learning_rate": 1.9753228202019966e-05,
+      "loss": 2.4697,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1977043578699136,
+      "grad_norm": 3.5964264735097284,
+      "learning_rate": 1.9752539347528973e-05,
+      "loss": 2.4387,
+      "step": 1027
+    },
+    {
+      "epoch": 0.197896864547489,
+      "grad_norm": 3.9519620118813883,
+      "learning_rate": 1.9751849544966385e-05,
+      "loss": 2.4901,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19808937122506437,
+      "grad_norm": 3.3338873628092025,
+      "learning_rate": 1.9751158794399255e-05,
+      "loss": 2.4591,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19828187790263974,
+      "grad_norm": 3.4342699536094448,
+      "learning_rate": 1.9750467095894738e-05,
+      "loss": 2.4724,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19847438458021513,
+      "grad_norm": 3.987252007022662,
+      "learning_rate": 1.9749774449520078e-05,
+      "loss": 2.4988,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "grad_norm": 3.2412066444484293,
+      "learning_rate": 1.97490808553426e-05,
+      "loss": 2.4671,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.4341,
+      "step": 1032,
+      "vm_loss": 0.1054
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.0009,
+      "step": 1032,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.0975,
+      "step": 1032,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.3541,
+      "step": 1032,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.2321,
+      "step": 1032,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.0613,
+      "step": 1032,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.4225,
+      "step": 1032,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.1986668912577905,
+      "lm_loss": 2.1777,
+      "step": 1032,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.1988593979353659,
+      "grad_norm": 3.4767540050777366,
+      "learning_rate": 1.974838631342974e-05,
+      "loss": 2.4279,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19905190461294126,
+      "grad_norm": 3.4765721191917813,
+      "learning_rate": 1.9747690823849012e-05,
+      "loss": 2.4845,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19924441129051665,
+      "grad_norm": 4.157545201488476,
+      "learning_rate": 1.9746994386668022e-05,
+      "loss": 2.4575,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19943691796809201,
+      "grad_norm": 3.5651942757837873,
+      "learning_rate": 1.974629700195448e-05,
+      "loss": 2.4731,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1996294246456674,
+      "grad_norm": 3.3897257561244434,
+      "learning_rate": 1.974559866977618e-05,
+      "loss": 2.496,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19982193132324277,
+      "grad_norm": 3.8318341664084294,
+      "learning_rate": 1.9744899390201006e-05,
+      "loss": 2.5046,
+      "step": 1038
+    },
+    {
+      "epoch": 0.20001443800081814,
+      "grad_norm": 3.417175940392068,
+      "learning_rate": 1.9744199163296934e-05,
+      "loss": 2.4673,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "grad_norm": 3.3985560683942406,
+      "learning_rate": 1.9743497989132045e-05,
+      "loss": 2.4647,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.0189,
+      "step": 1040,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.1004,
+      "step": 1040,
+      "vm_loss": 0.2506
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.2524,
+      "step": 1040,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.1861,
+      "step": 1040,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.1416,
+      "step": 1040,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.2216,
+      "step": 1040,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.3397,
+      "step": 1040,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 0.20020694467839353,
+      "lm_loss": 2.4533,
+      "step": 1040,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.2003994513559689,
+      "grad_norm": 3.8489824363798104,
+      "learning_rate": 1.974279586777449e-05,
+      "loss": 2.441,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2005919580335443,
+      "grad_norm": 3.9382097624886683,
+      "learning_rate": 1.974209279929253e-05,
+      "loss": 2.4659,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20078446471111966,
+      "grad_norm": 3.4128789112293547,
+      "learning_rate": 1.9741388783754515e-05,
+      "loss": 2.4814,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20097697138869505,
+      "grad_norm": 4.191056146380478,
+      "learning_rate": 1.974068382122888e-05,
+      "loss": 2.4609,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20116947806627042,
+      "grad_norm": 3.4651483243464045,
+      "learning_rate": 1.973997791178416e-05,
+      "loss": 2.49,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2013619847438458,
+      "grad_norm": 3.5084783935666466,
+      "learning_rate": 1.9739271055488972e-05,
+      "loss": 2.4891,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20155449142142118,
+      "grad_norm": 4.0241057682700925,
+      "learning_rate": 1.973856325241204e-05,
+      "loss": 2.4783,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "grad_norm": 3.6096856231782635,
+      "learning_rate": 1.9737854502622165e-05,
+      "loss": 2.4573,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 1.9424,
+      "step": 1048,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.4505,
+      "step": 1048,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.2513,
+      "step": 1048,
+      "vm_loss": 0.0989
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.18,
+      "step": 1048,
+      "vm_loss": 0.2511
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.4586,
+      "step": 1048,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.4717,
+      "step": 1048,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.295,
+      "step": 1048,
+      "vm_loss": 0.2642
+    },
+    {
+      "epoch": 0.20174699809899657,
+      "lm_loss": 2.149,
+      "step": 1048,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.20193950477657194,
+      "grad_norm": 3.6071298244581804,
+      "learning_rate": 1.9737144806188248e-05,
+      "loss": 2.4893,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2021320114541473,
+      "grad_norm": 3.6876354499888166,
+      "learning_rate": 1.9736434163179284e-05,
+      "loss": 2.5184,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2023245181317227,
+      "grad_norm": 3.7532035449555123,
+      "learning_rate": 1.973572257366435e-05,
+      "loss": 2.4624,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20251702480929806,
+      "grad_norm": 4.170890971005629,
+      "learning_rate": 1.9735010037712628e-05,
+      "loss": 2.4604,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20270953148687346,
+      "grad_norm": 4.349845014346077,
+      "learning_rate": 1.9734296555393385e-05,
+      "loss": 2.5229,
+      "step": 1053
+    },
+    {
+      "epoch": 0.20290203816444882,
+      "grad_norm": 3.7706674313057746,
+      "learning_rate": 1.9733582126775978e-05,
+      "loss": 2.4935,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20309454484202422,
+      "grad_norm": 3.66283713202524,
+      "learning_rate": 1.9732866751929856e-05,
+      "loss": 2.4703,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "grad_norm": 4.190262556624985,
+      "learning_rate": 1.973215043092457e-05,
+      "loss": 2.4534,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.0002,
+      "step": 1056,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.1351,
+      "step": 1056,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.374,
+      "step": 1056,
+      "vm_loss": 0.2017
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.3327,
+      "step": 1056,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.3892,
+      "step": 1056,
+      "vm_loss": 0.209
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.1823,
+      "step": 1056,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.6833,
+      "step": 1056,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.20328705151959958,
+      "lm_loss": 2.233,
+      "step": 1056,
+      "vm_loss": 0.0971
+    },
+    {
+      "epoch": 0.20347955819717498,
+      "grad_norm": 3.0310609474543178,
+      "learning_rate": 1.973143316382975e-05,
+      "loss": 2.4414,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20367206487475034,
+      "grad_norm": 3.6354270125806094,
+      "learning_rate": 1.9730714950715125e-05,
+      "loss": 2.4536,
+      "step": 1058
+    },
+    {
+      "epoch": 0.20386457155232574,
+      "grad_norm": 3.506230343601751,
+      "learning_rate": 1.9729995791650513e-05,
+      "loss": 2.4605,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2040570782299011,
+      "grad_norm": 3.05679376572762,
+      "learning_rate": 1.9729275686705832e-05,
+      "loss": 2.4478,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20424958490747647,
+      "grad_norm": 3.3411958432249764,
+      "learning_rate": 1.9728554635951077e-05,
+      "loss": 2.4445,
+      "step": 1061
+    },
+    {
+      "epoch": 0.20444209158505186,
+      "grad_norm": 3.0757403581456875,
+      "learning_rate": 1.9727832639456345e-05,
+      "loss": 2.4233,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20463459826262723,
+      "grad_norm": 3.3958910262660913,
+      "learning_rate": 1.9727109697291828e-05,
+      "loss": 2.4771,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "grad_norm": 3.5450974176658314,
+      "learning_rate": 1.9726385809527803e-05,
+      "loss": 2.4607,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.4008,
+      "step": 1064,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.5694,
+      "step": 1064,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.1236,
+      "step": 1064,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.3893,
+      "step": 1064,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.3014,
+      "step": 1064,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.2917,
+      "step": 1064,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.3791,
+      "step": 1064,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 0.20482710494020262,
+      "lm_loss": 2.2845,
+      "step": 1064,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.20501961161777799,
+      "grad_norm": 3.673255582470371,
+      "learning_rate": 1.9725660976234637e-05,
+      "loss": 2.4733,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20521211829535338,
+      "grad_norm": 3.5561683208373,
+      "learning_rate": 1.9724935197482798e-05,
+      "loss": 2.4681,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20540462497292875,
+      "grad_norm": 3.9106349950784396,
+      "learning_rate": 1.9724208473342843e-05,
+      "loss": 2.4398,
+      "step": 1067
+    },
+    {
+      "epoch": 0.20559713165050414,
+      "grad_norm": 3.5814670866695906,
+      "learning_rate": 1.972348080388541e-05,
+      "loss": 2.4605,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2057896383280795,
+      "grad_norm": 4.154509313324284,
+      "learning_rate": 1.9722752189181247e-05,
+      "loss": 2.5055,
+      "step": 1069
+    },
+    {
+      "epoch": 0.20598214500565487,
+      "grad_norm": 3.541309031296491,
+      "learning_rate": 1.9722022629301182e-05,
+      "loss": 2.4832,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20617465168323026,
+      "grad_norm": 2.9521296949090616,
+      "learning_rate": 1.972129212431613e-05,
+      "loss": 2.519,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "grad_norm": 3.652116900961156,
+      "learning_rate": 1.9720560674297118e-05,
+      "loss": 2.466,
+      "step": 1072
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.3107,
+      "step": 1072,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.3423,
+      "step": 1072,
+      "vm_loss": 0.1525
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.1638,
+      "step": 1072,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 1.954,
+      "step": 1072,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.1565,
+      "step": 1072,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.2434,
+      "step": 1072,
+      "vm_loss": 0.2372
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.3942,
+      "step": 1072,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 0.20636715836080563,
+      "lm_loss": 2.4193,
+      "step": 1072,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.20655966503838102,
+      "grad_norm": 3.462325461702056,
+      "learning_rate": 1.9719828279315244e-05,
+      "loss": 2.4689,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2067521717159564,
+      "grad_norm": 3.4853023310226603,
+      "learning_rate": 1.9719094939441707e-05,
+      "loss": 2.4959,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20694467839353178,
+      "grad_norm": 3.8084676541661566,
+      "learning_rate": 1.9718360654747803e-05,
+      "loss": 2.4487,
+      "step": 1075
+    },
+    {
+      "epoch": 0.20713718507110715,
+      "grad_norm": 3.307081892423532,
+      "learning_rate": 1.9717625425304903e-05,
+      "loss": 2.4767,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20732969174868254,
+      "grad_norm": 3.156421642407463,
+      "learning_rate": 1.971688925118449e-05,
+      "loss": 2.4697,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2075221984262579,
+      "grad_norm": 3.348708082298858,
+      "learning_rate": 1.971615213245813e-05,
+      "loss": 2.4388,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2077147051038333,
+      "grad_norm": 2.9728931602673563,
+      "learning_rate": 1.971541406919747e-05,
+      "loss": 2.4722,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "grad_norm": 3.5960490334354507,
+      "learning_rate": 1.9714675061474274e-05,
+      "loss": 2.5265,
+      "step": 1080
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.5364,
+      "step": 1080,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.1915,
+      "step": 1080,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.2793,
+      "step": 1080,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.1183,
+      "step": 1080,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.3568,
+      "step": 1080,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.4381,
+      "step": 1080,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.3992,
+      "step": 1080,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.20790721178140867,
+      "lm_loss": 2.3177,
+      "step": 1080,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.20809971845898403,
+      "grad_norm": 3.838037734841467,
+      "learning_rate": 1.9713935109360366e-05,
+      "loss": 2.4503,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20829222513655943,
+      "grad_norm": 3.3836744697301504,
+      "learning_rate": 1.9713194212927694e-05,
+      "loss": 2.5003,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2084847318141348,
+      "grad_norm": 3.268022060949484,
+      "learning_rate": 1.9712452372248277e-05,
+      "loss": 2.4584,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2086772384917102,
+      "grad_norm": 3.0971479536525734,
+      "learning_rate": 1.9711709587394226e-05,
+      "loss": 2.4506,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20886974516928555,
+      "grad_norm": 3.098899409918459,
+      "learning_rate": 1.9710965858437757e-05,
+      "loss": 2.4733,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20906225184686095,
+      "grad_norm": 2.973243507493669,
+      "learning_rate": 1.9710221185451166e-05,
+      "loss": 2.4794,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2092547585244363,
+      "grad_norm": 2.983289905551904,
+      "learning_rate": 1.9709475568506847e-05,
+      "loss": 2.4554,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "grad_norm": 3.5919574904853717,
+      "learning_rate": 1.9708729007677282e-05,
+      "loss": 2.4788,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.3375,
+      "step": 1088,
+      "vm_loss": 0.2891
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.3548,
+      "step": 1088,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.1963,
+      "step": 1088,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.3831,
+      "step": 1088,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.0638,
+      "step": 1088,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.2605,
+      "step": 1088,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.4435,
+      "step": 1088,
+      "vm_loss": 0.2177
+    },
+    {
+      "epoch": 0.2094472652020117,
+      "lm_loss": 2.3499,
+      "step": 1088,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.20963977187958707,
+      "grad_norm": 3.107849031762314,
+      "learning_rate": 1.970798150303505e-05,
+      "loss": 2.4858,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20983227855716244,
+      "grad_norm": 3.9196148085746296,
+      "learning_rate": 1.970723305465281e-05,
+      "loss": 2.4468,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21002478523473783,
+      "grad_norm": 2.832563062531479,
+      "learning_rate": 1.9706483662603322e-05,
+      "loss": 2.4646,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2102172919123132,
+      "grad_norm": 3.8235081517885434,
+      "learning_rate": 1.9705733326959448e-05,
+      "loss": 2.4678,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2104097985898886,
+      "grad_norm": 3.7969224590107458,
+      "learning_rate": 1.9704982047794118e-05,
+      "loss": 2.4776,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21060230526746396,
+      "grad_norm": 3.6819414000725974,
+      "learning_rate": 1.970422982518037e-05,
+      "loss": 2.4803,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21079481194503935,
+      "grad_norm": 3.6814781999598116,
+      "learning_rate": 1.9703476659191326e-05,
+      "loss": 2.4354,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "grad_norm": 3.3036012385776443,
+      "learning_rate": 1.9702722549900214e-05,
+      "loss": 2.4561,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.2784,
+      "step": 1096,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.6026,
+      "step": 1096,
+      "vm_loss": 0.1025
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.4877,
+      "step": 1096,
+      "vm_loss": 0.2261
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.5044,
+      "step": 1096,
+      "vm_loss": 0.2337
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.3991,
+      "step": 1096,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.0255,
+      "step": 1096,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 1.956,
+      "step": 1096,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.21098731862261472,
+      "lm_loss": 2.2897,
+      "step": 1096,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 0.2111798253001901,
+      "grad_norm": 3.5654036696633677,
+      "learning_rate": 1.9701967497380334e-05,
+      "loss": 2.468,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21137233197776548,
+      "grad_norm": 3.22316907138247,
+      "learning_rate": 1.9701211501705083e-05,
+      "loss": 2.4887,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21156483865534087,
+      "grad_norm": 3.043080031212153,
+      "learning_rate": 1.9700454562947966e-05,
+      "loss": 2.4256,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21175734533291624,
+      "grad_norm": 3.6046942557381687,
+      "learning_rate": 1.9699696681182557e-05,
+      "loss": 2.467,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2119498520104916,
+      "grad_norm": 3.623433266963806,
+      "learning_rate": 1.9698937856482538e-05,
+      "loss": 2.4551,
+      "step": 1101
+    },
+    {
+      "epoch": 0.212142358688067,
+      "grad_norm": 3.690269380113225,
+      "learning_rate": 1.969817808892167e-05,
+      "loss": 2.4576,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21233486536564236,
+      "grad_norm": 3.5394811084859117,
+      "learning_rate": 1.969741737857382e-05,
+      "loss": 2.4842,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "grad_norm": 3.1776357190012714,
+      "learning_rate": 1.9696655725512933e-05,
+      "loss": 2.4614,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.3948,
+      "step": 1104,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.4987,
+      "step": 1104,
+      "vm_loss": 0.1417
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.3013,
+      "step": 1104,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.173,
+      "step": 1104,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 1.949,
+      "step": 1104,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.0258,
+      "step": 1104,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.3912,
+      "step": 1104,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.21252737204321775,
+      "lm_loss": 2.2166,
+      "step": 1104,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.21271987872079312,
+      "grad_norm": 3.657961006377713,
+      "learning_rate": 1.9695893129813054e-05,
+      "loss": 2.4767,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21291238539836851,
+      "grad_norm": 3.4801132398158554,
+      "learning_rate": 1.9695129591548315e-05,
+      "loss": 2.5115,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21310489207594388,
+      "grad_norm": 4.708400754862967,
+      "learning_rate": 1.9694365110792943e-05,
+      "loss": 2.478,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21329739875351927,
+      "grad_norm": 3.7583293280439047,
+      "learning_rate": 1.9693599687621253e-05,
+      "loss": 2.4364,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21348990543109464,
+      "grad_norm": 4.242514516222299,
+      "learning_rate": 1.969283332210766e-05,
+      "loss": 2.4977,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21368241210867,
+      "grad_norm": 3.4472666671048606,
+      "learning_rate": 1.9692066014326655e-05,
+      "loss": 2.5105,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2138749187862454,
+      "grad_norm": 3.8129388078345627,
+      "learning_rate": 1.969129776435284e-05,
+      "loss": 2.414,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "grad_norm": 3.9965781838270242,
+      "learning_rate": 1.9690528572260894e-05,
+      "loss": 2.4672,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.1861,
+      "step": 1112,
+      "vm_loss": 0.1096
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.3975,
+      "step": 1112,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.4695,
+      "step": 1112,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.341,
+      "step": 1112,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.2028,
+      "step": 1112,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.4051,
+      "step": 1112,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.3279,
+      "step": 1112,
+      "vm_loss": 0.2397
+    },
+    {
+      "epoch": 0.21406742546382077,
+      "lm_loss": 2.2707,
+      "step": 1112,
+      "vm_loss": 0.2604
+    },
+    {
+      "epoch": 0.21425993214139616,
+      "grad_norm": 3.425090927850954,
+      "learning_rate": 1.968975843812559e-05,
+      "loss": 2.4717,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21445243881897152,
+      "grad_norm": 3.961245949081223,
+      "learning_rate": 1.9688987362021802e-05,
+      "loss": 2.4859,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21464494549654692,
+      "grad_norm": 3.3767167640424223,
+      "learning_rate": 1.968821534402448e-05,
+      "loss": 2.4119,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21483745217412228,
+      "grad_norm": 3.884816746855957,
+      "learning_rate": 1.9687442384208682e-05,
+      "loss": 2.4755,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21502995885169768,
+      "grad_norm": 3.545492428730258,
+      "learning_rate": 1.9686668482649543e-05,
+      "loss": 2.5134,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21522246552927304,
+      "grad_norm": 3.6174149838197707,
+      "learning_rate": 1.96858936394223e-05,
+      "loss": 2.5237,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21541497220684844,
+      "grad_norm": 3.446055327133417,
+      "learning_rate": 1.9685117854602276e-05,
+      "loss": 2.4808,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "grad_norm": 2.9068861070416987,
+      "learning_rate": 1.9684341128264884e-05,
+      "loss": 2.4478,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.5983,
+      "step": 1120,
+      "vm_loss": 0.2049
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.331,
+      "step": 1120,
+      "vm_loss": 0.2151
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.4922,
+      "step": 1120,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.3474,
+      "step": 1120,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.2739,
+      "step": 1120,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.2397,
+      "step": 1120,
+      "vm_loss": 0.2368
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.4236,
+      "step": 1120,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 0.2156074788844238,
+      "lm_loss": 2.1614,
+      "step": 1120,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.21579998556199917,
+      "grad_norm": 3.288904935672784,
+      "learning_rate": 1.968356346048564e-05,
+      "loss": 2.452,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21599249223957456,
+      "grad_norm": 3.310553622738488,
+      "learning_rate": 1.9682784851340137e-05,
+      "loss": 2.4679,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21618499891714993,
+      "grad_norm": 3.584333329972139,
+      "learning_rate": 1.968200530090407e-05,
+      "loss": 2.4584,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21637750559472532,
+      "grad_norm": 3.4314883097095725,
+      "learning_rate": 1.9681224809253217e-05,
+      "loss": 2.4644,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2165700122723007,
+      "grad_norm": 3.5942601337019435,
+      "learning_rate": 1.968044337646345e-05,
+      "loss": 2.469,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21676251894987608,
+      "grad_norm": 4.162594894638387,
+      "learning_rate": 1.9679661002610743e-05,
+      "loss": 2.4726,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21695502562745145,
+      "grad_norm": 3.0808734276942604,
+      "learning_rate": 1.967887768777115e-05,
+      "loss": 2.4474,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "grad_norm": 3.837693798700385,
+      "learning_rate": 1.967809343202081e-05,
+      "loss": 2.4836,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.2449,
+      "step": 1128,
+      "vm_loss": 0.2727
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.244,
+      "step": 1128,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.3455,
+      "step": 1128,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.2377,
+      "step": 1128,
+      "vm_loss": 0.2255
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.2009,
+      "step": 1128,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.3181,
+      "step": 1128,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.1726,
+      "step": 1128,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 0.21714753230502684,
+      "lm_loss": 2.1383,
+      "step": 1128,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.2173400389826022,
+      "grad_norm": 3.24752334731843,
+      "learning_rate": 1.967730823543597e-05,
+      "loss": 2.4561,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2175325456601776,
+      "grad_norm": 3.3269744160512085,
+      "learning_rate": 1.9676522098092965e-05,
+      "loss": 2.4966,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21772505233775297,
+      "grad_norm": 2.9074985667778734,
+      "learning_rate": 1.967573502006821e-05,
+      "loss": 2.4921,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21791755901532833,
+      "grad_norm": 3.5730107471420745,
+      "learning_rate": 1.967494700143822e-05,
+      "loss": 2.5049,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21811006569290373,
+      "grad_norm": 3.313305700715054,
+      "learning_rate": 1.9674158042279606e-05,
+      "loss": 2.4919,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2183025723704791,
+      "grad_norm": 3.4326954502787537,
+      "learning_rate": 1.9673368142669066e-05,
+      "loss": 2.4775,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21849507904805449,
+      "grad_norm": 3.3122493441262395,
+      "learning_rate": 1.9672577302683374e-05,
+      "loss": 2.4351,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "grad_norm": 3.543085246451503,
+      "learning_rate": 1.9671785522399428e-05,
+      "loss": 2.4401,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.3683,
+      "step": 1136,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.4835,
+      "step": 1136,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.241,
+      "step": 1136,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.4631,
+      "step": 1136,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.1926,
+      "step": 1136,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.0073,
+      "step": 1136,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.5343,
+      "step": 1136,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 0.21868758572562985,
+      "lm_loss": 2.3711,
+      "step": 1136,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 0.21888009240320525,
+      "grad_norm": 3.3855325951831343,
+      "learning_rate": 1.9670992801894185e-05,
+      "loss": 2.4079,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2190725990807806,
+      "grad_norm": 3.3264073223945982,
+      "learning_rate": 1.967019914124472e-05,
+      "loss": 2.4761,
+      "step": 1138
+    },
+    {
+      "epoch": 0.219265105758356,
+      "grad_norm": 3.0395092523524445,
+      "learning_rate": 1.9669404540528176e-05,
+      "loss": 2.476,
+      "step": 1139
+    },
+    {
+      "epoch": 0.21945761243593137,
+      "grad_norm": 3.460025319791979,
+      "learning_rate": 1.9668608999821806e-05,
+      "loss": 2.4243,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21965011911350674,
+      "grad_norm": 3.4865871033475417,
+      "learning_rate": 1.966781251920294e-05,
+      "loss": 2.468,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21984262579108213,
+      "grad_norm": 3.660505950427934,
+      "learning_rate": 1.9667015098749013e-05,
+      "loss": 2.4833,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2200351324686575,
+      "grad_norm": 2.7877308478794673,
+      "learning_rate": 1.966621673853754e-05,
+      "loss": 2.4424,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "grad_norm": 3.4580754913773006,
+      "learning_rate": 1.9665417438646132e-05,
+      "loss": 2.4281,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.3236,
+      "step": 1144,
+      "vm_loss": 0.2032
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.3738,
+      "step": 1144,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.4211,
+      "step": 1144,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.2454,
+      "step": 1144,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.3133,
+      "step": 1144,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.2163,
+      "step": 1144,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.3465,
+      "step": 1144,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.2202276391462329,
+      "lm_loss": 2.1948,
+      "step": 1144,
+      "vm_loss": 0.2123
+    },
+    {
+      "epoch": 0.22042014582380826,
+      "grad_norm": 3.9941520772190584,
+      "learning_rate": 1.966461719915249e-05,
+      "loss": 2.4858,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22061265250138365,
+      "grad_norm": 2.673214965013801,
+      "learning_rate": 1.966381602013441e-05,
+      "loss": 2.4529,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22080515917895902,
+      "grad_norm": 3.525819465385896,
+      "learning_rate": 1.966301390166978e-05,
+      "loss": 2.4796,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2209976658565344,
+      "grad_norm": 4.141749473586178,
+      "learning_rate": 1.9662210843836574e-05,
+      "loss": 2.5134,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22119017253410977,
+      "grad_norm": 3.5395805230836235,
+      "learning_rate": 1.9661406846712854e-05,
+      "loss": 2.439,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22138267921168517,
+      "grad_norm": 3.4650741859323415,
+      "learning_rate": 1.9660601910376784e-05,
+      "loss": 2.4731,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22157518588926053,
+      "grad_norm": 3.663834711108934,
+      "learning_rate": 1.9659796034906614e-05,
+      "loss": 2.4649,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "grad_norm": 4.219484135337479,
+      "learning_rate": 1.9658989220380686e-05,
+      "loss": 2.4461,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.5942,
+      "step": 1152,
+      "vm_loss": 0.2674
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.3894,
+      "step": 1152,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.4629,
+      "step": 1152,
+      "vm_loss": 0.2307
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.3915,
+      "step": 1152,
+      "vm_loss": 0.1001
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.0616,
+      "step": 1152,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.5497,
+      "step": 1152,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.2942,
+      "step": 1152,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 0.2217676925668359,
+      "lm_loss": 2.2603,
+      "step": 1152,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 0.2219601992444113,
+      "grad_norm": 3.439145279154436,
+      "learning_rate": 1.9658181466877427e-05,
+      "loss": 2.491,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22215270592198666,
+      "grad_norm": 3.2252892303841874,
+      "learning_rate": 1.965737277447537e-05,
+      "loss": 2.4905,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22234521259956205,
+      "grad_norm": 3.554339049029039,
+      "learning_rate": 1.9656563143253122e-05,
+      "loss": 2.4758,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22253771927713742,
+      "grad_norm": 3.7321968669043954,
+      "learning_rate": 1.9655752573289396e-05,
+      "loss": 2.4721,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2227302259547128,
+      "grad_norm": 3.306356652848731,
+      "learning_rate": 1.9654941064662986e-05,
+      "loss": 2.5052,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22292273263228818,
+      "grad_norm": 4.375134518763874,
+      "learning_rate": 1.965412861745278e-05,
+      "loss": 2.4728,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22311523930986357,
+      "grad_norm": 3.961325960685138,
+      "learning_rate": 1.965331523173776e-05,
+      "loss": 2.451,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "grad_norm": 3.467604030714786,
+      "learning_rate": 1.9652500907596998e-05,
+      "loss": 2.4754,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.3335,
+      "step": 1160,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.1967,
+      "step": 1160,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.2758,
+      "step": 1160,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.266,
+      "step": 1160,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.1162,
+      "step": 1160,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.3065,
+      "step": 1160,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.0585,
+      "step": 1160,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 0.22330774598743894,
+      "lm_loss": 2.433,
+      "step": 1160,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.2235002526650143,
+      "grad_norm": 4.073247633569005,
+      "learning_rate": 1.9651685645109657e-05,
+      "loss": 2.4406,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2236927593425897,
+      "grad_norm": 3.813135705064817,
+      "learning_rate": 1.965086944435499e-05,
+      "loss": 2.4289,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22388526602016506,
+      "grad_norm": 3.518834975932944,
+      "learning_rate": 1.965005230541234e-05,
+      "loss": 2.4556,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22407777269774046,
+      "grad_norm": 3.5368983980092423,
+      "learning_rate": 1.9649234228361148e-05,
+      "loss": 2.4827,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22427027937531582,
+      "grad_norm": 4.64997992121861,
+      "learning_rate": 1.964841521328094e-05,
+      "loss": 2.4536,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22446278605289122,
+      "grad_norm": 3.879797133723813,
+      "learning_rate": 1.964759526025133e-05,
+      "loss": 2.41,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22465529273046658,
+      "grad_norm": 3.1871897758966106,
+      "learning_rate": 1.964677436935204e-05,
+      "loss": 2.4376,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "grad_norm": 4.147788638776459,
+      "learning_rate": 1.9645952540662855e-05,
+      "loss": 2.4666,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 1.9165,
+      "step": 1168,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 1.8994,
+      "step": 1168,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.3829,
+      "step": 1168,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.2215,
+      "step": 1168,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.1421,
+      "step": 1168,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.0762,
+      "step": 1168,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.4662,
+      "step": 1168,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 0.22484779940804198,
+      "lm_loss": 2.4899,
+      "step": 1168,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 0.22504030608561734,
+      "grad_norm": 4.075723294234439,
+      "learning_rate": 1.9645129774263678e-05,
+      "loss": 2.4739,
+      "step": 1169
+    },
+    {
+      "epoch": 0.22523281276319274,
+      "grad_norm": 3.6593286814891504,
+      "learning_rate": 1.9644306070234492e-05,
+      "loss": 2.4516,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2254253194407681,
+      "grad_norm": 3.3902425689510305,
+      "learning_rate": 1.9643481428655366e-05,
+      "loss": 2.4688,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22561782611834347,
+      "grad_norm": 3.7166742638096237,
+      "learning_rate": 1.9642655849606468e-05,
+      "loss": 2.5134,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22581033279591886,
+      "grad_norm": 3.4153498463214,
+      "learning_rate": 1.9641829333168058e-05,
+      "loss": 2.4675,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22600283947349423,
+      "grad_norm": 3.7397391809765352,
+      "learning_rate": 1.9641001879420482e-05,
+      "loss": 2.4392,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22619534615106962,
+      "grad_norm": 3.911807991772737,
+      "learning_rate": 1.964017348844418e-05,
+      "loss": 2.5032,
+      "step": 1175
+    },
+    {
+      "epoch": 0.226387852828645,
+      "grad_norm": 3.2511768379737562,
+      "learning_rate": 1.9639344160319678e-05,
+      "loss": 2.4762,
+      "step": 1176
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.1043,
+      "step": 1176,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.4644,
+      "step": 1176,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.5693,
+      "step": 1176,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 1.8619,
+      "step": 1176,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.4452,
+      "step": 1176,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.5079,
+      "step": 1176,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.4046,
+      "step": 1176,
+      "vm_loss": 0.2209
+    },
+    {
+      "epoch": 0.226387852828645,
+      "lm_loss": 2.2349,
+      "step": 1176,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.22658035950622038,
+      "grad_norm": 3.178634843921187,
+      "learning_rate": 1.9638513895127604e-05,
+      "loss": 2.4444,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22677286618379575,
+      "grad_norm": 3.9327577031615304,
+      "learning_rate": 1.9637682692948666e-05,
+      "loss": 2.4303,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22696537286137114,
+      "grad_norm": 3.47990577513902,
+      "learning_rate": 1.963685055386367e-05,
+      "loss": 2.4917,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2271578795389465,
+      "grad_norm": 3.943248486056169,
+      "learning_rate": 1.9636017477953502e-05,
+      "loss": 2.4956,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2273503862165219,
+      "grad_norm": 3.290729248661274,
+      "learning_rate": 1.9635183465299157e-05,
+      "loss": 2.4889,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22754289289409727,
+      "grad_norm": 4.1554701885405025,
+      "learning_rate": 1.963434851598171e-05,
+      "loss": 2.473,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22773539957167263,
+      "grad_norm": 3.616210835013554,
+      "learning_rate": 1.963351263008233e-05,
+      "loss": 2.4482,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "grad_norm": 3.4742629093571877,
+      "learning_rate": 1.963267580768227e-05,
+      "loss": 2.5033,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.3939,
+      "step": 1184,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.3575,
+      "step": 1184,
+      "vm_loss": 0.2145
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.1724,
+      "step": 1184,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.2225,
+      "step": 1184,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.3885,
+      "step": 1184,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.2578,
+      "step": 1184,
+      "vm_loss": 0.2011
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.4048,
+      "step": 1184,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.22792790624924802,
+      "lm_loss": 2.3607,
+      "step": 1184,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 0.2281204129268234,
+      "grad_norm": 3.7167764647989325,
+      "learning_rate": 1.963183804886289e-05,
+      "loss": 2.4591,
+      "step": 1185
+    },
+    {
+      "epoch": 0.22831291960439878,
+      "grad_norm": 4.118404926502813,
+      "learning_rate": 1.9630999353705617e-05,
+      "loss": 2.5048,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22850542628197415,
+      "grad_norm": 3.267249100361888,
+      "learning_rate": 1.9630159722291996e-05,
+      "loss": 2.4638,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22869793295954954,
+      "grad_norm": 3.3634596526994938,
+      "learning_rate": 1.962931915470364e-05,
+      "loss": 2.4699,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2288904396371249,
+      "grad_norm": 4.438569797110638,
+      "learning_rate": 1.962847765102227e-05,
+      "loss": 2.508,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2290829463147003,
+      "grad_norm": 3.0322702860915447,
+      "learning_rate": 1.9627635211329686e-05,
+      "loss": 2.4761,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22927545299227567,
+      "grad_norm": 3.823828437286329,
+      "learning_rate": 1.9626791835707787e-05,
+      "loss": 2.4543,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "grad_norm": 3.9025126813697057,
+      "learning_rate": 1.9625947524238564e-05,
+      "loss": 2.4187,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.2605,
+      "step": 1192,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.0292,
+      "step": 1192,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.465,
+      "step": 1192,
+      "vm_loss": 0.2515
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.3289,
+      "step": 1192,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.2172,
+      "step": 1192,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.4247,
+      "step": 1192,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.1336,
+      "step": 1192,
+      "vm_loss": 0.1359
+    },
+    {
+      "epoch": 0.22946795966985103,
+      "lm_loss": 2.3907,
+      "step": 1192,
+      "vm_loss": 0.2451
+    },
+    {
+      "epoch": 0.22966046634742643,
+      "grad_norm": 3.2003435388077692,
+      "learning_rate": 1.9625102277004085e-05,
+      "loss": 2.4428,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2298529730250018,
+      "grad_norm": 4.138706738787077,
+      "learning_rate": 1.9624256094086523e-05,
+      "loss": 2.4618,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2300454797025772,
+      "grad_norm": 4.437828298246324,
+      "learning_rate": 1.9623408975568138e-05,
+      "loss": 2.4434,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23023798638015255,
+      "grad_norm": 3.468856888317049,
+      "learning_rate": 1.9622560921531285e-05,
+      "loss": 2.4399,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23043049305772795,
+      "grad_norm": 3.6938786693141674,
+      "learning_rate": 1.9621711932058395e-05,
+      "loss": 2.4629,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2306229997353033,
+      "grad_norm": 3.5439846936354193,
+      "learning_rate": 1.9620862007232015e-05,
+      "loss": 2.4251,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2308155064128787,
+      "grad_norm": 3.080250781285466,
+      "learning_rate": 1.9620011147134758e-05,
+      "loss": 2.4156,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "grad_norm": 4.249715445311841,
+      "learning_rate": 1.961915935184934e-05,
+      "loss": 2.5074,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.428,
+      "step": 1200,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.2714,
+      "step": 1200,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.3676,
+      "step": 1200,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.2548,
+      "step": 1200,
+      "vm_loss": 0.2473
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.6291,
+      "step": 1200,
+      "vm_loss": 0.2803
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 1.759,
+      "step": 1200,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.2732,
+      "step": 1200,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 0.23100801309045407,
+      "lm_loss": 2.1425,
+      "step": 1200,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.23120051976802947,
+      "grad_norm": 3.0513229314448282,
+      "learning_rate": 1.961830662145857e-05,
+      "loss": 2.4736,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23139302644560483,
+      "grad_norm": 3.081842396441924,
+      "learning_rate": 1.9617452956045344e-05,
+      "loss": 2.4947,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2315855331231802,
+      "grad_norm": 3.7129843148889603,
+      "learning_rate": 1.9616598355692644e-05,
+      "loss": 2.4964,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2317780398007556,
+      "grad_norm": 4.0149708384232,
+      "learning_rate": 1.9615742820483554e-05,
+      "loss": 2.4421,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23197054647833096,
+      "grad_norm": 3.3334216835742896,
+      "learning_rate": 1.961488635050124e-05,
+      "loss": 2.3777,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23216305315590635,
+      "grad_norm": 3.7976150665394166,
+      "learning_rate": 1.961402894582896e-05,
+      "loss": 2.4144,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23235555983348172,
+      "grad_norm": 4.484708449826713,
+      "learning_rate": 1.961317060655007e-05,
+      "loss": 2.4671,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "grad_norm": 3.305011549556479,
+      "learning_rate": 1.9612311332748008e-05,
+      "loss": 2.46,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.1799,
+      "step": 1208,
+      "vm_loss": 0.2228
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.0738,
+      "step": 1208,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.3666,
+      "step": 1208,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.4713,
+      "step": 1208,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.387,
+      "step": 1208,
+      "vm_loss": 0.2495
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.447,
+      "step": 1208,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 1.5992,
+      "step": 1208,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 0.2325480665110571,
+      "lm_loss": 2.3106,
+      "step": 1208,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 0.23274057318863248,
+      "grad_norm": 3.7303729944956676,
+      "learning_rate": 1.9611451124506308e-05,
+      "loss": 2.4799,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23293307986620787,
+      "grad_norm": 3.961349763793609,
+      "learning_rate": 1.9610589981908593e-05,
+      "loss": 2.474,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23312558654378324,
+      "grad_norm": 3.607499882312154,
+      "learning_rate": 1.9609727905038573e-05,
+      "loss": 2.5005,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2333180932213586,
+      "grad_norm": 4.2850567074387484,
+      "learning_rate": 1.9608864893980056e-05,
+      "loss": 2.5114,
+      "step": 1212
+    },
+    {
+      "epoch": 0.233510599898934,
+      "grad_norm": 3.6643182984245413,
+      "learning_rate": 1.9608000948816942e-05,
+      "loss": 2.5217,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23370310657650936,
+      "grad_norm": 3.1842880662237607,
+      "learning_rate": 1.9607136069633212e-05,
+      "loss": 2.4342,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23389561325408476,
+      "grad_norm": 3.4572435575787273,
+      "learning_rate": 1.9606270256512947e-05,
+      "loss": 2.5053,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "grad_norm": 3.8211503219535814,
+      "learning_rate": 1.9605403509540308e-05,
+      "loss": 2.4401,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 1.9969,
+      "step": 1216,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.138,
+      "step": 1216,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.1311,
+      "step": 1216,
+      "vm_loss": 0.2112
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.2109,
+      "step": 1216,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.2883,
+      "step": 1216,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.4944,
+      "step": 1216,
+      "vm_loss": 0.1258
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 2.2771,
+      "step": 1216,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 0.23408811993166012,
+      "lm_loss": 1.9629,
+      "step": 1216,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 0.23428062660923551,
+      "grad_norm": 2.918449772579497,
+      "learning_rate": 1.960453582879956e-05,
+      "loss": 2.4251,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23447313328681088,
+      "grad_norm": 3.5844067289270836,
+      "learning_rate": 1.9603667214375053e-05,
+      "loss": 2.4489,
+      "step": 1218
+    },
+    {
+      "epoch": 0.23466563996438627,
+      "grad_norm": 3.5047086080765086,
+      "learning_rate": 1.960279766635123e-05,
+      "loss": 2.4884,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23485814664196164,
+      "grad_norm": 3.333499859532291,
+      "learning_rate": 1.960192718481261e-05,
+      "loss": 2.4472,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23505065331953703,
+      "grad_norm": 3.149896738681449,
+      "learning_rate": 1.9601055769843832e-05,
+      "loss": 2.4542,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2352431599971124,
+      "grad_norm": 3.453267975528268,
+      "learning_rate": 1.960018342152959e-05,
+      "loss": 2.4386,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23543566667468777,
+      "grad_norm": 2.994502295368568,
+      "learning_rate": 1.9599310139954706e-05,
+      "loss": 2.4791,
+      "step": 1223
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "grad_norm": 3.272488889736858,
+      "learning_rate": 1.959843592520406e-05,
+      "loss": 2.4356,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.0075,
+      "step": 1224,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.4588,
+      "step": 1224,
+      "vm_loss": 0.2488
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.4585,
+      "step": 1224,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.3119,
+      "step": 1224,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.4838,
+      "step": 1224,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.3504,
+      "step": 1224,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 1.8162,
+      "step": 1224,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.23562817335226316,
+      "lm_loss": 2.2534,
+      "step": 1224,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.23582068002983853,
+      "grad_norm": 2.9811635532020957,
+      "learning_rate": 1.9597560777362645e-05,
+      "loss": 2.5131,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23601318670741392,
+      "grad_norm": 3.276735449151304,
+      "learning_rate": 1.9596684696515533e-05,
+      "loss": 2.4349,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23620569338498928,
+      "grad_norm": 3.202355412954144,
+      "learning_rate": 1.959580768274789e-05,
+      "loss": 2.4488,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23639820006256468,
+      "grad_norm": 2.9707272558132978,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 2.4522,
+      "step": 1228
+    },
+    {
+      "epoch": 0.23659070674014004,
+      "grad_norm": 3.180961673277115,
+      "learning_rate": 1.9594050856792135e-05,
+      "loss": 2.4415,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23678321341771544,
+      "grad_norm": 3.6708411494666424,
+      "learning_rate": 1.9593171044774807e-05,
+      "loss": 2.4995,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2369757200952908,
+      "grad_norm": 3.1780451718881895,
+      "learning_rate": 1.959229030017852e-05,
+      "loss": 2.4377,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "grad_norm": 3.954831781737448,
+      "learning_rate": 1.95914086230889e-05,
+      "loss": 2.4723,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.4703,
+      "step": 1232,
+      "vm_loss": 0.2084
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 1.9716,
+      "step": 1232,
+      "vm_loss": 0.2874
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.3523,
+      "step": 1232,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.1043,
+      "step": 1232,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.1198,
+      "step": 1232,
+      "vm_loss": 0.2386
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 1.6781,
+      "step": 1232,
+      "vm_loss": 0.114
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.5147,
+      "step": 1232,
+      "vm_loss": 0.2229
+    },
+    {
+      "epoch": 0.2371682267728662,
+      "lm_loss": 2.3105,
+      "step": 1232,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.23736073345044156,
+      "grad_norm": 3.4541065634328603,
+      "learning_rate": 1.9590526013591646e-05,
+      "loss": 2.4773,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23755324012801693,
+      "grad_norm": 3.5129467842005,
+      "learning_rate": 1.9589642471772565e-05,
+      "loss": 2.4527,
+      "step": 1234
+    },
+    {
+      "epoch": 0.23774574680559232,
+      "grad_norm": 3.8413181697399685,
+      "learning_rate": 1.958875799771755e-05,
+      "loss": 2.4038,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2379382534831677,
+      "grad_norm": 3.6433119169877033,
+      "learning_rate": 1.9587872591512583e-05,
+      "loss": 2.4911,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23813076016074308,
+      "grad_norm": 3.312951522437006,
+      "learning_rate": 1.9586986253243734e-05,
+      "loss": 2.4925,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23832326683831845,
+      "grad_norm": 3.2298244398229254,
+      "learning_rate": 1.9586098982997168e-05,
+      "loss": 2.5106,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23851577351589384,
+      "grad_norm": 3.33885121791172,
+      "learning_rate": 1.958521078085914e-05,
+      "loss": 2.4396,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "grad_norm": 3.7790297546295704,
+      "learning_rate": 1.9584321646915988e-05,
+      "loss": 2.4562,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.6071,
+      "step": 1240,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 1.871,
+      "step": 1240,
+      "vm_loss": 0.2463
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.339,
+      "step": 1240,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.3169,
+      "step": 1240,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.5389,
+      "step": 1240,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.4738,
+      "step": 1240,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.4138,
+      "step": 1240,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.2387082801934692,
+      "lm_loss": 2.4076,
+      "step": 1240,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.2389007868710446,
+      "grad_norm": 4.21761014010306,
+      "learning_rate": 1.958343158125416e-05,
+      "loss": 2.4804,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23909329354861997,
+      "grad_norm": 3.253780108906917,
+      "learning_rate": 1.958254058396017e-05,
+      "loss": 2.4486,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23928580022619533,
+      "grad_norm": 3.6916334941405746,
+      "learning_rate": 1.958164865512064e-05,
+      "loss": 2.4559,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23947830690377073,
+      "grad_norm": 2.9460102699291153,
+      "learning_rate": 1.9580755794822278e-05,
+      "loss": 2.3986,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2396708135813461,
+      "grad_norm": 3.5712419605370385,
+      "learning_rate": 1.9579862003151875e-05,
+      "loss": 2.458,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2398633202589215,
+      "grad_norm": 3.85224886013358,
+      "learning_rate": 1.9578967280196327e-05,
+      "loss": 2.4255,
+      "step": 1246
+    },
+    {
+      "epoch": 0.24005582693649685,
+      "grad_norm": 3.28352008255624,
+      "learning_rate": 1.9578071626042602e-05,
+      "loss": 2.4651,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "grad_norm": 3.9129803607234668,
+      "learning_rate": 1.957717504077778e-05,
+      "loss": 2.4547,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.5251,
+      "step": 1248,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 1.8931,
+      "step": 1248,
+      "vm_loss": 0.2147
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.4392,
+      "step": 1248,
+      "vm_loss": 0.2412
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.4346,
+      "step": 1248,
+      "vm_loss": 0.1223
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.3947,
+      "step": 1248,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.3603,
+      "step": 1248,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.3085,
+      "step": 1248,
+      "vm_loss": 0.2455
+    },
+    {
+      "epoch": 0.24024833361407225,
+      "lm_loss": 2.4174,
+      "step": 1248,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.2404408402916476,
+      "grad_norm": 3.9670120959047233,
+      "learning_rate": 1.957627752448902e-05,
+      "loss": 2.4589,
+      "step": 1249
+    },
+    {
+      "epoch": 0.240633346969223,
+      "grad_norm": 3.583767071366644,
+      "learning_rate": 1.9575379077263562e-05,
+      "loss": 2.461,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24082585364679837,
+      "grad_norm": 3.966554281561264,
+      "learning_rate": 1.9574479699188757e-05,
+      "loss": 2.4281,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24101836032437376,
+      "grad_norm": 3.551034488151081,
+      "learning_rate": 1.9573579390352028e-05,
+      "loss": 2.5,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24121086700194913,
+      "grad_norm": 3.6095440710159004,
+      "learning_rate": 1.9572678150840904e-05,
+      "loss": 2.4647,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2414033736795245,
+      "grad_norm": 3.372422147812847,
+      "learning_rate": 1.957177598074299e-05,
+      "loss": 2.4433,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2415958803570999,
+      "grad_norm": 3.2500028917779464,
+      "learning_rate": 1.9570872880145995e-05,
+      "loss": 2.4731,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "grad_norm": 3.0608720031020376,
+      "learning_rate": 1.9569968849137705e-05,
+      "loss": 2.4711,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.2472,
+      "step": 1256,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.3767,
+      "step": 1256,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.3176,
+      "step": 1256,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.2331,
+      "step": 1256,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.1911,
+      "step": 1256,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.3035,
+      "step": 1256,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.2489,
+      "step": 1256,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.24178838703467526,
+      "lm_loss": 2.2317,
+      "step": 1256,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.24198089371225065,
+      "grad_norm": 3.7562562147713137,
+      "learning_rate": 1.9569063887806014e-05,
+      "loss": 2.4475,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24217340038982602,
+      "grad_norm": 3.69405227034211,
+      "learning_rate": 1.9568157996238884e-05,
+      "loss": 2.4218,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2423659070674014,
+      "grad_norm": 3.4793440501009294,
+      "learning_rate": 1.9567251174524384e-05,
+      "loss": 2.4519,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24255841374497678,
+      "grad_norm": 3.210379678646978,
+      "learning_rate": 1.956634342275067e-05,
+      "loss": 2.4733,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24275092042255217,
+      "grad_norm": 3.47679159835551,
+      "learning_rate": 1.9565434741005988e-05,
+      "loss": 2.4401,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24294342710012753,
+      "grad_norm": 3.968353171325261,
+      "learning_rate": 1.956452512937867e-05,
+      "loss": 2.419,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2431359337777029,
+      "grad_norm": 3.2161095814685465,
+      "learning_rate": 1.956361458795714e-05,
+      "loss": 2.4344,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "grad_norm": 3.4924038956391836,
+      "learning_rate": 1.9562703116829923e-05,
+      "loss": 2.4614,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.1035,
+      "step": 1264,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.4302,
+      "step": 1264,
+      "vm_loss": 0.2315
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.25,
+      "step": 1264,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.2083,
+      "step": 1264,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.1024,
+      "step": 1264,
+      "vm_loss": 0.2034
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.1332,
+      "step": 1264,
+      "vm_loss": 0.21
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.3157,
+      "step": 1264,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 0.2433284404552783,
+      "lm_loss": 2.4478,
+      "step": 1264,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.24352094713285366,
+      "grad_norm": 3.533861837805012,
+      "learning_rate": 1.9561790716085617e-05,
+      "loss": 2.4888,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24371345381042905,
+      "grad_norm": 3.2918500400473456,
+      "learning_rate": 1.9560877385812924e-05,
+      "loss": 2.4721,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24390596048800442,
+      "grad_norm": 3.447267206266168,
+      "learning_rate": 1.955996312610063e-05,
+      "loss": 2.4205,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2440984671655798,
+      "grad_norm": 3.3159754695482224,
+      "learning_rate": 1.9559047937037613e-05,
+      "loss": 2.4601,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24429097384315518,
+      "grad_norm": 3.477322062869611,
+      "learning_rate": 1.9558131818712838e-05,
+      "loss": 2.4085,
+      "step": 1269
+    },
+    {
+      "epoch": 0.24448348052073057,
+      "grad_norm": 3.216674401765448,
+      "learning_rate": 1.955721477121537e-05,
+      "loss": 2.4391,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24467598719830594,
+      "grad_norm": 3.393032976277194,
+      "learning_rate": 1.9556296794634348e-05,
+      "loss": 2.4743,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "grad_norm": 3.230922177210364,
+      "learning_rate": 1.9555377889059022e-05,
+      "loss": 2.4189,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.0877,
+      "step": 1272,
+      "vm_loss": 0.2495
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.3314,
+      "step": 1272,
+      "vm_loss": 0.3155
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 1.8549,
+      "step": 1272,
+      "vm_loss": 0.2071
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.3719,
+      "step": 1272,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.4522,
+      "step": 1272,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.4764,
+      "step": 1272,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.5035,
+      "step": 1272,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 0.24486849387588133,
+      "lm_loss": 2.133,
+      "step": 1272,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.2450610005534567,
+      "grad_norm": 3.05452318873155,
+      "learning_rate": 1.9554458054578713e-05,
+      "loss": 2.4764,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24525350723103206,
+      "grad_norm": 3.6540772139994715,
+      "learning_rate": 1.9553537291282846e-05,
+      "loss": 2.4683,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24544601390860746,
+      "grad_norm": 3.1383818005000097,
+      "learning_rate": 1.9552615599260927e-05,
+      "loss": 2.4941,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24563852058618282,
+      "grad_norm": 3.7826958435246265,
+      "learning_rate": 1.9551692978602557e-05,
+      "loss": 2.4663,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24583102726375822,
+      "grad_norm": 3.500809634231439,
+      "learning_rate": 1.955076942939743e-05,
+      "loss": 2.4188,
+      "step": 1277
+    },
+    {
+      "epoch": 0.24602353394133358,
+      "grad_norm": 3.139786347407025,
+      "learning_rate": 1.9549844951735328e-05,
+      "loss": 2.5235,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24621604061890898,
+      "grad_norm": 3.715662791454268,
+      "learning_rate": 1.9548919545706113e-05,
+      "loss": 2.4494,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "grad_norm": 3.5537681789566737,
+      "learning_rate": 1.9547993211399753e-05,
+      "loss": 2.5021,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.2823,
+      "step": 1280,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.583,
+      "step": 1280,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.0475,
+      "step": 1280,
+      "vm_loss": 0.2781
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.2312,
+      "step": 1280,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.2127,
+      "step": 1280,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.2479,
+      "step": 1280,
+      "vm_loss": 0.1445
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.3117,
+      "step": 1280,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 0.24640854729648434,
+      "lm_loss": 2.4109,
+      "step": 1280,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.24660105397405974,
+      "grad_norm": 3.3442705721691133,
+      "learning_rate": 1.95470659489063e-05,
+      "loss": 2.468,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2467935606516351,
+      "grad_norm": 3.5013485807753497,
+      "learning_rate": 1.954613775831589e-05,
+      "loss": 2.5184,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2469860673292105,
+      "grad_norm": 3.3422968392427834,
+      "learning_rate": 1.9545208639718766e-05,
+      "loss": 2.4301,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24717857400678586,
+      "grad_norm": 3.2711482106164613,
+      "learning_rate": 1.954427859320524e-05,
+      "loss": 2.5182,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24737108068436123,
+      "grad_norm": 4.07606390410819,
+      "learning_rate": 1.9543347618865726e-05,
+      "loss": 2.427,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24756358736193662,
+      "grad_norm": 2.9925954544734834,
+      "learning_rate": 1.9542415716790727e-05,
+      "loss": 2.4628,
+      "step": 1286
+    },
+    {
+      "epoch": 0.247756094039512,
+      "grad_norm": 3.7457811769328333,
+      "learning_rate": 1.9541482887070843e-05,
+      "loss": 2.4389,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "grad_norm": 3.422721224804485,
+      "learning_rate": 1.9540549129796745e-05,
+      "loss": 2.4672,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.4687,
+      "step": 1288,
+      "vm_loss": 0.2037
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.397,
+      "step": 1288,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 1.786,
+      "step": 1288,
+      "vm_loss": 0.2771
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.4285,
+      "step": 1288,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.4056,
+      "step": 1288,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.3301,
+      "step": 1288,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.3434,
+      "step": 1288,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 0.24794860071708738,
+      "lm_loss": 2.1714,
+      "step": 1288,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 0.24814110739466275,
+      "grad_norm": 2.9637292844826795,
+      "learning_rate": 1.9539614445059215e-05,
+      "loss": 2.5068,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24833361407223814,
+      "grad_norm": 4.205406904343167,
+      "learning_rate": 1.9538678832949113e-05,
+      "loss": 2.4976,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2485261207498135,
+      "grad_norm": 3.7337525949892374,
+      "learning_rate": 1.9537742293557397e-05,
+      "loss": 2.4688,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2487186274273889,
+      "grad_norm": 3.407423884312617,
+      "learning_rate": 1.9536804826975105e-05,
+      "loss": 2.4414,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24891113410496427,
+      "grad_norm": 3.7277354429840126,
+      "learning_rate": 1.953586643329337e-05,
+      "loss": 2.4665,
+      "step": 1293
+    },
+    {
+      "epoch": 0.24910364078253963,
+      "grad_norm": 3.4387483267838443,
+      "learning_rate": 1.9534927112603417e-05,
+      "loss": 2.4555,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24929614746011503,
+      "grad_norm": 4.057550120381568,
+      "learning_rate": 1.9533986864996563e-05,
+      "loss": 2.4908,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "grad_norm": 3.9799671789650453,
+      "learning_rate": 1.9533045690564214e-05,
+      "loss": 2.4363,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.0481,
+      "step": 1296,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.4215,
+      "step": 1296,
+      "vm_loss": 0.2068
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.3141,
+      "step": 1296,
+      "vm_loss": 0.2174
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.2247,
+      "step": 1296,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 1.9348,
+      "step": 1296,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.2618,
+      "step": 1296,
+      "vm_loss": 0.221
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.4166,
+      "step": 1296,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 0.2494886541376904,
+      "lm_loss": 2.2877,
+      "step": 1296,
+      "vm_loss": 0.165
+    },
+    {
+      "epoch": 0.24968116081526578,
+      "grad_norm": 3.553468080904293,
+      "learning_rate": 1.9532103589397858e-05,
+      "loss": 2.438,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24987366749284115,
+      "grad_norm": 3.3366699203215067,
+      "learning_rate": 1.953116056158908e-05,
+      "loss": 2.4579,
+      "step": 1298
+    },
+    {
+      "epoch": 0.25006617417041654,
+      "grad_norm": 3.549534831579184,
+      "learning_rate": 1.9530216607229557e-05,
+      "loss": 2.4496,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2502586808479919,
+      "grad_norm": 2.9416537435647623,
+      "learning_rate": 1.9529271726411055e-05,
+      "loss": 2.486,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2504511875255673,
+      "grad_norm": 3.549767881868953,
+      "learning_rate": 1.9528325919225426e-05,
+      "loss": 2.4643,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2506436942031427,
+      "grad_norm": 3.0193358813288893,
+      "learning_rate": 1.9527379185764613e-05,
+      "loss": 2.405,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25083620088071806,
+      "grad_norm": 3.270055103360828,
+      "learning_rate": 1.9526431526120656e-05,
+      "loss": 2.4652,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "grad_norm": 2.6596369083952967,
+      "learning_rate": 1.952548294038567e-05,
+      "loss": 2.439,
+      "step": 1304
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.3055,
+      "step": 1304,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.2558,
+      "step": 1304,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.2685,
+      "step": 1304,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 1.9447,
+      "step": 1304,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.4669,
+      "step": 1304,
+      "vm_loss": 0.1086
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.3587,
+      "step": 1304,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 1.7869,
+      "step": 1304,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 0.25102870755829343,
+      "lm_loss": 2.4312,
+      "step": 1304,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.2512212142358688,
+      "grad_norm": 3.2679977975074297,
+      "learning_rate": 1.9524533428651878e-05,
+      "loss": 2.4326,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25141372091344416,
+      "grad_norm": 3.1348361909953466,
+      "learning_rate": 1.9523582991011585e-05,
+      "loss": 2.4319,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2516062275910196,
+      "grad_norm": 3.1906128122893254,
+      "learning_rate": 1.952263162755718e-05,
+      "loss": 2.478,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25179873426859495,
+      "grad_norm": 3.512905127957672,
+      "learning_rate": 1.952167933838115e-05,
+      "loss": 2.4409,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2519912409461703,
+      "grad_norm": 2.8724466614974973,
+      "learning_rate": 1.9520726123576075e-05,
+      "loss": 2.4589,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2521837476237457,
+      "grad_norm": 3.077330917970436,
+      "learning_rate": 1.951977198323461e-05,
+      "loss": 2.4494,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2523762543013211,
+      "grad_norm": 3.2137246894170035,
+      "learning_rate": 1.9518816917449517e-05,
+      "loss": 2.4251,
+      "step": 1311
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "grad_norm": 3.4836684105392846,
+      "learning_rate": 1.9517860926313644e-05,
+      "loss": 2.4533,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.1867,
+      "step": 1312,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.4755,
+      "step": 1312,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.4383,
+      "step": 1312,
+      "vm_loss": 0.1179
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 1.8094,
+      "step": 1312,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.0335,
+      "step": 1312,
+      "vm_loss": 0.2433
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.6474,
+      "step": 1312,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.3003,
+      "step": 1312,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 0.25256876097889647,
+      "lm_loss": 2.7781,
+      "step": 1312,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 0.25276126765647183,
+      "grad_norm": 3.340385836767393,
+      "learning_rate": 1.951690400991991e-05,
+      "loss": 2.4892,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2529537743340472,
+      "grad_norm": 3.069045119976783,
+      "learning_rate": 1.9515946168361357e-05,
+      "loss": 2.4618,
+      "step": 1314
+    },
+    {
+      "epoch": 0.25314628101162256,
+      "grad_norm": 3.3100112809818834,
+      "learning_rate": 1.9514987401731088e-05,
+      "loss": 2.4564,
+      "step": 1315
+    },
+    {
+      "epoch": 0.253338787689198,
+      "grad_norm": 2.997859414486016,
+      "learning_rate": 1.9514027710122313e-05,
+      "loss": 2.445,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25353129436677335,
+      "grad_norm": 3.6982507774384494,
+      "learning_rate": 1.9513067093628323e-05,
+      "loss": 2.4981,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2537238010443487,
+      "grad_norm": 2.586676960021598,
+      "learning_rate": 1.9512105552342507e-05,
+      "loss": 2.415,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2539163077219241,
+      "grad_norm": 3.6333136142341704,
+      "learning_rate": 1.9511143086358334e-05,
+      "loss": 2.4571,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "grad_norm": 3.483185217523218,
+      "learning_rate": 1.9510179695769367e-05,
+      "loss": 2.4789,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.492,
+      "step": 1320,
+      "vm_loss": 0.1001
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.3933,
+      "step": 1320,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.6844,
+      "step": 1320,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.2821,
+      "step": 1320,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.3256,
+      "step": 1320,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.3802,
+      "step": 1320,
+      "vm_loss": 0.1134
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.1818,
+      "step": 1320,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 0.2541088143994995,
+      "lm_loss": 2.3754,
+      "step": 1320,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 0.25430132107707487,
+      "grad_norm": 2.826169024345817,
+      "learning_rate": 1.950921538066927e-05,
+      "loss": 2.4473,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25449382775465024,
+      "grad_norm": 3.327700483085308,
+      "learning_rate": 1.9508250141151773e-05,
+      "loss": 2.4782,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2546863344322256,
+      "grad_norm": 3.225481627547399,
+      "learning_rate": 1.9507283977310716e-05,
+      "loss": 2.3944,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25487884110980097,
+      "grad_norm": 3.1353652483189,
+      "learning_rate": 1.9506316889240027e-05,
+      "loss": 2.4734,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2550713477873764,
+      "grad_norm": 3.2868717470963733,
+      "learning_rate": 1.9505348877033715e-05,
+      "loss": 2.4282,
+      "step": 1325
+    },
+    {
+      "epoch": 0.25526385446495176,
+      "grad_norm": 2.876685465658946,
+      "learning_rate": 1.950437994078588e-05,
+      "loss": 2.4599,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2554563611425271,
+      "grad_norm": 3.3548750590399443,
+      "learning_rate": 1.950341008059072e-05,
+      "loss": 2.4458,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "grad_norm": 3.788430963411669,
+      "learning_rate": 1.9502439296542516e-05,
+      "loss": 2.4228,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.2098,
+      "step": 1328,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 1.9991,
+      "step": 1328,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.4696,
+      "step": 1328,
+      "vm_loss": 0.2352
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.272,
+      "step": 1328,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.3064,
+      "step": 1328,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.2167,
+      "step": 1328,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.2243,
+      "step": 1328,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.2556488678201025,
+      "lm_loss": 2.3176,
+      "step": 1328,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.2558413744976779,
+      "grad_norm": 3.157049014916354,
+      "learning_rate": 1.950146758873564e-05,
+      "loss": 2.4171,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2560338811752533,
+      "grad_norm": 3.217759995974044,
+      "learning_rate": 1.9500494957264556e-05,
+      "loss": 2.4448,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25622638785282864,
+      "grad_norm": 3.3655286097439756,
+      "learning_rate": 1.9499521402223816e-05,
+      "loss": 2.4174,
+      "step": 1331
+    },
+    {
+      "epoch": 0.256418894530404,
+      "grad_norm": 3.065707467387859,
+      "learning_rate": 1.949854692370806e-05,
+      "loss": 2.3962,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25661140120797943,
+      "grad_norm": 3.3265994165512383,
+      "learning_rate": 1.949757152181202e-05,
+      "loss": 2.4279,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2568039078855548,
+      "grad_norm": 3.047502649025238,
+      "learning_rate": 1.949659519663052e-05,
+      "loss": 2.4237,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25699641456313016,
+      "grad_norm": 3.81421034363058,
+      "learning_rate": 1.9495617948258473e-05,
+      "loss": 2.44,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "grad_norm": 3.4803572909974507,
+      "learning_rate": 1.949463977679087e-05,
+      "loss": 2.465,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.1166,
+      "step": 1336,
+      "vm_loss": 0.2559
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.0812,
+      "step": 1336,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 1.9693,
+      "step": 1336,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.096,
+      "step": 1336,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.21,
+      "step": 1336,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.2473,
+      "step": 1336,
+      "vm_loss": 0.2201
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.0484,
+      "step": 1336,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 0.2571889212407055,
+      "lm_loss": 2.4665,
+      "step": 1336,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.2573814279182809,
+      "grad_norm": 3.033006132100957,
+      "learning_rate": 1.9493660682322813e-05,
+      "loss": 2.4679,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2575739345958563,
+      "grad_norm": 3.205101448085657,
+      "learning_rate": 1.949268066494948e-05,
+      "loss": 2.4853,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2577664412734317,
+      "grad_norm": 3.831261203767552,
+      "learning_rate": 1.9491699724766136e-05,
+      "loss": 2.4555,
+      "step": 1339
+    },
+    {
+      "epoch": 0.25795894795100704,
+      "grad_norm": 2.973340038440222,
+      "learning_rate": 1.9490717861868146e-05,
+      "loss": 2.4245,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2581514546285824,
+      "grad_norm": 3.86195014991082,
+      "learning_rate": 1.948973507635096e-05,
+      "loss": 2.4634,
+      "step": 1341
+    },
+    {
+      "epoch": 0.25834396130615783,
+      "grad_norm": 3.4165985199006323,
+      "learning_rate": 1.9488751368310113e-05,
+      "loss": 2.4493,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2585364679837332,
+      "grad_norm": 3.2956956895908105,
+      "learning_rate": 1.9487766737841236e-05,
+      "loss": 2.4661,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "grad_norm": 3.1985477918133864,
+      "learning_rate": 1.9486781185040052e-05,
+      "loss": 2.4651,
+      "step": 1344
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 1.876,
+      "step": 1344,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 1.9058,
+      "step": 1344,
+      "vm_loss": 0.2471
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.281,
+      "step": 1344,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.579,
+      "step": 1344,
+      "vm_loss": 0.2354
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.5787,
+      "step": 1344,
+      "vm_loss": 0.2315
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.202,
+      "step": 1344,
+      "vm_loss": 0.099
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.4344,
+      "step": 1344,
+      "vm_loss": 0.2358
+    },
+    {
+      "epoch": 0.25872897466130856,
+      "lm_loss": 2.3623,
+      "step": 1344,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 0.25892148133888393,
+      "grad_norm": 3.765784192387748,
+      "learning_rate": 1.948579471000236e-05,
+      "loss": 2.4244,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2591139880164593,
+      "grad_norm": 3.79114609538082,
+      "learning_rate": 1.9484807312824066e-05,
+      "loss": 2.4661,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2593064946940347,
+      "grad_norm": 3.811668255443407,
+      "learning_rate": 1.9483818993601157e-05,
+      "loss": 2.4524,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2594990013716101,
+      "grad_norm": 3.8470479091873258,
+      "learning_rate": 1.9482829752429703e-05,
+      "loss": 2.4474,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25969150804918545,
+      "grad_norm": 2.8976283748457243,
+      "learning_rate": 1.948183958940588e-05,
+      "loss": 2.4393,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2598840147267608,
+      "grad_norm": 3.341725874920753,
+      "learning_rate": 1.9480848504625943e-05,
+      "loss": 2.4413,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26007652140433624,
+      "grad_norm": 3.8403607779619704,
+      "learning_rate": 1.9479856498186232e-05,
+      "loss": 2.4955,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "grad_norm": 3.2913681487474244,
+      "learning_rate": 1.947886357018319e-05,
+      "loss": 2.4731,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.0142,
+      "step": 1352,
+      "vm_loss": 0.2126
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.4495,
+      "step": 1352,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.1474,
+      "step": 1352,
+      "vm_loss": 0.2119
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.3824,
+      "step": 1352,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.2786,
+      "step": 1352,
+      "vm_loss": 0.2129
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.2902,
+      "step": 1352,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.0131,
+      "step": 1352,
+      "vm_loss": 0.1014
+    },
+    {
+      "epoch": 0.2602690280819116,
+      "lm_loss": 2.254,
+      "step": 1352,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 0.26046153475948697,
+      "grad_norm": 3.3369202959380155,
+      "learning_rate": 1.9477869720713335e-05,
+      "loss": 2.427,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26065404143706233,
+      "grad_norm": 4.0912223931257214,
+      "learning_rate": 1.9476874949873286e-05,
+      "loss": 2.4289,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2608465481146377,
+      "grad_norm": 3.6923710272736114,
+      "learning_rate": 1.9475879257759753e-05,
+      "loss": 2.455,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2610390547922131,
+      "grad_norm": 3.709818370142207,
+      "learning_rate": 1.9474882644469517e-05,
+      "loss": 2.4257,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2612315614697885,
+      "grad_norm": 3.4055689067583432,
+      "learning_rate": 1.947388511009947e-05,
+      "loss": 2.3915,
+      "step": 1357
+    },
+    {
+      "epoch": 0.26142406814736385,
+      "grad_norm": 3.9842501473511356,
+      "learning_rate": 1.9472886654746588e-05,
+      "loss": 2.4562,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2616165748249392,
+      "grad_norm": 2.952188124461196,
+      "learning_rate": 1.947188727850793e-05,
+      "loss": 2.4583,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "grad_norm": 3.953323780222254,
+      "learning_rate": 1.9470886981480642e-05,
+      "loss": 2.4813,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.3377,
+      "step": 1360,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.4006,
+      "step": 1360,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.3592,
+      "step": 1360,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.3334,
+      "step": 1360,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 1.6674,
+      "step": 1360,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.1496,
+      "step": 1360,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.1673,
+      "step": 1360,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 0.26180908150251464,
+      "lm_loss": 2.4051,
+      "step": 1360,
+      "vm_loss": 0.1144
+    },
+    {
+      "epoch": 0.26200158818009,
+      "grad_norm": 3.6364914509651243,
+      "learning_rate": 1.9469885763761975e-05,
+      "loss": 2.4193,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26219409485766537,
+      "grad_norm": 3.3224858959836725,
+      "learning_rate": 1.9468883625449256e-05,
+      "loss": 2.4438,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26238660153524074,
+      "grad_norm": 3.4044018426757514,
+      "learning_rate": 1.9467880566639908e-05,
+      "loss": 2.4129,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26257910821281616,
+      "grad_norm": 3.260301998937309,
+      "learning_rate": 1.9466876587431436e-05,
+      "loss": 2.463,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2627716148903915,
+      "grad_norm": 3.3845445974390356,
+      "learning_rate": 1.9465871687921444e-05,
+      "loss": 2.4312,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2629641215679669,
+      "grad_norm": 3.9139235265343135,
+      "learning_rate": 1.946486586820762e-05,
+      "loss": 2.3797,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26315662824554226,
+      "grad_norm": 3.106114386511253,
+      "learning_rate": 1.9463859128387745e-05,
+      "loss": 2.4967,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "grad_norm": 3.5513330707919453,
+      "learning_rate": 1.946285146855968e-05,
+      "loss": 2.5088,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.2648,
+      "step": 1368,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.1635,
+      "step": 1368,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.4339,
+      "step": 1368,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.4929,
+      "step": 1368,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.3344,
+      "step": 1368,
+      "vm_loss": 0.2022
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.3458,
+      "step": 1368,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.3191,
+      "step": 1368,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.2633491349231176,
+      "lm_loss": 2.2249,
+      "step": 1368,
+      "vm_loss": 0.1018
+    },
+    {
+      "epoch": 0.26354164160069304,
+      "grad_norm": 3.596890405237465,
+      "learning_rate": 1.946184288882139e-05,
+      "loss": 2.4508,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2637341482782684,
+      "grad_norm": 2.888969322878278,
+      "learning_rate": 1.946083338927092e-05,
+      "loss": 2.4324,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2639266549558438,
+      "grad_norm": 3.5671345359860647,
+      "learning_rate": 1.9459822970006407e-05,
+      "loss": 2.4462,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26411916163341914,
+      "grad_norm": 3.473380576223788,
+      "learning_rate": 1.945881163112607e-05,
+      "loss": 2.4195,
+      "step": 1372
+    },
+    {
+      "epoch": 0.26431166831099456,
+      "grad_norm": 3.4430168173110456,
+      "learning_rate": 1.945779937272823e-05,
+      "loss": 2.4703,
+      "step": 1373
+    },
+    {
+      "epoch": 0.26450417498856993,
+      "grad_norm": 3.4405994117511294,
+      "learning_rate": 1.9456786194911292e-05,
+      "loss": 2.4396,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2646966816661453,
+      "grad_norm": 3.115956749334274,
+      "learning_rate": 1.9455772097773746e-05,
+      "loss": 2.435,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "grad_norm": 3.3354418666958927,
+      "learning_rate": 1.945475708141418e-05,
+      "loss": 2.3856,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 1.8427,
+      "step": 1376,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.3545,
+      "step": 1376,
+      "vm_loss": 0.1236
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.474,
+      "step": 1376,
+      "vm_loss": 0.2317
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.6104,
+      "step": 1376,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.4971,
+      "step": 1376,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.3644,
+      "step": 1376,
+      "vm_loss": 0.2272
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.3845,
+      "step": 1376,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.26488918834372066,
+      "lm_loss": 2.2048,
+      "step": 1376,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 0.265081695021296,
+      "grad_norm": 3.283703394033432,
+      "learning_rate": 1.9453741145931263e-05,
+      "loss": 2.5054,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26527420169887145,
+      "grad_norm": 3.591374093888235,
+      "learning_rate": 1.9452724291423756e-05,
+      "loss": 2.4856,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2654667083764468,
+      "grad_norm": 3.104396396592494,
+      "learning_rate": 1.9451706517990516e-05,
+      "loss": 2.4176,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2656592150540222,
+      "grad_norm": 3.5614668340707767,
+      "learning_rate": 1.9450687825730476e-05,
+      "loss": 2.4932,
+      "step": 1380
+    },
+    {
+      "epoch": 0.26585172173159755,
+      "grad_norm": 2.9171927414336616,
+      "learning_rate": 1.9449668214742673e-05,
+      "loss": 2.4851,
+      "step": 1381
+    },
+    {
+      "epoch": 0.26604422840917297,
+      "grad_norm": 2.8508008003582366,
+      "learning_rate": 1.9448647685126222e-05,
+      "loss": 2.421,
+      "step": 1382
+    },
+    {
+      "epoch": 0.26623673508674833,
+      "grad_norm": 3.5220959827087968,
+      "learning_rate": 1.944762623698033e-05,
+      "loss": 2.487,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "grad_norm": 3.164982774887524,
+      "learning_rate": 1.94466038704043e-05,
+      "loss": 2.4083,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.43,
+      "step": 1384,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.1552,
+      "step": 1384,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.3703,
+      "step": 1384,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.3977,
+      "step": 1384,
+      "vm_loss": 0.2207
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.1224,
+      "step": 1384,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.366,
+      "step": 1384,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 1.9032,
+      "step": 1384,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 0.2664292417643237,
+      "lm_loss": 2.2334,
+      "step": 1384,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.26662174844189906,
+      "grad_norm": 2.750466161768323,
+      "learning_rate": 1.9445580585497516e-05,
+      "loss": 2.423,
+      "step": 1385
+    },
+    {
+      "epoch": 0.26681425511947443,
+      "grad_norm": 3.9582665977384788,
+      "learning_rate": 1.9444556382359454e-05,
+      "loss": 2.4138,
+      "step": 1386
+    },
+    {
+      "epoch": 0.26700676179704985,
+      "grad_norm": 3.1158391955155476,
+      "learning_rate": 1.9443531261089682e-05,
+      "loss": 2.4077,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2671992684746252,
+      "grad_norm": 3.0296666561383345,
+      "learning_rate": 1.944250522178785e-05,
+      "loss": 2.4164,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2673917751522006,
+      "grad_norm": 3.465929984323705,
+      "learning_rate": 1.944147826455371e-05,
+      "loss": 2.4274,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26758428182977595,
+      "grad_norm": 3.5465496574462967,
+      "learning_rate": 1.944045038948709e-05,
+      "loss": 2.4907,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26777678850735137,
+      "grad_norm": 3.156895609755201,
+      "learning_rate": 1.943942159668791e-05,
+      "loss": 2.4799,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "grad_norm": 3.4785568902455517,
+      "learning_rate": 1.943839188625619e-05,
+      "loss": 2.4296,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.3289,
+      "step": 1392,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.4526,
+      "step": 1392,
+      "vm_loss": 0.0885
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.3605,
+      "step": 1392,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.225,
+      "step": 1392,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.2868,
+      "step": 1392,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.3788,
+      "step": 1392,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.0234,
+      "step": 1392,
+      "vm_loss": 0.1437
+    },
+    {
+      "epoch": 0.26796929518492674,
+      "lm_loss": 2.0114,
+      "step": 1392,
+      "vm_loss": 0.2148
+    },
+    {
+      "epoch": 0.2681618018625021,
+      "grad_norm": 3.341224126619619,
+      "learning_rate": 1.9437361258292023e-05,
+      "loss": 2.4348,
+      "step": 1393
+    },
+    {
+      "epoch": 0.26835430854007747,
+      "grad_norm": 3.4269511921820475,
+      "learning_rate": 1.9436329712895605e-05,
+      "loss": 2.4465,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2685468152176529,
+      "grad_norm": 3.70836013637945,
+      "learning_rate": 1.943529725016721e-05,
+      "loss": 2.4456,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26873932189522826,
+      "grad_norm": 3.6695991073139846,
+      "learning_rate": 1.943426387020721e-05,
+      "loss": 2.466,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2689318285728036,
+      "grad_norm": 3.5660145683217275,
+      "learning_rate": 1.9433229573116067e-05,
+      "loss": 2.4619,
+      "step": 1397
+    },
+    {
+      "epoch": 0.269124335250379,
+      "grad_norm": 3.2775950533856335,
+      "learning_rate": 1.943219435899432e-05,
+      "loss": 2.4565,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26931684192795435,
+      "grad_norm": 3.6747508314816986,
+      "learning_rate": 1.9431158227942607e-05,
+      "loss": 2.4536,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "grad_norm": 3.035128726047039,
+      "learning_rate": 1.9430121180061654e-05,
+      "loss": 2.4246,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.3388,
+      "step": 1400,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.3608,
+      "step": 1400,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.2271,
+      "step": 1400,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.1778,
+      "step": 1400,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 1.8123,
+      "step": 1400,
+      "vm_loss": 0.2525
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.3682,
+      "step": 1400,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.2861,
+      "step": 1400,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 0.2695093486055298,
+      "lm_loss": 2.3043,
+      "step": 1400,
+      "vm_loss": 0.2017
+    },
+    {
+      "epoch": 0.26970185528310514,
+      "grad_norm": 3.505646957946543,
+      "learning_rate": 1.9429083215452276e-05,
+      "loss": 2.4131,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2698943619606805,
+      "grad_norm": 2.9293006954186085,
+      "learning_rate": 1.942804433421538e-05,
+      "loss": 2.4436,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2700868686382559,
+      "grad_norm": 3.884522514501745,
+      "learning_rate": 1.9427004536451952e-05,
+      "loss": 2.4756,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2702793753158313,
+      "grad_norm": 3.1622845713526013,
+      "learning_rate": 1.9425963822263077e-05,
+      "loss": 2.4612,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27047188199340666,
+      "grad_norm": 3.271885497090578,
+      "learning_rate": 1.9424922191749927e-05,
+      "loss": 2.3933,
+      "step": 1405
+    },
+    {
+      "epoch": 0.270664388670982,
+      "grad_norm": 3.0806264069388924,
+      "learning_rate": 1.9423879645013758e-05,
+      "loss": 2.4557,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2708568953485574,
+      "grad_norm": 3.158411007166911,
+      "learning_rate": 1.942283618215592e-05,
+      "loss": 2.4362,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "grad_norm": 3.442724779764118,
+      "learning_rate": 1.942179180327785e-05,
+      "loss": 2.4523,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.1825,
+      "step": 1408,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.425,
+      "step": 1408,
+      "vm_loss": 0.1232
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.5548,
+      "step": 1408,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.3955,
+      "step": 1408,
+      "vm_loss": 0.2327
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.2641,
+      "step": 1408,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 1.8005,
+      "step": 1408,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 1.9291,
+      "step": 1408,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 0.27104940202613276,
+      "lm_loss": 2.0288,
+      "step": 1408,
+      "vm_loss": 0.1916
+    },
+    {
+      "epoch": 0.2712419087037082,
+      "grad_norm": 3.6250834309428543,
+      "learning_rate": 1.9420746508481083e-05,
+      "loss": 2.4288,
+      "step": 1409
+    },
+    {
+      "epoch": 0.27143441538128354,
+      "grad_norm": 3.2004605722111075,
+      "learning_rate": 1.9419700297867227e-05,
+      "loss": 2.4353,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2716269220588589,
+      "grad_norm": 3.467379498683046,
+      "learning_rate": 1.9418653171537984e-05,
+      "loss": 2.4311,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2718194287364343,
+      "grad_norm": 3.892785395896328,
+      "learning_rate": 1.941760512959516e-05,
+      "loss": 2.4796,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2720119354140097,
+      "grad_norm": 3.8812363990587655,
+      "learning_rate": 1.9416556172140625e-05,
+      "loss": 2.4402,
+      "step": 1413
+    },
+    {
+      "epoch": 0.27220444209158506,
+      "grad_norm": 3.9808887010098664,
+      "learning_rate": 1.941550629927636e-05,
+      "loss": 2.3967,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27239694876916043,
+      "grad_norm": 2.971071485447531,
+      "learning_rate": 1.9414455511104423e-05,
+      "loss": 2.4938,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "grad_norm": 4.34825013055807,
+      "learning_rate": 1.9413403807726964e-05,
+      "loss": 2.4766,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.455,
+      "step": 1416,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.1661,
+      "step": 1416,
+      "vm_loss": 0.2036
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.1425,
+      "step": 1416,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 1.7711,
+      "step": 1416,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.3659,
+      "step": 1416,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.2913,
+      "step": 1416,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.3871,
+      "step": 1416,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.2725894554467358,
+      "lm_loss": 2.4161,
+      "step": 1416,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 0.27278196212431116,
+      "grad_norm": 3.80510396481745,
+      "learning_rate": 1.9412351189246222e-05,
+      "loss": 2.4145,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2729744688018866,
+      "grad_norm": 3.324689546315768,
+      "learning_rate": 1.9411297655764526e-05,
+      "loss": 2.4223,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27316697547946195,
+      "grad_norm": 3.2339236068441934,
+      "learning_rate": 1.9410243207384294e-05,
+      "loss": 2.4133,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2733594821570373,
+      "grad_norm": 3.3032509446274823,
+      "learning_rate": 1.9409187844208026e-05,
+      "loss": 2.4938,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2735519888346127,
+      "grad_norm": 3.4105288699708822,
+      "learning_rate": 1.940813156633832e-05,
+      "loss": 2.4174,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2737444955121881,
+      "grad_norm": 2.981732450523626,
+      "learning_rate": 1.9407074373877868e-05,
+      "loss": 2.4556,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27393700218976347,
+      "grad_norm": 3.483751064534896,
+      "learning_rate": 1.940601626692943e-05,
+      "loss": 2.4673,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "grad_norm": 3.637206778259412,
+      "learning_rate": 1.9404957245595872e-05,
+      "loss": 2.4044,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.5646,
+      "step": 1424,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.2616,
+      "step": 1424,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.2578,
+      "step": 1424,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.4856,
+      "step": 1424,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.0239,
+      "step": 1424,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.4159,
+      "step": 1424,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.0632,
+      "step": 1424,
+      "vm_loss": 0.2392
+    },
+    {
+      "epoch": 0.27412950886733883,
+      "lm_loss": 2.3561,
+      "step": 1424,
+      "vm_loss": 0.118
+    },
+    {
+      "epoch": 0.2743220155449142,
+      "grad_norm": 3.334940308214593,
+      "learning_rate": 1.9403897309980148e-05,
+      "loss": 2.4394,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27451452222248957,
+      "grad_norm": 3.359629194681667,
+      "learning_rate": 1.9402836460185294e-05,
+      "loss": 2.4494,
+      "step": 1426
+    },
+    {
+      "epoch": 0.274707028900065,
+      "grad_norm": 3.584232008094124,
+      "learning_rate": 1.9401774696314436e-05,
+      "loss": 2.4618,
+      "step": 1427
+    },
+    {
+      "epoch": 0.27489953557764035,
+      "grad_norm": 3.32365386990465,
+      "learning_rate": 1.9400712018470795e-05,
+      "loss": 2.4542,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2750920422552157,
+      "grad_norm": 3.3221741105192013,
+      "learning_rate": 1.9399648426757677e-05,
+      "loss": 2.4543,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2752845489327911,
+      "grad_norm": 3.5807113365247,
+      "learning_rate": 1.9398583921278473e-05,
+      "loss": 2.4822,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2754770556103665,
+      "grad_norm": 3.8576944735433862,
+      "learning_rate": 1.939751850213667e-05,
+      "loss": 2.4871,
+      "step": 1431
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "grad_norm": 3.3826688525455495,
+      "learning_rate": 1.939645216943584e-05,
+      "loss": 2.4167,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.2886,
+      "step": 1432,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.331,
+      "step": 1432,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.0088,
+      "step": 1432,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.0556,
+      "step": 1432,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.4222,
+      "step": 1432,
+      "vm_loss": 0.2471
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.308,
+      "step": 1432,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.2325,
+      "step": 1432,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 0.27566956228794187,
+      "lm_loss": 2.2871,
+      "step": 1432,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 3.1797972697553796,
+      "learning_rate": 1.9395384923279645e-05,
+      "loss": 2.4508,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2760545756430926,
+      "grad_norm": 3.2526859642537445,
+      "learning_rate": 1.939431676377183e-05,
+      "loss": 2.4567,
+      "step": 1434
+    },
+    {
+      "epoch": 0.276247082320668,
+      "grad_norm": 3.4052983077775942,
+      "learning_rate": 1.939324769101624e-05,
+      "loss": 2.4154,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2764395889982434,
+      "grad_norm": 3.4141383180878973,
+      "learning_rate": 1.9392177705116796e-05,
+      "loss": 2.4501,
+      "step": 1436
+    },
+    {
+      "epoch": 0.27663209567581876,
+      "grad_norm": 3.215217216481154,
+      "learning_rate": 1.939110680617752e-05,
+      "loss": 2.4618,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2768246023533941,
+      "grad_norm": 3.477741379150412,
+      "learning_rate": 1.9390034994302516e-05,
+      "loss": 2.4218,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2770171090309695,
+      "grad_norm": 3.2436896574811778,
+      "learning_rate": 1.9388962269595977e-05,
+      "loss": 2.4259,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "grad_norm": 3.2570097703567935,
+      "learning_rate": 1.9387888632162185e-05,
+      "loss": 2.4457,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.3304,
+      "step": 1440,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.2604,
+      "step": 1440,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.3713,
+      "step": 1440,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.3503,
+      "step": 1440,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 1.9513,
+      "step": 1440,
+      "vm_loss": 0.1283
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.3759,
+      "step": 1440,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.3531,
+      "step": 1440,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 0.2772096157085449,
+      "lm_loss": 2.2514,
+      "step": 1440,
+      "vm_loss": 0.0969
+    },
+    {
+      "epoch": 0.2774021223861203,
+      "grad_norm": 3.3567399709409247,
+      "learning_rate": 1.938681408210551e-05,
+      "loss": 2.4071,
+      "step": 1441
+    },
+    {
+      "epoch": 0.27759462906369564,
+      "grad_norm": 3.2675266602580595,
+      "learning_rate": 1.9385738619530415e-05,
+      "loss": 2.4457,
+      "step": 1442
+    },
+    {
+      "epoch": 0.277787135741271,
+      "grad_norm": 3.405524210116633,
+      "learning_rate": 1.9384662244541448e-05,
+      "loss": 2.4597,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27797964241884643,
+      "grad_norm": 3.7974134684640584,
+      "learning_rate": 1.9383584957243245e-05,
+      "loss": 2.3995,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2781721490964218,
+      "grad_norm": 2.9300876578354194,
+      "learning_rate": 1.9382506757740532e-05,
+      "loss": 2.4493,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27836465577399716,
+      "grad_norm": 4.225220921943164,
+      "learning_rate": 1.938142764613813e-05,
+      "loss": 2.4154,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2785571624515725,
+      "grad_norm": 3.418211995090257,
+      "learning_rate": 1.9380347622540932e-05,
+      "loss": 2.447,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "grad_norm": 3.706350884167747,
+      "learning_rate": 1.937926668705394e-05,
+      "loss": 2.4166,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.2721,
+      "step": 1448,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.0539,
+      "step": 1448,
+      "vm_loss": 0.2276
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.3414,
+      "step": 1448,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.2531,
+      "step": 1448,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.1248,
+      "step": 1448,
+      "vm_loss": 0.2533
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.3428,
+      "step": 1448,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.1109,
+      "step": 1448,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.2787496691291479,
+      "lm_loss": 2.3689,
+      "step": 1448,
+      "vm_loss": 0.0945
+    },
+    {
+      "epoch": 0.2789421758067233,
+      "grad_norm": 3.30134788915163,
+      "learning_rate": 1.9378184839782226e-05,
+      "loss": 2.4076,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2791346824842987,
+      "grad_norm": 3.2799994295793544,
+      "learning_rate": 1.937710208083096e-05,
+      "loss": 2.4107,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27932718916187405,
+      "grad_norm": 3.1642706198204884,
+      "learning_rate": 1.9376018410305414e-05,
+      "loss": 2.4124,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2795196958394494,
+      "grad_norm": 3.558069638639298,
+      "learning_rate": 1.9374933828310917e-05,
+      "loss": 2.4536,
+      "step": 1452
+    },
+    {
+      "epoch": 0.27971220251702483,
+      "grad_norm": 3.9670991485634928,
+      "learning_rate": 1.9373848334952913e-05,
+      "loss": 2.4421,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2799047091946002,
+      "grad_norm": 2.934392409520615,
+      "learning_rate": 1.9372761930336925e-05,
+      "loss": 2.4224,
+      "step": 1454
+    },
+    {
+      "epoch": 0.28009721587217556,
+      "grad_norm": 3.0701092166421446,
+      "learning_rate": 1.9371674614568565e-05,
+      "loss": 2.4721,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "grad_norm": 3.4830205583229037,
+      "learning_rate": 1.9370586387753532e-05,
+      "loss": 2.4002,
+      "step": 1456
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.1842,
+      "step": 1456,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.3036,
+      "step": 1456,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.3776,
+      "step": 1456,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.3612,
+      "step": 1456,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.0333,
+      "step": 1456,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.639,
+      "step": 1456,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.1852,
+      "step": 1456,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 0.28028972254975093,
+      "lm_loss": 2.0453,
+      "step": 1456,
+      "vm_loss": 0.2606
+    },
+    {
+      "epoch": 0.2804822292273263,
+      "grad_norm": 3.4796728604936105,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 2.4067,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2806747359049017,
+      "grad_norm": 3.4553376674331338,
+      "learning_rate": 1.93684072014067e-05,
+      "loss": 2.4537,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2808672425824771,
+      "grad_norm": 3.5616219282649104,
+      "learning_rate": 1.9367316242086747e-05,
+      "loss": 2.4485,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28105974926005245,
+      "grad_norm": 3.831548443895444,
+      "learning_rate": 1.936622437214381e-05,
+      "loss": 2.4071,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2812522559376278,
+      "grad_norm": 3.7627822599846765,
+      "learning_rate": 1.9365131591684037e-05,
+      "loss": 2.4595,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28144476261520324,
+      "grad_norm": 3.3081301772927865,
+      "learning_rate": 1.936403790081366e-05,
+      "loss": 2.4481,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2816372692927786,
+      "grad_norm": 3.333841573553775,
+      "learning_rate": 1.9362943299638994e-05,
+      "loss": 2.4035,
+      "step": 1463
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "grad_norm": 4.0084181972830075,
+      "learning_rate": 1.9361847788266455e-05,
+      "loss": 2.4311,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.1462,
+      "step": 1464,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.2539,
+      "step": 1464,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.0586,
+      "step": 1464,
+      "vm_loss": 0.1986
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.3023,
+      "step": 1464,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.4273,
+      "step": 1464,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.1696,
+      "step": 1464,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.3907,
+      "step": 1464,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 0.28182977597035397,
+      "lm_loss": 2.0218,
+      "step": 1464,
+      "vm_loss": 0.2175
+    },
+    {
+      "epoch": 0.28202228264792933,
+      "grad_norm": 3.3816599717158793,
+      "learning_rate": 1.9360751366802534e-05,
+      "loss": 2.4262,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28221478932550476,
+      "grad_norm": 3.031942496017371,
+      "learning_rate": 1.9359654035353825e-05,
+      "loss": 2.3665,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2824072960030801,
+      "grad_norm": 3.0897395198458635,
+      "learning_rate": 1.9358555794027e-05,
+      "loss": 2.4771,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2825998026806555,
+      "grad_norm": 3.0227620930025934,
+      "learning_rate": 1.9357456642928817e-05,
+      "loss": 2.3985,
+      "step": 1468
+    },
+    {
+      "epoch": 0.28279230935823085,
+      "grad_norm": 3.3676118116721767,
+      "learning_rate": 1.9356356582166134e-05,
+      "loss": 2.3991,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2829848160358062,
+      "grad_norm": 3.157686084310653,
+      "learning_rate": 1.9355255611845885e-05,
+      "loss": 2.4119,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28317732271338164,
+      "grad_norm": 3.118219953202858,
+      "learning_rate": 1.9354153732075106e-05,
+      "loss": 2.4515,
+      "step": 1471
+    },
+    {
+      "epoch": 0.283369829390957,
+      "grad_norm": 2.8497602634085433,
+      "learning_rate": 1.935305094296091e-05,
+      "loss": 2.4476,
+      "step": 1472
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.065,
+      "step": 1472,
+      "vm_loss": 0.2413
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.0688,
+      "step": 1472,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.1513,
+      "step": 1472,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.4929,
+      "step": 1472,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.2769,
+      "step": 1472,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.2175,
+      "step": 1472,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 1.8374,
+      "step": 1472,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.283369829390957,
+      "lm_loss": 2.2375,
+      "step": 1472,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.2835623360685324,
+      "grad_norm": 2.750610238396522,
+      "learning_rate": 1.93519472446105e-05,
+      "loss": 2.3855,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28375484274610774,
+      "grad_norm": 3.167429351489139,
+      "learning_rate": 1.9350842637131172e-05,
+      "loss": 2.4213,
+      "step": 1474
+    },
+    {
+      "epoch": 0.28394734942368316,
+      "grad_norm": 2.688577038309872,
+      "learning_rate": 1.9349737120630302e-05,
+      "loss": 2.4289,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2841398561012585,
+      "grad_norm": 3.8624307457942,
+      "learning_rate": 1.934863069521537e-05,
+      "loss": 2.3499,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2843323627788339,
+      "grad_norm": 3.1920537613772146,
+      "learning_rate": 1.9347523360993933e-05,
+      "loss": 2.4393,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28452486945640926,
+      "grad_norm": 3.4810979008869802,
+      "learning_rate": 1.9346415118073634e-05,
+      "loss": 2.451,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2847173761339846,
+      "grad_norm": 3.6482575452682764,
+      "learning_rate": 1.934530596656221e-05,
+      "loss": 2.4376,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "grad_norm": 3.0546001127709097,
+      "learning_rate": 1.9344195906567488e-05,
+      "loss": 2.407,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.3662,
+      "step": 1480,
+      "vm_loss": 0.1893
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.3432,
+      "step": 1480,
+      "vm_loss": 0.2485
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.451,
+      "step": 1480,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.423,
+      "step": 1480,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.2159,
+      "step": 1480,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 1.9316,
+      "step": 1480,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.3108,
+      "step": 1480,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.28490988281156004,
+      "lm_loss": 2.2624,
+      "step": 1480,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.2851023894891354,
+      "grad_norm": 3.2276755655245597,
+      "learning_rate": 1.934308493819737e-05,
+      "loss": 2.4757,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2852948961667108,
+      "grad_norm": 3.0527062119395536,
+      "learning_rate": 1.9341973061559866e-05,
+      "loss": 2.3694,
+      "step": 1482
+    },
+    {
+      "epoch": 0.28548740284428614,
+      "grad_norm": 2.7659991155426846,
+      "learning_rate": 1.9340860276763062e-05,
+      "loss": 2.4144,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28567990952186156,
+      "grad_norm": 3.934423341342164,
+      "learning_rate": 1.9339746583915135e-05,
+      "loss": 2.4713,
+      "step": 1484
+    },
+    {
+      "epoch": 0.28587241619943693,
+      "grad_norm": 2.911922614405069,
+      "learning_rate": 1.9338631983124348e-05,
+      "loss": 2.412,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2860649228770123,
+      "grad_norm": 3.2962340737937397,
+      "learning_rate": 1.9337516474499058e-05,
+      "loss": 2.4345,
+      "step": 1486
+    },
+    {
+      "epoch": 0.28625742955458766,
+      "grad_norm": 3.534613269749138,
+      "learning_rate": 1.9336400058147707e-05,
+      "loss": 2.4425,
+      "step": 1487
+    },
+    {
+      "epoch": 0.286449936232163,
+      "grad_norm": 2.925708784736393,
+      "learning_rate": 1.9335282734178816e-05,
+      "loss": 2.4239,
+      "step": 1488
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.0142,
+      "step": 1488,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.2237,
+      "step": 1488,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.2606,
+      "step": 1488,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.0704,
+      "step": 1488,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.3242,
+      "step": 1488,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.1596,
+      "step": 1488,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.6922,
+      "step": 1488,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.286449936232163,
+      "lm_loss": 2.0314,
+      "step": 1488,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.28664244290973845,
+      "grad_norm": 3.432725225137702,
+      "learning_rate": 1.9334164502701017e-05,
+      "loss": 2.3839,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2868349495873138,
+      "grad_norm": 3.6911435427842907,
+      "learning_rate": 1.9333045363823005e-05,
+      "loss": 2.4592,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2870274562648892,
+      "grad_norm": 2.9310357638767326,
+      "learning_rate": 1.9331925317653583e-05,
+      "loss": 2.4692,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28721996294246455,
+      "grad_norm": 3.263490254763622,
+      "learning_rate": 1.933080436430163e-05,
+      "loss": 2.4359,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28741246962003997,
+      "grad_norm": 3.2613532343481295,
+      "learning_rate": 1.9329682503876115e-05,
+      "loss": 2.4698,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28760497629761533,
+      "grad_norm": 3.7611595173855354,
+      "learning_rate": 1.93285597364861e-05,
+      "loss": 2.39,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2877974829751907,
+      "grad_norm": 3.0550631689166075,
+      "learning_rate": 1.932743606224073e-05,
+      "loss": 2.4212,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "grad_norm": 3.1079278386130396,
+      "learning_rate": 1.932631148124924e-05,
+      "loss": 2.4433,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.0973,
+      "step": 1496,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.1244,
+      "step": 1496,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.4423,
+      "step": 1496,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.1986,
+      "step": 1496,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.4945,
+      "step": 1496,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.4617,
+      "step": 1496,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.3319,
+      "step": 1496,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 0.28798998965276607,
+      "lm_loss": 2.3622,
+      "step": 1496,
+      "vm_loss": 0.103
+    },
+    {
+      "epoch": 0.28818249633034143,
+      "grad_norm": 3.408789973939458,
+      "learning_rate": 1.9325185993620962e-05,
+      "loss": 2.4686,
+      "step": 1497
+    },
+    {
+      "epoch": 0.28837500300791685,
+      "grad_norm": 3.553674151823099,
+      "learning_rate": 1.9324059599465296e-05,
+      "loss": 2.4361,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2885675096854922,
+      "grad_norm": 3.80201610175442,
+      "learning_rate": 1.932293229889175e-05,
+      "loss": 2.4753,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2887600163630676,
+      "grad_norm": 3.150315227405042,
+      "learning_rate": 1.932180409200991e-05,
+      "loss": 2.44,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28895252304064295,
+      "grad_norm": 3.0200214473873155,
+      "learning_rate": 1.9320674978929447e-05,
+      "loss": 2.4222,
+      "step": 1501
+    },
+    {
+      "epoch": 0.28914502971821837,
+      "grad_norm": 3.5690659587874842,
+      "learning_rate": 1.9319544959760135e-05,
+      "loss": 2.4609,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28933753639579374,
+      "grad_norm": 3.0061943532336786,
+      "learning_rate": 1.931841403461182e-05,
+      "loss": 2.4438,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "grad_norm": 3.5862230381876707,
+      "learning_rate": 1.9317282203594438e-05,
+      "loss": 2.4467,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.4456,
+      "step": 1504,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.4268,
+      "step": 1504,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.2311,
+      "step": 1504,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.253,
+      "step": 1504,
+      "vm_loss": 0.2262
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.2247,
+      "step": 1504,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.3179,
+      "step": 1504,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 1.9328,
+      "step": 1504,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 0.2895300430733691,
+      "lm_loss": 2.2715,
+      "step": 1504,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.28972254975094447,
+      "grad_norm": 3.6304841617843415,
+      "learning_rate": 1.931614946681803e-05,
+      "loss": 2.4679,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2899150564285199,
+      "grad_norm": 3.647313916477359,
+      "learning_rate": 1.9315015824392703e-05,
+      "loss": 2.4313,
+      "step": 1506
+    },
+    {
+      "epoch": 0.29010756310609526,
+      "grad_norm": 3.466466129136221,
+      "learning_rate": 1.9313881276428663e-05,
+      "loss": 2.415,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2903000697836706,
+      "grad_norm": 3.071139527153352,
+      "learning_rate": 1.9312745823036204e-05,
+      "loss": 2.4442,
+      "step": 1508
+    },
+    {
+      "epoch": 0.290492576461246,
+      "grad_norm": 3.048578618265307,
+      "learning_rate": 1.931160946432571e-05,
+      "loss": 2.4066,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29068508313882135,
+      "grad_norm": 3.292399401220064,
+      "learning_rate": 1.931047220040764e-05,
+      "loss": 2.4343,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2908775898163968,
+      "grad_norm": 3.02061935349102,
+      "learning_rate": 1.930933403139256e-05,
+      "loss": 2.4333,
+      "step": 1511
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "grad_norm": 3.246994924559873,
+      "learning_rate": 1.9308194957391108e-05,
+      "loss": 2.4321,
+      "step": 1512
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.0297,
+      "step": 1512,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.309,
+      "step": 1512,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.1463,
+      "step": 1512,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.4213,
+      "step": 1512,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.3265,
+      "step": 1512,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 1.9666,
+      "step": 1512,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.3875,
+      "step": 1512,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 0.29107009649397214,
+      "lm_loss": 2.4664,
+      "step": 1512,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 0.2912626031715475,
+      "grad_norm": 3.2806053534608557,
+      "learning_rate": 1.930705497851402e-05,
+      "loss": 2.3972,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2914551098491229,
+      "grad_norm": 2.8277551831830317,
+      "learning_rate": 1.930591409487212e-05,
+      "loss": 2.469,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2916476165266983,
+      "grad_norm": 3.0564067918013795,
+      "learning_rate": 1.9304772306576313e-05,
+      "loss": 2.4073,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29184012320427366,
+      "grad_norm": 3.3798323473530063,
+      "learning_rate": 1.9303629613737592e-05,
+      "loss": 2.5061,
+      "step": 1516
+    },
+    {
+      "epoch": 0.292032629881849,
+      "grad_norm": 3.141003667262534,
+      "learning_rate": 1.9302486016467044e-05,
+      "loss": 2.4846,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2922251365594244,
+      "grad_norm": 4.086867886740118,
+      "learning_rate": 1.930134151487584e-05,
+      "loss": 2.4152,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29241764323699976,
+      "grad_norm": 3.1399383187177756,
+      "learning_rate": 1.9300196109075246e-05,
+      "loss": 2.3922,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "grad_norm": 2.9732394632524817,
+      "learning_rate": 1.9299049799176602e-05,
+      "loss": 2.478,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.2772,
+      "step": 1520,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.0747,
+      "step": 1520,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.0432,
+      "step": 1520,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.4983,
+      "step": 1520,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.2147,
+      "step": 1520,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.2581,
+      "step": 1520,
+      "vm_loss": 0.1976
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.3986,
+      "step": 1520,
+      "vm_loss": 0.2364
+    },
+    {
+      "epoch": 0.2926101499145752,
+      "lm_loss": 2.2722,
+      "step": 1520,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 0.29280265659215055,
+      "grad_norm": 3.491632012710889,
+      "learning_rate": 1.9297902585291353e-05,
+      "loss": 2.4289,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2929951632697259,
+      "grad_norm": 3.1107432971832485,
+      "learning_rate": 1.9296754467531015e-05,
+      "loss": 2.4146,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2931876699473013,
+      "grad_norm": 3.3132383473598597,
+      "learning_rate": 1.92956054460072e-05,
+      "loss": 2.3997,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2933801766248767,
+      "grad_norm": 3.2747616389183585,
+      "learning_rate": 1.929445552083161e-05,
+      "loss": 2.4231,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29357268330245206,
+      "grad_norm": 2.6655935791159786,
+      "learning_rate": 1.9293304692116037e-05,
+      "loss": 2.4428,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29376518998002743,
+      "grad_norm": 2.983970671287051,
+      "learning_rate": 1.929215295997235e-05,
+      "loss": 2.4056,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2939576966576028,
+      "grad_norm": 3.090315462183964,
+      "learning_rate": 1.9291000324512514e-05,
+      "loss": 2.3686,
+      "step": 1527
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "grad_norm": 2.69929539644961,
+      "learning_rate": 1.928984678584858e-05,
+      "loss": 2.4185,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.4492,
+      "step": 1528,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.2989,
+      "step": 1528,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.1769,
+      "step": 1528,
+      "vm_loss": 0.1225
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.2638,
+      "step": 1528,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.5292,
+      "step": 1528,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.2534,
+      "step": 1528,
+      "vm_loss": 0.1369
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.1655,
+      "step": 1528,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.29415020333517816,
+      "lm_loss": 2.466,
+      "step": 1528,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.2943427100127536,
+      "grad_norm": 3.3696144018030414,
+      "learning_rate": 1.9288692344092686e-05,
+      "loss": 2.4349,
+      "step": 1529
+    },
+    {
+      "epoch": 0.29453521669032895,
+      "grad_norm": 3.1253590698791527,
+      "learning_rate": 1.928753699935706e-05,
+      "loss": 2.4333,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2947277233679043,
+      "grad_norm": 3.0896832279657245,
+      "learning_rate": 1.9286380751754013e-05,
+      "loss": 2.4346,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2949202300454797,
+      "grad_norm": 3.374611174377119,
+      "learning_rate": 1.9285223601395953e-05,
+      "loss": 2.4372,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2951127367230551,
+      "grad_norm": 3.1184706610359454,
+      "learning_rate": 1.9284065548395362e-05,
+      "loss": 2.4198,
+      "step": 1533
+    },
+    {
+      "epoch": 0.29530524340063047,
+      "grad_norm": 2.650637065578818,
+      "learning_rate": 1.9282906592864826e-05,
+      "loss": 2.4497,
+      "step": 1534
+    },
+    {
+      "epoch": 0.29549775007820583,
+      "grad_norm": 3.56945496321312,
+      "learning_rate": 1.9281746734917005e-05,
+      "loss": 2.4252,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "grad_norm": 3.5283418770831028,
+      "learning_rate": 1.9280585974664653e-05,
+      "loss": 2.3906,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.2659,
+      "step": 1536,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.3326,
+      "step": 1536,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.1007,
+      "step": 1536,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.1906,
+      "step": 1536,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.3567,
+      "step": 1536,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.241,
+      "step": 1536,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.5681,
+      "step": 1536,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.2956902567557812,
+      "lm_loss": 2.2083,
+      "step": 1536,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 0.2958827634333566,
+      "grad_norm": 3.4897159460884124,
+      "learning_rate": 1.927942431222061e-05,
+      "loss": 2.4126,
+      "step": 1537
+    },
+    {
+      "epoch": 0.296075270110932,
+      "grad_norm": 3.3246150847859552,
+      "learning_rate": 1.9278261747697804e-05,
+      "loss": 2.4071,
+      "step": 1538
+    },
+    {
+      "epoch": 0.29626777678850735,
+      "grad_norm": 3.3105966071770156,
+      "learning_rate": 1.9277098281209253e-05,
+      "loss": 2.4308,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2964602834660827,
+      "grad_norm": 2.9720637859340067,
+      "learning_rate": 1.9275933912868065e-05,
+      "loss": 2.465,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2966527901436581,
+      "grad_norm": 3.774152386581912,
+      "learning_rate": 1.927476864278742e-05,
+      "loss": 2.4356,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2968452968212335,
+      "grad_norm": 3.2326649121042825,
+      "learning_rate": 1.927360247108061e-05,
+      "loss": 2.4369,
+      "step": 1542
+    },
+    {
+      "epoch": 0.29703780349880887,
+      "grad_norm": 3.482446958667124,
+      "learning_rate": 1.9272435397860992e-05,
+      "loss": 2.3995,
+      "step": 1543
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "grad_norm": 3.528865631555095,
+      "learning_rate": 1.9271267423242028e-05,
+      "loss": 2.4558,
+      "step": 1544
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.3403,
+      "step": 1544,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 1.8689,
+      "step": 1544,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.3338,
+      "step": 1544,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.3295,
+      "step": 1544,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.2525,
+      "step": 1544,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.4347,
+      "step": 1544,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.3204,
+      "step": 1544,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 0.29723031017638424,
+      "lm_loss": 2.1844,
+      "step": 1544,
+      "vm_loss": 0.2673
+    },
+    {
+      "epoch": 0.2974228168539596,
+      "grad_norm": 3.2098407381664082,
+      "learning_rate": 1.927009854733725e-05,
+      "loss": 2.4554,
+      "step": 1545
+    },
+    {
+      "epoch": 0.297615323531535,
+      "grad_norm": 3.4178964317192593,
+      "learning_rate": 1.9268928770260298e-05,
+      "loss": 2.4251,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2978078302091104,
+      "grad_norm": 2.9512264174928124,
+      "learning_rate": 1.9267758092124883e-05,
+      "loss": 2.4219,
+      "step": 1547
+    },
+    {
+      "epoch": 0.29800033688668576,
+      "grad_norm": 3.112352823869454,
+      "learning_rate": 1.9266586513044818e-05,
+      "loss": 2.4294,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2981928435642611,
+      "grad_norm": 3.579117818917497,
+      "learning_rate": 1.9265414033133986e-05,
+      "loss": 2.4237,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2983853502418365,
+      "grad_norm": 3.3559863716959075,
+      "learning_rate": 1.926424065250637e-05,
+      "loss": 2.4521,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2985778569194119,
+      "grad_norm": 2.72048828887211,
+      "learning_rate": 1.9263066371276037e-05,
+      "loss": 2.4223,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "grad_norm": 3.2101721761303756,
+      "learning_rate": 1.9261891189557146e-05,
+      "loss": 2.4387,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.2197,
+      "step": 1552,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.4019,
+      "step": 1552,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.2678,
+      "step": 1552,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.3116,
+      "step": 1552,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.105,
+      "step": 1552,
+      "vm_loss": 0.2016
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.0397,
+      "step": 1552,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.3848,
+      "step": 1552,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.2987703635969873,
+      "lm_loss": 2.1193,
+      "step": 1552,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 0.29896287027456264,
+      "grad_norm": 3.314986146789833,
+      "learning_rate": 1.9260715107463935e-05,
+      "loss": 2.4265,
+      "step": 1553
+    },
+    {
+      "epoch": 0.299155376952138,
+      "grad_norm": 2.782186426301867,
+      "learning_rate": 1.925953812511074e-05,
+      "loss": 2.3966,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29934788362971343,
+      "grad_norm": 2.9291941363969816,
+      "learning_rate": 1.9258360242611973e-05,
+      "loss": 2.4862,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2995403903072888,
+      "grad_norm": 2.9244554928875646,
+      "learning_rate": 1.925718146008214e-05,
+      "loss": 2.388,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29973289698486416,
+      "grad_norm": 2.8639366877979575,
+      "learning_rate": 1.925600177763584e-05,
+      "loss": 2.4432,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2999254036624395,
+      "grad_norm": 3.5674673086339546,
+      "learning_rate": 1.9254821195387744e-05,
+      "loss": 2.3614,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3001179103400149,
+      "grad_norm": 3.061340156963764,
+      "learning_rate": 1.9253639713452624e-05,
+      "loss": 2.4346,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "grad_norm": 3.3531783522059557,
+      "learning_rate": 1.9252457331945336e-05,
+      "loss": 2.4009,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.0283,
+      "step": 1560,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.334,
+      "step": 1560,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.1629,
+      "step": 1560,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.0658,
+      "step": 1560,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.3044,
+      "step": 1560,
+      "vm_loss": 0.2496
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.354,
+      "step": 1560,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.1524,
+      "step": 1560,
+      "vm_loss": 0.2335
+    },
+    {
+      "epoch": 0.3003104170175903,
+      "lm_loss": 2.1873,
+      "step": 1560,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.3005029236951657,
+      "grad_norm": 3.167637171389351,
+      "learning_rate": 1.9251274050980825e-05,
+      "loss": 2.412,
+      "step": 1561
+    },
+    {
+      "epoch": 0.30069543037274105,
+      "grad_norm": 3.0716152116676283,
+      "learning_rate": 1.9250089870674115e-05,
+      "loss": 2.397,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3008879370503164,
+      "grad_norm": 3.015825236503033,
+      "learning_rate": 1.9248904791140325e-05,
+      "loss": 2.4303,
+      "step": 1563
+    },
+    {
+      "epoch": 0.30108044372789183,
+      "grad_norm": 3.393406559313713,
+      "learning_rate": 1.9247718812494662e-05,
+      "loss": 2.4367,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3012729504054672,
+      "grad_norm": 3.283385971741707,
+      "learning_rate": 1.924653193485242e-05,
+      "loss": 2.3824,
+      "step": 1565
+    },
+    {
+      "epoch": 0.30146545708304257,
+      "grad_norm": 2.945418391453257,
+      "learning_rate": 1.9245344158328975e-05,
+      "loss": 2.4084,
+      "step": 1566
+    },
+    {
+      "epoch": 0.30165796376061793,
+      "grad_norm": 2.659057019934935,
+      "learning_rate": 1.9244155483039794e-05,
+      "loss": 2.4126,
+      "step": 1567
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "grad_norm": 2.99504473620692,
+      "learning_rate": 1.924296590910043e-05,
+      "loss": 2.4431,
+      "step": 1568
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 1.8937,
+      "step": 1568,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.0376,
+      "step": 1568,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.1112,
+      "step": 1568,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.2003,
+      "step": 1568,
+      "vm_loss": 0.1013
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.2269,
+      "step": 1568,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.4779,
+      "step": 1568,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.1138,
+      "step": 1568,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 0.30185047043819335,
+      "lm_loss": 2.1316,
+      "step": 1568,
+      "vm_loss": 0.1095
+    },
+    {
+      "epoch": 0.3020429771157687,
+      "grad_norm": 2.584119312003221,
+      "learning_rate": 1.9241775436626535e-05,
+      "loss": 2.4011,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3022354837933441,
+      "grad_norm": 2.7809270869327407,
+      "learning_rate": 1.924058406573383e-05,
+      "loss": 2.4035,
+      "step": 1570
+    },
+    {
+      "epoch": 0.30242799047091945,
+      "grad_norm": 3.1652180231530807,
+      "learning_rate": 1.9239391796538125e-05,
+      "loss": 2.4615,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3026204971484948,
+      "grad_norm": 3.122009999204937,
+      "learning_rate": 1.9238198629155337e-05,
+      "loss": 2.4567,
+      "step": 1572
+    },
+    {
+      "epoch": 0.30281300382607024,
+      "grad_norm": 2.919014482300737,
+      "learning_rate": 1.9237004563701446e-05,
+      "loss": 2.4131,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3030055105036456,
+      "grad_norm": 3.4029298421946277,
+      "learning_rate": 1.9235809600292542e-05,
+      "loss": 2.4223,
+      "step": 1574
+    },
+    {
+      "epoch": 0.30319801718122097,
+      "grad_norm": 3.1332074317736507,
+      "learning_rate": 1.923461373904478e-05,
+      "loss": 2.4395,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "grad_norm": 2.648814162106668,
+      "learning_rate": 1.923341698007442e-05,
+      "loss": 2.4365,
+      "step": 1576
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 2.3981,
+      "step": 1576,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 2.1836,
+      "step": 1576,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 2.0677,
+      "step": 1576,
+      "vm_loss": 0.1046
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 1.6026,
+      "step": 1576,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 1.8092,
+      "step": 1576,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 2.376,
+      "step": 1576,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 2.3269,
+      "step": 1576,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.30339052385879633,
+      "lm_loss": 1.8972,
+      "step": 1576,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 0.30358303053637176,
+      "grad_norm": 3.3762109369248328,
+      "learning_rate": 1.9232219323497797e-05,
+      "loss": 2.3929,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3037755372139471,
+      "grad_norm": 3.4984837775593918,
+      "learning_rate": 1.9231020769431344e-05,
+      "loss": 2.4075,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3039680438915225,
+      "grad_norm": 2.8195248783062508,
+      "learning_rate": 1.9229821317991572e-05,
+      "loss": 2.4866,
+      "step": 1579
+    },
+    {
+      "epoch": 0.30416055056909785,
+      "grad_norm": 3.624294792634172,
+      "learning_rate": 1.922862096929508e-05,
+      "loss": 2.4038,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3043530572466732,
+      "grad_norm": 3.0880074995300597,
+      "learning_rate": 1.9227419723458565e-05,
+      "loss": 2.4406,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30454556392424864,
+      "grad_norm": 2.9737859978674805,
+      "learning_rate": 1.9226217580598804e-05,
+      "loss": 2.4169,
+      "step": 1582
+    },
+    {
+      "epoch": 0.304738070601824,
+      "grad_norm": 3.093610857310697,
+      "learning_rate": 1.922501454083265e-05,
+      "loss": 2.4537,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "grad_norm": 3.520144681125478,
+      "learning_rate": 1.9223810604277063e-05,
+      "loss": 2.4117,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.4783,
+      "step": 1584,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.4083,
+      "step": 1584,
+      "vm_loss": 0.2393
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.5333,
+      "step": 1584,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.2042,
+      "step": 1584,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.4878,
+      "step": 1584,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.4179,
+      "step": 1584,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.0712,
+      "step": 1584,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.3049305772793994,
+      "lm_loss": 2.2991,
+      "step": 1584,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.30512308395697474,
+      "grad_norm": 3.00242119123407,
+      "learning_rate": 1.9222605771049076e-05,
+      "loss": 2.4927,
+      "step": 1585
+    },
+    {
+      "epoch": 0.30531559063455016,
+      "grad_norm": 3.147923982339829,
+      "learning_rate": 1.9221400041265818e-05,
+      "loss": 2.3832,
+      "step": 1586
+    },
+    {
+      "epoch": 0.3055080973121255,
+      "grad_norm": 3.000091035285722,
+      "learning_rate": 1.9220193415044498e-05,
+      "loss": 2.4266,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3057006039897009,
+      "grad_norm": 3.1592738723604588,
+      "learning_rate": 1.921898589250242e-05,
+      "loss": 2.4278,
+      "step": 1588
+    },
+    {
+      "epoch": 0.30589311066727626,
+      "grad_norm": 3.075560665763926,
+      "learning_rate": 1.9217777473756967e-05,
+      "loss": 2.394,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3060856173448516,
+      "grad_norm": 2.7858440510575746,
+      "learning_rate": 1.921656815892561e-05,
+      "loss": 2.4336,
+      "step": 1590
+    },
+    {
+      "epoch": 0.30627812402242705,
+      "grad_norm": 3.535930254340447,
+      "learning_rate": 1.9215357948125918e-05,
+      "loss": 2.4183,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "grad_norm": 3.1682366203268915,
+      "learning_rate": 1.9214146841475536e-05,
+      "loss": 2.4321,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.1046,
+      "step": 1592,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.2586,
+      "step": 1592,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.1536,
+      "step": 1592,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.2727,
+      "step": 1592,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.2912,
+      "step": 1592,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.1954,
+      "step": 1592,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.0684,
+      "step": 1592,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 0.3064706307000024,
+      "lm_loss": 2.0347,
+      "step": 1592,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.3066631373775778,
+      "grad_norm": 3.252129372031478,
+      "learning_rate": 1.9212934839092196e-05,
+      "loss": 2.4038,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30685564405515314,
+      "grad_norm": 3.111933590591457,
+      "learning_rate": 1.9211721941093722e-05,
+      "loss": 2.4108,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30704815073272856,
+      "grad_norm": 2.8854291095017355,
+      "learning_rate": 1.9210508147598022e-05,
+      "loss": 2.4226,
+      "step": 1595
+    },
+    {
+      "epoch": 0.30724065741030393,
+      "grad_norm": 3.5117309962573633,
+      "learning_rate": 1.9209293458723093e-05,
+      "loss": 2.4552,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3074331640878793,
+      "grad_norm": 3.6809097376630766,
+      "learning_rate": 1.9208077874587017e-05,
+      "loss": 2.4058,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30762567076545466,
+      "grad_norm": 3.324244241560003,
+      "learning_rate": 1.9206861395307973e-05,
+      "loss": 2.4333,
+      "step": 1598
+    },
+    {
+      "epoch": 0.30781817744303,
+      "grad_norm": 3.6698747241911995,
+      "learning_rate": 1.9205644021004203e-05,
+      "loss": 2.4491,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "grad_norm": 3.0467367154505864,
+      "learning_rate": 1.9204425751794064e-05,
+      "loss": 2.367,
+      "step": 1600
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.3085,
+      "step": 1600,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.2527,
+      "step": 1600,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 1.8658,
+      "step": 1600,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.4182,
+      "step": 1600,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.3723,
+      "step": 1600,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.579,
+      "step": 1600,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.4558,
+      "step": 1600,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.30801068412060545,
+      "lm_loss": 2.2496,
+      "step": 1600,
+      "vm_loss": 0.2593
+    },
+    {
+      "epoch": 0.3082031907981808,
+      "grad_norm": 3.9531101788806238,
+      "learning_rate": 1.9203206587795982e-05,
+      "loss": 2.3899,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3083956974757562,
+      "grad_norm": 3.637420396432896,
+      "learning_rate": 1.9201986529128477e-05,
+      "loss": 2.3767,
+      "step": 1602
+    },
+    {
+      "epoch": 0.30858820415333155,
+      "grad_norm": 3.0524578861739937,
+      "learning_rate": 1.920076557591015e-05,
+      "loss": 2.3756,
+      "step": 1603
+    },
+    {
+      "epoch": 0.30878071083090697,
+      "grad_norm": 3.6819749716823393,
+      "learning_rate": 1.91995437282597e-05,
+      "loss": 2.4645,
+      "step": 1604
+    },
+    {
+      "epoch": 0.30897321750848233,
+      "grad_norm": 3.780474648887843,
+      "learning_rate": 1.9198320986295904e-05,
+      "loss": 2.4132,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3091657241860577,
+      "grad_norm": 3.8836029892814876,
+      "learning_rate": 1.9197097350137625e-05,
+      "loss": 2.4527,
+      "step": 1606
+    },
+    {
+      "epoch": 0.30935823086363307,
+      "grad_norm": 3.1565123554083008,
+      "learning_rate": 1.9195872819903817e-05,
+      "loss": 2.4009,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "grad_norm": 3.1515098808985527,
+      "learning_rate": 1.9194647395713525e-05,
+      "loss": 2.405,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.5084,
+      "step": 1608,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.2548,
+      "step": 1608,
+      "vm_loss": 0.2256
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 1.6081,
+      "step": 1608,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.4594,
+      "step": 1608,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.3055,
+      "step": 1608,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.0867,
+      "step": 1608,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.2439,
+      "step": 1608,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.3095507375412085,
+      "lm_loss": 2.2648,
+      "step": 1608,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.30974324421878385,
+      "grad_norm": 3.757860921384033,
+      "learning_rate": 1.919342107768587e-05,
+      "loss": 2.4035,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3099357508963592,
+      "grad_norm": 3.741681337777263,
+      "learning_rate": 1.919219386594007e-05,
+      "loss": 2.4419,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3101282575739346,
+      "grad_norm": 2.8799619899211075,
+      "learning_rate": 1.919096576059542e-05,
+      "loss": 2.408,
+      "step": 1611
+    },
+    {
+      "epoch": 0.31032076425150995,
+      "grad_norm": 3.5398337734615213,
+      "learning_rate": 1.9189736761771318e-05,
+      "loss": 2.4725,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31051327092908537,
+      "grad_norm": 3.7278409587860595,
+      "learning_rate": 1.9188506869587223e-05,
+      "loss": 2.4305,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31070577760666074,
+      "grad_norm": 2.931471786718931,
+      "learning_rate": 1.9187276084162714e-05,
+      "loss": 2.4303,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3108982842842361,
+      "grad_norm": 3.3217212956114266,
+      "learning_rate": 1.9186044405617423e-05,
+      "loss": 2.3905,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "grad_norm": 3.3395900084231633,
+      "learning_rate": 1.9184811834071097e-05,
+      "loss": 2.4765,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.3183,
+      "step": 1616,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.3912,
+      "step": 1616,
+      "vm_loss": 0.2341
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 1.9993,
+      "step": 1616,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.2254,
+      "step": 1616,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.1837,
+      "step": 1616,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.1318,
+      "step": 1616,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.0301,
+      "step": 1616,
+      "vm_loss": 0.2629
+    },
+    {
+      "epoch": 0.31109079096181147,
+      "lm_loss": 2.2213,
+      "step": 1616,
+      "vm_loss": 0.2409
+    },
+    {
+      "epoch": 0.3112832976393869,
+      "grad_norm": 2.857989138150822,
+      "learning_rate": 1.918357836964355e-05,
+      "loss": 2.3996,
+      "step": 1617
+    },
+    {
+      "epoch": 0.31147580431696226,
+      "grad_norm": 3.414176573864841,
+      "learning_rate": 1.918234401245469e-05,
+      "loss": 2.4288,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3116683109945376,
+      "grad_norm": 3.470435252438815,
+      "learning_rate": 1.9181108762624522e-05,
+      "loss": 2.4368,
+      "step": 1619
+    },
+    {
+      "epoch": 0.311860817672113,
+      "grad_norm": 3.3013239279231708,
+      "learning_rate": 1.9179872620273116e-05,
+      "loss": 2.4539,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31205332434968835,
+      "grad_norm": 2.8064524571948204,
+      "learning_rate": 1.9178635585520648e-05,
+      "loss": 2.404,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3122458310272638,
+      "grad_norm": 3.6435992475179266,
+      "learning_rate": 1.9177397658487376e-05,
+      "loss": 2.3827,
+      "step": 1622
+    },
+    {
+      "epoch": 0.31243833770483914,
+      "grad_norm": 2.9119566889862902,
+      "learning_rate": 1.9176158839293634e-05,
+      "loss": 2.3334,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "grad_norm": 3.013198522022594,
+      "learning_rate": 1.9174919128059856e-05,
+      "loss": 2.3938,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 1.9192,
+      "step": 1624,
+      "vm_loss": 0.0947
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.3788,
+      "step": 1624,
+      "vm_loss": 0.1457
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.0472,
+      "step": 1624,
+      "vm_loss": 0.2036
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.2758,
+      "step": 1624,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.1313,
+      "step": 1624,
+      "vm_loss": 0.2324
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.1901,
+      "step": 1624,
+      "vm_loss": 0.2184
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.0816,
+      "step": 1624,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 0.3126308443824145,
+      "lm_loss": 2.2704,
+      "step": 1624,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 0.3128233510599899,
+      "grad_norm": 3.2380900769520746,
+      "learning_rate": 1.9173678524906557e-05,
+      "loss": 2.3892,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3130158577375653,
+      "grad_norm": 2.982499010435214,
+      "learning_rate": 1.9172437029954342e-05,
+      "loss": 2.4341,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31320836441514066,
+      "grad_norm": 2.7673419695399404,
+      "learning_rate": 1.9171194643323896e-05,
+      "loss": 2.4548,
+      "step": 1627
+    },
+    {
+      "epoch": 0.313400871092716,
+      "grad_norm": 3.0025337500252505,
+      "learning_rate": 1.9169951365136e-05,
+      "loss": 2.4074,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3135933777702914,
+      "grad_norm": 3.0798793716423205,
+      "learning_rate": 1.9168707195511512e-05,
+      "loss": 2.4034,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31378588444786676,
+      "grad_norm": 2.995941162377776,
+      "learning_rate": 1.916746213457138e-05,
+      "loss": 2.4033,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3139783911254422,
+      "grad_norm": 2.9132283498345637,
+      "learning_rate": 1.9166216182436643e-05,
+      "loss": 2.4115,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "grad_norm": 2.758941529694054,
+      "learning_rate": 1.9164969339228424e-05,
+      "loss": 2.4151,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.4015,
+      "step": 1632,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.3777,
+      "step": 1632,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.0602,
+      "step": 1632,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.354,
+      "step": 1632,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.1065,
+      "step": 1632,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.1425,
+      "step": 1632,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.5505,
+      "step": 1632,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 0.31417089780301755,
+      "lm_loss": 2.1149,
+      "step": 1632,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.3143634044805929,
+      "grad_norm": 2.820036208327963,
+      "learning_rate": 1.916372160506793e-05,
+      "loss": 2.4471,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3145559111581683,
+      "grad_norm": 2.785639301197258,
+      "learning_rate": 1.916247298007646e-05,
+      "loss": 2.4152,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3147484178357437,
+      "grad_norm": 3.034776279384953,
+      "learning_rate": 1.9161223464375393e-05,
+      "loss": 2.4076,
+      "step": 1635
+    },
+    {
+      "epoch": 0.31494092451331906,
+      "grad_norm": 2.843864870634454,
+      "learning_rate": 1.9159973058086198e-05,
+      "loss": 2.398,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31513343119089443,
+      "grad_norm": 3.0798090836111034,
+      "learning_rate": 1.9158721761330432e-05,
+      "loss": 2.3938,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3153259378684698,
+      "grad_norm": 2.9122660704742147,
+      "learning_rate": 1.9157469574229736e-05,
+      "loss": 2.4125,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3155184445460452,
+      "grad_norm": 3.228181786720458,
+      "learning_rate": 1.915621649690584e-05,
+      "loss": 2.3899,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "grad_norm": 3.1463712301543514,
+      "learning_rate": 1.915496252948056e-05,
+      "loss": 2.4277,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.0584,
+      "step": 1640,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.3983,
+      "step": 1640,
+      "vm_loss": 0.214
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.1277,
+      "step": 1640,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.6039,
+      "step": 1640,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.4743,
+      "step": 1640,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.2705,
+      "step": 1640,
+      "vm_loss": 0.2481
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.0972,
+      "step": 1640,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 0.3157109512236206,
+      "lm_loss": 2.4121,
+      "step": 1640,
+      "vm_loss": 0.2644
+    },
+    {
+      "epoch": 0.31590345790119595,
+      "grad_norm": 2.874555616646559,
+      "learning_rate": 1.9153707672075793e-05,
+      "loss": 2.4827,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3160959645787713,
+      "grad_norm": 3.545184904839079,
+      "learning_rate": 1.9152451924813532e-05,
+      "loss": 2.3885,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3162884712563467,
+      "grad_norm": 3.224058729403147,
+      "learning_rate": 1.915119528781585e-05,
+      "loss": 2.4576,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3164809779339221,
+      "grad_norm": 3.136075283475528,
+      "learning_rate": 1.9149937761204907e-05,
+      "loss": 2.4034,
+      "step": 1644
+    },
+    {
+      "epoch": 0.31667348461149747,
+      "grad_norm": 2.9142663182231536,
+      "learning_rate": 1.9148679345102952e-05,
+      "loss": 2.4013,
+      "step": 1645
+    },
+    {
+      "epoch": 0.31686599128907283,
+      "grad_norm": 3.1876018007775753,
+      "learning_rate": 1.914742003963232e-05,
+      "loss": 2.4584,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3170584979666482,
+      "grad_norm": 3.419294524485164,
+      "learning_rate": 1.9146159844915427e-05,
+      "loss": 2.4318,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "grad_norm": 2.9730271290774373,
+      "learning_rate": 1.9144898761074785e-05,
+      "loss": 2.3417,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 1.842,
+      "step": 1648,
+      "vm_loss": 0.2541
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.4948,
+      "step": 1648,
+      "vm_loss": 0.1793
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.05,
+      "step": 1648,
+      "vm_loss": 0.2094
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.2142,
+      "step": 1648,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.013,
+      "step": 1648,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.0508,
+      "step": 1648,
+      "vm_loss": 0.228
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.2192,
+      "step": 1648,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 0.3172510046442236,
+      "lm_loss": 2.3756,
+      "step": 1648,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.317443511321799,
+      "grad_norm": 2.7837210529558276,
+      "learning_rate": 1.914363678823299e-05,
+      "loss": 2.4242,
+      "step": 1649
+    },
+    {
+      "epoch": 0.31763601799937435,
+      "grad_norm": 3.0424197234439223,
+      "learning_rate": 1.9142373926512713e-05,
+      "loss": 2.3841,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3178285246769497,
+      "grad_norm": 2.9039461576908883,
+      "learning_rate": 1.9141110176036726e-05,
+      "loss": 2.4165,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3180210313545251,
+      "grad_norm": 2.8248227439757554,
+      "learning_rate": 1.9139845536927882e-05,
+      "loss": 2.392,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3182135380321005,
+      "grad_norm": 2.8970420013193574,
+      "learning_rate": 1.9138580009309118e-05,
+      "loss": 2.4086,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3184060447096759,
+      "grad_norm": 3.1914263169991117,
+      "learning_rate": 1.9137313593303465e-05,
+      "loss": 2.4908,
+      "step": 1654
+    },
+    {
+      "epoch": 0.31859855138725124,
+      "grad_norm": 2.829729838111605,
+      "learning_rate": 1.9136046289034025e-05,
+      "loss": 2.3981,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "grad_norm": 2.9959100645433243,
+      "learning_rate": 1.9134778096624e-05,
+      "loss": 2.4189,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.175,
+      "step": 1656,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.4143,
+      "step": 1656,
+      "vm_loss": 0.2296
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.2002,
+      "step": 1656,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 1.8123,
+      "step": 1656,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.4234,
+      "step": 1656,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.0486,
+      "step": 1656,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.6399,
+      "step": 1656,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.3187910580648266,
+      "lm_loss": 2.4223,
+      "step": 1656,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 0.318983564742402,
+      "grad_norm": 3.190890267865329,
+      "learning_rate": 1.9133509016196683e-05,
+      "loss": 2.4414,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3191760714199774,
+      "grad_norm": 3.1852361419316733,
+      "learning_rate": 1.9132239047875433e-05,
+      "loss": 2.4034,
+      "step": 1658
+    },
+    {
+      "epoch": 0.31936857809755276,
+      "grad_norm": 3.1914000050905473,
+      "learning_rate": 1.9130968191783715e-05,
+      "loss": 2.441,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3195610847751281,
+      "grad_norm": 3.310642752320234,
+      "learning_rate": 1.9129696448045067e-05,
+      "loss": 2.4157,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3197535914527035,
+      "grad_norm": 2.817753501061679,
+      "learning_rate": 1.912842381678312e-05,
+      "loss": 2.4377,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3199460981302789,
+      "grad_norm": 3.1035160019109282,
+      "learning_rate": 1.9127150298121596e-05,
+      "loss": 2.4038,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3201386048078543,
+      "grad_norm": 3.122888134091364,
+      "learning_rate": 1.912587589218429e-05,
+      "loss": 2.4088,
+      "step": 1663
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "grad_norm": 3.019935114378441,
+      "learning_rate": 1.9124600599095094e-05,
+      "loss": 2.3971,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.081,
+      "step": 1664,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.3194,
+      "step": 1664,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.2441,
+      "step": 1664,
+      "vm_loss": 0.2384
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.1281,
+      "step": 1664,
+      "vm_loss": 0.2268
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.3737,
+      "step": 1664,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 1.9669,
+      "step": 1664,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.0802,
+      "step": 1664,
+      "vm_loss": 0.2034
+    },
+    {
+      "epoch": 0.32033111148542964,
+      "lm_loss": 2.6208,
+      "step": 1664,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.320523618163005,
+      "grad_norm": 3.0228353955625904,
+      "learning_rate": 1.9123324418977982e-05,
+      "loss": 2.4683,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32071612484058043,
+      "grad_norm": 3.0205956839577044,
+      "learning_rate": 1.9122047351957012e-05,
+      "loss": 2.3904,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3209086315181558,
+      "grad_norm": 3.2464196347789547,
+      "learning_rate": 1.9120769398156338e-05,
+      "loss": 2.3831,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32110113819573116,
+      "grad_norm": 2.7621650225226544,
+      "learning_rate": 1.9119490557700192e-05,
+      "loss": 2.3747,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3212936448733065,
+      "grad_norm": 2.898171373720304,
+      "learning_rate": 1.9118210830712886e-05,
+      "loss": 2.4245,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32148615155088195,
+      "grad_norm": 3.2810832839521313,
+      "learning_rate": 1.9116930217318834e-05,
+      "loss": 2.4284,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3216786582284573,
+      "grad_norm": 2.6934133638520765,
+      "learning_rate": 1.9115648717642527e-05,
+      "loss": 2.4347,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "grad_norm": 3.0350892141694956,
+      "learning_rate": 1.911436633180854e-05,
+      "loss": 2.4417,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.4947,
+      "step": 1672,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 1.78,
+      "step": 1672,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.3463,
+      "step": 1672,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.0045,
+      "step": 1672,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.379,
+      "step": 1672,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.1747,
+      "step": 1672,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.2924,
+      "step": 1672,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.3218711649060327,
+      "lm_loss": 2.1637,
+      "step": 1672,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 0.32206367158360805,
+      "grad_norm": 3.2101618268595367,
+      "learning_rate": 1.911308305994154e-05,
+      "loss": 2.4393,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3222561782611834,
+      "grad_norm": 2.5739350618931116,
+      "learning_rate": 1.911179890216627e-05,
+      "loss": 2.4652,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32244868493875883,
+      "grad_norm": 3.0075596556350295,
+      "learning_rate": 1.9110513858607576e-05,
+      "loss": 2.4356,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3226411916163342,
+      "grad_norm": 2.862037629948916,
+      "learning_rate": 1.9109227929390378e-05,
+      "loss": 2.4667,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32283369829390957,
+      "grad_norm": 2.776365763194087,
+      "learning_rate": 1.9107941114639685e-05,
+      "loss": 2.413,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32302620497148493,
+      "grad_norm": 2.988130151602014,
+      "learning_rate": 1.9106653414480593e-05,
+      "loss": 2.3915,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32321871164906035,
+      "grad_norm": 3.2591702958630417,
+      "learning_rate": 1.9105364829038278e-05,
+      "loss": 2.4165,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "grad_norm": 2.794659666012045,
+      "learning_rate": 1.910407535843801e-05,
+      "loss": 2.4127,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.2186,
+      "step": 1680,
+      "vm_loss": 0.265
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.1581,
+      "step": 1680,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.2775,
+      "step": 1680,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.3566,
+      "step": 1680,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 1.9381,
+      "step": 1680,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.157,
+      "step": 1680,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.1727,
+      "step": 1680,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 0.3234112183266357,
+      "lm_loss": 2.0669,
+      "step": 1680,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.3236037250042111,
+      "grad_norm": 3.073408880660531,
+      "learning_rate": 1.9102785002805145e-05,
+      "loss": 2.3909,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32379623168178645,
+      "grad_norm": 2.9220658268757975,
+      "learning_rate": 1.9101493762265116e-05,
+      "loss": 2.414,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3239887383593618,
+      "grad_norm": 2.9992045034805628,
+      "learning_rate": 1.9100201636943453e-05,
+      "loss": 2.38,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32418124503693724,
+      "grad_norm": 2.9880082163340265,
+      "learning_rate": 1.909890862696576e-05,
+      "loss": 2.4029,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3243737517145126,
+      "grad_norm": 3.0729355375353853,
+      "learning_rate": 1.9097614732457745e-05,
+      "loss": 2.4072,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32456625839208797,
+      "grad_norm": 3.068605934741322,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 2.4419,
+      "step": 1686
+    },
+    {
+      "epoch": 0.32475876506966334,
+      "grad_norm": 3.3148660484575436,
+      "learning_rate": 1.9095024290353948e-05,
+      "loss": 2.4111,
+      "step": 1687
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "grad_norm": 3.229534212884842,
+      "learning_rate": 1.9093727743009993e-05,
+      "loss": 2.4089,
+      "step": 1688
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.3924,
+      "step": 1688,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.2844,
+      "step": 1688,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.3729,
+      "step": 1688,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.2958,
+      "step": 1688,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.3885,
+      "step": 1688,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.1698,
+      "step": 1688,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.3188,
+      "step": 1688,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.32495127174723876,
+      "lm_loss": 2.1556,
+      "step": 1688,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.3251437784248141,
+      "grad_norm": 3.014794639758744,
+      "learning_rate": 1.9092430311639355e-05,
+      "loss": 2.4353,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3253362851023895,
+      "grad_norm": 3.070224386406534,
+      "learning_rate": 1.909113199636817e-05,
+      "loss": 2.4172,
+      "step": 1690
+    },
+    {
+      "epoch": 0.32552879177996485,
+      "grad_norm": 2.832930561833768,
+      "learning_rate": 1.908983279732264e-05,
+      "loss": 2.4414,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3257212984575402,
+      "grad_norm": 3.228483190648978,
+      "learning_rate": 1.908853271462907e-05,
+      "loss": 2.4296,
+      "step": 1692
+    },
+    {
+      "epoch": 0.32591380513511564,
+      "grad_norm": 3.259311739741599,
+      "learning_rate": 1.9087231748413847e-05,
+      "loss": 2.3699,
+      "step": 1693
+    },
+    {
+      "epoch": 0.326106311812691,
+      "grad_norm": 2.6838979497458655,
+      "learning_rate": 1.9085929898803437e-05,
+      "loss": 2.4361,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3262988184902664,
+      "grad_norm": 3.615176605280966,
+      "learning_rate": 1.9084627165924397e-05,
+      "loss": 2.4544,
+      "step": 1695
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "grad_norm": 2.963390452799069,
+      "learning_rate": 1.908332354990337e-05,
+      "loss": 2.4173,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.2288,
+      "step": 1696,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.2827,
+      "step": 1696,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.5743,
+      "step": 1696,
+      "vm_loss": 0.1049
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.0604,
+      "step": 1696,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.3513,
+      "step": 1696,
+      "vm_loss": 0.2499
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.0254,
+      "step": 1696,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.0627,
+      "step": 1696,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 0.32649132516784174,
+      "lm_loss": 2.0476,
+      "step": 1696,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 0.32668383184541716,
+      "grad_norm": 3.059725368462857,
+      "learning_rate": 1.9082019050867086e-05,
+      "loss": 2.3985,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3268763385229925,
+      "grad_norm": 3.0129163406099533,
+      "learning_rate": 1.9080713668942356e-05,
+      "loss": 2.4074,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3270688452005679,
+      "grad_norm": 3.1096098924196807,
+      "learning_rate": 1.907940740425608e-05,
+      "loss": 2.4537,
+      "step": 1699
+    },
+    {
+      "epoch": 0.32726135187814326,
+      "grad_norm": 2.890239355800731,
+      "learning_rate": 1.9078100256935247e-05,
+      "loss": 2.424,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3274538585557186,
+      "grad_norm": 3.170518539649865,
+      "learning_rate": 1.9076792227106926e-05,
+      "loss": 2.3731,
+      "step": 1701
+    },
+    {
+      "epoch": 0.32764636523329405,
+      "grad_norm": 3.4547990948980143,
+      "learning_rate": 1.9075483314898275e-05,
+      "loss": 2.4053,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3278388719108694,
+      "grad_norm": 2.568403546763582,
+      "learning_rate": 1.9074173520436535e-05,
+      "loss": 2.3553,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "grad_norm": 2.852690727468107,
+      "learning_rate": 1.9072862843849036e-05,
+      "loss": 2.4605,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.1918,
+      "step": 1704,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.2387,
+      "step": 1704,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.3272,
+      "step": 1704,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.2215,
+      "step": 1704,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.3218,
+      "step": 1704,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.3165,
+      "step": 1704,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.4307,
+      "step": 1704,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.3280313785884448,
+      "lm_loss": 2.1909,
+      "step": 1704,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.32822388526602014,
+      "grad_norm": 3.05549450215224,
+      "learning_rate": 1.9071551285263195e-05,
+      "loss": 2.4405,
+      "step": 1705
+    },
+    {
+      "epoch": 0.32841639194359556,
+      "grad_norm": 3.1269760108753033,
+      "learning_rate": 1.9070238844806507e-05,
+      "loss": 2.3716,
+      "step": 1706
+    },
+    {
+      "epoch": 0.32860889862117093,
+      "grad_norm": 3.097977814972827,
+      "learning_rate": 1.906892552260656e-05,
+      "loss": 2.4153,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3288014052987463,
+      "grad_norm": 3.2565685454359348,
+      "learning_rate": 1.906761131879103e-05,
+      "loss": 2.4378,
+      "step": 1708
+    },
+    {
+      "epoch": 0.32899391197632166,
+      "grad_norm": 3.5405687730250266,
+      "learning_rate": 1.906629623348767e-05,
+      "loss": 2.4818,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3291864186538971,
+      "grad_norm": 2.9302540013991396,
+      "learning_rate": 1.9064980266824323e-05,
+      "loss": 2.446,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32937892533147245,
+      "grad_norm": 3.2652875543626037,
+      "learning_rate": 1.9063663418928923e-05,
+      "loss": 2.4593,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "grad_norm": 3.460715665643919,
+      "learning_rate": 1.906234568992948e-05,
+      "loss": 2.4212,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.3905,
+      "step": 1712,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.4512,
+      "step": 1712,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.0487,
+      "step": 1712,
+      "vm_loss": 0.238
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.1492,
+      "step": 1712,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.2014,
+      "step": 1712,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.4585,
+      "step": 1712,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.2054,
+      "step": 1712,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 0.3295714320090478,
+      "lm_loss": 2.2655,
+      "step": 1712,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.3297639386866232,
+      "grad_norm": 3.1561803387048624,
+      "learning_rate": 1.9061027079954093e-05,
+      "loss": 2.4079,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32995644536419855,
+      "grad_norm": 4.209720800062066,
+      "learning_rate": 1.9059707589130946e-05,
+      "loss": 2.4078,
+      "step": 1714
+    },
+    {
+      "epoch": 0.33014895204177397,
+      "grad_norm": 3.0566142302859673,
+      "learning_rate": 1.9058387217588322e-05,
+      "loss": 2.3901,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33034145871934933,
+      "grad_norm": 3.8504528922520103,
+      "learning_rate": 1.9057065965454562e-05,
+      "loss": 2.3737,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3305339653969247,
+      "grad_norm": 3.5168665645160204,
+      "learning_rate": 1.905574383285812e-05,
+      "loss": 2.3657,
+      "step": 1717
+    },
+    {
+      "epoch": 0.33072647207450007,
+      "grad_norm": 3.154826785041629,
+      "learning_rate": 1.905442081992752e-05,
+      "loss": 2.4031,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3309189787520755,
+      "grad_norm": 3.2774373440695252,
+      "learning_rate": 1.9053096926791377e-05,
+      "loss": 2.5024,
+      "step": 1719
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "grad_norm": 3.676863375596527,
+      "learning_rate": 1.905177215357839e-05,
+      "loss": 2.4191,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.5633,
+      "step": 1720,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.5542,
+      "step": 1720,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 1.9477,
+      "step": 1720,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.1229,
+      "step": 1720,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 1.9846,
+      "step": 1720,
+      "vm_loss": 0.1195
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.0449,
+      "step": 1720,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.0597,
+      "step": 1720,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 0.33111148542965085,
+      "lm_loss": 2.4556,
+      "step": 1720,
+      "vm_loss": 0.09
+    },
+    {
+      "epoch": 0.3313039921072262,
+      "grad_norm": 3.4302961132949092,
+      "learning_rate": 1.905044650041734e-05,
+      "loss": 2.3374,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3314964987848016,
+      "grad_norm": 2.9757935556438113,
+      "learning_rate": 1.9049119967437108e-05,
+      "loss": 2.388,
+      "step": 1722
+    },
+    {
+      "epoch": 0.33168900546237695,
+      "grad_norm": 3.131951025886107,
+      "learning_rate": 1.9047792554766637e-05,
+      "loss": 2.3646,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3318815121399524,
+      "grad_norm": 3.1522615196499633,
+      "learning_rate": 1.9046464262534978e-05,
+      "loss": 2.4163,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33207401881752774,
+      "grad_norm": 3.205922169167952,
+      "learning_rate": 1.9045135090871254e-05,
+      "loss": 2.3618,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3322665254951031,
+      "grad_norm": 3.5196443172186838,
+      "learning_rate": 1.9043805039904677e-05,
+      "loss": 2.4858,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33245903217267847,
+      "grad_norm": 2.987227039786243,
+      "learning_rate": 1.9042474109764545e-05,
+      "loss": 2.3571,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "grad_norm": 3.3088066481019505,
+      "learning_rate": 1.9041142300580247e-05,
+      "loss": 2.3551,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.2981,
+      "step": 1728,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.4517,
+      "step": 1728,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.0383,
+      "step": 1728,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 1.8719,
+      "step": 1728,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.1821,
+      "step": 1728,
+      "vm_loss": 0.2171
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.31,
+      "step": 1728,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 2.3241,
+      "step": 1728,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 0.3326515388502539,
+      "lm_loss": 1.9967,
+      "step": 1728,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.33284404552782926,
+      "grad_norm": 3.088237616719538,
+      "learning_rate": 1.9039809612481245e-05,
+      "loss": 2.3887,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3330365522054046,
+      "grad_norm": 3.119223866435456,
+      "learning_rate": 1.9038476045597093e-05,
+      "loss": 2.4067,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33322905888298,
+      "grad_norm": 2.6996055623863056,
+      "learning_rate": 1.903714160005744e-05,
+      "loss": 2.3805,
+      "step": 1731
+    },
+    {
+      "epoch": 0.33342156556055536,
+      "grad_norm": 3.385718510407307,
+      "learning_rate": 1.9035806275991996e-05,
+      "loss": 2.3792,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3336140722381308,
+      "grad_norm": 2.7826967696854075,
+      "learning_rate": 1.9034470073530588e-05,
+      "loss": 2.3799,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33380657891570614,
+      "grad_norm": 2.890236769159508,
+      "learning_rate": 1.90331329928031e-05,
+      "loss": 2.4318,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3339990855932815,
+      "grad_norm": 2.906373585337404,
+      "learning_rate": 1.903179503393952e-05,
+      "loss": 2.3809,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "grad_norm": 3.358284100078741,
+      "learning_rate": 1.903045619706991e-05,
+      "loss": 2.4329,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.1365,
+      "step": 1736,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 1.9399,
+      "step": 1736,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 1.9106,
+      "step": 1736,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.3076,
+      "step": 1736,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.2196,
+      "step": 1736,
+      "vm_loss": 0.2207
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.174,
+      "step": 1736,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.4505,
+      "step": 1736,
+      "vm_loss": 0.2018
+    },
+    {
+      "epoch": 0.3341915922708569,
+      "lm_loss": 2.2559,
+      "step": 1736,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.3343840989484323,
+      "grad_norm": 3.0287045228388507,
+      "learning_rate": 1.9029116482324427e-05,
+      "loss": 2.3799,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33457660562600766,
+      "grad_norm": 3.146579728110227,
+      "learning_rate": 1.90277758898333e-05,
+      "loss": 2.4323,
+      "step": 1738
+    },
+    {
+      "epoch": 0.334769112303583,
+      "grad_norm": 3.300009059371359,
+      "learning_rate": 1.9026434419726866e-05,
+      "loss": 2.3945,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3349616189811584,
+      "grad_norm": 3.2676158409395524,
+      "learning_rate": 1.9025092072135514e-05,
+      "loss": 2.4328,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3351541256587338,
+      "grad_norm": 3.602263620536653,
+      "learning_rate": 1.9023748847189757e-05,
+      "loss": 2.3842,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3353466323363092,
+      "grad_norm": 3.0877383327328305,
+      "learning_rate": 1.9022404745020165e-05,
+      "loss": 2.3776,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33553913901388455,
+      "grad_norm": 3.7428783919632225,
+      "learning_rate": 1.9021059765757396e-05,
+      "loss": 2.4108,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "grad_norm": 2.7777833462674875,
+      "learning_rate": 1.9019713909532207e-05,
+      "loss": 2.3975,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.1749,
+      "step": 1744,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.2814,
+      "step": 1744,
+      "vm_loss": 0.2063
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 1.8623,
+      "step": 1744,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.4433,
+      "step": 1744,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.3347,
+      "step": 1744,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.12,
+      "step": 1744,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.6005,
+      "step": 1744,
+      "vm_loss": 0.2413
+    },
+    {
+      "epoch": 0.3357316456914599,
+      "lm_loss": 2.1044,
+      "step": 1744,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.3359241523690353,
+      "grad_norm": 2.889883464809344,
+      "learning_rate": 1.901836717647543e-05,
+      "loss": 2.4433,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3361166590466107,
+      "grad_norm": 3.50359113674953,
+      "learning_rate": 1.9017019566717984e-05,
+      "loss": 2.3941,
+      "step": 1746
+    },
+    {
+      "epoch": 0.33630916572418607,
+      "grad_norm": 3.4836649085891516,
+      "learning_rate": 1.9015671080390876e-05,
+      "loss": 2.4313,
+      "step": 1747
+    },
+    {
+      "epoch": 0.33650167240176143,
+      "grad_norm": 2.6283468362876437,
+      "learning_rate": 1.9014321717625198e-05,
+      "loss": 2.4079,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3366941790793368,
+      "grad_norm": 2.9780936113384917,
+      "learning_rate": 1.9012971478552117e-05,
+      "loss": 2.4014,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3368866857569122,
+      "grad_norm": 3.3658841309946586,
+      "learning_rate": 1.9011620363302898e-05,
+      "loss": 2.4677,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3370791924344876,
+      "grad_norm": 3.1426882548979176,
+      "learning_rate": 1.901026837200889e-05,
+      "loss": 2.388,
+      "step": 1751
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "grad_norm": 2.594427955619498,
+      "learning_rate": 1.9008915504801524e-05,
+      "loss": 2.3904,
+      "step": 1752
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.0674,
+      "step": 1752,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.2917,
+      "step": 1752,
+      "vm_loss": 0.2054
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.0626,
+      "step": 1752,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.2077,
+      "step": 1752,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.3863,
+      "step": 1752,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.0476,
+      "step": 1752,
+      "vm_loss": 0.0952
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.317,
+      "step": 1752,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 0.33727169911206295,
+      "lm_loss": 2.4646,
+      "step": 1752,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 0.3374642057896383,
+      "grad_norm": 3.528213143762084,
+      "learning_rate": 1.900756176181231e-05,
+      "loss": 2.4266,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3376567124672137,
+      "grad_norm": 3.2015170544565463,
+      "learning_rate": 1.900620714317285e-05,
+      "loss": 2.3885,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3378492191447891,
+      "grad_norm": 2.8546790721533655,
+      "learning_rate": 1.900485164901484e-05,
+      "loss": 2.3852,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33804172582236447,
+      "grad_norm": 3.589044908522458,
+      "learning_rate": 1.9003495279470035e-05,
+      "loss": 2.4071,
+      "step": 1756
+    },
+    {
+      "epoch": 0.33823423249993984,
+      "grad_norm": 3.2932653210014444,
+      "learning_rate": 1.9002138034670308e-05,
+      "loss": 2.4403,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3384267391775152,
+      "grad_norm": 3.0482781234838994,
+      "learning_rate": 1.900077991474759e-05,
+      "loss": 2.4359,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3386192458550906,
+      "grad_norm": 3.3508684315845363,
+      "learning_rate": 1.8999420919833912e-05,
+      "loss": 2.3675,
+      "step": 1759
+    },
+    {
+      "epoch": 0.338811752532666,
+      "grad_norm": 3.3243098362628505,
+      "learning_rate": 1.899806105006138e-05,
+      "loss": 2.4104,
+      "step": 1760
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.2646,
+      "step": 1760,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.2976,
+      "step": 1760,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.1831,
+      "step": 1760,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 1.9797,
+      "step": 1760,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.4605,
+      "step": 1760,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.4575,
+      "step": 1760,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.3641,
+      "step": 1760,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 0.338811752532666,
+      "lm_loss": 2.2425,
+      "step": 1760,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.33900425921024135,
+      "grad_norm": 3.642381160922666,
+      "learning_rate": 1.89967003055622e-05,
+      "loss": 2.3777,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3391967658878167,
+      "grad_norm": 2.8564030083501253,
+      "learning_rate": 1.8995338686468648e-05,
+      "loss": 2.3715,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3393892725653921,
+      "grad_norm": 3.233731017390185,
+      "learning_rate": 1.899397619291309e-05,
+      "loss": 2.3613,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3395817792429675,
+      "grad_norm": 2.982087407599961,
+      "learning_rate": 1.8992612825027978e-05,
+      "loss": 2.394,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3397742859205429,
+      "grad_norm": 3.3766804340379717,
+      "learning_rate": 1.8991248582945855e-05,
+      "loss": 2.4355,
+      "step": 1765
+    },
+    {
+      "epoch": 0.33996679259811824,
+      "grad_norm": 3.0277486123598254,
+      "learning_rate": 1.8989883466799335e-05,
+      "loss": 2.3914,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3401592992756936,
+      "grad_norm": 3.3329987674067856,
+      "learning_rate": 1.898851747672113e-05,
+      "loss": 2.3859,
+      "step": 1767
+    },
+    {
+      "epoch": 0.340351805953269,
+      "grad_norm": 3.249698568758816,
+      "learning_rate": 1.8987150612844028e-05,
+      "loss": 2.4515,
+      "step": 1768
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.0811,
+      "step": 1768,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.4001,
+      "step": 1768,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.3949,
+      "step": 1768,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.4494,
+      "step": 1768,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.3516,
+      "step": 1768,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.0027,
+      "step": 1768,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.2244,
+      "step": 1768,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.340351805953269,
+      "lm_loss": 2.3771,
+      "step": 1768,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.3405443126308444,
+      "grad_norm": 3.468558048038586,
+      "learning_rate": 1.8985782875300908e-05,
+      "loss": 2.3769,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34073681930841976,
+      "grad_norm": 3.0027107780383906,
+      "learning_rate": 1.8984414264224727e-05,
+      "loss": 2.459,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3409293259859951,
+      "grad_norm": 3.0096247219000176,
+      "learning_rate": 1.898304477974854e-05,
+      "loss": 2.383,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3411218326635705,
+      "grad_norm": 3.6875624244607264,
+      "learning_rate": 1.8981674422005472e-05,
+      "loss": 2.398,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3413143393411459,
+      "grad_norm": 3.3956076388785212,
+      "learning_rate": 1.898030319112874e-05,
+      "loss": 2.4397,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3415068460187213,
+      "grad_norm": 3.277790075803118,
+      "learning_rate": 1.8978931087251648e-05,
+      "loss": 2.4013,
+      "step": 1774
+    },
+    {
+      "epoch": 0.34169935269629664,
+      "grad_norm": 3.2445901587391934,
+      "learning_rate": 1.897755811050758e-05,
+      "loss": 2.4201,
+      "step": 1775
+    },
+    {
+      "epoch": 0.341891859373872,
+      "grad_norm": 3.4711477530604067,
+      "learning_rate": 1.8976184261030007e-05,
+      "loss": 2.4039,
+      "step": 1776
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.5154,
+      "step": 1776,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 1.9922,
+      "step": 1776,
+      "vm_loss": 0.201
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.6143,
+      "step": 1776,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.3017,
+      "step": 1776,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.2973,
+      "step": 1776,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.3998,
+      "step": 1776,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.0749,
+      "step": 1776,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 0.341891859373872,
+      "lm_loss": 2.3958,
+      "step": 1776,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.34208436605144743,
+      "grad_norm": 3.2357586471960693,
+      "learning_rate": 1.8974809538952483e-05,
+      "loss": 2.4398,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3422768727290228,
+      "grad_norm": 3.531348848539006,
+      "learning_rate": 1.897343394440865e-05,
+      "loss": 2.3907,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34246937940659816,
+      "grad_norm": 3.0828718583120547,
+      "learning_rate": 1.897205747753224e-05,
+      "loss": 2.3617,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34266188608417353,
+      "grad_norm": 3.0926075152284924,
+      "learning_rate": 1.8970680138457052e-05,
+      "loss": 2.383,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34285439276174895,
+      "grad_norm": 3.2922114486366545,
+      "learning_rate": 1.8969301927316984e-05,
+      "loss": 2.4189,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3430468994393243,
+      "grad_norm": 3.4601428580471794,
+      "learning_rate": 1.8967922844246025e-05,
+      "loss": 2.4457,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3432394061168997,
+      "grad_norm": 2.540788546760732,
+      "learning_rate": 1.8966542889378225e-05,
+      "loss": 2.4001,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "grad_norm": 3.1119921856114297,
+      "learning_rate": 1.8965162062847743e-05,
+      "loss": 2.4513,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.0376,
+      "step": 1784,
+      "vm_loss": 0.2298
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.2602,
+      "step": 1784,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.2952,
+      "step": 1784,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.4144,
+      "step": 1784,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.3366,
+      "step": 1784,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.3469,
+      "step": 1784,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.2138,
+      "step": 1784,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 0.34343191279447505,
+      "lm_loss": 2.1255,
+      "step": 1784,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.3436244194720504,
+      "grad_norm": 3.356424725142414,
+      "learning_rate": 1.8963780364788815e-05,
+      "loss": 2.3973,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34381692614962583,
+      "grad_norm": 2.8489180512515246,
+      "learning_rate": 1.896239779533575e-05,
+      "loss": 2.384,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3440094328272012,
+      "grad_norm": 3.487823571010014,
+      "learning_rate": 1.896101435462296e-05,
+      "loss": 2.4175,
+      "step": 1787
+    },
+    {
+      "epoch": 0.34420193950477657,
+      "grad_norm": 3.362377925537072,
+      "learning_rate": 1.8959630042784928e-05,
+      "loss": 2.3878,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34439444618235193,
+      "grad_norm": 3.371671022224365,
+      "learning_rate": 1.895824485995623e-05,
+      "loss": 2.423,
+      "step": 1789
+    },
+    {
+      "epoch": 0.34458695285992735,
+      "grad_norm": 3.5484698135490906,
+      "learning_rate": 1.895685880627152e-05,
+      "loss": 2.4268,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3447794595375027,
+      "grad_norm": 3.1922204183491356,
+      "learning_rate": 1.895547188186554e-05,
+      "loss": 2.4092,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "grad_norm": 3.7753734759492965,
+      "learning_rate": 1.8954084086873124e-05,
+      "loss": 2.4394,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.4267,
+      "step": 1792,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.2821,
+      "step": 1792,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.3579,
+      "step": 1792,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.2514,
+      "step": 1792,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.3238,
+      "step": 1792,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.251,
+      "step": 1792,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.344,
+      "step": 1792,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 0.3449719662150781,
+      "lm_loss": 2.3249,
+      "step": 1792,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.34516447289265345,
+      "grad_norm": 3.032929745047781,
+      "learning_rate": 1.8952695421429176e-05,
+      "loss": 2.4199,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3453569795702288,
+      "grad_norm": 3.9556924058657357,
+      "learning_rate": 1.8951305885668695e-05,
+      "loss": 2.3523,
+      "step": 1794
+    },
+    {
+      "epoch": 0.34554948624780424,
+      "grad_norm": 3.3988968425269697,
+      "learning_rate": 1.8949915479726763e-05,
+      "loss": 2.3717,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3457419929253796,
+      "grad_norm": 3.0620316604197964,
+      "learning_rate": 1.8948524203738545e-05,
+      "loss": 2.4092,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34593449960295497,
+      "grad_norm": 3.847200885087783,
+      "learning_rate": 1.8947132057839287e-05,
+      "loss": 2.3864,
+      "step": 1797
+    },
+    {
+      "epoch": 0.34612700628053034,
+      "grad_norm": 3.5311669754787562,
+      "learning_rate": 1.8945739042164325e-05,
+      "loss": 2.4425,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34631951295810576,
+      "grad_norm": 3.706395456300257,
+      "learning_rate": 1.894434515684908e-05,
+      "loss": 2.3992,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "grad_norm": 3.172605360763211,
+      "learning_rate": 1.894295040202905e-05,
+      "loss": 2.3147,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.0065,
+      "step": 1800,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.1158,
+      "step": 1800,
+      "vm_loss": 0.2535
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.2855,
+      "step": 1800,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.0779,
+      "step": 1800,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.3025,
+      "step": 1800,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.0014,
+      "step": 1800,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.1758,
+      "step": 1800,
+      "vm_loss": 0.2182
+    },
+    {
+      "epoch": 0.3465120196356811,
+      "lm_loss": 2.1542,
+      "step": 1800,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.3467045263132565,
+      "grad_norm": 2.9197881276349613,
+      "learning_rate": 1.894155477783983e-05,
+      "loss": 2.3709,
+      "step": 1801
+    },
+    {
+      "epoch": 0.34689703299083186,
+      "grad_norm": 3.4664267219060725,
+      "learning_rate": 1.8940158284417093e-05,
+      "loss": 2.4017,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3470895396684072,
+      "grad_norm": 3.2253861471370797,
+      "learning_rate": 1.8938760921896592e-05,
+      "loss": 2.3781,
+      "step": 1803
+    },
+    {
+      "epoch": 0.34728204634598264,
+      "grad_norm": 2.875584731666018,
+      "learning_rate": 1.8937362690414165e-05,
+      "loss": 2.4179,
+      "step": 1804
+    },
+    {
+      "epoch": 0.347474553023558,
+      "grad_norm": 3.0420742296641143,
+      "learning_rate": 1.8935963590105745e-05,
+      "loss": 2.3615,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3476670597011334,
+      "grad_norm": 3.3754393151986926,
+      "learning_rate": 1.893456362110734e-05,
+      "loss": 2.4128,
+      "step": 1806
+    },
+    {
+      "epoch": 0.34785956637870874,
+      "grad_norm": 3.080489929696284,
+      "learning_rate": 1.8933162783555044e-05,
+      "loss": 2.4182,
+      "step": 1807
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "grad_norm": 3.151003955574368,
+      "learning_rate": 1.8931761077585037e-05,
+      "loss": 2.3303,
+      "step": 1808
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.422,
+      "step": 1808,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.2289,
+      "step": 1808,
+      "vm_loss": 0.2409
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.4526,
+      "step": 1808,
+      "vm_loss": 0.1422
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.0436,
+      "step": 1808,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.2206,
+      "step": 1808,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.3812,
+      "step": 1808,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.4423,
+      "step": 1808,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.34805207305628416,
+      "lm_loss": 2.4581,
+      "step": 1808,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 0.3482445797338595,
+      "grad_norm": 3.316248251295914,
+      "learning_rate": 1.8930358503333584e-05,
+      "loss": 2.4207,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3484370864114349,
+      "grad_norm": 3.4721530021081173,
+      "learning_rate": 1.892895506093703e-05,
+      "loss": 2.3729,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34862959308901026,
+      "grad_norm": 3.141333593403813,
+      "learning_rate": 1.892755075053181e-05,
+      "loss": 2.372,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3488220997665857,
+      "grad_norm": 2.937787306669567,
+      "learning_rate": 1.8926145572254437e-05,
+      "loss": 2.3917,
+      "step": 1812
+    },
+    {
+      "epoch": 0.34901460644416105,
+      "grad_norm": 3.3638315082157693,
+      "learning_rate": 1.892473952624152e-05,
+      "loss": 2.4192,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3492071131217364,
+      "grad_norm": 3.254866099193651,
+      "learning_rate": 1.892333261262974e-05,
+      "loss": 2.3966,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3493996197993118,
+      "grad_norm": 3.0544369888150924,
+      "learning_rate": 1.8921924831555865e-05,
+      "loss": 2.3895,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "grad_norm": 2.860844842503817,
+      "learning_rate": 1.8920516183156753e-05,
+      "loss": 2.3796,
+      "step": 1816
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.3789,
+      "step": 1816,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.4993,
+      "step": 1816,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.1931,
+      "step": 1816,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.212,
+      "step": 1816,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.19,
+      "step": 1816,
+      "vm_loss": 0.2078
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.533,
+      "step": 1816,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.1039,
+      "step": 1816,
+      "vm_loss": 0.1133
+    },
+    {
+      "epoch": 0.34959212647688714,
+      "lm_loss": 2.2825,
+      "step": 1816,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 0.34978463315446257,
+      "grad_norm": 2.6868805256373354,
+      "learning_rate": 1.891910666756934e-05,
+      "loss": 2.3841,
+      "step": 1817
+    },
+    {
+      "epoch": 0.34997713983203793,
+      "grad_norm": 2.9523439883969744,
+      "learning_rate": 1.8917696284930654e-05,
+      "loss": 2.3594,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3501696465096133,
+      "grad_norm": 3.1216210261934156,
+      "learning_rate": 1.8916285035377794e-05,
+      "loss": 2.4226,
+      "step": 1819
+    },
+    {
+      "epoch": 0.35036215318718866,
+      "grad_norm": 2.971388675922259,
+      "learning_rate": 1.8914872919047957e-05,
+      "loss": 2.3859,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3505546598647641,
+      "grad_norm": 2.588473122039049,
+      "learning_rate": 1.8913459936078416e-05,
+      "loss": 2.4038,
+      "step": 1821
+    },
+    {
+      "epoch": 0.35074716654233945,
+      "grad_norm": 2.8638933386115832,
+      "learning_rate": 1.8912046086606537e-05,
+      "loss": 2.4036,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3509396732199148,
+      "grad_norm": 3.2413608139304357,
+      "learning_rate": 1.891063137076976e-05,
+      "loss": 2.348,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "grad_norm": 2.9059829753499127,
+      "learning_rate": 1.890921578870561e-05,
+      "loss": 2.4431,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.1095,
+      "step": 1824,
+      "vm_loss": 0.11
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.2551,
+      "step": 1824,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.468,
+      "step": 1824,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 1.9787,
+      "step": 1824,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.1936,
+      "step": 1824,
+      "vm_loss": 0.2466
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.2338,
+      "step": 1824,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.3164,
+      "step": 1824,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 0.3511321798974902,
+      "lm_loss": 2.2622,
+      "step": 1824,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.35132468657506555,
+      "grad_norm": 2.9135405139075585,
+      "learning_rate": 1.8907799340551707e-05,
+      "loss": 2.3874,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35151719325264097,
+      "grad_norm": 2.8260875008078696,
+      "learning_rate": 1.890638202644574e-05,
+      "loss": 2.4063,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35170969993021634,
+      "grad_norm": 3.2235505478452797,
+      "learning_rate": 1.8904963846525498e-05,
+      "loss": 2.4551,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3519022066077917,
+      "grad_norm": 3.3452384596793023,
+      "learning_rate": 1.8903544800928842e-05,
+      "loss": 2.4453,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35209471328536707,
+      "grad_norm": 3.0456168870735483,
+      "learning_rate": 1.8902124889793717e-05,
+      "loss": 2.4042,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3522872199629425,
+      "grad_norm": 3.1266663814464515,
+      "learning_rate": 1.8900704113258166e-05,
+      "loss": 2.3673,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35247972664051785,
+      "grad_norm": 2.853481701145706,
+      "learning_rate": 1.88992824714603e-05,
+      "loss": 2.4409,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "grad_norm": 3.3333952572076164,
+      "learning_rate": 1.8897859964538323e-05,
+      "loss": 2.4424,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.0206,
+      "step": 1832,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 1.9559,
+      "step": 1832,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.1728,
+      "step": 1832,
+      "vm_loss": 0.2192
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.4046,
+      "step": 1832,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.0467,
+      "step": 1832,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.0041,
+      "step": 1832,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.2573,
+      "step": 1832,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.3526722333180932,
+      "lm_loss": 2.0069,
+      "step": 1832,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 0.3528647399956686,
+      "grad_norm": 3.1756662783733796,
+      "learning_rate": 1.8896436592630524e-05,
+      "loss": 2.367,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35305724667324395,
+      "grad_norm": 2.768869024649846,
+      "learning_rate": 1.8895012355875262e-05,
+      "loss": 2.396,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3532497533508194,
+      "grad_norm": 2.603034053391109,
+      "learning_rate": 1.8893587254411004e-05,
+      "loss": 2.3929,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35344226002839474,
+      "grad_norm": 2.8764151681111314,
+      "learning_rate": 1.8892161288376282e-05,
+      "loss": 2.4061,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3536347667059701,
+      "grad_norm": 3.5656811328659135,
+      "learning_rate": 1.889073445790972e-05,
+      "loss": 2.4311,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35382727338354547,
+      "grad_norm": 3.0329038542118147,
+      "learning_rate": 1.8889306763150023e-05,
+      "loss": 2.3644,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3540197800611209,
+      "grad_norm": 2.6652974066488095,
+      "learning_rate": 1.8887878204235977e-05,
+      "loss": 2.4213,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "grad_norm": 3.1406848904227873,
+      "learning_rate": 1.8886448781306464e-05,
+      "loss": 2.3764,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.2437,
+      "step": 1840,
+      "vm_loss": 0.2457
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.2813,
+      "step": 1840,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 1.8656,
+      "step": 1840,
+      "vm_loss": 0.1354
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.3191,
+      "step": 1840,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.3323,
+      "step": 1840,
+      "vm_loss": 0.2042
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.399,
+      "step": 1840,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.1422,
+      "step": 1840,
+      "vm_loss": 0.2358
+    },
+    {
+      "epoch": 0.35421228673869626,
+      "lm_loss": 2.3123,
+      "step": 1840,
+      "vm_loss": 0.21
+    },
+    {
+      "epoch": 0.3544047934162716,
+      "grad_norm": 2.9327422211932563,
+      "learning_rate": 1.8885018494500436e-05,
+      "loss": 2.3661,
+      "step": 1841
+    },
+    {
+      "epoch": 0.354597300093847,
+      "grad_norm": 3.2333966435578896,
+      "learning_rate": 1.888358734395694e-05,
+      "loss": 2.385,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3547898067714224,
+      "grad_norm": 3.1461269216384915,
+      "learning_rate": 1.88821553298151e-05,
+      "loss": 2.375,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3549823134489978,
+      "grad_norm": 3.0833457080814863,
+      "learning_rate": 1.8880722452214123e-05,
+      "loss": 2.4091,
+      "step": 1844
+    },
+    {
+      "epoch": 0.35517482012657314,
+      "grad_norm": 3.2730325133341727,
+      "learning_rate": 1.887928871129331e-05,
+      "loss": 2.4039,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3553673268041485,
+      "grad_norm": 2.680616521521122,
+      "learning_rate": 1.887785410719203e-05,
+      "loss": 2.4179,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3555598334817239,
+      "grad_norm": 2.8682883067424667,
+      "learning_rate": 1.8876418640049757e-05,
+      "loss": 2.3778,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "grad_norm": 3.032017642586472,
+      "learning_rate": 1.8874982310006024e-05,
+      "loss": 2.367,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.186,
+      "step": 1848,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.4609,
+      "step": 1848,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.2255,
+      "step": 1848,
+      "vm_loss": 0.2218
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.2172,
+      "step": 1848,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.3671,
+      "step": 1848,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.2448,
+      "step": 1848,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.3494,
+      "step": 1848,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.3557523401592993,
+      "lm_loss": 2.3271,
+      "step": 1848,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 0.35594484683687466,
+      "grad_norm": 2.9426966384262343,
+      "learning_rate": 1.8873545117200467e-05,
+      "loss": 2.4336,
+      "step": 1849
+    },
+    {
+      "epoch": 0.35613735351445003,
+      "grad_norm": 3.3873083524636898,
+      "learning_rate": 1.88721070617728e-05,
+      "loss": 2.401,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3563298601920254,
+      "grad_norm": 3.0028306949601373,
+      "learning_rate": 1.887066814386282e-05,
+      "loss": 2.4453,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3565223668696008,
+      "grad_norm": 3.2221959961558078,
+      "learning_rate": 1.8869228363610406e-05,
+      "loss": 2.4398,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3567148735471762,
+      "grad_norm": 2.793047686194737,
+      "learning_rate": 1.8867787721155527e-05,
+      "loss": 2.4019,
+      "step": 1853
+    },
+    {
+      "epoch": 0.35690738022475155,
+      "grad_norm": 3.553907347466848,
+      "learning_rate": 1.8866346216638226e-05,
+      "loss": 2.3934,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3570998869023269,
+      "grad_norm": 3.3494787533447306,
+      "learning_rate": 1.886490385019864e-05,
+      "loss": 2.3857,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "grad_norm": 2.833512966230157,
+      "learning_rate": 1.8863460621976985e-05,
+      "loss": 2.3305,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.0537,
+      "step": 1856,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.144,
+      "step": 1856,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.2537,
+      "step": 1856,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.5965,
+      "step": 1856,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.0588,
+      "step": 1856,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.2977,
+      "step": 1856,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.3968,
+      "step": 1856,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.3572923935799023,
+      "lm_loss": 2.1899,
+      "step": 1856,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.3574849002574777,
+      "grad_norm": 2.798688270336243,
+      "learning_rate": 1.8862016532113562e-05,
+      "loss": 2.3998,
+      "step": 1857
+    },
+    {
+      "epoch": 0.35767740693505307,
+      "grad_norm": 3.070572149767001,
+      "learning_rate": 1.886057158074875e-05,
+      "loss": 2.3942,
+      "step": 1858
+    },
+    {
+      "epoch": 0.35786991361262843,
+      "grad_norm": 3.302594386145543,
+      "learning_rate": 1.8859125768023024e-05,
+      "loss": 2.385,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3580624202902038,
+      "grad_norm": 2.6300784393289405,
+      "learning_rate": 1.885767909407693e-05,
+      "loss": 2.3233,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3582549269677792,
+      "grad_norm": 3.1732526080158125,
+      "learning_rate": 1.8856231559051106e-05,
+      "loss": 2.3801,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3584474336453546,
+      "grad_norm": 2.9475980974253666,
+      "learning_rate": 1.885478316308627e-05,
+      "loss": 2.3744,
+      "step": 1862
+    },
+    {
+      "epoch": 0.35863994032292995,
+      "grad_norm": 2.9845057808900037,
+      "learning_rate": 1.8853333906323226e-05,
+      "loss": 2.4119,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "grad_norm": 3.122668779938046,
+      "learning_rate": 1.885188378890286e-05,
+      "loss": 2.3693,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 2.2618,
+      "step": 1864,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 2.1943,
+      "step": 1864,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 2.1947,
+      "step": 1864,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 2.395,
+      "step": 1864,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 1.795,
+      "step": 1864,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 1.8709,
+      "step": 1864,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 2.2653,
+      "step": 1864,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.3588324470005053,
+      "lm_loss": 1.7913,
+      "step": 1864,
+      "vm_loss": 0.2729
+    },
+    {
+      "epoch": 0.3590249536780807,
+      "grad_norm": 2.758639760471608,
+      "learning_rate": 1.8850432810966138e-05,
+      "loss": 2.3818,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3592174603556561,
+      "grad_norm": 2.8130805823246354,
+      "learning_rate": 1.8848980972654117e-05,
+      "loss": 2.3793,
+      "step": 1866
+    },
+    {
+      "epoch": 0.35940996703323147,
+      "grad_norm": 2.9681859596450773,
+      "learning_rate": 1.8847528274107934e-05,
+      "loss": 2.4219,
+      "step": 1867
+    },
+    {
+      "epoch": 0.35960247371080684,
+      "grad_norm": 2.9066682466205656,
+      "learning_rate": 1.8846074715468813e-05,
+      "loss": 2.3847,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3597949803883822,
+      "grad_norm": 2.7409136167481067,
+      "learning_rate": 1.8844620296878052e-05,
+      "loss": 2.4063,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3599874870659576,
+      "grad_norm": 3.4945430884721254,
+      "learning_rate": 1.8843165018477045e-05,
+      "loss": 2.4089,
+      "step": 1870
+    },
+    {
+      "epoch": 0.360179993743533,
+      "grad_norm": 3.2025875184671793,
+      "learning_rate": 1.884170888040726e-05,
+      "loss": 2.4408,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "grad_norm": 3.417333771236262,
+      "learning_rate": 1.884025188281025e-05,
+      "loss": 2.3907,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.1101,
+      "step": 1872,
+      "vm_loss": 0.0883
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.4935,
+      "step": 1872,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.2495,
+      "step": 1872,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 1.9439,
+      "step": 1872,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.3628,
+      "step": 1872,
+      "vm_loss": 0.2658
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.2243,
+      "step": 1872,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 1.8425,
+      "step": 1872,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 0.36037250042110835,
+      "lm_loss": 2.1349,
+      "step": 1872,
+      "vm_loss": 0.1281
+    },
+    {
+      "epoch": 0.3605650070986837,
+      "grad_norm": 3.1327989390307103,
+      "learning_rate": 1.883879402582766e-05,
+      "loss": 2.4281,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3607575137762591,
+      "grad_norm": 2.8734236311134635,
+      "learning_rate": 1.8837335309601214e-05,
+      "loss": 2.4036,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3609500204538345,
+      "grad_norm": 3.49351557950339,
+      "learning_rate": 1.8835875734272705e-05,
+      "loss": 2.3593,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3611425271314099,
+      "grad_norm": 4.007147016689313,
+      "learning_rate": 1.8834415299984035e-05,
+      "loss": 2.3566,
+      "step": 1876
+    },
+    {
+      "epoch": 0.36133503380898524,
+      "grad_norm": 3.2609412662507213,
+      "learning_rate": 1.8832954006877174e-05,
+      "loss": 2.3452,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3615275404865606,
+      "grad_norm": 3.5884454466667903,
+      "learning_rate": 1.8831491855094176e-05,
+      "loss": 2.3751,
+      "step": 1878
+    },
+    {
+      "epoch": 0.361720047164136,
+      "grad_norm": 3.193687345270011,
+      "learning_rate": 1.883002884477718e-05,
+      "loss": 2.3745,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "grad_norm": 3.2010730910497385,
+      "learning_rate": 1.8828564976068414e-05,
+      "loss": 2.3465,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.3082,
+      "step": 1880,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.5152,
+      "step": 1880,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.2222,
+      "step": 1880,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.3972,
+      "step": 1880,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.4112,
+      "step": 1880,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.4877,
+      "step": 1880,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 1.8775,
+      "step": 1880,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.3619125538417114,
+      "lm_loss": 2.4603,
+      "step": 1880,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 0.36210506051928676,
+      "grad_norm": 3.414759674740355,
+      "learning_rate": 1.882710024911018e-05,
+      "loss": 2.4158,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3622975671968621,
+      "grad_norm": 3.0290105832648715,
+      "learning_rate": 1.882563466404487e-05,
+      "loss": 2.3991,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36249007387443755,
+      "grad_norm": 3.3146882372720246,
+      "learning_rate": 1.882416822101496e-05,
+      "loss": 2.4032,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3626825805520129,
+      "grad_norm": 3.746256597796705,
+      "learning_rate": 1.8822700920163002e-05,
+      "loss": 2.4028,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3628750872295883,
+      "grad_norm": 2.7888523675049153,
+      "learning_rate": 1.882123276163164e-05,
+      "loss": 2.3824,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36306759390716364,
+      "grad_norm": 3.098190781130727,
+      "learning_rate": 1.88197637455636e-05,
+      "loss": 2.3873,
+      "step": 1886
+    },
+    {
+      "epoch": 0.363260100584739,
+      "grad_norm": 3.1384152081597523,
+      "learning_rate": 1.8818293872101683e-05,
+      "loss": 2.4482,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "grad_norm": 2.8336697049162636,
+      "learning_rate": 1.881682314138878e-05,
+      "loss": 2.361,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.0311,
+      "step": 1888,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.038,
+      "step": 1888,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.2507,
+      "step": 1888,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.2721,
+      "step": 1888,
+      "vm_loss": 0.2198
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.3505,
+      "step": 1888,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.1526,
+      "step": 1888,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.4873,
+      "step": 1888,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.36345260726231443,
+      "lm_loss": 2.3849,
+      "step": 1888,
+      "vm_loss": 0.1184
+    },
+    {
+      "epoch": 0.3636451139398898,
+      "grad_norm": 2.8860805308762987,
+      "learning_rate": 1.8815351553567875e-05,
+      "loss": 2.3837,
+      "step": 1889
+    },
+    {
+      "epoch": 0.36383762061746516,
+      "grad_norm": 3.0820414318235,
+      "learning_rate": 1.8813879108782014e-05,
+      "loss": 2.3988,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36403012729504053,
+      "grad_norm": 3.0225315714033343,
+      "learning_rate": 1.881240580717434e-05,
+      "loss": 2.3823,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36422263397261595,
+      "grad_norm": 3.10262796845587,
+      "learning_rate": 1.881093164888808e-05,
+      "loss": 2.3791,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3644151406501913,
+      "grad_norm": 3.1554255793581354,
+      "learning_rate": 1.880945663406654e-05,
+      "loss": 2.3871,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3646076473277667,
+      "grad_norm": 2.747965661804524,
+      "learning_rate": 1.8807980762853108e-05,
+      "loss": 2.3625,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36480015400534205,
+      "grad_norm": 3.3691012981847837,
+      "learning_rate": 1.8806504035391257e-05,
+      "loss": 2.4316,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "grad_norm": 3.2147784256728933,
+      "learning_rate": 1.8805026451824547e-05,
+      "loss": 2.3692,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 1.8805,
+      "step": 1896,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 1.9792,
+      "step": 1896,
+      "vm_loss": 0.1085
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 2.3455,
+      "step": 1896,
+      "vm_loss": 0.2345
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 1.9959,
+      "step": 1896,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 2.2495,
+      "step": 1896,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 2.4827,
+      "step": 1896,
+      "vm_loss": 0.1281
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 2.1369,
+      "step": 1896,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 0.3649926606829174,
+      "lm_loss": 2.3965,
+      "step": 1896,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.36518516736049283,
+      "grad_norm": 2.6187823658075997,
+      "learning_rate": 1.8803548012296616e-05,
+      "loss": 2.374,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3653776740380682,
+      "grad_norm": 3.320039766282272,
+      "learning_rate": 1.8802068716951187e-05,
+      "loss": 2.3727,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36557018071564357,
+      "grad_norm": 3.675441337628201,
+      "learning_rate": 1.8800588565932068e-05,
+      "loss": 2.3737,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36576268739321893,
+      "grad_norm": 3.007094943657547,
+      "learning_rate": 1.879910755938315e-05,
+      "loss": 2.3961,
+      "step": 1900
+    },
+    {
+      "epoch": 0.36595519407079435,
+      "grad_norm": 3.0405979683581363,
+      "learning_rate": 1.8797625697448402e-05,
+      "loss": 2.3969,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3661477007483697,
+      "grad_norm": 3.8330291231058227,
+      "learning_rate": 1.8796142980271887e-05,
+      "loss": 2.3956,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3663402074259451,
+      "grad_norm": 3.4382474826056484,
+      "learning_rate": 1.879465940799773e-05,
+      "loss": 2.3945,
+      "step": 1903
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "grad_norm": 2.9126509643226517,
+      "learning_rate": 1.8793174980770167e-05,
+      "loss": 2.3739,
+      "step": 1904
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.3646,
+      "step": 1904,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 1.9501,
+      "step": 1904,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.2726,
+      "step": 1904,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.3145,
+      "step": 1904,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.343,
+      "step": 1904,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 1.8461,
+      "step": 1904,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.1248,
+      "step": 1904,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.36653271410352045,
+      "lm_loss": 2.1239,
+      "step": 1904,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 0.3667252207810958,
+      "grad_norm": 3.5683443964942034,
+      "learning_rate": 1.87916896987335e-05,
+      "loss": 2.4038,
+      "step": 1905
+    },
+    {
+      "epoch": 0.36691772745867124,
+      "grad_norm": 3.1788744561231446,
+      "learning_rate": 1.8790203562032114e-05,
+      "loss": 2.3457,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3671102341362466,
+      "grad_norm": 3.0644531180815897,
+      "learning_rate": 1.8788716570810478e-05,
+      "loss": 2.3909,
+      "step": 1907
+    },
+    {
+      "epoch": 0.36730274081382197,
+      "grad_norm": 3.5609014826730947,
+      "learning_rate": 1.8787228725213155e-05,
+      "loss": 2.3211,
+      "step": 1908
+    },
+    {
+      "epoch": 0.36749524749139734,
+      "grad_norm": 3.3539129723707437,
+      "learning_rate": 1.878574002538478e-05,
+      "loss": 2.3694,
+      "step": 1909
+    },
+    {
+      "epoch": 0.36768775416897276,
+      "grad_norm": 2.8579089054359677,
+      "learning_rate": 1.878425047147007e-05,
+      "loss": 2.3609,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3678802608465481,
+      "grad_norm": 3.226354542860252,
+      "learning_rate": 1.8782760063613827e-05,
+      "loss": 2.3683,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "grad_norm": 3.847371245493553,
+      "learning_rate": 1.8781268801960947e-05,
+      "loss": 2.3816,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.2523,
+      "step": 1912,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.2065,
+      "step": 1912,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.0961,
+      "step": 1912,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.2932,
+      "step": 1912,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.2491,
+      "step": 1912,
+      "vm_loss": 0.1338
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.1999,
+      "step": 1912,
+      "vm_loss": 0.1184
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 1.9894,
+      "step": 1912,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 0.3680727675241235,
+      "lm_loss": 2.4337,
+      "step": 1912,
+      "vm_loss": 0.152
+    },
+    {
+      "epoch": 0.36826527420169886,
+      "grad_norm": 3.2629487262153583,
+      "learning_rate": 1.877977668665639e-05,
+      "loss": 2.3773,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3684577808792743,
+      "grad_norm": 3.4941922701432473,
+      "learning_rate": 1.8778283717845216e-05,
+      "loss": 2.3957,
+      "step": 1914
+    },
+    {
+      "epoch": 0.36865028755684964,
+      "grad_norm": 3.367971513105956,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 2.3977,
+      "step": 1915
+    },
+    {
+      "epoch": 0.368842794234425,
+      "grad_norm": 3.0785093524779445,
+      "learning_rate": 1.8775295220283633e-05,
+      "loss": 2.4138,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3690353009120004,
+      "grad_norm": 3.163153285453088,
+      "learning_rate": 1.8773799691823745e-05,
+      "loss": 2.3586,
+      "step": 1917
+    },
+    {
+      "epoch": 0.36922780758957574,
+      "grad_norm": 3.066781181171109,
+      "learning_rate": 1.8772303310438275e-05,
+      "loss": 2.3674,
+      "step": 1918
+    },
+    {
+      "epoch": 0.36942031426715116,
+      "grad_norm": 3.140621907030957,
+      "learning_rate": 1.8770806076272694e-05,
+      "loss": 2.3615,
+      "step": 1919
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "grad_norm": 3.3015796703934024,
+      "learning_rate": 1.8769307989472548e-05,
+      "loss": 2.4008,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.2325,
+      "step": 1920,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 1.9864,
+      "step": 1920,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 1.994,
+      "step": 1920,
+      "vm_loss": 0.2722
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.336,
+      "step": 1920,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.1098,
+      "step": 1920,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.2577,
+      "step": 1920,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.2293,
+      "step": 1920,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 0.36961282094472653,
+      "lm_loss": 2.423,
+      "step": 1920,
+      "vm_loss": 0.1074
+    },
+    {
+      "epoch": 0.3698053276223019,
+      "grad_norm": 3.8041246052027056,
+      "learning_rate": 1.8767809050183478e-05,
+      "loss": 2.3708,
+      "step": 1921
+    },
+    {
+      "epoch": 0.36999783429987726,
+      "grad_norm": 3.2521784935503453,
+      "learning_rate": 1.8766309258551195e-05,
+      "loss": 2.3597,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3701903409774527,
+      "grad_norm": 3.248778469862273,
+      "learning_rate": 1.8764808614721495e-05,
+      "loss": 2.3936,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37038284765502805,
+      "grad_norm": 3.560934149460666,
+      "learning_rate": 1.8763307118840264e-05,
+      "loss": 2.379,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3705753543326034,
+      "grad_norm": 3.041505869883774,
+      "learning_rate": 1.8761804771053467e-05,
+      "loss": 2.3626,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3707678610101788,
+      "grad_norm": 2.9401687744776828,
+      "learning_rate": 1.876030157150715e-05,
+      "loss": 2.348,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37096036768775414,
+      "grad_norm": 3.0663044410933664,
+      "learning_rate": 1.8758797520347443e-05,
+      "loss": 2.4106,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "grad_norm": 3.3873653068142526,
+      "learning_rate": 1.8757292617720558e-05,
+      "loss": 2.3837,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.3013,
+      "step": 1928,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.3144,
+      "step": 1928,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.2618,
+      "step": 1928,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 1.9892,
+      "step": 1928,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.1486,
+      "step": 1928,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.2018,
+      "step": 1928,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.3237,
+      "step": 1928,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 0.37115287436532957,
+      "lm_loss": 2.2211,
+      "step": 1928,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 0.37134538104290493,
+      "grad_norm": 3.377385777248594,
+      "learning_rate": 1.875578686377279e-05,
+      "loss": 2.3632,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3715378877204803,
+      "grad_norm": 3.2211923182640727,
+      "learning_rate": 1.8754280258650522e-05,
+      "loss": 2.3739,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37173039439805566,
+      "grad_norm": 2.889498804532783,
+      "learning_rate": 1.8752772802500215e-05,
+      "loss": 2.3582,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3719229010756311,
+      "grad_norm": 2.5217446913551798,
+      "learning_rate": 1.8751264495468412e-05,
+      "loss": 2.3939,
+      "step": 1932
+    },
+    {
+      "epoch": 0.37211540775320645,
+      "grad_norm": 3.1654377622202854,
+      "learning_rate": 1.8749755337701738e-05,
+      "loss": 2.329,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3723079144307818,
+      "grad_norm": 2.6859384383505804,
+      "learning_rate": 1.8748245329346895e-05,
+      "loss": 2.3646,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3725004211083572,
+      "grad_norm": 2.7187052960278155,
+      "learning_rate": 1.8746734470550693e-05,
+      "loss": 2.3761,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "grad_norm": 2.6963051840829473,
+      "learning_rate": 1.8745222761459995e-05,
+      "loss": 2.3997,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 1.9126,
+      "step": 1936,
+      "vm_loss": 0.1559
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.2587,
+      "step": 1936,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.1454,
+      "step": 1936,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.0813,
+      "step": 1936,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.2616,
+      "step": 1936,
+      "vm_loss": 0.2044
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.4239,
+      "step": 1936,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 1.7507,
+      "step": 1936,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.37269292778593255,
+      "lm_loss": 2.3021,
+      "step": 1936,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.37288543446350797,
+      "grad_norm": 3.0469395150127694,
+      "learning_rate": 1.8743710202221756e-05,
+      "loss": 2.3373,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37307794114108334,
+      "grad_norm": 3.0424517645989617,
+      "learning_rate": 1.8742196792983023e-05,
+      "loss": 2.4125,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3732704478186587,
+      "grad_norm": 2.9758849352586116,
+      "learning_rate": 1.874068253389092e-05,
+      "loss": 2.3958,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37346295449623407,
+      "grad_norm": 2.847104432681737,
+      "learning_rate": 1.8739167425092644e-05,
+      "loss": 2.3616,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3736554611738095,
+      "grad_norm": 2.712239491342634,
+      "learning_rate": 1.8737651466735487e-05,
+      "loss": 2.3634,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37384796785138485,
+      "grad_norm": 2.9040744783936203,
+      "learning_rate": 1.8736134658966823e-05,
+      "loss": 2.3839,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3740404745289602,
+      "grad_norm": 3.056239027324488,
+      "learning_rate": 1.8734617001934103e-05,
+      "loss": 2.3674,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "grad_norm": 3.0675740117235444,
+      "learning_rate": 1.8733098495784864e-05,
+      "loss": 2.3601,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.1866,
+      "step": 1944,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.3248,
+      "step": 1944,
+      "vm_loss": 0.2112
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 1.9464,
+      "step": 1944,
+      "vm_loss": 0.1195
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.2396,
+      "step": 1944,
+      "vm_loss": 0.2181
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.1292,
+      "step": 1944,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.1945,
+      "step": 1944,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.0653,
+      "step": 1944,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 0.3742329812065356,
+      "lm_loss": 2.05,
+      "step": 1944,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 0.37442548788411095,
+      "grad_norm": 2.8227348357938826,
+      "learning_rate": 1.873157914066672e-05,
+      "loss": 2.3979,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3746179945616864,
+      "grad_norm": 2.8041852272865286,
+      "learning_rate": 1.8730058936727367e-05,
+      "loss": 2.4049,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37481050123926174,
+      "grad_norm": 3.1714044280707805,
+      "learning_rate": 1.8728537884114603e-05,
+      "loss": 2.3971,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3750030079168371,
+      "grad_norm": 2.8849558772183075,
+      "learning_rate": 1.8727015982976287e-05,
+      "loss": 2.3108,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37519551459441247,
+      "grad_norm": 2.7366992104351677,
+      "learning_rate": 1.8725493233460362e-05,
+      "loss": 2.3581,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3753880212719879,
+      "grad_norm": 2.9404370867883247,
+      "learning_rate": 1.8723969635714863e-05,
+      "loss": 2.3992,
+      "step": 1950
+    },
+    {
+      "epoch": 0.37558052794956326,
+      "grad_norm": 3.140588166744062,
+      "learning_rate": 1.872244518988791e-05,
+      "loss": 2.4254,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "grad_norm": 2.780856850779236,
+      "learning_rate": 1.8720919896127686e-05,
+      "loss": 2.346,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.0607,
+      "step": 1952,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.2176,
+      "step": 1952,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.4069,
+      "step": 1952,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.4947,
+      "step": 1952,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.1389,
+      "step": 1952,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.1236,
+      "step": 1952,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.2782,
+      "step": 1952,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.3757730346271386,
+      "lm_loss": 2.1217,
+      "step": 1952,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 0.375965541304714,
+      "grad_norm": 3.3152655201256227,
+      "learning_rate": 1.8719393754582476e-05,
+      "loss": 2.3488,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3761580479822894,
+      "grad_norm": 3.2171576874458947,
+      "learning_rate": 1.8717866765400638e-05,
+      "loss": 2.3665,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3763505546598648,
+      "grad_norm": 3.11254153086239,
+      "learning_rate": 1.871633892873062e-05,
+      "loss": 2.3429,
+      "step": 1955
+    },
+    {
+      "epoch": 0.37654306133744014,
+      "grad_norm": 2.8596833908224295,
+      "learning_rate": 1.871481024472094e-05,
+      "loss": 2.3372,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3767355680150155,
+      "grad_norm": 3.804049061089317,
+      "learning_rate": 1.871328071352021e-05,
+      "loss": 2.353,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3769280746925909,
+      "grad_norm": 3.2778849941357224,
+      "learning_rate": 1.871175033527712e-05,
+      "loss": 2.3991,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3771205813701663,
+      "grad_norm": 3.5397498020860474,
+      "learning_rate": 1.871021911014044e-05,
+      "loss": 2.4213,
+      "step": 1959
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "grad_norm": 3.1265352783859597,
+      "learning_rate": 1.870868703825903e-05,
+      "loss": 2.3497,
+      "step": 1960
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.3613,
+      "step": 1960,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.0751,
+      "step": 1960,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 1.8618,
+      "step": 1960,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.4324,
+      "step": 1960,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.4073,
+      "step": 1960,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.2026,
+      "step": 1960,
+      "vm_loss": 0.2031
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.5163,
+      "step": 1960,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 0.37731308804774166,
+      "lm_loss": 2.3539,
+      "step": 1960,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 0.37750559472531703,
+      "grad_norm": 3.244911354964728,
+      "learning_rate": 1.870715411978182e-05,
+      "loss": 2.3894,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3776981014028924,
+      "grad_norm": 3.333701896025753,
+      "learning_rate": 1.8705620354857833e-05,
+      "loss": 2.4247,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3778906080804678,
+      "grad_norm": 3.123706656005585,
+      "learning_rate": 1.8704085743636172e-05,
+      "loss": 2.4053,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3780831147580432,
+      "grad_norm": 3.0089791592640722,
+      "learning_rate": 1.8702550286266016e-05,
+      "loss": 2.3529,
+      "step": 1964
+    },
+    {
+      "epoch": 0.37827562143561855,
+      "grad_norm": 3.13980250888995,
+      "learning_rate": 1.8701013982896637e-05,
+      "loss": 2.3809,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3784681281131939,
+      "grad_norm": 2.9609601146988185,
+      "learning_rate": 1.869947683367738e-05,
+      "loss": 2.3352,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3786606347907693,
+      "grad_norm": 3.4079749990468704,
+      "learning_rate": 1.8697938838757676e-05,
+      "loss": 2.3317,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "grad_norm": 3.007512275197292,
+      "learning_rate": 1.8696399998287035e-05,
+      "loss": 2.4,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.272,
+      "step": 1968,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.1801,
+      "step": 1968,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.2943,
+      "step": 1968,
+      "vm_loss": 0.2011
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.1813,
+      "step": 1968,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.1738,
+      "step": 1968,
+      "vm_loss": 0.1232
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.1781,
+      "step": 1968,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.374,
+      "step": 1968,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.3788531414683447,
+      "lm_loss": 2.0512,
+      "step": 1968,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 0.37904564814592007,
+      "grad_norm": 3.369135889053919,
+      "learning_rate": 1.869486031241505e-05,
+      "loss": 2.3859,
+      "step": 1969
+    },
+    {
+      "epoch": 0.37923815482349543,
+      "grad_norm": 3.1435506605616523,
+      "learning_rate": 1.869331978129141e-05,
+      "loss": 2.3138,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3794306615010708,
+      "grad_norm": 3.02540120868267,
+      "learning_rate": 1.869177840506587e-05,
+      "loss": 2.367,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3796231681786462,
+      "grad_norm": 2.862784629608692,
+      "learning_rate": 1.8690236183888266e-05,
+      "loss": 2.3668,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3798156748562216,
+      "grad_norm": 2.883959863544139,
+      "learning_rate": 1.868869311790852e-05,
+      "loss": 2.35,
+      "step": 1973
+    },
+    {
+      "epoch": 0.38000818153379695,
+      "grad_norm": 3.2223476977375585,
+      "learning_rate": 1.868714920727665e-05,
+      "loss": 2.4115,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3802006882113723,
+      "grad_norm": 2.843032120928401,
+      "learning_rate": 1.8685604452142732e-05,
+      "loss": 2.3609,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "grad_norm": 2.9367593648056785,
+      "learning_rate": 1.868405885265694e-05,
+      "loss": 2.3769,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.2925,
+      "step": 1976,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.3714,
+      "step": 1976,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.2931,
+      "step": 1976,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.4703,
+      "step": 1976,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.291,
+      "step": 1976,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.0574,
+      "step": 1976,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.1833,
+      "step": 1976,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 0.3803931948889477,
+      "lm_loss": 2.3593,
+      "step": 1976,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 0.3805857015665231,
+      "grad_norm": 3.402916162567342,
+      "learning_rate": 1.8682512408969522e-05,
+      "loss": 2.3653,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38077820824409847,
+      "grad_norm": 3.0036919034170952,
+      "learning_rate": 1.8680965121230822e-05,
+      "loss": 2.3696,
+      "step": 1978
+    },
+    {
+      "epoch": 0.38097071492167384,
+      "grad_norm": 3.0463269897308956,
+      "learning_rate": 1.8679416989591245e-05,
+      "loss": 2.3773,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3811632215992492,
+      "grad_norm": 3.264879139699,
+      "learning_rate": 1.86778680142013e-05,
+      "loss": 2.3906,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3813557282768246,
+      "grad_norm": 3.1101290285189256,
+      "learning_rate": 1.867631819521156e-05,
+      "loss": 2.3463,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3815482349544,
+      "grad_norm": 2.689597183845588,
+      "learning_rate": 1.8674767532772688e-05,
+      "loss": 2.3953,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38174074163197536,
+      "grad_norm": 2.9753959196268536,
+      "learning_rate": 1.8673216027035425e-05,
+      "loss": 2.41,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "grad_norm": 3.2863766862537886,
+      "learning_rate": 1.8671663678150605e-05,
+      "loss": 2.3829,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.6174,
+      "step": 1984,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.2023,
+      "step": 1984,
+      "vm_loss": 0.1073
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.225,
+      "step": 1984,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.1383,
+      "step": 1984,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.1718,
+      "step": 1984,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.2941,
+      "step": 1984,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.1302,
+      "step": 1984,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 0.3819332483095507,
+      "lm_loss": 2.1689,
+      "step": 1984,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.38212575498712614,
+      "grad_norm": 3.1817050865598833,
+      "learning_rate": 1.8670110486269137e-05,
+      "loss": 2.3583,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3823182616647015,
+      "grad_norm": 2.6313037644499824,
+      "learning_rate": 1.8668556451542e-05,
+      "loss": 2.3742,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3825107683422769,
+      "grad_norm": 2.7610656914380067,
+      "learning_rate": 1.8667001574120272e-05,
+      "loss": 2.2977,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38270327501985224,
+      "grad_norm": 3.34762927773237,
+      "learning_rate": 1.8665445854155114e-05,
+      "loss": 2.3691,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3828957816974276,
+      "grad_norm": 3.0795798803149097,
+      "learning_rate": 1.8663889291797754e-05,
+      "loss": 2.4007,
+      "step": 1989
+    },
+    {
+      "epoch": 0.383088288375003,
+      "grad_norm": 3.4091518069598172,
+      "learning_rate": 1.8662331887199508e-05,
+      "loss": 2.4175,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3832807950525784,
+      "grad_norm": 3.1702664908227343,
+      "learning_rate": 1.8660773640511782e-05,
+      "loss": 2.4171,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "grad_norm": 3.0870067427358654,
+      "learning_rate": 1.8659214551886052e-05,
+      "loss": 2.3238,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.3383,
+      "step": 1992,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.3296,
+      "step": 1992,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.2698,
+      "step": 1992,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.4423,
+      "step": 1992,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.3427,
+      "step": 1992,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.5469,
+      "step": 1992,
+      "vm_loss": 0.201
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.3079,
+      "step": 1992,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 0.38347330173015376,
+      "lm_loss": 2.2981,
+      "step": 1992,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 0.3836658084077291,
+      "grad_norm": 2.839682502415458,
+      "learning_rate": 1.8657654621473884e-05,
+      "loss": 2.3972,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38385831508530455,
+      "grad_norm": 3.6968356296057094,
+      "learning_rate": 1.8656093849426927e-05,
+      "loss": 2.3528,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3840508217628799,
+      "grad_norm": 2.899987226556135,
+      "learning_rate": 1.8654532235896897e-05,
+      "loss": 2.3644,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3842433284404553,
+      "grad_norm": 3.2858672919331573,
+      "learning_rate": 1.8652969781035617e-05,
+      "loss": 2.3945,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38443583511803064,
+      "grad_norm": 2.8878937760027266,
+      "learning_rate": 1.8651406484994967e-05,
+      "loss": 2.3974,
+      "step": 1997
+    },
+    {
+      "epoch": 0.384628341795606,
+      "grad_norm": 3.3099261677660308,
+      "learning_rate": 1.8649842347926922e-05,
+      "loss": 2.3715,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38482084847318143,
+      "grad_norm": 3.4595333570132487,
+      "learning_rate": 1.8648277369983537e-05,
+      "loss": 2.3914,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "grad_norm": 2.7680952978850693,
+      "learning_rate": 1.8646711551316953e-05,
+      "loss": 2.3656,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.407,
+      "step": 2000,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.353,
+      "step": 2000,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.3701,
+      "step": 2000,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.2167,
+      "step": 2000,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.3023,
+      "step": 2000,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.458,
+      "step": 2000,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 2.421,
+      "step": 2000,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.3850133551507568,
+      "lm_loss": 1.949,
+      "step": 2000,
+      "vm_loss": 0.2042
+    },
+    {
+      "epoch": 0.38520586182833216,
+      "grad_norm": 2.8440346852863283,
+      "learning_rate": 1.8645144892079375e-05,
+      "loss": 2.3992,
+      "step": 2001
+    },
+    {
+      "epoch": 0.38539836850590753,
+      "grad_norm": 3.0163884086557418,
+      "learning_rate": 1.8643577392423117e-05,
+      "loss": 2.3894,
+      "step": 2002
+    },
+    {
+      "epoch": 0.38559087518348295,
+      "grad_norm": 3.246043913748404,
+      "learning_rate": 1.8642009052500545e-05,
+      "loss": 2.4099,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3857833818610583,
+      "grad_norm": 2.839426449676454,
+      "learning_rate": 1.8640439872464134e-05,
+      "loss": 2.3676,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3859758885386337,
+      "grad_norm": 3.343379803114167,
+      "learning_rate": 1.8638869852466427e-05,
+      "loss": 2.3827,
+      "step": 2005
+    },
+    {
+      "epoch": 0.38616839521620905,
+      "grad_norm": 3.0302767493948917,
+      "learning_rate": 1.8637298992660042e-05,
+      "loss": 2.412,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3863609018937844,
+      "grad_norm": 2.9078652856877047,
+      "learning_rate": 1.8635727293197692e-05,
+      "loss": 2.3806,
+      "step": 2007
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "grad_norm": 3.0201214146653195,
+      "learning_rate": 1.8634154754232166e-05,
+      "loss": 2.4042,
+      "step": 2008
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.2746,
+      "step": 2008,
+      "vm_loss": 0.0923
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.3249,
+      "step": 2008,
+      "vm_loss": 0.2573
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.2309,
+      "step": 2008,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.1614,
+      "step": 2008,
+      "vm_loss": 0.2161
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.1619,
+      "step": 2008,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.1641,
+      "step": 2008,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 1.9329,
+      "step": 2008,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 0.38655340857135984,
+      "lm_loss": 2.0703,
+      "step": 2008,
+      "vm_loss": 0.2889
+    },
+    {
+      "epoch": 0.3867459152489352,
+      "grad_norm": 3.0136423275193147,
+      "learning_rate": 1.863258137591634e-05,
+      "loss": 2.3728,
+      "step": 2009
+    },
+    {
+      "epoch": 0.38693842192651057,
+      "grad_norm": 2.879018694904631,
+      "learning_rate": 1.8631007158403155e-05,
+      "loss": 2.3785,
+      "step": 2010
+    },
+    {
+      "epoch": 0.38713092860408593,
+      "grad_norm": 2.85103774703889,
+      "learning_rate": 1.8629432101845655e-05,
+      "loss": 2.3931,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38732343528166135,
+      "grad_norm": 2.8668263695269784,
+      "learning_rate": 1.862785620639695e-05,
+      "loss": 2.3724,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3875159419592367,
+      "grad_norm": 2.5701259852923335,
+      "learning_rate": 1.8626279472210237e-05,
+      "loss": 2.3411,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3877084486368121,
+      "grad_norm": 3.2875170339065534,
+      "learning_rate": 1.8624701899438802e-05,
+      "loss": 2.3735,
+      "step": 2014
+    },
+    {
+      "epoch": 0.38790095531438745,
+      "grad_norm": 3.126993072709185,
+      "learning_rate": 1.8623123488235995e-05,
+      "loss": 2.3321,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "grad_norm": 2.6647001464043987,
+      "learning_rate": 1.862154423875527e-05,
+      "loss": 2.3769,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.1819,
+      "step": 2016,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.2146,
+      "step": 2016,
+      "vm_loss": 0.2051
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.2675,
+      "step": 2016,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.137,
+      "step": 2016,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.2228,
+      "step": 2016,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.199,
+      "step": 2016,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.2623,
+      "step": 2016,
+      "vm_loss": 0.263
+    },
+    {
+      "epoch": 0.3880934619919629,
+      "lm_loss": 2.4624,
+      "step": 2016,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.38828596866953824,
+      "grad_norm": 3.3470613889497596,
+      "learning_rate": 1.861996415115014e-05,
+      "loss": 2.4091,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3884784753471136,
+      "grad_norm": 3.616999162685489,
+      "learning_rate": 1.861838322557421e-05,
+      "loss": 2.3646,
+      "step": 2018
+    },
+    {
+      "epoch": 0.38867098202468897,
+      "grad_norm": 3.0673270312334022,
+      "learning_rate": 1.861680146218117e-05,
+      "loss": 2.3448,
+      "step": 2019
+    },
+    {
+      "epoch": 0.38886348870226434,
+      "grad_norm": 3.1477514897276997,
+      "learning_rate": 1.8615218861124787e-05,
+      "loss": 2.3796,
+      "step": 2020
+    },
+    {
+      "epoch": 0.38905599537983976,
+      "grad_norm": 3.2590006522776997,
+      "learning_rate": 1.8613635422558913e-05,
+      "loss": 2.3036,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3892485020574151,
+      "grad_norm": 3.2819230314377688,
+      "learning_rate": 1.8612051146637477e-05,
+      "loss": 2.4464,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3894410087349905,
+      "grad_norm": 3.4984747289433997,
+      "learning_rate": 1.8610466033514483e-05,
+      "loss": 2.3916,
+      "step": 2023
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "grad_norm": 2.9443748339448943,
+      "learning_rate": 1.8608880083344033e-05,
+      "loss": 2.4302,
+      "step": 2024
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.5331,
+      "step": 2024,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.1304,
+      "step": 2024,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.4083,
+      "step": 2024,
+      "vm_loss": 0.1096
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.3711,
+      "step": 2024,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.2704,
+      "step": 2024,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.5538,
+      "step": 2024,
+      "vm_loss": 0.2058
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.3603,
+      "step": 2024,
+      "vm_loss": 0.2367
+    },
+    {
+      "epoch": 0.38963351541256586,
+      "lm_loss": 2.1702,
+      "step": 2024,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 0.3898260220901413,
+      "grad_norm": 3.4604129325753603,
+      "learning_rate": 1.86072932962803e-05,
+      "loss": 2.384,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39001852876771664,
+      "grad_norm": 3.145886776532001,
+      "learning_rate": 1.860570567247754e-05,
+      "loss": 2.3878,
+      "step": 2026
+    },
+    {
+      "epoch": 0.390211035445292,
+      "grad_norm": 2.583739171946118,
+      "learning_rate": 1.8604117212090084e-05,
+      "loss": 2.3341,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3904035421228674,
+      "grad_norm": 3.353559198885944,
+      "learning_rate": 1.860252791527236e-05,
+      "loss": 2.4004,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39059604880044274,
+      "grad_norm": 3.029609564913458,
+      "learning_rate": 1.8600937782178868e-05,
+      "loss": 2.3855,
+      "step": 2029
+    },
+    {
+      "epoch": 0.39078855547801816,
+      "grad_norm": 2.9129425309627703,
+      "learning_rate": 1.859934681296418e-05,
+      "loss": 2.3812,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39098106215559353,
+      "grad_norm": 3.5733235480970036,
+      "learning_rate": 1.8597755007782966e-05,
+      "loss": 2.3406,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "grad_norm": 3.238856079367312,
+      "learning_rate": 1.859616236678997e-05,
+      "loss": 2.3559,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.1058,
+      "step": 2032,
+      "vm_loss": 0.2438
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.3157,
+      "step": 2032,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.1296,
+      "step": 2032,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.4608,
+      "step": 2032,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.1726,
+      "step": 2032,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.1331,
+      "step": 2032,
+      "vm_loss": 0.1576
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.225,
+      "step": 2032,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.3911735688331689,
+      "lm_loss": 2.1458,
+      "step": 2032,
+      "vm_loss": 0.2434
+    },
+    {
+      "epoch": 0.39136607551074426,
+      "grad_norm": 2.9747930290792097,
+      "learning_rate": 1.859456889014001e-05,
+      "loss": 2.3842,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3915585821883197,
+      "grad_norm": 3.4932337159423663,
+      "learning_rate": 1.8592974577988e-05,
+      "loss": 2.358,
+      "step": 2034
+    },
+    {
+      "epoch": 0.39175108886589505,
+      "grad_norm": 2.7743135368373615,
+      "learning_rate": 1.8591379430488922e-05,
+      "loss": 2.3519,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3919435955434704,
+      "grad_norm": 3.2273189259824573,
+      "learning_rate": 1.8589783447797852e-05,
+      "loss": 2.3795,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3921361022210458,
+      "grad_norm": 3.3361305306755966,
+      "learning_rate": 1.8588186630069935e-05,
+      "loss": 2.3414,
+      "step": 2037
+    },
+    {
+      "epoch": 0.39232860889862115,
+      "grad_norm": 2.892889310143435,
+      "learning_rate": 1.85865889774604e-05,
+      "loss": 2.3361,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39252111557619657,
+      "grad_norm": 3.444314411455052,
+      "learning_rate": 1.858499049012456e-05,
+      "loss": 2.3996,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "grad_norm": 3.217486455718521,
+      "learning_rate": 1.8583391168217814e-05,
+      "loss": 2.3229,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.1637,
+      "step": 2040,
+      "vm_loss": 0.2215
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.4445,
+      "step": 2040,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.495,
+      "step": 2040,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.4946,
+      "step": 2040,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.2974,
+      "step": 2040,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.4617,
+      "step": 2040,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.228,
+      "step": 2040,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 0.39271362225377193,
+      "lm_loss": 2.2925,
+      "step": 2040,
+      "vm_loss": 0.2802
+    },
+    {
+      "epoch": 0.3929061289313473,
+      "grad_norm": 2.6109603581382097,
+      "learning_rate": 1.8581791011895633e-05,
+      "loss": 2.4388,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39309863560892266,
+      "grad_norm": 2.790514869338661,
+      "learning_rate": 1.858019002131357e-05,
+      "loss": 2.3692,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3932911422864981,
+      "grad_norm": 3.0620125357572796,
+      "learning_rate": 1.8578588196627266e-05,
+      "loss": 2.3519,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39348364896407345,
+      "grad_norm": 2.524050758725505,
+      "learning_rate": 1.857698553799244e-05,
+      "loss": 2.3478,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3936761556416488,
+      "grad_norm": 2.922462216205388,
+      "learning_rate": 1.8575382045564883e-05,
+      "loss": 2.3428,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3938686623192242,
+      "grad_norm": 3.1158354683087017,
+      "learning_rate": 1.857377771950048e-05,
+      "loss": 2.3585,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39406116899679955,
+      "grad_norm": 3.046297916046909,
+      "learning_rate": 1.8572172559955194e-05,
+      "loss": 2.3604,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "grad_norm": 3.4195501727544704,
+      "learning_rate": 1.8570566567085067e-05,
+      "loss": 2.317,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.1239,
+      "step": 2048,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 1.8804,
+      "step": 2048,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.1542,
+      "step": 2048,
+      "vm_loss": 0.2163
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.1105,
+      "step": 2048,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 1.8584,
+      "step": 2048,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.1507,
+      "step": 2048,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.1636,
+      "step": 2048,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 0.39425367567437497,
+      "lm_loss": 2.3265,
+      "step": 2048,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 0.39444618235195034,
+      "grad_norm": 3.1153836592471915,
+      "learning_rate": 1.8568959741046217e-05,
+      "loss": 2.3245,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3946386890295257,
+      "grad_norm": 2.3565713206396324,
+      "learning_rate": 1.8567352081994852e-05,
+      "loss": 2.356,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39483119570710107,
+      "grad_norm": 3.095627990215364,
+      "learning_rate": 1.8565743590087256e-05,
+      "loss": 2.3211,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3950237023846765,
+      "grad_norm": 3.0688479603071634,
+      "learning_rate": 1.8564134265479798e-05,
+      "loss": 2.3572,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39521620906225186,
+      "grad_norm": 2.773118438442102,
+      "learning_rate": 1.856252410832892e-05,
+      "loss": 2.3896,
+      "step": 2053
+    },
+    {
+      "epoch": 0.3954087157398272,
+      "grad_norm": 2.9907654759929025,
+      "learning_rate": 1.856091311879115e-05,
+      "loss": 2.3259,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3956012224174026,
+      "grad_norm": 2.593740916753053,
+      "learning_rate": 1.85593012970231e-05,
+      "loss": 2.3194,
+      "step": 2055
+    },
+    {
+      "epoch": 0.395793729094978,
+      "grad_norm": 2.881627558220902,
+      "learning_rate": 1.855768864318146e-05,
+      "loss": 2.371,
+      "step": 2056
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.1765,
+      "step": 2056,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 1.94,
+      "step": 2056,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.0098,
+      "step": 2056,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.2663,
+      "step": 2056,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.1563,
+      "step": 2056,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.6052,
+      "step": 2056,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.2295,
+      "step": 2056,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.395793729094978,
+      "lm_loss": 2.3169,
+      "step": 2056,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.3959862357725534,
+      "grad_norm": 3.16326948920532,
+      "learning_rate": 1.8556075157422996e-05,
+      "loss": 2.3685,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39617874245012874,
+      "grad_norm": 3.034727417399684,
+      "learning_rate": 1.8554460839904566e-05,
+      "loss": 2.3517,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3963712491277041,
+      "grad_norm": 2.743832446696198,
+      "learning_rate": 1.8552845690783096e-05,
+      "loss": 2.397,
+      "step": 2059
+    },
+    {
+      "epoch": 0.39656375580527947,
+      "grad_norm": 2.665699129525352,
+      "learning_rate": 1.8551229710215604e-05,
+      "loss": 2.3751,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3967562624828549,
+      "grad_norm": 2.870773403906124,
+      "learning_rate": 1.8549612898359185e-05,
+      "loss": 2.3075,
+      "step": 2061
+    },
+    {
+      "epoch": 0.39694876916043026,
+      "grad_norm": 2.718877430868723,
+      "learning_rate": 1.8547995255371007e-05,
+      "loss": 2.3559,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3971412758380056,
+      "grad_norm": 3.003171338836357,
+      "learning_rate": 1.8546376781408328e-05,
+      "loss": 2.4221,
+      "step": 2063
+    },
+    {
+      "epoch": 0.397333782515581,
+      "grad_norm": 2.69965600956393,
+      "learning_rate": 1.8544757476628487e-05,
+      "loss": 2.3154,
+      "step": 2064
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.2021,
+      "step": 2064,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.075,
+      "step": 2064,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 1.9951,
+      "step": 2064,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.2588,
+      "step": 2064,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.3122,
+      "step": 2064,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.1324,
+      "step": 2064,
+      "vm_loss": 0.2003
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.1604,
+      "step": 2064,
+      "vm_loss": 0.1192
+    },
+    {
+      "epoch": 0.397333782515581,
+      "lm_loss": 2.1421,
+      "step": 2064,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.3975262891931564,
+      "grad_norm": 2.988678085262676,
+      "learning_rate": 1.85431373411889e-05,
+      "loss": 2.3549,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3977187958707318,
+      "grad_norm": 2.8748570422625446,
+      "learning_rate": 1.8541516375247063e-05,
+      "loss": 2.3505,
+      "step": 2066
+    },
+    {
+      "epoch": 0.39791130254830714,
+      "grad_norm": 2.9680860132199283,
+      "learning_rate": 1.853989457896056e-05,
+      "loss": 2.3902,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3981038092258825,
+      "grad_norm": 3.32396089252872,
+      "learning_rate": 1.8538271952487044e-05,
+      "loss": 2.3977,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3982963159034579,
+      "grad_norm": 3.1991700906287055,
+      "learning_rate": 1.853664849598426e-05,
+      "loss": 2.3517,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3984888225810333,
+      "grad_norm": 3.1527773494010525,
+      "learning_rate": 1.8535024209610026e-05,
+      "loss": 2.3576,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39868132925860866,
+      "grad_norm": 3.3321723724755192,
+      "learning_rate": 1.8533399093522244e-05,
+      "loss": 2.3578,
+      "step": 2071
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "grad_norm": 2.9962683363846616,
+      "learning_rate": 1.8531773147878895e-05,
+      "loss": 2.3266,
+      "step": 2072
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.2826,
+      "step": 2072,
+      "vm_loss": 0.2171
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.434,
+      "step": 2072,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.2177,
+      "step": 2072,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 1.7106,
+      "step": 2072,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.3132,
+      "step": 2072,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.3104,
+      "step": 2072,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.1642,
+      "step": 2072,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 0.39887383593618403,
+      "lm_loss": 2.1532,
+      "step": 2072,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.3990663426137594,
+      "grad_norm": 3.2315090828305393,
+      "learning_rate": 1.8530146372838045e-05,
+      "loss": 2.3467,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3992588492913348,
+      "grad_norm": 2.956983044209395,
+      "learning_rate": 1.852851876855783e-05,
+      "loss": 2.335,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3994513559689102,
+      "grad_norm": 3.2636455416558703,
+      "learning_rate": 1.8526890335196483e-05,
+      "loss": 2.3525,
+      "step": 2075
+    },
+    {
+      "epoch": 0.39964386264648555,
+      "grad_norm": 3.173723207522457,
+      "learning_rate": 1.8525261072912305e-05,
+      "loss": 2.3799,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3998363693240609,
+      "grad_norm": 3.2814888328996474,
+      "learning_rate": 1.852363098186368e-05,
+      "loss": 2.3786,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4000288760016363,
+      "grad_norm": 3.6152925232572755,
+      "learning_rate": 1.8522000062209078e-05,
+      "loss": 2.3779,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4002213826792117,
+      "grad_norm": 3.020599153252269,
+      "learning_rate": 1.8520368314107036e-05,
+      "loss": 2.3436,
+      "step": 2079
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "grad_norm": 3.05394719935269,
+      "learning_rate": 1.8518735737716188e-05,
+      "loss": 2.3577,
+      "step": 2080
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 1.9162,
+      "step": 2080,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 2.2728,
+      "step": 2080,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 2.2192,
+      "step": 2080,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 2.2454,
+      "step": 2080,
+      "vm_loss": 0.2285
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 2.2141,
+      "step": 2080,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 1.9505,
+      "step": 2080,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 1.9737,
+      "step": 2080,
+      "vm_loss": 0.1511
+    },
+    {
+      "epoch": 0.40041388935678707,
+      "lm_loss": 2.1599,
+      "step": 2080,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 0.40060639603436243,
+      "grad_norm": 3.0858368231494193,
+      "learning_rate": 1.851710233319524e-05,
+      "loss": 2.3258,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4007989027119378,
+      "grad_norm": 3.422540185958113,
+      "learning_rate": 1.8515468100702983e-05,
+      "loss": 2.4107,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4009914093895132,
+      "grad_norm": 2.821456428774914,
+      "learning_rate": 1.851383304039828e-05,
+      "loss": 2.3222,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4011839160670886,
+      "grad_norm": 3.142853241007964,
+      "learning_rate": 1.851219715244008e-05,
+      "loss": 2.3429,
+      "step": 2084
+    },
+    {
+      "epoch": 0.40137642274466395,
+      "grad_norm": 3.7750412960300346,
+      "learning_rate": 1.8510560436987414e-05,
+      "loss": 2.3645,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4015689294222393,
+      "grad_norm": 3.029553677458068,
+      "learning_rate": 1.8508922894199393e-05,
+      "loss": 2.3588,
+      "step": 2086
+    },
+    {
+      "epoch": 0.40176143609981474,
+      "grad_norm": 3.1623164441063927,
+      "learning_rate": 1.8507284524235208e-05,
+      "loss": 2.3256,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "grad_norm": 3.152002749911262,
+      "learning_rate": 1.850564532725412e-05,
+      "loss": 2.4107,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.2248,
+      "step": 2088,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 1.8678,
+      "step": 2088,
+      "vm_loss": 0.1804
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.1745,
+      "step": 2088,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 1.9515,
+      "step": 2088,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.129,
+      "step": 2088,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.1459,
+      "step": 2088,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.2922,
+      "step": 2088,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.4019539427773901,
+      "lm_loss": 2.3448,
+      "step": 2088,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.40214644945496547,
+      "grad_norm": 2.773597977322333,
+      "learning_rate": 1.8504005303415494e-05,
+      "loss": 2.349,
+      "step": 2089
+    },
+    {
+      "epoch": 0.40233895613254084,
+      "grad_norm": 3.139764150050187,
+      "learning_rate": 1.8502364452878755e-05,
+      "loss": 2.3327,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4025314628101162,
+      "grad_norm": 2.895376636997464,
+      "learning_rate": 1.850072277580341e-05,
+      "loss": 2.405,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4027239694876916,
+      "grad_norm": 3.6007083949359076,
+      "learning_rate": 1.8499080272349057e-05,
+      "loss": 2.3429,
+      "step": 2092
+    },
+    {
+      "epoch": 0.402916476165267,
+      "grad_norm": 3.3544394363705448,
+      "learning_rate": 1.8497436942675366e-05,
+      "loss": 2.3403,
+      "step": 2093
+    },
+    {
+      "epoch": 0.40310898284284236,
+      "grad_norm": 3.386412673761896,
+      "learning_rate": 1.849579278694209e-05,
+      "loss": 2.3836,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4033014895204177,
+      "grad_norm": 3.526287711151043,
+      "learning_rate": 1.849414780530906e-05,
+      "loss": 2.3513,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "grad_norm": 3.4130453504031113,
+      "learning_rate": 1.8492501997936196e-05,
+      "loss": 2.3317,
+      "step": 2096
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.2077,
+      "step": 2096,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.07,
+      "step": 2096,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.3461,
+      "step": 2096,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.2929,
+      "step": 2096,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.23,
+      "step": 2096,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.0502,
+      "step": 2096,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.3213,
+      "step": 2096,
+      "vm_loss": 0.2232
+    },
+    {
+      "epoch": 0.40349399619799314,
+      "lm_loss": 2.2544,
+      "step": 2096,
+      "vm_loss": 0.2123
+    },
+    {
+      "epoch": 0.4036865028755685,
+      "grad_norm": 2.754506936894068,
+      "learning_rate": 1.8490855364983488e-05,
+      "loss": 2.3744,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4038790095531439,
+      "grad_norm": 3.260158849495847,
+      "learning_rate": 1.8489207906611003e-05,
+      "loss": 2.3482,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40407151623071924,
+      "grad_norm": 3.0051348691845177,
+      "learning_rate": 1.8487559622978906e-05,
+      "loss": 2.3223,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4042640229082946,
+      "grad_norm": 2.9106647062416706,
+      "learning_rate": 1.8485910514247424e-05,
+      "loss": 2.3617,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40445652958587003,
+      "grad_norm": 3.040003452725519,
+      "learning_rate": 1.8484260580576873e-05,
+      "loss": 2.3809,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4046490362634454,
+      "grad_norm": 3.277968031109553,
+      "learning_rate": 1.8482609822127647e-05,
+      "loss": 2.3713,
+      "step": 2102
+    },
+    {
+      "epoch": 0.40484154294102076,
+      "grad_norm": 2.5418297864609767,
+      "learning_rate": 1.8480958239060226e-05,
+      "loss": 2.3364,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "grad_norm": 3.045313253427444,
+      "learning_rate": 1.8479305831535156e-05,
+      "loss": 2.3705,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.0037,
+      "step": 2104,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.1825,
+      "step": 2104,
+      "vm_loss": 0.1171
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.2291,
+      "step": 2104,
+      "vm_loss": 0.2034
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.0108,
+      "step": 2104,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.4695,
+      "step": 2104,
+      "vm_loss": 0.2458
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 1.902,
+      "step": 2104,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.0359,
+      "step": 2104,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 0.4050340496185961,
+      "lm_loss": 2.1429,
+      "step": 2104,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 0.40522655629617155,
+      "grad_norm": 2.6914989609059177,
+      "learning_rate": 1.8477652599713082e-05,
+      "loss": 2.3515,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4054190629737469,
+      "grad_norm": 2.9839354131602387,
+      "learning_rate": 1.8475998543754713e-05,
+      "loss": 2.3393,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4056115696513223,
+      "grad_norm": 2.9286941595123963,
+      "learning_rate": 1.8474343663820844e-05,
+      "loss": 2.3487,
+      "step": 2107
+    },
+    {
+      "epoch": 0.40580407632889764,
+      "grad_norm": 2.7218229552518807,
+      "learning_rate": 1.847268796007235e-05,
+      "loss": 2.3624,
+      "step": 2108
+    },
+    {
+      "epoch": 0.405996583006473,
+      "grad_norm": 2.5674365360169564,
+      "learning_rate": 1.8471031432670194e-05,
+      "loss": 2.344,
+      "step": 2109
+    },
+    {
+      "epoch": 0.40618908968404843,
+      "grad_norm": 3.198034870853174,
+      "learning_rate": 1.8469374081775405e-05,
+      "loss": 2.388,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4063815963616238,
+      "grad_norm": 3.1875221751385023,
+      "learning_rate": 1.84677159075491e-05,
+      "loss": 2.3866,
+      "step": 2111
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "grad_norm": 2.9758315548135417,
+      "learning_rate": 1.8466056910152476e-05,
+      "loss": 2.3774,
+      "step": 2112
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.4807,
+      "step": 2112,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 1.9244,
+      "step": 2112,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.3026,
+      "step": 2112,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 1.8861,
+      "step": 2112,
+      "vm_loss": 0.2457
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.5146,
+      "step": 2112,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.1579,
+      "step": 2112,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.178,
+      "step": 2112,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 0.40657410303919916,
+      "lm_loss": 2.0944,
+      "step": 2112,
+      "vm_loss": 0.2179
+    },
+    {
+      "epoch": 0.40676660971677453,
+      "grad_norm": 3.2055877847242895,
+      "learning_rate": 1.8464397089746807e-05,
+      "loss": 2.3455,
+      "step": 2113
+    },
+    {
+      "epoch": 0.40695911639434995,
+      "grad_norm": 2.9327102257208355,
+      "learning_rate": 1.8462736446493452e-05,
+      "loss": 2.3763,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4071516230719253,
+      "grad_norm": 2.6493428633871234,
+      "learning_rate": 1.846107498055384e-05,
+      "loss": 2.3452,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4073441297495007,
+      "grad_norm": 3.0305954613896686,
+      "learning_rate": 1.8459412692089497e-05,
+      "loss": 2.3569,
+      "step": 2116
+    },
+    {
+      "epoch": 0.40753663642707605,
+      "grad_norm": 3.408555878596119,
+      "learning_rate": 1.8457749581262006e-05,
+      "loss": 2.3873,
+      "step": 2117
+    },
+    {
+      "epoch": 0.40772914310465147,
+      "grad_norm": 2.8994020568100716,
+      "learning_rate": 1.8456085648233054e-05,
+      "loss": 2.3589,
+      "step": 2118
+    },
+    {
+      "epoch": 0.40792164978222684,
+      "grad_norm": 3.2053508022916324,
+      "learning_rate": 1.845442089316439e-05,
+      "loss": 2.3044,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "grad_norm": 3.389674238590442,
+      "learning_rate": 1.8452755316217854e-05,
+      "loss": 2.3681,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.2301,
+      "step": 2120,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.1089,
+      "step": 2120,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.3274,
+      "step": 2120,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.4197,
+      "step": 2120,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.0335,
+      "step": 2120,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.3224,
+      "step": 2120,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.0235,
+      "step": 2120,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.4081141564598022,
+      "lm_loss": 2.2433,
+      "step": 2120,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 0.40830666313737757,
+      "grad_norm": 2.808924740374487,
+      "learning_rate": 1.8451088917555356e-05,
+      "loss": 2.3595,
+      "step": 2121
+    },
+    {
+      "epoch": 0.40849916981495293,
+      "grad_norm": 3.212233776272681,
+      "learning_rate": 1.8449421697338896e-05,
+      "loss": 2.3535,
+      "step": 2122
+    },
+    {
+      "epoch": 0.40869167649252836,
+      "grad_norm": 3.0112316348161388,
+      "learning_rate": 1.8447753655730547e-05,
+      "loss": 2.3898,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4088841831701037,
+      "grad_norm": 2.5425130979155326,
+      "learning_rate": 1.8446084792892464e-05,
+      "loss": 2.381,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4090766898476791,
+      "grad_norm": 2.9916612674600174,
+      "learning_rate": 1.8444415108986882e-05,
+      "loss": 2.3354,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40926919652525445,
+      "grad_norm": 3.2384512062452933,
+      "learning_rate": 1.844274460417612e-05,
+      "loss": 2.3252,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4094617032028299,
+      "grad_norm": 2.765098258363053,
+      "learning_rate": 1.8441073278622562e-05,
+      "loss": 2.3449,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "grad_norm": 3.532905199614185,
+      "learning_rate": 1.843940113248869e-05,
+      "loss": 2.3256,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.3837,
+      "step": 2128,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.0457,
+      "step": 2128,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.3777,
+      "step": 2128,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.0095,
+      "step": 2128,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.357,
+      "step": 2128,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.0287,
+      "step": 2128,
+      "vm_loss": 0.2898
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.2274,
+      "step": 2128,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.40965420988040524,
+      "lm_loss": 2.2978,
+      "step": 2128,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 0.4098467165579806,
+      "grad_norm": 3.4233780188970067,
+      "learning_rate": 1.8437728165937056e-05,
+      "loss": 2.3334,
+      "step": 2129
+    },
+    {
+      "epoch": 0.41003922323555597,
+      "grad_norm": 3.036727315418898,
+      "learning_rate": 1.8436054379130296e-05,
+      "loss": 2.3704,
+      "step": 2130
+    },
+    {
+      "epoch": 0.41023172991313134,
+      "grad_norm": 3.473673653482789,
+      "learning_rate": 1.8434379772231122e-05,
+      "loss": 2.3206,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41042423659070676,
+      "grad_norm": 3.3594688254290332,
+      "learning_rate": 1.8432704345402324e-05,
+      "loss": 2.3723,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4106167432682821,
+      "grad_norm": 2.8684637280670726,
+      "learning_rate": 1.8431028098806783e-05,
+      "loss": 2.4048,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4108092499458575,
+      "grad_norm": 2.9474018999457883,
+      "learning_rate": 1.8429351032607444e-05,
+      "loss": 2.3326,
+      "step": 2134
+    },
+    {
+      "epoch": 0.41100175662343286,
+      "grad_norm": 4.100170445873237,
+      "learning_rate": 1.842767314696734e-05,
+      "loss": 2.3102,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "grad_norm": 2.994105072789109,
+      "learning_rate": 1.8425994442049585e-05,
+      "loss": 2.3435,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.0731,
+      "step": 2136,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.1314,
+      "step": 2136,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.1192,
+      "step": 2136,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 1.971,
+      "step": 2136,
+      "vm_loss": 0.2431
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.3416,
+      "step": 2136,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.1558,
+      "step": 2136,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.1965,
+      "step": 2136,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 0.4111942633010083,
+      "lm_loss": 2.0486,
+      "step": 2136,
+      "vm_loss": 0.2119
+    },
+    {
+      "epoch": 0.41138676997858364,
+      "grad_norm": 3.303565617203994,
+      "learning_rate": 1.8424314918017372e-05,
+      "loss": 2.3768,
+      "step": 2137
+    },
+    {
+      "epoch": 0.411579276656159,
+      "grad_norm": 3.202682378535108,
+      "learning_rate": 1.842263457503397e-05,
+      "loss": 2.3502,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4117717833337344,
+      "grad_norm": 2.7570099050112837,
+      "learning_rate": 1.8420953413262727e-05,
+      "loss": 2.3837,
+      "step": 2139
+    },
+    {
+      "epoch": 0.41196429001130974,
+      "grad_norm": 3.021769115981808,
+      "learning_rate": 1.841927143286708e-05,
+      "loss": 2.3498,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41215679668888516,
+      "grad_norm": 2.836439852485431,
+      "learning_rate": 1.8417588634010536e-05,
+      "loss": 2.3951,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41234930336646053,
+      "grad_norm": 3.1906417927306014,
+      "learning_rate": 1.8415905016856682e-05,
+      "loss": 2.3617,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4125418100440359,
+      "grad_norm": 2.802803268464032,
+      "learning_rate": 1.841422058156919e-05,
+      "loss": 2.3259,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "grad_norm": 2.814776057707198,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 2.383,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.146,
+      "step": 2144,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.2029,
+      "step": 2144,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.4944,
+      "step": 2144,
+      "vm_loss": 0.1191
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.3076,
+      "step": 2144,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.2968,
+      "step": 2144,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.2323,
+      "step": 2144,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.1842,
+      "step": 2144,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 0.41273431672161126,
+      "lm_loss": 2.2253,
+      "step": 2144,
+      "vm_loss": 0.2586
+    },
+    {
+      "epoch": 0.4129268233991867,
+      "grad_norm": 2.8833750937905935,
+      "learning_rate": 1.8410849257248372e-05,
+      "loss": 2.3451,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41311933007676205,
+      "grad_norm": 3.1989008487113386,
+      "learning_rate": 1.8409162368542778e-05,
+      "loss": 2.3593,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4133118367543374,
+      "grad_norm": 2.686935078478039,
+      "learning_rate": 1.8407474662359014e-05,
+      "loss": 2.3545,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4135043434319128,
+      "grad_norm": 2.917875871207584,
+      "learning_rate": 1.840578613886115e-05,
+      "loss": 2.3165,
+      "step": 2148
+    },
+    {
+      "epoch": 0.41369685010948815,
+      "grad_norm": 2.9899575980229938,
+      "learning_rate": 1.8404096798213334e-05,
+      "loss": 2.3807,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41388935678706357,
+      "grad_norm": 3.3527018619678497,
+      "learning_rate": 1.8402406640579788e-05,
+      "loss": 2.3335,
+      "step": 2150
+    },
+    {
+      "epoch": 0.41408186346463893,
+      "grad_norm": 3.041375448295186,
+      "learning_rate": 1.8400715666124815e-05,
+      "loss": 2.3613,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "grad_norm": 3.012256771374496,
+      "learning_rate": 1.8399023875012808e-05,
+      "loss": 2.2652,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.1856,
+      "step": 2152,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.3424,
+      "step": 2152,
+      "vm_loss": 0.1178
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.4276,
+      "step": 2152,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 1.8783,
+      "step": 2152,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.3399,
+      "step": 2152,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.1739,
+      "step": 2152,
+      "vm_loss": 0.2735
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.1701,
+      "step": 2152,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.4142743701422143,
+      "lm_loss": 2.3573,
+      "step": 2152,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 0.41446687681978966,
+      "grad_norm": 3.2521429507103083,
+      "learning_rate": 1.8397331267408222e-05,
+      "loss": 2.3721,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4146593834973651,
+      "grad_norm": 2.8432046010965757,
+      "learning_rate": 1.8395637843475607e-05,
+      "loss": 2.3389,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41485189017494045,
+      "grad_norm": 3.0496752744846938,
+      "learning_rate": 1.839394360337958e-05,
+      "loss": 2.3211,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4150443968525158,
+      "grad_norm": 3.036044526875652,
+      "learning_rate": 1.8392248547284844e-05,
+      "loss": 2.3149,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4152369035300912,
+      "grad_norm": 2.9626809902230207,
+      "learning_rate": 1.8390552675356183e-05,
+      "loss": 2.3476,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4154294102076666,
+      "grad_norm": 2.6944664731149395,
+      "learning_rate": 1.8388855987758455e-05,
+      "loss": 2.3466,
+      "step": 2158
+    },
+    {
+      "epoch": 0.41562191688524197,
+      "grad_norm": 2.9261309163208478,
+      "learning_rate": 1.8387158484656603e-05,
+      "loss": 2.3725,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "grad_norm": 2.70597099804516,
+      "learning_rate": 1.838546016621564e-05,
+      "loss": 2.4081,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.4882,
+      "step": 2160,
+      "vm_loss": 0.2385
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.1829,
+      "step": 2160,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.3456,
+      "step": 2160,
+      "vm_loss": 0.2613
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 1.9341,
+      "step": 2160,
+      "vm_loss": 0.1964
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.2251,
+      "step": 2160,
+      "vm_loss": 0.165
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.4087,
+      "step": 2160,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.0562,
+      "step": 2160,
+      "vm_loss": 0.2364
+    },
+    {
+      "epoch": 0.41581442356281734,
+      "lm_loss": 2.2579,
+      "step": 2160,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.4160069302403927,
+      "grad_norm": 3.08532306126558,
+      "learning_rate": 1.8383761032600663e-05,
+      "loss": 2.3714,
+      "step": 2161
+    },
+    {
+      "epoch": 0.41619943691796807,
+      "grad_norm": 2.9651120591167808,
+      "learning_rate": 1.838206108397686e-05,
+      "loss": 2.363,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4163919435955435,
+      "grad_norm": 3.1146431347665966,
+      "learning_rate": 1.838036032050948e-05,
+      "loss": 2.3569,
+      "step": 2163
+    },
+    {
+      "epoch": 0.41658445027311886,
+      "grad_norm": 3.288019650218014,
+      "learning_rate": 1.8378658742363862e-05,
+      "loss": 2.3746,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4167769569506942,
+      "grad_norm": 2.9221016590665103,
+      "learning_rate": 1.837695634970542e-05,
+      "loss": 2.3388,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4169694636282696,
+      "grad_norm": 2.841662853804467,
+      "learning_rate": 1.8375253142699646e-05,
+      "loss": 2.3571,
+      "step": 2166
+    },
+    {
+      "epoch": 0.417161970305845,
+      "grad_norm": 2.732405558606511,
+      "learning_rate": 1.8373549121512118e-05,
+      "loss": 2.4483,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "grad_norm": 3.1732136627728647,
+      "learning_rate": 1.8371844286308486e-05,
+      "loss": 2.3182,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.1019,
+      "step": 2168,
+      "vm_loss": 0.2212
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.1206,
+      "step": 2168,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.4115,
+      "step": 2168,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.3951,
+      "step": 2168,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.4041,
+      "step": 2168,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.2038,
+      "step": 2168,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.0637,
+      "step": 2168,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.4173544769834204,
+      "lm_loss": 2.1314,
+      "step": 2168,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 0.41754698366099574,
+      "grad_norm": 2.9700775384797655,
+      "learning_rate": 1.837013863725448e-05,
+      "loss": 2.3298,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4177394903385711,
+      "grad_norm": 2.7890134844403875,
+      "learning_rate": 1.8368432174515915e-05,
+      "loss": 2.34,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4179319970161465,
+      "grad_norm": 2.5921923600620347,
+      "learning_rate": 1.8366724898258678e-05,
+      "loss": 2.2839,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4181245036937219,
+      "grad_norm": 3.014915369427207,
+      "learning_rate": 1.8365016808648743e-05,
+      "loss": 2.3425,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41831701037129726,
+      "grad_norm": 3.0982445714073212,
+      "learning_rate": 1.8363307905852152e-05,
+      "loss": 2.3935,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4185095170488726,
+      "grad_norm": 2.802939594824342,
+      "learning_rate": 1.836159819003503e-05,
+      "loss": 2.3589,
+      "step": 2174
+    },
+    {
+      "epoch": 0.418702023726448,
+      "grad_norm": 3.0903480637656253,
+      "learning_rate": 1.8359887661363594e-05,
+      "loss": 2.3021,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "grad_norm": 3.066508193015433,
+      "learning_rate": 1.835817632000412e-05,
+      "loss": 2.3484,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 1.724,
+      "step": 2176,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 2.0409,
+      "step": 2176,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 1.9736,
+      "step": 2176,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 2.2024,
+      "step": 2176,
+      "vm_loss": 0.0899
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 1.7809,
+      "step": 2176,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 2.1523,
+      "step": 2176,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 2.161,
+      "step": 2176,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.4188945304040234,
+      "lm_loss": 2.2636,
+      "step": 2176,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.4190870370815988,
+      "grad_norm": 2.642592897201478,
+      "learning_rate": 1.8356464166122976e-05,
+      "loss": 2.3381,
+      "step": 2177
+    },
+    {
+      "epoch": 0.41927954375917414,
+      "grad_norm": 3.254076077578234,
+      "learning_rate": 1.8354751199886608e-05,
+      "loss": 2.3801,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4194720504367495,
+      "grad_norm": 2.6657664927068643,
+      "learning_rate": 1.8353037421461527e-05,
+      "loss": 2.3021,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4196645571143249,
+      "grad_norm": 2.7003895211298263,
+      "learning_rate": 1.835132283101435e-05,
+      "loss": 2.3323,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4198570637919003,
+      "grad_norm": 2.9357196498250473,
+      "learning_rate": 1.8349607428711747e-05,
+      "loss": 2.3956,
+      "step": 2181
+    },
+    {
+      "epoch": 0.42004957046947566,
+      "grad_norm": 3.4185324094915193,
+      "learning_rate": 1.8347891214720477e-05,
+      "loss": 2.3317,
+      "step": 2182
+    },
+    {
+      "epoch": 0.42024207714705103,
+      "grad_norm": 2.6587801175604144,
+      "learning_rate": 1.834617418920738e-05,
+      "loss": 2.3195,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "grad_norm": 3.0276757496206477,
+      "learning_rate": 1.8344456352339377e-05,
+      "loss": 2.3734,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.2437,
+      "step": 2184,
+      "vm_loss": 0.2399
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.0155,
+      "step": 2184,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.2087,
+      "step": 2184,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.1535,
+      "step": 2184,
+      "vm_loss": 0.1153
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 1.9832,
+      "step": 2184,
+      "vm_loss": 0.2594
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.1844,
+      "step": 2184,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.0304,
+      "step": 2184,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.4204345838246264,
+      "lm_loss": 2.2383,
+      "step": 2184,
+      "vm_loss": 0.2531
+    },
+    {
+      "epoch": 0.4206270905022018,
+      "grad_norm": 2.695100763528606,
+      "learning_rate": 1.8342737704283458e-05,
+      "loss": 2.3624,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4208195971797772,
+      "grad_norm": 2.761006169969045,
+      "learning_rate": 1.8341018245206696e-05,
+      "loss": 2.3488,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42101210385735255,
+      "grad_norm": 3.156939382528445,
+      "learning_rate": 1.8339297975276254e-05,
+      "loss": 2.3282,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4212046105349279,
+      "grad_norm": 2.8258156355453723,
+      "learning_rate": 1.8337576894659353e-05,
+      "loss": 2.3752,
+      "step": 2188
+    },
+    {
+      "epoch": 0.42139711721250334,
+      "grad_norm": 2.8959402619317336,
+      "learning_rate": 1.833585500352331e-05,
+      "loss": 2.3784,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4215896238900787,
+      "grad_norm": 3.1410318379606497,
+      "learning_rate": 1.8334132302035515e-05,
+      "loss": 2.36,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42178213056765407,
+      "grad_norm": 2.7508664077296543,
+      "learning_rate": 1.8332408790363437e-05,
+      "loss": 2.3315,
+      "step": 2191
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "grad_norm": 2.859582371003272,
+      "learning_rate": 1.833068446867462e-05,
+      "loss": 2.3644,
+      "step": 2192
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.3192,
+      "step": 2192,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.294,
+      "step": 2192,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.0959,
+      "step": 2192,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 1.9382,
+      "step": 2192,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.3338,
+      "step": 2192,
+      "vm_loss": 0.2247
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.2903,
+      "step": 2192,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.1545,
+      "step": 2192,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.42197463724522943,
+      "lm_loss": 2.1422,
+      "step": 2192,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.4221671439228048,
+      "grad_norm": 2.8583333800590656,
+      "learning_rate": 1.8328959337136692e-05,
+      "loss": 2.3448,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4223596506003802,
+      "grad_norm": 2.98799330395155,
+      "learning_rate": 1.832723339591736e-05,
+      "loss": 2.3576,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4225521572779556,
+      "grad_norm": 2.6870081808167705,
+      "learning_rate": 1.8325506645184402e-05,
+      "loss": 2.3232,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42274466395553095,
+      "grad_norm": 3.203068497228845,
+      "learning_rate": 1.8323779085105684e-05,
+      "loss": 2.3231,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4229371706331063,
+      "grad_norm": 3.0855690506625506,
+      "learning_rate": 1.8322050715849152e-05,
+      "loss": 2.3168,
+      "step": 2197
+    },
+    {
+      "epoch": 0.42312967731068174,
+      "grad_norm": 2.655088944444877,
+      "learning_rate": 1.8320321537582815e-05,
+      "loss": 2.3434,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4233221839882571,
+      "grad_norm": 3.078639542604538,
+      "learning_rate": 1.831859155047478e-05,
+      "loss": 2.3247,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "grad_norm": 3.3473765595244505,
+      "learning_rate": 1.8316860754693215e-05,
+      "loss": 2.327,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.075,
+      "step": 2200,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 1.7723,
+      "step": 2200,
+      "vm_loss": 0.0993
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.2,
+      "step": 2200,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.2439,
+      "step": 2200,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.2557,
+      "step": 2200,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.1867,
+      "step": 2200,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 2.3099,
+      "step": 2200,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.42351469066583247,
+      "lm_loss": 1.9626,
+      "step": 2200,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 0.42370719734340784,
+      "grad_norm": 2.8337305175869747,
+      "learning_rate": 1.8315129150406386e-05,
+      "loss": 2.3027,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4238997040209832,
+      "grad_norm": 2.9449433634328037,
+      "learning_rate": 1.831339673778262e-05,
+      "loss": 2.255,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4240922106985586,
+      "grad_norm": 2.905475099765251,
+      "learning_rate": 1.8311663516990332e-05,
+      "loss": 2.3381,
+      "step": 2203
+    },
+    {
+      "epoch": 0.424284717376134,
+      "grad_norm": 2.9464403100852996,
+      "learning_rate": 1.8309929488198012e-05,
+      "loss": 2.3072,
+      "step": 2204
+    },
+    {
+      "epoch": 0.42447722405370936,
+      "grad_norm": 3.0311573399879497,
+      "learning_rate": 1.8308194651574232e-05,
+      "loss": 2.3421,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4246697307312847,
+      "grad_norm": 2.7371020985877963,
+      "learning_rate": 1.8306459007287637e-05,
+      "loss": 2.3742,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42486223740886014,
+      "grad_norm": 2.6068674367419216,
+      "learning_rate": 1.830472255550696e-05,
+      "loss": 2.3886,
+      "step": 2207
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "grad_norm": 2.9373892553700953,
+      "learning_rate": 1.8302985296401e-05,
+      "loss": 2.3709,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.0728,
+      "step": 2208,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.2847,
+      "step": 2208,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.5463,
+      "step": 2208,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 1.9467,
+      "step": 2208,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.021,
+      "step": 2208,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.2198,
+      "step": 2208,
+      "vm_loss": 0.1148
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 1.8601,
+      "step": 2208,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.4250547440864355,
+      "lm_loss": 2.129,
+      "step": 2208,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 0.4252472507640109,
+      "grad_norm": 2.9983140048511654,
+      "learning_rate": 1.830124723013864e-05,
+      "loss": 2.2861,
+      "step": 2209
+    },
+    {
+      "epoch": 0.42543975744158624,
+      "grad_norm": 2.730833474461183,
+      "learning_rate": 1.829950835688885e-05,
+      "loss": 2.3471,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4256322641191616,
+      "grad_norm": 2.862369895340129,
+      "learning_rate": 1.8297768676820665e-05,
+      "loss": 2.3215,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42582477079673703,
+      "grad_norm": 3.6304801019321498,
+      "learning_rate": 1.8296028190103205e-05,
+      "loss": 2.2975,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4260172774743124,
+      "grad_norm": 3.3409308437570027,
+      "learning_rate": 1.8294286896905665e-05,
+      "loss": 2.3725,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42620978415188776,
+      "grad_norm": 2.8043646591601634,
+      "learning_rate": 1.8292544797397327e-05,
+      "loss": 2.3411,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4264022908294631,
+      "grad_norm": 2.7328741821076625,
+      "learning_rate": 1.829080189174754e-05,
+      "loss": 2.3302,
+      "step": 2215
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "grad_norm": 2.900396750536101,
+      "learning_rate": 1.828905818012574e-05,
+      "loss": 2.3147,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.2344,
+      "step": 2216,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.4235,
+      "step": 2216,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.2181,
+      "step": 2216,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.3286,
+      "step": 2216,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.1762,
+      "step": 2216,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.0343,
+      "step": 2216,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.3713,
+      "step": 2216,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.42659479750703855,
+      "lm_loss": 2.3332,
+      "step": 2216,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.4267873041846139,
+      "grad_norm": 2.8823395153204,
+      "learning_rate": 1.8287313662701436e-05,
+      "loss": 2.3646,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4269798108621893,
+      "grad_norm": 3.139119355230146,
+      "learning_rate": 1.828556833964422e-05,
+      "loss": 2.3537,
+      "step": 2218
+    },
+    {
+      "epoch": 0.42717231753976465,
+      "grad_norm": 2.4723202448121073,
+      "learning_rate": 1.8283822211123758e-05,
+      "loss": 2.324,
+      "step": 2219
+    },
+    {
+      "epoch": 0.42736482421734,
+      "grad_norm": 3.01517865352798,
+      "learning_rate": 1.8282075277309792e-05,
+      "loss": 2.3434,
+      "step": 2220
+    },
+    {
+      "epoch": 0.42755733089491543,
+      "grad_norm": 2.9302152898677787,
+      "learning_rate": 1.8280327538372155e-05,
+      "loss": 2.3382,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4277498375724908,
+      "grad_norm": 2.6665277903917004,
+      "learning_rate": 1.827857899448074e-05,
+      "loss": 2.2625,
+      "step": 2222
+    },
+    {
+      "epoch": 0.42794234425006616,
+      "grad_norm": 2.8370088248688377,
+      "learning_rate": 1.8276829645805536e-05,
+      "loss": 2.3409,
+      "step": 2223
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "grad_norm": 2.7110791506392107,
+      "learning_rate": 1.8275079492516597e-05,
+      "loss": 2.3519,
+      "step": 2224
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 2.1767,
+      "step": 2224,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 1.8086,
+      "step": 2224,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 1.9442,
+      "step": 2224,
+      "vm_loss": 0.1005
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 2.2649,
+      "step": 2224,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 2.1648,
+      "step": 2224,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 2.4827,
+      "step": 2224,
+      "vm_loss": 0.1812
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 2.0415,
+      "step": 2224,
+      "vm_loss": 0.2364
+    },
+    {
+      "epoch": 0.42813485092764153,
+      "lm_loss": 1.9691,
+      "step": 2224,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 0.42832735760521695,
+      "grad_norm": 2.9401484778077047,
+      "learning_rate": 1.8273328534784067e-05,
+      "loss": 2.3657,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4285198642827923,
+      "grad_norm": 2.83229555133251,
+      "learning_rate": 1.8271576772778154e-05,
+      "loss": 2.3054,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4287123709603677,
+      "grad_norm": 3.1111944626892387,
+      "learning_rate": 1.8269824206669153e-05,
+      "loss": 2.3393,
+      "step": 2227
+    },
+    {
+      "epoch": 0.42890487763794305,
+      "grad_norm": 3.3099832420602393,
+      "learning_rate": 1.826807083662744e-05,
+      "loss": 2.3056,
+      "step": 2228
+    },
+    {
+      "epoch": 0.42909738431551847,
+      "grad_norm": 2.605118641407932,
+      "learning_rate": 1.8266316662823456e-05,
+      "loss": 2.3289,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42928989099309384,
+      "grad_norm": 3.0330197853056746,
+      "learning_rate": 1.826456168542774e-05,
+      "loss": 2.3253,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4294823976706692,
+      "grad_norm": 3.017107128608619,
+      "learning_rate": 1.8262805904610894e-05,
+      "loss": 2.3361,
+      "step": 2231
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "grad_norm": 2.9297229241573666,
+      "learning_rate": 1.82610493205436e-05,
+      "loss": 2.2937,
+      "step": 2232
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 1.8694,
+      "step": 2232,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 2.0562,
+      "step": 2232,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 1.9866,
+      "step": 2232,
+      "vm_loss": 0.085
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 2.0144,
+      "step": 2232,
+      "vm_loss": 0.1182
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 2.4143,
+      "step": 2232,
+      "vm_loss": 0.1122
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 2.2187,
+      "step": 2232,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 1.9456,
+      "step": 2232,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.42967490434824457,
+      "lm_loss": 2.1749,
+      "step": 2232,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 0.42986741102581993,
+      "grad_norm": 3.4703806471897813,
+      "learning_rate": 1.8259291933396625e-05,
+      "loss": 2.31,
+      "step": 2233
+    },
+    {
+      "epoch": 0.43005991770339536,
+      "grad_norm": 3.2460382714743212,
+      "learning_rate": 1.8257533743340806e-05,
+      "loss": 2.3453,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4302524243809707,
+      "grad_norm": 2.7981051424157295,
+      "learning_rate": 1.825577475054706e-05,
+      "loss": 2.3402,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4304449310585461,
+      "grad_norm": 3.0666500194026325,
+      "learning_rate": 1.8254014955186386e-05,
+      "loss": 2.3343,
+      "step": 2236
+    },
+    {
+      "epoch": 0.43063743773612145,
+      "grad_norm": 3.473243301255162,
+      "learning_rate": 1.825225435742986e-05,
+      "loss": 2.3174,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4308299444136969,
+      "grad_norm": 2.825915484416556,
+      "learning_rate": 1.8250492957448634e-05,
+      "loss": 2.3386,
+      "step": 2238
+    },
+    {
+      "epoch": 0.43102245109127224,
+      "grad_norm": 2.813152685775438,
+      "learning_rate": 1.824873075541394e-05,
+      "loss": 2.2796,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "grad_norm": 3.644551671361753,
+      "learning_rate": 1.8246967751497083e-05,
+      "loss": 2.3409,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 2.2338,
+      "step": 2240,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 2.4392,
+      "step": 2240,
+      "vm_loss": 0.1084
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 1.8853,
+      "step": 2240,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 2.2702,
+      "step": 2240,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 2.4741,
+      "step": 2240,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 1.8013,
+      "step": 2240,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 1.9907,
+      "step": 2240,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 0.4312149577688476,
+      "lm_loss": 2.1314,
+      "step": 2240,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 0.431407464446423,
+      "grad_norm": 2.5496307794075985,
+      "learning_rate": 1.824520394586945e-05,
+      "loss": 2.3183,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43159997112399834,
+      "grad_norm": 3.1005403403092853,
+      "learning_rate": 1.8243439338702504e-05,
+      "loss": 2.3601,
+      "step": 2242
+    },
+    {
+      "epoch": 0.43179247780157376,
+      "grad_norm": 2.880706386710158,
+      "learning_rate": 1.8241673930167792e-05,
+      "loss": 2.334,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4319849844791491,
+      "grad_norm": 2.7386247784669826,
+      "learning_rate": 1.8239907720436935e-05,
+      "loss": 2.3091,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4321774911567245,
+      "grad_norm": 2.720028088706721,
+      "learning_rate": 1.8238140709681625e-05,
+      "loss": 2.2935,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43236999783429986,
+      "grad_norm": 2.783310393714633,
+      "learning_rate": 1.8236372898073642e-05,
+      "loss": 2.3403,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4325625045118753,
+      "grad_norm": 3.0276080413209896,
+      "learning_rate": 1.823460428578484e-05,
+      "loss": 2.384,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "grad_norm": 2.9225361694585104,
+      "learning_rate": 1.8232834872987147e-05,
+      "loss": 2.2671,
+      "step": 2248
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.2045,
+      "step": 2248,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.195,
+      "step": 2248,
+      "vm_loss": 0.1143
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 1.9849,
+      "step": 2248,
+      "vm_loss": 0.1108
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 1.9011,
+      "step": 2248,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.3294,
+      "step": 2248,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.4616,
+      "step": 2248,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.1318,
+      "step": 2248,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 0.43275501118945064,
+      "lm_loss": 2.3315,
+      "step": 2248,
+      "vm_loss": 0.1121
+    },
+    {
+      "epoch": 0.432947517867026,
+      "grad_norm": 2.662969001880157,
+      "learning_rate": 1.823106465985258e-05,
+      "loss": 2.3476,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4331400245446014,
+      "grad_norm": 2.8890921929828735,
+      "learning_rate": 1.8229293646553224e-05,
+      "loss": 2.3723,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43333253122217674,
+      "grad_norm": 2.954219283955891,
+      "learning_rate": 1.8227521833261237e-05,
+      "loss": 2.3179,
+      "step": 2251
+    },
+    {
+      "epoch": 0.43352503789975216,
+      "grad_norm": 2.7749063882076124,
+      "learning_rate": 1.822574922014887e-05,
+      "loss": 2.3006,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43371754457732753,
+      "grad_norm": 2.743081860071622,
+      "learning_rate": 1.822397580738844e-05,
+      "loss": 2.313,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4339100512549029,
+      "grad_norm": 2.5719047175625467,
+      "learning_rate": 1.822220159515234e-05,
+      "loss": 2.2858,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43410255793247826,
+      "grad_norm": 2.9392031442512336,
+      "learning_rate": 1.8220426583613063e-05,
+      "loss": 2.3376,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "grad_norm": 3.0591530218417593,
+      "learning_rate": 1.8218650772943144e-05,
+      "loss": 2.3579,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.0416,
+      "step": 2256,
+      "vm_loss": 0.1077
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.0622,
+      "step": 2256,
+      "vm_loss": 0.2354
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.0739,
+      "step": 2256,
+      "vm_loss": 0.2172
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.121,
+      "step": 2256,
+      "vm_loss": 0.0901
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.3271,
+      "step": 2256,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.247,
+      "step": 2256,
+      "vm_loss": 0.2204
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.4194,
+      "step": 2256,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.4342950646100537,
+      "lm_loss": 2.342,
+      "step": 2256,
+      "vm_loss": 0.2323
+    },
+    {
+      "epoch": 0.43448757128762905,
+      "grad_norm": 2.7995848004046953,
+      "learning_rate": 1.8216874163315227e-05,
+      "loss": 2.336,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4346800779652044,
+      "grad_norm": 2.763459608904795,
+      "learning_rate": 1.8215096754902014e-05,
+      "loss": 2.3168,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4348725846427798,
+      "grad_norm": 3.2027606616981004,
+      "learning_rate": 1.82133185478763e-05,
+      "loss": 2.3105,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4350650913203552,
+      "grad_norm": 2.9535160926563018,
+      "learning_rate": 1.8211539542410938e-05,
+      "loss": 2.33,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43525759799793057,
+      "grad_norm": 2.765769406785263,
+      "learning_rate": 1.8209759738678877e-05,
+      "loss": 2.3622,
+      "step": 2261
+    },
+    {
+      "epoch": 0.43545010467550593,
+      "grad_norm": 2.8346286786172663,
+      "learning_rate": 1.8207979136853136e-05,
+      "loss": 2.3126,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4356426113530813,
+      "grad_norm": 2.7849536479320354,
+      "learning_rate": 1.8206197737106818e-05,
+      "loss": 2.3096,
+      "step": 2263
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "grad_norm": 3.103087349009926,
+      "learning_rate": 1.8204415539613085e-05,
+      "loss": 2.2904,
+      "step": 2264
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.2558,
+      "step": 2264,
+      "vm_loss": 0.2054
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.0647,
+      "step": 2264,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.2168,
+      "step": 2264,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.2076,
+      "step": 2264,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.1873,
+      "step": 2264,
+      "vm_loss": 0.2243
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.2121,
+      "step": 2264,
+      "vm_loss": 0.237
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.4772,
+      "step": 2264,
+      "vm_loss": 0.1023
+    },
+    {
+      "epoch": 0.43583511803065667,
+      "lm_loss": 2.1969,
+      "step": 2264,
+      "vm_loss": 0.101
+    },
+    {
+      "epoch": 0.4360276247082321,
+      "grad_norm": 2.9321325294492033,
+      "learning_rate": 1.82026325445452e-05,
+      "loss": 2.3591,
+      "step": 2265
+    },
+    {
+      "epoch": 0.43622013138580745,
+      "grad_norm": 2.7059144075188124,
+      "learning_rate": 1.8200848752076486e-05,
+      "loss": 2.329,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4364126380633828,
+      "grad_norm": 2.980462799239122,
+      "learning_rate": 1.8199064162380358e-05,
+      "loss": 2.3358,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4366051447409582,
+      "grad_norm": 2.7277992230655874,
+      "learning_rate": 1.8197278775630295e-05,
+      "loss": 2.26,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4367976514185336,
+      "grad_norm": 2.8509979052657264,
+      "learning_rate": 1.819549259199986e-05,
+      "loss": 2.2831,
+      "step": 2269
+    },
+    {
+      "epoch": 0.43699015809610897,
+      "grad_norm": 3.0805412306199207,
+      "learning_rate": 1.8193705611662697e-05,
+      "loss": 2.3273,
+      "step": 2270
+    },
+    {
+      "epoch": 0.43718266477368434,
+      "grad_norm": 2.8817601843232405,
+      "learning_rate": 1.8191917834792516e-05,
+      "loss": 2.3425,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "grad_norm": 2.785495287162447,
+      "learning_rate": 1.819012926156312e-05,
+      "loss": 2.3611,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 1.9766,
+      "step": 2272,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 2.276,
+      "step": 2272,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 1.8094,
+      "step": 2272,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 2.1238,
+      "step": 2272,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 1.9581,
+      "step": 2272,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 1.9666,
+      "step": 2272,
+      "vm_loss": 0.1793
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 2.2968,
+      "step": 2272,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 0.4373751714512597,
+      "lm_loss": 2.2719,
+      "step": 2272,
+      "vm_loss": 0.1804
+    },
+    {
+      "epoch": 0.43756767812883507,
+      "grad_norm": 3.075200229363919,
+      "learning_rate": 1.8188339892148375e-05,
+      "loss": 2.3117,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4377601848064105,
+      "grad_norm": 2.450610527956838,
+      "learning_rate": 1.8186549726722234e-05,
+      "loss": 2.2776,
+      "step": 2274
+    },
+    {
+      "epoch": 0.43795269148398586,
+      "grad_norm": 2.6649109372800104,
+      "learning_rate": 1.8184758765458722e-05,
+      "loss": 2.3495,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4381451981615612,
+      "grad_norm": 2.797810230039664,
+      "learning_rate": 1.8182967008531948e-05,
+      "loss": 2.3035,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4383377048391366,
+      "grad_norm": 3.0747072507211497,
+      "learning_rate": 1.8181174456116085e-05,
+      "loss": 2.3527,
+      "step": 2277
+    },
+    {
+      "epoch": 0.438530211516712,
+      "grad_norm": 3.1805540217005284,
+      "learning_rate": 1.81793811083854e-05,
+      "loss": 2.2976,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4387227181942874,
+      "grad_norm": 2.8465572133598633,
+      "learning_rate": 1.817758696551423e-05,
+      "loss": 2.2874,
+      "step": 2279
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "grad_norm": 2.966299399160946,
+      "learning_rate": 1.817579202767698e-05,
+      "loss": 2.3541,
+      "step": 2280
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.1517,
+      "step": 2280,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.3433,
+      "step": 2280,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.0524,
+      "step": 2280,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 1.9284,
+      "step": 2280,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.0675,
+      "step": 2280,
+      "vm_loss": 0.2603
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.1306,
+      "step": 2280,
+      "vm_loss": 0.2119
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 2.1933,
+      "step": 2280,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 0.43891522487186274,
+      "lm_loss": 1.9605,
+      "step": 2280,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 0.4391077315494381,
+      "grad_norm": 2.90105025988649,
+      "learning_rate": 1.8173996295048147e-05,
+      "loss": 2.2911,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4393002382270135,
+      "grad_norm": 3.066794446216245,
+      "learning_rate": 1.8172199767802298e-05,
+      "loss": 2.318,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4394927449045889,
+      "grad_norm": 2.920750217805513,
+      "learning_rate": 1.817040244611408e-05,
+      "loss": 2.362,
+      "step": 2283
+    },
+    {
+      "epoch": 0.43968525158216426,
+      "grad_norm": 2.709972371828956,
+      "learning_rate": 1.816860433015821e-05,
+      "loss": 2.3416,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4398777582597396,
+      "grad_norm": 2.6873289268547413,
+      "learning_rate": 1.8166805420109498e-05,
+      "loss": 2.3161,
+      "step": 2285
+    },
+    {
+      "epoch": 0.440070264937315,
+      "grad_norm": 3.1716789181957035,
+      "learning_rate": 1.8165005716142813e-05,
+      "loss": 2.2905,
+      "step": 2286
+    },
+    {
+      "epoch": 0.4402627716148904,
+      "grad_norm": 2.8882577886129845,
+      "learning_rate": 1.816320521843311e-05,
+      "loss": 2.3165,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "grad_norm": 2.9140995543471773,
+      "learning_rate": 1.8161403927155424e-05,
+      "loss": 2.2568,
+      "step": 2288
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.363,
+      "step": 2288,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.1881,
+      "step": 2288,
+      "vm_loss": 0.1181
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.0624,
+      "step": 2288,
+      "vm_loss": 0.2538
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 1.8617,
+      "step": 2288,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.3127,
+      "step": 2288,
+      "vm_loss": 0.1178
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.3468,
+      "step": 2288,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 2.2677,
+      "step": 2288,
+      "vm_loss": 0.0928
+    },
+    {
+      "epoch": 0.4404552782924658,
+      "lm_loss": 1.846,
+      "step": 2288,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.44064778497004115,
+      "grad_norm": 2.674997057544507,
+      "learning_rate": 1.815960184248486e-05,
+      "loss": 2.2859,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4408402916476165,
+      "grad_norm": 2.95699712330038,
+      "learning_rate": 1.8157798964596606e-05,
+      "loss": 2.336,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44103279832519193,
+      "grad_norm": 2.773987803306304,
+      "learning_rate": 1.8155995293665923e-05,
+      "loss": 2.31,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4412253050027673,
+      "grad_norm": 2.8267279313377456,
+      "learning_rate": 1.8154190829868152e-05,
+      "loss": 2.2767,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44141781168034266,
+      "grad_norm": 3.316062036862609,
+      "learning_rate": 1.815238557337871e-05,
+      "loss": 2.3168,
+      "step": 2293
+    },
+    {
+      "epoch": 0.44161031835791803,
+      "grad_norm": 2.975626667407499,
+      "learning_rate": 1.815057952437309e-05,
+      "loss": 2.3367,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4418028250354934,
+      "grad_norm": 3.1478180106540403,
+      "learning_rate": 1.8148772683026865e-05,
+      "loss": 2.3538,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "grad_norm": 3.1141203666257544,
+      "learning_rate": 1.8146965049515684e-05,
+      "loss": 2.2515,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.4991,
+      "step": 2296,
+      "vm_loss": 0.2327
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.552,
+      "step": 2296,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 1.3974,
+      "step": 2296,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.2506,
+      "step": 2296,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.2569,
+      "step": 2296,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.1856,
+      "step": 2296,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.3442,
+      "step": 2296,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 0.4419953317130688,
+      "lm_loss": 2.2605,
+      "step": 2296,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 0.4421878383906442,
+      "grad_norm": 2.592237520048315,
+      "learning_rate": 1.8145156624015267e-05,
+      "loss": 2.2863,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44238034506821955,
+      "grad_norm": 3.112368654795366,
+      "learning_rate": 1.8143347406701418e-05,
+      "loss": 2.2812,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4425728517457949,
+      "grad_norm": 3.0672581311328173,
+      "learning_rate": 1.814153739775002e-05,
+      "loss": 2.34,
+      "step": 2299
+    },
+    {
+      "epoch": 0.44276535842337034,
+      "grad_norm": 2.7153923012881567,
+      "learning_rate": 1.8139726597337026e-05,
+      "loss": 2.289,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4429578651009457,
+      "grad_norm": 3.0565394236394954,
+      "learning_rate": 1.813791500563847e-05,
+      "loss": 2.2871,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44315037177852107,
+      "grad_norm": 3.367596758507889,
+      "learning_rate": 1.8136102622830458e-05,
+      "loss": 2.3192,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44334287845609643,
+      "grad_norm": 2.8463625264033836,
+      "learning_rate": 1.8134289449089184e-05,
+      "loss": 2.3273,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "grad_norm": 2.698690460089316,
+      "learning_rate": 1.8132475484590905e-05,
+      "loss": 2.357,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.1319,
+      "step": 2304,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 1.8273,
+      "step": 2304,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.1133,
+      "step": 2304,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.1563,
+      "step": 2304,
+      "vm_loss": 0.1183
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.2225,
+      "step": 2304,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.2088,
+      "step": 2304,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 1.6983,
+      "step": 2304,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 0.4435353851336718,
+      "lm_loss": 2.0593,
+      "step": 2304,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.4437278918112472,
+      "grad_norm": 3.00803392508199,
+      "learning_rate": 1.8130660729511964e-05,
+      "loss": 2.2784,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4439203984888226,
+      "grad_norm": 3.049338533735284,
+      "learning_rate": 1.812884518402878e-05,
+      "loss": 2.3778,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44411290516639795,
+      "grad_norm": 2.5404105702401845,
+      "learning_rate": 1.8127028848317843e-05,
+      "loss": 2.3092,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4443054118439733,
+      "grad_norm": 3.0215819518567986,
+      "learning_rate": 1.812521172255573e-05,
+      "loss": 2.2888,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44449791852154874,
+      "grad_norm": 2.9535979655659346,
+      "learning_rate": 1.8123393806919084e-05,
+      "loss": 2.3081,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4446904251991241,
+      "grad_norm": 2.895735213659374,
+      "learning_rate": 1.8121575101584632e-05,
+      "loss": 2.3651,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4448829318766995,
+      "grad_norm": 2.687726853708876,
+      "learning_rate": 1.8119755606729174e-05,
+      "loss": 2.2924,
+      "step": 2311
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "grad_norm": 2.693461560074217,
+      "learning_rate": 1.811793532252959e-05,
+      "loss": 2.294,
+      "step": 2312
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.2555,
+      "step": 2312,
+      "vm_loss": 0.2395
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 1.9768,
+      "step": 2312,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.1486,
+      "step": 2312,
+      "vm_loss": 0.1173
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.0378,
+      "step": 2312,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.1273,
+      "step": 2312,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.331,
+      "step": 2312,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.2891,
+      "step": 2312,
+      "vm_loss": 0.116
+    },
+    {
+      "epoch": 0.44507543855427484,
+      "lm_loss": 2.2291,
+      "step": 2312,
+      "vm_loss": 0.1412
+    },
+    {
+      "epoch": 0.4452679452318502,
+      "grad_norm": 3.2736052081237696,
+      "learning_rate": 1.811611424916283e-05,
+      "loss": 2.3059,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4454604519094256,
+      "grad_norm": 2.881535923317366,
+      "learning_rate": 1.8114292386805935e-05,
+      "loss": 2.3288,
+      "step": 2314
+    },
+    {
+      "epoch": 0.445652958587001,
+      "grad_norm": 3.0517008846081013,
+      "learning_rate": 1.8112469735636007e-05,
+      "loss": 2.2992,
+      "step": 2315
+    },
+    {
+      "epoch": 0.44584546526457636,
+      "grad_norm": 3.195731863253457,
+      "learning_rate": 1.8110646295830233e-05,
+      "loss": 2.313,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4460379719421517,
+      "grad_norm": 2.78787429459504,
+      "learning_rate": 1.810882206756587e-05,
+      "loss": 2.3179,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44623047861972714,
+      "grad_norm": 3.1160825668708614,
+      "learning_rate": 1.810699705102026e-05,
+      "loss": 2.2568,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4464229852973025,
+      "grad_norm": 3.008668873403132,
+      "learning_rate": 1.810517124637082e-05,
+      "loss": 2.2848,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "grad_norm": 2.9730995722507965,
+      "learning_rate": 1.8103344653795042e-05,
+      "loss": 2.3607,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.2469,
+      "step": 2320,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.3266,
+      "step": 2320,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.2016,
+      "step": 2320,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 1.9088,
+      "step": 2320,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.1452,
+      "step": 2320,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.2678,
+      "step": 2320,
+      "vm_loss": 0.2041
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 2.3045,
+      "step": 2320,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.4466154919748779,
+      "lm_loss": 1.8293,
+      "step": 2320,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 0.44680799865245324,
+      "grad_norm": 3.138343527926431,
+      "learning_rate": 1.8101517273470485e-05,
+      "loss": 2.3157,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4470005053300286,
+      "grad_norm": 2.817627156353456,
+      "learning_rate": 1.8099689105574805e-05,
+      "loss": 2.33,
+      "step": 2322
+    },
+    {
+      "epoch": 0.44719301200760403,
+      "grad_norm": 3.0224937962314007,
+      "learning_rate": 1.8097860150285717e-05,
+      "loss": 2.2621,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4473855186851794,
+      "grad_norm": 3.0258571758902186,
+      "learning_rate": 1.809603040778102e-05,
+      "loss": 2.2688,
+      "step": 2324
+    },
+    {
+      "epoch": 0.44757802536275476,
+      "grad_norm": 2.8483829903936484,
+      "learning_rate": 1.809419987823859e-05,
+      "loss": 2.2751,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4477705320403301,
+      "grad_norm": 2.80689706768817,
+      "learning_rate": 1.8092368561836377e-05,
+      "loss": 2.2937,
+      "step": 2326
+    },
+    {
+      "epoch": 0.44796303871790555,
+      "grad_norm": 3.277412781913552,
+      "learning_rate": 1.809053645875241e-05,
+      "loss": 2.2877,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "grad_norm": 2.9159806121008414,
+      "learning_rate": 1.808870356916479e-05,
+      "loss": 2.3153,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 2.2425,
+      "step": 2328,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 1.8858,
+      "step": 2328,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 2.0504,
+      "step": 2328,
+      "vm_loss": 0.1112
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 1.9213,
+      "step": 2328,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 1.8681,
+      "step": 2328,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 2.1767,
+      "step": 2328,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 2.2315,
+      "step": 2328,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 0.4481555453954809,
+      "lm_loss": 2.3613,
+      "step": 2328,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 0.4483480520730563,
+      "grad_norm": 3.2099408324439365,
+      "learning_rate": 1.8086869893251693e-05,
+      "loss": 2.3188,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44854055875063165,
+      "grad_norm": 2.7266249885205487,
+      "learning_rate": 1.8085035431191388e-05,
+      "loss": 2.2785,
+      "step": 2330
+    },
+    {
+      "epoch": 0.44873306542820707,
+      "grad_norm": 3.1390339205874875,
+      "learning_rate": 1.8083200183162198e-05,
+      "loss": 2.3798,
+      "step": 2331
+    },
+    {
+      "epoch": 0.44892557210578243,
+      "grad_norm": 2.9238563664288826,
+      "learning_rate": 1.8081364149342535e-05,
+      "loss": 2.3034,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4491180787833578,
+      "grad_norm": 2.9322900488499895,
+      "learning_rate": 1.8079527329910885e-05,
+      "loss": 2.2561,
+      "step": 2333
+    },
+    {
+      "epoch": 0.44931058546093317,
+      "grad_norm": 2.8551518564765472,
+      "learning_rate": 1.8077689725045818e-05,
+      "loss": 2.3207,
+      "step": 2334
+    },
+    {
+      "epoch": 0.44950309213850853,
+      "grad_norm": 2.8620785893899843,
+      "learning_rate": 1.807585133492596e-05,
+      "loss": 2.2502,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "grad_norm": 3.229618433456999,
+      "learning_rate": 1.8074012159730034e-05,
+      "loss": 2.3092,
+      "step": 2336
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.1694,
+      "step": 2336,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.0757,
+      "step": 2336,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.3814,
+      "step": 2336,
+      "vm_loss": 0.1233
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.3128,
+      "step": 2336,
+      "vm_loss": 0.3002
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.1168,
+      "step": 2336,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.3629,
+      "step": 2336,
+      "vm_loss": 0.1138
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.3175,
+      "step": 2336,
+      "vm_loss": 0.2315
+    },
+    {
+      "epoch": 0.44969559881608395,
+      "lm_loss": 2.116,
+      "step": 2336,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.4498881054936593,
+      "grad_norm": 2.8240984073782736,
+      "learning_rate": 1.807217219963683e-05,
+      "loss": 2.3496,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4500806121712347,
+      "grad_norm": 2.7334664084967812,
+      "learning_rate": 1.807033145482521e-05,
+      "loss": 2.2459,
+      "step": 2338
+    },
+    {
+      "epoch": 0.45027311884881005,
+      "grad_norm": 2.994307892206342,
+      "learning_rate": 1.806848992547413e-05,
+      "loss": 2.3126,
+      "step": 2339
+    },
+    {
+      "epoch": 0.45046562552638547,
+      "grad_norm": 2.926334002817865,
+      "learning_rate": 1.8066647611762596e-05,
+      "loss": 2.2942,
+      "step": 2340
+    },
+    {
+      "epoch": 0.45065813220396084,
+      "grad_norm": 2.6010397355706996,
+      "learning_rate": 1.8064804513869716e-05,
+      "loss": 2.2984,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4508506388815362,
+      "grad_norm": 2.9548169531187147,
+      "learning_rate": 1.8062960631974657e-05,
+      "loss": 2.3312,
+      "step": 2342
+    },
+    {
+      "epoch": 0.45104314555911157,
+      "grad_norm": 2.9293199107374934,
+      "learning_rate": 1.806111596625667e-05,
+      "loss": 2.3716,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "grad_norm": 3.0609017607682056,
+      "learning_rate": 1.805927051689508e-05,
+      "loss": 2.2608,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 1.8285,
+      "step": 2344,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.1332,
+      "step": 2344,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.3643,
+      "step": 2344,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.1365,
+      "step": 2344,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.1722,
+      "step": 2344,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.2516,
+      "step": 2344,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.0841,
+      "step": 2344,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.45123565223668693,
+      "lm_loss": 2.2478,
+      "step": 2344,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.45142815891426236,
+      "grad_norm": 2.848659581249523,
+      "learning_rate": 1.8057424284069287e-05,
+      "loss": 2.2831,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4516206655918377,
+      "grad_norm": 3.1172548747918056,
+      "learning_rate": 1.805557726795877e-05,
+      "loss": 2.3554,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4518131722694131,
+      "grad_norm": 2.766966260579383,
+      "learning_rate": 1.805372946874308e-05,
+      "loss": 2.2855,
+      "step": 2347
+    },
+    {
+      "epoch": 0.45200567894698845,
+      "grad_norm": 3.1466402905754913,
+      "learning_rate": 1.8051880886601845e-05,
+      "loss": 2.2876,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4521981856245639,
+      "grad_norm": 2.8340130616752646,
+      "learning_rate": 1.805003152171478e-05,
+      "loss": 2.3182,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45239069230213924,
+      "grad_norm": 3.1059274593784623,
+      "learning_rate": 1.8048181374261656e-05,
+      "loss": 2.256,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4525831989797146,
+      "grad_norm": 2.868401927654121,
+      "learning_rate": 1.804633044442234e-05,
+      "loss": 2.2972,
+      "step": 2351
+    },
+    {
+      "epoch": 0.45277570565729,
+      "grad_norm": 2.855277595725172,
+      "learning_rate": 1.804447873237676e-05,
+      "loss": 2.2832,
+      "step": 2352
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 1.975,
+      "step": 2352,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 2.1758,
+      "step": 2352,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 1.6455,
+      "step": 2352,
+      "vm_loss": 0.1977
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 2.3612,
+      "step": 2352,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 1.8856,
+      "step": 2352,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 2.1037,
+      "step": 2352,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 2.2165,
+      "step": 2352,
+      "vm_loss": 0.1122
+    },
+    {
+      "epoch": 0.45277570565729,
+      "lm_loss": 2.2202,
+      "step": 2352,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 0.45296821233486534,
+      "grad_norm": 2.930887479217807,
+      "learning_rate": 1.804262623830493e-05,
+      "loss": 2.3317,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45316071901244076,
+      "grad_norm": 2.8845706527768464,
+      "learning_rate": 1.8040772962386934e-05,
+      "loss": 2.2809,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4533532256900161,
+      "grad_norm": 2.483947272889894,
+      "learning_rate": 1.8038918904802932e-05,
+      "loss": 2.304,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4535457323675915,
+      "grad_norm": 3.103766313101901,
+      "learning_rate": 1.803706406573317e-05,
+      "loss": 2.2961,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45373823904516686,
+      "grad_norm": 2.773846275925031,
+      "learning_rate": 1.8035208445357955e-05,
+      "loss": 2.3007,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4539307457227423,
+      "grad_norm": 2.931230848773706,
+      "learning_rate": 1.8033352043857677e-05,
+      "loss": 2.2822,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45412325240031765,
+      "grad_norm": 3.020776139813076,
+      "learning_rate": 1.8031494861412804e-05,
+      "loss": 2.2981,
+      "step": 2359
+    },
+    {
+      "epoch": 0.454315759077893,
+      "grad_norm": 3.122936422808837,
+      "learning_rate": 1.8029636898203877e-05,
+      "loss": 2.3055,
+      "step": 2360
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 1.6894,
+      "step": 2360,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.0346,
+      "step": 2360,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.4783,
+      "step": 2360,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.2976,
+      "step": 2360,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 1.8729,
+      "step": 2360,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.1979,
+      "step": 2360,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.3412,
+      "step": 2360,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 0.454315759077893,
+      "lm_loss": 2.0552,
+      "step": 2360,
+      "vm_loss": 0.0952
+    },
+    {
+      "epoch": 0.4545082657554684,
+      "grad_norm": 2.7638142005128223,
+      "learning_rate": 1.8027778154411515e-05,
+      "loss": 2.3154,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4547007724330438,
+      "grad_norm": 3.064304999190845,
+      "learning_rate": 1.802591863021641e-05,
+      "loss": 2.3077,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45489327911061916,
+      "grad_norm": 5.562591732814892,
+      "learning_rate": 1.8024058325799335e-05,
+      "loss": 2.3113,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45508578578819453,
+      "grad_norm": 22.97063855165341,
+      "learning_rate": 1.8022197241341134e-05,
+      "loss": 2.5621,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4552782924657699,
+      "grad_norm": 8.15658944492121,
+      "learning_rate": 1.8020335377022723e-05,
+      "loss": 2.666,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45547079914334526,
+      "grad_norm": 6.85679448293632,
+      "learning_rate": 1.8018472733025108e-05,
+      "loss": 2.7253,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4556633058209207,
+      "grad_norm": 5.327370792129126,
+      "learning_rate": 1.8016609309529353e-05,
+      "loss": 2.6148,
+      "step": 2367
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "grad_norm": 5.658883351664112,
+      "learning_rate": 1.8014745106716612e-05,
+      "loss": 2.5471,
+      "step": 2368
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.2808,
+      "step": 2368,
+      "vm_loss": 0.1296
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.1064,
+      "step": 2368,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.4046,
+      "step": 2368,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.391,
+      "step": 2368,
+      "vm_loss": 0.1135
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.2072,
+      "step": 2368,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.4615,
+      "step": 2368,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.3372,
+      "step": 2368,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 0.45585581249849605,
+      "lm_loss": 2.2365,
+      "step": 2368,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.4560483191760714,
+      "grad_norm": 3.916581168435844,
+      "learning_rate": 1.801288012476811e-05,
+      "loss": 2.4711,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4562408258536468,
+      "grad_norm": 17.491146549436323,
+      "learning_rate": 1.8011014363865143e-05,
+      "loss": 2.5565,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4564333325312222,
+      "grad_norm": 3.8577337976652992,
+      "learning_rate": 1.800914782418909e-05,
+      "loss": 2.4503,
+      "step": 2371
+    },
+    {
+      "epoch": 0.45662583920879757,
+      "grad_norm": 4.307679763305614,
+      "learning_rate": 1.8007280505921404e-05,
+      "loss": 2.4555,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45681834588637293,
+      "grad_norm": 4.4976377989056715,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 2.495,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4570108525639483,
+      "grad_norm": 3.9387534614636563,
+      "learning_rate": 1.8003543534337303e-05,
+      "loss": 2.5048,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45720335924152367,
+      "grad_norm": 3.0372727514126185,
+      "learning_rate": 1.8001673881384176e-05,
+      "loss": 2.4397,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "grad_norm": 4.210620707311464,
+      "learning_rate": 1.7999803450565974e-05,
+      "loss": 2.3988,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.1933,
+      "step": 2376,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.2809,
+      "step": 2376,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.4831,
+      "step": 2376,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.2587,
+      "step": 2376,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.0197,
+      "step": 2376,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.1841,
+      "step": 2376,
+      "vm_loss": 0.1157
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.3972,
+      "step": 2376,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 0.4573958659190991,
+      "lm_loss": 2.2493,
+      "step": 2376,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 0.45758837259667445,
+      "grad_norm": 6.104813099053376,
+      "learning_rate": 1.7997932242064527e-05,
+      "loss": 2.3671,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4577808792742498,
+      "grad_norm": 3.7715778172645056,
+      "learning_rate": 1.799606025606175e-05,
+      "loss": 2.4,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4579733859518252,
+      "grad_norm": 3.218804062332221,
+      "learning_rate": 1.799418749273961e-05,
+      "loss": 2.4755,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4581658926294006,
+      "grad_norm": 3.1883713762000196,
+      "learning_rate": 1.7992313952280175e-05,
+      "loss": 2.4082,
+      "step": 2380
+    },
+    {
+      "epoch": 0.45835839930697597,
+      "grad_norm": 3.889008555999607,
+      "learning_rate": 1.7990439634865566e-05,
+      "loss": 2.3559,
+      "step": 2381
+    },
+    {
+      "epoch": 0.45855090598455134,
+      "grad_norm": 2.930844843482591,
+      "learning_rate": 1.7988564540678e-05,
+      "loss": 2.3576,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4587434126621267,
+      "grad_norm": 2.995541205198694,
+      "learning_rate": 1.798668866989976e-05,
+      "loss": 2.3986,
+      "step": 2383
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "grad_norm": 3.4163952463172556,
+      "learning_rate": 1.79848120227132e-05,
+      "loss": 2.3668,
+      "step": 2384
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.095,
+      "step": 2384,
+      "vm_loss": 0.2456
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.1814,
+      "step": 2384,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.2911,
+      "step": 2384,
+      "vm_loss": 0.2083
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.3935,
+      "step": 2384,
+      "vm_loss": 0.2145
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.0076,
+      "step": 2384,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.0577,
+      "step": 2384,
+      "vm_loss": 0.2581
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.2319,
+      "step": 2384,
+      "vm_loss": 0.1164
+    },
+    {
+      "epoch": 0.45893591933970207,
+      "lm_loss": 2.0758,
+      "step": 2384,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.4591284260172775,
+      "grad_norm": 3.083416169116607,
+      "learning_rate": 1.7982934599300754e-05,
+      "loss": 2.4253,
+      "step": 2385
+    },
+    {
+      "epoch": 0.45932093269485286,
+      "grad_norm": 3.1549298832566333,
+      "learning_rate": 1.7981056399844934e-05,
+      "loss": 2.3548,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4595134393724282,
+      "grad_norm": 3.6372264184634973,
+      "learning_rate": 1.7979177424528325e-05,
+      "loss": 2.35,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4597059460500036,
+      "grad_norm": 2.9445602397701194,
+      "learning_rate": 1.7977297673533586e-05,
+      "loss": 2.3203,
+      "step": 2388
+    },
+    {
+      "epoch": 0.459898452727579,
+      "grad_norm": 2.979596609096832,
+      "learning_rate": 1.7975417147043453e-05,
+      "loss": 2.3105,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4600909594051544,
+      "grad_norm": 3.1092741789944376,
+      "learning_rate": 1.7973535845240743e-05,
+      "loss": 2.3725,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46028346608272974,
+      "grad_norm": 2.960922411807894,
+      "learning_rate": 1.7971653768308334e-05,
+      "loss": 2.31,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "grad_norm": 2.80248806718247,
+      "learning_rate": 1.796977091642919e-05,
+      "loss": 2.3358,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.1382,
+      "step": 2392,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.03,
+      "step": 2392,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.3969,
+      "step": 2392,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.2939,
+      "step": 2392,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.1927,
+      "step": 2392,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 1.9511,
+      "step": 2392,
+      "vm_loss": 0.2459
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.4046,
+      "step": 2392,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 0.4604759727603051,
+      "lm_loss": 2.1387,
+      "step": 2392,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 0.46066847943788053,
+      "grad_norm": 2.990245336043424,
+      "learning_rate": 1.7967887289786352e-05,
+      "loss": 2.3125,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4608609861154559,
+      "grad_norm": 3.2394163591528478,
+      "learning_rate": 1.796600288856293e-05,
+      "loss": 2.3457,
+      "step": 2394
+    },
+    {
+      "epoch": 0.46105349279303126,
+      "grad_norm": 2.779711196596348,
+      "learning_rate": 1.7964117712942112e-05,
+      "loss": 2.3019,
+      "step": 2395
+    },
+    {
+      "epoch": 0.4612459994706066,
+      "grad_norm": 2.776740214777698,
+      "learning_rate": 1.7962231763107158e-05,
+      "loss": 2.3017,
+      "step": 2396
+    },
+    {
+      "epoch": 0.461438506148182,
+      "grad_norm": 3.017331508243643,
+      "learning_rate": 1.7960345039241415e-05,
+      "loss": 2.3713,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4616310128257574,
+      "grad_norm": 3.03999712944687,
+      "learning_rate": 1.7958457541528292e-05,
+      "loss": 2.3442,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4618235195033328,
+      "grad_norm": 3.0063769137489795,
+      "learning_rate": 1.7956569270151274e-05,
+      "loss": 2.3298,
+      "step": 2399
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "grad_norm": 2.8908918455742194,
+      "learning_rate": 1.795468022529393e-05,
+      "loss": 2.3644,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.3922,
+      "step": 2400,
+      "vm_loss": 0.2353
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.0621,
+      "step": 2400,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.3499,
+      "step": 2400,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 1.9638,
+      "step": 2400,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 1.9578,
+      "step": 2400,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.4211,
+      "step": 2400,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.3138,
+      "step": 2400,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.46201602618090815,
+      "lm_loss": 2.3924,
+      "step": 2400,
+      "vm_loss": 0.226
+    },
+    {
+      "epoch": 0.4622085328584835,
+      "grad_norm": 2.865274471298513,
+      "learning_rate": 1.79527904071399e-05,
+      "loss": 2.3759,
+      "step": 2401
+    },
+    {
+      "epoch": 0.46240103953605893,
+      "grad_norm": 2.9435011174637213,
+      "learning_rate": 1.7950899815872894e-05,
+      "loss": 2.339,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4625935462136343,
+      "grad_norm": 2.9114939429828786,
+      "learning_rate": 1.7949008451676703e-05,
+      "loss": 2.3239,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46278605289120966,
+      "grad_norm": 3.07677194247858,
+      "learning_rate": 1.7947116314735197e-05,
+      "loss": 2.3558,
+      "step": 2404
+    },
+    {
+      "epoch": 0.46297855956878503,
+      "grad_norm": 3.125604945990555,
+      "learning_rate": 1.7945223405232307e-05,
+      "loss": 2.3249,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4631710662463604,
+      "grad_norm": 3.262357826726187,
+      "learning_rate": 1.7943329723352057e-05,
+      "loss": 2.2219,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4633635729239358,
+      "grad_norm": 2.7454123096868996,
+      "learning_rate": 1.7941435269278533e-05,
+      "loss": 2.3578,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "grad_norm": 2.817129068134592,
+      "learning_rate": 1.7939540043195898e-05,
+      "loss": 2.2635,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.2138,
+      "step": 2408,
+      "vm_loss": 0.2145
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.4109,
+      "step": 2408,
+      "vm_loss": 0.1977
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.4497,
+      "step": 2408,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.198,
+      "step": 2408,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 1.9394,
+      "step": 2408,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.0549,
+      "step": 2408,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.2088,
+      "step": 2408,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.4635560796015112,
+      "lm_loss": 2.2197,
+      "step": 2408,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 0.46374858627908655,
+      "grad_norm": 3.11112440172383,
+      "learning_rate": 1.7937644045288395e-05,
+      "loss": 2.2915,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4639410929566619,
+      "grad_norm": 3.125630245160032,
+      "learning_rate": 1.793574727574034e-05,
+      "loss": 2.3692,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46413359963423734,
+      "grad_norm": 3.077932733601034,
+      "learning_rate": 1.7933849734736123e-05,
+      "loss": 2.3446,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4643261063118127,
+      "grad_norm": 2.735712827532773,
+      "learning_rate": 1.7931951422460203e-05,
+      "loss": 2.3003,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46451861298938807,
+      "grad_norm": 2.6751770647790414,
+      "learning_rate": 1.793005233909713e-05,
+      "loss": 2.3522,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46471111966696343,
+      "grad_norm": 2.5434138639147617,
+      "learning_rate": 1.7928152484831512e-05,
+      "loss": 2.2867,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4649036263445388,
+      "grad_norm": 2.853816866835236,
+      "learning_rate": 1.792625185984804e-05,
+      "loss": 2.3234,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "grad_norm": 2.8692653804710972,
+      "learning_rate": 1.792435046433149e-05,
+      "loss": 2.3467,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.7357,
+      "step": 2416,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.9325,
+      "step": 2416,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.9983,
+      "step": 2416,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.971,
+      "step": 2416,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 2.07,
+      "step": 2416,
+      "vm_loss": 0.2543
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.9344,
+      "step": 2416,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 1.9694,
+      "step": 2416,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 0.4650961330221142,
+      "lm_loss": 2.1749,
+      "step": 2416,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.4652886396996896,
+      "grad_norm": 2.691369060497828,
+      "learning_rate": 1.7922448298466687e-05,
+      "loss": 2.3143,
+      "step": 2417
+    },
+    {
+      "epoch": 0.46548114637726495,
+      "grad_norm": 3.0007583602816896,
+      "learning_rate": 1.792054536243855e-05,
+      "loss": 2.2396,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4656736530548403,
+      "grad_norm": 2.5770219085681996,
+      "learning_rate": 1.7918641656432073e-05,
+      "loss": 2.3234,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46586615973241574,
+      "grad_norm": 2.5414521653131903,
+      "learning_rate": 1.7916737180632316e-05,
+      "loss": 2.2853,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4660586664099911,
+      "grad_norm": 2.8400366264995167,
+      "learning_rate": 1.791483193522442e-05,
+      "loss": 2.2633,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4662511730875665,
+      "grad_norm": 2.9830389865983182,
+      "learning_rate": 1.79129259203936e-05,
+      "loss": 2.3212,
+      "step": 2422
+    },
+    {
+      "epoch": 0.46644367976514184,
+      "grad_norm": 2.5753937513011347,
+      "learning_rate": 1.791101913632515e-05,
+      "loss": 2.2955,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "grad_norm": 2.962506863682127,
+      "learning_rate": 1.790911158320442e-05,
+      "loss": 2.3183,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 2.151,
+      "step": 2424,
+      "vm_loss": 0.2178
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 1.882,
+      "step": 2424,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 2.3452,
+      "step": 2424,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 1.9555,
+      "step": 2424,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 2.0452,
+      "step": 2424,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 2.1511,
+      "step": 2424,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 2.1451,
+      "step": 2424,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 0.4666361864427172,
+      "lm_loss": 1.926,
+      "step": 2424,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.4668286931202926,
+      "grad_norm": 2.790728340621961,
+      "learning_rate": 1.7907203261216862e-05,
+      "loss": 2.2839,
+      "step": 2425
+    },
+    {
+      "epoch": 0.467021199797868,
+      "grad_norm": 2.8447247087147938,
+      "learning_rate": 1.7905294170547983e-05,
+      "loss": 2.245,
+      "step": 2426
+    },
+    {
+      "epoch": 0.46721370647544336,
+      "grad_norm": 2.78874672244569,
+      "learning_rate": 1.7903384311383374e-05,
+      "loss": 2.2752,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4674062131530187,
+      "grad_norm": 2.8421213752418777,
+      "learning_rate": 1.7901473683908694e-05,
+      "loss": 2.2709,
+      "step": 2428
+    },
+    {
+      "epoch": 0.46759871983059415,
+      "grad_norm": 2.7556359850468835,
+      "learning_rate": 1.7899562288309683e-05,
+      "loss": 2.2624,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4677912265081695,
+      "grad_norm": 2.614723145076032,
+      "learning_rate": 1.7897650124772154e-05,
+      "loss": 2.3186,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4679837331857449,
+      "grad_norm": 2.871992409114372,
+      "learning_rate": 1.7895737193481993e-05,
+      "loss": 2.3152,
+      "step": 2431
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "grad_norm": 2.783729285624364,
+      "learning_rate": 1.7893823494625158e-05,
+      "loss": 2.3507,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.1246,
+      "step": 2432,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.3636,
+      "step": 2432,
+      "vm_loss": 0.2153
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.1422,
+      "step": 2432,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.1018,
+      "step": 2432,
+      "vm_loss": 0.1812
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 1.8501,
+      "step": 2432,
+      "vm_loss": 0.1071
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.3932,
+      "step": 2432,
+      "vm_loss": 0.1056
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 2.2017,
+      "step": 2432,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 0.46817623986332024,
+      "lm_loss": 1.8816,
+      "step": 2432,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.46836874654089566,
+      "grad_norm": 2.683320317098717,
+      "learning_rate": 1.7891909028387688e-05,
+      "loss": 2.2729,
+      "step": 2433
+    },
+    {
+      "epoch": 0.46856125321847103,
+      "grad_norm": 2.6615637362099966,
+      "learning_rate": 1.7889993794955695e-05,
+      "loss": 2.3065,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4687537598960464,
+      "grad_norm": 2.7101461328072403,
+      "learning_rate": 1.7888077794515362e-05,
+      "loss": 2.2897,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46894626657362176,
+      "grad_norm": 2.895318792334533,
+      "learning_rate": 1.788616102725295e-05,
+      "loss": 2.246,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4691387732511971,
+      "grad_norm": 3.284331367802261,
+      "learning_rate": 1.7884243493354792e-05,
+      "loss": 2.3085,
+      "step": 2437
+    },
+    {
+      "epoch": 0.46933127992877255,
+      "grad_norm": 2.742929348158021,
+      "learning_rate": 1.78823251930073e-05,
+      "loss": 2.2718,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4695237866063479,
+      "grad_norm": 2.838975070062055,
+      "learning_rate": 1.7880406126396952e-05,
+      "loss": 2.2585,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "grad_norm": 3.21180468749498,
+      "learning_rate": 1.7878486293710308e-05,
+      "loss": 2.3511,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.2502,
+      "step": 2440,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 1.7412,
+      "step": 2440,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.3208,
+      "step": 2440,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.3066,
+      "step": 2440,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.0328,
+      "step": 2440,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 1.981,
+      "step": 2440,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.2021,
+      "step": 2440,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.4697162932839233,
+      "lm_loss": 2.3943,
+      "step": 2440,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.46990879996149865,
+      "grad_norm": 2.6299559480312054,
+      "learning_rate": 1.7876565695134002e-05,
+      "loss": 2.3137,
+      "step": 2441
+    },
+    {
+      "epoch": 0.47010130663907407,
+      "grad_norm": 2.9379007138784874,
+      "learning_rate": 1.7874644330854742e-05,
+      "loss": 2.3026,
+      "step": 2442
+    },
+    {
+      "epoch": 0.47029381331664943,
+      "grad_norm": 2.8774306668916507,
+      "learning_rate": 1.7872722201059305e-05,
+      "loss": 2.3318,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4704863199942248,
+      "grad_norm": 2.7792748659879907,
+      "learning_rate": 1.7870799305934547e-05,
+      "loss": 2.3365,
+      "step": 2444
+    },
+    {
+      "epoch": 0.47067882667180017,
+      "grad_norm": 3.0319054358484507,
+      "learning_rate": 1.7868875645667404e-05,
+      "loss": 2.261,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47087133334937553,
+      "grad_norm": 2.7070302029013407,
+      "learning_rate": 1.786695122044487e-05,
+      "loss": 2.2547,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47106384002695095,
+      "grad_norm": 3.440815425482522,
+      "learning_rate": 1.7865026030454037e-05,
+      "loss": 2.314,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "grad_norm": 2.9166625202947967,
+      "learning_rate": 1.786310007588204e-05,
+      "loss": 2.3376,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 2.3058,
+      "step": 2448,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 2.3001,
+      "step": 2448,
+      "vm_loss": 0.111
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 1.9594,
+      "step": 2448,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 1.8172,
+      "step": 2448,
+      "vm_loss": 0.1096
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 2.3719,
+      "step": 2448,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 1.8892,
+      "step": 2448,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 2.1904,
+      "step": 2448,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.4712563467045263,
+      "lm_loss": 1.9629,
+      "step": 2448,
+      "vm_loss": 0.1281
+    },
+    {
+      "epoch": 0.4714488533821017,
+      "grad_norm": 2.93973245685304,
+      "learning_rate": 1.7861173356916127e-05,
+      "loss": 2.2693,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47164136005967705,
+      "grad_norm": 3.467385242028855,
+      "learning_rate": 1.7859245873743586e-05,
+      "loss": 2.2733,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47183386673725247,
+      "grad_norm": 2.774660294996914,
+      "learning_rate": 1.7857317626551798e-05,
+      "loss": 2.3051,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47202637341482784,
+      "grad_norm": 2.858592246269067,
+      "learning_rate": 1.7855388615528213e-05,
+      "loss": 2.3226,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4722188800924032,
+      "grad_norm": 3.25872473287099,
+      "learning_rate": 1.7853458840860353e-05,
+      "loss": 2.2839,
+      "step": 2453
+    },
+    {
+      "epoch": 0.47241138676997857,
+      "grad_norm": 3.2214287691275176,
+      "learning_rate": 1.785152830273582e-05,
+      "loss": 2.3369,
+      "step": 2454
+    },
+    {
+      "epoch": 0.47260389344755394,
+      "grad_norm": 3.0807183857014273,
+      "learning_rate": 1.7849597001342284e-05,
+      "loss": 2.2634,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "grad_norm": 3.0540556998661263,
+      "learning_rate": 1.7847664936867493e-05,
+      "loss": 2.3348,
+      "step": 2456
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 1.7261,
+      "step": 2456,
+      "vm_loss": 0.2499
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 1.7832,
+      "step": 2456,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 2.0767,
+      "step": 2456,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 2.071,
+      "step": 2456,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 2.5267,
+      "step": 2456,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 2.0225,
+      "step": 2456,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 1.9774,
+      "step": 2456,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.47279640012512936,
+      "lm_loss": 1.8656,
+      "step": 2456,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 0.4729889068027047,
+      "grad_norm": 2.9824302351413934,
+      "learning_rate": 1.7845732109499275e-05,
+      "loss": 2.2793,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4731814134802801,
+      "grad_norm": 3.0379791470252595,
+      "learning_rate": 1.7843798519425514e-05,
+      "loss": 2.2933,
+      "step": 2458
+    },
+    {
+      "epoch": 0.47337392015785545,
+      "grad_norm": 2.7802610923880504,
+      "learning_rate": 1.7841864166834185e-05,
+      "loss": 2.3005,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4735664268354309,
+      "grad_norm": 2.6795933679970974,
+      "learning_rate": 1.7839929051913336e-05,
+      "loss": 2.2457,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47375893351300624,
+      "grad_norm": 3.345450928143097,
+      "learning_rate": 1.783799317485108e-05,
+      "loss": 2.27,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4739514401905816,
+      "grad_norm": 2.759777826130828,
+      "learning_rate": 1.7836056535835607e-05,
+      "loss": 2.3254,
+      "step": 2462
+    },
+    {
+      "epoch": 0.474143946868157,
+      "grad_norm": 2.991248081618804,
+      "learning_rate": 1.783411913505519e-05,
+      "loss": 2.287,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "grad_norm": 2.72177921061463,
+      "learning_rate": 1.7832180972698166e-05,
+      "loss": 2.3354,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 2.2283,
+      "step": 2464,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 1.9308,
+      "step": 2464,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 2.1478,
+      "step": 2464,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 1.8642,
+      "step": 2464,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 1.9318,
+      "step": 2464,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 2.0204,
+      "step": 2464,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 2.0637,
+      "step": 2464,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 0.4743364535457324,
+      "lm_loss": 1.6153,
+      "step": 2464,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 0.47452896022330776,
+      "grad_norm": 2.8921298143247993,
+      "learning_rate": 1.7830242048952947e-05,
+      "loss": 2.2717,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4747214669008831,
+      "grad_norm": 2.9135477026611243,
+      "learning_rate": 1.7828302364008024e-05,
+      "loss": 2.3188,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4749139735784585,
+      "grad_norm": 2.76850667840445,
+      "learning_rate": 1.7826361918051956e-05,
+      "loss": 2.3106,
+      "step": 2467
+    },
+    {
+      "epoch": 0.47510648025603386,
+      "grad_norm": 2.6380143727757974,
+      "learning_rate": 1.782442071127338e-05,
+      "loss": 2.3306,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4752989869336093,
+      "grad_norm": 2.7172113234387756,
+      "learning_rate": 1.7822478743861008e-05,
+      "loss": 2.2874,
+      "step": 2469
+    },
+    {
+      "epoch": 0.47549149361118465,
+      "grad_norm": 2.820359645241799,
+      "learning_rate": 1.7820536016003625e-05,
+      "loss": 2.2874,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47568400028876,
+      "grad_norm": 2.7271457941493678,
+      "learning_rate": 1.781859252789009e-05,
+      "loss": 2.2711,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "grad_norm": 2.7686974400428874,
+      "learning_rate": 1.7816648279709327e-05,
+      "loss": 2.2821,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 1.7014,
+      "step": 2472,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.0505,
+      "step": 2472,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.3833,
+      "step": 2472,
+      "vm_loss": 0.2545
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.3397,
+      "step": 2472,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.1906,
+      "step": 2472,
+      "vm_loss": 0.2012
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.3236,
+      "step": 2472,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 1.89,
+      "step": 2472,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.4758765069663354,
+      "lm_loss": 2.2237,
+      "step": 2472,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.4760690136439108,
+      "grad_norm": 2.8731253245191786,
+      "learning_rate": 1.781470327165035e-05,
+      "loss": 2.3367,
+      "step": 2473
+    },
+    {
+      "epoch": 0.47626152032148616,
+      "grad_norm": 2.9764662480395754,
+      "learning_rate": 1.7812757503902234e-05,
+      "loss": 2.3115,
+      "step": 2474
+    },
+    {
+      "epoch": 0.47645402699906153,
+      "grad_norm": 2.6349584211140864,
+      "learning_rate": 1.7810810976654135e-05,
+      "loss": 2.2839,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4766465336766369,
+      "grad_norm": 3.2707711685971934,
+      "learning_rate": 1.7808863690095283e-05,
+      "loss": 2.3148,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47683904035421226,
+      "grad_norm": 2.8203970868055928,
+      "learning_rate": 1.780691564441497e-05,
+      "loss": 2.324,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4770315470317877,
+      "grad_norm": 2.7007971964851913,
+      "learning_rate": 1.780496683980258e-05,
+      "loss": 2.2756,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47722405370936305,
+      "grad_norm": 3.022028469303706,
+      "learning_rate": 1.7803017276447558e-05,
+      "loss": 2.3281,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "grad_norm": 3.0015100776468806,
+      "learning_rate": 1.7801066954539424e-05,
+      "loss": 2.2613,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 2.1357,
+      "step": 2480,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 1.6637,
+      "step": 2480,
+      "vm_loss": 0.1793
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 1.87,
+      "step": 2480,
+      "vm_loss": 0.2122
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 2.2797,
+      "step": 2480,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 2.4752,
+      "step": 2480,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 2.2192,
+      "step": 2480,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 1.976,
+      "step": 2480,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 0.4774165603869384,
+      "lm_loss": 2.2177,
+      "step": 2480,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 0.4776090670645138,
+      "grad_norm": 2.9507337386080583,
+      "learning_rate": 1.7799115874267782e-05,
+      "loss": 2.2796,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4778015737420892,
+      "grad_norm": 3.273729245309616,
+      "learning_rate": 1.7797164035822295e-05,
+      "loss": 2.3087,
+      "step": 2482
+    },
+    {
+      "epoch": 0.47799408041966457,
+      "grad_norm": 2.9390748708887457,
+      "learning_rate": 1.779521143939271e-05,
+      "loss": 2.2512,
+      "step": 2483
+    },
+    {
+      "epoch": 0.47818658709723993,
+      "grad_norm": 3.0189722052742503,
+      "learning_rate": 1.7793258085168844e-05,
+      "loss": 2.278,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4783790937748153,
+      "grad_norm": 2.867311762530984,
+      "learning_rate": 1.7791303973340584e-05,
+      "loss": 2.3131,
+      "step": 2485
+    },
+    {
+      "epoch": 0.47857160045239067,
+      "grad_norm": 2.8110016070823174,
+      "learning_rate": 1.77893491040979e-05,
+      "loss": 2.2791,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4787641071299661,
+      "grad_norm": 3.054189715388609,
+      "learning_rate": 1.7787393477630826e-05,
+      "loss": 2.3342,
+      "step": 2487
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "grad_norm": 2.6569225068447717,
+      "learning_rate": 1.778543709412948e-05,
+      "loss": 2.2865,
+      "step": 2488
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 2.3942,
+      "step": 2488,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 1.8804,
+      "step": 2488,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 1.9992,
+      "step": 2488,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 2.1193,
+      "step": 2488,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 2.0918,
+      "step": 2488,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 1.9725,
+      "step": 2488,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 2.1409,
+      "step": 2488,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.47895661380754145,
+      "lm_loss": 2.5955,
+      "step": 2488,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 0.4791491204851168,
+      "grad_norm": 2.8484718637776267,
+      "learning_rate": 1.778347995378404e-05,
+      "loss": 2.287,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4793416271626922,
+      "grad_norm": 2.842830316607992,
+      "learning_rate": 1.778152205678477e-05,
+      "loss": 2.3335,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4795341338402676,
+      "grad_norm": 2.8669813223498863,
+      "learning_rate": 1.7779563403322006e-05,
+      "loss": 2.2747,
+      "step": 2491
+    },
+    {
+      "epoch": 0.479726640517843,
+      "grad_norm": 2.6786687075178657,
+      "learning_rate": 1.777760399358614e-05,
+      "loss": 2.2678,
+      "step": 2492
+    },
+    {
+      "epoch": 0.47991914719541834,
+      "grad_norm": 2.7338259002136684,
+      "learning_rate": 1.7775643827767668e-05,
+      "loss": 2.341,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4801116538729937,
+      "grad_norm": 2.827619438066888,
+      "learning_rate": 1.7773682906057133e-05,
+      "loss": 2.3136,
+      "step": 2494
+    },
+    {
+      "epoch": 0.48030416055056907,
+      "grad_norm": 2.7764024789498998,
+      "learning_rate": 1.777172122864517e-05,
+      "loss": 2.2741,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "grad_norm": 2.5973000744295325,
+      "learning_rate": 1.776975879572247e-05,
+      "loss": 2.3121,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 2.2804,
+      "step": 2496,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 2.0859,
+      "step": 2496,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 1.9264,
+      "step": 2496,
+      "vm_loss": 0.2008
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 2.1099,
+      "step": 2496,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 1.9701,
+      "step": 2496,
+      "vm_loss": 0.2113
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 2.1673,
+      "step": 2496,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 1.8768,
+      "step": 2496,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 0.4804966672281445,
+      "lm_loss": 2.1848,
+      "step": 2496,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.48068917390571986,
+      "grad_norm": 2.758407370499133,
+      "learning_rate": 1.776779560747981e-05,
+      "loss": 2.3027,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4808816805832952,
+      "grad_norm": 2.8175415601978298,
+      "learning_rate": 1.776583166410804e-05,
+      "loss": 2.2574,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4810741872608706,
+      "grad_norm": 3.1160858530931383,
+      "learning_rate": 1.776386696579808e-05,
+      "loss": 2.2968,
+      "step": 2499
+    },
+    {
+      "epoch": 0.481266693938446,
+      "grad_norm": 2.9813271165326176,
+      "learning_rate": 1.7761901512740918e-05,
+      "loss": 2.2009,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4814592006160214,
+      "grad_norm": 3.0056322173375865,
+      "learning_rate": 1.775993530512763e-05,
+      "loss": 2.2438,
+      "step": 2501
+    },
+    {
+      "epoch": 0.48165170729359674,
+      "grad_norm": 2.902724321473114,
+      "learning_rate": 1.7757968343149348e-05,
+      "loss": 2.2238,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4818442139711721,
+      "grad_norm": 2.9153237156705303,
+      "learning_rate": 1.775600062699729e-05,
+      "loss": 2.2917,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "grad_norm": 3.076326841136655,
+      "learning_rate": 1.7754032156862744e-05,
+      "loss": 2.2813,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.106,
+      "step": 2504,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.2349,
+      "step": 2504,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.273,
+      "step": 2504,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 1.9146,
+      "step": 2504,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 1.9185,
+      "step": 2504,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.3888,
+      "step": 2504,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.1171,
+      "step": 2504,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.48203672064874753,
+      "lm_loss": 2.0008,
+      "step": 2504,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 0.4822292273263229,
+      "grad_norm": 2.9137155613882237,
+      "learning_rate": 1.7752062932937072e-05,
+      "loss": 2.2365,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48242173400389826,
+      "grad_norm": 2.8470226444867173,
+      "learning_rate": 1.7750092955411697e-05,
+      "loss": 2.3559,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4826142406814736,
+      "grad_norm": 2.992200474800163,
+      "learning_rate": 1.7748122224478138e-05,
+      "loss": 2.2447,
+      "step": 2507
+    },
+    {
+      "epoch": 0.482806747359049,
+      "grad_norm": 3.332323118584798,
+      "learning_rate": 1.7746150740327978e-05,
+      "loss": 2.298,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4829992540366244,
+      "grad_norm": 3.2181560702089893,
+      "learning_rate": 1.7744178503152856e-05,
+      "loss": 2.2955,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4831917607141998,
+      "grad_norm": 3.3009885697712225,
+      "learning_rate": 1.7742205513144507e-05,
+      "loss": 2.2493,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48338426739177515,
+      "grad_norm": 2.9802031659739896,
+      "learning_rate": 1.7740231770494733e-05,
+      "loss": 2.2029,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "grad_norm": 2.702075454671211,
+      "learning_rate": 1.7738257275395404e-05,
+      "loss": 2.2539,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 1.7886,
+      "step": 2512,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 1.9571,
+      "step": 2512,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.1742,
+      "step": 2512,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.2608,
+      "step": 2512,
+      "vm_loss": 0.2197
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.0741,
+      "step": 2512,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.2156,
+      "step": 2512,
+      "vm_loss": 0.1177
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.3894,
+      "step": 2512,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.4835767740693505,
+      "lm_loss": 2.3169,
+      "step": 2512,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.48376928074692593,
+      "grad_norm": 2.656867607246949,
+      "learning_rate": 1.7736282028038467e-05,
+      "loss": 2.2801,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4839617874245013,
+      "grad_norm": 2.990666451512867,
+      "learning_rate": 1.773430602861594e-05,
+      "loss": 2.3087,
+      "step": 2514
+    },
+    {
+      "epoch": 0.48415429410207667,
+      "grad_norm": 2.7272215355453886,
+      "learning_rate": 1.7732329277319914e-05,
+      "loss": 2.2135,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48434680077965203,
+      "grad_norm": 2.8236867862317725,
+      "learning_rate": 1.773035177434256e-05,
+      "loss": 2.278,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4845393074572274,
+      "grad_norm": 2.854742291753803,
+      "learning_rate": 1.772837351987611e-05,
+      "loss": 2.3181,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4847318141348028,
+      "grad_norm": 3.1824360146062967,
+      "learning_rate": 1.7726394514112885e-05,
+      "loss": 2.2925,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4849243208123782,
+      "grad_norm": 2.6495974401136557,
+      "learning_rate": 1.772441475724526e-05,
+      "loss": 2.2998,
+      "step": 2519
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "grad_norm": 2.901398825129236,
+      "learning_rate": 1.7722434249465695e-05,
+      "loss": 2.3034,
+      "step": 2520
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.1055,
+      "step": 2520,
+      "vm_loss": 0.2231
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.2893,
+      "step": 2520,
+      "vm_loss": 0.2163
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.0676,
+      "step": 2520,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.2021,
+      "step": 2520,
+      "vm_loss": 0.2222
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.1718,
+      "step": 2520,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.2904,
+      "step": 2520,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.141,
+      "step": 2520,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.48511682748995355,
+      "lm_loss": 2.0731,
+      "step": 2520,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 0.4853093341675289,
+      "grad_norm": 2.7079131059338817,
+      "learning_rate": 1.7720452990966728e-05,
+      "loss": 2.262,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48550184084510434,
+      "grad_norm": 2.9569668962994458,
+      "learning_rate": 1.7718470981940953e-05,
+      "loss": 2.272,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4856943475226797,
+      "grad_norm": 3.128942120995433,
+      "learning_rate": 1.771648822258105e-05,
+      "loss": 2.3106,
+      "step": 2523
+    },
+    {
+      "epoch": 0.48588685420025507,
+      "grad_norm": 3.0672495369227493,
+      "learning_rate": 1.771450471307977e-05,
+      "loss": 2.2554,
+      "step": 2524
+    },
+    {
+      "epoch": 0.48607936087783044,
+      "grad_norm": 2.748026796346768,
+      "learning_rate": 1.7712520453629933e-05,
+      "loss": 2.2535,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4862718675554058,
+      "grad_norm": 2.5591060358947355,
+      "learning_rate": 1.771053544442444e-05,
+      "loss": 2.3121,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4864643742329812,
+      "grad_norm": 2.507063965197852,
+      "learning_rate": 1.770854968565625e-05,
+      "loss": 2.2214,
+      "step": 2527
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "grad_norm": 2.778943604490944,
+      "learning_rate": 1.7706563177518414e-05,
+      "loss": 2.2849,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 2.386,
+      "step": 2528,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 1.9305,
+      "step": 2528,
+      "vm_loss": 0.0614
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 2.2808,
+      "step": 2528,
+      "vm_loss": 0.2107
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 2.1614,
+      "step": 2528,
+      "vm_loss": 0.2191
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 1.9934,
+      "step": 2528,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 2.1757,
+      "step": 2528,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 1.9836,
+      "step": 2528,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.4866568809105566,
+      "lm_loss": 1.9253,
+      "step": 2528,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.48684938758813195,
+      "grad_norm": 2.8527012368064257,
+      "learning_rate": 1.770457592020404e-05,
+      "loss": 2.2983,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4870418942657073,
+      "grad_norm": 2.548903435725304,
+      "learning_rate": 1.7702587913906314e-05,
+      "loss": 2.2299,
+      "step": 2530
+    },
+    {
+      "epoch": 0.48723440094328274,
+      "grad_norm": 2.7534766543717537,
+      "learning_rate": 1.77005991588185e-05,
+      "loss": 2.2199,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4874269076208581,
+      "grad_norm": 2.831929858131102,
+      "learning_rate": 1.7698609655133925e-05,
+      "loss": 2.243,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4876194142984335,
+      "grad_norm": 2.8308981229969206,
+      "learning_rate": 1.7696619403046003e-05,
+      "loss": 2.2362,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48781192097600884,
+      "grad_norm": 2.817520067738735,
+      "learning_rate": 1.7694628402748203e-05,
+      "loss": 2.2854,
+      "step": 2534
+    },
+    {
+      "epoch": 0.48800442765358426,
+      "grad_norm": 3.114029941678181,
+      "learning_rate": 1.769263665443408e-05,
+      "loss": 2.2538,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "grad_norm": 2.95465485247302,
+      "learning_rate": 1.7690644158297258e-05,
+      "loss": 2.2373,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.1383,
+      "step": 2536,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.0472,
+      "step": 2536,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.1432,
+      "step": 2536,
+      "vm_loss": 0.2142
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.0421,
+      "step": 2536,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.0239,
+      "step": 2536,
+      "vm_loss": 0.2181
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.3476,
+      "step": 2536,
+      "vm_loss": 0.1197
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.1245,
+      "step": 2536,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 0.4881969343311596,
+      "lm_loss": 2.1739,
+      "step": 2536,
+      "vm_loss": 0.1057
+    },
+    {
+      "epoch": 0.488389441008735,
+      "grad_norm": 2.886846240006783,
+      "learning_rate": 1.768865091453143e-05,
+      "loss": 2.26,
+      "step": 2537
+    },
+    {
+      "epoch": 0.48858194768631036,
+      "grad_norm": 2.7188743383619447,
+      "learning_rate": 1.7686656923330366e-05,
+      "loss": 2.2457,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4887744543638857,
+      "grad_norm": 2.879436959628215,
+      "learning_rate": 1.768466218488791e-05,
+      "loss": 2.3291,
+      "step": 2539
+    },
+    {
+      "epoch": 0.48896696104146115,
+      "grad_norm": 2.6543654927251783,
+      "learning_rate": 1.7682666699397974e-05,
+      "loss": 2.2987,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4891594677190365,
+      "grad_norm": 2.988649180985386,
+      "learning_rate": 1.7680670467054543e-05,
+      "loss": 2.2739,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4893519743966119,
+      "grad_norm": 2.4518801462538455,
+      "learning_rate": 1.7678673488051677e-05,
+      "loss": 2.3312,
+      "step": 2542
+    },
+    {
+      "epoch": 0.48954448107418724,
+      "grad_norm": 2.9967463023262577,
+      "learning_rate": 1.767667576258351e-05,
+      "loss": 2.2245,
+      "step": 2543
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "grad_norm": 2.9455019108387908,
+      "learning_rate": 1.7674677290844245e-05,
+      "loss": 2.3238,
+      "step": 2544
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 2.2408,
+      "step": 2544,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 2.3415,
+      "step": 2544,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 2.1294,
+      "step": 2544,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 1.7967,
+      "step": 2544,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 1.762,
+      "step": 2544,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 2.1862,
+      "step": 2544,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 1.6257,
+      "step": 2544,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 0.48973698775176266,
+      "lm_loss": 2.1738,
+      "step": 2544,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 0.48992949442933803,
+      "grad_norm": 3.0964093031507316,
+      "learning_rate": 1.7672678073028163e-05,
+      "loss": 2.2093,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4901220011069134,
+      "grad_norm": 2.9368734050734098,
+      "learning_rate": 1.767067810932961e-05,
+      "loss": 2.218,
+      "step": 2546
+    },
+    {
+      "epoch": 0.49031450778448876,
+      "grad_norm": 2.74702750810346,
+      "learning_rate": 1.7668677399943004e-05,
+      "loss": 2.2907,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49050701446206413,
+      "grad_norm": 3.312589435932395,
+      "learning_rate": 1.7666675945062848e-05,
+      "loss": 2.2803,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49069952113963955,
+      "grad_norm": 3.0587406364477436,
+      "learning_rate": 1.76646737448837e-05,
+      "loss": 2.3107,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4908920278172149,
+      "grad_norm": 2.777917692566353,
+      "learning_rate": 1.7662670799600207e-05,
+      "loss": 2.263,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4910845344947903,
+      "grad_norm": 2.9732451609985864,
+      "learning_rate": 1.766066710940708e-05,
+      "loss": 2.3238,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "grad_norm": 3.036282775155767,
+      "learning_rate": 1.76586626744991e-05,
+      "loss": 2.233,
+      "step": 2552
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 1.8737,
+      "step": 2552,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 2.5071,
+      "step": 2552,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 1.9605,
+      "step": 2552,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 1.9639,
+      "step": 2552,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 2.1104,
+      "step": 2552,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 2.2875,
+      "step": 2552,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 2.0941,
+      "step": 2552,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.49127704117236565,
+      "lm_loss": 1.6552,
+      "step": 2552,
+      "vm_loss": 0.1164
+    },
+    {
+      "epoch": 0.49146954784994107,
+      "grad_norm": 2.972581653046284,
+      "learning_rate": 1.7656657495071122e-05,
+      "loss": 2.2559,
+      "step": 2553
+    },
+    {
+      "epoch": 0.49166205452751643,
+      "grad_norm": 2.9562968463803183,
+      "learning_rate": 1.765465157131808e-05,
+      "loss": 2.2865,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4918545612050918,
+      "grad_norm": 2.72488944187822,
+      "learning_rate": 1.7652644903434975e-05,
+      "loss": 2.2532,
+      "step": 2555
+    },
+    {
+      "epoch": 0.49204706788266717,
+      "grad_norm": 3.1832018923207177,
+      "learning_rate": 1.765063749161688e-05,
+      "loss": 2.2995,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49223957456024253,
+      "grad_norm": 2.785982255207764,
+      "learning_rate": 1.764862933605894e-05,
+      "loss": 2.2413,
+      "step": 2557
+    },
+    {
+      "epoch": 0.49243208123781795,
+      "grad_norm": 2.6755032984446565,
+      "learning_rate": 1.7646620436956372e-05,
+      "loss": 2.2768,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4926245879153933,
+      "grad_norm": 3.097254628498717,
+      "learning_rate": 1.764461079450447e-05,
+      "loss": 2.2866,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "grad_norm": 2.972208832970011,
+      "learning_rate": 1.76426004088986e-05,
+      "loss": 2.2834,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.0279,
+      "step": 2560,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.2367,
+      "step": 2560,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.0743,
+      "step": 2560,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 1.7309,
+      "step": 2560,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.115,
+      "step": 2560,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.2572,
+      "step": 2560,
+      "vm_loss": 0.2288
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 1.9989,
+      "step": 2560,
+      "vm_loss": 0.2201
+    },
+    {
+      "epoch": 0.4928170945929687,
+      "lm_loss": 2.2212,
+      "step": 2560,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 0.49300960127054405,
+      "grad_norm": 2.9174311662819754,
+      "learning_rate": 1.7640589280334185e-05,
+      "loss": 2.2437,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4932021079481195,
+      "grad_norm": 2.880484016044867,
+      "learning_rate": 1.7638577409006746e-05,
+      "loss": 2.2711,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49339461462569484,
+      "grad_norm": 2.6745391159531273,
+      "learning_rate": 1.763656479511185e-05,
+      "loss": 2.2332,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4935871213032702,
+      "grad_norm": 3.4460242205312097,
+      "learning_rate": 1.7634551438845166e-05,
+      "loss": 2.2399,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49377962798084557,
+      "grad_norm": 3.05609538339663,
+      "learning_rate": 1.76325373404024e-05,
+      "loss": 2.2649,
+      "step": 2565
+    },
+    {
+      "epoch": 0.493972134658421,
+      "grad_norm": 2.9606522693095676,
+      "learning_rate": 1.7630522499979358e-05,
+      "loss": 2.2873,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49416464133599636,
+      "grad_norm": 3.0479777234997627,
+      "learning_rate": 1.7628506917771906e-05,
+      "loss": 2.2346,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "grad_norm": 3.2481405774702137,
+      "learning_rate": 1.7626490593975987e-05,
+      "loss": 2.2538,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.205,
+      "step": 2568,
+      "vm_loss": 0.184
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.2148,
+      "step": 2568,
+      "vm_loss": 0.0979
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.1169,
+      "step": 2568,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.2488,
+      "step": 2568,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.0137,
+      "step": 2568,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.0691,
+      "step": 2568,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.1003,
+      "step": 2568,
+      "vm_loss": 0.2279
+    },
+    {
+      "epoch": 0.4943571480135717,
+      "lm_loss": 2.2928,
+      "step": 2568,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.4945496546911471,
+      "grad_norm": 2.6619418250407727,
+      "learning_rate": 1.7624473528787605e-05,
+      "loss": 2.2489,
+      "step": 2569
+    },
+    {
+      "epoch": 0.49474216136872246,
+      "grad_norm": 2.9907414740006932,
+      "learning_rate": 1.7622455722402857e-05,
+      "loss": 2.2233,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4949346680462979,
+      "grad_norm": 3.3050848080361033,
+      "learning_rate": 1.762043717501789e-05,
+      "loss": 2.2595,
+      "step": 2571
+    },
+    {
+      "epoch": 0.49512717472387324,
+      "grad_norm": 2.9552003063155086,
+      "learning_rate": 1.7618417886828934e-05,
+      "loss": 2.2703,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4953196814014486,
+      "grad_norm": 3.1171436658654432,
+      "learning_rate": 1.7616397858032295e-05,
+      "loss": 2.2244,
+      "step": 2573
+    },
+    {
+      "epoch": 0.495512188079024,
+      "grad_norm": 2.776066597500756,
+      "learning_rate": 1.761437708882434e-05,
+      "loss": 2.1886,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4957046947565994,
+      "grad_norm": 3.094713604254845,
+      "learning_rate": 1.7612355579401518e-05,
+      "loss": 2.2518,
+      "step": 2575
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "grad_norm": 3.1307204060496607,
+      "learning_rate": 1.761033332996034e-05,
+      "loss": 2.2356,
+      "step": 2576
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.1416,
+      "step": 2576,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.0645,
+      "step": 2576,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.2717,
+      "step": 2576,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.1734,
+      "step": 2576,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.248,
+      "step": 2576,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 1.9946,
+      "step": 2576,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.0347,
+      "step": 2576,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 0.49589720143417476,
+      "lm_loss": 2.1461,
+      "step": 2576,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.4960897081117501,
+      "grad_norm": 3.020120593297418,
+      "learning_rate": 1.7608310340697396e-05,
+      "loss": 2.2175,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4962822147893255,
+      "grad_norm": 2.683664208066447,
+      "learning_rate": 1.7606286611809353e-05,
+      "loss": 2.2577,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49647472146690086,
+      "grad_norm": 2.9558971024250775,
+      "learning_rate": 1.7604262143492937e-05,
+      "loss": 2.2257,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4966672281444763,
+      "grad_norm": 2.803176100642997,
+      "learning_rate": 1.7602236935944953e-05,
+      "loss": 2.2715,
+      "step": 2580
+    },
+    {
+      "epoch": 0.49685973482205165,
+      "grad_norm": 2.6343832298658683,
+      "learning_rate": 1.7600210989362277e-05,
+      "loss": 2.2643,
+      "step": 2581
+    },
+    {
+      "epoch": 0.497052241499627,
+      "grad_norm": 2.9775843277826706,
+      "learning_rate": 1.7598184303941862e-05,
+      "loss": 2.2377,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4972447481772024,
+      "grad_norm": 2.770780143857951,
+      "learning_rate": 1.7596156879880718e-05,
+      "loss": 2.2485,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "grad_norm": 3.011594554533267,
+      "learning_rate": 1.7594128717375945e-05,
+      "loss": 2.2445,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 1.9146,
+      "step": 2584,
+      "vm_loss": 0.126
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 1.9133,
+      "step": 2584,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 2.1006,
+      "step": 2584,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 1.8866,
+      "step": 2584,
+      "vm_loss": 0.1397
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 2.3141,
+      "step": 2584,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 2.0353,
+      "step": 2584,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 2.2654,
+      "step": 2584,
+      "vm_loss": 0.1907
+    },
+    {
+      "epoch": 0.4974372548547778,
+      "lm_loss": 1.5699,
+      "step": 2584,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 0.49762976153235317,
+      "grad_norm": 2.733499147524477,
+      "learning_rate": 1.7592099816624703e-05,
+      "loss": 2.2869,
+      "step": 2585
+    },
+    {
+      "epoch": 0.49782226820992853,
+      "grad_norm": 2.680883007818799,
+      "learning_rate": 1.7590070177824228e-05,
+      "loss": 2.2111,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4980147748875039,
+      "grad_norm": 2.7995595381169385,
+      "learning_rate": 1.7588039801171827e-05,
+      "loss": 2.3103,
+      "step": 2587
+    },
+    {
+      "epoch": 0.49820728156507926,
+      "grad_norm": 3.053585236789676,
+      "learning_rate": 1.7586008686864874e-05,
+      "loss": 2.2993,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4983997882426547,
+      "grad_norm": 2.9638680394664774,
+      "learning_rate": 1.758397683510083e-05,
+      "loss": 2.2384,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49859229492023005,
+      "grad_norm": 2.748332135849734,
+      "learning_rate": 1.7581944246077202e-05,
+      "loss": 2.2757,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4987848015978054,
+      "grad_norm": 2.737966006045177,
+      "learning_rate": 1.7579910919991596e-05,
+      "loss": 2.2066,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "grad_norm": 2.890117448056369,
+      "learning_rate": 1.7577876857041672e-05,
+      "loss": 2.2362,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.1523,
+      "step": 2592,
+      "vm_loss": 0.2719
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.0649,
+      "step": 2592,
+      "vm_loss": 0.2127
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.0689,
+      "step": 2592,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.2836,
+      "step": 2592,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 1.8959,
+      "step": 2592,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.3432,
+      "step": 2592,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.2886,
+      "step": 2592,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 0.4989773082753808,
+      "lm_loss": 2.1799,
+      "step": 2592,
+      "vm_loss": 0.2484
+    },
+    {
+      "epoch": 0.4991698149529562,
+      "grad_norm": 2.802861383384065,
+      "learning_rate": 1.757584205742517e-05,
+      "loss": 2.2816,
+      "step": 2593
+    },
+    {
+      "epoch": 0.49936232163053157,
+      "grad_norm": 2.892799416827955,
+      "learning_rate": 1.75738065213399e-05,
+      "loss": 2.2579,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49955482830810694,
+      "grad_norm": 2.463193662151769,
+      "learning_rate": 1.7571770248983734e-05,
+      "loss": 2.2585,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4997473349856823,
+      "grad_norm": 2.815154361430906,
+      "learning_rate": 1.756973324055463e-05,
+      "loss": 2.2424,
+      "step": 2596
+    },
+    {
+      "epoch": 0.49993984166325767,
+      "grad_norm": 2.989821253217763,
+      "learning_rate": 1.756769549625061e-05,
+      "loss": 2.2718,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5001323483408331,
+      "grad_norm": 2.940453260654077,
+      "learning_rate": 1.7565657016269767e-05,
+      "loss": 2.2653,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5003248550184084,
+      "grad_norm": 2.699491163905135,
+      "learning_rate": 1.756361780081027e-05,
+      "loss": 2.2537,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "grad_norm": 2.797466387789946,
+      "learning_rate": 1.7561577850070355e-05,
+      "loss": 2.2796,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 2.2221,
+      "step": 2600,
+      "vm_loss": 0.1416
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 2.2654,
+      "step": 2600,
+      "vm_loss": 0.118
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 1.8445,
+      "step": 2600,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 1.8306,
+      "step": 2600,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 1.7497,
+      "step": 2600,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 1.7837,
+      "step": 2600,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 1.836,
+      "step": 2600,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 0.5005173616959838,
+      "lm_loss": 2.0453,
+      "step": 2600,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 0.5007098683735592,
+      "grad_norm": 2.7794518667196986,
+      "learning_rate": 1.7559537164248333e-05,
+      "loss": 2.2319,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5009023750511346,
+      "grad_norm": 2.8899109317913867,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 2.2734,
+      "step": 2602
+    },
+    {
+      "epoch": 0.50109488172871,
+      "grad_norm": 3.0605169807087447,
+      "learning_rate": 1.755545358815156e-05,
+      "loss": 2.2713,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5012873884062854,
+      "grad_norm": 2.7685890938663333,
+      "learning_rate": 1.7553410698273786e-05,
+      "loss": 2.2534,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5014798950838607,
+      "grad_norm": 2.718212449509204,
+      "learning_rate": 1.7551367074107854e-05,
+      "loss": 2.212,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5016724017614361,
+      "grad_norm": 2.9281001237128073,
+      "learning_rate": 1.754932271585243e-05,
+      "loss": 2.2476,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5018649084390114,
+      "grad_norm": 2.8061292565480263,
+      "learning_rate": 1.754727762370626e-05,
+      "loss": 2.24,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "grad_norm": 2.542839139968227,
+      "learning_rate": 1.7545231797868144e-05,
+      "loss": 2.2834,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.2796,
+      "step": 2608,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 1.9135,
+      "step": 2608,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.4112,
+      "step": 2608,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 1.9448,
+      "step": 2608,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.3266,
+      "step": 2608,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.0836,
+      "step": 2608,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.1216,
+      "step": 2608,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 0.5020574151165869,
+      "lm_loss": 2.3413,
+      "step": 2608,
+      "vm_loss": 0.229
+    },
+    {
+      "epoch": 0.5022499217941623,
+      "grad_norm": 2.719356695124388,
+      "learning_rate": 1.7543185238536967e-05,
+      "loss": 2.2869,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5024424284717376,
+      "grad_norm": 2.730083865426326,
+      "learning_rate": 1.7541137945911676e-05,
+      "loss": 2.257,
+      "step": 2610
+    },
+    {
+      "epoch": 0.502634935149313,
+      "grad_norm": 2.785481551920978,
+      "learning_rate": 1.7539089920191298e-05,
+      "loss": 2.2287,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5028274418268883,
+      "grad_norm": 2.8220640238962504,
+      "learning_rate": 1.753704116157493e-05,
+      "loss": 2.2565,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5030199485044637,
+      "grad_norm": 2.9564037523078004,
+      "learning_rate": 1.753499167026173e-05,
+      "loss": 2.2344,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5032124551820392,
+      "grad_norm": 2.637009324707848,
+      "learning_rate": 1.7532941446450943e-05,
+      "loss": 2.2918,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5034049618596145,
+      "grad_norm": 3.0999628395536387,
+      "learning_rate": 1.753089049034187e-05,
+      "loss": 2.2758,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "grad_norm": 2.8104568800068903,
+      "learning_rate": 1.752883880213389e-05,
+      "loss": 2.2073,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.0801,
+      "step": 2616,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.1941,
+      "step": 2616,
+      "vm_loss": 0.1049
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 1.9485,
+      "step": 2616,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 1.8634,
+      "step": 2616,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.1711,
+      "step": 2616,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.2103,
+      "step": 2616,
+      "vm_loss": 0.2481
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.1504,
+      "step": 2616,
+      "vm_loss": 0.2305
+    },
+    {
+      "epoch": 0.5035974685371899,
+      "lm_loss": 2.0238,
+      "step": 2616,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 0.5037899752147653,
+      "grad_norm": 2.861550022259338,
+      "learning_rate": 1.7526786382026463e-05,
+      "loss": 2.2016,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5039824818923406,
+      "grad_norm": 2.782844768672414,
+      "learning_rate": 1.7524733230219102e-05,
+      "loss": 2.2173,
+      "step": 2618
+    },
+    {
+      "epoch": 0.504174988569916,
+      "grad_norm": 3.18662327250915,
+      "learning_rate": 1.75226793469114e-05,
+      "loss": 2.3004,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5043674952474914,
+      "grad_norm": 2.9953073631385543,
+      "learning_rate": 1.7520624732303024e-05,
+      "loss": 2.279,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5045600019250668,
+      "grad_norm": 2.8556103251754617,
+      "learning_rate": 1.7518569386593708e-05,
+      "loss": 2.2484,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5047525086026422,
+      "grad_norm": 2.7715476509028685,
+      "learning_rate": 1.7516513309983253e-05,
+      "loss": 2.3008,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5049450152802175,
+      "grad_norm": 3.0071218031797,
+      "learning_rate": 1.7514456502671543e-05,
+      "loss": 2.2333,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "grad_norm": 2.81686704029729,
+      "learning_rate": 1.7512398964858523e-05,
+      "loss": 2.2838,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.3301,
+      "step": 2624,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.2091,
+      "step": 2624,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.0045,
+      "step": 2624,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 1.7932,
+      "step": 2624,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.0679,
+      "step": 2624,
+      "vm_loss": 0.2389
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.3463,
+      "step": 2624,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.1037,
+      "step": 2624,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.5051375219577929,
+      "lm_loss": 2.0524,
+      "step": 2624,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.5053300286353682,
+      "grad_norm": 2.9623821652152973,
+      "learning_rate": 1.7510340696744213e-05,
+      "loss": 2.2559,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5055225353129437,
+      "grad_norm": 2.879773491619362,
+      "learning_rate": 1.75082816985287e-05,
+      "loss": 2.2652,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5057150419905191,
+      "grad_norm": 2.8174046980388385,
+      "learning_rate": 1.7506221970412145e-05,
+      "loss": 2.2643,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5059075486680944,
+      "grad_norm": 2.7382722430931645,
+      "learning_rate": 1.7504161512594782e-05,
+      "loss": 2.2383,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5061000553456698,
+      "grad_norm": 2.799785111468018,
+      "learning_rate": 1.7502100325276918e-05,
+      "loss": 2.2455,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5062925620232451,
+      "grad_norm": 2.903663411893443,
+      "learning_rate": 1.7500038408658922e-05,
+      "loss": 2.2606,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5064850687008206,
+      "grad_norm": 2.400303835842584,
+      "learning_rate": 1.7497975762941237e-05,
+      "loss": 2.2132,
+      "step": 2631
+    },
+    {
+      "epoch": 0.506677575378396,
+      "grad_norm": 2.74395779442082,
+      "learning_rate": 1.749591238832438e-05,
+      "loss": 2.1893,
+      "step": 2632
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 1.8266,
+      "step": 2632,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 2.358,
+      "step": 2632,
+      "vm_loss": 0.2234
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 2.102,
+      "step": 2632,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 1.7411,
+      "step": 2632,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 1.9312,
+      "step": 2632,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 1.8583,
+      "step": 2632,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 1.6884,
+      "step": 2632,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.506677575378396,
+      "lm_loss": 2.2887,
+      "step": 2632,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 0.5068700820559713,
+      "grad_norm": 2.7243347315537183,
+      "learning_rate": 1.7493848285008938e-05,
+      "loss": 2.2111,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5070625887335467,
+      "grad_norm": 2.644455835470837,
+      "learning_rate": 1.749178345319557e-05,
+      "loss": 2.2662,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5072550954111221,
+      "grad_norm": 2.982752342627713,
+      "learning_rate": 1.7489717893085007e-05,
+      "loss": 2.2514,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5074476020886974,
+      "grad_norm": 2.966604760254571,
+      "learning_rate": 1.7487651604878037e-05,
+      "loss": 2.2519,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5076401087662729,
+      "grad_norm": 2.655046520402533,
+      "learning_rate": 1.7485584588775536e-05,
+      "loss": 2.228,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5078326154438482,
+      "grad_norm": 2.891107355714645,
+      "learning_rate": 1.7483516844978445e-05,
+      "loss": 2.2483,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5080251221214236,
+      "grad_norm": 2.8454734626339753,
+      "learning_rate": 1.7481448373687775e-05,
+      "loss": 2.2529,
+      "step": 2639
+    },
+    {
+      "epoch": 0.508217628798999,
+      "grad_norm": 3.021217908752849,
+      "learning_rate": 1.747937917510461e-05,
+      "loss": 2.2402,
+      "step": 2640
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 2.3467,
+      "step": 2640,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 1.6598,
+      "step": 2640,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 1.8667,
+      "step": 2640,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 1.8767,
+      "step": 2640,
+      "vm_loss": 0.2188
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 1.8069,
+      "step": 2640,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 2.0007,
+      "step": 2640,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 1.7902,
+      "step": 2640,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 0.508217628798999,
+      "lm_loss": 2.229,
+      "step": 2640,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 0.5084101354765743,
+      "grad_norm": 2.7388572369836663,
+      "learning_rate": 1.7477309249430096e-05,
+      "loss": 2.2098,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5086026421541497,
+      "grad_norm": 2.976865901748515,
+      "learning_rate": 1.747523859686546e-05,
+      "loss": 2.2313,
+      "step": 2642
+    },
+    {
+      "epoch": 0.508795148831725,
+      "grad_norm": 3.249671008927816,
+      "learning_rate": 1.7473167217612e-05,
+      "loss": 2.2461,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5089876555093005,
+      "grad_norm": 2.5726386083371025,
+      "learning_rate": 1.7471095111871076e-05,
+      "loss": 2.2406,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5091801621868759,
+      "grad_norm": 2.726950050216297,
+      "learning_rate": 1.7469022279844123e-05,
+      "loss": 2.2778,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5093726688644512,
+      "grad_norm": 3.1198540195543933,
+      "learning_rate": 1.7466948721732646e-05,
+      "loss": 2.2201,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5095651755420266,
+      "grad_norm": 2.7510537634054866,
+      "learning_rate": 1.7464874437738223e-05,
+      "loss": 2.2555,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "grad_norm": 3.3430748962075296,
+      "learning_rate": 1.7462799428062505e-05,
+      "loss": 2.2254,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 1.9903,
+      "step": 2648,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 2.2131,
+      "step": 2648,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 2.0546,
+      "step": 2648,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 2.2459,
+      "step": 2648,
+      "vm_loss": 0.114
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 1.952,
+      "step": 2648,
+      "vm_loss": 0.2214
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 1.9912,
+      "step": 2648,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 1.7982,
+      "step": 2648,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.5097576822196019,
+      "lm_loss": 2.3354,
+      "step": 2648,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.5099501888971774,
+      "grad_norm": 2.9979161019239604,
+      "learning_rate": 1.74607236929072e-05,
+      "loss": 2.2241,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5101426955747528,
+      "grad_norm": 2.665715609670079,
+      "learning_rate": 1.7458647232474106e-05,
+      "loss": 2.2793,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5103352022523281,
+      "grad_norm": 2.885459234109262,
+      "learning_rate": 1.745657004696508e-05,
+      "loss": 2.2464,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5105277089299035,
+      "grad_norm": 2.6116294195633816,
+      "learning_rate": 1.7454492136582042e-05,
+      "loss": 2.2282,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5107202156074789,
+      "grad_norm": 2.9933898549905487,
+      "learning_rate": 1.7452413501527e-05,
+      "loss": 2.2348,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5109127222850542,
+      "grad_norm": 2.7557118029733236,
+      "learning_rate": 1.7450334142002022e-05,
+      "loss": 2.2753,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5111052289626297,
+      "grad_norm": 2.610904800870158,
+      "learning_rate": 1.7448254058209244e-05,
+      "loss": 2.2483,
+      "step": 2655
+    },
+    {
+      "epoch": 0.511297735640205,
+      "grad_norm": 2.951826491871495,
+      "learning_rate": 1.7446173250350883e-05,
+      "loss": 2.2551,
+      "step": 2656
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 2.2202,
+      "step": 2656,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 2.053,
+      "step": 2656,
+      "vm_loss": 0.247
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 2.1313,
+      "step": 2656,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 1.8586,
+      "step": 2656,
+      "vm_loss": 0.2448
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 2.1471,
+      "step": 2656,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 1.7696,
+      "step": 2656,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 1.9407,
+      "step": 2656,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 0.511297735640205,
+      "lm_loss": 1.9062,
+      "step": 2656,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.5114902423177804,
+      "grad_norm": 2.790749191287158,
+      "learning_rate": 1.7444091718629217e-05,
+      "loss": 2.2562,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5116827489953558,
+      "grad_norm": 2.973697486881595,
+      "learning_rate": 1.74420094632466e-05,
+      "loss": 2.2418,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5118752556729311,
+      "grad_norm": 2.6160886356582522,
+      "learning_rate": 1.743992648440545e-05,
+      "loss": 2.276,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5120677623505066,
+      "grad_norm": 2.741550296601859,
+      "learning_rate": 1.7437842782308262e-05,
+      "loss": 2.2539,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5122602690280819,
+      "grad_norm": 2.882234469762646,
+      "learning_rate": 1.7435758357157597e-05,
+      "loss": 2.2583,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5124527757056573,
+      "grad_norm": 2.7153662882756144,
+      "learning_rate": 1.743367320915609e-05,
+      "loss": 2.2076,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5126452823832327,
+      "grad_norm": 2.8950592504870962,
+      "learning_rate": 1.7431587338506443e-05,
+      "loss": 2.2255,
+      "step": 2663
+    },
+    {
+      "epoch": 0.512837789060808,
+      "grad_norm": 2.932371797674889,
+      "learning_rate": 1.7429500745411426e-05,
+      "loss": 2.2391,
+      "step": 2664
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 2.0825,
+      "step": 2664,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 2.3607,
+      "step": 2664,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 1.8591,
+      "step": 2664,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 2.3858,
+      "step": 2664,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 2.4122,
+      "step": 2664,
+      "vm_loss": 0.2299
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 1.7084,
+      "step": 2664,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 2.2069,
+      "step": 2664,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 0.512837789060808,
+      "lm_loss": 1.9393,
+      "step": 2664,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.5130302957383834,
+      "grad_norm": 2.8143085477673124,
+      "learning_rate": 1.742741343007389e-05,
+      "loss": 2.2604,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5132228024159589,
+      "grad_norm": 2.915248151555896,
+      "learning_rate": 1.7425325392696742e-05,
+      "loss": 2.2219,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5134153090935342,
+      "grad_norm": 2.7949596383277533,
+      "learning_rate": 1.7423236633482966e-05,
+      "loss": 2.1937,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5136078157711096,
+      "grad_norm": 2.956405234343142,
+      "learning_rate": 1.7421147152635622e-05,
+      "loss": 2.2288,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5138003224486849,
+      "grad_norm": 2.695272293516905,
+      "learning_rate": 1.7419056950357828e-05,
+      "loss": 2.197,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5139928291262603,
+      "grad_norm": 2.891992805176657,
+      "learning_rate": 1.7416966026852784e-05,
+      "loss": 2.195,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5141853358038357,
+      "grad_norm": 2.949105197723159,
+      "learning_rate": 1.7414874382323747e-05,
+      "loss": 2.2368,
+      "step": 2671
+    },
+    {
+      "epoch": 0.514377842481411,
+      "grad_norm": 2.6435504260712013,
+      "learning_rate": 1.741278201697406e-05,
+      "loss": 2.2133,
+      "step": 2672
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 2.03,
+      "step": 2672,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 2.121,
+      "step": 2672,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 1.8565,
+      "step": 2672,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 1.8095,
+      "step": 2672,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 1.9122,
+      "step": 2672,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 2.29,
+      "step": 2672,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 2.2422,
+      "step": 2672,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.514377842481411,
+      "lm_loss": 2.3059,
+      "step": 2672,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.5145703491589865,
+      "grad_norm": 2.7479058763492015,
+      "learning_rate": 1.7410688931007122e-05,
+      "loss": 2.2694,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5147628558365618,
+      "grad_norm": 2.8164772207909947,
+      "learning_rate": 1.740859512462641e-05,
+      "loss": 2.2171,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5149553625141372,
+      "grad_norm": 2.7092746900203504,
+      "learning_rate": 1.7406500598035466e-05,
+      "loss": 2.2407,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5151478691917126,
+      "grad_norm": 2.6261928048335776,
+      "learning_rate": 1.7404405351437913e-05,
+      "loss": 2.1885,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5153403758692879,
+      "grad_norm": 2.677192099785163,
+      "learning_rate": 1.7402309385037423e-05,
+      "loss": 2.2378,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5155328825468634,
+      "grad_norm": 2.718749277700357,
+      "learning_rate": 1.740021269903776e-05,
+      "loss": 2.2922,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5157253892244387,
+      "grad_norm": 2.9050086699200204,
+      "learning_rate": 1.7398115293642748e-05,
+      "loss": 2.2505,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "grad_norm": 3.051972292329932,
+      "learning_rate": 1.7396017169056278e-05,
+      "loss": 2.2532,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 1.8824,
+      "step": 2680,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 1.8886,
+      "step": 2680,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 2.0481,
+      "step": 2680,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 2.2578,
+      "step": 2680,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 2.1413,
+      "step": 2680,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 2.1407,
+      "step": 2680,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 2.0647,
+      "step": 2680,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.5159178959020141,
+      "lm_loss": 1.9486,
+      "step": 2680,
+      "vm_loss": 0.2478
+    },
+    {
+      "epoch": 0.5161104025795895,
+      "grad_norm": 2.903780041059459,
+      "learning_rate": 1.7393918325482316e-05,
+      "loss": 2.2337,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5163029092571648,
+      "grad_norm": 3.034083796354088,
+      "learning_rate": 1.7391818763124902e-05,
+      "loss": 2.2708,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5164954159347402,
+      "grad_norm": 2.6052578950476204,
+      "learning_rate": 1.7389718482188128e-05,
+      "loss": 2.2051,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5166879226123157,
+      "grad_norm": 2.8678113364109756,
+      "learning_rate": 1.738761748287618e-05,
+      "loss": 2.2004,
+      "step": 2684
+    },
+    {
+      "epoch": 0.516880429289891,
+      "grad_norm": 3.099690877183125,
+      "learning_rate": 1.73855157653933e-05,
+      "loss": 2.2109,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5170729359674664,
+      "grad_norm": 3.0547528797715455,
+      "learning_rate": 1.738341332994379e-05,
+      "loss": 2.1762,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5172654426450417,
+      "grad_norm": 2.856551937302945,
+      "learning_rate": 1.7381310176732052e-05,
+      "loss": 2.2181,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "grad_norm": 2.835751527725957,
+      "learning_rate": 1.7379206305962525e-05,
+      "loss": 2.226,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 1.9812,
+      "step": 2688,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 1.6525,
+      "step": 2688,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 2.2495,
+      "step": 2688,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 2.166,
+      "step": 2688,
+      "vm_loss": 0.2457
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 2.1551,
+      "step": 2688,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 2.059,
+      "step": 2688,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 2.0375,
+      "step": 2688,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 0.5174579493226171,
+      "lm_loss": 1.7353,
+      "step": 2688,
+      "vm_loss": 0.1242
+    },
+    {
+      "epoch": 0.5176504560001925,
+      "grad_norm": 2.9119493379293977,
+      "learning_rate": 1.737710171783974e-05,
+      "loss": 2.2171,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5178429626777679,
+      "grad_norm": 2.9504745826748042,
+      "learning_rate": 1.7374996412568286e-05,
+      "loss": 2.2538,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5180354693553433,
+      "grad_norm": 2.904252616732406,
+      "learning_rate": 1.7372890390352827e-05,
+      "loss": 2.1859,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5182279760329186,
+      "grad_norm": 2.907993300936597,
+      "learning_rate": 1.7370783651398094e-05,
+      "loss": 2.1808,
+      "step": 2692
+    },
+    {
+      "epoch": 0.518420482710494,
+      "grad_norm": 2.690731702715627,
+      "learning_rate": 1.7368676195908895e-05,
+      "loss": 2.1896,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5186129893880694,
+      "grad_norm": 2.882498152916489,
+      "learning_rate": 1.7366568024090095e-05,
+      "loss": 2.2449,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5188054960656447,
+      "grad_norm": 3.0117456470892465,
+      "learning_rate": 1.7364459136146635e-05,
+      "loss": 2.2272,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "grad_norm": 3.041114105517781,
+      "learning_rate": 1.736234953228353e-05,
+      "loss": 2.2337,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.0561,
+      "step": 2696,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.0122,
+      "step": 2696,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.1078,
+      "step": 2696,
+      "vm_loss": 0.1099
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.1573,
+      "step": 2696,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 1.8558,
+      "step": 2696,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.2854,
+      "step": 2696,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 1.943,
+      "step": 2696,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 0.5189980027432202,
+      "lm_loss": 2.3349,
+      "step": 2696,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.5191905094207956,
+      "grad_norm": 2.6538177271964964,
+      "learning_rate": 1.7360239212705857e-05,
+      "loss": 2.252,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5193830160983709,
+      "grad_norm": 2.9722479288517825,
+      "learning_rate": 1.7358128177618766e-05,
+      "loss": 2.2332,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5195755227759463,
+      "grad_norm": 3.1149764375503075,
+      "learning_rate": 1.735601642722748e-05,
+      "loss": 2.2374,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5197680294535216,
+      "grad_norm": 2.907294994411981,
+      "learning_rate": 1.735390396173729e-05,
+      "loss": 2.2109,
+      "step": 2700
+    },
+    {
+      "epoch": 0.519960536131097,
+      "grad_norm": 2.935189947020246,
+      "learning_rate": 1.7351790781353543e-05,
+      "loss": 2.2964,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5201530428086725,
+      "grad_norm": 2.993038795619102,
+      "learning_rate": 1.734967688628168e-05,
+      "loss": 2.2453,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5203455494862478,
+      "grad_norm": 3.220951353637244,
+      "learning_rate": 1.7347562276727194e-05,
+      "loss": 2.2849,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "grad_norm": 2.8945886064359874,
+      "learning_rate": 1.734544695289565e-05,
+      "loss": 2.2433,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 2.0627,
+      "step": 2704,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 1.9644,
+      "step": 2704,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 2.1807,
+      "step": 2704,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 1.9107,
+      "step": 2704,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 1.859,
+      "step": 2704,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 1.711,
+      "step": 2704,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 2.3529,
+      "step": 2704,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.5205380561638232,
+      "lm_loss": 2.3153,
+      "step": 2704,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 0.5207305628413985,
+      "grad_norm": 3.0467579619785106,
+      "learning_rate": 1.734333091499269e-05,
+      "loss": 2.2089,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5209230695189739,
+      "grad_norm": 3.2824128906538563,
+      "learning_rate": 1.7341214163224016e-05,
+      "loss": 2.228,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5211155761965494,
+      "grad_norm": 3.191283673167816,
+      "learning_rate": 1.7339096697795402e-05,
+      "loss": 2.2052,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5213080828741247,
+      "grad_norm": 3.336251698360061,
+      "learning_rate": 1.7336978518912695e-05,
+      "loss": 2.2402,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5215005895517001,
+      "grad_norm": 2.5873556729138545,
+      "learning_rate": 1.733485962678181e-05,
+      "loss": 2.2573,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5216930962292754,
+      "grad_norm": 2.819207031019723,
+      "learning_rate": 1.7332740021608722e-05,
+      "loss": 2.2262,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5218856029068508,
+      "grad_norm": 2.7390418135056707,
+      "learning_rate": 1.73306197035995e-05,
+      "loss": 2.2327,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "grad_norm": 2.7870993764546146,
+      "learning_rate": 1.7328498672960253e-05,
+      "loss": 2.243,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 2.1499,
+      "step": 2712,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 1.8771,
+      "step": 2712,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 1.8958,
+      "step": 2712,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 1.8842,
+      "step": 2712,
+      "vm_loss": 0.2084
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 2.0225,
+      "step": 2712,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 1.8018,
+      "step": 2712,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 2.153,
+      "step": 2712,
+      "vm_loss": 0.1089
+    },
+    {
+      "epoch": 0.5220781095844262,
+      "lm_loss": 2.3027,
+      "step": 2712,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.5222706162620016,
+      "grad_norm": 2.800823858479726,
+      "learning_rate": 1.7326376929897173e-05,
+      "loss": 2.2074,
+      "step": 2713
+    },
+    {
+      "epoch": 0.522463122939577,
+      "grad_norm": 2.983643744166364,
+      "learning_rate": 1.7324254474616528e-05,
+      "loss": 2.2719,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5226556296171524,
+      "grad_norm": 2.6158176885582503,
+      "learning_rate": 1.7322131307324644e-05,
+      "loss": 2.1819,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5228481362947277,
+      "grad_norm": 3.2589706095271462,
+      "learning_rate": 1.7320007428227918e-05,
+      "loss": 2.2301,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5230406429723031,
+      "grad_norm": 3.0743230770092116,
+      "learning_rate": 1.731788283753282e-05,
+      "loss": 2.2033,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5232331496498784,
+      "grad_norm": 2.796594062480214,
+      "learning_rate": 1.7315757535445884e-05,
+      "loss": 2.1677,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5234256563274539,
+      "grad_norm": 2.9089157334218063,
+      "learning_rate": 1.7313631522173723e-05,
+      "loss": 2.1832,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "grad_norm": 2.789662930529021,
+      "learning_rate": 1.7311504797923006e-05,
+      "loss": 2.2068,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 2.0746,
+      "step": 2720,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 1.8341,
+      "step": 2720,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 2.3011,
+      "step": 2720,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 1.905,
+      "step": 2720,
+      "vm_loss": 0.2093
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 2.3392,
+      "step": 2720,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 1.8519,
+      "step": 2720,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 1.9364,
+      "step": 2720,
+      "vm_loss": 0.1111
+    },
+    {
+      "epoch": 0.5236181630050293,
+      "lm_loss": 2.2202,
+      "step": 2720,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.5238106696826046,
+      "grad_norm": 2.786669712127074,
+      "learning_rate": 1.7309377362900486e-05,
+      "loss": 2.2292,
+      "step": 2721
+    },
+    {
+      "epoch": 0.52400317636018,
+      "grad_norm": 2.8935011917470583,
+      "learning_rate": 1.730724921731297e-05,
+      "loss": 2.2302,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5241956830377553,
+      "grad_norm": 2.6175762482978264,
+      "learning_rate": 1.730512036136734e-05,
+      "loss": 2.246,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5243881897153307,
+      "grad_norm": 2.850558122727029,
+      "learning_rate": 1.7302990795270557e-05,
+      "loss": 2.2079,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5245806963929062,
+      "grad_norm": 2.891895735027978,
+      "learning_rate": 1.730086051922963e-05,
+      "loss": 2.1856,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5247732030704815,
+      "grad_norm": 2.9181011661840923,
+      "learning_rate": 1.729872953345166e-05,
+      "loss": 2.157,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5249657097480569,
+      "grad_norm": 2.8679221618913235,
+      "learning_rate": 1.72965978381438e-05,
+      "loss": 2.281,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "grad_norm": 2.7988322499357987,
+      "learning_rate": 1.7294465433513278e-05,
+      "loss": 2.2143,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 2.127,
+      "step": 2728,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 2.2974,
+      "step": 2728,
+      "vm_loss": 0.2194
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.8778,
+      "step": 2728,
+      "vm_loss": 0.2451
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.9747,
+      "step": 2728,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.9019,
+      "step": 2728,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.7037,
+      "step": 2728,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.9436,
+      "step": 2728,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 0.5251582164256323,
+      "lm_loss": 1.9532,
+      "step": 2728,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 0.5253507231032076,
+      "grad_norm": 2.682672642624575,
+      "learning_rate": 1.7292332319767394e-05,
+      "loss": 2.2503,
+      "step": 2729
+    },
+    {
+      "epoch": 0.525543229780783,
+      "grad_norm": 2.8102973003861247,
+      "learning_rate": 1.7290198497113513e-05,
+      "loss": 2.2087,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5257357364583584,
+      "grad_norm": 2.828460901822948,
+      "learning_rate": 1.728806396575907e-05,
+      "loss": 2.2568,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5259282431359338,
+      "grad_norm": 2.8097666507916483,
+      "learning_rate": 1.7285928725911562e-05,
+      "loss": 2.2266,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5261207498135092,
+      "grad_norm": 2.8892431448880913,
+      "learning_rate": 1.7283792777778575e-05,
+      "loss": 2.2068,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5263132564910845,
+      "grad_norm": 3.0437306974051848,
+      "learning_rate": 1.728165612156774e-05,
+      "loss": 2.2506,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5265057631686599,
+      "grad_norm": 2.6083167184893927,
+      "learning_rate": 1.7279518757486767e-05,
+      "loss": 2.2036,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "grad_norm": 3.3566222342520606,
+      "learning_rate": 1.7277380685743444e-05,
+      "loss": 2.235,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.1634,
+      "step": 2736,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.1888,
+      "step": 2736,
+      "vm_loss": 0.2093
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.0889,
+      "step": 2736,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.1414,
+      "step": 2736,
+      "vm_loss": 0.2849
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.4522,
+      "step": 2736,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 1.9322,
+      "step": 2736,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 1.735,
+      "step": 2736,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.5266982698462352,
+      "lm_loss": 2.0324,
+      "step": 2736,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.5268907765238107,
+      "grad_norm": 2.8248197200189247,
+      "learning_rate": 1.727524190654561e-05,
+      "loss": 2.3045,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5270832832013861,
+      "grad_norm": 2.578205757388353,
+      "learning_rate": 1.7273102420101187e-05,
+      "loss": 2.2532,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5272757898789614,
+      "grad_norm": 2.8895454269757854,
+      "learning_rate": 1.7270962226618156e-05,
+      "loss": 2.1591,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5274682965565368,
+      "grad_norm": 2.9172714154375092,
+      "learning_rate": 1.726882132630458e-05,
+      "loss": 2.1599,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5276608032341121,
+      "grad_norm": 2.9225337437434593,
+      "learning_rate": 1.726667971936857e-05,
+      "loss": 2.2545,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5278533099116876,
+      "grad_norm": 2.8455547028252535,
+      "learning_rate": 1.7264537406018323e-05,
+      "loss": 2.232,
+      "step": 2742
+    },
+    {
+      "epoch": 0.528045816589263,
+      "grad_norm": 2.91143726475438,
+      "learning_rate": 1.7262394386462102e-05,
+      "loss": 2.2398,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "grad_norm": 2.9939401927064826,
+      "learning_rate": 1.7260250660908233e-05,
+      "loss": 2.2467,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 1.624,
+      "step": 2744,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 2.0168,
+      "step": 2744,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 1.7295,
+      "step": 2744,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 2.048,
+      "step": 2744,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 1.7678,
+      "step": 2744,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 2.3112,
+      "step": 2744,
+      "vm_loss": 0.2289
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 2.3074,
+      "step": 2744,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 0.5282383232668383,
+      "lm_loss": 1.7058,
+      "step": 2744,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.5284308299444137,
+      "grad_norm": 2.8288647038892303,
+      "learning_rate": 1.7258106229565115e-05,
+      "loss": 2.215,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5286233366219891,
+      "grad_norm": 2.7066053185053094,
+      "learning_rate": 1.7255961092641212e-05,
+      "loss": 2.1805,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5288158432995644,
+      "grad_norm": 2.984640230345748,
+      "learning_rate": 1.7253815250345058e-05,
+      "loss": 2.2483,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5290083499771399,
+      "grad_norm": 2.8337366341359393,
+      "learning_rate": 1.7251668702885262e-05,
+      "loss": 2.2069,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5292008566547152,
+      "grad_norm": 2.759195177892417,
+      "learning_rate": 1.724952145047049e-05,
+      "loss": 2.2198,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5293933633322906,
+      "grad_norm": 3.2688140959407845,
+      "learning_rate": 1.7247373493309484e-05,
+      "loss": 2.2207,
+      "step": 2750
+    },
+    {
+      "epoch": 0.529585870009866,
+      "grad_norm": 3.1063217777207184,
+      "learning_rate": 1.7245224831611057e-05,
+      "loss": 2.193,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "grad_norm": 2.868409794919299,
+      "learning_rate": 1.724307546558408e-05,
+      "loss": 2.2015,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.2212,
+      "step": 2752,
+      "vm_loss": 0.1071
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.0574,
+      "step": 2752,
+      "vm_loss": 0.2019
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.1518,
+      "step": 2752,
+      "vm_loss": 0.2377
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.2163,
+      "step": 2752,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.1522,
+      "step": 2752,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 1.8956,
+      "step": 2752,
+      "vm_loss": 0.2013
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.2918,
+      "step": 2752,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 0.5297783766874413,
+      "lm_loss": 2.1732,
+      "step": 2752,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.5299708833650167,
+      "grad_norm": 2.6304106171264325,
+      "learning_rate": 1.72409253954375e-05,
+      "loss": 2.2464,
+      "step": 2753
+    },
+    {
+      "epoch": 0.530163390042592,
+      "grad_norm": 2.9630607681296386,
+      "learning_rate": 1.7238774621380335e-05,
+      "loss": 2.2125,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5303558967201675,
+      "grad_norm": 2.8777381660679406,
+      "learning_rate": 1.7236623143621666e-05,
+      "loss": 2.2109,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5305484033977429,
+      "grad_norm": 2.981110883310165,
+      "learning_rate": 1.7234470962370646e-05,
+      "loss": 2.2314,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5307409100753182,
+      "grad_norm": 3.030164749753389,
+      "learning_rate": 1.723231807783649e-05,
+      "loss": 2.2369,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5309334167528936,
+      "grad_norm": 2.857420219750154,
+      "learning_rate": 1.723016449022849e-05,
+      "loss": 2.1863,
+      "step": 2758
+    },
+    {
+      "epoch": 0.531125923430469,
+      "grad_norm": 2.752568237645256,
+      "learning_rate": 1.7228010199756e-05,
+      "loss": 2.2047,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "grad_norm": 2.812141531803607,
+      "learning_rate": 1.722585520662844e-05,
+      "loss": 2.278,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 2.1925,
+      "step": 2760,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 2.1859,
+      "step": 2760,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 1.9906,
+      "step": 2760,
+      "vm_loss": 0.1416
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 2.2004,
+      "step": 2760,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 1.8358,
+      "step": 2760,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 1.9115,
+      "step": 2760,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 2.009,
+      "step": 2760,
+      "vm_loss": 0.2204
+    },
+    {
+      "epoch": 0.5313184301080444,
+      "lm_loss": 1.5956,
+      "step": 2760,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 0.5315109367856198,
+      "grad_norm": 3.1278518058623375,
+      "learning_rate": 1.7223699511055316e-05,
+      "loss": 2.1944,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5317034434631951,
+      "grad_norm": 2.7367063073803077,
+      "learning_rate": 1.722154311324618e-05,
+      "loss": 2.2048,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5318959501407705,
+      "grad_norm": 2.7763989944417538,
+      "learning_rate": 1.721938601341066e-05,
+      "loss": 2.2284,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5320884568183459,
+      "grad_norm": 2.8416085685011425,
+      "learning_rate": 1.721722821175846e-05,
+      "loss": 2.1799,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5322809634959212,
+      "grad_norm": 2.562367566254436,
+      "learning_rate": 1.721506970849934e-05,
+      "loss": 2.1748,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5324734701734967,
+      "grad_norm": 2.7750132557249616,
+      "learning_rate": 1.7212910503843135e-05,
+      "loss": 2.1865,
+      "step": 2766
+    },
+    {
+      "epoch": 0.532665976851072,
+      "grad_norm": 2.941758721362244,
+      "learning_rate": 1.7210750597999753e-05,
+      "loss": 2.1543,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "grad_norm": 3.0968797342934566,
+      "learning_rate": 1.7208589991179157e-05,
+      "loss": 2.1927,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 1.7928,
+      "step": 2768,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 2.2464,
+      "step": 2768,
+      "vm_loss": 0.2022
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 2.2616,
+      "step": 2768,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 2.0468,
+      "step": 2768,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 2.3645,
+      "step": 2768,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 1.9536,
+      "step": 2768,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 2.3235,
+      "step": 2768,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.5328584835286474,
+      "lm_loss": 1.8999,
+      "step": 2768,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 0.5330509902062228,
+      "grad_norm": 3.035133047702225,
+      "learning_rate": 1.7206428683591393e-05,
+      "loss": 2.2184,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5332434968837981,
+      "grad_norm": 3.0391767015149704,
+      "learning_rate": 1.720426667544656e-05,
+      "loss": 2.1596,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5334360035613736,
+      "grad_norm": 3.010789242514809,
+      "learning_rate": 1.720210396695484e-05,
+      "loss": 2.1463,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5336285102389489,
+      "grad_norm": 2.8928833649492063,
+      "learning_rate": 1.719994055832647e-05,
+      "loss": 2.2712,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5338210169165243,
+      "grad_norm": 3.0970783685947745,
+      "learning_rate": 1.7197776449771765e-05,
+      "loss": 2.2125,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5340135235940997,
+      "grad_norm": 3.1253087245775246,
+      "learning_rate": 1.7195611641501103e-05,
+      "loss": 2.2247,
+      "step": 2774
+    },
+    {
+      "epoch": 0.534206030271675,
+      "grad_norm": 2.66580061465288,
+      "learning_rate": 1.719344613372493e-05,
+      "loss": 2.1911,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "grad_norm": 2.8029580271758467,
+      "learning_rate": 1.719127992665376e-05,
+      "loss": 2.2206,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.9056,
+      "step": 2776,
+      "vm_loss": 0.1242
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.9529,
+      "step": 2776,
+      "vm_loss": 0.1151
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 2.3989,
+      "step": 2776,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.9788,
+      "step": 2776,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 2.0475,
+      "step": 2776,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.7666,
+      "step": 2776,
+      "vm_loss": 0.1359
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.6481,
+      "step": 2776,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.5343985369492504,
+      "lm_loss": 1.8702,
+      "step": 2776,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.5345910436268259,
+      "grad_norm": 2.982385439722535,
+      "learning_rate": 1.718911302049818e-05,
+      "loss": 2.2108,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5347835503044012,
+      "grad_norm": 2.9907747946050103,
+      "learning_rate": 1.7186945415468837e-05,
+      "loss": 2.1914,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5349760569819766,
+      "grad_norm": 3.1081625210274915,
+      "learning_rate": 1.7184777111776455e-05,
+      "loss": 2.2222,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5351685636595519,
+      "grad_norm": 2.886793828612081,
+      "learning_rate": 1.7182608109631816e-05,
+      "loss": 2.1566,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5353610703371273,
+      "grad_norm": 3.1565866275434655,
+      "learning_rate": 1.7180438409245774e-05,
+      "loss": 2.2052,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5355535770147027,
+      "grad_norm": 2.9984097833804024,
+      "learning_rate": 1.7178268010829256e-05,
+      "loss": 2.1325,
+      "step": 2782
+    },
+    {
+      "epoch": 0.535746083692278,
+      "grad_norm": 2.969263911422326,
+      "learning_rate": 1.717609691459325e-05,
+      "loss": 2.1218,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "grad_norm": 3.202129451523313,
+      "learning_rate": 1.7173925120748816e-05,
+      "loss": 2.143,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.3837,
+      "step": 2784,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.3103,
+      "step": 2784,
+      "vm_loss": 0.2541
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 1.8356,
+      "step": 2784,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.1483,
+      "step": 2784,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 1.8324,
+      "step": 2784,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.2568,
+      "step": 2784,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.2442,
+      "step": 2784,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.5359385903698535,
+      "lm_loss": 2.0604,
+      "step": 2784,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 0.5361310970474288,
+      "grad_norm": 2.86582319485646,
+      "learning_rate": 1.7171752629507076e-05,
+      "loss": 2.2504,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5363236037250042,
+      "grad_norm": 2.99909770256116,
+      "learning_rate": 1.716957944107923e-05,
+      "loss": 2.2329,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5365161104025796,
+      "grad_norm": 2.9525908257221847,
+      "learning_rate": 1.7167405555676535e-05,
+      "loss": 2.1521,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5367086170801549,
+      "grad_norm": 2.9226957819073816,
+      "learning_rate": 1.7165230973510323e-05,
+      "loss": 2.2156,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5369011237577304,
+      "grad_norm": 2.8647890667620475,
+      "learning_rate": 1.7163055694791987e-05,
+      "loss": 2.2148,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5370936304353058,
+      "grad_norm": 2.736339506663662,
+      "learning_rate": 1.7160879719733002e-05,
+      "loss": 2.1976,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5372861371128811,
+      "grad_norm": 3.3970062916063815,
+      "learning_rate": 1.715870304854489e-05,
+      "loss": 2.2683,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "grad_norm": 2.8899381446193178,
+      "learning_rate": 1.7156525681439252e-05,
+      "loss": 2.199,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 2.022,
+      "step": 2792,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 1.9709,
+      "step": 2792,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 1.962,
+      "step": 2792,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 2.1104,
+      "step": 2792,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 2.4044,
+      "step": 2792,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 2.016,
+      "step": 2792,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 2.1678,
+      "step": 2792,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 0.5374786437904565,
+      "lm_loss": 1.7413,
+      "step": 2792,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 0.5376711504680318,
+      "grad_norm": 2.877157366137414,
+      "learning_rate": 1.7154347618627763e-05,
+      "loss": 2.1965,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5378636571456072,
+      "grad_norm": 2.812799175092424,
+      "learning_rate": 1.7152168860322156e-05,
+      "loss": 2.1877,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5380561638231827,
+      "grad_norm": 2.9351006381214706,
+      "learning_rate": 1.714998940673423e-05,
+      "loss": 2.2072,
+      "step": 2795
+    },
+    {
+      "epoch": 0.538248670500758,
+      "grad_norm": 3.1849733400213878,
+      "learning_rate": 1.7147809258075862e-05,
+      "loss": 2.1958,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5384411771783334,
+      "grad_norm": 2.9396554295358626,
+      "learning_rate": 1.7145628414558982e-05,
+      "loss": 2.2508,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5386336838559087,
+      "grad_norm": 2.724302442394363,
+      "learning_rate": 1.71434468763956e-05,
+      "loss": 2.1736,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5388261905334841,
+      "grad_norm": 2.9456581261913795,
+      "learning_rate": 1.7141264643797796e-05,
+      "loss": 2.1665,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "grad_norm": 3.3403082063681304,
+      "learning_rate": 1.7139081716977704e-05,
+      "loss": 2.1858,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.0765,
+      "step": 2800,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.2672,
+      "step": 2800,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.2168,
+      "step": 2800,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 1.8619,
+      "step": 2800,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 1.931,
+      "step": 2800,
+      "vm_loss": 0.246
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.0091,
+      "step": 2800,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.437,
+      "step": 2800,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.5390186972110595,
+      "lm_loss": 2.2046,
+      "step": 2800,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 0.5392112038886349,
+      "grad_norm": 2.6705208808738887,
+      "learning_rate": 1.7136898096147533e-05,
+      "loss": 2.1814,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5394037105662103,
+      "grad_norm": 3.0522969891492693,
+      "learning_rate": 1.7134713781519556e-05,
+      "loss": 2.2477,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5395962172437856,
+      "grad_norm": 3.5958958471394062,
+      "learning_rate": 1.7132528773306124e-05,
+      "loss": 2.2008,
+      "step": 2803
+    },
+    {
+      "epoch": 0.539788723921361,
+      "grad_norm": 2.8131822172768723,
+      "learning_rate": 1.713034307171964e-05,
+      "loss": 2.1877,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5399812305989364,
+      "grad_norm": 3.153468776070917,
+      "learning_rate": 1.7128156676972587e-05,
+      "loss": 2.2204,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5401737372765117,
+      "grad_norm": 3.482230010017003,
+      "learning_rate": 1.712596958927751e-05,
+      "loss": 2.2671,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5403662439540872,
+      "grad_norm": 2.71188788581524,
+      "learning_rate": 1.7123781808847023e-05,
+      "loss": 2.209,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "grad_norm": 3.2041032310884523,
+      "learning_rate": 1.71215933358938e-05,
+      "loss": 2.2697,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 1.9634,
+      "step": 2808,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 2.0582,
+      "step": 2808,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 1.9243,
+      "step": 2808,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 1.8375,
+      "step": 2808,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 1.8677,
+      "step": 2808,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 2.0188,
+      "step": 2808,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 2.1829,
+      "step": 2808,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 0.5405587506316626,
+      "lm_loss": 2.0511,
+      "step": 2808,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 0.5407512573092379,
+      "grad_norm": 2.9209754893485624,
+      "learning_rate": 1.7119404170630594e-05,
+      "loss": 2.1637,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5409437639868133,
+      "grad_norm": 2.8498416561913658,
+      "learning_rate": 1.7117214313270224e-05,
+      "loss": 2.1383,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5411362706643886,
+      "grad_norm": 3.304679376655516,
+      "learning_rate": 1.7115023764025562e-05,
+      "loss": 2.2082,
+      "step": 2811
+    },
+    {
+      "epoch": 0.541328777341964,
+      "grad_norm": 2.8835366396841406,
+      "learning_rate": 1.711283252310956e-05,
+      "loss": 2.19,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5415212840195395,
+      "grad_norm": 2.690694900171086,
+      "learning_rate": 1.7110640590735243e-05,
+      "loss": 2.1414,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5417137906971148,
+      "grad_norm": 3.153531385476327,
+      "learning_rate": 1.7108447967115686e-05,
+      "loss": 2.2194,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5419062973746902,
+      "grad_norm": 2.801025006421387,
+      "learning_rate": 1.7106254652464045e-05,
+      "loss": 2.1248,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "grad_norm": 3.000140822298546,
+      "learning_rate": 1.7104060646993535e-05,
+      "loss": 2.2304,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 2.0456,
+      "step": 2816,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 2.2187,
+      "step": 2816,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 2.0623,
+      "step": 2816,
+      "vm_loss": 0.2225
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 1.976,
+      "step": 2816,
+      "vm_loss": 0.1916
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 2.1747,
+      "step": 2816,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 1.9973,
+      "step": 2816,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 2.0629,
+      "step": 2816,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 0.5420988040522655,
+      "lm_loss": 1.9232,
+      "step": 2816,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 0.5422913107298409,
+      "grad_norm": 3.155359265008571,
+      "learning_rate": 1.7101865950917443e-05,
+      "loss": 2.1552,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5424838174074164,
+      "grad_norm": 2.651478041574105,
+      "learning_rate": 1.709967056444912e-05,
+      "loss": 2.1768,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5426763240849917,
+      "grad_norm": 2.620746514356997,
+      "learning_rate": 1.7097474487801987e-05,
+      "loss": 2.1803,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5428688307625671,
+      "grad_norm": 2.7817344526616523,
+      "learning_rate": 1.709527772118953e-05,
+      "loss": 2.19,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5430613374401424,
+      "grad_norm": 2.861634875619585,
+      "learning_rate": 1.7093080264825304e-05,
+      "loss": 2.2197,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5432538441177178,
+      "grad_norm": 2.66975632541193,
+      "learning_rate": 1.709088211892293e-05,
+      "loss": 2.1843,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5434463507952932,
+      "grad_norm": 2.8147879522962307,
+      "learning_rate": 1.7088683283696096e-05,
+      "loss": 2.2398,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "grad_norm": 2.7414116432024893,
+      "learning_rate": 1.7086483759358556e-05,
+      "loss": 2.101,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 2.2263,
+      "step": 2824,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 1.8173,
+      "step": 2824,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 2.098,
+      "step": 2824,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 1.7867,
+      "step": 2824,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 1.8203,
+      "step": 2824,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 2.0425,
+      "step": 2824,
+      "vm_loss": 0.1308
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 2.0885,
+      "step": 2824,
+      "vm_loss": 0.254
+    },
+    {
+      "epoch": 0.5436388574728686,
+      "lm_loss": 2.3214,
+      "step": 2824,
+      "vm_loss": 0.2387
+    },
+    {
+      "epoch": 0.543831364150444,
+      "grad_norm": 3.080524081987845,
+      "learning_rate": 1.708428354612413e-05,
+      "loss": 2.1746,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5440238708280194,
+      "grad_norm": 3.1100842289094266,
+      "learning_rate": 1.7082082644206708e-05,
+      "loss": 2.1619,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5442163775055947,
+      "grad_norm": 2.8263010921321268,
+      "learning_rate": 1.707988105382025e-05,
+      "loss": 2.2335,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5444088841831701,
+      "grad_norm": 2.8722930245985814,
+      "learning_rate": 1.7077678775178774e-05,
+      "loss": 2.1514,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5446013908607454,
+      "grad_norm": 2.7052072708399395,
+      "learning_rate": 1.7075475808496368e-05,
+      "loss": 2.1564,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5447938975383209,
+      "grad_norm": 2.9833975971555566,
+      "learning_rate": 1.7073272153987196e-05,
+      "loss": 2.1882,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5449864042158963,
+      "grad_norm": 2.848153359786647,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 2.138,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "grad_norm": 2.7681203239902206,
+      "learning_rate": 1.70688627823455e-05,
+      "loss": 2.1482,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 1.8498,
+      "step": 2832,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 1.7863,
+      "step": 2832,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 2.1311,
+      "step": 2832,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 1.9125,
+      "step": 2832,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 2.0878,
+      "step": 2832,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 1.5625,
+      "step": 2832,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 2.14,
+      "step": 2832,
+      "vm_loss": 0.2575
+    },
+    {
+      "epoch": 0.5451789108934716,
+      "lm_loss": 2.0169,
+      "step": 2832,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 0.545371417571047,
+      "grad_norm": 2.8729993302896286,
+      "learning_rate": 1.7066657065641624e-05,
+      "loss": 2.2296,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5455639242486223,
+      "grad_norm": 2.904684301348938,
+      "learning_rate": 1.706445066196827e-05,
+      "loss": 2.2374,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5457564309261977,
+      "grad_norm": 3.0279667340139156,
+      "learning_rate": 1.7062243571539935e-05,
+      "loss": 2.1113,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5459489376037732,
+      "grad_norm": 2.83401379299048,
+      "learning_rate": 1.706003579457117e-05,
+      "loss": 2.1863,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5461414442813485,
+      "grad_norm": 2.681881971783011,
+      "learning_rate": 1.70578273312766e-05,
+      "loss": 2.2348,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5463339509589239,
+      "grad_norm": 2.9269528542419914,
+      "learning_rate": 1.705561818187092e-05,
+      "loss": 2.1534,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5465264576364993,
+      "grad_norm": 2.8765052311848227,
+      "learning_rate": 1.7053408346568884e-05,
+      "loss": 2.2024,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "grad_norm": 2.74906735945929,
+      "learning_rate": 1.7051197825585316e-05,
+      "loss": 2.1726,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 2.137,
+      "step": 2840,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 2.4126,
+      "step": 2840,
+      "vm_loss": 0.1139
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 1.8027,
+      "step": 2840,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 1.9414,
+      "step": 2840,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 2.0159,
+      "step": 2840,
+      "vm_loss": 0.2042
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 2.03,
+      "step": 2840,
+      "vm_loss": 0.2136
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 1.883,
+      "step": 2840,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 0.5467189643140746,
+      "lm_loss": 1.8566,
+      "step": 2840,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 0.54691147099165,
+      "grad_norm": 2.8649960595451627,
+      "learning_rate": 1.7048986619135108e-05,
+      "loss": 2.1805,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5471039776692254,
+      "grad_norm": 2.943181826477159,
+      "learning_rate": 1.7046774727433223e-05,
+      "loss": 2.2107,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5472964843468008,
+      "grad_norm": 2.700998685676341,
+      "learning_rate": 1.7044562150694675e-05,
+      "loss": 2.1649,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5474889910243762,
+      "grad_norm": 2.7662927302752514,
+      "learning_rate": 1.704234888913456e-05,
+      "loss": 2.1768,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5476814977019515,
+      "grad_norm": 3.134918485389084,
+      "learning_rate": 1.7040134942968037e-05,
+      "loss": 2.2271,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5478740043795269,
+      "grad_norm": 2.6568599048952963,
+      "learning_rate": 1.7037920312410325e-05,
+      "loss": 2.2275,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5480665110571022,
+      "grad_norm": 2.716705190357507,
+      "learning_rate": 1.7035704997676718e-05,
+      "loss": 2.1732,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "grad_norm": 2.877005783063368,
+      "learning_rate": 1.7033488998982573e-05,
+      "loss": 2.2098,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 2.2362,
+      "step": 2848,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 1.9952,
+      "step": 2848,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 2.1946,
+      "step": 2848,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 2.178,
+      "step": 2848,
+      "vm_loss": 0.0952
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 1.4631,
+      "step": 2848,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 2.1477,
+      "step": 2848,
+      "vm_loss": 0.2079
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 2.0217,
+      "step": 2848,
+      "vm_loss": 0.1804
+    },
+    {
+      "epoch": 0.5482590177346777,
+      "lm_loss": 1.9593,
+      "step": 2848,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.5484515244122531,
+      "grad_norm": 2.805428232674696,
+      "learning_rate": 1.703127231654331e-05,
+      "loss": 2.1781,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5486440310898284,
+      "grad_norm": 2.825724704161364,
+      "learning_rate": 1.702905495057443e-05,
+      "loss": 2.1874,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5488365377674038,
+      "grad_norm": 2.9679952407920496,
+      "learning_rate": 1.7026836901291475e-05,
+      "loss": 2.2046,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5490290444449791,
+      "grad_norm": 2.5453465836182896,
+      "learning_rate": 1.7024618168910075e-05,
+      "loss": 2.1409,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5492215511225546,
+      "grad_norm": 2.90415321555807,
+      "learning_rate": 1.7022398753645918e-05,
+      "loss": 2.1689,
+      "step": 2853
+    },
+    {
+      "epoch": 0.54941405780013,
+      "grad_norm": 2.699163942425459,
+      "learning_rate": 1.7020178655714757e-05,
+      "loss": 2.1864,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5496065644777053,
+      "grad_norm": 2.6752299095346834,
+      "learning_rate": 1.7017957875332417e-05,
+      "loss": 2.1632,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "grad_norm": 2.4777289329560186,
+      "learning_rate": 1.7015736412714786e-05,
+      "loss": 2.1717,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 2.3689,
+      "step": 2856,
+      "vm_loss": 0.094
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 1.9439,
+      "step": 2856,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 2.0768,
+      "step": 2856,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 1.8754,
+      "step": 2856,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 2.3018,
+      "step": 2856,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 2.1182,
+      "step": 2856,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 1.9846,
+      "step": 2856,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 0.5497990711552807,
+      "lm_loss": 1.9328,
+      "step": 2856,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.5499915778328561,
+      "grad_norm": 2.858328302557514,
+      "learning_rate": 1.701351426807782e-05,
+      "loss": 2.1587,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5501840845104314,
+      "grad_norm": 2.903914100249301,
+      "learning_rate": 1.7011291441637537e-05,
+      "loss": 2.2268,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5503765911880069,
+      "grad_norm": 3.0091860471084755,
+      "learning_rate": 1.7009067933610027e-05,
+      "loss": 2.1963,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5505690978655822,
+      "grad_norm": 2.503947266375995,
+      "learning_rate": 1.7006843744211437e-05,
+      "loss": 2.2147,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5507616045431576,
+      "grad_norm": 2.7778365266132807,
+      "learning_rate": 1.7004618873657994e-05,
+      "loss": 2.1082,
+      "step": 2861
+    },
+    {
+      "epoch": 0.550954111220733,
+      "grad_norm": 3.128715291659576,
+      "learning_rate": 1.7002393322165982e-05,
+      "loss": 2.2089,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5511466178983083,
+      "grad_norm": 2.7948260365575583,
+      "learning_rate": 1.7000167089951755e-05,
+      "loss": 2.1869,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "grad_norm": 2.7680988132528066,
+      "learning_rate": 1.6997940177231722e-05,
+      "loss": 2.2366,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 2.0758,
+      "step": 2864,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 2.3273,
+      "step": 2864,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 1.9972,
+      "step": 2864,
+      "vm_loss": 0.2417
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 2.1159,
+      "step": 2864,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 1.7565,
+      "step": 2864,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 2.1037,
+      "step": 2864,
+      "vm_loss": 0.2238
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 1.7659,
+      "step": 2864,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.5513391245758837,
+      "lm_loss": 2.0511,
+      "step": 2864,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.551531631253459,
+      "grad_norm": 2.880083912273121,
+      "learning_rate": 1.6995712584222378e-05,
+      "loss": 2.1681,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 3.0684582465130017,
+      "learning_rate": 1.699348431114027e-05,
+      "loss": 2.2094,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5519166446086099,
+      "grad_norm": 2.734666545633309,
+      "learning_rate": 1.6991255358202012e-05,
+      "loss": 2.1669,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5521091512861852,
+      "grad_norm": 2.9895071283120935,
+      "learning_rate": 1.698902572562429e-05,
+      "loss": 2.1927,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5523016579637606,
+      "grad_norm": 2.906917339030937,
+      "learning_rate": 1.698679541362385e-05,
+      "loss": 2.1975,
+      "step": 2869
+    },
+    {
+      "epoch": 0.552494164641336,
+      "grad_norm": 2.9262093702459535,
+      "learning_rate": 1.698456442241751e-05,
+      "loss": 2.1954,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5526866713189114,
+      "grad_norm": 2.8019443629237895,
+      "learning_rate": 1.698233275222215e-05,
+      "loss": 2.2183,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "grad_norm": 3.077509499146404,
+      "learning_rate": 1.6980100403254714e-05,
+      "loss": 2.1781,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.0781,
+      "step": 2872,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.0759,
+      "step": 2872,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.194,
+      "step": 2872,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.2238,
+      "step": 2872,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.0502,
+      "step": 2872,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.0838,
+      "step": 2872,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 1.7142,
+      "step": 2872,
+      "vm_loss": 0.123
+    },
+    {
+      "epoch": 0.5528791779964868,
+      "lm_loss": 2.3354,
+      "step": 2872,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.5530716846740621,
+      "grad_norm": 2.890288961610733,
+      "learning_rate": 1.6977867375732223e-05,
+      "loss": 2.2113,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5532641913516375,
+      "grad_norm": 2.7646624113582723,
+      "learning_rate": 1.6975633669871748e-05,
+      "loss": 2.2386,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5534566980292129,
+      "grad_norm": 3.209625539138657,
+      "learning_rate": 1.697339928589043e-05,
+      "loss": 2.1852,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5536492047067882,
+      "grad_norm": 2.835435847009742,
+      "learning_rate": 1.6971164224005492e-05,
+      "loss": 2.1425,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5538417113843637,
+      "grad_norm": 2.8377171924653344,
+      "learning_rate": 1.69689284844342e-05,
+      "loss": 2.1973,
+      "step": 2877
+    },
+    {
+      "epoch": 0.554034218061939,
+      "grad_norm": 2.8293848207634245,
+      "learning_rate": 1.6966692067393905e-05,
+      "loss": 2.1897,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5542267247395144,
+      "grad_norm": 2.858469689350339,
+      "learning_rate": 1.6964454973102008e-05,
+      "loss": 2.1411,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "grad_norm": 2.758531640707539,
+      "learning_rate": 1.6962217201775986e-05,
+      "loss": 2.2343,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 2.0407,
+      "step": 2880,
+      "vm_loss": 0.1175
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 1.9243,
+      "step": 2880,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 2.1767,
+      "step": 2880,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 2.2357,
+      "step": 2880,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 2.3737,
+      "step": 2880,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 1.9462,
+      "step": 2880,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 2.3412,
+      "step": 2880,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.5544192314170898,
+      "lm_loss": 1.9693,
+      "step": 2880,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 0.5546117380946651,
+      "grad_norm": 2.7825699052644626,
+      "learning_rate": 1.6959978753633382e-05,
+      "loss": 2.1886,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5548042447722406,
+      "grad_norm": 2.878408221535546,
+      "learning_rate": 1.6957739628891797e-05,
+      "loss": 2.2125,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5549967514498159,
+      "grad_norm": 2.8326487998668726,
+      "learning_rate": 1.6955499827768905e-05,
+      "loss": 2.1505,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5551892581273913,
+      "grad_norm": 2.779430971347055,
+      "learning_rate": 1.6953259350482444e-05,
+      "loss": 2.2568,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5553817648049667,
+      "grad_norm": 2.63230265808158,
+      "learning_rate": 1.695101819725021e-05,
+      "loss": 2.0978,
+      "step": 2885
+    },
+    {
+      "epoch": 0.555574271482542,
+      "grad_norm": 2.80406006509171,
+      "learning_rate": 1.6948776368290085e-05,
+      "loss": 2.1764,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5557667781601174,
+      "grad_norm": 2.646791310225615,
+      "learning_rate": 1.6946533863819993e-05,
+      "loss": 2.1503,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "grad_norm": 2.669685720958,
+      "learning_rate": 1.694429068405794e-05,
+      "loss": 2.1791,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.0098,
+      "step": 2888,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.0938,
+      "step": 2888,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.1374,
+      "step": 2888,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 1.973,
+      "step": 2888,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.0896,
+      "step": 2888,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.0662,
+      "step": 2888,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.6298,
+      "step": 2888,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 0.5559592848376929,
+      "lm_loss": 2.1503,
+      "step": 2888,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.5561517915152682,
+      "grad_norm": 2.854749072449087,
+      "learning_rate": 1.6942046829221986e-05,
+      "loss": 2.251,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5563442981928436,
+      "grad_norm": 2.9419846354740224,
+      "learning_rate": 1.6939802299530266e-05,
+      "loss": 2.1314,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5565368048704189,
+      "grad_norm": 2.593903165665498,
+      "learning_rate": 1.6937557095200978e-05,
+      "loss": 2.1317,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5567293115479943,
+      "grad_norm": 2.9585810463133506,
+      "learning_rate": 1.6935311216452382e-05,
+      "loss": 2.1902,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5569218182255697,
+      "grad_norm": 2.714056102608393,
+      "learning_rate": 1.6933064663502805e-05,
+      "loss": 2.1536,
+      "step": 2893
+    },
+    {
+      "epoch": 0.557114324903145,
+      "grad_norm": 2.636195004657239,
+      "learning_rate": 1.6930817436570644e-05,
+      "loss": 2.2033,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5573068315807205,
+      "grad_norm": 2.7343012068680674,
+      "learning_rate": 1.692856953587436e-05,
+      "loss": 2.1655,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "grad_norm": 2.766015931399634,
+      "learning_rate": 1.6926320961632475e-05,
+      "loss": 2.1219,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 1.6178,
+      "step": 2896,
+      "vm_loss": 0.2169
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.216,
+      "step": 2896,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.1938,
+      "step": 2896,
+      "vm_loss": 0.2207
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.2074,
+      "step": 2896,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.2695,
+      "step": 2896,
+      "vm_loss": 0.1932
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.2981,
+      "step": 2896,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 2.1422,
+      "step": 2896,
+      "vm_loss": 0.2147
+    },
+    {
+      "epoch": 0.5574993382582958,
+      "lm_loss": 1.9907,
+      "step": 2896,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.5576918449358712,
+      "grad_norm": 2.847538388995897,
+      "learning_rate": 1.6924071714063574e-05,
+      "loss": 2.2251,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5578843516134466,
+      "grad_norm": 2.8123256232025966,
+      "learning_rate": 1.6921821793386324e-05,
+      "loss": 2.1261,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5580768582910219,
+      "grad_norm": 2.8942544463366713,
+      "learning_rate": 1.6919571199819437e-05,
+      "loss": 2.1939,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5582693649685974,
+      "grad_norm": 2.841089072911991,
+      "learning_rate": 1.69173199335817e-05,
+      "loss": 2.1738,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5584618716461728,
+      "grad_norm": 2.8796405858181204,
+      "learning_rate": 1.6915067994891967e-05,
+      "loss": 2.1771,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5586543783237481,
+      "grad_norm": 2.923301911642831,
+      "learning_rate": 1.6912815383969155e-05,
+      "loss": 2.1657,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5588468850013235,
+      "grad_norm": 2.6916849586351534,
+      "learning_rate": 1.6910562101032253e-05,
+      "loss": 2.1127,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "grad_norm": 2.6608379909050535,
+      "learning_rate": 1.6908308146300297e-05,
+      "loss": 2.1615,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 2.1971,
+      "step": 2904,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 2.5155,
+      "step": 2904,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 2.2244,
+      "step": 2904,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 1.8414,
+      "step": 2904,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 1.6498,
+      "step": 2904,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 2.0944,
+      "step": 2904,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 1.7682,
+      "step": 2904,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 0.5590393916788988,
+      "lm_loss": 1.3593,
+      "step": 2904,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.5592318983564742,
+      "grad_norm": 3.1174771459968276,
+      "learning_rate": 1.690605351999241e-05,
+      "loss": 2.1874,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5594244050340497,
+      "grad_norm": 2.9656710821677392,
+      "learning_rate": 1.6903798222327765e-05,
+      "loss": 2.1838,
+      "step": 2906
+    },
+    {
+      "epoch": 0.559616911711625,
+      "grad_norm": 2.6594457304989936,
+      "learning_rate": 1.6901542253525607e-05,
+      "loss": 2.229,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5598094183892004,
+      "grad_norm": 2.8047743270863275,
+      "learning_rate": 1.6899285613805246e-05,
+      "loss": 2.1749,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5600019250667757,
+      "grad_norm": 2.9705526417826555,
+      "learning_rate": 1.6897028303386057e-05,
+      "loss": 2.1882,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5601944317443511,
+      "grad_norm": 2.7318849990106604,
+      "learning_rate": 1.6894770322487474e-05,
+      "loss": 2.2005,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5603869384219266,
+      "grad_norm": 2.5160188269679615,
+      "learning_rate": 1.6892511671329008e-05,
+      "loss": 2.1305,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "grad_norm": 2.763951952753921,
+      "learning_rate": 1.689025235013023e-05,
+      "loss": 2.1323,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 2.2732,
+      "step": 2912,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 2.1882,
+      "step": 2912,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 1.9254,
+      "step": 2912,
+      "vm_loss": 0.2294
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 1.8028,
+      "step": 2912,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 2.0102,
+      "step": 2912,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 1.9107,
+      "step": 2912,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 2.2454,
+      "step": 2912,
+      "vm_loss": 0.2492
+    },
+    {
+      "epoch": 0.5605794450995019,
+      "lm_loss": 1.7294,
+      "step": 2912,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 0.5607719517770773,
+      "grad_norm": 3.019655143388101,
+      "learning_rate": 1.6887992359110765e-05,
+      "loss": 2.1836,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5609644584546526,
+      "grad_norm": 2.892364744751732,
+      "learning_rate": 1.6885731698490322e-05,
+      "loss": 2.1564,
+      "step": 2914
+    },
+    {
+      "epoch": 0.561156965132228,
+      "grad_norm": 2.771983307646824,
+      "learning_rate": 1.6883470368488666e-05,
+      "loss": 2.1416,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5613494718098034,
+      "grad_norm": 3.039943481023494,
+      "learning_rate": 1.6881208369325622e-05,
+      "loss": 2.1488,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5615419784873787,
+      "grad_norm": 2.8725679981974985,
+      "learning_rate": 1.6878945701221092e-05,
+      "loss": 2.1768,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5617344851649542,
+      "grad_norm": 3.00403542377559,
+      "learning_rate": 1.687668236439503e-05,
+      "loss": 2.1777,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5619269918425296,
+      "grad_norm": 2.7069314019052078,
+      "learning_rate": 1.6874418359067463e-05,
+      "loss": 2.2706,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "grad_norm": 2.991886744904629,
+      "learning_rate": 1.6872153685458484e-05,
+      "loss": 2.1828,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 2.0101,
+      "step": 2920,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 1.9062,
+      "step": 2920,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 2.3523,
+      "step": 2920,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 1.8348,
+      "step": 2920,
+      "vm_loss": 0.1246
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 2.1886,
+      "step": 2920,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 2.1179,
+      "step": 2920,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 2.0022,
+      "step": 2920,
+      "vm_loss": 0.2107
+    },
+    {
+      "epoch": 0.5621194985201049,
+      "lm_loss": 1.7119,
+      "step": 2920,
+      "vm_loss": 0.2178
+    },
+    {
+      "epoch": 0.5623120051976803,
+      "grad_norm": 2.78633237825189,
+      "learning_rate": 1.6869888343788244e-05,
+      "loss": 2.1791,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5625045118752556,
+      "grad_norm": 2.953817775717228,
+      "learning_rate": 1.6867622334276966e-05,
+      "loss": 2.09,
+      "step": 2922
+    },
+    {
+      "epoch": 0.562697018552831,
+      "grad_norm": 2.90390007602645,
+      "learning_rate": 1.6865355657144935e-05,
+      "loss": 2.1343,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5628895252304065,
+      "grad_norm": 2.678492383665957,
+      "learning_rate": 1.68630883126125e-05,
+      "loss": 2.1936,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5630820319079818,
+      "grad_norm": 2.9692553787278317,
+      "learning_rate": 1.6860820300900077e-05,
+      "loss": 2.167,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5632745385855572,
+      "grad_norm": 2.778151163224821,
+      "learning_rate": 1.6858551622228144e-05,
+      "loss": 2.0763,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5634670452631325,
+      "grad_norm": 2.9616010830806627,
+      "learning_rate": 1.685628227681725e-05,
+      "loss": 2.1559,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "grad_norm": 2.8145982251849286,
+      "learning_rate": 1.6854012264888e-05,
+      "loss": 2.1599,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 2.0825,
+      "step": 2928,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 2.1065,
+      "step": 2928,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 1.75,
+      "step": 2928,
+      "vm_loss": 0.1011
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 1.3023,
+      "step": 2928,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 2.2106,
+      "step": 2928,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 1.954,
+      "step": 2928,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 1.8126,
+      "step": 2928,
+      "vm_loss": 0.2232
+    },
+    {
+      "epoch": 0.5636595519407079,
+      "lm_loss": 1.8773,
+      "step": 2928,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.5638520586182834,
+      "grad_norm": 2.940348634021967,
+      "learning_rate": 1.685174158666107e-05,
+      "loss": 2.2084,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5640445652958587,
+      "grad_norm": 2.770211973360341,
+      "learning_rate": 1.6849470242357197e-05,
+      "loss": 2.1725,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5642370719734341,
+      "grad_norm": 2.9379068487275615,
+      "learning_rate": 1.6847198232197184e-05,
+      "loss": 2.2058,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5644295786510095,
+      "grad_norm": 3.0554261235661193,
+      "learning_rate": 1.6844925556401906e-05,
+      "loss": 2.1566,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5646220853285848,
+      "grad_norm": 3.0900310723957425,
+      "learning_rate": 1.6842652215192287e-05,
+      "loss": 2.1493,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5648145920061602,
+      "grad_norm": 3.157840594279967,
+      "learning_rate": 1.6840378208789334e-05,
+      "loss": 2.1677,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5650070986837356,
+      "grad_norm": 2.9556818086969416,
+      "learning_rate": 1.6838103537414103e-05,
+      "loss": 2.1675,
+      "step": 2935
+    },
+    {
+      "epoch": 0.565199605361311,
+      "grad_norm": 2.7230589166538777,
+      "learning_rate": 1.6835828201287723e-05,
+      "loss": 2.13,
+      "step": 2936
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 2.2277,
+      "step": 2936,
+      "vm_loss": 0.2264
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 1.8808,
+      "step": 2936,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 1.9247,
+      "step": 2936,
+      "vm_loss": 0.242
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 1.9168,
+      "step": 2936,
+      "vm_loss": 0.2063
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 1.9045,
+      "step": 2936,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 2.1572,
+      "step": 2936,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 2.1727,
+      "step": 2936,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 0.565199605361311,
+      "lm_loss": 2.1261,
+      "step": 2936,
+      "vm_loss": 0.201
+    },
+    {
+      "epoch": 0.5653921120388864,
+      "grad_norm": 2.9386711762826883,
+      "learning_rate": 1.6833552200631388e-05,
+      "loss": 2.1348,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5655846187164617,
+      "grad_norm": 2.920716062415379,
+      "learning_rate": 1.683127553566635e-05,
+      "loss": 2.1997,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5657771253940371,
+      "grad_norm": 2.909017783735447,
+      "learning_rate": 1.6828998206613937e-05,
+      "loss": 2.2063,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5659696320716124,
+      "grad_norm": 3.0186595681504036,
+      "learning_rate": 1.682672021369552e-05,
+      "loss": 2.1127,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5661621387491879,
+      "grad_norm": 2.7402832013425633,
+      "learning_rate": 1.682444155713257e-05,
+      "loss": 2.1851,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5663546454267633,
+      "grad_norm": 2.861908479368446,
+      "learning_rate": 1.6822162237146585e-05,
+      "loss": 2.1627,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5665471521043386,
+      "grad_norm": 3.093710581956597,
+      "learning_rate": 1.681988225395915e-05,
+      "loss": 2.1038,
+      "step": 2943
+    },
+    {
+      "epoch": 0.566739658781914,
+      "grad_norm": 2.8830932876222426,
+      "learning_rate": 1.681760160779191e-05,
+      "loss": 2.159,
+      "step": 2944
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.2174,
+      "step": 2944,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.1765,
+      "step": 2944,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.1171,
+      "step": 2944,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.0809,
+      "step": 2944,
+      "vm_loss": 0.2162
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.4125,
+      "step": 2944,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 1.9601,
+      "step": 2944,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 2.2961,
+      "step": 2944,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 0.566739658781914,
+      "lm_loss": 1.9124,
+      "step": 2944,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.5669321654594893,
+      "grad_norm": 2.956744303687958,
+      "learning_rate": 1.6815320298866566e-05,
+      "loss": 2.2233,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5671246721370647,
+      "grad_norm": 2.804794520316079,
+      "learning_rate": 1.6813038327404903e-05,
+      "loss": 2.1572,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5673171788146402,
+      "grad_norm": 2.716289944626574,
+      "learning_rate": 1.6810755693628747e-05,
+      "loss": 2.1972,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5675096854922155,
+      "grad_norm": 2.8787525521346584,
+      "learning_rate": 1.680847239776e-05,
+      "loss": 2.1898,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5677021921697909,
+      "grad_norm": 3.0569900637485343,
+      "learning_rate": 1.6806188440020635e-05,
+      "loss": 2.1302,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5678946988473663,
+      "grad_norm": 2.9551687345044777,
+      "learning_rate": 1.6803903820632674e-05,
+      "loss": 2.1972,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5680872055249416,
+      "grad_norm": 2.9781623382651516,
+      "learning_rate": 1.6801618539818215e-05,
+      "loss": 2.2128,
+      "step": 2951
+    },
+    {
+      "epoch": 0.568279712202517,
+      "grad_norm": 2.952464885218521,
+      "learning_rate": 1.6799332597799413e-05,
+      "loss": 2.204,
+      "step": 2952
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 1.9899,
+      "step": 2952,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 1.7825,
+      "step": 2952,
+      "vm_loss": 0.2561
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 1.7164,
+      "step": 2952,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 2.0708,
+      "step": 2952,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 2.2554,
+      "step": 2952,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 1.7811,
+      "step": 2952,
+      "vm_loss": 0.2239
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 2.3576,
+      "step": 2952,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 0.568279712202517,
+      "lm_loss": 1.6213,
+      "step": 2952,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.5684722188800924,
+      "grad_norm": 3.111689000237369,
+      "learning_rate": 1.6797045994798497e-05,
+      "loss": 2.1154,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5686647255576678,
+      "grad_norm": 2.9058315009195947,
+      "learning_rate": 1.679475873103775e-05,
+      "loss": 2.0712,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5688572322352432,
+      "grad_norm": 2.5687230547297975,
+      "learning_rate": 1.6792470806739522e-05,
+      "loss": 2.1406,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5690497389128185,
+      "grad_norm": 3.2409232169140587,
+      "learning_rate": 1.6790182222126237e-05,
+      "loss": 2.1877,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5692422455903939,
+      "grad_norm": 2.798838308512494,
+      "learning_rate": 1.678789297742036e-05,
+      "loss": 2.1487,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5694347522679692,
+      "grad_norm": 3.0632329001100884,
+      "learning_rate": 1.6785603072844446e-05,
+      "loss": 2.1333,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5696272589455447,
+      "grad_norm": 2.857039426648582,
+      "learning_rate": 1.6783312508621105e-05,
+      "loss": 2.1017,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "grad_norm": 3.066759338733449,
+      "learning_rate": 1.6781021284972998e-05,
+      "loss": 2.1359,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 1.8409,
+      "step": 2960,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 2.053,
+      "step": 2960,
+      "vm_loss": 0.2123
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 2.0793,
+      "step": 2960,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 1.9363,
+      "step": 2960,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 2.1015,
+      "step": 2960,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 1.9023,
+      "step": 2960,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 2.4383,
+      "step": 2960,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 0.5698197656231201,
+      "lm_loss": 1.9681,
+      "step": 2960,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.5700122723006954,
+      "grad_norm": 2.909279826468173,
+      "learning_rate": 1.677872940212287e-05,
+      "loss": 2.1331,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5702047789782708,
+      "grad_norm": 2.588324998584431,
+      "learning_rate": 1.677643686029352e-05,
+      "loss": 2.1342,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5703972856558462,
+      "grad_norm": 2.808406413835792,
+      "learning_rate": 1.677414365970781e-05,
+      "loss": 2.162,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5705897923334216,
+      "grad_norm": 3.0400637773965977,
+      "learning_rate": 1.6771849800588665e-05,
+      "loss": 2.2225,
+      "step": 2964
+    },
+    {
+      "epoch": 0.570782299010997,
+      "grad_norm": 3.140516871478318,
+      "learning_rate": 1.676955528315909e-05,
+      "loss": 2.1632,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5709748056885723,
+      "grad_norm": 2.953478609795563,
+      "learning_rate": 1.6767260107642128e-05,
+      "loss": 2.1879,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5711673123661477,
+      "grad_norm": 2.9332075927894175,
+      "learning_rate": 1.6764964274260908e-05,
+      "loss": 2.1155,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "grad_norm": 2.847837865025105,
+      "learning_rate": 1.676266778323861e-05,
+      "loss": 2.1645,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.1508,
+      "step": 2968,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.1353,
+      "step": 2968,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.0621,
+      "step": 2968,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 1.7121,
+      "step": 2968,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.2223,
+      "step": 2968,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.2704,
+      "step": 2968,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 2.3551,
+      "step": 2968,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 0.5713598190437231,
+      "lm_loss": 1.8017,
+      "step": 2968,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 0.5715523257212984,
+      "grad_norm": 2.9490681437606456,
+      "learning_rate": 1.6760370634798486e-05,
+      "loss": 2.1815,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5717448323988739,
+      "grad_norm": 3.068313561260416,
+      "learning_rate": 1.6758072829163843e-05,
+      "loss": 2.1503,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5719373390764492,
+      "grad_norm": 2.860764696646835,
+      "learning_rate": 1.6755774366558064e-05,
+      "loss": 2.1224,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5721298457540246,
+      "grad_norm": 2.8943423145574476,
+      "learning_rate": 1.6753475247204577e-05,
+      "loss": 2.1219,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5723223524316,
+      "grad_norm": 2.9453364746513673,
+      "learning_rate": 1.67511754713269e-05,
+      "loss": 2.1063,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5725148591091753,
+      "grad_norm": 3.022413204512645,
+      "learning_rate": 1.6748875039148592e-05,
+      "loss": 2.2033,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5727073657867507,
+      "grad_norm": 3.1293250943764863,
+      "learning_rate": 1.674657395089329e-05,
+      "loss": 2.1785,
+      "step": 2975
+    },
+    {
+      "epoch": 0.572899872464326,
+      "grad_norm": 2.7538465108666155,
+      "learning_rate": 1.6744272206784685e-05,
+      "loss": 2.1158,
+      "step": 2976
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 1.6389,
+      "step": 2976,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 2.0219,
+      "step": 2976,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 1.8265,
+      "step": 2976,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 1.739,
+      "step": 2976,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 2.0564,
+      "step": 2976,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 1.9844,
+      "step": 2976,
+      "vm_loss": 0.0958
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 2.5677,
+      "step": 2976,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 0.572899872464326,
+      "lm_loss": 1.9938,
+      "step": 2976,
+      "vm_loss": 0.1352
+    },
+    {
+      "epoch": 0.5730923791419015,
+      "grad_norm": 2.65350249387357,
+      "learning_rate": 1.6741969807046533e-05,
+      "loss": 2.1802,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5732848858194769,
+      "grad_norm": 2.619306264450932,
+      "learning_rate": 1.673966675190267e-05,
+      "loss": 2.107,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5734773924970522,
+      "grad_norm": 2.8808668419166326,
+      "learning_rate": 1.6737363041576963e-05,
+      "loss": 2.176,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5736698991746276,
+      "grad_norm": 2.9735296681186147,
+      "learning_rate": 1.673505867629338e-05,
+      "loss": 2.1654,
+      "step": 2980
+    },
+    {
+      "epoch": 0.573862405852203,
+      "grad_norm": 2.603425908930411,
+      "learning_rate": 1.673275365627592e-05,
+      "loss": 2.1068,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5740549125297784,
+      "grad_norm": 2.9192074879905476,
+      "learning_rate": 1.6730447981748672e-05,
+      "loss": 2.1754,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5742474192073538,
+      "grad_norm": 2.7902590181730185,
+      "learning_rate": 1.6728141652935773e-05,
+      "loss": 2.1504,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "grad_norm": 3.0914333151363387,
+      "learning_rate": 1.6725834670061427e-05,
+      "loss": 2.1831,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 1.9061,
+      "step": 2984,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 1.942,
+      "step": 2984,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 2.0901,
+      "step": 2984,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 1.9405,
+      "step": 2984,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 2.0959,
+      "step": 2984,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 2.287,
+      "step": 2984,
+      "vm_loss": 0.244
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 2.0566,
+      "step": 2984,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.5744399258849291,
+      "lm_loss": 2.1861,
+      "step": 2984,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 0.5746324325625045,
+      "grad_norm": 2.946787245867567,
+      "learning_rate": 1.67235270333499e-05,
+      "loss": 2.1674,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5748249392400799,
+      "grad_norm": 3.0645177447167526,
+      "learning_rate": 1.672121874302553e-05,
+      "loss": 2.1586,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5750174459176552,
+      "grad_norm": 2.9410114812127044,
+      "learning_rate": 1.6718909799312705e-05,
+      "loss": 2.1419,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5752099525952307,
+      "grad_norm": 4.221624039449748,
+      "learning_rate": 1.671660020243589e-05,
+      "loss": 2.1172,
+      "step": 2988
+    },
+    {
+      "epoch": 0.575402459272806,
+      "grad_norm": 3.193419327774961,
+      "learning_rate": 1.6714289952619602e-05,
+      "loss": 2.1837,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5755949659503814,
+      "grad_norm": 2.9341434918566116,
+      "learning_rate": 1.6711979050088434e-05,
+      "loss": 2.1157,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5757874726279568,
+      "grad_norm": 2.776906643403077,
+      "learning_rate": 1.6709667495067027e-05,
+      "loss": 2.1757,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "grad_norm": 2.9222000246874096,
+      "learning_rate": 1.67073552877801e-05,
+      "loss": 2.0911,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.9962,
+      "step": 2992,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.7887,
+      "step": 2992,
+      "vm_loss": 0.2351
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.9677,
+      "step": 2992,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 2.0608,
+      "step": 2992,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.9229,
+      "step": 2992,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.9755,
+      "step": 2992,
+      "vm_loss": 0.222
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.7754,
+      "step": 2992,
+      "vm_loss": 0.1242
+    },
+    {
+      "epoch": 0.5759799793055321,
+      "lm_loss": 1.5932,
+      "step": 2992,
+      "vm_loss": 0.2292
+    },
+    {
+      "epoch": 0.5761724859831076,
+      "grad_norm": 3.077133222557676,
+      "learning_rate": 1.6705042428452428e-05,
+      "loss": 2.1646,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5763649926606829,
+      "grad_norm": 3.196324918071841,
+      "learning_rate": 1.6702728917308847e-05,
+      "loss": 2.1238,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5765574993382583,
+      "grad_norm": 2.674037565069713,
+      "learning_rate": 1.6700414754574265e-05,
+      "loss": 2.1051,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5767500060158337,
+      "grad_norm": 2.88322051954885,
+      "learning_rate": 1.6698099940473644e-05,
+      "loss": 2.1877,
+      "step": 2996
+    },
+    {
+      "epoch": 0.576942512693409,
+      "grad_norm": 3.042383218322858,
+      "learning_rate": 1.6695784475232012e-05,
+      "loss": 2.0791,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5771350193709844,
+      "grad_norm": 3.012044312350685,
+      "learning_rate": 1.669346835907447e-05,
+      "loss": 2.1168,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5773275260485599,
+      "grad_norm": 2.9135181808340525,
+      "learning_rate": 1.6691151592226168e-05,
+      "loss": 2.2086,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "grad_norm": 3.0163349891080897,
+      "learning_rate": 1.6688834174912326e-05,
+      "loss": 2.1839,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 1.8148,
+      "step": 3000,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 2.0436,
+      "step": 3000,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 1.7701,
+      "step": 3000,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 1.7789,
+      "step": 3000,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 2.4492,
+      "step": 3000,
+      "vm_loss": 0.1039
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 2.2197,
+      "step": 3000,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 2.2634,
+      "step": 3000,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.5775200327261352,
+      "lm_loss": 2.1622,
+      "step": 3000,
+      "vm_loss": 0.2332
+    },
+    {
+      "epoch": 0.5777125394037106,
+      "grad_norm": 2.618486873029357,
+      "learning_rate": 1.6686516107358225e-05,
+      "loss": 2.1332,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5779050460812859,
+      "grad_norm": 2.7642437850936843,
+      "learning_rate": 1.6684197389789214e-05,
+      "loss": 2.1207,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5780975527588613,
+      "grad_norm": 2.841408811008923,
+      "learning_rate": 1.66818780224307e-05,
+      "loss": 2.2031,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5782900594364367,
+      "grad_norm": 3.198645434761591,
+      "learning_rate": 1.6679558005508157e-05,
+      "loss": 2.1408,
+      "step": 3004
+    },
+    {
+      "epoch": 0.578482566114012,
+      "grad_norm": 2.7723782081090635,
+      "learning_rate": 1.667723733924712e-05,
+      "loss": 2.1571,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5786750727915875,
+      "grad_norm": 2.853256361346693,
+      "learning_rate": 1.6674916023873186e-05,
+      "loss": 2.155,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5788675794691628,
+      "grad_norm": 2.9502615222138635,
+      "learning_rate": 1.6672594059612018e-05,
+      "loss": 2.1177,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "grad_norm": 3.1401898369039767,
+      "learning_rate": 1.667027144668934e-05,
+      "loss": 2.1398,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 1.7484,
+      "step": 3008,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 2.686,
+      "step": 3008,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 2.0838,
+      "step": 3008,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 1.9938,
+      "step": 3008,
+      "vm_loss": 0.1258
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 1.6312,
+      "step": 3008,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 2.0108,
+      "step": 3008,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 2.2867,
+      "step": 3008,
+      "vm_loss": 0.184
+    },
+    {
+      "epoch": 0.5790600861467382,
+      "lm_loss": 2.1133,
+      "step": 3008,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.5792525928243136,
+      "grad_norm": 3.137967040854821,
+      "learning_rate": 1.666794818533094e-05,
+      "loss": 2.1429,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5794450995018889,
+      "grad_norm": 2.6415181296743824,
+      "learning_rate": 1.6665624275762663e-05,
+      "loss": 2.1185,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5796376061794644,
+      "grad_norm": 2.8699039466992473,
+      "learning_rate": 1.6663299718210435e-05,
+      "loss": 2.1734,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5798301128570398,
+      "grad_norm": 3.009270528154217,
+      "learning_rate": 1.6660974512900226e-05,
+      "loss": 2.1735,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5800226195346151,
+      "grad_norm": 2.710147199758013,
+      "learning_rate": 1.6658648660058073e-05,
+      "loss": 2.1425,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5802151262121905,
+      "grad_norm": 2.9133236226274435,
+      "learning_rate": 1.665632215991008e-05,
+      "loss": 2.1326,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5804076328897658,
+      "grad_norm": 2.8180354871077666,
+      "learning_rate": 1.6653995012682414e-05,
+      "loss": 2.1402,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "grad_norm": 2.7642663465213557,
+      "learning_rate": 1.6651667218601306e-05,
+      "loss": 2.1586,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.9077,
+      "step": 3016,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.786,
+      "step": 3016,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.523,
+      "step": 3016,
+      "vm_loss": 0.1128
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 2.0013,
+      "step": 3016,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.9704,
+      "step": 3016,
+      "vm_loss": 0.2197
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.8691,
+      "step": 3016,
+      "vm_loss": 0.1074
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 2.0971,
+      "step": 3016,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.5806001395673412,
+      "lm_loss": 1.9733,
+      "step": 3016,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.5807926462449167,
+      "grad_norm": 3.371445079678996,
+      "learning_rate": 1.6649338777893043e-05,
+      "loss": 2.1184,
+      "step": 3017
+    },
+    {
+      "epoch": 0.580985152922492,
+      "grad_norm": 2.7310585057137735,
+      "learning_rate": 1.664700969078398e-05,
+      "loss": 2.1046,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5811776596000674,
+      "grad_norm": 2.6919097245860715,
+      "learning_rate": 1.6644679957500537e-05,
+      "loss": 2.1296,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5813701662776427,
+      "grad_norm": 2.730289319124603,
+      "learning_rate": 1.6642349578269192e-05,
+      "loss": 2.1523,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5815626729552181,
+      "grad_norm": 2.745369352029728,
+      "learning_rate": 1.6640018553316486e-05,
+      "loss": 2.1834,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5817551796327936,
+      "grad_norm": 2.6476482993923125,
+      "learning_rate": 1.6637686882869027e-05,
+      "loss": 2.1435,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5819476863103689,
+      "grad_norm": 2.913432247629352,
+      "learning_rate": 1.6635354567153482e-05,
+      "loss": 2.1485,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "grad_norm": 3.0774270333031573,
+      "learning_rate": 1.663302160639658e-05,
+      "loss": 2.091,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 1.9161,
+      "step": 3024,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 2.298,
+      "step": 3024,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 2.1439,
+      "step": 3024,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 1.6152,
+      "step": 3024,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 1.9405,
+      "step": 3024,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 2.1522,
+      "step": 3024,
+      "vm_loss": 0.1028
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 2.1761,
+      "step": 3024,
+      "vm_loss": 0.1147
+    },
+    {
+      "epoch": 0.5821401929879443,
+      "lm_loss": 1.6991,
+      "step": 3024,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 0.5823326996655196,
+      "grad_norm": 2.748384072711639,
+      "learning_rate": 1.663068800082512e-05,
+      "loss": 2.0845,
+      "step": 3025
+    },
+    {
+      "epoch": 0.582525206343095,
+      "grad_norm": 3.312325788378188,
+      "learning_rate": 1.6628353750665954e-05,
+      "loss": 2.0968,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5827177130206704,
+      "grad_norm": 3.1289334873310395,
+      "learning_rate": 1.6626018856146e-05,
+      "loss": 2.1256,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5829102196982457,
+      "grad_norm": 2.7340214554749327,
+      "learning_rate": 1.6623683317492247e-05,
+      "loss": 2.1634,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5831027263758212,
+      "grad_norm": 3.048497445691234,
+      "learning_rate": 1.662134713493173e-05,
+      "loss": 2.1602,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5832952330533966,
+      "grad_norm": 3.167323346776916,
+      "learning_rate": 1.6619010308691558e-05,
+      "loss": 2.1507,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5834877397309719,
+      "grad_norm": 2.7814775701561465,
+      "learning_rate": 1.6616672838998904e-05,
+      "loss": 2.2123,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "grad_norm": 2.773021395966257,
+      "learning_rate": 1.6614334726080997e-05,
+      "loss": 2.1371,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 1.6373,
+      "step": 3032,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 1.6518,
+      "step": 3032,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 2.0603,
+      "step": 3032,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 2.199,
+      "step": 3032,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 2.1185,
+      "step": 3032,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 1.6362,
+      "step": 3032,
+      "vm_loss": 0.2011
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 2.0823,
+      "step": 3032,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 0.5836802464085473,
+      "lm_loss": 1.6695,
+      "step": 3032,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.5838727530861226,
+      "grad_norm": 3.151178804919212,
+      "learning_rate": 1.6611995970165137e-05,
+      "loss": 2.1769,
+      "step": 3033
+    },
+    {
+      "epoch": 0.584065259763698,
+      "grad_norm": 2.7766689792106063,
+      "learning_rate": 1.6609656571478673e-05,
+      "loss": 2.1652,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5842577664412735,
+      "grad_norm": 2.993634674309778,
+      "learning_rate": 1.660731653024903e-05,
+      "loss": 2.112,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5844502731188488,
+      "grad_norm": 2.923656336549419,
+      "learning_rate": 1.6604975846703685e-05,
+      "loss": 2.1523,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5846427797964242,
+      "grad_norm": 3.101385612578588,
+      "learning_rate": 1.6602634521070185e-05,
+      "loss": 2.164,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5848352864739995,
+      "grad_norm": 2.873618086647502,
+      "learning_rate": 1.6600292553576142e-05,
+      "loss": 2.1445,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5850277931515749,
+      "grad_norm": 3.2203575151247286,
+      "learning_rate": 1.6597949944449214e-05,
+      "loss": 2.0717,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "grad_norm": 2.7598521391898214,
+      "learning_rate": 1.659560669391714e-05,
+      "loss": 2.1171,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 1.9191,
+      "step": 3040,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 2.2079,
+      "step": 3040,
+      "vm_loss": 0.0775
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 1.8304,
+      "step": 3040,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 2.173,
+      "step": 3040,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 2.1856,
+      "step": 3040,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 2.1718,
+      "step": 3040,
+      "vm_loss": 0.2608
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 1.809,
+      "step": 3040,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 0.5852202998291504,
+      "lm_loss": 1.8568,
+      "step": 3040,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 0.5854128065067257,
+      "grad_norm": 2.83765343751536,
+      "learning_rate": 1.6593262802207718e-05,
+      "loss": 2.1945,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5856053131843011,
+      "grad_norm": 3.210067274546048,
+      "learning_rate": 1.6590918269548798e-05,
+      "loss": 2.1865,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5857978198618765,
+      "grad_norm": 3.0551239908079775,
+      "learning_rate": 1.6588573096168298e-05,
+      "loss": 2.1211,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5859903265394518,
+      "grad_norm": 2.9643746658660515,
+      "learning_rate": 1.65862272822942e-05,
+      "loss": 2.1489,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5861828332170272,
+      "grad_norm": 3.0571060599250917,
+      "learning_rate": 1.6583880828154545e-05,
+      "loss": 2.1398,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5863753398946026,
+      "grad_norm": 2.8324813461231853,
+      "learning_rate": 1.6581533733977448e-05,
+      "loss": 2.1551,
+      "step": 3046
+    },
+    {
+      "epoch": 0.586567846572178,
+      "grad_norm": 2.6363970533204686,
+      "learning_rate": 1.6579185999991066e-05,
+      "loss": 2.0789,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "grad_norm": 3.1069228929429467,
+      "learning_rate": 1.657683762642363e-05,
+      "loss": 2.1731,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.9754,
+      "step": 3048,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.7068,
+      "step": 3048,
+      "vm_loss": 0.1165
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 2.2438,
+      "step": 3048,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 2.1466,
+      "step": 3048,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.9538,
+      "step": 3048,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.7407,
+      "step": 3048,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.7298,
+      "step": 3048,
+      "vm_loss": 0.1137
+    },
+    {
+      "epoch": 0.5867603532497534,
+      "lm_loss": 1.5824,
+      "step": 3048,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 0.5869528599273287,
+      "grad_norm": 3.0057477794377547,
+      "learning_rate": 1.657448861350344e-05,
+      "loss": 2.1495,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5871453666049041,
+      "grad_norm": 2.5440215261832875,
+      "learning_rate": 1.6572138961458842e-05,
+      "loss": 2.0865,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5873378732824794,
+      "grad_norm": 2.9842927925728486,
+      "learning_rate": 1.6569788670518253e-05,
+      "loss": 2.093,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5875303799600549,
+      "grad_norm": 3.0767755972023614,
+      "learning_rate": 1.6567437740910156e-05,
+      "loss": 2.1278,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5877228866376303,
+      "grad_norm": 2.9736798317030186,
+      "learning_rate": 1.6565086172863084e-05,
+      "loss": 2.0497,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5879153933152056,
+      "grad_norm": 3.118449357781708,
+      "learning_rate": 1.6562733966605648e-05,
+      "loss": 2.1532,
+      "step": 3054
+    },
+    {
+      "epoch": 0.588107899992781,
+      "grad_norm": 3.107290224926494,
+      "learning_rate": 1.656038112236651e-05,
+      "loss": 2.0866,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "grad_norm": 2.979031705867233,
+      "learning_rate": 1.655802764037439e-05,
+      "loss": 2.0899,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 2.0332,
+      "step": 3056,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 1.6015,
+      "step": 3056,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 1.8559,
+      "step": 3056,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 1.7194,
+      "step": 3056,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 1.4751,
+      "step": 3056,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 1.4304,
+      "step": 3056,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 2.4113,
+      "step": 3056,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 0.5883004066703563,
+      "lm_loss": 2.0223,
+      "step": 3056,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.5884929133479317,
+      "grad_norm": 2.756892024960601,
+      "learning_rate": 1.6555673520858084e-05,
+      "loss": 2.101,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5886854200255072,
+      "grad_norm": 2.7666622277764503,
+      "learning_rate": 1.655331876404644e-05,
+      "loss": 2.0925,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5888779267030825,
+      "grad_norm": 2.9767820145843316,
+      "learning_rate": 1.655096337016837e-05,
+      "loss": 2.1422,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5890704333806579,
+      "grad_norm": 3.016400898066739,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 2.1109,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5892629400582333,
+      "grad_norm": 2.8856736593884977,
+      "learning_rate": 1.6546250672128915e-05,
+      "loss": 2.0473,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5894554467358086,
+      "grad_norm": 2.809008754399697,
+      "learning_rate": 1.6543893368425664e-05,
+      "loss": 2.0572,
+      "step": 3062
+    },
+    {
+      "epoch": 0.589647953413384,
+      "grad_norm": 2.815648995782425,
+      "learning_rate": 1.654153542857226e-05,
+      "loss": 2.1103,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "grad_norm": 2.81728503708321,
+      "learning_rate": 1.653917685279792e-05,
+      "loss": 2.1614,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 1.4508,
+      "step": 3064,
+      "vm_loss": 0.2674
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 2.4672,
+      "step": 3064,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 1.9315,
+      "step": 3064,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 1.5932,
+      "step": 3064,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 1.7128,
+      "step": 3064,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 2.3497,
+      "step": 3064,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 2.122,
+      "step": 3064,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 0.5898404600909594,
+      "lm_loss": 1.7593,
+      "step": 3064,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 0.5900329667685348,
+      "grad_norm": 2.819667052656945,
+      "learning_rate": 1.6536817641331928e-05,
+      "loss": 2.0813,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5902254734461102,
+      "grad_norm": 2.786566063637792,
+      "learning_rate": 1.653445779440363e-05,
+      "loss": 2.1434,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5904179801236855,
+      "grad_norm": 2.7336344675602042,
+      "learning_rate": 1.653209731224244e-05,
+      "loss": 2.1545,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5906104868012609,
+      "grad_norm": 2.82808881389369,
+      "learning_rate": 1.6529736195077814e-05,
+      "loss": 2.1404,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5908029934788362,
+      "grad_norm": 2.985314118518069,
+      "learning_rate": 1.652737444313929e-05,
+      "loss": 2.1506,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5909955001564117,
+      "grad_norm": 2.6801192766996844,
+      "learning_rate": 1.6525012056656463e-05,
+      "loss": 2.1268,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5911880068339871,
+      "grad_norm": 2.847983360219452,
+      "learning_rate": 1.6522649035858988e-05,
+      "loss": 2.1326,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "grad_norm": 2.8784764803182115,
+      "learning_rate": 1.6520285380976576e-05,
+      "loss": 2.147,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 2.1412,
+      "step": 3072,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 1.5153,
+      "step": 3072,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 1.7926,
+      "step": 3072,
+      "vm_loss": 0.2407
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 1.969,
+      "step": 3072,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 2.1109,
+      "step": 3072,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 2.158,
+      "step": 3072,
+      "vm_loss": 0.1194
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 1.6353,
+      "step": 3072,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.5913805135115624,
+      "lm_loss": 2.1399,
+      "step": 3072,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.5915730201891378,
+      "grad_norm": 2.7136733115355316,
+      "learning_rate": 1.6517921092239002e-05,
+      "loss": 2.1694,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5917655268667132,
+      "grad_norm": 2.846406856509497,
+      "learning_rate": 1.6515556169876115e-05,
+      "loss": 2.1305,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5919580335442886,
+      "grad_norm": 2.7493678118332023,
+      "learning_rate": 1.6513190614117805e-05,
+      "loss": 2.1621,
+      "step": 3075
+    },
+    {
+      "epoch": 0.592150540221864,
+      "grad_norm": 2.8434055330118486,
+      "learning_rate": 1.6510824425194043e-05,
+      "loss": 2.1271,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5923430468994393,
+      "grad_norm": 3.08171570849402,
+      "learning_rate": 1.6508457603334848e-05,
+      "loss": 2.1416,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5925355535770147,
+      "grad_norm": 2.8191036195925223,
+      "learning_rate": 1.6506090148770306e-05,
+      "loss": 2.155,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5927280602545901,
+      "grad_norm": 2.777495879092757,
+      "learning_rate": 1.6503722061730565e-05,
+      "loss": 2.0914,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "grad_norm": 2.76847626487161,
+      "learning_rate": 1.650135334244583e-05,
+      "loss": 2.156,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 1.6757,
+      "step": 3080,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 2.1201,
+      "step": 3080,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 1.7977,
+      "step": 3080,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 1.7977,
+      "step": 3080,
+      "vm_loss": 0.1093
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 2.2029,
+      "step": 3080,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 2.2853,
+      "step": 3080,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 2.039,
+      "step": 3080,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.5929205669321654,
+      "lm_loss": 1.8213,
+      "step": 3080,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 0.5931130736097409,
+      "grad_norm": 2.7624164999638565,
+      "learning_rate": 1.6498983991146377e-05,
+      "loss": 2.0748,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5933055802873162,
+      "grad_norm": 2.7991924711373293,
+      "learning_rate": 1.6496614008062533e-05,
+      "loss": 2.0823,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5934980869648916,
+      "grad_norm": 3.0474521586445165,
+      "learning_rate": 1.6494243393424693e-05,
+      "loss": 2.1214,
+      "step": 3083
+    },
+    {
+      "epoch": 0.593690593642467,
+      "grad_norm": 3.251477309664315,
+      "learning_rate": 1.6491872147463307e-05,
+      "loss": 2.0788,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5938831003200423,
+      "grad_norm": 2.9523167013991145,
+      "learning_rate": 1.6489500270408894e-05,
+      "loss": 2.1514,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5940756069976177,
+      "grad_norm": 3.1621505174689855,
+      "learning_rate": 1.648712776249203e-05,
+      "loss": 2.1231,
+      "step": 3086
+    },
+    {
+      "epoch": 0.594268113675193,
+      "grad_norm": 2.7068999145353048,
+      "learning_rate": 1.648475462394335e-05,
+      "loss": 2.07,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "grad_norm": 2.7609146520409897,
+      "learning_rate": 1.6482380854993558e-05,
+      "loss": 2.1109,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 2.3017,
+      "step": 3088,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 1.8878,
+      "step": 3088,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 1.7726,
+      "step": 3088,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 1.7092,
+      "step": 3088,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 1.9233,
+      "step": 3088,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 1.9601,
+      "step": 3088,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 2.1384,
+      "step": 3088,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 0.5944606203527685,
+      "lm_loss": 2.1962,
+      "step": 3088,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 0.5946531270303439,
+      "grad_norm": 3.006288736919539,
+      "learning_rate": 1.6480006455873415e-05,
+      "loss": 2.1321,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5948456337079192,
+      "grad_norm": 3.0979170778252105,
+      "learning_rate": 1.6477631426813737e-05,
+      "loss": 2.0971,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5950381403854946,
+      "grad_norm": 3.1360765057743705,
+      "learning_rate": 1.6475255768045415e-05,
+      "loss": 2.1783,
+      "step": 3091
+    },
+    {
+      "epoch": 0.59523064706307,
+      "grad_norm": 2.9230329805705573,
+      "learning_rate": 1.647287947979939e-05,
+      "loss": 2.0912,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5954231537406454,
+      "grad_norm": 2.927314575830423,
+      "learning_rate": 1.6470502562306664e-05,
+      "loss": 2.1638,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5956156604182208,
+      "grad_norm": 3.1089013455394894,
+      "learning_rate": 1.6468125015798307e-05,
+      "loss": 2.1538,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5958081670957961,
+      "grad_norm": 2.9433711613829057,
+      "learning_rate": 1.646574684050545e-05,
+      "loss": 2.0776,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "grad_norm": 2.774132946873311,
+      "learning_rate": 1.646336803665928e-05,
+      "loss": 2.0849,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 2.1088,
+      "step": 3096,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.8132,
+      "step": 3096,
+      "vm_loss": 0.2243
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.6197,
+      "step": 3096,
+      "vm_loss": 0.2126
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.9509,
+      "step": 3096,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.8054,
+      "step": 3096,
+      "vm_loss": 0.2375
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.9674,
+      "step": 3096,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.6279,
+      "step": 3096,
+      "vm_loss": 0.234
+    },
+    {
+      "epoch": 0.5960006737733715,
+      "lm_loss": 1.5834,
+      "step": 3096,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.5961931804509469,
+      "grad_norm": 3.1102912681314456,
+      "learning_rate": 1.6460988604491044e-05,
+      "loss": 2.0974,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5963856871285222,
+      "grad_norm": 3.209904410802056,
+      "learning_rate": 1.6458608544232052e-05,
+      "loss": 2.098,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5965781938060977,
+      "grad_norm": 2.959313334858762,
+      "learning_rate": 1.6456227856113683e-05,
+      "loss": 2.0795,
+      "step": 3099
+    },
+    {
+      "epoch": 0.596770700483673,
+      "grad_norm": 2.7199332191624537,
+      "learning_rate": 1.645384654036737e-05,
+      "loss": 2.138,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5969632071612484,
+      "grad_norm": 2.759879116157722,
+      "learning_rate": 1.6451464597224603e-05,
+      "loss": 2.0507,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5971557138388238,
+      "grad_norm": 3.119661728528989,
+      "learning_rate": 1.6449082026916934e-05,
+      "loss": 2.229,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5973482205163991,
+      "grad_norm": 3.0034052779210647,
+      "learning_rate": 1.6446698829675986e-05,
+      "loss": 2.1487,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "grad_norm": 2.863377385051883,
+      "learning_rate": 1.644431500573344e-05,
+      "loss": 2.108,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 2.2092,
+      "step": 3104,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 2.1008,
+      "step": 3104,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 2.213,
+      "step": 3104,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 2.0819,
+      "step": 3104,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 2.1218,
+      "step": 3104,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 1.9068,
+      "step": 3104,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 1.9765,
+      "step": 3104,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.5975407271939746,
+      "lm_loss": 1.92,
+      "step": 3104,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 0.59773323387155,
+      "grad_norm": 2.9689068731455723,
+      "learning_rate": 1.6441930555321026e-05,
+      "loss": 2.1197,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5979257405491253,
+      "grad_norm": 3.097564050558102,
+      "learning_rate": 1.6439545478670543e-05,
+      "loss": 2.1217,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5981182472267007,
+      "grad_norm": 2.9485061072165055,
+      "learning_rate": 1.6437159776013856e-05,
+      "loss": 2.1186,
+      "step": 3107
+    },
+    {
+      "epoch": 0.598310753904276,
+      "grad_norm": 2.9449434150118288,
+      "learning_rate": 1.6434773447582882e-05,
+      "loss": 2.1196,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5985032605818514,
+      "grad_norm": 2.892442618301742,
+      "learning_rate": 1.6432386493609606e-05,
+      "loss": 2.1639,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5986957672594269,
+      "grad_norm": 2.9301560663595287,
+      "learning_rate": 1.6429998914326064e-05,
+      "loss": 2.1238,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5988882739370022,
+      "grad_norm": 2.5957474368595737,
+      "learning_rate": 1.6427610709964367e-05,
+      "loss": 2.0595,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "grad_norm": 3.0215715144038473,
+      "learning_rate": 1.6425221880756677e-05,
+      "loss": 2.1745,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 2.3693,
+      "step": 3112,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 1.8474,
+      "step": 3112,
+      "vm_loss": 0.098
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 1.991,
+      "step": 3112,
+      "vm_loss": 0.2447
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 2.0397,
+      "step": 3112,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 2.028,
+      "step": 3112,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 2.1633,
+      "step": 3112,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 2.1718,
+      "step": 3112,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.5990807806145776,
+      "lm_loss": 1.8395,
+      "step": 3112,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 0.5992732872921529,
+      "grad_norm": 2.6427740186479283,
+      "learning_rate": 1.6422832426935216e-05,
+      "loss": 2.1091,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5994657939697283,
+      "grad_norm": 2.7658436602417034,
+      "learning_rate": 1.642044234873227e-05,
+      "loss": 2.1472,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5996583006473037,
+      "grad_norm": 2.791107423450398,
+      "learning_rate": 1.6418051646380187e-05,
+      "loss": 2.1144,
+      "step": 3115
+    },
+    {
+      "epoch": 0.599850807324879,
+      "grad_norm": 2.9253078480863177,
+      "learning_rate": 1.6415660320111376e-05,
+      "loss": 2.1129,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6000433140024545,
+      "grad_norm": 2.749328051887258,
+      "learning_rate": 1.64132683701583e-05,
+      "loss": 2.1051,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6002358206800298,
+      "grad_norm": 2.6775405828032395,
+      "learning_rate": 1.641087579675349e-05,
+      "loss": 2.1119,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6004283273576052,
+      "grad_norm": 2.8670121783474016,
+      "learning_rate": 1.6408482600129526e-05,
+      "loss": 2.1357,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "grad_norm": 2.6600516360354214,
+      "learning_rate": 1.640608878051907e-05,
+      "loss": 2.1157,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.7266,
+      "step": 3120,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.7653,
+      "step": 3120,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.8918,
+      "step": 3120,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.9661,
+      "step": 3120,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.7109,
+      "step": 3120,
+      "vm_loss": 0.1042
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 2.0747,
+      "step": 3120,
+      "vm_loss": 0.2553
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 2.0993,
+      "step": 3120,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.6006208340351806,
+      "lm_loss": 1.8689,
+      "step": 3120,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.6008133407127559,
+      "grad_norm": 2.8378894612568004,
+      "learning_rate": 1.6403694338154828e-05,
+      "loss": 2.1019,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6010058473903314,
+      "grad_norm": 2.964683498745593,
+      "learning_rate": 1.6401299273269572e-05,
+      "loss": 2.1403,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6011983540679068,
+      "grad_norm": 2.9847529592062187,
+      "learning_rate": 1.6398903586096128e-05,
+      "loss": 2.1044,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6013908607454821,
+      "grad_norm": 2.7649675136474383,
+      "learning_rate": 1.6396507276867386e-05,
+      "loss": 2.1079,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6015833674230575,
+      "grad_norm": 2.700862714423676,
+      "learning_rate": 1.6394110345816302e-05,
+      "loss": 2.1276,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6017758741006328,
+      "grad_norm": 2.8538236861747146,
+      "learning_rate": 1.639171279317589e-05,
+      "loss": 2.1021,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6019683807782082,
+      "grad_norm": 2.782355506492004,
+      "learning_rate": 1.638931461917922e-05,
+      "loss": 2.1307,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "grad_norm": 2.715795496543863,
+      "learning_rate": 1.6386915824059427e-05,
+      "loss": 2.1066,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 2.254,
+      "step": 3128,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 2.2265,
+      "step": 3128,
+      "vm_loss": 0.2451
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 1.8418,
+      "step": 3128,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 1.7923,
+      "step": 3128,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 1.6608,
+      "step": 3128,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 1.5838,
+      "step": 3128,
+      "vm_loss": 0.2162
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 1.7733,
+      "step": 3128,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.6021608874557837,
+      "lm_loss": 2.1976,
+      "step": 3128,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 0.602353394133359,
+      "grad_norm": 2.556823839721738,
+      "learning_rate": 1.63845164080497e-05,
+      "loss": 2.1295,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6025459008109344,
+      "grad_norm": 2.758001356529022,
+      "learning_rate": 1.63821163713833e-05,
+      "loss": 2.0889,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6027384074885097,
+      "grad_norm": 2.758340263093989,
+      "learning_rate": 1.6379715714293536e-05,
+      "loss": 2.0958,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6029309141660851,
+      "grad_norm": 2.7591547941026775,
+      "learning_rate": 1.6377314437013786e-05,
+      "loss": 2.0781,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6031234208436606,
+      "grad_norm": 2.8910216216542612,
+      "learning_rate": 1.6374912539777483e-05,
+      "loss": 2.0632,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6033159275212359,
+      "grad_norm": 2.973880034324225,
+      "learning_rate": 1.637251002281812e-05,
+      "loss": 2.1312,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6035084341988113,
+      "grad_norm": 2.8086452247351206,
+      "learning_rate": 1.6370106886369253e-05,
+      "loss": 2.1051,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "grad_norm": 2.993744282956302,
+      "learning_rate": 1.63677031306645e-05,
+      "loss": 2.0856,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.9913,
+      "step": 3136,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.6417,
+      "step": 3136,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.9158,
+      "step": 3136,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.5652,
+      "step": 3136,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.8118,
+      "step": 3136,
+      "vm_loss": 0.0926
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 2.135,
+      "step": 3136,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 2.1011,
+      "step": 3136,
+      "vm_loss": 0.2212
+    },
+    {
+      "epoch": 0.6037009408763867,
+      "lm_loss": 1.9596,
+      "step": 3136,
+      "vm_loss": 0.111
+    },
+    {
+      "epoch": 0.603893447553962,
+      "grad_norm": 2.9773326560686337,
+      "learning_rate": 1.6365298755937534e-05,
+      "loss": 2.0918,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6040859542315374,
+      "grad_norm": 2.7659726024628633,
+      "learning_rate": 1.63628937624221e-05,
+      "loss": 2.0297,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6042784609091127,
+      "grad_norm": 3.1548979907377968,
+      "learning_rate": 1.6360488150351984e-05,
+      "loss": 2.0924,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6044709675866882,
+      "grad_norm": 2.8928207432105726,
+      "learning_rate": 1.6358081919961044e-05,
+      "loss": 2.1029,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6046634742642636,
+      "grad_norm": 2.823551057318725,
+      "learning_rate": 1.63556750714832e-05,
+      "loss": 2.1177,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6048559809418389,
+      "grad_norm": 2.846249474094296,
+      "learning_rate": 1.6353267605152424e-05,
+      "loss": 2.1025,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6050484876194143,
+      "grad_norm": 3.054820514672883,
+      "learning_rate": 1.6350859521202755e-05,
+      "loss": 2.096,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "grad_norm": 3.0712096143284207,
+      "learning_rate": 1.634845081986829e-05,
+      "loss": 2.0949,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.9218,
+      "step": 3144,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 2.0666,
+      "step": 3144,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.6691,
+      "step": 3144,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.6648,
+      "step": 3144,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.3433,
+      "step": 3144,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.6114,
+      "step": 3144,
+      "vm_loss": 0.1696
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 1.8505,
+      "step": 3144,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 0.6052409942969896,
+      "lm_loss": 2.4911,
+      "step": 3144,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.605433500974565,
+      "grad_norm": 2.835328495524342,
+      "learning_rate": 1.6346041501383184e-05,
+      "loss": 2.0802,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6056260076521405,
+      "grad_norm": 2.92849519481618,
+      "learning_rate": 1.6343631565981654e-05,
+      "loss": 2.0441,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6058185143297158,
+      "grad_norm": 3.0463062663964324,
+      "learning_rate": 1.634122101389798e-05,
+      "loss": 2.1568,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6060110210072912,
+      "grad_norm": 3.1554149138705214,
+      "learning_rate": 1.6338809845366493e-05,
+      "loss": 2.0897,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6062035276848665,
+      "grad_norm": 2.731802671555569,
+      "learning_rate": 1.6336398060621592e-05,
+      "loss": 2.1514,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6063960343624419,
+      "grad_norm": 3.0152084637803873,
+      "learning_rate": 1.6333985659897737e-05,
+      "loss": 2.1413,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6065885410400174,
+      "grad_norm": 3.0776264682704255,
+      "learning_rate": 1.6331572643429433e-05,
+      "loss": 2.1152,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "grad_norm": 2.963824143166886,
+      "learning_rate": 1.6329159011451273e-05,
+      "loss": 2.1244,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 1.6866,
+      "step": 3152,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 2.1177,
+      "step": 3152,
+      "vm_loss": 0.1355
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 1.9209,
+      "step": 3152,
+      "vm_loss": 0.1008
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 1.9096,
+      "step": 3152,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 2.2478,
+      "step": 3152,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 2.1336,
+      "step": 3152,
+      "vm_loss": 0.1122
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 1.64,
+      "step": 3152,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 0.6067810477175927,
+      "lm_loss": 2.1637,
+      "step": 3152,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 0.6069735543951681,
+      "grad_norm": 3.1438201711506903,
+      "learning_rate": 1.6326744764197876e-05,
+      "loss": 2.1044,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6071660610727435,
+      "grad_norm": 3.0859865381780835,
+      "learning_rate": 1.6324329901903946e-05,
+      "loss": 2.0864,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6073585677503188,
+      "grad_norm": 3.2253098221943093,
+      "learning_rate": 1.632191442480424e-05,
+      "loss": 2.1114,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6075510744278942,
+      "grad_norm": 2.8904717869222023,
+      "learning_rate": 1.6319498333133572e-05,
+      "loss": 2.1729,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6077435811054696,
+      "grad_norm": 2.9691321538621867,
+      "learning_rate": 1.6317081627126817e-05,
+      "loss": 2.1507,
+      "step": 3157
+    },
+    {
+      "epoch": 0.607936087783045,
+      "grad_norm": 3.1396755992246854,
+      "learning_rate": 1.6314664307018902e-05,
+      "loss": 2.0873,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6081285944606204,
+      "grad_norm": 3.1259666135796147,
+      "learning_rate": 1.6312246373044835e-05,
+      "loss": 2.0331,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "grad_norm": 3.1071483225402625,
+      "learning_rate": 1.630982782543966e-05,
+      "loss": 2.1273,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.8931,
+      "step": 3160,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.8999,
+      "step": 3160,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.9565,
+      "step": 3160,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.5077,
+      "step": 3160,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.7355,
+      "step": 3160,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.6954,
+      "step": 3160,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.415,
+      "step": 3160,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 0.6083211011381957,
+      "lm_loss": 1.5183,
+      "step": 3160,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.6085136078157711,
+      "grad_norm": 3.0467831333516613,
+      "learning_rate": 1.6307408664438496e-05,
+      "loss": 2.0497,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6087061144933464,
+      "grad_norm": 3.4622780167619815,
+      "learning_rate": 1.6304988890276515e-05,
+      "loss": 2.0739,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6088986211709219,
+      "grad_norm": 2.8572980280905367,
+      "learning_rate": 1.6302568503188946e-05,
+      "loss": 2.1324,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6090911278484973,
+      "grad_norm": 2.8955750566187457,
+      "learning_rate": 1.630014750341109e-05,
+      "loss": 2.1081,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6092836345260726,
+      "grad_norm": 3.022213257586785,
+      "learning_rate": 1.629772589117829e-05,
+      "loss": 2.0795,
+      "step": 3165
+    },
+    {
+      "epoch": 0.609476141203648,
+      "grad_norm": 2.911666092157024,
+      "learning_rate": 1.6295303666725963e-05,
+      "loss": 2.0625,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6096686478812233,
+      "grad_norm": 2.987814984399549,
+      "learning_rate": 1.629288083028958e-05,
+      "loss": 2.1602,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "grad_norm": 2.762750417446928,
+      "learning_rate": 1.629045738210467e-05,
+      "loss": 2.0978,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 1.8585,
+      "step": 3168,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 2.1456,
+      "step": 3168,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 1.8984,
+      "step": 3168,
+      "vm_loss": 0.2566
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 1.7668,
+      "step": 3168,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 1.9968,
+      "step": 3168,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 2.0106,
+      "step": 3168,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 2.0301,
+      "step": 3168,
+      "vm_loss": 0.2122
+    },
+    {
+      "epoch": 0.6098611545587987,
+      "lm_loss": 2.0269,
+      "step": 3168,
+      "vm_loss": 0.2335
+    },
+    {
+      "epoch": 0.6100536612363742,
+      "grad_norm": 3.19094204683108,
+      "learning_rate": 1.6288033322406826e-05,
+      "loss": 2.0942,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6102461679139495,
+      "grad_norm": 3.0645364923540077,
+      "learning_rate": 1.6285608651431692e-05,
+      "loss": 2.0482,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6104386745915249,
+      "grad_norm": 3.2824346827346496,
+      "learning_rate": 1.6283183369414986e-05,
+      "loss": 2.0982,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6106311812691003,
+      "grad_norm": 2.9755716710347073,
+      "learning_rate": 1.6280757476592467e-05,
+      "loss": 2.1002,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6108236879466756,
+      "grad_norm": 3.0240372525686907,
+      "learning_rate": 1.6278330973199963e-05,
+      "loss": 2.0837,
+      "step": 3173
+    },
+    {
+      "epoch": 0.611016194624251,
+      "grad_norm": 2.782826203392463,
+      "learning_rate": 1.6275903859473374e-05,
+      "loss": 2.1048,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6112087013018264,
+      "grad_norm": 2.827268432431336,
+      "learning_rate": 1.627347613564863e-05,
+      "loss": 2.1476,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "grad_norm": 3.0814375790338935,
+      "learning_rate": 1.627104780196175e-05,
+      "loss": 2.0954,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 1.8808,
+      "step": 3176,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 2.0095,
+      "step": 3176,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 2.1868,
+      "step": 3176,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 1.96,
+      "step": 3176,
+      "vm_loss": 0.2162
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 2.0044,
+      "step": 3176,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 1.9393,
+      "step": 3176,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 1.902,
+      "step": 3176,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 0.6114012079794018,
+      "lm_loss": 2.1005,
+      "step": 3176,
+      "vm_loss": 0.2449
+    },
+    {
+      "epoch": 0.6115937146569772,
+      "grad_norm": 3.014359561279996,
+      "learning_rate": 1.626861885864879e-05,
+      "loss": 2.0651,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6117862213345525,
+      "grad_norm": 2.676790591742261,
+      "learning_rate": 1.6266189305945878e-05,
+      "loss": 2.0786,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6119787280121279,
+      "grad_norm": 2.9025690876455963,
+      "learning_rate": 1.62637591440892e-05,
+      "loss": 2.1289,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6121712346897032,
+      "grad_norm": 2.9615786598423557,
+      "learning_rate": 1.6261328373314995e-05,
+      "loss": 2.0853,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6123637413672787,
+      "grad_norm": 2.9624895726333076,
+      "learning_rate": 1.6258896993859566e-05,
+      "loss": 2.0707,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6125562480448541,
+      "grad_norm": 3.1354391140772133,
+      "learning_rate": 1.6256465005959277e-05,
+      "loss": 2.0831,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6127487547224294,
+      "grad_norm": 3.215196700630077,
+      "learning_rate": 1.625403240985054e-05,
+      "loss": 2.1399,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "grad_norm": 3.1360356171676966,
+      "learning_rate": 1.6251599205769845e-05,
+      "loss": 2.0937,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.941,
+      "step": 3184,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.8345,
+      "step": 3184,
+      "vm_loss": 0.1036
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.9953,
+      "step": 3184,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 2.1133,
+      "step": 3184,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.6006,
+      "step": 3184,
+      "vm_loss": 0.1144
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 2.1592,
+      "step": 3184,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.5276,
+      "step": 3184,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.6129412614000048,
+      "lm_loss": 1.8149,
+      "step": 3184,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.6131337680775802,
+      "grad_norm": 2.9057593517280487,
+      "learning_rate": 1.6249165393953727e-05,
+      "loss": 2.0647,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6133262747551556,
+      "grad_norm": 2.8131783955620966,
+      "learning_rate": 1.624673097463878e-05,
+      "loss": 2.1041,
+      "step": 3186
+    },
+    {
+      "epoch": 0.613518781432731,
+      "grad_norm": 2.9169879856427743,
+      "learning_rate": 1.6244295948061663e-05,
+      "loss": 2.0291,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6137112881103063,
+      "grad_norm": 2.9397036118165714,
+      "learning_rate": 1.6241860314459096e-05,
+      "loss": 2.1053,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6139037947878817,
+      "grad_norm": 3.0678595748184203,
+      "learning_rate": 1.6239424074067845e-05,
+      "loss": 2.0949,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6140963014654571,
+      "grad_norm": 2.9734960974328,
+      "learning_rate": 1.6236987227124748e-05,
+      "loss": 2.1509,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6142888081430324,
+      "grad_norm": 3.043773111444623,
+      "learning_rate": 1.62345497738667e-05,
+      "loss": 2.0506,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "grad_norm": 2.893973337494174,
+      "learning_rate": 1.623211171453065e-05,
+      "loss": 2.0905,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.7668,
+      "step": 3192,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.7622,
+      "step": 3192,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.4562,
+      "step": 3192,
+      "vm_loss": 0.2336
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 2.215,
+      "step": 3192,
+      "vm_loss": 0.244
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.8997,
+      "step": 3192,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.9953,
+      "step": 3192,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.7962,
+      "step": 3192,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 0.6144813148206079,
+      "lm_loss": 1.9989,
+      "step": 3192,
+      "vm_loss": 0.2018
+    },
+    {
+      "epoch": 0.6146738214981832,
+      "grad_norm": 2.9494806129924522,
+      "learning_rate": 1.6229673049353612e-05,
+      "loss": 2.0856,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6148663281757586,
+      "grad_norm": 2.8932857741551157,
+      "learning_rate": 1.622723377857265e-05,
+      "loss": 2.1079,
+      "step": 3194
+    },
+    {
+      "epoch": 0.615058834853334,
+      "grad_norm": 2.7985344129187144,
+      "learning_rate": 1.62247939024249e-05,
+      "loss": 2.0593,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6152513415309093,
+      "grad_norm": 3.2135061394917948,
+      "learning_rate": 1.622235342114754e-05,
+      "loss": 2.1201,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6154438482084847,
+      "grad_norm": 2.756267579527573,
+      "learning_rate": 1.621991233497782e-05,
+      "loss": 2.0738,
+      "step": 3197
+    },
+    {
+      "epoch": 0.61563635488606,
+      "grad_norm": 2.743092268777963,
+      "learning_rate": 1.6217470644153045e-05,
+      "loss": 2.1584,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6158288615636355,
+      "grad_norm": 2.7047178972930674,
+      "learning_rate": 1.6215028348910577e-05,
+      "loss": 2.0681,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "grad_norm": 2.8421509802042193,
+      "learning_rate": 1.6212585449487842e-05,
+      "loss": 2.0451,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 2.2161,
+      "step": 3200,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.9087,
+      "step": 3200,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.67,
+      "step": 3200,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 2.2056,
+      "step": 3200,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.8152,
+      "step": 3200,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.9843,
+      "step": 3200,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.3233,
+      "step": 3200,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.6160213682412109,
+      "lm_loss": 1.9735,
+      "step": 3200,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.6162138749187862,
+      "grad_norm": 3.070108243125282,
+      "learning_rate": 1.6210141946122322e-05,
+      "loss": 2.061,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6164063815963616,
+      "grad_norm": 2.9327218614405948,
+      "learning_rate": 1.620769783905155e-05,
+      "loss": 2.155,
+      "step": 3202
+    },
+    {
+      "epoch": 0.616598888273937,
+      "grad_norm": 2.8836796347423186,
+      "learning_rate": 1.6205253128513133e-05,
+      "loss": 2.0599,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6167913949515124,
+      "grad_norm": 2.7829783265112447,
+      "learning_rate": 1.620280781474472e-05,
+      "loss": 2.0259,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6169839016290878,
+      "grad_norm": 2.9808814874585265,
+      "learning_rate": 1.620036189798403e-05,
+      "loss": 2.1368,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6171764083066631,
+      "grad_norm": 2.971563304864646,
+      "learning_rate": 1.619791537846884e-05,
+      "loss": 2.0435,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6173689149842385,
+      "grad_norm": 2.7751986637485477,
+      "learning_rate": 1.619546825643698e-05,
+      "loss": 2.0856,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "grad_norm": 2.693359645171731,
+      "learning_rate": 1.6193020532126342e-05,
+      "loss": 2.0517,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.5563,
+      "step": 3208,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.7835,
+      "step": 3208,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.714,
+      "step": 3208,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.6305,
+      "step": 3208,
+      "vm_loss": 0.2519
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.7705,
+      "step": 3208,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.8373,
+      "step": 3208,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.822,
+      "step": 3208,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.6175614216618139,
+      "lm_loss": 1.5346,
+      "step": 3208,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 0.6177539283393892,
+      "grad_norm": 3.310893855951052,
+      "learning_rate": 1.619057220577487e-05,
+      "loss": 2.0618,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6179464350169647,
+      "grad_norm": 2.928554604886684,
+      "learning_rate": 1.618812327762059e-05,
+      "loss": 2.1162,
+      "step": 3210
+    },
+    {
+      "epoch": 0.61813894169454,
+      "grad_norm": 2.942721976499006,
+      "learning_rate": 1.6185673747901554e-05,
+      "loss": 2.1128,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6183314483721154,
+      "grad_norm": 2.820986691806293,
+      "learning_rate": 1.6183223616855888e-05,
+      "loss": 2.1592,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6185239550496908,
+      "grad_norm": 2.6620656782849816,
+      "learning_rate": 1.6180772884721786e-05,
+      "loss": 2.1516,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6187164617272661,
+      "grad_norm": 2.834615064973116,
+      "learning_rate": 1.617832155173748e-05,
+      "loss": 2.0964,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6189089684048416,
+      "grad_norm": 2.770172171950774,
+      "learning_rate": 1.6175869618141278e-05,
+      "loss": 2.0133,
+      "step": 3215
+    },
+    {
+      "epoch": 0.619101475082417,
+      "grad_norm": 2.9966203528062665,
+      "learning_rate": 1.6173417084171537e-05,
+      "loss": 2.1524,
+      "step": 3216
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 2.1031,
+      "step": 3216,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 2.1308,
+      "step": 3216,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 1.4265,
+      "step": 3216,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 1.8085,
+      "step": 3216,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 1.9499,
+      "step": 3216,
+      "vm_loss": 0.2371
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 1.7959,
+      "step": 3216,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 1.7617,
+      "step": 3216,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.619101475082417,
+      "lm_loss": 2.1991,
+      "step": 3216,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 0.6192939817599923,
+      "grad_norm": 2.984699383955704,
+      "learning_rate": 1.6170963950066672e-05,
+      "loss": 2.0398,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6194864884375677,
+      "grad_norm": 3.0863065514460186,
+      "learning_rate": 1.6168510216065162e-05,
+      "loss": 2.1241,
+      "step": 3218
+    },
+    {
+      "epoch": 0.619678995115143,
+      "grad_norm": 3.0806461972237273,
+      "learning_rate": 1.616605588240554e-05,
+      "loss": 2.1218,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6198715017927184,
+      "grad_norm": 2.874351171395082,
+      "learning_rate": 1.6163600949326405e-05,
+      "loss": 2.0915,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6200640084702939,
+      "grad_norm": 2.7559362122654227,
+      "learning_rate": 1.6161145417066392e-05,
+      "loss": 2.0673,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6202565151478692,
+      "grad_norm": 2.885909144866961,
+      "learning_rate": 1.615868928586423e-05,
+      "loss": 2.0847,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6204490218254446,
+      "grad_norm": 2.8390260663978815,
+      "learning_rate": 1.615623255595867e-05,
+      "loss": 2.1185,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "grad_norm": 2.829431338923187,
+      "learning_rate": 1.6153775227588547e-05,
+      "loss": 2.1076,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 2.3484,
+      "step": 3224,
+      "vm_loss": 0.1346
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 1.7484,
+      "step": 3224,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 2.1724,
+      "step": 3224,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 1.62,
+      "step": 3224,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 2.007,
+      "step": 3224,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 1.5406,
+      "step": 3224,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 2.0054,
+      "step": 3224,
+      "vm_loss": 0.2293
+    },
+    {
+      "epoch": 0.6206415285030199,
+      "lm_loss": 1.9702,
+      "step": 3224,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 0.6208340351805953,
+      "grad_norm": 2.721669568085743,
+      "learning_rate": 1.6151317300992743e-05,
+      "loss": 2.0792,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6210265418581707,
+      "grad_norm": 2.793934930081318,
+      "learning_rate": 1.61488587764102e-05,
+      "loss": 2.0452,
+      "step": 3226
+    },
+    {
+      "epoch": 0.621219048535746,
+      "grad_norm": 2.9110488755819075,
+      "learning_rate": 1.6146399654079915e-05,
+      "loss": 2.0789,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6214115552133215,
+      "grad_norm": 2.7810142670961495,
+      "learning_rate": 1.614393993424095e-05,
+      "loss": 2.1153,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6216040618908968,
+      "grad_norm": 2.8281314844694174,
+      "learning_rate": 1.6141479617132417e-05,
+      "loss": 1.9844,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6217965685684722,
+      "grad_norm": 2.723547867862937,
+      "learning_rate": 1.6139018702993495e-05,
+      "loss": 2.003,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6219890752460476,
+      "grad_norm": 2.720139820849798,
+      "learning_rate": 1.613655719206341e-05,
+      "loss": 2.1188,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "grad_norm": 2.797496042123596,
+      "learning_rate": 1.613409508458146e-05,
+      "loss": 2.0817,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.4603,
+      "step": 3232,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.97,
+      "step": 3232,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.8605,
+      "step": 3232,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 2.1665,
+      "step": 3232,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.7954,
+      "step": 3232,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 2.0923,
+      "step": 3232,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.5658,
+      "step": 3232,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 0.6221815819236229,
+      "lm_loss": 1.6446,
+      "step": 3232,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 0.6223740886011984,
+      "grad_norm": 2.836820240361788,
+      "learning_rate": 1.613163238078699e-05,
+      "loss": 2.053,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6225665952787738,
+      "grad_norm": 2.8689611065856404,
+      "learning_rate": 1.6129169080919405e-05,
+      "loss": 2.0844,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6227591019563491,
+      "grad_norm": 2.958303885276219,
+      "learning_rate": 1.612670518521817e-05,
+      "loss": 2.1158,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6229516086339245,
+      "grad_norm": 3.1440409362555144,
+      "learning_rate": 1.6124240693922808e-05,
+      "loss": 2.0957,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6231441153114998,
+      "grad_norm": 2.72055694405428,
+      "learning_rate": 1.61217756072729e-05,
+      "loss": 2.033,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6233366219890752,
+      "grad_norm": 3.1464460416021787,
+      "learning_rate": 1.611930992550808e-05,
+      "loss": 2.0671,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6235291286666507,
+      "grad_norm": 3.2932572541966527,
+      "learning_rate": 1.6116843648868047e-05,
+      "loss": 2.0354,
+      "step": 3239
+    },
+    {
+      "epoch": 0.623721635344226,
+      "grad_norm": 3.0321716212657552,
+      "learning_rate": 1.6114376777592556e-05,
+      "loss": 2.079,
+      "step": 3240
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.8431,
+      "step": 3240,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.5771,
+      "step": 3240,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 2.0431,
+      "step": 3240,
+      "vm_loss": 0.1179
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.5818,
+      "step": 3240,
+      "vm_loss": 0.2094
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 2.2321,
+      "step": 3240,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.7004,
+      "step": 3240,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.543,
+      "step": 3240,
+      "vm_loss": 0.1183
+    },
+    {
+      "epoch": 0.623721635344226,
+      "lm_loss": 1.6946,
+      "step": 3240,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 0.6239141420218014,
+      "grad_norm": 2.9879939493693977,
+      "learning_rate": 1.6111909311921413e-05,
+      "loss": 2.0229,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6241066486993767,
+      "grad_norm": 3.22187543906383,
+      "learning_rate": 1.6109441252094495e-05,
+      "loss": 2.0647,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6242991553769521,
+      "grad_norm": 3.0232401925477914,
+      "learning_rate": 1.610697259835172e-05,
+      "loss": 2.0914,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6244916620545276,
+      "grad_norm": 3.062352805921721,
+      "learning_rate": 1.610450335093308e-05,
+      "loss": 2.1114,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6246841687321029,
+      "grad_norm": 3.4633648377938755,
+      "learning_rate": 1.610203351007862e-05,
+      "loss": 2.0415,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6248766754096783,
+      "grad_norm": 3.042531011877521,
+      "learning_rate": 1.6099563076028427e-05,
+      "loss": 2.0638,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6250691820872537,
+      "grad_norm": 2.9386519431627427,
+      "learning_rate": 1.609709204902267e-05,
+      "loss": 2.0361,
+      "step": 3247
+    },
+    {
+      "epoch": 0.625261688764829,
+      "grad_norm": 3.4657798473256043,
+      "learning_rate": 1.609462042930156e-05,
+      "loss": 2.1306,
+      "step": 3248
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 2.0203,
+      "step": 3248,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.8791,
+      "step": 3248,
+      "vm_loss": 0.1747
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.4903,
+      "step": 3248,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.8407,
+      "step": 3248,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.6741,
+      "step": 3248,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.9715,
+      "step": 3248,
+      "vm_loss": 0.2259
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.5132,
+      "step": 3248,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.625261688764829,
+      "lm_loss": 1.8687,
+      "step": 3248,
+      "vm_loss": 0.2442
+    },
+    {
+      "epoch": 0.6254541954424044,
+      "grad_norm": 3.160398971061847,
+      "learning_rate": 1.6092148217105372e-05,
+      "loss": 2.0839,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6256467021199797,
+      "grad_norm": 2.8940007770699894,
+      "learning_rate": 1.6089675412674436e-05,
+      "loss": 2.0199,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6258392087975552,
+      "grad_norm": 2.6958379679164155,
+      "learning_rate": 1.608720201624914e-05,
+      "loss": 2.0245,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6260317154751306,
+      "grad_norm": 3.306982120459445,
+      "learning_rate": 1.6084728028069935e-05,
+      "loss": 2.0307,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6262242221527059,
+      "grad_norm": 3.1510069115974915,
+      "learning_rate": 1.6082253448377312e-05,
+      "loss": 2.0957,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6264167288302813,
+      "grad_norm": 2.9267431955331085,
+      "learning_rate": 1.6079778277411843e-05,
+      "loss": 2.0544,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6266092355078566,
+      "grad_norm": 2.8269267507829356,
+      "learning_rate": 1.6077302515414142e-05,
+      "loss": 2.0425,
+      "step": 3255
+    },
+    {
+      "epoch": 0.626801742185432,
+      "grad_norm": 2.8171361752175192,
+      "learning_rate": 1.6074826162624883e-05,
+      "loss": 2.1019,
+      "step": 3256
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.6816,
+      "step": 3256,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.9142,
+      "step": 3256,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.7568,
+      "step": 3256,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 2.1106,
+      "step": 3256,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.9164,
+      "step": 3256,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.7232,
+      "step": 3256,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 2.0661,
+      "step": 3256,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.626801742185432,
+      "lm_loss": 1.93,
+      "step": 3256,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 0.6269942488630075,
+      "grad_norm": 2.958991963577431,
+      "learning_rate": 1.6072349219284805e-05,
+      "loss": 2.0596,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6271867555405828,
+      "grad_norm": 2.7804274531826523,
+      "learning_rate": 1.606987168563469e-05,
+      "loss": 2.0335,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6273792622181582,
+      "grad_norm": 2.879610627506035,
+      "learning_rate": 1.6067393561915395e-05,
+      "loss": 2.0597,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6275717688957335,
+      "grad_norm": 2.9926630797216034,
+      "learning_rate": 1.6064914848367818e-05,
+      "loss": 2.0735,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6277642755733089,
+      "grad_norm": 2.8666754652569235,
+      "learning_rate": 1.606243554523293e-05,
+      "loss": 2.054,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6279567822508844,
+      "grad_norm": 3.275475330705778,
+      "learning_rate": 1.605995565275174e-05,
+      "loss": 2.1118,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6281492889284597,
+      "grad_norm": 2.835854144413279,
+      "learning_rate": 1.6057475171165333e-05,
+      "loss": 2.0543,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "grad_norm": 3.0033977622092176,
+      "learning_rate": 1.605499410071484e-05,
+      "loss": 2.094,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.9847,
+      "step": 3264,
+      "vm_loss": 0.1007
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.4738,
+      "step": 3264,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.9652,
+      "step": 3264,
+      "vm_loss": 0.1837
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.4214,
+      "step": 3264,
+      "vm_loss": 0.1369
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.7637,
+      "step": 3264,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 2.0176,
+      "step": 3264,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.7222,
+      "step": 3264,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.6283417956060351,
+      "lm_loss": 1.7601,
+      "step": 3264,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.6285343022836105,
+      "grad_norm": 2.659107301330133,
+      "learning_rate": 1.6052512441641456e-05,
+      "loss": 2.0407,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6287268089611858,
+      "grad_norm": 2.8663836480113534,
+      "learning_rate": 1.605003019418643e-05,
+      "loss": 2.0407,
+      "step": 3266
+    },
+    {
+      "epoch": 0.6289193156387612,
+      "grad_norm": 3.0307360465573834,
+      "learning_rate": 1.6047547358591063e-05,
+      "loss": 2.0345,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6291118223163366,
+      "grad_norm": 2.9942403075182065,
+      "learning_rate": 1.604506393509673e-05,
+      "loss": 2.045,
+      "step": 3268
+    },
+    {
+      "epoch": 0.629304328993912,
+      "grad_norm": 2.9463247531711736,
+      "learning_rate": 1.6042579923944834e-05,
+      "loss": 2.012,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6294968356714874,
+      "grad_norm": 2.9749780526600307,
+      "learning_rate": 1.6040095325376867e-05,
+      "loss": 2.076,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6296893423490627,
+      "grad_norm": 2.9523475624585664,
+      "learning_rate": 1.6037610139634358e-05,
+      "loss": 2.0609,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "grad_norm": 2.7609148414533142,
+      "learning_rate": 1.60351243669589e-05,
+      "loss": 2.0429,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.7073,
+      "step": 3272,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.8577,
+      "step": 3272,
+      "vm_loss": 0.2083
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.9248,
+      "step": 3272,
+      "vm_loss": 0.1104
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.7696,
+      "step": 3272,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 2.3402,
+      "step": 3272,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 2.0146,
+      "step": 3272,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.5198,
+      "step": 3272,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 0.6298818490266381,
+      "lm_loss": 1.6701,
+      "step": 3272,
+      "vm_loss": 0.1037
+    },
+    {
+      "epoch": 0.6300743557042134,
+      "grad_norm": 2.9176825605467607,
+      "learning_rate": 1.6032638007592143e-05,
+      "loss": 2.0413,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6302668623817889,
+      "grad_norm": 2.978385302217501,
+      "learning_rate": 1.603015106177579e-05,
+      "loss": 1.9534,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6304593690593643,
+      "grad_norm": 3.019605757575759,
+      "learning_rate": 1.60276635297516e-05,
+      "loss": 2.0332,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6306518757369396,
+      "grad_norm": 2.978349274068598,
+      "learning_rate": 1.602517541176141e-05,
+      "loss": 2.0575,
+      "step": 3276
+    },
+    {
+      "epoch": 0.630844382414515,
+      "grad_norm": 2.928080702894649,
+      "learning_rate": 1.602268670804708e-05,
+      "loss": 2.0351,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6310368890920904,
+      "grad_norm": 2.8851228608228907,
+      "learning_rate": 1.602019741885055e-05,
+      "loss": 2.0468,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6312293957696657,
+      "grad_norm": 2.934821805893519,
+      "learning_rate": 1.6017707544413808e-05,
+      "loss": 2.0145,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "grad_norm": 2.6787353699381624,
+      "learning_rate": 1.6015217084978908e-05,
+      "loss": 2.052,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.7433,
+      "step": 3280,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.6783,
+      "step": 3280,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.638,
+      "step": 3280,
+      "vm_loss": 0.2398
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.8944,
+      "step": 3280,
+      "vm_loss": 0.1291
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.9888,
+      "step": 3280,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.7269,
+      "step": 3280,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 1.8964,
+      "step": 3280,
+      "vm_loss": 0.1907
+    },
+    {
+      "epoch": 0.6314219024472412,
+      "lm_loss": 2.2046,
+      "step": 3280,
+      "vm_loss": 0.1276
+    },
+    {
+      "epoch": 0.6316144091248165,
+      "grad_norm": 2.8850037555915695,
+      "learning_rate": 1.6012726040787944e-05,
+      "loss": 2.054,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6318069158023919,
+      "grad_norm": 2.951420455515323,
+      "learning_rate": 1.601023441208309e-05,
+      "loss": 2.1242,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6319994224799673,
+      "grad_norm": 3.196164145490763,
+      "learning_rate": 1.600774219910655e-05,
+      "loss": 2.077,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6321919291575426,
+      "grad_norm": 2.82576519307112,
+      "learning_rate": 1.6005249402100612e-05,
+      "loss": 1.9973,
+      "step": 3284
+    },
+    {
+      "epoch": 0.632384435835118,
+      "grad_norm": 2.9895702645153532,
+      "learning_rate": 1.6002756021307603e-05,
+      "loss": 2.0559,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6325769425126934,
+      "grad_norm": 3.3972380122932937,
+      "learning_rate": 1.600026205696991e-05,
+      "loss": 2.0727,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6327694491902688,
+      "grad_norm": 2.8123570879748407,
+      "learning_rate": 1.599776750932998e-05,
+      "loss": 2.074,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "grad_norm": 3.016210079875932,
+      "learning_rate": 1.5995272378630313e-05,
+      "loss": 2.0213,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.8798,
+      "step": 3288,
+      "vm_loss": 0.2196
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.6479,
+      "step": 3288,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.615,
+      "step": 3288,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.9822,
+      "step": 3288,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.799,
+      "step": 3288,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.7904,
+      "step": 3288,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 2.207,
+      "step": 3288,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 0.6329619558678442,
+      "lm_loss": 1.9522,
+      "step": 3288,
+      "vm_loss": 0.2228
+    },
+    {
+      "epoch": 0.6331544625454195,
+      "grad_norm": 3.6456296755612674,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 2.0451,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6333469692229949,
+      "grad_norm": 3.0173959761456217,
+      "learning_rate": 1.5990280369022064e-05,
+      "loss": 2.0896,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6335394759005702,
+      "grad_norm": 3.003685282374375,
+      "learning_rate": 1.598778349059877e-05,
+      "loss": 2.1061,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6337319825781457,
+      "grad_norm": 3.467649304205802,
+      "learning_rate": 1.5985286030086315e-05,
+      "loss": 2.0255,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6339244892557211,
+      "grad_norm": 3.118770581639357,
+      "learning_rate": 1.598278798772748e-05,
+      "loss": 2.0783,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6341169959332964,
+      "grad_norm": 2.7747227595558503,
+      "learning_rate": 1.5980289363765113e-05,
+      "loss": 2.0201,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6343095026108718,
+      "grad_norm": 2.976046922116868,
+      "learning_rate": 1.597779015844211e-05,
+      "loss": 2.0429,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "grad_norm": 2.9661040809587047,
+      "learning_rate": 1.5975290372001423e-05,
+      "loss": 2.0405,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 1.8267,
+      "step": 3296,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 1.6175,
+      "step": 3296,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 1.7879,
+      "step": 3296,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 1.5495,
+      "step": 3296,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 2.0817,
+      "step": 3296,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 2.0249,
+      "step": 3296,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 2.1368,
+      "step": 3296,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 0.6345020092884472,
+      "lm_loss": 1.75,
+      "step": 3296,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.6346945159660226,
+      "grad_norm": 2.7107119940730566,
+      "learning_rate": 1.597279000468607e-05,
+      "loss": 2.034,
+      "step": 3297
+    },
+    {
+      "epoch": 0.634887022643598,
+      "grad_norm": 2.888205903036789,
+      "learning_rate": 1.597028905673911e-05,
+      "loss": 2.0562,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6350795293211733,
+      "grad_norm": 3.327241549349959,
+      "learning_rate": 1.596778752840367e-05,
+      "loss": 2.0407,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6352720359987487,
+      "grad_norm": 3.053432659277193,
+      "learning_rate": 1.5965285419922936e-05,
+      "loss": 2.0587,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6354645426763241,
+      "grad_norm": 2.760345374328317,
+      "learning_rate": 1.596278273154014e-05,
+      "loss": 2.0274,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6356570493538994,
+      "grad_norm": 2.9083317304313994,
+      "learning_rate": 1.5960279463498576e-05,
+      "loss": 2.0866,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6358495560314749,
+      "grad_norm": 2.669141789540415,
+      "learning_rate": 1.5957775616041594e-05,
+      "loss": 2.0148,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "grad_norm": 2.8077454795126546,
+      "learning_rate": 1.5955271189412596e-05,
+      "loss": 2.0561,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 2.231,
+      "step": 3304,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 2.0571,
+      "step": 3304,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 2.3381,
+      "step": 3304,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 1.4315,
+      "step": 3304,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 1.6932,
+      "step": 3304,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 1.7046,
+      "step": 3304,
+      "vm_loss": 0.2253
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 1.2636,
+      "step": 3304,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.6360420627090502,
+      "lm_loss": 2.1153,
+      "step": 3304,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.6362345693866256,
+      "grad_norm": 2.9974686567284006,
+      "learning_rate": 1.595276618385505e-05,
+      "loss": 2.07,
+      "step": 3305
+    },
+    {
+      "epoch": 0.636427076064201,
+      "grad_norm": 2.983821918017044,
+      "learning_rate": 1.5950260599612478e-05,
+      "loss": 2.0981,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6366195827417763,
+      "grad_norm": 2.795410831641941,
+      "learning_rate": 1.5947754436928445e-05,
+      "loss": 2.069,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6368120894193517,
+      "grad_norm": 2.9708374381594767,
+      "learning_rate": 1.594524769604659e-05,
+      "loss": 2.054,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6370045960969272,
+      "grad_norm": 2.8290560136821172,
+      "learning_rate": 1.5942740377210597e-05,
+      "loss": 2.0266,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6371971027745025,
+      "grad_norm": 2.6833896258540526,
+      "learning_rate": 1.5940232480664205e-05,
+      "loss": 2.1062,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6373896094520779,
+      "grad_norm": 2.8616296115912934,
+      "learning_rate": 1.5937724006651222e-05,
+      "loss": 2.0091,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "grad_norm": 2.9638169926847704,
+      "learning_rate": 1.5935214955415496e-05,
+      "loss": 2.0339,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 2.0585,
+      "step": 3312,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 1.6663,
+      "step": 3312,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 2.0045,
+      "step": 3312,
+      "vm_loss": 0.2354
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 1.53,
+      "step": 3312,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 2.0119,
+      "step": 3312,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 1.8432,
+      "step": 3312,
+      "vm_loss": 0.1152
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 1.9417,
+      "step": 3312,
+      "vm_loss": 0.2161
+    },
+    {
+      "epoch": 0.6375821161296532,
+      "lm_loss": 1.7011,
+      "step": 3312,
+      "vm_loss": 0.1962
+    },
+    {
+      "epoch": 0.6377746228072286,
+      "grad_norm": 2.9680209701670823,
+      "learning_rate": 1.5932705327200948e-05,
+      "loss": 2.0567,
+      "step": 3313
+    },
+    {
+      "epoch": 0.637967129484804,
+      "grad_norm": 3.145858056178398,
+      "learning_rate": 1.593019512225154e-05,
+      "loss": 2.064,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6381596361623794,
+      "grad_norm": 2.8657016073930173,
+      "learning_rate": 1.5927684340811292e-05,
+      "loss": 2.0391,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6383521428399548,
+      "grad_norm": 2.753424023140314,
+      "learning_rate": 1.5925172983124292e-05,
+      "loss": 2.0104,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6385446495175301,
+      "grad_norm": 2.670779550129038,
+      "learning_rate": 1.592266104943467e-05,
+      "loss": 2.0245,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6387371561951055,
+      "grad_norm": 2.8361908175389394,
+      "learning_rate": 1.5920148539986623e-05,
+      "loss": 2.016,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6389296628726809,
+      "grad_norm": 2.802222203007585,
+      "learning_rate": 1.59176354550244e-05,
+      "loss": 2.0571,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "grad_norm": 3.0332668143014265,
+      "learning_rate": 1.5915121794792302e-05,
+      "loss": 2.0137,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 2.1169,
+      "step": 3320,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 1.4978,
+      "step": 3320,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 2.097,
+      "step": 3320,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 1.794,
+      "step": 3320,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 1.6489,
+      "step": 3320,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 2.1894,
+      "step": 3320,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 1.2987,
+      "step": 3320,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 0.6391221695502562,
+      "lm_loss": 1.9577,
+      "step": 3320,
+      "vm_loss": 0.0952
+    },
+    {
+      "epoch": 0.6393146762278317,
+      "grad_norm": 3.0584574826512205,
+      "learning_rate": 1.5912607559534684e-05,
+      "loss": 2.0532,
+      "step": 3321
+    },
+    {
+      "epoch": 0.639507182905407,
+      "grad_norm": 2.833461610120119,
+      "learning_rate": 1.5910092749495972e-05,
+      "loss": 2.0738,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6396996895829824,
+      "grad_norm": 2.6985401222281262,
+      "learning_rate": 1.590757736492063e-05,
+      "loss": 2.0767,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6398921962605578,
+      "grad_norm": 2.883571231355534,
+      "learning_rate": 1.5905061406053186e-05,
+      "loss": 1.9904,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6400847029381331,
+      "grad_norm": 3.0584918647258514,
+      "learning_rate": 1.5902544873138228e-05,
+      "loss": 2.0095,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6402772096157086,
+      "grad_norm": 2.8503687868954666,
+      "learning_rate": 1.5900027766420396e-05,
+      "loss": 2.044,
+      "step": 3326
+    },
+    {
+      "epoch": 0.640469716293284,
+      "grad_norm": 2.750696515497026,
+      "learning_rate": 1.5897510086144374e-05,
+      "loss": 2.0355,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "grad_norm": 2.808866913889281,
+      "learning_rate": 1.5894991832554926e-05,
+      "loss": 2.0258,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.6966,
+      "step": 3328,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 2.1342,
+      "step": 3328,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.7212,
+      "step": 3328,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.7517,
+      "step": 3328,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.4576,
+      "step": 3328,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.7404,
+      "step": 3328,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 2.0482,
+      "step": 3328,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 0.6406622229708593,
+      "lm_loss": 1.7131,
+      "step": 3328,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.6408547296484347,
+      "grad_norm": 3.042957276070317,
+      "learning_rate": 1.589247300589685e-05,
+      "loss": 2.1082,
+      "step": 3329
+    },
+    {
+      "epoch": 0.64104723632601,
+      "grad_norm": 2.935121582805986,
+      "learning_rate": 1.5889953606415014e-05,
+      "loss": 2.0608,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6412397430035854,
+      "grad_norm": 3.0704924786474366,
+      "learning_rate": 1.5887433634354335e-05,
+      "loss": 2.0611,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6414322496811609,
+      "grad_norm": 2.8471577273649773,
+      "learning_rate": 1.5884913089959778e-05,
+      "loss": 2.0239,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6416247563587362,
+      "grad_norm": 2.8555693950657925,
+      "learning_rate": 1.588239197347638e-05,
+      "loss": 2.0545,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6418172630363116,
+      "grad_norm": 3.0491748486611576,
+      "learning_rate": 1.5879870285149232e-05,
+      "loss": 2.0432,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6420097697138869,
+      "grad_norm": 2.822007576139157,
+      "learning_rate": 1.5877348025223458e-05,
+      "loss": 2.0804,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "grad_norm": 2.9273675253440583,
+      "learning_rate": 1.5874825193944264e-05,
+      "loss": 2.0267,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.7186,
+      "step": 3336,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 2.0483,
+      "step": 3336,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.8428,
+      "step": 3336,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.7132,
+      "step": 3336,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.639,
+      "step": 3336,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.7427,
+      "step": 3336,
+      "vm_loss": 0.2607
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.6197,
+      "step": 3336,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.6422022763914623,
+      "lm_loss": 1.4923,
+      "step": 3336,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.6423947830690377,
+      "grad_norm": 2.9028976799538664,
+      "learning_rate": 1.5872301791556904e-05,
+      "loss": 2.0061,
+      "step": 3337
+    },
+    {
+      "epoch": 0.642587289746613,
+      "grad_norm": 3.106169298332078,
+      "learning_rate": 1.586977781830668e-05,
+      "loss": 2.0524,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6427797964241885,
+      "grad_norm": 3.0360476401193814,
+      "learning_rate": 1.5867253274438954e-05,
+      "loss": 2.1155,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6429723031017639,
+      "grad_norm": 2.9732666727001673,
+      "learning_rate": 1.5864728160199145e-05,
+      "loss": 2.0105,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6431648097793392,
+      "grad_norm": 3.057200243234311,
+      "learning_rate": 1.586220247583273e-05,
+      "loss": 2.0101,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6433573164569146,
+      "grad_norm": 2.93237720166904,
+      "learning_rate": 1.5859676221585233e-05,
+      "loss": 2.0244,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6435498231344899,
+      "grad_norm": 2.7617260927879292,
+      "learning_rate": 1.585714939770224e-05,
+      "loss": 2.0318,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "grad_norm": 2.8552648221139925,
+      "learning_rate": 1.5854622004429392e-05,
+      "loss": 2.0279,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.2019,
+      "step": 3344,
+      "vm_loss": 0.1113
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.5573,
+      "step": 3344,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.3952,
+      "step": 3344,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.523,
+      "step": 3344,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.6737,
+      "step": 3344,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.987,
+      "step": 3344,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.3552,
+      "step": 3344,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 0.6437423298120654,
+      "lm_loss": 1.9727,
+      "step": 3344,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 0.6439348364896408,
+      "grad_norm": 2.8998247702239968,
+      "learning_rate": 1.5852094042012384e-05,
+      "loss": 1.9637,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6441273431672161,
+      "grad_norm": 3.0268684376395307,
+      "learning_rate": 1.584956551069696e-05,
+      "loss": 2.0307,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6443198498447915,
+      "grad_norm": 2.8674829263791044,
+      "learning_rate": 1.5847036410728933e-05,
+      "loss": 2.0548,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6445123565223668,
+      "grad_norm": 2.874074737392052,
+      "learning_rate": 1.5844506742354163e-05,
+      "loss": 1.9918,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6447048631999422,
+      "grad_norm": 2.8648274573427064,
+      "learning_rate": 1.584197650581856e-05,
+      "loss": 2.0863,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6448973698775177,
+      "grad_norm": 2.7674233658501364,
+      "learning_rate": 1.5839445701368104e-05,
+      "loss": 2.0943,
+      "step": 3350
+    },
+    {
+      "epoch": 0.645089876555093,
+      "grad_norm": 2.8961242482338214,
+      "learning_rate": 1.5836914329248817e-05,
+      "loss": 2.0335,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "grad_norm": 2.9747271683162566,
+      "learning_rate": 1.583438238970678e-05,
+      "loss": 2.0065,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 1.688,
+      "step": 3352,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 2.1008,
+      "step": 3352,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 1.8394,
+      "step": 3352,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 1.4674,
+      "step": 3352,
+      "vm_loss": 0.2615
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 1.8466,
+      "step": 3352,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 2.0024,
+      "step": 3352,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 2.1023,
+      "step": 3352,
+      "vm_loss": 0.2063
+    },
+    {
+      "epoch": 0.6452823832326684,
+      "lm_loss": 2.0437,
+      "step": 3352,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 0.6454748899102437,
+      "grad_norm": 2.9917526492886415,
+      "learning_rate": 1.5831849882988138e-05,
+      "loss": 2.1314,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6456673965878191,
+      "grad_norm": 2.758164479447144,
+      "learning_rate": 1.5829316809339068e-05,
+      "loss": 1.9959,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6458599032653946,
+      "grad_norm": 2.8806748694174558,
+      "learning_rate": 1.5826783169005827e-05,
+      "loss": 2.0012,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6460524099429699,
+      "grad_norm": 2.92702907914621,
+      "learning_rate": 1.5824248962234715e-05,
+      "loss": 1.9718,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6462449166205453,
+      "grad_norm": 2.9915276550254633,
+      "learning_rate": 1.5821714189272093e-05,
+      "loss": 2.0118,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6464374232981207,
+      "grad_norm": 3.071158921315843,
+      "learning_rate": 1.581917885036437e-05,
+      "loss": 2.0091,
+      "step": 3358
+    },
+    {
+      "epoch": 0.646629929975696,
+      "grad_norm": 2.9865223380529216,
+      "learning_rate": 1.581664294575801e-05,
+      "loss": 2.081,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "grad_norm": 2.874942712172753,
+      "learning_rate": 1.5814106475699546e-05,
+      "loss": 2.0613,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.6293,
+      "step": 3360,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.3367,
+      "step": 3360,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 2.2137,
+      "step": 3360,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.873,
+      "step": 3360,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.7447,
+      "step": 3360,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.9527,
+      "step": 3360,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 2.1401,
+      "step": 3360,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.6468224366532714,
+      "lm_loss": 1.7832,
+      "step": 3360,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.6470149433308467,
+      "grad_norm": 2.965084940060805,
+      "learning_rate": 1.581156944043554e-05,
+      "loss": 2.0341,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6472074500084222,
+      "grad_norm": 2.9535612731752185,
+      "learning_rate": 1.580903184021264e-05,
+      "loss": 2.0015,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6473999566859976,
+      "grad_norm": 3.0264711841637504,
+      "learning_rate": 1.580649367527752e-05,
+      "loss": 2.0291,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6475924633635729,
+      "grad_norm": 3.0282319074475303,
+      "learning_rate": 1.5803954945876933e-05,
+      "loss": 2.0478,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6477849700411483,
+      "grad_norm": 3.0511656046216085,
+      "learning_rate": 1.580141565225767e-05,
+      "loss": 2.016,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6479774767187236,
+      "grad_norm": 3.0070316032997515,
+      "learning_rate": 1.5798875794666582e-05,
+      "loss": 2.0287,
+      "step": 3366
+    },
+    {
+      "epoch": 0.648169983396299,
+      "grad_norm": 2.9736436104394195,
+      "learning_rate": 1.5796335373350577e-05,
+      "loss": 1.9982,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "grad_norm": 2.851670807729729,
+      "learning_rate": 1.5793794388556617e-05,
+      "loss": 1.9933,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.6781,
+      "step": 3368,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.9421,
+      "step": 3368,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.5757,
+      "step": 3368,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.7888,
+      "step": 3368,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.523,
+      "step": 3368,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.9573,
+      "step": 3368,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.7266,
+      "step": 3368,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.6483624900738745,
+      "lm_loss": 1.8996,
+      "step": 3368,
+      "vm_loss": 0.185
+    },
+    {
+      "epoch": 0.6485549967514498,
+      "grad_norm": 2.8822716894955556,
+      "learning_rate": 1.579125284053172e-05,
+      "loss": 1.9607,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6487475034290252,
+      "grad_norm": 2.969764721719698,
+      "learning_rate": 1.5788710729522953e-05,
+      "loss": 1.9954,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6489400101066005,
+      "grad_norm": 3.0955256143874283,
+      "learning_rate": 1.5786168055777445e-05,
+      "loss": 2.065,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6491325167841759,
+      "grad_norm": 3.0266748730683055,
+      "learning_rate": 1.578362481954238e-05,
+      "loss": 1.9835,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6493250234617514,
+      "grad_norm": 2.81747653422955,
+      "learning_rate": 1.5781081021064983e-05,
+      "loss": 2.0244,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6495175301393267,
+      "grad_norm": 2.735177166904523,
+      "learning_rate": 1.5778536660592545e-05,
+      "loss": 1.9463,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6497100368169021,
+      "grad_norm": 2.8850286110875585,
+      "learning_rate": 1.577599173837242e-05,
+      "loss": 2.0777,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "grad_norm": 3.108746343431595,
+      "learning_rate": 1.5773446254652002e-05,
+      "loss": 2.0573,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 2.055,
+      "step": 3376,
+      "vm_loss": 0.2294
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 2.026,
+      "step": 3376,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 2.3197,
+      "step": 3376,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 1.7229,
+      "step": 3376,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 1.871,
+      "step": 3376,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 2.1273,
+      "step": 3376,
+      "vm_loss": 0.2058
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 1.6647,
+      "step": 3376,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 0.6499025434944775,
+      "lm_loss": 1.9205,
+      "step": 3376,
+      "vm_loss": 0.2001
+    },
+    {
+      "epoch": 0.6500950501720528,
+      "grad_norm": 2.8810753317904245,
+      "learning_rate": 1.5770900209678742e-05,
+      "loss": 1.9721,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6502875568496282,
+      "grad_norm": 3.2565079248723743,
+      "learning_rate": 1.5768353603700152e-05,
+      "loss": 1.9951,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6504800635272036,
+      "grad_norm": 3.4762835316162803,
+      "learning_rate": 1.5765806436963793e-05,
+      "loss": 2.0728,
+      "step": 3379
+    },
+    {
+      "epoch": 0.650672570204779,
+      "grad_norm": 2.819181062391101,
+      "learning_rate": 1.5763258709717282e-05,
+      "loss": 2.044,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6508650768823544,
+      "grad_norm": 2.7087932661598733,
+      "learning_rate": 1.576071042220829e-05,
+      "loss": 2.0344,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6510575835599297,
+      "grad_norm": 2.8468133262027124,
+      "learning_rate": 1.5758161574684544e-05,
+      "loss": 2.0382,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6512500902375051,
+      "grad_norm": 3.1302616853320964,
+      "learning_rate": 1.5755612167393827e-05,
+      "loss": 2.0711,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "grad_norm": 3.073042992911792,
+      "learning_rate": 1.5753062200583967e-05,
+      "loss": 2.1246,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.4069,
+      "step": 3384,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.5835,
+      "step": 3384,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.9258,
+      "step": 3384,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.3249,
+      "step": 3384,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 2.1316,
+      "step": 3384,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.77,
+      "step": 3384,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.8358,
+      "step": 3384,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 0.6514425969150804,
+      "lm_loss": 1.9636,
+      "step": 3384,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 0.6516351035926559,
+      "grad_norm": 2.7719716067245286,
+      "learning_rate": 1.575051167450286e-05,
+      "loss": 1.9522,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6518276102702313,
+      "grad_norm": 3.0614920779669066,
+      "learning_rate": 1.5747960589398447e-05,
+      "loss": 2.0054,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6520201169478066,
+      "grad_norm": 2.9202469318987765,
+      "learning_rate": 1.574540894551873e-05,
+      "loss": 1.9906,
+      "step": 3387
+    },
+    {
+      "epoch": 0.652212623625382,
+      "grad_norm": 3.1519819154667212,
+      "learning_rate": 1.5742856743111753e-05,
+      "loss": 2.0739,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6524051303029574,
+      "grad_norm": 2.928325350516434,
+      "learning_rate": 1.574030398242563e-05,
+      "loss": 1.9817,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6525976369805327,
+      "grad_norm": 3.397517726884812,
+      "learning_rate": 1.5737750663708526e-05,
+      "loss": 2.055,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6527901436581082,
+      "grad_norm": 3.0333572272710354,
+      "learning_rate": 1.573519678720864e-05,
+      "loss": 2.0125,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "grad_norm": 2.7685766341728137,
+      "learning_rate": 1.5732642353174257e-05,
+      "loss": 2.0549,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.9299,
+      "step": 3392,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.5882,
+      "step": 3392,
+      "vm_loss": 0.2296
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.9038,
+      "step": 3392,
+      "vm_loss": 0.1091
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.8111,
+      "step": 3392,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.9049,
+      "step": 3392,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 2.0374,
+      "step": 3392,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 1.7034,
+      "step": 3392,
+      "vm_loss": 0.1239
+    },
+    {
+      "epoch": 0.6529826503356835,
+      "lm_loss": 2.21,
+      "step": 3392,
+      "vm_loss": 0.2505
+    },
+    {
+      "epoch": 0.6531751570132589,
+      "grad_norm": 2.972010163136967,
+      "learning_rate": 1.5730087361853696e-05,
+      "loss": 2.0188,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6533676636908343,
+      "grad_norm": 3.3182017029799837,
+      "learning_rate": 1.5727531813495335e-05,
+      "loss": 2.0668,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6535601703684096,
+      "grad_norm": 3.03414851839463,
+      "learning_rate": 1.5724975708347603e-05,
+      "loss": 2.014,
+      "step": 3395
+    },
+    {
+      "epoch": 0.653752677045985,
+      "grad_norm": 2.8983963533454644,
+      "learning_rate": 1.572241904665899e-05,
+      "loss": 2.0301,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6539451837235604,
+      "grad_norm": 2.9062250834043186,
+      "learning_rate": 1.5719861828678033e-05,
+      "loss": 2.0651,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6541376904011358,
+      "grad_norm": 2.8714890842775183,
+      "learning_rate": 1.5717304054653327e-05,
+      "loss": 2.0567,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6543301970787112,
+      "grad_norm": 2.968658438168832,
+      "learning_rate": 1.5714745724833524e-05,
+      "loss": 1.996,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "grad_norm": 2.747876452270412,
+      "learning_rate": 1.571218683946732e-05,
+      "loss": 1.9906,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.7027,
+      "step": 3400,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.7728,
+      "step": 3400,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.5423,
+      "step": 3400,
+      "vm_loss": 0.1291
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.661,
+      "step": 3400,
+      "vm_loss": 0.2097
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.7011,
+      "step": 3400,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 2.2959,
+      "step": 3400,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 1.7793,
+      "step": 3400,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.6545227037562865,
+      "lm_loss": 2.2202,
+      "step": 3400,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 0.6547152104338619,
+      "grad_norm": 2.9835769459797272,
+      "learning_rate": 1.570962739880348e-05,
+      "loss": 2.0532,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6549077171114372,
+      "grad_norm": 3.150741838620774,
+      "learning_rate": 1.5707067403090803e-05,
+      "loss": 2.0205,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6551002237890127,
+      "grad_norm": 3.4170029251909724,
+      "learning_rate": 1.5704506852578165e-05,
+      "loss": 2.0636,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6552927304665881,
+      "grad_norm": 3.110253786012647,
+      "learning_rate": 1.5701945747514476e-05,
+      "loss": 2.0175,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6554852371441634,
+      "grad_norm": 2.68279732289107,
+      "learning_rate": 1.5699384088148713e-05,
+      "loss": 2.0557,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6556777438217388,
+      "grad_norm": 2.7271677324414854,
+      "learning_rate": 1.56968218747299e-05,
+      "loss": 2.0537,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6558702504993142,
+      "grad_norm": 2.697876411616252,
+      "learning_rate": 1.5694259107507117e-05,
+      "loss": 2.051,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "grad_norm": 2.9578479773799264,
+      "learning_rate": 1.5691695786729498e-05,
+      "loss": 1.9997,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.941,
+      "step": 3408,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.7293,
+      "step": 3408,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 2.207,
+      "step": 3408,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.6512,
+      "step": 3408,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.556,
+      "step": 3408,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.4449,
+      "step": 3408,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 1.6831,
+      "step": 3408,
+      "vm_loss": 0.2399
+    },
+    {
+      "epoch": 0.6560627571768896,
+      "lm_loss": 2.4147,
+      "step": 3408,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 0.656255263854465,
+      "grad_norm": 3.0444447262285594,
+      "learning_rate": 1.5689131912646227e-05,
+      "loss": 2.0458,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6564477705320403,
+      "grad_norm": 3.006755721771176,
+      "learning_rate": 1.5686567485506552e-05,
+      "loss": 2.0522,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6566402772096157,
+      "grad_norm": 2.6689143571135006,
+      "learning_rate": 1.5684002505559766e-05,
+      "loss": 2.0127,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6568327838871911,
+      "grad_norm": 2.7023171450616124,
+      "learning_rate": 1.5681436973055214e-05,
+      "loss": 2.0361,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6570252905647664,
+      "grad_norm": 2.709842390603092,
+      "learning_rate": 1.5678870888242304e-05,
+      "loss": 2.0,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6572177972423419,
+      "grad_norm": 2.922918978783161,
+      "learning_rate": 1.567630425137049e-05,
+      "loss": 2.0158,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6574103039199172,
+      "grad_norm": 2.8302159315183886,
+      "learning_rate": 1.567373706268928e-05,
+      "loss": 1.9985,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "grad_norm": 3.027700207435978,
+      "learning_rate": 1.567116932244824e-05,
+      "loss": 2.0694,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 1.9269,
+      "step": 3416,
+      "vm_loss": 0.2624
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 2.2354,
+      "step": 3416,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 2.1576,
+      "step": 3416,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 1.8478,
+      "step": 3416,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 2.3587,
+      "step": 3416,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 1.9866,
+      "step": 3416,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 2.0243,
+      "step": 3416,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.6576028105974926,
+      "lm_loss": 1.775,
+      "step": 3416,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.657795317275068,
+      "grad_norm": 2.847442502518517,
+      "learning_rate": 1.5668601030896987e-05,
+      "loss": 2.0471,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6579878239526433,
+      "grad_norm": 2.9557472795179085,
+      "learning_rate": 1.5666032188285188e-05,
+      "loss": 1.8998,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6581803306302187,
+      "grad_norm": 2.901403000409162,
+      "learning_rate": 1.5663462794862575e-05,
+      "loss": 2.0332,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6583728373077942,
+      "grad_norm": 2.8418445004034814,
+      "learning_rate": 1.566089285087892e-05,
+      "loss": 2.0205,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6585653439853695,
+      "grad_norm": 2.826850492192919,
+      "learning_rate": 1.565832235658405e-05,
+      "loss": 2.0542,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6587578506629449,
+      "grad_norm": 2.7836412097920373,
+      "learning_rate": 1.5655751312227863e-05,
+      "loss": 2.0041,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6589503573405202,
+      "grad_norm": 2.80879958368805,
+      "learning_rate": 1.565317971806029e-05,
+      "loss": 2.0289,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "grad_norm": 2.725278332808272,
+      "learning_rate": 1.565060757433132e-05,
+      "loss": 1.9269,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.9155,
+      "step": 3424,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.6846,
+      "step": 3424,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.2645,
+      "step": 3424,
+      "vm_loss": 0.2341
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.5602,
+      "step": 3424,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.5456,
+      "step": 3424,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.817,
+      "step": 3424,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.5246,
+      "step": 3424,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.6591428640180956,
+      "lm_loss": 1.4716,
+      "step": 3424,
+      "vm_loss": 0.2201
+    },
+    {
+      "epoch": 0.659335370695671,
+      "grad_norm": 2.987240946372525,
+      "learning_rate": 1.5648034881291005e-05,
+      "loss": 1.973,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6595278773732464,
+      "grad_norm": 2.8525445130621248,
+      "learning_rate": 1.5645461639189437e-05,
+      "loss": 1.9432,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6597203840508218,
+      "grad_norm": 2.8161380609631794,
+      "learning_rate": 1.5642887848276775e-05,
+      "loss": 2.0197,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6599128907283971,
+      "grad_norm": 2.805475129561156,
+      "learning_rate": 1.5640313508803217e-05,
+      "loss": 1.9881,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6601053974059725,
+      "grad_norm": 2.88272232888688,
+      "learning_rate": 1.5637738621019026e-05,
+      "loss": 2.0012,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6602979040835479,
+      "grad_norm": 2.8680119710289067,
+      "learning_rate": 1.5635163185174512e-05,
+      "loss": 1.9932,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6604904107611232,
+      "grad_norm": 2.9984468936883126,
+      "learning_rate": 1.5632587201520042e-05,
+      "loss": 1.9723,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "grad_norm": 2.7987091532425934,
+      "learning_rate": 1.5630010670306037e-05,
+      "loss": 2.0323,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.9288,
+      "step": 3432,
+      "vm_loss": 0.1695
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.6718,
+      "step": 3432,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 2.2938,
+      "step": 3432,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.6039,
+      "step": 3432,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 2.0922,
+      "step": 3432,
+      "vm_loss": 0.11
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.8396,
+      "step": 3432,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.5361,
+      "step": 3432,
+      "vm_loss": 0.0889
+    },
+    {
+      "epoch": 0.6606829174386987,
+      "lm_loss": 1.941,
+      "step": 3432,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.660875424116274,
+      "grad_norm": 2.8881863040351314,
+      "learning_rate": 1.5627433591782963e-05,
+      "loss": 2.0371,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6610679307938494,
+      "grad_norm": 2.8017031745412315,
+      "learning_rate": 1.562485596620135e-05,
+      "loss": 1.9789,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6612604374714248,
+      "grad_norm": 2.8428717915391193,
+      "learning_rate": 1.562227779381177e-05,
+      "loss": 1.9797,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6614529441490001,
+      "grad_norm": 3.083704515675064,
+      "learning_rate": 1.5619699074864864e-05,
+      "loss": 2.0164,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6616454508265756,
+      "grad_norm": 2.855110285247169,
+      "learning_rate": 1.5617119809611308e-05,
+      "loss": 2.0349,
+      "step": 3437
+    },
+    {
+      "epoch": 0.661837957504151,
+      "grad_norm": 2.944567963600729,
+      "learning_rate": 1.561453999830184e-05,
+      "loss": 2.0145,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6620304641817263,
+      "grad_norm": 3.1023614657863194,
+      "learning_rate": 1.561195964118726e-05,
+      "loss": 1.988,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "grad_norm": 2.8243815231457674,
+      "learning_rate": 1.56093787385184e-05,
+      "loss": 1.9731,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 2.1885,
+      "step": 3440,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 1.4917,
+      "step": 3440,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 1.6558,
+      "step": 3440,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 2.1221,
+      "step": 3440,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 2.1218,
+      "step": 3440,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 1.8281,
+      "step": 3440,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 1.7897,
+      "step": 3440,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 0.6622229708593017,
+      "lm_loss": 2.1482,
+      "step": 3440,
+      "vm_loss": 0.103
+    },
+    {
+      "epoch": 0.662415477536877,
+      "grad_norm": 2.7397043699901813,
+      "learning_rate": 1.560679729054616e-05,
+      "loss": 1.991,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6626079842144524,
+      "grad_norm": 3.0580459675589022,
+      "learning_rate": 1.5604215297521498e-05,
+      "loss": 2.0069,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6628004908920279,
+      "grad_norm": 3.041994247593464,
+      "learning_rate": 1.5601632759695408e-05,
+      "loss": 1.979,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6629929975696032,
+      "grad_norm": 2.9349032068403567,
+      "learning_rate": 1.5599049677318944e-05,
+      "loss": 1.9642,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6631855042471786,
+      "grad_norm": 3.024445542281148,
+      "learning_rate": 1.5596466050643225e-05,
+      "loss": 2.0155,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6633780109247539,
+      "grad_norm": 3.1267719054774523,
+      "learning_rate": 1.5593881879919403e-05,
+      "loss": 1.952,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6635705176023293,
+      "grad_norm": 3.117584784320607,
+      "learning_rate": 1.5591297165398693e-05,
+      "loss": 2.028,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "grad_norm": 2.9688443606527937,
+      "learning_rate": 1.5588711907332368e-05,
+      "loss": 2.0514,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.8242,
+      "step": 3448,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.5338,
+      "step": 3448,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.9291,
+      "step": 3448,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 2.2259,
+      "step": 3448,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.7583,
+      "step": 3448,
+      "vm_loss": 0.1308
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.8037,
+      "step": 3448,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.8267,
+      "step": 3448,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.6637630242799047,
+      "lm_loss": 1.9707,
+      "step": 3448,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 0.6639555309574801,
+      "grad_norm": 3.0209883449226234,
+      "learning_rate": 1.5586126105971744e-05,
+      "loss": 1.9607,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6641480376350555,
+      "grad_norm": 3.1492199274650203,
+      "learning_rate": 1.5583539761568194e-05,
+      "loss": 1.9427,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6643405443126309,
+      "grad_norm": 2.9140437650949655,
+      "learning_rate": 1.558095287437315e-05,
+      "loss": 2.0092,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6645330509902062,
+      "grad_norm": 3.1676452757072386,
+      "learning_rate": 1.5578365444638078e-05,
+      "loss": 1.9852,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6647255576677816,
+      "grad_norm": 3.010447397114762,
+      "learning_rate": 1.557577747261452e-05,
+      "loss": 2.0513,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6649180643453569,
+      "grad_norm": 3.022548399018991,
+      "learning_rate": 1.5573188958554056e-05,
+      "loss": 2.0029,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6651105710229324,
+      "grad_norm": 2.845997979383061,
+      "learning_rate": 1.5570599902708325e-05,
+      "loss": 2.01,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "grad_norm": 3.0622108945712543,
+      "learning_rate": 1.5568010305329012e-05,
+      "loss": 2.0546,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 2.2959,
+      "step": 3456,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 1.7777,
+      "step": 3456,
+      "vm_loss": 0.1111
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 1.508,
+      "step": 3456,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 1.6771,
+      "step": 3456,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 2.1553,
+      "step": 3456,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 1.6149,
+      "step": 3456,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 2.0503,
+      "step": 3456,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 0.6653030777005078,
+      "lm_loss": 1.5738,
+      "step": 3456,
+      "vm_loss": 0.1171
+    },
+    {
+      "epoch": 0.6654955843780831,
+      "grad_norm": 3.052977472133687,
+      "learning_rate": 1.5565420166667867e-05,
+      "loss": 2.0091,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6656880910556585,
+      "grad_norm": 3.1487828529891964,
+      "learning_rate": 1.5562829486976675e-05,
+      "loss": 2.0333,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6658805977332338,
+      "grad_norm": 2.8968394397514294,
+      "learning_rate": 1.5560238266507287e-05,
+      "loss": 2.005,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6660731044108092,
+      "grad_norm": 2.877036579317055,
+      "learning_rate": 1.5557646505511605e-05,
+      "loss": 2.0145,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6662656110883847,
+      "grad_norm": 2.7400517292784876,
+      "learning_rate": 1.555505420424158e-05,
+      "loss": 2.019,
+      "step": 3461
+    },
+    {
+      "epoch": 0.66645811776596,
+      "grad_norm": 3.2083964470678517,
+      "learning_rate": 1.5552461362949214e-05,
+      "loss": 1.9642,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6666506244435354,
+      "grad_norm": 2.747255053376632,
+      "learning_rate": 1.5549867981886567e-05,
+      "loss": 2.0075,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "grad_norm": 2.8665802575812127,
+      "learning_rate": 1.5547274061305753e-05,
+      "loss": 1.9852,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 2.3415,
+      "step": 3464,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.8594,
+      "step": 3464,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.8048,
+      "step": 3464,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.9043,
+      "step": 3464,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.9511,
+      "step": 3464,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.7978,
+      "step": 3464,
+      "vm_loss": 0.2501
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.6212,
+      "step": 3464,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 0.6668431311211107,
+      "lm_loss": 1.9876,
+      "step": 3464,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 0.6670356377986861,
+      "grad_norm": 2.9883267350361997,
+      "learning_rate": 1.5544679601458925e-05,
+      "loss": 2.01,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6672281444762616,
+      "grad_norm": 3.021619780823117,
+      "learning_rate": 1.5542084602598305e-05,
+      "loss": 1.9351,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6674206511538369,
+      "grad_norm": 3.0649062889123835,
+      "learning_rate": 1.5539489064976157e-05,
+      "loss": 2.0582,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6676131578314123,
+      "grad_norm": 2.880175633373562,
+      "learning_rate": 1.55368929888448e-05,
+      "loss": 2.0809,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6678056645089877,
+      "grad_norm": 2.7253570766901136,
+      "learning_rate": 1.553429637445661e-05,
+      "loss": 1.9596,
+      "step": 3469
+    },
+    {
+      "epoch": 0.667998171186563,
+      "grad_norm": 2.821924131951273,
+      "learning_rate": 1.5531699222064005e-05,
+      "loss": 2.0052,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6681906778641384,
+      "grad_norm": 2.7434439103369708,
+      "learning_rate": 1.5529101531919465e-05,
+      "loss": 1.9824,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "grad_norm": 3.0967602915104013,
+      "learning_rate": 1.5526503304275518e-05,
+      "loss": 2.0319,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.6055,
+      "step": 3472,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.9154,
+      "step": 3472,
+      "vm_loss": 0.1283
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.571,
+      "step": 3472,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.6757,
+      "step": 3472,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.7838,
+      "step": 3472,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.9148,
+      "step": 3472,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.8315,
+      "step": 3472,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 0.6683831845417137,
+      "lm_loss": 1.6614,
+      "step": 3472,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.6685756912192892,
+      "grad_norm": 3.081452169159797,
+      "learning_rate": 1.5523904539384746e-05,
+      "loss": 1.9643,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6687681978968646,
+      "grad_norm": 3.2058894146864403,
+      "learning_rate": 1.5521305237499785e-05,
+      "loss": 2.0005,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6689607045744399,
+      "grad_norm": 2.9795203766456693,
+      "learning_rate": 1.5518705398873312e-05,
+      "loss": 1.9782,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6691532112520153,
+      "grad_norm": 2.7351741695000906,
+      "learning_rate": 1.551610502375807e-05,
+      "loss": 2.0507,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6693457179295906,
+      "grad_norm": 2.67702214854956,
+      "learning_rate": 1.551350411240685e-05,
+      "loss": 2.0183,
+      "step": 3477
+    },
+    {
+      "epoch": 0.669538224607166,
+      "grad_norm": 2.7198653300284925,
+      "learning_rate": 1.551090266507249e-05,
+      "loss": 1.9985,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6697307312847415,
+      "grad_norm": 2.8843065969743185,
+      "learning_rate": 1.5508300682007885e-05,
+      "loss": 1.9538,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "grad_norm": 2.967227707552942,
+      "learning_rate": 1.5505698163465986e-05,
+      "loss": 2.0143,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 1.4283,
+      "step": 3480,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 1.9501,
+      "step": 3480,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 1.8329,
+      "step": 3480,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 1.8189,
+      "step": 3480,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 2.2648,
+      "step": 3480,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 2.2595,
+      "step": 3480,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 2.2414,
+      "step": 3480,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 0.6699232379623168,
+      "lm_loss": 1.6074,
+      "step": 3480,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 0.6701157446398922,
+      "grad_norm": 3.412154765084055,
+      "learning_rate": 1.5503095109699783e-05,
+      "loss": 1.9997,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6703082513174676,
+      "grad_norm": 3.0734874352038335,
+      "learning_rate": 1.5500491520962333e-05,
+      "loss": 2.0571,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6705007579950429,
+      "grad_norm": 2.8309247108213307,
+      "learning_rate": 1.5497887397506735e-05,
+      "loss": 2.0329,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6706932646726184,
+      "grad_norm": 2.9852449930169342,
+      "learning_rate": 1.549528273958614e-05,
+      "loss": 1.9516,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6708857713501937,
+      "grad_norm": 2.751853182656892,
+      "learning_rate": 1.549267754745376e-05,
+      "loss": 1.9794,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6710782780277691,
+      "grad_norm": 2.8104850029035804,
+      "learning_rate": 1.5490071821362855e-05,
+      "loss": 1.9668,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6712707847053445,
+      "grad_norm": 3.1035222028992684,
+      "learning_rate": 1.5487465561566725e-05,
+      "loss": 1.9581,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "grad_norm": 3.1953144326479324,
+      "learning_rate": 1.5484858768318738e-05,
+      "loss": 1.9561,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.86,
+      "step": 3488,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.9647,
+      "step": 3488,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.919,
+      "step": 3488,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.9231,
+      "step": 3488,
+      "vm_loss": 0.2286
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.3727,
+      "step": 3488,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 1.5676,
+      "step": 3488,
+      "vm_loss": 0.1194
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 2.2973,
+      "step": 3488,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.6714632913829198,
+      "lm_loss": 2.233,
+      "step": 3488,
+      "vm_loss": 0.2083
+    },
+    {
+      "epoch": 0.6716557980604952,
+      "grad_norm": 3.04166470809487,
+      "learning_rate": 1.5482251441872306e-05,
+      "loss": 2.0468,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6718483047380706,
+      "grad_norm": 3.145946815675012,
+      "learning_rate": 1.54796435824809e-05,
+      "loss": 2.0161,
+      "step": 3490
+    },
+    {
+      "epoch": 0.672040811415646,
+      "grad_norm": 2.822444943732931,
+      "learning_rate": 1.5477035190398028e-05,
+      "loss": 1.9597,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6722333180932214,
+      "grad_norm": 2.9098668100064637,
+      "learning_rate": 1.5474426265877266e-05,
+      "loss": 1.9746,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6724258247707967,
+      "grad_norm": 3.013575404485257,
+      "learning_rate": 1.5471816809172228e-05,
+      "loss": 1.9756,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6726183314483721,
+      "grad_norm": 2.9444460624530175,
+      "learning_rate": 1.5469206820536595e-05,
+      "loss": 2.0041,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6728108381259474,
+      "grad_norm": 2.979265390403357,
+      "learning_rate": 1.5466596300224087e-05,
+      "loss": 2.0074,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "grad_norm": 2.9324685917280675,
+      "learning_rate": 1.5463985248488484e-05,
+      "loss": 1.9547,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 1.9443,
+      "step": 3496,
+      "vm_loss": 0.0981
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 2.3861,
+      "step": 3496,
+      "vm_loss": 0.2585
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 1.8024,
+      "step": 3496,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 1.7392,
+      "step": 3496,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 2.0842,
+      "step": 3496,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 1.962,
+      "step": 3496,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 2.1859,
+      "step": 3496,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 0.6730033448035229,
+      "lm_loss": 2.2246,
+      "step": 3496,
+      "vm_loss": 0.1247
+    },
+    {
+      "epoch": 0.6731958514810983,
+      "grad_norm": 3.1319824903009676,
+      "learning_rate": 1.5461373665583606e-05,
+      "loss": 2.033,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6733883581586736,
+      "grad_norm": 2.936784868205389,
+      "learning_rate": 1.545876155176334e-05,
+      "loss": 2.0058,
+      "step": 3498
+    },
+    {
+      "epoch": 0.673580864836249,
+      "grad_norm": 3.0620857184280483,
+      "learning_rate": 1.545614890728161e-05,
+      "loss": 1.9886,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6737733715138244,
+      "grad_norm": 2.8170381406077,
+      "learning_rate": 1.5453535732392397e-05,
+      "loss": 1.9922,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6739658781913997,
+      "grad_norm": 2.86332818896035,
+      "learning_rate": 1.5450922027349747e-05,
+      "loss": 2.0093,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6741583848689752,
+      "grad_norm": 2.7930804059793273,
+      "learning_rate": 1.5448307792407737e-05,
+      "loss": 2.0342,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6743508915465505,
+      "grad_norm": 2.9960072303819256,
+      "learning_rate": 1.5445693027820503e-05,
+      "loss": 2.0626,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "grad_norm": 3.110512417480102,
+      "learning_rate": 1.544307773384224e-05,
+      "loss": 1.9193,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.8012,
+      "step": 3504,
+      "vm_loss": 0.2022
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.7025,
+      "step": 3504,
+      "vm_loss": 0.1008
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.9561,
+      "step": 3504,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.6308,
+      "step": 3504,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.6486,
+      "step": 3504,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.7085,
+      "step": 3504,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.7881,
+      "step": 3504,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.6745433982241259,
+      "lm_loss": 1.8465,
+      "step": 3504,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.6747359049017013,
+      "grad_norm": 3.150743693277032,
+      "learning_rate": 1.544046191072718e-05,
+      "loss": 1.9663,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6749284115792766,
+      "grad_norm": 2.8846130321638426,
+      "learning_rate": 1.543784555872962e-05,
+      "loss": 1.9214,
+      "step": 3506
+    },
+    {
+      "epoch": 0.675120918256852,
+      "grad_norm": 3.0299567165442176,
+      "learning_rate": 1.54352286781039e-05,
+      "loss": 1.9829,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6753134249344274,
+      "grad_norm": 2.981331500565479,
+      "learning_rate": 1.5432611269104416e-05,
+      "loss": 1.9782,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6755059316120028,
+      "grad_norm": 2.8495231835086816,
+      "learning_rate": 1.5429993331985613e-05,
+      "loss": 1.9637,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6756984382895782,
+      "grad_norm": 2.9049199773040524,
+      "learning_rate": 1.5427374867001993e-05,
+      "loss": 1.9932,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6758909449671535,
+      "grad_norm": 3.1018396989524217,
+      "learning_rate": 1.5424755874408095e-05,
+      "loss": 1.9572,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "grad_norm": 2.872796161759113,
+      "learning_rate": 1.5422136354458522e-05,
+      "loss": 2.0115,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.3793,
+      "step": 3512,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.3485,
+      "step": 3512,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.6604,
+      "step": 3512,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.3108,
+      "step": 3512,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.8638,
+      "step": 3512,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.6561,
+      "step": 3512,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 1.938,
+      "step": 3512,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 0.6760834516447289,
+      "lm_loss": 2.223,
+      "step": 3512,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 0.6762759583223044,
+      "grad_norm": 2.7047768775248153,
+      "learning_rate": 1.5419516307407925e-05,
+      "loss": 1.9297,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6764684649998797,
+      "grad_norm": 2.7927486272194435,
+      "learning_rate": 1.541689573351101e-05,
+      "loss": 2.1054,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6766609716774551,
+      "grad_norm": 3.0514903264849935,
+      "learning_rate": 1.5414274633022533e-05,
+      "loss": 1.9215,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6768534783550304,
+      "grad_norm": 2.7581916942521696,
+      "learning_rate": 1.5411653006197284e-05,
+      "loss": 2.0527,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6770459850326058,
+      "grad_norm": 2.9838382743238934,
+      "learning_rate": 1.540903085329013e-05,
+      "loss": 2.0438,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6772384917101812,
+      "grad_norm": 2.872790819795221,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 1.9362,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6774309983877566,
+      "grad_norm": 2.8707493577267456,
+      "learning_rate": 1.540378497024978e-05,
+      "loss": 1.9643,
+      "step": 3519
+    },
+    {
+      "epoch": 0.677623505065332,
+      "grad_norm": 2.89725179771934,
+      "learning_rate": 1.540116124062655e-05,
+      "loss": 2.0444,
+      "step": 3520
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.7288,
+      "step": 3520,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.9892,
+      "step": 3520,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 2.0788,
+      "step": 3520,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.7419,
+      "step": 3520,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.8654,
+      "step": 3520,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.9472,
+      "step": 3520,
+      "vm_loss": 0.2365
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 1.8347,
+      "step": 3520,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 0.677623505065332,
+      "lm_loss": 2.0136,
+      "step": 3520,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 0.6778160117429073,
+      "grad_norm": 2.8276560865305944,
+      "learning_rate": 1.5398536985941347e-05,
+      "loss": 2.0026,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6780085184204827,
+      "grad_norm": 2.9067846424046593,
+      "learning_rate": 1.539591220644928e-05,
+      "loss": 1.9606,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6782010250980581,
+      "grad_norm": 2.8439958339052245,
+      "learning_rate": 1.5393286902405513e-05,
+      "loss": 1.9776,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6783935317756334,
+      "grad_norm": 2.8976549312203863,
+      "learning_rate": 1.5390661074065257e-05,
+      "loss": 2.0439,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6785860384532089,
+      "grad_norm": 3.1006514730406867,
+      "learning_rate": 1.5388034721683783e-05,
+      "loss": 1.9781,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6787785451307842,
+      "grad_norm": 2.7939585585356217,
+      "learning_rate": 1.5385407845516395e-05,
+      "loss": 1.9368,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6789710518083596,
+      "grad_norm": 2.8590422138281264,
+      "learning_rate": 1.5382780445818468e-05,
+      "loss": 1.955,
+      "step": 3527
+    },
+    {
+      "epoch": 0.679163558485935,
+      "grad_norm": 3.065389616390679,
+      "learning_rate": 1.5380152522845416e-05,
+      "loss": 2.039,
+      "step": 3528
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 1.68,
+      "step": 3528,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 1.7364,
+      "step": 3528,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 2.5211,
+      "step": 3528,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 2.186,
+      "step": 3528,
+      "vm_loss": 0.2242
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 1.5331,
+      "step": 3528,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 1.727,
+      "step": 3528,
+      "vm_loss": 0.2105
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 2.0293,
+      "step": 3528,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.679163558485935,
+      "lm_loss": 1.4385,
+      "step": 3528,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.6793560651635103,
+      "grad_norm": 3.14617451915695,
+      "learning_rate": 1.5377524076852704e-05,
+      "loss": 1.948,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6795485718410857,
+      "grad_norm": 2.876815650812825,
+      "learning_rate": 1.5374895108095858e-05,
+      "loss": 1.9959,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6797410785186612,
+      "grad_norm": 2.995124562180894,
+      "learning_rate": 1.537226561683044e-05,
+      "loss": 2.0018,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6799335851962365,
+      "grad_norm": 2.8711812447192915,
+      "learning_rate": 1.5369635603312073e-05,
+      "loss": 1.9582,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6801260918738119,
+      "grad_norm": 2.8699348550468917,
+      "learning_rate": 1.5367005067796426e-05,
+      "loss": 2.0119,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6803185985513872,
+      "grad_norm": 2.7712292799927036,
+      "learning_rate": 1.5364374010539225e-05,
+      "loss": 1.9118,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6805111052289626,
+      "grad_norm": 2.7486627795720633,
+      "learning_rate": 1.5361742431796235e-05,
+      "loss": 2.0,
+      "step": 3535
+    },
+    {
+      "epoch": 0.680703611906538,
+      "grad_norm": 2.8599852390444416,
+      "learning_rate": 1.535911033182329e-05,
+      "loss": 1.8943,
+      "step": 3536
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 1.753,
+      "step": 3536,
+      "vm_loss": 0.1128
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 2.017,
+      "step": 3536,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 1.8967,
+      "step": 3536,
+      "vm_loss": 0.1209
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 1.8915,
+      "step": 3536,
+      "vm_loss": 0.2178
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 1.6509,
+      "step": 3536,
+      "vm_loss": 0.2328
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 2.1272,
+      "step": 3536,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 2.0965,
+      "step": 3536,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 0.680703611906538,
+      "lm_loss": 1.7545,
+      "step": 3536,
+      "vm_loss": 0.0953
+    },
+    {
+      "epoch": 0.6808961185841134,
+      "grad_norm": 3.0179392952066126,
+      "learning_rate": 1.5356477710876255e-05,
+      "loss": 1.9832,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6810886252616888,
+      "grad_norm": 3.0031548743576773,
+      "learning_rate": 1.5353844569211056e-05,
+      "loss": 1.9833,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6812811319392641,
+      "grad_norm": 2.765648008151935,
+      "learning_rate": 1.535121090708367e-05,
+      "loss": 1.8995,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6814736386168395,
+      "grad_norm": 2.9755962524505675,
+      "learning_rate": 1.5348576724750123e-05,
+      "loss": 1.9483,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6816661452944149,
+      "grad_norm": 2.9369714167325784,
+      "learning_rate": 1.5345942022466488e-05,
+      "loss": 2.0139,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6818586519719902,
+      "grad_norm": 3.1180432527990414,
+      "learning_rate": 1.5343306800488896e-05,
+      "loss": 1.9913,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6820511586495657,
+      "grad_norm": 3.0379557253781018,
+      "learning_rate": 1.534067105907352e-05,
+      "loss": 1.9596,
+      "step": 3543
+    },
+    {
+      "epoch": 0.682243665327141,
+      "grad_norm": 3.054172199435693,
+      "learning_rate": 1.533803479847659e-05,
+      "loss": 1.8881,
+      "step": 3544
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 2.1323,
+      "step": 3544,
+      "vm_loss": 0.2054
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.752,
+      "step": 3544,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.884,
+      "step": 3544,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 2.2649,
+      "step": 3544,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.831,
+      "step": 3544,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.8256,
+      "step": 3544,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.8808,
+      "step": 3544,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 0.682243665327141,
+      "lm_loss": 1.8405,
+      "step": 3544,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 0.6824361720047164,
+      "grad_norm": 3.122498022349623,
+      "learning_rate": 1.5335398018954385e-05,
+      "loss": 1.9769,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6826286786822918,
+      "grad_norm": 2.8998094669325027,
+      "learning_rate": 1.5332760720763232e-05,
+      "loss": 1.9797,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6828211853598671,
+      "grad_norm": 2.94157799970006,
+      "learning_rate": 1.5330122904159513e-05,
+      "loss": 1.9511,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6830136920374426,
+      "grad_norm": 3.0153973250992,
+      "learning_rate": 1.5327484569399655e-05,
+      "loss": 1.9852,
+      "step": 3548
+    },
+    {
+      "epoch": 0.683206198715018,
+      "grad_norm": 3.2771290967265823,
+      "learning_rate": 1.532484571674014e-05,
+      "loss": 1.9962,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6833987053925933,
+      "grad_norm": 2.877853758691398,
+      "learning_rate": 1.5322206346437492e-05,
+      "loss": 1.9261,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6835912120701687,
+      "grad_norm": 2.9096314659106577,
+      "learning_rate": 1.5319566458748303e-05,
+      "loss": 2.0331,
+      "step": 3551
+    },
+    {
+      "epoch": 0.683783718747744,
+      "grad_norm": 2.93229005971163,
+      "learning_rate": 1.5316926053929192e-05,
+      "loss": 1.9389,
+      "step": 3552
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.6662,
+      "step": 3552,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.4992,
+      "step": 3552,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.9278,
+      "step": 3552,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 2.0457,
+      "step": 3552,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.9813,
+      "step": 3552,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.6245,
+      "step": 3552,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 2.0912,
+      "step": 3552,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 0.683783718747744,
+      "lm_loss": 1.5839,
+      "step": 3552,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 0.6839762254253194,
+      "grad_norm": 3.093072813815383,
+      "learning_rate": 1.5314285132236845e-05,
+      "loss": 1.9847,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6841687321028949,
+      "grad_norm": 2.848386084421865,
+      "learning_rate": 1.5311643693927997e-05,
+      "loss": 1.8948,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6843612387804702,
+      "grad_norm": 3.191335796802072,
+      "learning_rate": 1.530900173925942e-05,
+      "loss": 1.9773,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6845537454580456,
+      "grad_norm": 2.990210028114377,
+      "learning_rate": 1.5306359268487958e-05,
+      "loss": 1.9985,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6847462521356209,
+      "grad_norm": 2.9093156860972815,
+      "learning_rate": 1.5303716281870485e-05,
+      "loss": 2.0383,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6849387588131963,
+      "grad_norm": 2.8978387948492315,
+      "learning_rate": 1.5301072779663936e-05,
+      "loss": 1.9715,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6851312654907717,
+      "grad_norm": 3.018804339661139,
+      "learning_rate": 1.5298428762125293e-05,
+      "loss": 1.9616,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "grad_norm": 3.381703127663243,
+      "learning_rate": 1.5295784229511583e-05,
+      "loss": 1.9815,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.7722,
+      "step": 3560,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 2.1165,
+      "step": 3560,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.4734,
+      "step": 3560,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.6724,
+      "step": 3560,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 2.1123,
+      "step": 3560,
+      "vm_loss": 0.1092
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.6136,
+      "step": 3560,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.9916,
+      "step": 3560,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.6853237721683471,
+      "lm_loss": 1.5897,
+      "step": 3560,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 0.6855162788459225,
+      "grad_norm": 2.830539865804742,
+      "learning_rate": 1.5293139182079894e-05,
+      "loss": 1.9684,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6857087855234979,
+      "grad_norm": 2.9754972639752015,
+      "learning_rate": 1.5290493620087362e-05,
+      "loss": 1.9758,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6859012922010732,
+      "grad_norm": 3.093228135430268,
+      "learning_rate": 1.5287847543791162e-05,
+      "loss": 2.0164,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6860937988786486,
+      "grad_norm": 2.8552487592072344,
+      "learning_rate": 1.528520095344853e-05,
+      "loss": 1.9806,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6862863055562239,
+      "grad_norm": 2.9246862871822805,
+      "learning_rate": 1.5282553849316744e-05,
+      "loss": 2.0243,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6864788122337994,
+      "grad_norm": 3.1427723491893453,
+      "learning_rate": 1.5279906231653146e-05,
+      "loss": 1.989,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6866713189113748,
+      "grad_norm": 2.9267215787326077,
+      "learning_rate": 1.527725810071511e-05,
+      "loss": 1.9616,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "grad_norm": 3.178720792001327,
+      "learning_rate": 1.5274609456760073e-05,
+      "loss": 1.9679,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 2.0215,
+      "step": 3568,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.7344,
+      "step": 3568,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.9781,
+      "step": 3568,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.6267,
+      "step": 3568,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.6278,
+      "step": 3568,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.4253,
+      "step": 3568,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.8945,
+      "step": 3568,
+      "vm_loss": 0.2569
+    },
+    {
+      "epoch": 0.6868638255889501,
+      "lm_loss": 1.9701,
+      "step": 3568,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 0.6870563322665255,
+      "grad_norm": 3.105526323629938,
+      "learning_rate": 1.5271960300045514e-05,
+      "loss": 1.9754,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6872488389441008,
+      "grad_norm": 3.3532885915928636,
+      "learning_rate": 1.5269310630828963e-05,
+      "loss": 1.9227,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6874413456216762,
+      "grad_norm": 3.043739984183519,
+      "learning_rate": 1.526666044936801e-05,
+      "loss": 2.012,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6876338522992517,
+      "grad_norm": 2.9659747073530442,
+      "learning_rate": 1.5264009755920276e-05,
+      "loss": 2.0128,
+      "step": 3572
+    },
+    {
+      "epoch": 0.687826358976827,
+      "grad_norm": 3.355471471281819,
+      "learning_rate": 1.526135855074345e-05,
+      "loss": 1.9744,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6880188656544024,
+      "grad_norm": 3.1248765784788493,
+      "learning_rate": 1.5258706834095256e-05,
+      "loss": 1.9422,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6882113723319777,
+      "grad_norm": 2.857522659304052,
+      "learning_rate": 1.5256054606233481e-05,
+      "loss": 1.9944,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "grad_norm": 3.1002121909964195,
+      "learning_rate": 1.5253401867415958e-05,
+      "loss": 1.9869,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.6013,
+      "step": 3576,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 2.1497,
+      "step": 3576,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 2.248,
+      "step": 3576,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.8658,
+      "step": 3576,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.5132,
+      "step": 3576,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.2832,
+      "step": 3576,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.7683,
+      "step": 3576,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.6884038790095531,
+      "lm_loss": 1.7649,
+      "step": 3576,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 0.6885963856871286,
+      "grad_norm": 2.900853932298349,
+      "learning_rate": 1.525074861790056e-05,
+      "loss": 1.9473,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6887888923647039,
+      "grad_norm": 2.836583621256632,
+      "learning_rate": 1.524809485794522e-05,
+      "loss": 1.9117,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6889813990422793,
+      "grad_norm": 3.113230410017473,
+      "learning_rate": 1.5245440587807917e-05,
+      "loss": 1.9043,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6891739057198547,
+      "grad_norm": 3.1593715705332692,
+      "learning_rate": 1.5242785807746678e-05,
+      "loss": 1.9086,
+      "step": 3580
+    },
+    {
+      "epoch": 0.68936641239743,
+      "grad_norm": 3.181101193388769,
+      "learning_rate": 1.5240130518019588e-05,
+      "loss": 1.9745,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6895589190750054,
+      "grad_norm": 2.8789257814271627,
+      "learning_rate": 1.5237474718884767e-05,
+      "loss": 1.9761,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6897514257525807,
+      "grad_norm": 3.1152786611065135,
+      "learning_rate": 1.5234818410600396e-05,
+      "loss": 1.936,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "grad_norm": 2.876959467751351,
+      "learning_rate": 1.5232161593424704e-05,
+      "loss": 1.9802,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.9767,
+      "step": 3584,
+      "vm_loss": 0.2543
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.8907,
+      "step": 3584,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.5442,
+      "step": 3584,
+      "vm_loss": 0.1138
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.6459,
+      "step": 3584,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.7727,
+      "step": 3584,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 2.1747,
+      "step": 3584,
+      "vm_loss": 0.1148
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.8457,
+      "step": 3584,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 0.6899439324301562,
+      "lm_loss": 1.8709,
+      "step": 3584,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 0.6901364391077316,
+      "grad_norm": 2.8579243740427773,
+      "learning_rate": 1.5229504267615967e-05,
+      "loss": 1.9501,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6903289457853069,
+      "grad_norm": 3.1640023759376334,
+      "learning_rate": 1.522684643343251e-05,
+      "loss": 1.9739,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6905214524628823,
+      "grad_norm": 3.0110607080883125,
+      "learning_rate": 1.5224188091132708e-05,
+      "loss": 1.9444,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6907139591404576,
+      "grad_norm": 2.99224492667875,
+      "learning_rate": 1.5221529240974987e-05,
+      "loss": 1.9935,
+      "step": 3588
+    },
+    {
+      "epoch": 0.690906465818033,
+      "grad_norm": 3.088337571848301,
+      "learning_rate": 1.5218869883217822e-05,
+      "loss": 1.9547,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6910989724956085,
+      "grad_norm": 2.8448920643109803,
+      "learning_rate": 1.5216210018119735e-05,
+      "loss": 1.9707,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6912914791731838,
+      "grad_norm": 2.8230195051109,
+      "learning_rate": 1.5213549645939299e-05,
+      "loss": 1.969,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "grad_norm": 2.866308876996519,
+      "learning_rate": 1.5210888766935138e-05,
+      "loss": 1.9058,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 2.1078,
+      "step": 3592,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 1.6051,
+      "step": 3592,
+      "vm_loss": 0.1977
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 1.7529,
+      "step": 3592,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 1.5033,
+      "step": 3592,
+      "vm_loss": 0.0956
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 2.3358,
+      "step": 3592,
+      "vm_loss": 0.1135
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 1.9542,
+      "step": 3592,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 1.6266,
+      "step": 3592,
+      "vm_loss": 0.1122
+    },
+    {
+      "epoch": 0.6914839858507592,
+      "lm_loss": 2.1033,
+      "step": 3592,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.6916764925283346,
+      "grad_norm": 3.021169320973257,
+      "learning_rate": 1.5208227381365919e-05,
+      "loss": 1.9119,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6918689992059099,
+      "grad_norm": 3.090184515345662,
+      "learning_rate": 1.5205565489490369e-05,
+      "loss": 1.9892,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6920615058834854,
+      "grad_norm": 3.086627243066056,
+      "learning_rate": 1.5202903091567253e-05,
+      "loss": 1.931,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6922540125610607,
+      "grad_norm": 2.9501333184256193,
+      "learning_rate": 1.5200240187855392e-05,
+      "loss": 1.92,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6924465192386361,
+      "grad_norm": 2.824361407462446,
+      "learning_rate": 1.5197576778613657e-05,
+      "loss": 1.9343,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6926390259162115,
+      "grad_norm": 2.8894982685009403,
+      "learning_rate": 1.519491286410096e-05,
+      "loss": 2.036,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6928315325937868,
+      "grad_norm": 2.9326869747406703,
+      "learning_rate": 1.519224844457627e-05,
+      "loss": 1.9683,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "grad_norm": 2.924726968764479,
+      "learning_rate": 1.5189583520298605e-05,
+      "loss": 1.898,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.5069,
+      "step": 3600,
+      "vm_loss": 0.1117
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.7374,
+      "step": 3600,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.9252,
+      "step": 3600,
+      "vm_loss": 0.2247
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.9606,
+      "step": 3600,
+      "vm_loss": 0.1098
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.3729,
+      "step": 3600,
+      "vm_loss": 0.1437
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.607,
+      "step": 3600,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 1.6964,
+      "step": 3600,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.6930240392713622,
+      "lm_loss": 2.0102,
+      "step": 3600,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.6932165459489376,
+      "grad_norm": 3.0372860382335305,
+      "learning_rate": 1.5186918091527025e-05,
+      "loss": 1.9145,
+      "step": 3601
+    },
+    {
+      "epoch": 0.693409052626513,
+      "grad_norm": 2.9607599928161368,
+      "learning_rate": 1.5184252158520647e-05,
+      "loss": 1.8524,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6936015593040884,
+      "grad_norm": 2.9089548754353394,
+      "learning_rate": 1.5181585721538636e-05,
+      "loss": 1.9669,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6937940659816637,
+      "grad_norm": 2.819319271679664,
+      "learning_rate": 1.5178918780840198e-05,
+      "loss": 1.9424,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6939865726592391,
+      "grad_norm": 2.974853590223117,
+      "learning_rate": 1.5176251336684597e-05,
+      "loss": 1.9874,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6941790793368144,
+      "grad_norm": 2.8289683115907707,
+      "learning_rate": 1.5173583389331144e-05,
+      "loss": 1.9164,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6943715860143899,
+      "grad_norm": 2.9868943352681194,
+      "learning_rate": 1.5170914939039195e-05,
+      "loss": 2.016,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "grad_norm": 2.9452229250098387,
+      "learning_rate": 1.516824598606816e-05,
+      "loss": 1.987,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 1.8154,
+      "step": 3608,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 2.1617,
+      "step": 3608,
+      "vm_loss": 0.2547
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 1.9443,
+      "step": 3608,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 2.0248,
+      "step": 3608,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 2.1246,
+      "step": 3608,
+      "vm_loss": 0.1125
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 1.8375,
+      "step": 3608,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 2.3115,
+      "step": 3608,
+      "vm_loss": 0.1181
+    },
+    {
+      "epoch": 0.6945640926919653,
+      "lm_loss": 0.9842,
+      "step": 3608,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 0.6947565993695406,
+      "grad_norm": 2.948513754884626,
+      "learning_rate": 1.5165576530677492e-05,
+      "loss": 2.0207,
+      "step": 3609
+    },
+    {
+      "epoch": 0.694949106047116,
+      "grad_norm": 2.960162170971852,
+      "learning_rate": 1.5162906573126698e-05,
+      "loss": 1.9101,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6951416127246914,
+      "grad_norm": 2.909200606908281,
+      "learning_rate": 1.5160236113675333e-05,
+      "loss": 1.9988,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6953341194022667,
+      "grad_norm": 3.007786461709295,
+      "learning_rate": 1.5157565152583002e-05,
+      "loss": 1.9468,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6955266260798422,
+      "grad_norm": 2.81787849759017,
+      "learning_rate": 1.5154893690109349e-05,
+      "loss": 2.0016,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6957191327574175,
+      "grad_norm": 3.038201465904211,
+      "learning_rate": 1.5152221726514082e-05,
+      "loss": 1.9454,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6959116394349929,
+      "grad_norm": 3.0036994997596316,
+      "learning_rate": 1.5149549262056946e-05,
+      "loss": 1.9501,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "grad_norm": 3.1216605876343593,
+      "learning_rate": 1.5146876296997739e-05,
+      "loss": 1.9862,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 2.1122,
+      "step": 3616,
+      "vm_loss": 0.2121
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.6196,
+      "step": 3616,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.8496,
+      "step": 3616,
+      "vm_loss": 0.2051
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 2.2242,
+      "step": 3616,
+      "vm_loss": 0.2501
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.4669,
+      "step": 3616,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.8816,
+      "step": 3616,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.359,
+      "step": 3616,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 0.6961041461125683,
+      "lm_loss": 1.988,
+      "step": 3616,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.6962966527901436,
+      "grad_norm": 3.048908171464537,
+      "learning_rate": 1.514420283159631e-05,
+      "loss": 1.9869,
+      "step": 3617
+    },
+    {
+      "epoch": 0.696489159467719,
+      "grad_norm": 2.780863729468389,
+      "learning_rate": 1.5141528866112551e-05,
+      "loss": 1.8739,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6966816661452944,
+      "grad_norm": 2.915142335212109,
+      "learning_rate": 1.5138854400806408e-05,
+      "loss": 1.9529,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6968741728228698,
+      "grad_norm": 3.060927670706388,
+      "learning_rate": 1.5136179435937871e-05,
+      "loss": 1.9732,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6970666795004452,
+      "grad_norm": 2.7912706153539464,
+      "learning_rate": 1.513350397176698e-05,
+      "loss": 1.9412,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6972591861780205,
+      "grad_norm": 2.868448117584272,
+      "learning_rate": 1.513082800855383e-05,
+      "loss": 2.0046,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6974516928555959,
+      "grad_norm": 2.81072155310048,
+      "learning_rate": 1.5128151546558554e-05,
+      "loss": 1.9713,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "grad_norm": 3.11880247067057,
+      "learning_rate": 1.5125474586041337e-05,
+      "loss": 1.9949,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 1.8441,
+      "step": 3624,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 1.8873,
+      "step": 3624,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 2.231,
+      "step": 3624,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 2.0389,
+      "step": 3624,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 2.0288,
+      "step": 3624,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 1.7233,
+      "step": 3624,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 1.4897,
+      "step": 3624,
+      "vm_loss": 0.21
+    },
+    {
+      "epoch": 0.6976441995331714,
+      "lm_loss": 1.7322,
+      "step": 3624,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 0.6978367062107467,
+      "grad_norm": 2.8830523554256655,
+      "learning_rate": 1.5122797127262422e-05,
+      "loss": 1.9565,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6980292128883221,
+      "grad_norm": 2.890140601798388,
+      "learning_rate": 1.5120119170482079e-05,
+      "loss": 1.9505,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6982217195658974,
+      "grad_norm": 2.7371199214020074,
+      "learning_rate": 1.511744071596065e-05,
+      "loss": 1.9619,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6984142262434728,
+      "grad_norm": 2.839582880806902,
+      "learning_rate": 1.5114761763958513e-05,
+      "loss": 1.9895,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6986067329210482,
+      "grad_norm": 2.907063297765613,
+      "learning_rate": 1.5112082314736096e-05,
+      "loss": 2.0068,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6987992395986236,
+      "grad_norm": 2.9998919856400037,
+      "learning_rate": 1.5109402368553876e-05,
+      "loss": 2.0009,
+      "step": 3630
+    },
+    {
+      "epoch": 0.698991746276199,
+      "grad_norm": 2.948567062565866,
+      "learning_rate": 1.5106721925672379e-05,
+      "loss": 1.9757,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "grad_norm": 2.8687635027866394,
+      "learning_rate": 1.5104040986352173e-05,
+      "loss": 1.8674,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 2.153,
+      "step": 3632,
+      "vm_loss": 0.2083
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.9258,
+      "step": 3632,
+      "vm_loss": 0.1191
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.4406,
+      "step": 3632,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.7789,
+      "step": 3632,
+      "vm_loss": 0.2111
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.9998,
+      "step": 3632,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.6037,
+      "step": 3632,
+      "vm_loss": 0.2228
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.806,
+      "step": 3632,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 0.6991842529537743,
+      "lm_loss": 1.4199,
+      "step": 3632,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.6993767596313497,
+      "grad_norm": 3.1255405483968186,
+      "learning_rate": 1.5101359550853883e-05,
+      "loss": 1.9726,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6995692663089251,
+      "grad_norm": 3.0628078020029395,
+      "learning_rate": 1.5098677619438183e-05,
+      "loss": 1.9393,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6997617729865004,
+      "grad_norm": 2.804392237668242,
+      "learning_rate": 1.5095995192365784e-05,
+      "loss": 1.9323,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6999542796640759,
+      "grad_norm": 3.0333090747620948,
+      "learning_rate": 1.5093312269897458e-05,
+      "loss": 1.9089,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7001467863416512,
+      "grad_norm": 3.0138606987179255,
+      "learning_rate": 1.5090628852294014e-05,
+      "loss": 1.9429,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7003392930192266,
+      "grad_norm": 2.9190957374143776,
+      "learning_rate": 1.5087944939816322e-05,
+      "loss": 1.9847,
+      "step": 3638
+    },
+    {
+      "epoch": 0.700531799696802,
+      "grad_norm": 2.8850694022946577,
+      "learning_rate": 1.508526053272528e-05,
+      "loss": 1.9084,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "grad_norm": 2.959422989661412,
+      "learning_rate": 1.5082575631281862e-05,
+      "loss": 1.8962,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.8689,
+      "step": 3640,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.8218,
+      "step": 3640,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.7596,
+      "step": 3640,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.3301,
+      "step": 3640,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.7032,
+      "step": 3640,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.1456,
+      "step": 3640,
+      "vm_loss": 0.2242
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 1.5863,
+      "step": 3640,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.7007243063743773,
+      "lm_loss": 2.0456,
+      "step": 3640,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 0.7009168130519527,
+      "grad_norm": 2.9769855647541705,
+      "learning_rate": 1.5079890235747066e-05,
+      "loss": 1.9123,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7011093197295282,
+      "grad_norm": 3.2811469805953513,
+      "learning_rate": 1.5077204346381946e-05,
+      "loss": 1.9549,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7013018264071035,
+      "grad_norm": 3.0379382934444403,
+      "learning_rate": 1.507451796344761e-05,
+      "loss": 1.9615,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7014943330846789,
+      "grad_norm": 2.9636436235644736,
+      "learning_rate": 1.5071831087205205e-05,
+      "loss": 1.9491,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7016868397622542,
+      "grad_norm": 3.036526421297367,
+      "learning_rate": 1.5069143717915928e-05,
+      "loss": 1.9216,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7018793464398296,
+      "grad_norm": 3.1912969356836975,
+      "learning_rate": 1.5066455855841028e-05,
+      "loss": 1.9395,
+      "step": 3646
+    },
+    {
+      "epoch": 0.702071853117405,
+      "grad_norm": 2.9034727887408738,
+      "learning_rate": 1.5063767501241804e-05,
+      "loss": 1.9549,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "grad_norm": 3.185919849681255,
+      "learning_rate": 1.5061078654379593e-05,
+      "loss": 1.9251,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.7484,
+      "step": 3648,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.675,
+      "step": 3648,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.8368,
+      "step": 3648,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.7594,
+      "step": 3648,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.6704,
+      "step": 3648,
+      "vm_loss": 0.209
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.7657,
+      "step": 3648,
+      "vm_loss": 0.2118
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 2.1663,
+      "step": 3648,
+      "vm_loss": 0.2122
+    },
+    {
+      "epoch": 0.7022643597949804,
+      "lm_loss": 1.9826,
+      "step": 3648,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.7024568664725558,
+      "grad_norm": 3.1924538518851757,
+      "learning_rate": 1.5058389315515786e-05,
+      "loss": 2.0014,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7026493731501311,
+      "grad_norm": 3.3276572144474654,
+      "learning_rate": 1.5055699484911822e-05,
+      "loss": 2.0373,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7028418798277065,
+      "grad_norm": 3.0158872153573943,
+      "learning_rate": 1.5053009162829188e-05,
+      "loss": 1.9828,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7030343865052819,
+      "grad_norm": 2.879433544701226,
+      "learning_rate": 1.5050318349529416e-05,
+      "loss": 1.9595,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7032268931828572,
+      "grad_norm": 2.9908264220168177,
+      "learning_rate": 1.504762704527409e-05,
+      "loss": 1.962,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7034193998604327,
+      "grad_norm": 3.292653240593635,
+      "learning_rate": 1.5044935250324832e-05,
+      "loss": 1.9893,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7036119065380081,
+      "grad_norm": 2.9446634675233496,
+      "learning_rate": 1.504224296494333e-05,
+      "loss": 1.8828,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "grad_norm": 2.8583238934066952,
+      "learning_rate": 1.50395501893913e-05,
+      "loss": 1.9728,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.9844,
+      "step": 3656,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.6128,
+      "step": 3656,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.2618,
+      "step": 3656,
+      "vm_loss": 0.1098
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.8605,
+      "step": 3656,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.905,
+      "step": 3656,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 2.2919,
+      "step": 3656,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 1.7497,
+      "step": 3656,
+      "vm_loss": 0.2163
+    },
+    {
+      "epoch": 0.7038044132155834,
+      "lm_loss": 2.012,
+      "step": 3656,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 0.7039969198931588,
+      "grad_norm": 3.153595845368003,
+      "learning_rate": 1.5036856923930514e-05,
+      "loss": 1.9101,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7041894265707341,
+      "grad_norm": 2.9442524495402043,
+      "learning_rate": 1.5034163168822797e-05,
+      "loss": 1.8937,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7043819332483096,
+      "grad_norm": 3.1614976300585127,
+      "learning_rate": 1.5031468924330015e-05,
+      "loss": 1.9508,
+      "step": 3659
+    },
+    {
+      "epoch": 0.704574439925885,
+      "grad_norm": 2.8529734534221487,
+      "learning_rate": 1.5028774190714082e-05,
+      "loss": 1.9771,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7047669466034603,
+      "grad_norm": 2.910387990854519,
+      "learning_rate": 1.5026078968236962e-05,
+      "loss": 1.9205,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7049594532810357,
+      "grad_norm": 2.8678511724490803,
+      "learning_rate": 1.5023383257160662e-05,
+      "loss": 1.9159,
+      "step": 3662
+    },
+    {
+      "epoch": 0.705151959958611,
+      "grad_norm": 2.9130142694048184,
+      "learning_rate": 1.5020687057747241e-05,
+      "loss": 1.9766,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "grad_norm": 3.0179284921192084,
+      "learning_rate": 1.5017990370258806e-05,
+      "loss": 1.8681,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.7657,
+      "step": 3664,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.5371,
+      "step": 3664,
+      "vm_loss": 0.1272
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.8717,
+      "step": 3664,
+      "vm_loss": 0.2319
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.9186,
+      "step": 3664,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.7543,
+      "step": 3664,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 2.0273,
+      "step": 3664,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.8591,
+      "step": 3664,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.7053444666361864,
+      "lm_loss": 1.3255,
+      "step": 3664,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.7055369733137619,
+      "grad_norm": 2.953865498704835,
+      "learning_rate": 1.501529319495751e-05,
+      "loss": 1.9549,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7057294799913372,
+      "grad_norm": 2.9171848906970625,
+      "learning_rate": 1.5012595532105547e-05,
+      "loss": 1.9099,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7059219866689126,
+      "grad_norm": 2.933804306347362,
+      "learning_rate": 1.5009897381965171e-05,
+      "loss": 1.9525,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7061144933464879,
+      "grad_norm": 2.732231035676331,
+      "learning_rate": 1.5007198744798673e-05,
+      "loss": 1.8986,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7063070000240633,
+      "grad_norm": 2.9131324974146975,
+      "learning_rate": 1.5004499620868395e-05,
+      "loss": 2.0361,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7064995067016387,
+      "grad_norm": 3.2010181370604514,
+      "learning_rate": 1.5001800010436731e-05,
+      "loss": 2.0474,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7066920133792141,
+      "grad_norm": 2.800951628496817,
+      "learning_rate": 1.4999099913766113e-05,
+      "loss": 1.9389,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "grad_norm": 2.9597387330200924,
+      "learning_rate": 1.4996399331119026e-05,
+      "loss": 1.9196,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.9407,
+      "step": 3672,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.8135,
+      "step": 3672,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.2727,
+      "step": 3672,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.646,
+      "step": 3672,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.6671,
+      "step": 3672,
+      "vm_loss": 0.2486
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 2.1241,
+      "step": 3672,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 1.658,
+      "step": 3672,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 0.7068845200567895,
+      "lm_loss": 2.0151,
+      "step": 3672,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.7070770267343649,
+      "grad_norm": 2.9422197321951202,
+      "learning_rate": 1.4993698262758e-05,
+      "loss": 1.8897,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7072695334119402,
+      "grad_norm": 2.982046999938329,
+      "learning_rate": 1.4990996708945618e-05,
+      "loss": 1.9786,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7074620400895156,
+      "grad_norm": 2.993585891994863,
+      "learning_rate": 1.4988294669944501e-05,
+      "loss": 1.9215,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7076545467670909,
+      "grad_norm": 3.4963731692640985,
+      "learning_rate": 1.4985592146017322e-05,
+      "loss": 2.0173,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7078470534446664,
+      "grad_norm": 2.712323519776265,
+      "learning_rate": 1.4982889137426803e-05,
+      "loss": 1.9098,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7080395601222418,
+      "grad_norm": 2.7697458932977037,
+      "learning_rate": 1.4980185644435712e-05,
+      "loss": 2.0103,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7082320667998171,
+      "grad_norm": 3.3143028085309223,
+      "learning_rate": 1.4977481667306857e-05,
+      "loss": 1.9511,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "grad_norm": 3.1932280932825714,
+      "learning_rate": 1.4974777206303106e-05,
+      "loss": 1.8988,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 1.2301,
+      "step": 3680,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 1.8124,
+      "step": 3680,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 1.8047,
+      "step": 3680,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 2.0136,
+      "step": 3680,
+      "vm_loss": 0.2167
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 1.9321,
+      "step": 3680,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 1.8036,
+      "step": 3680,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 2.2365,
+      "step": 3680,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 0.7084245734773925,
+      "lm_loss": 2.0735,
+      "step": 3680,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.7086170801549678,
+      "grad_norm": 2.892149321167442,
+      "learning_rate": 1.4972072261687365e-05,
+      "loss": 1.8933,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7088095868325432,
+      "grad_norm": 2.9787878894004103,
+      "learning_rate": 1.4969366833722587e-05,
+      "loss": 1.959,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7090020935101187,
+      "grad_norm": 2.8597305354027767,
+      "learning_rate": 1.4966660922671779e-05,
+      "loss": 1.8298,
+      "step": 3683
+    },
+    {
+      "epoch": 0.709194600187694,
+      "grad_norm": 3.0913370448707815,
+      "learning_rate": 1.4963954528797983e-05,
+      "loss": 1.9603,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7093871068652694,
+      "grad_norm": 2.6438987003671577,
+      "learning_rate": 1.4961247652364303e-05,
+      "loss": 1.9356,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7095796135428448,
+      "grad_norm": 3.056020359556225,
+      "learning_rate": 1.4958540293633875e-05,
+      "loss": 1.9182,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7097721202204201,
+      "grad_norm": 2.8993099628532093,
+      "learning_rate": 1.4955832452869893e-05,
+      "loss": 1.9854,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "grad_norm": 2.8790834945258923,
+      "learning_rate": 1.4953124130335594e-05,
+      "loss": 1.8998,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.5317,
+      "step": 3688,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.9062,
+      "step": 3688,
+      "vm_loss": 0.2458
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.8765,
+      "step": 3688,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.9494,
+      "step": 3688,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 2.3855,
+      "step": 3688,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 2.3481,
+      "step": 3688,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.7073,
+      "step": 3688,
+      "vm_loss": 0.2296
+    },
+    {
+      "epoch": 0.7099646268979956,
+      "lm_loss": 1.6671,
+      "step": 3688,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 0.7101571335755709,
+      "grad_norm": 2.8292398192039285,
+      "learning_rate": 1.495041532629426e-05,
+      "loss": 1.954,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7103496402531463,
+      "grad_norm": 3.0200014870401515,
+      "learning_rate": 1.4947706041009223e-05,
+      "loss": 1.8534,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7105421469307217,
+      "grad_norm": 3.1136084865061946,
+      "learning_rate": 1.4944996274743857e-05,
+      "loss": 1.8933,
+      "step": 3691
+    },
+    {
+      "epoch": 0.710734653608297,
+      "grad_norm": 2.9356339267602305,
+      "learning_rate": 1.494228602776159e-05,
+      "loss": 1.9205,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7109271602858724,
+      "grad_norm": 3.172558410462714,
+      "learning_rate": 1.493957530032589e-05,
+      "loss": 1.944,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7111196669634477,
+      "grad_norm": 3.0320596542526688,
+      "learning_rate": 1.4936864092700278e-05,
+      "loss": 1.9274,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7113121736410232,
+      "grad_norm": 2.9182609554141945,
+      "learning_rate": 1.4934152405148312e-05,
+      "loss": 1.9629,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "grad_norm": 3.2744116488572588,
+      "learning_rate": 1.4931440237933605e-05,
+      "loss": 1.9073,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.8795,
+      "step": 3696,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.7848,
+      "step": 3696,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 2.033,
+      "step": 3696,
+      "vm_loss": 0.1511
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.95,
+      "step": 3696,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.7633,
+      "step": 3696,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.8678,
+      "step": 3696,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.9616,
+      "step": 3696,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 0.7115046803185986,
+      "lm_loss": 1.4825,
+      "step": 3696,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 0.7116971869961739,
+      "grad_norm": 2.8536925931608743,
+      "learning_rate": 1.492872759131982e-05,
+      "loss": 1.9543,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7118896936737493,
+      "grad_norm": 3.0120423071835685,
+      "learning_rate": 1.4926014465570652e-05,
+      "loss": 1.9256,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7120822003513246,
+      "grad_norm": 2.8958635643689137,
+      "learning_rate": 1.492330086094986e-05,
+      "loss": 1.9489,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7122747070289001,
+      "grad_norm": 2.8932758814879502,
+      "learning_rate": 1.4920586777721231e-05,
+      "loss": 1.973,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7124672137064755,
+      "grad_norm": 3.2130394399389415,
+      "learning_rate": 1.491787221614862e-05,
+      "loss": 1.9698,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7126597203840508,
+      "grad_norm": 2.9738723358837187,
+      "learning_rate": 1.491515717649591e-05,
+      "loss": 1.9653,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7128522270616262,
+      "grad_norm": 2.769676264659778,
+      "learning_rate": 1.4912441659027037e-05,
+      "loss": 1.8621,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "grad_norm": 2.816529435197119,
+      "learning_rate": 1.4909725664005987e-05,
+      "loss": 1.8816,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.9877,
+      "step": 3704,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.5424,
+      "step": 3704,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.5501,
+      "step": 3704,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.6442,
+      "step": 3704,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.0074,
+      "step": 3704,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.7407,
+      "step": 3704,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.5802,
+      "step": 3704,
+      "vm_loss": 0.1058
+    },
+    {
+      "epoch": 0.7130447337392016,
+      "lm_loss": 1.6994,
+      "step": 3704,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.7132372404167769,
+      "grad_norm": 3.0258604172335963,
+      "learning_rate": 1.4907009191696791e-05,
+      "loss": 1.8583,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7134297470943524,
+      "grad_norm": 3.120601881167479,
+      "learning_rate": 1.4904292242363521e-05,
+      "loss": 1.9304,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7136222537719277,
+      "grad_norm": 3.0398261737868744,
+      "learning_rate": 1.4901574816270302e-05,
+      "loss": 1.933,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7138147604495031,
+      "grad_norm": 2.911962354152422,
+      "learning_rate": 1.48988569136813e-05,
+      "loss": 1.9485,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7140072671270785,
+      "grad_norm": 3.2596608325013667,
+      "learning_rate": 1.4896138534860728e-05,
+      "loss": 1.9451,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7141997738046538,
+      "grad_norm": 2.996590040777344,
+      "learning_rate": 1.4893419680072856e-05,
+      "loss": 1.964,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7143922804822292,
+      "grad_norm": 2.864444338506274,
+      "learning_rate": 1.4890700349581986e-05,
+      "loss": 1.9124,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "grad_norm": 3.1503096212407846,
+      "learning_rate": 1.4887980543652467e-05,
+      "loss": 1.883,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.647,
+      "step": 3712,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.8455,
+      "step": 3712,
+      "vm_loss": 0.2255
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.4691,
+      "step": 3712,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.759,
+      "step": 3712,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.8049,
+      "step": 3712,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.5675,
+      "step": 3712,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.955,
+      "step": 3712,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.7145847871598046,
+      "lm_loss": 1.9605,
+      "step": 3712,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.71477729383738,
+      "grad_norm": 3.121243343215388,
+      "learning_rate": 1.4885260262548706e-05,
+      "loss": 1.9278,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7149698005149554,
+      "grad_norm": 2.9712268915660704,
+      "learning_rate": 1.488253950653515e-05,
+      "loss": 2.0209,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7151623071925307,
+      "grad_norm": 2.973055936590317,
+      "learning_rate": 1.4879818275876283e-05,
+      "loss": 1.922,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7153548138701061,
+      "grad_norm": 2.832989712910767,
+      "learning_rate": 1.4877096570836651e-05,
+      "loss": 1.9317,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7155473205476814,
+      "grad_norm": 3.028559499643474,
+      "learning_rate": 1.4874374391680837e-05,
+      "loss": 1.9111,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7157398272252569,
+      "grad_norm": 3.115262017561496,
+      "learning_rate": 1.4871651738673469e-05,
+      "loss": 2.0183,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7159323339028323,
+      "grad_norm": 3.1181027224898594,
+      "learning_rate": 1.4868928612079226e-05,
+      "loss": 1.8976,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "grad_norm": 3.1881361420256735,
+      "learning_rate": 1.486620501216283e-05,
+      "loss": 1.9471,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.5344,
+      "step": 3720,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.9682,
+      "step": 3720,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 2.2396,
+      "step": 3720,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.8707,
+      "step": 3720,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.6082,
+      "step": 3720,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.3282,
+      "step": 3720,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.8681,
+      "step": 3720,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 0.7161248405804076,
+      "lm_loss": 1.8214,
+      "step": 3720,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 0.716317347257983,
+      "grad_norm": 2.963801681292483,
+      "learning_rate": 1.486348093918905e-05,
+      "loss": 1.935,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7165098539355584,
+      "grad_norm": 2.7421865251275,
+      "learning_rate": 1.4860756393422699e-05,
+      "loss": 1.8775,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7167023606131337,
+      "grad_norm": 2.7773370709149146,
+      "learning_rate": 1.4858031375128643e-05,
+      "loss": 1.8892,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7168948672907092,
+      "grad_norm": 3.0049953609481146,
+      "learning_rate": 1.4855305884571782e-05,
+      "loss": 1.8382,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7170873739682845,
+      "grad_norm": 3.0525509380576734,
+      "learning_rate": 1.4852579922017072e-05,
+      "loss": 1.8724,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7172798806458599,
+      "grad_norm": 3.2518079410622778,
+      "learning_rate": 1.4849853487729512e-05,
+      "loss": 1.903,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7174723873234353,
+      "grad_norm": 3.002528005269556,
+      "learning_rate": 1.4847126581974146e-05,
+      "loss": 1.907,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "grad_norm": 2.870813502564371,
+      "learning_rate": 1.4844399205016062e-05,
+      "loss": 1.9513,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.9327,
+      "step": 3728,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.5341,
+      "step": 3728,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.844,
+      "step": 3728,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.7995,
+      "step": 3728,
+      "vm_loss": 0.1147
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.8214,
+      "step": 3728,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.8343,
+      "step": 3728,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 2.0218,
+      "step": 3728,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.7176648940010106,
+      "lm_loss": 1.5754,
+      "step": 3728,
+      "vm_loss": 0.1213
+    },
+    {
+      "epoch": 0.717857400678586,
+      "grad_norm": 2.889644092271477,
+      "learning_rate": 1.4841671357120395e-05,
+      "loss": 1.8583,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7180499073561614,
+      "grad_norm": 2.909230833785038,
+      "learning_rate": 1.483894303855233e-05,
+      "loss": 1.935,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7182424140337368,
+      "grad_norm": 3.035777813753681,
+      "learning_rate": 1.4836214249577094e-05,
+      "loss": 1.9072,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7184349207113122,
+      "grad_norm": 2.9401793345334917,
+      "learning_rate": 1.4833484990459963e-05,
+      "loss": 1.9445,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7186274273888875,
+      "grad_norm": 3.157828467371138,
+      "learning_rate": 1.4830755261466249e-05,
+      "loss": 1.934,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7188199340664629,
+      "grad_norm": 3.342609333773877,
+      "learning_rate": 1.482802506286132e-05,
+      "loss": 1.907,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7190124407440384,
+      "grad_norm": 3.2489124115875008,
+      "learning_rate": 1.482529439491059e-05,
+      "loss": 1.9053,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "grad_norm": 2.8670753083817777,
+      "learning_rate": 1.4822563257879508e-05,
+      "loss": 1.8917,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 1.6906,
+      "step": 3736,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 1.597,
+      "step": 3736,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 2.1388,
+      "step": 3736,
+      "vm_loss": 0.1934
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 1.6235,
+      "step": 3736,
+      "vm_loss": 0.2226
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 1.9806,
+      "step": 3736,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 2.1045,
+      "step": 3736,
+      "vm_loss": 0.2493
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 2.2695,
+      "step": 3736,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 0.7192049474216137,
+      "lm_loss": 1.6723,
+      "step": 3736,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.7193974540991891,
+      "grad_norm": 3.1072404818057557,
+      "learning_rate": 1.4819831652033579e-05,
+      "loss": 1.9802,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7195899607767644,
+      "grad_norm": 3.049431558234614,
+      "learning_rate": 1.481709957763835e-05,
+      "loss": 1.9003,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7197824674543398,
+      "grad_norm": 3.08415414804755,
+      "learning_rate": 1.4814367034959414e-05,
+      "loss": 1.894,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7199749741319152,
+      "grad_norm": 3.1817186645565174,
+      "learning_rate": 1.4811634024262409e-05,
+      "loss": 1.9126,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7201674808094906,
+      "grad_norm": 3.229228692319169,
+      "learning_rate": 1.4808900545813016e-05,
+      "loss": 1.9417,
+      "step": 3741
+    },
+    {
+      "epoch": 0.720359987487066,
+      "grad_norm": 3.0200083591959395,
+      "learning_rate": 1.4806166599876967e-05,
+      "loss": 1.885,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7205524941646413,
+      "grad_norm": 3.152478698419477,
+      "learning_rate": 1.4803432186720037e-05,
+      "loss": 1.8664,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "grad_norm": 2.928741057067477,
+      "learning_rate": 1.4800697306608043e-05,
+      "loss": 1.896,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.1524,
+      "step": 3744,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.4933,
+      "step": 3744,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.7221,
+      "step": 3744,
+      "vm_loss": 0.2314
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.7901,
+      "step": 3744,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.9902,
+      "step": 3744,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.7596,
+      "step": 3744,
+      "vm_loss": 0.1246
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 1.555,
+      "step": 3744,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 0.7207450008422167,
+      "lm_loss": 2.3212,
+      "step": 3744,
+      "vm_loss": 0.2247
+    },
+    {
+      "epoch": 0.7209375075197921,
+      "grad_norm": 3.023002099864317,
+      "learning_rate": 1.4797961959806854e-05,
+      "loss": 1.885,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7211300141973674,
+      "grad_norm": 3.2507530108483493,
+      "learning_rate": 1.4795226146582378e-05,
+      "loss": 1.8696,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7213225208749429,
+      "grad_norm": 2.9215267991045475,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 1.9525,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7215150275525182,
+      "grad_norm": 2.9915521516980466,
+      "learning_rate": 1.4789753121927435e-05,
+      "loss": 1.8833,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7217075342300936,
+      "grad_norm": 2.9456974182390123,
+      "learning_rate": 1.4787015911029014e-05,
+      "loss": 1.9252,
+      "step": 3749
+    },
+    {
+      "epoch": 0.721900040907669,
+      "grad_norm": 3.1366470952785486,
+      "learning_rate": 1.4784278234771407e-05,
+      "loss": 1.9243,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7220925475852443,
+      "grad_norm": 3.289474461723117,
+      "learning_rate": 1.4781540093420746e-05,
+      "loss": 1.9456,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "grad_norm": 2.7924391041033676,
+      "learning_rate": 1.4778801487243214e-05,
+      "loss": 1.9359,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 2.18,
+      "step": 3752,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.9218,
+      "step": 3752,
+      "vm_loss": 0.2452
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.5757,
+      "step": 3752,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.4875,
+      "step": 3752,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.2811,
+      "step": 3752,
+      "vm_loss": 0.1184
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.8495,
+      "step": 3752,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.9507,
+      "step": 3752,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.7222850542628197,
+      "lm_loss": 1.4793,
+      "step": 3752,
+      "vm_loss": 0.2049
+    },
+    {
+      "epoch": 0.7224775609403952,
+      "grad_norm": 3.0610919750334586,
+      "learning_rate": 1.4776062416505035e-05,
+      "loss": 1.8408,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7226700676179705,
+      "grad_norm": 2.8901879349015247,
+      "learning_rate": 1.477332288147249e-05,
+      "loss": 1.9165,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7228625742955459,
+      "grad_norm": 2.911078905028071,
+      "learning_rate": 1.477058288241189e-05,
+      "loss": 1.8403,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7230550809731212,
+      "grad_norm": 3.1773776189665774,
+      "learning_rate": 1.47678424195896e-05,
+      "loss": 1.8859,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7232475876506966,
+      "grad_norm": 3.322724287841504,
+      "learning_rate": 1.4765101493272032e-05,
+      "loss": 1.9475,
+      "step": 3757
+    },
+    {
+      "epoch": 0.723440094328272,
+      "grad_norm": 3.065576238962563,
+      "learning_rate": 1.476236010372563e-05,
+      "loss": 1.944,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7236326010058474,
+      "grad_norm": 3.1019181158658453,
+      "learning_rate": 1.4759618251216903e-05,
+      "loss": 1.8806,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "grad_norm": 2.9596534839491175,
+      "learning_rate": 1.4756875936012385e-05,
+      "loss": 1.9602,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.5672,
+      "step": 3760,
+      "vm_loss": 0.2213
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.4406,
+      "step": 3760,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.8501,
+      "step": 3760,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.8334,
+      "step": 3760,
+      "vm_loss": 0.2294
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.7821,
+      "step": 3760,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 1.9379,
+      "step": 3760,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 2.3439,
+      "step": 3760,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 0.7238251076834228,
+      "lm_loss": 2.0827,
+      "step": 3760,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 0.7240176143609981,
+      "grad_norm": 2.8078062475850794,
+      "learning_rate": 1.4754133158378668e-05,
+      "loss": 1.9215,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7242101210385735,
+      "grad_norm": 2.8732470476455836,
+      "learning_rate": 1.4751389918582387e-05,
+      "loss": 1.966,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7244026277161489,
+      "grad_norm": 2.9614104763428,
+      "learning_rate": 1.4748646216890215e-05,
+      "loss": 1.9411,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7245951343937242,
+      "grad_norm": 3.131225361159734,
+      "learning_rate": 1.4745902053568882e-05,
+      "loss": 1.9054,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7247876410712997,
+      "grad_norm": 2.91588752534843,
+      "learning_rate": 1.4743157428885147e-05,
+      "loss": 1.9081,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7249801477488751,
+      "grad_norm": 2.9327040501890687,
+      "learning_rate": 1.4740412343105828e-05,
+      "loss": 1.8678,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7251726544264504,
+      "grad_norm": 2.74747089193638,
+      "learning_rate": 1.4737666796497784e-05,
+      "loss": 1.8966,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "grad_norm": 2.855983359239089,
+      "learning_rate": 1.4734920789327913e-05,
+      "loss": 1.8764,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.6491,
+      "step": 3768,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.6772,
+      "step": 3768,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.3075,
+      "step": 3768,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.464,
+      "step": 3768,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.6724,
+      "step": 3768,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.9063,
+      "step": 3768,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 1.744,
+      "step": 3768,
+      "vm_loss": 0.2026
+    },
+    {
+      "epoch": 0.7253651611040258,
+      "lm_loss": 2.2858,
+      "step": 3768,
+      "vm_loss": 0.2235
+    },
+    {
+      "epoch": 0.7255576677816011,
+      "grad_norm": 2.994524235639348,
+      "learning_rate": 1.4732174321863167e-05,
+      "loss": 1.9165,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7257501744591766,
+      "grad_norm": 2.8683545229264475,
+      "learning_rate": 1.4729427394370533e-05,
+      "loss": 1.8537,
+      "step": 3770
+    },
+    {
+      "epoch": 0.725942681136752,
+      "grad_norm": 2.865619534192569,
+      "learning_rate": 1.4726680007117046e-05,
+      "loss": 1.8509,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7261351878143273,
+      "grad_norm": 2.8866391106728604,
+      "learning_rate": 1.4723932160369794e-05,
+      "loss": 1.9106,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7263276944919027,
+      "grad_norm": 3.0757442050545394,
+      "learning_rate": 1.4721183854395897e-05,
+      "loss": 1.947,
+      "step": 3773
+    },
+    {
+      "epoch": 0.726520201169478,
+      "grad_norm": 2.9798541110704146,
+      "learning_rate": 1.4718435089462528e-05,
+      "loss": 1.8744,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7267127078470534,
+      "grad_norm": 2.813690311602499,
+      "learning_rate": 1.4715685865836904e-05,
+      "loss": 1.9043,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "grad_norm": 2.881585400846457,
+      "learning_rate": 1.4712936183786281e-05,
+      "loss": 1.9092,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 2.0198,
+      "step": 3776,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.7665,
+      "step": 3776,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.6736,
+      "step": 3776,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 2.1747,
+      "step": 3776,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.9061,
+      "step": 3776,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.6564,
+      "step": 3776,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.1195,
+      "step": 3776,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.7269052145246289,
+      "lm_loss": 1.967,
+      "step": 3776,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 0.7270977212022042,
+      "grad_norm": 2.8651637905957283,
+      "learning_rate": 1.4710186043577965e-05,
+      "loss": 1.9225,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7272902278797796,
+      "grad_norm": 2.823024656375641,
+      "learning_rate": 1.4707435445479306e-05,
+      "loss": 1.8473,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7274827345573549,
+      "grad_norm": 2.9268825113603874,
+      "learning_rate": 1.4704684389757696e-05,
+      "loss": 1.9245,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7276752412349303,
+      "grad_norm": 2.929838647278052,
+      "learning_rate": 1.470193287668057e-05,
+      "loss": 1.871,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7278677479125057,
+      "grad_norm": 3.044792363818208,
+      "learning_rate": 1.4699180906515418e-05,
+      "loss": 1.8977,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7280602545900811,
+      "grad_norm": 2.8703812272871043,
+      "learning_rate": 1.4696428479529758e-05,
+      "loss": 1.8681,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7282527612676565,
+      "grad_norm": 2.984227209164664,
+      "learning_rate": 1.4693675595991163e-05,
+      "loss": 1.9523,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "grad_norm": 2.9747978746757924,
+      "learning_rate": 1.4690922256167256e-05,
+      "loss": 1.985,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.3529,
+      "step": 3784,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.5735,
+      "step": 3784,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.2311,
+      "step": 3784,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.2852,
+      "step": 3784,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.714,
+      "step": 3784,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.721,
+      "step": 3784,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.3125,
+      "step": 3784,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.7284452679452319,
+      "lm_loss": 1.9266,
+      "step": 3784,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 0.7286377746228072,
+      "grad_norm": 2.8770598699520784,
+      "learning_rate": 1.4688168460325686e-05,
+      "loss": 1.8648,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7288302813003826,
+      "grad_norm": 2.8249853074431477,
+      "learning_rate": 1.4685414208734168e-05,
+      "loss": 1.8871,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7290227879779579,
+      "grad_norm": 2.6915223035627687,
+      "learning_rate": 1.468265950166044e-05,
+      "loss": 1.858,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7292152946555334,
+      "grad_norm": 2.9186050017578653,
+      "learning_rate": 1.4679904339372301e-05,
+      "loss": 1.8947,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7294078013331088,
+      "grad_norm": 3.0135445207936495,
+      "learning_rate": 1.467714872213759e-05,
+      "loss": 1.9631,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7296003080106841,
+      "grad_norm": 2.9510883750314276,
+      "learning_rate": 1.467439265022418e-05,
+      "loss": 1.928,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7297928146882595,
+      "grad_norm": 2.979859919913248,
+      "learning_rate": 1.4671636123900006e-05,
+      "loss": 1.9898,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "grad_norm": 2.8783329180705644,
+      "learning_rate": 1.4668879143433031e-05,
+      "loss": 1.8716,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 2.201,
+      "step": 3792,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.986,
+      "step": 3792,
+      "vm_loss": 0.2028
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.8175,
+      "step": 3792,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.7665,
+      "step": 3792,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.974,
+      "step": 3792,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.44,
+      "step": 3792,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.4445,
+      "step": 3792,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.7299853213658348,
+      "lm_loss": 1.5639,
+      "step": 3792,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.7301778280434102,
+      "grad_norm": 2.897848340764509,
+      "learning_rate": 1.466612170909127e-05,
+      "loss": 1.8876,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7303703347209857,
+      "grad_norm": 2.855486079063728,
+      "learning_rate": 1.4663363821142784e-05,
+      "loss": 1.8899,
+      "step": 3794
+    },
+    {
+      "epoch": 0.730562841398561,
+      "grad_norm": 2.890471014035725,
+      "learning_rate": 1.466060547985567e-05,
+      "loss": 1.8745,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7307553480761364,
+      "grad_norm": 3.120574012481081,
+      "learning_rate": 1.4657846685498079e-05,
+      "loss": 1.9247,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7309478547537118,
+      "grad_norm": 3.1286212358505394,
+      "learning_rate": 1.4655087438338196e-05,
+      "loss": 1.9042,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7311403614312871,
+      "grad_norm": 3.032055545764561,
+      "learning_rate": 1.465232773864426e-05,
+      "loss": 1.8474,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7313328681088626,
+      "grad_norm": 2.961280707772565,
+      "learning_rate": 1.4649567586684548e-05,
+      "loss": 1.9814,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "grad_norm": 2.825872236590409,
+      "learning_rate": 1.464680698272738e-05,
+      "loss": 1.8708,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.5248,
+      "step": 3800,
+      "vm_loss": 0.1962
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 2.0103,
+      "step": 3800,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.3539,
+      "step": 3800,
+      "vm_loss": 0.0982
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.7909,
+      "step": 3800,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.8984,
+      "step": 3800,
+      "vm_loss": 0.2291
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.7355,
+      "step": 3800,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.7069,
+      "step": 3800,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.7315253747864379,
+      "lm_loss": 1.6806,
+      "step": 3800,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 0.7317178814640133,
+      "grad_norm": 3.210465647848649,
+      "learning_rate": 1.4644045927041124e-05,
+      "loss": 1.8633,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7319103881415887,
+      "grad_norm": 3.1170651130741187,
+      "learning_rate": 1.464128441989419e-05,
+      "loss": 1.859,
+      "step": 3802
+    },
+    {
+      "epoch": 0.732102894819164,
+      "grad_norm": 2.8536383287067286,
+      "learning_rate": 1.463852246155503e-05,
+      "loss": 1.8636,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7322954014967394,
+      "grad_norm": 3.0796702239268665,
+      "learning_rate": 1.463576005229214e-05,
+      "loss": 1.8913,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7324879081743148,
+      "grad_norm": 3.3263877220154847,
+      "learning_rate": 1.463299719237407e-05,
+      "loss": 1.8516,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7326804148518902,
+      "grad_norm": 3.375887425072847,
+      "learning_rate": 1.4630233882069396e-05,
+      "loss": 1.859,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7328729215294656,
+      "grad_norm": 2.9099849025767135,
+      "learning_rate": 1.462747012164675e-05,
+      "loss": 1.8462,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "grad_norm": 2.8266158078709815,
+      "learning_rate": 1.4624705911374808e-05,
+      "loss": 1.8635,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.4052,
+      "step": 3808,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.5563,
+      "step": 3808,
+      "vm_loss": 0.24
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.9506,
+      "step": 3808,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 2.1204,
+      "step": 3808,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.6784,
+      "step": 3808,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.585,
+      "step": 3808,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.5776,
+      "step": 3808,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.7330654282070409,
+      "lm_loss": 1.858,
+      "step": 3808,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 0.7332579348846163,
+      "grad_norm": 3.1287871877030957,
+      "learning_rate": 1.462194125152228e-05,
+      "loss": 1.9462,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7334504415621916,
+      "grad_norm": 2.783954028897125,
+      "learning_rate": 1.4619176142357936e-05,
+      "loss": 1.9007,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7336429482397671,
+      "grad_norm": 3.04246002898484,
+      "learning_rate": 1.4616410584150572e-05,
+      "loss": 1.9307,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7338354549173425,
+      "grad_norm": 2.8826782971176583,
+      "learning_rate": 1.4613644577169035e-05,
+      "loss": 1.8615,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7340279615949178,
+      "grad_norm": 2.910438167504957,
+      "learning_rate": 1.4610878121682223e-05,
+      "loss": 1.8768,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7342204682724932,
+      "grad_norm": 2.898400065854732,
+      "learning_rate": 1.4608111217959067e-05,
+      "loss": 1.8516,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7344129749500686,
+      "grad_norm": 3.0642227593692035,
+      "learning_rate": 1.4605343866268544e-05,
+      "loss": 1.9445,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "grad_norm": 2.850021427137016,
+      "learning_rate": 1.460257606687968e-05,
+      "loss": 1.8347,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.6676,
+      "step": 3816,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.7034,
+      "step": 3816,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.8896,
+      "step": 3816,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.8827,
+      "step": 3816,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.7872,
+      "step": 3816,
+      "vm_loss": 0.1964
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.9646,
+      "step": 3816,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.5278,
+      "step": 3816,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.7346054816276439,
+      "lm_loss": 1.2605,
+      "step": 3816,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 0.7347979883052194,
+      "grad_norm": 2.9055510497597026,
+      "learning_rate": 1.4599807820061536e-05,
+      "loss": 1.8787,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7349904949827947,
+      "grad_norm": 2.971811352824652,
+      "learning_rate": 1.4597039126083224e-05,
+      "loss": 1.8528,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7351830016603701,
+      "grad_norm": 2.9539883288609943,
+      "learning_rate": 1.4594269985213893e-05,
+      "loss": 1.8676,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7353755083379455,
+      "grad_norm": 2.9868512124543294,
+      "learning_rate": 1.4591500397722746e-05,
+      "loss": 1.8643,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7355680150155208,
+      "grad_norm": 3.101406742443786,
+      "learning_rate": 1.4588730363879015e-05,
+      "loss": 1.9055,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7357605216930962,
+      "grad_norm": 2.9113109890318625,
+      "learning_rate": 1.458595988395199e-05,
+      "loss": 1.9581,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7359530283706716,
+      "grad_norm": 2.8932218278534862,
+      "learning_rate": 1.4583188958210987e-05,
+      "loss": 1.9013,
+      "step": 3823
+    },
+    {
+      "epoch": 0.736145535048247,
+      "grad_norm": 2.8870778768085663,
+      "learning_rate": 1.4580417586925384e-05,
+      "loss": 1.9068,
+      "step": 3824
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.8975,
+      "step": 3824,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.419,
+      "step": 3824,
+      "vm_loss": 0.2197
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.5307,
+      "step": 3824,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 2.175,
+      "step": 3824,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 2.3087,
+      "step": 3824,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.5409,
+      "step": 3824,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.7161,
+      "step": 3824,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.736145535048247,
+      "lm_loss": 1.7306,
+      "step": 3824,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 0.7363380417258224,
+      "grad_norm": 2.8516293532175463,
+      "learning_rate": 1.457764577036459e-05,
+      "loss": 1.9099,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7365305484033977,
+      "grad_norm": 3.1458646036580364,
+      "learning_rate": 1.4574873508798063e-05,
+      "loss": 1.8971,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7367230550809731,
+      "grad_norm": 2.9776755579716236,
+      "learning_rate": 1.4572100802495302e-05,
+      "loss": 1.8959,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7369155617585486,
+      "grad_norm": 2.9628437232221123,
+      "learning_rate": 1.4569327651725847e-05,
+      "loss": 1.8474,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7371080684361239,
+      "grad_norm": 2.910023063689976,
+      "learning_rate": 1.4566554056759286e-05,
+      "loss": 1.8876,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7373005751136993,
+      "grad_norm": 2.846094873646071,
+      "learning_rate": 1.4563780017865248e-05,
+      "loss": 1.8749,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7374930817912746,
+      "grad_norm": 2.942425683678951,
+      "learning_rate": 1.4561005535313402e-05,
+      "loss": 1.88,
+      "step": 3831
+    },
+    {
+      "epoch": 0.73768558846885,
+      "grad_norm": 2.8362740868832317,
+      "learning_rate": 1.4558230609373469e-05,
+      "loss": 1.9019,
+      "step": 3832
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.9086,
+      "step": 3832,
+      "vm_loss": 0.2219
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.8401,
+      "step": 3832,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.2908,
+      "step": 3832,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.7236,
+      "step": 3832,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.5062,
+      "step": 3832,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 2.0799,
+      "step": 3832,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 2.094,
+      "step": 3832,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 0.73768558846885,
+      "lm_loss": 1.2925,
+      "step": 3832,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 0.7378780951464254,
+      "grad_norm": 3.1971256387533082,
+      "learning_rate": 1.4555455240315202e-05,
+      "loss": 1.8861,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7380706018240007,
+      "grad_norm": 2.881995289066406,
+      "learning_rate": 1.4552679428408402e-05,
+      "loss": 1.8704,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7382631085015762,
+      "grad_norm": 2.9762213116119525,
+      "learning_rate": 1.4549903173922921e-05,
+      "loss": 1.9293,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7384556151791515,
+      "grad_norm": 2.9306629177899377,
+      "learning_rate": 1.4547126477128637e-05,
+      "loss": 1.8696,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7386481218567269,
+      "grad_norm": 3.077156437479805,
+      "learning_rate": 1.4544349338295487e-05,
+      "loss": 1.8942,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7388406285343023,
+      "grad_norm": 3.164571209751457,
+      "learning_rate": 1.454157175769344e-05,
+      "loss": 1.922,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7390331352118776,
+      "grad_norm": 2.8618423982224286,
+      "learning_rate": 1.4538793735592518e-05,
+      "loss": 1.8946,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "grad_norm": 3.0244413345367693,
+      "learning_rate": 1.4536015272262773e-05,
+      "loss": 1.9509,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 2.3827,
+      "step": 3840,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.9091,
+      "step": 3840,
+      "vm_loss": 0.2034
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.7933,
+      "step": 3840,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.955,
+      "step": 3840,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.9445,
+      "step": 3840,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 2.1779,
+      "step": 3840,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.8821,
+      "step": 3840,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.7392256418894531,
+      "lm_loss": 1.8644,
+      "step": 3840,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.7394181485670284,
+      "grad_norm": 3.139005253671557,
+      "learning_rate": 1.4533236367974313e-05,
+      "loss": 1.8266,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7396106552446038,
+      "grad_norm": 3.0658324622634305,
+      "learning_rate": 1.4530457022997282e-05,
+      "loss": 1.8826,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7398031619221792,
+      "grad_norm": 2.971648309669812,
+      "learning_rate": 1.4527677237601865e-05,
+      "loss": 1.8877,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7399956685997545,
+      "grad_norm": 3.111543085055416,
+      "learning_rate": 1.4524897012058297e-05,
+      "loss": 1.8498,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7401881752773299,
+      "grad_norm": 2.95760332318267,
+      "learning_rate": 1.4522116346636846e-05,
+      "loss": 1.8323,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7403806819549054,
+      "grad_norm": 2.8049768098510253,
+      "learning_rate": 1.4519335241607833e-05,
+      "loss": 1.8212,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7405731886324807,
+      "grad_norm": 2.8587699703344644,
+      "learning_rate": 1.4516553697241615e-05,
+      "loss": 1.87,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "grad_norm": 2.978140868099243,
+      "learning_rate": 1.4513771713808594e-05,
+      "loss": 1.9335,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.6118,
+      "step": 3848,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.3833,
+      "step": 3848,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.0515,
+      "step": 3848,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.2491,
+      "step": 3848,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.8356,
+      "step": 3848,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 2.0404,
+      "step": 3848,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.4947,
+      "step": 3848,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 0.7407656953100561,
+      "lm_loss": 1.6058,
+      "step": 3848,
+      "vm_loss": 0.1189
+    },
+    {
+      "epoch": 0.7409582019876314,
+      "grad_norm": 2.9531610415798357,
+      "learning_rate": 1.4510989291579214e-05,
+      "loss": 1.883,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7411507086652068,
+      "grad_norm": 3.157199596048571,
+      "learning_rate": 1.4508206430823963e-05,
+      "loss": 1.9209,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7413432153427822,
+      "grad_norm": 3.053226852390783,
+      "learning_rate": 1.4505423131813371e-05,
+      "loss": 1.8842,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7415357220203576,
+      "grad_norm": 3.2702361176148953,
+      "learning_rate": 1.4502639394818007e-05,
+      "loss": 1.8606,
+      "step": 3852
+    },
+    {
+      "epoch": 0.741728228697933,
+      "grad_norm": 3.104752763861082,
+      "learning_rate": 1.449985522010849e-05,
+      "loss": 1.8445,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7419207353755083,
+      "grad_norm": 2.9520776078384436,
+      "learning_rate": 1.4497070607955477e-05,
+      "loss": 1.9598,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7421132420530837,
+      "grad_norm": 3.0018615102635247,
+      "learning_rate": 1.4494285558629664e-05,
+      "loss": 1.8486,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "grad_norm": 3.091778112936236,
+      "learning_rate": 1.4491500072401796e-05,
+      "loss": 1.9188,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 2.0952,
+      "step": 3856,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.563,
+      "step": 3856,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.8727,
+      "step": 3856,
+      "vm_loss": 0.2032
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.4588,
+      "step": 3856,
+      "vm_loss": 0.1934
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.7593,
+      "step": 3856,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.4849,
+      "step": 3856,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 0.7698,
+      "step": 3856,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 0.7423057487306591,
+      "lm_loss": 1.7999,
+      "step": 3856,
+      "vm_loss": 0.2298
+    },
+    {
+      "epoch": 0.7424982554082344,
+      "grad_norm": 3.2231424573590006,
+      "learning_rate": 1.448871414954266e-05,
+      "loss": 1.8714,
+      "step": 3857
+    },
+    {
+      "epoch": 0.7426907620858099,
+      "grad_norm": 3.0774949500867144,
+      "learning_rate": 1.448592779032308e-05,
+      "loss": 1.9052,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7428832687633853,
+      "grad_norm": 2.8867989946937986,
+      "learning_rate": 1.4483140995013924e-05,
+      "loss": 1.8708,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7430757754409606,
+      "grad_norm": 3.187271409887001,
+      "learning_rate": 1.448035376388611e-05,
+      "loss": 1.8576,
+      "step": 3860
+    },
+    {
+      "epoch": 0.743268282118536,
+      "grad_norm": 3.016348859284071,
+      "learning_rate": 1.4477566097210592e-05,
+      "loss": 1.8046,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7434607887961113,
+      "grad_norm": 2.9555569253122562,
+      "learning_rate": 1.4474777995258363e-05,
+      "loss": 1.8916,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7436532954736867,
+      "grad_norm": 2.8893921198800854,
+      "learning_rate": 1.4471989458300462e-05,
+      "loss": 1.9013,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "grad_norm": 2.8337064941075525,
+      "learning_rate": 1.4469200486607972e-05,
+      "loss": 1.9004,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.9101,
+      "step": 3864,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.2254,
+      "step": 3864,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.6179,
+      "step": 3864,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.4599,
+      "step": 3864,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.709,
+      "step": 3864,
+      "vm_loss": 0.1001
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 2.0548,
+      "step": 3864,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.8772,
+      "step": 3864,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 0.7438458021512622,
+      "lm_loss": 1.6035,
+      "step": 3864,
+      "vm_loss": 0.1123
+    },
+    {
+      "epoch": 0.7440383088288375,
+      "grad_norm": 2.813384024534517,
+      "learning_rate": 1.4466411080452019e-05,
+      "loss": 1.8162,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7442308155064129,
+      "grad_norm": 2.9652358880743184,
+      "learning_rate": 1.4463621240103768e-05,
+      "loss": 1.8161,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7444233221839882,
+      "grad_norm": 2.9258324606987824,
+      "learning_rate": 1.4460830965834425e-05,
+      "loss": 1.8261,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7446158288615636,
+      "grad_norm": 3.2039421090786178,
+      "learning_rate": 1.445804025791524e-05,
+      "loss": 1.8737,
+      "step": 3868
+    },
+    {
+      "epoch": 0.744808335539139,
+      "grad_norm": 2.9797814709557016,
+      "learning_rate": 1.4455249116617506e-05,
+      "loss": 1.8664,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7450008422167144,
+      "grad_norm": 2.849677007744024,
+      "learning_rate": 1.4452457542212561e-05,
+      "loss": 1.8846,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7451933488942898,
+      "grad_norm": 2.9659264675401205,
+      "learning_rate": 1.444966553497178e-05,
+      "loss": 1.9239,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "grad_norm": 2.904541785937833,
+      "learning_rate": 1.4446873095166578e-05,
+      "loss": 1.8107,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.1513,
+      "step": 3872,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.2626,
+      "step": 3872,
+      "vm_loss": 0.2322
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.163,
+      "step": 3872,
+      "vm_loss": 0.232
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.6007,
+      "step": 3872,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.802,
+      "step": 3872,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.5019,
+      "step": 3872,
+      "vm_loss": 0.0985
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.6935,
+      "step": 3872,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 0.7453858555718651,
+      "lm_loss": 1.921,
+      "step": 3872,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 0.7455783622494405,
+      "grad_norm": 2.9990768639811685,
+      "learning_rate": 1.4444080223068424e-05,
+      "loss": 1.8768,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7457708689270159,
+      "grad_norm": 3.1413771295448405,
+      "learning_rate": 1.4441286918948811e-05,
+      "loss": 1.8594,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7459633756045913,
+      "grad_norm": 3.0373578971492057,
+      "learning_rate": 1.443849318307929e-05,
+      "loss": 1.8458,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7461558822821667,
+      "grad_norm": 2.944574147954027,
+      "learning_rate": 1.4435699015731449e-05,
+      "loss": 1.8835,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7463483889597421,
+      "grad_norm": 3.14360339600945,
+      "learning_rate": 1.4432904417176911e-05,
+      "loss": 1.9239,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7465408956373174,
+      "grad_norm": 2.877200407612022,
+      "learning_rate": 1.4430109387687352e-05,
+      "loss": 1.8334,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7467334023148928,
+      "grad_norm": 2.9756564385533437,
+      "learning_rate": 1.4427313927534483e-05,
+      "loss": 1.7689,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "grad_norm": 2.904140260457494,
+      "learning_rate": 1.4424518036990059e-05,
+      "loss": 1.9144,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 2.1963,
+      "step": 3880,
+      "vm_loss": 0.1126
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 1.3887,
+      "step": 3880,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 0.936,
+      "step": 3880,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 2.0121,
+      "step": 3880,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 1.6809,
+      "step": 3880,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 1.8965,
+      "step": 3880,
+      "vm_loss": 0.1147
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 2.0161,
+      "step": 3880,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 0.7469259089924681,
+      "lm_loss": 1.5197,
+      "step": 3880,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 0.7471184156700436,
+      "grad_norm": 2.912253714852121,
+      "learning_rate": 1.4421721716325874e-05,
+      "loss": 1.8549,
+      "step": 3881
+    },
+    {
+      "epoch": 0.747310922347619,
+      "grad_norm": 2.7843082722882886,
+      "learning_rate": 1.4418924965813767e-05,
+      "loss": 1.8798,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7475034290251943,
+      "grad_norm": 2.820575624027667,
+      "learning_rate": 1.4416127785725622e-05,
+      "loss": 1.9084,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7476959357027697,
+      "grad_norm": 2.8362621376225086,
+      "learning_rate": 1.4413330176333356e-05,
+      "loss": 1.8479,
+      "step": 3884
+    },
+    {
+      "epoch": 0.747888442380345,
+      "grad_norm": 2.914174112542245,
+      "learning_rate": 1.4410532137908935e-05,
+      "loss": 1.8347,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7480809490579204,
+      "grad_norm": 2.9674830287574023,
+      "learning_rate": 1.440773367072436e-05,
+      "loss": 1.8337,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7482734557354959,
+      "grad_norm": 3.235735026248816,
+      "learning_rate": 1.4404934775051682e-05,
+      "loss": 1.8931,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "grad_norm": 3.012888010875251,
+      "learning_rate": 1.440213545116299e-05,
+      "loss": 1.8514,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.8733,
+      "step": 3888,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.6941,
+      "step": 3888,
+      "vm_loss": 0.2008
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.5611,
+      "step": 3888,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.9527,
+      "step": 3888,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 2.2498,
+      "step": 3888,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.5923,
+      "step": 3888,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 2.1298,
+      "step": 3888,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.7484659624130712,
+      "lm_loss": 1.4539,
+      "step": 3888,
+      "vm_loss": 0.2041
+    },
+    {
+      "epoch": 0.7486584690906466,
+      "grad_norm": 3.0391284543297714,
+      "learning_rate": 1.4399335699330412e-05,
+      "loss": 1.9166,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7488509757682219,
+      "grad_norm": 2.8420235929792352,
+      "learning_rate": 1.439653551982612e-05,
+      "loss": 1.8831,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7490434824457973,
+      "grad_norm": 2.8722009357184386,
+      "learning_rate": 1.4393734912922326e-05,
+      "loss": 1.8296,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7492359891233727,
+      "grad_norm": 2.9941890579370383,
+      "learning_rate": 1.4390933878891286e-05,
+      "loss": 1.8771,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7494284958009481,
+      "grad_norm": 3.2520967650133987,
+      "learning_rate": 1.4388132418005299e-05,
+      "loss": 1.868,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7496210024785235,
+      "grad_norm": 3.2932001252044913,
+      "learning_rate": 1.43853305305367e-05,
+      "loss": 1.8571,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7498135091560989,
+      "grad_norm": 2.9912662393601996,
+      "learning_rate": 1.4382528216757867e-05,
+      "loss": 1.8718,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "grad_norm": 3.029563285262763,
+      "learning_rate": 1.4379725476941222e-05,
+      "loss": 1.867,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.4198,
+      "step": 3896,
+      "vm_loss": 0.2212
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.8307,
+      "step": 3896,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.3123,
+      "step": 3896,
+      "vm_loss": 0.185
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 2.0322,
+      "step": 3896,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.7277,
+      "step": 3896,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.5033,
+      "step": 3896,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.7703,
+      "step": 3896,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 0.7500060158336742,
+      "lm_loss": 1.7678,
+      "step": 3896,
+      "vm_loss": 0.2395
+    },
+    {
+      "epoch": 0.7501985225112496,
+      "grad_norm": 3.031277070056288,
+      "learning_rate": 1.437692231135923e-05,
+      "loss": 1.7953,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7503910291888249,
+      "grad_norm": 3.0646516595423634,
+      "learning_rate": 1.4374118720284388e-05,
+      "loss": 1.8635,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7505835358664004,
+      "grad_norm": 3.094125218287347,
+      "learning_rate": 1.437131470398925e-05,
+      "loss": 1.885,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7507760425439758,
+      "grad_norm": 3.0173812164233795,
+      "learning_rate": 1.4368510262746393e-05,
+      "loss": 1.8942,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7509685492215511,
+      "grad_norm": 3.1860832458680757,
+      "learning_rate": 1.4365705396828452e-05,
+      "loss": 1.8622,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7511610558991265,
+      "grad_norm": 3.126314852927878,
+      "learning_rate": 1.4362900106508094e-05,
+      "loss": 1.8794,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7513535625767018,
+      "grad_norm": 3.0744961197882485,
+      "learning_rate": 1.4360094392058024e-05,
+      "loss": 1.8516,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "grad_norm": 2.9634641082677713,
+      "learning_rate": 1.4357288253750996e-05,
+      "loss": 1.9332,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 1.2288,
+      "step": 3904,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 1.4124,
+      "step": 3904,
+      "vm_loss": 0.1998
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 2.2023,
+      "step": 3904,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 2.0974,
+      "step": 3904,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 1.9227,
+      "step": 3904,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 2.2358,
+      "step": 3904,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 1.5846,
+      "step": 3904,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 0.7515460692542772,
+      "lm_loss": 1.7607,
+      "step": 3904,
+      "vm_loss": 0.2853
+    },
+    {
+      "epoch": 0.7517385759318527,
+      "grad_norm": 2.978005621161641,
+      "learning_rate": 1.4354481691859809e-05,
+      "loss": 1.9035,
+      "step": 3905
+    },
+    {
+      "epoch": 0.751931082609428,
+      "grad_norm": 3.0099502006903487,
+      "learning_rate": 1.4351674706657288e-05,
+      "loss": 1.9248,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7521235892870034,
+      "grad_norm": 2.9986193380978032,
+      "learning_rate": 1.4348867298416313e-05,
+      "loss": 1.8944,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7523160959645788,
+      "grad_norm": 2.9627537913807522,
+      "learning_rate": 1.4346059467409798e-05,
+      "loss": 1.8015,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7525086026421541,
+      "grad_norm": 3.2489308260690675,
+      "learning_rate": 1.4343251213910698e-05,
+      "loss": 1.7759,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7527011093197296,
+      "grad_norm": 3.1126730832924987,
+      "learning_rate": 1.4340442538192018e-05,
+      "loss": 1.8406,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7528936159973049,
+      "grad_norm": 2.972358642088698,
+      "learning_rate": 1.4337633440526791e-05,
+      "loss": 1.901,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "grad_norm": 2.95847612504498,
+      "learning_rate": 1.4334823921188098e-05,
+      "loss": 1.869,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.6565,
+      "step": 3912,
+      "vm_loss": 0.2607
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.7178,
+      "step": 3912,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 2.1061,
+      "step": 3912,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 2.3604,
+      "step": 3912,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.7426,
+      "step": 3912,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.8165,
+      "step": 3912,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.9397,
+      "step": 3912,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 0.7530861226748803,
+      "lm_loss": 1.2593,
+      "step": 3912,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.7532786293524557,
+      "grad_norm": 3.089997570161952,
+      "learning_rate": 1.4332013980449063e-05,
+      "loss": 1.8888,
+      "step": 3913
+    },
+    {
+      "epoch": 0.753471136030031,
+      "grad_norm": 2.948105995137009,
+      "learning_rate": 1.4329203618582845e-05,
+      "loss": 1.8735,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7536636427076064,
+      "grad_norm": 2.9142879797968564,
+      "learning_rate": 1.4326392835862648e-05,
+      "loss": 1.8487,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7538561493851818,
+      "grad_norm": 2.833880698658168,
+      "learning_rate": 1.4323581632561716e-05,
+      "loss": 1.8807,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7540486560627572,
+      "grad_norm": 3.0335384496629163,
+      "learning_rate": 1.4320770008953338e-05,
+      "loss": 1.8224,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7542411627403326,
+      "grad_norm": 3.352566358195461,
+      "learning_rate": 1.4317957965310834e-05,
+      "loss": 1.8276,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7544336694179079,
+      "grad_norm": 3.091554004826876,
+      "learning_rate": 1.4315145501907575e-05,
+      "loss": 1.8271,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "grad_norm": 3.2014966592167564,
+      "learning_rate": 1.4312332619016964e-05,
+      "loss": 1.7813,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.5071,
+      "step": 3920,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.4232,
+      "step": 3920,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.5671,
+      "step": 3920,
+      "vm_loss": 0.2175
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.7572,
+      "step": 3920,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.9343,
+      "step": 3920,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.7613,
+      "step": 3920,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.9382,
+      "step": 3920,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.7546261760954833,
+      "lm_loss": 1.8083,
+      "step": 3920,
+      "vm_loss": 0.2572
+    },
+    {
+      "epoch": 0.7548186827730586,
+      "grad_norm": 2.8945642993549776,
+      "learning_rate": 1.430951931691245e-05,
+      "loss": 1.8143,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7550111894506341,
+      "grad_norm": 3.0719264506978403,
+      "learning_rate": 1.4306705595867528e-05,
+      "loss": 1.852,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7552036961282095,
+      "grad_norm": 3.3224304368249666,
+      "learning_rate": 1.4303891456155718e-05,
+      "loss": 1.8697,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7553962028057848,
+      "grad_norm": 3.1948690992466373,
+      "learning_rate": 1.4301076898050601e-05,
+      "loss": 1.8788,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7555887094833602,
+      "grad_norm": 3.1131279809084775,
+      "learning_rate": 1.429826192182578e-05,
+      "loss": 1.9402,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7557812161609356,
+      "grad_norm": 2.959955782238859,
+      "learning_rate": 1.429544652775491e-05,
+      "loss": 1.8269,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7559737228385109,
+      "grad_norm": 2.9504320946450777,
+      "learning_rate": 1.4292630716111682e-05,
+      "loss": 1.9012,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "grad_norm": 3.212225613502672,
+      "learning_rate": 1.428981448716983e-05,
+      "loss": 1.9018,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.6364,
+      "step": 3928,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.6102,
+      "step": 3928,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.6576,
+      "step": 3928,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.2704,
+      "step": 3928,
+      "vm_loss": 0.207
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.701,
+      "step": 3928,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.9657,
+      "step": 3928,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 1.5115,
+      "step": 3928,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 0.7561662295160864,
+      "lm_loss": 2.0605,
+      "step": 3928,
+      "vm_loss": 0.2228
+    },
+    {
+      "epoch": 0.7563587361936617,
+      "grad_norm": 3.1072399237488777,
+      "learning_rate": 1.4286997841203127e-05,
+      "loss": 1.901,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7565512428712371,
+      "grad_norm": 2.993995786318162,
+      "learning_rate": 1.4284180778485388e-05,
+      "loss": 1.8361,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7567437495488125,
+      "grad_norm": 2.801774156807687,
+      "learning_rate": 1.4281363299290466e-05,
+      "loss": 1.824,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7569362562263878,
+      "grad_norm": 2.82552839984301,
+      "learning_rate": 1.4278545403892257e-05,
+      "loss": 1.8591,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7571287629039632,
+      "grad_norm": 2.942922458721903,
+      "learning_rate": 1.4275727092564696e-05,
+      "loss": 1.8325,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7573212695815386,
+      "grad_norm": 2.9027549912041546,
+      "learning_rate": 1.4272908365581758e-05,
+      "loss": 1.7916,
+      "step": 3934
+    },
+    {
+      "epoch": 0.757513776259114,
+      "grad_norm": 2.883983175187057,
+      "learning_rate": 1.427008922321746e-05,
+      "loss": 1.8013,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "grad_norm": 3.143747270042312,
+      "learning_rate": 1.4267269665745862e-05,
+      "loss": 1.863,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.8463,
+      "step": 3936,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.2841,
+      "step": 3936,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.7682,
+      "step": 3936,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.6743,
+      "step": 3936,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.6931,
+      "step": 3936,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.6102,
+      "step": 3936,
+      "vm_loss": 0.2155
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.7691,
+      "step": 3936,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 0.7577062829366894,
+      "lm_loss": 1.3207,
+      "step": 3936,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 0.7578987896142647,
+      "grad_norm": 3.1518217098390675,
+      "learning_rate": 1.4264449693441057e-05,
+      "loss": 1.8609,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7580912962918401,
+      "grad_norm": 2.967204088171265,
+      "learning_rate": 1.4261629306577182e-05,
+      "loss": 1.8501,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7582838029694156,
+      "grad_norm": 2.896278825295079,
+      "learning_rate": 1.4258808505428419e-05,
+      "loss": 1.8782,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7584763096469909,
+      "grad_norm": 2.951709637457461,
+      "learning_rate": 1.425598729026898e-05,
+      "loss": 1.8666,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7586688163245663,
+      "grad_norm": 2.8596163288953886,
+      "learning_rate": 1.4253165661373129e-05,
+      "loss": 1.7891,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7588613230021416,
+      "grad_norm": 2.9194011470989087,
+      "learning_rate": 1.4250343619015162e-05,
+      "loss": 1.8329,
+      "step": 3942
+    },
+    {
+      "epoch": 0.759053829679717,
+      "grad_norm": 3.2166618225609045,
+      "learning_rate": 1.4247521163469413e-05,
+      "loss": 1.9026,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "grad_norm": 3.163432110421066,
+      "learning_rate": 1.4244698295010268e-05,
+      "loss": 1.8788,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.9078,
+      "step": 3944,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.1845,
+      "step": 3944,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.6857,
+      "step": 3944,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.0323,
+      "step": 3944,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.9001,
+      "step": 3944,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.6769,
+      "step": 3944,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.372,
+      "step": 3944,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.7592463363572924,
+      "lm_loss": 1.7638,
+      "step": 3944,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 0.7594388430348677,
+      "grad_norm": 2.757870121723556,
+      "learning_rate": 1.4241875013912143e-05,
+      "loss": 1.8452,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7596313497124432,
+      "grad_norm": 3.0363678141840293,
+      "learning_rate": 1.4239051320449501e-05,
+      "loss": 1.8401,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7598238563900185,
+      "grad_norm": 3.180450693408773,
+      "learning_rate": 1.4236227214896831e-05,
+      "loss": 1.839,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7600163630675939,
+      "grad_norm": 2.8819630276964983,
+      "learning_rate": 1.4233402697528685e-05,
+      "loss": 1.9173,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7602088697451693,
+      "grad_norm": 2.987702219284944,
+      "learning_rate": 1.423057776861963e-05,
+      "loss": 1.7845,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7604013764227446,
+      "grad_norm": 3.340641470073444,
+      "learning_rate": 1.4227752428444294e-05,
+      "loss": 1.9255,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7605938831003201,
+      "grad_norm": 2.978987478726501,
+      "learning_rate": 1.4224926677277336e-05,
+      "loss": 1.8569,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "grad_norm": 2.845046259713777,
+      "learning_rate": 1.4222100515393447e-05,
+      "loss": 1.8437,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.2545,
+      "step": 3952,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.1551,
+      "step": 3952,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.9904,
+      "step": 3952,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.3759,
+      "step": 3952,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 2.0572,
+      "step": 3952,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.9234,
+      "step": 3952,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.4797,
+      "step": 3952,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.7607863897778954,
+      "lm_loss": 1.8059,
+      "step": 3952,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 0.7609788964554708,
+      "grad_norm": 2.985387755989666,
+      "learning_rate": 1.4219273943067375e-05,
+      "loss": 1.8321,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7611714031330462,
+      "grad_norm": 3.085324422170434,
+      "learning_rate": 1.4216446960573893e-05,
+      "loss": 1.808,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7613639098106215,
+      "grad_norm": 3.0042291326829376,
+      "learning_rate": 1.4213619568187827e-05,
+      "loss": 1.8534,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7615564164881969,
+      "grad_norm": 3.1357714629379165,
+      "learning_rate": 1.4210791766184025e-05,
+      "loss": 1.8321,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7617489231657724,
+      "grad_norm": 3.1411157814644577,
+      "learning_rate": 1.4207963554837391e-05,
+      "loss": 1.7995,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7619414298433477,
+      "grad_norm": 3.0650157266410982,
+      "learning_rate": 1.420513493442287e-05,
+      "loss": 1.8373,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7621339365209231,
+      "grad_norm": 2.989938974532215,
+      "learning_rate": 1.420230590521543e-05,
+      "loss": 1.8302,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "grad_norm": 2.895483167815223,
+      "learning_rate": 1.4199476467490097e-05,
+      "loss": 1.7878,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.4025,
+      "step": 3960,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 2.04,
+      "step": 3960,
+      "vm_loss": 0.2038
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.3072,
+      "step": 3960,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.4161,
+      "step": 3960,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.302,
+      "step": 3960,
+      "vm_loss": 0.2173
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.7175,
+      "step": 3960,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 2.1627,
+      "step": 3960,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 0.7623264431984984,
+      "lm_loss": 1.6751,
+      "step": 3960,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 0.7625189498760738,
+      "grad_norm": 3.1139646239547734,
+      "learning_rate": 1.4196646621521919e-05,
+      "loss": 1.8937,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7627114565536492,
+      "grad_norm": 3.0772541018928634,
+      "learning_rate": 1.4193816367586e-05,
+      "loss": 1.8611,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7629039632312246,
+      "grad_norm": 2.859952195349067,
+      "learning_rate": 1.4190985705957476e-05,
+      "loss": 1.8489,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7630964699088,
+      "grad_norm": 3.1744025105175466,
+      "learning_rate": 1.4188154636911524e-05,
+      "loss": 1.9318,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7632889765863753,
+      "grad_norm": 3.3165222272553185,
+      "learning_rate": 1.4185323160723357e-05,
+      "loss": 1.8166,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7634814832639507,
+      "grad_norm": 3.0765716137629027,
+      "learning_rate": 1.4182491277668233e-05,
+      "loss": 1.7699,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7636739899415261,
+      "grad_norm": 3.0423554445344227,
+      "learning_rate": 1.4179658988021447e-05,
+      "loss": 1.8542,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "grad_norm": 3.2152753769254403,
+      "learning_rate": 1.4176826292058338e-05,
+      "loss": 1.8537,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.6772,
+      "step": 3968,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.2117,
+      "step": 3968,
+      "vm_loss": 0.1119
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.5107,
+      "step": 3968,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.5143,
+      "step": 3968,
+      "vm_loss": 0.1174
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 2.1202,
+      "step": 3968,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.851,
+      "step": 3968,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.7877,
+      "step": 3968,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.7638664966191014,
+      "lm_loss": 1.5673,
+      "step": 3968,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.7640590032966769,
+      "grad_norm": 3.31072984280053,
+      "learning_rate": 1.4173993190054272e-05,
+      "loss": 1.7911,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7642515099742523,
+      "grad_norm": 3.0398249054418818,
+      "learning_rate": 1.417115968228467e-05,
+      "loss": 1.8015,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7644440166518276,
+      "grad_norm": 2.8167325835752663,
+      "learning_rate": 1.416832576902498e-05,
+      "loss": 1.8447,
+      "step": 3971
+    },
+    {
+      "epoch": 0.764636523329403,
+      "grad_norm": 2.9676542282376155,
+      "learning_rate": 1.41654914505507e-05,
+      "loss": 1.8211,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7648290300069783,
+      "grad_norm": 2.961992646510966,
+      "learning_rate": 1.4162656727137353e-05,
+      "loss": 1.8905,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7650215366845537,
+      "grad_norm": 3.034034752343732,
+      "learning_rate": 1.4159821599060518e-05,
+      "loss": 1.8651,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7652140433621292,
+      "grad_norm": 2.898247826972282,
+      "learning_rate": 1.4156986066595807e-05,
+      "loss": 1.8471,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "grad_norm": 2.9700540489161766,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 1.797,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.7395,
+      "step": 3976,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.7074,
+      "step": 3976,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.4981,
+      "step": 3976,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.5866,
+      "step": 3976,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.8493,
+      "step": 3976,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.7097,
+      "step": 3976,
+      "vm_loss": 0.2271
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.4475,
+      "step": 3976,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 0.7654065500397045,
+      "lm_loss": 1.832,
+      "step": 3976,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 0.7655990567172799,
+      "grad_norm": 2.8183329341015435,
+      "learning_rate": 1.4151313789605386e-05,
+      "loss": 1.8533,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7657915633948552,
+      "grad_norm": 2.9791558971756613,
+      "learning_rate": 1.4148477045631094e-05,
+      "loss": 1.8373,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7659840700724306,
+      "grad_norm": 2.8972751134585657,
+      "learning_rate": 1.4145639898371755e-05,
+      "loss": 1.8384,
+      "step": 3979
+    },
+    {
+      "epoch": 0.766176576750006,
+      "grad_norm": 3.1050748687162018,
+      "learning_rate": 1.4142802348103184e-05,
+      "loss": 1.7923,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7663690834275814,
+      "grad_norm": 3.0763978947148254,
+      "learning_rate": 1.4139964395101222e-05,
+      "loss": 1.8703,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7665615901051568,
+      "grad_norm": 2.945435291021433,
+      "learning_rate": 1.4137126039641755e-05,
+      "loss": 1.828,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7667540967827321,
+      "grad_norm": 2.79987981235688,
+      "learning_rate": 1.4134287282000708e-05,
+      "loss": 1.8288,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "grad_norm": 2.8943327544633495,
+      "learning_rate": 1.4131448122454042e-05,
+      "loss": 1.8372,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.8878,
+      "step": 3984,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.9113,
+      "step": 3984,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.3748,
+      "step": 3984,
+      "vm_loss": 0.2142
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.637,
+      "step": 3984,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.5189,
+      "step": 3984,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.5993,
+      "step": 3984,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 2.1075,
+      "step": 3984,
+      "vm_loss": 0.1115
+    },
+    {
+      "epoch": 0.7669466034603075,
+      "lm_loss": 1.5463,
+      "step": 3984,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 0.7671391101378829,
+      "grad_norm": 2.924952715945339,
+      "learning_rate": 1.412860856127776e-05,
+      "loss": 1.8579,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7673316168154583,
+      "grad_norm": 3.1058606435304266,
+      "learning_rate": 1.412576859874791e-05,
+      "loss": 1.748,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7675241234930337,
+      "grad_norm": 2.905379502230526,
+      "learning_rate": 1.412292823514057e-05,
+      "loss": 1.8257,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7677166301706091,
+      "grad_norm": 3.4764941502537035,
+      "learning_rate": 1.4120087470731854e-05,
+      "loss": 1.8689,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7679091368481844,
+      "grad_norm": 3.133069201261247,
+      "learning_rate": 1.4117246305797925e-05,
+      "loss": 1.8739,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7681016435257598,
+      "grad_norm": 2.9230297544358126,
+      "learning_rate": 1.411440474061498e-05,
+      "loss": 1.8838,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7682941502033351,
+      "grad_norm": 2.9363653302656885,
+      "learning_rate": 1.4111562775459256e-05,
+      "loss": 1.8304,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "grad_norm": 2.859654908042205,
+      "learning_rate": 1.4108720410607028e-05,
+      "loss": 1.8379,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 1.143,
+      "step": 3992,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 2.0982,
+      "step": 3992,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 0.9737,
+      "step": 3992,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 1.7117,
+      "step": 3992,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 1.9586,
+      "step": 3992,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 1.5372,
+      "step": 3992,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 1.2196,
+      "step": 3992,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 0.7684866568809106,
+      "lm_loss": 2.2931,
+      "step": 3992,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.768679163558486,
+      "grad_norm": 2.8818544641674846,
+      "learning_rate": 1.4105877646334611e-05,
+      "loss": 1.9196,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7688716702360613,
+      "grad_norm": 2.8991168237385208,
+      "learning_rate": 1.4103034482918359e-05,
+      "loss": 1.7729,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7690641769136367,
+      "grad_norm": 2.9582439988838196,
+      "learning_rate": 1.4100190920634657e-05,
+      "loss": 1.7994,
+      "step": 3995
+    },
+    {
+      "epoch": 0.769256683591212,
+      "grad_norm": 3.3426038447491124,
+      "learning_rate": 1.4097346959759943e-05,
+      "loss": 1.8557,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7694491902687874,
+      "grad_norm": 3.1638795121153818,
+      "learning_rate": 1.4094502600570687e-05,
+      "loss": 1.7578,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7696416969463629,
+      "grad_norm": 2.9947585256683027,
+      "learning_rate": 1.4091657843343392e-05,
+      "loss": 1.8195,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7698342036239382,
+      "grad_norm": 2.850048292803185,
+      "learning_rate": 1.4088812688354607e-05,
+      "loss": 1.7869,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "grad_norm": 3.1299515256116215,
+      "learning_rate": 1.4085967135880915e-05,
+      "loss": 1.7451,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.0991,
+      "step": 4000,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.5031,
+      "step": 4000,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.4119,
+      "step": 4000,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.5303,
+      "step": 4000,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.7049,
+      "step": 4000,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.8045,
+      "step": 4000,
+      "vm_loss": 0.1059
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.8893,
+      "step": 4000,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 0.7700267103015136,
+      "lm_loss": 1.9339,
+      "step": 4000,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 0.770219216979089,
+      "grad_norm": 2.88009095925609,
+      "learning_rate": 1.408312118619895e-05,
+      "loss": 1.7356,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7704117236566643,
+      "grad_norm": 2.987194684422851,
+      "learning_rate": 1.408027483958536e-05,
+      "loss": 1.7879,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7706042303342397,
+      "grad_norm": 3.1587514781539126,
+      "learning_rate": 1.4077428096316857e-05,
+      "loss": 1.8439,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7707967370118151,
+      "grad_norm": 3.1832997675628163,
+      "learning_rate": 1.4074580956670176e-05,
+      "loss": 1.8227,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7709892436893905,
+      "grad_norm": 3.1088278440263855,
+      "learning_rate": 1.4071733420922096e-05,
+      "loss": 1.8454,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7711817503669659,
+      "grad_norm": 2.9423191545091947,
+      "learning_rate": 1.4068885489349437e-05,
+      "loss": 1.8288,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7713742570445412,
+      "grad_norm": 3.1755948417808475,
+      "learning_rate": 1.406603716222905e-05,
+      "loss": 1.8897,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "grad_norm": 2.913291665844025,
+      "learning_rate": 1.4063188439837831e-05,
+      "loss": 1.7643,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.5608,
+      "step": 4008,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.5677,
+      "step": 4008,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.7777,
+      "step": 4008,
+      "vm_loss": 0.2226
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.5992,
+      "step": 4008,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.9004,
+      "step": 4008,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.4711,
+      "step": 4008,
+      "vm_loss": 0.2121
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.5642,
+      "step": 4008,
+      "vm_loss": 0.2166
+    },
+    {
+      "epoch": 0.7715667637221166,
+      "lm_loss": 1.3571,
+      "step": 4008,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.7717592703996919,
+      "grad_norm": 2.9424246623209163,
+      "learning_rate": 1.4060339322452717e-05,
+      "loss": 1.8602,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7719517770772674,
+      "grad_norm": 3.137520034929048,
+      "learning_rate": 1.4057489810350668e-05,
+      "loss": 1.8616,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7721442837548428,
+      "grad_norm": 3.1707206302899587,
+      "learning_rate": 1.4054639903808704e-05,
+      "loss": 1.8335,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7723367904324181,
+      "grad_norm": 3.0661467867274075,
+      "learning_rate": 1.4051789603103865e-05,
+      "loss": 1.77,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7725292971099935,
+      "grad_norm": 3.019270131821078,
+      "learning_rate": 1.4048938908513239e-05,
+      "loss": 1.7895,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7727218037875688,
+      "grad_norm": 3.0046318497102473,
+      "learning_rate": 1.4046087820313949e-05,
+      "loss": 1.8895,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7729143104651442,
+      "grad_norm": 2.936466085236794,
+      "learning_rate": 1.404323633878316e-05,
+      "loss": 1.8369,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "grad_norm": 2.781151789315632,
+      "learning_rate": 1.4040384464198072e-05,
+      "loss": 1.7959,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 1.6099,
+      "step": 4016,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 1.2724,
+      "step": 4016,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 2.2063,
+      "step": 4016,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 1.7715,
+      "step": 4016,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 1.9839,
+      "step": 4016,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 1.52,
+      "step": 4016,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 2.1398,
+      "step": 4016,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 0.7731068171427197,
+      "lm_loss": 2.0469,
+      "step": 4016,
+      "vm_loss": 0.0888
+    },
+    {
+      "epoch": 0.773299323820295,
+      "grad_norm": 2.9666955011730693,
+      "learning_rate": 1.4037532196835918e-05,
+      "loss": 1.8401,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7734918304978704,
+      "grad_norm": 3.125914517258326,
+      "learning_rate": 1.4034679536973985e-05,
+      "loss": 1.8088,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7736843371754458,
+      "grad_norm": 3.2431790134318126,
+      "learning_rate": 1.4031826484889579e-05,
+      "loss": 1.8709,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7738768438530211,
+      "grad_norm": 3.139605941407338,
+      "learning_rate": 1.4028973040860056e-05,
+      "loss": 1.8715,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7740693505305966,
+      "grad_norm": 2.8109364526210823,
+      "learning_rate": 1.402611920516281e-05,
+      "loss": 1.8227,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7742618572081719,
+      "grad_norm": 2.853754697606584,
+      "learning_rate": 1.4023264978075266e-05,
+      "loss": 1.8375,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7744543638857473,
+      "grad_norm": 3.010463561666103,
+      "learning_rate": 1.4020410359874897e-05,
+      "loss": 1.8411,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "grad_norm": 2.901744673093344,
+      "learning_rate": 1.40175553508392e-05,
+      "loss": 1.8887,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.4194,
+      "step": 4024,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.8114,
+      "step": 4024,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.4711,
+      "step": 4024,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.5008,
+      "step": 4024,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.7318,
+      "step": 4024,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.6722,
+      "step": 4024,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.8567,
+      "step": 4024,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.7746468705633227,
+      "lm_loss": 1.5928,
+      "step": 4024,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 0.774839377240898,
+      "grad_norm": 3.0331698143575947,
+      "learning_rate": 1.4014699951245725e-05,
+      "loss": 1.8565,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7750318839184734,
+      "grad_norm": 2.9077554051636714,
+      "learning_rate": 1.4011844161372054e-05,
+      "loss": 1.8223,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7752243905960488,
+      "grad_norm": 3.0577890206872067,
+      "learning_rate": 1.4008987981495801e-05,
+      "loss": 1.8245,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7754168972736242,
+      "grad_norm": 3.3130318926779316,
+      "learning_rate": 1.4006131411894627e-05,
+      "loss": 1.8063,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7756094039511996,
+      "grad_norm": 3.3150186099020016,
+      "learning_rate": 1.4003274452846227e-05,
+      "loss": 1.8869,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7758019106287749,
+      "grad_norm": 3.073970282576432,
+      "learning_rate": 1.4000417104628332e-05,
+      "loss": 1.828,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7759944173063503,
+      "grad_norm": 2.8507468939052707,
+      "learning_rate": 1.3997559367518718e-05,
+      "loss": 1.7164,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "grad_norm": 2.872735085083444,
+      "learning_rate": 1.3994701241795188e-05,
+      "loss": 1.7941,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.594,
+      "step": 4032,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.7514,
+      "step": 4032,
+      "vm_loss": 0.2017
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.86,
+      "step": 4032,
+      "vm_loss": 0.0957
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 2.5574,
+      "step": 4032,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.7639,
+      "step": 4032,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.9776,
+      "step": 4032,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.8151,
+      "step": 4032,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.7761869239839257,
+      "lm_loss": 1.5379,
+      "step": 4032,
+      "vm_loss": 0.2461
+    },
+    {
+      "epoch": 0.7763794306615011,
+      "grad_norm": 3.0256713042689265,
+      "learning_rate": 1.3991842727735589e-05,
+      "loss": 1.8689,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7765719373390765,
+      "grad_norm": 2.962435975496588,
+      "learning_rate": 1.3988983825617806e-05,
+      "loss": 1.7957,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7767644440166518,
+      "grad_norm": 3.1034203355082535,
+      "learning_rate": 1.3986124535719766e-05,
+      "loss": 1.8842,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7769569506942272,
+      "grad_norm": 3.0571419352254394,
+      "learning_rate": 1.3983264858319416e-05,
+      "loss": 1.8052,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7771494573718026,
+      "grad_norm": 2.9462460798693204,
+      "learning_rate": 1.3980404793694766e-05,
+      "loss": 1.7405,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7773419640493779,
+      "grad_norm": 2.813958994943593,
+      "learning_rate": 1.3977544342123846e-05,
+      "loss": 1.8,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7775344707269534,
+      "grad_norm": 2.7774017658501795,
+      "learning_rate": 1.3974683503884725e-05,
+      "loss": 1.8311,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "grad_norm": 2.8141368449901516,
+      "learning_rate": 1.3971822279255518e-05,
+      "loss": 1.843,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.8025,
+      "step": 4040,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.6357,
+      "step": 4040,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.8278,
+      "step": 4040,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 2.1172,
+      "step": 4040,
+      "vm_loss": 0.2529
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.5749,
+      "step": 4040,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.357,
+      "step": 4040,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 1.0937,
+      "step": 4040,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 0.7777269774045287,
+      "lm_loss": 0.9817,
+      "step": 4040,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 0.7779194840821041,
+      "grad_norm": 2.7949606433976886,
+      "learning_rate": 1.396896066851437e-05,
+      "loss": 1.8226,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7781119907596795,
+      "grad_norm": 3.024343789695567,
+      "learning_rate": 1.396609867193947e-05,
+      "loss": 1.7849,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7783044974372548,
+      "grad_norm": 3.2153926302851623,
+      "learning_rate": 1.3963236289809037e-05,
+      "loss": 1.8482,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7784970041148302,
+      "grad_norm": 3.116901715360136,
+      "learning_rate": 1.396037352240133e-05,
+      "loss": 1.766,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7786895107924056,
+      "grad_norm": 2.9049954421487496,
+      "learning_rate": 1.3957510369994651e-05,
+      "loss": 1.7699,
+      "step": 4045
+    },
+    {
+      "epoch": 0.778882017469981,
+      "grad_norm": 2.9345498375797137,
+      "learning_rate": 1.3954646832867332e-05,
+      "loss": 1.7634,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7790745241475564,
+      "grad_norm": 2.8354739448950137,
+      "learning_rate": 1.3951782911297745e-05,
+      "loss": 1.766,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "grad_norm": 2.94549970705335,
+      "learning_rate": 1.3948918605564302e-05,
+      "loss": 1.788,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.7841,
+      "step": 4048,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 2.1371,
+      "step": 4048,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.2237,
+      "step": 4048,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.4411,
+      "step": 4048,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 0.9289,
+      "step": 4048,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.8343,
+      "step": 4048,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.6576,
+      "step": 4048,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 0.7792670308251317,
+      "lm_loss": 1.1951,
+      "step": 4048,
+      "vm_loss": 0.232
+    },
+    {
+      "epoch": 0.7794595375027071,
+      "grad_norm": 2.9205606952664933,
+      "learning_rate": 1.3946053915945447e-05,
+      "loss": 1.8279,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7796520441802826,
+      "grad_norm": 2.8229355954617112,
+      "learning_rate": 1.3943188842719669e-05,
+      "loss": 1.7886,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7798445508578579,
+      "grad_norm": 2.8820103434268547,
+      "learning_rate": 1.3940323386165486e-05,
+      "loss": 1.8028,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7800370575354333,
+      "grad_norm": 3.111333756648554,
+      "learning_rate": 1.393745754656146e-05,
+      "loss": 1.8146,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7802295642130086,
+      "grad_norm": 2.9772298504003363,
+      "learning_rate": 1.3934591324186184e-05,
+      "loss": 1.7845,
+      "step": 4053
+    },
+    {
+      "epoch": 0.780422070890584,
+      "grad_norm": 3.0515547437708492,
+      "learning_rate": 1.3931724719318293e-05,
+      "loss": 1.7499,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7806145775681594,
+      "grad_norm": 2.956722880524745,
+      "learning_rate": 1.3928857732236461e-05,
+      "loss": 1.7989,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "grad_norm": 2.778985714413757,
+      "learning_rate": 1.3925990363219391e-05,
+      "loss": 1.9,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.3524,
+      "step": 4056,
+      "vm_loss": 0.1024
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.5343,
+      "step": 4056,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 0.8962,
+      "step": 4056,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.4084,
+      "step": 4056,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.666,
+      "step": 4056,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.5609,
+      "step": 4056,
+      "vm_loss": 0.1231
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 2.0982,
+      "step": 4056,
+      "vm_loss": 0.103
+    },
+    {
+      "epoch": 0.7808070842457348,
+      "lm_loss": 1.8253,
+      "step": 4056,
+      "vm_loss": 0.1976
+    },
+    {
+      "epoch": 0.7809995909233102,
+      "grad_norm": 3.0039588319775876,
+      "learning_rate": 1.3923122612545829e-05,
+      "loss": 1.7997,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7811920976008855,
+      "grad_norm": 2.9983927081128234,
+      "learning_rate": 1.3920254480494558e-05,
+      "loss": 1.7633,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7813846042784609,
+      "grad_norm": 3.067163532516728,
+      "learning_rate": 1.3917385967344395e-05,
+      "loss": 1.8634,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7815771109560363,
+      "grad_norm": 3.0466989589705262,
+      "learning_rate": 1.39145170733742e-05,
+      "loss": 1.817,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7817696176336116,
+      "grad_norm": 3.0451178017885723,
+      "learning_rate": 1.3911647798862864e-05,
+      "loss": 1.7693,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7819621243111871,
+      "grad_norm": 3.011886601372958,
+      "learning_rate": 1.3908778144089318e-05,
+      "loss": 1.7836,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7821546309887625,
+      "grad_norm": 2.829280743382815,
+      "learning_rate": 1.3905908109332526e-05,
+      "loss": 1.7758,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "grad_norm": 2.9431053026682843,
+      "learning_rate": 1.39030376948715e-05,
+      "loss": 1.8196,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 2.0483,
+      "step": 4064,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 2.1775,
+      "step": 4064,
+      "vm_loss": 0.2173
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.1851,
+      "step": 4064,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.4748,
+      "step": 4064,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.0178,
+      "step": 4064,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.0924,
+      "step": 4064,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.8808,
+      "step": 4064,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.7823471376663378,
+      "lm_loss": 1.1702,
+      "step": 4064,
+      "vm_loss": 0.1893
+    },
+    {
+      "epoch": 0.7825396443439132,
+      "grad_norm": 3.0483740041496254,
+      "learning_rate": 1.390016690098527e-05,
+      "loss": 1.8378,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7827321510214885,
+      "grad_norm": 2.8271357025801698,
+      "learning_rate": 1.3897295727952922e-05,
+      "loss": 1.8298,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7829246576990639,
+      "grad_norm": 3.0999878284764697,
+      "learning_rate": 1.3894424176053566e-05,
+      "loss": 1.7901,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7831171643766394,
+      "grad_norm": 3.0944559937144858,
+      "learning_rate": 1.3891552245566357e-05,
+      "loss": 1.7819,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7833096710542147,
+      "grad_norm": 3.06537944330935,
+      "learning_rate": 1.3888679936770485e-05,
+      "loss": 1.7079,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7835021777317901,
+      "grad_norm": 3.1509569988191917,
+      "learning_rate": 1.3885807249945166e-05,
+      "loss": 1.8096,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7836946844093654,
+      "grad_norm": 3.1595710943026174,
+      "learning_rate": 1.3882934185369674e-05,
+      "loss": 1.8333,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "grad_norm": 3.1486737483786733,
+      "learning_rate": 1.38800607433233e-05,
+      "loss": 1.8572,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.7733,
+      "step": 4072,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.7505,
+      "step": 4072,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.2405,
+      "step": 4072,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.2191,
+      "step": 4072,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.6826,
+      "step": 4072,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.2599,
+      "step": 4072,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.3386,
+      "step": 4072,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.7838871910869408,
+      "lm_loss": 1.304,
+      "step": 4072,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 0.7840796977645162,
+      "grad_norm": 2.942846956341721,
+      "learning_rate": 1.3877186924085381e-05,
+      "loss": 1.8078,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7842722044420916,
+      "grad_norm": 2.9717820726789856,
+      "learning_rate": 1.3874312727935293e-05,
+      "loss": 1.799,
+      "step": 4074
+    },
+    {
+      "epoch": 0.784464711119667,
+      "grad_norm": 2.909170511089149,
+      "learning_rate": 1.3871438155152437e-05,
+      "loss": 1.7596,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7846572177972423,
+      "grad_norm": 3.1610936927974502,
+      "learning_rate": 1.3868563206016263e-05,
+      "loss": 1.8184,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7848497244748177,
+      "grad_norm": 3.012412494238963,
+      "learning_rate": 1.3865687880806253e-05,
+      "loss": 1.7573,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7850422311523931,
+      "grad_norm": 3.0813443893217487,
+      "learning_rate": 1.3862812179801922e-05,
+      "loss": 1.7904,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7852347378299684,
+      "grad_norm": 2.9588761143031785,
+      "learning_rate": 1.385993610328283e-05,
+      "loss": 1.7988,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "grad_norm": 3.050427421099969,
+      "learning_rate": 1.3857059651528564e-05,
+      "loss": 1.7342,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.9363,
+      "step": 4080,
+      "vm_loss": 0.1232
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.9776,
+      "step": 4080,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.5176,
+      "step": 4080,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.8757,
+      "step": 4080,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.4139,
+      "step": 4080,
+      "vm_loss": 0.1283
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.684,
+      "step": 4080,
+      "vm_loss": 0.2414
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.7236,
+      "step": 4080,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 0.7854272445075439,
+      "lm_loss": 1.3506,
+      "step": 4080,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.7856197511851193,
+      "grad_norm": 3.1097761182725088,
+      "learning_rate": 1.3854182824818755e-05,
+      "loss": 1.7873,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7858122578626946,
+      "grad_norm": 3.1260086699133427,
+      "learning_rate": 1.3851305623433064e-05,
+      "loss": 1.7808,
+      "step": 4082
+    },
+    {
+      "epoch": 0.78600476454027,
+      "grad_norm": 2.9690424458336833,
+      "learning_rate": 1.3848428047651195e-05,
+      "loss": 1.8014,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7861972712178453,
+      "grad_norm": 2.9957559993132397,
+      "learning_rate": 1.3845550097752884e-05,
+      "loss": 1.7166,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7863897778954207,
+      "grad_norm": 3.374527272864488,
+      "learning_rate": 1.3842671774017905e-05,
+      "loss": 1.8519,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7865822845729962,
+      "grad_norm": 3.178359363226118,
+      "learning_rate": 1.3839793076726069e-05,
+      "loss": 1.9157,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7867747912505715,
+      "grad_norm": 2.9233902786787405,
+      "learning_rate": 1.3836914006157218e-05,
+      "loss": 1.8245,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "grad_norm": 2.938397457704189,
+      "learning_rate": 1.3834034562591236e-05,
+      "loss": 1.7517,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.2485,
+      "step": 4088,
+      "vm_loss": 0.2384
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.7862,
+      "step": 4088,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.7462,
+      "step": 4088,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.3506,
+      "step": 4088,
+      "vm_loss": 0.1964
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.8381,
+      "step": 4088,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.7815,
+      "step": 4088,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.444,
+      "step": 4088,
+      "vm_loss": 0.2113
+    },
+    {
+      "epoch": 0.7869672979281469,
+      "lm_loss": 1.5421,
+      "step": 4088,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.7871598046057222,
+      "grad_norm": 2.9556575549687976,
+      "learning_rate": 1.3831154746308046e-05,
+      "loss": 1.8388,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7873523112832976,
+      "grad_norm": 3.0504332699491536,
+      "learning_rate": 1.38282745575876e-05,
+      "loss": 1.809,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7875448179608731,
+      "grad_norm": 3.325100100646713,
+      "learning_rate": 1.3825393996709885e-05,
+      "loss": 1.8219,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7877373246384484,
+      "grad_norm": 3.1646913595687076,
+      "learning_rate": 1.3822513063954934e-05,
+      "loss": 1.8466,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7879298313160238,
+      "grad_norm": 3.2056820159609902,
+      "learning_rate": 1.3819631759602812e-05,
+      "loss": 1.857,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7881223379935991,
+      "grad_norm": 3.032242283613238,
+      "learning_rate": 1.3816750083933612e-05,
+      "loss": 1.8223,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7883148446711745,
+      "grad_norm": 3.113099355250901,
+      "learning_rate": 1.3813868037227477e-05,
+      "loss": 1.7623,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "grad_norm": 3.0948790490464617,
+      "learning_rate": 1.3810985619764573e-05,
+      "loss": 1.8179,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.4658,
+      "step": 4096,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.7345,
+      "step": 4096,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.6974,
+      "step": 4096,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.7658,
+      "step": 4096,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.9589,
+      "step": 4096,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.6987,
+      "step": 4096,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.6775,
+      "step": 4096,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 0.7885073513487499,
+      "lm_loss": 1.3665,
+      "step": 4096,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.7886998580263253,
+      "grad_norm": 3.112798483159566,
+      "learning_rate": 1.3808102831825109e-05,
+      "loss": 1.8462,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7888923647039007,
+      "grad_norm": 3.0224642032864613,
+      "learning_rate": 1.3805219673689333e-05,
+      "loss": 1.7617,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7890848713814761,
+      "grad_norm": 3.0285955634333965,
+      "learning_rate": 1.3802336145637519e-05,
+      "loss": 1.8123,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7892773780590514,
+      "grad_norm": 2.93467580979092,
+      "learning_rate": 1.3799452247949985e-05,
+      "loss": 1.8467,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7894698847366268,
+      "grad_norm": 2.99445117237434,
+      "learning_rate": 1.3796567980907085e-05,
+      "loss": 1.8077,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7896623914142021,
+      "grad_norm": 3.0410283214470564,
+      "learning_rate": 1.3793683344789206e-05,
+      "loss": 1.7431,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7898548980917776,
+      "grad_norm": 3.2777335966472214,
+      "learning_rate": 1.3790798339876769e-05,
+      "loss": 1.8565,
+      "step": 4103
+    },
+    {
+      "epoch": 0.790047404769353,
+      "grad_norm": 2.8317291643216564,
+      "learning_rate": 1.3787912966450234e-05,
+      "loss": 1.7599,
+      "step": 4104
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.6602,
+      "step": 4104,
+      "vm_loss": 0.1152
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.4861,
+      "step": 4104,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.5513,
+      "step": 4104,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.5775,
+      "step": 4104,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.2776,
+      "step": 4104,
+      "vm_loss": 0.0974
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.1686,
+      "step": 4104,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 2.1077,
+      "step": 4104,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.790047404769353,
+      "lm_loss": 1.5043,
+      "step": 4104,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 0.7902399114469283,
+      "grad_norm": 3.0348646008479756,
+      "learning_rate": 1.3785027224790098e-05,
+      "loss": 1.7403,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7904324181245037,
+      "grad_norm": 2.9492514307203472,
+      "learning_rate": 1.3782141115176893e-05,
+      "loss": 1.8339,
+      "step": 4106
+    },
+    {
+      "epoch": 0.790624924802079,
+      "grad_norm": 2.953620654582367,
+      "learning_rate": 1.3779254637891181e-05,
+      "loss": 1.7783,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7908174314796544,
+      "grad_norm": 3.028874213426188,
+      "learning_rate": 1.377636779321357e-05,
+      "loss": 1.8489,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7910099381572299,
+      "grad_norm": 3.086760055031574,
+      "learning_rate": 1.3773480581424694e-05,
+      "loss": 1.8296,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7912024448348052,
+      "grad_norm": 2.9565670864497684,
+      "learning_rate": 1.3770593002805228e-05,
+      "loss": 1.8049,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7913949515123806,
+      "grad_norm": 2.9900317204590183,
+      "learning_rate": 1.3767705057635885e-05,
+      "loss": 1.7794,
+      "step": 4111
+    },
+    {
+      "epoch": 0.791587458189956,
+      "grad_norm": 2.8574744503109986,
+      "learning_rate": 1.3764816746197405e-05,
+      "loss": 1.7062,
+      "step": 4112
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.818,
+      "step": 4112,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.8306,
+      "step": 4112,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.937,
+      "step": 4112,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.4639,
+      "step": 4112,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.2267,
+      "step": 4112,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.5697,
+      "step": 4112,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 0.883,
+      "step": 4112,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.791587458189956,
+      "lm_loss": 1.985,
+      "step": 4112,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 0.7917799648675313,
+      "grad_norm": 3.0439372590121456,
+      "learning_rate": 1.3761928068770569e-05,
+      "loss": 1.8224,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7919724715451067,
+      "grad_norm": 3.060132207663992,
+      "learning_rate": 1.3759039025636199e-05,
+      "loss": 1.8291,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7921649782226821,
+      "grad_norm": 2.969659313034043,
+      "learning_rate": 1.3756149617075143e-05,
+      "loss": 1.7905,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7923574849002575,
+      "grad_norm": 3.1349497200103285,
+      "learning_rate": 1.3753259843368287e-05,
+      "loss": 1.8445,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7925499915778329,
+      "grad_norm": 3.032116286133235,
+      "learning_rate": 1.3750369704796557e-05,
+      "loss": 1.8869,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7927424982554082,
+      "grad_norm": 2.961375721578509,
+      "learning_rate": 1.3747479201640914e-05,
+      "loss": 1.8675,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7929350049329836,
+      "grad_norm": 2.9345590134424944,
+      "learning_rate": 1.3744588334182346e-05,
+      "loss": 1.7738,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "grad_norm": 2.789534675391815,
+      "learning_rate": 1.3741697102701884e-05,
+      "loss": 1.7528,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.2631,
+      "step": 4120,
+      "vm_loss": 0.2968
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.7093,
+      "step": 4120,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.8607,
+      "step": 4120,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.6359,
+      "step": 4120,
+      "vm_loss": 0.1152
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.6032,
+      "step": 4120,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.3036,
+      "step": 4120,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.4906,
+      "step": 4120,
+      "vm_loss": 0.1171
+    },
+    {
+      "epoch": 0.7931275116105589,
+      "lm_loss": 1.2669,
+      "step": 4120,
+      "vm_loss": 0.2248
+    },
+    {
+      "epoch": 0.7933200182881344,
+      "grad_norm": 2.8590303458393143,
+      "learning_rate": 1.3738805507480593e-05,
+      "loss": 1.7927,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7935125249657098,
+      "grad_norm": 3.172233154223333,
+      "learning_rate": 1.3735913548799576e-05,
+      "loss": 1.9059,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7937050316432851,
+      "grad_norm": 2.9152528967846534,
+      "learning_rate": 1.3733021226939962e-05,
+      "loss": 1.8546,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7938975383208605,
+      "grad_norm": 3.0002474134073394,
+      "learning_rate": 1.3730128542182926e-05,
+      "loss": 1.8077,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7940900449984358,
+      "grad_norm": 2.9023444029749377,
+      "learning_rate": 1.3727235494809677e-05,
+      "loss": 1.7438,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7942825516760113,
+      "grad_norm": 2.9158866311670293,
+      "learning_rate": 1.372434208510145e-05,
+      "loss": 1.7853,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7944750583535867,
+      "grad_norm": 3.1289160815928168,
+      "learning_rate": 1.3721448313339525e-05,
+      "loss": 1.7448,
+      "step": 4127
+    },
+    {
+      "epoch": 0.794667565031162,
+      "grad_norm": 3.0904899443513556,
+      "learning_rate": 1.371855417980521e-05,
+      "loss": 1.819,
+      "step": 4128
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.3426,
+      "step": 4128,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.8798,
+      "step": 4128,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.2844,
+      "step": 4128,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.5478,
+      "step": 4128,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.8707,
+      "step": 4128,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.6675,
+      "step": 4128,
+      "vm_loss": 0.2084
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.7918,
+      "step": 4128,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.794667565031162,
+      "lm_loss": 1.8239,
+      "step": 4128,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.7948600717087374,
+      "grad_norm": 2.9179317679232053,
+      "learning_rate": 1.3715659684779857e-05,
+      "loss": 1.8219,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7950525783863128,
+      "grad_norm": 3.071604802587858,
+      "learning_rate": 1.3712764828544844e-05,
+      "loss": 1.8166,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7952450850638881,
+      "grad_norm": 2.961709357463004,
+      "learning_rate": 1.3709869611381592e-05,
+      "loss": 1.8102,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7954375917414636,
+      "grad_norm": 2.8831109008537092,
+      "learning_rate": 1.3706974033571548e-05,
+      "loss": 1.7393,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7956300984190389,
+      "grad_norm": 3.1013353062502684,
+      "learning_rate": 1.3704078095396204e-05,
+      "loss": 1.7627,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7958226050966143,
+      "grad_norm": 3.0922116208491586,
+      "learning_rate": 1.3701181797137077e-05,
+      "loss": 1.7066,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7960151117741897,
+      "grad_norm": 3.0261397418312312,
+      "learning_rate": 1.3698285139075732e-05,
+      "loss": 1.7395,
+      "step": 4135
+    },
+    {
+      "epoch": 0.796207618451765,
+      "grad_norm": 2.930750769491821,
+      "learning_rate": 1.369538812149375e-05,
+      "loss": 1.8324,
+      "step": 4136
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.4476,
+      "step": 4136,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.7678,
+      "step": 4136,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.0921,
+      "step": 4136,
+      "vm_loss": 0.2071
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 0.7299,
+      "step": 4136,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.1343,
+      "step": 4136,
+      "vm_loss": 0.0957
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 2.0802,
+      "step": 4136,
+      "vm_loss": 0.0988
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.4803,
+      "step": 4136,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.796207618451765,
+      "lm_loss": 1.8327,
+      "step": 4136,
+      "vm_loss": 0.1967
+    },
+    {
+      "epoch": 0.7964001251293404,
+      "grad_norm": 3.1196954415221576,
+      "learning_rate": 1.3692490744672772e-05,
+      "loss": 1.7881,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7965926318069158,
+      "grad_norm": 3.0863368487562424,
+      "learning_rate": 1.3689593008894448e-05,
+      "loss": 1.8187,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7967851384844912,
+      "grad_norm": 2.8681528478976723,
+      "learning_rate": 1.3686694914440484e-05,
+      "loss": 1.8407,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7969776451620666,
+      "grad_norm": 3.058069903730238,
+      "learning_rate": 1.3683796461592604e-05,
+      "loss": 1.7561,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7971701518396419,
+      "grad_norm": 3.095041929547832,
+      "learning_rate": 1.3680897650632581e-05,
+      "loss": 1.8094,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7973626585172173,
+      "grad_norm": 3.091448095377173,
+      "learning_rate": 1.3677998481842213e-05,
+      "loss": 1.8349,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7975551651947927,
+      "grad_norm": 3.1095072101732737,
+      "learning_rate": 1.3675098955503336e-05,
+      "loss": 1.8109,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "grad_norm": 3.040109112713848,
+      "learning_rate": 1.3672199071897825e-05,
+      "loss": 1.8036,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.3263,
+      "step": 4144,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.5142,
+      "step": 4144,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 2.0414,
+      "step": 4144,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.7033,
+      "step": 4144,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.4974,
+      "step": 4144,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.8169,
+      "step": 4144,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.4401,
+      "step": 4144,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.7977476718723681,
+      "lm_loss": 1.0588,
+      "step": 4144,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 0.7979401785499435,
+      "grad_norm": 3.1020639310752025,
+      "learning_rate": 1.366929883130758e-05,
+      "loss": 1.7401,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7981326852275188,
+      "grad_norm": 3.1023362498290865,
+      "learning_rate": 1.3666398234014546e-05,
+      "loss": 1.7683,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7983251919050942,
+      "grad_norm": 2.9222792817742733,
+      "learning_rate": 1.3663497280300698e-05,
+      "loss": 1.6984,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7985176985826696,
+      "grad_norm": 3.1709375783402236,
+      "learning_rate": 1.3660595970448044e-05,
+      "loss": 1.8345,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7987102052602449,
+      "grad_norm": 3.2405450862529377,
+      "learning_rate": 1.3657694304738627e-05,
+      "loss": 1.7698,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7989027119378204,
+      "grad_norm": 3.0996143666495724,
+      "learning_rate": 1.3654792283454529e-05,
+      "loss": 1.7334,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7990952186153957,
+      "grad_norm": 3.153233711735116,
+      "learning_rate": 1.3651889906877865e-05,
+      "loss": 1.807,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "grad_norm": 3.0355731091566382,
+      "learning_rate": 1.3648987175290777e-05,
+      "loss": 1.7163,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.4009,
+      "step": 4152,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 2.0995,
+      "step": 4152,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.5432,
+      "step": 4152,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.8983,
+      "step": 4152,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.7048,
+      "step": 4152,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.7232,
+      "step": 4152,
+      "vm_loss": 0.2077
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.8925,
+      "step": 4152,
+      "vm_loss": 0.2076
+    },
+    {
+      "epoch": 0.7992877252929711,
+      "lm_loss": 1.6771,
+      "step": 4152,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.7994802319705465,
+      "grad_norm": 3.0203741570733946,
+      "learning_rate": 1.364608408897545e-05,
+      "loss": 1.7903,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7996727386481218,
+      "grad_norm": 3.0844672072280517,
+      "learning_rate": 1.3643180648214105e-05,
+      "loss": 1.7259,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7998652453256972,
+      "grad_norm": 3.0718964121654198,
+      "learning_rate": 1.3640276853288992e-05,
+      "loss": 1.7457,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8000577520032726,
+      "grad_norm": 3.0013264252849163,
+      "learning_rate": 1.3637372704482397e-05,
+      "loss": 1.7349,
+      "step": 4156
+    },
+    {
+      "epoch": 0.800250258680848,
+      "grad_norm": 3.166518840818398,
+      "learning_rate": 1.3634468202076639e-05,
+      "loss": 1.8091,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8004427653584234,
+      "grad_norm": 3.1510012996782413,
+      "learning_rate": 1.3631563346354072e-05,
+      "loss": 1.7817,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8006352720359987,
+      "grad_norm": 2.9929104043703023,
+      "learning_rate": 1.3628658137597086e-05,
+      "loss": 1.7993,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "grad_norm": 2.9075421072370107,
+      "learning_rate": 1.3625752576088105e-05,
+      "loss": 1.7418,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.62,
+      "step": 4160,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.3552,
+      "step": 4160,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.6661,
+      "step": 4160,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.0421,
+      "step": 4160,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.8515,
+      "step": 4160,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.8931,
+      "step": 4160,
+      "vm_loss": 0.2289
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.0161,
+      "step": 4160,
+      "vm_loss": 0.1207
+    },
+    {
+      "epoch": 0.8008277787135741,
+      "lm_loss": 1.0134,
+      "step": 4160,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 0.8010202853911496,
+      "grad_norm": 3.006315087639913,
+      "learning_rate": 1.3622846662109589e-05,
+      "loss": 1.8408,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8012127920687249,
+      "grad_norm": 2.8956403409845257,
+      "learning_rate": 1.3619940395944027e-05,
+      "loss": 1.7977,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8014052987463003,
+      "grad_norm": 2.8891525324244802,
+      "learning_rate": 1.3617033777873943e-05,
+      "loss": 1.7503,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8015978054238756,
+      "grad_norm": 3.021975654259892,
+      "learning_rate": 1.3614126808181904e-05,
+      "loss": 1.781,
+      "step": 4164
+    },
+    {
+      "epoch": 0.801790312101451,
+      "grad_norm": 3.143310442513092,
+      "learning_rate": 1.36112194871505e-05,
+      "loss": 1.7302,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8019828187790264,
+      "grad_norm": 3.2191722146098134,
+      "learning_rate": 1.3608311815062361e-05,
+      "loss": 1.7207,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8021753254566018,
+      "grad_norm": 3.170107194061057,
+      "learning_rate": 1.360540379220015e-05,
+      "loss": 1.7594,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "grad_norm": 3.0484191505845732,
+      "learning_rate": 1.3602495418846564e-05,
+      "loss": 1.774,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.2763,
+      "step": 4168,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.174,
+      "step": 4168,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.1064,
+      "step": 4168,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.2594,
+      "step": 4168,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 2.0125,
+      "step": 4168,
+      "vm_loss": 0.1338
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.9577,
+      "step": 4168,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.4861,
+      "step": 4168,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.8023678321341772,
+      "lm_loss": 1.7579,
+      "step": 4168,
+      "vm_loss": 0.1159
+    },
+    {
+      "epoch": 0.8025603388117525,
+      "grad_norm": 3.050749151806815,
+      "learning_rate": 1.3599586695284331e-05,
+      "loss": 1.694,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8027528454893279,
+      "grad_norm": 2.86913173466317,
+      "learning_rate": 1.3596677621796221e-05,
+      "loss": 1.7799,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8029453521669033,
+      "grad_norm": 2.8261896613352726,
+      "learning_rate": 1.3593768198665031e-05,
+      "loss": 1.743,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8031378588444786,
+      "grad_norm": 2.9453427834323627,
+      "learning_rate": 1.3590858426173595e-05,
+      "loss": 1.7357,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8033303655220541,
+      "grad_norm": 3.1042981432954364,
+      "learning_rate": 1.3587948304604781e-05,
+      "loss": 1.8153,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8035228721996295,
+      "grad_norm": 3.1985162848087674,
+      "learning_rate": 1.3585037834241485e-05,
+      "loss": 1.7691,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8037153788772048,
+      "grad_norm": 3.1839762209239124,
+      "learning_rate": 1.3582127015366649e-05,
+      "loss": 1.8489,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "grad_norm": 3.1115510513237528,
+      "learning_rate": 1.3579215848263234e-05,
+      "loss": 1.7104,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.4175,
+      "step": 4176,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.4508,
+      "step": 4176,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.7968,
+      "step": 4176,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.9964,
+      "step": 4176,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.0855,
+      "step": 4176,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.7647,
+      "step": 4176,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.7331,
+      "step": 4176,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 0.8039078855547802,
+      "lm_loss": 1.94,
+      "step": 4176,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 0.8041003922323555,
+      "grad_norm": 3.0471759909201683,
+      "learning_rate": 1.357630433321425e-05,
+      "loss": 1.7945,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8042928989099309,
+      "grad_norm": 3.1023322126403063,
+      "learning_rate": 1.3573392470502729e-05,
+      "loss": 1.7121,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8044854055875064,
+      "grad_norm": 3.1005887719437983,
+      "learning_rate": 1.3570480260411745e-05,
+      "loss": 1.8209,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8046779122650817,
+      "grad_norm": 2.8865554987876743,
+      "learning_rate": 1.3567567703224398e-05,
+      "loss": 1.7591,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8048704189426571,
+      "grad_norm": 2.9767501269679117,
+      "learning_rate": 1.356465479922383e-05,
+      "loss": 1.7614,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8050629256202324,
+      "grad_norm": 2.9145814771687113,
+      "learning_rate": 1.3561741548693208e-05,
+      "loss": 1.7493,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8052554322978078,
+      "grad_norm": 3.19190644211542,
+      "learning_rate": 1.355882795191574e-05,
+      "loss": 1.7349,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "grad_norm": 3.2383860814875027,
+      "learning_rate": 1.3555914009174665e-05,
+      "loss": 1.8099,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.7311,
+      "step": 4184,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.6794,
+      "step": 4184,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.6862,
+      "step": 4184,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.3711,
+      "step": 4184,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.3464,
+      "step": 4184,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.8787,
+      "step": 4184,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 2.037,
+      "step": 4184,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 0.8054479389753832,
+      "lm_loss": 1.5833,
+      "step": 4184,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 0.8056404456529586,
+      "grad_norm": 3.0350628682987915,
+      "learning_rate": 1.3552999720753255e-05,
+      "loss": 1.7292,
+      "step": 4185
+    },
+    {
+      "epoch": 0.805832952330534,
+      "grad_norm": 2.8679878356224813,
+      "learning_rate": 1.3550085086934815e-05,
+      "loss": 1.6892,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8060254590081093,
+      "grad_norm": 2.930655824720837,
+      "learning_rate": 1.3547170108002685e-05,
+      "loss": 1.8168,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8062179656856847,
+      "grad_norm": 3.0648528032094515,
+      "learning_rate": 1.3544254784240243e-05,
+      "loss": 1.7858,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8064104723632601,
+      "grad_norm": 3.1633340868402433,
+      "learning_rate": 1.3541339115930887e-05,
+      "loss": 1.7398,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8066029790408354,
+      "grad_norm": 3.034954374668336,
+      "learning_rate": 1.3538423103358066e-05,
+      "loss": 1.7034,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8067954857184109,
+      "grad_norm": 3.095265537520487,
+      "learning_rate": 1.3535506746805247e-05,
+      "loss": 1.7476,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "grad_norm": 3.268188239359184,
+      "learning_rate": 1.3532590046555939e-05,
+      "loss": 1.7765,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.5298,
+      "step": 4192,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.5119,
+      "step": 4192,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.3325,
+      "step": 4192,
+      "vm_loss": 0.1154
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.9065,
+      "step": 4192,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.3308,
+      "step": 4192,
+      "vm_loss": 0.2217
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.3178,
+      "step": 4192,
+      "vm_loss": 0.13
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.5023,
+      "step": 4192,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.8069879923959863,
+      "lm_loss": 1.5513,
+      "step": 4192,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.8071804990735616,
+      "grad_norm": 3.230598820652005,
+      "learning_rate": 1.3529673002893686e-05,
+      "loss": 1.7385,
+      "step": 4193
+    },
+    {
+      "epoch": 0.807373005751137,
+      "grad_norm": 3.060410825098863,
+      "learning_rate": 1.3526755616102058e-05,
+      "loss": 1.7631,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8075655124287123,
+      "grad_norm": 2.9509573237221067,
+      "learning_rate": 1.3523837886464664e-05,
+      "loss": 1.7553,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8077580191062878,
+      "grad_norm": 3.0285404122721484,
+      "learning_rate": 1.3520919814265146e-05,
+      "loss": 1.7242,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8079505257838632,
+      "grad_norm": 3.168218591535499,
+      "learning_rate": 1.3518001399787174e-05,
+      "loss": 1.6952,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8081430324614385,
+      "grad_norm": 2.9021182303180666,
+      "learning_rate": 1.3515082643314457e-05,
+      "loss": 1.7218,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8083355391390139,
+      "grad_norm": 3.0740080244492978,
+      "learning_rate": 1.3512163545130736e-05,
+      "loss": 1.7938,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "grad_norm": 3.185109915921439,
+      "learning_rate": 1.3509244105519788e-05,
+      "loss": 1.7589,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.5822,
+      "step": 4200,
+      "vm_loss": 0.2174
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.1601,
+      "step": 4200,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 2.2057,
+      "step": 4200,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.3627,
+      "step": 4200,
+      "vm_loss": 0.1052
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.9234,
+      "step": 4200,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 0.8681,
+      "step": 4200,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.2193,
+      "step": 4200,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.8085280458165892,
+      "lm_loss": 1.3911,
+      "step": 4200,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 0.8087205524941646,
+      "grad_norm": 2.8091591189525618,
+      "learning_rate": 1.3506324324765414e-05,
+      "loss": 1.6959,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8089130591717401,
+      "grad_norm": 3.0075184815212177,
+      "learning_rate": 1.3503404203151458e-05,
+      "loss": 1.766,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8091055658493154,
+      "grad_norm": 2.950334567951103,
+      "learning_rate": 1.3500483740961792e-05,
+      "loss": 1.749,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8092980725268908,
+      "grad_norm": 2.9699652426456447,
+      "learning_rate": 1.349756293848032e-05,
+      "loss": 1.7783,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8094905792044662,
+      "grad_norm": 3.103827831584143,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 1.7509,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8096830858820415,
+      "grad_norm": 3.0540475358481642,
+      "learning_rate": 1.3491720313777756e-05,
+      "loss": 1.7321,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8098755925596169,
+      "grad_norm": 3.1431337847449012,
+      "learning_rate": 1.3488798492124645e-05,
+      "loss": 1.697,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "grad_norm": 3.036980372111481,
+      "learning_rate": 1.3485876331315682e-05,
+      "loss": 1.7405,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.4505,
+      "step": 4208,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.515,
+      "step": 4208,
+      "vm_loss": 0.2041
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.8511,
+      "step": 4208,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.4691,
+      "step": 4208,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.7261,
+      "step": 4208,
+      "vm_loss": 0.2225
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.8041,
+      "step": 4208,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.55,
+      "step": 4208,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 0.8100680992371923,
+      "lm_loss": 1.7651,
+      "step": 4208,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 0.8102606059147677,
+      "grad_norm": 2.9630933375614137,
+      "learning_rate": 1.3482953831634941e-05,
+      "loss": 1.7285,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8104531125923431,
+      "grad_norm": 3.1595450969622054,
+      "learning_rate": 1.348003099336653e-05,
+      "loss": 1.727,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8106456192699184,
+      "grad_norm": 3.4164469045954564,
+      "learning_rate": 1.3477107816794583e-05,
+      "loss": 1.7991,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8108381259474938,
+      "grad_norm": 3.092609435794148,
+      "learning_rate": 1.3474184302203275e-05,
+      "loss": 1.7423,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8110306326250691,
+      "grad_norm": 2.970812730307828,
+      "learning_rate": 1.34712604498768e-05,
+      "loss": 1.7448,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8112231393026446,
+      "grad_norm": 3.135328031365485,
+      "learning_rate": 1.3468336260099401e-05,
+      "loss": 1.7889,
+      "step": 4214
+    },
+    {
+      "epoch": 0.81141564598022,
+      "grad_norm": 3.434772059259396,
+      "learning_rate": 1.3465411733155342e-05,
+      "loss": 1.8172,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "grad_norm": 3.2510891722237387,
+      "learning_rate": 1.346248686932893e-05,
+      "loss": 1.8063,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.8768,
+      "step": 4216,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.9212,
+      "step": 4216,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.9487,
+      "step": 4216,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 2.128,
+      "step": 4216,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.7591,
+      "step": 4216,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.5731,
+      "step": 4216,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.5061,
+      "step": 4216,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.8116081526577953,
+      "lm_loss": 1.5745,
+      "step": 4216,
+      "vm_loss": 0.2486
+    },
+    {
+      "epoch": 0.8118006593353707,
+      "grad_norm": 3.087688855434372,
+      "learning_rate": 1.3459561668904496e-05,
+      "loss": 1.8176,
+      "step": 4217
+    },
+    {
+      "epoch": 0.811993166012946,
+      "grad_norm": 3.0048426480157793,
+      "learning_rate": 1.3456636132166406e-05,
+      "loss": 1.7865,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8121856726905214,
+      "grad_norm": 2.9385449217627073,
+      "learning_rate": 1.345371025939906e-05,
+      "loss": 1.8323,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8123781793680969,
+      "grad_norm": 3.11194397351124,
+      "learning_rate": 1.3450784050886891e-05,
+      "loss": 1.7165,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8125706860456722,
+      "grad_norm": 3.10133339124572,
+      "learning_rate": 1.3447857506914363e-05,
+      "loss": 1.7359,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8127631927232476,
+      "grad_norm": 3.1377905145886444,
+      "learning_rate": 1.3444930627765978e-05,
+      "loss": 1.7256,
+      "step": 4222
+    },
+    {
+      "epoch": 0.812955699400823,
+      "grad_norm": 3.135955377110187,
+      "learning_rate": 1.344200341372626e-05,
+      "loss": 1.7197,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "grad_norm": 2.995005178114232,
+      "learning_rate": 1.3439075865079777e-05,
+      "loss": 1.6991,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.5297,
+      "step": 4224,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.4745,
+      "step": 4224,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.798,
+      "step": 4224,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.1874,
+      "step": 4224,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.6013,
+      "step": 4224,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 2.045,
+      "step": 4224,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 0.9502,
+      "step": 4224,
+      "vm_loss": 0.2246
+    },
+    {
+      "epoch": 0.8131482060783983,
+      "lm_loss": 1.5346,
+      "step": 4224,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.8133407127559737,
+      "grad_norm": 3.004449221976845,
+      "learning_rate": 1.3436147982111116e-05,
+      "loss": 1.6784,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8135332194335491,
+      "grad_norm": 3.1928208966914564,
+      "learning_rate": 1.3433219765104913e-05,
+      "loss": 1.7894,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8137257261111245,
+      "grad_norm": 3.025728406355041,
+      "learning_rate": 1.3430291214345825e-05,
+      "loss": 1.7178,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8139182327886999,
+      "grad_norm": 3.010848551962657,
+      "learning_rate": 1.3427362330118542e-05,
+      "loss": 1.7906,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8141107394662752,
+      "grad_norm": 3.077800366060747,
+      "learning_rate": 1.3424433112707798e-05,
+      "loss": 1.7662,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8143032461438506,
+      "grad_norm": 2.9697349009555674,
+      "learning_rate": 1.3421503562398339e-05,
+      "loss": 1.8232,
+      "step": 4230
+    },
+    {
+      "epoch": 0.814495752821426,
+      "grad_norm": 2.961893574534323,
+      "learning_rate": 1.3418573679474962e-05,
+      "loss": 1.7537,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "grad_norm": 3.0719664336014314,
+      "learning_rate": 1.3415643464222487e-05,
+      "loss": 1.7107,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.7374,
+      "step": 4232,
+      "vm_loss": 0.2452
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.1992,
+      "step": 4232,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.6857,
+      "step": 4232,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.5649,
+      "step": 4232,
+      "vm_loss": 0.1123
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.6239,
+      "step": 4232,
+      "vm_loss": 0.2647
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.8705,
+      "step": 4232,
+      "vm_loss": 0.226
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.4023,
+      "step": 4232,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 0.8146882594990014,
+      "lm_loss": 1.3925,
+      "step": 4232,
+      "vm_loss": 0.1695
+    },
+    {
+      "epoch": 0.8148807661765768,
+      "grad_norm": 3.056474152000144,
+      "learning_rate": 1.3412712916925768e-05,
+      "loss": 1.7672,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8150732728541521,
+      "grad_norm": 3.118553472919914,
+      "learning_rate": 1.3409782037869692e-05,
+      "loss": 1.7945,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8152657795317275,
+      "grad_norm": 2.9588114611600616,
+      "learning_rate": 1.3406850827339177e-05,
+      "loss": 1.7608,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8154582862093029,
+      "grad_norm": 2.952878484538887,
+      "learning_rate": 1.3403919285619179e-05,
+      "loss": 1.7732,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8156507928868783,
+      "grad_norm": 3.139813173629082,
+      "learning_rate": 1.3400987412994673e-05,
+      "loss": 1.7564,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8158432995644537,
+      "grad_norm": 3.113546562427499,
+      "learning_rate": 1.339805520975068e-05,
+      "loss": 1.7142,
+      "step": 4238
+    },
+    {
+      "epoch": 0.816035806242029,
+      "grad_norm": 3.0329534417999144,
+      "learning_rate": 1.3395122676172248e-05,
+      "loss": 1.8003,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "grad_norm": 2.9078464008822977,
+      "learning_rate": 1.3392189812544454e-05,
+      "loss": 1.7531,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.3255,
+      "step": 4240,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.9207,
+      "step": 4240,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.3671,
+      "step": 4240,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.8623,
+      "step": 4240,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.2443,
+      "step": 4240,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.7799,
+      "step": 4240,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.9387,
+      "step": 4240,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 0.8162283129196044,
+      "lm_loss": 1.9082,
+      "step": 4240,
+      "vm_loss": 0.2437
+    },
+    {
+      "epoch": 0.8164208195971798,
+      "grad_norm": 3.0017982685155107,
+      "learning_rate": 1.3389256619152416e-05,
+      "loss": 1.8217,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8166133262747551,
+      "grad_norm": 3.0487075049373074,
+      "learning_rate": 1.3386323096281268e-05,
+      "loss": 1.7352,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8168058329523306,
+      "grad_norm": 2.977176962346174,
+      "learning_rate": 1.3383389244216193e-05,
+      "loss": 1.7995,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8169983396299059,
+      "grad_norm": 2.8916049415010425,
+      "learning_rate": 1.3380455063242399e-05,
+      "loss": 1.7063,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8171908463074813,
+      "grad_norm": 2.9897527401065895,
+      "learning_rate": 1.3377520553645122e-05,
+      "loss": 1.7536,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8173833529850567,
+      "grad_norm": 2.97455430838434,
+      "learning_rate": 1.3374585715709638e-05,
+      "loss": 1.7192,
+      "step": 4246
+    },
+    {
+      "epoch": 0.817575859662632,
+      "grad_norm": 3.022859891085978,
+      "learning_rate": 1.3371650549721246e-05,
+      "loss": 1.7041,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "grad_norm": 2.926727736452646,
+      "learning_rate": 1.3368715055965289e-05,
+      "loss": 1.6899,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.7005,
+      "step": 4248,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.0538,
+      "step": 4248,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.6263,
+      "step": 4248,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.5526,
+      "step": 4248,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.5682,
+      "step": 4248,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.5879,
+      "step": 4248,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 1.8125,
+      "step": 4248,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 0.8177683663402074,
+      "lm_loss": 2.2098,
+      "step": 4248,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 0.8179608730177828,
+      "grad_norm": 2.9584910565929032,
+      "learning_rate": 1.3365779234727127e-05,
+      "loss": 1.7387,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8181533796953582,
+      "grad_norm": 3.195450526833678,
+      "learning_rate": 1.336284308629216e-05,
+      "loss": 1.654,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8183458863729336,
+      "grad_norm": 3.1213635065641743,
+      "learning_rate": 1.3359906610945829e-05,
+      "loss": 1.7943,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8185383930505089,
+      "grad_norm": 3.060303564759357,
+      "learning_rate": 1.3356969808973583e-05,
+      "loss": 1.8135,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8187308997280843,
+      "grad_norm": 2.907618765168843,
+      "learning_rate": 1.3354032680660928e-05,
+      "loss": 1.7595,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8189234064056597,
+      "grad_norm": 3.0830533502737145,
+      "learning_rate": 1.3351095226293383e-05,
+      "loss": 1.7521,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8191159130832351,
+      "grad_norm": 3.002790591094632,
+      "learning_rate": 1.3348157446156508e-05,
+      "loss": 1.7921,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "grad_norm": 3.12771378595863,
+      "learning_rate": 1.3345219340535897e-05,
+      "loss": 1.7093,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.5638,
+      "step": 4256,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.5537,
+      "step": 4256,
+      "vm_loss": 0.1042
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.8795,
+      "step": 4256,
+      "vm_loss": 0.1811
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.833,
+      "step": 4256,
+      "vm_loss": 0.1003
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.6799,
+      "step": 4256,
+      "vm_loss": 0.2417
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 2.0418,
+      "step": 4256,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.5613,
+      "step": 4256,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 0.8193084197608105,
+      "lm_loss": 1.3785,
+      "step": 4256,
+      "vm_loss": 0.0856
+    },
+    {
+      "epoch": 0.8195009264383858,
+      "grad_norm": 3.087553954059264,
+      "learning_rate": 1.3342280909717166e-05,
+      "loss": 1.7217,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8196934331159612,
+      "grad_norm": 3.297333111638422,
+      "learning_rate": 1.3339342153985973e-05,
+      "loss": 1.7545,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8198859397935366,
+      "grad_norm": 3.126109433302256,
+      "learning_rate": 1.3336403073627997e-05,
+      "loss": 1.7496,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8200784464711119,
+      "grad_norm": 3.0400943295068803,
+      "learning_rate": 1.3333463668928958e-05,
+      "loss": 1.7762,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8202709531486874,
+      "grad_norm": 3.0445408784785633,
+      "learning_rate": 1.3330523940174603e-05,
+      "loss": 1.7057,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8204634598262627,
+      "grad_norm": 2.933298159888886,
+      "learning_rate": 1.332758388765071e-05,
+      "loss": 1.7476,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8206559665038381,
+      "grad_norm": 2.967201689873884,
+      "learning_rate": 1.3324643511643091e-05,
+      "loss": 1.7618,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "grad_norm": 3.1038049182627345,
+      "learning_rate": 1.3321702812437591e-05,
+      "loss": 1.7328,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.3352,
+      "step": 4264,
+      "vm_loss": 0.2115
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.6949,
+      "step": 4264,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.4472,
+      "step": 4264,
+      "vm_loss": 0.1165
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 2.0302,
+      "step": 4264,
+      "vm_loss": 0.1158
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.5004,
+      "step": 4264,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 2.2003,
+      "step": 4264,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.6877,
+      "step": 4264,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.8208484731814135,
+      "lm_loss": 1.4109,
+      "step": 4264,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 0.8210409798589888,
+      "grad_norm": 3.4524913755032887,
+      "learning_rate": 1.3318761790320078e-05,
+      "loss": 1.7804,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8212334865365643,
+      "grad_norm": 3.084612718140637,
+      "learning_rate": 1.3315820445576461e-05,
+      "loss": 1.7303,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8214259932141396,
+      "grad_norm": 2.86453479820018,
+      "learning_rate": 1.3312878778492673e-05,
+      "loss": 1.7261,
+      "step": 4267
+    },
+    {
+      "epoch": 0.821618499891715,
+      "grad_norm": 2.9642940257281603,
+      "learning_rate": 1.3309936789354687e-05,
+      "loss": 1.7786,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8218110065692904,
+      "grad_norm": 2.872544224702129,
+      "learning_rate": 1.3306994478448497e-05,
+      "loss": 1.6908,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8220035132468657,
+      "grad_norm": 3.24797991298083,
+      "learning_rate": 1.3304051846060135e-05,
+      "loss": 1.7209,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8221960199244411,
+      "grad_norm": 3.2505037606017484,
+      "learning_rate": 1.3301108892475662e-05,
+      "loss": 1.6965,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "grad_norm": 3.31943131891016,
+      "learning_rate": 1.3298165617981171e-05,
+      "loss": 1.7208,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.3379,
+      "step": 4272,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.7845,
+      "step": 4272,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.5951,
+      "step": 4272,
+      "vm_loss": 0.2236
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.9505,
+      "step": 4272,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.632,
+      "step": 4272,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.5017,
+      "step": 4272,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.7023,
+      "step": 4272,
+      "vm_loss": 0.2229
+    },
+    {
+      "epoch": 0.8223885266020166,
+      "lm_loss": 1.5454,
+      "step": 4272,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 0.8225810332795919,
+      "grad_norm": 3.173916383370582,
+      "learning_rate": 1.3295222022862786e-05,
+      "loss": 1.7939,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8227735399571673,
+      "grad_norm": 3.020100243496548,
+      "learning_rate": 1.3292278107406662e-05,
+      "loss": 1.7385,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8229660466347426,
+      "grad_norm": 2.8670773424380767,
+      "learning_rate": 1.3289333871898985e-05,
+      "loss": 1.6237,
+      "step": 4275
+    },
+    {
+      "epoch": 0.823158553312318,
+      "grad_norm": 3.078847320816505,
+      "learning_rate": 1.3286389316625974e-05,
+      "loss": 1.8541,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8233510599898934,
+      "grad_norm": 3.267465954184543,
+      "learning_rate": 1.3283444441873879e-05,
+      "loss": 1.7641,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8235435666674688,
+      "grad_norm": 3.0722589517473287,
+      "learning_rate": 1.3280499247928971e-05,
+      "loss": 1.7111,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8237360733450442,
+      "grad_norm": 2.972695817002317,
+      "learning_rate": 1.3277553735077568e-05,
+      "loss": 1.6193,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "grad_norm": 3.469561820609016,
+      "learning_rate": 1.3274607903606011e-05,
+      "loss": 1.7081,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.8575,
+      "step": 4280,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.5696,
+      "step": 4280,
+      "vm_loss": 0.1104
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.828,
+      "step": 4280,
+      "vm_loss": 0.11
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.9287,
+      "step": 4280,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.2578,
+      "step": 4280,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.8995,
+      "step": 4280,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.0758,
+      "step": 4280,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 0.8239285800226195,
+      "lm_loss": 1.5689,
+      "step": 4280,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 0.8241210867001949,
+      "grad_norm": 3.3610551697349105,
+      "learning_rate": 1.3271661753800671e-05,
+      "loss": 1.7621,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8243135933777703,
+      "grad_norm": 3.206184853890844,
+      "learning_rate": 1.326871528594795e-05,
+      "loss": 1.7821,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8245061000553456,
+      "grad_norm": 3.2065666190871056,
+      "learning_rate": 1.3265768500334286e-05,
+      "loss": 1.7724,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8246986067329211,
+      "grad_norm": 3.0557994407050213,
+      "learning_rate": 1.3262821397246142e-05,
+      "loss": 1.7552,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8248911134104965,
+      "grad_norm": 3.16455530963429,
+      "learning_rate": 1.3259873976970017e-05,
+      "loss": 1.766,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8250836200880718,
+      "grad_norm": 3.11495912863915,
+      "learning_rate": 1.3256926239792432e-05,
+      "loss": 1.7142,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8252761267656472,
+      "grad_norm": 3.1639425631823235,
+      "learning_rate": 1.325397818599995e-05,
+      "loss": 1.7724,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "grad_norm": 3.2540000858537304,
+      "learning_rate": 1.3251029815879157e-05,
+      "loss": 1.7853,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.0671,
+      "step": 4288,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.7359,
+      "step": 4288,
+      "vm_loss": 0.19
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.3392,
+      "step": 4288,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.7199,
+      "step": 4288,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.3968,
+      "step": 4288,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.7895,
+      "step": 4288,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.5899,
+      "step": 4288,
+      "vm_loss": 0.2258
+    },
+    {
+      "epoch": 0.8254686334432225,
+      "lm_loss": 1.4685,
+      "step": 4288,
+      "vm_loss": 0.1074
+    },
+    {
+      "epoch": 0.8256611401207979,
+      "grad_norm": 3.3643625865291633,
+      "learning_rate": 1.3248081129716674e-05,
+      "loss": 1.714,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8258536467983734,
+      "grad_norm": 3.1606649244447365,
+      "learning_rate": 1.3245132127799147e-05,
+      "loss": 1.6943,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8260461534759487,
+      "grad_norm": 3.130702444656028,
+      "learning_rate": 1.3242182810413262e-05,
+      "loss": 1.6464,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8262386601535241,
+      "grad_norm": 3.1671571621051915,
+      "learning_rate": 1.3239233177845732e-05,
+      "loss": 1.5863,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8264311668310994,
+      "grad_norm": 3.1401928995314443,
+      "learning_rate": 1.323628323038329e-05,
+      "loss": 1.7095,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8266236735086748,
+      "grad_norm": 3.2568144055943256,
+      "learning_rate": 1.3233332968312715e-05,
+      "loss": 1.7271,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8268161801862502,
+      "grad_norm": 3.3262138501821514,
+      "learning_rate": 1.3230382391920808e-05,
+      "loss": 1.7909,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "grad_norm": 3.147954434809932,
+      "learning_rate": 1.3227431501494406e-05,
+      "loss": 1.7132,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.2794,
+      "step": 4296,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.3709,
+      "step": 4296,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.6473,
+      "step": 4296,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.554,
+      "step": 4296,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.2883,
+      "step": 4296,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.7367,
+      "step": 4296,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.7747,
+      "step": 4296,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 0.8270086868638256,
+      "lm_loss": 1.2201,
+      "step": 4296,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.827201193541401,
+      "grad_norm": 3.098110333123799,
+      "learning_rate": 1.3224480297320373e-05,
+      "loss": 1.7194,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8273937002189763,
+      "grad_norm": 2.954003103505225,
+      "learning_rate": 1.3221528779685597e-05,
+      "loss": 1.7169,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 3.1261572280040575,
+      "learning_rate": 1.3218576948877013e-05,
+      "loss": 1.7307,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8277787135741271,
+      "grad_norm": 2.9410075458142826,
+      "learning_rate": 1.3215624805181568e-05,
+      "loss": 1.7621,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8279712202517024,
+      "grad_norm": 3.15622291770706,
+      "learning_rate": 1.3212672348886254e-05,
+      "loss": 1.7768,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8281637269292779,
+      "grad_norm": 3.045413340542613,
+      "learning_rate": 1.3209719580278085e-05,
+      "loss": 1.6844,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8283562336068533,
+      "grad_norm": 3.0723424443103045,
+      "learning_rate": 1.3206766499644111e-05,
+      "loss": 1.7145,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "grad_norm": 3.080632806858383,
+      "learning_rate": 1.3203813107271401e-05,
+      "loss": 1.7135,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.4784,
+      "step": 4304,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.44,
+      "step": 4304,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.9401,
+      "step": 4304,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 2.0472,
+      "step": 4304,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.7871,
+      "step": 4304,
+      "vm_loss": 0.2297
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.8191,
+      "step": 4304,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.7695,
+      "step": 4304,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 0.8285487402844286,
+      "lm_loss": 1.858,
+      "step": 4304,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.828741246962004,
+      "grad_norm": 2.997829906303288,
+      "learning_rate": 1.3200859403447072e-05,
+      "loss": 1.7718,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8289337536395793,
+      "grad_norm": 3.0669689279731003,
+      "learning_rate": 1.319790538845826e-05,
+      "loss": 1.817,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8291262603171548,
+      "grad_norm": 3.024933959494137,
+      "learning_rate": 1.3194951062592127e-05,
+      "loss": 1.7663,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8293187669947302,
+      "grad_norm": 2.9890220690854803,
+      "learning_rate": 1.3191996426135876e-05,
+      "loss": 1.7223,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8295112736723055,
+      "grad_norm": 3.0407935122446905,
+      "learning_rate": 1.3189041479376739e-05,
+      "loss": 1.6666,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8297037803498809,
+      "grad_norm": 3.1377110576316087,
+      "learning_rate": 1.3186086222601965e-05,
+      "loss": 1.7573,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8298962870274562,
+      "grad_norm": 2.880648423818378,
+      "learning_rate": 1.3183130656098852e-05,
+      "loss": 1.7631,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "grad_norm": 3.100730411465285,
+      "learning_rate": 1.3180174780154712e-05,
+      "loss": 1.6916,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.3353,
+      "step": 4312,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.6853,
+      "step": 4312,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.286,
+      "step": 4312,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.7319,
+      "step": 4312,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.436,
+      "step": 4312,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.5393,
+      "step": 4312,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.4342,
+      "step": 4312,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 0.8300887937050316,
+      "lm_loss": 1.9062,
+      "step": 4312,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 0.8302813003826071,
+      "grad_norm": 2.9849119995428435,
+      "learning_rate": 1.3177218595056899e-05,
+      "loss": 1.665,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8304738070601824,
+      "grad_norm": 3.1273888132548513,
+      "learning_rate": 1.3174262101092792e-05,
+      "loss": 1.7917,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8306663137377578,
+      "grad_norm": 2.916513814222911,
+      "learning_rate": 1.3171305298549797e-05,
+      "loss": 1.6309,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8308588204153332,
+      "grad_norm": 3.0272047536354023,
+      "learning_rate": 1.3168348187715353e-05,
+      "loss": 1.7967,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8310513270929085,
+      "grad_norm": 2.924545959133155,
+      "learning_rate": 1.316539076887693e-05,
+      "loss": 1.7042,
+      "step": 4317
+    },
+    {
+      "epoch": 0.8312438337704839,
+      "grad_norm": 3.0858666430461996,
+      "learning_rate": 1.3162433042322026e-05,
+      "loss": 1.6617,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8314363404480593,
+      "grad_norm": 3.1526376925494084,
+      "learning_rate": 1.3159475008338174e-05,
+      "loss": 1.7445,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "grad_norm": 3.058549450276619,
+      "learning_rate": 1.3156516667212929e-05,
+      "loss": 1.6957,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.0864,
+      "step": 4320,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.7334,
+      "step": 4320,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.6165,
+      "step": 4320,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.5458,
+      "step": 4320,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.9312,
+      "step": 4320,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.7114,
+      "step": 4320,
+      "vm_loss": 0.1881
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.6733,
+      "step": 4320,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 0.8316288471256347,
+      "lm_loss": 1.4645,
+      "step": 4320,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 0.8318213538032101,
+      "grad_norm": 2.9855443525388656,
+      "learning_rate": 1.315355801923388e-05,
+      "loss": 1.7789,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8320138604807854,
+      "grad_norm": 3.1908073180382623,
+      "learning_rate": 1.3150599064688643e-05,
+      "loss": 1.7147,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8322063671583608,
+      "grad_norm": 3.1022150465040834,
+      "learning_rate": 1.3147639803864871e-05,
+      "loss": 1.76,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8323988738359361,
+      "grad_norm": 2.9471853463082156,
+      "learning_rate": 1.314468023705024e-05,
+      "loss": 1.6761,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8325913805135116,
+      "grad_norm": 2.854576162752735,
+      "learning_rate": 1.3141720364532458e-05,
+      "loss": 1.6797,
+      "step": 4325
+    },
+    {
+      "epoch": 0.832783887191087,
+      "grad_norm": 2.8976015643510107,
+      "learning_rate": 1.3138760186599262e-05,
+      "loss": 1.7644,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8329763938686623,
+      "grad_norm": 3.042699031256594,
+      "learning_rate": 1.3135799703538418e-05,
+      "loss": 1.7433,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "grad_norm": 3.112085598050744,
+      "learning_rate": 1.3132838915637727e-05,
+      "loss": 1.7964,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.6855,
+      "step": 4328,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.7287,
+      "step": 4328,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.6022,
+      "step": 4328,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 2.0756,
+      "step": 4328,
+      "vm_loss": 0.2051
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.2478,
+      "step": 4328,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.4121,
+      "step": 4328,
+      "vm_loss": 0.2168
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.7495,
+      "step": 4328,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 0.8331689005462377,
+      "lm_loss": 1.1925,
+      "step": 4328,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.833361407223813,
+      "grad_norm": 3.081484070977461,
+      "learning_rate": 1.312987782318501e-05,
+      "loss": 1.7265,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8335539139013884,
+      "grad_norm": 3.041756979727332,
+      "learning_rate": 1.3126916426468128e-05,
+      "loss": 1.747,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8337464205789639,
+      "grad_norm": 2.9962928499382917,
+      "learning_rate": 1.3123954725774964e-05,
+      "loss": 1.6656,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8339389272565392,
+      "grad_norm": 3.219265615435624,
+      "learning_rate": 1.3120992721393433e-05,
+      "loss": 1.6059,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8341314339341146,
+      "grad_norm": 2.929623078330894,
+      "learning_rate": 1.3118030413611483e-05,
+      "loss": 1.6296,
+      "step": 4333
+    },
+    {
+      "epoch": 0.83432394061169,
+      "grad_norm": 2.9423185987076317,
+      "learning_rate": 1.311506780271708e-05,
+      "loss": 1.6675,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8345164472892653,
+      "grad_norm": 3.079786539206392,
+      "learning_rate": 1.3112104888998238e-05,
+      "loss": 1.7253,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "grad_norm": 3.2597172240869536,
+      "learning_rate": 1.3109141672742983e-05,
+      "loss": 1.7855,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.9719,
+      "step": 4336,
+      "vm_loss": 0.2307
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.8905,
+      "step": 4336,
+      "vm_loss": 0.2066
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 2.231,
+      "step": 4336,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.2111,
+      "step": 4336,
+      "vm_loss": 0.2329
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.8987,
+      "step": 4336,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.567,
+      "step": 4336,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.356,
+      "step": 4336,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.8347089539668407,
+      "lm_loss": 1.0883,
+      "step": 4336,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 0.8349014606444161,
+      "grad_norm": 3.0634122335302343,
+      "learning_rate": 1.310617815423938e-05,
+      "loss": 1.679,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8350939673219915,
+      "grad_norm": 3.163017311305193,
+      "learning_rate": 1.3103214333775522e-05,
+      "loss": 1.6907,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8352864739995669,
+      "grad_norm": 3.1631439299909396,
+      "learning_rate": 1.3100250211639527e-05,
+      "loss": 1.685,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8354789806771422,
+      "grad_norm": 2.970612701036097,
+      "learning_rate": 1.309728578811955e-05,
+      "loss": 1.5722,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8356714873547176,
+      "grad_norm": 3.166437765091976,
+      "learning_rate": 1.3094321063503768e-05,
+      "loss": 1.7154,
+      "step": 4341
+    },
+    {
+      "epoch": 0.835863994032293,
+      "grad_norm": 3.0212925378787374,
+      "learning_rate": 1.309135603808039e-05,
+      "loss": 1.6008,
+      "step": 4342
+    },
+    {
+      "epoch": 0.8360565007098684,
+      "grad_norm": 3.231279639494313,
+      "learning_rate": 1.3088390712137656e-05,
+      "loss": 1.6666,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "grad_norm": 2.9996254254066503,
+      "learning_rate": 1.308542508596383e-05,
+      "loss": 1.6851,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.5815,
+      "step": 4344,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.2942,
+      "step": 4344,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.1068,
+      "step": 4344,
+      "vm_loss": 0.1301
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.8251,
+      "step": 4344,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.4435,
+      "step": 4344,
+      "vm_loss": 0.1083
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 2.0858,
+      "step": 4344,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.4153,
+      "step": 4344,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.8362490073874438,
+      "lm_loss": 1.0147,
+      "step": 4344,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.8364415140650191,
+      "grad_norm": 2.9387825127261356,
+      "learning_rate": 1.3082459159847215e-05,
+      "loss": 1.6416,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8366340207425945,
+      "grad_norm": 3.252453269638057,
+      "learning_rate": 1.307949293407613e-05,
+      "loss": 1.7184,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8368265274201699,
+      "grad_norm": 3.0950856828521642,
+      "learning_rate": 1.3076526408938937e-05,
+      "loss": 1.7215,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8370190340977453,
+      "grad_norm": 3.0720351857356163,
+      "learning_rate": 1.3073559584724014e-05,
+      "loss": 1.7436,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8372115407753207,
+      "grad_norm": 2.9544535763763604,
+      "learning_rate": 1.3070592461719778e-05,
+      "loss": 1.7351,
+      "step": 4349
+    },
+    {
+      "epoch": 0.837404047452896,
+      "grad_norm": 2.9724622137453083,
+      "learning_rate": 1.3067625040214673e-05,
+      "loss": 1.6959,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8375965541304714,
+      "grad_norm": 3.1319990036676733,
+      "learning_rate": 1.3064657320497165e-05,
+      "loss": 1.706,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "grad_norm": 2.880729076499695,
+      "learning_rate": 1.3061689302855758e-05,
+      "loss": 1.7127,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.0865,
+      "step": 4352,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 2.0064,
+      "step": 4352,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.1778,
+      "step": 4352,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.3265,
+      "step": 4352,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.6751,
+      "step": 4352,
+      "vm_loss": 0.2303
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.9797,
+      "step": 4352,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.3836,
+      "step": 4352,
+      "vm_loss": 0.1184
+    },
+    {
+      "epoch": 0.8377890608080468,
+      "lm_loss": 1.9283,
+      "step": 4352,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.8379815674856221,
+      "grad_norm": 3.0943276550785894,
+      "learning_rate": 1.3058720987578978e-05,
+      "loss": 1.72,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8381740741631976,
+      "grad_norm": 3.0955715215633486,
+      "learning_rate": 1.3055752374955389e-05,
+      "loss": 1.6804,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8383665808407729,
+      "grad_norm": 3.236203418403068,
+      "learning_rate": 1.3052783465273573e-05,
+      "loss": 1.7305,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8385590875183483,
+      "grad_norm": 3.0711202566535984,
+      "learning_rate": 1.3049814258822147e-05,
+      "loss": 1.6848,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8387515941959237,
+      "grad_norm": 3.219756738286842,
+      "learning_rate": 1.3046844755889758e-05,
+      "loss": 1.7334,
+      "step": 4357
+    },
+    {
+      "epoch": 0.838944100873499,
+      "grad_norm": 3.1958736261796847,
+      "learning_rate": 1.3043874956765077e-05,
+      "loss": 1.7674,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8391366075510744,
+      "grad_norm": 2.9699152078715993,
+      "learning_rate": 1.3040904861736809e-05,
+      "loss": 1.7144,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "grad_norm": 3.2269581974200032,
+      "learning_rate": 1.3037934471093683e-05,
+      "loss": 1.712,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.3242,
+      "step": 4360,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.4208,
+      "step": 4360,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.3043,
+      "step": 4360,
+      "vm_loss": 0.1132
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.5691,
+      "step": 4360,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.0469,
+      "step": 4360,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.6433,
+      "step": 4360,
+      "vm_loss": 0.2061
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.6181,
+      "step": 4360,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 0.8393291142286498,
+      "lm_loss": 1.5244,
+      "step": 4360,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 0.8395216209062252,
+      "grad_norm": 3.0851228569188915,
+      "learning_rate": 1.303496378512446e-05,
+      "loss": 1.6432,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8397141275838006,
+      "grad_norm": 3.0392999551247155,
+      "learning_rate": 1.303199280411793e-05,
+      "loss": 1.6857,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8399066342613759,
+      "grad_norm": 3.1267006693387573,
+      "learning_rate": 1.3029021528362906e-05,
+      "loss": 1.6084,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8400991409389513,
+      "grad_norm": 3.4465845069229832,
+      "learning_rate": 1.3026049958148236e-05,
+      "loss": 1.7086,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8402916476165267,
+      "grad_norm": 3.036159481809941,
+      "learning_rate": 1.3023078093762802e-05,
+      "loss": 1.7075,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8404841542941021,
+      "grad_norm": 2.9602071606780065,
+      "learning_rate": 1.3020105935495496e-05,
+      "loss": 1.6513,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8406766609716775,
+      "grad_norm": 3.168469049745241,
+      "learning_rate": 1.3017133483635253e-05,
+      "loss": 1.7288,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "grad_norm": 3.117941451459316,
+      "learning_rate": 1.3014160738471037e-05,
+      "loss": 1.6945,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.5681,
+      "step": 4368,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.3728,
+      "step": 4368,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.4545,
+      "step": 4368,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.5959,
+      "step": 4368,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.4277,
+      "step": 4368,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.0381,
+      "step": 4368,
+      "vm_loss": 0.093
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.368,
+      "step": 4368,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 0.8408691676492528,
+      "lm_loss": 1.4613,
+      "step": 4368,
+      "vm_loss": 0.165
+    },
+    {
+      "epoch": 0.8410616743268282,
+      "grad_norm": 3.1687155861974152,
+      "learning_rate": 1.301118770029184e-05,
+      "loss": 1.7255,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8412541810044036,
+      "grad_norm": 3.020417425428901,
+      "learning_rate": 1.3008214369386667e-05,
+      "loss": 1.7248,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8414466876819789,
+      "grad_norm": 3.15466560116373,
+      "learning_rate": 1.3005240746044572e-05,
+      "loss": 1.7554,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8416391943595544,
+      "grad_norm": 3.133746879438481,
+      "learning_rate": 1.300226683055463e-05,
+      "loss": 1.6209,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8418317010371297,
+      "grad_norm": 2.9935772664744937,
+      "learning_rate": 1.2999292623205942e-05,
+      "loss": 1.7297,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8420242077147051,
+      "grad_norm": 3.114097538564018,
+      "learning_rate": 1.2996318124287638e-05,
+      "loss": 1.7117,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8422167143922805,
+      "grad_norm": 2.9472148398573204,
+      "learning_rate": 1.2993343334088879e-05,
+      "loss": 1.7152,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "grad_norm": 3.1484104763609326,
+      "learning_rate": 1.299036825289885e-05,
+      "loss": 1.7502,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.9482,
+      "step": 4376,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.4684,
+      "step": 4376,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.6806,
+      "step": 4376,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.6284,
+      "step": 4376,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.7249,
+      "step": 4376,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 2.0307,
+      "step": 4376,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.7772,
+      "step": 4376,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 0.8424092210698558,
+      "lm_loss": 1.747,
+      "step": 4376,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 0.8426017277474313,
+      "grad_norm": 3.057821633445439,
+      "learning_rate": 1.2987392881006773e-05,
+      "loss": 1.6721,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8427942344250067,
+      "grad_norm": 2.9228310185035684,
+      "learning_rate": 1.2984417218701883e-05,
+      "loss": 1.6497,
+      "step": 4378
+    },
+    {
+      "epoch": 0.842986741102582,
+      "grad_norm": 2.9792571412327282,
+      "learning_rate": 1.2981441266273462e-05,
+      "loss": 1.6332,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8431792477801574,
+      "grad_norm": 3.1047203335827795,
+      "learning_rate": 1.2978465024010803e-05,
+      "loss": 1.6713,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8433717544577327,
+      "grad_norm": 3.046873451531447,
+      "learning_rate": 1.2975488492203241e-05,
+      "loss": 1.6209,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8435642611353081,
+      "grad_norm": 3.0713850487777656,
+      "learning_rate": 1.2972511671140127e-05,
+      "loss": 1.6902,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8437567678128836,
+      "grad_norm": 3.5186504370725973,
+      "learning_rate": 1.2969534561110848e-05,
+      "loss": 1.7662,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "grad_norm": 2.9124896326540086,
+      "learning_rate": 1.2966557162404823e-05,
+      "loss": 1.6847,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.5537,
+      "step": 4384,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.4561,
+      "step": 4384,
+      "vm_loss": 0.1327
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.2491,
+      "step": 4384,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 2.3442,
+      "step": 4384,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.5548,
+      "step": 4384,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.2072,
+      "step": 4384,
+      "vm_loss": 0.088
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.6889,
+      "step": 4384,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 0.8439492744904589,
+      "lm_loss": 1.057,
+      "step": 4384,
+      "vm_loss": 0.2097
+    },
+    {
+      "epoch": 0.8441417811680343,
+      "grad_norm": 3.112590158695623,
+      "learning_rate": 1.2963579475311485e-05,
+      "loss": 1.6484,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8443342878456096,
+      "grad_norm": 3.062370038617411,
+      "learning_rate": 1.2960601500120307e-05,
+      "loss": 1.7194,
+      "step": 4386
+    },
+    {
+      "epoch": 0.844526794523185,
+      "grad_norm": 3.1758525210826174,
+      "learning_rate": 1.2957623237120784e-05,
+      "loss": 1.7476,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8447193012007604,
+      "grad_norm": 3.1485560540709527,
+      "learning_rate": 1.2954644686602449e-05,
+      "loss": 1.7309,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8449118078783358,
+      "grad_norm": 3.310210577452421,
+      "learning_rate": 1.2951665848854845e-05,
+      "loss": 1.7006,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8451043145559112,
+      "grad_norm": 3.105159334962832,
+      "learning_rate": 1.2948686724167559e-05,
+      "loss": 1.6738,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8452968212334865,
+      "grad_norm": 3.003040600761736,
+      "learning_rate": 1.2945707312830197e-05,
+      "loss": 1.6846,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "grad_norm": 2.914091679808134,
+      "learning_rate": 1.29427276151324e-05,
+      "loss": 1.7869,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.3304,
+      "step": 4392,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.5971,
+      "step": 4392,
+      "vm_loss": 0.1559
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.6343,
+      "step": 4392,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.1903,
+      "step": 4392,
+      "vm_loss": 0.108
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.493,
+      "step": 4392,
+      "vm_loss": 0.1283
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.6251,
+      "step": 4392,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.8112,
+      "step": 4392,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 0.8454893279110619,
+      "lm_loss": 1.5236,
+      "step": 4392,
+      "vm_loss": 0.0975
+    },
+    {
+      "epoch": 0.8456818345886373,
+      "grad_norm": 3.1125788233275964,
+      "learning_rate": 1.293974763136383e-05,
+      "loss": 1.6903,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8458743412662126,
+      "grad_norm": 3.179429058715911,
+      "learning_rate": 1.293676736181418e-05,
+      "loss": 1.7269,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8460668479437881,
+      "grad_norm": 2.959998430945532,
+      "learning_rate": 1.293378680677317e-05,
+      "loss": 1.7335,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8462593546213635,
+      "grad_norm": 3.1487827553199015,
+      "learning_rate": 1.2930805966530548e-05,
+      "loss": 1.7915,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8464518612989388,
+      "grad_norm": 3.0411303880509957,
+      "learning_rate": 1.292782484137609e-05,
+      "loss": 1.7142,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8466443679765142,
+      "grad_norm": 3.080035179409326,
+      "learning_rate": 1.29248434315996e-05,
+      "loss": 1.6986,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8468368746540895,
+      "grad_norm": 3.0322106595129226,
+      "learning_rate": 1.2921861737490908e-05,
+      "loss": 1.7125,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "grad_norm": 3.002032605090812,
+      "learning_rate": 1.2918879759339874e-05,
+      "loss": 1.6556,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.6954,
+      "step": 4400,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.1661,
+      "step": 4400,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.6242,
+      "step": 4400,
+      "vm_loss": 0.2181
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.3823,
+      "step": 4400,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.5291,
+      "step": 4400,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.3078,
+      "step": 4400,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.3767,
+      "step": 4400,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.8470293813316649,
+      "lm_loss": 1.4285,
+      "step": 4400,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 0.8472218880092404,
+      "grad_norm": 3.028198960753014,
+      "learning_rate": 1.2915897497436385e-05,
+      "loss": 1.7108,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8474143946868157,
+      "grad_norm": 3.102521513788809,
+      "learning_rate": 1.2912914952070352e-05,
+      "loss": 1.7153,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8476069013643911,
+      "grad_norm": 3.038183005438477,
+      "learning_rate": 1.2909932123531721e-05,
+      "loss": 1.726,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8477994080419664,
+      "grad_norm": 2.997786049380668,
+      "learning_rate": 1.2906949012110456e-05,
+      "loss": 1.6443,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8479919147195418,
+      "grad_norm": 3.065878758309054,
+      "learning_rate": 1.2903965618096557e-05,
+      "loss": 1.6612,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8481844213971172,
+      "grad_norm": 3.108680060707114,
+      "learning_rate": 1.2900981941780049e-05,
+      "loss": 1.694,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8483769280746926,
+      "grad_norm": 3.0741516778288434,
+      "learning_rate": 1.289799798345098e-05,
+      "loss": 1.7056,
+      "step": 4407
+    },
+    {
+      "epoch": 0.848569434752268,
+      "grad_norm": 2.8878425363227334,
+      "learning_rate": 1.2895013743399432e-05,
+      "loss": 1.6273,
+      "step": 4408
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.2965,
+      "step": 4408,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.7567,
+      "step": 4408,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.3998,
+      "step": 4408,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.0901,
+      "step": 4408,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.4649,
+      "step": 4408,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.6219,
+      "step": 4408,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.1227,
+      "step": 4408,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 0.848569434752268,
+      "lm_loss": 1.0629,
+      "step": 4408,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 0.8487619414298434,
+      "grad_norm": 3.0432887729568208,
+      "learning_rate": 1.2892029221915508e-05,
+      "loss": 1.6767,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8489544481074187,
+      "grad_norm": 2.9597448142345173,
+      "learning_rate": 1.2889044419289347e-05,
+      "loss": 1.6346,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8491469547849941,
+      "grad_norm": 2.9543102717920604,
+      "learning_rate": 1.2886059335811106e-05,
+      "loss": 1.6848,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8493394614625694,
+      "grad_norm": 2.9453317598785005,
+      "learning_rate": 1.2883073971770971e-05,
+      "loss": 1.6299,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8495319681401449,
+      "grad_norm": 3.1508828939284115,
+      "learning_rate": 1.2880088327459164e-05,
+      "loss": 1.6633,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8497244748177203,
+      "grad_norm": 2.9977843801287003,
+      "learning_rate": 1.2877102403165922e-05,
+      "loss": 1.7089,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8499169814952956,
+      "grad_norm": 3.1787254448781677,
+      "learning_rate": 1.2874116199181522e-05,
+      "loss": 1.6849,
+      "step": 4415
+    },
+    {
+      "epoch": 0.850109488172871,
+      "grad_norm": 3.026261071217002,
+      "learning_rate": 1.2871129715796253e-05,
+      "loss": 1.6622,
+      "step": 4416
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.2626,
+      "step": 4416,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.1737,
+      "step": 4416,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.3827,
+      "step": 4416,
+      "vm_loss": 0.2169
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.6548,
+      "step": 4416,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.4932,
+      "step": 4416,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.5004,
+      "step": 4416,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.6964,
+      "step": 4416,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.850109488172871,
+      "lm_loss": 1.5854,
+      "step": 4416,
+      "vm_loss": 0.21
+    },
+    {
+      "epoch": 0.8503019948504463,
+      "grad_norm": 2.967167164075282,
+      "learning_rate": 1.2868142953300442e-05,
+      "loss": 1.6123,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8504945015280218,
+      "grad_norm": 3.055304588664331,
+      "learning_rate": 1.2865155911984445e-05,
+      "loss": 1.6461,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8506870082055972,
+      "grad_norm": 2.992944752205052,
+      "learning_rate": 1.2862168592138635e-05,
+      "loss": 1.7252,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8508795148831725,
+      "grad_norm": 3.0834994495829675,
+      "learning_rate": 1.2859180994053424e-05,
+      "loss": 1.6663,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8510720215607479,
+      "grad_norm": 2.961857124765966,
+      "learning_rate": 1.2856193118019235e-05,
+      "loss": 1.6574,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8512645282383232,
+      "grad_norm": 3.0558728721543202,
+      "learning_rate": 1.2853204964326536e-05,
+      "loss": 1.6458,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8514570349158986,
+      "grad_norm": 3.03977260308226,
+      "learning_rate": 1.2850216533265814e-05,
+      "loss": 1.6869,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "grad_norm": 2.9482612800103607,
+      "learning_rate": 1.2847227825127578e-05,
+      "loss": 1.6941,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.7082,
+      "step": 4424,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.6234,
+      "step": 4424,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.8249,
+      "step": 4424,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.7877,
+      "step": 4424,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.4533,
+      "step": 4424,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.3378,
+      "step": 4424,
+      "vm_loss": 0.1233
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.687,
+      "step": 4424,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.8516495415934741,
+      "lm_loss": 1.2291,
+      "step": 4424,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 0.8518420482710494,
+      "grad_norm": 3.050494491986149,
+      "learning_rate": 1.284423884020237e-05,
+      "loss": 1.7177,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8520345549486248,
+      "grad_norm": 2.9041037013114965,
+      "learning_rate": 1.2841249578780756e-05,
+      "loss": 1.6738,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8522270616262002,
+      "grad_norm": 3.0444074352669577,
+      "learning_rate": 1.2838260041153339e-05,
+      "loss": 1.6578,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8524195683037755,
+      "grad_norm": 2.9682919905362106,
+      "learning_rate": 1.283527022761073e-05,
+      "loss": 1.7867,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8526120749813509,
+      "grad_norm": 3.0700905398642218,
+      "learning_rate": 1.283228013844358e-05,
+      "loss": 1.6722,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8528045816589263,
+      "grad_norm": 3.055669392950635,
+      "learning_rate": 1.2829289773942567e-05,
+      "loss": 1.6533,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8529970883365017,
+      "grad_norm": 3.119651451324948,
+      "learning_rate": 1.2826299134398387e-05,
+      "loss": 1.6535,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "grad_norm": 3.0988419364495665,
+      "learning_rate": 1.2823308220101777e-05,
+      "loss": 1.807,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.3772,
+      "step": 4432,
+      "vm_loss": 0.2209
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.2338,
+      "step": 4432,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.5472,
+      "step": 4432,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.5292,
+      "step": 4432,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.5988,
+      "step": 4432,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.8606,
+      "step": 4432,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 1.2638,
+      "step": 4432,
+      "vm_loss": 0.2367
+    },
+    {
+      "epoch": 0.8531895950140771,
+      "lm_loss": 0.9159,
+      "step": 4432,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 0.8533821016916524,
+      "grad_norm": 3.0419804980930594,
+      "learning_rate": 1.2820317031343485e-05,
+      "loss": 1.6427,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8535746083692278,
+      "grad_norm": 2.8795738351575393,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 1.6269,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8537671150468031,
+      "grad_norm": 2.992128438235043,
+      "learning_rate": 1.2814333831605019e-05,
+      "loss": 1.7017,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8539596217243786,
+      "grad_norm": 3.000331001767118,
+      "learning_rate": 1.2811341821206487e-05,
+      "loss": 1.6946,
+      "step": 4436
+    },
+    {
+      "epoch": 0.854152128401954,
+      "grad_norm": 3.074197489173797,
+      "learning_rate": 1.2808349537509564e-05,
+      "loss": 1.6514,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8543446350795293,
+      "grad_norm": 3.158141085384541,
+      "learning_rate": 1.2805356980805132e-05,
+      "loss": 1.661,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8545371417571047,
+      "grad_norm": 3.161881857328982,
+      "learning_rate": 1.2802364151384116e-05,
+      "loss": 1.6844,
+      "step": 4439
+    },
+    {
+      "epoch": 0.85472964843468,
+      "grad_norm": 3.1013498278455054,
+      "learning_rate": 1.2799371049537448e-05,
+      "loss": 1.6821,
+      "step": 4440
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.4488,
+      "step": 4440,
+      "vm_loss": 0.0951
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.118,
+      "step": 4440,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.933,
+      "step": 4440,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 0.9792,
+      "step": 4440,
+      "vm_loss": 0.1073
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.7091,
+      "step": 4440,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.3559,
+      "step": 4440,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 1.5367,
+      "step": 4440,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 0.85472964843468,
+      "lm_loss": 0.9572,
+      "step": 4440,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.8549221551122554,
+      "grad_norm": 2.9380934733636392,
+      "learning_rate": 1.27963776755561e-05,
+      "loss": 1.6732,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8551146617898309,
+      "grad_norm": 2.9949646721292864,
+      "learning_rate": 1.2793384029731066e-05,
+      "loss": 1.7065,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8553071684674062,
+      "grad_norm": 3.0680098851192463,
+      "learning_rate": 1.2790390112353369e-05,
+      "loss": 1.712,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8554996751449816,
+      "grad_norm": 3.091282238202021,
+      "learning_rate": 1.2787395923714049e-05,
+      "loss": 1.7368,
+      "step": 4444
+    },
+    {
+      "epoch": 0.855692181822557,
+      "grad_norm": 2.96441454518578,
+      "learning_rate": 1.2784401464104189e-05,
+      "loss": 1.6559,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8558846885001323,
+      "grad_norm": 3.334169924401074,
+      "learning_rate": 1.2781406733814879e-05,
+      "loss": 1.7048,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8560771951777078,
+      "grad_norm": 3.17574344447504,
+      "learning_rate": 1.2778411733137253e-05,
+      "loss": 1.698,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "grad_norm": 3.0851685450318844,
+      "learning_rate": 1.2775416462362458e-05,
+      "loss": 1.7089,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.2059,
+      "step": 4448,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.2959,
+      "step": 4448,
+      "vm_loss": 0.2538
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.36,
+      "step": 4448,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.4915,
+      "step": 4448,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.697,
+      "step": 4448,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.7197,
+      "step": 4448,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.8211,
+      "step": 4448,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 0.8562697018552831,
+      "lm_loss": 1.2567,
+      "step": 4448,
+      "vm_loss": 0.2277
+    },
+    {
+      "epoch": 0.8564622085328585,
+      "grad_norm": 2.986701654283633,
+      "learning_rate": 1.2772420921781677e-05,
+      "loss": 1.6713,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8566547152104339,
+      "grad_norm": 2.9389940716368033,
+      "learning_rate": 1.276942511168611e-05,
+      "loss": 1.6625,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8568472218880092,
+      "grad_norm": 3.0107046795792947,
+      "learning_rate": 1.2766429032366994e-05,
+      "loss": 1.6501,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8570397285655846,
+      "grad_norm": 3.064247953908551,
+      "learning_rate": 1.2763432684115582e-05,
+      "loss": 1.6622,
+      "step": 4452
+    },
+    {
+      "epoch": 0.85723223524316,
+      "grad_norm": 3.106108831727385,
+      "learning_rate": 1.2760436067223156e-05,
+      "loss": 1.69,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8574247419207354,
+      "grad_norm": 2.9829775742102895,
+      "learning_rate": 1.2757439181981033e-05,
+      "loss": 1.6909,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8576172485983108,
+      "grad_norm": 3.0164088771005284,
+      "learning_rate": 1.275444202868054e-05,
+      "loss": 1.7258,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "grad_norm": 3.015194334952046,
+      "learning_rate": 1.2751444607613044e-05,
+      "loss": 1.6578,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.5486,
+      "step": 4456,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.3585,
+      "step": 4456,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.7845,
+      "step": 4456,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.6806,
+      "step": 4456,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.7309,
+      "step": 4456,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.5521,
+      "step": 4456,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.2181,
+      "step": 4456,
+      "vm_loss": 0.2328
+    },
+    {
+      "epoch": 0.8578097552758861,
+      "lm_loss": 1.3713,
+      "step": 4456,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.8580022619534615,
+      "grad_norm": 2.9504431217016114,
+      "learning_rate": 1.274844691906993e-05,
+      "loss": 1.7202,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8581947686310369,
+      "grad_norm": 2.9642851500204817,
+      "learning_rate": 1.2745448963342612e-05,
+      "loss": 1.5875,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8583872753086123,
+      "grad_norm": 3.2369529928116125,
+      "learning_rate": 1.2742450740722532e-05,
+      "loss": 1.6929,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8585797819861877,
+      "grad_norm": 3.1381033212467173,
+      "learning_rate": 1.2739452251501152e-05,
+      "loss": 1.6784,
+      "step": 4460
+    },
+    {
+      "epoch": 0.858772288663763,
+      "grad_norm": 3.0593129256530194,
+      "learning_rate": 1.2736453495969967e-05,
+      "loss": 1.632,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8589647953413384,
+      "grad_norm": 3.089659793123969,
+      "learning_rate": 1.2733454474420493e-05,
+      "loss": 1.6274,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8591573020189138,
+      "grad_norm": 3.066609905530649,
+      "learning_rate": 1.2730455187144269e-05,
+      "loss": 1.6699,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "grad_norm": 3.0534938405455674,
+      "learning_rate": 1.2727455634432873e-05,
+      "loss": 1.6807,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.2368,
+      "step": 4464,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.3027,
+      "step": 4464,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.1642,
+      "step": 4464,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.4768,
+      "step": 4464,
+      "vm_loss": 0.2516
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 0.9715,
+      "step": 4464,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.4413,
+      "step": 4464,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.6963,
+      "step": 4464,
+      "vm_loss": 0.209
+    },
+    {
+      "epoch": 0.8593498086964891,
+      "lm_loss": 1.7286,
+      "step": 4464,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 0.8595423153740646,
+      "grad_norm": 3.0154793954165835,
+      "learning_rate": 1.2724455816577891e-05,
+      "loss": 1.6409,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8597348220516399,
+      "grad_norm": 2.9078586481620374,
+      "learning_rate": 1.2721455733870952e-05,
+      "loss": 1.7146,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8599273287292153,
+      "grad_norm": 3.078662308753128,
+      "learning_rate": 1.2718455386603695e-05,
+      "loss": 1.7405,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8601198354067907,
+      "grad_norm": 3.034324836966756,
+      "learning_rate": 1.2715454775067798e-05,
+      "loss": 1.7353,
+      "step": 4468
+    },
+    {
+      "epoch": 0.860312342084366,
+      "grad_norm": 2.9498075382587996,
+      "learning_rate": 1.271245389955495e-05,
+      "loss": 1.7237,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8605048487619414,
+      "grad_norm": 3.0301071703289217,
+      "learning_rate": 1.2709452760356884e-05,
+      "loss": 1.6602,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8606973554395168,
+      "grad_norm": 3.0409535418330638,
+      "learning_rate": 1.2706451357765347e-05,
+      "loss": 1.6519,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "grad_norm": 3.040719003145782,
+      "learning_rate": 1.2703449692072112e-05,
+      "loss": 1.7106,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.6528,
+      "step": 4472,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.3673,
+      "step": 4472,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.6238,
+      "step": 4472,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.6757,
+      "step": 4472,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.7708,
+      "step": 4472,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.2154,
+      "step": 4472,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.5934,
+      "step": 4472,
+      "vm_loss": 0.2166
+    },
+    {
+      "epoch": 0.8608898621170922,
+      "lm_loss": 1.2965,
+      "step": 4472,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.8610823687946676,
+      "grad_norm": 2.9392567593614833,
+      "learning_rate": 1.2700447763568977e-05,
+      "loss": 1.6685,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8612748754722429,
+      "grad_norm": 3.0407914173970685,
+      "learning_rate": 1.2697445572547773e-05,
+      "loss": 1.72,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8614673821498183,
+      "grad_norm": 3.1670146643943125,
+      "learning_rate": 1.2694443119300348e-05,
+      "loss": 1.6622,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8616598888273937,
+      "grad_norm": 2.9350261507142448,
+      "learning_rate": 1.2691440404118582e-05,
+      "loss": 1.6374,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8618523955049691,
+      "grad_norm": 3.016237493766547,
+      "learning_rate": 1.2688437427294371e-05,
+      "loss": 1.6382,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8620449021825445,
+      "grad_norm": 3.0005654231300007,
+      "learning_rate": 1.2685434189119652e-05,
+      "loss": 1.6675,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8622374088601198,
+      "grad_norm": 3.1857688717745827,
+      "learning_rate": 1.268243068988637e-05,
+      "loss": 1.6929,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "grad_norm": 3.1171575576813715,
+      "learning_rate": 1.2679426929886507e-05,
+      "loss": 1.682,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.5832,
+      "step": 4480,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.8804,
+      "step": 4480,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.5709,
+      "step": 4480,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 2.0507,
+      "step": 4480,
+      "vm_loss": 0.2352
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 2.0701,
+      "step": 4480,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.1112,
+      "step": 4480,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.2841,
+      "step": 4480,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 0.8624299155376952,
+      "lm_loss": 1.8705,
+      "step": 4480,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 0.8626224222152706,
+      "grad_norm": 3.0897309797569377,
+      "learning_rate": 1.2676422909412067e-05,
+      "loss": 1.6388,
+      "step": 4481
+    },
+    {
+      "epoch": 0.862814928892846,
+      "grad_norm": 3.3078437971865875,
+      "learning_rate": 1.2673418628755078e-05,
+      "loss": 1.806,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8630074355704214,
+      "grad_norm": 3.094966719813121,
+      "learning_rate": 1.2670414088207597e-05,
+      "loss": 1.7651,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8631999422479967,
+      "grad_norm": 2.9244838672595446,
+      "learning_rate": 1.26674092880617e-05,
+      "loss": 1.6751,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8633924489255721,
+      "grad_norm": 3.03126661454168,
+      "learning_rate": 1.2664404228609498e-05,
+      "loss": 1.6308,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8635849556031475,
+      "grad_norm": 3.058533701565695,
+      "learning_rate": 1.2661398910143112e-05,
+      "loss": 1.6744,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8637774622807228,
+      "grad_norm": 3.086312651540744,
+      "learning_rate": 1.2658393332954704e-05,
+      "loss": 1.6787,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "grad_norm": 3.06225690621572,
+      "learning_rate": 1.2655387497336456e-05,
+      "loss": 1.6254,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.1454,
+      "step": 4488,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.9375,
+      "step": 4488,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.9362,
+      "step": 4488,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.9991,
+      "step": 4488,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.9301,
+      "step": 4488,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.5812,
+      "step": 4488,
+      "vm_loss": 0.2221
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.355,
+      "step": 4488,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.8639699689582983,
+      "lm_loss": 1.3635,
+      "step": 4488,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 0.8641624756358737,
+      "grad_norm": 3.2148140541238646,
+      "learning_rate": 1.265238140358057e-05,
+      "loss": 1.7138,
+      "step": 4489
+    },
+    {
+      "epoch": 0.864354982313449,
+      "grad_norm": 3.017382667130634,
+      "learning_rate": 1.2649375051979276e-05,
+      "loss": 1.6194,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8645474889910244,
+      "grad_norm": 3.0050034680899467,
+      "learning_rate": 1.2646368442824832e-05,
+      "loss": 1.707,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8647399956685997,
+      "grad_norm": 3.020961298784208,
+      "learning_rate": 1.2643361576409517e-05,
+      "loss": 1.6307,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8649325023461751,
+      "grad_norm": 2.8863948839624225,
+      "learning_rate": 1.264035445302564e-05,
+      "loss": 1.6951,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8651250090237506,
+      "grad_norm": 2.9371411689720572,
+      "learning_rate": 1.263734707296553e-05,
+      "loss": 1.6998,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8653175157013259,
+      "grad_norm": 3.0191043075879636,
+      "learning_rate": 1.2634339436521543e-05,
+      "loss": 1.6333,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "grad_norm": 3.0311774510797136,
+      "learning_rate": 1.263133154398606e-05,
+      "loss": 1.7016,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 1.3431,
+      "step": 4496,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 0.9305,
+      "step": 4496,
+      "vm_loss": 0.1038
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 1.4762,
+      "step": 4496,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 0.8536,
+      "step": 4496,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 1.2865,
+      "step": 4496,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 1.4997,
+      "step": 4496,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 1.7362,
+      "step": 4496,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 0.8655100223789013,
+      "lm_loss": 0.8027,
+      "step": 4496,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 0.8657025290564766,
+      "grad_norm": 3.032642485101621,
+      "learning_rate": 1.2628323395651485e-05,
+      "loss": 1.6102,
+      "step": 4497
+    },
+    {
+      "epoch": 0.865895035734052,
+      "grad_norm": 3.1145686090229643,
+      "learning_rate": 1.2625314991810254e-05,
+      "loss": 1.6627,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8660875424116274,
+      "grad_norm": 3.1790364780462492,
+      "learning_rate": 1.2622306332754812e-05,
+      "loss": 1.7303,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8662800490892028,
+      "grad_norm": 3.0131872076240596,
+      "learning_rate": 1.2619297418777651e-05,
+      "loss": 1.7053,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8664725557667782,
+      "grad_norm": 2.9860416847197535,
+      "learning_rate": 1.2616288250171268e-05,
+      "loss": 1.7084,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8666650624443535,
+      "grad_norm": 3.018242603209765,
+      "learning_rate": 1.2613278827228197e-05,
+      "loss": 1.6736,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8668575691219289,
+      "grad_norm": 3.0926621651436963,
+      "learning_rate": 1.2610269150240985e-05,
+      "loss": 1.6979,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "grad_norm": 3.0550182199135905,
+      "learning_rate": 1.2607259219502222e-05,
+      "loss": 1.6118,
+      "step": 4504
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.1916,
+      "step": 4504,
+      "vm_loss": 0.1236
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.8031,
+      "step": 4504,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.8845,
+      "step": 4504,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.3913,
+      "step": 4504,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.7355,
+      "step": 4504,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 0.8855,
+      "step": 4504,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.6043,
+      "step": 4504,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 0.8670500757995043,
+      "lm_loss": 1.8649,
+      "step": 4504,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.8672425824770796,
+      "grad_norm": 3.040848271698335,
+      "learning_rate": 1.2604249035304505e-05,
+      "loss": 1.683,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8674350891546551,
+      "grad_norm": 3.2964680800870596,
+      "learning_rate": 1.2601238597940462e-05,
+      "loss": 1.6255,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8676275958322305,
+      "grad_norm": 3.2444759612907106,
+      "learning_rate": 1.2598227907702749e-05,
+      "loss": 1.6381,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8678201025098058,
+      "grad_norm": 3.0614446983141534,
+      "learning_rate": 1.2595216964884043e-05,
+      "loss": 1.7674,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8680126091873812,
+      "grad_norm": 3.05410010044593,
+      "learning_rate": 1.2592205769777044e-05,
+      "loss": 1.7364,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8682051158649565,
+      "grad_norm": 3.0570537511777602,
+      "learning_rate": 1.258919432267448e-05,
+      "loss": 1.7035,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8683976225425319,
+      "grad_norm": 3.166324911497316,
+      "learning_rate": 1.2586182623869102e-05,
+      "loss": 1.6489,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "grad_norm": 3.061439217901443,
+      "learning_rate": 1.2583170673653688e-05,
+      "loss": 1.6768,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.7003,
+      "step": 4512,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.0974,
+      "step": 4512,
+      "vm_loss": 0.1254
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.581,
+      "step": 4512,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 2.0384,
+      "step": 4512,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.8247,
+      "step": 4512,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.5041,
+      "step": 4512,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.5306,
+      "step": 4512,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 0.8685901292201074,
+      "lm_loss": 1.4329,
+      "step": 4512,
+      "vm_loss": 0.2135
+    },
+    {
+      "epoch": 0.8687826358976827,
+      "grad_norm": 3.138110276354647,
+      "learning_rate": 1.2580158472321035e-05,
+      "loss": 1.6968,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8689751425752581,
+      "grad_norm": 3.244684812037457,
+      "learning_rate": 1.2577146020163968e-05,
+      "loss": 1.6605,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8691676492528334,
+      "grad_norm": 3.2220205622214824,
+      "learning_rate": 1.2574133317475337e-05,
+      "loss": 1.6659,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8693601559304088,
+      "grad_norm": 3.126389018681553,
+      "learning_rate": 1.2571120364548015e-05,
+      "loss": 1.6461,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8695526626079843,
+      "grad_norm": 3.0977651156155077,
+      "learning_rate": 1.25681071616749e-05,
+      "loss": 1.6534,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8697451692855596,
+      "grad_norm": 2.9236833692927497,
+      "learning_rate": 1.2565093709148913e-05,
+      "loss": 1.6396,
+      "step": 4518
+    },
+    {
+      "epoch": 0.869937675963135,
+      "grad_norm": 2.97654897421359,
+      "learning_rate": 1.2562080007263e-05,
+      "loss": 1.6218,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "grad_norm": 2.981091271799177,
+      "learning_rate": 1.2559066056310132e-05,
+      "loss": 1.6319,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 0.9042,
+      "step": 4520,
+      "vm_loss": 0.1029
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.1927,
+      "step": 4520,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.0412,
+      "step": 4520,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.3148,
+      "step": 4520,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.4818,
+      "step": 4520,
+      "vm_loss": 0.2433
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.555,
+      "step": 4520,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.3902,
+      "step": 4520,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.8701301826407104,
+      "lm_loss": 1.3759,
+      "step": 4520,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.8703226893182857,
+      "grad_norm": 3.114404094929091,
+      "learning_rate": 1.2556051856583302e-05,
+      "loss": 1.6005,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8705151959958611,
+      "grad_norm": 3.0898387301273296,
+      "learning_rate": 1.2553037408375532e-05,
+      "loss": 1.704,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8707077026734364,
+      "grad_norm": 3.416529763203889,
+      "learning_rate": 1.2550022711979861e-05,
+      "loss": 1.7238,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8709002093510119,
+      "grad_norm": 3.1046861715040777,
+      "learning_rate": 1.2547007767689361e-05,
+      "loss": 1.7134,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8710927160285873,
+      "grad_norm": 3.0865199969765102,
+      "learning_rate": 1.2543992575797118e-05,
+      "loss": 1.6264,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8712852227061626,
+      "grad_norm": 3.041407015619365,
+      "learning_rate": 1.254097713659625e-05,
+      "loss": 1.6276,
+      "step": 4526
+    },
+    {
+      "epoch": 0.871477729383738,
+      "grad_norm": 2.97918325936035,
+      "learning_rate": 1.2537961450379896e-05,
+      "loss": 1.6819,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "grad_norm": 3.1665083925795163,
+      "learning_rate": 1.2534945517441217e-05,
+      "loss": 1.705,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.262,
+      "step": 4528,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.6075,
+      "step": 4528,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.837,
+      "step": 4528,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.4592,
+      "step": 4528,
+      "vm_loss": 0.2896
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.9992,
+      "step": 4528,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 1.3408,
+      "step": 4528,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 0.9686,
+      "step": 4528,
+      "vm_loss": 0.0975
+    },
+    {
+      "epoch": 0.8716702360613133,
+      "lm_loss": 0.9556,
+      "step": 4528,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.8718627427388888,
+      "grad_norm": 3.2669378526116875,
+      "learning_rate": 1.2531929338073404e-05,
+      "loss": 1.6542,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8720552494164642,
+      "grad_norm": 3.057730215153544,
+      "learning_rate": 1.2528912912569666e-05,
+      "loss": 1.6103,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8722477560940395,
+      "grad_norm": 3.04827502993877,
+      "learning_rate": 1.252589624122324e-05,
+      "loss": 1.5632,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8724402627716149,
+      "grad_norm": 3.1293689322359013,
+      "learning_rate": 1.2522879324327382e-05,
+      "loss": 1.6759,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8726327694491902,
+      "grad_norm": 3.0097413213092943,
+      "learning_rate": 1.2519862162175377e-05,
+      "loss": 1.6904,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8728252761267656,
+      "grad_norm": 2.9590036337945134,
+      "learning_rate": 1.2516844755060533e-05,
+      "loss": 1.6119,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8730177828043411,
+      "grad_norm": 3.1314755033065547,
+      "learning_rate": 1.2513827103276177e-05,
+      "loss": 1.7016,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "grad_norm": 2.99572618480157,
+      "learning_rate": 1.2510809207115666e-05,
+      "loss": 1.6597,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 1.6401,
+      "step": 4536,
+      "vm_loss": 0.1153
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 0.9012,
+      "step": 4536,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 0.8623,
+      "step": 4536,
+      "vm_loss": 0.1239
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 2.1662,
+      "step": 4536,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 1.6903,
+      "step": 4536,
+      "vm_loss": 0.1063
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 1.0089,
+      "step": 4536,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 1.3925,
+      "step": 4536,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 0.8732102894819164,
+      "lm_loss": 1.4715,
+      "step": 4536,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.8734027961594918,
+      "grad_norm": 2.8922712583563372,
+      "learning_rate": 1.2507791066872377e-05,
+      "loss": 1.6018,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8735953028370672,
+      "grad_norm": 2.995805259893655,
+      "learning_rate": 1.2504772682839713e-05,
+      "loss": 1.6936,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8737878095146425,
+      "grad_norm": 2.909318212187558,
+      "learning_rate": 1.2501754055311098e-05,
+      "loss": 1.6816,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8739803161922179,
+      "grad_norm": 2.9946738866273748,
+      "learning_rate": 1.2498735184579982e-05,
+      "loss": 1.6868,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8741728228697933,
+      "grad_norm": 2.99397166880558,
+      "learning_rate": 1.2495716070939841e-05,
+      "loss": 1.6723,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8743653295473687,
+      "grad_norm": 3.102729386804844,
+      "learning_rate": 1.2492696714684165e-05,
+      "loss": 1.6424,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8745578362249441,
+      "grad_norm": 3.000925337211229,
+      "learning_rate": 1.2489677116106478e-05,
+      "loss": 1.6273,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "grad_norm": 3.1244958420367603,
+      "learning_rate": 1.2486657275500322e-05,
+      "loss": 1.6249,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 2.1406,
+      "step": 4544,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.6883,
+      "step": 4544,
+      "vm_loss": 0.1354
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.7223,
+      "step": 4544,
+      "vm_loss": 0.0714
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.6763,
+      "step": 4544,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.6168,
+      "step": 4544,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.1973,
+      "step": 4544,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.3835,
+      "step": 4544,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 0.8747503429025194,
+      "lm_loss": 1.2531,
+      "step": 4544,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.8749428495800948,
+      "grad_norm": 3.0721203185954367,
+      "learning_rate": 1.2483637193159268e-05,
+      "loss": 1.6778,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8751353562576701,
+      "grad_norm": 3.0051601609676575,
+      "learning_rate": 1.2480616869376903e-05,
+      "loss": 1.6284,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8753278629352456,
+      "grad_norm": 3.0086585252084923,
+      "learning_rate": 1.247759630444684e-05,
+      "loss": 1.6178,
+      "step": 4547
+    },
+    {
+      "epoch": 0.875520369612821,
+      "grad_norm": 2.9493966968582845,
+      "learning_rate": 1.2474575498662722e-05,
+      "loss": 1.6195,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8757128762903963,
+      "grad_norm": 3.073232391234651,
+      "learning_rate": 1.2471554452318201e-05,
+      "loss": 1.669,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8759053829679717,
+      "grad_norm": 3.1985464544251907,
+      "learning_rate": 1.2468533165706971e-05,
+      "loss": 1.6932,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8760978896455471,
+      "grad_norm": 3.0270338607373413,
+      "learning_rate": 1.2465511639122737e-05,
+      "loss": 1.6856,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "grad_norm": 3.1794867515628527,
+      "learning_rate": 1.2462489872859225e-05,
+      "loss": 1.6008,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 0.8905,
+      "step": 4552,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.4828,
+      "step": 4552,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.149,
+      "step": 4552,
+      "vm_loss": 0.1048
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.513,
+      "step": 4552,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.3674,
+      "step": 4552,
+      "vm_loss": 0.1132
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.6657,
+      "step": 4552,
+      "vm_loss": 0.2663
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.4919,
+      "step": 4552,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.8762903963231224,
+      "lm_loss": 1.6948,
+      "step": 4552,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.8764829030006979,
+      "grad_norm": 2.898069556738775,
+      "learning_rate": 1.2459467867210195e-05,
+      "loss": 1.6365,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8766754096782732,
+      "grad_norm": 3.001681802908033,
+      "learning_rate": 1.2456445622469422e-05,
+      "loss": 1.6694,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8768679163558486,
+      "grad_norm": 3.1156492602116796,
+      "learning_rate": 1.2453423138930704e-05,
+      "loss": 1.6596,
+      "step": 4555
+    },
+    {
+      "epoch": 0.877060423033424,
+      "grad_norm": 3.039170441785524,
+      "learning_rate": 1.2450400416887872e-05,
+      "loss": 1.6374,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8772529297109993,
+      "grad_norm": 3.0219659222375914,
+      "learning_rate": 1.244737745663477e-05,
+      "loss": 1.649,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8774454363885748,
+      "grad_norm": 3.08620213967334,
+      "learning_rate": 1.2444354258465268e-05,
+      "loss": 1.6217,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8776379430661501,
+      "grad_norm": 3.103126411561901,
+      "learning_rate": 1.2441330822673257e-05,
+      "loss": 1.6227,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "grad_norm": 3.009849825563307,
+      "learning_rate": 1.2438307149552659e-05,
+      "loss": 1.5811,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.7564,
+      "step": 4560,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.1507,
+      "step": 4560,
+      "vm_loss": 0.2036
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.2605,
+      "step": 4560,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.2455,
+      "step": 4560,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.5245,
+      "step": 4560,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 2.0303,
+      "step": 4560,
+      "vm_loss": 0.1204
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.2958,
+      "step": 4560,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 0.8778304497437255,
+      "lm_loss": 1.6348,
+      "step": 4560,
+      "vm_loss": 0.2421
+    },
+    {
+      "epoch": 0.8780229564213009,
+      "grad_norm": 3.0874447260341302,
+      "learning_rate": 1.2435283239397411e-05,
+      "loss": 1.6386,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8782154630988762,
+      "grad_norm": 2.9737312909336198,
+      "learning_rate": 1.2432259092501474e-05,
+      "loss": 1.5963,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8784079697764516,
+      "grad_norm": 3.01348030067374,
+      "learning_rate": 1.2429234709158837e-05,
+      "loss": 1.6532,
+      "step": 4563
+    },
+    {
+      "epoch": 0.878600476454027,
+      "grad_norm": 3.161406167177918,
+      "learning_rate": 1.2426210089663507e-05,
+      "loss": 1.691,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8787929831316024,
+      "grad_norm": 3.190201401742706,
+      "learning_rate": 1.2423185234309516e-05,
+      "loss": 1.7028,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8789854898091778,
+      "grad_norm": 3.2318175696114064,
+      "learning_rate": 1.2420160143390918e-05,
+      "loss": 1.6397,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8791779964867531,
+      "grad_norm": 3.1098491512355766,
+      "learning_rate": 1.2417134817201792e-05,
+      "loss": 1.6048,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "grad_norm": 3.0554103452449124,
+      "learning_rate": 1.2414109256036237e-05,
+      "loss": 1.7338,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.7922,
+      "step": 4568,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 0.8861,
+      "step": 4568,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.3009,
+      "step": 4568,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.2766,
+      "step": 4568,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.4508,
+      "step": 4568,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.1622,
+      "step": 4568,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.5518,
+      "step": 4568,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 0.8793705031643285,
+      "lm_loss": 1.9879,
+      "step": 4568,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 0.8795630098419039,
+      "grad_norm": 3.049499359721112,
+      "learning_rate": 1.2411083460188378e-05,
+      "loss": 1.6687,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8797555165194793,
+      "grad_norm": 2.900962754963383,
+      "learning_rate": 1.240805742995236e-05,
+      "loss": 1.6325,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8799480231970547,
+      "grad_norm": 3.08890790992242,
+      "learning_rate": 1.240503116562235e-05,
+      "loss": 1.6275,
+      "step": 4571
+    },
+    {
+      "epoch": 0.88014052987463,
+      "grad_norm": 3.0867500115031805,
+      "learning_rate": 1.240200466749254e-05,
+      "loss": 1.6601,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8803330365522054,
+      "grad_norm": 3.145114025357283,
+      "learning_rate": 1.239897793585715e-05,
+      "loss": 1.68,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8805255432297808,
+      "grad_norm": 3.1196334446703586,
+      "learning_rate": 1.2395950971010408e-05,
+      "loss": 1.604,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8807180499073561,
+      "grad_norm": 3.09137320033844,
+      "learning_rate": 1.2392923773246583e-05,
+      "loss": 1.6026,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "grad_norm": 3.3147088740344417,
+      "learning_rate": 1.238989634285995e-05,
+      "loss": 1.6865,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.809,
+      "step": 4576,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.4648,
+      "step": 4576,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.6058,
+      "step": 4576,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.6269,
+      "step": 4576,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.376,
+      "step": 4576,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.7382,
+      "step": 4576,
+      "vm_loss": 0.1007
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.6333,
+      "step": 4576,
+      "vm_loss": 0.2567
+    },
+    {
+      "epoch": 0.8809105565849316,
+      "lm_loss": 1.4625,
+      "step": 4576,
+      "vm_loss": 0.2702
+    },
+    {
+      "epoch": 0.8811030632625069,
+      "grad_norm": 3.19647188730665,
+      "learning_rate": 1.2386868680144816e-05,
+      "loss": 1.7354,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8812955699400823,
+      "grad_norm": 2.928682338156493,
+      "learning_rate": 1.2383840785395511e-05,
+      "loss": 1.5272,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8814880766176577,
+      "grad_norm": 2.9378246150967513,
+      "learning_rate": 1.2380812658906385e-05,
+      "loss": 1.6514,
+      "step": 4579
+    },
+    {
+      "epoch": 0.881680583295233,
+      "grad_norm": 3.1033376933264236,
+      "learning_rate": 1.2377784300971807e-05,
+      "loss": 1.6931,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8818730899728084,
+      "grad_norm": 3.1889888253104086,
+      "learning_rate": 1.2374755711886173e-05,
+      "loss": 1.6055,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8820655966503839,
+      "grad_norm": 3.0050036021158197,
+      "learning_rate": 1.2371726891943905e-05,
+      "loss": 1.5516,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8822581033279592,
+      "grad_norm": 3.1669435906135748,
+      "learning_rate": 1.2368697841439439e-05,
+      "loss": 1.5779,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "grad_norm": 3.182860087338219,
+      "learning_rate": 1.2365668560667237e-05,
+      "loss": 1.6214,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.6302,
+      "step": 4584,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.5885,
+      "step": 4584,
+      "vm_loss": 0.2405
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.5876,
+      "step": 4584,
+      "vm_loss": 0.2252
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.3184,
+      "step": 4584,
+      "vm_loss": 0.2534
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.2578,
+      "step": 4584,
+      "vm_loss": 0.0995
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.3042,
+      "step": 4584,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 0.9598,
+      "step": 4584,
+      "vm_loss": 0.2077
+    },
+    {
+      "epoch": 0.8824506100055346,
+      "lm_loss": 1.0264,
+      "step": 4584,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.8826431166831099,
+      "grad_norm": 3.3881675973074805,
+      "learning_rate": 1.2362639049921785e-05,
+      "loss": 1.6034,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8828356233606853,
+      "grad_norm": 3.3858077264681414,
+      "learning_rate": 1.2359609309497591e-05,
+      "loss": 1.5744,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8830281300382608,
+      "grad_norm": 3.203102517177925,
+      "learning_rate": 1.2356579339689186e-05,
+      "loss": 1.704,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8832206367158361,
+      "grad_norm": 3.1306568130015515,
+      "learning_rate": 1.2353549140791116e-05,
+      "loss": 1.567,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8834131433934115,
+      "grad_norm": 3.2228695926530846,
+      "learning_rate": 1.2350518713097964e-05,
+      "loss": 1.6452,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8836056500709868,
+      "grad_norm": 3.0623422546782244,
+      "learning_rate": 1.2347488056904316e-05,
+      "loss": 1.5413,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8837981567485622,
+      "grad_norm": 3.2335550336654526,
+      "learning_rate": 1.2344457172504796e-05,
+      "loss": 1.697,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "grad_norm": 3.0967465520611213,
+      "learning_rate": 1.2341426060194049e-05,
+      "loss": 1.5974,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.5585,
+      "step": 4592,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.5586,
+      "step": 4592,
+      "vm_loss": 0.1143
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.0075,
+      "step": 4592,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.4814,
+      "step": 4592,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.3402,
+      "step": 4592,
+      "vm_loss": 0.2109
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.4366,
+      "step": 4592,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.7803,
+      "step": 4592,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 0.8839906634261376,
+      "lm_loss": 1.3822,
+      "step": 4592,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 0.884183170103713,
+      "grad_norm": 3.315556386594215,
+      "learning_rate": 1.2338394720266729e-05,
+      "loss": 1.6535,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8843756767812884,
+      "grad_norm": 3.1962428399109504,
+      "learning_rate": 1.2335363153017528e-05,
+      "loss": 1.635,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8845681834588637,
+      "grad_norm": 3.0978866346938694,
+      "learning_rate": 1.2332331358741147e-05,
+      "loss": 1.692,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8847606901364391,
+      "grad_norm": 3.0753693627008856,
+      "learning_rate": 1.232929933773232e-05,
+      "loss": 1.5809,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8849531968140145,
+      "grad_norm": 3.149636508210439,
+      "learning_rate": 1.23262670902858e-05,
+      "loss": 1.7049,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8851457034915898,
+      "grad_norm": 2.988612280970963,
+      "learning_rate": 1.2323234616696353e-05,
+      "loss": 1.6249,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8853382101691653,
+      "grad_norm": 3.1006811744798806,
+      "learning_rate": 1.232020191725878e-05,
+      "loss": 1.6025,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "grad_norm": 3.1546596868266668,
+      "learning_rate": 1.2317168992267897e-05,
+      "loss": 1.6837,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.6674,
+      "step": 4600,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 2.0395,
+      "step": 4600,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.7314,
+      "step": 4600,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.026,
+      "step": 4600,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.4702,
+      "step": 4600,
+      "vm_loss": 0.1194
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.5369,
+      "step": 4600,
+      "vm_loss": 0.2163
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.2585,
+      "step": 4600,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.8855307168467407,
+      "lm_loss": 1.6183,
+      "step": 4600,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 0.885723223524316,
+      "grad_norm": 3.0676064950398714,
+      "learning_rate": 1.2314135842018541e-05,
+      "loss": 1.6357,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8859157302018914,
+      "grad_norm": 3.4160529813487814,
+      "learning_rate": 1.2311102466805578e-05,
+      "loss": 1.6635,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8861082368794667,
+      "grad_norm": 3.0827882577358348,
+      "learning_rate": 1.2308068866923885e-05,
+      "loss": 1.6261,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8863007435570421,
+      "grad_norm": 3.1139809583041567,
+      "learning_rate": 1.2305035042668372e-05,
+      "loss": 1.6624,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8864932502346176,
+      "grad_norm": 3.2010709841406575,
+      "learning_rate": 1.230200099433396e-05,
+      "loss": 1.6666,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8866857569121929,
+      "grad_norm": 3.159535980287746,
+      "learning_rate": 1.2298966722215604e-05,
+      "loss": 1.6099,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8868782635897683,
+      "grad_norm": 3.1742782734120714,
+      "learning_rate": 1.2295932226608267e-05,
+      "loss": 1.65,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "grad_norm": 3.1888948681243336,
+      "learning_rate": 1.2292897507806942e-05,
+      "loss": 1.567,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.5239,
+      "step": 4608,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.1824,
+      "step": 4608,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.5153,
+      "step": 4608,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.562,
+      "step": 4608,
+      "vm_loss": 0.0847
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.2188,
+      "step": 4608,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.5364,
+      "step": 4608,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 1.7758,
+      "step": 4608,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 0.8870707702673436,
+      "lm_loss": 2.0222,
+      "step": 4608,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 0.887263276944919,
+      "grad_norm": 3.3027548119621435,
+      "learning_rate": 1.2289862566106649e-05,
+      "loss": 1.6529,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8874557836224944,
+      "grad_norm": 3.1453994806310086,
+      "learning_rate": 1.2286827401802418e-05,
+      "loss": 1.4935,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8876482903000698,
+      "grad_norm": 3.227512629656992,
+      "learning_rate": 1.2283792015189304e-05,
+      "loss": 1.6087,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8878407969776452,
+      "grad_norm": 3.0819205992772725,
+      "learning_rate": 1.2280756406562388e-05,
+      "loss": 1.6505,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8880333036552205,
+      "grad_norm": 3.1243158083345848,
+      "learning_rate": 1.2277720576216772e-05,
+      "loss": 1.6221,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8882258103327959,
+      "grad_norm": 3.1095502589038673,
+      "learning_rate": 1.2274684524447575e-05,
+      "loss": 1.6685,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8884183170103713,
+      "grad_norm": 3.253054573823724,
+      "learning_rate": 1.227164825154994e-05,
+      "loss": 1.6932,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "grad_norm": 3.0614054837713764,
+      "learning_rate": 1.2268611757819034e-05,
+      "loss": 1.6843,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 0.6115,
+      "step": 4616,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.635,
+      "step": 4616,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.0678,
+      "step": 4616,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 2.0023,
+      "step": 4616,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.3538,
+      "step": 4616,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.0059,
+      "step": 4616,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.0953,
+      "step": 4616,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.8886108236879466,
+      "lm_loss": 1.3482,
+      "step": 4616,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 0.8888033303655221,
+      "grad_norm": 3.10550399722038,
+      "learning_rate": 1.2265575043550038e-05,
+      "loss": 1.565,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8889958370430975,
+      "grad_norm": 3.100190892774192,
+      "learning_rate": 1.2262538109038162e-05,
+      "loss": 1.6939,
+      "step": 4618
+    },
+    {
+      "epoch": 0.8891883437206728,
+      "grad_norm": 3.0403326081130655,
+      "learning_rate": 1.225950095457864e-05,
+      "loss": 1.6972,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8893808503982482,
+      "grad_norm": 3.128885863206769,
+      "learning_rate": 1.2256463580466713e-05,
+      "loss": 1.6154,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8895733570758235,
+      "grad_norm": 3.0510610344879048,
+      "learning_rate": 1.2253425986997658e-05,
+      "loss": 1.6542,
+      "step": 4621
+    },
+    {
+      "epoch": 0.889765863753399,
+      "grad_norm": 2.9153665077793716,
+      "learning_rate": 1.225038817446677e-05,
+      "loss": 1.6346,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8899583704309744,
+      "grad_norm": 2.942641990042011,
+      "learning_rate": 1.2247350143169358e-05,
+      "loss": 1.5289,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "grad_norm": 3.1305139392744055,
+      "learning_rate": 1.2244311893400761e-05,
+      "loss": 1.6499,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.9097,
+      "step": 4624,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.4179,
+      "step": 4624,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.5429,
+      "step": 4624,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.9257,
+      "step": 4624,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.5695,
+      "step": 4624,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.3959,
+      "step": 4624,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.358,
+      "step": 4624,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 0.8901508771085497,
+      "lm_loss": 1.1129,
+      "step": 4624,
+      "vm_loss": 0.1069
+    },
+    {
+      "epoch": 0.8903433837861251,
+      "grad_norm": 3.0393107727471476,
+      "learning_rate": 1.2241273425456335e-05,
+      "loss": 1.6413,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8905358904637004,
+      "grad_norm": 3.0958053591355994,
+      "learning_rate": 1.223823473963146e-05,
+      "loss": 1.6782,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8907283971412758,
+      "grad_norm": 3.2182620886126374,
+      "learning_rate": 1.223519583622153e-05,
+      "loss": 1.6827,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8909209038188513,
+      "grad_norm": 3.179706211206332,
+      "learning_rate": 1.223215671552197e-05,
+      "loss": 1.6415,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8911134104964266,
+      "grad_norm": 3.244708835716366,
+      "learning_rate": 1.2229117377828219e-05,
+      "loss": 1.6588,
+      "step": 4629
+    },
+    {
+      "epoch": 0.891305917174002,
+      "grad_norm": 3.1654432788061952,
+      "learning_rate": 1.2226077823435738e-05,
+      "loss": 1.6757,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8914984238515774,
+      "grad_norm": 3.10313238038243,
+      "learning_rate": 1.2223038052640016e-05,
+      "loss": 1.6096,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "grad_norm": 3.052738658824864,
+      "learning_rate": 1.2219998065736554e-05,
+      "loss": 1.65,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.8079,
+      "step": 4632,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.6441,
+      "step": 4632,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.2368,
+      "step": 4632,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.2051,
+      "step": 4632,
+      "vm_loss": 0.2239
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.6459,
+      "step": 4632,
+      "vm_loss": 0.098
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.3861,
+      "step": 4632,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 1.1626,
+      "step": 4632,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 0.8916909305291527,
+      "lm_loss": 2.021,
+      "step": 4632,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.8918834372067281,
+      "grad_norm": 3.2697272247795865,
+      "learning_rate": 1.2216957863020878e-05,
+      "loss": 1.5705,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8920759438843034,
+      "grad_norm": 2.9708022167190373,
+      "learning_rate": 1.2213917444788532e-05,
+      "loss": 1.5708,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8922684505618789,
+      "grad_norm": 3.0047296594198984,
+      "learning_rate": 1.2210876811335089e-05,
+      "loss": 1.5814,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8924609572394543,
+      "grad_norm": 3.1603834086492255,
+      "learning_rate": 1.220783596295613e-05,
+      "loss": 1.6645,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8926534639170296,
+      "grad_norm": 3.023515673247715,
+      "learning_rate": 1.2204794899947272e-05,
+      "loss": 1.5917,
+      "step": 4637
+    },
+    {
+      "epoch": 0.892845970594605,
+      "grad_norm": 3.1204589083766807,
+      "learning_rate": 1.2201753622604145e-05,
+      "loss": 1.6147,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8930384772721803,
+      "grad_norm": 3.2572988148402566,
+      "learning_rate": 1.2198712131222395e-05,
+      "loss": 1.6018,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "grad_norm": 3.2353028219647286,
+      "learning_rate": 1.2195670426097692e-05,
+      "loss": 1.5814,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.9769,
+      "step": 4640,
+      "vm_loss": 0.2061
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.3061,
+      "step": 4640,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.0332,
+      "step": 4640,
+      "vm_loss": 0.2129
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.294,
+      "step": 4640,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 2.0536,
+      "step": 4640,
+      "vm_loss": 0.1034
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.6611,
+      "step": 4640,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.6921,
+      "step": 4640,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.8932309839497558,
+      "lm_loss": 1.5264,
+      "step": 4640,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 0.8934234906273312,
+      "grad_norm": 3.366885372667514,
+      "learning_rate": 1.2192628507525737e-05,
+      "loss": 1.668,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8936159973049065,
+      "grad_norm": 3.0493219596915204,
+      "learning_rate": 1.218958637580224e-05,
+      "loss": 1.586,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8938085039824819,
+      "grad_norm": 3.18845784404462,
+      "learning_rate": 1.2186544031222932e-05,
+      "loss": 1.652,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8940010106600572,
+      "grad_norm": 3.178947272327734,
+      "learning_rate": 1.218350147408357e-05,
+      "loss": 1.6399,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8941935173376326,
+      "grad_norm": 3.00309772751451,
+      "learning_rate": 1.2180458704679931e-05,
+      "loss": 1.5833,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8943860240152081,
+      "grad_norm": 3.0841302452372537,
+      "learning_rate": 1.2177415723307808e-05,
+      "loss": 1.6665,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8945785306927834,
+      "grad_norm": 3.216045703054112,
+      "learning_rate": 1.2174372530263018e-05,
+      "loss": 1.645,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "grad_norm": 3.05202555598051,
+      "learning_rate": 1.2171329125841404e-05,
+      "loss": 1.6664,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.8764,
+      "step": 4648,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.2324,
+      "step": 4648,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.0173,
+      "step": 4648,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.1536,
+      "step": 4648,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.2665,
+      "step": 4648,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.3861,
+      "step": 4648,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.4213,
+      "step": 4648,
+      "vm_loss": 0.2204
+    },
+    {
+      "epoch": 0.8947710373703588,
+      "lm_loss": 1.4593,
+      "step": 4648,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 0.8949635440479342,
+      "grad_norm": 3.2904167436694265,
+      "learning_rate": 1.2168285510338817e-05,
+      "loss": 1.6271,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8951560507255095,
+      "grad_norm": 3.1996784863490455,
+      "learning_rate": 1.216524168405114e-05,
+      "loss": 1.6564,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8953485574030849,
+      "grad_norm": 3.0093466192511378,
+      "learning_rate": 1.2162197647274269e-05,
+      "loss": 1.6286,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8955410640806603,
+      "grad_norm": 3.190197375782913,
+      "learning_rate": 1.2159153400304124e-05,
+      "loss": 1.6191,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8957335707582357,
+      "grad_norm": 3.095108948285983,
+      "learning_rate": 1.2156108943436644e-05,
+      "loss": 1.6293,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8959260774358111,
+      "grad_norm": 3.169867952570169,
+      "learning_rate": 1.215306427696779e-05,
+      "loss": 1.6101,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8961185841133864,
+      "grad_norm": 3.064936468378051,
+      "learning_rate": 1.2150019401193545e-05,
+      "loss": 1.6075,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "grad_norm": 3.31709284902921,
+      "learning_rate": 1.2146974316409902e-05,
+      "loss": 1.6022,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.4719,
+      "step": 4656,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.3698,
+      "step": 4656,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.2543,
+      "step": 4656,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.6949,
+      "step": 4656,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.6257,
+      "step": 4656,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.2,
+      "step": 4656,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.6329,
+      "step": 4656,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 0.8963110907909618,
+      "lm_loss": 1.4026,
+      "step": 4656,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 0.8965035974685371,
+      "grad_norm": 3.1366733945081773,
+      "learning_rate": 1.2143929022912895e-05,
+      "loss": 1.5494,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8966961041461126,
+      "grad_norm": 3.1187682095136795,
+      "learning_rate": 1.2140883520998552e-05,
+      "loss": 1.5779,
+      "step": 4658
+    },
+    {
+      "epoch": 0.896888610823688,
+      "grad_norm": 3.0839650711338087,
+      "learning_rate": 1.2137837810962942e-05,
+      "loss": 1.6034,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8970811175012633,
+      "grad_norm": 3.0239628653720616,
+      "learning_rate": 1.213479189310215e-05,
+      "loss": 1.622,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8972736241788387,
+      "grad_norm": 3.0494806224712994,
+      "learning_rate": 1.2131745767712268e-05,
+      "loss": 1.5916,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8974661308564141,
+      "grad_norm": 3.076872414199497,
+      "learning_rate": 1.2128699435089429e-05,
+      "loss": 1.6399,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8976586375339894,
+      "grad_norm": 3.0832554094676636,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 1.6499,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "grad_norm": 3.132694545451436,
+      "learning_rate": 1.2122606149329452e-05,
+      "loss": 1.6629,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.4086,
+      "step": 4664,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.6571,
+      "step": 4664,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.744,
+      "step": 4664,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.4371,
+      "step": 4664,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.5094,
+      "step": 4664,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 0.94,
+      "step": 4664,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.7128,
+      "step": 4664,
+      "vm_loss": 0.2407
+    },
+    {
+      "epoch": 0.8978511442115649,
+      "lm_loss": 1.5537,
+      "step": 4664,
+      "vm_loss": 0.2256
+    },
+    {
+      "epoch": 0.8980436508891402,
+      "grad_norm": 3.3090151082549486,
+      "learning_rate": 1.211955919678466e-05,
+      "loss": 1.6759,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8982361575667156,
+      "grad_norm": 3.08192270973342,
+      "learning_rate": 1.2116512038191599e-05,
+      "loss": 1.5975,
+      "step": 4666
+    },
+    {
+      "epoch": 0.898428664244291,
+      "grad_norm": 3.1936096077047322,
+      "learning_rate": 1.2113464673846492e-05,
+      "loss": 1.6439,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8986211709218663,
+      "grad_norm": 3.0011780195168476,
+      "learning_rate": 1.2110417104045575e-05,
+      "loss": 1.6437,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8988136775994418,
+      "grad_norm": 3.037762702966735,
+      "learning_rate": 1.210736932908512e-05,
+      "loss": 1.6962,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8990061842770171,
+      "grad_norm": 3.0322050904902,
+      "learning_rate": 1.2104321349261401e-05,
+      "loss": 1.5935,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8991986909545925,
+      "grad_norm": 3.158070329535032,
+      "learning_rate": 1.2101273164870726e-05,
+      "loss": 1.6979,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "grad_norm": 3.00476197813505,
+      "learning_rate": 1.2098224776209417e-05,
+      "loss": 1.5603,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.0689,
+      "step": 4672,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 2.5474,
+      "step": 4672,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.631,
+      "step": 4672,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.789,
+      "step": 4672,
+      "vm_loss": 0.2855
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.5229,
+      "step": 4672,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.6291,
+      "step": 4672,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.2529,
+      "step": 4672,
+      "vm_loss": 0.1437
+    },
+    {
+      "epoch": 0.8993911976321679,
+      "lm_loss": 1.237,
+      "step": 4672,
+      "vm_loss": 0.1077
+    },
+    {
+      "epoch": 0.8995837043097432,
+      "grad_norm": 3.2543579634043627,
+      "learning_rate": 1.2095176183573815e-05,
+      "loss": 1.6247,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8997762109873186,
+      "grad_norm": 3.148133273728255,
+      "learning_rate": 1.2092127387260283e-05,
+      "loss": 1.5773,
+      "step": 4674
+    },
+    {
+      "epoch": 0.899968717664894,
+      "grad_norm": 3.073996355350252,
+      "learning_rate": 1.2089078387565204e-05,
+      "loss": 1.6061,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9001612243424694,
+      "grad_norm": 3.175612370012723,
+      "learning_rate": 1.2086029184784978e-05,
+      "loss": 1.6827,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9003537310200448,
+      "grad_norm": 3.035258058385415,
+      "learning_rate": 1.2082979779216027e-05,
+      "loss": 1.6511,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9005462376976201,
+      "grad_norm": 3.0649198727208358,
+      "learning_rate": 1.2079930171154792e-05,
+      "loss": 1.5653,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9007387443751955,
+      "grad_norm": 3.055381101298269,
+      "learning_rate": 1.2076880360897737e-05,
+      "loss": 1.6153,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "grad_norm": 3.0158721380168196,
+      "learning_rate": 1.2073830348741335e-05,
+      "loss": 1.5756,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 2.2093,
+      "step": 4680,
+      "vm_loss": 0.2302
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 0.9889,
+      "step": 4680,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 1.5358,
+      "step": 4680,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 1.434,
+      "step": 4680,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 1.7574,
+      "step": 4680,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 1.0625,
+      "step": 4680,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 1.8779,
+      "step": 4680,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 0.9009312510527709,
+      "lm_loss": 0.9325,
+      "step": 4680,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 0.9011237577303463,
+      "grad_norm": 3.2162631844118277,
+      "learning_rate": 1.2070780134982095e-05,
+      "loss": 1.7162,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9013162644079217,
+      "grad_norm": 3.1308054971656745,
+      "learning_rate": 1.206772971991653e-05,
+      "loss": 1.6107,
+      "step": 4682
+    },
+    {
+      "epoch": 0.901508771085497,
+      "grad_norm": 2.941418197703706,
+      "learning_rate": 1.2064679103841182e-05,
+      "loss": 1.5446,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9017012777630724,
+      "grad_norm": 3.193208413458499,
+      "learning_rate": 1.2061628287052611e-05,
+      "loss": 1.6469,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9018937844406478,
+      "grad_norm": 3.2828086615883274,
+      "learning_rate": 1.2058577269847394e-05,
+      "loss": 1.6149,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9020862911182231,
+      "grad_norm": 3.1134899536792666,
+      "learning_rate": 1.2055526052522132e-05,
+      "loss": 1.6555,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9022787977957986,
+      "grad_norm": 2.9969836519682254,
+      "learning_rate": 1.2052474635373436e-05,
+      "loss": 1.6475,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "grad_norm": 2.990386236690656,
+      "learning_rate": 1.204942301869795e-05,
+      "loss": 1.6913,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.898,
+      "step": 4688,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.1247,
+      "step": 4688,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.6261,
+      "step": 4688,
+      "vm_loss": 0.2368
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 2.0588,
+      "step": 4688,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.422,
+      "step": 4688,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.116,
+      "step": 4688,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 1.6583,
+      "step": 4688,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 0.9024713044733739,
+      "lm_loss": 2.0492,
+      "step": 4688,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 0.9026638111509493,
+      "grad_norm": 3.1122430373755576,
+      "learning_rate": 1.2046371202792326e-05,
+      "loss": 1.661,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9028563178285247,
+      "grad_norm": 3.0778606178917554,
+      "learning_rate": 1.2043319187953242e-05,
+      "loss": 1.5298,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9030488245061,
+      "grad_norm": 3.107991430557598,
+      "learning_rate": 1.204026697447739e-05,
+      "loss": 1.5297,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9032413311836754,
+      "grad_norm": 3.1109362734329538,
+      "learning_rate": 1.2037214562661483e-05,
+      "loss": 1.6395,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9034338378612509,
+      "grad_norm": 3.1776933412960404,
+      "learning_rate": 1.2034161952802259e-05,
+      "loss": 1.6126,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9036263445388262,
+      "grad_norm": 3.1758622802049192,
+      "learning_rate": 1.2031109145196469e-05,
+      "loss": 1.6582,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9038188512164016,
+      "grad_norm": 3.0104186612043264,
+      "learning_rate": 1.2028056140140885e-05,
+      "loss": 1.6123,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "grad_norm": 3.047143656417212,
+      "learning_rate": 1.2025002937932298e-05,
+      "loss": 1.6624,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.1889,
+      "step": 4696,
+      "vm_loss": 0.2658
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.362,
+      "step": 4696,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.1682,
+      "step": 4696,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.6901,
+      "step": 4696,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 0.6715,
+      "step": 4696,
+      "vm_loss": 0.1095
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.5931,
+      "step": 4696,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.4708,
+      "step": 4696,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 0.9040113578939769,
+      "lm_loss": 1.3216,
+      "step": 4696,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 0.9042038645715523,
+      "grad_norm": 2.997997235687829,
+      "learning_rate": 1.2021949538867514e-05,
+      "loss": 1.5484,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9043963712491278,
+      "grad_norm": 2.922027475024749,
+      "learning_rate": 1.2018895943243372e-05,
+      "loss": 1.5326,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9045888779267031,
+      "grad_norm": 3.009076763808899,
+      "learning_rate": 1.2015842151356711e-05,
+      "loss": 1.5415,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9047813846042785,
+      "grad_norm": 3.1825166629669934,
+      "learning_rate": 1.2012788163504405e-05,
+      "loss": 1.6531,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9049738912818538,
+      "grad_norm": 3.0516416241714244,
+      "learning_rate": 1.2009733979983338e-05,
+      "loss": 1.5371,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9051663979594292,
+      "grad_norm": 3.0795298313082937,
+      "learning_rate": 1.2006679601090415e-05,
+      "loss": 1.6377,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9053589046370046,
+      "grad_norm": 3.242419716000162,
+      "learning_rate": 1.2003625027122562e-05,
+      "loss": 1.615,
+      "step": 4703
+    },
+    {
+      "epoch": 0.90555141131458,
+      "grad_norm": 3.0705280806131863,
+      "learning_rate": 1.2000570258376724e-05,
+      "loss": 1.5279,
+      "step": 4704
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 2.1423,
+      "step": 4704,
+      "vm_loss": 0.1696
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.7026,
+      "step": 4704,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.4717,
+      "step": 4704,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.9995,
+      "step": 4704,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.612,
+      "step": 4704,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.9819,
+      "step": 4704,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.5629,
+      "step": 4704,
+      "vm_loss": 0.1346
+    },
+    {
+      "epoch": 0.90555141131458,
+      "lm_loss": 1.5307,
+      "step": 4704,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 0.9057439179921554,
+      "grad_norm": 3.2109250004982073,
+      "learning_rate": 1.1997515295149861e-05,
+      "loss": 1.6352,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9059364246697307,
+      "grad_norm": 3.120886181270095,
+      "learning_rate": 1.1994460137738954e-05,
+      "loss": 1.6207,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9061289313473061,
+      "grad_norm": 3.244670606071805,
+      "learning_rate": 1.1991404786441007e-05,
+      "loss": 1.6012,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9063214380248815,
+      "grad_norm": 3.385320122090827,
+      "learning_rate": 1.1988349241553038e-05,
+      "loss": 1.6448,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9065139447024568,
+      "grad_norm": 3.1747655067058607,
+      "learning_rate": 1.1985293503372085e-05,
+      "loss": 1.5946,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9067064513800323,
+      "grad_norm": 2.932729733122712,
+      "learning_rate": 1.1982237572195203e-05,
+      "loss": 1.56,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9068989580576077,
+      "grad_norm": 3.0855871978194744,
+      "learning_rate": 1.1979181448319469e-05,
+      "loss": 1.6133,
+      "step": 4711
+    },
+    {
+      "epoch": 0.907091464735183,
+      "grad_norm": 3.386528743654921,
+      "learning_rate": 1.1976125132041974e-05,
+      "loss": 1.7261,
+      "step": 4712
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.0121,
+      "step": 4712,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.5156,
+      "step": 4712,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.7849,
+      "step": 4712,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.9754,
+      "step": 4712,
+      "vm_loss": 0.2184
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.2805,
+      "step": 4712,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.4139,
+      "step": 4712,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.0879,
+      "step": 4712,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 0.907091464735183,
+      "lm_loss": 1.1509,
+      "step": 4712,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 0.9072839714127584,
+      "grad_norm": 3.199990428689835,
+      "learning_rate": 1.1973068623659839e-05,
+      "loss": 1.6055,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9074764780903337,
+      "grad_norm": 3.018276693329278,
+      "learning_rate": 1.1970011923470189e-05,
+      "loss": 1.538,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9076689847679091,
+      "grad_norm": 3.0560692992183722,
+      "learning_rate": 1.1966955031770176e-05,
+      "loss": 1.6219,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9078614914454846,
+      "grad_norm": 3.0018417725136213,
+      "learning_rate": 1.196389794885697e-05,
+      "loss": 1.6411,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9080539981230599,
+      "grad_norm": 3.1087391739088837,
+      "learning_rate": 1.1960840675027754e-05,
+      "loss": 1.6407,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9082465048006353,
+      "grad_norm": 3.0829636796115616,
+      "learning_rate": 1.195778321057974e-05,
+      "loss": 1.6297,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9084390114782106,
+      "grad_norm": 3.08512040251026,
+      "learning_rate": 1.1954725555810151e-05,
+      "loss": 1.5259,
+      "step": 4719
+    },
+    {
+      "epoch": 0.908631518155786,
+      "grad_norm": 3.1531988826165147,
+      "learning_rate": 1.1951667711016233e-05,
+      "loss": 1.5728,
+      "step": 4720
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.7369,
+      "step": 4720,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.233,
+      "step": 4720,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.0099,
+      "step": 4720,
+      "vm_loss": 0.1276
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.0384,
+      "step": 4720,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.4658,
+      "step": 4720,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.0697,
+      "step": 4720,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.7965,
+      "step": 4720,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 0.908631518155786,
+      "lm_loss": 1.9753,
+      "step": 4720,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.9088240248333614,
+      "grad_norm": 3.325567806677158,
+      "learning_rate": 1.194860967649524e-05,
+      "loss": 1.5958,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9090165315109368,
+      "grad_norm": 3.2019621959431266,
+      "learning_rate": 1.1945551452544458e-05,
+      "loss": 1.5239,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9092090381885122,
+      "grad_norm": 3.309722748894992,
+      "learning_rate": 1.1942493039461185e-05,
+      "loss": 1.6618,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9094015448660876,
+      "grad_norm": 3.3072880613639044,
+      "learning_rate": 1.1939434437542734e-05,
+      "loss": 1.6342,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9095940515436629,
+      "grad_norm": 3.049943449060783,
+      "learning_rate": 1.1936375647086444e-05,
+      "loss": 1.6307,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9097865582212383,
+      "grad_norm": 3.3162130404318253,
+      "learning_rate": 1.1933316668389669e-05,
+      "loss": 1.6491,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9099790648988136,
+      "grad_norm": 3.223514727480635,
+      "learning_rate": 1.1930257501749777e-05,
+      "loss": 1.6575,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "grad_norm": 2.977685932359518,
+      "learning_rate": 1.1927198147464163e-05,
+      "loss": 1.6196,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.7089,
+      "step": 4728,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 0.9011,
+      "step": 4728,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.3424,
+      "step": 4728,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.3399,
+      "step": 4728,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.7321,
+      "step": 4728,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.7444,
+      "step": 4728,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.1951,
+      "step": 4728,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 0.9101715715763891,
+      "lm_loss": 1.9315,
+      "step": 4728,
+      "vm_loss": 0.1422
+    },
+    {
+      "epoch": 0.9103640782539645,
+      "grad_norm": 3.1635544184550577,
+      "learning_rate": 1.1924138605830231e-05,
+      "loss": 1.5845,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9105565849315398,
+      "grad_norm": 3.021581326141417,
+      "learning_rate": 1.1921078877145414e-05,
+      "loss": 1.5498,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9107490916091152,
+      "grad_norm": 3.434839780280532,
+      "learning_rate": 1.1918018961707147e-05,
+      "loss": 1.6179,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9109415982866905,
+      "grad_norm": 3.2198833407364114,
+      "learning_rate": 1.1914958859812901e-05,
+      "loss": 1.5567,
+      "step": 4732
+    },
+    {
+      "epoch": 0.911134104964266,
+      "grad_norm": 3.1224098506504623,
+      "learning_rate": 1.1911898571760155e-05,
+      "loss": 1.5817,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9113266116418414,
+      "grad_norm": 3.0629613940052614,
+      "learning_rate": 1.1908838097846404e-05,
+      "loss": 1.6571,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9115191183194167,
+      "grad_norm": 3.1317016752153846,
+      "learning_rate": 1.1905777438369175e-05,
+      "loss": 1.6147,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "grad_norm": 3.05355189921111,
+      "learning_rate": 1.1902716593625991e-05,
+      "loss": 1.6229,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.5109,
+      "step": 4736,
+      "vm_loss": 0.0979
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.4057,
+      "step": 4736,
+      "vm_loss": 0.2678
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.2949,
+      "step": 4736,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.5492,
+      "step": 4736,
+      "vm_loss": 0.2187
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 0.9376,
+      "step": 4736,
+      "vm_loss": 0.1047
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.8476,
+      "step": 4736,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.7221,
+      "step": 4736,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 0.9117116249969921,
+      "lm_loss": 1.4249,
+      "step": 4736,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 0.9119041316745674,
+      "grad_norm": 3.078609919769739,
+      "learning_rate": 1.1899655563914413e-05,
+      "loss": 1.5567,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9120966383521428,
+      "grad_norm": 3.1494639008782683,
+      "learning_rate": 1.1896594349532014e-05,
+      "loss": 1.6325,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9122891450297183,
+      "grad_norm": 3.1079321919728846,
+      "learning_rate": 1.1893532950776373e-05,
+      "loss": 1.5214,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9124816517072936,
+      "grad_norm": 3.296226850772759,
+      "learning_rate": 1.1890471367945108e-05,
+      "loss": 1.5918,
+      "step": 4740
+    },
+    {
+      "epoch": 0.912674158384869,
+      "grad_norm": 3.314991454319871,
+      "learning_rate": 1.188740960133584e-05,
+      "loss": 1.6301,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9128666650624444,
+      "grad_norm": 3.304338307960673,
+      "learning_rate": 1.1884347651246214e-05,
+      "loss": 1.6048,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9130591717400197,
+      "grad_norm": 3.3300366135285278,
+      "learning_rate": 1.1881285517973885e-05,
+      "loss": 1.6116,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "grad_norm": 3.163511130898148,
+      "learning_rate": 1.1878223201816534e-05,
+      "loss": 1.685,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.419,
+      "step": 4744,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.1975,
+      "step": 4744,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.4845,
+      "step": 4744,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.5721,
+      "step": 4744,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.3634,
+      "step": 4744,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.2942,
+      "step": 4744,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.5083,
+      "step": 4744,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 0.9132516784175951,
+      "lm_loss": 1.0931,
+      "step": 4744,
+      "vm_loss": 0.2037
+    },
+    {
+      "epoch": 0.9134441850951704,
+      "grad_norm": 3.096686639377927,
+      "learning_rate": 1.187516070307186e-05,
+      "loss": 1.64,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9136366917727459,
+      "grad_norm": 2.9079387174667533,
+      "learning_rate": 1.1872098022037575e-05,
+      "loss": 1.5758,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9138291984503213,
+      "grad_norm": 3.061758563851062,
+      "learning_rate": 1.186903515901141e-05,
+      "loss": 1.5934,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9140217051278966,
+      "grad_norm": 3.1823371658554103,
+      "learning_rate": 1.1865972114291116e-05,
+      "loss": 1.5801,
+      "step": 4748
+    },
+    {
+      "epoch": 0.914214211805472,
+      "grad_norm": 3.484095623995338,
+      "learning_rate": 1.1862908888174462e-05,
+      "loss": 1.6215,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9144067184830473,
+      "grad_norm": 3.3834394171611537,
+      "learning_rate": 1.1859845480959229e-05,
+      "loss": 1.5714,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9145992251606228,
+      "grad_norm": 3.107808325801354,
+      "learning_rate": 1.1856781892943218e-05,
+      "loss": 1.546,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "grad_norm": 3.092399057497426,
+      "learning_rate": 1.1853718124424258e-05,
+      "loss": 1.6522,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.3328,
+      "step": 4752,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.3626,
+      "step": 4752,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.2497,
+      "step": 4752,
+      "vm_loss": 0.2243
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.4696,
+      "step": 4752,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.528,
+      "step": 4752,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 2.1192,
+      "step": 4752,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.4258,
+      "step": 4752,
+      "vm_loss": 0.1793
+    },
+    {
+      "epoch": 0.9147917318381982,
+      "lm_loss": 1.5073,
+      "step": 4752,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.9149842385157735,
+      "grad_norm": 3.118080759115116,
+      "learning_rate": 1.1850654175700175e-05,
+      "loss": 1.5939,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9151767451933489,
+      "grad_norm": 3.0603491941346945,
+      "learning_rate": 1.1847590047068834e-05,
+      "loss": 1.5756,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9153692518709243,
+      "grad_norm": 3.054071426864371,
+      "learning_rate": 1.1844525738828098e-05,
+      "loss": 1.5916,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9155617585484996,
+      "grad_norm": 3.1115514042059838,
+      "learning_rate": 1.1841461251275868e-05,
+      "loss": 1.629,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9157542652260751,
+      "grad_norm": 3.156577782343747,
+      "learning_rate": 1.183839658471004e-05,
+      "loss": 1.5943,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9159467719036504,
+      "grad_norm": 2.96346855001217,
+      "learning_rate": 1.1835331739428545e-05,
+      "loss": 1.5604,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9161392785812258,
+      "grad_norm": 3.0608833534009663,
+      "learning_rate": 1.1832266715729324e-05,
+      "loss": 1.5823,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "grad_norm": 3.213660262891851,
+      "learning_rate": 1.1829201513910339e-05,
+      "loss": 1.5654,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 0.971,
+      "step": 4760,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.468,
+      "step": 4760,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.0862,
+      "step": 4760,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.3217,
+      "step": 4760,
+      "vm_loss": 0.2288
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.7517,
+      "step": 4760,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.3305,
+      "step": 4760,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.3218,
+      "step": 4760,
+      "vm_loss": 0.1109
+    },
+    {
+      "epoch": 0.9163317852588012,
+      "lm_loss": 1.4472,
+      "step": 4760,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 0.9165242919363765,
+      "grad_norm": 3.111033909598755,
+      "learning_rate": 1.182613613426956e-05,
+      "loss": 1.555,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9167167986139519,
+      "grad_norm": 3.0694053880828096,
+      "learning_rate": 1.1823070577104989e-05,
+      "loss": 1.6144,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9169093052915273,
+      "grad_norm": 3.0911651853895354,
+      "learning_rate": 1.1820004842714634e-05,
+      "loss": 1.6094,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9171018119691027,
+      "grad_norm": 3.1440997607893215,
+      "learning_rate": 1.1816938931396525e-05,
+      "loss": 1.5272,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9172943186466781,
+      "grad_norm": 3.2112795550665725,
+      "learning_rate": 1.18138728434487e-05,
+      "loss": 1.619,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9174868253242534,
+      "grad_norm": 3.037661841808173,
+      "learning_rate": 1.1810806579169235e-05,
+      "loss": 1.6242,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9176793320018288,
+      "grad_norm": 3.0503935184502766,
+      "learning_rate": 1.1807740138856203e-05,
+      "loss": 1.6127,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "grad_norm": 3.013279554216417,
+      "learning_rate": 1.1804673522807699e-05,
+      "loss": 1.5879,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.0693,
+      "step": 4768,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 0.9161,
+      "step": 4768,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.3118,
+      "step": 4768,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.497,
+      "step": 4768,
+      "vm_loss": 0.2291
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.3785,
+      "step": 4768,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.2063,
+      "step": 4768,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.3182,
+      "step": 4768,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 0.9178718386794041,
+      "lm_loss": 1.4246,
+      "step": 4768,
+      "vm_loss": 0.1194
+    },
+    {
+      "epoch": 0.9180643453569796,
+      "grad_norm": 3.133762994320165,
+      "learning_rate": 1.180160673132184e-05,
+      "loss": 1.581,
+      "step": 4769
+    },
+    {
+      "epoch": 0.918256852034555,
+      "grad_norm": 3.272012218669178,
+      "learning_rate": 1.1798539764696756e-05,
+      "loss": 1.6411,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9184493587121303,
+      "grad_norm": 3.183984436062204,
+      "learning_rate": 1.17954726232306e-05,
+      "loss": 1.6584,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9186418653897057,
+      "grad_norm": 3.1608099526418023,
+      "learning_rate": 1.1792405307221535e-05,
+      "loss": 1.6111,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9188343720672811,
+      "grad_norm": 3.094614310245401,
+      "learning_rate": 1.1789337816967744e-05,
+      "loss": 1.6069,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9190268787448564,
+      "grad_norm": 3.113612205834525,
+      "learning_rate": 1.178627015276742e-05,
+      "loss": 1.6155,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9192193854224319,
+      "grad_norm": 3.096776503243558,
+      "learning_rate": 1.1783202314918787e-05,
+      "loss": 1.5997,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "grad_norm": 3.1498093917458228,
+      "learning_rate": 1.1780134303720074e-05,
+      "loss": 1.5243,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.2499,
+      "step": 4776,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.0967,
+      "step": 4776,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.3902,
+      "step": 4776,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.7551,
+      "step": 4776,
+      "vm_loss": 0.2257
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 0.8441,
+      "step": 4776,
+      "vm_loss": 0.2517
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.4534,
+      "step": 4776,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 1.457,
+      "step": 4776,
+      "vm_loss": 0.1164
+    },
+    {
+      "epoch": 0.9194118921000072,
+      "lm_loss": 0.8831,
+      "step": 4776,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 0.9196043987775826,
+      "grad_norm": 3.0232594043394485,
+      "learning_rate": 1.1777066119469533e-05,
+      "loss": 1.6018,
+      "step": 4777
+    },
+    {
+      "epoch": 0.919796905455158,
+      "grad_norm": 3.300903977480852,
+      "learning_rate": 1.177399776246543e-05,
+      "loss": 1.508,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9199894121327333,
+      "grad_norm": 3.184781032146618,
+      "learning_rate": 1.1770929233006047e-05,
+      "loss": 1.5742,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9201819188103088,
+      "grad_norm": 3.269287148303652,
+      "learning_rate": 1.1767860531389686e-05,
+      "loss": 1.5955,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9203744254878841,
+      "grad_norm": 3.2165907530184694,
+      "learning_rate": 1.1764791657914664e-05,
+      "loss": 1.6458,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9205669321654595,
+      "grad_norm": 3.0867529396215097,
+      "learning_rate": 1.1761722612879317e-05,
+      "loss": 1.5739,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9207594388430349,
+      "grad_norm": 3.060016522500291,
+      "learning_rate": 1.1758653396581991e-05,
+      "loss": 1.643,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "grad_norm": 3.029460367165218,
+      "learning_rate": 1.1755584009321055e-05,
+      "loss": 1.5523,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.4024,
+      "step": 4784,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.754,
+      "step": 4784,
+      "vm_loss": 0.1331
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.55,
+      "step": 4784,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.4865,
+      "step": 4784,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.0225,
+      "step": 4784,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.6498,
+      "step": 4784,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.1726,
+      "step": 4784,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 0.9209519455206102,
+      "lm_loss": 1.3623,
+      "step": 4784,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 0.9211444521981856,
+      "grad_norm": 2.913313625258337,
+      "learning_rate": 1.1752514451394894e-05,
+      "loss": 1.6203,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9213369588757611,
+      "grad_norm": 3.166308522409298,
+      "learning_rate": 1.1749444723101907e-05,
+      "loss": 1.5781,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9215294655533364,
+      "grad_norm": 3.216965026399961,
+      "learning_rate": 1.1746374824740508e-05,
+      "loss": 1.5486,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9217219722309118,
+      "grad_norm": 3.1735804589544108,
+      "learning_rate": 1.1743304756609133e-05,
+      "loss": 1.5616,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9219144789084871,
+      "grad_norm": 3.2929616053865858,
+      "learning_rate": 1.1740234519006238e-05,
+      "loss": 1.5232,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9221069855860625,
+      "grad_norm": 3.0941134779164434,
+      "learning_rate": 1.173716411223028e-05,
+      "loss": 1.5588,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9222994922636379,
+      "grad_norm": 3.2018675263601333,
+      "learning_rate": 1.1734093536579742e-05,
+      "loss": 1.6539,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "grad_norm": 3.1293930189346195,
+      "learning_rate": 1.173102279235313e-05,
+      "loss": 1.534,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.4114,
+      "step": 4792,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 0.817,
+      "step": 4792,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.6947,
+      "step": 4792,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.7196,
+      "step": 4792,
+      "vm_loss": 0.1193
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.0684,
+      "step": 4792,
+      "vm_loss": 0.0944
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.5546,
+      "step": 4792,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.1797,
+      "step": 4792,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.9224919989412133,
+      "lm_loss": 1.5089,
+      "step": 4792,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.9226845056187887,
+      "grad_norm": 3.2066999457012457,
+      "learning_rate": 1.172795187984896e-05,
+      "loss": 1.5393,
+      "step": 4793
+    },
+    {
+      "epoch": 0.922877012296364,
+      "grad_norm": 3.2279126103638798,
+      "learning_rate": 1.1724880799365757e-05,
+      "loss": 1.6209,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9230695189739394,
+      "grad_norm": 3.0127613118616288,
+      "learning_rate": 1.1721809551202072e-05,
+      "loss": 1.532,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9232620256515148,
+      "grad_norm": 3.215697232902078,
+      "learning_rate": 1.1718738135656471e-05,
+      "loss": 1.6465,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9234545323290901,
+      "grad_norm": 3.3188923915044146,
+      "learning_rate": 1.1715666553027534e-05,
+      "loss": 1.5811,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9236470390066656,
+      "grad_norm": 3.1733251299750815,
+      "learning_rate": 1.171259480361386e-05,
+      "loss": 1.5342,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9238395456842409,
+      "grad_norm": 3.1543324824781265,
+      "learning_rate": 1.1709522887714058e-05,
+      "loss": 1.5265,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "grad_norm": 3.2246702387975525,
+      "learning_rate": 1.1706450805626762e-05,
+      "loss": 1.5997,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.4951,
+      "step": 4800,
+      "vm_loss": 0.1984
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.3107,
+      "step": 4800,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.4419,
+      "step": 4800,
+      "vm_loss": 0.2452
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.6896,
+      "step": 4800,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.3192,
+      "step": 4800,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 0.8301,
+      "step": 4800,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.6884,
+      "step": 4800,
+      "vm_loss": 0.2358
+    },
+    {
+      "epoch": 0.9240320523618163,
+      "lm_loss": 1.5382,
+      "step": 4800,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 0.9242245590393917,
+      "grad_norm": 3.1085704583271836,
+      "learning_rate": 1.1703378557650612e-05,
+      "loss": 1.6157,
+      "step": 4801
+    },
+    {
+      "epoch": 0.924417065716967,
+      "grad_norm": 3.2843450494121336,
+      "learning_rate": 1.1700306144084279e-05,
+      "loss": 1.6467,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9246095723945424,
+      "grad_norm": 3.1540049787069013,
+      "learning_rate": 1.1697233565226433e-05,
+      "loss": 1.6082,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9248020790721179,
+      "grad_norm": 3.0073162271969256,
+      "learning_rate": 1.1694160821375768e-05,
+      "loss": 1.4865,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9249945857496932,
+      "grad_norm": 3.109294811338232,
+      "learning_rate": 1.1691087912831004e-05,
+      "loss": 1.5301,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9251870924272686,
+      "grad_norm": 3.2121240194224856,
+      "learning_rate": 1.168801483989085e-05,
+      "loss": 1.6176,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9253795991048439,
+      "grad_norm": 3.3680376309864997,
+      "learning_rate": 1.1684941602854065e-05,
+      "loss": 1.596,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "grad_norm": 3.156682359555501,
+      "learning_rate": 1.1681868202019397e-05,
+      "loss": 1.5878,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.8091,
+      "step": 4808,
+      "vm_loss": 0.2077
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.4879,
+      "step": 4808,
+      "vm_loss": 0.237
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.499,
+      "step": 4808,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.3911,
+      "step": 4808,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.0723,
+      "step": 4808,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.2265,
+      "step": 4808,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.7926,
+      "step": 4808,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 0.9255721057824193,
+      "lm_loss": 1.7457,
+      "step": 4808,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 0.9257646124599948,
+      "grad_norm": 3.0406175944253895,
+      "learning_rate": 1.167879463768562e-05,
+      "loss": 1.6384,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9259571191375701,
+      "grad_norm": 3.1469707634999065,
+      "learning_rate": 1.1675720910151529e-05,
+      "loss": 1.5512,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9261496258151455,
+      "grad_norm": 3.0160858366002006,
+      "learning_rate": 1.1672647019715926e-05,
+      "loss": 1.5781,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9263421324927208,
+      "grad_norm": 3.125319667772183,
+      "learning_rate": 1.1669572966677634e-05,
+      "loss": 1.5595,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9265346391702962,
+      "grad_norm": 3.2166400520842777,
+      "learning_rate": 1.1666498751335487e-05,
+      "loss": 1.6011,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9267271458478716,
+      "grad_norm": 3.180651338229441,
+      "learning_rate": 1.1663424373988341e-05,
+      "loss": 1.5731,
+      "step": 4814
+    },
+    {
+      "epoch": 0.926919652525447,
+      "grad_norm": 3.146645207159717,
+      "learning_rate": 1.166034983493507e-05,
+      "loss": 1.5267,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "grad_norm": 3.479080493797216,
+      "learning_rate": 1.1657275134474547e-05,
+      "loss": 1.6216,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.814,
+      "step": 4816,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.2522,
+      "step": 4816,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.0369,
+      "step": 4816,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.9353,
+      "step": 4816,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.3556,
+      "step": 4816,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.6112,
+      "step": 4816,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.7312,
+      "step": 4816,
+      "vm_loss": 0.2253
+    },
+    {
+      "epoch": 0.9271121592030224,
+      "lm_loss": 1.4574,
+      "step": 4816,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 0.9273046658805977,
+      "grad_norm": 3.2409739248579377,
+      "learning_rate": 1.165420027290568e-05,
+      "loss": 1.5664,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9274971725581731,
+      "grad_norm": 3.153166675923825,
+      "learning_rate": 1.1651125250527385e-05,
+      "loss": 1.5624,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9276896792357485,
+      "grad_norm": 3.05655847899456,
+      "learning_rate": 1.164805006763859e-05,
+      "loss": 1.5636,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9278821859133238,
+      "grad_norm": 3.056913551557172,
+      "learning_rate": 1.1644974724538243e-05,
+      "loss": 1.4699,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9280746925908993,
+      "grad_norm": 3.1345778526856387,
+      "learning_rate": 1.1641899221525306e-05,
+      "loss": 1.6289,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9282671992684747,
+      "grad_norm": 3.0436407536550805,
+      "learning_rate": 1.1638823558898762e-05,
+      "loss": 1.5319,
+      "step": 4822
+    },
+    {
+      "epoch": 0.92845970594605,
+      "grad_norm": 3.135359867776693,
+      "learning_rate": 1.1635747736957598e-05,
+      "loss": 1.592,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "grad_norm": 3.1211437400886526,
+      "learning_rate": 1.1632671756000831e-05,
+      "loss": 1.5556,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.7428,
+      "step": 4824,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.6523,
+      "step": 4824,
+      "vm_loss": 0.2443
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.4588,
+      "step": 4824,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 0.9358,
+      "step": 4824,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.2836,
+      "step": 4824,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.2809,
+      "step": 4824,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 2.1211,
+      "step": 4824,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 0.9286522126236254,
+      "lm_loss": 1.6074,
+      "step": 4824,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 0.9288447193012007,
+      "grad_norm": 3.3336391041839124,
+      "learning_rate": 1.1629595616327479e-05,
+      "loss": 1.5579,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9290372259787761,
+      "grad_norm": 3.2202924660440546,
+      "learning_rate": 1.1626519318236582e-05,
+      "loss": 1.536,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9292297326563516,
+      "grad_norm": 3.138997658731568,
+      "learning_rate": 1.1623442862027204e-05,
+      "loss": 1.5365,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9294222393339269,
+      "grad_norm": 2.917488871004586,
+      "learning_rate": 1.1620366247998404e-05,
+      "loss": 1.5782,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9296147460115023,
+      "grad_norm": 2.9983484636208986,
+      "learning_rate": 1.1617289476449278e-05,
+      "loss": 1.5491,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9298072526890776,
+      "grad_norm": 3.1563516285528626,
+      "learning_rate": 1.1614212547678923e-05,
+      "loss": 1.5611,
+      "step": 4830
+    },
+    {
+      "epoch": 0.929999759366653,
+      "grad_norm": 3.263707280856024,
+      "learning_rate": 1.1611135461986454e-05,
+      "loss": 1.5301,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "grad_norm": 3.121231101363883,
+      "learning_rate": 1.1608058219671005e-05,
+      "loss": 1.6169,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.1342,
+      "step": 4832,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.4738,
+      "step": 4832,
+      "vm_loss": 0.2279
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.3081,
+      "step": 4832,
+      "vm_loss": 0.1039
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.1317,
+      "step": 4832,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.0903,
+      "step": 4832,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.1748,
+      "step": 4832,
+      "vm_loss": 0.2013
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.3884,
+      "step": 4832,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 0.9301922660442284,
+      "lm_loss": 1.3137,
+      "step": 4832,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 0.9303847727218038,
+      "grad_norm": 3.0644765110699805,
+      "learning_rate": 1.1604980821031726e-05,
+      "loss": 1.5475,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9305772793993792,
+      "grad_norm": 3.1326240378006034,
+      "learning_rate": 1.160190326636778e-05,
+      "loss": 1.5892,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9307697860769546,
+      "grad_norm": 3.2592622504085185,
+      "learning_rate": 1.1598825555978339e-05,
+      "loss": 1.5569,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9309622927545299,
+      "grad_norm": 3.0310499672192215,
+      "learning_rate": 1.1595747690162601e-05,
+      "loss": 1.5168,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9311547994321053,
+      "grad_norm": 3.062408033812157,
+      "learning_rate": 1.1592669669219771e-05,
+      "loss": 1.4959,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9313473061096806,
+      "grad_norm": 3.2068415478092516,
+      "learning_rate": 1.1589591493449073e-05,
+      "loss": 1.5672,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9315398127872561,
+      "grad_norm": 3.264087274465582,
+      "learning_rate": 1.1586513163149746e-05,
+      "loss": 1.5339,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "grad_norm": 3.029278121521542,
+      "learning_rate": 1.158343467862104e-05,
+      "loss": 1.5309,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.5672,
+      "step": 4840,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.3623,
+      "step": 4840,
+      "vm_loss": 0.0931
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.633,
+      "step": 4840,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 0.9986,
+      "step": 4840,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.2554,
+      "step": 4840,
+      "vm_loss": 0.1026
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.0123,
+      "step": 4840,
+      "vm_loss": 0.1055
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.3729,
+      "step": 4840,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 0.9317323194648315,
+      "lm_loss": 1.4821,
+      "step": 4840,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 0.9319248261424068,
+      "grad_norm": 3.063626137379955,
+      "learning_rate": 1.1580356040162228e-05,
+      "loss": 1.5395,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9321173328199822,
+      "grad_norm": 3.0821042931468456,
+      "learning_rate": 1.1577277248072586e-05,
+      "loss": 1.4663,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9323098394975575,
+      "grad_norm": 3.0416556570901956,
+      "learning_rate": 1.1574198302651421e-05,
+      "loss": 1.5184,
+      "step": 4843
+    },
+    {
+      "epoch": 0.932502346175133,
+      "grad_norm": 3.208528869496445,
+      "learning_rate": 1.1571119204198038e-05,
+      "loss": 1.5918,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9326948528527084,
+      "grad_norm": 3.192967532999396,
+      "learning_rate": 1.156803995301177e-05,
+      "loss": 1.5279,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9328873595302837,
+      "grad_norm": 3.1194253738676374,
+      "learning_rate": 1.1564960549391957e-05,
+      "loss": 1.4991,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9330798662078591,
+      "grad_norm": 3.134893381699097,
+      "learning_rate": 1.1561880993637956e-05,
+      "loss": 1.531,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "grad_norm": 3.050254221272552,
+      "learning_rate": 1.155880128604914e-05,
+      "loss": 1.5547,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.5866,
+      "step": 4848,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.1025,
+      "step": 4848,
+      "vm_loss": 0.13
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.2048,
+      "step": 4848,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.7769,
+      "step": 4848,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.4299,
+      "step": 4848,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.48,
+      "step": 4848,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.14,
+      "step": 4848,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 0.9332723728854344,
+      "lm_loss": 1.1291,
+      "step": 4848,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 0.9334648795630098,
+      "grad_norm": 3.1736928114218843,
+      "learning_rate": 1.1555721426924895e-05,
+      "loss": 1.5718,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9336573862405853,
+      "grad_norm": 3.1994502425083837,
+      "learning_rate": 1.1552641416564626e-05,
+      "loss": 1.6227,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9338498929181606,
+      "grad_norm": 3.0221815357792976,
+      "learning_rate": 1.1549561255267746e-05,
+      "loss": 1.5033,
+      "step": 4851
+    },
+    {
+      "epoch": 0.934042399595736,
+      "grad_norm": 3.164211916836797,
+      "learning_rate": 1.1546480943333687e-05,
+      "loss": 1.5602,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9342349062733114,
+      "grad_norm": 3.1552740665362506,
+      "learning_rate": 1.1543400481061895e-05,
+      "loss": 1.5503,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9344274129508867,
+      "grad_norm": 3.2818367107285193,
+      "learning_rate": 1.154031986875183e-05,
+      "loss": 1.6392,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9346199196284621,
+      "grad_norm": 3.156085070543579,
+      "learning_rate": 1.1537239106702968e-05,
+      "loss": 1.4913,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "grad_norm": 3.2001590277133745,
+      "learning_rate": 1.15341581952148e-05,
+      "loss": 1.5564,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.8592,
+      "step": 4856,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.4133,
+      "step": 4856,
+      "vm_loss": 0.1063
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.786,
+      "step": 4856,
+      "vm_loss": 0.1412
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.2183,
+      "step": 4856,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.1631,
+      "step": 4856,
+      "vm_loss": 0.1327
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.0705,
+      "step": 4856,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.1934,
+      "step": 4856,
+      "vm_loss": 0.0829
+    },
+    {
+      "epoch": 0.9348124263060374,
+      "lm_loss": 1.7506,
+      "step": 4856,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 0.9350049329836129,
+      "grad_norm": 3.168393377303047,
+      "learning_rate": 1.1531077134586826e-05,
+      "loss": 1.497,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9351974396611883,
+      "grad_norm": 3.0979508166853726,
+      "learning_rate": 1.1527995925118567e-05,
+      "loss": 1.58,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9353899463387636,
+      "grad_norm": 3.2282771126891285,
+      "learning_rate": 1.1524914567109553e-05,
+      "loss": 1.457,
+      "step": 4859
+    },
+    {
+      "epoch": 0.935582453016339,
+      "grad_norm": 3.1004618219242768,
+      "learning_rate": 1.1521833060859338e-05,
+      "loss": 1.5539,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9357749596939143,
+      "grad_norm": 3.165984828973839,
+      "learning_rate": 1.1518751406667477e-05,
+      "loss": 1.5899,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9359674663714898,
+      "grad_norm": 3.0278481131988086,
+      "learning_rate": 1.1515669604833546e-05,
+      "loss": 1.5914,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9361599730490652,
+      "grad_norm": 3.099379791712217,
+      "learning_rate": 1.1512587655657144e-05,
+      "loss": 1.4938,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "grad_norm": 3.143625551503428,
+      "learning_rate": 1.1509505559437866e-05,
+      "loss": 1.5861,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.5652,
+      "step": 4864,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.3394,
+      "step": 4864,
+      "vm_loss": 0.1083
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.3699,
+      "step": 4864,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.7026,
+      "step": 4864,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.9699,
+      "step": 4864,
+      "vm_loss": 0.2315
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.6658,
+      "step": 4864,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.456,
+      "step": 4864,
+      "vm_loss": 0.2297
+    },
+    {
+      "epoch": 0.9363524797266405,
+      "lm_loss": 1.4579,
+      "step": 4864,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.9365449864042159,
+      "grad_norm": 3.0087095550939846,
+      "learning_rate": 1.1506423316475336e-05,
+      "loss": 1.6279,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9367374930817913,
+      "grad_norm": 3.080304278923635,
+      "learning_rate": 1.1503340927069189e-05,
+      "loss": 1.6414,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9369299997593666,
+      "grad_norm": 3.0244907233347624,
+      "learning_rate": 1.1500258391519068e-05,
+      "loss": 1.5328,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9371225064369421,
+      "grad_norm": 3.0172207372915207,
+      "learning_rate": 1.149717571012464e-05,
+      "loss": 1.5867,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9373150131145174,
+      "grad_norm": 3.171742951196958,
+      "learning_rate": 1.1494092883185574e-05,
+      "loss": 1.5866,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9375075197920928,
+      "grad_norm": 3.0540213613615474,
+      "learning_rate": 1.1491009911001572e-05,
+      "loss": 1.4671,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9377000264696682,
+      "grad_norm": 3.0893632063474312,
+      "learning_rate": 1.1487926793872326e-05,
+      "loss": 1.5704,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "grad_norm": 3.1753938354348143,
+      "learning_rate": 1.1484843532097558e-05,
+      "loss": 1.5911,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.5414,
+      "step": 4872,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.4615,
+      "step": 4872,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.1818,
+      "step": 4872,
+      "vm_loss": 0.2019
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 0.9545,
+      "step": 4872,
+      "vm_loss": 0.1239
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.4708,
+      "step": 4872,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.3805,
+      "step": 4872,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.0126,
+      "step": 4872,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 0.9378925331472435,
+      "lm_loss": 1.9353,
+      "step": 4872,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 0.938085039824819,
+      "grad_norm": 3.167477045695026,
+      "learning_rate": 1.1481760125977006e-05,
+      "loss": 1.5535,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9382775465023943,
+      "grad_norm": 3.0832071739105547,
+      "learning_rate": 1.147867657581041e-05,
+      "loss": 1.5618,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9384700531799697,
+      "grad_norm": 3.063530755053782,
+      "learning_rate": 1.1475592881897532e-05,
+      "loss": 1.5471,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9386625598575451,
+      "grad_norm": 3.0586462970968653,
+      "learning_rate": 1.147250904453815e-05,
+      "loss": 1.5151,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9388550665351204,
+      "grad_norm": 3.209239511768028,
+      "learning_rate": 1.1469425064032051e-05,
+      "loss": 1.6254,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9390475732126958,
+      "grad_norm": 3.1461683265244993,
+      "learning_rate": 1.1466340940679033e-05,
+      "loss": 1.6053,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9392400798902711,
+      "grad_norm": 3.2032952209867203,
+      "learning_rate": 1.1463256674778918e-05,
+      "loss": 1.5755,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "grad_norm": 3.1882747477272666,
+      "learning_rate": 1.1460172266631532e-05,
+      "loss": 1.5079,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.0817,
+      "step": 4880,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 0.8867,
+      "step": 4880,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.3817,
+      "step": 4880,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.8459,
+      "step": 4880,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.4768,
+      "step": 4880,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.5947,
+      "step": 4880,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.4836,
+      "step": 4880,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 0.9394325865678466,
+      "lm_loss": 1.4317,
+      "step": 4880,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 0.939625093245422,
+      "grad_norm": 3.0943695939502724,
+      "learning_rate": 1.145708771653672e-05,
+      "loss": 1.5843,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9398175999229973,
+      "grad_norm": 3.05885073975982,
+      "learning_rate": 1.1454003024794344e-05,
+      "loss": 1.5338,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9400101066005727,
+      "grad_norm": 3.283124191345047,
+      "learning_rate": 1.145091819170427e-05,
+      "loss": 1.5193,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9402026132781481,
+      "grad_norm": 3.245001455358806,
+      "learning_rate": 1.1447833217566383e-05,
+      "loss": 1.6128,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9403951199557234,
+      "grad_norm": 3.1659155142818802,
+      "learning_rate": 1.144474810268059e-05,
+      "loss": 1.4824,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9405876266332989,
+      "grad_norm": 3.0214319883676155,
+      "learning_rate": 1.1441662847346794e-05,
+      "loss": 1.4376,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9407801333108742,
+      "grad_norm": 3.1081845228797085,
+      "learning_rate": 1.1438577451864923e-05,
+      "loss": 1.537,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "grad_norm": 3.0459142778145534,
+      "learning_rate": 1.1435491916534919e-05,
+      "loss": 1.6073,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.1136,
+      "step": 4888,
+      "vm_loss": 0.2038
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.8127,
+      "step": 4888,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.3128,
+      "step": 4888,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 0.8779,
+      "step": 4888,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.897,
+      "step": 4888,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.0679,
+      "step": 4888,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.0779,
+      "step": 4888,
+      "vm_loss": 0.1063
+    },
+    {
+      "epoch": 0.9409726399884496,
+      "lm_loss": 1.1264,
+      "step": 4888,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.941165146666025,
+      "grad_norm": 3.1837593792149765,
+      "learning_rate": 1.1432406241656737e-05,
+      "loss": 1.4771,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9413576533436003,
+      "grad_norm": 3.0957834234190247,
+      "learning_rate": 1.1429320427530343e-05,
+      "loss": 1.533,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9415501600211758,
+      "grad_norm": 3.3753160246525225,
+      "learning_rate": 1.1426234474455718e-05,
+      "loss": 1.5847,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9417426666987511,
+      "grad_norm": 3.2750259226281333,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 1.5888,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9419351733763265,
+      "grad_norm": 3.042623538360436,
+      "learning_rate": 1.142006215266176e-05,
+      "loss": 1.5139,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9421276800539019,
+      "grad_norm": 3.0984437638331106,
+      "learning_rate": 1.1416975784542454e-05,
+      "loss": 1.587,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9423201867314772,
+      "grad_norm": 3.1520703222636204,
+      "learning_rate": 1.1413889278674977e-05,
+      "loss": 1.5697,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "grad_norm": 3.029448965933574,
+      "learning_rate": 1.1410802635359373e-05,
+      "loss": 1.4825,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.4535,
+      "step": 4896,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.6739,
+      "step": 4896,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.1151,
+      "step": 4896,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.7162,
+      "step": 4896,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.3903,
+      "step": 4896,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 0.9864,
+      "step": 4896,
+      "vm_loss": 0.2469
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.5197,
+      "step": 4896,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.9425126934090526,
+      "lm_loss": 1.5114,
+      "step": 4896,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 0.9427052000866281,
+      "grad_norm": 3.0435527137690266,
+      "learning_rate": 1.14077158548957e-05,
+      "loss": 1.6032,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9428977067642034,
+      "grad_norm": 3.162895857751957,
+      "learning_rate": 1.1404628937584036e-05,
+      "loss": 1.5725,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9430902134417788,
+      "grad_norm": 3.143355252823054,
+      "learning_rate": 1.1401541883724474e-05,
+      "loss": 1.5566,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9432827201193541,
+      "grad_norm": 3.001866948364369,
+      "learning_rate": 1.1398454693617107e-05,
+      "loss": 1.5556,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9434752267969295,
+      "grad_norm": 3.311727949517917,
+      "learning_rate": 1.139536736756205e-05,
+      "loss": 1.6002,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9436677334745049,
+      "grad_norm": 3.148587097344572,
+      "learning_rate": 1.1392279905859438e-05,
+      "loss": 1.5546,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9438602401520803,
+      "grad_norm": 3.2665995722471854,
+      "learning_rate": 1.1389192308809402e-05,
+      "loss": 1.521,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "grad_norm": 3.144212409101037,
+      "learning_rate": 1.1386104576712105e-05,
+      "loss": 1.6015,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.116,
+      "step": 4904,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 0.913,
+      "step": 4904,
+      "vm_loss": 0.1163
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 0.9958,
+      "step": 4904,
+      "vm_loss": 0.2041
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.2078,
+      "step": 4904,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.2221,
+      "step": 4904,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.5112,
+      "step": 4904,
+      "vm_loss": 0.1252
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.4032,
+      "step": 4904,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 0.9440527468296557,
+      "lm_loss": 1.0256,
+      "step": 4904,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 0.944245253507231,
+      "grad_norm": 3.235582188083785,
+      "learning_rate": 1.1383016709867706e-05,
+      "loss": 1.5676,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9444377601848064,
+      "grad_norm": 3.058808523747482,
+      "learning_rate": 1.1379928708576394e-05,
+      "loss": 1.5706,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9446302668623818,
+      "grad_norm": 3.064475529218613,
+      "learning_rate": 1.137684057313835e-05,
+      "loss": 1.4793,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9448227735399571,
+      "grad_norm": 3.0214177007846708,
+      "learning_rate": 1.1373752303853791e-05,
+      "loss": 1.5653,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9450152802175326,
+      "grad_norm": 3.1990952707037126,
+      "learning_rate": 1.1370663901022934e-05,
+      "loss": 1.5365,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9452077868951079,
+      "grad_norm": 3.250619324455714,
+      "learning_rate": 1.1367575364946006e-05,
+      "loss": 1.5833,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9454002935726833,
+      "grad_norm": 3.2049096288589527,
+      "learning_rate": 1.136448669592326e-05,
+      "loss": 1.4932,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "grad_norm": 3.2144282272893228,
+      "learning_rate": 1.1361397894254949e-05,
+      "loss": 1.5363,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 0.8845,
+      "step": 4912,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.5232,
+      "step": 4912,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.5202,
+      "step": 4912,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.5089,
+      "step": 4912,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 0.9956,
+      "step": 4912,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.3623,
+      "step": 4912,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.3111,
+      "step": 4912,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 0.9455928002502587,
+      "lm_loss": 1.616,
+      "step": 4912,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 0.945785306927834,
+      "grad_norm": 3.0988056937748083,
+      "learning_rate": 1.1358308960241341e-05,
+      "loss": 1.5214,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9459778136054094,
+      "grad_norm": 3.05750733391802,
+      "learning_rate": 1.1355219894182733e-05,
+      "loss": 1.4986,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9461703202829849,
+      "grad_norm": 3.0882525545721253,
+      "learning_rate": 1.1352130696379406e-05,
+      "loss": 1.5015,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9463628269605602,
+      "grad_norm": 3.013030011802002,
+      "learning_rate": 1.1349041367131677e-05,
+      "loss": 1.5253,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9465553336381356,
+      "grad_norm": 3.278662237320305,
+      "learning_rate": 1.134595190673987e-05,
+      "loss": 1.5729,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9467478403157109,
+      "grad_norm": 3.201480421866968,
+      "learning_rate": 1.1342862315504317e-05,
+      "loss": 1.5561,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9469403469932863,
+      "grad_norm": 3.1123366547274314,
+      "learning_rate": 1.1339772593725367e-05,
+      "loss": 1.5354,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "grad_norm": 3.2351873894798415,
+      "learning_rate": 1.133668274170338e-05,
+      "loss": 1.5292,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.3133,
+      "step": 4920,
+      "vm_loss": 0.1023
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.2972,
+      "step": 4920,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.823,
+      "step": 4920,
+      "vm_loss": 0.187
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.1321,
+      "step": 4920,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.3792,
+      "step": 4920,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.3652,
+      "step": 4920,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.8067,
+      "step": 4920,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 0.9471328536708618,
+      "lm_loss": 1.2198,
+      "step": 4920,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 0.9473253603484371,
+      "grad_norm": 3.2310810068224383,
+      "learning_rate": 1.133359275973873e-05,
+      "loss": 1.5215,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9475178670260125,
+      "grad_norm": 3.2755841817760167,
+      "learning_rate": 1.13305026481318e-05,
+      "loss": 1.5601,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9477103737035878,
+      "grad_norm": 2.910984151233412,
+      "learning_rate": 1.1327412407182994e-05,
+      "loss": 1.4745,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9479028803811632,
+      "grad_norm": 2.8948142549989857,
+      "learning_rate": 1.1324322037192716e-05,
+      "loss": 1.523,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9480953870587386,
+      "grad_norm": 3.139939339963268,
+      "learning_rate": 1.1321231538461395e-05,
+      "loss": 1.5119,
+      "step": 4925
+    },
+    {
+      "epoch": 0.948287893736314,
+      "grad_norm": 3.0586700870669947,
+      "learning_rate": 1.1318140911289465e-05,
+      "loss": 1.6178,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9484804004138894,
+      "grad_norm": 3.177624029359829,
+      "learning_rate": 1.1315050155977374e-05,
+      "loss": 1.5625,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "grad_norm": 3.13465913917213,
+      "learning_rate": 1.1311959272825586e-05,
+      "loss": 1.5137,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.0588,
+      "step": 4928,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.5185,
+      "step": 4928,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.8094,
+      "step": 4928,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.2392,
+      "step": 4928,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.3031,
+      "step": 4928,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.3994,
+      "step": 4928,
+      "vm_loss": 0.2161
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.8366,
+      "step": 4928,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.9486729070914648,
+      "lm_loss": 1.0128,
+      "step": 4928,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 0.9488654137690401,
+      "grad_norm": 3.063110072336763,
+      "learning_rate": 1.1308868262134566e-05,
+      "loss": 1.5292,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9490579204466155,
+      "grad_norm": 3.2467589947739204,
+      "learning_rate": 1.1305777124204812e-05,
+      "loss": 1.5051,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9492504271241908,
+      "grad_norm": 3.1607679070207437,
+      "learning_rate": 1.1302685859336812e-05,
+      "loss": 1.52,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9494429338017663,
+      "grad_norm": 3.3017046972525717,
+      "learning_rate": 1.1299594467831079e-05,
+      "loss": 1.5853,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9496354404793417,
+      "grad_norm": 3.090866892150214,
+      "learning_rate": 1.1296502949988142e-05,
+      "loss": 1.5195,
+      "step": 4933
+    },
+    {
+      "epoch": 0.949827947156917,
+      "grad_norm": 3.1211796579681583,
+      "learning_rate": 1.1293411306108526e-05,
+      "loss": 1.5689,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9500204538344924,
+      "grad_norm": 3.2692116963309683,
+      "learning_rate": 1.1290319536492786e-05,
+      "loss": 1.5567,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "grad_norm": 3.1415776875048205,
+      "learning_rate": 1.1287227641441478e-05,
+      "loss": 1.5173,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 0.7196,
+      "step": 4936,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.6534,
+      "step": 4936,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.4973,
+      "step": 4936,
+      "vm_loss": 0.1301
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.3946,
+      "step": 4936,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.5871,
+      "step": 4936,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.4575,
+      "step": 4936,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.3059,
+      "step": 4936,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 0.9502129605120677,
+      "lm_loss": 1.4286,
+      "step": 4936,
+      "vm_loss": 0.1103
+    },
+    {
+      "epoch": 0.9504054671896431,
+      "grad_norm": 3.222931344598351,
+      "learning_rate": 1.1284135621255177e-05,
+      "loss": 1.5321,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9505979738672186,
+      "grad_norm": 3.1184862502659505,
+      "learning_rate": 1.1281043476234461e-05,
+      "loss": 1.4561,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9507904805447939,
+      "grad_norm": 3.317785224622354,
+      "learning_rate": 1.127795120667993e-05,
+      "loss": 1.5369,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9509829872223693,
+      "grad_norm": 3.2377981450902524,
+      "learning_rate": 1.1274858812892198e-05,
+      "loss": 1.5141,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9511754938999446,
+      "grad_norm": 3.3142560982923057,
+      "learning_rate": 1.1271766295171873e-05,
+      "loss": 1.5112,
+      "step": 4941
+    },
+    {
+      "epoch": 0.95136800057752,
+      "grad_norm": 3.4039945829062193,
+      "learning_rate": 1.1268673653819594e-05,
+      "loss": 1.5644,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9515605072550954,
+      "grad_norm": 3.1422287825617863,
+      "learning_rate": 1.1265580889136007e-05,
+      "loss": 1.5406,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "grad_norm": 3.018636398312615,
+      "learning_rate": 1.1262488001421765e-05,
+      "loss": 1.4895,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 0.9212,
+      "step": 4944,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.36,
+      "step": 4944,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.2078,
+      "step": 4944,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.5963,
+      "step": 4944,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.5913,
+      "step": 4944,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.3362,
+      "step": 4944,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 0.9535,
+      "step": 4944,
+      "vm_loss": 0.2216
+    },
+    {
+      "epoch": 0.9517530139326708,
+      "lm_loss": 1.0879,
+      "step": 4944,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 0.9519455206102462,
+      "grad_norm": 3.2544283354247736,
+      "learning_rate": 1.1259394990977541e-05,
+      "loss": 1.472,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9521380272878216,
+      "grad_norm": 3.2095407748938976,
+      "learning_rate": 1.1256301858104007e-05,
+      "loss": 1.4981,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9523305339653969,
+      "grad_norm": 3.524097672988114,
+      "learning_rate": 1.125320860310186e-05,
+      "loss": 1.6084,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9525230406429723,
+      "grad_norm": 3.2341711913449536,
+      "learning_rate": 1.1250115226271804e-05,
+      "loss": 1.4711,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9527155473205476,
+      "grad_norm": 3.4667260649290346,
+      "learning_rate": 1.1247021727914555e-05,
+      "loss": 1.5321,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9529080539981231,
+      "grad_norm": 3.0880161582827643,
+      "learning_rate": 1.124392810833084e-05,
+      "loss": 1.5284,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9531005606756985,
+      "grad_norm": 3.0167625032502743,
+      "learning_rate": 1.1240834367821398e-05,
+      "loss": 1.5366,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "grad_norm": 3.127109342620602,
+      "learning_rate": 1.1237740506686983e-05,
+      "loss": 1.5798,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 2.0237,
+      "step": 4952,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.2665,
+      "step": 4952,
+      "vm_loss": 0.218
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.6865,
+      "step": 4952,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.0832,
+      "step": 4952,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.2294,
+      "step": 4952,
+      "vm_loss": 0.2231
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 2.0328,
+      "step": 4952,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.4052,
+      "step": 4952,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 0.9532930673532738,
+      "lm_loss": 1.5694,
+      "step": 4952,
+      "vm_loss": 0.2345
+    },
+    {
+      "epoch": 0.9534855740308492,
+      "grad_norm": 3.07905336068598,
+      "learning_rate": 1.1234646525228353e-05,
+      "loss": 1.5861,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9536780807084245,
+      "grad_norm": 3.0700224772077807,
+      "learning_rate": 1.1231552423746284e-05,
+      "loss": 1.5368,
+      "step": 4954
+    },
+    {
+      "epoch": 0.953870587386,
+      "grad_norm": 3.1117391854579144,
+      "learning_rate": 1.1228458202541566e-05,
+      "loss": 1.4671,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9540630940635754,
+      "grad_norm": 3.1579746964044366,
+      "learning_rate": 1.122536386191499e-05,
+      "loss": 1.56,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9542556007411507,
+      "grad_norm": 3.288425003716889,
+      "learning_rate": 1.1222269402167372e-05,
+      "loss": 1.5579,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9544481074187261,
+      "grad_norm": 3.1795499598907906,
+      "learning_rate": 1.1219174823599532e-05,
+      "loss": 1.5184,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9546406140963015,
+      "grad_norm": 3.24444769593709,
+      "learning_rate": 1.1216080126512298e-05,
+      "loss": 1.5073,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "grad_norm": 3.150687090032631,
+      "learning_rate": 1.1212985311206522e-05,
+      "loss": 1.5568,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.6337,
+      "step": 4960,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.7092,
+      "step": 4960,
+      "vm_loss": 0.2044
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 0.8714,
+      "step": 4960,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.6399,
+      "step": 4960,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.3171,
+      "step": 4960,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 0.9431,
+      "step": 4960,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.8818,
+      "step": 4960,
+      "vm_loss": 0.2111
+    },
+    {
+      "epoch": 0.9548331207738768,
+      "lm_loss": 1.3153,
+      "step": 4960,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 0.9550256274514523,
+      "grad_norm": 3.070799484841203,
+      "learning_rate": 1.120989037798305e-05,
+      "loss": 1.5702,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9552181341290276,
+      "grad_norm": 3.0812185508135452,
+      "learning_rate": 1.120679532714276e-05,
+      "loss": 1.5618,
+      "step": 4962
+    },
+    {
+      "epoch": 0.955410640806603,
+      "grad_norm": 3.0407337386431874,
+      "learning_rate": 1.120370015898652e-05,
+      "loss": 1.5118,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9556031474841784,
+      "grad_norm": 3.1324404134633306,
+      "learning_rate": 1.1200604873815226e-05,
+      "loss": 1.5251,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9557956541617537,
+      "grad_norm": 3.177728013772863,
+      "learning_rate": 1.1197509471929779e-05,
+      "loss": 1.5302,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9559881608393291,
+      "grad_norm": 3.3034508925586885,
+      "learning_rate": 1.119441395363109e-05,
+      "loss": 1.592,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9561806675169044,
+      "grad_norm": 3.2302100345526634,
+      "learning_rate": 1.1191318319220085e-05,
+      "loss": 1.5793,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "grad_norm": 3.205608358848487,
+      "learning_rate": 1.1188222568997697e-05,
+      "loss": 1.5227,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.8363,
+      "step": 4968,
+      "vm_loss": 0.1191
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.2392,
+      "step": 4968,
+      "vm_loss": 0.1407
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.4104,
+      "step": 4968,
+      "vm_loss": 0.2182
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.4452,
+      "step": 4968,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.5054,
+      "step": 4968,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.0966,
+      "step": 4968,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.5982,
+      "step": 4968,
+      "vm_loss": 0.2177
+    },
+    {
+      "epoch": 0.9563731741944799,
+      "lm_loss": 1.3861,
+      "step": 4968,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 0.9565656808720553,
+      "grad_norm": 2.9280178432610175,
+      "learning_rate": 1.1185126703264875e-05,
+      "loss": 1.4848,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9567581875496306,
+      "grad_norm": 3.051466697417576,
+      "learning_rate": 1.1182030722322571e-05,
+      "loss": 1.4295,
+      "step": 4970
+    },
+    {
+      "epoch": 0.956950694227206,
+      "grad_norm": 3.047311201224443,
+      "learning_rate": 1.1178934626471762e-05,
+      "loss": 1.3757,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9571432009047813,
+      "grad_norm": 2.9806130932304327,
+      "learning_rate": 1.1175838416013425e-05,
+      "loss": 1.53,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9573357075823568,
+      "grad_norm": 3.1935678923291047,
+      "learning_rate": 1.1172742091248548e-05,
+      "loss": 1.5851,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9575282142599322,
+      "grad_norm": 3.3291942217519286,
+      "learning_rate": 1.1169645652478138e-05,
+      "loss": 1.5695,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9577207209375075,
+      "grad_norm": 3.262252693044439,
+      "learning_rate": 1.1166549100003203e-05,
+      "loss": 1.585,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "grad_norm": 3.397908129430199,
+      "learning_rate": 1.1163452434124773e-05,
+      "loss": 1.5269,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.2388,
+      "step": 4976,
+      "vm_loss": 0.1189
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.6814,
+      "step": 4976,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.6699,
+      "step": 4976,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.2209,
+      "step": 4976,
+      "vm_loss": 0.2209
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.4817,
+      "step": 4976,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.4436,
+      "step": 4976,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.0202,
+      "step": 4976,
+      "vm_loss": 0.2728
+    },
+    {
+      "epoch": 0.9579132276150829,
+      "lm_loss": 1.3588,
+      "step": 4976,
+      "vm_loss": 0.22
+    },
+    {
+      "epoch": 0.9581057342926583,
+      "grad_norm": 3.1657081991723137,
+      "learning_rate": 1.1160355655143883e-05,
+      "loss": 1.6135,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9582982409702336,
+      "grad_norm": 2.99265792139987,
+      "learning_rate": 1.1157258763361577e-05,
+      "loss": 1.4804,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9584907476478091,
+      "grad_norm": 3.2195053117486006,
+      "learning_rate": 1.1154161759078912e-05,
+      "loss": 1.487,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9586832543253844,
+      "grad_norm": 3.0686529605655664,
+      "learning_rate": 1.1151064642596958e-05,
+      "loss": 1.4882,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9588757610029598,
+      "grad_norm": 3.2883065855922937,
+      "learning_rate": 1.1147967414216796e-05,
+      "loss": 1.5837,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9590682676805352,
+      "grad_norm": 3.2242174201478404,
+      "learning_rate": 1.1144870074239514e-05,
+      "loss": 1.5435,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9592607743581105,
+      "grad_norm": 3.161845572759309,
+      "learning_rate": 1.1141772622966214e-05,
+      "loss": 1.5549,
+      "step": 4983
+    },
+    {
+      "epoch": 0.959453281035686,
+      "grad_norm": 3.2259888525380815,
+      "learning_rate": 1.1138675060698007e-05,
+      "loss": 1.5718,
+      "step": 4984
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.0418,
+      "step": 4984,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.0052,
+      "step": 4984,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.2592,
+      "step": 4984,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.4809,
+      "step": 4984,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.409,
+      "step": 4984,
+      "vm_loss": 0.0994
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.356,
+      "step": 4984,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.2268,
+      "step": 4984,
+      "vm_loss": 0.1177
+    },
+    {
+      "epoch": 0.959453281035686,
+      "lm_loss": 1.9241,
+      "step": 4984,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 0.9596457877132613,
+      "grad_norm": 3.274063562252294,
+      "learning_rate": 1.1135577387736016e-05,
+      "loss": 1.444,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9598382943908367,
+      "grad_norm": 3.0000261767060614,
+      "learning_rate": 1.1132479604381377e-05,
+      "loss": 1.6082,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9600308010684121,
+      "grad_norm": 3.049413535990162,
+      "learning_rate": 1.1129381710935229e-05,
+      "loss": 1.5844,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9602233077459874,
+      "grad_norm": 3.132131404265152,
+      "learning_rate": 1.112628370769873e-05,
+      "loss": 1.5485,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9604158144235628,
+      "grad_norm": 2.999358898174001,
+      "learning_rate": 1.1123185594973047e-05,
+      "loss": 1.4486,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9606083211011381,
+      "grad_norm": 3.255676728811577,
+      "learning_rate": 1.1120087373059351e-05,
+      "loss": 1.5545,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9608008277787136,
+      "grad_norm": 3.311115513310606,
+      "learning_rate": 1.1116989042258837e-05,
+      "loss": 1.5843,
+      "step": 4991
+    },
+    {
+      "epoch": 0.960993334456289,
+      "grad_norm": 3.2665485107454475,
+      "learning_rate": 1.1113890602872695e-05,
+      "loss": 1.5794,
+      "step": 4992
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 0.9666,
+      "step": 4992,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.5262,
+      "step": 4992,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.4755,
+      "step": 4992,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.5749,
+      "step": 4992,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.5446,
+      "step": 4992,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.9874,
+      "step": 4992,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.4392,
+      "step": 4992,
+      "vm_loss": 0.1407
+    },
+    {
+      "epoch": 0.960993334456289,
+      "lm_loss": 1.1821,
+      "step": 4992,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 0.9611858411338643,
+      "grad_norm": 3.1126080864229255,
+      "learning_rate": 1.1110792055202135e-05,
+      "loss": 1.4613,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9613783478114397,
+      "grad_norm": 3.277587209054993,
+      "learning_rate": 1.1107693399548378e-05,
+      "loss": 1.5535,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9615708544890151,
+      "grad_norm": 3.131997486153474,
+      "learning_rate": 1.110459463621265e-05,
+      "loss": 1.5459,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9617633611665904,
+      "grad_norm": 3.165541572953028,
+      "learning_rate": 1.110149576549619e-05,
+      "loss": 1.5136,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9619558678441659,
+      "grad_norm": 3.098776074179078,
+      "learning_rate": 1.1098396787700252e-05,
+      "loss": 1.4477,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9621483745217412,
+      "grad_norm": 3.104081992397341,
+      "learning_rate": 1.1095297703126094e-05,
+      "loss": 1.5055,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9623408811993166,
+      "grad_norm": 3.111710358114047,
+      "learning_rate": 1.1092198512074983e-05,
+      "loss": 1.5488,
+      "step": 4999
+    },
+    {
+      "epoch": 0.962533387876892,
+      "grad_norm": 3.1759162888321497,
+      "learning_rate": 1.1089099214848202e-05,
+      "loss": 1.4918,
+      "step": 5000
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.7775,
+      "step": 5000,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 0.6308,
+      "step": 5000,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.2442,
+      "step": 5000,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.1558,
+      "step": 5000,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.215,
+      "step": 5000,
+      "vm_loss": 0.1017
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.4159,
+      "step": 5000,
+      "vm_loss": 0.2236
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.6202,
+      "step": 5000,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.962533387876892,
+      "lm_loss": 1.4315,
+      "step": 5000,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 0.9627258945544673,
+      "grad_norm": 3.1277312529275623,
+      "learning_rate": 1.1085999811747048e-05,
+      "loss": 1.4734,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9629184012320428,
+      "grad_norm": 3.0358493979105043,
+      "learning_rate": 1.1082900303072814e-05,
+      "loss": 1.4905,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9631109079096181,
+      "grad_norm": 3.009859005626985,
+      "learning_rate": 1.107980068912682e-05,
+      "loss": 1.4066,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9633034145871935,
+      "grad_norm": 3.2689938698285483,
+      "learning_rate": 1.1076700970210383e-05,
+      "loss": 1.5161,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9634959212647689,
+      "grad_norm": 3.2371356512914633,
+      "learning_rate": 1.1073601146624837e-05,
+      "loss": 1.5081,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9636884279423442,
+      "grad_norm": 3.2563042849276025,
+      "learning_rate": 1.1070501218671526e-05,
+      "loss": 1.5896,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9638809346199196,
+      "grad_norm": 3.1266000777353633,
+      "learning_rate": 1.1067401186651799e-05,
+      "loss": 1.6047,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "grad_norm": 3.0572994685307306,
+      "learning_rate": 1.1064301050867022e-05,
+      "loss": 1.4774,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.0822,
+      "step": 5008,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.0384,
+      "step": 5008,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.4654,
+      "step": 5008,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.6399,
+      "step": 5008,
+      "vm_loss": 0.1397
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.1153,
+      "step": 5008,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.116,
+      "step": 5008,
+      "vm_loss": 0.1136
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.4629,
+      "step": 5008,
+      "vm_loss": 0.214
+    },
+    {
+      "epoch": 0.9640734412974951,
+      "lm_loss": 1.4472,
+      "step": 5008,
+      "vm_loss": 0.2289
+    },
+    {
+      "epoch": 0.9642659479750704,
+      "grad_norm": 3.0707765392144863,
+      "learning_rate": 1.1061200811618564e-05,
+      "loss": 1.5074,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9644584546526458,
+      "grad_norm": 3.11302522345523,
+      "learning_rate": 1.1058100469207811e-05,
+      "loss": 1.5559,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9646509613302211,
+      "grad_norm": 3.1106507784028286,
+      "learning_rate": 1.1055000023936155e-05,
+      "loss": 1.5193,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9648434680077965,
+      "grad_norm": 3.0946780148701527,
+      "learning_rate": 1.1051899476105001e-05,
+      "loss": 1.5627,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9650359746853719,
+      "grad_norm": 3.006655990895208,
+      "learning_rate": 1.104879882601576e-05,
+      "loss": 1.5304,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9652284813629473,
+      "grad_norm": 3.0771532295837507,
+      "learning_rate": 1.1045698073969855e-05,
+      "loss": 1.4615,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9654209880405227,
+      "grad_norm": 3.056591900513941,
+      "learning_rate": 1.1042597220268717e-05,
+      "loss": 1.4858,
+      "step": 5015
+    },
+    {
+      "epoch": 0.965613494718098,
+      "grad_norm": 3.1226024561361325,
+      "learning_rate": 1.1039496265213792e-05,
+      "loss": 1.5535,
+      "step": 5016
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.5655,
+      "step": 5016,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.4469,
+      "step": 5016,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.0383,
+      "step": 5016,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.0489,
+      "step": 5016,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.0831,
+      "step": 5016,
+      "vm_loss": 0.2744
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.2477,
+      "step": 5016,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 2.135,
+      "step": 5016,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 0.965613494718098,
+      "lm_loss": 1.4885,
+      "step": 5016,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 0.9658060013956734,
+      "grad_norm": 3.227004644371464,
+      "learning_rate": 1.103639520910653e-05,
+      "loss": 1.4961,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9659985080732488,
+      "grad_norm": 3.13013353110654,
+      "learning_rate": 1.1033294052248396e-05,
+      "loss": 1.4988,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9661910147508241,
+      "grad_norm": 3.101879817042813,
+      "learning_rate": 1.103019279494086e-05,
+      "loss": 1.4998,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9663835214283996,
+      "grad_norm": 3.197413848021874,
+      "learning_rate": 1.1027091437485404e-05,
+      "loss": 1.5503,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9665760281059749,
+      "grad_norm": 3.0283583462031967,
+      "learning_rate": 1.1023989980183518e-05,
+      "loss": 1.5142,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9667685347835503,
+      "grad_norm": 3.302367679081862,
+      "learning_rate": 1.1020888423336706e-05,
+      "loss": 1.5303,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9669610414611257,
+      "grad_norm": 3.192947791422235,
+      "learning_rate": 1.1017786767246479e-05,
+      "loss": 1.5018,
+      "step": 5023
+    },
+    {
+      "epoch": 0.967153548138701,
+      "grad_norm": 3.164070760564195,
+      "learning_rate": 1.1014685012214356e-05,
+      "loss": 1.5219,
+      "step": 5024
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 0.9907,
+      "step": 5024,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.4281,
+      "step": 5024,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.162,
+      "step": 5024,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.2192,
+      "step": 5024,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.3241,
+      "step": 5024,
+      "vm_loss": 0.2218
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.4035,
+      "step": 5024,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.3924,
+      "step": 5024,
+      "vm_loss": 0.2563
+    },
+    {
+      "epoch": 0.967153548138701,
+      "lm_loss": 1.7133,
+      "step": 5024,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 0.9673460548162764,
+      "grad_norm": 3.1564790637080673,
+      "learning_rate": 1.1011583158541872e-05,
+      "loss": 1.5896,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9675385614938519,
+      "grad_norm": 3.110982167192257,
+      "learning_rate": 1.1008481206530559e-05,
+      "loss": 1.5237,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9677310681714272,
+      "grad_norm": 3.066232119994031,
+      "learning_rate": 1.1005379156481975e-05,
+      "loss": 1.4477,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9679235748490026,
+      "grad_norm": 3.1831212206810964,
+      "learning_rate": 1.1002277008697668e-05,
+      "loss": 1.5055,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9681160815265779,
+      "grad_norm": 3.008984222314179,
+      "learning_rate": 1.0999174763479217e-05,
+      "loss": 1.5004,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9683085882041533,
+      "grad_norm": 3.0763207553316767,
+      "learning_rate": 1.0996072421128196e-05,
+      "loss": 1.4857,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9685010948817288,
+      "grad_norm": 3.1765247155977567,
+      "learning_rate": 1.0992969981946192e-05,
+      "loss": 1.4765,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "grad_norm": 3.220280058020591,
+      "learning_rate": 1.0989867446234801e-05,
+      "loss": 1.4952,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.1544,
+      "step": 5032,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.2017,
+      "step": 5032,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 0.8891,
+      "step": 5032,
+      "vm_loss": 0.1029
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.2234,
+      "step": 5032,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.0949,
+      "step": 5032,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.2767,
+      "step": 5032,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.3779,
+      "step": 5032,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 0.9686936015593041,
+      "lm_loss": 1.2076,
+      "step": 5032,
+      "vm_loss": 0.2111
+    },
+    {
+      "epoch": 0.9688861082368795,
+      "grad_norm": 3.263019198728408,
+      "learning_rate": 1.098676481429563e-05,
+      "loss": 1.47,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9690786149144548,
+      "grad_norm": 3.09992119773287,
+      "learning_rate": 1.0983662086430294e-05,
+      "loss": 1.4738,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9692711215920302,
+      "grad_norm": 3.2311374956387824,
+      "learning_rate": 1.098055926294042e-05,
+      "loss": 1.5493,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9694636282696056,
+      "grad_norm": 3.176599893392964,
+      "learning_rate": 1.0977456344127639e-05,
+      "loss": 1.5388,
+      "step": 5036
+    },
+    {
+      "epoch": 0.969656134947181,
+      "grad_norm": 3.2598170299157867,
+      "learning_rate": 1.0974353330293599e-05,
+      "loss": 1.5156,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9698486416247564,
+      "grad_norm": 3.2021674598402416,
+      "learning_rate": 1.0971250221739947e-05,
+      "loss": 1.5872,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9700411483023318,
+      "grad_norm": 3.0781674527047254,
+      "learning_rate": 1.0968147018768347e-05,
+      "loss": 1.505,
+      "step": 5039
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "grad_norm": 3.173571737677203,
+      "learning_rate": 1.0965043721680472e-05,
+      "loss": 1.538,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.1073,
+      "step": 5040,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.4669,
+      "step": 5040,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.3553,
+      "step": 5040,
+      "vm_loss": 0.2674
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.5128,
+      "step": 5040,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.1608,
+      "step": 5040,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.4648,
+      "step": 5040,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.2252,
+      "step": 5040,
+      "vm_loss": 0.1099
+    },
+    {
+      "epoch": 0.9702336549799071,
+      "lm_loss": 1.45,
+      "step": 5040,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 0.9704261616574825,
+      "grad_norm": 3.309162253381353,
+      "learning_rate": 1.0961940330778001e-05,
+      "loss": 1.5304,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9706186683350578,
+      "grad_norm": 3.0690111469541512,
+      "learning_rate": 1.0958836846362622e-05,
+      "loss": 1.4746,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9708111750126333,
+      "grad_norm": 2.9626854571825785,
+      "learning_rate": 1.0955733268736037e-05,
+      "loss": 1.4372,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9710036816902087,
+      "grad_norm": 3.106790003973013,
+      "learning_rate": 1.0952629598199946e-05,
+      "loss": 1.4666,
+      "step": 5044
+    },
+    {
+      "epoch": 0.971196188367784,
+      "grad_norm": 3.108409914143002,
+      "learning_rate": 1.0949525835056073e-05,
+      "loss": 1.4984,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9713886950453594,
+      "grad_norm": 3.0322054387270554,
+      "learning_rate": 1.0946421979606141e-05,
+      "loss": 1.5261,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9715812017229347,
+      "grad_norm": 3.1495917482279365,
+      "learning_rate": 1.0943318032151887e-05,
+      "loss": 1.5153,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "grad_norm": 3.1222659581799515,
+      "learning_rate": 1.0940213992995048e-05,
+      "loss": 1.4472,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.6739,
+      "step": 5048,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.0821,
+      "step": 5048,
+      "vm_loss": 0.1218
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.2419,
+      "step": 5048,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.1225,
+      "step": 5048,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.1091,
+      "step": 5048,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.3134,
+      "step": 5048,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 1.3475,
+      "step": 5048,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 0.9717737084005101,
+      "lm_loss": 2.0123,
+      "step": 5048,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 0.9719662150780856,
+      "grad_norm": 3.255306351038778,
+      "learning_rate": 1.0937109862437384e-05,
+      "loss": 1.4684,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9721587217556609,
+      "grad_norm": 3.3634241186903435,
+      "learning_rate": 1.0934005640780648e-05,
+      "loss": 1.5599,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9723512284332363,
+      "grad_norm": 3.184275517596426,
+      "learning_rate": 1.0930901328326613e-05,
+      "loss": 1.509,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9725437351108116,
+      "grad_norm": 3.1789682559324217,
+      "learning_rate": 1.0927796925377064e-05,
+      "loss": 1.5125,
+      "step": 5052
+    },
+    {
+      "epoch": 0.972736241788387,
+      "grad_norm": 3.080556359503767,
+      "learning_rate": 1.0924692432233781e-05,
+      "loss": 1.4913,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9729287484659624,
+      "grad_norm": 3.089755182054428,
+      "learning_rate": 1.0921587849198565e-05,
+      "loss": 1.4843,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9731212551435378,
+      "grad_norm": 2.993585810769113,
+      "learning_rate": 1.0918483176573217e-05,
+      "loss": 1.4952,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "grad_norm": 3.324191859105468,
+      "learning_rate": 1.0915378414659557e-05,
+      "loss": 1.5028,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.5365,
+      "step": 5056,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.4559,
+      "step": 5056,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.3338,
+      "step": 5056,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.6344,
+      "step": 5056,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.3296,
+      "step": 5056,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.3883,
+      "step": 5056,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.2136,
+      "step": 5056,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 0.9733137618211132,
+      "lm_loss": 1.4391,
+      "step": 5056,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 0.9735062684986886,
+      "grad_norm": 3.3766287783857307,
+      "learning_rate": 1.0912273563759402e-05,
+      "loss": 1.5543,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9736987751762639,
+      "grad_norm": 3.272054564204573,
+      "learning_rate": 1.0909168624174584e-05,
+      "loss": 1.4719,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9738912818538393,
+      "grad_norm": 3.055140788789017,
+      "learning_rate": 1.0906063596206946e-05,
+      "loss": 1.4603,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9740837885314146,
+      "grad_norm": 3.1514589908614883,
+      "learning_rate": 1.0902958480158335e-05,
+      "loss": 1.5118,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9742762952089901,
+      "grad_norm": 3.1827112557344504,
+      "learning_rate": 1.0899853276330605e-05,
+      "loss": 1.5618,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9744688018865655,
+      "grad_norm": 3.0904213088786245,
+      "learning_rate": 1.0896747985025626e-05,
+      "loss": 1.5089,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9746613085641408,
+      "grad_norm": 3.015654347582122,
+      "learning_rate": 1.0893642606545269e-05,
+      "loss": 1.4205,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "grad_norm": 3.153906222085746,
+      "learning_rate": 1.0890537141191417e-05,
+      "loss": 1.4576,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.5905,
+      "step": 5064,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 0.5789,
+      "step": 5064,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.4248,
+      "step": 5064,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 0.9698,
+      "step": 5064,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.1604,
+      "step": 5064,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.3652,
+      "step": 5064,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.6127,
+      "step": 5064,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 0.9748538152417162,
+      "lm_loss": 1.0081,
+      "step": 5064,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 0.9750463219192915,
+      "grad_norm": 3.246451313208057,
+      "learning_rate": 1.0887431589265964e-05,
+      "loss": 1.5028,
+      "step": 5065
+    },
+    {
+      "epoch": 0.975238828596867,
+      "grad_norm": 3.04927508526419,
+      "learning_rate": 1.0884325951070804e-05,
+      "loss": 1.4178,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9754313352744424,
+      "grad_norm": 3.277779996302365,
+      "learning_rate": 1.088122022690785e-05,
+      "loss": 1.5067,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9756238419520177,
+      "grad_norm": 3.0843807309244653,
+      "learning_rate": 1.0878114417079013e-05,
+      "loss": 1.4937,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9758163486295931,
+      "grad_norm": 3.064278174023516,
+      "learning_rate": 1.0875008521886223e-05,
+      "loss": 1.4749,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9760088553071685,
+      "grad_norm": 3.1611696501653164,
+      "learning_rate": 1.0871902541631409e-05,
+      "loss": 1.4488,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9762013619847438,
+      "grad_norm": 3.1971325522563454,
+      "learning_rate": 1.0868796476616516e-05,
+      "loss": 1.5234,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "grad_norm": 3.1726027096461182,
+      "learning_rate": 1.0865690327143488e-05,
+      "loss": 1.506,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.4105,
+      "step": 5072,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.1371,
+      "step": 5072,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.5114,
+      "step": 5072,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.6416,
+      "step": 5072,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.6405,
+      "step": 5072,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.2148,
+      "step": 5072,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.204,
+      "step": 5072,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 0.9763938686623193,
+      "lm_loss": 1.2776,
+      "step": 5072,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 0.9765863753398946,
+      "grad_norm": 3.2158857158618477,
+      "learning_rate": 1.0862584093514285e-05,
+      "loss": 1.4987,
+      "step": 5073
+    },
+    {
+      "epoch": 0.97677888201747,
+      "grad_norm": 3.256693245507279,
+      "learning_rate": 1.0859477776030872e-05,
+      "loss": 1.4808,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9769713886950454,
+      "grad_norm": 3.197236321317148,
+      "learning_rate": 1.0856371374995222e-05,
+      "loss": 1.5149,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9771638953726207,
+      "grad_norm": 3.1370566335817274,
+      "learning_rate": 1.085326489070932e-05,
+      "loss": 1.54,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9773564020501961,
+      "grad_norm": 3.160503226869559,
+      "learning_rate": 1.0850158323475152e-05,
+      "loss": 1.4793,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9775489087277714,
+      "grad_norm": 3.183313832897267,
+      "learning_rate": 1.0847051673594724e-05,
+      "loss": 1.5053,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9777414154053469,
+      "grad_norm": 3.118247783485568,
+      "learning_rate": 1.084394494137003e-05,
+      "loss": 1.5068,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "grad_norm": 3.089225277455859,
+      "learning_rate": 1.0840838127103093e-05,
+      "loss": 1.4629,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 1.4794,
+      "step": 5080,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 1.2263,
+      "step": 5080,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 2.344,
+      "step": 5080,
+      "vm_loss": 0.1984
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 1.3033,
+      "step": 5080,
+      "vm_loss": 0.1038
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 1.8235,
+      "step": 5080,
+      "vm_loss": 0.1119
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 0.925,
+      "step": 5080,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 1.4049,
+      "step": 5080,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 0.9779339220829223,
+      "lm_loss": 2.0164,
+      "step": 5080,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 0.9781264287604976,
+      "grad_norm": 3.1444033522273056,
+      "learning_rate": 1.0837731231095934e-05,
+      "loss": 1.504,
+      "step": 5081
+    },
+    {
+      "epoch": 0.978318935438073,
+      "grad_norm": 3.2710760827059326,
+      "learning_rate": 1.0834624253650579e-05,
+      "loss": 1.5012,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9785114421156483,
+      "grad_norm": 3.18729257021845,
+      "learning_rate": 1.083151719506907e-05,
+      "loss": 1.5112,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9787039487932238,
+      "grad_norm": 3.148707069033162,
+      "learning_rate": 1.0828410055653451e-05,
+      "loss": 1.5381,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9788964554707992,
+      "grad_norm": 3.160382129589161,
+      "learning_rate": 1.0825302835705776e-05,
+      "loss": 1.4748,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9790889621483745,
+      "grad_norm": 3.2375711105249776,
+      "learning_rate": 1.0822195535528106e-05,
+      "loss": 1.4919,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9792814688259499,
+      "grad_norm": 3.020912128934804,
+      "learning_rate": 1.081908815542251e-05,
+      "loss": 1.4839,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "grad_norm": 3.0581526085618225,
+      "learning_rate": 1.0815980695691068e-05,
+      "loss": 1.4536,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 0.8236,
+      "step": 5088,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 1.5163,
+      "step": 5088,
+      "vm_loss": 0.1286
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 0.983,
+      "step": 5088,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 0.9062,
+      "step": 5088,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 1.8458,
+      "step": 5088,
+      "vm_loss": 0.1258
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 1.5874,
+      "step": 5088,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 1.6175,
+      "step": 5088,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 0.9794739755035253,
+      "lm_loss": 0.9154,
+      "step": 5088,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 0.9796664821811006,
+      "grad_norm": 3.1865684570246646,
+      "learning_rate": 1.0812873156635864e-05,
+      "loss": 1.4647,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9798589888586761,
+      "grad_norm": 3.257020889524908,
+      "learning_rate": 1.080976553855899e-05,
+      "loss": 1.5563,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9800514955362514,
+      "grad_norm": 3.2491158883260223,
+      "learning_rate": 1.0806657841762542e-05,
+      "loss": 1.5032,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9802440022138268,
+      "grad_norm": 3.2575254063268955,
+      "learning_rate": 1.0803550066548634e-05,
+      "loss": 1.4804,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9804365088914022,
+      "grad_norm": 3.116673690439317,
+      "learning_rate": 1.080044221321938e-05,
+      "loss": 1.5548,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9806290155689775,
+      "grad_norm": 3.0864541256514744,
+      "learning_rate": 1.0797334282076898e-05,
+      "loss": 1.4788,
+      "step": 5094
+    },
+    {
+      "epoch": 0.980821522246553,
+      "grad_norm": 3.0842548618909973,
+      "learning_rate": 1.079422627342333e-05,
+      "loss": 1.4056,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "grad_norm": 3.083940344140672,
+      "learning_rate": 1.0791118187560801e-05,
+      "loss": 1.4139,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 0.8556,
+      "step": 5096,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.2532,
+      "step": 5096,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.1542,
+      "step": 5096,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.4572,
+      "step": 5096,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 2.0382,
+      "step": 5096,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.1248,
+      "step": 5096,
+      "vm_loss": 0.2109
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.5335,
+      "step": 5096,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 0.9810140289241283,
+      "lm_loss": 1.4494,
+      "step": 5096,
+      "vm_loss": 0.1073
+    },
+    {
+      "epoch": 0.9812065356017037,
+      "grad_norm": 3.1219095160790324,
+      "learning_rate": 1.0788010024791465e-05,
+      "loss": 1.452,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9813990422792791,
+      "grad_norm": 3.491937895268139,
+      "learning_rate": 1.0784901785417474e-05,
+      "loss": 1.481,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9815915489568544,
+      "grad_norm": 3.3704852163736874,
+      "learning_rate": 1.0781793469740984e-05,
+      "loss": 1.498,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9817840556344298,
+      "grad_norm": 3.229131580133439,
+      "learning_rate": 1.0778685078064172e-05,
+      "loss": 1.4592,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9819765623120053,
+      "grad_norm": 3.0934918125098454,
+      "learning_rate": 1.0775576610689205e-05,
+      "loss": 1.4983,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9821690689895806,
+      "grad_norm": 3.175488113458974,
+      "learning_rate": 1.077246806791827e-05,
+      "loss": 1.4828,
+      "step": 5102
+    },
+    {
+      "epoch": 0.982361575667156,
+      "grad_norm": 3.0483937907567658,
+      "learning_rate": 1.0769359450053553e-05,
+      "loss": 1.4747,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "grad_norm": 3.094902182125298,
+      "learning_rate": 1.0766250757397256e-05,
+      "loss": 1.5029,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.6889,
+      "step": 5104,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.0876,
+      "step": 5104,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.4541,
+      "step": 5104,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.2922,
+      "step": 5104,
+      "vm_loss": 0.2192
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.6232,
+      "step": 5104,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.1543,
+      "step": 5104,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.5891,
+      "step": 5104,
+      "vm_loss": 0.2756
+    },
+    {
+      "epoch": 0.9825540823447313,
+      "lm_loss": 1.3451,
+      "step": 5104,
+      "vm_loss": 0.1246
+    },
+    {
+      "epoch": 0.9827465890223067,
+      "grad_norm": 3.1152313142280277,
+      "learning_rate": 1.0763141990251586e-05,
+      "loss": 1.4765,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9829390956998821,
+      "grad_norm": 3.1994604687130623,
+      "learning_rate": 1.0760033148918747e-05,
+      "loss": 1.4895,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9831316023774574,
+      "grad_norm": 3.280600630824215,
+      "learning_rate": 1.0756924233700963e-05,
+      "loss": 1.4686,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9833241090550329,
+      "grad_norm": 3.23383752018144,
+      "learning_rate": 1.0753815244900459e-05,
+      "loss": 1.4649,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9835166157326082,
+      "grad_norm": 3.2192692766516626,
+      "learning_rate": 1.0750706182819468e-05,
+      "loss": 1.5051,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9837091224101836,
+      "grad_norm": 3.2542010151470926,
+      "learning_rate": 1.0747597047760236e-05,
+      "loss": 1.5459,
+      "step": 5110
+    },
+    {
+      "epoch": 0.983901629087759,
+      "grad_norm": 3.14894053192367,
+      "learning_rate": 1.0744487840025001e-05,
+      "loss": 1.4758,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "grad_norm": 3.043445294354467,
+      "learning_rate": 1.0741378559916029e-05,
+      "loss": 1.5055,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.6875,
+      "step": 5112,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.4693,
+      "step": 5112,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 0.8308,
+      "step": 5112,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.071,
+      "step": 5112,
+      "vm_loss": 0.1125
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.404,
+      "step": 5112,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.4635,
+      "step": 5112,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.0517,
+      "step": 5112,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 0.9840941357653343,
+      "lm_loss": 1.3387,
+      "step": 5112,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 0.9842866424429098,
+      "grad_norm": 3.273247435364176,
+      "learning_rate": 1.0738269207735573e-05,
+      "loss": 1.4612,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9844791491204851,
+      "grad_norm": 3.0481708944911214,
+      "learning_rate": 1.0735159783785905e-05,
+      "loss": 1.4686,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9846716557980605,
+      "grad_norm": 3.1539878163603987,
+      "learning_rate": 1.07320502883693e-05,
+      "loss": 1.4064,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9848641624756359,
+      "grad_norm": 3.3934956241595144,
+      "learning_rate": 1.0728940721788041e-05,
+      "loss": 1.4699,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9850566691532112,
+      "grad_norm": 3.282275792838578,
+      "learning_rate": 1.072583108434442e-05,
+      "loss": 1.47,
+      "step": 5117
+    },
+    {
+      "epoch": 0.9852491758307866,
+      "grad_norm": 3.2967793734759714,
+      "learning_rate": 1.072272137634073e-05,
+      "loss": 1.4415,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9854416825083621,
+      "grad_norm": 3.275878257788189,
+      "learning_rate": 1.0719611598079278e-05,
+      "loss": 1.4493,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "grad_norm": 3.26253730246854,
+      "learning_rate": 1.0716501749862369e-05,
+      "loss": 1.4776,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.2419,
+      "step": 5120,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 0.8267,
+      "step": 5120,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 0.7665,
+      "step": 5120,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.5169,
+      "step": 5120,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.1724,
+      "step": 5120,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.3158,
+      "step": 5120,
+      "vm_loss": 0.2295
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.6846,
+      "step": 5120,
+      "vm_loss": 0.2286
+    },
+    {
+      "epoch": 0.9856341891859374,
+      "lm_loss": 1.248,
+      "step": 5120,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 0.9858266958635128,
+      "grad_norm": 3.2041187833608653,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 1.4751,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9860192025410881,
+      "grad_norm": 3.1781097369895694,
+      "learning_rate": 1.0710281844771469e-05,
+      "loss": 1.5337,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9862117092186635,
+      "grad_norm": 3.043691286169084,
+      "learning_rate": 1.0707171788502129e-05,
+      "loss": 1.4937,
+      "step": 5123
+    },
+    {
+      "epoch": 0.986404215896239,
+      "grad_norm": 3.044828485902192,
+      "learning_rate": 1.0704061663486645e-05,
+      "loss": 1.4685,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9865967225738143,
+      "grad_norm": 3.264003149302052,
+      "learning_rate": 1.0700951470027357e-05,
+      "loss": 1.5052,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9867892292513897,
+      "grad_norm": 3.2005612480115397,
+      "learning_rate": 1.0697841208426617e-05,
+      "loss": 1.4816,
+      "step": 5126
+    },
+    {
+      "epoch": 0.986981735928965,
+      "grad_norm": 3.3183210382633277,
+      "learning_rate": 1.069473087898679e-05,
+      "loss": 1.5739,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "grad_norm": 3.3945775008555152,
+      "learning_rate": 1.0691620482010227e-05,
+      "loss": 1.5584,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 0.9012,
+      "step": 5128,
+      "vm_loss": 0.2294
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.6728,
+      "step": 5128,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.6825,
+      "step": 5128,
+      "vm_loss": 0.2086
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.4753,
+      "step": 5128,
+      "vm_loss": 0.2063
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.7834,
+      "step": 5128,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 0.9395,
+      "step": 5128,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.5501,
+      "step": 5128,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 0.9871742426065404,
+      "lm_loss": 1.4829,
+      "step": 5128,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 0.9873667492841158,
+      "grad_norm": 3.100445528016157,
+      "learning_rate": 1.068851001779931e-05,
+      "loss": 1.5566,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9875592559616911,
+      "grad_norm": 3.0946163510110374,
+      "learning_rate": 1.0685399486656407e-05,
+      "loss": 1.527,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9877517626392666,
+      "grad_norm": 3.067301694698744,
+      "learning_rate": 1.0682288888883905e-05,
+      "loss": 1.5129,
+      "step": 5131
+    },
+    {
+      "epoch": 0.987944269316842,
+      "grad_norm": 3.1427012755918624,
+      "learning_rate": 1.0679178224784197e-05,
+      "loss": 1.4775,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9881367759944173,
+      "grad_norm": 3.108769437991071,
+      "learning_rate": 1.0676067494659673e-05,
+      "loss": 1.4521,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9883292826719927,
+      "grad_norm": 3.1054006768686877,
+      "learning_rate": 1.0672956698812741e-05,
+      "loss": 1.478,
+      "step": 5134
+    },
+    {
+      "epoch": 0.988521789349568,
+      "grad_norm": 3.0890832467756337,
+      "learning_rate": 1.066984583754581e-05,
+      "loss": 1.4382,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "grad_norm": 3.232718631336568,
+      "learning_rate": 1.0666734911161292e-05,
+      "loss": 1.4512,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.6239,
+      "step": 5136,
+      "vm_loss": 0.1259
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.3203,
+      "step": 5136,
+      "vm_loss": 0.2441
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.3079,
+      "step": 5136,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.604,
+      "step": 5136,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.3442,
+      "step": 5136,
+      "vm_loss": 0.1976
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.2042,
+      "step": 5136,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.2235,
+      "step": 5136,
+      "vm_loss": 0.1346
+    },
+    {
+      "epoch": 0.9887142960271434,
+      "lm_loss": 1.1693,
+      "step": 5136,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 0.9889068027047189,
+      "grad_norm": 3.2850293869554874,
+      "learning_rate": 1.0663623919961608e-05,
+      "loss": 1.4995,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9890993093822942,
+      "grad_norm": 3.1394844876116466,
+      "learning_rate": 1.066051286424919e-05,
+      "loss": 1.415,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9892918160598696,
+      "grad_norm": 3.15755217535084,
+      "learning_rate": 1.0657401744326474e-05,
+      "loss": 1.4169,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9894843227374449,
+      "grad_norm": 3.323771989734763,
+      "learning_rate": 1.0654290560495897e-05,
+      "loss": 1.4747,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9896768294150203,
+      "grad_norm": 3.252469853828255,
+      "learning_rate": 1.0651179313059904e-05,
+      "loss": 1.4618,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9898693360925958,
+      "grad_norm": 3.309364710083231,
+      "learning_rate": 1.0648068002320953e-05,
+      "loss": 1.5262,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9900618427701711,
+      "grad_norm": 3.183395597115284,
+      "learning_rate": 1.06449566285815e-05,
+      "loss": 1.5184,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "grad_norm": 3.152919906818271,
+      "learning_rate": 1.0641845192144013e-05,
+      "loss": 1.4453,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.7308,
+      "step": 5144,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.5433,
+      "step": 5144,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.4702,
+      "step": 5144,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.0467,
+      "step": 5144,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.1147,
+      "step": 5144,
+      "vm_loss": 0.1218
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.1266,
+      "step": 5144,
+      "vm_loss": 0.2445
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 1.4071,
+      "step": 5144,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 0.9902543494477465,
+      "lm_loss": 0.7921,
+      "step": 5144,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 0.9904468561253218,
+      "grad_norm": 3.1507326274152048,
+      "learning_rate": 1.063873369331096e-05,
+      "loss": 1.4981,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9906393628028972,
+      "grad_norm": 3.1033121527883765,
+      "learning_rate": 1.0635622132384821e-05,
+      "loss": 1.508,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9908318694804726,
+      "grad_norm": 3.0980205790453987,
+      "learning_rate": 1.0632510509668079e-05,
+      "loss": 1.4269,
+      "step": 5147
+    },
+    {
+      "epoch": 0.991024376158048,
+      "grad_norm": 3.262012530974657,
+      "learning_rate": 1.0629398825463223e-05,
+      "loss": 1.5174,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9912168828356234,
+      "grad_norm": 3.139943289625485,
+      "learning_rate": 1.0626287080072747e-05,
+      "loss": 1.4301,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9914093895131988,
+      "grad_norm": 3.255552846527656,
+      "learning_rate": 1.0623175273799158e-05,
+      "loss": 1.4833,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9916018961907741,
+      "grad_norm": 3.3293575832645,
+      "learning_rate": 1.0620063406944956e-05,
+      "loss": 1.5011,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "grad_norm": 3.17686847821999,
+      "learning_rate": 1.0616951479812658e-05,
+      "loss": 1.4806,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.3465,
+      "step": 5152,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.4512,
+      "step": 5152,
+      "vm_loss": 0.2219
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 0.9472,
+      "step": 5152,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.7034,
+      "step": 5152,
+      "vm_loss": 0.2329
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 0.9384,
+      "step": 5152,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.446,
+      "step": 5152,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.068,
+      "step": 5152,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 0.9917944028683495,
+      "lm_loss": 1.4045,
+      "step": 5152,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 0.9919869095459248,
+      "grad_norm": 3.1996735775395897,
+      "learning_rate": 1.0613839492704784e-05,
+      "loss": 1.4944,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9921794162235003,
+      "grad_norm": 3.155055301257671,
+      "learning_rate": 1.0610727445923858e-05,
+      "loss": 1.3565,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9923719229010757,
+      "grad_norm": 3.2128393370425674,
+      "learning_rate": 1.0607615339772412e-05,
+      "loss": 1.4292,
+      "step": 5155
+    },
+    {
+      "epoch": 0.992564429578651,
+      "grad_norm": 3.1610813521395653,
+      "learning_rate": 1.0604503174552982e-05,
+      "loss": 1.4889,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9927569362562264,
+      "grad_norm": 3.2208654860446315,
+      "learning_rate": 1.0601390950568106e-05,
+      "loss": 1.5059,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9929494429338017,
+      "grad_norm": 3.0908580941570305,
+      "learning_rate": 1.059827866812034e-05,
+      "loss": 1.503,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9931419496113771,
+      "grad_norm": 3.1712177645381994,
+      "learning_rate": 1.059516632751223e-05,
+      "loss": 1.4529,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "grad_norm": 3.2481163389266205,
+      "learning_rate": 1.059205392904634e-05,
+      "loss": 1.5239,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.1123,
+      "step": 5160,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.4215,
+      "step": 5160,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.8061,
+      "step": 5160,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 0.9452,
+      "step": 5160,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.2034,
+      "step": 5160,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.1324,
+      "step": 5160,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.3233,
+      "step": 5160,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 0.9933344562889526,
+      "lm_loss": 1.3686,
+      "step": 5160,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 0.9935269629665279,
+      "grad_norm": 3.0479866632881305,
+      "learning_rate": 1.0588941473025234e-05,
+      "loss": 1.4826,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9937194696441033,
+      "grad_norm": 3.311144798373948,
+      "learning_rate": 1.0585828959751478e-05,
+      "loss": 1.5256,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9939119763216786,
+      "grad_norm": 3.316950033865326,
+      "learning_rate": 1.0582716389527659e-05,
+      "loss": 1.4482,
+      "step": 5163
+    },
+    {
+      "epoch": 0.994104482999254,
+      "grad_norm": 3.1207980520784546,
+      "learning_rate": 1.0579603762656347e-05,
+      "loss": 1.4965,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9942969896768294,
+      "grad_norm": 3.1699648952579875,
+      "learning_rate": 1.0576491079440137e-05,
+      "loss": 1.4857,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9944894963544048,
+      "grad_norm": 3.231213567919864,
+      "learning_rate": 1.0573378340181614e-05,
+      "loss": 1.4901,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9946820030319802,
+      "grad_norm": 3.071529170750346,
+      "learning_rate": 1.0570265545183382e-05,
+      "loss": 1.396,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "grad_norm": 3.130870206394583,
+      "learning_rate": 1.0567152694748046e-05,
+      "loss": 1.4962,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.4961,
+      "step": 5168,
+      "vm_loss": 0.1937
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 0.9975,
+      "step": 5168,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.04,
+      "step": 5168,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.3609,
+      "step": 5168,
+      "vm_loss": 0.098
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.1595,
+      "step": 5168,
+      "vm_loss": 0.2534
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.3601,
+      "step": 5168,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.3202,
+      "step": 5168,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 0.9948745097095556,
+      "lm_loss": 1.0364,
+      "step": 5168,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.9950670163871309,
+      "grad_norm": 2.933246845790367,
+      "learning_rate": 1.0564039789178211e-05,
+      "loss": 1.4263,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9952595230647063,
+      "grad_norm": 3.1811132991656295,
+      "learning_rate": 1.056092682877649e-05,
+      "loss": 1.5109,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9954520297422816,
+      "grad_norm": 3.123891655430106,
+      "learning_rate": 1.0557813813845508e-05,
+      "loss": 1.3532,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9956445364198571,
+      "grad_norm": 3.2588478318132807,
+      "learning_rate": 1.0554700744687885e-05,
+      "loss": 1.4477,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9958370430974325,
+      "grad_norm": 3.249772397889448,
+      "learning_rate": 1.0551587621606253e-05,
+      "loss": 1.4698,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9960295497750078,
+      "grad_norm": 3.23628835620586,
+      "learning_rate": 1.0548474444903247e-05,
+      "loss": 1.5267,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9962220564525832,
+      "grad_norm": 3.2840124788427523,
+      "learning_rate": 1.054536121488151e-05,
+      "loss": 1.4961,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "grad_norm": 3.1689419482787464,
+      "learning_rate": 1.0542247931843684e-05,
+      "loss": 1.4285,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 0.7649,
+      "step": 5176,
+      "vm_loss": 0.2461
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 1.8616,
+      "step": 5176,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 1.1569,
+      "step": 5176,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 0.8808,
+      "step": 5176,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 1.2465,
+      "step": 5176,
+      "vm_loss": 0.1164
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 1.4974,
+      "step": 5176,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 2.4756,
+      "step": 5176,
+      "vm_loss": 0.0899
+    },
+    {
+      "epoch": 0.9964145631301585,
+      "lm_loss": 1.595,
+      "step": 5176,
+      "vm_loss": 0.0963
+    },
+    {
+      "epoch": 0.996607069807734,
+      "grad_norm": 3.0695706645997234,
+      "learning_rate": 1.0539134596092423e-05,
+      "loss": 1.4913,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9967995764853094,
+      "grad_norm": 3.039550839618325,
+      "learning_rate": 1.0536021207930381e-05,
+      "loss": 1.4326,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9969920831628847,
+      "grad_norm": 3.110007854319775,
+      "learning_rate": 1.0532907767660218e-05,
+      "loss": 1.4257,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9971845898404601,
+      "grad_norm": 3.1108761420049773,
+      "learning_rate": 1.0529794275584606e-05,
+      "loss": 1.4324,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9973770965180355,
+      "grad_norm": 3.3111678742001107,
+      "learning_rate": 1.0526680732006209e-05,
+      "loss": 1.5098,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9975696031956108,
+      "grad_norm": 3.30252037218265,
+      "learning_rate": 1.0523567137227708e-05,
+      "loss": 1.4985,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9977621098731863,
+      "grad_norm": 3.326749085967109,
+      "learning_rate": 1.0520453491551783e-05,
+      "loss": 1.4911,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "grad_norm": 3.1907896097177093,
+      "learning_rate": 1.0517339795281118e-05,
+      "loss": 1.3981,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.1123,
+      "step": 5184,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.6794,
+      "step": 5184,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.7063,
+      "step": 5184,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.3821,
+      "step": 5184,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.7377,
+      "step": 5184,
+      "vm_loss": 0.1296
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.4997,
+      "step": 5184,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.0181,
+      "step": 5184,
+      "vm_loss": 0.1181
+    },
+    {
+      "epoch": 0.9979546165507616,
+      "lm_loss": 1.3014,
+      "step": 5184,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 0.998147123228337,
+      "grad_norm": 3.0400737235446122,
+      "learning_rate": 1.051422604871841e-05,
+      "loss": 1.4413,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9983396299059124,
+      "grad_norm": 3.038496730884325,
+      "learning_rate": 1.051111225216635e-05,
+      "loss": 1.4707,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9985321365834877,
+      "grad_norm": 3.1326005174540743,
+      "learning_rate": 1.0507998405927638e-05,
+      "loss": 1.4119,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9987246432610631,
+      "grad_norm": 3.2447187292621544,
+      "learning_rate": 1.0504884510304982e-05,
+      "loss": 1.4558,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9989171499386384,
+      "grad_norm": 3.1455350352179012,
+      "learning_rate": 1.0501770565601093e-05,
+      "loss": 1.4679,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9991096566162139,
+      "grad_norm": 3.2838123218271376,
+      "learning_rate": 1.0498656572118687e-05,
+      "loss": 1.488,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9993021632937893,
+      "grad_norm": 3.2137620754875096,
+      "learning_rate": 1.0495542530160481e-05,
+      "loss": 1.4699,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "grad_norm": 3.0667918137777845,
+      "learning_rate": 1.04924284400292e-05,
+      "loss": 1.4652,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.365,
+      "step": 5192,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.5673,
+      "step": 5192,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.4192,
+      "step": 5192,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.2579,
+      "step": 5192,
+      "vm_loss": 0.2228
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 0.6447,
+      "step": 5192,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.5067,
+      "step": 5192,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.3401,
+      "step": 5192,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 0.9994946699713646,
+      "lm_loss": 1.6762,
+      "step": 5192,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 0.99968717664894,
+      "grad_norm": 3.1994135536860604,
+      "learning_rate": 1.0489314302027576e-05,
+      "loss": 1.4936,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9998796833265153,
+      "grad_norm": 3.22642452404787,
+      "learning_rate": 1.0486200116458343e-05,
+      "loss": 1.5552,
+      "step": 5194
+    },
+    {
+      "epoch": 1.0000721900040908,
+      "grad_norm": 3.162386536178728,
+      "learning_rate": 1.0483085883624236e-05,
+      "loss": 1.4202,
+      "step": 5195
+    },
+    {
+      "epoch": 1.0002646966816662,
+      "grad_norm": 3.179228071581397,
+      "learning_rate": 1.0479971603828001e-05,
+      "loss": 1.4598,
+      "step": 5196
+    },
+    {
+      "epoch": 1.0004572033592416,
+      "grad_norm": 3.184312446038585,
+      "learning_rate": 1.0476857277372387e-05,
+      "loss": 1.4779,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0006497100368168,
+      "grad_norm": 3.157349199681182,
+      "learning_rate": 1.0473742904560143e-05,
+      "loss": 1.472,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0008422167143922,
+      "grad_norm": 3.0979039251442155,
+      "learning_rate": 1.047062848569403e-05,
+      "loss": 1.4494,
+      "step": 5199
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "grad_norm": 3.112808485950145,
+      "learning_rate": 1.0467514021076805e-05,
+      "loss": 1.3441,
+      "step": 5200
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.0705,
+      "step": 5200,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.3485,
+      "step": 5200,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.2488,
+      "step": 5200,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.5624,
+      "step": 5200,
+      "vm_loss": 0.2688
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.4236,
+      "step": 5200,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.4954,
+      "step": 5200,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.5853,
+      "step": 5200,
+      "vm_loss": 0.2351
+    },
+    {
+      "epoch": 1.0010347233919676,
+      "lm_loss": 1.7917,
+      "step": 5200,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 1.001227230069543,
+      "grad_norm": 3.1661471596884248,
+      "learning_rate": 1.0464399511011236e-05,
+      "loss": 1.4379,
+      "step": 5201
+    },
+    {
+      "epoch": 1.0014197367471185,
+      "grad_norm": 3.496955929590711,
+      "learning_rate": 1.0461284955800094e-05,
+      "loss": 1.4888,
+      "step": 5202
+    },
+    {
+      "epoch": 1.001612243424694,
+      "grad_norm": 3.265446181662951,
+      "learning_rate": 1.0458170355746154e-05,
+      "loss": 1.4981,
+      "step": 5203
+    },
+    {
+      "epoch": 1.001804750102269,
+      "grad_norm": 3.0717926778926294,
+      "learning_rate": 1.0455055711152194e-05,
+      "loss": 1.3417,
+      "step": 5204
+    },
+    {
+      "epoch": 1.0019972567798445,
+      "grad_norm": 3.083236944411091,
+      "learning_rate": 1.0451941022320993e-05,
+      "loss": 1.3909,
+      "step": 5205
+    },
+    {
+      "epoch": 1.00218976345742,
+      "grad_norm": 3.133559481934841,
+      "learning_rate": 1.0448826289555345e-05,
+      "loss": 1.4288,
+      "step": 5206
+    },
+    {
+      "epoch": 1.0023822701349954,
+      "grad_norm": 3.1973796323832393,
+      "learning_rate": 1.0445711513158039e-05,
+      "loss": 1.3973,
+      "step": 5207
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "grad_norm": 3.2553336060458156,
+      "learning_rate": 1.0442596693431872e-05,
+      "loss": 1.4556,
+      "step": 5208
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 1.0739,
+      "step": 5208,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 1.4179,
+      "step": 5208,
+      "vm_loss": 0.2097
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 1.2275,
+      "step": 5208,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 0.7862,
+      "step": 5208,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 0.9683,
+      "step": 5208,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 1.2458,
+      "step": 5208,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 1.0941,
+      "step": 5208,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.0025747768125708,
+      "lm_loss": 0.6557,
+      "step": 5208,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 1.002767283490146,
+      "grad_norm": 3.1589191316239043,
+      "learning_rate": 1.0439481830679645e-05,
+      "loss": 1.3939,
+      "step": 5209
+    },
+    {
+      "epoch": 1.0029597901677214,
+      "grad_norm": 3.2808017576085167,
+      "learning_rate": 1.043636692520416e-05,
+      "loss": 1.5028,
+      "step": 5210
+    },
+    {
+      "epoch": 1.0031522968452968,
+      "grad_norm": 3.566053230283462,
+      "learning_rate": 1.0433251977308227e-05,
+      "loss": 1.4628,
+      "step": 5211
+    },
+    {
+      "epoch": 1.0033448035228723,
+      "grad_norm": 3.221899998514581,
+      "learning_rate": 1.0430136987294661e-05,
+      "loss": 1.4336,
+      "step": 5212
+    },
+    {
+      "epoch": 1.0035373102004477,
+      "grad_norm": 3.1795297123867585,
+      "learning_rate": 1.0427021955466274e-05,
+      "loss": 1.3693,
+      "step": 5213
+    },
+    {
+      "epoch": 1.0037298168780229,
+      "grad_norm": 3.1485205426822103,
+      "learning_rate": 1.042390688212589e-05,
+      "loss": 1.3793,
+      "step": 5214
+    },
+    {
+      "epoch": 1.0039223235555983,
+      "grad_norm": 3.0701711755441674,
+      "learning_rate": 1.0420791767576334e-05,
+      "loss": 1.4546,
+      "step": 5215
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "grad_norm": 3.0752007514669315,
+      "learning_rate": 1.0417676612120438e-05,
+      "loss": 1.4924,
+      "step": 5216
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.0657,
+      "step": 5216,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.1886,
+      "step": 5216,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 0.9246,
+      "step": 5216,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.2511,
+      "step": 5216,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.5134,
+      "step": 5216,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 0.6979,
+      "step": 5216,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.0754,
+      "step": 5216,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.0041148302331737,
+      "lm_loss": 1.2109,
+      "step": 5216,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 1.0043073369107491,
+      "grad_norm": 3.0531474288076037,
+      "learning_rate": 1.0414561416061027e-05,
+      "loss": 1.3911,
+      "step": 5217
+    },
+    {
+      "epoch": 1.0044998435883246,
+      "grad_norm": 3.2132300977004653,
+      "learning_rate": 1.0411446179700944e-05,
+      "loss": 1.5404,
+      "step": 5218
+    },
+    {
+      "epoch": 1.0046923502658998,
+      "grad_norm": 3.0146976069540776,
+      "learning_rate": 1.0408330903343029e-05,
+      "loss": 1.4114,
+      "step": 5219
+    },
+    {
+      "epoch": 1.0048848569434752,
+      "grad_norm": 3.171097654113657,
+      "learning_rate": 1.0405215587290126e-05,
+      "loss": 1.4515,
+      "step": 5220
+    },
+    {
+      "epoch": 1.0050773636210506,
+      "grad_norm": 3.2455155961267215,
+      "learning_rate": 1.0402100231845084e-05,
+      "loss": 1.4366,
+      "step": 5221
+    },
+    {
+      "epoch": 1.005269870298626,
+      "grad_norm": 3.22901564982213,
+      "learning_rate": 1.0398984837310755e-05,
+      "loss": 1.4834,
+      "step": 5222
+    },
+    {
+      "epoch": 1.0054623769762014,
+      "grad_norm": 3.3496168743749046,
+      "learning_rate": 1.0395869403989998e-05,
+      "loss": 1.468,
+      "step": 5223
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "grad_norm": 3.2536169087370728,
+      "learning_rate": 1.0392753932185667e-05,
+      "loss": 1.4217,
+      "step": 5224
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.8638,
+      "step": 5224,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 0.9168,
+      "step": 5224,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.3041,
+      "step": 5224,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.2719,
+      "step": 5224,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.4,
+      "step": 5224,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.0455,
+      "step": 5224,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.0459,
+      "step": 5224,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.0056548836537766,
+      "lm_loss": 1.2135,
+      "step": 5224,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 1.005847390331352,
+      "grad_norm": 3.2048423208939028,
+      "learning_rate": 1.0389638422200627e-05,
+      "loss": 1.3887,
+      "step": 5225
+    },
+    {
+      "epoch": 1.0060398970089275,
+      "grad_norm": 3.0113042110121984,
+      "learning_rate": 1.0386522874337752e-05,
+      "loss": 1.3036,
+      "step": 5226
+    },
+    {
+      "epoch": 1.006232403686503,
+      "grad_norm": 3.072168645649257,
+      "learning_rate": 1.0383407288899904e-05,
+      "loss": 1.4374,
+      "step": 5227
+    },
+    {
+      "epoch": 1.0064249103640783,
+      "grad_norm": 3.118892777007722,
+      "learning_rate": 1.0380291666189968e-05,
+      "loss": 1.4333,
+      "step": 5228
+    },
+    {
+      "epoch": 1.0066174170416535,
+      "grad_norm": 3.264699659391858,
+      "learning_rate": 1.0377176006510812e-05,
+      "loss": 1.4752,
+      "step": 5229
+    },
+    {
+      "epoch": 1.006809923719229,
+      "grad_norm": 3.2665980015175116,
+      "learning_rate": 1.0374060310165324e-05,
+      "loss": 1.4826,
+      "step": 5230
+    },
+    {
+      "epoch": 1.0070024303968044,
+      "grad_norm": 3.155272863651309,
+      "learning_rate": 1.0370944577456388e-05,
+      "loss": 1.4027,
+      "step": 5231
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "grad_norm": 3.150861512859124,
+      "learning_rate": 1.0367828808686892e-05,
+      "loss": 1.4328,
+      "step": 5232
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.3273,
+      "step": 5232,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.3858,
+      "step": 5232,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.6511,
+      "step": 5232,
+      "vm_loss": 0.2237
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.338,
+      "step": 5232,
+      "vm_loss": 0.1812
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.4826,
+      "step": 5232,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.6167,
+      "step": 5232,
+      "vm_loss": 0.1457
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 0.8488,
+      "step": 5232,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.0071949370743798,
+      "lm_loss": 1.508,
+      "step": 5232,
+      "vm_loss": 0.1258
+    },
+    {
+      "epoch": 1.0073874437519552,
+      "grad_norm": 3.359076758798513,
+      "learning_rate": 1.0364713004159733e-05,
+      "loss": 1.4635,
+      "step": 5233
+    },
+    {
+      "epoch": 1.0075799504295304,
+      "grad_norm": 3.1923613989881563,
+      "learning_rate": 1.0361597164177802e-05,
+      "loss": 1.4511,
+      "step": 5234
+    },
+    {
+      "epoch": 1.0077724571071058,
+      "grad_norm": 3.1955137645426013,
+      "learning_rate": 1.0358481289044002e-05,
+      "loss": 1.438,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0079649637846813,
+      "grad_norm": 3.113364889754319,
+      "learning_rate": 1.035536537906123e-05,
+      "loss": 1.4117,
+      "step": 5236
+    },
+    {
+      "epoch": 1.0081574704622567,
+      "grad_norm": 3.0817782710398642,
+      "learning_rate": 1.03522494345324e-05,
+      "loss": 1.4212,
+      "step": 5237
+    },
+    {
+      "epoch": 1.008349977139832,
+      "grad_norm": 3.04033494996302,
+      "learning_rate": 1.034913345576042e-05,
+      "loss": 1.4207,
+      "step": 5238
+    },
+    {
+      "epoch": 1.0085424838174075,
+      "grad_norm": 3.317204646936617,
+      "learning_rate": 1.03460174430482e-05,
+      "loss": 1.461,
+      "step": 5239
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "grad_norm": 3.091562523290644,
+      "learning_rate": 1.0342901396698658e-05,
+      "loss": 1.4214,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.1824,
+      "step": 5240,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.258,
+      "step": 5240,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.4302,
+      "step": 5240,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.3365,
+      "step": 5240,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 0.6611,
+      "step": 5240,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.0134,
+      "step": 5240,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 1.2141,
+      "step": 5240,
+      "vm_loss": 0.2307
+    },
+    {
+      "epoch": 1.0087349904949827,
+      "lm_loss": 0.7424,
+      "step": 5240,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 1.0089274971725581,
+      "grad_norm": 3.187239188299685,
+      "learning_rate": 1.0339785317014713e-05,
+      "loss": 1.4711,
+      "step": 5241
+    },
+    {
+      "epoch": 1.0091200038501336,
+      "grad_norm": 3.156670805972283,
+      "learning_rate": 1.0336669204299286e-05,
+      "loss": 1.3994,
+      "step": 5242
+    },
+    {
+      "epoch": 1.009312510527709,
+      "grad_norm": 3.482825333960884,
+      "learning_rate": 1.0333553058855309e-05,
+      "loss": 1.5643,
+      "step": 5243
+    },
+    {
+      "epoch": 1.0095050172052844,
+      "grad_norm": 3.193519212219886,
+      "learning_rate": 1.0330436880985704e-05,
+      "loss": 1.465,
+      "step": 5244
+    },
+    {
+      "epoch": 1.0096975238828596,
+      "grad_norm": 3.119312141286633,
+      "learning_rate": 1.0327320670993409e-05,
+      "loss": 1.4434,
+      "step": 5245
+    },
+    {
+      "epoch": 1.009890030560435,
+      "grad_norm": 3.027848459519069,
+      "learning_rate": 1.0324204429181355e-05,
+      "loss": 1.4603,
+      "step": 5246
+    },
+    {
+      "epoch": 1.0100825372380104,
+      "grad_norm": 3.12083207215003,
+      "learning_rate": 1.0321088155852484e-05,
+      "loss": 1.4209,
+      "step": 5247
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "grad_norm": 3.1428892058115285,
+      "learning_rate": 1.0317971851309736e-05,
+      "loss": 1.4654,
+      "step": 5248
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 1.0534,
+      "step": 5248,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 1.0712,
+      "step": 5248,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 0.8561,
+      "step": 5248,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 0.9943,
+      "step": 5248,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 1.2064,
+      "step": 5248,
+      "vm_loss": 0.2008
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 0.8936,
+      "step": 5248,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 1.7405,
+      "step": 5248,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.0102750439155859,
+      "lm_loss": 1.2537,
+      "step": 5248,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 1.0104675505931613,
+      "grad_norm": 3.1568219977954275,
+      "learning_rate": 1.0314855515856055e-05,
+      "loss": 1.407,
+      "step": 5249
+    },
+    {
+      "epoch": 1.0106600572707365,
+      "grad_norm": 3.0545804169416173,
+      "learning_rate": 1.0311739149794394e-05,
+      "loss": 1.3806,
+      "step": 5250
+    },
+    {
+      "epoch": 1.010852563948312,
+      "grad_norm": 3.2868937988084896,
+      "learning_rate": 1.0308622753427692e-05,
+      "loss": 1.4317,
+      "step": 5251
+    },
+    {
+      "epoch": 1.0110450706258873,
+      "grad_norm": 3.279832328674485,
+      "learning_rate": 1.0305506327058916e-05,
+      "loss": 1.4319,
+      "step": 5252
+    },
+    {
+      "epoch": 1.0112375773034628,
+      "grad_norm": 3.1806904948089523,
+      "learning_rate": 1.0302389870991015e-05,
+      "loss": 1.3886,
+      "step": 5253
+    },
+    {
+      "epoch": 1.0114300839810382,
+      "grad_norm": 3.245406812470036,
+      "learning_rate": 1.0299273385526949e-05,
+      "loss": 1.4277,
+      "step": 5254
+    },
+    {
+      "epoch": 1.0116225906586134,
+      "grad_norm": 3.220546266751897,
+      "learning_rate": 1.0296156870969681e-05,
+      "loss": 1.4215,
+      "step": 5255
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "grad_norm": 3.183225485454342,
+      "learning_rate": 1.0293040327622175e-05,
+      "loss": 1.4502,
+      "step": 5256
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.5459,
+      "step": 5256,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 0.8177,
+      "step": 5256,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.3076,
+      "step": 5256,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.0591,
+      "step": 5256,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.1081,
+      "step": 5256,
+      "vm_loss": 0.0943
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.2002,
+      "step": 5256,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.3334,
+      "step": 5256,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 1.0118150973361888,
+      "lm_loss": 1.6397,
+      "step": 5256,
+      "vm_loss": 0.2071
+    },
+    {
+      "epoch": 1.0120076040137642,
+      "grad_norm": 3.047694369454868,
+      "learning_rate": 1.0289923755787401e-05,
+      "loss": 1.3061,
+      "step": 5257
+    },
+    {
+      "epoch": 1.0122001106913396,
+      "grad_norm": 3.312699106019508,
+      "learning_rate": 1.0286807155768331e-05,
+      "loss": 1.4754,
+      "step": 5258
+    },
+    {
+      "epoch": 1.012392617368915,
+      "grad_norm": 3.3788030383218257,
+      "learning_rate": 1.0283690527867931e-05,
+      "loss": 1.4469,
+      "step": 5259
+    },
+    {
+      "epoch": 1.0125851240464903,
+      "grad_norm": 3.3165415825726896,
+      "learning_rate": 1.0280573872389187e-05,
+      "loss": 1.3928,
+      "step": 5260
+    },
+    {
+      "epoch": 1.0127776307240657,
+      "grad_norm": 3.3838556072765424,
+      "learning_rate": 1.0277457189635071e-05,
+      "loss": 1.3587,
+      "step": 5261
+    },
+    {
+      "epoch": 1.012970137401641,
+      "grad_norm": 3.466213026247168,
+      "learning_rate": 1.0274340479908568e-05,
+      "loss": 1.4747,
+      "step": 5262
+    },
+    {
+      "epoch": 1.0131626440792165,
+      "grad_norm": 3.2428300917265664,
+      "learning_rate": 1.027122374351266e-05,
+      "loss": 1.4718,
+      "step": 5263
+    },
+    {
+      "epoch": 1.013355150756792,
+      "grad_norm": 3.16897949739538,
+      "learning_rate": 1.0268106980750334e-05,
+      "loss": 1.3461,
+      "step": 5264
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.1759,
+      "step": 5264,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.4544,
+      "step": 5264,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.2957,
+      "step": 5264,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.0837,
+      "step": 5264,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.3465,
+      "step": 5264,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.1683,
+      "step": 5264,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 1.303,
+      "step": 5264,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 1.013355150756792,
+      "lm_loss": 0.817,
+      "step": 5264,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 1.0135476574343674,
+      "grad_norm": 3.2342853446683413,
+      "learning_rate": 1.026499019192458e-05,
+      "loss": 1.4262,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0137401641119426,
+      "grad_norm": 3.0567282752690628,
+      "learning_rate": 1.0261873377338387e-05,
+      "loss": 1.3825,
+      "step": 5266
+    },
+    {
+      "epoch": 1.013932670789518,
+      "grad_norm": 3.5180534312384153,
+      "learning_rate": 1.0258756537294754e-05,
+      "loss": 1.4764,
+      "step": 5267
+    },
+    {
+      "epoch": 1.0141251774670934,
+      "grad_norm": 3.2360901212068582,
+      "learning_rate": 1.0255639672096673e-05,
+      "loss": 1.3408,
+      "step": 5268
+    },
+    {
+      "epoch": 1.0143176841446688,
+      "grad_norm": 3.27572343652706,
+      "learning_rate": 1.0252522782047147e-05,
+      "loss": 1.4576,
+      "step": 5269
+    },
+    {
+      "epoch": 1.0145101908222443,
+      "grad_norm": 3.2020992380999926,
+      "learning_rate": 1.0249405867449176e-05,
+      "loss": 1.3733,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0147026974998195,
+      "grad_norm": 3.20208234847383,
+      "learning_rate": 1.0246288928605762e-05,
+      "loss": 1.3456,
+      "step": 5271
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "grad_norm": 3.1915674510142016,
+      "learning_rate": 1.0243171965819914e-05,
+      "loss": 1.4144,
+      "step": 5272
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.0684,
+      "step": 5272,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.3565,
+      "step": 5272,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.0271,
+      "step": 5272,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.4733,
+      "step": 5272,
+      "vm_loss": 0.1148
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.2065,
+      "step": 5272,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.2886,
+      "step": 5272,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.2064,
+      "step": 5272,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.0148952041773949,
+      "lm_loss": 1.1793,
+      "step": 5272,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.0150877108549703,
+      "grad_norm": 3.0389727232566557,
+      "learning_rate": 1.024005497939464e-05,
+      "loss": 1.3706,
+      "step": 5273
+    },
+    {
+      "epoch": 1.0152802175325457,
+      "grad_norm": 3.1730530785317783,
+      "learning_rate": 1.0236937969632951e-05,
+      "loss": 1.3743,
+      "step": 5274
+    },
+    {
+      "epoch": 1.0154727242101211,
+      "grad_norm": 3.241070767558403,
+      "learning_rate": 1.0233820936837859e-05,
+      "loss": 1.4349,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0156652308876963,
+      "grad_norm": 3.163541255353608,
+      "learning_rate": 1.023070388131238e-05,
+      "loss": 1.39,
+      "step": 5276
+    },
+    {
+      "epoch": 1.0158577375652718,
+      "grad_norm": 3.279424677110529,
+      "learning_rate": 1.0227586803359532e-05,
+      "loss": 1.4857,
+      "step": 5277
+    },
+    {
+      "epoch": 1.0160502442428472,
+      "grad_norm": 3.0392038365640337,
+      "learning_rate": 1.0224469703282336e-05,
+      "loss": 1.3688,
+      "step": 5278
+    },
+    {
+      "epoch": 1.0162427509204226,
+      "grad_norm": 3.107688767971896,
+      "learning_rate": 1.0221352581383815e-05,
+      "loss": 1.3962,
+      "step": 5279
+    },
+    {
+      "epoch": 1.016435257597998,
+      "grad_norm": 3.229133483622222,
+      "learning_rate": 1.021823543796699e-05,
+      "loss": 1.3811,
+      "step": 5280
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.3626,
+      "step": 5280,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.2113,
+      "step": 5280,
+      "vm_loss": 0.2032
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.2378,
+      "step": 5280,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.3814,
+      "step": 5280,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.1118,
+      "step": 5280,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 1.238,
+      "step": 5280,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 0.9976,
+      "step": 5280,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 1.016435257597998,
+      "lm_loss": 0.4437,
+      "step": 5280,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 1.0166277642755732,
+      "grad_norm": 3.1895405736270965,
+      "learning_rate": 1.0215118273334885e-05,
+      "loss": 1.3868,
+      "step": 5281
+    },
+    {
+      "epoch": 1.0168202709531486,
+      "grad_norm": 3.3588583179219142,
+      "learning_rate": 1.0212001087790538e-05,
+      "loss": 1.4782,
+      "step": 5282
+    },
+    {
+      "epoch": 1.017012777630724,
+      "grad_norm": 3.1984493432076433,
+      "learning_rate": 1.0208883881636969e-05,
+      "loss": 1.339,
+      "step": 5283
+    },
+    {
+      "epoch": 1.0172052843082995,
+      "grad_norm": 3.318610884941977,
+      "learning_rate": 1.0205766655177217e-05,
+      "loss": 1.4481,
+      "step": 5284
+    },
+    {
+      "epoch": 1.017397790985875,
+      "grad_norm": 3.1703718895832793,
+      "learning_rate": 1.0202649408714309e-05,
+      "loss": 1.4301,
+      "step": 5285
+    },
+    {
+      "epoch": 1.01759029766345,
+      "grad_norm": 3.2425917172108045,
+      "learning_rate": 1.019953214255129e-05,
+      "loss": 1.4198,
+      "step": 5286
+    },
+    {
+      "epoch": 1.0177828043410255,
+      "grad_norm": 3.2128255289844456,
+      "learning_rate": 1.019641485699119e-05,
+      "loss": 1.4475,
+      "step": 5287
+    },
+    {
+      "epoch": 1.017975311018601,
+      "grad_norm": 3.269415254382278,
+      "learning_rate": 1.0193297552337055e-05,
+      "loss": 1.424,
+      "step": 5288
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 0.871,
+      "step": 5288,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 1.3448,
+      "step": 5288,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 1.4553,
+      "step": 5288,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 1.0574,
+      "step": 5288,
+      "vm_loss": 0.2351
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 0.8342,
+      "step": 5288,
+      "vm_loss": 0.1225
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 1.4492,
+      "step": 5288,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 0.9951,
+      "step": 5288,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.017975311018601,
+      "lm_loss": 0.9214,
+      "step": 5288,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 1.0181678176961764,
+      "grad_norm": 3.1809889748977906,
+      "learning_rate": 1.0190180228891928e-05,
+      "loss": 1.4325,
+      "step": 5289
+    },
+    {
+      "epoch": 1.0183603243737518,
+      "grad_norm": 3.161266610240595,
+      "learning_rate": 1.0187062886958845e-05,
+      "loss": 1.4522,
+      "step": 5290
+    },
+    {
+      "epoch": 1.018552831051327,
+      "grad_norm": 3.13937199218909,
+      "learning_rate": 1.018394552684086e-05,
+      "loss": 1.425,
+      "step": 5291
+    },
+    {
+      "epoch": 1.0187453377289024,
+      "grad_norm": 3.0696475000092236,
+      "learning_rate": 1.0180828148841013e-05,
+      "loss": 1.3128,
+      "step": 5292
+    },
+    {
+      "epoch": 1.0189378444064778,
+      "grad_norm": 3.2629891648249334,
+      "learning_rate": 1.0177710753262358e-05,
+      "loss": 1.4148,
+      "step": 5293
+    },
+    {
+      "epoch": 1.0191303510840533,
+      "grad_norm": 3.487317270262764,
+      "learning_rate": 1.0174593340407945e-05,
+      "loss": 1.4459,
+      "step": 5294
+    },
+    {
+      "epoch": 1.0193228577616287,
+      "grad_norm": 3.223753277963404,
+      "learning_rate": 1.0171475910580824e-05,
+      "loss": 1.3645,
+      "step": 5295
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "grad_norm": 3.1659643364483814,
+      "learning_rate": 1.0168358464084052e-05,
+      "loss": 1.4245,
+      "step": 5296
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.2782,
+      "step": 5296,
+      "vm_loss": 0.0919
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.207,
+      "step": 5296,
+      "vm_loss": 0.2118
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.6258,
+      "step": 5296,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.4703,
+      "step": 5296,
+      "vm_loss": 0.2561
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.4977,
+      "step": 5296,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.5082,
+      "step": 5296,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.0653,
+      "step": 5296,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.0195153644392039,
+      "lm_loss": 1.4842,
+      "step": 5296,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.0197078711167793,
+      "grad_norm": 3.309585063882577,
+      "learning_rate": 1.0165241001220682e-05,
+      "loss": 1.4798,
+      "step": 5297
+    },
+    {
+      "epoch": 1.0199003777943547,
+      "grad_norm": 3.098357780652905,
+      "learning_rate": 1.0162123522293775e-05,
+      "loss": 1.3605,
+      "step": 5298
+    },
+    {
+      "epoch": 1.0200928844719301,
+      "grad_norm": 3.285065831004035,
+      "learning_rate": 1.0159006027606385e-05,
+      "loss": 1.453,
+      "step": 5299
+    },
+    {
+      "epoch": 1.0202853911495056,
+      "grad_norm": 3.1451059635329015,
+      "learning_rate": 1.0155888517461575e-05,
+      "loss": 1.4094,
+      "step": 5300
+    },
+    {
+      "epoch": 1.020477897827081,
+      "grad_norm": 3.2605192736138937,
+      "learning_rate": 1.0152770992162407e-05,
+      "loss": 1.4339,
+      "step": 5301
+    },
+    {
+      "epoch": 1.0206704045046562,
+      "grad_norm": 3.1099133662654515,
+      "learning_rate": 1.0149653452011943e-05,
+      "loss": 1.4244,
+      "step": 5302
+    },
+    {
+      "epoch": 1.0208629111822316,
+      "grad_norm": 3.1149443589911003,
+      "learning_rate": 1.0146535897313253e-05,
+      "loss": 1.4468,
+      "step": 5303
+    },
+    {
+      "epoch": 1.021055417859807,
+      "grad_norm": 3.1183860664006313,
+      "learning_rate": 1.0143418328369394e-05,
+      "loss": 1.4413,
+      "step": 5304
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.0358,
+      "step": 5304,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.2513,
+      "step": 5304,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.2671,
+      "step": 5304,
+      "vm_loss": 0.1004
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.8601,
+      "step": 5304,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.3223,
+      "step": 5304,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.0145,
+      "step": 5304,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 0.8064,
+      "step": 5304,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 1.021055417859807,
+      "lm_loss": 1.3773,
+      "step": 5304,
+      "vm_loss": 0.2019
+    },
+    {
+      "epoch": 1.0212479245373824,
+      "grad_norm": 3.1876945926345357,
+      "learning_rate": 1.014030074548344e-05,
+      "loss": 1.3092,
+      "step": 5305
+    },
+    {
+      "epoch": 1.0214404312149579,
+      "grad_norm": 3.160294845799342,
+      "learning_rate": 1.0137183148958462e-05,
+      "loss": 1.4389,
+      "step": 5306
+    },
+    {
+      "epoch": 1.021632937892533,
+      "grad_norm": 3.1488466954151177,
+      "learning_rate": 1.0134065539097524e-05,
+      "loss": 1.3584,
+      "step": 5307
+    },
+    {
+      "epoch": 1.0218254445701085,
+      "grad_norm": 3.2183981746042423,
+      "learning_rate": 1.0130947916203701e-05,
+      "loss": 1.4,
+      "step": 5308
+    },
+    {
+      "epoch": 1.022017951247684,
+      "grad_norm": 3.309119695013585,
+      "learning_rate": 1.0127830280580064e-05,
+      "loss": 1.4486,
+      "step": 5309
+    },
+    {
+      "epoch": 1.0222104579252593,
+      "grad_norm": 3.2917959183037135,
+      "learning_rate": 1.012471263252969e-05,
+      "loss": 1.4806,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0224029646028348,
+      "grad_norm": 3.041168110733001,
+      "learning_rate": 1.0121594972355653e-05,
+      "loss": 1.3193,
+      "step": 5311
+    },
+    {
+      "epoch": 1.02259547128041,
+      "grad_norm": 3.1217711058701334,
+      "learning_rate": 1.0118477300361027e-05,
+      "loss": 1.4177,
+      "step": 5312
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.1922,
+      "step": 5312,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.8132,
+      "step": 5312,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.3263,
+      "step": 5312,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.2304,
+      "step": 5312,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.4567,
+      "step": 5312,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.2695,
+      "step": 5312,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.0845,
+      "step": 5312,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.02259547128041,
+      "lm_loss": 1.31,
+      "step": 5312,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.0227879779579854,
+      "grad_norm": 3.209033675576354,
+      "learning_rate": 1.0115359616848894e-05,
+      "loss": 1.4252,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0229804846355608,
+      "grad_norm": 3.1966544527759972,
+      "learning_rate": 1.0112241922122328e-05,
+      "loss": 1.4263,
+      "step": 5314
+    },
+    {
+      "epoch": 1.0231729913131362,
+      "grad_norm": 3.2289260728185973,
+      "learning_rate": 1.0109124216484413e-05,
+      "loss": 1.4179,
+      "step": 5315
+    },
+    {
+      "epoch": 1.0233654979907116,
+      "grad_norm": 3.2165733141536985,
+      "learning_rate": 1.0106006500238225e-05,
+      "loss": 1.394,
+      "step": 5316
+    },
+    {
+      "epoch": 1.0235580046682868,
+      "grad_norm": 3.2782471043668746,
+      "learning_rate": 1.0102888773686852e-05,
+      "loss": 1.4332,
+      "step": 5317
+    },
+    {
+      "epoch": 1.0237505113458623,
+      "grad_norm": 3.310818491794284,
+      "learning_rate": 1.0099771037133372e-05,
+      "loss": 1.4755,
+      "step": 5318
+    },
+    {
+      "epoch": 1.0239430180234377,
+      "grad_norm": 3.135705805781398,
+      "learning_rate": 1.0096653290880873e-05,
+      "loss": 1.3867,
+      "step": 5319
+    },
+    {
+      "epoch": 1.024135524701013,
+      "grad_norm": 3.2555511824794707,
+      "learning_rate": 1.0093535535232435e-05,
+      "loss": 1.4129,
+      "step": 5320
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.2205,
+      "step": 5320,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.2717,
+      "step": 5320,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.4631,
+      "step": 5320,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.2231,
+      "step": 5320,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.1232,
+      "step": 5320,
+      "vm_loss": 0.2094
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.2995,
+      "step": 5320,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.3768,
+      "step": 5320,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.024135524701013,
+      "lm_loss": 1.2309,
+      "step": 5320,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.0243280313785885,
+      "grad_norm": 3.205451557131525,
+      "learning_rate": 1.0090417770491148e-05,
+      "loss": 1.4128,
+      "step": 5321
+    },
+    {
+      "epoch": 1.0245205380561637,
+      "grad_norm": 3.0897210498051084,
+      "learning_rate": 1.0087299996960096e-05,
+      "loss": 1.3047,
+      "step": 5322
+    },
+    {
+      "epoch": 1.0247130447337391,
+      "grad_norm": 3.1896416551894,
+      "learning_rate": 1.0084182214942369e-05,
+      "loss": 1.4053,
+      "step": 5323
+    },
+    {
+      "epoch": 1.0249055514113146,
+      "grad_norm": 3.348943987544542,
+      "learning_rate": 1.0081064424741052e-05,
+      "loss": 1.3919,
+      "step": 5324
+    },
+    {
+      "epoch": 1.02509805808889,
+      "grad_norm": 3.286855495325973,
+      "learning_rate": 1.0077946626659236e-05,
+      "loss": 1.4559,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0252905647664654,
+      "grad_norm": 3.2149820459269023,
+      "learning_rate": 1.0074828821000012e-05,
+      "loss": 1.4314,
+      "step": 5326
+    },
+    {
+      "epoch": 1.0254830714440408,
+      "grad_norm": 3.1558744476761666,
+      "learning_rate": 1.0071711008066466e-05,
+      "loss": 1.4334,
+      "step": 5327
+    },
+    {
+      "epoch": 1.025675578121616,
+      "grad_norm": 2.9632586118212982,
+      "learning_rate": 1.0068593188161698e-05,
+      "loss": 1.3293,
+      "step": 5328
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.707,
+      "step": 5328,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.2245,
+      "step": 5328,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.0639,
+      "step": 5328,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.4365,
+      "step": 5328,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.4362,
+      "step": 5328,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.0708,
+      "step": 5328,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 1.0495,
+      "step": 5328,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.025675578121616,
+      "lm_loss": 0.8801,
+      "step": 5328,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.0258680847991914,
+      "grad_norm": 3.1651625019448097,
+      "learning_rate": 1.006547536158879e-05,
+      "loss": 1.3872,
+      "step": 5329
+    },
+    {
+      "epoch": 1.0260605914767669,
+      "grad_norm": 3.182450699949572,
+      "learning_rate": 1.0062357528650841e-05,
+      "loss": 1.4508,
+      "step": 5330
+    },
+    {
+      "epoch": 1.0262530981543423,
+      "grad_norm": 3.133293572588481,
+      "learning_rate": 1.0059239689650941e-05,
+      "loss": 1.4227,
+      "step": 5331
+    },
+    {
+      "epoch": 1.0264456048319177,
+      "grad_norm": 3.1361916643565504,
+      "learning_rate": 1.0056121844892188e-05,
+      "loss": 1.3474,
+      "step": 5332
+    },
+    {
+      "epoch": 1.026638111509493,
+      "grad_norm": 3.1388939111256695,
+      "learning_rate": 1.0053003994677673e-05,
+      "loss": 1.4219,
+      "step": 5333
+    },
+    {
+      "epoch": 1.0268306181870683,
+      "grad_norm": 3.26017446039176,
+      "learning_rate": 1.004988613931049e-05,
+      "loss": 1.369,
+      "step": 5334
+    },
+    {
+      "epoch": 1.0270231248646438,
+      "grad_norm": 3.3943827858384217,
+      "learning_rate": 1.0046768279093737e-05,
+      "loss": 1.4274,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "grad_norm": 3.3231513281638834,
+      "learning_rate": 1.0043650414330508e-05,
+      "loss": 1.424,
+      "step": 5336
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.6288,
+      "step": 5336,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 0.6245,
+      "step": 5336,
+      "vm_loss": 0.1128
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 0.7742,
+      "step": 5336,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.1095,
+      "step": 5336,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.1431,
+      "step": 5336,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.198,
+      "step": 5336,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.5549,
+      "step": 5336,
+      "vm_loss": 0.2308
+    },
+    {
+      "epoch": 1.0272156315422192,
+      "lm_loss": 1.7397,
+      "step": 5336,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.0274081382197946,
+      "grad_norm": 3.249015798947832,
+      "learning_rate": 1.0040532545323902e-05,
+      "loss": 1.4466,
+      "step": 5337
+    },
+    {
+      "epoch": 1.0276006448973698,
+      "grad_norm": 3.0782205659937323,
+      "learning_rate": 1.0037414672377013e-05,
+      "loss": 1.3859,
+      "step": 5338
+    },
+    {
+      "epoch": 1.0277931515749452,
+      "grad_norm": 3.135001391128802,
+      "learning_rate": 1.0034296795792936e-05,
+      "loss": 1.443,
+      "step": 5339
+    },
+    {
+      "epoch": 1.0279856582525206,
+      "grad_norm": 3.139373370248466,
+      "learning_rate": 1.0031178915874775e-05,
+      "loss": 1.4479,
+      "step": 5340
+    },
+    {
+      "epoch": 1.028178164930096,
+      "grad_norm": 3.108944125463726,
+      "learning_rate": 1.0028061032925622e-05,
+      "loss": 1.4183,
+      "step": 5341
+    },
+    {
+      "epoch": 1.0283706716076715,
+      "grad_norm": 3.0654320462283726,
+      "learning_rate": 1.0024943147248582e-05,
+      "loss": 1.3772,
+      "step": 5342
+    },
+    {
+      "epoch": 1.0285631782852467,
+      "grad_norm": 3.1211612241691045,
+      "learning_rate": 1.0021825259146747e-05,
+      "loss": 1.4313,
+      "step": 5343
+    },
+    {
+      "epoch": 1.028755684962822,
+      "grad_norm": 3.1410284394050207,
+      "learning_rate": 1.0018707368923217e-05,
+      "loss": 1.4232,
+      "step": 5344
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.3754,
+      "step": 5344,
+      "vm_loss": 0.2186
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.3445,
+      "step": 5344,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.1443,
+      "step": 5344,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 0.819,
+      "step": 5344,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.1697,
+      "step": 5344,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.1347,
+      "step": 5344,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 1.4737,
+      "step": 5344,
+      "vm_loss": 0.1247
+    },
+    {
+      "epoch": 1.028755684962822,
+      "lm_loss": 0.9111,
+      "step": 5344,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.0289481916403975,
+      "grad_norm": 3.094445676840102,
+      "learning_rate": 1.0015589476881092e-05,
+      "loss": 1.3898,
+      "step": 5345
+    },
+    {
+      "epoch": 1.029140698317973,
+      "grad_norm": 3.1060274800819236,
+      "learning_rate": 1.001247158332347e-05,
+      "loss": 1.4405,
+      "step": 5346
+    },
+    {
+      "epoch": 1.0293332049955484,
+      "grad_norm": 3.0758249715764703,
+      "learning_rate": 1.0009353688553451e-05,
+      "loss": 1.4121,
+      "step": 5347
+    },
+    {
+      "epoch": 1.0295257116731236,
+      "grad_norm": 3.284471717421342,
+      "learning_rate": 1.0006235792874133e-05,
+      "loss": 1.3676,
+      "step": 5348
+    },
+    {
+      "epoch": 1.029718218350699,
+      "grad_norm": 3.249339545384867,
+      "learning_rate": 1.0003117896588619e-05,
+      "loss": 1.4919,
+      "step": 5349
+    },
+    {
+      "epoch": 1.0299107250282744,
+      "grad_norm": 3.1553690020413434,
+      "learning_rate": 1e-05,
+      "loss": 1.3574,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0301032317058498,
+      "grad_norm": 3.132590582717691,
+      "learning_rate": 9.996882103411388e-06,
+      "loss": 1.3704,
+      "step": 5351
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "grad_norm": 3.250763919568954,
+      "learning_rate": 9.993764207125868e-06,
+      "loss": 1.3527,
+      "step": 5352
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.4645,
+      "step": 5352,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.3112,
+      "step": 5352,
+      "vm_loss": 0.1424
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.3946,
+      "step": 5352,
+      "vm_loss": 0.2251
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.0482,
+      "step": 5352,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.222,
+      "step": 5352,
+      "vm_loss": 0.2189
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 0.9827,
+      "step": 5352,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.2148,
+      "step": 5352,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 1.0302957383834253,
+      "lm_loss": 1.3597,
+      "step": 5352,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.0304882450610005,
+      "grad_norm": 3.244305837429493,
+      "learning_rate": 9.990646311446552e-06,
+      "loss": 1.3827,
+      "step": 5353
+    },
+    {
+      "epoch": 1.0306807517385759,
+      "grad_norm": 3.2438888594859816,
+      "learning_rate": 9.987528416676534e-06,
+      "loss": 1.4393,
+      "step": 5354
+    },
+    {
+      "epoch": 1.0308732584161513,
+      "grad_norm": 3.2527600413831905,
+      "learning_rate": 9.984410523118913e-06,
+      "loss": 1.3645,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0310657650937267,
+      "grad_norm": 3.1793339085471053,
+      "learning_rate": 9.981292631076785e-06,
+      "loss": 1.3937,
+      "step": 5356
+    },
+    {
+      "epoch": 1.0312582717713021,
+      "grad_norm": 3.2898567611297924,
+      "learning_rate": 9.978174740853257e-06,
+      "loss": 1.3974,
+      "step": 5357
+    },
+    {
+      "epoch": 1.0314507784488773,
+      "grad_norm": 3.202726762246539,
+      "learning_rate": 9.975056852751421e-06,
+      "loss": 1.4186,
+      "step": 5358
+    },
+    {
+      "epoch": 1.0316432851264528,
+      "grad_norm": 3.1197452600849216,
+      "learning_rate": 9.971938967074378e-06,
+      "loss": 1.367,
+      "step": 5359
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "grad_norm": 3.2109037077170477,
+      "learning_rate": 9.968821084125228e-06,
+      "loss": 1.393,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.1128,
+      "step": 5360,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.4715,
+      "step": 5360,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.216,
+      "step": 5360,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 0.8798,
+      "step": 5360,
+      "vm_loss": 0.1901
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.4744,
+      "step": 5360,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.2315,
+      "step": 5360,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 1.4365,
+      "step": 5360,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.0318357918040282,
+      "lm_loss": 0.9068,
+      "step": 5360,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.0320282984816036,
+      "grad_norm": 3.193875723913729,
+      "learning_rate": 9.965703204207065e-06,
+      "loss": 1.3459,
+      "step": 5361
+    },
+    {
+      "epoch": 1.032220805159179,
+      "grad_norm": 3.199794479122657,
+      "learning_rate": 9.962585327622992e-06,
+      "loss": 1.3892,
+      "step": 5362
+    },
+    {
+      "epoch": 1.0324133118367544,
+      "grad_norm": 3.148181520641845,
+      "learning_rate": 9.959467454676102e-06,
+      "loss": 1.4401,
+      "step": 5363
+    },
+    {
+      "epoch": 1.0326058185143296,
+      "grad_norm": 3.209367370680021,
+      "learning_rate": 9.956349585669495e-06,
+      "loss": 1.4013,
+      "step": 5364
+    },
+    {
+      "epoch": 1.032798325191905,
+      "grad_norm": 3.322079761855245,
+      "learning_rate": 9.953231720906268e-06,
+      "loss": 1.4014,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0329908318694805,
+      "grad_norm": 3.231172852821185,
+      "learning_rate": 9.950113860689512e-06,
+      "loss": 1.3964,
+      "step": 5366
+    },
+    {
+      "epoch": 1.033183338547056,
+      "grad_norm": 3.1515599259032987,
+      "learning_rate": 9.946996005322332e-06,
+      "loss": 1.3612,
+      "step": 5367
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "grad_norm": 3.277200308074966,
+      "learning_rate": 9.943878155107817e-06,
+      "loss": 1.3995,
+      "step": 5368
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.2137,
+      "step": 5368,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.2698,
+      "step": 5368,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.303,
+      "step": 5368,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.4011,
+      "step": 5368,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.4055,
+      "step": 5368,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.2743,
+      "step": 5368,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 0.9325,
+      "step": 5368,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.0333758452246313,
+      "lm_loss": 1.0772,
+      "step": 5368,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 1.0335683519022065,
+      "grad_norm": 3.1449388887374754,
+      "learning_rate": 9.940760310349059e-06,
+      "loss": 1.3498,
+      "step": 5369
+    },
+    {
+      "epoch": 1.033760858579782,
+      "grad_norm": 3.310454158858666,
+      "learning_rate": 9.937642471349162e-06,
+      "loss": 1.4872,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0339533652573574,
+      "grad_norm": 3.1365198713540816,
+      "learning_rate": 9.934524638411213e-06,
+      "loss": 1.4136,
+      "step": 5371
+    },
+    {
+      "epoch": 1.0341458719349328,
+      "grad_norm": 3.2456231230900103,
+      "learning_rate": 9.931406811838307e-06,
+      "loss": 1.3062,
+      "step": 5372
+    },
+    {
+      "epoch": 1.0343383786125082,
+      "grad_norm": 3.2647305859061784,
+      "learning_rate": 9.928288991933534e-06,
+      "loss": 1.4442,
+      "step": 5373
+    },
+    {
+      "epoch": 1.0345308852900834,
+      "grad_norm": 3.1062637771244574,
+      "learning_rate": 9.925171178999992e-06,
+      "loss": 1.3525,
+      "step": 5374
+    },
+    {
+      "epoch": 1.0347233919676588,
+      "grad_norm": 3.2955391344579996,
+      "learning_rate": 9.922053373340767e-06,
+      "loss": 1.3956,
+      "step": 5375
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "grad_norm": 3.3745698577763057,
+      "learning_rate": 9.918935575258948e-06,
+      "loss": 1.4915,
+      "step": 5376
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.6253,
+      "step": 5376,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.1861,
+      "step": 5376,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.4872,
+      "step": 5376,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.209,
+      "step": 5376,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.4559,
+      "step": 5376,
+      "vm_loss": 0.2011
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.7794,
+      "step": 5376,
+      "vm_loss": 0.1445
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 1.0959,
+      "step": 5376,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 1.0349158986452343,
+      "lm_loss": 0.8984,
+      "step": 5376,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 1.0351084053228097,
+      "grad_norm": 3.300685691166521,
+      "learning_rate": 9.915817785057634e-06,
+      "loss": 1.4182,
+      "step": 5377
+    },
+    {
+      "epoch": 1.035300912000385,
+      "grad_norm": 3.165662753519481,
+      "learning_rate": 9.912700003039907e-06,
+      "loss": 1.3947,
+      "step": 5378
+    },
+    {
+      "epoch": 1.0354934186779603,
+      "grad_norm": 3.190423305945623,
+      "learning_rate": 9.909582229508852e-06,
+      "loss": 1.3968,
+      "step": 5379
+    },
+    {
+      "epoch": 1.0356859253555357,
+      "grad_norm": 3.1729390048145407,
+      "learning_rate": 9.906464464767567e-06,
+      "loss": 1.352,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0358784320331111,
+      "grad_norm": 3.1471749043384367,
+      "learning_rate": 9.903346709119129e-06,
+      "loss": 1.387,
+      "step": 5381
+    },
+    {
+      "epoch": 1.0360709387106866,
+      "grad_norm": 3.1515754598937433,
+      "learning_rate": 9.900228962866631e-06,
+      "loss": 1.3318,
+      "step": 5382
+    },
+    {
+      "epoch": 1.036263445388262,
+      "grad_norm": 3.3230505120830216,
+      "learning_rate": 9.89711122631315e-06,
+      "loss": 1.4003,
+      "step": 5383
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "grad_norm": 3.3918096353558775,
+      "learning_rate": 9.893993499761776e-06,
+      "loss": 1.3778,
+      "step": 5384
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.5019,
+      "step": 5384,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.1124,
+      "step": 5384,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.0978,
+      "step": 5384,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.6393,
+      "step": 5384,
+      "vm_loss": 0.1272
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.2078,
+      "step": 5384,
+      "vm_loss": 0.2061
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.2014,
+      "step": 5384,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.2002,
+      "step": 5384,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.0364559520658372,
+      "lm_loss": 1.9068,
+      "step": 5384,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.0366484587434126,
+      "grad_norm": 3.4488107236522176,
+      "learning_rate": 9.890875783515592e-06,
+      "loss": 1.4161,
+      "step": 5385
+    },
+    {
+      "epoch": 1.036840965420988,
+      "grad_norm": 3.2760276075089916,
+      "learning_rate": 9.887758077877673e-06,
+      "loss": 1.3369,
+      "step": 5386
+    },
+    {
+      "epoch": 1.0370334720985634,
+      "grad_norm": 3.254576549458216,
+      "learning_rate": 9.88464038315111e-06,
+      "loss": 1.3624,
+      "step": 5387
+    },
+    {
+      "epoch": 1.0372259787761389,
+      "grad_norm": 3.276148975684649,
+      "learning_rate": 9.881522699638976e-06,
+      "loss": 1.3929,
+      "step": 5388
+    },
+    {
+      "epoch": 1.0374184854537143,
+      "grad_norm": 3.26106298682237,
+      "learning_rate": 9.878405027644352e-06,
+      "loss": 1.3947,
+      "step": 5389
+    },
+    {
+      "epoch": 1.0376109921312895,
+      "grad_norm": 3.2202465582721866,
+      "learning_rate": 9.87528736747031e-06,
+      "loss": 1.3397,
+      "step": 5390
+    },
+    {
+      "epoch": 1.037803498808865,
+      "grad_norm": 3.2293015598232446,
+      "learning_rate": 9.872169719419938e-06,
+      "loss": 1.3353,
+      "step": 5391
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "grad_norm": 3.5487671987950495,
+      "learning_rate": 9.869052083796302e-06,
+      "loss": 1.4769,
+      "step": 5392
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.285,
+      "step": 5392,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.4873,
+      "step": 5392,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.2664,
+      "step": 5392,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.6228,
+      "step": 5392,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.3204,
+      "step": 5392,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.4353,
+      "step": 5392,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.1501,
+      "step": 5392,
+      "vm_loss": 0.2052
+    },
+    {
+      "epoch": 1.0379960054864403,
+      "lm_loss": 1.3081,
+      "step": 5392,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.0381885121640158,
+      "grad_norm": 3.108028384484025,
+      "learning_rate": 9.865934460902476e-06,
+      "loss": 1.4117,
+      "step": 5393
+    },
+    {
+      "epoch": 1.0383810188415912,
+      "grad_norm": 3.0883729779283806,
+      "learning_rate": 9.862816851041541e-06,
+      "loss": 1.3842,
+      "step": 5394
+    },
+    {
+      "epoch": 1.0385735255191664,
+      "grad_norm": 3.1929596088229384,
+      "learning_rate": 9.859699254516562e-06,
+      "loss": 1.3792,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0387660321967418,
+      "grad_norm": 3.2863250794078804,
+      "learning_rate": 9.856581671630609e-06,
+      "loss": 1.4054,
+      "step": 5396
+    },
+    {
+      "epoch": 1.0389585388743172,
+      "grad_norm": 3.1925937920225405,
+      "learning_rate": 9.85346410268675e-06,
+      "loss": 1.3145,
+      "step": 5397
+    },
+    {
+      "epoch": 1.0391510455518926,
+      "grad_norm": 3.262136834044595,
+      "learning_rate": 9.850346547988059e-06,
+      "loss": 1.3324,
+      "step": 5398
+    },
+    {
+      "epoch": 1.039343552229468,
+      "grad_norm": 3.1923735629517824,
+      "learning_rate": 9.847229007837598e-06,
+      "loss": 1.3257,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "grad_norm": 3.2093886194811674,
+      "learning_rate": 9.844111482538427e-06,
+      "loss": 1.3253,
+      "step": 5400
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 1.1906,
+      "step": 5400,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 1.3638,
+      "step": 5400,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 0.9292,
+      "step": 5400,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 1.1142,
+      "step": 5400,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 1.6241,
+      "step": 5400,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 0.6412,
+      "step": 5400,
+      "vm_loss": 0.1233
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 0.977,
+      "step": 5400,
+      "vm_loss": 0.0819
+    },
+    {
+      "epoch": 1.0395360589070433,
+      "lm_loss": 1.7489,
+      "step": 5400,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 1.0397285655846187,
+      "grad_norm": 3.20446940176017,
+      "learning_rate": 9.84099397239362e-06,
+      "loss": 1.4377,
+      "step": 5401
+    },
+    {
+      "epoch": 1.039921072262194,
+      "grad_norm": 3.3628851219098705,
+      "learning_rate": 9.83787647770623e-06,
+      "loss": 1.3535,
+      "step": 5402
+    },
+    {
+      "epoch": 1.0401135789397695,
+      "grad_norm": 3.1598105163862638,
+      "learning_rate": 9.83475899877932e-06,
+      "loss": 1.3382,
+      "step": 5403
+    },
+    {
+      "epoch": 1.040306085617345,
+      "grad_norm": 3.1081425579274735,
+      "learning_rate": 9.831641535915953e-06,
+      "loss": 1.2984,
+      "step": 5404
+    },
+    {
+      "epoch": 1.0404985922949201,
+      "grad_norm": 3.337070743893075,
+      "learning_rate": 9.82852408941918e-06,
+      "loss": 1.4279,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0406910989724956,
+      "grad_norm": 3.192534793407697,
+      "learning_rate": 9.82540665959206e-06,
+      "loss": 1.445,
+      "step": 5406
+    },
+    {
+      "epoch": 1.040883605650071,
+      "grad_norm": 3.2348921877666372,
+      "learning_rate": 9.822289246737642e-06,
+      "loss": 1.3463,
+      "step": 5407
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "grad_norm": 3.137003976646942,
+      "learning_rate": 9.81917185115899e-06,
+      "loss": 1.3859,
+      "step": 5408
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 0.791,
+      "step": 5408,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.2919,
+      "step": 5408,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.1612,
+      "step": 5408,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.574,
+      "step": 5408,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.1696,
+      "step": 5408,
+      "vm_loss": 0.1126
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 0.9024,
+      "step": 5408,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.286,
+      "step": 5408,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.0410761123276464,
+      "lm_loss": 1.1437,
+      "step": 5408,
+      "vm_loss": 0.1033
+    },
+    {
+      "epoch": 1.0412686190052218,
+      "grad_norm": 3.195562145687875,
+      "learning_rate": 9.816054473159145e-06,
+      "loss": 1.289,
+      "step": 5409
+    },
+    {
+      "epoch": 1.041461125682797,
+      "grad_norm": 3.501157576003049,
+      "learning_rate": 9.812937113041155e-06,
+      "loss": 1.4108,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0416536323603725,
+      "grad_norm": 3.2252080430300305,
+      "learning_rate": 9.809819771108076e-06,
+      "loss": 1.345,
+      "step": 5411
+    },
+    {
+      "epoch": 1.0418461390379479,
+      "grad_norm": 3.1686936266317023,
+      "learning_rate": 9.806702447662946e-06,
+      "loss": 1.3906,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0420386457155233,
+      "grad_norm": 3.1699792283707944,
+      "learning_rate": 9.803585143008814e-06,
+      "loss": 1.3431,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0422311523930987,
+      "grad_norm": 3.149811572761568,
+      "learning_rate": 9.800467857448713e-06,
+      "loss": 1.3529,
+      "step": 5414
+    },
+    {
+      "epoch": 1.042423659070674,
+      "grad_norm": 3.1949878323993324,
+      "learning_rate": 9.797350591285694e-06,
+      "loss": 1.4059,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "grad_norm": 3.42104176058902,
+      "learning_rate": 9.79423334482279e-06,
+      "loss": 1.4167,
+      "step": 5416
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.2595,
+      "step": 5416,
+      "vm_loss": 0.2135
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.567,
+      "step": 5416,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.2305,
+      "step": 5416,
+      "vm_loss": 0.2172
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.1163,
+      "step": 5416,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.5555,
+      "step": 5416,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.2243,
+      "step": 5416,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.9649,
+      "step": 5416,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 1.0426161657482493,
+      "lm_loss": 1.0061,
+      "step": 5416,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 1.0428086724258248,
+      "grad_norm": 3.3125370344067804,
+      "learning_rate": 9.791116118363033e-06,
+      "loss": 1.3985,
+      "step": 5417
+    },
+    {
+      "epoch": 1.0430011791034002,
+      "grad_norm": 3.255981493727488,
+      "learning_rate": 9.787998912209466e-06,
+      "loss": 1.394,
+      "step": 5418
+    },
+    {
+      "epoch": 1.0431936857809756,
+      "grad_norm": 3.244088810004539,
+      "learning_rate": 9.784881726665117e-06,
+      "loss": 1.4061,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0433861924585508,
+      "grad_norm": 3.2455077384495183,
+      "learning_rate": 9.781764562033011e-06,
+      "loss": 1.398,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0435786991361262,
+      "grad_norm": 3.2520388112240517,
+      "learning_rate": 9.778647418616187e-06,
+      "loss": 1.336,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0437712058137016,
+      "grad_norm": 3.295193023030233,
+      "learning_rate": 9.775530296717666e-06,
+      "loss": 1.3657,
+      "step": 5422
+    },
+    {
+      "epoch": 1.043963712491277,
+      "grad_norm": 3.370179588026639,
+      "learning_rate": 9.77241319664047e-06,
+      "loss": 1.327,
+      "step": 5423
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "grad_norm": 3.192534190234887,
+      "learning_rate": 9.769296118687623e-06,
+      "loss": 1.3379,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.4854,
+      "step": 5424,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.796,
+      "step": 5424,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.0299,
+      "step": 5424,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.554,
+      "step": 5424,
+      "vm_loss": 0.1009
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.3919,
+      "step": 5424,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 1.0757,
+      "step": 5424,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 0.8189,
+      "step": 5424,
+      "vm_loss": 0.2273
+    },
+    {
+      "epoch": 1.0441562191688525,
+      "lm_loss": 0.7756,
+      "step": 5424,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 1.044348725846428,
+      "grad_norm": 3.4266427025458777,
+      "learning_rate": 9.766179063162143e-06,
+      "loss": 1.4116,
+      "step": 5425
+    },
+    {
+      "epoch": 1.044541232524003,
+      "grad_norm": 3.3775988136593296,
+      "learning_rate": 9.763062030367054e-06,
+      "loss": 1.3926,
+      "step": 5426
+    },
+    {
+      "epoch": 1.0447337392015785,
+      "grad_norm": 3.2975218995845075,
+      "learning_rate": 9.759945020605363e-06,
+      "loss": 1.4077,
+      "step": 5427
+    },
+    {
+      "epoch": 1.044926245879154,
+      "grad_norm": 3.3322400263594063,
+      "learning_rate": 9.756828034180088e-06,
+      "loss": 1.4358,
+      "step": 5428
+    },
+    {
+      "epoch": 1.0451187525567294,
+      "grad_norm": 3.1956105877185625,
+      "learning_rate": 9.753711071394242e-06,
+      "loss": 1.414,
+      "step": 5429
+    },
+    {
+      "epoch": 1.0453112592343048,
+      "grad_norm": 3.209203842790889,
+      "learning_rate": 9.75059413255083e-06,
+      "loss": 1.4038,
+      "step": 5430
+    },
+    {
+      "epoch": 1.04550376591188,
+      "grad_norm": 3.1365457548324613,
+      "learning_rate": 9.747477217952855e-06,
+      "loss": 1.3633,
+      "step": 5431
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "grad_norm": 3.161884700784253,
+      "learning_rate": 9.74436032790333e-06,
+      "loss": 1.4195,
+      "step": 5432
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.2009,
+      "step": 5432,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.0991,
+      "step": 5432,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.2694,
+      "step": 5432,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.6871,
+      "step": 5432,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.324,
+      "step": 5432,
+      "vm_loss": 0.1079
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 0.8793,
+      "step": 5432,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.1131,
+      "step": 5432,
+      "vm_loss": 0.0766
+    },
+    {
+      "epoch": 1.0456962725894554,
+      "lm_loss": 1.8577,
+      "step": 5432,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.0458887792670308,
+      "grad_norm": 3.268027301590538,
+      "learning_rate": 9.741243462705251e-06,
+      "loss": 1.3532,
+      "step": 5433
+    },
+    {
+      "epoch": 1.0460812859446063,
+      "grad_norm": 3.359714236854639,
+      "learning_rate": 9.738126622661613e-06,
+      "loss": 1.3868,
+      "step": 5434
+    },
+    {
+      "epoch": 1.0462737926221817,
+      "grad_norm": 3.3355513954355325,
+      "learning_rate": 9.735009808075422e-06,
+      "loss": 1.3391,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0464662992997569,
+      "grad_norm": 3.2965579103438074,
+      "learning_rate": 9.731893019249669e-06,
+      "loss": 1.3524,
+      "step": 5436
+    },
+    {
+      "epoch": 1.0466588059773323,
+      "grad_norm": 3.4936833672772427,
+      "learning_rate": 9.72877625648734e-06,
+      "loss": 1.4113,
+      "step": 5437
+    },
+    {
+      "epoch": 1.0468513126549077,
+      "grad_norm": 3.2731217216358446,
+      "learning_rate": 9.725659520091433e-06,
+      "loss": 1.4354,
+      "step": 5438
+    },
+    {
+      "epoch": 1.0470438193324831,
+      "grad_norm": 3.11045425622832,
+      "learning_rate": 9.72254281036493e-06,
+      "loss": 1.3663,
+      "step": 5439
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "grad_norm": 3.073939093369468,
+      "learning_rate": 9.719426127610818e-06,
+      "loss": 1.3429,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 0.8708,
+      "step": 5440,
+      "vm_loss": 0.2349
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.0635,
+      "step": 5440,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.7092,
+      "step": 5440,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.1106,
+      "step": 5440,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.2037,
+      "step": 5440,
+      "vm_loss": 0.1038
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.2045,
+      "step": 5440,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 1.0665,
+      "step": 5440,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.0472363260100586,
+      "lm_loss": 0.8388,
+      "step": 5440,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 1.0474288326876338,
+      "grad_norm": 3.1840058326940213,
+      "learning_rate": 9.716309472132069e-06,
+      "loss": 1.3188,
+      "step": 5441
+    },
+    {
+      "epoch": 1.0476213393652092,
+      "grad_norm": 3.073987586177332,
+      "learning_rate": 9.713192844231674e-06,
+      "loss": 1.2815,
+      "step": 5442
+    },
+    {
+      "epoch": 1.0478138460427846,
+      "grad_norm": 3.1318191979639263,
+      "learning_rate": 9.7100762442126e-06,
+      "loss": 1.3444,
+      "step": 5443
+    },
+    {
+      "epoch": 1.04800635272036,
+      "grad_norm": 3.2230700538362544,
+      "learning_rate": 9.706959672377826e-06,
+      "loss": 1.2817,
+      "step": 5444
+    },
+    {
+      "epoch": 1.0481988593979354,
+      "grad_norm": 3.454312697421651,
+      "learning_rate": 9.703843129030324e-06,
+      "loss": 1.3167,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0483913660755106,
+      "grad_norm": 3.4157576555049696,
+      "learning_rate": 9.700726614473055e-06,
+      "loss": 1.3455,
+      "step": 5446
+    },
+    {
+      "epoch": 1.048583872753086,
+      "grad_norm": 3.363796833928215,
+      "learning_rate": 9.69761012900899e-06,
+      "loss": 1.3621,
+      "step": 5447
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "grad_norm": 3.3421109094411996,
+      "learning_rate": 9.694493672941086e-06,
+      "loss": 1.3925,
+      "step": 5448
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.8137,
+      "step": 5448,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.0316,
+      "step": 5448,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.6715,
+      "step": 5448,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.2971,
+      "step": 5448,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.3823,
+      "step": 5448,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.7679,
+      "step": 5448,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 1.4051,
+      "step": 5448,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.0487763794306615,
+      "lm_loss": 0.9337,
+      "step": 5448,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.048968886108237,
+      "grad_norm": 3.2010205556991176,
+      "learning_rate": 9.69137724657231e-06,
+      "loss": 1.4144,
+      "step": 5449
+    },
+    {
+      "epoch": 1.0491613927858123,
+      "grad_norm": 3.163163683838556,
+      "learning_rate": 9.688260850205613e-06,
+      "loss": 1.3706,
+      "step": 5450
+    },
+    {
+      "epoch": 1.0493538994633875,
+      "grad_norm": 3.2024519043361077,
+      "learning_rate": 9.685144484143945e-06,
+      "loss": 1.3649,
+      "step": 5451
+    },
+    {
+      "epoch": 1.049546406140963,
+      "grad_norm": 3.2186631444263765,
+      "learning_rate": 9.682028148690267e-06,
+      "loss": 1.3479,
+      "step": 5452
+    },
+    {
+      "epoch": 1.0497389128185384,
+      "grad_norm": 3.1590699901924046,
+      "learning_rate": 9.67891184414752e-06,
+      "loss": 1.2377,
+      "step": 5453
+    },
+    {
+      "epoch": 1.0499314194961138,
+      "grad_norm": 3.4004263535349124,
+      "learning_rate": 9.675795570818645e-06,
+      "loss": 1.361,
+      "step": 5454
+    },
+    {
+      "epoch": 1.0501239261736892,
+      "grad_norm": 3.3908322976794607,
+      "learning_rate": 9.672679329006593e-06,
+      "loss": 1.3957,
+      "step": 5455
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "grad_norm": 3.582504419373602,
+      "learning_rate": 9.669563119014299e-06,
+      "loss": 1.3794,
+      "step": 5456
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.1451,
+      "step": 5456,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.2296,
+      "step": 5456,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.8296,
+      "step": 5456,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.6063,
+      "step": 5456,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.1651,
+      "step": 5456,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.1,
+      "step": 5456,
+      "vm_loss": 0.1004
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 1.3713,
+      "step": 5456,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 1.0503164328512646,
+      "lm_loss": 0.8877,
+      "step": 5456,
+      "vm_loss": 0.1139
+    },
+    {
+      "epoch": 1.0505089395288398,
+      "grad_norm": 3.258457749628126,
+      "learning_rate": 9.666446941144696e-06,
+      "loss": 1.3779,
+      "step": 5457
+    },
+    {
+      "epoch": 1.0507014462064153,
+      "grad_norm": 3.13787964693258,
+      "learning_rate": 9.663330795700714e-06,
+      "loss": 1.356,
+      "step": 5458
+    },
+    {
+      "epoch": 1.0508939528839907,
+      "grad_norm": 3.0945007688429027,
+      "learning_rate": 9.66021468298529e-06,
+      "loss": 1.3179,
+      "step": 5459
+    },
+    {
+      "epoch": 1.051086459561566,
+      "grad_norm": 3.1067042166063388,
+      "learning_rate": 9.657098603301347e-06,
+      "loss": 1.3659,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0512789662391415,
+      "grad_norm": 3.1037167090908944,
+      "learning_rate": 9.653982556951802e-06,
+      "loss": 1.3527,
+      "step": 5461
+    },
+    {
+      "epoch": 1.0514714729167167,
+      "grad_norm": 3.302168792952475,
+      "learning_rate": 9.650866544239582e-06,
+      "loss": 1.3317,
+      "step": 5462
+    },
+    {
+      "epoch": 1.0516639795942921,
+      "grad_norm": 3.2818875272593484,
+      "learning_rate": 9.647750565467602e-06,
+      "loss": 1.3724,
+      "step": 5463
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "grad_norm": 3.290746685639992,
+      "learning_rate": 9.644634620938771e-06,
+      "loss": 1.3587,
+      "step": 5464
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.1375,
+      "step": 5464,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.0107,
+      "step": 5464,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 0.808,
+      "step": 5464,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.3957,
+      "step": 5464,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.4544,
+      "step": 5464,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.144,
+      "step": 5464,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.3339,
+      "step": 5464,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.0518564862718676,
+      "lm_loss": 1.6451,
+      "step": 5464,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 1.052048992949443,
+      "grad_norm": 3.1326206069678846,
+      "learning_rate": 9.641518710956001e-06,
+      "loss": 1.3477,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0522414996270184,
+      "grad_norm": 3.204884544063744,
+      "learning_rate": 9.638402835822203e-06,
+      "loss": 1.3998,
+      "step": 5466
+    },
+    {
+      "epoch": 1.0524340063045936,
+      "grad_norm": 3.2836470642143762,
+      "learning_rate": 9.63528699584027e-06,
+      "loss": 1.4714,
+      "step": 5467
+    },
+    {
+      "epoch": 1.052626512982169,
+      "grad_norm": 3.24579415508618,
+      "learning_rate": 9.63217119131311e-06,
+      "loss": 1.3421,
+      "step": 5468
+    },
+    {
+      "epoch": 1.0528190196597444,
+      "grad_norm": 3.204660668685535,
+      "learning_rate": 9.629055422543613e-06,
+      "loss": 1.3596,
+      "step": 5469
+    },
+    {
+      "epoch": 1.0530115263373199,
+      "grad_norm": 3.232675305449826,
+      "learning_rate": 9.62593968983468e-06,
+      "loss": 1.3669,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0532040330148953,
+      "grad_norm": 3.0967130902059075,
+      "learning_rate": 9.622823993489193e-06,
+      "loss": 1.3964,
+      "step": 5471
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "grad_norm": 3.1226377764884528,
+      "learning_rate": 9.619708333810035e-06,
+      "loss": 1.3929,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.5588,
+      "step": 5472,
+      "vm_loss": 0.2292
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 0.894,
+      "step": 5472,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.4592,
+      "step": 5472,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.4189,
+      "step": 5472,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.0472,
+      "step": 5472,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.2052,
+      "step": 5472,
+      "vm_loss": 0.1119
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 0.9274,
+      "step": 5472,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.0533965396924705,
+      "lm_loss": 1.6927,
+      "step": 5472,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.053589046370046,
+      "grad_norm": 3.1968328058448,
+      "learning_rate": 9.616592711100097e-06,
+      "loss": 1.424,
+      "step": 5473
+    },
+    {
+      "epoch": 1.0537815530476213,
+      "grad_norm": 3.1657660770776648,
+      "learning_rate": 9.613477125662255e-06,
+      "loss": 1.3533,
+      "step": 5474
+    },
+    {
+      "epoch": 1.0539740597251968,
+      "grad_norm": 3.267773534815087,
+      "learning_rate": 9.610361577799373e-06,
+      "loss": 1.3784,
+      "step": 5475
+    },
+    {
+      "epoch": 1.0541665664027722,
+      "grad_norm": 3.2379643399334026,
+      "learning_rate": 9.607246067814338e-06,
+      "loss": 1.3694,
+      "step": 5476
+    },
+    {
+      "epoch": 1.0543590730803474,
+      "grad_norm": 3.2997355524433685,
+      "learning_rate": 9.604130596010009e-06,
+      "loss": 1.3979,
+      "step": 5477
+    },
+    {
+      "epoch": 1.0545515797579228,
+      "grad_norm": 3.3484553664352545,
+      "learning_rate": 9.601015162689245e-06,
+      "loss": 1.4398,
+      "step": 5478
+    },
+    {
+      "epoch": 1.0547440864354982,
+      "grad_norm": 3.2034867993858227,
+      "learning_rate": 9.597899768154917e-06,
+      "loss": 1.336,
+      "step": 5479
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "grad_norm": 3.247468190994917,
+      "learning_rate": 9.594784412709876e-06,
+      "loss": 1.3769,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.2273,
+      "step": 5480,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 0.8006,
+      "step": 5480,
+      "vm_loss": 0.2182
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.2139,
+      "step": 5480,
+      "vm_loss": 0.1171
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.1277,
+      "step": 5480,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 0.8925,
+      "step": 5480,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.2873,
+      "step": 5480,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.7226,
+      "step": 5480,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.0549365931130736,
+      "lm_loss": 1.1429,
+      "step": 5480,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 1.055129099790649,
+      "grad_norm": 3.1462705924497696,
+      "learning_rate": 9.591669096656975e-06,
+      "loss": 1.3087,
+      "step": 5481
+    },
+    {
+      "epoch": 1.0553216064682243,
+      "grad_norm": 3.1607823724074673,
+      "learning_rate": 9.588553820299057e-06,
+      "loss": 1.401,
+      "step": 5482
+    },
+    {
+      "epoch": 1.0555141131457997,
+      "grad_norm": 3.1466229700315536,
+      "learning_rate": 9.585438583938976e-06,
+      "loss": 1.3744,
+      "step": 5483
+    },
+    {
+      "epoch": 1.055706619823375,
+      "grad_norm": 3.1055546860091576,
+      "learning_rate": 9.582323387879569e-06,
+      "loss": 1.2741,
+      "step": 5484
+    },
+    {
+      "epoch": 1.0558991265009505,
+      "grad_norm": 3.1687489440084318,
+      "learning_rate": 9.579208232423667e-06,
+      "loss": 1.3701,
+      "step": 5485
+    },
+    {
+      "epoch": 1.056091633178526,
+      "grad_norm": 3.2362720401673,
+      "learning_rate": 9.576093117874112e-06,
+      "loss": 1.4174,
+      "step": 5486
+    },
+    {
+      "epoch": 1.0562841398561011,
+      "grad_norm": 3.2887288922445426,
+      "learning_rate": 9.572978044533729e-06,
+      "loss": 1.416,
+      "step": 5487
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "grad_norm": 3.202312231641938,
+      "learning_rate": 9.569863012705345e-06,
+      "loss": 1.3477,
+      "step": 5488
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 1.2546,
+      "step": 5488,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 0.9154,
+      "step": 5488,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 1.4171,
+      "step": 5488,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 0.7775,
+      "step": 5488,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 1.4801,
+      "step": 5488,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 0.9349,
+      "step": 5488,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 0.7298,
+      "step": 5488,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 1.0564766465336766,
+      "lm_loss": 1.5569,
+      "step": 5488,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.056669153211252,
+      "grad_norm": 3.354804592943215,
+      "learning_rate": 9.566748022691776e-06,
+      "loss": 1.324,
+      "step": 5489
+    },
+    {
+      "epoch": 1.0568616598888274,
+      "grad_norm": 3.1340332711030747,
+      "learning_rate": 9.563633074795843e-06,
+      "loss": 1.3168,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0570541665664028,
+      "grad_norm": 3.19717365800173,
+      "learning_rate": 9.56051816932036e-06,
+      "loss": 1.3949,
+      "step": 5491
+    },
+    {
+      "epoch": 1.0572466732439783,
+      "grad_norm": 3.1980309002715295,
+      "learning_rate": 9.557403306568128e-06,
+      "loss": 1.3398,
+      "step": 5492
+    },
+    {
+      "epoch": 1.0574391799215535,
+      "grad_norm": 3.1294102643117334,
+      "learning_rate": 9.554288486841963e-06,
+      "loss": 1.3647,
+      "step": 5493
+    },
+    {
+      "epoch": 1.0576316865991289,
+      "grad_norm": 3.1787080791133797,
+      "learning_rate": 9.551173710444658e-06,
+      "loss": 1.3546,
+      "step": 5494
+    },
+    {
+      "epoch": 1.0578241932767043,
+      "grad_norm": 3.1608070672607878,
+      "learning_rate": 9.548058977679009e-06,
+      "loss": 1.4116,
+      "step": 5495
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "grad_norm": 3.0895406186405743,
+      "learning_rate": 9.544944288847811e-06,
+      "loss": 1.2737,
+      "step": 5496
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.1827,
+      "step": 5496,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.018,
+      "step": 5496,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.7915,
+      "step": 5496,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.2585,
+      "step": 5496,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.1598,
+      "step": 5496,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.2979,
+      "step": 5496,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.3715,
+      "step": 5496,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.0580166999542797,
+      "lm_loss": 1.0014,
+      "step": 5496,
+      "vm_loss": 0.0986
+    },
+    {
+      "epoch": 1.0582092066318551,
+      "grad_norm": 3.1990702556244517,
+      "learning_rate": 9.541829644253851e-06,
+      "loss": 1.3976,
+      "step": 5497
+    },
+    {
+      "epoch": 1.0584017133094303,
+      "grad_norm": 3.2423345582479715,
+      "learning_rate": 9.53871504419991e-06,
+      "loss": 1.4015,
+      "step": 5498
+    },
+    {
+      "epoch": 1.0585942199870058,
+      "grad_norm": 3.234646377479755,
+      "learning_rate": 9.535600488988765e-06,
+      "loss": 1.3359,
+      "step": 5499
+    },
+    {
+      "epoch": 1.0587867266645812,
+      "grad_norm": 3.283801487146299,
+      "learning_rate": 9.532485978923199e-06,
+      "loss": 1.3731,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0589792333421566,
+      "grad_norm": 3.4712007327410648,
+      "learning_rate": 9.529371514305976e-06,
+      "loss": 1.3819,
+      "step": 5501
+    },
+    {
+      "epoch": 1.059171740019732,
+      "grad_norm": 3.250546190873005,
+      "learning_rate": 9.526257095439859e-06,
+      "loss": 1.3562,
+      "step": 5502
+    },
+    {
+      "epoch": 1.0593642466973072,
+      "grad_norm": 3.1773738322890086,
+      "learning_rate": 9.523142722627616e-06,
+      "loss": 1.3108,
+      "step": 5503
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "grad_norm": 3.1614215400053007,
+      "learning_rate": 9.520028396172002e-06,
+      "loss": 1.3152,
+      "step": 5504
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.0594,
+      "step": 5504,
+      "vm_loss": 0.1881
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.1106,
+      "step": 5504,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.6877,
+      "step": 5504,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.1166,
+      "step": 5504,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.3703,
+      "step": 5504,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 0.9894,
+      "step": 5504,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.0837,
+      "step": 5504,
+      "vm_loss": 0.2254
+    },
+    {
+      "epoch": 1.0595567533748826,
+      "lm_loss": 1.198,
+      "step": 5504,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.059749260052458,
+      "grad_norm": 3.14156539454186,
+      "learning_rate": 9.516914116375768e-06,
+      "loss": 1.3315,
+      "step": 5505
+    },
+    {
+      "epoch": 1.0599417667300335,
+      "grad_norm": 3.131932772847044,
+      "learning_rate": 9.51379988354166e-06,
+      "loss": 1.3831,
+      "step": 5506
+    },
+    {
+      "epoch": 1.060134273407609,
+      "grad_norm": 3.107728989187229,
+      "learning_rate": 9.510685697972429e-06,
+      "loss": 1.347,
+      "step": 5507
+    },
+    {
+      "epoch": 1.060326780085184,
+      "grad_norm": 3.27751325877077,
+      "learning_rate": 9.507571559970802e-06,
+      "loss": 1.4523,
+      "step": 5508
+    },
+    {
+      "epoch": 1.0605192867627595,
+      "grad_norm": 3.2564747721848604,
+      "learning_rate": 9.504457469839522e-06,
+      "loss": 1.3724,
+      "step": 5509
+    },
+    {
+      "epoch": 1.060711793440335,
+      "grad_norm": 3.1804606065999947,
+      "learning_rate": 9.501343427881317e-06,
+      "loss": 1.2729,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0609043001179104,
+      "grad_norm": 3.2315894137636527,
+      "learning_rate": 9.49822943439891e-06,
+      "loss": 1.4042,
+      "step": 5511
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "grad_norm": 3.2250274987852356,
+      "learning_rate": 9.49511548969502e-06,
+      "loss": 1.3667,
+      "step": 5512
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.3068,
+      "step": 5512,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 0.7451,
+      "step": 5512,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.3144,
+      "step": 5512,
+      "vm_loss": 0.2083
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.3018,
+      "step": 5512,
+      "vm_loss": 0.2
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.2121,
+      "step": 5512,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.469,
+      "step": 5512,
+      "vm_loss": 0.2339
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.5717,
+      "step": 5512,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.0610968067954858,
+      "lm_loss": 1.1969,
+      "step": 5512,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.061289313473061,
+      "grad_norm": 3.161972446545002,
+      "learning_rate": 9.492001594072364e-06,
+      "loss": 1.3499,
+      "step": 5513
+    },
+    {
+      "epoch": 1.0614818201506364,
+      "grad_norm": 3.1706424609459654,
+      "learning_rate": 9.488887747833655e-06,
+      "loss": 1.4026,
+      "step": 5514
+    },
+    {
+      "epoch": 1.0616743268282118,
+      "grad_norm": 3.2230485621414475,
+      "learning_rate": 9.485773951281594e-06,
+      "loss": 1.3903,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0618668335057873,
+      "grad_norm": 3.0607407945570997,
+      "learning_rate": 9.482660204718882e-06,
+      "loss": 1.3522,
+      "step": 5516
+    },
+    {
+      "epoch": 1.0620593401833627,
+      "grad_norm": 3.1370278677580616,
+      "learning_rate": 9.47954650844822e-06,
+      "loss": 1.3724,
+      "step": 5517
+    },
+    {
+      "epoch": 1.062251846860938,
+      "grad_norm": 3.1852560086999655,
+      "learning_rate": 9.476432862772297e-06,
+      "loss": 1.3959,
+      "step": 5518
+    },
+    {
+      "epoch": 1.0624443535385133,
+      "grad_norm": 3.1122660643811337,
+      "learning_rate": 9.473319267993793e-06,
+      "loss": 1.383,
+      "step": 5519
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "grad_norm": 3.115348971023863,
+      "learning_rate": 9.470205724415397e-06,
+      "loss": 1.3538,
+      "step": 5520
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 0.6927,
+      "step": 5520,
+      "vm_loss": 0.2302
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 1.4636,
+      "step": 5520,
+      "vm_loss": 0.2369
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 0.9029,
+      "step": 5520,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 1.2039,
+      "step": 5520,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 1.1532,
+      "step": 5520,
+      "vm_loss": 0.152
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 0.8009,
+      "step": 5520,
+      "vm_loss": 0.1272
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 0.8479,
+      "step": 5520,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 1.0626368602160887,
+      "lm_loss": 1.5394,
+      "step": 5520,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.0628293668936641,
+      "grad_norm": 3.214962216060597,
+      "learning_rate": 9.467092232339785e-06,
+      "loss": 1.3926,
+      "step": 5521
+    },
+    {
+      "epoch": 1.0630218735712396,
+      "grad_norm": 3.257986693392231,
+      "learning_rate": 9.463978792069624e-06,
+      "loss": 1.3756,
+      "step": 5522
+    },
+    {
+      "epoch": 1.063214380248815,
+      "grad_norm": 3.3293203971725145,
+      "learning_rate": 9.46086540390758e-06,
+      "loss": 1.3543,
+      "step": 5523
+    },
+    {
+      "epoch": 1.0634068869263902,
+      "grad_norm": 3.214518068462653,
+      "learning_rate": 9.457752068156318e-06,
+      "loss": 1.3075,
+      "step": 5524
+    },
+    {
+      "epoch": 1.0635993936039656,
+      "grad_norm": 3.30846353584957,
+      "learning_rate": 9.454638785118494e-06,
+      "loss": 1.3906,
+      "step": 5525
+    },
+    {
+      "epoch": 1.063791900281541,
+      "grad_norm": 3.1405904433221874,
+      "learning_rate": 9.451525555096753e-06,
+      "loss": 1.3856,
+      "step": 5526
+    },
+    {
+      "epoch": 1.0639844069591164,
+      "grad_norm": 3.1499927362033153,
+      "learning_rate": 9.44841237839375e-06,
+      "loss": 1.3788,
+      "step": 5527
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "grad_norm": 3.1332903103264544,
+      "learning_rate": 9.445299255312117e-06,
+      "loss": 1.3226,
+      "step": 5528
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 0.5048,
+      "step": 5528,
+      "vm_loss": 0.1027
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.3757,
+      "step": 5528,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.0306,
+      "step": 5528,
+      "vm_loss": 0.2452
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.5186,
+      "step": 5528,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.0104,
+      "step": 5528,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.0585,
+      "step": 5528,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.8206,
+      "step": 5528,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 1.0641769136366919,
+      "lm_loss": 1.0597,
+      "step": 5528,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 1.064369420314267,
+      "grad_norm": 3.0925379134096875,
+      "learning_rate": 9.442186186154493e-06,
+      "loss": 1.3705,
+      "step": 5529
+    },
+    {
+      "epoch": 1.0645619269918425,
+      "grad_norm": 3.3061318942947135,
+      "learning_rate": 9.439073171223511e-06,
+      "loss": 1.3784,
+      "step": 5530
+    },
+    {
+      "epoch": 1.064754433669418,
+      "grad_norm": 3.25435156961351,
+      "learning_rate": 9.435960210821792e-06,
+      "loss": 1.3472,
+      "step": 5531
+    },
+    {
+      "epoch": 1.0649469403469933,
+      "grad_norm": 3.26852388794344,
+      "learning_rate": 9.432847305251957e-06,
+      "loss": 1.3235,
+      "step": 5532
+    },
+    {
+      "epoch": 1.0651394470245688,
+      "grad_norm": 3.2186926054919303,
+      "learning_rate": 9.429734454816616e-06,
+      "loss": 1.3413,
+      "step": 5533
+    },
+    {
+      "epoch": 1.065331953702144,
+      "grad_norm": 3.342135743280298,
+      "learning_rate": 9.426621659818388e-06,
+      "loss": 1.346,
+      "step": 5534
+    },
+    {
+      "epoch": 1.0655244603797194,
+      "grad_norm": 3.2890832655300835,
+      "learning_rate": 9.42350892055987e-06,
+      "loss": 1.3569,
+      "step": 5535
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "grad_norm": 3.2046026266689958,
+      "learning_rate": 9.420396237343654e-06,
+      "loss": 1.3518,
+      "step": 5536
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.172,
+      "step": 5536,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.0842,
+      "step": 5536,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.028,
+      "step": 5536,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.0291,
+      "step": 5536,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.2232,
+      "step": 5536,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.5943,
+      "step": 5536,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.6264,
+      "step": 5536,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 1.0657169670572948,
+      "lm_loss": 1.5356,
+      "step": 5536,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 1.0659094737348702,
+      "grad_norm": 3.2119574975529157,
+      "learning_rate": 9.417283610472345e-06,
+      "loss": 1.3855,
+      "step": 5537
+    },
+    {
+      "epoch": 1.0661019804124456,
+      "grad_norm": 3.209459213728874,
+      "learning_rate": 9.414171040248523e-06,
+      "loss": 1.4334,
+      "step": 5538
+    },
+    {
+      "epoch": 1.0662944870900208,
+      "grad_norm": 3.2173801614183684,
+      "learning_rate": 9.41105852697477e-06,
+      "loss": 1.4036,
+      "step": 5539
+    },
+    {
+      "epoch": 1.0664869937675963,
+      "grad_norm": 3.0550331138758278,
+      "learning_rate": 9.407946070953662e-06,
+      "loss": 1.323,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0666795004451717,
+      "grad_norm": 3.356867240591646,
+      "learning_rate": 9.404833672487772e-06,
+      "loss": 1.3927,
+      "step": 5541
+    },
+    {
+      "epoch": 1.066872007122747,
+      "grad_norm": 3.3540095447689113,
+      "learning_rate": 9.401721331879665e-06,
+      "loss": 1.4108,
+      "step": 5542
+    },
+    {
+      "epoch": 1.0670645138003225,
+      "grad_norm": 3.2693585638277662,
+      "learning_rate": 9.398609049431894e-06,
+      "loss": 1.3606,
+      "step": 5543
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "grad_norm": 3.257143753258995,
+      "learning_rate": 9.395496825447022e-06,
+      "loss": 1.383,
+      "step": 5544
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 0.7556,
+      "step": 5544,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 0.8079,
+      "step": 5544,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 1.2367,
+      "step": 5544,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 1.1791,
+      "step": 5544,
+      "vm_loss": 0.2135
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 1.3971,
+      "step": 5544,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 1.512,
+      "step": 5544,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 1.8917,
+      "step": 5544,
+      "vm_loss": 0.1209
+    },
+    {
+      "epoch": 1.0672570204778977,
+      "lm_loss": 0.9322,
+      "step": 5544,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 1.0674495271554731,
+      "grad_norm": 3.2613928042766887,
+      "learning_rate": 9.392384660227591e-06,
+      "loss": 1.3759,
+      "step": 5545
+    },
+    {
+      "epoch": 1.0676420338330486,
+      "grad_norm": 3.1389333600898923,
+      "learning_rate": 9.389272554076142e-06,
+      "loss": 1.3723,
+      "step": 5546
+    },
+    {
+      "epoch": 1.067834540510624,
+      "grad_norm": 3.126299361772461,
+      "learning_rate": 9.386160507295217e-06,
+      "loss": 1.3228,
+      "step": 5547
+    },
+    {
+      "epoch": 1.0680270471881994,
+      "grad_norm": 3.2267531044216513,
+      "learning_rate": 9.383048520187344e-06,
+      "loss": 1.3486,
+      "step": 5548
+    },
+    {
+      "epoch": 1.0682195538657746,
+      "grad_norm": 3.1449322864087357,
+      "learning_rate": 9.379936593055048e-06,
+      "loss": 1.3331,
+      "step": 5549
+    },
+    {
+      "epoch": 1.06841206054335,
+      "grad_norm": 3.22693186688896,
+      "learning_rate": 9.376824726200847e-06,
+      "loss": 1.3342,
+      "step": 5550
+    },
+    {
+      "epoch": 1.0686045672209254,
+      "grad_norm": 3.3067000824509503,
+      "learning_rate": 9.373712919927255e-06,
+      "loss": 1.3576,
+      "step": 5551
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "grad_norm": 3.4724923628475,
+      "learning_rate": 9.37060117453678e-06,
+      "loss": 1.433,
+      "step": 5552
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 0.9841,
+      "step": 5552,
+      "vm_loss": 0.1145
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.5305,
+      "step": 5552,
+      "vm_loss": 0.0796
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.6985,
+      "step": 5552,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.5509,
+      "step": 5552,
+      "vm_loss": 0.1576
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 0.9366,
+      "step": 5552,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.1486,
+      "step": 5552,
+      "vm_loss": 0.2215
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.841,
+      "step": 5552,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.0687970738985009,
+      "lm_loss": 1.1481,
+      "step": 5552,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 1.0689895805760763,
+      "grad_norm": 3.24294221239512,
+      "learning_rate": 9.367489490331923e-06,
+      "loss": 1.3523,
+      "step": 5553
+    },
+    {
+      "epoch": 1.0691820872536517,
+      "grad_norm": 3.132357727758026,
+      "learning_rate": 9.36437786761518e-06,
+      "loss": 1.3414,
+      "step": 5554
+    },
+    {
+      "epoch": 1.069374593931227,
+      "grad_norm": 3.2264972217198933,
+      "learning_rate": 9.361266306689043e-06,
+      "loss": 1.3088,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0695671006088023,
+      "grad_norm": 3.1989211828158988,
+      "learning_rate": 9.358154807855992e-06,
+      "loss": 1.3214,
+      "step": 5556
+    },
+    {
+      "epoch": 1.0697596072863778,
+      "grad_norm": 3.248003844905881,
+      "learning_rate": 9.355043371418501e-06,
+      "loss": 1.3495,
+      "step": 5557
+    },
+    {
+      "epoch": 1.0699521139639532,
+      "grad_norm": 3.29257100010339,
+      "learning_rate": 9.351931997679049e-06,
+      "loss": 1.3572,
+      "step": 5558
+    },
+    {
+      "epoch": 1.0701446206415286,
+      "grad_norm": 3.273101648818866,
+      "learning_rate": 9.348820686940099e-06,
+      "loss": 1.3609,
+      "step": 5559
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "grad_norm": 3.3311418105670203,
+      "learning_rate": 9.345709439504104e-06,
+      "loss": 1.3484,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 1.2996,
+      "step": 5560,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 1.2805,
+      "step": 5560,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 0.7554,
+      "step": 5560,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 0.9952,
+      "step": 5560,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 1.5369,
+      "step": 5560,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 1.3448,
+      "step": 5560,
+      "vm_loss": 0.1013
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 1.0685,
+      "step": 5560,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 1.0703371273191038,
+      "lm_loss": 0.8945,
+      "step": 5560,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.0705296339966792,
+      "grad_norm": 3.2508346065296707,
+      "learning_rate": 9.342598255673528e-06,
+      "loss": 1.3502,
+      "step": 5561
+    },
+    {
+      "epoch": 1.0707221406742546,
+      "grad_norm": 3.2695752230137325,
+      "learning_rate": 9.339487135750811e-06,
+      "loss": 1.3456,
+      "step": 5562
+    },
+    {
+      "epoch": 1.07091464735183,
+      "grad_norm": 3.1136596016957325,
+      "learning_rate": 9.336376080038397e-06,
+      "loss": 1.2774,
+      "step": 5563
+    },
+    {
+      "epoch": 1.0711071540294055,
+      "grad_norm": 3.2071240217875294,
+      "learning_rate": 9.333265088838713e-06,
+      "loss": 1.3178,
+      "step": 5564
+    },
+    {
+      "epoch": 1.0712996607069807,
+      "grad_norm": 3.325872811973558,
+      "learning_rate": 9.330154162454194e-06,
+      "loss": 1.3159,
+      "step": 5565
+    },
+    {
+      "epoch": 1.071492167384556,
+      "grad_norm": 3.3147133497472145,
+      "learning_rate": 9.327043301187264e-06,
+      "loss": 1.4158,
+      "step": 5566
+    },
+    {
+      "epoch": 1.0716846740621315,
+      "grad_norm": 3.1779783457830755,
+      "learning_rate": 9.323932505340327e-06,
+      "loss": 1.356,
+      "step": 5567
+    },
+    {
+      "epoch": 1.071877180739707,
+      "grad_norm": 3.24784563060547,
+      "learning_rate": 9.320821775215807e-06,
+      "loss": 1.2972,
+      "step": 5568
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.3385,
+      "step": 5568,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 0.9812,
+      "step": 5568,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.2648,
+      "step": 5568,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.035,
+      "step": 5568,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.7109,
+      "step": 5568,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.3619,
+      "step": 5568,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 1.6522,
+      "step": 5568,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.071877180739707,
+      "lm_loss": 0.805,
+      "step": 5568,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 1.0720696874172824,
+      "grad_norm": 3.47390906500689,
+      "learning_rate": 9.317711111116098e-06,
+      "loss": 1.4401,
+      "step": 5569
+    },
+    {
+      "epoch": 1.0722621940948576,
+      "grad_norm": 3.0915878509853334,
+      "learning_rate": 9.314600513343595e-06,
+      "loss": 1.3178,
+      "step": 5570
+    },
+    {
+      "epoch": 1.072454700772433,
+      "grad_norm": 3.1851176812917505,
+      "learning_rate": 9.311489982200695e-06,
+      "loss": 1.3234,
+      "step": 5571
+    },
+    {
+      "epoch": 1.0726472074500084,
+      "grad_norm": 3.2647896275943356,
+      "learning_rate": 9.308379517989775e-06,
+      "loss": 1.3934,
+      "step": 5572
+    },
+    {
+      "epoch": 1.0728397141275838,
+      "grad_norm": 3.1518661523441134,
+      "learning_rate": 9.305269121013214e-06,
+      "loss": 1.3106,
+      "step": 5573
+    },
+    {
+      "epoch": 1.0730322208051593,
+      "grad_norm": 3.1579274267979756,
+      "learning_rate": 9.302158791573384e-06,
+      "loss": 1.308,
+      "step": 5574
+    },
+    {
+      "epoch": 1.0732247274827347,
+      "grad_norm": 3.294144867619082,
+      "learning_rate": 9.299048529972647e-06,
+      "loss": 1.3786,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "grad_norm": 3.2827801171325333,
+      "learning_rate": 9.295938336513362e-06,
+      "loss": 1.4,
+      "step": 5576
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.0909,
+      "step": 5576,
+      "vm_loss": 0.2028
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 0.8172,
+      "step": 5576,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.3225,
+      "step": 5576,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.0128,
+      "step": 5576,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.446,
+      "step": 5576,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 0.8409,
+      "step": 5576,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.2535,
+      "step": 5576,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.0734172341603099,
+      "lm_loss": 1.2566,
+      "step": 5576,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.0736097408378853,
+      "grad_norm": 3.3056015475490987,
+      "learning_rate": 9.292828211497873e-06,
+      "loss": 1.4011,
+      "step": 5577
+    },
+    {
+      "epoch": 1.0738022475154607,
+      "grad_norm": 3.191573499434425,
+      "learning_rate": 9.289718155228535e-06,
+      "loss": 1.345,
+      "step": 5578
+    },
+    {
+      "epoch": 1.0739947541930361,
+      "grad_norm": 3.2581016160008955,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 1.3927,
+      "step": 5579
+    },
+    {
+      "epoch": 1.0741872608706116,
+      "grad_norm": 3.1994116204612206,
+      "learning_rate": 9.283498250137636e-06,
+      "loss": 1.3218,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0743797675481868,
+      "grad_norm": 3.151571378839607,
+      "learning_rate": 9.280388401920724e-06,
+      "loss": 1.3232,
+      "step": 5581
+    },
+    {
+      "epoch": 1.0745722742257622,
+      "grad_norm": 3.165131733201607,
+      "learning_rate": 9.277278623659273e-06,
+      "loss": 1.3694,
+      "step": 5582
+    },
+    {
+      "epoch": 1.0747647809033376,
+      "grad_norm": 3.320884928187641,
+      "learning_rate": 9.274168915655583e-06,
+      "loss": 1.3723,
+      "step": 5583
+    },
+    {
+      "epoch": 1.074957287580913,
+      "grad_norm": 3.1834172221634387,
+      "learning_rate": 9.271059278211959e-06,
+      "loss": 1.3614,
+      "step": 5584
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.196,
+      "step": 5584,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.4323,
+      "step": 5584,
+      "vm_loss": 0.1916
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.7479,
+      "step": 5584,
+      "vm_loss": 0.185
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 0.8278,
+      "step": 5584,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.8402,
+      "step": 5584,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.0791,
+      "step": 5584,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.6947,
+      "step": 5584,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 1.074957287580913,
+      "lm_loss": 1.214,
+      "step": 5584,
+      "vm_loss": 0.1998
+    },
+    {
+      "epoch": 1.0751497942584884,
+      "grad_norm": 3.1486265201118853,
+      "learning_rate": 9.267949711630701e-06,
+      "loss": 1.3818,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0753423009360636,
+      "grad_norm": 3.366567619134176,
+      "learning_rate": 9.264840216214099e-06,
+      "loss": 1.3255,
+      "step": 5586
+    },
+    {
+      "epoch": 1.075534807613639,
+      "grad_norm": 3.125600940091369,
+      "learning_rate": 9.261730792264429e-06,
+      "loss": 1.2776,
+      "step": 5587
+    },
+    {
+      "epoch": 1.0757273142912145,
+      "grad_norm": 3.0778228260380596,
+      "learning_rate": 9.258621440083973e-06,
+      "loss": 1.3043,
+      "step": 5588
+    },
+    {
+      "epoch": 1.07591982096879,
+      "grad_norm": 3.128962069772777,
+      "learning_rate": 9.255512159975e-06,
+      "loss": 1.3145,
+      "step": 5589
+    },
+    {
+      "epoch": 1.0761123276463653,
+      "grad_norm": 3.201602437326969,
+      "learning_rate": 9.252402952239767e-06,
+      "loss": 1.3456,
+      "step": 5590
+    },
+    {
+      "epoch": 1.0763048343239405,
+      "grad_norm": 3.1683037764909714,
+      "learning_rate": 9.249293817180532e-06,
+      "loss": 1.3172,
+      "step": 5591
+    },
+    {
+      "epoch": 1.076497341001516,
+      "grad_norm": 3.334446013280881,
+      "learning_rate": 9.246184755099545e-06,
+      "loss": 1.4251,
+      "step": 5592
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.3221,
+      "step": 5592,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.1799,
+      "step": 5592,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.3996,
+      "step": 5592,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 0.9323,
+      "step": 5592,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.3081,
+      "step": 5592,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.3987,
+      "step": 5592,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 0.9182,
+      "step": 5592,
+      "vm_loss": 0.1243
+    },
+    {
+      "epoch": 1.076497341001516,
+      "lm_loss": 1.0443,
+      "step": 5592,
+      "vm_loss": 0.1355
+    },
+    {
+      "epoch": 1.0766898476790914,
+      "grad_norm": 3.0388159730339903,
+      "learning_rate": 9.243075766299042e-06,
+      "loss": 1.314,
+      "step": 5593
+    },
+    {
+      "epoch": 1.0768823543566668,
+      "grad_norm": 3.1361715749433436,
+      "learning_rate": 9.239966851081256e-06,
+      "loss": 1.3035,
+      "step": 5594
+    },
+    {
+      "epoch": 1.0770748610342422,
+      "grad_norm": 3.310299147627332,
+      "learning_rate": 9.236858009748417e-06,
+      "loss": 1.351,
+      "step": 5595
+    },
+    {
+      "epoch": 1.0772673677118174,
+      "grad_norm": 3.142324914262676,
+      "learning_rate": 9.233749242602745e-06,
+      "loss": 1.3388,
+      "step": 5596
+    },
+    {
+      "epoch": 1.0774598743893928,
+      "grad_norm": 3.1156432864499126,
+      "learning_rate": 9.23064054994645e-06,
+      "loss": 1.3499,
+      "step": 5597
+    },
+    {
+      "epoch": 1.0776523810669683,
+      "grad_norm": 3.2268799191935313,
+      "learning_rate": 9.227531932081732e-06,
+      "loss": 1.3481,
+      "step": 5598
+    },
+    {
+      "epoch": 1.0778448877445437,
+      "grad_norm": 3.293693605554753,
+      "learning_rate": 9.224423389310799e-06,
+      "loss": 1.2627,
+      "step": 5599
+    },
+    {
+      "epoch": 1.078037394422119,
+      "grad_norm": 3.20375506598723,
+      "learning_rate": 9.221314921935833e-06,
+      "loss": 1.3245,
+      "step": 5600
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.5035,
+      "step": 5600,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.3449,
+      "step": 5600,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.1092,
+      "step": 5600,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.2073,
+      "step": 5600,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 0.8621,
+      "step": 5600,
+      "vm_loss": 0.1113
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.3378,
+      "step": 5600,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.4005,
+      "step": 5600,
+      "vm_loss": 0.1096
+    },
+    {
+      "epoch": 1.078037394422119,
+      "lm_loss": 1.3804,
+      "step": 5600,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.0782299010996943,
+      "grad_norm": 3.4237851172723883,
+      "learning_rate": 9.218206530259016e-06,
+      "loss": 1.4015,
+      "step": 5601
+    },
+    {
+      "epoch": 1.0784224077772697,
+      "grad_norm": 3.220122858294109,
+      "learning_rate": 9.21509821458253e-06,
+      "loss": 1.3112,
+      "step": 5602
+    },
+    {
+      "epoch": 1.0786149144548451,
+      "grad_norm": 3.2244602976168664,
+      "learning_rate": 9.211989975208538e-06,
+      "loss": 1.3542,
+      "step": 5603
+    },
+    {
+      "epoch": 1.0788074211324206,
+      "grad_norm": 3.211157743767312,
+      "learning_rate": 9.208881812439199e-06,
+      "loss": 1.4079,
+      "step": 5604
+    },
+    {
+      "epoch": 1.078999927809996,
+      "grad_norm": 3.0995121417458336,
+      "learning_rate": 9.205773726576673e-06,
+      "loss": 1.3484,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0791924344875712,
+      "grad_norm": 3.4115777887664946,
+      "learning_rate": 9.202665717923103e-06,
+      "loss": 1.3965,
+      "step": 5606
+    },
+    {
+      "epoch": 1.0793849411651466,
+      "grad_norm": 3.1127732275476028,
+      "learning_rate": 9.199557786780626e-06,
+      "loss": 1.2936,
+      "step": 5607
+    },
+    {
+      "epoch": 1.079577447842722,
+      "grad_norm": 3.116278383862214,
+      "learning_rate": 9.196449933451366e-06,
+      "loss": 1.3132,
+      "step": 5608
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 1.3763,
+      "step": 5608,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 0.8764,
+      "step": 5608,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 0.8006,
+      "step": 5608,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 1.103,
+      "step": 5608,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 1.3528,
+      "step": 5608,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 0.9172,
+      "step": 5608,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 1.5046,
+      "step": 5608,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 1.079577447842722,
+      "lm_loss": 0.5672,
+      "step": 5608,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.0797699545202974,
+      "grad_norm": 3.1853642651186154,
+      "learning_rate": 9.193342158237461e-06,
+      "loss": 1.3346,
+      "step": 5609
+    },
+    {
+      "epoch": 1.0799624611978729,
+      "grad_norm": 3.2617039931768024,
+      "learning_rate": 9.190234461441015e-06,
+      "loss": 1.3073,
+      "step": 5610
+    },
+    {
+      "epoch": 1.080154967875448,
+      "grad_norm": 3.272132300401591,
+      "learning_rate": 9.187126843364138e-06,
+      "loss": 1.3388,
+      "step": 5611
+    },
+    {
+      "epoch": 1.0803474745530235,
+      "grad_norm": 3.410332008063213,
+      "learning_rate": 9.184019304308933e-06,
+      "loss": 1.369,
+      "step": 5612
+    },
+    {
+      "epoch": 1.080539981230599,
+      "grad_norm": 3.500265904636164,
+      "learning_rate": 9.180911844577492e-06,
+      "loss": 1.4171,
+      "step": 5613
+    },
+    {
+      "epoch": 1.0807324879081743,
+      "grad_norm": 3.1231867091659318,
+      "learning_rate": 9.177804464471897e-06,
+      "loss": 1.3271,
+      "step": 5614
+    },
+    {
+      "epoch": 1.0809249945857498,
+      "grad_norm": 3.1650242742528083,
+      "learning_rate": 9.174697164294228e-06,
+      "loss": 1.3551,
+      "step": 5615
+    },
+    {
+      "epoch": 1.081117501263325,
+      "grad_norm": 3.1344394035953944,
+      "learning_rate": 9.171589944346552e-06,
+      "loss": 1.3229,
+      "step": 5616
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.1695,
+      "step": 5616,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.6816,
+      "step": 5616,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.4095,
+      "step": 5616,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.3409,
+      "step": 5616,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.3467,
+      "step": 5616,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 0.7703,
+      "step": 5616,
+      "vm_loss": 0.1006
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 1.14,
+      "step": 5616,
+      "vm_loss": 0.089
+    },
+    {
+      "epoch": 1.081117501263325,
+      "lm_loss": 0.7698,
+      "step": 5616,
+      "vm_loss": 0.2383
+    },
+    {
+      "epoch": 1.0813100079409004,
+      "grad_norm": 3.2346473025052904,
+      "learning_rate": 9.168482804930935e-06,
+      "loss": 1.3788,
+      "step": 5617
+    },
+    {
+      "epoch": 1.0815025146184758,
+      "grad_norm": 3.2466149876167174,
+      "learning_rate": 9.165375746349424e-06,
+      "loss": 1.3781,
+      "step": 5618
+    },
+    {
+      "epoch": 1.0816950212960512,
+      "grad_norm": 3.2442343757848104,
+      "learning_rate": 9.16226876890407e-06,
+      "loss": 1.3038,
+      "step": 5619
+    },
+    {
+      "epoch": 1.0818875279736266,
+      "grad_norm": 3.188876304079787,
+      "learning_rate": 9.15916187289691e-06,
+      "loss": 1.2986,
+      "step": 5620
+    },
+    {
+      "epoch": 1.082080034651202,
+      "grad_norm": 3.4713123129512615,
+      "learning_rate": 9.15605505862997e-06,
+      "loss": 1.3816,
+      "step": 5621
+    },
+    {
+      "epoch": 1.0822725413287773,
+      "grad_norm": 3.160790565480621,
+      "learning_rate": 9.15294832640528e-06,
+      "loss": 1.2938,
+      "step": 5622
+    },
+    {
+      "epoch": 1.0824650480063527,
+      "grad_norm": 3.1861005575735044,
+      "learning_rate": 9.14984167652485e-06,
+      "loss": 1.3553,
+      "step": 5623
+    },
+    {
+      "epoch": 1.082657554683928,
+      "grad_norm": 3.242793708904605,
+      "learning_rate": 9.146735109290685e-06,
+      "loss": 1.3536,
+      "step": 5624
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 1.784,
+      "step": 5624,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 0.9895,
+      "step": 5624,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 1.4984,
+      "step": 5624,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 0.901,
+      "step": 5624,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 1.0796,
+      "step": 5624,
+      "vm_loss": 0.0945
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 0.9166,
+      "step": 5624,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 1.0005,
+      "step": 5624,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.082657554683928,
+      "lm_loss": 1.0859,
+      "step": 5624,
+      "vm_loss": 0.2272
+    },
+    {
+      "epoch": 1.0828500613615035,
+      "grad_norm": 3.2347946021845155,
+      "learning_rate": 9.14362862500478e-06,
+      "loss": 1.3152,
+      "step": 5625
+    },
+    {
+      "epoch": 1.083042568039079,
+      "grad_norm": 3.3486793454807176,
+      "learning_rate": 9.140522223969132e-06,
+      "loss": 1.3904,
+      "step": 5626
+    },
+    {
+      "epoch": 1.0832350747166541,
+      "grad_norm": 3.182503554964886,
+      "learning_rate": 9.13741590648572e-06,
+      "loss": 1.2623,
+      "step": 5627
+    },
+    {
+      "epoch": 1.0834275813942296,
+      "grad_norm": 3.238017989777943,
+      "learning_rate": 9.134309672856513e-06,
+      "loss": 1.3072,
+      "step": 5628
+    },
+    {
+      "epoch": 1.083620088071805,
+      "grad_norm": 3.216862328776496,
+      "learning_rate": 9.131203523383487e-06,
+      "loss": 1.3622,
+      "step": 5629
+    },
+    {
+      "epoch": 1.0838125947493804,
+      "grad_norm": 3.2114941720973147,
+      "learning_rate": 9.128097458368593e-06,
+      "loss": 1.3098,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0840051014269558,
+      "grad_norm": 3.1529689407249943,
+      "learning_rate": 9.12499147811378e-06,
+      "loss": 1.2471,
+      "step": 5631
+    },
+    {
+      "epoch": 1.084197608104531,
+      "grad_norm": 3.1505777818508,
+      "learning_rate": 9.121885582920987e-06,
+      "loss": 1.2947,
+      "step": 5632
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 0.8096,
+      "step": 5632,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 1.4151,
+      "step": 5632,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 1.5111,
+      "step": 5632,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 1.0143,
+      "step": 5632,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 1.0986,
+      "step": 5632,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 0.9617,
+      "step": 5632,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 0.9046,
+      "step": 5632,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.084197608104531,
+      "lm_loss": 1.2678,
+      "step": 5632,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.0843901147821065,
+      "grad_norm": 3.2413186979891457,
+      "learning_rate": 9.118779773092154e-06,
+      "loss": 1.2555,
+      "step": 5633
+    },
+    {
+      "epoch": 1.0845826214596819,
+      "grad_norm": 3.295569042665185,
+      "learning_rate": 9.115674048929199e-06,
+      "loss": 1.3234,
+      "step": 5634
+    },
+    {
+      "epoch": 1.0847751281372573,
+      "grad_norm": 3.426311384291926,
+      "learning_rate": 9.112568410734039e-06,
+      "loss": 1.3859,
+      "step": 5635
+    },
+    {
+      "epoch": 1.0849676348148327,
+      "grad_norm": 3.2377294893857496,
+      "learning_rate": 9.109462858808586e-06,
+      "loss": 1.2857,
+      "step": 5636
+    },
+    {
+      "epoch": 1.085160141492408,
+      "grad_norm": 3.242225588204388,
+      "learning_rate": 9.106357393454734e-06,
+      "loss": 1.3465,
+      "step": 5637
+    },
+    {
+      "epoch": 1.0853526481699833,
+      "grad_norm": 3.1525927900142503,
+      "learning_rate": 9.10325201497438e-06,
+      "loss": 1.3269,
+      "step": 5638
+    },
+    {
+      "epoch": 1.0855451548475588,
+      "grad_norm": 3.1941445231616745,
+      "learning_rate": 9.100146723669397e-06,
+      "loss": 1.3537,
+      "step": 5639
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "grad_norm": 3.1498306529255493,
+      "learning_rate": 9.09704151984167e-06,
+      "loss": 1.3055,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 0.7588,
+      "step": 5640,
+      "vm_loss": 0.1227
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 0.6631,
+      "step": 5640,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 1.1715,
+      "step": 5640,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 1.2777,
+      "step": 5640,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 1.0492,
+      "step": 5640,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 0.7675,
+      "step": 5640,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 0.9948,
+      "step": 5640,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.0857376615251342,
+      "lm_loss": 1.5775,
+      "step": 5640,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.0859301682027096,
+      "grad_norm": 3.146314131853108,
+      "learning_rate": 9.093936403793057e-06,
+      "loss": 1.2698,
+      "step": 5641
+    },
+    {
+      "epoch": 1.086122674880285,
+      "grad_norm": 3.1552392868556,
+      "learning_rate": 9.090831375825416e-06,
+      "loss": 1.2445,
+      "step": 5642
+    },
+    {
+      "epoch": 1.0863151815578602,
+      "grad_norm": 3.3170929025036187,
+      "learning_rate": 9.087726436240601e-06,
+      "loss": 1.316,
+      "step": 5643
+    },
+    {
+      "epoch": 1.0865076882354356,
+      "grad_norm": 3.3415596394813414,
+      "learning_rate": 9.084621585340448e-06,
+      "loss": 1.2846,
+      "step": 5644
+    },
+    {
+      "epoch": 1.086700194913011,
+      "grad_norm": 3.3089984096724074,
+      "learning_rate": 9.081516823426783e-06,
+      "loss": 1.2825,
+      "step": 5645
+    },
+    {
+      "epoch": 1.0868927015905865,
+      "grad_norm": 3.3521087164646284,
+      "learning_rate": 9.078412150801438e-06,
+      "loss": 1.3557,
+      "step": 5646
+    },
+    {
+      "epoch": 1.087085208268162,
+      "grad_norm": 3.2254473086601916,
+      "learning_rate": 9.075307567766222e-06,
+      "loss": 1.2782,
+      "step": 5647
+    },
+    {
+      "epoch": 1.087277714945737,
+      "grad_norm": 3.1440803048245423,
+      "learning_rate": 9.072203074622941e-06,
+      "loss": 1.2887,
+      "step": 5648
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 1.6449,
+      "step": 5648,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 1.3673,
+      "step": 5648,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 1.4146,
+      "step": 5648,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 0.9623,
+      "step": 5648,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 1.0349,
+      "step": 5648,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 0.9455,
+      "step": 5648,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 1.7619,
+      "step": 5648,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.087277714945737,
+      "lm_loss": 0.6367,
+      "step": 5648,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.0874702216233125,
+      "grad_norm": 3.2512703613371086,
+      "learning_rate": 9.069098671673387e-06,
+      "loss": 1.3273,
+      "step": 5649
+    },
+    {
+      "epoch": 1.087662728300888,
+      "grad_norm": 3.3341693502548595,
+      "learning_rate": 9.065994359219357e-06,
+      "loss": 1.3922,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0878552349784634,
+      "grad_norm": 3.2341100594876075,
+      "learning_rate": 9.062890137562623e-06,
+      "loss": 1.4104,
+      "step": 5651
+    },
+    {
+      "epoch": 1.0880477416560388,
+      "grad_norm": 3.148063103142531,
+      "learning_rate": 9.059786007004953e-06,
+      "loss": 1.249,
+      "step": 5652
+    },
+    {
+      "epoch": 1.088240248333614,
+      "grad_norm": 3.2259478266781065,
+      "learning_rate": 9.056681967848117e-06,
+      "loss": 1.3031,
+      "step": 5653
+    },
+    {
+      "epoch": 1.0884327550111894,
+      "grad_norm": 3.251268724052633,
+      "learning_rate": 9.053578020393862e-06,
+      "loss": 1.3143,
+      "step": 5654
+    },
+    {
+      "epoch": 1.0886252616887648,
+      "grad_norm": 3.3933464605283468,
+      "learning_rate": 9.050474164943928e-06,
+      "loss": 1.347,
+      "step": 5655
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "grad_norm": 3.2158435104353646,
+      "learning_rate": 9.047370401800055e-06,
+      "loss": 1.2745,
+      "step": 5656
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 0.8327,
+      "step": 5656,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.2655,
+      "step": 5656,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.2796,
+      "step": 5656,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.2851,
+      "step": 5656,
+      "vm_loss": 0.1352
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.1139,
+      "step": 5656,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.0567,
+      "step": 5656,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.0972,
+      "step": 5656,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 1.0888177683663403,
+      "lm_loss": 1.3835,
+      "step": 5656,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.0890102750439157,
+      "grad_norm": 3.365445421169781,
+      "learning_rate": 9.044266731263966e-06,
+      "loss": 1.358,
+      "step": 5657
+    },
+    {
+      "epoch": 1.0892027817214909,
+      "grad_norm": 3.1799178912596457,
+      "learning_rate": 9.041163153637382e-06,
+      "loss": 1.2963,
+      "step": 5658
+    },
+    {
+      "epoch": 1.0893952883990663,
+      "grad_norm": 3.2056908790131966,
+      "learning_rate": 9.038059669222e-06,
+      "loss": 1.2634,
+      "step": 5659
+    },
+    {
+      "epoch": 1.0895877950766417,
+      "grad_norm": 3.144224581238796,
+      "learning_rate": 9.03495627831953e-06,
+      "loss": 1.3538,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0897803017542171,
+      "grad_norm": 3.268945702475398,
+      "learning_rate": 9.031852981231656e-06,
+      "loss": 1.3097,
+      "step": 5661
+    },
+    {
+      "epoch": 1.0899728084317926,
+      "grad_norm": 3.283374190320138,
+      "learning_rate": 9.028749778260055e-06,
+      "loss": 1.4262,
+      "step": 5662
+    },
+    {
+      "epoch": 1.0901653151093678,
+      "grad_norm": 3.2891409008783263,
+      "learning_rate": 9.025646669706405e-06,
+      "loss": 1.3216,
+      "step": 5663
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "grad_norm": 3.3716704336312477,
+      "learning_rate": 9.022543655872363e-06,
+      "loss": 1.3586,
+      "step": 5664
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 1.0418,
+      "step": 5664,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 0.9155,
+      "step": 5664,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 1.4843,
+      "step": 5664,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 1.0664,
+      "step": 5664,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 0.6404,
+      "step": 5664,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 0.9507,
+      "step": 5664,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 1.1979,
+      "step": 5664,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 1.0903578217869432,
+      "lm_loss": 0.7537,
+      "step": 5664,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.0905503284645186,
+      "grad_norm": 3.2136032971325723,
+      "learning_rate": 9.019440737059585e-06,
+      "loss": 1.2997,
+      "step": 5665
+    },
+    {
+      "epoch": 1.090742835142094,
+      "grad_norm": 3.1334220885118995,
+      "learning_rate": 9.016337913569708e-06,
+      "loss": 1.2393,
+      "step": 5666
+    },
+    {
+      "epoch": 1.0909353418196694,
+      "grad_norm": 3.1517433062241,
+      "learning_rate": 9.013235185704374e-06,
+      "loss": 1.295,
+      "step": 5667
+    },
+    {
+      "epoch": 1.0911278484972446,
+      "grad_norm": 3.3172375820447426,
+      "learning_rate": 9.010132553765204e-06,
+      "loss": 1.2911,
+      "step": 5668
+    },
+    {
+      "epoch": 1.09132035517482,
+      "grad_norm": 3.272666488681282,
+      "learning_rate": 9.00703001805381e-06,
+      "loss": 1.313,
+      "step": 5669
+    },
+    {
+      "epoch": 1.0915128618523955,
+      "grad_norm": 3.16269257199317,
+      "learning_rate": 9.003927578871807e-06,
+      "loss": 1.3056,
+      "step": 5670
+    },
+    {
+      "epoch": 1.091705368529971,
+      "grad_norm": 3.1779747990065084,
+      "learning_rate": 9.000825236520786e-06,
+      "loss": 1.2796,
+      "step": 5671
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "grad_norm": 3.0754888218483245,
+      "learning_rate": 8.997722991302335e-06,
+      "loss": 1.3469,
+      "step": 5672
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 1.1048,
+      "step": 5672,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 1.3984,
+      "step": 5672,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 0.5434,
+      "step": 5672,
+      "vm_loss": 0.104
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 1.3016,
+      "step": 5672,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 0.7313,
+      "step": 5672,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 1.0544,
+      "step": 5672,
+      "vm_loss": 0.2133
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 1.3284,
+      "step": 5672,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.0918978752075463,
+      "lm_loss": 0.7356,
+      "step": 5672,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 1.0920903818851215,
+      "grad_norm": 3.1328898034537827,
+      "learning_rate": 8.99462084351803e-06,
+      "loss": 1.3003,
+      "step": 5673
+    },
+    {
+      "epoch": 1.092282888562697,
+      "grad_norm": 3.146352768171252,
+      "learning_rate": 8.991518793469444e-06,
+      "loss": 1.2844,
+      "step": 5674
+    },
+    {
+      "epoch": 1.0924753952402724,
+      "grad_norm": 3.2329772737455884,
+      "learning_rate": 8.988416841458131e-06,
+      "loss": 1.3251,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0926679019178478,
+      "grad_norm": 3.2238944957365754,
+      "learning_rate": 8.985314987785644e-06,
+      "loss": 1.3392,
+      "step": 5676
+    },
+    {
+      "epoch": 1.0928604085954232,
+      "grad_norm": 3.23393808417215,
+      "learning_rate": 8.982213232753523e-06,
+      "loss": 1.3148,
+      "step": 5677
+    },
+    {
+      "epoch": 1.0930529152729984,
+      "grad_norm": 3.3104508630892138,
+      "learning_rate": 8.979111576663295e-06,
+      "loss": 1.3799,
+      "step": 5678
+    },
+    {
+      "epoch": 1.0932454219505738,
+      "grad_norm": 3.475890470701071,
+      "learning_rate": 8.976010019816484e-06,
+      "loss": 1.3488,
+      "step": 5679
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "grad_norm": 3.192017942227653,
+      "learning_rate": 8.9729085625146e-06,
+      "loss": 1.3173,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.7857,
+      "step": 5680,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.9222,
+      "step": 5680,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.6439,
+      "step": 5680,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 0.9772,
+      "step": 5680,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.1941,
+      "step": 5680,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.1633,
+      "step": 5680,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 1.1305,
+      "step": 5680,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.0934379286281493,
+      "lm_loss": 0.9578,
+      "step": 5680,
+      "vm_loss": 0.2459
+    },
+    {
+      "epoch": 1.0936304353057247,
+      "grad_norm": 3.3268726040842975,
+      "learning_rate": 8.969807205059144e-06,
+      "loss": 1.3326,
+      "step": 5681
+    },
+    {
+      "epoch": 1.0938229419833,
+      "grad_norm": 3.3294333761238786,
+      "learning_rate": 8.966705947751608e-06,
+      "loss": 1.3391,
+      "step": 5682
+    },
+    {
+      "epoch": 1.0940154486608755,
+      "grad_norm": 3.1628002566022633,
+      "learning_rate": 8.96360479089347e-06,
+      "loss": 1.3703,
+      "step": 5683
+    },
+    {
+      "epoch": 1.0942079553384507,
+      "grad_norm": 3.2991841413351732,
+      "learning_rate": 8.96050373478621e-06,
+      "loss": 1.3413,
+      "step": 5684
+    },
+    {
+      "epoch": 1.0944004620160261,
+      "grad_norm": 3.1891290278593973,
+      "learning_rate": 8.957402779731288e-06,
+      "loss": 1.2926,
+      "step": 5685
+    },
+    {
+      "epoch": 1.0945929686936016,
+      "grad_norm": 3.1823080045827115,
+      "learning_rate": 8.954301926030146e-06,
+      "loss": 1.2862,
+      "step": 5686
+    },
+    {
+      "epoch": 1.094785475371177,
+      "grad_norm": 3.20346621057419,
+      "learning_rate": 8.951201173984243e-06,
+      "loss": 1.325,
+      "step": 5687
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "grad_norm": 3.2578293288530737,
+      "learning_rate": 8.948100523895002e-06,
+      "loss": 1.3479,
+      "step": 5688
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.4778,
+      "step": 5688,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.2696,
+      "step": 5688,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.0023,
+      "step": 5688,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.3103,
+      "step": 5688,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.2342,
+      "step": 5688,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.3477,
+      "step": 5688,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.4247,
+      "step": 5688,
+      "vm_loss": 0.1877
+    },
+    {
+      "epoch": 1.0949779820487524,
+      "lm_loss": 1.0773,
+      "step": 5688,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 1.0951704887263276,
+      "grad_norm": 3.2956872971471607,
+      "learning_rate": 8.944999976063848e-06,
+      "loss": 1.3785,
+      "step": 5689
+    },
+    {
+      "epoch": 1.095362995403903,
+      "grad_norm": 3.1829411772902527,
+      "learning_rate": 8.94189953079219e-06,
+      "loss": 1.2844,
+      "step": 5690
+    },
+    {
+      "epoch": 1.0955555020814784,
+      "grad_norm": 3.2735102528095523,
+      "learning_rate": 8.938799188381439e-06,
+      "loss": 1.3904,
+      "step": 5691
+    },
+    {
+      "epoch": 1.0957480087590539,
+      "grad_norm": 3.1418807431825573,
+      "learning_rate": 8.935698949132983e-06,
+      "loss": 1.3297,
+      "step": 5692
+    },
+    {
+      "epoch": 1.0959405154366293,
+      "grad_norm": 3.233228086065883,
+      "learning_rate": 8.932598813348203e-06,
+      "loss": 1.3426,
+      "step": 5693
+    },
+    {
+      "epoch": 1.0961330221142045,
+      "grad_norm": 3.074023225740406,
+      "learning_rate": 8.929498781328477e-06,
+      "loss": 1.2919,
+      "step": 5694
+    },
+    {
+      "epoch": 1.09632552879178,
+      "grad_norm": 3.307272176178498,
+      "learning_rate": 8.926398853375165e-06,
+      "loss": 1.3589,
+      "step": 5695
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "grad_norm": 3.2056346120917243,
+      "learning_rate": 8.923299029789617e-06,
+      "loss": 1.2958,
+      "step": 5696
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.0955,
+      "step": 5696,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.1629,
+      "step": 5696,
+      "vm_loss": 0.1113
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.9042,
+      "step": 5696,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.1895,
+      "step": 5696,
+      "vm_loss": 0.2194
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 0.9745,
+      "step": 5696,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.536,
+      "step": 5696,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 0.7094,
+      "step": 5696,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 1.0965180354693553,
+      "lm_loss": 1.4681,
+      "step": 5696,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 1.0967105421469308,
+      "grad_norm": 3.2061990960286058,
+      "learning_rate": 8.92019931087318e-06,
+      "loss": 1.3578,
+      "step": 5697
+    },
+    {
+      "epoch": 1.0969030488245062,
+      "grad_norm": 3.2176623245472333,
+      "learning_rate": 8.917099696927188e-06,
+      "loss": 1.3112,
+      "step": 5698
+    },
+    {
+      "epoch": 1.0970955555020814,
+      "grad_norm": 3.389623603965464,
+      "learning_rate": 8.914000188252955e-06,
+      "loss": 1.3546,
+      "step": 5699
+    },
+    {
+      "epoch": 1.0972880621796568,
+      "grad_norm": 3.347557446122881,
+      "learning_rate": 8.9109007851518e-06,
+      "loss": 1.2894,
+      "step": 5700
+    },
+    {
+      "epoch": 1.0974805688572322,
+      "grad_norm": 3.434783087036281,
+      "learning_rate": 8.90780148792502e-06,
+      "loss": 1.3188,
+      "step": 5701
+    },
+    {
+      "epoch": 1.0976730755348076,
+      "grad_norm": 3.313694495487245,
+      "learning_rate": 8.904702296873913e-06,
+      "loss": 1.3537,
+      "step": 5702
+    },
+    {
+      "epoch": 1.097865582212383,
+      "grad_norm": 3.3140814474179985,
+      "learning_rate": 8.901603212299751e-06,
+      "loss": 1.4102,
+      "step": 5703
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "grad_norm": 3.209801159029106,
+      "learning_rate": 8.898504234503811e-06,
+      "loss": 1.2962,
+      "step": 5704
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 1.3149,
+      "step": 5704,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 1.516,
+      "step": 5704,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 0.5353,
+      "step": 5704,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 1.3332,
+      "step": 5704,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 0.8764,
+      "step": 5704,
+      "vm_loss": 0.1747
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 0.833,
+      "step": 5704,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 1.2967,
+      "step": 5704,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.0980580888899585,
+      "lm_loss": 0.9802,
+      "step": 5704,
+      "vm_loss": 0.1277
+    },
+    {
+      "epoch": 1.0982505955675337,
+      "grad_norm": 3.194259831195042,
+      "learning_rate": 8.895405363787356e-06,
+      "loss": 1.3314,
+      "step": 5705
+    },
+    {
+      "epoch": 1.098443102245109,
+      "grad_norm": 3.198044851177277,
+      "learning_rate": 8.892306600451628e-06,
+      "loss": 1.2774,
+      "step": 5706
+    },
+    {
+      "epoch": 1.0986356089226845,
+      "grad_norm": 3.2487984183833034,
+      "learning_rate": 8.889207944797866e-06,
+      "loss": 1.3128,
+      "step": 5707
+    },
+    {
+      "epoch": 1.09882811560026,
+      "grad_norm": 3.2111271010134255,
+      "learning_rate": 8.886109397127309e-06,
+      "loss": 1.2663,
+      "step": 5708
+    },
+    {
+      "epoch": 1.0990206222778354,
+      "grad_norm": 3.299987832538869,
+      "learning_rate": 8.883010957741168e-06,
+      "loss": 1.2782,
+      "step": 5709
+    },
+    {
+      "epoch": 1.0992131289554106,
+      "grad_norm": 3.346647907029083,
+      "learning_rate": 8.879912626940647e-06,
+      "loss": 1.2763,
+      "step": 5710
+    },
+    {
+      "epoch": 1.099405635632986,
+      "grad_norm": 3.329141543332642,
+      "learning_rate": 8.876814405026955e-06,
+      "loss": 1.2916,
+      "step": 5711
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "grad_norm": 3.335237301203536,
+      "learning_rate": 8.873716292301273e-06,
+      "loss": 1.3337,
+      "step": 5712
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 0.9059,
+      "step": 5712,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 1.3644,
+      "step": 5712,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 1.0724,
+      "step": 5712,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 1.2328,
+      "step": 5712,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 0.9025,
+      "step": 5712,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 0.7712,
+      "step": 5712,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 0.9679,
+      "step": 5712,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 1.0995981423105614,
+      "lm_loss": 0.9716,
+      "step": 5712,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 1.0997906489881368,
+      "grad_norm": 3.3860530989647764,
+      "learning_rate": 8.870618289064776e-06,
+      "loss": 1.3245,
+      "step": 5713
+    },
+    {
+      "epoch": 1.0999831556657123,
+      "grad_norm": 3.116866692562132,
+      "learning_rate": 8.867520395618626e-06,
+      "loss": 1.2991,
+      "step": 5714
+    },
+    {
+      "epoch": 1.1001756623432875,
+      "grad_norm": 3.172194706063709,
+      "learning_rate": 8.864422612263986e-06,
+      "loss": 1.2737,
+      "step": 5715
+    },
+    {
+      "epoch": 1.1003681690208629,
+      "grad_norm": 3.275481186771253,
+      "learning_rate": 8.861324939301998e-06,
+      "loss": 1.2483,
+      "step": 5716
+    },
+    {
+      "epoch": 1.1005606756984383,
+      "grad_norm": 3.190524089796947,
+      "learning_rate": 8.858227377033787e-06,
+      "loss": 1.2717,
+      "step": 5717
+    },
+    {
+      "epoch": 1.1007531823760137,
+      "grad_norm": 3.2672227934924747,
+      "learning_rate": 8.855129925760488e-06,
+      "loss": 1.27,
+      "step": 5718
+    },
+    {
+      "epoch": 1.1009456890535891,
+      "grad_norm": 3.41309060278696,
+      "learning_rate": 8.852032585783206e-06,
+      "loss": 1.2894,
+      "step": 5719
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "grad_norm": 3.5509748306118847,
+      "learning_rate": 8.848935357403041e-06,
+      "loss": 1.3076,
+      "step": 5720
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.6671,
+      "step": 5720,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.0089,
+      "step": 5720,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 0.7442,
+      "step": 5720,
+      "vm_loss": 0.1025
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.1051,
+      "step": 5720,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.1958,
+      "step": 5720,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.062,
+      "step": 5720,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.1236,
+      "step": 5720,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.1011381957311643,
+      "lm_loss": 1.1452,
+      "step": 5720,
+      "vm_loss": 0.2024
+    },
+    {
+      "epoch": 1.1013307024087398,
+      "grad_norm": 3.2172550917825746,
+      "learning_rate": 8.845838240921091e-06,
+      "loss": 1.2263,
+      "step": 5721
+    },
+    {
+      "epoch": 1.1015232090863152,
+      "grad_norm": 3.2650362536324367,
+      "learning_rate": 8.842741236638427e-06,
+      "loss": 1.3122,
+      "step": 5722
+    },
+    {
+      "epoch": 1.1017157157638906,
+      "grad_norm": 3.3086706903377947,
+      "learning_rate": 8.839644344856122e-06,
+      "loss": 1.3099,
+      "step": 5723
+    },
+    {
+      "epoch": 1.101908222441466,
+      "grad_norm": 3.0866868531425062,
+      "learning_rate": 8.836547565875227e-06,
+      "loss": 1.2873,
+      "step": 5724
+    },
+    {
+      "epoch": 1.1021007291190412,
+      "grad_norm": 3.2058832697428983,
+      "learning_rate": 8.833450899996799e-06,
+      "loss": 1.259,
+      "step": 5725
+    },
+    {
+      "epoch": 1.1022932357966166,
+      "grad_norm": 3.0449475786265503,
+      "learning_rate": 8.830354347521867e-06,
+      "loss": 1.3154,
+      "step": 5726
+    },
+    {
+      "epoch": 1.102485742474192,
+      "grad_norm": 3.186909057301959,
+      "learning_rate": 8.827257908751454e-06,
+      "loss": 1.2959,
+      "step": 5727
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "grad_norm": 3.179916271433507,
+      "learning_rate": 8.824161583986578e-06,
+      "loss": 1.263,
+      "step": 5728
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 1.407,
+      "step": 5728,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 0.9904,
+      "step": 5728,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 0.8983,
+      "step": 5728,
+      "vm_loss": 0.1021
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 0.8898,
+      "step": 5728,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 1.5359,
+      "step": 5728,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 1.2492,
+      "step": 5728,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 1.6421,
+      "step": 5728,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 1.1026782491517675,
+      "lm_loss": 0.7561,
+      "step": 5728,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.102870755829343,
+      "grad_norm": 3.3544099807973398,
+      "learning_rate": 8.821065373528242e-06,
+      "loss": 1.3186,
+      "step": 5729
+    },
+    {
+      "epoch": 1.103063262506918,
+      "grad_norm": 3.2155243484799474,
+      "learning_rate": 8.817969277677434e-06,
+      "loss": 1.3368,
+      "step": 5730
+    },
+    {
+      "epoch": 1.1032557691844935,
+      "grad_norm": 3.2399010599616163,
+      "learning_rate": 8.814873296735129e-06,
+      "loss": 1.2513,
+      "step": 5731
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 3.411926884325892,
+      "learning_rate": 8.811777431002307e-06,
+      "loss": 1.3891,
+      "step": 5732
+    },
+    {
+      "epoch": 1.1036407825396444,
+      "grad_norm": 3.239426826739809,
+      "learning_rate": 8.80868168077992e-06,
+      "loss": 1.2687,
+      "step": 5733
+    },
+    {
+      "epoch": 1.1038332892172198,
+      "grad_norm": 3.2149869406082408,
+      "learning_rate": 8.80558604636891e-06,
+      "loss": 1.2915,
+      "step": 5734
+    },
+    {
+      "epoch": 1.104025795894795,
+      "grad_norm": 3.20867661109182,
+      "learning_rate": 8.802490528070223e-06,
+      "loss": 1.2929,
+      "step": 5735
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "grad_norm": 2.9938701382306245,
+      "learning_rate": 8.799395126184777e-06,
+      "loss": 1.266,
+      "step": 5736
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.2729,
+      "step": 5736,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.1825,
+      "step": 5736,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 0.9335,
+      "step": 5736,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.3149,
+      "step": 5736,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.2084,
+      "step": 5736,
+      "vm_loss": 0.2314
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.0026,
+      "step": 5736,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.2398,
+      "step": 5736,
+      "vm_loss": 0.1695
+    },
+    {
+      "epoch": 1.1042183025723704,
+      "lm_loss": 1.284,
+      "step": 5736,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.1044108092499458,
+      "grad_norm": 3.3294033524570925,
+      "learning_rate": 8.79629984101348e-06,
+      "loss": 1.3744,
+      "step": 5737
+    },
+    {
+      "epoch": 1.1046033159275213,
+      "grad_norm": 3.230884024334751,
+      "learning_rate": 8.793204672857244e-06,
+      "loss": 1.3228,
+      "step": 5738
+    },
+    {
+      "epoch": 1.1047958226050967,
+      "grad_norm": 3.1790858903114034,
+      "learning_rate": 8.790109622016953e-06,
+      "loss": 1.2911,
+      "step": 5739
+    },
+    {
+      "epoch": 1.1049883292826719,
+      "grad_norm": 3.1446978365129534,
+      "learning_rate": 8.787014688793483e-06,
+      "loss": 1.2558,
+      "step": 5740
+    },
+    {
+      "epoch": 1.1051808359602473,
+      "grad_norm": 3.3188546078909105,
+      "learning_rate": 8.783919873487703e-06,
+      "loss": 1.3439,
+      "step": 5741
+    },
+    {
+      "epoch": 1.1053733426378227,
+      "grad_norm": 3.178844859898877,
+      "learning_rate": 8.78082517640047e-06,
+      "loss": 1.2719,
+      "step": 5742
+    },
+    {
+      "epoch": 1.1055658493153981,
+      "grad_norm": 3.3052463644202037,
+      "learning_rate": 8.777730597832631e-06,
+      "loss": 1.293,
+      "step": 5743
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "grad_norm": 3.223868764354965,
+      "learning_rate": 8.774636138085013e-06,
+      "loss": 1.2274,
+      "step": 5744
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.1531,
+      "step": 5744,
+      "vm_loss": 0.1115
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.5379,
+      "step": 5744,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.4541,
+      "step": 5744,
+      "vm_loss": 0.2231
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.1609,
+      "step": 5744,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.0592,
+      "step": 5744,
+      "vm_loss": 0.2418
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 0.6771,
+      "step": 5744,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.2431,
+      "step": 5744,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.1057583559929736,
+      "lm_loss": 1.4037,
+      "step": 5744,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 1.105950862670549,
+      "grad_norm": 3.3372441862643623,
+      "learning_rate": 8.771541797458438e-06,
+      "loss": 1.3425,
+      "step": 5745
+    },
+    {
+      "epoch": 1.1061433693481242,
+      "grad_norm": 3.282351538931925,
+      "learning_rate": 8.76844757625372e-06,
+      "loss": 1.3096,
+      "step": 5746
+    },
+    {
+      "epoch": 1.1063358760256996,
+      "grad_norm": 3.255183448764697,
+      "learning_rate": 8.765353474771652e-06,
+      "loss": 1.303,
+      "step": 5747
+    },
+    {
+      "epoch": 1.106528382703275,
+      "grad_norm": 3.2298008842137027,
+      "learning_rate": 8.76225949331302e-06,
+      "loss": 1.3707,
+      "step": 5748
+    },
+    {
+      "epoch": 1.1067208893808504,
+      "grad_norm": 3.2003155492387556,
+      "learning_rate": 8.759165632178605e-06,
+      "loss": 1.3075,
+      "step": 5749
+    },
+    {
+      "epoch": 1.1069133960584259,
+      "grad_norm": 3.0844910498814087,
+      "learning_rate": 8.756071891669164e-06,
+      "loss": 1.2853,
+      "step": 5750
+    },
+    {
+      "epoch": 1.107105902736001,
+      "grad_norm": 3.200859508529741,
+      "learning_rate": 8.752978272085445e-06,
+      "loss": 1.4472,
+      "step": 5751
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "grad_norm": 3.1647514808431545,
+      "learning_rate": 8.749884773728197e-06,
+      "loss": 1.2861,
+      "step": 5752
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.1724,
+      "step": 5752,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.0232,
+      "step": 5752,
+      "vm_loss": 0.2277
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.339,
+      "step": 5752,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.7843,
+      "step": 5752,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 0.9503,
+      "step": 5752,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 0.7747,
+      "step": 5752,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.042,
+      "step": 5752,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.1072984094135765,
+      "lm_loss": 1.321,
+      "step": 5752,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.107490916091152,
+      "grad_norm": 3.2335417505779027,
+      "learning_rate": 8.746791396898142e-06,
+      "loss": 1.3539,
+      "step": 5753
+    },
+    {
+      "epoch": 1.1076834227687273,
+      "grad_norm": 3.129828256571666,
+      "learning_rate": 8.743698141895995e-06,
+      "loss": 1.2401,
+      "step": 5754
+    },
+    {
+      "epoch": 1.1078759294463028,
+      "grad_norm": 3.38379470508949,
+      "learning_rate": 8.740605009022462e-06,
+      "loss": 1.3669,
+      "step": 5755
+    },
+    {
+      "epoch": 1.108068436123878,
+      "grad_norm": 3.4064027881281316,
+      "learning_rate": 8.737511998578237e-06,
+      "loss": 1.2685,
+      "step": 5756
+    },
+    {
+      "epoch": 1.1082609428014534,
+      "grad_norm": 3.3082602514750192,
+      "learning_rate": 8.734419110863996e-06,
+      "loss": 1.3381,
+      "step": 5757
+    },
+    {
+      "epoch": 1.1084534494790288,
+      "grad_norm": 3.104944738599945,
+      "learning_rate": 8.731326346180406e-06,
+      "loss": 1.2362,
+      "step": 5758
+    },
+    {
+      "epoch": 1.1086459561566042,
+      "grad_norm": 3.2475149597922517,
+      "learning_rate": 8.728233704828129e-06,
+      "loss": 1.2941,
+      "step": 5759
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "grad_norm": 3.1182486341455573,
+      "learning_rate": 8.725141187107809e-06,
+      "loss": 1.2464,
+      "step": 5760
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 0.7673,
+      "step": 5760,
+      "vm_loss": 0.2399
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 1.2846,
+      "step": 5760,
+      "vm_loss": 0.1154
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 0.7412,
+      "step": 5760,
+      "vm_loss": 0.2398
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 1.2939,
+      "step": 5760,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 0.8116,
+      "step": 5760,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 1.1136,
+      "step": 5760,
+      "vm_loss": 0.1837
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 1.2673,
+      "step": 5760,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 1.1088384628341796,
+      "lm_loss": 0.8531,
+      "step": 5760,
+      "vm_loss": 0.2475
+    },
+    {
+      "epoch": 1.1090309695117548,
+      "grad_norm": 3.1310694188883486,
+      "learning_rate": 8.722048793320069e-06,
+      "loss": 1.257,
+      "step": 5761
+    },
+    {
+      "epoch": 1.1092234761893303,
+      "grad_norm": 3.1253915911173493,
+      "learning_rate": 8.718956523765542e-06,
+      "loss": 1.2536,
+      "step": 5762
+    },
+    {
+      "epoch": 1.1094159828669057,
+      "grad_norm": 3.1741424388362764,
+      "learning_rate": 8.715864378744827e-06,
+      "loss": 1.2745,
+      "step": 5763
+    },
+    {
+      "epoch": 1.109608489544481,
+      "grad_norm": 3.2240120514473882,
+      "learning_rate": 8.712772358558526e-06,
+      "loss": 1.3256,
+      "step": 5764
+    },
+    {
+      "epoch": 1.1098009962220565,
+      "grad_norm": 3.24387418275331,
+      "learning_rate": 8.709680463507217e-06,
+      "loss": 1.2493,
+      "step": 5765
+    },
+    {
+      "epoch": 1.109993502899632,
+      "grad_norm": 3.319137751666423,
+      "learning_rate": 8.706588693891475e-06,
+      "loss": 1.298,
+      "step": 5766
+    },
+    {
+      "epoch": 1.1101860095772071,
+      "grad_norm": 3.249226861485674,
+      "learning_rate": 8.703497050011865e-06,
+      "loss": 1.2556,
+      "step": 5767
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "grad_norm": 3.162930382574667,
+      "learning_rate": 8.700405532168921e-06,
+      "loss": 1.3077,
+      "step": 5768
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 0.8303,
+      "step": 5768,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.0244,
+      "step": 5768,
+      "vm_loss": 0.2202
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.1835,
+      "step": 5768,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.0347,
+      "step": 5768,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 0.8018,
+      "step": 5768,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.0218,
+      "step": 5768,
+      "vm_loss": 0.2613
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.0776,
+      "step": 5768,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.1103785162547826,
+      "lm_loss": 1.4667,
+      "step": 5768,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.110571022932358,
+      "grad_norm": 3.07642110724025,
+      "learning_rate": 8.697314140663191e-06,
+      "loss": 1.2263,
+      "step": 5769
+    },
+    {
+      "epoch": 1.1107635296099334,
+      "grad_norm": 3.249267159171443,
+      "learning_rate": 8.694222875795193e-06,
+      "loss": 1.2744,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1109560362875088,
+      "grad_norm": 3.13720427137784,
+      "learning_rate": 8.691131737865432e-06,
+      "loss": 1.2248,
+      "step": 5771
+    },
+    {
+      "epoch": 1.111148542965084,
+      "grad_norm": 3.2859805775860718,
+      "learning_rate": 8.688040727174419e-06,
+      "loss": 1.2715,
+      "step": 5772
+    },
+    {
+      "epoch": 1.1113410496426595,
+      "grad_norm": 3.2549457878110406,
+      "learning_rate": 8.684949844022629e-06,
+      "loss": 1.2662,
+      "step": 5773
+    },
+    {
+      "epoch": 1.1115335563202349,
+      "grad_norm": 3.3197574947459763,
+      "learning_rate": 8.681859088710538e-06,
+      "loss": 1.2975,
+      "step": 5774
+    },
+    {
+      "epoch": 1.1117260629978103,
+      "grad_norm": 3.3564377916550563,
+      "learning_rate": 8.678768461538605e-06,
+      "loss": 1.3153,
+      "step": 5775
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "grad_norm": 3.3079923980813457,
+      "learning_rate": 8.675677962807285e-06,
+      "loss": 1.3404,
+      "step": 5776
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 1.1684,
+      "step": 5776,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 0.7326,
+      "step": 5776,
+      "vm_loss": 0.1369
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 1.3145,
+      "step": 5776,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 1.2669,
+      "step": 5776,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 0.7184,
+      "step": 5776,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 1.6034,
+      "step": 5776,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 1.6335,
+      "step": 5776,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 1.1119185696753857,
+      "lm_loss": 0.5951,
+      "step": 5776,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.112111076352961,
+      "grad_norm": 3.296730670451773,
+      "learning_rate": 8.67258759281701e-06,
+      "loss": 1.2567,
+      "step": 5777
+    },
+    {
+      "epoch": 1.1123035830305363,
+      "grad_norm": 3.4502152249762306,
+      "learning_rate": 8.669497351868199e-06,
+      "loss": 1.367,
+      "step": 5778
+    },
+    {
+      "epoch": 1.1124960897081118,
+      "grad_norm": 3.236412433177419,
+      "learning_rate": 8.666407240261271e-06,
+      "loss": 1.2906,
+      "step": 5779
+    },
+    {
+      "epoch": 1.1126885963856872,
+      "grad_norm": 3.20910457197812,
+      "learning_rate": 8.663317258296624e-06,
+      "loss": 1.233,
+      "step": 5780
+    },
+    {
+      "epoch": 1.1128811030632626,
+      "grad_norm": 3.3524233100461003,
+      "learning_rate": 8.660227406274636e-06,
+      "loss": 1.331,
+      "step": 5781
+    },
+    {
+      "epoch": 1.1130736097408378,
+      "grad_norm": 3.3928806572816166,
+      "learning_rate": 8.657137684495685e-06,
+      "loss": 1.34,
+      "step": 5782
+    },
+    {
+      "epoch": 1.1132661164184132,
+      "grad_norm": 3.1867955984439162,
+      "learning_rate": 8.654048093260135e-06,
+      "loss": 1.2842,
+      "step": 5783
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "grad_norm": 3.375041312887529,
+      "learning_rate": 8.650958632868325e-06,
+      "loss": 1.285,
+      "step": 5784
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.0885,
+      "step": 5784,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 0.9434,
+      "step": 5784,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.1283,
+      "step": 5784,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.0804,
+      "step": 5784,
+      "vm_loss": 0.1216
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.1903,
+      "step": 5784,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.0224,
+      "step": 5784,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 0.4357,
+      "step": 5784,
+      "vm_loss": 0.2024
+    },
+    {
+      "epoch": 1.1134586230959886,
+      "lm_loss": 1.1196,
+      "step": 5784,
+      "vm_loss": 0.2381
+    },
+    {
+      "epoch": 1.113651129773564,
+      "grad_norm": 3.211773769195659,
+      "learning_rate": 8.647869303620597e-06,
+      "loss": 1.2346,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1138436364511395,
+      "grad_norm": 3.416427406500532,
+      "learning_rate": 8.644780105817272e-06,
+      "loss": 1.3322,
+      "step": 5786
+    },
+    {
+      "epoch": 1.1140361431287147,
+      "grad_norm": 3.194090236827222,
+      "learning_rate": 8.64169103975866e-06,
+      "loss": 1.2845,
+      "step": 5787
+    },
+    {
+      "epoch": 1.11422864980629,
+      "grad_norm": 3.2073354884866188,
+      "learning_rate": 8.638602105745058e-06,
+      "loss": 1.2709,
+      "step": 5788
+    },
+    {
+      "epoch": 1.1144211564838655,
+      "grad_norm": 3.248757242893497,
+      "learning_rate": 8.635513304076742e-06,
+      "loss": 1.3377,
+      "step": 5789
+    },
+    {
+      "epoch": 1.114613663161441,
+      "grad_norm": 3.103522565537857,
+      "learning_rate": 8.632424635053997e-06,
+      "loss": 1.2263,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1148061698390164,
+      "grad_norm": 3.0884328320560743,
+      "learning_rate": 8.629336098977073e-06,
+      "loss": 1.26,
+      "step": 5791
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "grad_norm": 3.383489323314709,
+      "learning_rate": 8.62624769614621e-06,
+      "loss": 1.3099,
+      "step": 5792
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 0.9667,
+      "step": 5792,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 1.3053,
+      "step": 5792,
+      "vm_loss": 0.2438
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 1.2457,
+      "step": 5792,
+      "vm_loss": 0.2179
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 0.7514,
+      "step": 5792,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 1.0549,
+      "step": 5792,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 0.7807,
+      "step": 5792,
+      "vm_loss": 0.1064
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 0.911,
+      "step": 5792,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 1.1149986765165916,
+      "lm_loss": 0.9711,
+      "step": 5792,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.115191183194167,
+      "grad_norm": 3.2417360932903425,
+      "learning_rate": 8.623159426861653e-06,
+      "loss": 1.2937,
+      "step": 5793
+    },
+    {
+      "epoch": 1.1153836898717424,
+      "grad_norm": 3.153840568085341,
+      "learning_rate": 8.620071291423613e-06,
+      "loss": 1.2251,
+      "step": 5794
+    },
+    {
+      "epoch": 1.1155761965493178,
+      "grad_norm": 3.255810899035514,
+      "learning_rate": 8.616983290132294e-06,
+      "loss": 1.2595,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1157687032268933,
+      "grad_norm": 3.1984720836336007,
+      "learning_rate": 8.613895423287899e-06,
+      "loss": 1.2035,
+      "step": 5796
+    },
+    {
+      "epoch": 1.1159612099044685,
+      "grad_norm": 3.3253300663742875,
+      "learning_rate": 8.6108076911906e-06,
+      "loss": 1.2823,
+      "step": 5797
+    },
+    {
+      "epoch": 1.1161537165820439,
+      "grad_norm": 3.2594200375673217,
+      "learning_rate": 8.607720094140567e-06,
+      "loss": 1.3097,
+      "step": 5798
+    },
+    {
+      "epoch": 1.1163462232596193,
+      "grad_norm": 3.337625101396846,
+      "learning_rate": 8.60463263243795e-06,
+      "loss": 1.2551,
+      "step": 5799
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "grad_norm": 3.29332614245858,
+      "learning_rate": 8.601545306382897e-06,
+      "loss": 1.3796,
+      "step": 5800
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 1.4544,
+      "step": 5800,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 0.7637,
+      "step": 5800,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 1.4044,
+      "step": 5800,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 1.1879,
+      "step": 5800,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 0.9673,
+      "step": 5800,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 1.3464,
+      "step": 5800,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 0.8967,
+      "step": 5800,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 1.1165387299371947,
+      "lm_loss": 0.9183,
+      "step": 5800,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 1.1167312366147701,
+      "grad_norm": 3.181126926582151,
+      "learning_rate": 8.59845811627553e-06,
+      "loss": 1.2678,
+      "step": 5801
+    },
+    {
+      "epoch": 1.1169237432923453,
+      "grad_norm": 3.2007875687220184,
+      "learning_rate": 8.595371062415962e-06,
+      "loss": 1.3012,
+      "step": 5802
+    },
+    {
+      "epoch": 1.1171162499699208,
+      "grad_norm": 3.3794410356754905,
+      "learning_rate": 8.592284145104303e-06,
+      "loss": 1.2866,
+      "step": 5803
+    },
+    {
+      "epoch": 1.1173087566474962,
+      "grad_norm": 3.270731015546087,
+      "learning_rate": 8.58919736464063e-06,
+      "loss": 1.2684,
+      "step": 5804
+    },
+    {
+      "epoch": 1.1175012633250716,
+      "grad_norm": 3.0871461124290884,
+      "learning_rate": 8.586110721325028e-06,
+      "loss": 1.2454,
+      "step": 5805
+    },
+    {
+      "epoch": 1.117693770002647,
+      "grad_norm": 3.196088735564785,
+      "learning_rate": 8.583024215457548e-06,
+      "loss": 1.249,
+      "step": 5806
+    },
+    {
+      "epoch": 1.1178862766802224,
+      "grad_norm": 3.232896060269576,
+      "learning_rate": 8.579937847338245e-06,
+      "loss": 1.3196,
+      "step": 5807
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "grad_norm": 3.2574655536993955,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 1.255,
+      "step": 5808
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 1.5429,
+      "step": 5808,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 0.8541,
+      "step": 5808,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 1.1003,
+      "step": 5808,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 1.0175,
+      "step": 5808,
+      "vm_loss": 0.1691
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 0.5826,
+      "step": 5808,
+      "vm_loss": 0.1146
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 1.3037,
+      "step": 5808,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 0.8394,
+      "step": 5808,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.1180787833577976,
+      "lm_loss": 1.0671,
+      "step": 5808,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.118271290035373,
+      "grad_norm": 3.1902434672686613,
+      "learning_rate": 8.573765525544282e-06,
+      "loss": 1.2234,
+      "step": 5809
+    },
+    {
+      "epoch": 1.1184637967129485,
+      "grad_norm": 3.3514099123226946,
+      "learning_rate": 8.570679572469659e-06,
+      "loss": 1.3134,
+      "step": 5810
+    },
+    {
+      "epoch": 1.118656303390524,
+      "grad_norm": 3.3315402318895324,
+      "learning_rate": 8.567593758343266e-06,
+      "loss": 1.3401,
+      "step": 5811
+    },
+    {
+      "epoch": 1.1188488100680993,
+      "grad_norm": 3.173129076311302,
+      "learning_rate": 8.56450808346508e-06,
+      "loss": 1.2829,
+      "step": 5812
+    },
+    {
+      "epoch": 1.1190413167456745,
+      "grad_norm": 3.1240676379334684,
+      "learning_rate": 8.561422548135079e-06,
+      "loss": 1.2446,
+      "step": 5813
+    },
+    {
+      "epoch": 1.11923382342325,
+      "grad_norm": 3.183081385208541,
+      "learning_rate": 8.558337152653211e-06,
+      "loss": 1.3134,
+      "step": 5814
+    },
+    {
+      "epoch": 1.1194263301008254,
+      "grad_norm": 3.2413698585858035,
+      "learning_rate": 8.555251897319417e-06,
+      "loss": 1.3259,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "grad_norm": 3.233607278621533,
+      "learning_rate": 8.552166782433617e-06,
+      "loss": 1.2863,
+      "step": 5816
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 1.1983,
+      "step": 5816,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 0.9664,
+      "step": 5816,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 1.0984,
+      "step": 5816,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 0.4811,
+      "step": 5816,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 0.6098,
+      "step": 5816,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 1.2023,
+      "step": 5816,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 1.0636,
+      "step": 5816,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 1.1196188367784008,
+      "lm_loss": 0.9158,
+      "step": 5816,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.1198113434559762,
+      "grad_norm": 3.189399888909901,
+      "learning_rate": 8.549081808295732e-06,
+      "loss": 1.2151,
+      "step": 5817
+    },
+    {
+      "epoch": 1.1200038501335514,
+      "grad_norm": 3.4421164148864336,
+      "learning_rate": 8.54599697520566e-06,
+      "loss": 1.4045,
+      "step": 5818
+    },
+    {
+      "epoch": 1.1201963568111268,
+      "grad_norm": 3.2848889591705737,
+      "learning_rate": 8.542912283463278e-06,
+      "loss": 1.3319,
+      "step": 5819
+    },
+    {
+      "epoch": 1.1203888634887023,
+      "grad_norm": 3.0149955803195025,
+      "learning_rate": 8.53982773336847e-06,
+      "loss": 1.2412,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1205813701662777,
+      "grad_norm": 3.3221193616576468,
+      "learning_rate": 8.536743325221087e-06,
+      "loss": 1.343,
+      "step": 5821
+    },
+    {
+      "epoch": 1.120773876843853,
+      "grad_norm": 3.1189604342214143,
+      "learning_rate": 8.53365905932097e-06,
+      "loss": 1.2208,
+      "step": 5822
+    },
+    {
+      "epoch": 1.1209663835214283,
+      "grad_norm": 3.2121365800015544,
+      "learning_rate": 8.530574935967952e-06,
+      "loss": 1.3094,
+      "step": 5823
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "grad_norm": 3.2384096282188666,
+      "learning_rate": 8.527490955461852e-06,
+      "loss": 1.2752,
+      "step": 5824
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.3402,
+      "step": 5824,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.5689,
+      "step": 5824,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.0668,
+      "step": 5824,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 0.8923,
+      "step": 5824,
+      "vm_loss": 0.1059
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.1477,
+      "step": 5824,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.4189,
+      "step": 5824,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 0.9885,
+      "step": 5824,
+      "vm_loss": 0.2499
+    },
+    {
+      "epoch": 1.1211588901990037,
+      "lm_loss": 1.1444,
+      "step": 5824,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 1.1213513968765791,
+      "grad_norm": 3.1174973130473465,
+      "learning_rate": 8.52440711810247e-06,
+      "loss": 1.2267,
+      "step": 5825
+    },
+    {
+      "epoch": 1.1215439035541546,
+      "grad_norm": 3.2690984713514513,
+      "learning_rate": 8.521323424189593e-06,
+      "loss": 1.3266,
+      "step": 5826
+    },
+    {
+      "epoch": 1.12173641023173,
+      "grad_norm": 3.100651494825096,
+      "learning_rate": 8.518239874022997e-06,
+      "loss": 1.2071,
+      "step": 5827
+    },
+    {
+      "epoch": 1.1219289169093054,
+      "grad_norm": 3.2724168318632745,
+      "learning_rate": 8.515156467902444e-06,
+      "loss": 1.2676,
+      "step": 5828
+    },
+    {
+      "epoch": 1.1221214235868806,
+      "grad_norm": 3.2248893350241414,
+      "learning_rate": 8.512073206127677e-06,
+      "loss": 1.2517,
+      "step": 5829
+    },
+    {
+      "epoch": 1.122313930264456,
+      "grad_norm": 3.2508027984971823,
+      "learning_rate": 8.508990088998431e-06,
+      "loss": 1.295,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1225064369420314,
+      "grad_norm": 3.2473741504206983,
+      "learning_rate": 8.505907116814427e-06,
+      "loss": 1.344,
+      "step": 5831
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "grad_norm": 3.151055139151268,
+      "learning_rate": 8.502824289875366e-06,
+      "loss": 1.223,
+      "step": 5832
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 1.6531,
+      "step": 5832,
+      "vm_loss": 0.2195
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 0.6116,
+      "step": 5832,
+      "vm_loss": 0.2436
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 1.3,
+      "step": 5832,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 0.9203,
+      "step": 5832,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 1.1451,
+      "step": 5832,
+      "vm_loss": 0.1276
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 0.8976,
+      "step": 5832,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 1.0667,
+      "step": 5832,
+      "vm_loss": 0.1296
+    },
+    {
+      "epoch": 1.1226989436196069,
+      "lm_loss": 1.4443,
+      "step": 5832,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.1228914502971823,
+      "grad_norm": 3.273362627432169,
+      "learning_rate": 8.499741608480932e-06,
+      "loss": 1.3474,
+      "step": 5833
+    },
+    {
+      "epoch": 1.1230839569747575,
+      "grad_norm": 3.1271279159256675,
+      "learning_rate": 8.496659072930815e-06,
+      "loss": 1.2773,
+      "step": 5834
+    },
+    {
+      "epoch": 1.123276463652333,
+      "grad_norm": 3.2203187817709824,
+      "learning_rate": 8.493576683524667e-06,
+      "loss": 1.3296,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1234689703299083,
+      "grad_norm": 3.2630457643465918,
+      "learning_rate": 8.490494440562136e-06,
+      "loss": 1.2938,
+      "step": 5836
+    },
+    {
+      "epoch": 1.1236614770074838,
+      "grad_norm": 3.2379638364155934,
+      "learning_rate": 8.48741234434286e-06,
+      "loss": 1.3148,
+      "step": 5837
+    },
+    {
+      "epoch": 1.1238539836850592,
+      "grad_norm": 3.170706942060178,
+      "learning_rate": 8.484330395166455e-06,
+      "loss": 1.3533,
+      "step": 5838
+    },
+    {
+      "epoch": 1.1240464903626344,
+      "grad_norm": 3.137551450835941,
+      "learning_rate": 8.48124859333253e-06,
+      "loss": 1.2266,
+      "step": 5839
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "grad_norm": 3.2694774073982154,
+      "learning_rate": 8.478166939140665e-06,
+      "loss": 1.3561,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 1.4241,
+      "step": 5840,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 1.2203,
+      "step": 5840,
+      "vm_loss": 0.2032
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 0.9619,
+      "step": 5840,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 0.9268,
+      "step": 5840,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 0.8854,
+      "step": 5840,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 1.368,
+      "step": 5840,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 1.1761,
+      "step": 5840,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.1242389970402098,
+      "lm_loss": 0.8146,
+      "step": 5840,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.1244315037177852,
+      "grad_norm": 3.0620694997931004,
+      "learning_rate": 8.475085432890449e-06,
+      "loss": 1.2172,
+      "step": 5841
+    },
+    {
+      "epoch": 1.1246240103953606,
+      "grad_norm": 3.237102027297433,
+      "learning_rate": 8.472004074881438e-06,
+      "loss": 1.3298,
+      "step": 5842
+    },
+    {
+      "epoch": 1.124816517072936,
+      "grad_norm": 3.2295491664366978,
+      "learning_rate": 8.468922865413176e-06,
+      "loss": 1.2595,
+      "step": 5843
+    },
+    {
+      "epoch": 1.1250090237505113,
+      "grad_norm": 3.1840226859224017,
+      "learning_rate": 8.465841804785202e-06,
+      "loss": 1.272,
+      "step": 5844
+    },
+    {
+      "epoch": 1.1252015304280867,
+      "grad_norm": 3.356093941009692,
+      "learning_rate": 8.462760893297033e-06,
+      "loss": 1.3238,
+      "step": 5845
+    },
+    {
+      "epoch": 1.125394037105662,
+      "grad_norm": 3.42043759963782,
+      "learning_rate": 8.459680131248168e-06,
+      "loss": 1.3248,
+      "step": 5846
+    },
+    {
+      "epoch": 1.1255865437832375,
+      "grad_norm": 3.299890605639256,
+      "learning_rate": 8.456599518938107e-06,
+      "loss": 1.32,
+      "step": 5847
+    },
+    {
+      "epoch": 1.125779050460813,
+      "grad_norm": 3.19825838761476,
+      "learning_rate": 8.453519056666314e-06,
+      "loss": 1.3599,
+      "step": 5848
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 1.5224,
+      "step": 5848,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 1.3296,
+      "step": 5848,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 0.5724,
+      "step": 5848,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 1.1999,
+      "step": 5848,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 0.8062,
+      "step": 5848,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 0.7877,
+      "step": 5848,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 0.9927,
+      "step": 5848,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 1.125779050460813,
+      "lm_loss": 1.9508,
+      "step": 5848,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 1.1259715571383881,
+      "grad_norm": 3.035662040743749,
+      "learning_rate": 8.450438744732259e-06,
+      "loss": 1.208,
+      "step": 5849
+    },
+    {
+      "epoch": 1.1261640638159636,
+      "grad_norm": 3.1088250859066857,
+      "learning_rate": 8.447358583435377e-06,
+      "loss": 1.2635,
+      "step": 5850
+    },
+    {
+      "epoch": 1.126356570493539,
+      "grad_norm": 3.21080519231543,
+      "learning_rate": 8.444278573075107e-06,
+      "loss": 1.3395,
+      "step": 5851
+    },
+    {
+      "epoch": 1.1265490771711144,
+      "grad_norm": 3.1600525452007626,
+      "learning_rate": 8.441198713950865e-06,
+      "loss": 1.2287,
+      "step": 5852
+    },
+    {
+      "epoch": 1.1267415838486898,
+      "grad_norm": 3.3862551949520965,
+      "learning_rate": 8.438119006362046e-06,
+      "loss": 1.3177,
+      "step": 5853
+    },
+    {
+      "epoch": 1.126934090526265,
+      "grad_norm": 3.1757911641365952,
+      "learning_rate": 8.435039450608046e-06,
+      "loss": 1.2492,
+      "step": 5854
+    },
+    {
+      "epoch": 1.1271265972038405,
+      "grad_norm": 3.2158351021886262,
+      "learning_rate": 8.431960046988235e-06,
+      "loss": 1.3062,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "grad_norm": 3.4285830689747785,
+      "learning_rate": 8.428880795801965e-06,
+      "loss": 1.2645,
+      "step": 5856
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 1.2031,
+      "step": 5856,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 0.834,
+      "step": 5856,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 0.9468,
+      "step": 5856,
+      "vm_loss": 0.2031
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 1.0256,
+      "step": 5856,
+      "vm_loss": 0.2021
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 0.8655,
+      "step": 5856,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 1.1139,
+      "step": 5856,
+      "vm_loss": 0.1018
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 1.0159,
+      "step": 5856,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.1273191038814159,
+      "lm_loss": 0.6217,
+      "step": 5856,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 1.1275116105589913,
+      "grad_norm": 3.140633895907306,
+      "learning_rate": 8.42580169734858e-06,
+      "loss": 1.2262,
+      "step": 5857
+    },
+    {
+      "epoch": 1.1277041172365667,
+      "grad_norm": 3.143927110954885,
+      "learning_rate": 8.422722751927415e-06,
+      "loss": 1.2241,
+      "step": 5858
+    },
+    {
+      "epoch": 1.127896623914142,
+      "grad_norm": 3.351408604636248,
+      "learning_rate": 8.419643959837777e-06,
+      "loss": 1.2718,
+      "step": 5859
+    },
+    {
+      "epoch": 1.1280891305917173,
+      "grad_norm": 3.3739720287411203,
+      "learning_rate": 8.416565321378962e-06,
+      "loss": 1.2684,
+      "step": 5860
+    },
+    {
+      "epoch": 1.1282816372692928,
+      "grad_norm": 3.2173463678522864,
+      "learning_rate": 8.413486836850258e-06,
+      "loss": 1.2844,
+      "step": 5861
+    },
+    {
+      "epoch": 1.1284741439468682,
+      "grad_norm": 3.1883903205622155,
+      "learning_rate": 8.41040850655093e-06,
+      "loss": 1.3117,
+      "step": 5862
+    },
+    {
+      "epoch": 1.1286666506244436,
+      "grad_norm": 3.1852220130364937,
+      "learning_rate": 8.407330330780234e-06,
+      "loss": 1.3195,
+      "step": 5863
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "grad_norm": 3.316046859845407,
+      "learning_rate": 8.4042523098374e-06,
+      "loss": 1.2692,
+      "step": 5864
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 1.8406,
+      "step": 5864,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.9737,
+      "step": 5864,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.5428,
+      "step": 5864,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.9614,
+      "step": 5864,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 1.2193,
+      "step": 5864,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.8916,
+      "step": 5864,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.9738,
+      "step": 5864,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 1.1288591573020188,
+      "lm_loss": 0.8708,
+      "step": 5864,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.1290516639795942,
+      "grad_norm": 3.2270768209537355,
+      "learning_rate": 8.401174444021663e-06,
+      "loss": 1.2638,
+      "step": 5865
+    },
+    {
+      "epoch": 1.1292441706571696,
+      "grad_norm": 3.3164734166666596,
+      "learning_rate": 8.398096733632223e-06,
+      "loss": 1.2329,
+      "step": 5866
+    },
+    {
+      "epoch": 1.129436677334745,
+      "grad_norm": 3.2916633833882223,
+      "learning_rate": 8.395019178968274e-06,
+      "loss": 1.2415,
+      "step": 5867
+    },
+    {
+      "epoch": 1.1296291840123205,
+      "grad_norm": 3.5267657621273036,
+      "learning_rate": 8.391941780328996e-06,
+      "loss": 1.2922,
+      "step": 5868
+    },
+    {
+      "epoch": 1.1298216906898957,
+      "grad_norm": 3.1315513936167827,
+      "learning_rate": 8.38886453801355e-06,
+      "loss": 1.2486,
+      "step": 5869
+    },
+    {
+      "epoch": 1.130014197367471,
+      "grad_norm": 3.1883547318831607,
+      "learning_rate": 8.38578745232108e-06,
+      "loss": 1.2458,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1302067040450465,
+      "grad_norm": 3.2144540780756685,
+      "learning_rate": 8.382710523550725e-06,
+      "loss": 1.2772,
+      "step": 5871
+    },
+    {
+      "epoch": 1.130399210722622,
+      "grad_norm": 3.0267568001221354,
+      "learning_rate": 8.379633752001597e-06,
+      "loss": 1.2134,
+      "step": 5872
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 0.8084,
+      "step": 5872,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 0.991,
+      "step": 5872,
+      "vm_loss": 0.1964
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 1.2142,
+      "step": 5872,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 1.483,
+      "step": 5872,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 1.1382,
+      "step": 5872,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 1.4086,
+      "step": 5872,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 1.0364,
+      "step": 5872,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.130399210722622,
+      "lm_loss": 0.8559,
+      "step": 5872,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 1.1305917174001974,
+      "grad_norm": 3.1448279535608665,
+      "learning_rate": 8.376557137972801e-06,
+      "loss": 1.2348,
+      "step": 5873
+    },
+    {
+      "epoch": 1.1307842240777728,
+      "grad_norm": 3.2773959062530764,
+      "learning_rate": 8.373480681763418e-06,
+      "loss": 1.3179,
+      "step": 5874
+    },
+    {
+      "epoch": 1.130976730755348,
+      "grad_norm": 3.2631498231279292,
+      "learning_rate": 8.370404383672525e-06,
+      "loss": 1.3622,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1311692374329234,
+      "grad_norm": 3.2126047187470883,
+      "learning_rate": 8.367328243999174e-06,
+      "loss": 1.2495,
+      "step": 5876
+    },
+    {
+      "epoch": 1.1313617441104988,
+      "grad_norm": 3.2704190844929886,
+      "learning_rate": 8.3642522630424e-06,
+      "loss": 1.3115,
+      "step": 5877
+    },
+    {
+      "epoch": 1.1315542507880743,
+      "grad_norm": 3.126034016582004,
+      "learning_rate": 8.36117644110124e-06,
+      "loss": 1.2036,
+      "step": 5878
+    },
+    {
+      "epoch": 1.1317467574656497,
+      "grad_norm": 3.319519104724287,
+      "learning_rate": 8.358100778474696e-06,
+      "loss": 1.3649,
+      "step": 5879
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "grad_norm": 3.389219887371969,
+      "learning_rate": 8.355025275461764e-06,
+      "loss": 1.3423,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 0.9885,
+      "step": 5880,
+      "vm_loss": 0.1037
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.0471,
+      "step": 5880,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.2167,
+      "step": 5880,
+      "vm_loss": 0.2535
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.006,
+      "step": 5880,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.4664,
+      "step": 5880,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.0578,
+      "step": 5880,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.1965,
+      "step": 5880,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.1319392641432249,
+      "lm_loss": 1.1448,
+      "step": 5880,
+      "vm_loss": 0.1074
+    },
+    {
+      "epoch": 1.1321317708208003,
+      "grad_norm": 3.0365568448688363,
+      "learning_rate": 8.351949932361413e-06,
+      "loss": 1.2472,
+      "step": 5881
+    },
+    {
+      "epoch": 1.1323242774983757,
+      "grad_norm": 3.1737492451190423,
+      "learning_rate": 8.34887474947262e-06,
+      "loss": 1.2445,
+      "step": 5882
+    },
+    {
+      "epoch": 1.1325167841759511,
+      "grad_norm": 3.155271224195876,
+      "learning_rate": 8.345799727094323e-06,
+      "loss": 1.3018,
+      "step": 5883
+    },
+    {
+      "epoch": 1.1327092908535266,
+      "grad_norm": 3.0813235955400273,
+      "learning_rate": 8.342724865525455e-06,
+      "loss": 1.3244,
+      "step": 5884
+    },
+    {
+      "epoch": 1.132901797531102,
+      "grad_norm": 3.2410567940081236,
+      "learning_rate": 8.339650165064935e-06,
+      "loss": 1.3045,
+      "step": 5885
+    },
+    {
+      "epoch": 1.1330943042086772,
+      "grad_norm": 3.1581908678685644,
+      "learning_rate": 8.33657562601166e-06,
+      "loss": 1.2405,
+      "step": 5886
+    },
+    {
+      "epoch": 1.1332868108862526,
+      "grad_norm": 3.1498172778383347,
+      "learning_rate": 8.333501248664513e-06,
+      "loss": 1.2199,
+      "step": 5887
+    },
+    {
+      "epoch": 1.133479317563828,
+      "grad_norm": 3.3202762759733138,
+      "learning_rate": 8.33042703332237e-06,
+      "loss": 1.2404,
+      "step": 5888
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 1.3709,
+      "step": 5888,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 0.8066,
+      "step": 5888,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 1.247,
+      "step": 5888,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 0.7363,
+      "step": 5888,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 1.52,
+      "step": 5888,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 0.8289,
+      "step": 5888,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 1.4974,
+      "step": 5888,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.133479317563828,
+      "lm_loss": 0.8929,
+      "step": 5888,
+      "vm_loss": 0.1811
+    },
+    {
+      "epoch": 1.1336718242414034,
+      "grad_norm": 3.2119194287590354,
+      "learning_rate": 8.327352980284076e-06,
+      "loss": 1.2301,
+      "step": 5889
+    },
+    {
+      "epoch": 1.1338643309189789,
+      "grad_norm": 3.367612511967922,
+      "learning_rate": 8.324279089848475e-06,
+      "loss": 1.2604,
+      "step": 5890
+    },
+    {
+      "epoch": 1.134056837596554,
+      "grad_norm": 3.393785170173505,
+      "learning_rate": 8.321205362314382e-06,
+      "loss": 1.3117,
+      "step": 5891
+    },
+    {
+      "epoch": 1.1342493442741295,
+      "grad_norm": 3.2545119201967827,
+      "learning_rate": 8.318131797980606e-06,
+      "loss": 1.2792,
+      "step": 5892
+    },
+    {
+      "epoch": 1.134441850951705,
+      "grad_norm": 3.2777173328509694,
+      "learning_rate": 8.31505839714594e-06,
+      "loss": 1.3164,
+      "step": 5893
+    },
+    {
+      "epoch": 1.1346343576292803,
+      "grad_norm": 3.2705333322198573,
+      "learning_rate": 8.311985160109148e-06,
+      "loss": 1.3177,
+      "step": 5894
+    },
+    {
+      "epoch": 1.1348268643068558,
+      "grad_norm": 3.112691149032303,
+      "learning_rate": 8.308912087169001e-06,
+      "loss": 1.2403,
+      "step": 5895
+    },
+    {
+      "epoch": 1.135019370984431,
+      "grad_norm": 3.1961449583180523,
+      "learning_rate": 8.305839178624233e-06,
+      "loss": 1.245,
+      "step": 5896
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.1306,
+      "step": 5896,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.5792,
+      "step": 5896,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.2191,
+      "step": 5896,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.7574,
+      "step": 5896,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.3805,
+      "step": 5896,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.5082,
+      "step": 5896,
+      "vm_loss": 0.121
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 0.7554,
+      "step": 5896,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.135019370984431,
+      "lm_loss": 1.0155,
+      "step": 5896,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.1352118776620064,
+      "grad_norm": 3.25801437423002,
+      "learning_rate": 8.302766434773572e-06,
+      "loss": 1.33,
+      "step": 5897
+    },
+    {
+      "epoch": 1.1354043843395818,
+      "grad_norm": 3.193130018969506,
+      "learning_rate": 8.299693855915723e-06,
+      "loss": 1.2436,
+      "step": 5898
+    },
+    {
+      "epoch": 1.1355968910171572,
+      "grad_norm": 3.2663209610979114,
+      "learning_rate": 8.29662144234939e-06,
+      "loss": 1.2687,
+      "step": 5899
+    },
+    {
+      "epoch": 1.1357893976947326,
+      "grad_norm": 3.3788371678679905,
+      "learning_rate": 8.293549194373243e-06,
+      "loss": 1.2267,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1359819043723078,
+      "grad_norm": 3.295447626922356,
+      "learning_rate": 8.290477112285944e-06,
+      "loss": 1.2073,
+      "step": 5901
+    },
+    {
+      "epoch": 1.1361744110498833,
+      "grad_norm": 3.3398406764439974,
+      "learning_rate": 8.287405196386144e-06,
+      "loss": 1.2767,
+      "step": 5902
+    },
+    {
+      "epoch": 1.1363669177274587,
+      "grad_norm": 3.2929088211476523,
+      "learning_rate": 8.284333446972471e-06,
+      "loss": 1.2379,
+      "step": 5903
+    },
+    {
+      "epoch": 1.136559424405034,
+      "grad_norm": 3.3768346170975154,
+      "learning_rate": 8.28126186434353e-06,
+      "loss": 1.2948,
+      "step": 5904
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.1123,
+      "step": 5904,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.1455,
+      "step": 5904,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 0.6919,
+      "step": 5904,
+      "vm_loss": 0.2131
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.0409,
+      "step": 5904,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.1416,
+      "step": 5904,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.0636,
+      "step": 5904,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 0.5159,
+      "step": 5904,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.136559424405034,
+      "lm_loss": 1.1402,
+      "step": 5904,
+      "vm_loss": 0.1266
+    },
+    {
+      "epoch": 1.1367519310826095,
+      "grad_norm": 3.347507969112467,
+      "learning_rate": 8.278190448797931e-06,
+      "loss": 1.2502,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1369444377601847,
+      "grad_norm": 3.220006291494801,
+      "learning_rate": 8.275119200634248e-06,
+      "loss": 1.2404,
+      "step": 5906
+    },
+    {
+      "epoch": 1.1371369444377601,
+      "grad_norm": 3.302165627611826,
+      "learning_rate": 8.272048120151045e-06,
+      "loss": 1.3117,
+      "step": 5907
+    },
+    {
+      "epoch": 1.1373294511153356,
+      "grad_norm": 3.222033805945584,
+      "learning_rate": 8.26897720764687e-06,
+      "loss": 1.2809,
+      "step": 5908
+    },
+    {
+      "epoch": 1.137521957792911,
+      "grad_norm": 3.255902778533384,
+      "learning_rate": 8.26590646342026e-06,
+      "loss": 1.2362,
+      "step": 5909
+    },
+    {
+      "epoch": 1.1377144644704864,
+      "grad_norm": 3.132923384517969,
+      "learning_rate": 8.262835887769725e-06,
+      "loss": 1.2725,
+      "step": 5910
+    },
+    {
+      "epoch": 1.1379069711480616,
+      "grad_norm": 3.247885768061465,
+      "learning_rate": 8.259765480993767e-06,
+      "loss": 1.2829,
+      "step": 5911
+    },
+    {
+      "epoch": 1.138099477825637,
+      "grad_norm": 3.225341105290348,
+      "learning_rate": 8.256695243390868e-06,
+      "loss": 1.258,
+      "step": 5912
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 0.8413,
+      "step": 5912,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 1.3644,
+      "step": 5912,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 1.3922,
+      "step": 5912,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 1.0489,
+      "step": 5912,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 0.9715,
+      "step": 5912,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 1.0801,
+      "step": 5912,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 0.8907,
+      "step": 5912,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.138099477825637,
+      "lm_loss": 0.8373,
+      "step": 5912,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 1.1382919845032125,
+      "grad_norm": 3.202713247720956,
+      "learning_rate": 8.253625175259496e-06,
+      "loss": 1.2277,
+      "step": 5913
+    },
+    {
+      "epoch": 1.1384844911807879,
+      "grad_norm": 3.274301887721809,
+      "learning_rate": 8.2505552768981e-06,
+      "loss": 1.3141,
+      "step": 5914
+    },
+    {
+      "epoch": 1.1386769978583633,
+      "grad_norm": 3.2081132999983923,
+      "learning_rate": 8.24748554860511e-06,
+      "loss": 1.2779,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1388695045359385,
+      "grad_norm": 3.432795170897406,
+      "learning_rate": 8.244415990678949e-06,
+      "loss": 1.3048,
+      "step": 5916
+    },
+    {
+      "epoch": 1.139062011213514,
+      "grad_norm": 3.3953901171098724,
+      "learning_rate": 8.241346603418014e-06,
+      "loss": 1.3193,
+      "step": 5917
+    },
+    {
+      "epoch": 1.1392545178910893,
+      "grad_norm": 3.2766238592456514,
+      "learning_rate": 8.238277387120685e-06,
+      "loss": 1.2205,
+      "step": 5918
+    },
+    {
+      "epoch": 1.1394470245686648,
+      "grad_norm": 3.333379257890828,
+      "learning_rate": 8.235208342085337e-06,
+      "loss": 1.3327,
+      "step": 5919
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "grad_norm": 3.2670742832098916,
+      "learning_rate": 8.232139468610316e-06,
+      "loss": 1.2547,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 1.4901,
+      "step": 5920,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 0.5242,
+      "step": 5920,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 0.9055,
+      "step": 5920,
+      "vm_loss": 0.2502
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 1.0267,
+      "step": 5920,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 0.9592,
+      "step": 5920,
+      "vm_loss": 0.1139
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 0.9626,
+      "step": 5920,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 1.075,
+      "step": 5920,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.1396395312462402,
+      "lm_loss": 0.773,
+      "step": 5920,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.1398320379238154,
+      "grad_norm": 3.128453319459127,
+      "learning_rate": 8.229070766993955e-06,
+      "loss": 1.256,
+      "step": 5921
+    },
+    {
+      "epoch": 1.1400245446013908,
+      "grad_norm": 3.074504414024722,
+      "learning_rate": 8.226002237534574e-06,
+      "loss": 1.2452,
+      "step": 5922
+    },
+    {
+      "epoch": 1.1402170512789662,
+      "grad_norm": 3.1236767517262582,
+      "learning_rate": 8.22293388053047e-06,
+      "loss": 1.2531,
+      "step": 5923
+    },
+    {
+      "epoch": 1.1404095579565416,
+      "grad_norm": 3.1929280601830228,
+      "learning_rate": 8.21986569627993e-06,
+      "loss": 1.2414,
+      "step": 5924
+    },
+    {
+      "epoch": 1.140602064634117,
+      "grad_norm": 3.3336778312718036,
+      "learning_rate": 8.216797685081214e-06,
+      "loss": 1.2975,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1407945713116923,
+      "grad_norm": 3.2610425030665087,
+      "learning_rate": 8.213729847232583e-06,
+      "loss": 1.2399,
+      "step": 5926
+    },
+    {
+      "epoch": 1.1409870779892677,
+      "grad_norm": 3.323426816433846,
+      "learning_rate": 8.210662183032262e-06,
+      "loss": 1.2964,
+      "step": 5927
+    },
+    {
+      "epoch": 1.141179584666843,
+      "grad_norm": 3.389401166964098,
+      "learning_rate": 8.207594692778464e-06,
+      "loss": 1.3115,
+      "step": 5928
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 0.6674,
+      "step": 5928,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 1.0127,
+      "step": 5928,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 0.7549,
+      "step": 5928,
+      "vm_loss": 0.2239
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 0.8433,
+      "step": 5928,
+      "vm_loss": 0.1168
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 1.461,
+      "step": 5928,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 0.849,
+      "step": 5928,
+      "vm_loss": 0.2322
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 0.9908,
+      "step": 5928,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.141179584666843,
+      "lm_loss": 1.1951,
+      "step": 5928,
+      "vm_loss": 0.1122
+    },
+    {
+      "epoch": 1.1413720913444185,
+      "grad_norm": 3.2606204289380565,
+      "learning_rate": 8.204527376769401e-06,
+      "loss": 1.269,
+      "step": 5929
+    },
+    {
+      "epoch": 1.141564598021994,
+      "grad_norm": 3.1432490902994736,
+      "learning_rate": 8.201460235303245e-06,
+      "loss": 1.312,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1417571046995691,
+      "grad_norm": 3.263003331658955,
+      "learning_rate": 8.198393268678163e-06,
+      "loss": 1.3229,
+      "step": 5931
+    },
+    {
+      "epoch": 1.1419496113771446,
+      "grad_norm": 3.105052795488984,
+      "learning_rate": 8.195326477192305e-06,
+      "loss": 1.2408,
+      "step": 5932
+    },
+    {
+      "epoch": 1.14214211805472,
+      "grad_norm": 3.201663012326055,
+      "learning_rate": 8.1922598611438e-06,
+      "loss": 1.3032,
+      "step": 5933
+    },
+    {
+      "epoch": 1.1423346247322954,
+      "grad_norm": 3.3066660747381014,
+      "learning_rate": 8.189193420830768e-06,
+      "loss": 1.3294,
+      "step": 5934
+    },
+    {
+      "epoch": 1.1425271314098708,
+      "grad_norm": 3.2277095117417303,
+      "learning_rate": 8.1861271565513e-06,
+      "loss": 1.3289,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "grad_norm": 3.204952925050108,
+      "learning_rate": 8.18306106860348e-06,
+      "loss": 1.2099,
+      "step": 5936
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 1.0924,
+      "step": 5936,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 1.06,
+      "step": 5936,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.8774,
+      "step": 5936,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.7256,
+      "step": 5936,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.945,
+      "step": 5936,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.8488,
+      "step": 5936,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.8559,
+      "step": 5936,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.1427196380874463,
+      "lm_loss": 0.732,
+      "step": 5936,
+      "vm_loss": 0.1127
+    },
+    {
+      "epoch": 1.1429121447650215,
+      "grad_norm": 3.159000653909086,
+      "learning_rate": 8.17999515728537e-06,
+      "loss": 1.2107,
+      "step": 5937
+    },
+    {
+      "epoch": 1.1431046514425969,
+      "grad_norm": 3.161568254489513,
+      "learning_rate": 8.17692942289501e-06,
+      "loss": 1.2297,
+      "step": 5938
+    },
+    {
+      "epoch": 1.1432971581201723,
+      "grad_norm": 3.2931193580244518,
+      "learning_rate": 8.173863865730441e-06,
+      "loss": 1.2056,
+      "step": 5939
+    },
+    {
+      "epoch": 1.1434896647977477,
+      "grad_norm": 3.3584679898396033,
+      "learning_rate": 8.170798486089666e-06,
+      "loss": 1.2431,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1436821714753231,
+      "grad_norm": 3.3366810272422684,
+      "learning_rate": 8.167733284270681e-06,
+      "loss": 1.2478,
+      "step": 5941
+    },
+    {
+      "epoch": 1.1438746781528983,
+      "grad_norm": 3.3192864453322706,
+      "learning_rate": 8.164668260571457e-06,
+      "loss": 1.2387,
+      "step": 5942
+    },
+    {
+      "epoch": 1.1440671848304738,
+      "grad_norm": 3.4568546688336435,
+      "learning_rate": 8.161603415289964e-06,
+      "loss": 1.2533,
+      "step": 5943
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "grad_norm": 3.4068916522800383,
+      "learning_rate": 8.158538748724139e-06,
+      "loss": 1.2816,
+      "step": 5944
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 1.4009,
+      "step": 5944,
+      "vm_loss": 0.246
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 0.9023,
+      "step": 5944,
+      "vm_loss": 0.1045
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 0.953,
+      "step": 5944,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 1.2306,
+      "step": 5944,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 0.9315,
+      "step": 5944,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 1.2075,
+      "step": 5944,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 0.9148,
+      "step": 5944,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.1442596915080492,
+      "lm_loss": 0.5915,
+      "step": 5944,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 1.1444521981856246,
+      "grad_norm": 3.215902215356208,
+      "learning_rate": 8.155474261171902e-06,
+      "loss": 1.2718,
+      "step": 5945
+    },
+    {
+      "epoch": 1.1446447048632,
+      "grad_norm": 3.3448726721315825,
+      "learning_rate": 8.152409952931168e-06,
+      "loss": 1.2588,
+      "step": 5946
+    },
+    {
+      "epoch": 1.1448372115407752,
+      "grad_norm": 3.188109293885731,
+      "learning_rate": 8.149345824299826e-06,
+      "loss": 1.2375,
+      "step": 5947
+    },
+    {
+      "epoch": 1.1450297182183506,
+      "grad_norm": 3.2546810917779494,
+      "learning_rate": 8.146281875575748e-06,
+      "loss": 1.277,
+      "step": 5948
+    },
+    {
+      "epoch": 1.145222224895926,
+      "grad_norm": 3.125567246071508,
+      "learning_rate": 8.14321810705678e-06,
+      "loss": 1.2126,
+      "step": 5949
+    },
+    {
+      "epoch": 1.1454147315735015,
+      "grad_norm": 3.1885004732648343,
+      "learning_rate": 8.140154519040775e-06,
+      "loss": 1.3086,
+      "step": 5950
+    },
+    {
+      "epoch": 1.145607238251077,
+      "grad_norm": 3.3430427028184564,
+      "learning_rate": 8.137091111825541e-06,
+      "loss": 1.371,
+      "step": 5951
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "grad_norm": 3.2094088458729457,
+      "learning_rate": 8.134027885708884e-06,
+      "loss": 1.2888,
+      "step": 5952
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 0.6678,
+      "step": 5952,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 0.9595,
+      "step": 5952,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 1.3242,
+      "step": 5952,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 0.9153,
+      "step": 5952,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 1.095,
+      "step": 5952,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 0.7651,
+      "step": 5952,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 1.0999,
+      "step": 5952,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.1457997449286523,
+      "lm_loss": 0.9255,
+      "step": 5952,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.1459922516062275,
+      "grad_norm": 3.120233319769545,
+      "learning_rate": 8.130964840988593e-06,
+      "loss": 1.223,
+      "step": 5953
+    },
+    {
+      "epoch": 1.146184758283803,
+      "grad_norm": 3.161250321376422,
+      "learning_rate": 8.127901977962428e-06,
+      "loss": 1.239,
+      "step": 5954
+    },
+    {
+      "epoch": 1.1463772649613784,
+      "grad_norm": 3.200221989792111,
+      "learning_rate": 8.124839296928144e-06,
+      "loss": 1.1959,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1465697716389538,
+      "grad_norm": 3.4369949980712606,
+      "learning_rate": 8.121776798183468e-06,
+      "loss": 1.271,
+      "step": 5956
+    },
+    {
+      "epoch": 1.1467622783165292,
+      "grad_norm": 3.3972309079232956,
+      "learning_rate": 8.11871448202612e-06,
+      "loss": 1.3276,
+      "step": 5957
+    },
+    {
+      "epoch": 1.1469547849941044,
+      "grad_norm": 3.5101688273837417,
+      "learning_rate": 8.115652348753791e-06,
+      "loss": 1.2771,
+      "step": 5958
+    },
+    {
+      "epoch": 1.1471472916716798,
+      "grad_norm": 3.3620678161460162,
+      "learning_rate": 8.11259039866416e-06,
+      "loss": 1.2147,
+      "step": 5959
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "grad_norm": 3.1849842004415088,
+      "learning_rate": 8.109528632054893e-06,
+      "loss": 1.2932,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 0.9591,
+      "step": 5960,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.7359,
+      "step": 5960,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.078,
+      "step": 5960,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 0.8653,
+      "step": 5960,
+      "vm_loss": 0.152
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.3356,
+      "step": 5960,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.6872,
+      "step": 5960,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.4369,
+      "step": 5960,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 1.1473397983492553,
+      "lm_loss": 1.3156,
+      "step": 5960,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.1475323050268307,
+      "grad_norm": 3.0988719823340025,
+      "learning_rate": 8.106467049223628e-06,
+      "loss": 1.2428,
+      "step": 5961
+    },
+    {
+      "epoch": 1.147724811704406,
+      "grad_norm": 3.1559597705391473,
+      "learning_rate": 8.103405650467988e-06,
+      "loss": 1.2601,
+      "step": 5962
+    },
+    {
+      "epoch": 1.1479173183819813,
+      "grad_norm": 3.218222551515623,
+      "learning_rate": 8.100344436085588e-06,
+      "loss": 1.2911,
+      "step": 5963
+    },
+    {
+      "epoch": 1.1481098250595567,
+      "grad_norm": 3.188426498694288,
+      "learning_rate": 8.097283406374012e-06,
+      "loss": 1.2724,
+      "step": 5964
+    },
+    {
+      "epoch": 1.1483023317371321,
+      "grad_norm": 3.1882918974843273,
+      "learning_rate": 8.094222561630832e-06,
+      "loss": 1.271,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1484948384147076,
+      "grad_norm": 3.247748487957767,
+      "learning_rate": 8.091161902153594e-06,
+      "loss": 1.2275,
+      "step": 5966
+    },
+    {
+      "epoch": 1.148687345092283,
+      "grad_norm": 3.235754295006389,
+      "learning_rate": 8.088101428239847e-06,
+      "loss": 1.2666,
+      "step": 5967
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "grad_norm": 3.2132228713360225,
+      "learning_rate": 8.085041140187102e-06,
+      "loss": 1.2314,
+      "step": 5968
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 0.7354,
+      "step": 5968,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 1.1644,
+      "step": 5968,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 1.0089,
+      "step": 5968,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 1.3632,
+      "step": 5968,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 1.1275,
+      "step": 5968,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 0.5409,
+      "step": 5968,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 0.6812,
+      "step": 5968,
+      "vm_loss": 0.2174
+    },
+    {
+      "epoch": 1.1488798517698582,
+      "lm_loss": 1.1398,
+      "step": 5968,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 1.1490723584474336,
+      "grad_norm": 3.1845594396762547,
+      "learning_rate": 8.081981038292853e-06,
+      "loss": 1.2379,
+      "step": 5969
+    },
+    {
+      "epoch": 1.149264865125009,
+      "grad_norm": 3.2295798992310862,
+      "learning_rate": 8.07892112285459e-06,
+      "loss": 1.2347,
+      "step": 5970
+    },
+    {
+      "epoch": 1.1494573718025844,
+      "grad_norm": 3.3013468623571653,
+      "learning_rate": 8.07586139416977e-06,
+      "loss": 1.2511,
+      "step": 5971
+    },
+    {
+      "epoch": 1.1496498784801599,
+      "grad_norm": 3.3203794304672014,
+      "learning_rate": 8.07280185253584e-06,
+      "loss": 1.1935,
+      "step": 5972
+    },
+    {
+      "epoch": 1.149842385157735,
+      "grad_norm": 3.338494989573673,
+      "learning_rate": 8.069742498250224e-06,
+      "loss": 1.2398,
+      "step": 5973
+    },
+    {
+      "epoch": 1.1500348918353105,
+      "grad_norm": 3.2671916194226425,
+      "learning_rate": 8.066683331610333e-06,
+      "loss": 1.2229,
+      "step": 5974
+    },
+    {
+      "epoch": 1.150227398512886,
+      "grad_norm": 3.2815867381174817,
+      "learning_rate": 8.06362435291356e-06,
+      "loss": 1.2301,
+      "step": 5975
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "grad_norm": 3.2420687979436207,
+      "learning_rate": 8.060565562457268e-06,
+      "loss": 1.2742,
+      "step": 5976
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 1.4417,
+      "step": 5976,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 1.1356,
+      "step": 5976,
+      "vm_loss": 0.1525
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 1.517,
+      "step": 5976,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 1.6541,
+      "step": 5976,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 1.1276,
+      "step": 5976,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 0.9718,
+      "step": 5976,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 0.9761,
+      "step": 5976,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.1504199051904613,
+      "lm_loss": 0.7615,
+      "step": 5976,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 1.1506124118680368,
+      "grad_norm": 3.217745115199923,
+      "learning_rate": 8.057506960538818e-06,
+      "loss": 1.2809,
+      "step": 5977
+    },
+    {
+      "epoch": 1.150804918545612,
+      "grad_norm": 3.135667888303514,
+      "learning_rate": 8.054448547455546e-06,
+      "loss": 1.2145,
+      "step": 5978
+    },
+    {
+      "epoch": 1.1509974252231874,
+      "grad_norm": 3.1731805177447994,
+      "learning_rate": 8.05139032350476e-06,
+      "loss": 1.2485,
+      "step": 5979
+    },
+    {
+      "epoch": 1.1511899319007628,
+      "grad_norm": 3.3240257737439616,
+      "learning_rate": 8.04833228898377e-06,
+      "loss": 1.2794,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1513824385783382,
+      "grad_norm": 3.339371548614613,
+      "learning_rate": 8.04527444418985e-06,
+      "loss": 1.2404,
+      "step": 5981
+    },
+    {
+      "epoch": 1.1515749452559136,
+      "grad_norm": 3.3516356036188597,
+      "learning_rate": 8.042216789420263e-06,
+      "loss": 1.2525,
+      "step": 5982
+    },
+    {
+      "epoch": 1.1517674519334888,
+      "grad_norm": 3.4654169910620967,
+      "learning_rate": 8.039159324972246e-06,
+      "loss": 1.2676,
+      "step": 5983
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "grad_norm": 3.320542548458265,
+      "learning_rate": 8.036102051143036e-06,
+      "loss": 1.3017,
+      "step": 5984
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.4839,
+      "step": 5984,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.3175,
+      "step": 5984,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.2177,
+      "step": 5984,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 0.8308,
+      "step": 5984,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.1378,
+      "step": 5984,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 0.9164,
+      "step": 5984,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.1588,
+      "step": 5984,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.1519599586110643,
+      "lm_loss": 1.0239,
+      "step": 5984,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.1521524652886397,
+      "grad_norm": 3.5122424690887044,
+      "learning_rate": 8.03304496822983e-06,
+      "loss": 1.3349,
+      "step": 5985
+    },
+    {
+      "epoch": 1.152344971966215,
+      "grad_norm": 3.186790881694689,
+      "learning_rate": 8.029988076529813e-06,
+      "loss": 1.2459,
+      "step": 5986
+    },
+    {
+      "epoch": 1.1525374786437905,
+      "grad_norm": 3.2618014604310317,
+      "learning_rate": 8.026931376340163e-06,
+      "loss": 1.289,
+      "step": 5987
+    },
+    {
+      "epoch": 1.1527299853213657,
+      "grad_norm": 3.2814227532493754,
+      "learning_rate": 8.023874867958027e-06,
+      "loss": 1.2508,
+      "step": 5988
+    },
+    {
+      "epoch": 1.1529224919989411,
+      "grad_norm": 3.2872809651638812,
+      "learning_rate": 8.020818551680536e-06,
+      "loss": 1.2185,
+      "step": 5989
+    },
+    {
+      "epoch": 1.1531149986765166,
+      "grad_norm": 3.111117769936296,
+      "learning_rate": 8.017762427804799e-06,
+      "loss": 1.2512,
+      "step": 5990
+    },
+    {
+      "epoch": 1.153307505354092,
+      "grad_norm": 3.232585397008882,
+      "learning_rate": 8.014706496627919e-06,
+      "loss": 1.2449,
+      "step": 5991
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "grad_norm": 3.4061255355504496,
+      "learning_rate": 8.011650758446967e-06,
+      "loss": 1.2965,
+      "step": 5992
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.1944,
+      "step": 5992,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.3503,
+      "step": 5992,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.222,
+      "step": 5992,
+      "vm_loss": 0.1209
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 0.94,
+      "step": 5992,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.4303,
+      "step": 5992,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 0.6596,
+      "step": 5992,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.3028,
+      "step": 5992,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 1.1535000120316674,
+      "lm_loss": 1.38,
+      "step": 5992,
+      "vm_loss": 0.2221
+    },
+    {
+      "epoch": 1.1536925187092426,
+      "grad_norm": 3.4460640069779753,
+      "learning_rate": 8.008595213558993e-06,
+      "loss": 1.2603,
+      "step": 5993
+    },
+    {
+      "epoch": 1.153885025386818,
+      "grad_norm": 3.1835671318564542,
+      "learning_rate": 8.005539862261048e-06,
+      "loss": 1.1515,
+      "step": 5994
+    },
+    {
+      "epoch": 1.1540775320643935,
+      "grad_norm": 3.2449092447379915,
+      "learning_rate": 8.002484704850142e-06,
+      "loss": 1.2132,
+      "step": 5995
+    },
+    {
+      "epoch": 1.1542700387419689,
+      "grad_norm": 3.4288666135643764,
+      "learning_rate": 7.999429741623278e-06,
+      "loss": 1.3617,
+      "step": 5996
+    },
+    {
+      "epoch": 1.1544625454195443,
+      "grad_norm": 3.072304956450402,
+      "learning_rate": 7.996374972877441e-06,
+      "loss": 1.1793,
+      "step": 5997
+    },
+    {
+      "epoch": 1.1546550520971195,
+      "grad_norm": 3.161630593528054,
+      "learning_rate": 7.993320398909587e-06,
+      "loss": 1.307,
+      "step": 5998
+    },
+    {
+      "epoch": 1.154847558774695,
+      "grad_norm": 3.179616861803642,
+      "learning_rate": 7.990266020016667e-06,
+      "loss": 1.263,
+      "step": 5999
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "grad_norm": 3.047692704123104,
+      "learning_rate": 7.987211836495597e-06,
+      "loss": 1.2191,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.2796,
+      "step": 6000,
+      "vm_loss": 0.2105
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.1836,
+      "step": 6000,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.3955,
+      "step": 6000,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 0.7848,
+      "step": 6000,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 0.8213,
+      "step": 6000,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.0619,
+      "step": 6000,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.3353,
+      "step": 6000,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 1.1550400654522703,
+      "lm_loss": 1.0139,
+      "step": 6000,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.1552325721298458,
+      "grad_norm": 3.2563527807845976,
+      "learning_rate": 7.984157848643292e-06,
+      "loss": 1.2719,
+      "step": 6001
+    },
+    {
+      "epoch": 1.1554250788074212,
+      "grad_norm": 3.2389211499290917,
+      "learning_rate": 7.981104056756633e-06,
+      "loss": 1.2772,
+      "step": 6002
+    },
+    {
+      "epoch": 1.1556175854849966,
+      "grad_norm": 3.2633922537389117,
+      "learning_rate": 7.978050461132486e-06,
+      "loss": 1.2615,
+      "step": 6003
+    },
+    {
+      "epoch": 1.1558100921625718,
+      "grad_norm": 3.2468537424803685,
+      "learning_rate": 7.974997062067707e-06,
+      "loss": 1.2582,
+      "step": 6004
+    },
+    {
+      "epoch": 1.1560025988401472,
+      "grad_norm": 3.1887455191950163,
+      "learning_rate": 7.97194385985912e-06,
+      "loss": 1.2566,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1561951055177226,
+      "grad_norm": 3.215568593250897,
+      "learning_rate": 7.968890854803536e-06,
+      "loss": 1.2216,
+      "step": 6006
+    },
+    {
+      "epoch": 1.156387612195298,
+      "grad_norm": 3.3918618929471656,
+      "learning_rate": 7.965838047197743e-06,
+      "loss": 1.2283,
+      "step": 6007
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "grad_norm": 3.2230562893790213,
+      "learning_rate": 7.96278543733852e-06,
+      "loss": 1.2536,
+      "step": 6008
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 1.6314,
+      "step": 6008,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 0.8353,
+      "step": 6008,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 0.8045,
+      "step": 6008,
+      "vm_loss": 0.1182
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 1.1367,
+      "step": 6008,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 0.9787,
+      "step": 6008,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 1.1859,
+      "step": 6008,
+      "vm_loss": 0.1986
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 0.7178,
+      "step": 6008,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 1.1565801188728735,
+      "lm_loss": 1.0924,
+      "step": 6008,
+      "vm_loss": 0.1008
+    },
+    {
+      "epoch": 1.1567726255504487,
+      "grad_norm": 3.1759366710330186,
+      "learning_rate": 7.959733025522615e-06,
+      "loss": 1.1742,
+      "step": 6009
+    },
+    {
+      "epoch": 1.156965132228024,
+      "grad_norm": 3.194702220347344,
+      "learning_rate": 7.95668081204676e-06,
+      "loss": 1.1728,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1571576389055995,
+      "grad_norm": 3.329988549158313,
+      "learning_rate": 7.953628797207675e-06,
+      "loss": 1.2232,
+      "step": 6011
+    },
+    {
+      "epoch": 1.157350145583175,
+      "grad_norm": 3.2319188046055993,
+      "learning_rate": 7.950576981302053e-06,
+      "loss": 1.199,
+      "step": 6012
+    },
+    {
+      "epoch": 1.1575426522607504,
+      "grad_norm": 3.1842137489807962,
+      "learning_rate": 7.947525364626562e-06,
+      "loss": 1.2606,
+      "step": 6013
+    },
+    {
+      "epoch": 1.1577351589383258,
+      "grad_norm": 3.245136394895245,
+      "learning_rate": 7.944473947477871e-06,
+      "loss": 1.2613,
+      "step": 6014
+    },
+    {
+      "epoch": 1.157927665615901,
+      "grad_norm": 3.1626588182355615,
+      "learning_rate": 7.941422730152607e-06,
+      "loss": 1.246,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "grad_norm": 3.2668529867141807,
+      "learning_rate": 7.938371712947392e-06,
+      "loss": 1.2052,
+      "step": 6016
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 0.6604,
+      "step": 6016,
+      "vm_loss": 0.1233
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 0.7238,
+      "step": 6016,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 0.9639,
+      "step": 6016,
+      "vm_loss": 0.2214
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 0.5416,
+      "step": 6016,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 1.1536,
+      "step": 6016,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 1.4388,
+      "step": 6016,
+      "vm_loss": 0.1133
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 1.2018,
+      "step": 6016,
+      "vm_loss": 0.1152
+    },
+    {
+      "epoch": 1.1581201722934764,
+      "lm_loss": 0.9978,
+      "step": 6016,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 1.1583126789710518,
+      "grad_norm": 3.2749108740168356,
+      "learning_rate": 7.93532089615882e-06,
+      "loss": 1.2607,
+      "step": 6017
+    },
+    {
+      "epoch": 1.1585051856486273,
+      "grad_norm": 3.2824637807279062,
+      "learning_rate": 7.932270280083473e-06,
+      "loss": 1.2124,
+      "step": 6018
+    },
+    {
+      "epoch": 1.1586976923262027,
+      "grad_norm": 3.251809751009011,
+      "learning_rate": 7.92921986501791e-06,
+      "loss": 1.2866,
+      "step": 6019
+    },
+    {
+      "epoch": 1.1588901990037779,
+      "grad_norm": 3.194500187230321,
+      "learning_rate": 7.926169651258668e-06,
+      "loss": 1.2563,
+      "step": 6020
+    },
+    {
+      "epoch": 1.1590827056813533,
+      "grad_norm": 3.2655303838480285,
+      "learning_rate": 7.923119639102268e-06,
+      "loss": 1.2299,
+      "step": 6021
+    },
+    {
+      "epoch": 1.1592752123589287,
+      "grad_norm": 3.0509843282729854,
+      "learning_rate": 7.920069828845211e-06,
+      "loss": 1.1647,
+      "step": 6022
+    },
+    {
+      "epoch": 1.1594677190365041,
+      "grad_norm": 3.137476953712626,
+      "learning_rate": 7.917020220783978e-06,
+      "loss": 1.2431,
+      "step": 6023
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "grad_norm": 3.1925441219277215,
+      "learning_rate": 7.913970815215024e-06,
+      "loss": 1.2213,
+      "step": 6024
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 0.9437,
+      "step": 6024,
+      "vm_loss": 0.1154
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 1.0937,
+      "step": 6024,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 1.3067,
+      "step": 6024,
+      "vm_loss": 0.0873
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 0.9509,
+      "step": 6024,
+      "vm_loss": 0.1001
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 0.9217,
+      "step": 6024,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 1.0277,
+      "step": 6024,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 0.9979,
+      "step": 6024,
+      "vm_loss": 0.2206
+    },
+    {
+      "epoch": 1.1596602257140796,
+      "lm_loss": 0.8125,
+      "step": 6024,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.1598527323916548,
+      "grad_norm": 3.177949817384393,
+      "learning_rate": 7.910921612434799e-06,
+      "loss": 1.268,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1600452390692302,
+      "grad_norm": 3.16292343862426,
+      "learning_rate": 7.90787261273972e-06,
+      "loss": 1.2446,
+      "step": 6026
+    },
+    {
+      "epoch": 1.1602377457468056,
+      "grad_norm": 3.226058390566725,
+      "learning_rate": 7.904823816426185e-06,
+      "loss": 1.166,
+      "step": 6027
+    },
+    {
+      "epoch": 1.160430252424381,
+      "grad_norm": 3.653607627770296,
+      "learning_rate": 7.901775223790585e-06,
+      "loss": 1.2908,
+      "step": 6028
+    },
+    {
+      "epoch": 1.1606227591019564,
+      "grad_norm": 3.395645850098499,
+      "learning_rate": 7.898726835129277e-06,
+      "loss": 1.2342,
+      "step": 6029
+    },
+    {
+      "epoch": 1.1608152657795316,
+      "grad_norm": 3.402042348628373,
+      "learning_rate": 7.895678650738602e-06,
+      "loss": 1.221,
+      "step": 6030
+    },
+    {
+      "epoch": 1.161007772457107,
+      "grad_norm": 3.3067696899484282,
+      "learning_rate": 7.892630670914882e-06,
+      "loss": 1.2748,
+      "step": 6031
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "grad_norm": 3.4454807452681817,
+      "learning_rate": 7.889582895954427e-06,
+      "loss": 1.326,
+      "step": 6032
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 0.9989,
+      "step": 6032,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 0.7525,
+      "step": 6032,
+      "vm_loss": 0.1093
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 0.5834,
+      "step": 6032,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 0.8088,
+      "step": 6032,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 1.1025,
+      "step": 6032,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 1.1073,
+      "step": 6032,
+      "vm_loss": 0.1193
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 0.6936,
+      "step": 6032,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 1.1612002791346825,
+      "lm_loss": 1.0145,
+      "step": 6032,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.161392785812258,
+      "grad_norm": 3.2488163594881927,
+      "learning_rate": 7.886535326153513e-06,
+      "loss": 1.2685,
+      "step": 6033
+    },
+    {
+      "epoch": 1.1615852924898333,
+      "grad_norm": 3.250922761189258,
+      "learning_rate": 7.883487961808401e-06,
+      "loss": 1.2791,
+      "step": 6034
+    },
+    {
+      "epoch": 1.1617777991674085,
+      "grad_norm": 3.1745511658623964,
+      "learning_rate": 7.880440803215341e-06,
+      "loss": 1.2866,
+      "step": 6035
+    },
+    {
+      "epoch": 1.161970305844984,
+      "grad_norm": 3.231440034221781,
+      "learning_rate": 7.87739385067055e-06,
+      "loss": 1.3415,
+      "step": 6036
+    },
+    {
+      "epoch": 1.1621628125225594,
+      "grad_norm": 3.1284013468259535,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 1.2376,
+      "step": 6037
+    },
+    {
+      "epoch": 1.1623553192001348,
+      "grad_norm": 3.2382772646271953,
+      "learning_rate": 7.871300564910576e-06,
+      "loss": 1.1916,
+      "step": 6038
+    },
+    {
+      "epoch": 1.1625478258777102,
+      "grad_norm": 3.2915301591193913,
+      "learning_rate": 7.868254232287733e-06,
+      "loss": 1.229,
+      "step": 6039
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "grad_norm": 3.412554706239831,
+      "learning_rate": 7.865208106897856e-06,
+      "loss": 1.2492,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.5306,
+      "step": 6040,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.3196,
+      "step": 6040,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 0.9985,
+      "step": 6040,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 0.6738,
+      "step": 6040,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.2792,
+      "step": 6040,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.245,
+      "step": 6040,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.3332,
+      "step": 6040,
+      "vm_loss": 0.2609
+    },
+    {
+      "epoch": 1.1627403325552854,
+      "lm_loss": 1.0222,
+      "step": 6040,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.1629328392328608,
+      "grad_norm": 3.52821450340791,
+      "learning_rate": 7.862162189037058e-06,
+      "loss": 1.3276,
+      "step": 6041
+    },
+    {
+      "epoch": 1.1631253459104363,
+      "grad_norm": 3.3089665786852183,
+      "learning_rate": 7.85911647900145e-06,
+      "loss": 1.1754,
+      "step": 6042
+    },
+    {
+      "epoch": 1.1633178525880117,
+      "grad_norm": 3.3398472211853987,
+      "learning_rate": 7.85607097708711e-06,
+      "loss": 1.2402,
+      "step": 6043
+    },
+    {
+      "epoch": 1.163510359265587,
+      "grad_norm": 3.306070251280547,
+      "learning_rate": 7.853025683590096e-06,
+      "loss": 1.2843,
+      "step": 6044
+    },
+    {
+      "epoch": 1.1637028659431623,
+      "grad_norm": 3.2407089086776066,
+      "learning_rate": 7.849980598806459e-06,
+      "loss": 1.2803,
+      "step": 6045
+    },
+    {
+      "epoch": 1.1638953726207377,
+      "grad_norm": 3.219550718434643,
+      "learning_rate": 7.846935723032213e-06,
+      "loss": 1.2197,
+      "step": 6046
+    },
+    {
+      "epoch": 1.1640878792983131,
+      "grad_norm": 3.1097994335672725,
+      "learning_rate": 7.843891056563361e-06,
+      "loss": 1.1946,
+      "step": 6047
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "grad_norm": 3.325837955519314,
+      "learning_rate": 7.840846599695878e-06,
+      "loss": 1.2862,
+      "step": 6048
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 1.158,
+      "step": 6048,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 1.3378,
+      "step": 6048,
+      "vm_loss": 0.1181
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 0.6989,
+      "step": 6048,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 0.8295,
+      "step": 6048,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 0.6564,
+      "step": 6048,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 1.0447,
+      "step": 6048,
+      "vm_loss": 0.1716
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 1.0332,
+      "step": 6048,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 1.1642803859758886,
+      "lm_loss": 1.3833,
+      "step": 6048,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.164472892653464,
+      "grad_norm": 3.2645210940418,
+      "learning_rate": 7.837802352725734e-06,
+      "loss": 1.1994,
+      "step": 6049
+    },
+    {
+      "epoch": 1.1646653993310392,
+      "grad_norm": 3.3228753397949546,
+      "learning_rate": 7.834758315948865e-06,
+      "loss": 1.247,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1648579060086146,
+      "grad_norm": 3.0723297345773775,
+      "learning_rate": 7.831714489661181e-06,
+      "loss": 1.1127,
+      "step": 6051
+    },
+    {
+      "epoch": 1.16505041268619,
+      "grad_norm": 3.1659719455570547,
+      "learning_rate": 7.828670874158598e-06,
+      "loss": 1.1899,
+      "step": 6052
+    },
+    {
+      "epoch": 1.1652429193637655,
+      "grad_norm": 3.33908858568199,
+      "learning_rate": 7.825627469736984e-06,
+      "loss": 1.206,
+      "step": 6053
+    },
+    {
+      "epoch": 1.1654354260413409,
+      "grad_norm": 3.4273867064880825,
+      "learning_rate": 7.822584276692192e-06,
+      "loss": 1.2833,
+      "step": 6054
+    },
+    {
+      "epoch": 1.165627932718916,
+      "grad_norm": 3.251674497138722,
+      "learning_rate": 7.819541295320072e-06,
+      "loss": 1.2254,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "grad_norm": 3.2769478411485764,
+      "learning_rate": 7.816498525916435e-06,
+      "loss": 1.246,
+      "step": 6056
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 0.8814,
+      "step": 6056,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 1.1356,
+      "step": 6056,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 0.9583,
+      "step": 6056,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 1.0268,
+      "step": 6056,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 0.7276,
+      "step": 6056,
+      "vm_loss": 0.2051
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 1.0062,
+      "step": 6056,
+      "vm_loss": 0.2236
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 0.9913,
+      "step": 6056,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.1658204393964915,
+      "lm_loss": 0.646,
+      "step": 6056,
+      "vm_loss": 0.0789
+    },
+    {
+      "epoch": 1.166012946074067,
+      "grad_norm": 3.1425701585453956,
+      "learning_rate": 7.813455968777072e-06,
+      "loss": 1.1428,
+      "step": 6057
+    },
+    {
+      "epoch": 1.1662054527516423,
+      "grad_norm": 3.255369180601749,
+      "learning_rate": 7.810413624197764e-06,
+      "loss": 1.175,
+      "step": 6058
+    },
+    {
+      "epoch": 1.1663979594292178,
+      "grad_norm": 3.164628359517738,
+      "learning_rate": 7.807371492474266e-06,
+      "loss": 1.2214,
+      "step": 6059
+    },
+    {
+      "epoch": 1.166590466106793,
+      "grad_norm": 3.293675121730721,
+      "learning_rate": 7.80432957390231e-06,
+      "loss": 1.2633,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1667829727843684,
+      "grad_norm": 3.2671911191419785,
+      "learning_rate": 7.80128786877761e-06,
+      "loss": 1.2637,
+      "step": 6061
+    },
+    {
+      "epoch": 1.1669754794619438,
+      "grad_norm": 3.2281453304153716,
+      "learning_rate": 7.798246377395858e-06,
+      "loss": 1.2298,
+      "step": 6062
+    },
+    {
+      "epoch": 1.1671679861395192,
+      "grad_norm": 3.1966186018123954,
+      "learning_rate": 7.79520510005273e-06,
+      "loss": 1.2091,
+      "step": 6063
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "grad_norm": 3.3494192065334074,
+      "learning_rate": 7.792164037043874e-06,
+      "loss": 1.2768,
+      "step": 6064
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 1.2898,
+      "step": 6064,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 1.0545,
+      "step": 6064,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 1.0931,
+      "step": 6064,
+      "vm_loss": 0.2095
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 0.8892,
+      "step": 6064,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 0.9904,
+      "step": 6064,
+      "vm_loss": 0.2161
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 0.6541,
+      "step": 6064,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 1.0412,
+      "step": 6064,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.1673604928170946,
+      "lm_loss": 0.9578,
+      "step": 6064,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 1.16755299949467,
+      "grad_norm": 3.3255017161699714,
+      "learning_rate": 7.789123188664914e-06,
+      "loss": 1.3037,
+      "step": 6065
+    },
+    {
+      "epoch": 1.1677455061722453,
+      "grad_norm": 3.2095186649069887,
+      "learning_rate": 7.786082555211471e-06,
+      "loss": 1.2326,
+      "step": 6066
+    },
+    {
+      "epoch": 1.1679380128498207,
+      "grad_norm": 3.116170883234849,
+      "learning_rate": 7.783042136979127e-06,
+      "loss": 1.1927,
+      "step": 6067
+    },
+    {
+      "epoch": 1.168130519527396,
+      "grad_norm": 3.265854493646826,
+      "learning_rate": 7.780001934263447e-06,
+      "loss": 1.2211,
+      "step": 6068
+    },
+    {
+      "epoch": 1.1683230262049715,
+      "grad_norm": 3.235423868772814,
+      "learning_rate": 7.776961947359987e-06,
+      "loss": 1.2312,
+      "step": 6069
+    },
+    {
+      "epoch": 1.168515532882547,
+      "grad_norm": 3.142208526937211,
+      "learning_rate": 7.773922176564264e-06,
+      "loss": 1.2295,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1687080395601221,
+      "grad_norm": 3.248111677667479,
+      "learning_rate": 7.770882622171783e-06,
+      "loss": 1.2616,
+      "step": 6071
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "grad_norm": 3.2229762653269094,
+      "learning_rate": 7.767843284478031e-06,
+      "loss": 1.2146,
+      "step": 6072
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 1.1333,
+      "step": 6072,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 1.0305,
+      "step": 6072,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 0.8646,
+      "step": 6072,
+      "vm_loss": 0.2033
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 1.1849,
+      "step": 6072,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 1.0135,
+      "step": 6072,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 0.8469,
+      "step": 6072,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 0.8382,
+      "step": 6072,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.1689005462376976,
+      "lm_loss": 0.8109,
+      "step": 6072,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.169093052915273,
+      "grad_norm": 3.415748192258437,
+      "learning_rate": 7.764804163778472e-06,
+      "loss": 1.2553,
+      "step": 6073
+    },
+    {
+      "epoch": 1.1692855595928484,
+      "grad_norm": 3.243437880832997,
+      "learning_rate": 7.761765260368546e-06,
+      "loss": 1.2042,
+      "step": 6074
+    },
+    {
+      "epoch": 1.1694780662704238,
+      "grad_norm": 3.2296092720287004,
+      "learning_rate": 7.758726574543665e-06,
+      "loss": 1.1945,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1696705729479993,
+      "grad_norm": 3.3953647599277943,
+      "learning_rate": 7.75568810659924e-06,
+      "loss": 1.3019,
+      "step": 6076
+    },
+    {
+      "epoch": 1.1698630796255745,
+      "grad_norm": 3.180131029246469,
+      "learning_rate": 7.752649856830645e-06,
+      "loss": 1.2322,
+      "step": 6077
+    },
+    {
+      "epoch": 1.1700555863031499,
+      "grad_norm": 3.1661847001069385,
+      "learning_rate": 7.749611825533231e-06,
+      "loss": 1.1808,
+      "step": 6078
+    },
+    {
+      "epoch": 1.1702480929807253,
+      "grad_norm": 3.35343023048648,
+      "learning_rate": 7.746574013002344e-06,
+      "loss": 1.2684,
+      "step": 6079
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "grad_norm": 3.2200019012286556,
+      "learning_rate": 7.743536419533289e-06,
+      "loss": 1.2512,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.0463,
+      "step": 6080,
+      "vm_loss": 0.2373
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 0.9312,
+      "step": 6080,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.09,
+      "step": 6080,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.3708,
+      "step": 6080,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.1168,
+      "step": 6080,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.0453,
+      "step": 6080,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.0371,
+      "step": 6080,
+      "vm_loss": 0.0974
+    },
+    {
+      "epoch": 1.1704405996583007,
+      "lm_loss": 1.1337,
+      "step": 6080,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 1.1706331063358761,
+      "grad_norm": 3.2311523348726685,
+      "learning_rate": 7.740499045421366e-06,
+      "loss": 1.2094,
+      "step": 6081
+    },
+    {
+      "epoch": 1.1708256130134513,
+      "grad_norm": 3.189549567738214,
+      "learning_rate": 7.737461890961841e-06,
+      "loss": 1.1445,
+      "step": 6082
+    },
+    {
+      "epoch": 1.1710181196910268,
+      "grad_norm": 3.399362893807536,
+      "learning_rate": 7.734424956449965e-06,
+      "loss": 1.2159,
+      "step": 6083
+    },
+    {
+      "epoch": 1.1712106263686022,
+      "grad_norm": 3.4431427439200064,
+      "learning_rate": 7.73138824218097e-06,
+      "loss": 1.2297,
+      "step": 6084
+    },
+    {
+      "epoch": 1.1714031330461776,
+      "grad_norm": 3.286141287738985,
+      "learning_rate": 7.728351748450062e-06,
+      "loss": 1.2609,
+      "step": 6085
+    },
+    {
+      "epoch": 1.171595639723753,
+      "grad_norm": 3.3056881954486785,
+      "learning_rate": 7.725315475552428e-06,
+      "loss": 1.2308,
+      "step": 6086
+    },
+    {
+      "epoch": 1.1717881464013282,
+      "grad_norm": 3.2401397785309753,
+      "learning_rate": 7.722279423783231e-06,
+      "loss": 1.2215,
+      "step": 6087
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "grad_norm": 3.3031077875404,
+      "learning_rate": 7.71924359343761e-06,
+      "loss": 1.2569,
+      "step": 6088
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 0.9556,
+      "step": 6088,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.4546,
+      "step": 6088,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.2894,
+      "step": 6088,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.2921,
+      "step": 6088,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.338,
+      "step": 6088,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 0.7538,
+      "step": 6088,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.363,
+      "step": 6088,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 1.1719806530789036,
+      "lm_loss": 1.1648,
+      "step": 6088,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 1.172173159756479,
+      "grad_norm": 3.177163103223128,
+      "learning_rate": 7.716207984810698e-06,
+      "loss": 1.2234,
+      "step": 6089
+    },
+    {
+      "epoch": 1.1723656664340545,
+      "grad_norm": 3.2012564040408074,
+      "learning_rate": 7.713172598197587e-06,
+      "loss": 1.2607,
+      "step": 6090
+    },
+    {
+      "epoch": 1.17255817311163,
+      "grad_norm": 3.171217374893182,
+      "learning_rate": 7.710137433893356e-06,
+      "loss": 1.2587,
+      "step": 6091
+    },
+    {
+      "epoch": 1.172750679789205,
+      "grad_norm": 3.2683926434896677,
+      "learning_rate": 7.707102492193058e-06,
+      "loss": 1.2019,
+      "step": 6092
+    },
+    {
+      "epoch": 1.1729431864667805,
+      "grad_norm": 3.2579314487402984,
+      "learning_rate": 7.704067773391736e-06,
+      "loss": 1.2201,
+      "step": 6093
+    },
+    {
+      "epoch": 1.173135693144356,
+      "grad_norm": 3.1222975049872645,
+      "learning_rate": 7.701033277784402e-06,
+      "loss": 1.2729,
+      "step": 6094
+    },
+    {
+      "epoch": 1.1733281998219314,
+      "grad_norm": 3.1749440792459844,
+      "learning_rate": 7.697999005666041e-06,
+      "loss": 1.2102,
+      "step": 6095
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "grad_norm": 3.200408144892804,
+      "learning_rate": 7.69496495733163e-06,
+      "loss": 1.231,
+      "step": 6096
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 1.315,
+      "step": 6096,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 1.0508,
+      "step": 6096,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 1.3423,
+      "step": 6096,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 0.856,
+      "step": 6096,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 0.7529,
+      "step": 6096,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 1.1407,
+      "step": 6096,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 0.6215,
+      "step": 6096,
+      "vm_loss": 0.119
+    },
+    {
+      "epoch": 1.1735207064995068,
+      "lm_loss": 0.8128,
+      "step": 6096,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.173713213177082,
+      "grad_norm": 3.272116317791104,
+      "learning_rate": 7.691931133076116e-06,
+      "loss": 1.227,
+      "step": 6097
+    },
+    {
+      "epoch": 1.1739057198546574,
+      "grad_norm": 3.2679713191529296,
+      "learning_rate": 7.688897533194423e-06,
+      "loss": 1.187,
+      "step": 6098
+    },
+    {
+      "epoch": 1.1740982265322328,
+      "grad_norm": 3.114035420061861,
+      "learning_rate": 7.68586415798146e-06,
+      "loss": 1.1658,
+      "step": 6099
+    },
+    {
+      "epoch": 1.1742907332098083,
+      "grad_norm": 3.0285070977139186,
+      "learning_rate": 7.682831007732105e-06,
+      "loss": 1.1892,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1744832398873837,
+      "grad_norm": 3.1505328288548453,
+      "learning_rate": 7.679798082741223e-06,
+      "loss": 1.2246,
+      "step": 6101
+    },
+    {
+      "epoch": 1.1746757465649589,
+      "grad_norm": 3.2339088438533117,
+      "learning_rate": 7.676765383303649e-06,
+      "loss": 1.2237,
+      "step": 6102
+    },
+    {
+      "epoch": 1.1748682532425343,
+      "grad_norm": 3.252943676747561,
+      "learning_rate": 7.673732909714204e-06,
+      "loss": 1.2636,
+      "step": 6103
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "grad_norm": 3.269614692316229,
+      "learning_rate": 7.670700662267683e-06,
+      "loss": 1.253,
+      "step": 6104
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 0.7834,
+      "step": 6104,
+      "vm_loss": 0.1104
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 1.1494,
+      "step": 6104,
+      "vm_loss": 0.2235
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 0.6054,
+      "step": 6104,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 1.0462,
+      "step": 6104,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 1.3916,
+      "step": 6104,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 0.9219,
+      "step": 6104,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 1.1053,
+      "step": 6104,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 1.1750607599201097,
+      "lm_loss": 1.1389,
+      "step": 6104,
+      "vm_loss": 0.1266
+    },
+    {
+      "epoch": 1.1752532665976851,
+      "grad_norm": 3.3297778299252276,
+      "learning_rate": 7.667668641258858e-06,
+      "loss": 1.2251,
+      "step": 6105
+    },
+    {
+      "epoch": 1.1754457732752606,
+      "grad_norm": 3.191894568523854,
+      "learning_rate": 7.664636846982476e-06,
+      "loss": 1.2583,
+      "step": 6106
+    },
+    {
+      "epoch": 1.1756382799528358,
+      "grad_norm": 3.4304407409016493,
+      "learning_rate": 7.661605279733275e-06,
+      "loss": 1.3137,
+      "step": 6107
+    },
+    {
+      "epoch": 1.1758307866304112,
+      "grad_norm": 3.3211786463611435,
+      "learning_rate": 7.658573939805958e-06,
+      "loss": 1.2498,
+      "step": 6108
+    },
+    {
+      "epoch": 1.1760232933079866,
+      "grad_norm": 3.2574586041382414,
+      "learning_rate": 7.655542827495205e-06,
+      "loss": 1.2671,
+      "step": 6109
+    },
+    {
+      "epoch": 1.176215799985562,
+      "grad_norm": 3.1480474455918177,
+      "learning_rate": 7.652511943095687e-06,
+      "loss": 1.1949,
+      "step": 6110
+    },
+    {
+      "epoch": 1.1764083066631374,
+      "grad_norm": 3.071867415544505,
+      "learning_rate": 7.649481286902043e-06,
+      "loss": 1.2004,
+      "step": 6111
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "grad_norm": 3.2445520926328917,
+      "learning_rate": 7.646450859208884e-06,
+      "loss": 1.2997,
+      "step": 6112
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 1.3768,
+      "step": 6112,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 1.1704,
+      "step": 6112,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 0.9388,
+      "step": 6112,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 0.405,
+      "step": 6112,
+      "vm_loss": 0.2655
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 0.8671,
+      "step": 6112,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 1.9855,
+      "step": 6112,
+      "vm_loss": 0.2259
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 1.0908,
+      "step": 6112,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 1.1766008133407126,
+      "lm_loss": 0.996,
+      "step": 6112,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 1.176793320018288,
+      "grad_norm": 3.121422349257162,
+      "learning_rate": 7.643420660310817e-06,
+      "loss": 1.2327,
+      "step": 6113
+    },
+    {
+      "epoch": 1.1769858266958635,
+      "grad_norm": 3.1655764602384417,
+      "learning_rate": 7.640390690502412e-06,
+      "loss": 1.2626,
+      "step": 6114
+    },
+    {
+      "epoch": 1.177178333373439,
+      "grad_norm": 3.078722753836169,
+      "learning_rate": 7.637360950078218e-06,
+      "loss": 1.1797,
+      "step": 6115
+    },
+    {
+      "epoch": 1.1773708400510143,
+      "grad_norm": 3.1830917346580567,
+      "learning_rate": 7.634331439332765e-06,
+      "loss": 1.2305,
+      "step": 6116
+    },
+    {
+      "epoch": 1.1775633467285895,
+      "grad_norm": 3.150364731662425,
+      "learning_rate": 7.631302158560564e-06,
+      "loss": 1.2067,
+      "step": 6117
+    },
+    {
+      "epoch": 1.177755853406165,
+      "grad_norm": 3.121972753890976,
+      "learning_rate": 7.628273108056099e-06,
+      "loss": 1.1705,
+      "step": 6118
+    },
+    {
+      "epoch": 1.1779483600837404,
+      "grad_norm": 3.3939227409414996,
+      "learning_rate": 7.625244288113827e-06,
+      "loss": 1.2799,
+      "step": 6119
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "grad_norm": 3.1952704291724796,
+      "learning_rate": 7.622215699028196e-06,
+      "loss": 1.1534,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 1.2247,
+      "step": 6120,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 0.9736,
+      "step": 6120,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 1.1532,
+      "step": 6120,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 1.4124,
+      "step": 6120,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 1.1035,
+      "step": 6120,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 1.3728,
+      "step": 6120,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 0.5252,
+      "step": 6120,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.1781408667613158,
+      "lm_loss": 0.8146,
+      "step": 6120,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 1.1783333734388912,
+      "grad_norm": 3.191346384253806,
+      "learning_rate": 7.619187341093617e-06,
+      "loss": 1.1461,
+      "step": 6121
+    },
+    {
+      "epoch": 1.1785258801164664,
+      "grad_norm": 3.2489607203935673,
+      "learning_rate": 7.616159214604491e-06,
+      "loss": 1.2574,
+      "step": 6122
+    },
+    {
+      "epoch": 1.1787183867940418,
+      "grad_norm": 3.359279493380381,
+      "learning_rate": 7.613131319855185e-06,
+      "loss": 1.2626,
+      "step": 6123
+    },
+    {
+      "epoch": 1.1789108934716173,
+      "grad_norm": 3.2811659969587996,
+      "learning_rate": 7.610103657140052e-06,
+      "loss": 1.2282,
+      "step": 6124
+    },
+    {
+      "epoch": 1.1791034001491927,
+      "grad_norm": 3.293820787791637,
+      "learning_rate": 7.60707622675342e-06,
+      "loss": 1.2547,
+      "step": 6125
+    },
+    {
+      "epoch": 1.179295906826768,
+      "grad_norm": 3.1523615927974533,
+      "learning_rate": 7.604049028989591e-06,
+      "loss": 1.1417,
+      "step": 6126
+    },
+    {
+      "epoch": 1.1794884135043435,
+      "grad_norm": 3.2177422652202012,
+      "learning_rate": 7.601022064142853e-06,
+      "loss": 1.2666,
+      "step": 6127
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "grad_norm": 3.141040953184416,
+      "learning_rate": 7.597995332507462e-06,
+      "loss": 1.2107,
+      "step": 6128
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.1697,
+      "step": 6128,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 0.888,
+      "step": 6128,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.2819,
+      "step": 6128,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.3237,
+      "step": 6128,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.1864,
+      "step": 6128,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 0.794,
+      "step": 6128,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.2782,
+      "step": 6128,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.1796809201819187,
+      "lm_loss": 1.1459,
+      "step": 6128,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 1.1798734268594941,
+      "grad_norm": 3.351149528770669,
+      "learning_rate": 7.594968834377652e-06,
+      "loss": 1.3111,
+      "step": 6129
+    },
+    {
+      "epoch": 1.1800659335370696,
+      "grad_norm": 3.1601437871982014,
+      "learning_rate": 7.591942570047643e-06,
+      "loss": 1.2143,
+      "step": 6130
+    },
+    {
+      "epoch": 1.180258440214645,
+      "grad_norm": 3.4056196629676387,
+      "learning_rate": 7.588916539811626e-06,
+      "loss": 1.2268,
+      "step": 6131
+    },
+    {
+      "epoch": 1.1804509468922204,
+      "grad_norm": 3.323165121755806,
+      "learning_rate": 7.585890743963767e-06,
+      "loss": 1.217,
+      "step": 6132
+    },
+    {
+      "epoch": 1.1806434535697956,
+      "grad_norm": 3.2725761268098443,
+      "learning_rate": 7.582865182798209e-06,
+      "loss": 1.182,
+      "step": 6133
+    },
+    {
+      "epoch": 1.180835960247371,
+      "grad_norm": 3.3592922734893054,
+      "learning_rate": 7.579839856609085e-06,
+      "loss": 1.2307,
+      "step": 6134
+    },
+    {
+      "epoch": 1.1810284669249465,
+      "grad_norm": 3.325530009096257,
+      "learning_rate": 7.576814765690488e-06,
+      "loss": 1.2526,
+      "step": 6135
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "grad_norm": 3.3165829033378604,
+      "learning_rate": 7.573789910336494e-06,
+      "loss": 1.2551,
+      "step": 6136
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 1.3722,
+      "step": 6136,
+      "vm_loss": 0.1144
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 0.7562,
+      "step": 6136,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 0.7106,
+      "step": 6136,
+      "vm_loss": 0.1063
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 0.7439,
+      "step": 6136,
+      "vm_loss": 0.2323
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 1.4404,
+      "step": 6136,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 1.3321,
+      "step": 6136,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 0.9439,
+      "step": 6136,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 1.1812209736025219,
+      "lm_loss": 0.7343,
+      "step": 6136,
+      "vm_loss": 0.1158
+    },
+    {
+      "epoch": 1.1814134802800973,
+      "grad_norm": 3.408472309339248,
+      "learning_rate": 7.570765290841165e-06,
+      "loss": 1.3067,
+      "step": 6137
+    },
+    {
+      "epoch": 1.1816059869576727,
+      "grad_norm": 3.260547138499601,
+      "learning_rate": 7.567740907498529e-06,
+      "loss": 1.296,
+      "step": 6138
+    },
+    {
+      "epoch": 1.181798493635248,
+      "grad_norm": 3.002320892257385,
+      "learning_rate": 7.564716760602595e-06,
+      "loss": 1.1626,
+      "step": 6139
+    },
+    {
+      "epoch": 1.1819910003128233,
+      "grad_norm": 3.2548984469588707,
+      "learning_rate": 7.561692850447343e-06,
+      "loss": 1.1986,
+      "step": 6140
+    },
+    {
+      "epoch": 1.1821835069903988,
+      "grad_norm": 3.0987378878945857,
+      "learning_rate": 7.558669177326745e-06,
+      "loss": 1.1862,
+      "step": 6141
+    },
+    {
+      "epoch": 1.1823760136679742,
+      "grad_norm": 3.0606482658399576,
+      "learning_rate": 7.555645741534736e-06,
+      "loss": 1.1741,
+      "step": 6142
+    },
+    {
+      "epoch": 1.1825685203455496,
+      "grad_norm": 3.1957034570874403,
+      "learning_rate": 7.552622543365233e-06,
+      "loss": 1.2273,
+      "step": 6143
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "grad_norm": 3.3300397456899695,
+      "learning_rate": 7.549599583112131e-06,
+      "loss": 1.2496,
+      "step": 6144
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 1.0396,
+      "step": 6144,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.9948,
+      "step": 6144,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.6277,
+      "step": 6144,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 1.3043,
+      "step": 6144,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.7678,
+      "step": 6144,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.7145,
+      "step": 6144,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.6558,
+      "step": 6144,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 1.1827610270231248,
+      "lm_loss": 0.9492,
+      "step": 6144,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.1829535337007002,
+      "grad_norm": 3.350568048814192,
+      "learning_rate": 7.546576861069297e-06,
+      "loss": 1.191,
+      "step": 6145
+    },
+    {
+      "epoch": 1.1831460403782756,
+      "grad_norm": 3.300789334439735,
+      "learning_rate": 7.543554377530582e-06,
+      "loss": 1.2334,
+      "step": 6146
+    },
+    {
+      "epoch": 1.183338547055851,
+      "grad_norm": 3.3580018816327524,
+      "learning_rate": 7.5405321327898075e-06,
+      "loss": 1.2344,
+      "step": 6147
+    },
+    {
+      "epoch": 1.1835310537334265,
+      "grad_norm": 3.4427880431031794,
+      "learning_rate": 7.537510127140778e-06,
+      "loss": 1.2696,
+      "step": 6148
+    },
+    {
+      "epoch": 1.1837235604110017,
+      "grad_norm": 3.290988138106618,
+      "learning_rate": 7.5344883608772675e-06,
+      "loss": 1.2615,
+      "step": 6149
+    },
+    {
+      "epoch": 1.183916067088577,
+      "grad_norm": 3.368063465406094,
+      "learning_rate": 7.531466834293029e-06,
+      "loss": 1.2394,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1841085737661525,
+      "grad_norm": 3.3293151424403273,
+      "learning_rate": 7.5284455476818e-06,
+      "loss": 1.2465,
+      "step": 6151
+    },
+    {
+      "epoch": 1.184301080443728,
+      "grad_norm": 3.1964368913305514,
+      "learning_rate": 7.525424501337284e-06,
+      "loss": 1.1736,
+      "step": 6152
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 1.0186,
+      "step": 6152,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 0.9387,
+      "step": 6152,
+      "vm_loss": 0.1977
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 1.0602,
+      "step": 6152,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 0.8528,
+      "step": 6152,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 1.2751,
+      "step": 6152,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 1.1112,
+      "step": 6152,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 1.018,
+      "step": 6152,
+      "vm_loss": 0.1174
+    },
+    {
+      "epoch": 1.184301080443728,
+      "lm_loss": 0.8559,
+      "step": 6152,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 1.1844935871213034,
+      "grad_norm": 3.1730839466709466,
+      "learning_rate": 7.52240369555316e-06,
+      "loss": 1.1976,
+      "step": 6153
+    },
+    {
+      "epoch": 1.1846860937988786,
+      "grad_norm": 3.1648035431815815,
+      "learning_rate": 7.5193831306231e-06,
+      "loss": 1.1902,
+      "step": 6154
+    },
+    {
+      "epoch": 1.184878600476454,
+      "grad_norm": 3.045988424538661,
+      "learning_rate": 7.516362806840736e-06,
+      "loss": 1.1579,
+      "step": 6155
+    },
+    {
+      "epoch": 1.1850711071540294,
+      "grad_norm": 3.261703573080538,
+      "learning_rate": 7.513342724499682e-06,
+      "loss": 1.2734,
+      "step": 6156
+    },
+    {
+      "epoch": 1.1852636138316048,
+      "grad_norm": 3.163168691374363,
+      "learning_rate": 7.510322883893524e-06,
+      "loss": 1.1692,
+      "step": 6157
+    },
+    {
+      "epoch": 1.1854561205091803,
+      "grad_norm": 3.3998879838271527,
+      "learning_rate": 7.507303285315839e-06,
+      "loss": 1.186,
+      "step": 6158
+    },
+    {
+      "epoch": 1.1856486271867555,
+      "grad_norm": 3.326055087265486,
+      "learning_rate": 7.5042839290601655e-06,
+      "loss": 1.1941,
+      "step": 6159
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "grad_norm": 3.2872318799370204,
+      "learning_rate": 7.501264815420019e-06,
+      "loss": 1.2264,
+      "step": 6160
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.0201,
+      "step": 6160,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.1475,
+      "step": 6160,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.2055,
+      "step": 6160,
+      "vm_loss": 0.1346
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.5111,
+      "step": 6160,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 0.9498,
+      "step": 6160,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 0.9566,
+      "step": 6160,
+      "vm_loss": 0.1018
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.1052,
+      "step": 6160,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 1.1858411338643309,
+      "lm_loss": 1.1775,
+      "step": 6160,
+      "vm_loss": 0.1006
+    },
+    {
+      "epoch": 1.1860336405419063,
+      "grad_norm": 3.354725456308886,
+      "learning_rate": 7.498245944688904e-06,
+      "loss": 1.2821,
+      "step": 6161
+    },
+    {
+      "epoch": 1.1862261472194817,
+      "grad_norm": 3.3477471998239663,
+      "learning_rate": 7.495227317160292e-06,
+      "loss": 1.2759,
+      "step": 6162
+    },
+    {
+      "epoch": 1.1864186538970571,
+      "grad_norm": 3.2420694946380273,
+      "learning_rate": 7.492208933127625e-06,
+      "loss": 1.207,
+      "step": 6163
+    },
+    {
+      "epoch": 1.1866111605746323,
+      "grad_norm": 3.234054388727203,
+      "learning_rate": 7.489190792884338e-06,
+      "loss": 1.187,
+      "step": 6164
+    },
+    {
+      "epoch": 1.1868036672522078,
+      "grad_norm": 3.295862072446685,
+      "learning_rate": 7.486172896723826e-06,
+      "loss": 1.2846,
+      "step": 6165
+    },
+    {
+      "epoch": 1.1869961739297832,
+      "grad_norm": 3.167120671281481,
+      "learning_rate": 7.483155244939471e-06,
+      "loss": 1.1728,
+      "step": 6166
+    },
+    {
+      "epoch": 1.1871886806073586,
+      "grad_norm": 3.1183057937221235,
+      "learning_rate": 7.480137837824626e-06,
+      "loss": 1.1969,
+      "step": 6167
+    },
+    {
+      "epoch": 1.187381187284934,
+      "grad_norm": 3.321682198204172,
+      "learning_rate": 7.47712067567262e-06,
+      "loss": 1.2728,
+      "step": 6168
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 1.1225,
+      "step": 6168,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 1.4882,
+      "step": 6168,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 0.8208,
+      "step": 6168,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 1.0842,
+      "step": 6168,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 0.8446,
+      "step": 6168,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 0.8169,
+      "step": 6168,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 0.8156,
+      "step": 6168,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.187381187284934,
+      "lm_loss": 0.888,
+      "step": 6168,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 1.1875736939625092,
+      "grad_norm": 3.491126090797365,
+      "learning_rate": 7.474103758776764e-06,
+      "loss": 1.2885,
+      "step": 6169
+    },
+    {
+      "epoch": 1.1877662006400846,
+      "grad_norm": 3.323535183665952,
+      "learning_rate": 7.471087087430334e-06,
+      "loss": 1.1791,
+      "step": 6170
+    },
+    {
+      "epoch": 1.18795870731766,
+      "grad_norm": 3.236288698500348,
+      "learning_rate": 7.468070661926599e-06,
+      "loss": 1.1956,
+      "step": 6171
+    },
+    {
+      "epoch": 1.1881512139952355,
+      "grad_norm": 3.2000631102171253,
+      "learning_rate": 7.4650544825587865e-06,
+      "loss": 1.2091,
+      "step": 6172
+    },
+    {
+      "epoch": 1.188343720672811,
+      "grad_norm": 3.379853037894804,
+      "learning_rate": 7.46203854962011e-06,
+      "loss": 1.2954,
+      "step": 6173
+    },
+    {
+      "epoch": 1.188536227350386,
+      "grad_norm": 3.2048383490268413,
+      "learning_rate": 7.459022863403753e-06,
+      "loss": 1.2266,
+      "step": 6174
+    },
+    {
+      "epoch": 1.1887287340279615,
+      "grad_norm": 3.126968254247338,
+      "learning_rate": 7.456007424202886e-06,
+      "loss": 1.2586,
+      "step": 6175
+    },
+    {
+      "epoch": 1.188921240705537,
+      "grad_norm": 3.2755282263856302,
+      "learning_rate": 7.452992232310645e-06,
+      "loss": 1.216,
+      "step": 6176
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 1.1578,
+      "step": 6176,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 1.0037,
+      "step": 6176,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 0.9248,
+      "step": 6176,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 1.1977,
+      "step": 6176,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 1.8386,
+      "step": 6176,
+      "vm_loss": 0.2537
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 0.8929,
+      "step": 6176,
+      "vm_loss": 0.2167
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 0.5072,
+      "step": 6176,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 1.188921240705537,
+      "lm_loss": 0.8753,
+      "step": 6176,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 1.1891137473831124,
+      "grad_norm": 3.16222103503267,
+      "learning_rate": 7.44997728802014e-06,
+      "loss": 1.228,
+      "step": 6177
+    },
+    {
+      "epoch": 1.1893062540606878,
+      "grad_norm": 3.2936959025738166,
+      "learning_rate": 7.446962591624471e-06,
+      "loss": 1.2465,
+      "step": 6178
+    },
+    {
+      "epoch": 1.189498760738263,
+      "grad_norm": 3.283102256055236,
+      "learning_rate": 7.443948143416702e-06,
+      "loss": 1.2371,
+      "step": 6179
+    },
+    {
+      "epoch": 1.1896912674158384,
+      "grad_norm": 3.3894301224994035,
+      "learning_rate": 7.440933943689874e-06,
+      "loss": 1.2532,
+      "step": 6180
+    },
+    {
+      "epoch": 1.1898837740934138,
+      "grad_norm": 3.2334481748368638,
+      "learning_rate": 7.437919992737003e-06,
+      "loss": 1.1663,
+      "step": 6181
+    },
+    {
+      "epoch": 1.1900762807709893,
+      "grad_norm": 3.2284119074789586,
+      "learning_rate": 7.434906290851091e-06,
+      "loss": 1.1445,
+      "step": 6182
+    },
+    {
+      "epoch": 1.1902687874485647,
+      "grad_norm": 3.277937001283184,
+      "learning_rate": 7.4318928383251035e-06,
+      "loss": 1.1553,
+      "step": 6183
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "grad_norm": 3.3448301168805092,
+      "learning_rate": 7.428879635451986e-06,
+      "loss": 1.1992,
+      "step": 6184
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 1.2329,
+      "step": 6184,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 1.2779,
+      "step": 6184,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 0.923,
+      "step": 6184,
+      "vm_loss": 0.1278
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 0.7622,
+      "step": 6184,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 1.171,
+      "step": 6184,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 1.2231,
+      "step": 6184,
+      "vm_loss": 0.0985
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 1.0454,
+      "step": 6184,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.1904612941261399,
+      "lm_loss": 0.9305,
+      "step": 6184,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.1906538008037153,
+      "grad_norm": 3.2702402175724035,
+      "learning_rate": 7.425866682524665e-06,
+      "loss": 1.1968,
+      "step": 6185
+    },
+    {
+      "epoch": 1.1908463074812907,
+      "grad_norm": 3.272444529644762,
+      "learning_rate": 7.422853979836035e-06,
+      "loss": 1.1913,
+      "step": 6186
+    },
+    {
+      "epoch": 1.1910388141588661,
+      "grad_norm": 3.3329941529459552,
+      "learning_rate": 7.419841527678967e-06,
+      "loss": 1.2529,
+      "step": 6187
+    },
+    {
+      "epoch": 1.1912313208364416,
+      "grad_norm": 3.106954940272689,
+      "learning_rate": 7.416829326346313e-06,
+      "loss": 1.175,
+      "step": 6188
+    },
+    {
+      "epoch": 1.191423827514017,
+      "grad_norm": 3.162673809376489,
+      "learning_rate": 7.413817376130901e-06,
+      "loss": 1.1783,
+      "step": 6189
+    },
+    {
+      "epoch": 1.1916163341915922,
+      "grad_norm": 3.2304730104542565,
+      "learning_rate": 7.4108056773255245e-06,
+      "loss": 1.2486,
+      "step": 6190
+    },
+    {
+      "epoch": 1.1918088408691676,
+      "grad_norm": 3.3115602211061734,
+      "learning_rate": 7.407794230222958e-06,
+      "loss": 1.2557,
+      "step": 6191
+    },
+    {
+      "epoch": 1.192001347546743,
+      "grad_norm": 3.2728645813007744,
+      "learning_rate": 7.404783035115961e-06,
+      "loss": 1.1979,
+      "step": 6192
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 1.2222,
+      "step": 6192,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 0.6885,
+      "step": 6192,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 1.0615,
+      "step": 6192,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 0.6434,
+      "step": 6192,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 0.7688,
+      "step": 6192,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 1.1392,
+      "step": 6192,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 0.5012,
+      "step": 6192,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 1.192001347546743,
+      "lm_loss": 0.7217,
+      "step": 6192,
+      "vm_loss": 0.1183
+    },
+    {
+      "epoch": 1.1921938542243185,
+      "grad_norm": 3.125565278870283,
+      "learning_rate": 7.401772092297254e-06,
+      "loss": 1.1268,
+      "step": 6193
+    },
+    {
+      "epoch": 1.1923863609018939,
+      "grad_norm": 3.3155124403649756,
+      "learning_rate": 7.3987614020595385e-06,
+      "loss": 1.1991,
+      "step": 6194
+    },
+    {
+      "epoch": 1.192578867579469,
+      "grad_norm": 3.236628006851464,
+      "learning_rate": 7.395750964695499e-06,
+      "loss": 1.2418,
+      "step": 6195
+    },
+    {
+      "epoch": 1.1927713742570445,
+      "grad_norm": 3.549167723460142,
+      "learning_rate": 7.3927407804977814e-06,
+      "loss": 1.2797,
+      "step": 6196
+    },
+    {
+      "epoch": 1.19296388093462,
+      "grad_norm": 3.1920525656594854,
+      "learning_rate": 7.389730849759018e-06,
+      "loss": 1.2399,
+      "step": 6197
+    },
+    {
+      "epoch": 1.1931563876121953,
+      "grad_norm": 3.1733956925620603,
+      "learning_rate": 7.386721172771807e-06,
+      "loss": 1.2363,
+      "step": 6198
+    },
+    {
+      "epoch": 1.1933488942897708,
+      "grad_norm": 3.176200652429145,
+      "learning_rate": 7.383711749828735e-06,
+      "loss": 1.183,
+      "step": 6199
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "grad_norm": 3.2896289527802676,
+      "learning_rate": 7.380702581222353e-06,
+      "loss": 1.2697,
+      "step": 6200
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 1.2398,
+      "step": 6200,
+      "vm_loss": 0.1035
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 1.7413,
+      "step": 6200,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 1.31,
+      "step": 6200,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 1.4241,
+      "step": 6200,
+      "vm_loss": 0.2191
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 0.7958,
+      "step": 6200,
+      "vm_loss": 0.1932
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 0.9645,
+      "step": 6200,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 0.5985,
+      "step": 6200,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.1935414009673462,
+      "lm_loss": 1.2084,
+      "step": 6200,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.1937339076449214,
+      "grad_norm": 3.2747695138186126,
+      "learning_rate": 7.377693667245187e-06,
+      "loss": 1.2394,
+      "step": 6201
+    },
+    {
+      "epoch": 1.1939264143224968,
+      "grad_norm": 3.443711273539638,
+      "learning_rate": 7.37468500818975e-06,
+      "loss": 1.2095,
+      "step": 6202
+    },
+    {
+      "epoch": 1.1941189210000722,
+      "grad_norm": 3.2764875825710256,
+      "learning_rate": 7.371676604348516e-06,
+      "loss": 1.2377,
+      "step": 6203
+    },
+    {
+      "epoch": 1.1943114276776476,
+      "grad_norm": 3.2969254826835095,
+      "learning_rate": 7.368668456013941e-06,
+      "loss": 1.2586,
+      "step": 6204
+    },
+    {
+      "epoch": 1.194503934355223,
+      "grad_norm": 3.318369219201361,
+      "learning_rate": 7.365660563478458e-06,
+      "loss": 1.2444,
+      "step": 6205
+    },
+    {
+      "epoch": 1.1946964410327983,
+      "grad_norm": 3.189390488154652,
+      "learning_rate": 7.362652927034474e-06,
+      "loss": 1.1092,
+      "step": 6206
+    },
+    {
+      "epoch": 1.1948889477103737,
+      "grad_norm": 3.290045806810415,
+      "learning_rate": 7.359645546974364e-06,
+      "loss": 1.2321,
+      "step": 6207
+    },
+    {
+      "epoch": 1.195081454387949,
+      "grad_norm": 3.0659480184230765,
+      "learning_rate": 7.3566384235904855e-06,
+      "loss": 1.147,
+      "step": 6208
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 1.2341,
+      "step": 6208,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 1.027,
+      "step": 6208,
+      "vm_loss": 0.1184
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 0.6748,
+      "step": 6208,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 1.2132,
+      "step": 6208,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 0.9857,
+      "step": 6208,
+      "vm_loss": 0.0964
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 0.8908,
+      "step": 6208,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 1.1228,
+      "step": 6208,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 1.195081454387949,
+      "lm_loss": 1.0035,
+      "step": 6208,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.1952739610655245,
+      "grad_norm": 3.3454839972767085,
+      "learning_rate": 7.353631557175172e-06,
+      "loss": 1.235,
+      "step": 6209
+    },
+    {
+      "epoch": 1.1954664677431,
+      "grad_norm": 3.154619047150957,
+      "learning_rate": 7.3506249480207284e-06,
+      "loss": 1.1767,
+      "step": 6210
+    },
+    {
+      "epoch": 1.1956589744206751,
+      "grad_norm": 3.2720044374901507,
+      "learning_rate": 7.3476185964194345e-06,
+      "loss": 1.1863,
+      "step": 6211
+    },
+    {
+      "epoch": 1.1958514810982506,
+      "grad_norm": 3.2507692137042667,
+      "learning_rate": 7.344612502663548e-06,
+      "loss": 1.2062,
+      "step": 6212
+    },
+    {
+      "epoch": 1.196043987775826,
+      "grad_norm": 3.2641337513735436,
+      "learning_rate": 7.341606667045299e-06,
+      "loss": 1.2614,
+      "step": 6213
+    },
+    {
+      "epoch": 1.1962364944534014,
+      "grad_norm": 3.250203841808071,
+      "learning_rate": 7.338601089856894e-06,
+      "loss": 1.2484,
+      "step": 6214
+    },
+    {
+      "epoch": 1.1964290011309768,
+      "grad_norm": 3.2265894007170215,
+      "learning_rate": 7.3355957713905065e-06,
+      "loss": 1.1716,
+      "step": 6215
+    },
+    {
+      "epoch": 1.196621507808552,
+      "grad_norm": 3.169694952108079,
+      "learning_rate": 7.332590711938303e-06,
+      "loss": 1.1967,
+      "step": 6216
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 0.7031,
+      "step": 6216,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 0.9983,
+      "step": 6216,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 1.1673,
+      "step": 6216,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 0.9241,
+      "step": 6216,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 0.6495,
+      "step": 6216,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 1.2762,
+      "step": 6216,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 1.0543,
+      "step": 6216,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.196621507808552,
+      "lm_loss": 1.0742,
+      "step": 6216,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 1.1968140144861275,
+      "grad_norm": 3.2277696827229225,
+      "learning_rate": 7.329585911792407e-06,
+      "loss": 1.1916,
+      "step": 6217
+    },
+    {
+      "epoch": 1.1970065211637029,
+      "grad_norm": 3.163722348409198,
+      "learning_rate": 7.326581371244923e-06,
+      "loss": 1.154,
+      "step": 6218
+    },
+    {
+      "epoch": 1.1971990278412783,
+      "grad_norm": 3.248891868485399,
+      "learning_rate": 7.323577090587936e-06,
+      "loss": 1.1905,
+      "step": 6219
+    },
+    {
+      "epoch": 1.1973915345188537,
+      "grad_norm": 3.227041808296925,
+      "learning_rate": 7.320573070113497e-06,
+      "loss": 1.2002,
+      "step": 6220
+    },
+    {
+      "epoch": 1.197584041196429,
+      "grad_norm": 3.21744906848506,
+      "learning_rate": 7.31756931011363e-06,
+      "loss": 1.1798,
+      "step": 6221
+    },
+    {
+      "epoch": 1.1977765478740043,
+      "grad_norm": 3.212013137666816,
+      "learning_rate": 7.314565810880351e-06,
+      "loss": 1.1972,
+      "step": 6222
+    },
+    {
+      "epoch": 1.1979690545515798,
+      "grad_norm": 3.2475371789276353,
+      "learning_rate": 7.311562572705631e-06,
+      "loss": 1.1855,
+      "step": 6223
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "grad_norm": 3.4267964686402532,
+      "learning_rate": 7.308559595881424e-06,
+      "loss": 1.2382,
+      "step": 6224
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 1.1657,
+      "step": 6224,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 0.9857,
+      "step": 6224,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 1.2079,
+      "step": 6224,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 1.3074,
+      "step": 6224,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 0.9671,
+      "step": 6224,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 1.5091,
+      "step": 6224,
+      "vm_loss": 0.2195
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 0.7788,
+      "step": 6224,
+      "vm_loss": 0.1169
+    },
+    {
+      "epoch": 1.1981615612291552,
+      "lm_loss": 0.6039,
+      "step": 6224,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 1.1983540679067306,
+      "grad_norm": 3.299078269091446,
+      "learning_rate": 7.305556880699654e-06,
+      "loss": 1.1843,
+      "step": 6225
+    },
+    {
+      "epoch": 1.1985465745843058,
+      "grad_norm": 3.320527666963726,
+      "learning_rate": 7.302554427452231e-06,
+      "loss": 1.2126,
+      "step": 6226
+    },
+    {
+      "epoch": 1.1987390812618812,
+      "grad_norm": 3.1985635946986153,
+      "learning_rate": 7.299552236431025e-06,
+      "loss": 1.2148,
+      "step": 6227
+    },
+    {
+      "epoch": 1.1989315879394566,
+      "grad_norm": 3.275271839408709,
+      "learning_rate": 7.296550307927892e-06,
+      "loss": 1.1551,
+      "step": 6228
+    },
+    {
+      "epoch": 1.199124094617032,
+      "grad_norm": 3.1898739436906096,
+      "learning_rate": 7.293548642234658e-06,
+      "loss": 1.1853,
+      "step": 6229
+    },
+    {
+      "epoch": 1.1993166012946075,
+      "grad_norm": 3.3066667012176336,
+      "learning_rate": 7.290547239643117e-06,
+      "loss": 1.1874,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1995091079721827,
+      "grad_norm": 3.343224828428256,
+      "learning_rate": 7.287546100445054e-06,
+      "loss": 1.2729,
+      "step": 6231
+    },
+    {
+      "epoch": 1.199701614649758,
+      "grad_norm": 3.285667226862546,
+      "learning_rate": 7.284545224932207e-06,
+      "loss": 1.2143,
+      "step": 6232
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 1.0936,
+      "step": 6232,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 0.8895,
+      "step": 6232,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 1.177,
+      "step": 6232,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 0.859,
+      "step": 6232,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 1.0927,
+      "step": 6232,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 1.0089,
+      "step": 6232,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 0.954,
+      "step": 6232,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 1.199701614649758,
+      "lm_loss": 1.2182,
+      "step": 6232,
+      "vm_loss": 0.0866
+    },
+    {
+      "epoch": 1.1998941213273335,
+      "grad_norm": 3.197548288639276,
+      "learning_rate": 7.281544613396309e-06,
+      "loss": 1.1323,
+      "step": 6233
+    },
+    {
+      "epoch": 1.200086628004909,
+      "grad_norm": 3.3397984686715962,
+      "learning_rate": 7.278544266129054e-06,
+      "loss": 1.2801,
+      "step": 6234
+    },
+    {
+      "epoch": 1.2002791346824844,
+      "grad_norm": 3.31671961022355,
+      "learning_rate": 7.2755441834221095e-06,
+      "loss": 1.2424,
+      "step": 6235
+    },
+    {
+      "epoch": 1.2004716413600596,
+      "grad_norm": 3.2645689585151794,
+      "learning_rate": 7.272544365567131e-06,
+      "loss": 1.1644,
+      "step": 6236
+    },
+    {
+      "epoch": 1.200664148037635,
+      "grad_norm": 3.2489844026446018,
+      "learning_rate": 7.269544812855734e-06,
+      "loss": 1.2024,
+      "step": 6237
+    },
+    {
+      "epoch": 1.2008566547152104,
+      "grad_norm": 3.3981200333694983,
+      "learning_rate": 7.266545525579509e-06,
+      "loss": 1.2401,
+      "step": 6238
+    },
+    {
+      "epoch": 1.2010491613927858,
+      "grad_norm": 3.188410754158432,
+      "learning_rate": 7.263546504030036e-06,
+      "loss": 1.1297,
+      "step": 6239
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "grad_norm": 3.25408926477826,
+      "learning_rate": 7.260547748498851e-06,
+      "loss": 1.2461,
+      "step": 6240
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 0.7198,
+      "step": 6240,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 1.2332,
+      "step": 6240,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 0.885,
+      "step": 6240,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 1.0609,
+      "step": 6240,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 0.6779,
+      "step": 6240,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 1.2458,
+      "step": 6240,
+      "vm_loss": 0.2552
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 0.8837,
+      "step": 6240,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 1.2012416680703613,
+      "lm_loss": 0.9928,
+      "step": 6240,
+      "vm_loss": 0.2173
+    },
+    {
+      "epoch": 1.2014341747479365,
+      "grad_norm": 3.3782687461738488,
+      "learning_rate": 7.257549259277472e-06,
+      "loss": 1.2634,
+      "step": 6241
+    },
+    {
+      "epoch": 1.2016266814255119,
+      "grad_norm": 3.340928059291423,
+      "learning_rate": 7.254551036657389e-06,
+      "loss": 1.2107,
+      "step": 6242
+    },
+    {
+      "epoch": 1.2018191881030873,
+      "grad_norm": 3.2448429759403434,
+      "learning_rate": 7.251553080930073e-06,
+      "loss": 1.1888,
+      "step": 6243
+    },
+    {
+      "epoch": 1.2020116947806627,
+      "grad_norm": 3.214846528176014,
+      "learning_rate": 7.24855539238696e-06,
+      "loss": 1.1641,
+      "step": 6244
+    },
+    {
+      "epoch": 1.2022042014582381,
+      "grad_norm": 3.2422322180149292,
+      "learning_rate": 7.245557971319461e-06,
+      "loss": 1.2084,
+      "step": 6245
+    },
+    {
+      "epoch": 1.2023967081358133,
+      "grad_norm": 3.2217925250756676,
+      "learning_rate": 7.24256081801897e-06,
+      "loss": 1.1843,
+      "step": 6246
+    },
+    {
+      "epoch": 1.2025892148133888,
+      "grad_norm": 3.2929749335830274,
+      "learning_rate": 7.239563932776845e-06,
+      "loss": 1.1837,
+      "step": 6247
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "grad_norm": 3.2701090684026175,
+      "learning_rate": 7.236567315884422e-06,
+      "loss": 1.2006,
+      "step": 6248
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.3905,
+      "step": 6248,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.171,
+      "step": 6248,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.0381,
+      "step": 6248,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.1331,
+      "step": 6248,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.5097,
+      "step": 6248,
+      "vm_loss": 0.1286
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 0.9433,
+      "step": 6248,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.1767,
+      "step": 6248,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.2027817214909642,
+      "lm_loss": 1.4988,
+      "step": 6248,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.2029742281685396,
+      "grad_norm": 3.310512096841856,
+      "learning_rate": 7.233570967633008e-06,
+      "loss": 1.1829,
+      "step": 6249
+    },
+    {
+      "epoch": 1.203166734846115,
+      "grad_norm": 3.2115793111578674,
+      "learning_rate": 7.23057488831389e-06,
+      "loss": 1.1287,
+      "step": 6250
+    },
+    {
+      "epoch": 1.2033592415236904,
+      "grad_norm": 3.4209483303028434,
+      "learning_rate": 7.227579078218328e-06,
+      "loss": 1.2503,
+      "step": 6251
+    },
+    {
+      "epoch": 1.2035517482012656,
+      "grad_norm": 3.3349988269434965,
+      "learning_rate": 7.224583537637544e-06,
+      "loss": 1.1451,
+      "step": 6252
+    },
+    {
+      "epoch": 1.203744254878841,
+      "grad_norm": 3.363070437375981,
+      "learning_rate": 7.221588266862749e-06,
+      "loss": 1.2628,
+      "step": 6253
+    },
+    {
+      "epoch": 1.2039367615564165,
+      "grad_norm": 3.1535084147021806,
+      "learning_rate": 7.218593266185125e-06,
+      "loss": 1.1363,
+      "step": 6254
+    },
+    {
+      "epoch": 1.204129268233992,
+      "grad_norm": 3.1279933342989863,
+      "learning_rate": 7.215598535895814e-06,
+      "loss": 1.1468,
+      "step": 6255
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "grad_norm": 3.090503879220856,
+      "learning_rate": 7.212604076285952e-06,
+      "loss": 1.1315,
+      "step": 6256
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.673,
+      "step": 6256,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.0383,
+      "step": 6256,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.2368,
+      "step": 6256,
+      "vm_loss": 0.1298
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 0.9823,
+      "step": 6256,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.2037,
+      "step": 6256,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.0105,
+      "step": 6256,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.2091,
+      "step": 6256,
+      "vm_loss": 0.1059
+    },
+    {
+      "epoch": 1.2043217749115673,
+      "lm_loss": 1.1998,
+      "step": 6256,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.2045142815891425,
+      "grad_norm": 3.1475802534019013,
+      "learning_rate": 7.2096098876466355e-06,
+      "loss": 1.2004,
+      "step": 6257
+    },
+    {
+      "epoch": 1.204706788266718,
+      "grad_norm": 3.190325446464954,
+      "learning_rate": 7.206615970268938e-06,
+      "loss": 1.135,
+      "step": 6258
+    },
+    {
+      "epoch": 1.2048992949442934,
+      "grad_norm": 3.313235756427296,
+      "learning_rate": 7.203622324443902e-06,
+      "loss": 1.2336,
+      "step": 6259
+    },
+    {
+      "epoch": 1.2050918016218688,
+      "grad_norm": 3.323245804434984,
+      "learning_rate": 7.200628950462556e-06,
+      "loss": 1.1946,
+      "step": 6260
+    },
+    {
+      "epoch": 1.2052843082994442,
+      "grad_norm": 3.2728561770340048,
+      "learning_rate": 7.1976358486158895e-06,
+      "loss": 1.2412,
+      "step": 6261
+    },
+    {
+      "epoch": 1.2054768149770196,
+      "grad_norm": 3.408384228137689,
+      "learning_rate": 7.194643019194867e-06,
+      "loss": 1.2359,
+      "step": 6262
+    },
+    {
+      "epoch": 1.2056693216545948,
+      "grad_norm": 3.244930210019856,
+      "learning_rate": 7.191650462490439e-06,
+      "loss": 1.1801,
+      "step": 6263
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "grad_norm": 3.401049784818652,
+      "learning_rate": 7.188658178793516e-06,
+      "loss": 1.1734,
+      "step": 6264
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 1.3469,
+      "step": 6264,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 1.0489,
+      "step": 6264,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 0.9177,
+      "step": 6264,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 0.8576,
+      "step": 6264,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 0.9937,
+      "step": 6264,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 1.0464,
+      "step": 6264,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 0.9471,
+      "step": 6264,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.2058618283321703,
+      "lm_loss": 1.0336,
+      "step": 6264,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.2060543350097457,
+      "grad_norm": 3.235153447841721,
+      "learning_rate": 7.185666168394985e-06,
+      "loss": 1.1642,
+      "step": 6265
+    },
+    {
+      "epoch": 1.206246841687321,
+      "grad_norm": 3.1741085592825358,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 1.1767,
+      "step": 6266
+    },
+    {
+      "epoch": 1.2064393483648965,
+      "grad_norm": 3.107578353071712,
+      "learning_rate": 7.179682968656516e-06,
+      "loss": 1.2062,
+      "step": 6267
+    },
+    {
+      "epoch": 1.2066318550424717,
+      "grad_norm": 3.236103594447218,
+      "learning_rate": 7.1766917798982236e-06,
+      "loss": 1.2483,
+      "step": 6268
+    },
+    {
+      "epoch": 1.2068243617200471,
+      "grad_norm": 3.13447657632835,
+      "learning_rate": 7.1737008656016124e-06,
+      "loss": 1.1698,
+      "step": 6269
+    },
+    {
+      "epoch": 1.2070168683976226,
+      "grad_norm": 3.1977006305842606,
+      "learning_rate": 7.1707102260574376e-06,
+      "loss": 1.1787,
+      "step": 6270
+    },
+    {
+      "epoch": 1.207209375075198,
+      "grad_norm": 3.181931735917009,
+      "learning_rate": 7.1677198615564235e-06,
+      "loss": 1.2092,
+      "step": 6271
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "grad_norm": 3.4251984634650205,
+      "learning_rate": 7.164729772389276e-06,
+      "loss": 1.3033,
+      "step": 6272
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 0.9955,
+      "step": 6272,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 1.0398,
+      "step": 6272,
+      "vm_loss": 0.2003
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 1.0763,
+      "step": 6272,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 1.0323,
+      "step": 6272,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 0.9877,
+      "step": 6272,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 0.9707,
+      "step": 6272,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 1.2018,
+      "step": 6272,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.2074018817527734,
+      "lm_loss": 0.6506,
+      "step": 6272,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 1.2075943884303486,
+      "grad_norm": 3.1262888896545094,
+      "learning_rate": 7.161739958846665e-06,
+      "loss": 1.1531,
+      "step": 6273
+    },
+    {
+      "epoch": 1.207786895107924,
+      "grad_norm": 3.3181949647175104,
+      "learning_rate": 7.158750421219245e-06,
+      "loss": 1.2043,
+      "step": 6274
+    },
+    {
+      "epoch": 1.2079794017854995,
+      "grad_norm": 3.306411371650362,
+      "learning_rate": 7.155761159797635e-06,
+      "loss": 1.1639,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2081719084630749,
+      "grad_norm": 3.344924594958964,
+      "learning_rate": 7.152772174872424e-06,
+      "loss": 1.2305,
+      "step": 6276
+    },
+    {
+      "epoch": 1.2083644151406503,
+      "grad_norm": 3.210511343625126,
+      "learning_rate": 7.14978346673419e-06,
+      "loss": 1.1543,
+      "step": 6277
+    },
+    {
+      "epoch": 1.2085569218182255,
+      "grad_norm": 3.2280306134212373,
+      "learning_rate": 7.146795035673466e-06,
+      "loss": 1.2473,
+      "step": 6278
+    },
+    {
+      "epoch": 1.208749428495801,
+      "grad_norm": 3.113007166867191,
+      "learning_rate": 7.143806881980765e-06,
+      "loss": 1.1412,
+      "step": 6279
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "grad_norm": 3.147264123615081,
+      "learning_rate": 7.14081900594658e-06,
+      "loss": 1.1732,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 1.057,
+      "step": 6280,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 1.1339,
+      "step": 6280,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 0.9748,
+      "step": 6280,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 1.3231,
+      "step": 6280,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 0.9505,
+      "step": 6280,
+      "vm_loss": 0.2016
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 0.9946,
+      "step": 6280,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 0.8218,
+      "step": 6280,
+      "vm_loss": 0.1087
+    },
+    {
+      "epoch": 1.2089419351733763,
+      "lm_loss": 1.046,
+      "step": 6280,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 1.2091344418509518,
+      "grad_norm": 3.204502297688718,
+      "learning_rate": 7.137831407861366e-06,
+      "loss": 1.1991,
+      "step": 6281
+    },
+    {
+      "epoch": 1.2093269485285272,
+      "grad_norm": 3.2239978975507104,
+      "learning_rate": 7.13484408801556e-06,
+      "loss": 1.188,
+      "step": 6282
+    },
+    {
+      "epoch": 1.2095194552061024,
+      "grad_norm": 3.3273701498540675,
+      "learning_rate": 7.131857046699557e-06,
+      "loss": 1.1982,
+      "step": 6283
+    },
+    {
+      "epoch": 1.2097119618836778,
+      "grad_norm": 3.1209299108318946,
+      "learning_rate": 7.12887028420375e-06,
+      "loss": 1.1021,
+      "step": 6284
+    },
+    {
+      "epoch": 1.2099044685612532,
+      "grad_norm": 3.4801166094611258,
+      "learning_rate": 7.125883800818482e-06,
+      "loss": 1.2455,
+      "step": 6285
+    },
+    {
+      "epoch": 1.2100969752388286,
+      "grad_norm": 3.2662891795085245,
+      "learning_rate": 7.122897596834078e-06,
+      "loss": 1.1795,
+      "step": 6286
+    },
+    {
+      "epoch": 1.210289481916404,
+      "grad_norm": 3.493900537356155,
+      "learning_rate": 7.1199116725408375e-06,
+      "loss": 1.2837,
+      "step": 6287
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "grad_norm": 3.503112137751488,
+      "learning_rate": 7.116926028229032e-06,
+      "loss": 1.2832,
+      "step": 6288
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 0.8538,
+      "step": 6288,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 1.2523,
+      "step": 6288,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 0.7868,
+      "step": 6288,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 1.4973,
+      "step": 6288,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 0.7266,
+      "step": 6288,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 0.8101,
+      "step": 6288,
+      "vm_loss": 0.1
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 1.1246,
+      "step": 6288,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.2104819885939793,
+      "lm_loss": 0.8861,
+      "step": 6288,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.2106744952715547,
+      "grad_norm": 3.2547246975430015,
+      "learning_rate": 7.113940664188898e-06,
+      "loss": 1.1725,
+      "step": 6289
+    },
+    {
+      "epoch": 1.21086700194913,
+      "grad_norm": 3.275173385223495,
+      "learning_rate": 7.110955580710656e-06,
+      "loss": 1.1757,
+      "step": 6290
+    },
+    {
+      "epoch": 1.2110595086267055,
+      "grad_norm": 3.108700331568262,
+      "learning_rate": 7.107970778084494e-06,
+      "loss": 1.1669,
+      "step": 6291
+    },
+    {
+      "epoch": 1.211252015304281,
+      "grad_norm": 3.1805530953219705,
+      "learning_rate": 7.1049862566005725e-06,
+      "loss": 1.208,
+      "step": 6292
+    },
+    {
+      "epoch": 1.2114445219818561,
+      "grad_norm": 3.0302930776576313,
+      "learning_rate": 7.102002016549022e-06,
+      "loss": 1.1075,
+      "step": 6293
+    },
+    {
+      "epoch": 1.2116370286594316,
+      "grad_norm": 3.1991494908193823,
+      "learning_rate": 7.099018058219953e-06,
+      "loss": 1.1518,
+      "step": 6294
+    },
+    {
+      "epoch": 1.211829535337007,
+      "grad_norm": 3.2709897240512604,
+      "learning_rate": 7.096034381903446e-06,
+      "loss": 1.2065,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "grad_norm": 3.210550255538842,
+      "learning_rate": 7.093050987889547e-06,
+      "loss": 1.17,
+      "step": 6296
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 0.6208,
+      "step": 6296,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 0.9971,
+      "step": 6296,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 0.7363,
+      "step": 6296,
+      "vm_loss": 0.1936
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 1.2485,
+      "step": 6296,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 1.2022,
+      "step": 6296,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 1.1878,
+      "step": 6296,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 0.9401,
+      "step": 6296,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 1.2120220420145824,
+      "lm_loss": 0.921,
+      "step": 6296,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 1.2122145486921578,
+      "grad_norm": 3.22985518637462,
+      "learning_rate": 7.090067876468283e-06,
+      "loss": 1.1598,
+      "step": 6297
+    },
+    {
+      "epoch": 1.212407055369733,
+      "grad_norm": 3.1225585748900393,
+      "learning_rate": 7.087085047929652e-06,
+      "loss": 1.0889,
+      "step": 6298
+    },
+    {
+      "epoch": 1.2125995620473085,
+      "grad_norm": 3.4134806278107845,
+      "learning_rate": 7.084102502563621e-06,
+      "loss": 1.2457,
+      "step": 6299
+    },
+    {
+      "epoch": 1.2127920687248839,
+      "grad_norm": 3.2778289792966766,
+      "learning_rate": 7.081120240660128e-06,
+      "loss": 1.1413,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2129845754024593,
+      "grad_norm": 3.5682183566647163,
+      "learning_rate": 7.078138262509096e-06,
+      "loss": 1.2339,
+      "step": 6301
+    },
+    {
+      "epoch": 1.2131770820800347,
+      "grad_norm": 3.2795777394160965,
+      "learning_rate": 7.075156568400406e-06,
+      "loss": 1.2264,
+      "step": 6302
+    },
+    {
+      "epoch": 1.21336958875761,
+      "grad_norm": 3.2023166526472666,
+      "learning_rate": 7.072175158623911e-06,
+      "loss": 1.1447,
+      "step": 6303
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "grad_norm": 3.3501058884750163,
+      "learning_rate": 7.069194033469455e-06,
+      "loss": 1.2536,
+      "step": 6304
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 0.6913,
+      "step": 6304,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 1.0513,
+      "step": 6304,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 1.051,
+      "step": 6304,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 1.2587,
+      "step": 6304,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 0.3811,
+      "step": 6304,
+      "vm_loss": 0.2193
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 0.9453,
+      "step": 6304,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 0.5829,
+      "step": 6304,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 1.2135620954351853,
+      "lm_loss": 1.4331,
+      "step": 6304,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 1.2137546021127608,
+      "grad_norm": 3.2077695690202193,
+      "learning_rate": 7.066213193226834e-06,
+      "loss": 1.1831,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2139471087903362,
+      "grad_norm": 3.1771782135477307,
+      "learning_rate": 7.063232638185825e-06,
+      "loss": 1.228,
+      "step": 6306
+    },
+    {
+      "epoch": 1.2141396154679116,
+      "grad_norm": 3.1878522167575043,
+      "learning_rate": 7.060252368636171e-06,
+      "loss": 1.2435,
+      "step": 6307
+    },
+    {
+      "epoch": 1.2143321221454868,
+      "grad_norm": 3.1864072252025073,
+      "learning_rate": 7.057272384867603e-06,
+      "loss": 1.132,
+      "step": 6308
+    },
+    {
+      "epoch": 1.2145246288230622,
+      "grad_norm": 3.2777745304531503,
+      "learning_rate": 7.054292687169806e-06,
+      "loss": 1.247,
+      "step": 6309
+    },
+    {
+      "epoch": 1.2147171355006376,
+      "grad_norm": 3.220336951292951,
+      "learning_rate": 7.051313275832443e-06,
+      "loss": 1.1932,
+      "step": 6310
+    },
+    {
+      "epoch": 1.214909642178213,
+      "grad_norm": 3.3229143777514314,
+      "learning_rate": 7.0483341511451576e-06,
+      "loss": 1.1798,
+      "step": 6311
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "grad_norm": 3.3720909089318867,
+      "learning_rate": 7.045355313397555e-06,
+      "loss": 1.2373,
+      "step": 6312
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 1.4526,
+      "step": 6312,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 1.2553,
+      "step": 6312,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 1.0319,
+      "step": 6312,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 0.7074,
+      "step": 6312,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 1.2028,
+      "step": 6312,
+      "vm_loss": 0.2683
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 0.9914,
+      "step": 6312,
+      "vm_loss": 0.127
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 0.682,
+      "step": 6312,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.2151021488557885,
+      "lm_loss": 1.0474,
+      "step": 6312,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 1.215294655533364,
+      "grad_norm": 3.4816528653689436,
+      "learning_rate": 7.042376762879215e-06,
+      "loss": 1.2038,
+      "step": 6313
+    },
+    {
+      "epoch": 1.215487162210939,
+      "grad_norm": 3.3319608500571545,
+      "learning_rate": 7.0393984998796976e-06,
+      "loss": 1.1567,
+      "step": 6314
+    },
+    {
+      "epoch": 1.2156796688885145,
+      "grad_norm": 3.3912542959794973,
+      "learning_rate": 7.036420524688519e-06,
+      "loss": 1.2343,
+      "step": 6315
+    },
+    {
+      "epoch": 1.21587217556609,
+      "grad_norm": 3.269818379282169,
+      "learning_rate": 7.033442837595183e-06,
+      "loss": 1.1185,
+      "step": 6316
+    },
+    {
+      "epoch": 1.2160646822436654,
+      "grad_norm": 3.240081399511524,
+      "learning_rate": 7.030465438889152e-06,
+      "loss": 1.1855,
+      "step": 6317
+    },
+    {
+      "epoch": 1.2162571889212408,
+      "grad_norm": 3.2836433953377497,
+      "learning_rate": 7.027488328859876e-06,
+      "loss": 1.1753,
+      "step": 6318
+    },
+    {
+      "epoch": 1.216449695598816,
+      "grad_norm": 3.1727387759277366,
+      "learning_rate": 7.024511507796764e-06,
+      "loss": 1.1328,
+      "step": 6319
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "grad_norm": 3.0946199516398156,
+      "learning_rate": 7.021534975989197e-06,
+      "loss": 1.1338,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 1.4044,
+      "step": 6320,
+      "vm_loss": 0.1276
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 0.699,
+      "step": 6320,
+      "vm_loss": 0.1354
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 1.157,
+      "step": 6320,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 1.1441,
+      "step": 6320,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 1.0404,
+      "step": 6320,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 1.1469,
+      "step": 6320,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 0.9252,
+      "step": 6320,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.2166422022763914,
+      "lm_loss": 0.8761,
+      "step": 6320,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.2168347089539668,
+      "grad_norm": 3.2000192647736236,
+      "learning_rate": 7.01855873372654e-06,
+      "loss": 1.1929,
+      "step": 6321
+    },
+    {
+      "epoch": 1.2170272156315423,
+      "grad_norm": 3.243909779380425,
+      "learning_rate": 7.015582781298118e-06,
+      "loss": 1.221,
+      "step": 6322
+    },
+    {
+      "epoch": 1.2172197223091177,
+      "grad_norm": 3.2608828606546645,
+      "learning_rate": 7.0126071189932336e-06,
+      "loss": 1.207,
+      "step": 6323
+    },
+    {
+      "epoch": 1.2174122289866929,
+      "grad_norm": 3.063443458622882,
+      "learning_rate": 7.00963174710115e-06,
+      "loss": 1.0766,
+      "step": 6324
+    },
+    {
+      "epoch": 1.2176047356642683,
+      "grad_norm": 3.181150767934691,
+      "learning_rate": 7.0066566659111246e-06,
+      "loss": 1.228,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2177972423418437,
+      "grad_norm": 3.249325003248595,
+      "learning_rate": 7.003681875712366e-06,
+      "loss": 1.1501,
+      "step": 6326
+    },
+    {
+      "epoch": 1.2179897490194191,
+      "grad_norm": 3.39165345313089,
+      "learning_rate": 7.000707376794059e-06,
+      "loss": 1.2915,
+      "step": 6327
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "grad_norm": 3.2178612272632887,
+      "learning_rate": 6.997733169445372e-06,
+      "loss": 1.1737,
+      "step": 6328
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 0.8076,
+      "step": 6328,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 1.3434,
+      "step": 6328,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 1.0055,
+      "step": 6328,
+      "vm_loss": 0.1391
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 1.3691,
+      "step": 6328,
+      "vm_loss": 0.1072
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 1.482,
+      "step": 6328,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 0.9006,
+      "step": 6328,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 0.6516,
+      "step": 6328,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.2181822556969946,
+      "lm_loss": 1.1731,
+      "step": 6328,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 1.21837476237457,
+      "grad_norm": 3.4710482151707986,
+      "learning_rate": 6.994759253955432e-06,
+      "loss": 1.2257,
+      "step": 6329
+    },
+    {
+      "epoch": 1.2185672690521452,
+      "grad_norm": 3.329881707321585,
+      "learning_rate": 6.991785630613334e-06,
+      "loss": 1.1587,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2187597757297206,
+      "grad_norm": 3.348408955253127,
+      "learning_rate": 6.988812299708165e-06,
+      "loss": 1.1833,
+      "step": 6331
+    },
+    {
+      "epoch": 1.218952282407296,
+      "grad_norm": 3.26228539356878,
+      "learning_rate": 6.985839261528966e-06,
+      "loss": 1.1814,
+      "step": 6332
+    },
+    {
+      "epoch": 1.2191447890848714,
+      "grad_norm": 3.158730978599338,
+      "learning_rate": 6.982866516364749e-06,
+      "loss": 1.1598,
+      "step": 6333
+    },
+    {
+      "epoch": 1.2193372957624469,
+      "grad_norm": 3.0713112136553327,
+      "learning_rate": 6.979894064504508e-06,
+      "loss": 1.1843,
+      "step": 6334
+    },
+    {
+      "epoch": 1.219529802440022,
+      "grad_norm": 3.3747343042436464,
+      "learning_rate": 6.976921906237202e-06,
+      "loss": 1.1887,
+      "step": 6335
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "grad_norm": 3.1122766112857003,
+      "learning_rate": 6.9739500418517645e-06,
+      "loss": 1.1417,
+      "step": 6336
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 1.1145,
+      "step": 6336,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 1.0423,
+      "step": 6336,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 0.7772,
+      "step": 6336,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 0.7666,
+      "step": 6336,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 0.7329,
+      "step": 6336,
+      "vm_loss": 0.2232
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 1.0783,
+      "step": 6336,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 0.8958,
+      "step": 6336,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 1.2197223091175975,
+      "lm_loss": 1.7043,
+      "step": 6336,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.219914815795173,
+      "grad_norm": 3.268007424360385,
+      "learning_rate": 6.970978471637097e-06,
+      "loss": 1.2128,
+      "step": 6337
+    },
+    {
+      "epoch": 1.2201073224727483,
+      "grad_norm": 3.2308761642508363,
+      "learning_rate": 6.968007195882074e-06,
+      "loss": 1.1716,
+      "step": 6338
+    },
+    {
+      "epoch": 1.2202998291503238,
+      "grad_norm": 3.139863694301323,
+      "learning_rate": 6.965036214875542e-06,
+      "loss": 1.0914,
+      "step": 6339
+    },
+    {
+      "epoch": 1.220492335827899,
+      "grad_norm": 3.3129502649341536,
+      "learning_rate": 6.962065528906321e-06,
+      "loss": 1.1707,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2206848425054744,
+      "grad_norm": 3.26907779981963,
+      "learning_rate": 6.9590951382631926e-06,
+      "loss": 1.1628,
+      "step": 6341
+    },
+    {
+      "epoch": 1.2208773491830498,
+      "grad_norm": 3.4099939337519665,
+      "learning_rate": 6.956125043234925e-06,
+      "loss": 1.2209,
+      "step": 6342
+    },
+    {
+      "epoch": 1.2210698558606252,
+      "grad_norm": 3.3565261744081485,
+      "learning_rate": 6.953155244110246e-06,
+      "loss": 1.2062,
+      "step": 6343
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "grad_norm": 3.267385174607392,
+      "learning_rate": 6.950185741177853e-06,
+      "loss": 1.1668,
+      "step": 6344
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.9226,
+      "step": 6344,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.9342,
+      "step": 6344,
+      "vm_loss": 0.2127
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.8098,
+      "step": 6344,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.8304,
+      "step": 6344,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 1.1625,
+      "step": 6344,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.5067,
+      "step": 6344,
+      "vm_loss": 0.2202
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 0.6432,
+      "step": 6344,
+      "vm_loss": 0.2229
+    },
+    {
+      "epoch": 1.2212623625382006,
+      "lm_loss": 1.0615,
+      "step": 6344,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.2214548692157758,
+      "grad_norm": 3.280510693864168,
+      "learning_rate": 6.947216534726428e-06,
+      "loss": 1.2158,
+      "step": 6345
+    },
+    {
+      "epoch": 1.2216473758933513,
+      "grad_norm": 3.077767741385324,
+      "learning_rate": 6.9442476250446156e-06,
+      "loss": 1.1494,
+      "step": 6346
+    },
+    {
+      "epoch": 1.2218398825709267,
+      "grad_norm": 3.162607141282282,
+      "learning_rate": 6.941279012421025e-06,
+      "loss": 1.1823,
+      "step": 6347
+    },
+    {
+      "epoch": 1.222032389248502,
+      "grad_norm": 3.006629932782205,
+      "learning_rate": 6.938310697144245e-06,
+      "loss": 1.0913,
+      "step": 6348
+    },
+    {
+      "epoch": 1.2222248959260775,
+      "grad_norm": 3.30982558084763,
+      "learning_rate": 6.935342679502839e-06,
+      "loss": 1.23,
+      "step": 6349
+    },
+    {
+      "epoch": 1.2224174026036527,
+      "grad_norm": 3.290391128679161,
+      "learning_rate": 6.932374959785333e-06,
+      "loss": 1.1852,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2226099092812281,
+      "grad_norm": 3.2484693269368234,
+      "learning_rate": 6.9294075382802215e-06,
+      "loss": 1.2044,
+      "step": 6351
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "grad_norm": 3.254706654428902,
+      "learning_rate": 6.926440415275988e-06,
+      "loss": 1.1562,
+      "step": 6352
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 0.5933,
+      "step": 6352,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 1.1309,
+      "step": 6352,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 0.9949,
+      "step": 6352,
+      "vm_loss": 0.2301
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 0.8022,
+      "step": 6352,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 1.3695,
+      "step": 6352,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 0.8494,
+      "step": 6352,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 0.9152,
+      "step": 6352,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.2228024159588036,
+      "lm_loss": 1.4296,
+      "step": 6352,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.222994922636379,
+      "grad_norm": 3.268875392274639,
+      "learning_rate": 6.923473591061067e-06,
+      "loss": 1.1637,
+      "step": 6353
+    },
+    {
+      "epoch": 1.2231874293139544,
+      "grad_norm": 3.249527857636542,
+      "learning_rate": 6.920507065923871e-06,
+      "loss": 1.1749,
+      "step": 6354
+    },
+    {
+      "epoch": 1.2233799359915296,
+      "grad_norm": 3.337548369712692,
+      "learning_rate": 6.91754084015279e-06,
+      "loss": 1.1657,
+      "step": 6355
+    },
+    {
+      "epoch": 1.223572442669105,
+      "grad_norm": 3.2975232294242676,
+      "learning_rate": 6.914574914036172e-06,
+      "loss": 1.214,
+      "step": 6356
+    },
+    {
+      "epoch": 1.2237649493466805,
+      "grad_norm": 3.3902446782440214,
+      "learning_rate": 6.9116092878623485e-06,
+      "loss": 1.2485,
+      "step": 6357
+    },
+    {
+      "epoch": 1.2239574560242559,
+      "grad_norm": 3.042917875445795,
+      "learning_rate": 6.908643961919614e-06,
+      "loss": 1.1429,
+      "step": 6358
+    },
+    {
+      "epoch": 1.2241499627018313,
+      "grad_norm": 3.204473968924032,
+      "learning_rate": 6.905678936496235e-06,
+      "loss": 1.1691,
+      "step": 6359
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "grad_norm": 3.127600914365728,
+      "learning_rate": 6.902714211880454e-06,
+      "loss": 1.163,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 0.3458,
+      "step": 6360,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 0.9782,
+      "step": 6360,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 1.0702,
+      "step": 6360,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 0.5649,
+      "step": 6360,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 1.158,
+      "step": 6360,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 0.7082,
+      "step": 6360,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 0.8934,
+      "step": 6360,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 1.2243424693794065,
+      "lm_loss": 1.1257,
+      "step": 6360,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 1.224534976056982,
+      "grad_norm": 3.272410678834247,
+      "learning_rate": 6.899749788360473e-06,
+      "loss": 1.1753,
+      "step": 6361
+    },
+    {
+      "epoch": 1.2247274827345573,
+      "grad_norm": 3.32810540797477,
+      "learning_rate": 6.896785666224482e-06,
+      "loss": 1.1767,
+      "step": 6362
+    },
+    {
+      "epoch": 1.2249199894121328,
+      "grad_norm": 3.22079894919327,
+      "learning_rate": 6.893821845760623e-06,
+      "loss": 1.1962,
+      "step": 6363
+    },
+    {
+      "epoch": 1.2251124960897082,
+      "grad_norm": 3.3457615918505397,
+      "learning_rate": 6.890858327257022e-06,
+      "loss": 1.1658,
+      "step": 6364
+    },
+    {
+      "epoch": 1.2253050027672834,
+      "grad_norm": 3.3464868265624204,
+      "learning_rate": 6.887895111001765e-06,
+      "loss": 1.1623,
+      "step": 6365
+    },
+    {
+      "epoch": 1.2254975094448588,
+      "grad_norm": 3.3999814426085737,
+      "learning_rate": 6.884932197282923e-06,
+      "loss": 1.1875,
+      "step": 6366
+    },
+    {
+      "epoch": 1.2256900161224342,
+      "grad_norm": 3.2790174787680284,
+      "learning_rate": 6.881969586388524e-06,
+      "loss": 1.1849,
+      "step": 6367
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "grad_norm": 3.2084795826150914,
+      "learning_rate": 6.879007278606569e-06,
+      "loss": 1.1953,
+      "step": 6368
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 0.9642,
+      "step": 6368,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 0.8978,
+      "step": 6368,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 0.6809,
+      "step": 6368,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 1.2326,
+      "step": 6368,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 0.7029,
+      "step": 6368,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 0.679,
+      "step": 6368,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 1.3457,
+      "step": 6368,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.2258825228000096,
+      "lm_loss": 1.1471,
+      "step": 6368,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 1.226075029477585,
+      "grad_norm": 3.2977343250752615,
+      "learning_rate": 6.8760452742250385e-06,
+      "loss": 1.1721,
+      "step": 6369
+    },
+    {
+      "epoch": 1.2262675361551603,
+      "grad_norm": 3.148962007475685,
+      "learning_rate": 6.873083573531876e-06,
+      "loss": 1.1561,
+      "step": 6370
+    },
+    {
+      "epoch": 1.2264600428327357,
+      "grad_norm": 3.1906776920193725,
+      "learning_rate": 6.870122176814991e-06,
+      "loss": 1.0846,
+      "step": 6371
+    },
+    {
+      "epoch": 1.226652549510311,
+      "grad_norm": 3.245159011983423,
+      "learning_rate": 6.867161084362275e-06,
+      "loss": 1.209,
+      "step": 6372
+    },
+    {
+      "epoch": 1.2268450561878865,
+      "grad_norm": 3.184375407596528,
+      "learning_rate": 6.864200296461585e-06,
+      "loss": 1.1654,
+      "step": 6373
+    },
+    {
+      "epoch": 1.227037562865462,
+      "grad_norm": 3.2360517733828567,
+      "learning_rate": 6.861239813400742e-06,
+      "loss": 1.2161,
+      "step": 6374
+    },
+    {
+      "epoch": 1.2272300695430371,
+      "grad_norm": 3.2703951014009562,
+      "learning_rate": 6.858279635467544e-06,
+      "loss": 1.1823,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "grad_norm": 3.16553047828405,
+      "learning_rate": 6.855319762949763e-06,
+      "loss": 1.119,
+      "step": 6376
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 1.026,
+      "step": 6376,
+      "vm_loss": 0.1029
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 1.0472,
+      "step": 6376,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 1.3632,
+      "step": 6376,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 0.9518,
+      "step": 6376,
+      "vm_loss": 0.0994
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 1.502,
+      "step": 6376,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 0.7018,
+      "step": 6376,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 0.6451,
+      "step": 6376,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 1.2274225762206126,
+      "lm_loss": 1.1055,
+      "step": 6376,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 1.227615082898188,
+      "grad_norm": 3.23252955467232,
+      "learning_rate": 6.852360196135133e-06,
+      "loss": 1.1875,
+      "step": 6377
+    },
+    {
+      "epoch": 1.2278075895757634,
+      "grad_norm": 3.122840578742874,
+      "learning_rate": 6.849400935311359e-06,
+      "loss": 1.1374,
+      "step": 6378
+    },
+    {
+      "epoch": 1.2280000962533388,
+      "grad_norm": 3.1446273771302784,
+      "learning_rate": 6.846441980766123e-06,
+      "loss": 1.0958,
+      "step": 6379
+    },
+    {
+      "epoch": 1.2281926029309143,
+      "grad_norm": 3.159614219358664,
+      "learning_rate": 6.843483332787076e-06,
+      "loss": 1.1042,
+      "step": 6380
+    },
+    {
+      "epoch": 1.2283851096084895,
+      "grad_norm": 3.319008321988319,
+      "learning_rate": 6.840524991661831e-06,
+      "loss": 1.1741,
+      "step": 6381
+    },
+    {
+      "epoch": 1.2285776162860649,
+      "grad_norm": 3.3108241669562153,
+      "learning_rate": 6.837566957677974e-06,
+      "loss": 1.1579,
+      "step": 6382
+    },
+    {
+      "epoch": 1.2287701229636403,
+      "grad_norm": 3.2603794702010216,
+      "learning_rate": 6.834609231123073e-06,
+      "loss": 1.13,
+      "step": 6383
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "grad_norm": 3.2775946754971588,
+      "learning_rate": 6.831651812284652e-06,
+      "loss": 1.1505,
+      "step": 6384
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 1.2223,
+      "step": 6384,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 0.8183,
+      "step": 6384,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 0.7273,
+      "step": 6384,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 1.2707,
+      "step": 6384,
+      "vm_loss": 0.0932
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 1.2897,
+      "step": 6384,
+      "vm_loss": 0.2024
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 0.9332,
+      "step": 6384,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 1.0724,
+      "step": 6384,
+      "vm_loss": 0.2008
+    },
+    {
+      "epoch": 1.2289626296412157,
+      "lm_loss": 0.6767,
+      "step": 6384,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.2291551363187911,
+      "grad_norm": 3.252813045993992,
+      "learning_rate": 6.8286947014502045e-06,
+      "loss": 1.1225,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2293476429963663,
+      "grad_norm": 3.2764823607593874,
+      "learning_rate": 6.82573789890721e-06,
+      "loss": 1.204,
+      "step": 6386
+    },
+    {
+      "epoch": 1.2295401496739418,
+      "grad_norm": 3.413546774064386,
+      "learning_rate": 6.822781404943103e-06,
+      "loss": 1.1847,
+      "step": 6387
+    },
+    {
+      "epoch": 1.2297326563515172,
+      "grad_norm": 3.204617337086198,
+      "learning_rate": 6.819825219845286e-06,
+      "loss": 1.1588,
+      "step": 6388
+    },
+    {
+      "epoch": 1.2299251630290926,
+      "grad_norm": 3.2242473502446565,
+      "learning_rate": 6.81686934390115e-06,
+      "loss": 1.1631,
+      "step": 6389
+    },
+    {
+      "epoch": 1.230117669706668,
+      "grad_norm": 3.16720224612367,
+      "learning_rate": 6.8139137773980364e-06,
+      "loss": 1.1456,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2303101763842434,
+      "grad_norm": 3.225978099954139,
+      "learning_rate": 6.810958520623267e-06,
+      "loss": 1.142,
+      "step": 6391
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "grad_norm": 3.330341542525514,
+      "learning_rate": 6.8080035738641236e-06,
+      "loss": 1.2045,
+      "step": 6392
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 0.6361,
+      "step": 6392,
+      "vm_loss": 0.2022
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 1.083,
+      "step": 6392,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 0.7344,
+      "step": 6392,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 1.1294,
+      "step": 6392,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 0.6235,
+      "step": 6392,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 0.9764,
+      "step": 6392,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 1.3719,
+      "step": 6392,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 1.2305026830618186,
+      "lm_loss": 1.014,
+      "step": 6392,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 1.230695189739394,
+      "grad_norm": 3.334779462170675,
+      "learning_rate": 6.805048937407875e-06,
+      "loss": 1.1901,
+      "step": 6393
+    },
+    {
+      "epoch": 1.2308876964169695,
+      "grad_norm": 3.2258686608963942,
+      "learning_rate": 6.802094611541745e-06,
+      "loss": 1.18,
+      "step": 6394
+    },
+    {
+      "epoch": 1.231080203094545,
+      "grad_norm": 3.3416646873915297,
+      "learning_rate": 6.799140596552929e-06,
+      "loss": 1.1748,
+      "step": 6395
+    },
+    {
+      "epoch": 1.2312727097721203,
+      "grad_norm": 3.301910263871618,
+      "learning_rate": 6.7961868927286005e-06,
+      "loss": 1.1572,
+      "step": 6396
+    },
+    {
+      "epoch": 1.2314652164496955,
+      "grad_norm": 3.288587862360865,
+      "learning_rate": 6.793233500355893e-06,
+      "loss": 1.1634,
+      "step": 6397
+    },
+    {
+      "epoch": 1.231657723127271,
+      "grad_norm": 3.275838883678048,
+      "learning_rate": 6.790280419721919e-06,
+      "loss": 1.192,
+      "step": 6398
+    },
+    {
+      "epoch": 1.2318502298048464,
+      "grad_norm": 3.2206118944199322,
+      "learning_rate": 6.787327651113749e-06,
+      "loss": 1.1614,
+      "step": 6399
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "grad_norm": 3.2060283515758496,
+      "learning_rate": 6.784375194818435e-06,
+      "loss": 1.1508,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 1.0285,
+      "step": 6400,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 1.1317,
+      "step": 6400,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 1.1628,
+      "step": 6400,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 0.8128,
+      "step": 6400,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 1.3343,
+      "step": 6400,
+      "vm_loss": 0.2293
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 0.817,
+      "step": 6400,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 0.6489,
+      "step": 6400,
+      "vm_loss": 0.2312
+    },
+    {
+      "epoch": 1.2320427364824218,
+      "lm_loss": 0.7846,
+      "step": 6400,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.2322352431599972,
+      "grad_norm": 3.1731773778458763,
+      "learning_rate": 6.7814230511229924e-06,
+      "loss": 1.1579,
+      "step": 6401
+    },
+    {
+      "epoch": 1.2324277498375724,
+      "grad_norm": 3.259987382322204,
+      "learning_rate": 6.778471220314403e-06,
+      "loss": 1.1926,
+      "step": 6402
+    },
+    {
+      "epoch": 1.2326202565151478,
+      "grad_norm": 3.2589509568281603,
+      "learning_rate": 6.775519702679631e-06,
+      "loss": 1.137,
+      "step": 6403
+    },
+    {
+      "epoch": 1.2328127631927233,
+      "grad_norm": 3.1838492740212248,
+      "learning_rate": 6.7725684985055965e-06,
+      "loss": 1.1618,
+      "step": 6404
+    },
+    {
+      "epoch": 1.2330052698702987,
+      "grad_norm": 3.385631725967053,
+      "learning_rate": 6.769617608079191e-06,
+      "loss": 1.2271,
+      "step": 6405
+    },
+    {
+      "epoch": 1.233197776547874,
+      "grad_norm": 3.272867752305655,
+      "learning_rate": 6.766667031687286e-06,
+      "loss": 1.1534,
+      "step": 6406
+    },
+    {
+      "epoch": 1.2333902832254493,
+      "grad_norm": 3.1986803444259,
+      "learning_rate": 6.763716769616714e-06,
+      "loss": 1.15,
+      "step": 6407
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "grad_norm": 3.2869641475152886,
+      "learning_rate": 6.760766822154274e-06,
+      "loss": 1.1922,
+      "step": 6408
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 0.7554,
+      "step": 6408,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 1.1554,
+      "step": 6408,
+      "vm_loss": 0.1601
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 0.8012,
+      "step": 6408,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 0.7747,
+      "step": 6408,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 1.5025,
+      "step": 6408,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 0.7256,
+      "step": 6408,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 1.2578,
+      "step": 6408,
+      "vm_loss": 0.187
+    },
+    {
+      "epoch": 1.2335827899030247,
+      "lm_loss": 1.2996,
+      "step": 6408,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 1.2337752965806001,
+      "grad_norm": 3.343995773366055,
+      "learning_rate": 6.7578171895867375e-06,
+      "loss": 1.1975,
+      "step": 6409
+    },
+    {
+      "epoch": 1.2339678032581756,
+      "grad_norm": 3.2356894446518942,
+      "learning_rate": 6.754867872200854e-06,
+      "loss": 1.144,
+      "step": 6410
+    },
+    {
+      "epoch": 1.234160309935751,
+      "grad_norm": 3.288407368586856,
+      "learning_rate": 6.751918870283331e-06,
+      "loss": 1.1831,
+      "step": 6411
+    },
+    {
+      "epoch": 1.2343528166133262,
+      "grad_norm": 3.245010164018591,
+      "learning_rate": 6.748970184120844e-06,
+      "loss": 1.1664,
+      "step": 6412
+    },
+    {
+      "epoch": 1.2345453232909016,
+      "grad_norm": 3.185605696019955,
+      "learning_rate": 6.7460218140000525e-06,
+      "loss": 1.1185,
+      "step": 6413
+    },
+    {
+      "epoch": 1.234737829968477,
+      "grad_norm": 3.2796317965871205,
+      "learning_rate": 6.743073760207573e-06,
+      "loss": 1.1644,
+      "step": 6414
+    },
+    {
+      "epoch": 1.2349303366460525,
+      "grad_norm": 3.302757000057848,
+      "learning_rate": 6.740126023029987e-06,
+      "loss": 1.2243,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "grad_norm": 3.2120315257251666,
+      "learning_rate": 6.737178602753859e-06,
+      "loss": 1.177,
+      "step": 6416
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 1.2269,
+      "step": 6416,
+      "vm_loss": 0.2367
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 1.2172,
+      "step": 6416,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 1.1004,
+      "step": 6416,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 1.0071,
+      "step": 6416,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 0.9398,
+      "step": 6416,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 0.9857,
+      "step": 6416,
+      "vm_loss": 0.2076
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 0.604,
+      "step": 6416,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.2351228433236279,
+      "lm_loss": 0.9268,
+      "step": 6416,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 1.235315350001203,
+      "grad_norm": 3.2166254039846454,
+      "learning_rate": 6.734231499665717e-06,
+      "loss": 1.1426,
+      "step": 6417
+    },
+    {
+      "epoch": 1.2355078566787785,
+      "grad_norm": 3.3645591744255614,
+      "learning_rate": 6.731284714052051e-06,
+      "loss": 1.1776,
+      "step": 6418
+    },
+    {
+      "epoch": 1.235700363356354,
+      "grad_norm": 3.3144847649924274,
+      "learning_rate": 6.728338246199331e-06,
+      "loss": 1.1761,
+      "step": 6419
+    },
+    {
+      "epoch": 1.2358928700339293,
+      "grad_norm": 3.396453192371253,
+      "learning_rate": 6.72539209639399e-06,
+      "loss": 1.2255,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2360853767115048,
+      "grad_norm": 3.207005840991673,
+      "learning_rate": 6.7224462649224345e-06,
+      "loss": 1.1429,
+      "step": 6421
+    },
+    {
+      "epoch": 1.23627788338908,
+      "grad_norm": 3.2747951875335053,
+      "learning_rate": 6.719500752071033e-06,
+      "loss": 1.199,
+      "step": 6422
+    },
+    {
+      "epoch": 1.2364703900666554,
+      "grad_norm": 3.1125882856998617,
+      "learning_rate": 6.716555558126126e-06,
+      "loss": 1.1517,
+      "step": 6423
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "grad_norm": 3.1624782067730504,
+      "learning_rate": 6.713610683374028e-06,
+      "loss": 1.1404,
+      "step": 6424
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 1.2258,
+      "step": 6424,
+      "vm_loss": 0.2141
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 1.2754,
+      "step": 6424,
+      "vm_loss": 0.2242
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 1.021,
+      "step": 6424,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 0.651,
+      "step": 6424,
+      "vm_loss": 0.1906
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 0.7658,
+      "step": 6424,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 0.9653,
+      "step": 6424,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 0.9538,
+      "step": 6424,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 1.2366628967442308,
+      "lm_loss": 0.8322,
+      "step": 6424,
+      "vm_loss": 0.2061
+    },
+    {
+      "epoch": 1.2368554034218062,
+      "grad_norm": 3.1776533782410645,
+      "learning_rate": 6.710666128101018e-06,
+      "loss": 1.1754,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2370479100993816,
+      "grad_norm": 3.223181011492992,
+      "learning_rate": 6.707721892593339e-06,
+      "loss": 1.1607,
+      "step": 6426
+    },
+    {
+      "epoch": 1.2372404167769568,
+      "grad_norm": 3.417475601262568,
+      "learning_rate": 6.704777977137217e-06,
+      "loss": 1.2818,
+      "step": 6427
+    },
+    {
+      "epoch": 1.2374329234545323,
+      "grad_norm": 3.418829666485802,
+      "learning_rate": 6.7018343820188324e-06,
+      "loss": 1.1741,
+      "step": 6428
+    },
+    {
+      "epoch": 1.2376254301321077,
+      "grad_norm": 3.3420881975265613,
+      "learning_rate": 6.698891107524339e-06,
+      "loss": 1.1949,
+      "step": 6429
+    },
+    {
+      "epoch": 1.237817936809683,
+      "grad_norm": 3.2995499156600494,
+      "learning_rate": 6.695948153939869e-06,
+      "loss": 1.2,
+      "step": 6430
+    },
+    {
+      "epoch": 1.2380104434872585,
+      "grad_norm": 3.349400151108026,
+      "learning_rate": 6.693005521551507e-06,
+      "loss": 1.187,
+      "step": 6431
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "grad_norm": 3.364339599392584,
+      "learning_rate": 6.690063210645318e-06,
+      "loss": 1.1951,
+      "step": 6432
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 0.8332,
+      "step": 6432,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 1.1286,
+      "step": 6432,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 1.1205,
+      "step": 6432,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 1.1866,
+      "step": 6432,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 0.9926,
+      "step": 6432,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 1.3529,
+      "step": 6432,
+      "vm_loss": 0.234
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 0.5125,
+      "step": 6432,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.2382029501648337,
+      "lm_loss": 0.5856,
+      "step": 6432,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 1.2383954568424091,
+      "grad_norm": 3.2981878803705125,
+      "learning_rate": 6.687121221507327e-06,
+      "loss": 1.2182,
+      "step": 6433
+    },
+    {
+      "epoch": 1.2385879635199846,
+      "grad_norm": 3.181671269021458,
+      "learning_rate": 6.684179554423542e-06,
+      "loss": 1.1663,
+      "step": 6434
+    },
+    {
+      "epoch": 1.23878047019756,
+      "grad_norm": 3.1483686192548923,
+      "learning_rate": 6.681238209679926e-06,
+      "loss": 1.1028,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2389729768751354,
+      "grad_norm": 3.2765035443444837,
+      "learning_rate": 6.6782971875624105e-06,
+      "loss": 1.2155,
+      "step": 6436
+    },
+    {
+      "epoch": 1.2391654835527106,
+      "grad_norm": 3.1288715823570956,
+      "learning_rate": 6.67535648835691e-06,
+      "loss": 1.1221,
+      "step": 6437
+    },
+    {
+      "epoch": 1.239357990230286,
+      "grad_norm": 3.236588441010955,
+      "learning_rate": 6.672416112349293e-06,
+      "loss": 1.1422,
+      "step": 6438
+    },
+    {
+      "epoch": 1.2395504969078615,
+      "grad_norm": 3.2081420535282974,
+      "learning_rate": 6.669476059825401e-06,
+      "loss": 1.1201,
+      "step": 6439
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "grad_norm": 3.2849340398857074,
+      "learning_rate": 6.666536331071045e-06,
+      "loss": 1.1917,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.1492,
+      "step": 6440,
+      "vm_loss": 0.2317
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.1589,
+      "step": 6440,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.3083,
+      "step": 6440,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.2385,
+      "step": 6440,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.2025,
+      "step": 6440,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.0232,
+      "step": 6440,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.582,
+      "step": 6440,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.2397430035854369,
+      "lm_loss": 1.1614,
+      "step": 6440,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.2399355102630123,
+      "grad_norm": 3.5703682908965857,
+      "learning_rate": 6.663596926372006e-06,
+      "loss": 1.2414,
+      "step": 6441
+    },
+    {
+      "epoch": 1.2401280169405877,
+      "grad_norm": 3.2380129483288442,
+      "learning_rate": 6.660657846014031e-06,
+      "loss": 1.1402,
+      "step": 6442
+    },
+    {
+      "epoch": 1.240320523618163,
+      "grad_norm": 3.4179083462183044,
+      "learning_rate": 6.6577190902828355e-06,
+      "loss": 1.2141,
+      "step": 6443
+    },
+    {
+      "epoch": 1.2405130302957383,
+      "grad_norm": 3.4118236295316784,
+      "learning_rate": 6.654780659464105e-06,
+      "loss": 1.1876,
+      "step": 6444
+    },
+    {
+      "epoch": 1.2407055369733138,
+      "grad_norm": 3.2447648945003484,
+      "learning_rate": 6.651842553843494e-06,
+      "loss": 1.213,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2408980436508892,
+      "grad_norm": 3.25890208548692,
+      "learning_rate": 6.648904773706617e-06,
+      "loss": 1.1507,
+      "step": 6446
+    },
+    {
+      "epoch": 1.2410905503284646,
+      "grad_norm": 3.1274992619924795,
+      "learning_rate": 6.645967319339075e-06,
+      "loss": 1.1667,
+      "step": 6447
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "grad_norm": 3.2156283476732366,
+      "learning_rate": 6.643030191026419e-06,
+      "loss": 1.1586,
+      "step": 6448
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 1.4617,
+      "step": 6448,
+      "vm_loss": 0.1913
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 1.0869,
+      "step": 6448,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 1.1408,
+      "step": 6448,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 0.7855,
+      "step": 6448,
+      "vm_loss": 0.2224
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 1.0465,
+      "step": 6448,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 0.7974,
+      "step": 6448,
+      "vm_loss": 0.2201
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 0.9376,
+      "step": 6448,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 1.2412830570060398,
+      "lm_loss": 0.8957,
+      "step": 6448,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.2414755636836152,
+      "grad_norm": 3.3060825377202665,
+      "learning_rate": 6.640093389054176e-06,
+      "loss": 1.2578,
+      "step": 6449
+    },
+    {
+      "epoch": 1.2416680703611906,
+      "grad_norm": 3.260136064612598,
+      "learning_rate": 6.637156913707839e-06,
+      "loss": 1.1925,
+      "step": 6450
+    },
+    {
+      "epoch": 1.241860577038766,
+      "grad_norm": 3.2069638338448416,
+      "learning_rate": 6.6342207652728765e-06,
+      "loss": 1.1439,
+      "step": 6451
+    },
+    {
+      "epoch": 1.2420530837163415,
+      "grad_norm": 3.297263510135486,
+      "learning_rate": 6.631284944034716e-06,
+      "loss": 1.1867,
+      "step": 6452
+    },
+    {
+      "epoch": 1.242245590393917,
+      "grad_norm": 3.342538461912467,
+      "learning_rate": 6.628349450278753e-06,
+      "loss": 1.1885,
+      "step": 6453
+    },
+    {
+      "epoch": 1.242438097071492,
+      "grad_norm": 3.192040432257448,
+      "learning_rate": 6.6254142842903644e-06,
+      "loss": 1.1086,
+      "step": 6454
+    },
+    {
+      "epoch": 1.2426306037490675,
+      "grad_norm": 3.467120863214339,
+      "learning_rate": 6.622479446354881e-06,
+      "loss": 1.2112,
+      "step": 6455
+    },
+    {
+      "epoch": 1.242823110426643,
+      "grad_norm": 3.316627894706597,
+      "learning_rate": 6.619544936757607e-06,
+      "loss": 1.1334,
+      "step": 6456
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.976,
+      "step": 6456,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.6858,
+      "step": 6456,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 1.2426,
+      "step": 6456,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.4044,
+      "step": 6456,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.5019,
+      "step": 6456,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.9979,
+      "step": 6456,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 0.7759,
+      "step": 6456,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 1.242823110426643,
+      "lm_loss": 1.1337,
+      "step": 6456,
+      "vm_loss": 0.2132
+    },
+    {
+      "epoch": 1.2430156171042184,
+      "grad_norm": 3.231262655764489,
+      "learning_rate": 6.6166107557838075e-06,
+      "loss": 1.14,
+      "step": 6457
+    },
+    {
+      "epoch": 1.2432081237817938,
+      "grad_norm": 3.509915664429392,
+      "learning_rate": 6.613676903718735e-06,
+      "loss": 1.2653,
+      "step": 6458
+    },
+    {
+      "epoch": 1.243400630459369,
+      "grad_norm": 3.2454478405907037,
+      "learning_rate": 6.610743380847588e-06,
+      "loss": 1.1575,
+      "step": 6459
+    },
+    {
+      "epoch": 1.2435931371369444,
+      "grad_norm": 3.368978239540738,
+      "learning_rate": 6.6078101874555465e-06,
+      "loss": 1.2271,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2437856438145198,
+      "grad_norm": 3.3039800714411065,
+      "learning_rate": 6.604877323827755e-06,
+      "loss": 1.1796,
+      "step": 6461
+    },
+    {
+      "epoch": 1.2439781504920953,
+      "grad_norm": 3.0066908771144853,
+      "learning_rate": 6.601944790249322e-06,
+      "loss": 1.1622,
+      "step": 6462
+    },
+    {
+      "epoch": 1.2441706571696707,
+      "grad_norm": 3.229101237328442,
+      "learning_rate": 6.599012587005329e-06,
+      "loss": 1.1785,
+      "step": 6463
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "grad_norm": 3.1910294129349004,
+      "learning_rate": 6.596080714380825e-06,
+      "loss": 1.2032,
+      "step": 6464
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 0.9415,
+      "step": 6464,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 0.8613,
+      "step": 6464,
+      "vm_loss": 0.2152
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 1.2942,
+      "step": 6464,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 0.8417,
+      "step": 6464,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 1.067,
+      "step": 6464,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 1.1984,
+      "step": 6464,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 0.8002,
+      "step": 6464,
+      "vm_loss": 0.0925
+    },
+    {
+      "epoch": 1.2443631638472459,
+      "lm_loss": 0.8312,
+      "step": 6464,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 1.2445556705248213,
+      "grad_norm": 3.2743376303242173,
+      "learning_rate": 6.593149172660825e-06,
+      "loss": 1.194,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2447481772023967,
+      "grad_norm": 3.27407654661791,
+      "learning_rate": 6.5902179621303115e-06,
+      "loss": 1.2164,
+      "step": 6466
+    },
+    {
+      "epoch": 1.2449406838799721,
+      "grad_norm": 3.305632762462827,
+      "learning_rate": 6.587287083074234e-06,
+      "loss": 1.1852,
+      "step": 6467
+    },
+    {
+      "epoch": 1.2451331905575476,
+      "grad_norm": 3.328347173440295,
+      "learning_rate": 6.5843565357775165e-06,
+      "loss": 1.1505,
+      "step": 6468
+    },
+    {
+      "epoch": 1.2453256972351228,
+      "grad_norm": 3.160554069273794,
+      "learning_rate": 6.581426320525042e-06,
+      "loss": 1.0902,
+      "step": 6469
+    },
+    {
+      "epoch": 1.2455182039126982,
+      "grad_norm": 3.426550377743918,
+      "learning_rate": 6.578496437601661e-06,
+      "loss": 1.1708,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2457107105902736,
+      "grad_norm": 3.3036999435897645,
+      "learning_rate": 6.575566887292205e-06,
+      "loss": 1.1476,
+      "step": 6471
+    },
+    {
+      "epoch": 1.245903217267849,
+      "grad_norm": 3.251543383195193,
+      "learning_rate": 6.572637669881459e-06,
+      "loss": 1.1237,
+      "step": 6472
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 0.4743,
+      "step": 6472,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 0.9738,
+      "step": 6472,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 0.9925,
+      "step": 6472,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 0.8035,
+      "step": 6472,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 1.4435,
+      "step": 6472,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 1.0071,
+      "step": 6472,
+      "vm_loss": 0.1105
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 0.9499,
+      "step": 6472,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.245903217267849,
+      "lm_loss": 1.0864,
+      "step": 6472,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 1.2460957239454244,
+      "grad_norm": 3.2760580026000072,
+      "learning_rate": 6.56970878565418e-06,
+      "loss": 1.1676,
+      "step": 6473
+    },
+    {
+      "epoch": 1.2462882306229996,
+      "grad_norm": 3.376827501506037,
+      "learning_rate": 6.566780234895089e-06,
+      "loss": 1.1636,
+      "step": 6474
+    },
+    {
+      "epoch": 1.246480737300575,
+      "grad_norm": 3.2769856178786783,
+      "learning_rate": 6.563852017888886e-06,
+      "loss": 1.1399,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2466732439781505,
+      "grad_norm": 3.2301724721179808,
+      "learning_rate": 6.56092413492023e-06,
+      "loss": 1.1573,
+      "step": 6476
+    },
+    {
+      "epoch": 1.246865750655726,
+      "grad_norm": 3.300203600316979,
+      "learning_rate": 6.557996586273741e-06,
+      "loss": 1.1503,
+      "step": 6477
+    },
+    {
+      "epoch": 1.2470582573333013,
+      "grad_norm": 3.249049738345709,
+      "learning_rate": 6.5550693722340255e-06,
+      "loss": 1.197,
+      "step": 6478
+    },
+    {
+      "epoch": 1.2472507640108765,
+      "grad_norm": 3.2054957071922523,
+      "learning_rate": 6.5521424930856395e-06,
+      "loss": 1.1749,
+      "step": 6479
+    },
+    {
+      "epoch": 1.247443270688452,
+      "grad_norm": 3.3373034030763664,
+      "learning_rate": 6.54921594911311e-06,
+      "loss": 1.1994,
+      "step": 6480
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 0.8822,
+      "step": 6480,
+      "vm_loss": 0.1511
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 1.0125,
+      "step": 6480,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 1.3557,
+      "step": 6480,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 1.0088,
+      "step": 6480,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 0.893,
+      "step": 6480,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 0.8262,
+      "step": 6480,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 0.7606,
+      "step": 6480,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.247443270688452,
+      "lm_loss": 1.307,
+      "step": 6480,
+      "vm_loss": 0.2147
+    },
+    {
+      "epoch": 1.2476357773660274,
+      "grad_norm": 3.1915285629620853,
+      "learning_rate": 6.546289740600943e-06,
+      "loss": 1.1186,
+      "step": 6481
+    },
+    {
+      "epoch": 1.2478282840436028,
+      "grad_norm": 3.2830967989858792,
+      "learning_rate": 6.543363867833597e-06,
+      "loss": 1.1439,
+      "step": 6482
+    },
+    {
+      "epoch": 1.2480207907211782,
+      "grad_norm": 3.1723211773780133,
+      "learning_rate": 6.540438331095509e-06,
+      "loss": 1.0864,
+      "step": 6483
+    },
+    {
+      "epoch": 1.2482132973987534,
+      "grad_norm": 3.393476732782133,
+      "learning_rate": 6.5375131306710725e-06,
+      "loss": 1.1336,
+      "step": 6484
+    },
+    {
+      "epoch": 1.2484058040763288,
+      "grad_norm": 3.3428417219994833,
+      "learning_rate": 6.534588266844659e-06,
+      "loss": 1.1682,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2485983107539043,
+      "grad_norm": 3.452901540380813,
+      "learning_rate": 6.531663739900603e-06,
+      "loss": 1.2042,
+      "step": 6486
+    },
+    {
+      "epoch": 1.2487908174314797,
+      "grad_norm": 3.213686504230132,
+      "learning_rate": 6.528739550123201e-06,
+      "loss": 1.1636,
+      "step": 6487
+    },
+    {
+      "epoch": 1.248983324109055,
+      "grad_norm": 3.2043499127739143,
+      "learning_rate": 6.525815697796729e-06,
+      "loss": 1.1478,
+      "step": 6488
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 0.9215,
+      "step": 6488,
+      "vm_loss": 0.2279
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 0.9209,
+      "step": 6488,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 1.3809,
+      "step": 6488,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 1.2241,
+      "step": 6488,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 0.6929,
+      "step": 6488,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 1.087,
+      "step": 6488,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 0.7616,
+      "step": 6488,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 1.248983324109055,
+      "lm_loss": 0.8344,
+      "step": 6488,
+      "vm_loss": 0.0977
+    },
+    {
+      "epoch": 1.2491758307866303,
+      "grad_norm": 3.2593989964342853,
+      "learning_rate": 6.522892183205418e-06,
+      "loss": 1.1662,
+      "step": 6489
+    },
+    {
+      "epoch": 1.2493683374642057,
+      "grad_norm": 3.2876302717363863,
+      "learning_rate": 6.519969006633473e-06,
+      "loss": 1.2049,
+      "step": 6490
+    },
+    {
+      "epoch": 1.2495608441417811,
+      "grad_norm": 3.124470262346882,
+      "learning_rate": 6.517046168365059e-06,
+      "loss": 1.1201,
+      "step": 6491
+    },
+    {
+      "epoch": 1.2497533508193566,
+      "grad_norm": 3.2184367519592882,
+      "learning_rate": 6.514123668684321e-06,
+      "loss": 1.1273,
+      "step": 6492
+    },
+    {
+      "epoch": 1.249945857496932,
+      "grad_norm": 3.2300728259813365,
+      "learning_rate": 6.5112015078753595e-06,
+      "loss": 1.1094,
+      "step": 6493
+    },
+    {
+      "epoch": 1.2501383641745072,
+      "grad_norm": 3.306738755137397,
+      "learning_rate": 6.508279686222243e-06,
+      "loss": 1.1508,
+      "step": 6494
+    },
+    {
+      "epoch": 1.2503308708520826,
+      "grad_norm": 3.2526034007987814,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 1.1393,
+      "step": 6495
+    },
+    {
+      "epoch": 1.250523377529658,
+      "grad_norm": 3.2872425365484683,
+      "learning_rate": 6.502437061519684e-06,
+      "loss": 1.0873,
+      "step": 6496
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.6758,
+      "step": 6496,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.478,
+      "step": 6496,
+      "vm_loss": 0.2281
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 1.2305,
+      "step": 6496,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 1.398,
+      "step": 6496,
+      "vm_loss": 0.1434
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.8182,
+      "step": 6496,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.6764,
+      "step": 6496,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.7359,
+      "step": 6496,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.250523377529658,
+      "lm_loss": 0.8572,
+      "step": 6496,
+      "vm_loss": 0.0932
+    },
+    {
+      "epoch": 1.2507158842072335,
+      "grad_norm": 3.4920727769145343,
+      "learning_rate": 6.499516259038215e-06,
+      "loss": 1.1678,
+      "step": 6497
+    },
+    {
+      "epoch": 1.2509083908848089,
+      "grad_norm": 3.370107449387861,
+      "learning_rate": 6.496595796848545e-06,
+      "loss": 1.0923,
+      "step": 6498
+    },
+    {
+      "epoch": 1.251100897562384,
+      "grad_norm": 3.458431927595557,
+      "learning_rate": 6.49367567523459e-06,
+      "loss": 1.2007,
+      "step": 6499
+    },
+    {
+      "epoch": 1.2512934042399595,
+      "grad_norm": 3.398886026824005,
+      "learning_rate": 6.490755894480218e-06,
+      "loss": 1.2365,
+      "step": 6500
+    },
+    {
+      "epoch": 1.251485910917535,
+      "grad_norm": 3.2190851593638405,
+      "learning_rate": 6.487836454869265e-06,
+      "loss": 1.1265,
+      "step": 6501
+    },
+    {
+      "epoch": 1.2516784175951103,
+      "grad_norm": 3.1946081299341498,
+      "learning_rate": 6.484917356685546e-06,
+      "loss": 1.1497,
+      "step": 6502
+    },
+    {
+      "epoch": 1.2518709242726858,
+      "grad_norm": 3.3309267200587995,
+      "learning_rate": 6.48199860021283e-06,
+      "loss": 1.2098,
+      "step": 6503
+    },
+    {
+      "epoch": 1.252063430950261,
+      "grad_norm": 3.1452551879982056,
+      "learning_rate": 6.479080185734859e-06,
+      "loss": 1.1249,
+      "step": 6504
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.3131,
+      "step": 6504,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.3693,
+      "step": 6504,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.3976,
+      "step": 6504,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.2337,
+      "step": 6504,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 0.9856,
+      "step": 6504,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.3357,
+      "step": 6504,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.1298,
+      "step": 6504,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.252063430950261,
+      "lm_loss": 1.1775,
+      "step": 6504,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 1.2522559376278366,
+      "grad_norm": 3.185797758241765,
+      "learning_rate": 6.47616211353534e-06,
+      "loss": 1.1672,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2524484443054118,
+      "grad_norm": 3.2365708025184077,
+      "learning_rate": 6.4732443838979454e-06,
+      "loss": 1.1654,
+      "step": 6506
+    },
+    {
+      "epoch": 1.2526409509829872,
+      "grad_norm": 3.1965223956458195,
+      "learning_rate": 6.470326997106319e-06,
+      "loss": 1.1631,
+      "step": 6507
+    },
+    {
+      "epoch": 1.2528334576605626,
+      "grad_norm": 3.2188409564472273,
+      "learning_rate": 6.467409953444061e-06,
+      "loss": 1.1702,
+      "step": 6508
+    },
+    {
+      "epoch": 1.2530259643381378,
+      "grad_norm": 3.285414402656991,
+      "learning_rate": 6.464493253194757e-06,
+      "loss": 1.2021,
+      "step": 6509
+    },
+    {
+      "epoch": 1.2532184710157135,
+      "grad_norm": 3.286313818782534,
+      "learning_rate": 6.461576896641939e-06,
+      "loss": 1.1634,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2534109776932887,
+      "grad_norm": 3.438837267570832,
+      "learning_rate": 6.458660884069113e-06,
+      "loss": 1.1417,
+      "step": 6511
+    },
+    {
+      "epoch": 1.253603484370864,
+      "grad_norm": 3.3402585244739473,
+      "learning_rate": 6.455745215759761e-06,
+      "loss": 1.1229,
+      "step": 6512
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 1.0565,
+      "step": 6512,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 1.1177,
+      "step": 6512,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 1.1827,
+      "step": 6512,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 0.6863,
+      "step": 6512,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 0.458,
+      "step": 6512,
+      "vm_loss": 0.095
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 1.2695,
+      "step": 6512,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 0.9193,
+      "step": 6512,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 1.253603484370864,
+      "lm_loss": 0.8516,
+      "step": 6512,
+      "vm_loss": 0.1407
+    },
+    {
+      "epoch": 1.2537959910484395,
+      "grad_norm": 3.5496681882128875,
+      "learning_rate": 6.4528298919973165e-06,
+      "loss": 1.1636,
+      "step": 6513
+    },
+    {
+      "epoch": 1.253988497726015,
+      "grad_norm": 3.3582581441125363,
+      "learning_rate": 6.449914913065189e-06,
+      "loss": 1.1348,
+      "step": 6514
+    },
+    {
+      "epoch": 1.2541810044035904,
+      "grad_norm": 3.2356981689330873,
+      "learning_rate": 6.447000279246747e-06,
+      "loss": 1.1071,
+      "step": 6515
+    },
+    {
+      "epoch": 1.2543735110811656,
+      "grad_norm": 3.124705165766025,
+      "learning_rate": 6.444085990825338e-06,
+      "loss": 1.1098,
+      "step": 6516
+    },
+    {
+      "epoch": 1.254566017758741,
+      "grad_norm": 3.1864643929781,
+      "learning_rate": 6.441172048084264e-06,
+      "loss": 1.1456,
+      "step": 6517
+    },
+    {
+      "epoch": 1.2547585244363164,
+      "grad_norm": 3.2600036936890247,
+      "learning_rate": 6.438258451306794e-06,
+      "loss": 1.1983,
+      "step": 6518
+    },
+    {
+      "epoch": 1.2549510311138918,
+      "grad_norm": 3.104826943219379,
+      "learning_rate": 6.435345200776173e-06,
+      "loss": 1.1166,
+      "step": 6519
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "grad_norm": 3.1496104537650393,
+      "learning_rate": 6.432432296775606e-06,
+      "loss": 1.1224,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.9849,
+      "step": 6520,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.9404,
+      "step": 6520,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.7559,
+      "step": 6520,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.8853,
+      "step": 6520,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.6761,
+      "step": 6520,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 1.6008,
+      "step": 6520,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.8952,
+      "step": 6520,
+      "vm_loss": 0.0932
+    },
+    {
+      "epoch": 1.2551435377914673,
+      "lm_loss": 0.9865,
+      "step": 6520,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 1.2553360444690425,
+      "grad_norm": 3.1629690884168897,
+      "learning_rate": 6.429519739588257e-06,
+      "loss": 1.0974,
+      "step": 6521
+    },
+    {
+      "epoch": 1.2555285511466179,
+      "grad_norm": 3.2150610273477063,
+      "learning_rate": 6.426607529497273e-06,
+      "loss": 1.1424,
+      "step": 6522
+    },
+    {
+      "epoch": 1.2557210578241933,
+      "grad_norm": 3.221647749435571,
+      "learning_rate": 6.4236956667857545e-06,
+      "loss": 1.1249,
+      "step": 6523
+    },
+    {
+      "epoch": 1.2559135645017687,
+      "grad_norm": 3.445590731600045,
+      "learning_rate": 6.42078415173677e-06,
+      "loss": 1.1316,
+      "step": 6524
+    },
+    {
+      "epoch": 1.2561060711793441,
+      "grad_norm": 3.580801887650186,
+      "learning_rate": 6.417872984633356e-06,
+      "loss": 1.2201,
+      "step": 6525
+    },
+    {
+      "epoch": 1.2562985778569193,
+      "grad_norm": 3.5668018182268058,
+      "learning_rate": 6.4149621657585174e-06,
+      "loss": 1.1706,
+      "step": 6526
+    },
+    {
+      "epoch": 1.2564910845344948,
+      "grad_norm": 3.4298768836944586,
+      "learning_rate": 6.412051695395225e-06,
+      "loss": 1.197,
+      "step": 6527
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "grad_norm": 3.195084452236325,
+      "learning_rate": 6.409141573826407e-06,
+      "loss": 1.1456,
+      "step": 6528
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.8882,
+      "step": 6528,
+      "vm_loss": 0.1879
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.8564,
+      "step": 6528,
+      "vm_loss": 0.1073
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.7425,
+      "step": 6528,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.6299,
+      "step": 6528,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 1.2098,
+      "step": 6528,
+      "vm_loss": 0.0902
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 1.3022,
+      "step": 6528,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.7828,
+      "step": 6528,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 1.2566835912120702,
+      "lm_loss": 0.5261,
+      "step": 6528,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.2568760978896456,
+      "grad_norm": 3.253970234346718,
+      "learning_rate": 6.40623180133497e-06,
+      "loss": 1.1322,
+      "step": 6529
+    },
+    {
+      "epoch": 1.257068604567221,
+      "grad_norm": 3.26142651185074,
+      "learning_rate": 6.403322378203782e-06,
+      "loss": 1.1368,
+      "step": 6530
+    },
+    {
+      "epoch": 1.2572611112447962,
+      "grad_norm": 3.216254755246491,
+      "learning_rate": 6.400413304715674e-06,
+      "loss": 1.1666,
+      "step": 6531
+    },
+    {
+      "epoch": 1.2574536179223716,
+      "grad_norm": 3.2099764398814794,
+      "learning_rate": 6.39750458115344e-06,
+      "loss": 1.1747,
+      "step": 6532
+    },
+    {
+      "epoch": 1.257646124599947,
+      "grad_norm": 3.1863892210984774,
+      "learning_rate": 6.394596207799855e-06,
+      "loss": 1.1669,
+      "step": 6533
+    },
+    {
+      "epoch": 1.2578386312775225,
+      "grad_norm": 3.1253549474940114,
+      "learning_rate": 6.391688184937643e-06,
+      "loss": 1.1274,
+      "step": 6534
+    },
+    {
+      "epoch": 1.258031137955098,
+      "grad_norm": 3.3188478197006397,
+      "learning_rate": 6.388780512849501e-06,
+      "loss": 1.1915,
+      "step": 6535
+    },
+    {
+      "epoch": 1.258223644632673,
+      "grad_norm": 3.336536343674728,
+      "learning_rate": 6.385873191818099e-06,
+      "loss": 1.1862,
+      "step": 6536
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 0.8618,
+      "step": 6536,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 1.066,
+      "step": 6536,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 0.6471,
+      "step": 6536,
+      "vm_loss": 0.1437
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 0.5755,
+      "step": 6536,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 0.6876,
+      "step": 6536,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 1.1159,
+      "step": 6536,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 1.5133,
+      "step": 6536,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.258223644632673,
+      "lm_loss": 0.8326,
+      "step": 6536,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.2584161513102485,
+      "grad_norm": 3.1439114082407924,
+      "learning_rate": 6.382966222126059e-06,
+      "loss": 1.1417,
+      "step": 6537
+    },
+    {
+      "epoch": 1.258608657987824,
+      "grad_norm": 3.198305721525765,
+      "learning_rate": 6.380059604055975e-06,
+      "loss": 1.1544,
+      "step": 6538
+    },
+    {
+      "epoch": 1.2588011646653994,
+      "grad_norm": 3.34381104771015,
+      "learning_rate": 6.377153337890414e-06,
+      "loss": 1.1539,
+      "step": 6539
+    },
+    {
+      "epoch": 1.2589936713429748,
+      "grad_norm": 3.2119355967724603,
+      "learning_rate": 6.374247423911897e-06,
+      "loss": 1.1326,
+      "step": 6540
+    },
+    {
+      "epoch": 1.25918617802055,
+      "grad_norm": 3.4657214946278008,
+      "learning_rate": 6.371341862402919e-06,
+      "loss": 1.2058,
+      "step": 6541
+    },
+    {
+      "epoch": 1.2593786846981254,
+      "grad_norm": 3.484099281784217,
+      "learning_rate": 6.368436653645931e-06,
+      "loss": 1.2004,
+      "step": 6542
+    },
+    {
+      "epoch": 1.2595711913757008,
+      "grad_norm": 3.352240398342553,
+      "learning_rate": 6.365531797923365e-06,
+      "loss": 1.1844,
+      "step": 6543
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "grad_norm": 3.2715147242041733,
+      "learning_rate": 6.362627295517606e-06,
+      "loss": 1.1137,
+      "step": 6544
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 1.0429,
+      "step": 6544,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 1.3993,
+      "step": 6544,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 0.6411,
+      "step": 6544,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 1.4759,
+      "step": 6544,
+      "vm_loss": 0.1218
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 0.7102,
+      "step": 6544,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 1.2725,
+      "step": 6544,
+      "vm_loss": 0.2001
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 1.3043,
+      "step": 6544,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 1.2597636980532763,
+      "lm_loss": 0.875,
+      "step": 6544,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.2599562047308517,
+      "grad_norm": 3.4137467969118354,
+      "learning_rate": 6.359723146711009e-06,
+      "loss": 1.2157,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2601487114084269,
+      "grad_norm": 3.243972711073635,
+      "learning_rate": 6.356819351785897e-06,
+      "loss": 1.1443,
+      "step": 6546
+    },
+    {
+      "epoch": 1.2603412180860023,
+      "grad_norm": 3.237286165514483,
+      "learning_rate": 6.3539159110245505e-06,
+      "loss": 1.1171,
+      "step": 6547
+    },
+    {
+      "epoch": 1.2605337247635777,
+      "grad_norm": 3.21306763724978,
+      "learning_rate": 6.351012824709228e-06,
+      "loss": 1.1475,
+      "step": 6548
+    },
+    {
+      "epoch": 1.2607262314411531,
+      "grad_norm": 3.1692013366104286,
+      "learning_rate": 6.34811009312214e-06,
+      "loss": 1.0841,
+      "step": 6549
+    },
+    {
+      "epoch": 1.2609187381187286,
+      "grad_norm": 3.340230730404477,
+      "learning_rate": 6.345207716545474e-06,
+      "loss": 1.2215,
+      "step": 6550
+    },
+    {
+      "epoch": 1.2611112447963038,
+      "grad_norm": 3.161544444710282,
+      "learning_rate": 6.342305695261377e-06,
+      "loss": 1.1384,
+      "step": 6551
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "grad_norm": 3.1783228514730375,
+      "learning_rate": 6.339404029551959e-06,
+      "loss": 1.0953,
+      "step": 6552
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 0.9166,
+      "step": 6552,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 0.9839,
+      "step": 6552,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 0.8046,
+      "step": 6552,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 1.4118,
+      "step": 6552,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 1.1313,
+      "step": 6552,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 1.0796,
+      "step": 6552,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 1.2278,
+      "step": 6552,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 1.2613037514738792,
+      "lm_loss": 1.3371,
+      "step": 6552,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.2614962581514546,
+      "grad_norm": 3.2332800857758834,
+      "learning_rate": 6.336502719699306e-06,
+      "loss": 1.1484,
+      "step": 6553
+    },
+    {
+      "epoch": 1.26168876482903,
+      "grad_norm": 3.23907174208649,
+      "learning_rate": 6.333601765985458e-06,
+      "loss": 1.0782,
+      "step": 6554
+    },
+    {
+      "epoch": 1.2618812715066055,
+      "grad_norm": 3.2853273514445953,
+      "learning_rate": 6.33070116869242e-06,
+      "loss": 1.1368,
+      "step": 6555
+    },
+    {
+      "epoch": 1.2620737781841807,
+      "grad_norm": 3.3695517665808423,
+      "learning_rate": 6.327800928102178e-06,
+      "loss": 1.1909,
+      "step": 6556
+    },
+    {
+      "epoch": 1.262266284861756,
+      "grad_norm": 3.3504843754424214,
+      "learning_rate": 6.324901044496667e-06,
+      "loss": 1.1614,
+      "step": 6557
+    },
+    {
+      "epoch": 1.2624587915393315,
+      "grad_norm": 3.2518660288040078,
+      "learning_rate": 6.322001518157792e-06,
+      "loss": 1.08,
+      "step": 6558
+    },
+    {
+      "epoch": 1.262651298216907,
+      "grad_norm": 3.20670504449457,
+      "learning_rate": 6.319102349367421e-06,
+      "loss": 1.1324,
+      "step": 6559
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "grad_norm": 3.088995480287488,
+      "learning_rate": 6.316203538407397e-06,
+      "loss": 1.0336,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 1.1798,
+      "step": 6560,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 1.4228,
+      "step": 6560,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 0.7189,
+      "step": 6560,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 0.5666,
+      "step": 6560,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 0.5334,
+      "step": 6560,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 0.9095,
+      "step": 6560,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 1.0363,
+      "step": 6560,
+      "vm_loss": 0.1086
+    },
+    {
+      "epoch": 1.2628438048944823,
+      "lm_loss": 0.7588,
+      "step": 6560,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.2630363115720575,
+      "grad_norm": 3.242657550232243,
+      "learning_rate": 6.31330508555952e-06,
+      "loss": 1.1563,
+      "step": 6561
+    },
+    {
+      "epoch": 1.263228818249633,
+      "grad_norm": 3.3398806685798044,
+      "learning_rate": 6.310406991105551e-06,
+      "loss": 1.1558,
+      "step": 6562
+    },
+    {
+      "epoch": 1.2634213249272084,
+      "grad_norm": 3.127694393940714,
+      "learning_rate": 6.307509255327231e-06,
+      "loss": 1.0772,
+      "step": 6563
+    },
+    {
+      "epoch": 1.2636138316047838,
+      "grad_norm": 3.1235472129727317,
+      "learning_rate": 6.30461187850625e-06,
+      "loss": 1.1906,
+      "step": 6564
+    },
+    {
+      "epoch": 1.2638063382823592,
+      "grad_norm": 3.176269750572582,
+      "learning_rate": 6.3017148609242725e-06,
+      "loss": 1.1497,
+      "step": 6565
+    },
+    {
+      "epoch": 1.2639988449599344,
+      "grad_norm": 3.022608979538775,
+      "learning_rate": 6.2988182028629245e-06,
+      "loss": 1.0713,
+      "step": 6566
+    },
+    {
+      "epoch": 1.26419135163751,
+      "grad_norm": 3.14718691820071,
+      "learning_rate": 6.295921904603799e-06,
+      "loss": 1.1319,
+      "step": 6567
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "grad_norm": 3.1692395702697858,
+      "learning_rate": 6.293025966428456e-06,
+      "loss": 1.1177,
+      "step": 6568
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 1.0072,
+      "step": 6568,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 0.8117,
+      "step": 6568,
+      "vm_loss": 0.2362
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 0.7021,
+      "step": 6568,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 0.7703,
+      "step": 6568,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 1.3175,
+      "step": 6568,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 0.8012,
+      "step": 6568,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 0.686,
+      "step": 6568,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.2643838583150853,
+      "lm_loss": 1.0748,
+      "step": 6568,
+      "vm_loss": 0.1213
+    },
+    {
+      "epoch": 1.2645763649926607,
+      "grad_norm": 3.2430794622913086,
+      "learning_rate": 6.290130388618412e-06,
+      "loss": 1.1208,
+      "step": 6569
+    },
+    {
+      "epoch": 1.264768871670236,
+      "grad_norm": 3.296057530080696,
+      "learning_rate": 6.287235171455157e-06,
+      "loss": 1.1312,
+      "step": 6570
+    },
+    {
+      "epoch": 1.2649613783478113,
+      "grad_norm": 3.4003208262780613,
+      "learning_rate": 6.284340315220146e-06,
+      "loss": 1.2242,
+      "step": 6571
+    },
+    {
+      "epoch": 1.265153885025387,
+      "grad_norm": 3.4538104824283096,
+      "learning_rate": 6.281445820194795e-06,
+      "loss": 1.1916,
+      "step": 6572
+    },
+    {
+      "epoch": 1.2653463917029621,
+      "grad_norm": 3.2820107919980726,
+      "learning_rate": 6.278551686660479e-06,
+      "loss": 1.1741,
+      "step": 6573
+    },
+    {
+      "epoch": 1.2655388983805376,
+      "grad_norm": 3.261920089294364,
+      "learning_rate": 6.275657914898554e-06,
+      "loss": 1.1182,
+      "step": 6574
+    },
+    {
+      "epoch": 1.265731405058113,
+      "grad_norm": 3.3961592496259656,
+      "learning_rate": 6.272764505190329e-06,
+      "loss": 1.1691,
+      "step": 6575
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "grad_norm": 3.4323843898187607,
+      "learning_rate": 6.269871457817074e-06,
+      "loss": 1.2213,
+      "step": 6576
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 0.8828,
+      "step": 6576,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 1.7807,
+      "step": 6576,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 1.4014,
+      "step": 6576,
+      "vm_loss": 0.2309
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 0.8518,
+      "step": 6576,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 0.986,
+      "step": 6576,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 0.5707,
+      "step": 6576,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 0.8639,
+      "step": 6576,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.2659239117356884,
+      "lm_loss": 1.0247,
+      "step": 6576,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.2661164184132638,
+      "grad_norm": 3.284454552025599,
+      "learning_rate": 6.2669787730600405e-06,
+      "loss": 1.1349,
+      "step": 6577
+    },
+    {
+      "epoch": 1.266308925090839,
+      "grad_norm": 3.2569641087487513,
+      "learning_rate": 6.264086451200431e-06,
+      "loss": 1.1411,
+      "step": 6578
+    },
+    {
+      "epoch": 1.2665014317684145,
+      "grad_norm": 3.1943833956424434,
+      "learning_rate": 6.261194492519408e-06,
+      "loss": 1.1579,
+      "step": 6579
+    },
+    {
+      "epoch": 1.2666939384459899,
+      "grad_norm": 3.3191513353771183,
+      "learning_rate": 6.258302897298119e-06,
+      "loss": 1.1858,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2668864451235653,
+      "grad_norm": 3.2433977811436794,
+      "learning_rate": 6.255411665817659e-06,
+      "loss": 1.0819,
+      "step": 6581
+    },
+    {
+      "epoch": 1.2670789518011407,
+      "grad_norm": 3.2376220008229986,
+      "learning_rate": 6.252520798359092e-06,
+      "loss": 1.1653,
+      "step": 6582
+    },
+    {
+      "epoch": 1.267271458478716,
+      "grad_norm": 3.2963903999601727,
+      "learning_rate": 6.249630295203443e-06,
+      "loss": 1.204,
+      "step": 6583
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "grad_norm": 3.3534538583674016,
+      "learning_rate": 6.246740156631714e-06,
+      "loss": 1.1484,
+      "step": 6584
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 1.0217,
+      "step": 6584,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.7434,
+      "step": 6584,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 1.1464,
+      "step": 6584,
+      "vm_loss": 0.2339
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.7456,
+      "step": 6584,
+      "vm_loss": 0.1196
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.9957,
+      "step": 6584,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.7101,
+      "step": 6584,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.9687,
+      "step": 6584,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 1.2674639651562913,
+      "lm_loss": 0.6178,
+      "step": 6584,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 1.2676564718338668,
+      "grad_norm": 3.172844712094401,
+      "learning_rate": 6.243850382924861e-06,
+      "loss": 1.0895,
+      "step": 6585
+    },
+    {
+      "epoch": 1.2678489785114422,
+      "grad_norm": 3.35070716741379,
+      "learning_rate": 6.240960974363803e-06,
+      "loss": 1.1604,
+      "step": 6586
+    },
+    {
+      "epoch": 1.2680414851890176,
+      "grad_norm": 3.374475484371833,
+      "learning_rate": 6.238071931229433e-06,
+      "loss": 1.2074,
+      "step": 6587
+    },
+    {
+      "epoch": 1.2682339918665928,
+      "grad_norm": 3.430219849201853,
+      "learning_rate": 6.2351832538025995e-06,
+      "loss": 1.1668,
+      "step": 6588
+    },
+    {
+      "epoch": 1.2684264985441682,
+      "grad_norm": 3.195259537000875,
+      "learning_rate": 6.232294942364121e-06,
+      "loss": 1.1391,
+      "step": 6589
+    },
+    {
+      "epoch": 1.2686190052217436,
+      "grad_norm": 3.0949180533736316,
+      "learning_rate": 6.229406997194774e-06,
+      "loss": 1.1023,
+      "step": 6590
+    },
+    {
+      "epoch": 1.268811511899319,
+      "grad_norm": 3.320149514600776,
+      "learning_rate": 6.2265194185753085e-06,
+      "loss": 1.2249,
+      "step": 6591
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "grad_norm": 3.1216746481851962,
+      "learning_rate": 6.223632206786434e-06,
+      "loss": 1.1303,
+      "step": 6592
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 0.7316,
+      "step": 6592,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 0.8746,
+      "step": 6592,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 0.8615,
+      "step": 6592,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 1.0139,
+      "step": 6592,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 1.0843,
+      "step": 6592,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 1.2468,
+      "step": 6592,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 0.4102,
+      "step": 6592,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.2690040185768945,
+      "lm_loss": 0.5111,
+      "step": 6592,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.2691965252544697,
+      "grad_norm": 3.1963748838766026,
+      "learning_rate": 6.22074536210882e-06,
+      "loss": 1.109,
+      "step": 6593
+    },
+    {
+      "epoch": 1.269389031932045,
+      "grad_norm": 3.1486434960178946,
+      "learning_rate": 6.217858884823109e-06,
+      "loss": 1.1106,
+      "step": 6594
+    },
+    {
+      "epoch": 1.2695815386096205,
+      "grad_norm": 3.131291270941672,
+      "learning_rate": 6.214972775209904e-06,
+      "loss": 1.0971,
+      "step": 6595
+    },
+    {
+      "epoch": 1.269774045287196,
+      "grad_norm": 3.1404795494912183,
+      "learning_rate": 6.212087033549765e-06,
+      "loss": 1.0937,
+      "step": 6596
+    },
+    {
+      "epoch": 1.2699665519647714,
+      "grad_norm": 3.2213465906715313,
+      "learning_rate": 6.209201660123233e-06,
+      "loss": 1.0695,
+      "step": 6597
+    },
+    {
+      "epoch": 1.2701590586423466,
+      "grad_norm": 3.432325534461375,
+      "learning_rate": 6.206316655210798e-06,
+      "loss": 1.121,
+      "step": 6598
+    },
+    {
+      "epoch": 1.270351565319922,
+      "grad_norm": 3.3563740045164807,
+      "learning_rate": 6.203432019092919e-06,
+      "loss": 1.1182,
+      "step": 6599
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "grad_norm": 3.256192106349421,
+      "learning_rate": 6.2005477520500145e-06,
+      "loss": 1.0479,
+      "step": 6600
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.8173,
+      "step": 6600,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.7971,
+      "step": 6600,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 1.109,
+      "step": 6600,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.9904,
+      "step": 6600,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 1.0624,
+      "step": 6600,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.6629,
+      "step": 6600,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.9223,
+      "step": 6600,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.2705440719974974,
+      "lm_loss": 0.4428,
+      "step": 6600,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 1.2707365786750728,
+      "grad_norm": 3.4592258035144985,
+      "learning_rate": 6.197663854362483e-06,
+      "loss": 1.1377,
+      "step": 6601
+    },
+    {
+      "epoch": 1.2709290853526483,
+      "grad_norm": 3.2579943152181716,
+      "learning_rate": 6.194780326310672e-06,
+      "loss": 1.1475,
+      "step": 6602
+    },
+    {
+      "epoch": 1.2711215920302235,
+      "grad_norm": 3.2807791125826937,
+      "learning_rate": 6.191897168174892e-06,
+      "loss": 1.0765,
+      "step": 6603
+    },
+    {
+      "epoch": 1.2713140987077989,
+      "grad_norm": 3.2623462572463593,
+      "learning_rate": 6.18901438023543e-06,
+      "loss": 1.133,
+      "step": 6604
+    },
+    {
+      "epoch": 1.2715066053853743,
+      "grad_norm": 3.2587894175631957,
+      "learning_rate": 6.186131962772528e-06,
+      "loss": 1.128,
+      "step": 6605
+    },
+    {
+      "epoch": 1.2716991120629497,
+      "grad_norm": 3.3095294744490227,
+      "learning_rate": 6.183249916066389e-06,
+      "loss": 1.1393,
+      "step": 6606
+    },
+    {
+      "epoch": 1.2718916187405251,
+      "grad_norm": 3.218013281888009,
+      "learning_rate": 6.18036824039719e-06,
+      "loss": 1.1353,
+      "step": 6607
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "grad_norm": 3.2932117842421555,
+      "learning_rate": 6.177486936045067e-06,
+      "loss": 1.114,
+      "step": 6608
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 1.0368,
+      "step": 6608,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 0.6892,
+      "step": 6608,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 1.0632,
+      "step": 6608,
+      "vm_loss": 0.195
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 1.5611,
+      "step": 6608,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 0.5957,
+      "step": 6608,
+      "vm_loss": 0.1916
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 0.9344,
+      "step": 6608,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 1.3944,
+      "step": 6608,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 1.2720841254181003,
+      "lm_loss": 1.1844,
+      "step": 6608,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.2722766320956758,
+      "grad_norm": 3.193454305440564,
+      "learning_rate": 6.174606003290118e-06,
+      "loss": 1.0823,
+      "step": 6609
+    },
+    {
+      "epoch": 1.2724691387732512,
+      "grad_norm": 3.2752768440607682,
+      "learning_rate": 6.171725442412404e-06,
+      "loss": 1.119,
+      "step": 6610
+    },
+    {
+      "epoch": 1.2726616454508266,
+      "grad_norm": 3.3589643097520345,
+      "learning_rate": 6.168845253691956e-06,
+      "loss": 1.1485,
+      "step": 6611
+    },
+    {
+      "epoch": 1.272854152128402,
+      "grad_norm": 3.299537038928917,
+      "learning_rate": 6.165965437408766e-06,
+      "loss": 1.0858,
+      "step": 6612
+    },
+    {
+      "epoch": 1.2730466588059772,
+      "grad_norm": 3.3592794693852746,
+      "learning_rate": 6.163085993842786e-06,
+      "loss": 1.1386,
+      "step": 6613
+    },
+    {
+      "epoch": 1.2732391654835526,
+      "grad_norm": 3.4163801863493464,
+      "learning_rate": 6.160206923273934e-06,
+      "loss": 1.1376,
+      "step": 6614
+    },
+    {
+      "epoch": 1.273431672161128,
+      "grad_norm": 3.314220347513823,
+      "learning_rate": 6.157328225982098e-06,
+      "loss": 1.1128,
+      "step": 6615
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "grad_norm": 3.2548111536942796,
+      "learning_rate": 6.154449902247119e-06,
+      "loss": 1.089,
+      "step": 6616
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 0.8305,
+      "step": 6616,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 1.0349,
+      "step": 6616,
+      "vm_loss": 0.2181
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 0.9645,
+      "step": 6616,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 0.9943,
+      "step": 6616,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 0.9281,
+      "step": 6616,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 1.3873,
+      "step": 6616,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 1.0504,
+      "step": 6616,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.2736241788387035,
+      "lm_loss": 0.6133,
+      "step": 6616,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 1.273816685516279,
+      "grad_norm": 3.2634543760431813,
+      "learning_rate": 6.151571952348805e-06,
+      "loss": 1.0846,
+      "step": 6617
+    },
+    {
+      "epoch": 1.2740091921938541,
+      "grad_norm": 3.1855053218224816,
+      "learning_rate": 6.148694376566937e-06,
+      "loss": 1.1264,
+      "step": 6618
+    },
+    {
+      "epoch": 1.2742016988714295,
+      "grad_norm": 3.2428850014688977,
+      "learning_rate": 6.1458171751812505e-06,
+      "loss": 1.1414,
+      "step": 6619
+    },
+    {
+      "epoch": 1.274394205549005,
+      "grad_norm": 3.3641234481072706,
+      "learning_rate": 6.142940348471437e-06,
+      "loss": 1.1699,
+      "step": 6620
+    },
+    {
+      "epoch": 1.2745867122265804,
+      "grad_norm": 3.2307835337366466,
+      "learning_rate": 6.140063896717172e-06,
+      "loss": 1.1299,
+      "step": 6621
+    },
+    {
+      "epoch": 1.2747792189041558,
+      "grad_norm": 3.2586165239277554,
+      "learning_rate": 6.137187820198081e-06,
+      "loss": 1.1018,
+      "step": 6622
+    },
+    {
+      "epoch": 1.274971725581731,
+      "grad_norm": 3.4141249758897985,
+      "learning_rate": 6.134312119193752e-06,
+      "loss": 1.1981,
+      "step": 6623
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "grad_norm": 3.2327333969691168,
+      "learning_rate": 6.131436793983738e-06,
+      "loss": 1.1014,
+      "step": 6624
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.8796,
+      "step": 6624,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 1.0196,
+      "step": 6624,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.966,
+      "step": 6624,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 1.0035,
+      "step": 6624,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.9303,
+      "step": 6624,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.8744,
+      "step": 6624,
+      "vm_loss": 0.2677
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.8096,
+      "step": 6624,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.2751642322593064,
+      "lm_loss": 0.4843,
+      "step": 6624,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 1.2753567389368818,
+      "grad_norm": 3.226464147486299,
+      "learning_rate": 6.128561844847567e-06,
+      "loss": 1.1454,
+      "step": 6625
+    },
+    {
+      "epoch": 1.2755492456144573,
+      "grad_norm": 3.1811055753109296,
+      "learning_rate": 6.125687272064713e-06,
+      "loss": 1.0637,
+      "step": 6626
+    },
+    {
+      "epoch": 1.2757417522920327,
+      "grad_norm": 3.1948663720014987,
+      "learning_rate": 6.122813075914619e-06,
+      "loss": 1.0897,
+      "step": 6627
+    },
+    {
+      "epoch": 1.2759342589696079,
+      "grad_norm": 3.330322912109146,
+      "learning_rate": 6.119939256676701e-06,
+      "loss": 1.1696,
+      "step": 6628
+    },
+    {
+      "epoch": 1.2761267656471835,
+      "grad_norm": 3.1985282596025346,
+      "learning_rate": 6.117065814630327e-06,
+      "loss": 1.0874,
+      "step": 6629
+    },
+    {
+      "epoch": 1.2763192723247587,
+      "grad_norm": 3.2024199818159347,
+      "learning_rate": 6.114192750054833e-06,
+      "loss": 1.1188,
+      "step": 6630
+    },
+    {
+      "epoch": 1.2765117790023341,
+      "grad_norm": 3.1585191029462982,
+      "learning_rate": 6.11132006322952e-06,
+      "loss": 1.1118,
+      "step": 6631
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "grad_norm": 3.2365758802566145,
+      "learning_rate": 6.108447754433644e-06,
+      "loss": 1.1111,
+      "step": 6632
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 0.9974,
+      "step": 6632,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 0.7836,
+      "step": 6632,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 1.1529,
+      "step": 6632,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 0.8055,
+      "step": 6632,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 1.0699,
+      "step": 6632,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 1.4923,
+      "step": 6632,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 1.107,
+      "step": 6632,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.2767042856799096,
+      "lm_loss": 1.187,
+      "step": 6632,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.2768967923574848,
+      "grad_norm": 3.278623149511034,
+      "learning_rate": 6.105575823946437e-06,
+      "loss": 1.1268,
+      "step": 6633
+    },
+    {
+      "epoch": 1.2770892990350604,
+      "grad_norm": 3.2865357228763497,
+      "learning_rate": 6.10270427204708e-06,
+      "loss": 1.1155,
+      "step": 6634
+    },
+    {
+      "epoch": 1.2772818057126356,
+      "grad_norm": 3.1019539250023143,
+      "learning_rate": 6.099833099014734e-06,
+      "loss": 1.0594,
+      "step": 6635
+    },
+    {
+      "epoch": 1.277474312390211,
+      "grad_norm": 3.2910216827595646,
+      "learning_rate": 6.096962305128506e-06,
+      "loss": 1.1217,
+      "step": 6636
+    },
+    {
+      "epoch": 1.2776668190677865,
+      "grad_norm": 3.369657710892053,
+      "learning_rate": 6.094091890667473e-06,
+      "loss": 1.213,
+      "step": 6637
+    },
+    {
+      "epoch": 1.2778593257453619,
+      "grad_norm": 3.3943682948687557,
+      "learning_rate": 6.091221855910685e-06,
+      "loss": 1.1327,
+      "step": 6638
+    },
+    {
+      "epoch": 1.2780518324229373,
+      "grad_norm": 3.317886043429176,
+      "learning_rate": 6.088352201137139e-06,
+      "loss": 1.1677,
+      "step": 6639
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "grad_norm": 3.288379790735745,
+      "learning_rate": 6.085482926625803e-06,
+      "loss": 1.1482,
+      "step": 6640
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.6361,
+      "step": 6640,
+      "vm_loss": 0.1043
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.4126,
+      "step": 6640,
+      "vm_loss": 0.2592
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 1.4527,
+      "step": 6640,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.9723,
+      "step": 6640,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.8676,
+      "step": 6640,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 1.1842,
+      "step": 6640,
+      "vm_loss": 0.2359
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.89,
+      "step": 6640,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 1.2782443391005125,
+      "lm_loss": 0.6887,
+      "step": 6640,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 1.278436845778088,
+      "grad_norm": 3.0646674234816405,
+      "learning_rate": 6.082614032655605e-06,
+      "loss": 1.0738,
+      "step": 6641
+    },
+    {
+      "epoch": 1.2786293524556633,
+      "grad_norm": 3.276718687991323,
+      "learning_rate": 6.079745519505446e-06,
+      "loss": 1.0999,
+      "step": 6642
+    },
+    {
+      "epoch": 1.2788218591332388,
+      "grad_norm": 3.2995560943464635,
+      "learning_rate": 6.076877387454176e-06,
+      "loss": 1.1246,
+      "step": 6643
+    },
+    {
+      "epoch": 1.2790143658108142,
+      "grad_norm": 3.2146849271138866,
+      "learning_rate": 6.074009636780612e-06,
+      "loss": 1.1043,
+      "step": 6644
+    },
+    {
+      "epoch": 1.2792068724883894,
+      "grad_norm": 3.390724218021271,
+      "learning_rate": 6.071142267763541e-06,
+      "loss": 1.1727,
+      "step": 6645
+    },
+    {
+      "epoch": 1.2793993791659648,
+      "grad_norm": 3.2144347061724132,
+      "learning_rate": 6.068275280681709e-06,
+      "loss": 1.1362,
+      "step": 6646
+    },
+    {
+      "epoch": 1.2795918858435402,
+      "grad_norm": 3.2317995678324802,
+      "learning_rate": 6.065408675813816e-06,
+      "loss": 1.1073,
+      "step": 6647
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "grad_norm": 3.185002066065946,
+      "learning_rate": 6.0625424534385425e-06,
+      "loss": 1.1697,
+      "step": 6648
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 0.92,
+      "step": 6648,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 1.0522,
+      "step": 6648,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 1.0565,
+      "step": 6648,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 0.9726,
+      "step": 6648,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 0.8785,
+      "step": 6648,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 1.0389,
+      "step": 6648,
+      "vm_loss": 0.1023
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 0.9572,
+      "step": 6648,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 1.2797843925211156,
+      "lm_loss": 0.5625,
+      "step": 6648,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.279976899198691,
+      "grad_norm": 3.325125241980314,
+      "learning_rate": 6.0596766138345176e-06,
+      "loss": 1.1034,
+      "step": 6649
+    },
+    {
+      "epoch": 1.2801694058762663,
+      "grad_norm": 3.362687132237075,
+      "learning_rate": 6.056811157280333e-06,
+      "loss": 1.2115,
+      "step": 6650
+    },
+    {
+      "epoch": 1.2803619125538417,
+      "grad_norm": 3.267660427938564,
+      "learning_rate": 6.053946084054554e-06,
+      "loss": 1.1128,
+      "step": 6651
+    },
+    {
+      "epoch": 1.280554419231417,
+      "grad_norm": 3.3618210873818697,
+      "learning_rate": 6.051081394435702e-06,
+      "loss": 1.1879,
+      "step": 6652
+    },
+    {
+      "epoch": 1.2807469259089925,
+      "grad_norm": 3.4180425164302513,
+      "learning_rate": 6.048217088702258e-06,
+      "loss": 1.1624,
+      "step": 6653
+    },
+    {
+      "epoch": 1.280939432586568,
+      "grad_norm": 3.3554021744807034,
+      "learning_rate": 6.045353167132671e-06,
+      "loss": 1.1347,
+      "step": 6654
+    },
+    {
+      "epoch": 1.2811319392641431,
+      "grad_norm": 3.1485502998727277,
+      "learning_rate": 6.042489630005352e-06,
+      "loss": 1.1071,
+      "step": 6655
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "grad_norm": 3.22438501287988,
+      "learning_rate": 6.039626477598673e-06,
+      "loss": 1.0869,
+      "step": 6656
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 0.9497,
+      "step": 6656,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 0.762,
+      "step": 6656,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 1.0301,
+      "step": 6656,
+      "vm_loss": 0.2305
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 0.7932,
+      "step": 6656,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 1.4329,
+      "step": 6656,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 0.8006,
+      "step": 6656,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 1.0734,
+      "step": 6656,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 1.2813244459417186,
+      "lm_loss": 0.8487,
+      "step": 6656,
+      "vm_loss": 0.2176
+    },
+    {
+      "epoch": 1.281516952619294,
+      "grad_norm": 3.3263745981245023,
+      "learning_rate": 6.036763710190968e-06,
+      "loss": 1.1661,
+      "step": 6657
+    },
+    {
+      "epoch": 1.2817094592968694,
+      "grad_norm": 3.3593013411646417,
+      "learning_rate": 6.033901328060532e-06,
+      "loss": 1.1653,
+      "step": 6658
+    },
+    {
+      "epoch": 1.2819019659744448,
+      "grad_norm": 3.113484541583496,
+      "learning_rate": 6.031039331485631e-06,
+      "loss": 1.0612,
+      "step": 6659
+    },
+    {
+      "epoch": 1.28209447265202,
+      "grad_norm": 3.1606516864254615,
+      "learning_rate": 6.028177720744486e-06,
+      "loss": 1.1231,
+      "step": 6660
+    },
+    {
+      "epoch": 1.2822869793295955,
+      "grad_norm": 3.257141326529588,
+      "learning_rate": 6.025316496115276e-06,
+      "loss": 1.1159,
+      "step": 6661
+    },
+    {
+      "epoch": 1.2824794860071709,
+      "grad_norm": 3.290057076531557,
+      "learning_rate": 6.022455657876158e-06,
+      "loss": 1.1142,
+      "step": 6662
+    },
+    {
+      "epoch": 1.2826719926847463,
+      "grad_norm": 3.377804369433278,
+      "learning_rate": 6.019595206305237e-06,
+      "loss": 1.1586,
+      "step": 6663
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "grad_norm": 3.2187176409292833,
+      "learning_rate": 6.016735141680587e-06,
+      "loss": 1.0753,
+      "step": 6664
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.8042,
+      "step": 6664,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.8567,
+      "step": 6664,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.7168,
+      "step": 6664,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.6396,
+      "step": 6664,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.8455,
+      "step": 6664,
+      "vm_loss": 0.1763
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 1.1142,
+      "step": 6664,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 0.8613,
+      "step": 6664,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.2828644993623217,
+      "lm_loss": 1.1116,
+      "step": 6664,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.283057006039897,
+      "grad_norm": 3.2248676846011546,
+      "learning_rate": 6.013875464280239e-06,
+      "loss": 1.0863,
+      "step": 6665
+    },
+    {
+      "epoch": 1.2832495127174723,
+      "grad_norm": 3.1852668351607805,
+      "learning_rate": 6.011016174382196e-06,
+      "loss": 1.075,
+      "step": 6666
+    },
+    {
+      "epoch": 1.2834420193950478,
+      "grad_norm": 3.4590980045580104,
+      "learning_rate": 6.008157272264415e-06,
+      "loss": 1.2116,
+      "step": 6667
+    },
+    {
+      "epoch": 1.2836345260726232,
+      "grad_norm": 3.294351295793464,
+      "learning_rate": 6.005298758204814e-06,
+      "loss": 1.1388,
+      "step": 6668
+    },
+    {
+      "epoch": 1.2838270327501986,
+      "grad_norm": 3.162108651038212,
+      "learning_rate": 6.002440632481285e-06,
+      "loss": 1.1085,
+      "step": 6669
+    },
+    {
+      "epoch": 1.2840195394277738,
+      "grad_norm": 3.3218715312943092,
+      "learning_rate": 5.999582895371669e-06,
+      "loss": 1.1459,
+      "step": 6670
+    },
+    {
+      "epoch": 1.2842120461053492,
+      "grad_norm": 3.2533756587593303,
+      "learning_rate": 5.996725547153773e-06,
+      "loss": 1.1284,
+      "step": 6671
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "grad_norm": 3.238942824356261,
+      "learning_rate": 5.993868588105375e-06,
+      "loss": 1.1226,
+      "step": 6672
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 0.7356,
+      "step": 6672,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 1.0102,
+      "step": 6672,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 1.9887,
+      "step": 6672,
+      "vm_loss": 0.0918
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 0.9675,
+      "step": 6672,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 0.7122,
+      "step": 6672,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 0.938,
+      "step": 6672,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 0.8966,
+      "step": 6672,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.2844045527829246,
+      "lm_loss": 1.2908,
+      "step": 6672,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 1.2845970594605,
+      "grad_norm": 3.2062820040120363,
+      "learning_rate": 5.991012018504201e-06,
+      "loss": 1.0981,
+      "step": 6673
+    },
+    {
+      "epoch": 1.2847895661380755,
+      "grad_norm": 3.244400001080694,
+      "learning_rate": 5.988155838627951e-06,
+      "loss": 1.1547,
+      "step": 6674
+    },
+    {
+      "epoch": 1.2849820728156507,
+      "grad_norm": 3.1113397184566853,
+      "learning_rate": 5.985300048754278e-06,
+      "loss": 1.052,
+      "step": 6675
+    },
+    {
+      "epoch": 1.285174579493226,
+      "grad_norm": 3.2833974257504477,
+      "learning_rate": 5.9824446491608035e-06,
+      "loss": 1.1253,
+      "step": 6676
+    },
+    {
+      "epoch": 1.2853670861708015,
+      "grad_norm": 3.152368089179338,
+      "learning_rate": 5.979589640125109e-06,
+      "loss": 1.0917,
+      "step": 6677
+    },
+    {
+      "epoch": 1.285559592848377,
+      "grad_norm": 3.2092598257476683,
+      "learning_rate": 5.976735021924735e-06,
+      "loss": 1.0606,
+      "step": 6678
+    },
+    {
+      "epoch": 1.2857520995259524,
+      "grad_norm": 3.477398016078983,
+      "learning_rate": 5.973880794837194e-06,
+      "loss": 1.176,
+      "step": 6679
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "grad_norm": 3.35540579708099,
+      "learning_rate": 5.971026959139947e-06,
+      "loss": 1.1069,
+      "step": 6680
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 1.0275,
+      "step": 6680,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 0.9473,
+      "step": 6680,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 1.066,
+      "step": 6680,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 0.7842,
+      "step": 6680,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 0.6261,
+      "step": 6680,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 0.5599,
+      "step": 6680,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 1.114,
+      "step": 6680,
+      "vm_loss": 0.1875
+    },
+    {
+      "epoch": 1.2859446062035276,
+      "lm_loss": 0.9851,
+      "step": 6680,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.286137112881103,
+      "grad_norm": 3.342037762460158,
+      "learning_rate": 5.968173515110427e-06,
+      "loss": 1.1149,
+      "step": 6681
+    },
+    {
+      "epoch": 1.2863296195586784,
+      "grad_norm": 3.1508544719142715,
+      "learning_rate": 5.965320463026019e-06,
+      "loss": 1.0458,
+      "step": 6682
+    },
+    {
+      "epoch": 1.2865221262362538,
+      "grad_norm": 3.277391870723528,
+      "learning_rate": 5.962467803164083e-06,
+      "loss": 1.0688,
+      "step": 6683
+    },
+    {
+      "epoch": 1.2867146329138293,
+      "grad_norm": 3.173183626546501,
+      "learning_rate": 5.959615535801933e-06,
+      "loss": 1.1194,
+      "step": 6684
+    },
+    {
+      "epoch": 1.2869071395914045,
+      "grad_norm": 3.207537977150246,
+      "learning_rate": 5.956763661216841e-06,
+      "loss": 1.0952,
+      "step": 6685
+    },
+    {
+      "epoch": 1.2870996462689799,
+      "grad_norm": 3.33364042366414,
+      "learning_rate": 5.953912179686052e-06,
+      "loss": 1.1525,
+      "step": 6686
+    },
+    {
+      "epoch": 1.2872921529465553,
+      "grad_norm": 3.2416636686667455,
+      "learning_rate": 5.951061091486765e-06,
+      "loss": 1.1358,
+      "step": 6687
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "grad_norm": 3.166505654927115,
+      "learning_rate": 5.948210396896137e-06,
+      "loss": 1.0706,
+      "step": 6688
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 1.3101,
+      "step": 6688,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.8622,
+      "step": 6688,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.893,
+      "step": 6688,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.8109,
+      "step": 6688,
+      "vm_loss": 0.1127
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.82,
+      "step": 6688,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.951,
+      "step": 6688,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 1.0102,
+      "step": 6688,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 1.2874846596241307,
+      "lm_loss": 0.7994,
+      "step": 6688,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.2876771663017061,
+      "grad_norm": 3.1673670985766935,
+      "learning_rate": 5.9453600961913e-06,
+      "loss": 1.0801,
+      "step": 6689
+    },
+    {
+      "epoch": 1.2878696729792813,
+      "grad_norm": 3.243468298626356,
+      "learning_rate": 5.942510189649333e-06,
+      "loss": 1.1216,
+      "step": 6690
+    },
+    {
+      "epoch": 1.2880621796568568,
+      "grad_norm": 3.302040344549393,
+      "learning_rate": 5.939660677547288e-06,
+      "loss": 1.1427,
+      "step": 6691
+    },
+    {
+      "epoch": 1.2882546863344322,
+      "grad_norm": 3.241573382388179,
+      "learning_rate": 5.936811560162169e-06,
+      "loss": 1.0603,
+      "step": 6692
+    },
+    {
+      "epoch": 1.2884471930120076,
+      "grad_norm": 3.214616070686676,
+      "learning_rate": 5.9339628377709525e-06,
+      "loss": 1.0731,
+      "step": 6693
+    },
+    {
+      "epoch": 1.288639699689583,
+      "grad_norm": 3.287339342192405,
+      "learning_rate": 5.931114510650566e-06,
+      "loss": 1.1013,
+      "step": 6694
+    },
+    {
+      "epoch": 1.2888322063671582,
+      "grad_norm": 3.356604233703808,
+      "learning_rate": 5.928266579077906e-06,
+      "loss": 1.1523,
+      "step": 6695
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "grad_norm": 3.325152500691127,
+      "learning_rate": 5.925419043329827e-06,
+      "loss": 1.1452,
+      "step": 6696
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.7924,
+      "step": 6696,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 1.1074,
+      "step": 6696,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.6835,
+      "step": 6696,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.813,
+      "step": 6696,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.6944,
+      "step": 6696,
+      "vm_loss": 0.2483
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.7973,
+      "step": 6696,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 0.6998,
+      "step": 6696,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.2890247130447339,
+      "lm_loss": 1.7891,
+      "step": 6696,
+      "vm_loss": 0.2293
+    },
+    {
+      "epoch": 1.289217219722309,
+      "grad_norm": 3.2846629066686273,
+      "learning_rate": 5.922571903683147e-06,
+      "loss": 1.1362,
+      "step": 6697
+    },
+    {
+      "epoch": 1.2894097263998845,
+      "grad_norm": 3.3136833032332187,
+      "learning_rate": 5.919725160414645e-06,
+      "loss": 1.1666,
+      "step": 6698
+    },
+    {
+      "epoch": 1.28960223307746,
+      "grad_norm": 3.1327110991150757,
+      "learning_rate": 5.9168788138010545e-06,
+      "loss": 1.0954,
+      "step": 6699
+    },
+    {
+      "epoch": 1.2897947397550353,
+      "grad_norm": 3.2799814769159092,
+      "learning_rate": 5.9140328641190855e-06,
+      "loss": 1.1077,
+      "step": 6700
+    },
+    {
+      "epoch": 1.2899872464326108,
+      "grad_norm": 3.253586257257136,
+      "learning_rate": 5.911187311645398e-06,
+      "loss": 1.153,
+      "step": 6701
+    },
+    {
+      "epoch": 1.290179753110186,
+      "grad_norm": 3.219647443264257,
+      "learning_rate": 5.9083421566566105e-06,
+      "loss": 1.1153,
+      "step": 6702
+    },
+    {
+      "epoch": 1.2903722597877614,
+      "grad_norm": 3.2309199762763443,
+      "learning_rate": 5.905497399429316e-06,
+      "loss": 1.1125,
+      "step": 6703
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "grad_norm": 3.1445968019228556,
+      "learning_rate": 5.902653040240059e-06,
+      "loss": 1.0809,
+      "step": 6704
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 1.1049,
+      "step": 6704,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 0.8855,
+      "step": 6704,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 0.6965,
+      "step": 6704,
+      "vm_loss": 0.1079
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 0.4566,
+      "step": 6704,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 1.2037,
+      "step": 6704,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 1.0137,
+      "step": 6704,
+      "vm_loss": 0.2202
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 0.9589,
+      "step": 6704,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 1.2905647664653368,
+      "lm_loss": 0.7211,
+      "step": 6704,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.2907572731429122,
+      "grad_norm": 3.163381937510988,
+      "learning_rate": 5.8998090793653425e-06,
+      "loss": 1.0859,
+      "step": 6705
+    },
+    {
+      "epoch": 1.2909497798204876,
+      "grad_norm": 3.234409170746116,
+      "learning_rate": 5.896965517081645e-06,
+      "loss": 1.1419,
+      "step": 6706
+    },
+    {
+      "epoch": 1.2911422864980628,
+      "grad_norm": 3.25906867065638,
+      "learning_rate": 5.894122353665393e-06,
+      "loss": 1.0985,
+      "step": 6707
+    },
+    {
+      "epoch": 1.2913347931756383,
+      "grad_norm": 3.225383743928816,
+      "learning_rate": 5.891279589392975e-06,
+      "loss": 1.0706,
+      "step": 6708
+    },
+    {
+      "epoch": 1.2915272998532137,
+      "grad_norm": 3.309608742120105,
+      "learning_rate": 5.888437224540745e-06,
+      "loss": 1.0958,
+      "step": 6709
+    },
+    {
+      "epoch": 1.291719806530789,
+      "grad_norm": 3.3990285576293666,
+      "learning_rate": 5.885595259385022e-06,
+      "loss": 1.1407,
+      "step": 6710
+    },
+    {
+      "epoch": 1.2919123132083645,
+      "grad_norm": 3.2756592560484794,
+      "learning_rate": 5.882753694202079e-06,
+      "loss": 1.0561,
+      "step": 6711
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "grad_norm": 3.3420990240590873,
+      "learning_rate": 5.879912529268148e-06,
+      "loss": 1.0662,
+      "step": 6712
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 1.0454,
+      "step": 6712,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 1.0143,
+      "step": 6712,
+      "vm_loss": 0.1177
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 1.5339,
+      "step": 6712,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 0.9684,
+      "step": 6712,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 0.6806,
+      "step": 6712,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 0.5913,
+      "step": 6712,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 0.819,
+      "step": 6712,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.2921048198859397,
+      "lm_loss": 0.8451,
+      "step": 6712,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 1.2922973265635151,
+      "grad_norm": 3.2263486408927813,
+      "learning_rate": 5.877071764859434e-06,
+      "loss": 1.0681,
+      "step": 6713
+    },
+    {
+      "epoch": 1.2924898332410906,
+      "grad_norm": 3.270108977460173,
+      "learning_rate": 5.874231401252091e-06,
+      "loss": 1.1024,
+      "step": 6714
+    },
+    {
+      "epoch": 1.292682339918666,
+      "grad_norm": 3.1513460724551288,
+      "learning_rate": 5.87139143872224e-06,
+      "loss": 1.0754,
+      "step": 6715
+    },
+    {
+      "epoch": 1.2928748465962414,
+      "grad_norm": 3.2395103295286867,
+      "learning_rate": 5.868551877545962e-06,
+      "loss": 1.1047,
+      "step": 6716
+    },
+    {
+      "epoch": 1.2930673532738166,
+      "grad_norm": 3.247469995101324,
+      "learning_rate": 5.865712717999297e-06,
+      "loss": 1.122,
+      "step": 6717
+    },
+    {
+      "epoch": 1.293259859951392,
+      "grad_norm": 3.2125923890321473,
+      "learning_rate": 5.862873960358251e-06,
+      "loss": 1.0786,
+      "step": 6718
+    },
+    {
+      "epoch": 1.2934523666289675,
+      "grad_norm": 3.2457073499096922,
+      "learning_rate": 5.86003560489878e-06,
+      "loss": 1.0713,
+      "step": 6719
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "grad_norm": 3.3374953157867826,
+      "learning_rate": 5.857197651896819e-06,
+      "loss": 1.1625,
+      "step": 6720
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 0.8841,
+      "step": 6720,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 0.9326,
+      "step": 6720,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 1.6497,
+      "step": 6720,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 0.9635,
+      "step": 6720,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 0.6027,
+      "step": 6720,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 1.0785,
+      "step": 6720,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 0.5596,
+      "step": 6720,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 1.2936448733065429,
+      "lm_loss": 1.061,
+      "step": 6720,
+      "vm_loss": 0.2327
+    },
+    {
+      "epoch": 1.2938373799841183,
+      "grad_norm": 3.297782511910469,
+      "learning_rate": 5.854360101628247e-06,
+      "loss": 1.1376,
+      "step": 6721
+    },
+    {
+      "epoch": 1.2940298866616935,
+      "grad_norm": 3.3118415906323686,
+      "learning_rate": 5.851522954368909e-06,
+      "loss": 1.1761,
+      "step": 6722
+    },
+    {
+      "epoch": 1.294222393339269,
+      "grad_norm": 3.2205943406359,
+      "learning_rate": 5.8486862103946176e-06,
+      "loss": 1.1086,
+      "step": 6723
+    },
+    {
+      "epoch": 1.2944149000168443,
+      "grad_norm": 3.175018010738224,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 1.0791,
+      "step": 6724
+    },
+    {
+      "epoch": 1.2946074066944198,
+      "grad_norm": 3.1853281329942096,
+      "learning_rate": 5.843013933404197e-06,
+      "loss": 1.0898,
+      "step": 6725
+    },
+    {
+      "epoch": 1.2947999133719952,
+      "grad_norm": 3.18817508751911,
+      "learning_rate": 5.8401784009394815e-06,
+      "loss": 1.0988,
+      "step": 6726
+    },
+    {
+      "epoch": 1.2949924200495704,
+      "grad_norm": 3.028986933563594,
+      "learning_rate": 5.837343272862649e-06,
+      "loss": 1.0568,
+      "step": 6727
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "grad_norm": 3.286719058039817,
+      "learning_rate": 5.834508549449305e-06,
+      "loss": 1.0896,
+      "step": 6728
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.8548,
+      "step": 6728,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.6273,
+      "step": 6728,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.8368,
+      "step": 6728,
+      "vm_loss": 0.2093
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 1.0741,
+      "step": 6728,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.8913,
+      "step": 6728,
+      "vm_loss": 0.2136
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 1.2988,
+      "step": 6728,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.9203,
+      "step": 6728,
+      "vm_loss": 0.1214
+    },
+    {
+      "epoch": 1.2951849267271458,
+      "lm_loss": 0.8856,
+      "step": 6728,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.2953774334047212,
+      "grad_norm": 3.3178976039301555,
+      "learning_rate": 5.831674230975021e-06,
+      "loss": 1.1736,
+      "step": 6729
+    },
+    {
+      "epoch": 1.2955699400822966,
+      "grad_norm": 3.2076307799395494,
+      "learning_rate": 5.828840317715331e-06,
+      "loss": 1.1022,
+      "step": 6730
+    },
+    {
+      "epoch": 1.295762446759872,
+      "grad_norm": 3.2225902895843124,
+      "learning_rate": 5.82600680994573e-06,
+      "loss": 1.1005,
+      "step": 6731
+    },
+    {
+      "epoch": 1.2959549534374473,
+      "grad_norm": 3.1994529657963153,
+      "learning_rate": 5.823173707941665e-06,
+      "loss": 1.0855,
+      "step": 6732
+    },
+    {
+      "epoch": 1.2961474601150227,
+      "grad_norm": 3.2928044386054993,
+      "learning_rate": 5.8203410119785515e-06,
+      "loss": 1.1189,
+      "step": 6733
+    },
+    {
+      "epoch": 1.296339966792598,
+      "grad_norm": 3.1860256568116854,
+      "learning_rate": 5.8175087223317674e-06,
+      "loss": 1.0791,
+      "step": 6734
+    },
+    {
+      "epoch": 1.2965324734701735,
+      "grad_norm": 3.3695415782671314,
+      "learning_rate": 5.814676839276648e-06,
+      "loss": 1.0887,
+      "step": 6735
+    },
+    {
+      "epoch": 1.296724980147749,
+      "grad_norm": 3.2316577594511116,
+      "learning_rate": 5.811845363088477e-06,
+      "loss": 1.07,
+      "step": 6736
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 1.2677,
+      "step": 6736,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 0.5743,
+      "step": 6736,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 0.8285,
+      "step": 6736,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 0.8711,
+      "step": 6736,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 0.66,
+      "step": 6736,
+      "vm_loss": 0.1026
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 0.9126,
+      "step": 6736,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 1.3324,
+      "step": 6736,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 1.296724980147749,
+      "lm_loss": 1.1221,
+      "step": 6736,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.2969174868253242,
+      "grad_norm": 3.1988470027673,
+      "learning_rate": 5.809014294042528e-06,
+      "loss": 1.0707,
+      "step": 6737
+    },
+    {
+      "epoch": 1.2971099935028996,
+      "grad_norm": 3.2796317783655953,
+      "learning_rate": 5.806183632414003e-06,
+      "loss": 1.1394,
+      "step": 6738
+    },
+    {
+      "epoch": 1.297302500180475,
+      "grad_norm": 3.345223345634217,
+      "learning_rate": 5.8033533784780885e-06,
+      "loss": 1.1356,
+      "step": 6739
+    },
+    {
+      "epoch": 1.2974950068580504,
+      "grad_norm": 3.3762879257540472,
+      "learning_rate": 5.800523532509907e-06,
+      "loss": 1.1424,
+      "step": 6740
+    },
+    {
+      "epoch": 1.2976875135356258,
+      "grad_norm": 3.315179752645858,
+      "learning_rate": 5.797694094784574e-06,
+      "loss": 1.1435,
+      "step": 6741
+    },
+    {
+      "epoch": 1.297880020213201,
+      "grad_norm": 3.155546932446427,
+      "learning_rate": 5.794865065577134e-06,
+      "loss": 1.052,
+      "step": 6742
+    },
+    {
+      "epoch": 1.2980725268907765,
+      "grad_norm": 3.1691607236808803,
+      "learning_rate": 5.79203644516261e-06,
+      "loss": 1.1403,
+      "step": 6743
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "grad_norm": 3.194397428183179,
+      "learning_rate": 5.7892082338159774e-06,
+      "loss": 1.121,
+      "step": 6744
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 0.8229,
+      "step": 6744,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 0.6625,
+      "step": 6744,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 1.0013,
+      "step": 6744,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 0.5418,
+      "step": 6744,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 1.0268,
+      "step": 6744,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 1.1461,
+      "step": 6744,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 0.8486,
+      "step": 6744,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 1.2982650335683519,
+      "lm_loss": 1.0587,
+      "step": 6744,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.2984575402459273,
+      "grad_norm": 3.070647441645855,
+      "learning_rate": 5.786380431812178e-06,
+      "loss": 1.0944,
+      "step": 6745
+    },
+    {
+      "epoch": 1.2986500469235027,
+      "grad_norm": 3.3018689465183826,
+      "learning_rate": 5.783553039426107e-06,
+      "loss": 1.1229,
+      "step": 6746
+    },
+    {
+      "epoch": 1.298842553601078,
+      "grad_norm": 3.319140900202501,
+      "learning_rate": 5.78072605693263e-06,
+      "loss": 1.1085,
+      "step": 6747
+    },
+    {
+      "epoch": 1.2990350602786533,
+      "grad_norm": 3.1612195114340134,
+      "learning_rate": 5.777899484606555e-06,
+      "loss": 1.0235,
+      "step": 6748
+    },
+    {
+      "epoch": 1.2992275669562288,
+      "grad_norm": 3.362802521212921,
+      "learning_rate": 5.775073322722669e-06,
+      "loss": 1.1054,
+      "step": 6749
+    },
+    {
+      "epoch": 1.2994200736338042,
+      "grad_norm": 3.1598349090684983,
+      "learning_rate": 5.7722475715557056e-06,
+      "loss": 1.0791,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2996125803113796,
+      "grad_norm": 3.3170068075520796,
+      "learning_rate": 5.769422231380369e-06,
+      "loss": 1.0847,
+      "step": 6751
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "grad_norm": 3.3054676873467272,
+      "learning_rate": 5.766597302471319e-06,
+      "loss": 1.1031,
+      "step": 6752
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.9478,
+      "step": 6752,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.9842,
+      "step": 6752,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 1.1728,
+      "step": 6752,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.643,
+      "step": 6752,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.8266,
+      "step": 6752,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.5803,
+      "step": 6752,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 0.5385,
+      "step": 6752,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.2998050869889548,
+      "lm_loss": 1.4366,
+      "step": 6752,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 1.2999975936665302,
+      "grad_norm": 3.2693039802953687,
+      "learning_rate": 5.763772785103166e-06,
+      "loss": 1.0795,
+      "step": 6753
+    },
+    {
+      "epoch": 1.3001901003441056,
+      "grad_norm": 3.453954912595241,
+      "learning_rate": 5.760948679550502e-06,
+      "loss": 1.1465,
+      "step": 6754
+    },
+    {
+      "epoch": 1.300382607021681,
+      "grad_norm": 3.210306459377581,
+      "learning_rate": 5.7581249860878566e-06,
+      "loss": 1.0641,
+      "step": 6755
+    },
+    {
+      "epoch": 1.3005751136992565,
+      "grad_norm": 3.2652307946047734,
+      "learning_rate": 5.755301704989735e-06,
+      "loss": 1.1324,
+      "step": 6756
+    },
+    {
+      "epoch": 1.3007676203768317,
+      "grad_norm": 3.337089800878578,
+      "learning_rate": 5.752478836530586e-06,
+      "loss": 1.1125,
+      "step": 6757
+    },
+    {
+      "epoch": 1.3009601270544073,
+      "grad_norm": 3.2797625564655655,
+      "learning_rate": 5.7496563809848434e-06,
+      "loss": 1.1189,
+      "step": 6758
+    },
+    {
+      "epoch": 1.3011526337319825,
+      "grad_norm": 3.139538214615867,
+      "learning_rate": 5.746834338626874e-06,
+      "loss": 1.0742,
+      "step": 6759
+    },
+    {
+      "epoch": 1.301345140409558,
+      "grad_norm": 3.221486001443372,
+      "learning_rate": 5.7440127097310214e-06,
+      "loss": 1.0935,
+      "step": 6760
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 1.3481,
+      "step": 6760,
+      "vm_loss": 0.2456
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 0.9318,
+      "step": 6760,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 1.2141,
+      "step": 6760,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 1.0681,
+      "step": 6760,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 0.7319,
+      "step": 6760,
+      "vm_loss": 0.0679
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 0.8683,
+      "step": 6760,
+      "vm_loss": 0.248
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 1.4273,
+      "step": 6760,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.301345140409558,
+      "lm_loss": 0.8325,
+      "step": 6760,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.3015376470871334,
+      "grad_norm": 3.1533545710551114,
+      "learning_rate": 5.741191494571583e-06,
+      "loss": 1.1233,
+      "step": 6761
+    },
+    {
+      "epoch": 1.3017301537647086,
+      "grad_norm": 3.272179948393788,
+      "learning_rate": 5.738370693422822e-06,
+      "loss": 1.0779,
+      "step": 6762
+    },
+    {
+      "epoch": 1.3019226604422842,
+      "grad_norm": 3.214616809835346,
+      "learning_rate": 5.735550306558943e-06,
+      "loss": 1.0909,
+      "step": 6763
+    },
+    {
+      "epoch": 1.3021151671198594,
+      "grad_norm": 3.3897524145909483,
+      "learning_rate": 5.7327303342541415e-06,
+      "loss": 1.083,
+      "step": 6764
+    },
+    {
+      "epoch": 1.3023076737974348,
+      "grad_norm": 3.240392478199875,
+      "learning_rate": 5.729910776782542e-06,
+      "loss": 1.1251,
+      "step": 6765
+    },
+    {
+      "epoch": 1.3025001804750103,
+      "grad_norm": 3.309930806180067,
+      "learning_rate": 5.727091634418243e-06,
+      "loss": 1.1355,
+      "step": 6766
+    },
+    {
+      "epoch": 1.3026926871525857,
+      "grad_norm": 3.1743223658690334,
+      "learning_rate": 5.724272907435307e-06,
+      "loss": 1.0664,
+      "step": 6767
+    },
+    {
+      "epoch": 1.302885193830161,
+      "grad_norm": 3.3237043681773386,
+      "learning_rate": 5.721454596107747e-06,
+      "loss": 1.0529,
+      "step": 6768
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 1.0997,
+      "step": 6768,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 1.0425,
+      "step": 6768,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 1.4424,
+      "step": 6768,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 0.7211,
+      "step": 6768,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 1.0533,
+      "step": 6768,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 0.6738,
+      "step": 6768,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 1.3093,
+      "step": 6768,
+      "vm_loss": 0.0997
+    },
+    {
+      "epoch": 1.302885193830161,
+      "lm_loss": 0.705,
+      "step": 6768,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 1.3030777005077363,
+      "grad_norm": 3.3347147320347985,
+      "learning_rate": 5.718636700709538e-06,
+      "loss": 1.1053,
+      "step": 6769
+    },
+    {
+      "epoch": 1.3032702071853117,
+      "grad_norm": 3.4127874156542446,
+      "learning_rate": 5.7158192215146154e-06,
+      "loss": 1.1561,
+      "step": 6770
+    },
+    {
+      "epoch": 1.3034627138628871,
+      "grad_norm": 3.1944048337543465,
+      "learning_rate": 5.713002158796874e-06,
+      "loss": 1.1098,
+      "step": 6771
+    },
+    {
+      "epoch": 1.3036552205404626,
+      "grad_norm": 3.2932699944967063,
+      "learning_rate": 5.710185512830171e-06,
+      "loss": 1.1432,
+      "step": 6772
+    },
+    {
+      "epoch": 1.303847727218038,
+      "grad_norm": 3.2294556580828027,
+      "learning_rate": 5.707369283888322e-06,
+      "loss": 1.0967,
+      "step": 6773
+    },
+    {
+      "epoch": 1.3040402338956132,
+      "grad_norm": 3.2082578451820116,
+      "learning_rate": 5.70455347224509e-06,
+      "loss": 1.0733,
+      "step": 6774
+    },
+    {
+      "epoch": 1.3042327405731886,
+      "grad_norm": 3.246994024496379,
+      "learning_rate": 5.701738078174224e-06,
+      "loss": 1.099,
+      "step": 6775
+    },
+    {
+      "epoch": 1.304425247250764,
+      "grad_norm": 3.2234213379107777,
+      "learning_rate": 5.6989231019494016e-06,
+      "loss": 1.1293,
+      "step": 6776
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 0.9508,
+      "step": 6776,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 1.07,
+      "step": 6776,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 0.7285,
+      "step": 6776,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 1.206,
+      "step": 6776,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 0.8687,
+      "step": 6776,
+      "vm_loss": 0.1639
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 1.5568,
+      "step": 6776,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 0.6493,
+      "step": 6776,
+      "vm_loss": 0.1538
+    },
+    {
+      "epoch": 1.304425247250764,
+      "lm_loss": 0.9439,
+      "step": 6776,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 1.3046177539283395,
+      "grad_norm": 3.3146946977391094,
+      "learning_rate": 5.696108543844282e-06,
+      "loss": 1.107,
+      "step": 6777
+    },
+    {
+      "epoch": 1.3048102606059149,
+      "grad_norm": 3.1934439487412423,
+      "learning_rate": 5.693294404132474e-06,
+      "loss": 1.0768,
+      "step": 6778
+    },
+    {
+      "epoch": 1.30500276728349,
+      "grad_norm": 3.1245360071591257,
+      "learning_rate": 5.690480683087554e-06,
+      "loss": 1.0278,
+      "step": 6779
+    },
+    {
+      "epoch": 1.3051952739610655,
+      "grad_norm": 3.4021738150462273,
+      "learning_rate": 5.687667380983037e-06,
+      "loss": 1.1556,
+      "step": 6780
+    },
+    {
+      "epoch": 1.305387780638641,
+      "grad_norm": 3.337973502226165,
+      "learning_rate": 5.68485449809243e-06,
+      "loss": 1.1215,
+      "step": 6781
+    },
+    {
+      "epoch": 1.3055802873162163,
+      "grad_norm": 3.341225219849192,
+      "learning_rate": 5.682042034689168e-06,
+      "loss": 1.1376,
+      "step": 6782
+    },
+    {
+      "epoch": 1.3057727939937918,
+      "grad_norm": 3.2288798743694858,
+      "learning_rate": 5.679229991046669e-06,
+      "loss": 1.0701,
+      "step": 6783
+    },
+    {
+      "epoch": 1.305965300671367,
+      "grad_norm": 3.201899431606587,
+      "learning_rate": 5.6764183674382834e-06,
+      "loss": 1.0944,
+      "step": 6784
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.9724,
+      "step": 6784,
+      "vm_loss": 0.1806
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.481,
+      "step": 6784,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 1.2175,
+      "step": 6784,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.9441,
+      "step": 6784,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.8773,
+      "step": 6784,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.8689,
+      "step": 6784,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.907,
+      "step": 6784,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 1.305965300671367,
+      "lm_loss": 0.6336,
+      "step": 6784,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 1.3061578073489424,
+      "grad_norm": 3.2952599340064017,
+      "learning_rate": 5.673607164137356e-06,
+      "loss": 1.0991,
+      "step": 6785
+    },
+    {
+      "epoch": 1.3063503140265178,
+      "grad_norm": 3.072607409178018,
+      "learning_rate": 5.67079638141716e-06,
+      "loss": 1.0673,
+      "step": 6786
+    },
+    {
+      "epoch": 1.3065428207040932,
+      "grad_norm": 3.2593801050639857,
+      "learning_rate": 5.6679860195509416e-06,
+      "loss": 1.1151,
+      "step": 6787
+    },
+    {
+      "epoch": 1.3067353273816686,
+      "grad_norm": 3.201958699434688,
+      "learning_rate": 5.665176078811905e-06,
+      "loss": 1.0717,
+      "step": 6788
+    },
+    {
+      "epoch": 1.3069278340592438,
+      "grad_norm": 3.376825724493138,
+      "learning_rate": 5.662366559473212e-06,
+      "loss": 1.0886,
+      "step": 6789
+    },
+    {
+      "epoch": 1.3071203407368193,
+      "grad_norm": 3.1774777512029666,
+      "learning_rate": 5.659557461807987e-06,
+      "loss": 1.08,
+      "step": 6790
+    },
+    {
+      "epoch": 1.3073128474143947,
+      "grad_norm": 3.1173406672633295,
+      "learning_rate": 5.656748786089304e-06,
+      "loss": 1.0255,
+      "step": 6791
+    },
+    {
+      "epoch": 1.30750535409197,
+      "grad_norm": 3.292747526043047,
+      "learning_rate": 5.653940532590205e-06,
+      "loss": 1.1313,
+      "step": 6792
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 1.0739,
+      "step": 6792,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 1.2381,
+      "step": 6792,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 0.8689,
+      "step": 6792,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 0.8472,
+      "step": 6792,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 1.0332,
+      "step": 6792,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 0.9619,
+      "step": 6792,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 0.8545,
+      "step": 6792,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 1.30750535409197,
+      "lm_loss": 0.9727,
+      "step": 6792,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.3076978607695455,
+      "grad_norm": 3.4114704574420416,
+      "learning_rate": 5.6511327015836895e-06,
+      "loss": 1.1314,
+      "step": 6793
+    },
+    {
+      "epoch": 1.3078903674471207,
+      "grad_norm": 3.2169803587369112,
+      "learning_rate": 5.648325293342714e-06,
+      "loss": 1.078,
+      "step": 6794
+    },
+    {
+      "epoch": 1.3080828741246961,
+      "grad_norm": 3.109197791591697,
+      "learning_rate": 5.645518308140192e-06,
+      "loss": 1.0197,
+      "step": 6795
+    },
+    {
+      "epoch": 1.3082753808022716,
+      "grad_norm": 3.2619593952993435,
+      "learning_rate": 5.642711746249006e-06,
+      "loss": 1.1091,
+      "step": 6796
+    },
+    {
+      "epoch": 1.308467887479847,
+      "grad_norm": 3.377584607874024,
+      "learning_rate": 5.639905607941976e-06,
+      "loss": 1.0918,
+      "step": 6797
+    },
+    {
+      "epoch": 1.3086603941574224,
+      "grad_norm": 3.2155422049561175,
+      "learning_rate": 5.637099893491911e-06,
+      "loss": 1.0899,
+      "step": 6798
+    },
+    {
+      "epoch": 1.3088529008349976,
+      "grad_norm": 3.2378673833520755,
+      "learning_rate": 5.634294603171551e-06,
+      "loss": 1.1037,
+      "step": 6799
+    },
+    {
+      "epoch": 1.309045407512573,
+      "grad_norm": 3.2521025421536254,
+      "learning_rate": 5.6314897372536115e-06,
+      "loss": 1.1077,
+      "step": 6800
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 0.759,
+      "step": 6800,
+      "vm_loss": 0.1856
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 0.5033,
+      "step": 6800,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 0.6912,
+      "step": 6800,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 1.1405,
+      "step": 6800,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 0.8219,
+      "step": 6800,
+      "vm_loss": 0.1068
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 1.0372,
+      "step": 6800,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 0.7955,
+      "step": 6800,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.309045407512573,
+      "lm_loss": 1.0129,
+      "step": 6800,
+      "vm_loss": 0.1516
+    },
+    {
+      "epoch": 1.3092379141901485,
+      "grad_norm": 3.151913377908758,
+      "learning_rate": 5.628685296010752e-06,
+      "loss": 1.0695,
+      "step": 6801
+    },
+    {
+      "epoch": 1.3094304208677239,
+      "grad_norm": 3.326215342257217,
+      "learning_rate": 5.625881279715615e-06,
+      "loss": 1.152,
+      "step": 6802
+    },
+    {
+      "epoch": 1.3096229275452993,
+      "grad_norm": 3.152581951826157,
+      "learning_rate": 5.6230776886407745e-06,
+      "loss": 1.0788,
+      "step": 6803
+    },
+    {
+      "epoch": 1.3098154342228745,
+      "grad_norm": 3.2537269140234204,
+      "learning_rate": 5.620274523058781e-06,
+      "loss": 1.0777,
+      "step": 6804
+    },
+    {
+      "epoch": 1.31000794090045,
+      "grad_norm": 3.14831837340861,
+      "learning_rate": 5.617471783242137e-06,
+      "loss": 1.0636,
+      "step": 6805
+    },
+    {
+      "epoch": 1.3102004475780253,
+      "grad_norm": 3.3643419637949425,
+      "learning_rate": 5.614669469463307e-06,
+      "loss": 1.1172,
+      "step": 6806
+    },
+    {
+      "epoch": 1.3103929542556008,
+      "grad_norm": 3.3022522755551864,
+      "learning_rate": 5.611867581994705e-06,
+      "loss": 1.1269,
+      "step": 6807
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "grad_norm": 3.3039057834503116,
+      "learning_rate": 5.609066121108717e-06,
+      "loss": 1.1253,
+      "step": 6808
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.9438,
+      "step": 6808,
+      "vm_loss": 0.2584
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.6468,
+      "step": 6808,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.723,
+      "step": 6808,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.6605,
+      "step": 6808,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.7814,
+      "step": 6808,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 1.0829,
+      "step": 6808,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 1.4473,
+      "step": 6808,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 1.3105854609331762,
+      "lm_loss": 0.6278,
+      "step": 6808,
+      "vm_loss": 0.2379
+    },
+    {
+      "epoch": 1.3107779676107514,
+      "grad_norm": 3.1239119720757897,
+      "learning_rate": 5.606265087077677e-06,
+      "loss": 1.0447,
+      "step": 6809
+    },
+    {
+      "epoch": 1.3109704742883268,
+      "grad_norm": 3.286694789022265,
+      "learning_rate": 5.603464480173884e-06,
+      "loss": 1.1099,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3111629809659022,
+      "grad_norm": 3.2785136526197416,
+      "learning_rate": 5.600664300669589e-06,
+      "loss": 1.0719,
+      "step": 6811
+    },
+    {
+      "epoch": 1.3113554876434776,
+      "grad_norm": 3.3207451556789422,
+      "learning_rate": 5.597864548837011e-06,
+      "loss": 1.0529,
+      "step": 6812
+    },
+    {
+      "epoch": 1.311547994321053,
+      "grad_norm": 3.270711211480238,
+      "learning_rate": 5.59506522494832e-06,
+      "loss": 1.129,
+      "step": 6813
+    },
+    {
+      "epoch": 1.3117405009986283,
+      "grad_norm": 3.354711886269982,
+      "learning_rate": 5.592266329275641e-06,
+      "loss": 1.0879,
+      "step": 6814
+    },
+    {
+      "epoch": 1.3119330076762037,
+      "grad_norm": 3.1895961410309415,
+      "learning_rate": 5.589467862091069e-06,
+      "loss": 1.0493,
+      "step": 6815
+    },
+    {
+      "epoch": 1.312125514353779,
+      "grad_norm": 3.2974611482813616,
+      "learning_rate": 5.586669823666646e-06,
+      "loss": 1.0905,
+      "step": 6816
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 0.4995,
+      "step": 6816,
+      "vm_loss": 0.2086
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 0.6886,
+      "step": 6816,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 1.0722,
+      "step": 6816,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 0.5194,
+      "step": 6816,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 0.748,
+      "step": 6816,
+      "vm_loss": 0.2343
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 0.8596,
+      "step": 6816,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 1.0437,
+      "step": 6816,
+      "vm_loss": 0.2174
+    },
+    {
+      "epoch": 1.312125514353779,
+      "lm_loss": 1.2521,
+      "step": 6816,
+      "vm_loss": 0.226
+    },
+    {
+      "epoch": 1.3123180210313545,
+      "grad_norm": 3.338580201701724,
+      "learning_rate": 5.583872214274383e-06,
+      "loss": 1.0967,
+      "step": 6817
+    },
+    {
+      "epoch": 1.31251052770893,
+      "grad_norm": 3.385269464304009,
+      "learning_rate": 5.581075034186232e-06,
+      "loss": 1.1223,
+      "step": 6818
+    },
+    {
+      "epoch": 1.3127030343865052,
+      "grad_norm": 3.3341249863889426,
+      "learning_rate": 5.57827828367413e-06,
+      "loss": 1.1528,
+      "step": 6819
+    },
+    {
+      "epoch": 1.3128955410640808,
+      "grad_norm": 3.3671197998520226,
+      "learning_rate": 5.575481963009945e-06,
+      "loss": 1.1225,
+      "step": 6820
+    },
+    {
+      "epoch": 1.313088047741656,
+      "grad_norm": 3.246298777987742,
+      "learning_rate": 5.572686072465519e-06,
+      "loss": 1.091,
+      "step": 6821
+    },
+    {
+      "epoch": 1.3132805544192314,
+      "grad_norm": 3.1937464693038455,
+      "learning_rate": 5.569890612312649e-06,
+      "loss": 1.0635,
+      "step": 6822
+    },
+    {
+      "epoch": 1.3134730610968068,
+      "grad_norm": 3.180069204686733,
+      "learning_rate": 5.567095582823094e-06,
+      "loss": 1.0844,
+      "step": 6823
+    },
+    {
+      "epoch": 1.313665567774382,
+      "grad_norm": 3.1679948678829835,
+      "learning_rate": 5.564300984268556e-06,
+      "loss": 1.0516,
+      "step": 6824
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.3728,
+      "step": 6824,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 1.3732,
+      "step": 6824,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.9333,
+      "step": 6824,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 1.1613,
+      "step": 6824,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.9209,
+      "step": 6824,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.8465,
+      "step": 6824,
+      "vm_loss": 0.234
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.4611,
+      "step": 6824,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 1.313665567774382,
+      "lm_loss": 0.887,
+      "step": 6824,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 1.3138580744519577,
+      "grad_norm": 3.1148225817967448,
+      "learning_rate": 5.561506816920713e-06,
+      "loss": 1.0379,
+      "step": 6825
+    },
+    {
+      "epoch": 1.3140505811295329,
+      "grad_norm": 3.3791816638663854,
+      "learning_rate": 5.558713081051192e-06,
+      "loss": 1.1062,
+      "step": 6826
+    },
+    {
+      "epoch": 1.3142430878071083,
+      "grad_norm": 3.309018968668874,
+      "learning_rate": 5.555919776931584e-06,
+      "loss": 1.0984,
+      "step": 6827
+    },
+    {
+      "epoch": 1.3144355944846837,
+      "grad_norm": 3.1616429580855407,
+      "learning_rate": 5.5531269048334216e-06,
+      "loss": 1.0309,
+      "step": 6828
+    },
+    {
+      "epoch": 1.3146281011622591,
+      "grad_norm": 3.271338691082631,
+      "learning_rate": 5.550334465028224e-06,
+      "loss": 1.0709,
+      "step": 6829
+    },
+    {
+      "epoch": 1.3148206078398346,
+      "grad_norm": 3.536727848506085,
+      "learning_rate": 5.547542457787442e-06,
+      "loss": 1.1801,
+      "step": 6830
+    },
+    {
+      "epoch": 1.3150131145174098,
+      "grad_norm": 3.1400289464688256,
+      "learning_rate": 5.544750883382495e-06,
+      "loss": 1.0285,
+      "step": 6831
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "grad_norm": 3.2304741462335214,
+      "learning_rate": 5.541959742084761e-06,
+      "loss": 1.0434,
+      "step": 6832
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 0.7938,
+      "step": 6832,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 0.9232,
+      "step": 6832,
+      "vm_loss": 0.1196
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 1.1548,
+      "step": 6832,
+      "vm_loss": 0.2505
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 0.9291,
+      "step": 6832,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 0.894,
+      "step": 6832,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 1.0663,
+      "step": 6832,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 1.131,
+      "step": 6832,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.3152056211949852,
+      "lm_loss": 1.0495,
+      "step": 6832,
+      "vm_loss": 0.13
+    },
+    {
+      "epoch": 1.3153981278725606,
+      "grad_norm": 3.3888403195494616,
+      "learning_rate": 5.539169034165577e-06,
+      "loss": 1.1579,
+      "step": 6833
+    },
+    {
+      "epoch": 1.315590634550136,
+      "grad_norm": 3.221327772775588,
+      "learning_rate": 5.536378759896237e-06,
+      "loss": 1.0684,
+      "step": 6834
+    },
+    {
+      "epoch": 1.3157831412277115,
+      "grad_norm": 3.303925706919441,
+      "learning_rate": 5.533588919547979e-06,
+      "loss": 1.1403,
+      "step": 6835
+    },
+    {
+      "epoch": 1.3159756479052866,
+      "grad_norm": 3.21299284286161,
+      "learning_rate": 5.530799513392029e-06,
+      "loss": 1.0419,
+      "step": 6836
+    },
+    {
+      "epoch": 1.316168154582862,
+      "grad_norm": 3.255937788119579,
+      "learning_rate": 5.52801054169954e-06,
+      "loss": 1.1186,
+      "step": 6837
+    },
+    {
+      "epoch": 1.3163606612604375,
+      "grad_norm": 3.3594248441002548,
+      "learning_rate": 5.525222004741639e-06,
+      "loss": 1.1198,
+      "step": 6838
+    },
+    {
+      "epoch": 1.316553167938013,
+      "grad_norm": 3.46863621840443,
+      "learning_rate": 5.522433902789408e-06,
+      "loss": 1.1787,
+      "step": 6839
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "grad_norm": 3.2245547220750344,
+      "learning_rate": 5.519646236113891e-06,
+      "loss": 1.1158,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 1.1142,
+      "step": 6840,
+      "vm_loss": 0.1907
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 1.0466,
+      "step": 6840,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 0.9665,
+      "step": 6840,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 0.7343,
+      "step": 6840,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 1.0654,
+      "step": 6840,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 0.8259,
+      "step": 6840,
+      "vm_loss": 0.1038
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 0.8645,
+      "step": 6840,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.3167456746155883,
+      "lm_loss": 1.0416,
+      "step": 6840,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.3169381812931635,
+      "grad_norm": 3.2847984864002413,
+      "learning_rate": 5.516859004986077e-06,
+      "loss": 1.1061,
+      "step": 6841
+    },
+    {
+      "epoch": 1.317130687970739,
+      "grad_norm": 3.3258484033524542,
+      "learning_rate": 5.514072209676924e-06,
+      "loss": 1.1381,
+      "step": 6842
+    },
+    {
+      "epoch": 1.3173231946483144,
+      "grad_norm": 3.2890677771544548,
+      "learning_rate": 5.511285850457343e-06,
+      "loss": 1.1269,
+      "step": 6843
+    },
+    {
+      "epoch": 1.3175157013258898,
+      "grad_norm": 3.1683016735834064,
+      "learning_rate": 5.508499927598209e-06,
+      "loss": 1.0501,
+      "step": 6844
+    },
+    {
+      "epoch": 1.3177082080034652,
+      "grad_norm": 3.2720195812542077,
+      "learning_rate": 5.505714441370338e-06,
+      "loss": 1.1158,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3179007146810404,
+      "grad_norm": 3.224613017112921,
+      "learning_rate": 5.502929392044528e-06,
+      "loss": 1.0876,
+      "step": 6846
+    },
+    {
+      "epoch": 1.3180932213586158,
+      "grad_norm": 3.2866542999664805,
+      "learning_rate": 5.500144779891513e-06,
+      "loss": 1.1203,
+      "step": 6847
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "grad_norm": 3.298827306964532,
+      "learning_rate": 5.497360605181998e-06,
+      "loss": 1.1171,
+      "step": 6848
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 0.9218,
+      "step": 6848,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 0.6899,
+      "step": 6848,
+      "vm_loss": 0.1137
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 1.1841,
+      "step": 6848,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 0.9574,
+      "step": 6848,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 0.8867,
+      "step": 6848,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 0.6012,
+      "step": 6848,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 1.0803,
+      "step": 6848,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 1.3182857280361913,
+      "lm_loss": 1.0218,
+      "step": 6848,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.3184782347137667,
+      "grad_norm": 3.392204290782093,
+      "learning_rate": 5.494576868186632e-06,
+      "loss": 1.0956,
+      "step": 6849
+    },
+    {
+      "epoch": 1.318670741391342,
+      "grad_norm": 3.2547383256188485,
+      "learning_rate": 5.491793569176042e-06,
+      "loss": 1.0644,
+      "step": 6850
+    },
+    {
+      "epoch": 1.3188632480689173,
+      "grad_norm": 3.2597836207170676,
+      "learning_rate": 5.48901070842079e-06,
+      "loss": 1.1214,
+      "step": 6851
+    },
+    {
+      "epoch": 1.3190557547464927,
+      "grad_norm": 3.1863533305785485,
+      "learning_rate": 5.4862282861914095e-06,
+      "loss": 1.0854,
+      "step": 6852
+    },
+    {
+      "epoch": 1.3192482614240681,
+      "grad_norm": 3.1722641521112207,
+      "learning_rate": 5.483446302758389e-06,
+      "loss": 1.0786,
+      "step": 6853
+    },
+    {
+      "epoch": 1.3194407681016436,
+      "grad_norm": 3.174585915549796,
+      "learning_rate": 5.480664758392169e-06,
+      "loss": 1.0427,
+      "step": 6854
+    },
+    {
+      "epoch": 1.319633274779219,
+      "grad_norm": 3.390440023339875,
+      "learning_rate": 5.477883653363155e-06,
+      "loss": 1.1294,
+      "step": 6855
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "grad_norm": 3.098003098638949,
+      "learning_rate": 5.475102987941705e-06,
+      "loss": 1.0458,
+      "step": 6856
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.5035,
+      "step": 6856,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 1.0625,
+      "step": 6856,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.8518,
+      "step": 6856,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.935,
+      "step": 6856,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.9772,
+      "step": 6856,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.6268,
+      "step": 6856,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 0.851,
+      "step": 6856,
+      "vm_loss": 0.2018
+    },
+    {
+      "epoch": 1.3198257814567942,
+      "lm_loss": 1.0988,
+      "step": 6856,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.3200182881343696,
+      "grad_norm": 3.255400202211938,
+      "learning_rate": 5.472322762398139e-06,
+      "loss": 1.0887,
+      "step": 6857
+    },
+    {
+      "epoch": 1.320210794811945,
+      "grad_norm": 3.411683348119307,
+      "learning_rate": 5.469542977002722e-06,
+      "loss": 1.1152,
+      "step": 6858
+    },
+    {
+      "epoch": 1.3204033014895205,
+      "grad_norm": 3.474602284948717,
+      "learning_rate": 5.466763632025689e-06,
+      "loss": 1.1206,
+      "step": 6859
+    },
+    {
+      "epoch": 1.3205958081670959,
+      "grad_norm": 3.5411638429401653,
+      "learning_rate": 5.463984727737227e-06,
+      "loss": 1.141,
+      "step": 6860
+    },
+    {
+      "epoch": 1.320788314844671,
+      "grad_norm": 3.537454154531341,
+      "learning_rate": 5.461206264407487e-06,
+      "loss": 1.173,
+      "step": 6861
+    },
+    {
+      "epoch": 1.3209808215222465,
+      "grad_norm": 3.3068033135474515,
+      "learning_rate": 5.45842824230656e-06,
+      "loss": 1.0896,
+      "step": 6862
+    },
+    {
+      "epoch": 1.321173328199822,
+      "grad_norm": 3.3003051997031054,
+      "learning_rate": 5.455650661704517e-06,
+      "loss": 1.1119,
+      "step": 6863
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "grad_norm": 3.278279944408037,
+      "learning_rate": 5.452873522871365e-06,
+      "loss": 1.1406,
+      "step": 6864
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.8291,
+      "step": 6864,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.7603,
+      "step": 6864,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.7874,
+      "step": 6864,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.7299,
+      "step": 6864,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.8268,
+      "step": 6864,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.7649,
+      "step": 6864,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.3361,
+      "step": 6864,
+      "vm_loss": 0.1226
+    },
+    {
+      "epoch": 1.3213658348773973,
+      "lm_loss": 0.5882,
+      "step": 6864,
+      "vm_loss": 0.2669
+    },
+    {
+      "epoch": 1.3215583415549728,
+      "grad_norm": 3.225164583265845,
+      "learning_rate": 5.450096826077086e-06,
+      "loss": 1.0848,
+      "step": 6865
+    },
+    {
+      "epoch": 1.321750848232548,
+      "grad_norm": 3.20575269499846,
+      "learning_rate": 5.447320571591598e-06,
+      "loss": 1.0497,
+      "step": 6866
+    },
+    {
+      "epoch": 1.3219433549101234,
+      "grad_norm": 3.1294913804345517,
+      "learning_rate": 5.4445447596848045e-06,
+      "loss": 1.1334,
+      "step": 6867
+    },
+    {
+      "epoch": 1.3221358615876988,
+      "grad_norm": 3.136796129316707,
+      "learning_rate": 5.441769390626537e-06,
+      "loss": 1.0217,
+      "step": 6868
+    },
+    {
+      "epoch": 1.3223283682652742,
+      "grad_norm": 3.3011942872919584,
+      "learning_rate": 5.438994464686601e-06,
+      "loss": 1.1015,
+      "step": 6869
+    },
+    {
+      "epoch": 1.3225208749428496,
+      "grad_norm": 3.0577145060756594,
+      "learning_rate": 5.436219982134756e-06,
+      "loss": 1.0318,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3227133816204248,
+      "grad_norm": 3.219533394784254,
+      "learning_rate": 5.433445943240721e-06,
+      "loss": 1.0691,
+      "step": 6871
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "grad_norm": 3.4439773962233744,
+      "learning_rate": 5.430672348274155e-06,
+      "loss": 1.1165,
+      "step": 6872
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 0.5442,
+      "step": 6872,
+      "vm_loss": 0.1242
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 0.8073,
+      "step": 6872,
+      "vm_loss": 0.2056
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 1.1496,
+      "step": 6872,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 1.0221,
+      "step": 6872,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 1.2131,
+      "step": 6872,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 0.718,
+      "step": 6872,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 0.6556,
+      "step": 6872,
+      "vm_loss": 0.1126
+    },
+    {
+      "epoch": 1.3229058882980003,
+      "lm_loss": 1.3355,
+      "step": 6872,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.3230983949755757,
+      "grad_norm": 3.347689661732486,
+      "learning_rate": 5.427899197504703e-06,
+      "loss": 1.0835,
+      "step": 6873
+    },
+    {
+      "epoch": 1.323290901653151,
+      "grad_norm": 3.434818005625957,
+      "learning_rate": 5.42512649120194e-06,
+      "loss": 1.088,
+      "step": 6874
+    },
+    {
+      "epoch": 1.3234834083307265,
+      "grad_norm": 3.3283256339375407,
+      "learning_rate": 5.422354229635412e-06,
+      "loss": 1.0945,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3236759150083017,
+      "grad_norm": 3.1716535917928477,
+      "learning_rate": 5.419582413074618e-06,
+      "loss": 1.0332,
+      "step": 6876
+    },
+    {
+      "epoch": 1.3238684216858772,
+      "grad_norm": 3.2547821717140892,
+      "learning_rate": 5.416811041789014e-06,
+      "loss": 1.0515,
+      "step": 6877
+    },
+    {
+      "epoch": 1.3240609283634526,
+      "grad_norm": 3.101051179705844,
+      "learning_rate": 5.414040116048016e-06,
+      "loss": 1.044,
+      "step": 6878
+    },
+    {
+      "epoch": 1.324253435041028,
+      "grad_norm": 3.2552167748632503,
+      "learning_rate": 5.411269636120984e-06,
+      "loss": 1.1026,
+      "step": 6879
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "grad_norm": 3.197140893983596,
+      "learning_rate": 5.408499602277256e-06,
+      "loss": 1.0838,
+      "step": 6880
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.4903,
+      "step": 6880,
+      "vm_loss": 0.1913
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 1.0475,
+      "step": 6880,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.8657,
+      "step": 6880,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.6593,
+      "step": 6880,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.8005,
+      "step": 6880,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.744,
+      "step": 6880,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 1.0617,
+      "step": 6880,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 1.3244459417186034,
+      "lm_loss": 0.884,
+      "step": 6880,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 1.3246384483961786,
+      "grad_norm": 3.196365095129799,
+      "learning_rate": 5.405730014786107e-06,
+      "loss": 1.0868,
+      "step": 6881
+    },
+    {
+      "epoch": 1.3248309550737543,
+      "grad_norm": 3.2250465796722105,
+      "learning_rate": 5.4029608739167815e-06,
+      "loss": 1.091,
+      "step": 6882
+    },
+    {
+      "epoch": 1.3250234617513295,
+      "grad_norm": 3.4015932440409182,
+      "learning_rate": 5.400192179938465e-06,
+      "loss": 1.0743,
+      "step": 6883
+    },
+    {
+      "epoch": 1.3252159684289049,
+      "grad_norm": 3.25180967039191,
+      "learning_rate": 5.397423933120325e-06,
+      "loss": 1.0503,
+      "step": 6884
+    },
+    {
+      "epoch": 1.3254084751064803,
+      "grad_norm": 3.172798232766363,
+      "learning_rate": 5.394656133731458e-06,
+      "loss": 1.084,
+      "step": 6885
+    },
+    {
+      "epoch": 1.3256009817840555,
+      "grad_norm": 3.310818218043609,
+      "learning_rate": 5.3918887820409364e-06,
+      "loss": 1.0884,
+      "step": 6886
+    },
+    {
+      "epoch": 1.3257934884616311,
+      "grad_norm": 3.2509000113295103,
+      "learning_rate": 5.389121878317778e-06,
+      "loss": 1.0864,
+      "step": 6887
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "grad_norm": 3.272075873522302,
+      "learning_rate": 5.386355422830969e-06,
+      "loss": 1.0847,
+      "step": 6888
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 0.3604,
+      "step": 6888,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 1.1042,
+      "step": 6888,
+      "vm_loss": 0.1072
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 1.2242,
+      "step": 6888,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 1.3887,
+      "step": 6888,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 0.6736,
+      "step": 6888,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 1.2054,
+      "step": 6888,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 0.6526,
+      "step": 6888,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 1.3259859951392063,
+      "lm_loss": 0.6192,
+      "step": 6888,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.3261785018167818,
+      "grad_norm": 3.2628570362596157,
+      "learning_rate": 5.383589415849434e-06,
+      "loss": 1.1237,
+      "step": 6889
+    },
+    {
+      "epoch": 1.3263710084943572,
+      "grad_norm": 3.2731584881707145,
+      "learning_rate": 5.3808238576420694e-06,
+      "loss": 1.0837,
+      "step": 6890
+    },
+    {
+      "epoch": 1.3265635151719326,
+      "grad_norm": 3.2213939239969513,
+      "learning_rate": 5.378058748477722e-06,
+      "loss": 1.0586,
+      "step": 6891
+    },
+    {
+      "epoch": 1.326756021849508,
+      "grad_norm": 3.3295600342354437,
+      "learning_rate": 5.375294088625196e-06,
+      "loss": 1.0885,
+      "step": 6892
+    },
+    {
+      "epoch": 1.3269485285270832,
+      "grad_norm": 3.2371652707263263,
+      "learning_rate": 5.372529878353252e-06,
+      "loss": 1.0947,
+      "step": 6893
+    },
+    {
+      "epoch": 1.3271410352046586,
+      "grad_norm": 3.2603293027875413,
+      "learning_rate": 5.36976611793061e-06,
+      "loss": 1.1055,
+      "step": 6894
+    },
+    {
+      "epoch": 1.327333541882234,
+      "grad_norm": 3.2627971521945747,
+      "learning_rate": 5.3670028076259355e-06,
+      "loss": 1.113,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "grad_norm": 3.316257141872321,
+      "learning_rate": 5.364239947707861e-06,
+      "loss": 1.0874,
+      "step": 6896
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 0.6665,
+      "step": 6896,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 1.4276,
+      "step": 6896,
+      "vm_loss": 0.2144
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 0.9414,
+      "step": 6896,
+      "vm_loss": 0.2129
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 1.4092,
+      "step": 6896,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 0.9723,
+      "step": 6896,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 0.8288,
+      "step": 6896,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 0.4871,
+      "step": 6896,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 1.3275260485598095,
+      "lm_loss": 1.109,
+      "step": 6896,
+      "vm_loss": 0.2111
+    },
+    {
+      "epoch": 1.327718555237385,
+      "grad_norm": 3.370314938121935,
+      "learning_rate": 5.361477538444973e-06,
+      "loss": 1.1607,
+      "step": 6897
+    },
+    {
+      "epoch": 1.3279110619149601,
+      "grad_norm": 3.2472941820543486,
+      "learning_rate": 5.358715580105813e-06,
+      "loss": 1.0772,
+      "step": 6898
+    },
+    {
+      "epoch": 1.3281035685925355,
+      "grad_norm": 3.1629097179210035,
+      "learning_rate": 5.35595407295888e-06,
+      "loss": 1.023,
+      "step": 6899
+    },
+    {
+      "epoch": 1.328296075270111,
+      "grad_norm": 3.0412218058228198,
+      "learning_rate": 5.3531930172726195e-06,
+      "loss": 0.9986,
+      "step": 6900
+    },
+    {
+      "epoch": 1.3284885819476864,
+      "grad_norm": 3.2993112553088766,
+      "learning_rate": 5.350432413315455e-06,
+      "loss": 1.088,
+      "step": 6901
+    },
+    {
+      "epoch": 1.3286810886252618,
+      "grad_norm": 3.3455929183723128,
+      "learning_rate": 5.347672261355742e-06,
+      "loss": 1.1175,
+      "step": 6902
+    },
+    {
+      "epoch": 1.328873595302837,
+      "grad_norm": 3.1966685556094387,
+      "learning_rate": 5.344912561661806e-06,
+      "loss": 1.026,
+      "step": 6903
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "grad_norm": 3.2649545462920186,
+      "learning_rate": 5.342153314501923e-06,
+      "loss": 1.0719,
+      "step": 6904
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 1.3055,
+      "step": 6904,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.8856,
+      "step": 6904,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.94,
+      "step": 6904,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 1.1542,
+      "step": 6904,
+      "vm_loss": 0.2463
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.9263,
+      "step": 6904,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.4832,
+      "step": 6904,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.4448,
+      "step": 6904,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 1.3290661019804124,
+      "lm_loss": 0.7291,
+      "step": 6904,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 1.3292586086579878,
+      "grad_norm": 3.2793473622909697,
+      "learning_rate": 5.339394520144334e-06,
+      "loss": 1.0711,
+      "step": 6905
+    },
+    {
+      "epoch": 1.3294511153355633,
+      "grad_norm": 3.3569902689525515,
+      "learning_rate": 5.3366361788572215e-06,
+      "loss": 1.1332,
+      "step": 6906
+    },
+    {
+      "epoch": 1.3296436220131387,
+      "grad_norm": 3.244764772000376,
+      "learning_rate": 5.333878290908734e-06,
+      "loss": 1.0646,
+      "step": 6907
+    },
+    {
+      "epoch": 1.3298361286907139,
+      "grad_norm": 3.113239379918662,
+      "learning_rate": 5.331120856566972e-06,
+      "loss": 1.0743,
+      "step": 6908
+    },
+    {
+      "epoch": 1.3300286353682893,
+      "grad_norm": 3.3118697135160167,
+      "learning_rate": 5.3283638761e-06,
+      "loss": 1.1498,
+      "step": 6909
+    },
+    {
+      "epoch": 1.3302211420458647,
+      "grad_norm": 3.191066365421342,
+      "learning_rate": 5.325607349775818e-06,
+      "loss": 1.0525,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3304136487234401,
+      "grad_norm": 3.2738481099074868,
+      "learning_rate": 5.3228512778624156e-06,
+      "loss": 1.1252,
+      "step": 6911
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "grad_norm": 3.1918435934599034,
+      "learning_rate": 5.3200956606277006e-06,
+      "loss": 1.0633,
+      "step": 6912
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 1.5916,
+      "step": 6912,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 0.9236,
+      "step": 6912,
+      "vm_loss": 0.2144
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 0.8234,
+      "step": 6912,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 0.969,
+      "step": 6912,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 1.1552,
+      "step": 6912,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 0.9247,
+      "step": 6912,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 1.4569,
+      "step": 6912,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.3306061554010156,
+      "lm_loss": 0.9717,
+      "step": 6912,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 1.3307986620785908,
+      "grad_norm": 3.2889234517132193,
+      "learning_rate": 5.317340498339561e-06,
+      "loss": 1.1296,
+      "step": 6913
+    },
+    {
+      "epoch": 1.3309911687561662,
+      "grad_norm": 3.335211868132565,
+      "learning_rate": 5.314585791265835e-06,
+      "loss": 1.1246,
+      "step": 6914
+    },
+    {
+      "epoch": 1.3311836754337416,
+      "grad_norm": 3.2264391913559094,
+      "learning_rate": 5.311831539674317e-06,
+      "loss": 1.0785,
+      "step": 6915
+    },
+    {
+      "epoch": 1.331376182111317,
+      "grad_norm": 3.2205110850075434,
+      "learning_rate": 5.309077743832749e-06,
+      "loss": 1.065,
+      "step": 6916
+    },
+    {
+      "epoch": 1.3315686887888925,
+      "grad_norm": 3.154567295654948,
+      "learning_rate": 5.306324404008838e-06,
+      "loss": 1.0318,
+      "step": 6917
+    },
+    {
+      "epoch": 1.3317611954664677,
+      "grad_norm": 3.2282866306913927,
+      "learning_rate": 5.303571520470245e-06,
+      "loss": 1.0897,
+      "step": 6918
+    },
+    {
+      "epoch": 1.331953702144043,
+      "grad_norm": 3.2235587594358948,
+      "learning_rate": 5.300819093484586e-06,
+      "loss": 1.0563,
+      "step": 6919
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "grad_norm": 3.343270087207791,
+      "learning_rate": 5.29806712331943e-06,
+      "loss": 1.1389,
+      "step": 6920
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 1.3698,
+      "step": 6920,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.9847,
+      "step": 6920,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.7467,
+      "step": 6920,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.5713,
+      "step": 6920,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.7602,
+      "step": 6920,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.9817,
+      "step": 6920,
+      "vm_loss": 0.1407
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.9181,
+      "step": 6920,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.3321462088216185,
+      "lm_loss": 0.8862,
+      "step": 6920,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.332338715499194,
+      "grad_norm": 3.2543245430100436,
+      "learning_rate": 5.295315610242305e-06,
+      "loss": 1.0613,
+      "step": 6921
+    },
+    {
+      "epoch": 1.3325312221767693,
+      "grad_norm": 3.1305727186729246,
+      "learning_rate": 5.292564554520698e-06,
+      "loss": 1.0366,
+      "step": 6922
+    },
+    {
+      "epoch": 1.3327237288543445,
+      "grad_norm": 3.320716713752849,
+      "learning_rate": 5.289813956422037e-06,
+      "loss": 1.0985,
+      "step": 6923
+    },
+    {
+      "epoch": 1.33291623553192,
+      "grad_norm": 3.3004947850582784,
+      "learning_rate": 5.28706381621372e-06,
+      "loss": 1.0944,
+      "step": 6924
+    },
+    {
+      "epoch": 1.3331087422094954,
+      "grad_norm": 3.2851821881022585,
+      "learning_rate": 5.284314134163098e-06,
+      "loss": 1.0328,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3333012488870708,
+      "grad_norm": 3.4455814871621526,
+      "learning_rate": 5.281564910537476e-06,
+      "loss": 1.091,
+      "step": 6926
+    },
+    {
+      "epoch": 1.3334937555646462,
+      "grad_norm": 3.383897664710934,
+      "learning_rate": 5.278816145604103e-06,
+      "loss": 1.0939,
+      "step": 6927
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "grad_norm": 3.2224026961725754,
+      "learning_rate": 5.27606783963021e-06,
+      "loss": 1.0291,
+      "step": 6928
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 0.9031,
+      "step": 6928,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 1.2411,
+      "step": 6928,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 1.0549,
+      "step": 6928,
+      "vm_loss": 0.1059
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 0.5974,
+      "step": 6928,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 0.8604,
+      "step": 6928,
+      "vm_loss": 0.1138
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 1.2005,
+      "step": 6928,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 1.5703,
+      "step": 6928,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.3336862622422214,
+      "lm_loss": 0.7812,
+      "step": 6928,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.3338787689197968,
+      "grad_norm": 3.199032374006238,
+      "learning_rate": 5.273319992882957e-06,
+      "loss": 1.0846,
+      "step": 6929
+    },
+    {
+      "epoch": 1.3340712755973723,
+      "grad_norm": 3.1363885753599736,
+      "learning_rate": 5.270572605629471e-06,
+      "loss": 1.0353,
+      "step": 6930
+    },
+    {
+      "epoch": 1.3342637822749477,
+      "grad_norm": 3.2030617130649355,
+      "learning_rate": 5.2678256781368364e-06,
+      "loss": 1.1086,
+      "step": 6931
+    },
+    {
+      "epoch": 1.334456288952523,
+      "grad_norm": 3.214798053725716,
+      "learning_rate": 5.2650792106720905e-06,
+      "loss": 1.0768,
+      "step": 6932
+    },
+    {
+      "epoch": 1.3346487956300983,
+      "grad_norm": 3.251462340548556,
+      "learning_rate": 5.26233320350222e-06,
+      "loss": 1.0658,
+      "step": 6933
+    },
+    {
+      "epoch": 1.3348413023076737,
+      "grad_norm": 3.3068442657570674,
+      "learning_rate": 5.259587656894174e-06,
+      "loss": 1.1081,
+      "step": 6934
+    },
+    {
+      "epoch": 1.3350338089852491,
+      "grad_norm": 3.3069805607946012,
+      "learning_rate": 5.256842571114855e-06,
+      "loss": 1.1131,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "grad_norm": 3.255382601246712,
+      "learning_rate": 5.254097946431121e-06,
+      "loss": 1.0815,
+      "step": 6936
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 1.2063,
+      "step": 6936,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.5287,
+      "step": 6936,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.9248,
+      "step": 6936,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.4554,
+      "step": 6936,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.6011,
+      "step": 6936,
+      "vm_loss": 0.1055
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.9562,
+      "step": 6936,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 0.7735,
+      "step": 6936,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 1.3352263156628246,
+      "lm_loss": 1.3065,
+      "step": 6936,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.3354188223404,
+      "grad_norm": 3.2176035999449604,
+      "learning_rate": 5.251353783109785e-06,
+      "loss": 1.0451,
+      "step": 6937
+    },
+    {
+      "epoch": 1.3356113290179752,
+      "grad_norm": 3.355347432293387,
+      "learning_rate": 5.248610081417617e-06,
+      "loss": 1.1227,
+      "step": 6938
+    },
+    {
+      "epoch": 1.3358038356955506,
+      "grad_norm": 3.2924429880155244,
+      "learning_rate": 5.245866841621335e-06,
+      "loss": 1.064,
+      "step": 6939
+    },
+    {
+      "epoch": 1.335996342373126,
+      "grad_norm": 3.3502522848516865,
+      "learning_rate": 5.243124063987618e-06,
+      "loss": 1.1232,
+      "step": 6940
+    },
+    {
+      "epoch": 1.3361888490507015,
+      "grad_norm": 3.2857219256752,
+      "learning_rate": 5.2403817487831e-06,
+      "loss": 1.0804,
+      "step": 6941
+    },
+    {
+      "epoch": 1.3363813557282769,
+      "grad_norm": 3.2776621987920365,
+      "learning_rate": 5.237639896274369e-06,
+      "loss": 1.0797,
+      "step": 6942
+    },
+    {
+      "epoch": 1.336573862405852,
+      "grad_norm": 3.3159521584187477,
+      "learning_rate": 5.234898506727973e-06,
+      "loss": 1.0998,
+      "step": 6943
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "grad_norm": 3.323041177276933,
+      "learning_rate": 5.232157580410397e-06,
+      "loss": 1.1238,
+      "step": 6944
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 1.4002,
+      "step": 6944,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 0.7547,
+      "step": 6944,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 1.0656,
+      "step": 6944,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 0.6765,
+      "step": 6944,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 0.6627,
+      "step": 6944,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 1.1616,
+      "step": 6944,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 0.5965,
+      "step": 6944,
+      "vm_loss": 0.1202
+    },
+    {
+      "epoch": 1.3367663690834277,
+      "lm_loss": 0.8376,
+      "step": 6944,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.336958875761003,
+      "grad_norm": 3.3570869409403774,
+      "learning_rate": 5.229417117588112e-06,
+      "loss": 1.0671,
+      "step": 6945
+    },
+    {
+      "epoch": 1.3371513824385783,
+      "grad_norm": 3.0934497499081917,
+      "learning_rate": 5.226677118527512e-06,
+      "loss": 1.0377,
+      "step": 6946
+    },
+    {
+      "epoch": 1.3373438891161538,
+      "grad_norm": 3.328485370491098,
+      "learning_rate": 5.223937583494964e-06,
+      "loss": 1.088,
+      "step": 6947
+    },
+    {
+      "epoch": 1.337536395793729,
+      "grad_norm": 3.4006204192044094,
+      "learning_rate": 5.221198512756789e-06,
+      "loss": 1.0992,
+      "step": 6948
+    },
+    {
+      "epoch": 1.3377289024713046,
+      "grad_norm": 3.2466918053787306,
+      "learning_rate": 5.218459906579259e-06,
+      "loss": 0.9971,
+      "step": 6949
+    },
+    {
+      "epoch": 1.3379214091488798,
+      "grad_norm": 3.3878006822575157,
+      "learning_rate": 5.215721765228596e-06,
+      "loss": 1.0872,
+      "step": 6950
+    },
+    {
+      "epoch": 1.3381139158264552,
+      "grad_norm": 3.4611325712261496,
+      "learning_rate": 5.212984088970987e-06,
+      "loss": 1.1445,
+      "step": 6951
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "grad_norm": 3.2497383088604823,
+      "learning_rate": 5.2102468780725675e-06,
+      "loss": 1.0885,
+      "step": 6952
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.619,
+      "step": 6952,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.9264,
+      "step": 6952,
+      "vm_loss": 0.1174
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.9728,
+      "step": 6952,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.8471,
+      "step": 6952,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.8807,
+      "step": 6952,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.9985,
+      "step": 6952,
+      "vm_loss": 0.1013
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.5578,
+      "step": 6952,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 1.3383064225040306,
+      "lm_loss": 0.9214,
+      "step": 6952,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 1.338498929181606,
+      "grad_norm": 3.129673353084442,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 1.0404,
+      "step": 6953
+    },
+    {
+      "epoch": 1.3386914358591815,
+      "grad_norm": 3.277355557917731,
+      "learning_rate": 5.204773853417624e-06,
+      "loss": 1.1306,
+      "step": 6954
+    },
+    {
+      "epoch": 1.3388839425367567,
+      "grad_norm": 3.0838139829589615,
+      "learning_rate": 5.20203804019315e-06,
+      "loss": 0.9889,
+      "step": 6955
+    },
+    {
+      "epoch": 1.339076449214332,
+      "grad_norm": 3.1098672808817702,
+      "learning_rate": 5.199302693391958e-06,
+      "loss": 0.9934,
+      "step": 6956
+    },
+    {
+      "epoch": 1.3392689558919075,
+      "grad_norm": 3.1955853435960506,
+      "learning_rate": 5.196567813279964e-06,
+      "loss": 1.0248,
+      "step": 6957
+    },
+    {
+      "epoch": 1.339461462569483,
+      "grad_norm": 3.243707155436304,
+      "learning_rate": 5.193833400123034e-06,
+      "loss": 1.0726,
+      "step": 6958
+    },
+    {
+      "epoch": 1.3396539692470584,
+      "grad_norm": 3.2248598012651035,
+      "learning_rate": 5.191099454186984e-06,
+      "loss": 1.1026,
+      "step": 6959
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "grad_norm": 3.3288835063635784,
+      "learning_rate": 5.188365975737595e-06,
+      "loss": 1.0695,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.7154,
+      "step": 6960,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.938,
+      "step": 6960,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.3825,
+      "step": 6960,
+      "vm_loss": 0.105
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.9634,
+      "step": 6960,
+      "vm_loss": 0.1038
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 1.1795,
+      "step": 6960,
+      "vm_loss": 0.2573
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.8148,
+      "step": 6960,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 1.2566,
+      "step": 6960,
+      "vm_loss": 0.2478
+    },
+    {
+      "epoch": 1.3398464759246336,
+      "lm_loss": 0.9161,
+      "step": 6960,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 1.340038982602209,
+      "grad_norm": 3.2204728302890913,
+      "learning_rate": 5.185632965040589e-06,
+      "loss": 1.0324,
+      "step": 6961
+    },
+    {
+      "epoch": 1.3402314892797844,
+      "grad_norm": 3.1720486745009273,
+      "learning_rate": 5.182900422361652e-06,
+      "loss": 1.0168,
+      "step": 6962
+    },
+    {
+      "epoch": 1.3404239959573598,
+      "grad_norm": 3.217575165751346,
+      "learning_rate": 5.180168347966422e-06,
+      "loss": 1.1317,
+      "step": 6963
+    },
+    {
+      "epoch": 1.3406165026349353,
+      "grad_norm": 3.2126670111959887,
+      "learning_rate": 5.1774367421204975e-06,
+      "loss": 1.0256,
+      "step": 6964
+    },
+    {
+      "epoch": 1.3408090093125105,
+      "grad_norm": 3.245098369950989,
+      "learning_rate": 5.174705605089412e-06,
+      "loss": 1.0561,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3410015159900859,
+      "grad_norm": 3.3202999568300204,
+      "learning_rate": 5.171974937138682e-06,
+      "loss": 1.1031,
+      "step": 6966
+    },
+    {
+      "epoch": 1.3411940226676613,
+      "grad_norm": 3.2245200417534234,
+      "learning_rate": 5.169244738533754e-06,
+      "loss": 0.9983,
+      "step": 6967
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "grad_norm": 3.3479766639087245,
+      "learning_rate": 5.1665150095400395e-06,
+      "loss": 1.0662,
+      "step": 6968
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.9483,
+      "step": 6968,
+      "vm_loss": 0.094
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.6806,
+      "step": 6968,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.9263,
+      "step": 6968,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.5996,
+      "step": 6968,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.7838,
+      "step": 6968,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.5778,
+      "step": 6968,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.8887,
+      "step": 6968,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 1.3413865293452367,
+      "lm_loss": 0.9038,
+      "step": 6968,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 1.3415790360228121,
+      "grad_norm": 3.2784356051830805,
+      "learning_rate": 5.163785750422905e-06,
+      "loss": 1.0513,
+      "step": 6969
+    },
+    {
+      "epoch": 1.3417715427003873,
+      "grad_norm": 3.3227694192099086,
+      "learning_rate": 5.1610569614476735e-06,
+      "loss": 1.0921,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3419640493779628,
+      "grad_norm": 3.310711693670043,
+      "learning_rate": 5.1583286428796045e-06,
+      "loss": 1.0513,
+      "step": 6971
+    },
+    {
+      "epoch": 1.3421565560555382,
+      "grad_norm": 3.266230708433354,
+      "learning_rate": 5.1556007949839435e-06,
+      "loss": 1.0617,
+      "step": 6972
+    },
+    {
+      "epoch": 1.3423490627331136,
+      "grad_norm": 3.443904753766547,
+      "learning_rate": 5.152873418025859e-06,
+      "loss": 1.1211,
+      "step": 6973
+    },
+    {
+      "epoch": 1.342541569410689,
+      "grad_norm": 3.238602411064094,
+      "learning_rate": 5.150146512270494e-06,
+      "loss": 1.0233,
+      "step": 6974
+    },
+    {
+      "epoch": 1.3427340760882642,
+      "grad_norm": 3.3151212479413306,
+      "learning_rate": 5.147420077982928e-06,
+      "loss": 1.0981,
+      "step": 6975
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "grad_norm": 3.240780362478548,
+      "learning_rate": 5.144694115428222e-06,
+      "loss": 1.0353,
+      "step": 6976
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 0.7008,
+      "step": 6976,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 0.5197,
+      "step": 6976,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 0.7226,
+      "step": 6976,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 1.0323,
+      "step": 6976,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 1.1335,
+      "step": 6976,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 1.3661,
+      "step": 6976,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 0.8359,
+      "step": 6976,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.3429265827658396,
+      "lm_loss": 0.9437,
+      "step": 6976,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 1.343119089443415,
+      "grad_norm": 3.361879349342819,
+      "learning_rate": 5.14196862487136e-06,
+      "loss": 1.0546,
+      "step": 6977
+    },
+    {
+      "epoch": 1.3433115961209905,
+      "grad_norm": 3.1506508763458676,
+      "learning_rate": 5.139243606577302e-06,
+      "loss": 1.0283,
+      "step": 6978
+    },
+    {
+      "epoch": 1.343504102798566,
+      "grad_norm": 3.242870749476808,
+      "learning_rate": 5.136519060810952e-06,
+      "loss": 1.0785,
+      "step": 6979
+    },
+    {
+      "epoch": 1.3436966094761411,
+      "grad_norm": 3.270217491963181,
+      "learning_rate": 5.1337949878371705e-06,
+      "loss": 1.0961,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3438891161537165,
+      "grad_norm": 3.10994473058476,
+      "learning_rate": 5.1310713879207786e-06,
+      "loss": 1.0358,
+      "step": 6981
+    },
+    {
+      "epoch": 1.344081622831292,
+      "grad_norm": 3.2441401483734804,
+      "learning_rate": 5.1283482613265305e-06,
+      "loss": 1.07,
+      "step": 6982
+    },
+    {
+      "epoch": 1.3442741295088674,
+      "grad_norm": 3.2136902284051723,
+      "learning_rate": 5.1256256083191665e-06,
+      "loss": 1.0585,
+      "step": 6983
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "grad_norm": 3.202221464912572,
+      "learning_rate": 5.1229034291633505e-06,
+      "loss": 1.0539,
+      "step": 6984
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.9509,
+      "step": 6984,
+      "vm_loss": 0.2163
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.5015,
+      "step": 6984,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.6393,
+      "step": 6984,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.9299,
+      "step": 6984,
+      "vm_loss": 0.2031
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.9546,
+      "step": 6984,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 1.2449,
+      "step": 6984,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 0.6238,
+      "step": 6984,
+      "vm_loss": 0.235
+    },
+    {
+      "epoch": 1.3444666361864428,
+      "lm_loss": 1.1046,
+      "step": 6984,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 1.344659142864018,
+      "grad_norm": 3.0768520183826746,
+      "learning_rate": 5.1201817241237175e-06,
+      "loss": 1.0507,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3448516495415934,
+      "grad_norm": 3.261797715322335,
+      "learning_rate": 5.117460493464852e-06,
+      "loss": 1.0569,
+      "step": 6986
+    },
+    {
+      "epoch": 1.3450441562191688,
+      "grad_norm": 3.2171081270955604,
+      "learning_rate": 5.114739737451297e-06,
+      "loss": 1.0383,
+      "step": 6987
+    },
+    {
+      "epoch": 1.3452366628967443,
+      "grad_norm": 3.2706009020803033,
+      "learning_rate": 5.1120194563475315e-06,
+      "loss": 1.0872,
+      "step": 6988
+    },
+    {
+      "epoch": 1.3454291695743197,
+      "grad_norm": 3.1403946841379753,
+      "learning_rate": 5.109299650418018e-06,
+      "loss": 1.0149,
+      "step": 6989
+    },
+    {
+      "epoch": 1.3456216762518949,
+      "grad_norm": 3.14067313338354,
+      "learning_rate": 5.106580319927146e-06,
+      "loss": 1.04,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3458141829294703,
+      "grad_norm": 3.1176850897260273,
+      "learning_rate": 5.103861465139275e-06,
+      "loss": 1.0332,
+      "step": 6991
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "grad_norm": 3.1652729859081563,
+      "learning_rate": 5.101143086318703e-06,
+      "loss": 1.0394,
+      "step": 6992
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 0.4395,
+      "step": 6992,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 0.7579,
+      "step": 6992,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 1.0986,
+      "step": 6992,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 1.1285,
+      "step": 6992,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 0.8946,
+      "step": 6992,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 0.8644,
+      "step": 6992,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 1.4697,
+      "step": 6992,
+      "vm_loss": 0.1036
+    },
+    {
+      "epoch": 1.3460066896070457,
+      "lm_loss": 0.874,
+      "step": 6992,
+      "vm_loss": 0.1193
+    },
+    {
+      "epoch": 1.3461991962846211,
+      "grad_norm": 3.2511624077096597,
+      "learning_rate": 5.098425183729704e-06,
+      "loss": 1.0265,
+      "step": 6993
+    },
+    {
+      "epoch": 1.3463917029621966,
+      "grad_norm": 3.313968484559467,
+      "learning_rate": 5.095707757636484e-06,
+      "loss": 1.0642,
+      "step": 6994
+    },
+    {
+      "epoch": 1.3465842096397718,
+      "grad_norm": 3.398714973198529,
+      "learning_rate": 5.092990808303213e-06,
+      "loss": 1.0852,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3467767163173472,
+      "grad_norm": 3.2429646263022707,
+      "learning_rate": 5.090274335994014e-06,
+      "loss": 1.066,
+      "step": 6996
+    },
+    {
+      "epoch": 1.3469692229949226,
+      "grad_norm": 3.226575785924758,
+      "learning_rate": 5.0875583409729676e-06,
+      "loss": 1.0688,
+      "step": 6997
+    },
+    {
+      "epoch": 1.347161729672498,
+      "grad_norm": 3.3889982820683042,
+      "learning_rate": 5.084842823504096e-06,
+      "loss": 1.1398,
+      "step": 6998
+    },
+    {
+      "epoch": 1.3473542363500735,
+      "grad_norm": 3.3836863269700888,
+      "learning_rate": 5.082127783851385e-06,
+      "loss": 1.1321,
+      "step": 6999
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "grad_norm": 3.1329966084118817,
+      "learning_rate": 5.07941322227877e-06,
+      "loss": 1.0241,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 1.3697,
+      "step": 7000,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 1.295,
+      "step": 7000,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 1.0614,
+      "step": 7000,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 1.4258,
+      "step": 7000,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 0.5607,
+      "step": 7000,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 0.6863,
+      "step": 7000,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 0.8159,
+      "step": 7000,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 1.3475467430276487,
+      "lm_loss": 0.8655,
+      "step": 7000,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.347739249705224,
+      "grad_norm": 3.1971140212292912,
+      "learning_rate": 5.076699139050144e-06,
+      "loss": 1.1068,
+      "step": 7001
+    },
+    {
+      "epoch": 1.3479317563827995,
+      "grad_norm": 3.2873754937552895,
+      "learning_rate": 5.073985534429349e-06,
+      "loss": 1.1125,
+      "step": 7002
+    },
+    {
+      "epoch": 1.348124263060375,
+      "grad_norm": 3.134492004716534,
+      "learning_rate": 5.071272408680181e-06,
+      "loss": 1.0287,
+      "step": 7003
+    },
+    {
+      "epoch": 1.3483167697379503,
+      "grad_norm": 3.3359563863463264,
+      "learning_rate": 5.068559762066396e-06,
+      "loss": 1.1014,
+      "step": 7004
+    },
+    {
+      "epoch": 1.3485092764155255,
+      "grad_norm": 3.1324659075162247,
+      "learning_rate": 5.065847594851692e-06,
+      "loss": 1.0521,
+      "step": 7005
+    },
+    {
+      "epoch": 1.348701783093101,
+      "grad_norm": 3.20751896207077,
+      "learning_rate": 5.063135907299726e-06,
+      "loss": 1.0577,
+      "step": 7006
+    },
+    {
+      "epoch": 1.3488942897706764,
+      "grad_norm": 3.4023816127728703,
+      "learning_rate": 5.060424699674109e-06,
+      "loss": 1.0772,
+      "step": 7007
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "grad_norm": 3.2637932288110423,
+      "learning_rate": 5.057713972238414e-06,
+      "loss": 1.0365,
+      "step": 7008
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 1.0385,
+      "step": 7008,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 1.0117,
+      "step": 7008,
+      "vm_loss": 0.1255
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 1.7242,
+      "step": 7008,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 1.1684,
+      "step": 7008,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 1.0217,
+      "step": 7008,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 0.6708,
+      "step": 7008,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 0.6978,
+      "step": 7008,
+      "vm_loss": 0.2026
+    },
+    {
+      "epoch": 1.3490867964482518,
+      "lm_loss": 0.9669,
+      "step": 7008,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.3492793031258272,
+      "grad_norm": 3.3081169392962564,
+      "learning_rate": 5.055003725256142e-06,
+      "loss": 1.0759,
+      "step": 7009
+    },
+    {
+      "epoch": 1.3494718098034024,
+      "grad_norm": 3.2859245884411665,
+      "learning_rate": 5.05229395899078e-06,
+      "loss": 1.0728,
+      "step": 7010
+    },
+    {
+      "epoch": 1.349664316480978,
+      "grad_norm": 3.401787157871432,
+      "learning_rate": 5.049584673705742e-06,
+      "loss": 1.0894,
+      "step": 7011
+    },
+    {
+      "epoch": 1.3498568231585533,
+      "grad_norm": 3.3308832663866927,
+      "learning_rate": 5.046875869664407e-06,
+      "loss": 1.0842,
+      "step": 7012
+    },
+    {
+      "epoch": 1.3500493298361287,
+      "grad_norm": 3.2625058610122983,
+      "learning_rate": 5.044167547130108e-06,
+      "loss": 1.0355,
+      "step": 7013
+    },
+    {
+      "epoch": 1.350241836513704,
+      "grad_norm": 3.262555258201198,
+      "learning_rate": 5.04145970636613e-06,
+      "loss": 1.0667,
+      "step": 7014
+    },
+    {
+      "epoch": 1.3504343431912795,
+      "grad_norm": 3.1523685896757234,
+      "learning_rate": 5.0387523476357024e-06,
+      "loss": 1.035,
+      "step": 7015
+    },
+    {
+      "epoch": 1.350626849868855,
+      "grad_norm": 3.2261397396233473,
+      "learning_rate": 5.03604547120202e-06,
+      "loss": 1.0417,
+      "step": 7016
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 0.7755,
+      "step": 7016,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 1.5116,
+      "step": 7016,
+      "vm_loss": 0.2167
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 1.3271,
+      "step": 7016,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 0.9609,
+      "step": 7016,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 0.854,
+      "step": 7016,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 0.5871,
+      "step": 7016,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 1.1658,
+      "step": 7016,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.350626849868855,
+      "lm_loss": 0.6783,
+      "step": 7016,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 1.3508193565464301,
+      "grad_norm": 3.276769872753125,
+      "learning_rate": 5.033339077328225e-06,
+      "loss": 1.083,
+      "step": 7017
+    },
+    {
+      "epoch": 1.3510118632240056,
+      "grad_norm": 3.1706901801997582,
+      "learning_rate": 5.0306331662774185e-06,
+      "loss": 1.0477,
+      "step": 7018
+    },
+    {
+      "epoch": 1.351204369901581,
+      "grad_norm": 3.2552947299878676,
+      "learning_rate": 5.027927738312637e-06,
+      "loss": 1.0454,
+      "step": 7019
+    },
+    {
+      "epoch": 1.3513968765791564,
+      "grad_norm": 3.407927209008117,
+      "learning_rate": 5.025222793696898e-06,
+      "loss": 1.1409,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3515893832567318,
+      "grad_norm": 3.1788750969842368,
+      "learning_rate": 5.022518332693145e-06,
+      "loss": 1.0454,
+      "step": 7021
+    },
+    {
+      "epoch": 1.351781889934307,
+      "grad_norm": 3.3358747674809104,
+      "learning_rate": 5.019814355564292e-06,
+      "loss": 1.0628,
+      "step": 7022
+    },
+    {
+      "epoch": 1.3519743966118825,
+      "grad_norm": 3.2604331967790743,
+      "learning_rate": 5.017110862573198e-06,
+      "loss": 1.0653,
+      "step": 7023
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "grad_norm": 3.242091255967768,
+      "learning_rate": 5.014407853982679e-06,
+      "loss": 1.0925,
+      "step": 7024
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 0.7482,
+      "step": 7024,
+      "vm_loss": 0.2584
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 1.0807,
+      "step": 7024,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 0.9074,
+      "step": 7024,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 1.0855,
+      "step": 7024,
+      "vm_loss": 0.2429
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 1.5356,
+      "step": 7024,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 0.7081,
+      "step": 7024,
+      "vm_loss": 0.2358
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 0.6794,
+      "step": 7024,
+      "vm_loss": 0.1327
+    },
+    {
+      "epoch": 1.3521669032894579,
+      "lm_loss": 0.9537,
+      "step": 7024,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.3523594099670333,
+      "grad_norm": 3.4084755133963758,
+      "learning_rate": 5.011705330055503e-06,
+      "loss": 1.1128,
+      "step": 7025
+    },
+    {
+      "epoch": 1.3525519166446087,
+      "grad_norm": 3.1586039917168343,
+      "learning_rate": 5.009003291054382e-06,
+      "loss": 1.0662,
+      "step": 7026
+    },
+    {
+      "epoch": 1.352744423322184,
+      "grad_norm": 3.1051612367691717,
+      "learning_rate": 5.006301737242001e-06,
+      "loss": 0.981,
+      "step": 7027
+    },
+    {
+      "epoch": 1.3529369299997593,
+      "grad_norm": 3.257798011986029,
+      "learning_rate": 5.003600668880975e-06,
+      "loss": 1.0241,
+      "step": 7028
+    },
+    {
+      "epoch": 1.3531294366773348,
+      "grad_norm": 3.1825731148571967,
+      "learning_rate": 5.000900086233888e-06,
+      "loss": 1.0742,
+      "step": 7029
+    },
+    {
+      "epoch": 1.3533219433549102,
+      "grad_norm": 3.3461707133581724,
+      "learning_rate": 4.998199989563269e-06,
+      "loss": 1.0969,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3535144500324856,
+      "grad_norm": 3.292655661511612,
+      "learning_rate": 4.995500379131607e-06,
+      "loss": 1.0967,
+      "step": 7031
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "grad_norm": 3.243348524788829,
+      "learning_rate": 4.99280125520133e-06,
+      "loss": 1.0691,
+      "step": 7032
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.4519,
+      "step": 7032,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.6597,
+      "step": 7032,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.7942,
+      "step": 7032,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.6231,
+      "step": 7032,
+      "vm_loss": 0.2301
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.6757,
+      "step": 7032,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 1.002,
+      "step": 7032,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.9637,
+      "step": 7032,
+      "vm_loss": 0.2254
+    },
+    {
+      "epoch": 1.3537069567100608,
+      "lm_loss": 0.7412,
+      "step": 7032,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.3538994633876362,
+      "grad_norm": 3.2221471019601835,
+      "learning_rate": 4.990102618034832e-06,
+      "loss": 1.0794,
+      "step": 7033
+    },
+    {
+      "epoch": 1.3540919700652116,
+      "grad_norm": 3.3941904765736495,
+      "learning_rate": 4.987404467894456e-06,
+      "loss": 1.0973,
+      "step": 7034
+    },
+    {
+      "epoch": 1.354284476742787,
+      "grad_norm": 3.461218689878161,
+      "learning_rate": 4.984706805042497e-06,
+      "loss": 1.0949,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3544769834203625,
+      "grad_norm": 3.2016768049193502,
+      "learning_rate": 4.9820096297411945e-06,
+      "loss": 1.0234,
+      "step": 7036
+    },
+    {
+      "epoch": 1.3546694900979377,
+      "grad_norm": 3.4973752911200022,
+      "learning_rate": 4.979312942252763e-06,
+      "loss": 1.185,
+      "step": 7037
+    },
+    {
+      "epoch": 1.3548619967755131,
+      "grad_norm": 3.0898950681327846,
+      "learning_rate": 4.976616742839342e-06,
+      "loss": 1.0132,
+      "step": 7038
+    },
+    {
+      "epoch": 1.3550545034530885,
+      "grad_norm": 3.2200748093646845,
+      "learning_rate": 4.973921031763042e-06,
+      "loss": 1.1154,
+      "step": 7039
+    },
+    {
+      "epoch": 1.355247010130664,
+      "grad_norm": 3.1017967673750433,
+      "learning_rate": 4.97122580928592e-06,
+      "loss": 1.0223,
+      "step": 7040
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 0.9618,
+      "step": 7040,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 0.9607,
+      "step": 7040,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 0.8893,
+      "step": 7040,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 0.8751,
+      "step": 7040,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 1.0891,
+      "step": 7040,
+      "vm_loss": 0.0878
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 1.3179,
+      "step": 7040,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 1.0796,
+      "step": 7040,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.355247010130664,
+      "lm_loss": 0.6549,
+      "step": 7040,
+      "vm_loss": 0.2317
+    },
+    {
+      "epoch": 1.3554395168082394,
+      "grad_norm": 3.228617061313726,
+      "learning_rate": 4.96853107566999e-06,
+      "loss": 1.0932,
+      "step": 7041
+    },
+    {
+      "epoch": 1.3556320234858146,
+      "grad_norm": 3.2204735756778393,
+      "learning_rate": 4.965836831177206e-06,
+      "loss": 1.1058,
+      "step": 7042
+    },
+    {
+      "epoch": 1.35582453016339,
+      "grad_norm": 3.203507371510655,
+      "learning_rate": 4.96314307606949e-06,
+      "loss": 1.0637,
+      "step": 7043
+    },
+    {
+      "epoch": 1.3560170368409654,
+      "grad_norm": 3.203881705389357,
+      "learning_rate": 4.960449810608705e-06,
+      "loss": 1.0911,
+      "step": 7044
+    },
+    {
+      "epoch": 1.3562095435185408,
+      "grad_norm": 3.1978205491362113,
+      "learning_rate": 4.957757035056674e-06,
+      "loss": 1.0314,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3564020501961163,
+      "grad_norm": 3.2535014130887236,
+      "learning_rate": 4.955064749675168e-06,
+      "loss": 1.0927,
+      "step": 7046
+    },
+    {
+      "epoch": 1.3565945568736915,
+      "grad_norm": 3.2249263651046705,
+      "learning_rate": 4.9523729547259125e-06,
+      "loss": 1.0469,
+      "step": 7047
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "grad_norm": 3.167104445490603,
+      "learning_rate": 4.949681650470587e-06,
+      "loss": 1.031,
+      "step": 7048
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 1.012,
+      "step": 7048,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.5898,
+      "step": 7048,
+      "vm_loss": 0.2261
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.8023,
+      "step": 7048,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.9925,
+      "step": 7048,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.726,
+      "step": 7048,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.8478,
+      "step": 7048,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 0.6665,
+      "step": 7048,
+      "vm_loss": 0.1177
+    },
+    {
+      "epoch": 1.3567870635512669,
+      "lm_loss": 1.1384,
+      "step": 7048,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.3569795702288423,
+      "grad_norm": 3.1222255461659945,
+      "learning_rate": 4.946990837170814e-06,
+      "loss": 1.0054,
+      "step": 7049
+    },
+    {
+      "epoch": 1.3571720769064177,
+      "grad_norm": 3.4422307586468976,
+      "learning_rate": 4.944300515088178e-06,
+      "loss": 1.1559,
+      "step": 7050
+    },
+    {
+      "epoch": 1.3573645835839931,
+      "grad_norm": 3.230864226898403,
+      "learning_rate": 4.941610684484214e-06,
+      "loss": 1.0781,
+      "step": 7051
+    },
+    {
+      "epoch": 1.3575570902615683,
+      "grad_norm": 3.2991112150014317,
+      "learning_rate": 4.938921345620411e-06,
+      "loss": 1.0897,
+      "step": 7052
+    },
+    {
+      "epoch": 1.3577495969391438,
+      "grad_norm": 3.363526726113174,
+      "learning_rate": 4.936232498758195e-06,
+      "loss": 1.0753,
+      "step": 7053
+    },
+    {
+      "epoch": 1.3579421036167192,
+      "grad_norm": 3.284159403744669,
+      "learning_rate": 4.933544144158972e-06,
+      "loss": 1.0254,
+      "step": 7054
+    },
+    {
+      "epoch": 1.3581346102942946,
+      "grad_norm": 3.209213872587984,
+      "learning_rate": 4.930856282084074e-06,
+      "loss": 1.0024,
+      "step": 7055
+    },
+    {
+      "epoch": 1.35832711697187,
+      "grad_norm": 3.232070739964759,
+      "learning_rate": 4.928168912794803e-06,
+      "loss": 1.0453,
+      "step": 7056
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 1.2744,
+      "step": 7056,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 0.7717,
+      "step": 7056,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 1.1349,
+      "step": 7056,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 0.9859,
+      "step": 7056,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 0.7024,
+      "step": 7056,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 0.8248,
+      "step": 7056,
+      "vm_loss": 0.1957
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 1.1315,
+      "step": 7056,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.35832711697187,
+      "lm_loss": 1.0956,
+      "step": 7056,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.3585196236494452,
+      "grad_norm": 3.332246641255352,
+      "learning_rate": 4.925482036552393e-06,
+      "loss": 1.0732,
+      "step": 7057
+    },
+    {
+      "epoch": 1.3587121303270207,
+      "grad_norm": 3.3276336153700465,
+      "learning_rate": 4.922795653618059e-06,
+      "loss": 1.044,
+      "step": 7058
+    },
+    {
+      "epoch": 1.358904637004596,
+      "grad_norm": 3.2191386126354686,
+      "learning_rate": 4.920109764252939e-06,
+      "loss": 1.0459,
+      "step": 7059
+    },
+    {
+      "epoch": 1.3590971436821715,
+      "grad_norm": 3.375590773088858,
+      "learning_rate": 4.917424368718141e-06,
+      "loss": 1.0765,
+      "step": 7060
+    },
+    {
+      "epoch": 1.359289650359747,
+      "grad_norm": 3.213202976080089,
+      "learning_rate": 4.914739467274721e-06,
+      "loss": 1.0104,
+      "step": 7061
+    },
+    {
+      "epoch": 1.3594821570373221,
+      "grad_norm": 3.3084010657113203,
+      "learning_rate": 4.912055060183687e-06,
+      "loss": 1.0538,
+      "step": 7062
+    },
+    {
+      "epoch": 1.3596746637148975,
+      "grad_norm": 3.1141084709878037,
+      "learning_rate": 4.909371147705987e-06,
+      "loss": 0.964,
+      "step": 7063
+    },
+    {
+      "epoch": 1.359867170392473,
+      "grad_norm": 3.3617157213971813,
+      "learning_rate": 4.906687730102547e-06,
+      "loss": 1.0662,
+      "step": 7064
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.8885,
+      "step": 7064,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.7341,
+      "step": 7064,
+      "vm_loss": 0.0982
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.8265,
+      "step": 7064,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.7617,
+      "step": 7064,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 1.1231,
+      "step": 7064,
+      "vm_loss": 0.1338
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.6978,
+      "step": 7064,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.864,
+      "step": 7064,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.359867170392473,
+      "lm_loss": 0.8986,
+      "step": 7064,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.3600596770700484,
+      "grad_norm": 3.1566727302445945,
+      "learning_rate": 4.9040048076342195e-06,
+      "loss": 1.0205,
+      "step": 7065
+    },
+    {
+      "epoch": 1.3602521837476238,
+      "grad_norm": 3.364266433699712,
+      "learning_rate": 4.90132238056182e-06,
+      "loss": 1.1213,
+      "step": 7066
+    },
+    {
+      "epoch": 1.360444690425199,
+      "grad_norm": 3.281674362027754,
+      "learning_rate": 4.898640449146118e-06,
+      "loss": 1.0511,
+      "step": 7067
+    },
+    {
+      "epoch": 1.3606371971027744,
+      "grad_norm": 3.3329300665891,
+      "learning_rate": 4.895959013647828e-06,
+      "loss": 1.0947,
+      "step": 7068
+    },
+    {
+      "epoch": 1.3608297037803498,
+      "grad_norm": 3.2369855966483914,
+      "learning_rate": 4.893278074327627e-06,
+      "loss": 1.0573,
+      "step": 7069
+    },
+    {
+      "epoch": 1.3610222104579253,
+      "grad_norm": 3.2681183810022216,
+      "learning_rate": 4.890597631446124e-06,
+      "loss": 1.1227,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3612147171355007,
+      "grad_norm": 3.2419842159483814,
+      "learning_rate": 4.887917685263906e-06,
+      "loss": 1.0745,
+      "step": 7071
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "grad_norm": 3.1726430750595567,
+      "learning_rate": 4.885238236041487e-06,
+      "loss": 1.0189,
+      "step": 7072
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.7693,
+      "step": 7072,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 1.1509,
+      "step": 7072,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.9204,
+      "step": 7072,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.775,
+      "step": 7072,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 1.1243,
+      "step": 7072,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.7317,
+      "step": 7072,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.7785,
+      "step": 7072,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.3614072238130759,
+      "lm_loss": 0.8687,
+      "step": 7072,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.3615997304906515,
+      "grad_norm": 3.1079160357638527,
+      "learning_rate": 4.882559284039353e-06,
+      "loss": 1.0343,
+      "step": 7073
+    },
+    {
+      "epoch": 1.3617922371682267,
+      "grad_norm": 3.347357138036633,
+      "learning_rate": 4.879880829517921e-06,
+      "loss": 1.0684,
+      "step": 7074
+    },
+    {
+      "epoch": 1.3619847438458021,
+      "grad_norm": 3.1897892769497442,
+      "learning_rate": 4.877202872737585e-06,
+      "loss": 1.0463,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3621772505233776,
+      "grad_norm": 3.1912344431291984,
+      "learning_rate": 4.874525413958665e-06,
+      "loss": 1.0251,
+      "step": 7076
+    },
+    {
+      "epoch": 1.362369757200953,
+      "grad_norm": 3.237945837556285,
+      "learning_rate": 4.871848453441449e-06,
+      "loss": 1.077,
+      "step": 7077
+    },
+    {
+      "epoch": 1.3625622638785284,
+      "grad_norm": 3.2507323025596366,
+      "learning_rate": 4.8691719914461725e-06,
+      "loss": 1.0511,
+      "step": 7078
+    },
+    {
+      "epoch": 1.3627547705561036,
+      "grad_norm": 3.3846922632320147,
+      "learning_rate": 4.866496028233023e-06,
+      "loss": 1.0757,
+      "step": 7079
+    },
+    {
+      "epoch": 1.362947277233679,
+      "grad_norm": 3.308047938623538,
+      "learning_rate": 4.86382056406213e-06,
+      "loss": 1.1051,
+      "step": 7080
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.848,
+      "step": 7080,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.7518,
+      "step": 7080,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 1.1171,
+      "step": 7080,
+      "vm_loss": 0.2214
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.846,
+      "step": 7080,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.7262,
+      "step": 7080,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.8489,
+      "step": 7080,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 0.7605,
+      "step": 7080,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 1.362947277233679,
+      "lm_loss": 1.07,
+      "step": 7080,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.3631397839112545,
+      "grad_norm": 3.314393392928482,
+      "learning_rate": 4.861145599193596e-06,
+      "loss": 1.0714,
+      "step": 7081
+    },
+    {
+      "epoch": 1.3633322905888299,
+      "grad_norm": 3.234185518230245,
+      "learning_rate": 4.858471133887451e-06,
+      "loss": 0.9809,
+      "step": 7082
+    },
+    {
+      "epoch": 1.3635247972664053,
+      "grad_norm": 3.2570231782122177,
+      "learning_rate": 4.855797168403692e-06,
+      "loss": 1.0603,
+      "step": 7083
+    },
+    {
+      "epoch": 1.3637173039439805,
+      "grad_norm": 3.306835201132162,
+      "learning_rate": 4.853123703002262e-06,
+      "loss": 1.1068,
+      "step": 7084
+    },
+    {
+      "epoch": 1.363909810621556,
+      "grad_norm": 3.225418981328935,
+      "learning_rate": 4.8504507379430584e-06,
+      "loss": 1.0754,
+      "step": 7085
+    },
+    {
+      "epoch": 1.3641023172991313,
+      "grad_norm": 3.289673204544399,
+      "learning_rate": 4.8477782734859215e-06,
+      "loss": 1.0843,
+      "step": 7086
+    },
+    {
+      "epoch": 1.3642948239767068,
+      "grad_norm": 3.145958458809384,
+      "learning_rate": 4.845106309890654e-06,
+      "loss": 1.049,
+      "step": 7087
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "grad_norm": 3.193507920095359,
+      "learning_rate": 4.842434847417001e-06,
+      "loss": 1.0174,
+      "step": 7088
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.6428,
+      "step": 7088,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.7582,
+      "step": 7088,
+      "vm_loss": 0.187
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 1.0783,
+      "step": 7088,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.6633,
+      "step": 7088,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 1.1022,
+      "step": 7088,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.9337,
+      "step": 7088,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.8289,
+      "step": 7088,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.3644873306542822,
+      "lm_loss": 0.6194,
+      "step": 7088,
+      "vm_loss": 0.1258
+    },
+    {
+      "epoch": 1.3646798373318574,
+      "grad_norm": 3.2094122593164616,
+      "learning_rate": 4.839763886324668e-06,
+      "loss": 1.0365,
+      "step": 7089
+    },
+    {
+      "epoch": 1.3648723440094328,
+      "grad_norm": 3.1914931873834833,
+      "learning_rate": 4.837093426873306e-06,
+      "loss": 1.0235,
+      "step": 7090
+    },
+    {
+      "epoch": 1.3650648506870082,
+      "grad_norm": 3.288741738240951,
+      "learning_rate": 4.834423469322509e-06,
+      "loss": 1.1068,
+      "step": 7091
+    },
+    {
+      "epoch": 1.3652573573645836,
+      "grad_norm": 3.2756853261272747,
+      "learning_rate": 4.831754013931844e-06,
+      "loss": 1.0786,
+      "step": 7092
+    },
+    {
+      "epoch": 1.365449864042159,
+      "grad_norm": 3.2630285015588445,
+      "learning_rate": 4.829085060960808e-06,
+      "loss": 1.0209,
+      "step": 7093
+    },
+    {
+      "epoch": 1.3656423707197343,
+      "grad_norm": 3.2784387953089786,
+      "learning_rate": 4.826416610668858e-06,
+      "loss": 1.0458,
+      "step": 7094
+    },
+    {
+      "epoch": 1.3658348773973097,
+      "grad_norm": 3.256025392665378,
+      "learning_rate": 4.823748663315404e-06,
+      "loss": 1.0558,
+      "step": 7095
+    },
+    {
+      "epoch": 1.366027384074885,
+      "grad_norm": 3.1385564305713367,
+      "learning_rate": 4.821081219159806e-06,
+      "loss": 0.9779,
+      "step": 7096
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 0.8178,
+      "step": 7096,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 0.8379,
+      "step": 7096,
+      "vm_loss": 0.2216
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 1.0425,
+      "step": 7096,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 0.5086,
+      "step": 7096,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 1.1079,
+      "step": 7096,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 1.2865,
+      "step": 7096,
+      "vm_loss": 0.2172
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 0.9702,
+      "step": 7096,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.366027384074885,
+      "lm_loss": 1.0967,
+      "step": 7096,
+      "vm_loss": 0.2411
+    },
+    {
+      "epoch": 1.3662198907524605,
+      "grad_norm": 3.2470761727469672,
+      "learning_rate": 4.818414278461365e-06,
+      "loss": 1.0857,
+      "step": 7097
+    },
+    {
+      "epoch": 1.366412397430036,
+      "grad_norm": 3.1784661220901542,
+      "learning_rate": 4.815747841479356e-06,
+      "loss": 1.0154,
+      "step": 7098
+    },
+    {
+      "epoch": 1.3666049041076112,
+      "grad_norm": 3.3479864030220448,
+      "learning_rate": 4.813081908472977e-06,
+      "loss": 1.063,
+      "step": 7099
+    },
+    {
+      "epoch": 1.3667974107851866,
+      "grad_norm": 3.329823176932758,
+      "learning_rate": 4.810416479701402e-06,
+      "loss": 1.0397,
+      "step": 7100
+    },
+    {
+      "epoch": 1.366989917462762,
+      "grad_norm": 3.362421904172485,
+      "learning_rate": 4.8077515554237304e-06,
+      "loss": 1.0458,
+      "step": 7101
+    },
+    {
+      "epoch": 1.3671824241403374,
+      "grad_norm": 3.44068970315734,
+      "learning_rate": 4.8050871358990446e-06,
+      "loss": 1.0953,
+      "step": 7102
+    },
+    {
+      "epoch": 1.3673749308179128,
+      "grad_norm": 3.1922532266400223,
+      "learning_rate": 4.802423221386347e-06,
+      "loss": 1.064,
+      "step": 7103
+    },
+    {
+      "epoch": 1.367567437495488,
+      "grad_norm": 3.1764295308848087,
+      "learning_rate": 4.79975981214461e-06,
+      "loss": 0.9904,
+      "step": 7104
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 0.955,
+      "step": 7104,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 0.6801,
+      "step": 7104,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 0.828,
+      "step": 7104,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 1.0659,
+      "step": 7104,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 0.7548,
+      "step": 7104,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 1.177,
+      "step": 7104,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 0.9026,
+      "step": 7104,
+      "vm_loss": 0.1029
+    },
+    {
+      "epoch": 1.367567437495488,
+      "lm_loss": 1.478,
+      "step": 7104,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.3677599441730635,
+      "grad_norm": 3.3017230584823722,
+      "learning_rate": 4.797096908432749e-06,
+      "loss": 1.03,
+      "step": 7105
+    },
+    {
+      "epoch": 1.3679524508506389,
+      "grad_norm": 3.224838174502266,
+      "learning_rate": 4.794434510509632e-06,
+      "loss": 1.0323,
+      "step": 7106
+    },
+    {
+      "epoch": 1.3681449575282143,
+      "grad_norm": 3.3641021800481,
+      "learning_rate": 4.791772618634084e-06,
+      "loss": 1.1041,
+      "step": 7107
+    },
+    {
+      "epoch": 1.3683374642057897,
+      "grad_norm": 3.242532957559337,
+      "learning_rate": 4.789111233064867e-06,
+      "loss": 1.0518,
+      "step": 7108
+    },
+    {
+      "epoch": 1.368529970883365,
+      "grad_norm": 3.3298553167355105,
+      "learning_rate": 4.786450354060703e-06,
+      "loss": 1.0286,
+      "step": 7109
+    },
+    {
+      "epoch": 1.3687224775609403,
+      "grad_norm": 3.3661302962865727,
+      "learning_rate": 4.783789981880267e-06,
+      "loss": 1.0656,
+      "step": 7110
+    },
+    {
+      "epoch": 1.3689149842385158,
+      "grad_norm": 3.0979804569039397,
+      "learning_rate": 4.781130116782179e-06,
+      "loss": 0.9882,
+      "step": 7111
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "grad_norm": 3.2378044900306406,
+      "learning_rate": 4.7784707590250124e-06,
+      "loss": 1.012,
+      "step": 7112
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 1.058,
+      "step": 7112,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 0.6313,
+      "step": 7112,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 1.413,
+      "step": 7112,
+      "vm_loss": 0.2613
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 1.1325,
+      "step": 7112,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 0.7352,
+      "step": 7112,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 0.7394,
+      "step": 7112,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 0.6363,
+      "step": 7112,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.3691074909160912,
+      "lm_loss": 0.7023,
+      "step": 7112,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.3692999975936666,
+      "grad_norm": 3.364143469123974,
+      "learning_rate": 4.775811908867295e-06,
+      "loss": 1.1084,
+      "step": 7113
+    },
+    {
+      "epoch": 1.3694925042712418,
+      "grad_norm": 3.2782866525261536,
+      "learning_rate": 4.773153566567489e-06,
+      "loss": 1.0511,
+      "step": 7114
+    },
+    {
+      "epoch": 1.3696850109488172,
+      "grad_norm": 3.193657171189352,
+      "learning_rate": 4.770495732384035e-06,
+      "loss": 1.0323,
+      "step": 7115
+    },
+    {
+      "epoch": 1.3698775176263926,
+      "grad_norm": 3.298515994385738,
+      "learning_rate": 4.767838406575297e-06,
+      "loss": 1.0548,
+      "step": 7116
+    },
+    {
+      "epoch": 1.370070024303968,
+      "grad_norm": 3.2968768391369414,
+      "learning_rate": 4.765181589399609e-06,
+      "loss": 1.0691,
+      "step": 7117
+    },
+    {
+      "epoch": 1.3702625309815435,
+      "grad_norm": 3.1937952089991013,
+      "learning_rate": 4.762525281115235e-06,
+      "loss": 1.0606,
+      "step": 7118
+    },
+    {
+      "epoch": 1.3704550376591187,
+      "grad_norm": 3.500103068445227,
+      "learning_rate": 4.759869481980418e-06,
+      "loss": 1.1315,
+      "step": 7119
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "grad_norm": 3.345112654935229,
+      "learning_rate": 4.7572141922533245e-06,
+      "loss": 1.0523,
+      "step": 7120
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.7338,
+      "step": 7120,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.7907,
+      "step": 7120,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.8778,
+      "step": 7120,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.9886,
+      "step": 7120,
+      "vm_loss": 0.0891
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 1.3416,
+      "step": 7120,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.6807,
+      "step": 7120,
+      "vm_loss": 0.1116
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 1.0028,
+      "step": 7120,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.3706475443366941,
+      "lm_loss": 0.4997,
+      "step": 7120,
+      "vm_loss": 0.0931
+    },
+    {
+      "epoch": 1.3708400510142695,
+      "grad_norm": 3.211014526309793,
+      "learning_rate": 4.754559412192087e-06,
+      "loss": 1.0034,
+      "step": 7121
+    },
+    {
+      "epoch": 1.371032557691845,
+      "grad_norm": 3.3156406872211184,
+      "learning_rate": 4.751905142054783e-06,
+      "loss": 1.0655,
+      "step": 7122
+    },
+    {
+      "epoch": 1.3712250643694204,
+      "grad_norm": 3.27702613506469,
+      "learning_rate": 4.7492513820994455e-06,
+      "loss": 1.0251,
+      "step": 7123
+    },
+    {
+      "epoch": 1.3714175710469956,
+      "grad_norm": 3.2672661785670485,
+      "learning_rate": 4.746598132584047e-06,
+      "loss": 1.0664,
+      "step": 7124
+    },
+    {
+      "epoch": 1.371610077724571,
+      "grad_norm": 3.3222556304415773,
+      "learning_rate": 4.74394539376652e-06,
+      "loss": 1.0387,
+      "step": 7125
+    },
+    {
+      "epoch": 1.3718025844021464,
+      "grad_norm": 3.2365192202430575,
+      "learning_rate": 4.741293165904747e-06,
+      "loss": 1.023,
+      "step": 7126
+    },
+    {
+      "epoch": 1.3719950910797218,
+      "grad_norm": 3.3319102237630367,
+      "learning_rate": 4.738641449256555e-06,
+      "loss": 1.101,
+      "step": 7127
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "grad_norm": 3.1907847592354575,
+      "learning_rate": 4.735990244079727e-06,
+      "loss": 1.0664,
+      "step": 7128
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.6782,
+      "step": 7128,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.8332,
+      "step": 7128,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.4181,
+      "step": 7128,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 1.1329,
+      "step": 7128,
+      "vm_loss": 0.0839
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.8547,
+      "step": 7128,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.9379,
+      "step": 7128,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 0.802,
+      "step": 7128,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 1.3721875977572973,
+      "lm_loss": 1.0139,
+      "step": 7128,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 1.3723801044348725,
+      "grad_norm": 3.1155307860371484,
+      "learning_rate": 4.733339550631993e-06,
+      "loss": 0.9863,
+      "step": 7129
+    },
+    {
+      "epoch": 1.3725726111124479,
+      "grad_norm": 3.1389427447816463,
+      "learning_rate": 4.73068936917104e-06,
+      "loss": 0.9647,
+      "step": 7130
+    },
+    {
+      "epoch": 1.3727651177900233,
+      "grad_norm": 3.218499440956164,
+      "learning_rate": 4.72803969995449e-06,
+      "loss": 1.0451,
+      "step": 7131
+    },
+    {
+      "epoch": 1.3729576244675987,
+      "grad_norm": 3.302689711491638,
+      "learning_rate": 4.7253905432399295e-06,
+      "loss": 1.0729,
+      "step": 7132
+    },
+    {
+      "epoch": 1.3731501311451741,
+      "grad_norm": 3.2381152662064507,
+      "learning_rate": 4.72274189928489e-06,
+      "loss": 1.0452,
+      "step": 7133
+    },
+    {
+      "epoch": 1.3733426378227493,
+      "grad_norm": 3.338192506894632,
+      "learning_rate": 4.720093768346858e-06,
+      "loss": 1.048,
+      "step": 7134
+    },
+    {
+      "epoch": 1.373535144500325,
+      "grad_norm": 3.276929344775282,
+      "learning_rate": 4.717446150683252e-06,
+      "loss": 1.0199,
+      "step": 7135
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "grad_norm": 3.3288586200508377,
+      "learning_rate": 4.714799046551474e-06,
+      "loss": 1.0117,
+      "step": 7136
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.9793,
+      "step": 7136,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.807,
+      "step": 7136,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.9508,
+      "step": 7136,
+      "vm_loss": 0.1881
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.8771,
+      "step": 7136,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.7458,
+      "step": 7136,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.6409,
+      "step": 7136,
+      "vm_loss": 0.1327
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.735,
+      "step": 7136,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.3737276511779002,
+      "lm_loss": 0.5206,
+      "step": 7136,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.3739201578554756,
+      "grad_norm": 3.1478304942148974,
+      "learning_rate": 4.712152456208841e-06,
+      "loss": 1.0212,
+      "step": 7137
+    },
+    {
+      "epoch": 1.374112664533051,
+      "grad_norm": 3.409246517272823,
+      "learning_rate": 4.709506379912641e-06,
+      "loss": 1.1143,
+      "step": 7138
+    },
+    {
+      "epoch": 1.3743051712106262,
+      "grad_norm": 3.379073869435661,
+      "learning_rate": 4.706860817920105e-06,
+      "loss": 1.0876,
+      "step": 7139
+    },
+    {
+      "epoch": 1.3744976778882019,
+      "grad_norm": 3.294287589027186,
+      "learning_rate": 4.704215770488422e-06,
+      "loss": 1.0562,
+      "step": 7140
+    },
+    {
+      "epoch": 1.374690184565777,
+      "grad_norm": 3.2621472671038028,
+      "learning_rate": 4.701571237874714e-06,
+      "loss": 1.0322,
+      "step": 7141
+    },
+    {
+      "epoch": 1.3748826912433525,
+      "grad_norm": 3.3430024728994443,
+      "learning_rate": 4.698927220336068e-06,
+      "loss": 1.0675,
+      "step": 7142
+    },
+    {
+      "epoch": 1.375075197920928,
+      "grad_norm": 3.2293080959707057,
+      "learning_rate": 4.696283718129517e-06,
+      "loss": 1.0128,
+      "step": 7143
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "grad_norm": 3.280628068137271,
+      "learning_rate": 4.6936407315120465e-06,
+      "loss": 1.0512,
+      "step": 7144
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 0.4082,
+      "step": 7144,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 1.0373,
+      "step": 7144,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 0.9889,
+      "step": 7144,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 0.8766,
+      "step": 7144,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 1.0124,
+      "step": 7144,
+      "vm_loss": 0.0992
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 0.6896,
+      "step": 7144,
+      "vm_loss": 0.1422
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 0.8161,
+      "step": 7144,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.3752677045985033,
+      "lm_loss": 1.1946,
+      "step": 7144,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 1.3754602112760788,
+      "grad_norm": 3.249676009271436,
+      "learning_rate": 4.690998260740579e-06,
+      "loss": 0.9981,
+      "step": 7145
+    },
+    {
+      "epoch": 1.375652717953654,
+      "grad_norm": 3.2617896197347966,
+      "learning_rate": 4.688356306072009e-06,
+      "loss": 1.0517,
+      "step": 7146
+    },
+    {
+      "epoch": 1.3758452246312294,
+      "grad_norm": 3.079311594261293,
+      "learning_rate": 4.685714867763158e-06,
+      "loss": 0.9999,
+      "step": 7147
+    },
+    {
+      "epoch": 1.3760377313088048,
+      "grad_norm": 3.126671100611527,
+      "learning_rate": 4.683073946070812e-06,
+      "loss": 0.9622,
+      "step": 7148
+    },
+    {
+      "epoch": 1.3762302379863802,
+      "grad_norm": 3.2983521018026587,
+      "learning_rate": 4.680433541251701e-06,
+      "loss": 1.0557,
+      "step": 7149
+    },
+    {
+      "epoch": 1.3764227446639556,
+      "grad_norm": 3.324486552324521,
+      "learning_rate": 4.677793653562507e-06,
+      "loss": 1.13,
+      "step": 7150
+    },
+    {
+      "epoch": 1.3766152513415308,
+      "grad_norm": 3.225894736096773,
+      "learning_rate": 4.6751542832598655e-06,
+      "loss": 1.0358,
+      "step": 7151
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "grad_norm": 3.376397167759517,
+      "learning_rate": 4.672515430600347e-06,
+      "loss": 1.1284,
+      "step": 7152
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.7219,
+      "step": 7152,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.7553,
+      "step": 7152,
+      "vm_loss": 0.2024
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.8683,
+      "step": 7152,
+      "vm_loss": 0.1417
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 1.2722,
+      "step": 7152,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.3439,
+      "step": 7152,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.7411,
+      "step": 7152,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 1.1973,
+      "step": 7152,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.3768077580191063,
+      "lm_loss": 0.9364,
+      "step": 7152,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.3770002646966817,
+      "grad_norm": 3.184742649793817,
+      "learning_rate": 4.669877095840488e-06,
+      "loss": 1.049,
+      "step": 7153
+    },
+    {
+      "epoch": 1.377192771374257,
+      "grad_norm": 3.236740749731261,
+      "learning_rate": 4.667239279236768e-06,
+      "loss": 1.0181,
+      "step": 7154
+    },
+    {
+      "epoch": 1.3773852780518325,
+      "grad_norm": 3.2962556154139815,
+      "learning_rate": 4.664601981045616e-06,
+      "loss": 1.0817,
+      "step": 7155
+    },
+    {
+      "epoch": 1.3775777847294077,
+      "grad_norm": 3.2590386163660128,
+      "learning_rate": 4.6619652015234095e-06,
+      "loss": 1.0095,
+      "step": 7156
+    },
+    {
+      "epoch": 1.3777702914069831,
+      "grad_norm": 3.199013012293532,
+      "learning_rate": 4.659328940926484e-06,
+      "loss": 1.0261,
+      "step": 7157
+    },
+    {
+      "epoch": 1.3779627980845586,
+      "grad_norm": 3.3087070396825284,
+      "learning_rate": 4.656693199511108e-06,
+      "loss": 1.0567,
+      "step": 7158
+    },
+    {
+      "epoch": 1.378155304762134,
+      "grad_norm": 3.068071030562416,
+      "learning_rate": 4.654057977533515e-06,
+      "loss": 1.0008,
+      "step": 7159
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "grad_norm": 3.0796367185420768,
+      "learning_rate": 4.65142327524988e-06,
+      "loss": 0.989,
+      "step": 7160
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.844,
+      "step": 7160,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.8576,
+      "step": 7160,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 1.0579,
+      "step": 7160,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.7478,
+      "step": 7160,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.8814,
+      "step": 7160,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.5713,
+      "step": 7160,
+      "vm_loss": 0.116
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.8367,
+      "step": 7160,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.3783478114397094,
+      "lm_loss": 0.9311,
+      "step": 7160,
+      "vm_loss": 0.1173
+    },
+    {
+      "epoch": 1.3785403181172846,
+      "grad_norm": 3.260286882713741,
+      "learning_rate": 4.648789092916335e-06,
+      "loss": 0.9486,
+      "step": 7161
+    },
+    {
+      "epoch": 1.37873282479486,
+      "grad_norm": 3.397740795172978,
+      "learning_rate": 4.646155430788944e-06,
+      "loss": 1.064,
+      "step": 7162
+    },
+    {
+      "epoch": 1.3789253314724355,
+      "grad_norm": 3.2767185316166674,
+      "learning_rate": 4.643522289123749e-06,
+      "loss": 1.024,
+      "step": 7163
+    },
+    {
+      "epoch": 1.3791178381500109,
+      "grad_norm": 3.2308683721170817,
+      "learning_rate": 4.640889668176713e-06,
+      "loss": 0.99,
+      "step": 7164
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 3.35784136839363,
+      "learning_rate": 4.6382575682037685e-06,
+      "loss": 1.0311,
+      "step": 7165
+    },
+    {
+      "epoch": 1.3795028515051615,
+      "grad_norm": 3.454245029992402,
+      "learning_rate": 4.635625989460778e-06,
+      "loss": 1.0897,
+      "step": 7166
+    },
+    {
+      "epoch": 1.379695358182737,
+      "grad_norm": 3.281500622963091,
+      "learning_rate": 4.632994932203579e-06,
+      "loss": 1.0475,
+      "step": 7167
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "grad_norm": 3.402296868399998,
+      "learning_rate": 4.630364396687932e-06,
+      "loss": 1.0621,
+      "step": 7168
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 0.826,
+      "step": 7168,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 0.6723,
+      "step": 7168,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 1.1915,
+      "step": 7168,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 0.9041,
+      "step": 7168,
+      "vm_loss": 0.1205
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 1.0895,
+      "step": 7168,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 0.9383,
+      "step": 7168,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 0.9183,
+      "step": 7168,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.3798878648603123,
+      "lm_loss": 1.2286,
+      "step": 7168,
+      "vm_loss": 0.0968
+    },
+    {
+      "epoch": 1.3800803715378878,
+      "grad_norm": 3.0971321077542577,
+      "learning_rate": 4.627734383169563e-06,
+      "loss": 0.9602,
+      "step": 7169
+    },
+    {
+      "epoch": 1.3802728782154632,
+      "grad_norm": 3.3412425059318887,
+      "learning_rate": 4.625104891904144e-06,
+      "loss": 1.0162,
+      "step": 7170
+    },
+    {
+      "epoch": 1.3804653848930384,
+      "grad_norm": 3.2311206880194283,
+      "learning_rate": 4.622475923147295e-06,
+      "loss": 1.0286,
+      "step": 7171
+    },
+    {
+      "epoch": 1.3806578915706138,
+      "grad_norm": 3.303908705157134,
+      "learning_rate": 4.619847477154584e-06,
+      "loss": 1.0514,
+      "step": 7172
+    },
+    {
+      "epoch": 1.3808503982481892,
+      "grad_norm": 3.1139753400666725,
+      "learning_rate": 4.617219554181531e-06,
+      "loss": 0.987,
+      "step": 7173
+    },
+    {
+      "epoch": 1.3810429049257646,
+      "grad_norm": 3.3128177591935604,
+      "learning_rate": 4.614592154483607e-06,
+      "loss": 1.0418,
+      "step": 7174
+    },
+    {
+      "epoch": 1.38123541160334,
+      "grad_norm": 3.2468533366440444,
+      "learning_rate": 4.611965278316221e-06,
+      "loss": 1.05,
+      "step": 7175
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "grad_norm": 3.2502978168802112,
+      "learning_rate": 4.609338925934743e-06,
+      "loss": 1.0678,
+      "step": 7176
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 1.1664,
+      "step": 7176,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 0.9134,
+      "step": 7176,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 1.6322,
+      "step": 7176,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 0.8187,
+      "step": 7176,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 0.741,
+      "step": 7176,
+      "vm_loss": 0.1233
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 1.3134,
+      "step": 7176,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 0.9905,
+      "step": 7176,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 1.3814279182809153,
+      "lm_loss": 0.8516,
+      "step": 7176,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.3816204249584907,
+      "grad_norm": 3.39920893618975,
+      "learning_rate": 4.606713097594488e-06,
+      "loss": 1.0945,
+      "step": 7177
+    },
+    {
+      "epoch": 1.381812931636066,
+      "grad_norm": 3.1672113566580884,
+      "learning_rate": 4.6040877935507245e-06,
+      "loss": 0.9937,
+      "step": 7178
+    },
+    {
+      "epoch": 1.3820054383136415,
+      "grad_norm": 3.2503382510793957,
+      "learning_rate": 4.6014630140586534e-06,
+      "loss": 1.0246,
+      "step": 7179
+    },
+    {
+      "epoch": 1.382197944991217,
+      "grad_norm": 3.216884506545426,
+      "learning_rate": 4.598838759373453e-06,
+      "loss": 0.9988,
+      "step": 7180
+    },
+    {
+      "epoch": 1.3823904516687922,
+      "grad_norm": 3.3271411537072213,
+      "learning_rate": 4.596215029750222e-06,
+      "loss": 1.0792,
+      "step": 7181
+    },
+    {
+      "epoch": 1.3825829583463676,
+      "grad_norm": 3.2775934137100413,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 1.0196,
+      "step": 7182
+    },
+    {
+      "epoch": 1.382775465023943,
+      "grad_norm": 3.128423763187306,
+      "learning_rate": 4.59096914670987e-06,
+      "loss": 0.9343,
+      "step": 7183
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "grad_norm": 3.2861258216949993,
+      "learning_rate": 4.588346993802719e-06,
+      "loss": 1.0357,
+      "step": 7184
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.7793,
+      "step": 7184,
+      "vm_loss": 0.2527
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.8046,
+      "step": 7184,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 1.1453,
+      "step": 7184,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.3852,
+      "step": 7184,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.875,
+      "step": 7184,
+      "vm_loss": 0.1397
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.6063,
+      "step": 7184,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.7424,
+      "step": 7184,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.3829679717015184,
+      "lm_loss": 0.6702,
+      "step": 7184,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.3831604783790938,
+      "grad_norm": 3.3159232427173193,
+      "learning_rate": 4.5857253669774725e-06,
+      "loss": 1.0276,
+      "step": 7185
+    },
+    {
+      "epoch": 1.383352985056669,
+      "grad_norm": 3.171683064633105,
+      "learning_rate": 4.583104266488989e-06,
+      "loss": 0.9991,
+      "step": 7186
+    },
+    {
+      "epoch": 1.3835454917342445,
+      "grad_norm": 3.307096398905604,
+      "learning_rate": 4.580483692592074e-06,
+      "loss": 1.0719,
+      "step": 7187
+    },
+    {
+      "epoch": 1.3837379984118199,
+      "grad_norm": 3.199174668208879,
+      "learning_rate": 4.577863645541483e-06,
+      "loss": 1.0335,
+      "step": 7188
+    },
+    {
+      "epoch": 1.3839305050893953,
+      "grad_norm": 3.301151267278536,
+      "learning_rate": 4.575244125591906e-06,
+      "loss": 1.019,
+      "step": 7189
+    },
+    {
+      "epoch": 1.3841230117669707,
+      "grad_norm": 3.3132592548908075,
+      "learning_rate": 4.572625132998013e-06,
+      "loss": 1.0205,
+      "step": 7190
+    },
+    {
+      "epoch": 1.384315518444546,
+      "grad_norm": 3.294109865959035,
+      "learning_rate": 4.570006668014388e-06,
+      "loss": 1.0324,
+      "step": 7191
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "grad_norm": 3.2962832572059053,
+      "learning_rate": 4.567388730895586e-06,
+      "loss": 1.0754,
+      "step": 7192
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 1.0252,
+      "step": 7192,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 0.9189,
+      "step": 7192,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 0.5341,
+      "step": 7192,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 1.2423,
+      "step": 7192,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 1.3822,
+      "step": 7192,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 1.5858,
+      "step": 7192,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 0.5449,
+      "step": 7192,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.3845080251221213,
+      "lm_loss": 0.9615,
+      "step": 7192,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.3847005317996968,
+      "grad_norm": 3.2869732233550604,
+      "learning_rate": 4.564771321896102e-06,
+      "loss": 1.0617,
+      "step": 7193
+    },
+    {
+      "epoch": 1.3848930384772722,
+      "grad_norm": 3.1752680099543724,
+      "learning_rate": 4.562154441270383e-06,
+      "loss": 1.025,
+      "step": 7194
+    },
+    {
+      "epoch": 1.3850855451548476,
+      "grad_norm": 3.9682637690835856,
+      "learning_rate": 4.559538089272825e-06,
+      "loss": 1.1528,
+      "step": 7195
+    },
+    {
+      "epoch": 1.3852780518324228,
+      "grad_norm": 3.3920053084676827,
+      "learning_rate": 4.556922266157762e-06,
+      "loss": 1.0521,
+      "step": 7196
+    },
+    {
+      "epoch": 1.3854705585099985,
+      "grad_norm": 3.1967024525321968,
+      "learning_rate": 4.5543069721795e-06,
+      "loss": 1.0219,
+      "step": 7197
+    },
+    {
+      "epoch": 1.3856630651875737,
+      "grad_norm": 3.427403027422847,
+      "learning_rate": 4.551692207592265e-06,
+      "loss": 1.0914,
+      "step": 7198
+    },
+    {
+      "epoch": 1.385855571865149,
+      "grad_norm": 3.147041901278296,
+      "learning_rate": 4.549077972650257e-06,
+      "loss": 0.9766,
+      "step": 7199
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "grad_norm": 3.193079541884317,
+      "learning_rate": 4.546464267607601e-06,
+      "loss": 1.0413,
+      "step": 7200
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.5371,
+      "step": 7200,
+      "vm_loss": 0.2162
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.9204,
+      "step": 7200,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.8246,
+      "step": 7200,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.8937,
+      "step": 7200,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 1.0023,
+      "step": 7200,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.6684,
+      "step": 7200,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.8016,
+      "step": 7200,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.3860480785427245,
+      "lm_loss": 0.7231,
+      "step": 7200,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 1.3862405852202997,
+      "grad_norm": 3.396665186516052,
+      "learning_rate": 4.5438510927183964e-06,
+      "loss": 1.0713,
+      "step": 7201
+    },
+    {
+      "epoch": 1.3864330918978753,
+      "grad_norm": 3.27578743961322,
+      "learning_rate": 4.5412384482366665e-06,
+      "loss": 1.0203,
+      "step": 7202
+    },
+    {
+      "epoch": 1.3866255985754505,
+      "grad_norm": 3.359348201211265,
+      "learning_rate": 4.538626334416396e-06,
+      "loss": 1.0193,
+      "step": 7203
+    },
+    {
+      "epoch": 1.386818105253026,
+      "grad_norm": 3.2376379344104627,
+      "learning_rate": 4.536014751511518e-06,
+      "loss": 0.9915,
+      "step": 7204
+    },
+    {
+      "epoch": 1.3870106119306014,
+      "grad_norm": 3.23029642648798,
+      "learning_rate": 4.533403699775916e-06,
+      "loss": 1.0353,
+      "step": 7205
+    },
+    {
+      "epoch": 1.3872031186081768,
+      "grad_norm": 3.383302059490082,
+      "learning_rate": 4.530793179463407e-06,
+      "loss": 1.1001,
+      "step": 7206
+    },
+    {
+      "epoch": 1.3873956252857522,
+      "grad_norm": 3.371371020774235,
+      "learning_rate": 4.528183190827773e-06,
+      "loss": 1.0552,
+      "step": 7207
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "grad_norm": 3.285004809659229,
+      "learning_rate": 4.525573734122738e-06,
+      "loss": 1.0414,
+      "step": 7208
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.7833,
+      "step": 7208,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 1.4645,
+      "step": 7208,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.9078,
+      "step": 7208,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.81,
+      "step": 7208,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.6973,
+      "step": 7208,
+      "vm_loss": 0.1416
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 1.0551,
+      "step": 7208,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.5329,
+      "step": 7208,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 1.3875881319633274,
+      "lm_loss": 0.6353,
+      "step": 7208,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 1.3877806386409028,
+      "grad_norm": 3.29047756916702,
+      "learning_rate": 4.522964809601978e-06,
+      "loss": 1.0515,
+      "step": 7209
+    },
+    {
+      "epoch": 1.3879731453184783,
+      "grad_norm": 3.327147942865889,
+      "learning_rate": 4.520356417519104e-06,
+      "loss": 1.0796,
+      "step": 7210
+    },
+    {
+      "epoch": 1.3881656519960537,
+      "grad_norm": 3.4521941405127112,
+      "learning_rate": 4.517748558127697e-06,
+      "loss": 1.0523,
+      "step": 7211
+    },
+    {
+      "epoch": 1.388358158673629,
+      "grad_norm": 3.3160959637144964,
+      "learning_rate": 4.5151412316812656e-06,
+      "loss": 1.0242,
+      "step": 7212
+    },
+    {
+      "epoch": 1.3885506653512043,
+      "grad_norm": 3.1856021908245724,
+      "learning_rate": 4.512534438433278e-06,
+      "loss": 1.0152,
+      "step": 7213
+    },
+    {
+      "epoch": 1.3887431720287797,
+      "grad_norm": 3.2348098983557727,
+      "learning_rate": 4.509928178637149e-06,
+      "loss": 1.0597,
+      "step": 7214
+    },
+    {
+      "epoch": 1.3889356787063551,
+      "grad_norm": 3.3849927104429693,
+      "learning_rate": 4.5073224525462386e-06,
+      "loss": 1.1143,
+      "step": 7215
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "grad_norm": 3.2040462044368385,
+      "learning_rate": 4.504717260413863e-06,
+      "loss": 1.0179,
+      "step": 7216
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.9291,
+      "step": 7216,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.6274,
+      "step": 7216,
+      "vm_loss": 0.0889
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.9705,
+      "step": 7216,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.902,
+      "step": 7216,
+      "vm_loss": 0.1286
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.6908,
+      "step": 7216,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 0.9793,
+      "step": 7216,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 1.2459,
+      "step": 7216,
+      "vm_loss": 0.2207
+    },
+    {
+      "epoch": 1.3891281853839306,
+      "lm_loss": 1.0427,
+      "step": 7216,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.389320692061506,
+      "grad_norm": 3.141025219087616,
+      "learning_rate": 4.502112602493266e-06,
+      "loss": 0.9903,
+      "step": 7217
+    },
+    {
+      "epoch": 1.3895131987390812,
+      "grad_norm": 3.2037202653091104,
+      "learning_rate": 4.49950847903767e-06,
+      "loss": 1.014,
+      "step": 7218
+    },
+    {
+      "epoch": 1.3897057054166566,
+      "grad_norm": 3.318574421679347,
+      "learning_rate": 4.496904890300219e-06,
+      "loss": 1.0174,
+      "step": 7219
+    },
+    {
+      "epoch": 1.389898212094232,
+      "grad_norm": 3.235564291228873,
+      "learning_rate": 4.494301836534016e-06,
+      "loss": 1.0587,
+      "step": 7220
+    },
+    {
+      "epoch": 1.3900907187718075,
+      "grad_norm": 3.2361534504003164,
+      "learning_rate": 4.491699317992114e-06,
+      "loss": 1.0199,
+      "step": 7221
+    },
+    {
+      "epoch": 1.3902832254493829,
+      "grad_norm": 3.4226133330254522,
+      "learning_rate": 4.489097334927514e-06,
+      "loss": 1.1493,
+      "step": 7222
+    },
+    {
+      "epoch": 1.390475732126958,
+      "grad_norm": 3.3597589452218757,
+      "learning_rate": 4.4864958875931545e-06,
+      "loss": 1.0947,
+      "step": 7223
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "grad_norm": 3.219695591833373,
+      "learning_rate": 4.483894976241933e-06,
+      "loss": 1.0185,
+      "step": 7224
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 1.2147,
+      "step": 7224,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 1.1027,
+      "step": 7224,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 1.3044,
+      "step": 7224,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 1.2123,
+      "step": 7224,
+      "vm_loss": 0.19
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 0.8591,
+      "step": 7224,
+      "vm_loss": 0.2573
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 0.8682,
+      "step": 7224,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 0.5287,
+      "step": 7224,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.3906682388045335,
+      "lm_loss": 1.0067,
+      "step": 7224,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.390860745482109,
+      "grad_norm": 3.2375185692310846,
+      "learning_rate": 4.48129460112669e-06,
+      "loss": 1.0076,
+      "step": 7225
+    },
+    {
+      "epoch": 1.3910532521596843,
+      "grad_norm": 3.411162459282257,
+      "learning_rate": 4.478694762500222e-06,
+      "loss": 1.0823,
+      "step": 7226
+    },
+    {
+      "epoch": 1.3912457588372598,
+      "grad_norm": 3.238380728340277,
+      "learning_rate": 4.476095460615253e-06,
+      "loss": 0.9842,
+      "step": 7227
+    },
+    {
+      "epoch": 1.391438265514835,
+      "grad_norm": 3.252013012438708,
+      "learning_rate": 4.473496695724484e-06,
+      "loss": 1.0752,
+      "step": 7228
+    },
+    {
+      "epoch": 1.3916307721924104,
+      "grad_norm": 3.2129535095421287,
+      "learning_rate": 4.470898468080538e-06,
+      "loss": 0.9922,
+      "step": 7229
+    },
+    {
+      "epoch": 1.3918232788699858,
+      "grad_norm": 3.25915344119369,
+      "learning_rate": 4.4683007779359975e-06,
+      "loss": 1.0342,
+      "step": 7230
+    },
+    {
+      "epoch": 1.3920157855475612,
+      "grad_norm": 3.187012010329501,
+      "learning_rate": 4.465703625543393e-06,
+      "loss": 0.9991,
+      "step": 7231
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "grad_norm": 3.14853267219442,
+      "learning_rate": 4.463107011155204e-06,
+      "loss": 0.979,
+      "step": 7232
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 1.2003,
+      "step": 7232,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 0.7273,
+      "step": 7232,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 0.9422,
+      "step": 7232,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 1.155,
+      "step": 7232,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 0.8178,
+      "step": 7232,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 0.8455,
+      "step": 7232,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 0.6767,
+      "step": 7232,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 1.3922082922251366,
+      "lm_loss": 1.0292,
+      "step": 7232,
+      "vm_loss": 0.1062
+    },
+    {
+      "epoch": 1.3924007989027118,
+      "grad_norm": 3.137616420048383,
+      "learning_rate": 4.460510935023848e-06,
+      "loss": 0.9778,
+      "step": 7233
+    },
+    {
+      "epoch": 1.3925933055802873,
+      "grad_norm": 3.1467254630452586,
+      "learning_rate": 4.4579153974016986e-06,
+      "loss": 1.0096,
+      "step": 7234
+    },
+    {
+      "epoch": 1.3927858122578627,
+      "grad_norm": 3.0698213665525205,
+      "learning_rate": 4.455320398541078e-06,
+      "loss": 0.9647,
+      "step": 7235
+    },
+    {
+      "epoch": 1.392978318935438,
+      "grad_norm": 3.3548613982649824,
+      "learning_rate": 4.4527259386942515e-06,
+      "loss": 0.9872,
+      "step": 7236
+    },
+    {
+      "epoch": 1.3931708256130135,
+      "grad_norm": 3.439315685990862,
+      "learning_rate": 4.450132018113434e-06,
+      "loss": 1.0785,
+      "step": 7237
+    },
+    {
+      "epoch": 1.3933633322905887,
+      "grad_norm": 3.220278425434203,
+      "learning_rate": 4.447538637050787e-06,
+      "loss": 0.9896,
+      "step": 7238
+    },
+    {
+      "epoch": 1.3935558389681642,
+      "grad_norm": 3.3971823209419956,
+      "learning_rate": 4.4449457957584255e-06,
+      "loss": 1.0137,
+      "step": 7239
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "grad_norm": 3.4165500885908857,
+      "learning_rate": 4.4423534944883984e-06,
+      "loss": 1.0091,
+      "step": 7240
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.6314,
+      "step": 7240,
+      "vm_loss": 0.1298
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.9667,
+      "step": 7240,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 1.0063,
+      "step": 7240,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.7289,
+      "step": 7240,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.5011,
+      "step": 7240,
+      "vm_loss": 0.1464
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 1.1109,
+      "step": 7240,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.7889,
+      "step": 7240,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 1.3937483456457396,
+      "lm_loss": 0.9704,
+      "step": 7240,
+      "vm_loss": 0.1903
+    },
+    {
+      "epoch": 1.393940852323315,
+      "grad_norm": 3.448420545946613,
+      "learning_rate": 4.439761733492716e-06,
+      "loss": 1.0251,
+      "step": 7241
+    },
+    {
+      "epoch": 1.3941333590008904,
+      "grad_norm": 3.1351231087001294,
+      "learning_rate": 4.437170513023328e-06,
+      "loss": 0.9533,
+      "step": 7242
+    },
+    {
+      "epoch": 1.3943258656784656,
+      "grad_norm": 3.236364256875787,
+      "learning_rate": 4.43457983333214e-06,
+      "loss": 1.0286,
+      "step": 7243
+    },
+    {
+      "epoch": 1.394518372356041,
+      "grad_norm": 3.3104200425336785,
+      "learning_rate": 4.4319896946709875e-06,
+      "loss": 1.0756,
+      "step": 7244
+    },
+    {
+      "epoch": 1.3947108790336165,
+      "grad_norm": 3.2787601550931895,
+      "learning_rate": 4.4294000972916785e-06,
+      "loss": 1.0304,
+      "step": 7245
+    },
+    {
+      "epoch": 1.3949033857111919,
+      "grad_norm": 3.1918141296807145,
+      "learning_rate": 4.426811041445946e-06,
+      "loss": 1.0365,
+      "step": 7246
+    },
+    {
+      "epoch": 1.3950958923887673,
+      "grad_norm": 3.25795711637226,
+      "learning_rate": 4.424222527385482e-06,
+      "loss": 1.0663,
+      "step": 7247
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "grad_norm": 3.3175391760229354,
+      "learning_rate": 4.421634555361923e-06,
+      "loss": 1.0736,
+      "step": 7248
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.8402,
+      "step": 7248,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.817,
+      "step": 7248,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.6167,
+      "step": 7248,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.8723,
+      "step": 7248,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.5982,
+      "step": 7248,
+      "vm_loss": 0.1037
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.6458,
+      "step": 7248,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 1.2121,
+      "step": 7248,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.3952883990663425,
+      "lm_loss": 0.822,
+      "step": 7248,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.395480905743918,
+      "grad_norm": 3.2214795420753637,
+      "learning_rate": 4.419047125626859e-06,
+      "loss": 1.0083,
+      "step": 7249
+    },
+    {
+      "epoch": 1.3956734124214933,
+      "grad_norm": 3.326936723226034,
+      "learning_rate": 4.416460238431809e-06,
+      "loss": 1.0895,
+      "step": 7250
+    },
+    {
+      "epoch": 1.3958659190990688,
+      "grad_norm": 3.0973144320397346,
+      "learning_rate": 4.41387389402826e-06,
+      "loss": 0.9925,
+      "step": 7251
+    },
+    {
+      "epoch": 1.3960584257766442,
+      "grad_norm": 3.159966760526034,
+      "learning_rate": 4.411288092667636e-06,
+      "loss": 1.0237,
+      "step": 7252
+    },
+    {
+      "epoch": 1.3962509324542194,
+      "grad_norm": 3.1974776230684165,
+      "learning_rate": 4.408702834601309e-06,
+      "loss": 1.0049,
+      "step": 7253
+    },
+    {
+      "epoch": 1.3964434391317948,
+      "grad_norm": 3.126783721051538,
+      "learning_rate": 4.4061181200806e-06,
+      "loss": 0.9581,
+      "step": 7254
+    },
+    {
+      "epoch": 1.3966359458093702,
+      "grad_norm": 3.2599328173979245,
+      "learning_rate": 4.403533949356781e-06,
+      "loss": 1.0062,
+      "step": 7255
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "grad_norm": 3.3935221452812177,
+      "learning_rate": 4.4009503226810576e-06,
+      "loss": 1.0422,
+      "step": 7256
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 1.0591,
+      "step": 7256,
+      "vm_loss": 0.1384
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.6725,
+      "step": 7256,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.4944,
+      "step": 7256,
+      "vm_loss": 0.126
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.7794,
+      "step": 7256,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.8425,
+      "step": 7256,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.7313,
+      "step": 7256,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.6278,
+      "step": 7256,
+      "vm_loss": 0.201
+    },
+    {
+      "epoch": 1.3968284524869456,
+      "lm_loss": 0.9993,
+      "step": 7256,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.397020959164521,
+      "grad_norm": 3.241265329943556,
+      "learning_rate": 4.3983672403045954e-06,
+      "loss": 1.0021,
+      "step": 7257
+    },
+    {
+      "epoch": 1.3972134658420963,
+      "grad_norm": 3.1852523397834274,
+      "learning_rate": 4.395784702478504e-06,
+      "loss": 1.0148,
+      "step": 7258
+    },
+    {
+      "epoch": 1.397405972519672,
+      "grad_norm": 3.300321343363957,
+      "learning_rate": 4.393202709453838e-06,
+      "loss": 1.024,
+      "step": 7259
+    },
+    {
+      "epoch": 1.3975984791972471,
+      "grad_norm": 3.2757612445628346,
+      "learning_rate": 4.390621261481603e-06,
+      "loss": 1.0259,
+      "step": 7260
+    },
+    {
+      "epoch": 1.3977909858748225,
+      "grad_norm": 3.4123510058924746,
+      "learning_rate": 4.388040358812741e-06,
+      "loss": 1.014,
+      "step": 7261
+    },
+    {
+      "epoch": 1.397983492552398,
+      "grad_norm": 3.1793761199222534,
+      "learning_rate": 4.38546000169816e-06,
+      "loss": 0.9836,
+      "step": 7262
+    },
+    {
+      "epoch": 1.3981759992299732,
+      "grad_norm": 3.2269409992976286,
+      "learning_rate": 4.3828801903886944e-06,
+      "loss": 0.9891,
+      "step": 7263
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "grad_norm": 3.2880405892395603,
+      "learning_rate": 4.380300925135138e-06,
+      "loss": 1.0335,
+      "step": 7264
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.7709,
+      "step": 7264,
+      "vm_loss": 0.1326
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.8882,
+      "step": 7264,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.7829,
+      "step": 7264,
+      "vm_loss": 0.1472
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.8806,
+      "step": 7264,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.8995,
+      "step": 7264,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.5841,
+      "step": 7264,
+      "vm_loss": 0.0822
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 0.879,
+      "step": 7264,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 1.3983685059075488,
+      "lm_loss": 1.0041,
+      "step": 7264,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.398561012585124,
+      "grad_norm": 3.3126867405250153,
+      "learning_rate": 4.377722206188229e-06,
+      "loss": 1.0654,
+      "step": 7265
+    },
+    {
+      "epoch": 1.3987535192626994,
+      "grad_norm": 3.2267603303595034,
+      "learning_rate": 4.375144033798655e-06,
+      "loss": 0.9996,
+      "step": 7266
+    },
+    {
+      "epoch": 1.3989460259402748,
+      "grad_norm": 3.411733525371951,
+      "learning_rate": 4.372566408217041e-06,
+      "loss": 1.0859,
+      "step": 7267
+    },
+    {
+      "epoch": 1.3991385326178503,
+      "grad_norm": 3.26290067537603,
+      "learning_rate": 4.369989329693967e-06,
+      "loss": 1.0456,
+      "step": 7268
+    },
+    {
+      "epoch": 1.3993310392954257,
+      "grad_norm": 3.3516593555548324,
+      "learning_rate": 4.367412798479959e-06,
+      "loss": 1.0432,
+      "step": 7269
+    },
+    {
+      "epoch": 1.3995235459730009,
+      "grad_norm": 3.223393793471439,
+      "learning_rate": 4.364836814825493e-06,
+      "loss": 0.997,
+      "step": 7270
+    },
+    {
+      "epoch": 1.3997160526505763,
+      "grad_norm": 3.1981313427234417,
+      "learning_rate": 4.362261378980977e-06,
+      "loss": 1.0133,
+      "step": 7271
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "grad_norm": 3.154688550220936,
+      "learning_rate": 4.359686491196789e-06,
+      "loss": 1.0019,
+      "step": 7272
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.9545,
+      "step": 7272,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.7766,
+      "step": 7272,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.8152,
+      "step": 7272,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.6654,
+      "step": 7272,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.6261,
+      "step": 7272,
+      "vm_loss": 0.1542
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.9085,
+      "step": 7272,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 0.3901,
+      "step": 7272,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 1.3999085593281517,
+      "lm_loss": 1.1798,
+      "step": 7272,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.4001010660057271,
+      "grad_norm": 3.329802661399455,
+      "learning_rate": 4.35711215172323e-06,
+      "loss": 0.9845,
+      "step": 7273
+    },
+    {
+      "epoch": 1.4002935726833026,
+      "grad_norm": 3.339628740789547,
+      "learning_rate": 4.354538360810564e-06,
+      "loss": 1.0145,
+      "step": 7274
+    },
+    {
+      "epoch": 1.4004860793608778,
+      "grad_norm": 3.5029633162411025,
+      "learning_rate": 4.351965118708996e-06,
+      "loss": 1.0453,
+      "step": 7275
+    },
+    {
+      "epoch": 1.4006785860384532,
+      "grad_norm": 3.546755808969498,
+      "learning_rate": 4.349392425668683e-06,
+      "loss": 1.1019,
+      "step": 7276
+    },
+    {
+      "epoch": 1.4008710927160286,
+      "grad_norm": 3.305754008367797,
+      "learning_rate": 4.346820281939713e-06,
+      "loss": 0.9926,
+      "step": 7277
+    },
+    {
+      "epoch": 1.401063599393604,
+      "grad_norm": 3.4479443017814604,
+      "learning_rate": 4.344248687772138e-06,
+      "loss": 1.0128,
+      "step": 7278
+    },
+    {
+      "epoch": 1.4012561060711795,
+      "grad_norm": 3.196193341633772,
+      "learning_rate": 4.341677643415949e-06,
+      "loss": 0.9909,
+      "step": 7279
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "grad_norm": 3.163949879077566,
+      "learning_rate": 4.3391071491210834e-06,
+      "loss": 0.9313,
+      "step": 7280
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 1.121,
+      "step": 7280,
+      "vm_loss": 0.2672
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 1.1644,
+      "step": 7280,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 1.225,
+      "step": 7280,
+      "vm_loss": 0.1143
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 0.8647,
+      "step": 7280,
+      "vm_loss": 0.2452
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 1.1583,
+      "step": 7280,
+      "vm_loss": 0.2238
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 0.6586,
+      "step": 7280,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 1.2907,
+      "step": 7280,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 1.4014486127487547,
+      "lm_loss": 0.8495,
+      "step": 7280,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.40164111942633,
+      "grad_norm": 3.182995788297079,
+      "learning_rate": 4.336537205137431e-06,
+      "loss": 1.0415,
+      "step": 7281
+    },
+    {
+      "epoch": 1.4018336261039055,
+      "grad_norm": 3.136485722934549,
+      "learning_rate": 4.333967811714811e-06,
+      "loss": 0.9886,
+      "step": 7282
+    },
+    {
+      "epoch": 1.402026132781481,
+      "grad_norm": 3.2764522996051997,
+      "learning_rate": 4.3313989691030175e-06,
+      "loss": 0.9838,
+      "step": 7283
+    },
+    {
+      "epoch": 1.4022186394590563,
+      "grad_norm": 3.1908097624066807,
+      "learning_rate": 4.328830677551763e-06,
+      "loss": 1.0193,
+      "step": 7284
+    },
+    {
+      "epoch": 1.4024111461366315,
+      "grad_norm": 3.281168306914215,
+      "learning_rate": 4.3262629373107214e-06,
+      "loss": 1.081,
+      "step": 7285
+    },
+    {
+      "epoch": 1.402603652814207,
+      "grad_norm": 3.2179464085052722,
+      "learning_rate": 4.3236957486295115e-06,
+      "loss": 1.0098,
+      "step": 7286
+    },
+    {
+      "epoch": 1.4027961594917824,
+      "grad_norm": 3.2386755498098285,
+      "learning_rate": 4.321129111757699e-06,
+      "loss": 1.0356,
+      "step": 7287
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "grad_norm": 3.31189265735079,
+      "learning_rate": 4.318563026944784e-06,
+      "loss": 1.046,
+      "step": 7288
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.8486,
+      "step": 7288,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.7606,
+      "step": 7288,
+      "vm_loss": 0.2218
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.573,
+      "step": 7288,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 1.015,
+      "step": 7288,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.9283,
+      "step": 7288,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.7314,
+      "step": 7288,
+      "vm_loss": 0.1671
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.6255,
+      "step": 7288,
+      "vm_loss": 0.165
+    },
+    {
+      "epoch": 1.4029886661693578,
+      "lm_loss": 0.7905,
+      "step": 7288,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 1.4031811728469332,
+      "grad_norm": 3.335457634076318,
+      "learning_rate": 4.315997494440236e-06,
+      "loss": 1.0416,
+      "step": 7289
+    },
+    {
+      "epoch": 1.4033736795245084,
+      "grad_norm": 3.258655833881981,
+      "learning_rate": 4.3134325144934484e-06,
+      "loss": 1.0135,
+      "step": 7290
+    },
+    {
+      "epoch": 1.4035661862020838,
+      "grad_norm": 3.2833775289960925,
+      "learning_rate": 4.3108680873537765e-06,
+      "loss": 1.0548,
+      "step": 7291
+    },
+    {
+      "epoch": 1.4037586928796593,
+      "grad_norm": 3.2690774628374206,
+      "learning_rate": 4.308304213270504e-06,
+      "loss": 1.0575,
+      "step": 7292
+    },
+    {
+      "epoch": 1.4039511995572347,
+      "grad_norm": 3.0720614054330238,
+      "learning_rate": 4.3057408924928874e-06,
+      "loss": 0.9774,
+      "step": 7293
+    },
+    {
+      "epoch": 1.40414370623481,
+      "grad_norm": 3.320562023041671,
+      "learning_rate": 4.303178125270104e-06,
+      "loss": 1.0264,
+      "step": 7294
+    },
+    {
+      "epoch": 1.4043362129123853,
+      "grad_norm": 3.211414790894887,
+      "learning_rate": 4.30061591185129e-06,
+      "loss": 1.0217,
+      "step": 7295
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "grad_norm": 3.15356726324523,
+      "learning_rate": 4.298054252485525e-06,
+      "loss": 0.9973,
+      "step": 7296
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.7628,
+      "step": 7296,
+      "vm_loss": 0.2266
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.8077,
+      "step": 7296,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.7567,
+      "step": 7296,
+      "vm_loss": 0.2034
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.8168,
+      "step": 7296,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.8702,
+      "step": 7296,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.9975,
+      "step": 7296,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 0.8532,
+      "step": 7296,
+      "vm_loss": 0.1811
+    },
+    {
+      "epoch": 1.4045287195899607,
+      "lm_loss": 1.1108,
+      "step": 7296,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 1.4047212262675361,
+      "grad_norm": 3.170620425114536,
+      "learning_rate": 4.295493147421836e-06,
+      "loss": 1.0224,
+      "step": 7297
+    },
+    {
+      "epoch": 1.4049137329451116,
+      "grad_norm": 3.309635710505961,
+      "learning_rate": 4.292932596909199e-06,
+      "loss": 1.0348,
+      "step": 7298
+    },
+    {
+      "epoch": 1.405106239622687,
+      "grad_norm": 3.2120827317240046,
+      "learning_rate": 4.2903726011965265e-06,
+      "loss": 0.9676,
+      "step": 7299
+    },
+    {
+      "epoch": 1.4052987463002622,
+      "grad_norm": 3.1528441126651416,
+      "learning_rate": 4.287813160532682e-06,
+      "loss": 0.9883,
+      "step": 7300
+    },
+    {
+      "epoch": 1.4054912529778376,
+      "grad_norm": 3.3699919996954524,
+      "learning_rate": 4.28525427516648e-06,
+      "loss": 1.06,
+      "step": 7301
+    },
+    {
+      "epoch": 1.405683759655413,
+      "grad_norm": 3.213543604286197,
+      "learning_rate": 4.282695945346674e-06,
+      "loss": 1.0053,
+      "step": 7302
+    },
+    {
+      "epoch": 1.4058762663329885,
+      "grad_norm": 3.4365591393205195,
+      "learning_rate": 4.2801381713219695e-06,
+      "loss": 1.0726,
+      "step": 7303
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "grad_norm": 3.322222742331046,
+      "learning_rate": 4.277580953341015e-06,
+      "loss": 1.0318,
+      "step": 7304
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.5521,
+      "step": 7304,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.4944,
+      "step": 7304,
+      "vm_loss": 0.2401
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.6581,
+      "step": 7304,
+      "vm_loss": 0.114
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.8147,
+      "step": 7304,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.8176,
+      "step": 7304,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 1.498,
+      "step": 7304,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.6341,
+      "step": 7304,
+      "vm_loss": 0.1062
+    },
+    {
+      "epoch": 1.4060687730105639,
+      "lm_loss": 0.7056,
+      "step": 7304,
+      "vm_loss": 0.1837
+    },
+    {
+      "epoch": 1.406261279688139,
+      "grad_norm": 3.1682524016922904,
+      "learning_rate": 4.2750242916523965e-06,
+      "loss": 0.9833,
+      "step": 7305
+    },
+    {
+      "epoch": 1.4064537863657145,
+      "grad_norm": 3.395791082098268,
+      "learning_rate": 4.272468186504668e-06,
+      "loss": 1.0588,
+      "step": 7306
+    },
+    {
+      "epoch": 1.40664629304329,
+      "grad_norm": 3.4898333003568442,
+      "learning_rate": 4.269912638146305e-06,
+      "loss": 1.1293,
+      "step": 7307
+    },
+    {
+      "epoch": 1.4068387997208653,
+      "grad_norm": 3.27212011522272,
+      "learning_rate": 4.267357646825746e-06,
+      "loss": 1.0164,
+      "step": 7308
+    },
+    {
+      "epoch": 1.4070313063984408,
+      "grad_norm": 3.2014180161296957,
+      "learning_rate": 4.264803212791359e-06,
+      "loss": 1.0056,
+      "step": 7309
+    },
+    {
+      "epoch": 1.407223813076016,
+      "grad_norm": 3.1555948868316146,
+      "learning_rate": 4.262249336291481e-06,
+      "loss": 1.0083,
+      "step": 7310
+    },
+    {
+      "epoch": 1.4074163197535914,
+      "grad_norm": 3.243827283181842,
+      "learning_rate": 4.259696017574371e-06,
+      "loss": 0.9986,
+      "step": 7311
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "grad_norm": 3.098891302867036,
+      "learning_rate": 4.257143256888248e-06,
+      "loss": 0.9651,
+      "step": 7312
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.4941,
+      "step": 7312,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.9744,
+      "step": 7312,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.9488,
+      "step": 7312,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.9772,
+      "step": 7312,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 1.1317,
+      "step": 7312,
+      "vm_loss": 0.1047
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.9345,
+      "step": 7312,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 1.1215,
+      "step": 7312,
+      "vm_loss": 0.1017
+    },
+    {
+      "epoch": 1.4076088264311668,
+      "lm_loss": 0.8664,
+      "step": 7312,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.4078013331087422,
+      "grad_norm": 3.23687498459985,
+      "learning_rate": 4.254591054481274e-06,
+      "loss": 1.0063,
+      "step": 7313
+    },
+    {
+      "epoch": 1.4079938397863176,
+      "grad_norm": 3.2610700527557173,
+      "learning_rate": 4.2520394106015574e-06,
+      "loss": 1.0057,
+      "step": 7314
+    },
+    {
+      "epoch": 1.4081863464638928,
+      "grad_norm": 3.3554675937861194,
+      "learning_rate": 4.2494883254971444e-06,
+      "loss": 1.0421,
+      "step": 7315
+    },
+    {
+      "epoch": 1.4083788531414683,
+      "grad_norm": 3.393335455574113,
+      "learning_rate": 4.246937799416037e-06,
+      "loss": 0.9989,
+      "step": 7316
+    },
+    {
+      "epoch": 1.4085713598190437,
+      "grad_norm": 3.379727730482624,
+      "learning_rate": 4.244387832606179e-06,
+      "loss": 1.0274,
+      "step": 7317
+    },
+    {
+      "epoch": 1.408763866496619,
+      "grad_norm": 3.3099376666367046,
+      "learning_rate": 4.2418384253154585e-06,
+      "loss": 0.9735,
+      "step": 7318
+    },
+    {
+      "epoch": 1.4089563731741945,
+      "grad_norm": 3.436342655011878,
+      "learning_rate": 4.239289577791712e-06,
+      "loss": 1.0782,
+      "step": 7319
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "grad_norm": 3.2924206949159482,
+      "learning_rate": 4.2367412902827185e-06,
+      "loss": 1.0684,
+      "step": 7320
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.836,
+      "step": 7320,
+      "vm_loss": 0.1112
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.4804,
+      "step": 7320,
+      "vm_loss": 0.1989
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 1.0266,
+      "step": 7320,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.7001,
+      "step": 7320,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.5321,
+      "step": 7320,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.9073,
+      "step": 7320,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.7403,
+      "step": 7320,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 1.4091488798517697,
+      "lm_loss": 0.4705,
+      "step": 7320,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 1.4093413865293454,
+      "grad_norm": 3.0987218544910435,
+      "learning_rate": 4.234193563036211e-06,
+      "loss": 0.9552,
+      "step": 7321
+    },
+    {
+      "epoch": 1.4095338932069206,
+      "grad_norm": 3.1630195661630913,
+      "learning_rate": 4.23164639629985e-06,
+      "loss": 0.952,
+      "step": 7322
+    },
+    {
+      "epoch": 1.409726399884496,
+      "grad_norm": 3.100224102880909,
+      "learning_rate": 4.229099790321259e-06,
+      "loss": 0.9906,
+      "step": 7323
+    },
+    {
+      "epoch": 1.4099189065620714,
+      "grad_norm": 3.2992946372833343,
+      "learning_rate": 4.226553745347999e-06,
+      "loss": 1.0069,
+      "step": 7324
+    },
+    {
+      "epoch": 1.4101114132396466,
+      "grad_norm": 3.2068404973815348,
+      "learning_rate": 4.224008261627583e-06,
+      "loss": 1.0178,
+      "step": 7325
+    },
+    {
+      "epoch": 1.4103039199172223,
+      "grad_norm": 3.359014259833218,
+      "learning_rate": 4.221463339407454e-06,
+      "loss": 1.0193,
+      "step": 7326
+    },
+    {
+      "epoch": 1.4104964265947975,
+      "grad_norm": 3.358961999615674,
+      "learning_rate": 4.218918978935022e-06,
+      "loss": 0.994,
+      "step": 7327
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "grad_norm": 3.3486145025532803,
+      "learning_rate": 4.216375180457626e-06,
+      "loss": 1.0524,
+      "step": 7328
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.9263,
+      "step": 7328,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.5902,
+      "step": 7328,
+      "vm_loss": 0.1169
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.4089,
+      "step": 7328,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.9828,
+      "step": 7328,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.6561,
+      "step": 7328,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.5534,
+      "step": 7328,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.3088,
+      "step": 7328,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.4106889332723729,
+      "lm_loss": 0.545,
+      "step": 7328,
+      "vm_loss": 0.1097
+    },
+    {
+      "epoch": 1.4108814399499483,
+      "grad_norm": 3.165090778978362,
+      "learning_rate": 4.213831944222556e-06,
+      "loss": 0.9587,
+      "step": 7329
+    },
+    {
+      "epoch": 1.4110739466275237,
+      "grad_norm": 3.22520948138797,
+      "learning_rate": 4.211289270477047e-06,
+      "loss": 0.9913,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4112664533050991,
+      "grad_norm": 3.29524439109055,
+      "learning_rate": 4.2087471594682835e-06,
+      "loss": 1.0009,
+      "step": 7331
+    },
+    {
+      "epoch": 1.4114589599826743,
+      "grad_norm": 3.155685687100094,
+      "learning_rate": 4.206205611443386e-06,
+      "loss": 0.9773,
+      "step": 7332
+    },
+    {
+      "epoch": 1.4116514666602498,
+      "grad_norm": 3.1631723233831446,
+      "learning_rate": 4.203664626649426e-06,
+      "loss": 1.0305,
+      "step": 7333
+    },
+    {
+      "epoch": 1.4118439733378252,
+      "grad_norm": 3.263985432511573,
+      "learning_rate": 4.201124205333421e-06,
+      "loss": 1.077,
+      "step": 7334
+    },
+    {
+      "epoch": 1.4120364800154006,
+      "grad_norm": 3.3198038015606812,
+      "learning_rate": 4.198584347742337e-06,
+      "loss": 1.0166,
+      "step": 7335
+    },
+    {
+      "epoch": 1.412228986692976,
+      "grad_norm": 3.3581877619581135,
+      "learning_rate": 4.196045054123068e-06,
+      "loss": 1.0288,
+      "step": 7336
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.5751,
+      "step": 7336,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.7448,
+      "step": 7336,
+      "vm_loss": 0.1971
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.8573,
+      "step": 7336,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.7549,
+      "step": 7336,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.8223,
+      "step": 7336,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.6548,
+      "step": 7336,
+      "vm_loss": 0.1291
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 1.2283,
+      "step": 7336,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.412228986692976,
+      "lm_loss": 0.6998,
+      "step": 7336,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.4124214933705512,
+      "grad_norm": 3.0879554842673844,
+      "learning_rate": 4.193506324722481e-06,
+      "loss": 0.9324,
+      "step": 7337
+    },
+    {
+      "epoch": 1.4126140000481267,
+      "grad_norm": 3.315599772061835,
+      "learning_rate": 4.190968159787364e-06,
+      "loss": 1.0635,
+      "step": 7338
+    },
+    {
+      "epoch": 1.412806506725702,
+      "grad_norm": 3.136467155455047,
+      "learning_rate": 4.18843055956446e-06,
+      "loss": 1.0225,
+      "step": 7339
+    },
+    {
+      "epoch": 1.4129990134032775,
+      "grad_norm": 3.152344002246507,
+      "learning_rate": 4.185893524300458e-06,
+      "loss": 0.9572,
+      "step": 7340
+    },
+    {
+      "epoch": 1.413191520080853,
+      "grad_norm": 3.359363349712454,
+      "learning_rate": 4.183357054241989e-06,
+      "loss": 1.0496,
+      "step": 7341
+    },
+    {
+      "epoch": 1.4133840267584281,
+      "grad_norm": 3.413428952585428,
+      "learning_rate": 4.180821149635634e-06,
+      "loss": 1.0504,
+      "step": 7342
+    },
+    {
+      "epoch": 1.4135765334360035,
+      "grad_norm": 3.1090198648209566,
+      "learning_rate": 4.178285810727906e-06,
+      "loss": 0.9559,
+      "step": 7343
+    },
+    {
+      "epoch": 1.413769040113579,
+      "grad_norm": 3.3678127079420777,
+      "learning_rate": 4.175751037765286e-06,
+      "loss": 1.0858,
+      "step": 7344
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.5224,
+      "step": 7344,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.6963,
+      "step": 7344,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.7576,
+      "step": 7344,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.9183,
+      "step": 7344,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.8531,
+      "step": 7344,
+      "vm_loss": 0.2048
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 1.0998,
+      "step": 7344,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.5426,
+      "step": 7344,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.413769040113579,
+      "lm_loss": 0.443,
+      "step": 7344,
+      "vm_loss": 0.2192
+    },
+    {
+      "epoch": 1.4139615467911544,
+      "grad_norm": 3.084931466295408,
+      "learning_rate": 4.173216830994175e-06,
+      "loss": 0.986,
+      "step": 7345
+    },
+    {
+      "epoch": 1.4141540534687298,
+      "grad_norm": 3.1417702184589587,
+      "learning_rate": 4.170683190660934e-06,
+      "loss": 0.9746,
+      "step": 7346
+    },
+    {
+      "epoch": 1.414346560146305,
+      "grad_norm": 3.3208463954550624,
+      "learning_rate": 4.168150117011866e-06,
+      "loss": 1.1104,
+      "step": 7347
+    },
+    {
+      "epoch": 1.4145390668238804,
+      "grad_norm": 3.2567999363989935,
+      "learning_rate": 4.16561761029322e-06,
+      "loss": 1.0384,
+      "step": 7348
+    },
+    {
+      "epoch": 1.4147315735014558,
+      "grad_norm": 3.1159148533776615,
+      "learning_rate": 4.1630856707511834e-06,
+      "loss": 1.0272,
+      "step": 7349
+    },
+    {
+      "epoch": 1.4149240801790313,
+      "grad_norm": 3.171403810309266,
+      "learning_rate": 4.1605542986318955e-06,
+      "loss": 0.9976,
+      "step": 7350
+    },
+    {
+      "epoch": 1.4151165868566067,
+      "grad_norm": 3.278656649292173,
+      "learning_rate": 4.158023494181439e-06,
+      "loss": 1.0282,
+      "step": 7351
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "grad_norm": 3.3617582419369936,
+      "learning_rate": 4.155493257645842e-06,
+      "loss": 1.0848,
+      "step": 7352
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.7457,
+      "step": 7352,
+      "vm_loss": 0.1416
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 1.0191,
+      "step": 7352,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 1.3959,
+      "step": 7352,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.6692,
+      "step": 7352,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.9548,
+      "step": 7352,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.6711,
+      "step": 7352,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.8017,
+      "step": 7352,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 1.4153090935341819,
+      "lm_loss": 0.5682,
+      "step": 7352,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.4155016002117573,
+      "grad_norm": 3.296179896864732,
+      "learning_rate": 4.152963589271067e-06,
+      "loss": 1.0208,
+      "step": 7353
+    },
+    {
+      "epoch": 1.4156941068893327,
+      "grad_norm": 3.0462229680183643,
+      "learning_rate": 4.150434489303043e-06,
+      "loss": 0.9255,
+      "step": 7354
+    },
+    {
+      "epoch": 1.4158866135669081,
+      "grad_norm": 3.3604111462995485,
+      "learning_rate": 4.147905957987621e-06,
+      "loss": 1.0757,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4160791202444836,
+      "grad_norm": 3.346826409899582,
+      "learning_rate": 4.145377995570611e-06,
+      "loss": 1.0275,
+      "step": 7356
+    },
+    {
+      "epoch": 1.4162716269220588,
+      "grad_norm": 3.0982447370218082,
+      "learning_rate": 4.142850602297761e-06,
+      "loss": 0.932,
+      "step": 7357
+    },
+    {
+      "epoch": 1.4164641335996342,
+      "grad_norm": 3.2149632167768187,
+      "learning_rate": 4.140323778414772e-06,
+      "loss": 0.9685,
+      "step": 7358
+    },
+    {
+      "epoch": 1.4166566402772096,
+      "grad_norm": 3.3380226432358673,
+      "learning_rate": 4.137797524167274e-06,
+      "loss": 1.0108,
+      "step": 7359
+    },
+    {
+      "epoch": 1.416849146954785,
+      "grad_norm": 3.1673162551324245,
+      "learning_rate": 4.135271839800857e-06,
+      "loss": 0.9908,
+      "step": 7360
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 0.5839,
+      "step": 7360,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 1.2805,
+      "step": 7360,
+      "vm_loss": 0.225
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 1.0768,
+      "step": 7360,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 0.7986,
+      "step": 7360,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 1.5303,
+      "step": 7360,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 0.7462,
+      "step": 7360,
+      "vm_loss": 0.1018
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 1.0603,
+      "step": 7360,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.416849146954785,
+      "lm_loss": 0.5246,
+      "step": 7360,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 1.4170416536323605,
+      "grad_norm": 3.351910542398462,
+      "learning_rate": 4.132746725561049e-06,
+      "loss": 1.0354,
+      "step": 7361
+    },
+    {
+      "epoch": 1.4172341603099357,
+      "grad_norm": 3.1473049693140553,
+      "learning_rate": 4.130222181693323e-06,
+      "loss": 0.9857,
+      "step": 7362
+    },
+    {
+      "epoch": 1.417426666987511,
+      "grad_norm": 3.179922747859307,
+      "learning_rate": 4.127698208443097e-06,
+      "loss": 0.974,
+      "step": 7363
+    },
+    {
+      "epoch": 1.4176191736650865,
+      "grad_norm": 3.320696069472839,
+      "learning_rate": 4.125174806055735e-06,
+      "loss": 1.0389,
+      "step": 7364
+    },
+    {
+      "epoch": 1.417811680342662,
+      "grad_norm": 3.2790493766282345,
+      "learning_rate": 4.122651974776546e-06,
+      "loss": 1.0249,
+      "step": 7365
+    },
+    {
+      "epoch": 1.4180041870202373,
+      "grad_norm": 3.105329561705002,
+      "learning_rate": 4.120129714850773e-06,
+      "loss": 0.9381,
+      "step": 7366
+    },
+    {
+      "epoch": 1.4181966936978125,
+      "grad_norm": 3.0690335902779435,
+      "learning_rate": 4.11760802652362e-06,
+      "loss": 0.9104,
+      "step": 7367
+    },
+    {
+      "epoch": 1.418389200375388,
+      "grad_norm": 3.226350529488512,
+      "learning_rate": 4.115086910040222e-06,
+      "loss": 1.0282,
+      "step": 7368
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 1.6842,
+      "step": 7368,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 0.9523,
+      "step": 7368,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 0.529,
+      "step": 7368,
+      "vm_loss": 0.1143
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 0.5949,
+      "step": 7368,
+      "vm_loss": 0.1205
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 1.0147,
+      "step": 7368,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 0.9286,
+      "step": 7368,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 1.0407,
+      "step": 7368,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 1.418389200375388,
+      "lm_loss": 0.6375,
+      "step": 7368,
+      "vm_loss": 0.1422
+    },
+    {
+      "epoch": 1.4185817070529634,
+      "grad_norm": 3.4301240144104828,
+      "learning_rate": 4.112566365645671e-06,
+      "loss": 1.0072,
+      "step": 7369
+    },
+    {
+      "epoch": 1.4187742137305388,
+      "grad_norm": 3.3228689913713416,
+      "learning_rate": 4.110046393584985e-06,
+      "loss": 1.0226,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4189667204081142,
+      "grad_norm": 3.4187041171908503,
+      "learning_rate": 4.107526994103151e-06,
+      "loss": 1.069,
+      "step": 7371
+    },
+    {
+      "epoch": 1.4191592270856894,
+      "grad_norm": 3.19833901106283,
+      "learning_rate": 4.105008167445076e-06,
+      "loss": 0.9771,
+      "step": 7372
+    },
+    {
+      "epoch": 1.4193517337632648,
+      "grad_norm": 3.3456634701639083,
+      "learning_rate": 4.10248991385563e-06,
+      "loss": 1.0419,
+      "step": 7373
+    },
+    {
+      "epoch": 1.4195442404408403,
+      "grad_norm": 3.2372282551924627,
+      "learning_rate": 4.099972233579608e-06,
+      "loss": 1.0027,
+      "step": 7374
+    },
+    {
+      "epoch": 1.4197367471184157,
+      "grad_norm": 3.2251809983981645,
+      "learning_rate": 4.097455126861776e-06,
+      "loss": 0.9876,
+      "step": 7375
+    },
+    {
+      "epoch": 1.419929253795991,
+      "grad_norm": 3.3391858143242152,
+      "learning_rate": 4.094938593946817e-06,
+      "loss": 1.0079,
+      "step": 7376
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 0.6546,
+      "step": 7376,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 0.6136,
+      "step": 7376,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 0.7492,
+      "step": 7376,
+      "vm_loss": 0.2151
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 1.1023,
+      "step": 7376,
+      "vm_loss": 0.1368
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 0.6279,
+      "step": 7376,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 1.6072,
+      "step": 7376,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 0.6872,
+      "step": 7376,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.419929253795991,
+      "lm_loss": 1.1598,
+      "step": 7376,
+      "vm_loss": 0.207
+    },
+    {
+      "epoch": 1.4201217604735663,
+      "grad_norm": 3.4046885308325536,
+      "learning_rate": 4.092422635079375e-06,
+      "loss": 1.0126,
+      "step": 7377
+    },
+    {
+      "epoch": 1.4203142671511417,
+      "grad_norm": 3.446814180320635,
+      "learning_rate": 4.089907250504032e-06,
+      "loss": 1.0212,
+      "step": 7378
+    },
+    {
+      "epoch": 1.4205067738287172,
+      "grad_norm": 3.4339173009741746,
+      "learning_rate": 4.087392440465321e-06,
+      "loss": 1.0277,
+      "step": 7379
+    },
+    {
+      "epoch": 1.4206992805062926,
+      "grad_norm": 3.242526341651682,
+      "learning_rate": 4.084878205207702e-06,
+      "loss": 1.029,
+      "step": 7380
+    },
+    {
+      "epoch": 1.420891787183868,
+      "grad_norm": 3.4595850655916522,
+      "learning_rate": 4.082364544975604e-06,
+      "loss": 1.0997,
+      "step": 7381
+    },
+    {
+      "epoch": 1.4210842938614432,
+      "grad_norm": 3.183047256656175,
+      "learning_rate": 4.079851460013377e-06,
+      "loss": 1.03,
+      "step": 7382
+    },
+    {
+      "epoch": 1.4212768005390186,
+      "grad_norm": 3.1627443896582776,
+      "learning_rate": 4.07733895056533e-06,
+      "loss": 1.0262,
+      "step": 7383
+    },
+    {
+      "epoch": 1.421469307216594,
+      "grad_norm": 3.158894113957345,
+      "learning_rate": 4.074827016875711e-06,
+      "loss": 1.0214,
+      "step": 7384
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 0.8131,
+      "step": 7384,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 1.016,
+      "step": 7384,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 0.584,
+      "step": 7384,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 0.767,
+      "step": 7384,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 1.0661,
+      "step": 7384,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 0.8573,
+      "step": 7384,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 1.0721,
+      "step": 7384,
+      "vm_loss": 0.1553
+    },
+    {
+      "epoch": 1.421469307216594,
+      "lm_loss": 1.0122,
+      "step": 7384,
+      "vm_loss": 0.1163
+    },
+    {
+      "epoch": 1.4216618138941695,
+      "grad_norm": 3.350069305976011,
+      "learning_rate": 4.072315659188709e-06,
+      "loss": 1.0566,
+      "step": 7385
+    },
+    {
+      "epoch": 1.4218543205717449,
+      "grad_norm": 3.2171703777806897,
+      "learning_rate": 4.069804877748468e-06,
+      "loss": 1.0285,
+      "step": 7386
+    },
+    {
+      "epoch": 1.42204682724932,
+      "grad_norm": 3.190166627259227,
+      "learning_rate": 4.067294672799053e-06,
+      "loss": 1.0668,
+      "step": 7387
+    },
+    {
+      "epoch": 1.4222393339268957,
+      "grad_norm": 3.3119135660117944,
+      "learning_rate": 4.064785044584505e-06,
+      "loss": 1.0169,
+      "step": 7388
+    },
+    {
+      "epoch": 1.422431840604471,
+      "grad_norm": 3.207983359737915,
+      "learning_rate": 4.062275993348781e-06,
+      "loss": 0.9527,
+      "step": 7389
+    },
+    {
+      "epoch": 1.4226243472820463,
+      "grad_norm": 3.258990378604529,
+      "learning_rate": 4.0597675193357995e-06,
+      "loss": 1.0314,
+      "step": 7390
+    },
+    {
+      "epoch": 1.4228168539596218,
+      "grad_norm": 3.1921282043953663,
+      "learning_rate": 4.0572596227894054e-06,
+      "loss": 0.9862,
+      "step": 7391
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "grad_norm": 3.340513437604555,
+      "learning_rate": 4.0547523039534135e-06,
+      "loss": 1.0409,
+      "step": 7392
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.5914,
+      "step": 7392,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.9033,
+      "step": 7392,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.6052,
+      "step": 7392,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 1.1184,
+      "step": 7392,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.5199,
+      "step": 7392,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.5446,
+      "step": 7392,
+      "vm_loss": 0.1292
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.6481,
+      "step": 7392,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.4230093606371972,
+      "lm_loss": 0.7845,
+      "step": 7392,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 1.4232018673147726,
+      "grad_norm": 3.20556945845834,
+      "learning_rate": 4.052245563071557e-06,
+      "loss": 1.0324,
+      "step": 7393
+    },
+    {
+      "epoch": 1.4233943739923478,
+      "grad_norm": 3.4175801825526073,
+      "learning_rate": 4.049739400387524e-06,
+      "loss": 1.0217,
+      "step": 7394
+    },
+    {
+      "epoch": 1.4235868806699232,
+      "grad_norm": 3.3716952623686627,
+      "learning_rate": 4.0472338161449475e-06,
+      "loss": 1.0453,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4237793873474986,
+      "grad_norm": 3.244058303981847,
+      "learning_rate": 4.044728810587406e-06,
+      "loss": 1.0354,
+      "step": 7396
+    },
+    {
+      "epoch": 1.423971894025074,
+      "grad_norm": 3.2972306629488317,
+      "learning_rate": 4.042224383958408e-06,
+      "loss": 1.0326,
+      "step": 7397
+    },
+    {
+      "epoch": 1.4241644007026495,
+      "grad_norm": 3.2273620322360648,
+      "learning_rate": 4.039720536501428e-06,
+      "loss": 0.9559,
+      "step": 7398
+    },
+    {
+      "epoch": 1.4243569073802247,
+      "grad_norm": 3.3238210969343887,
+      "learning_rate": 4.037217268459863e-06,
+      "loss": 1.046,
+      "step": 7399
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "grad_norm": 3.2923909523340944,
+      "learning_rate": 4.034714580077066e-06,
+      "loss": 0.9389,
+      "step": 7400
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.6158,
+      "step": 7400,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 1.0048,
+      "step": 7400,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.7914,
+      "step": 7400,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.5698,
+      "step": 7400,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.9156,
+      "step": 7400,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.768,
+      "step": 7400,
+      "vm_loss": 0.2151
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.3554,
+      "step": 7400,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.4245494140578001,
+      "lm_loss": 0.9098,
+      "step": 7400,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 1.4247419207353755,
+      "grad_norm": 3.2288875348629715,
+      "learning_rate": 4.03221247159633e-06,
+      "loss": 1.0197,
+      "step": 7401
+    },
+    {
+      "epoch": 1.424934427412951,
+      "grad_norm": 3.115225700658362,
+      "learning_rate": 4.029710943260895e-06,
+      "loss": 0.9632,
+      "step": 7402
+    },
+    {
+      "epoch": 1.4251269340905264,
+      "grad_norm": 3.1840041920195,
+      "learning_rate": 4.027209995313935e-06,
+      "loss": 1.0211,
+      "step": 7403
+    },
+    {
+      "epoch": 1.4253194407681016,
+      "grad_norm": 3.2325860128206916,
+      "learning_rate": 4.024709627998579e-06,
+      "loss": 0.9853,
+      "step": 7404
+    },
+    {
+      "epoch": 1.425511947445677,
+      "grad_norm": 3.2774833124718556,
+      "learning_rate": 4.022209841557892e-06,
+      "loss": 0.9616,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4257044541232524,
+      "grad_norm": 3.3981942344028373,
+      "learning_rate": 4.019710636234888e-06,
+      "loss": 1.0437,
+      "step": 7406
+    },
+    {
+      "epoch": 1.4258969608008278,
+      "grad_norm": 3.301355253386875,
+      "learning_rate": 4.017212012272524e-06,
+      "loss": 1.0403,
+      "step": 7407
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "grad_norm": 3.22398247074651,
+      "learning_rate": 4.014713969913686e-06,
+      "loss": 0.9735,
+      "step": 7408
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 0.8611,
+      "step": 7408,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 0.5353,
+      "step": 7408,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 1.1884,
+      "step": 7408,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 1.1233,
+      "step": 7408,
+      "vm_loss": 0.1503
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 1.0175,
+      "step": 7408,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 0.771,
+      "step": 7408,
+      "vm_loss": 0.2339
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 0.7138,
+      "step": 7408,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.4260894674784033,
+      "lm_loss": 1.0238,
+      "step": 7408,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.4262819741559785,
+      "grad_norm": 3.303176690818786,
+      "learning_rate": 4.0122165094012335e-06,
+      "loss": 1.0029,
+      "step": 7409
+    },
+    {
+      "epoch": 1.4264744808335539,
+      "grad_norm": 3.256662276211901,
+      "learning_rate": 4.009719630977937e-06,
+      "loss": 1.0179,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4266669875111293,
+      "grad_norm": 3.1580133047293906,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 1.0115,
+      "step": 7411
+    },
+    {
+      "epoch": 1.4268594941887047,
+      "grad_norm": 3.1874531311439176,
+      "learning_rate": 4.004727621369688e-06,
+      "loss": 0.9649,
+      "step": 7412
+    },
+    {
+      "epoch": 1.4270520008662801,
+      "grad_norm": 3.13307885663222,
+      "learning_rate": 4.002232490670024e-06,
+      "loss": 1.0101,
+      "step": 7413
+    },
+    {
+      "epoch": 1.4272445075438553,
+      "grad_norm": 3.4095476491296495,
+      "learning_rate": 3.999737943030088e-06,
+      "loss": 1.0812,
+      "step": 7414
+    },
+    {
+      "epoch": 1.4274370142214308,
+      "grad_norm": 3.1072478471573235,
+      "learning_rate": 3.997243978692399e-06,
+      "loss": 0.9711,
+      "step": 7415
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "grad_norm": 3.255597463997609,
+      "learning_rate": 3.994750597899389e-06,
+      "loss": 1.0046,
+      "step": 7416
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 1.1918,
+      "step": 7416,
+      "vm_loss": 0.1391
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.5311,
+      "step": 7416,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 1.0459,
+      "step": 7416,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.5959,
+      "step": 7416,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.611,
+      "step": 7416,
+      "vm_loss": 0.2386
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.4351,
+      "step": 7416,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.8348,
+      "step": 7416,
+      "vm_loss": 0.2063
+    },
+    {
+      "epoch": 1.4276295208990062,
+      "lm_loss": 0.5412,
+      "step": 7416,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 1.4278220275765816,
+      "grad_norm": 3.123729056411103,
+      "learning_rate": 3.992257800893453e-06,
+      "loss": 0.9645,
+      "step": 7417
+    },
+    {
+      "epoch": 1.428014534254157,
+      "grad_norm": 3.2674029856602966,
+      "learning_rate": 3.989765587916914e-06,
+      "loss": 0.9322,
+      "step": 7418
+    },
+    {
+      "epoch": 1.4282070409317322,
+      "grad_norm": 3.188003582916186,
+      "learning_rate": 3.987273959212059e-06,
+      "loss": 0.9531,
+      "step": 7419
+    },
+    {
+      "epoch": 1.4283995476093077,
+      "grad_norm": 3.5679165307966625,
+      "learning_rate": 3.984782915021098e-06,
+      "loss": 1.0613,
+      "step": 7420
+    },
+    {
+      "epoch": 1.428592054286883,
+      "grad_norm": 3.420518370485407,
+      "learning_rate": 3.982292455586195e-06,
+      "loss": 1.0468,
+      "step": 7421
+    },
+    {
+      "epoch": 1.4287845609644585,
+      "grad_norm": 3.329595565137712,
+      "learning_rate": 3.979802581149453e-06,
+      "loss": 0.9915,
+      "step": 7422
+    },
+    {
+      "epoch": 1.428977067642034,
+      "grad_norm": 3.3412893140265023,
+      "learning_rate": 3.977313291952926e-06,
+      "loss": 1.0029,
+      "step": 7423
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "grad_norm": 3.2661413648901907,
+      "learning_rate": 3.974824588238595e-06,
+      "loss": 0.9889,
+      "step": 7424
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 1.0655,
+      "step": 7424,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 0.578,
+      "step": 7424,
+      "vm_loss": 0.2205
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 1.0376,
+      "step": 7424,
+      "vm_loss": 0.1147
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 1.2093,
+      "step": 7424,
+      "vm_loss": 0.1047
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 0.819,
+      "step": 7424,
+      "vm_loss": 0.0999
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 0.5052,
+      "step": 7424,
+      "vm_loss": 0.1983
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 0.7423,
+      "step": 7424,
+      "vm_loss": 0.1182
+    },
+    {
+      "epoch": 1.4291695743196091,
+      "lm_loss": 1.009,
+      "step": 7424,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.4293620809971845,
+      "grad_norm": 3.252995902185338,
+      "learning_rate": 3.972336470248399e-06,
+      "loss": 1.0009,
+      "step": 7425
+    },
+    {
+      "epoch": 1.42955458767476,
+      "grad_norm": 3.2873847364206763,
+      "learning_rate": 3.969848938224215e-06,
+      "loss": 1.0308,
+      "step": 7426
+    },
+    {
+      "epoch": 1.4297470943523354,
+      "grad_norm": 3.245030794213246,
+      "learning_rate": 3.967361992407861e-06,
+      "loss": 0.9839,
+      "step": 7427
+    },
+    {
+      "epoch": 1.4299396010299108,
+      "grad_norm": 3.291314172424728,
+      "learning_rate": 3.964875633041102e-06,
+      "loss": 1.0538,
+      "step": 7428
+    },
+    {
+      "epoch": 1.430132107707486,
+      "grad_norm": 3.2161676801464143,
+      "learning_rate": 3.962389860365643e-06,
+      "loss": 0.9574,
+      "step": 7429
+    },
+    {
+      "epoch": 1.4303246143850614,
+      "grad_norm": 3.0914050654666987,
+      "learning_rate": 3.959904674623137e-06,
+      "loss": 0.9535,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4305171210626368,
+      "grad_norm": 3.212896007320673,
+      "learning_rate": 3.957420076055165e-06,
+      "loss": 0.9593,
+      "step": 7431
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "grad_norm": 3.404286849714347,
+      "learning_rate": 3.954936064903276e-06,
+      "loss": 1.0614,
+      "step": 7432
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.7622,
+      "step": 7432,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.9301,
+      "step": 7432,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.8515,
+      "step": 7432,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.6838,
+      "step": 7432,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 1.0809,
+      "step": 7432,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.8926,
+      "step": 7432,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 1.3521,
+      "step": 7432,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.4307096277402123,
+      "lm_loss": 0.9398,
+      "step": 7432,
+      "vm_loss": 0.0998
+    },
+    {
+      "epoch": 1.4309021344177877,
+      "grad_norm": 3.345257249499339,
+      "learning_rate": 3.952452641408937e-06,
+      "loss": 1.0511,
+      "step": 7433
+    },
+    {
+      "epoch": 1.4310946410953629,
+      "grad_norm": 3.2836892473193955,
+      "learning_rate": 3.949969805813575e-06,
+      "loss": 0.9989,
+      "step": 7434
+    },
+    {
+      "epoch": 1.4312871477729383,
+      "grad_norm": 3.3396982441324647,
+      "learning_rate": 3.947487558358544e-06,
+      "loss": 1.0021,
+      "step": 7435
+    },
+    {
+      "epoch": 1.4314796544505137,
+      "grad_norm": 3.4172474902373176,
+      "learning_rate": 3.945005899285163e-06,
+      "loss": 1.0002,
+      "step": 7436
+    },
+    {
+      "epoch": 1.4316721611280891,
+      "grad_norm": 3.201080773306174,
+      "learning_rate": 3.94252482883467e-06,
+      "loss": 0.9697,
+      "step": 7437
+    },
+    {
+      "epoch": 1.4318646678056646,
+      "grad_norm": 3.2652406216236427,
+      "learning_rate": 3.940044347248263e-06,
+      "loss": 0.9904,
+      "step": 7438
+    },
+    {
+      "epoch": 1.4320571744832398,
+      "grad_norm": 3.351456416168999,
+      "learning_rate": 3.937564454767074e-06,
+      "loss": 1.0224,
+      "step": 7439
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "grad_norm": 3.244925645364108,
+      "learning_rate": 3.935085151632185e-06,
+      "loss": 1.0115,
+      "step": 7440
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 0.4762,
+      "step": 7440,
+      "vm_loss": 0.1246
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 1.3445,
+      "step": 7440,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 1.1248,
+      "step": 7440,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 0.8252,
+      "step": 7440,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 0.7972,
+      "step": 7440,
+      "vm_loss": 0.1218
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 0.5959,
+      "step": 7440,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 0.7924,
+      "step": 7440,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.4322496811608152,
+      "lm_loss": 1.1844,
+      "step": 7440,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 1.4324421878383906,
+      "grad_norm": 3.220571344953075,
+      "learning_rate": 3.932606438084609e-06,
+      "loss": 0.9602,
+      "step": 7441
+    },
+    {
+      "epoch": 1.432634694515966,
+      "grad_norm": 3.2605862014377633,
+      "learning_rate": 3.9301283143653125e-06,
+      "loss": 0.9803,
+      "step": 7442
+    },
+    {
+      "epoch": 1.4328272011935415,
+      "grad_norm": 3.3283590305902697,
+      "learning_rate": 3.9276507807151985e-06,
+      "loss": 1.0042,
+      "step": 7443
+    },
+    {
+      "epoch": 1.4330197078711167,
+      "grad_norm": 3.3350422244063975,
+      "learning_rate": 3.925173837375118e-06,
+      "loss": 1.0209,
+      "step": 7444
+    },
+    {
+      "epoch": 1.433212214548692,
+      "grad_norm": 3.2043577402483088,
+      "learning_rate": 3.92269748458586e-06,
+      "loss": 0.9655,
+      "step": 7445
+    },
+    {
+      "epoch": 1.4334047212262675,
+      "grad_norm": 3.31921524712407,
+      "learning_rate": 3.920221722588161e-06,
+      "loss": 0.9912,
+      "step": 7446
+    },
+    {
+      "epoch": 1.433597227903843,
+      "grad_norm": 3.124418309022242,
+      "learning_rate": 3.91774655162269e-06,
+      "loss": 0.9347,
+      "step": 7447
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "grad_norm": 3.3104114389444033,
+      "learning_rate": 3.915271971930069e-06,
+      "loss": 0.989,
+      "step": 7448
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 1.1521,
+      "step": 7448,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.4039,
+      "step": 7448,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.8436,
+      "step": 7448,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 1.0746,
+      "step": 7448,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.7943,
+      "step": 7448,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.7272,
+      "step": 7448,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.6191,
+      "step": 7448,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 1.4337897345814183,
+      "lm_loss": 0.5419,
+      "step": 7448,
+      "vm_loss": 0.1123
+    },
+    {
+      "epoch": 1.4339822412589935,
+      "grad_norm": 3.214734386182521,
+      "learning_rate": 3.912797983750859e-06,
+      "loss": 1.0084,
+      "step": 7449
+    },
+    {
+      "epoch": 1.4341747479365692,
+      "grad_norm": 3.236118347486923,
+      "learning_rate": 3.910324587325564e-06,
+      "loss": 0.933,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4343672546141444,
+      "grad_norm": 3.2760855075400297,
+      "learning_rate": 3.90785178289463e-06,
+      "loss": 0.9781,
+      "step": 7451
+    },
+    {
+      "epoch": 1.4345597612917198,
+      "grad_norm": 3.4453944172252378,
+      "learning_rate": 3.905379570698439e-06,
+      "loss": 1.0589,
+      "step": 7452
+    },
+    {
+      "epoch": 1.4347522679692952,
+      "grad_norm": 3.286427005157244,
+      "learning_rate": 3.902907950977333e-06,
+      "loss": 1.027,
+      "step": 7453
+    },
+    {
+      "epoch": 1.4349447746468704,
+      "grad_norm": 3.350066963710581,
+      "learning_rate": 3.900436923971576e-06,
+      "loss": 1.0257,
+      "step": 7454
+    },
+    {
+      "epoch": 1.435137281324446,
+      "grad_norm": 3.2163408955439383,
+      "learning_rate": 3.897966489921384e-06,
+      "loss": 0.9931,
+      "step": 7455
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "grad_norm": 3.1868054016766503,
+      "learning_rate": 3.895496649066919e-06,
+      "loss": 0.9636,
+      "step": 7456
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.8351,
+      "step": 7456,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.6497,
+      "step": 7456,
+      "vm_loss": 0.1041
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.5105,
+      "step": 7456,
+      "vm_loss": 0.1195
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.788,
+      "step": 7456,
+      "vm_loss": 0.1804
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.6853,
+      "step": 7456,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.8977,
+      "step": 7456,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 1.2236,
+      "step": 7456,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.4353297880020213,
+      "lm_loss": 0.8895,
+      "step": 7456,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 1.4355222946795967,
+      "grad_norm": 3.280423662805804,
+      "learning_rate": 3.893027401648281e-06,
+      "loss": 0.9728,
+      "step": 7457
+    },
+    {
+      "epoch": 1.435714801357172,
+      "grad_norm": 3.3222574346585723,
+      "learning_rate": 3.89055874790551e-06,
+      "loss": 0.9605,
+      "step": 7458
+    },
+    {
+      "epoch": 1.4359073080347475,
+      "grad_norm": 3.218268050577204,
+      "learning_rate": 3.888090688078589e-06,
+      "loss": 0.9926,
+      "step": 7459
+    },
+    {
+      "epoch": 1.436099814712323,
+      "grad_norm": 3.317277203857952,
+      "learning_rate": 3.885623222407448e-06,
+      "loss": 0.985,
+      "step": 7460
+    },
+    {
+      "epoch": 1.4362923213898982,
+      "grad_norm": 3.2611957253280526,
+      "learning_rate": 3.8831563511319585e-06,
+      "loss": 1.0346,
+      "step": 7461
+    },
+    {
+      "epoch": 1.4364848280674736,
+      "grad_norm": 3.33529861871596,
+      "learning_rate": 3.8806900744919205e-06,
+      "loss": 1.0221,
+      "step": 7462
+    },
+    {
+      "epoch": 1.436677334745049,
+      "grad_norm": 3.2641392981184447,
+      "learning_rate": 3.878224392727106e-06,
+      "loss": 1.002,
+      "step": 7463
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "grad_norm": 3.308520639128367,
+      "learning_rate": 3.875759306077196e-06,
+      "loss": 1.0438,
+      "step": 7464
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.6734,
+      "step": 7464,
+      "vm_loss": 0.1295
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.8854,
+      "step": 7464,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.4941,
+      "step": 7464,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.6,
+      "step": 7464,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.8256,
+      "step": 7464,
+      "vm_loss": 0.251
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 1.0003,
+      "step": 7464,
+      "vm_loss": 0.2095
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.6397,
+      "step": 7464,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.4368698414226244,
+      "lm_loss": 0.6008,
+      "step": 7464,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 1.4370623481001998,
+      "grad_norm": 3.2277235291794386,
+      "learning_rate": 3.8732948147818324e-06,
+      "loss": 0.9642,
+      "step": 7465
+    },
+    {
+      "epoch": 1.437254854777775,
+      "grad_norm": 3.203874049249102,
+      "learning_rate": 3.870830919080598e-06,
+      "loss": 0.9526,
+      "step": 7466
+    },
+    {
+      "epoch": 1.4374473614553505,
+      "grad_norm": 3.236389700174982,
+      "learning_rate": 3.868367619213012e-06,
+      "loss": 0.9435,
+      "step": 7467
+    },
+    {
+      "epoch": 1.4376398681329259,
+      "grad_norm": 3.2450882075874237,
+      "learning_rate": 3.865904915418544e-06,
+      "loss": 1.01,
+      "step": 7468
+    },
+    {
+      "epoch": 1.4378323748105013,
+      "grad_norm": 3.235749810889186,
+      "learning_rate": 3.863442807936592e-06,
+      "loss": 0.9921,
+      "step": 7469
+    },
+    {
+      "epoch": 1.4380248814880767,
+      "grad_norm": 3.3633187079149254,
+      "learning_rate": 3.86098129700651e-06,
+      "loss": 1.023,
+      "step": 7470
+    },
+    {
+      "epoch": 1.438217388165652,
+      "grad_norm": 3.4290620440258244,
+      "learning_rate": 3.858520382867586e-06,
+      "loss": 1.0356,
+      "step": 7471
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "grad_norm": 3.4227166008932906,
+      "learning_rate": 3.856060065759053e-06,
+      "loss": 1.0122,
+      "step": 7472
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.4834,
+      "step": 7472,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.4899,
+      "step": 7472,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.9298,
+      "step": 7472,
+      "vm_loss": 0.1239
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.7222,
+      "step": 7472,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.6443,
+      "step": 7472,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 1.0035,
+      "step": 7472,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 0.9978,
+      "step": 7472,
+      "vm_loss": 0.2305
+    },
+    {
+      "epoch": 1.4384098948432273,
+      "lm_loss": 1.0775,
+      "step": 7472,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 1.4386024015208028,
+      "grad_norm": 3.2644956050559304,
+      "learning_rate": 3.853600345920087e-06,
+      "loss": 0.9965,
+      "step": 7473
+    },
+    {
+      "epoch": 1.4387949081983782,
+      "grad_norm": 3.2573708745019103,
+      "learning_rate": 3.851141223589805e-06,
+      "loss": 0.985,
+      "step": 7474
+    },
+    {
+      "epoch": 1.4389874148759536,
+      "grad_norm": 3.202165768748596,
+      "learning_rate": 3.848682699007259e-06,
+      "loss": 1.0116,
+      "step": 7475
+    },
+    {
+      "epoch": 1.4391799215535288,
+      "grad_norm": 3.3032648931040574,
+      "learning_rate": 3.846224772411454e-06,
+      "loss": 1.0396,
+      "step": 7476
+    },
+    {
+      "epoch": 1.4393724282311042,
+      "grad_norm": 3.2751949325642595,
+      "learning_rate": 3.8437674440413306e-06,
+      "loss": 1.0157,
+      "step": 7477
+    },
+    {
+      "epoch": 1.4395649349086796,
+      "grad_norm": 3.219988408009005,
+      "learning_rate": 3.8413107141357765e-06,
+      "loss": 0.9703,
+      "step": 7478
+    },
+    {
+      "epoch": 1.439757441586255,
+      "grad_norm": 3.1296905556426107,
+      "learning_rate": 3.838854582933605e-06,
+      "loss": 0.9741,
+      "step": 7479
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "grad_norm": 3.2371632550965828,
+      "learning_rate": 3.836399050673601e-06,
+      "loss": 0.9632,
+      "step": 7480
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.7624,
+      "step": 7480,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.7381,
+      "step": 7480,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.9242,
+      "step": 7480,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.647,
+      "step": 7480,
+      "vm_loss": 0.233
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.9376,
+      "step": 7480,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.8964,
+      "step": 7480,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.8888,
+      "step": 7480,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 1.4399499482638305,
+      "lm_loss": 0.9,
+      "step": 7480,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 1.4401424549414057,
+      "grad_norm": 3.2498923138714635,
+      "learning_rate": 3.833944117594462e-06,
+      "loss": 0.9978,
+      "step": 7481
+    },
+    {
+      "epoch": 1.4403349616189811,
+      "grad_norm": 3.262256771287942,
+      "learning_rate": 3.831489783934843e-06,
+      "loss": 0.9932,
+      "step": 7482
+    },
+    {
+      "epoch": 1.4405274682965565,
+      "grad_norm": 3.4630054564912576,
+      "learning_rate": 3.829036049933329e-06,
+      "loss": 1.0508,
+      "step": 7483
+    },
+    {
+      "epoch": 1.440719974974132,
+      "grad_norm": 3.4278737502812477,
+      "learning_rate": 3.826582915828468e-06,
+      "loss": 1.0402,
+      "step": 7484
+    },
+    {
+      "epoch": 1.4409124816517074,
+      "grad_norm": 3.372472709575327,
+      "learning_rate": 3.824130381858725e-06,
+      "loss": 1.0717,
+      "step": 7485
+    },
+    {
+      "epoch": 1.4411049883292826,
+      "grad_norm": 3.1609286457287773,
+      "learning_rate": 3.821678448262522e-06,
+      "loss": 0.9934,
+      "step": 7486
+    },
+    {
+      "epoch": 1.441297495006858,
+      "grad_norm": 3.146847478090736,
+      "learning_rate": 3.8192271152782164e-06,
+      "loss": 0.962,
+      "step": 7487
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "grad_norm": 3.296196532032015,
+      "learning_rate": 3.816776383144111e-06,
+      "loss": 1.0363,
+      "step": 7488
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.7895,
+      "step": 7488,
+      "vm_loss": 0.2194
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.7976,
+      "step": 7488,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 1.1302,
+      "step": 7488,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.7889,
+      "step": 7488,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.9545,
+      "step": 7488,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.4963,
+      "step": 7488,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 1.1055,
+      "step": 7488,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.4414900016844334,
+      "lm_loss": 0.9651,
+      "step": 7488,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.4416825083620088,
+      "grad_norm": 3.3643801054859632,
+      "learning_rate": 3.814326252098447e-06,
+      "loss": 1.0321,
+      "step": 7489
+    },
+    {
+      "epoch": 1.4418750150395843,
+      "grad_norm": 3.226943747711484,
+      "learning_rate": 3.8118767223794105e-06,
+      "loss": 0.9387,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4420675217171595,
+      "grad_norm": 3.1776805066055407,
+      "learning_rate": 3.809427794225128e-06,
+      "loss": 0.9575,
+      "step": 7491
+    },
+    {
+      "epoch": 1.4422600283947349,
+      "grad_norm": 3.210129592832735,
+      "learning_rate": 3.8069794678736614e-06,
+      "loss": 0.9619,
+      "step": 7492
+    },
+    {
+      "epoch": 1.4424525350723103,
+      "grad_norm": 3.260087344804152,
+      "learning_rate": 3.8045317435630226e-06,
+      "loss": 0.9999,
+      "step": 7493
+    },
+    {
+      "epoch": 1.4426450417498857,
+      "grad_norm": 3.237644060631388,
+      "learning_rate": 3.8020846215311612e-06,
+      "loss": 1.0142,
+      "step": 7494
+    },
+    {
+      "epoch": 1.4428375484274611,
+      "grad_norm": 3.2343796105376588,
+      "learning_rate": 3.7996381020159733e-06,
+      "loss": 0.9763,
+      "step": 7495
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "grad_norm": 3.2320392764616517,
+      "learning_rate": 3.7971921852552807e-06,
+      "loss": 0.9766,
+      "step": 7496
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.3994,
+      "step": 7496,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.7851,
+      "step": 7496,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 1.002,
+      "step": 7496,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.7841,
+      "step": 7496,
+      "vm_loss": 0.2111
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.7084,
+      "step": 7496,
+      "vm_loss": 0.1516
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 1.0134,
+      "step": 7496,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.737,
+      "step": 7496,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 1.4430300551050363,
+      "lm_loss": 0.745,
+      "step": 7496,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 1.4432225617826118,
+      "grad_norm": 3.3132321566950442,
+      "learning_rate": 3.794746871486871e-06,
+      "loss": 1.0172,
+      "step": 7497
+    },
+    {
+      "epoch": 1.4434150684601872,
+      "grad_norm": 3.2689003122046763,
+      "learning_rate": 3.79230216094845e-06,
+      "loss": 0.9648,
+      "step": 7498
+    },
+    {
+      "epoch": 1.4436075751377626,
+      "grad_norm": 3.2688826987465878,
+      "learning_rate": 3.789858053877683e-06,
+      "loss": 0.9893,
+      "step": 7499
+    },
+    {
+      "epoch": 1.443800081815338,
+      "grad_norm": 3.3077582825469998,
+      "learning_rate": 3.7874145505121563e-06,
+      "loss": 1.0048,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4439925884929132,
+      "grad_norm": 3.2895767499573365,
+      "learning_rate": 3.784971651089425e-06,
+      "loss": 1.017,
+      "step": 7501
+    },
+    {
+      "epoch": 1.4441850951704887,
+      "grad_norm": 3.2483683485805295,
+      "learning_rate": 3.782529355846959e-06,
+      "loss": 0.974,
+      "step": 7502
+    },
+    {
+      "epoch": 1.444377601848064,
+      "grad_norm": 3.3250399042741337,
+      "learning_rate": 3.780087665022184e-06,
+      "loss": 1.0322,
+      "step": 7503
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "grad_norm": 3.089754405345806,
+      "learning_rate": 3.777646578852464e-06,
+      "loss": 0.9597,
+      "step": 7504
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.9068,
+      "step": 7504,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.6463,
+      "step": 7504,
+      "vm_loss": 0.2443
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.8423,
+      "step": 7504,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.7481,
+      "step": 7504,
+      "vm_loss": 0.229
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.8449,
+      "step": 7504,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 1.4216,
+      "step": 7504,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 0.6287,
+      "step": 7504,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 1.4445701085256395,
+      "lm_loss": 1.1786,
+      "step": 7504,
+      "vm_loss": 0.2206
+    },
+    {
+      "epoch": 1.444762615203215,
+      "grad_norm": 3.2833387844393664,
+      "learning_rate": 3.775206097575108e-06,
+      "loss": 1.0412,
+      "step": 7505
+    },
+    {
+      "epoch": 1.4449551218807901,
+      "grad_norm": 3.324964519919435,
+      "learning_rate": 3.7727662214273496e-06,
+      "loss": 1.023,
+      "step": 7506
+    },
+    {
+      "epoch": 1.4451476285583655,
+      "grad_norm": 3.2779973115036776,
+      "learning_rate": 3.7703269506463914e-06,
+      "loss": 1.0081,
+      "step": 7507
+    },
+    {
+      "epoch": 1.445340135235941,
+      "grad_norm": 3.1481965703019603,
+      "learning_rate": 3.767888285469351e-06,
+      "loss": 0.9619,
+      "step": 7508
+    },
+    {
+      "epoch": 1.4455326419135164,
+      "grad_norm": 3.2253701021785544,
+      "learning_rate": 3.7654502261333017e-06,
+      "loss": 0.9839,
+      "step": 7509
+    },
+    {
+      "epoch": 1.4457251485910918,
+      "grad_norm": 3.2457753010481505,
+      "learning_rate": 3.763012772875253e-06,
+      "loss": 0.9837,
+      "step": 7510
+    },
+    {
+      "epoch": 1.445917655268667,
+      "grad_norm": 3.445943380043002,
+      "learning_rate": 3.760575925932157e-06,
+      "loss": 1.0641,
+      "step": 7511
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "grad_norm": 3.369873365835705,
+      "learning_rate": 3.75813968554091e-06,
+      "loss": 1.0189,
+      "step": 7512
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 1.0479,
+      "step": 7512,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 0.3101,
+      "step": 7512,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 0.9104,
+      "step": 7512,
+      "vm_loss": 0.113
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 0.9212,
+      "step": 7512,
+      "vm_loss": 0.1012
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 1.2068,
+      "step": 7512,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 0.8678,
+      "step": 7512,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 1.1408,
+      "step": 7512,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 1.4461101619462426,
+      "lm_loss": 1.1058,
+      "step": 7512,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.4463026686238178,
+      "grad_norm": 3.3482323034896657,
+      "learning_rate": 3.755704051938339e-06,
+      "loss": 1.0005,
+      "step": 7513
+    },
+    {
+      "epoch": 1.4464951753013933,
+      "grad_norm": 3.1836188944426587,
+      "learning_rate": 3.7532690253612226e-06,
+      "loss": 1.0087,
+      "step": 7514
+    },
+    {
+      "epoch": 1.4466876819789687,
+      "grad_norm": 3.293230993070609,
+      "learning_rate": 3.7508346060462752e-06,
+      "loss": 1.0309,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4468801886565439,
+      "grad_norm": 3.1495133136119797,
+      "learning_rate": 3.748400794230158e-06,
+      "loss": 0.9427,
+      "step": 7516
+    },
+    {
+      "epoch": 1.4470726953341195,
+      "grad_norm": 3.230017133284445,
+      "learning_rate": 3.7459675901494587e-06,
+      "loss": 0.9685,
+      "step": 7517
+    },
+    {
+      "epoch": 1.4472652020116947,
+      "grad_norm": 3.328659724741726,
+      "learning_rate": 3.7435349940407274e-06,
+      "loss": 1.0522,
+      "step": 7518
+    },
+    {
+      "epoch": 1.4474577086892702,
+      "grad_norm": 3.1261749518043116,
+      "learning_rate": 3.741103006140436e-06,
+      "loss": 0.9525,
+      "step": 7519
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "grad_norm": 3.271125639666758,
+      "learning_rate": 3.7386716266850066e-06,
+      "loss": 0.9951,
+      "step": 7520
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.4959,
+      "step": 7520,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.8442,
+      "step": 7520,
+      "vm_loss": 0.1457
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.7844,
+      "step": 7520,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 1.0791,
+      "step": 7520,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 1.194,
+      "step": 7520,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.8037,
+      "step": 7520,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.8036,
+      "step": 7520,
+      "vm_loss": 0.1968
+    },
+    {
+      "epoch": 1.4476502153668456,
+      "lm_loss": 0.8202,
+      "step": 7520,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.447842722044421,
+      "grad_norm": 3.306374288929964,
+      "learning_rate": 3.736240855910801e-06,
+      "loss": 1.0315,
+      "step": 7521
+    },
+    {
+      "epoch": 1.4480352287219964,
+      "grad_norm": 3.2919471301750813,
+      "learning_rate": 3.733810694054124e-06,
+      "loss": 1.0287,
+      "step": 7522
+    },
+    {
+      "epoch": 1.4482277353995716,
+      "grad_norm": 3.27774068438974,
+      "learning_rate": 3.7313811413512134e-06,
+      "loss": 0.9701,
+      "step": 7523
+    },
+    {
+      "epoch": 1.448420242077147,
+      "grad_norm": 3.1863582032918623,
+      "learning_rate": 3.7289521980382535e-06,
+      "loss": 0.924,
+      "step": 7524
+    },
+    {
+      "epoch": 1.4486127487547225,
+      "grad_norm": 3.566581678943995,
+      "learning_rate": 3.7265238643513713e-06,
+      "loss": 1.0628,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4488052554322979,
+      "grad_norm": 3.1991363347929767,
+      "learning_rate": 3.724096140526633e-06,
+      "loss": 0.9763,
+      "step": 7526
+    },
+    {
+      "epoch": 1.4489977621098733,
+      "grad_norm": 3.3299171200360282,
+      "learning_rate": 3.7216690268000357e-06,
+      "loss": 1.0277,
+      "step": 7527
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "grad_norm": 3.1947564410213483,
+      "learning_rate": 3.719242523407539e-06,
+      "loss": 0.9573,
+      "step": 7528
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 1.2342,
+      "step": 7528,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.798,
+      "step": 7528,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.426,
+      "step": 7528,
+      "vm_loss": 0.1083
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.9921,
+      "step": 7528,
+      "vm_loss": 0.2058
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.4276,
+      "step": 7528,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.9316,
+      "step": 7528,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.8508,
+      "step": 7528,
+      "vm_loss": 0.2739
+    },
+    {
+      "epoch": 1.4491902687874485,
+      "lm_loss": 0.9222,
+      "step": 7528,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.449382775465024,
+      "grad_norm": 3.2972334841729443,
+      "learning_rate": 3.7168166305850208e-06,
+      "loss": 0.9857,
+      "step": 7529
+    },
+    {
+      "epoch": 1.4495752821425993,
+      "grad_norm": 3.488950617590216,
+      "learning_rate": 3.7143913485683103e-06,
+      "loss": 1.0115,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4497677888201748,
+      "grad_norm": 3.2191520296394294,
+      "learning_rate": 3.711966677593177e-06,
+      "loss": 0.9946,
+      "step": 7531
+    },
+    {
+      "epoch": 1.4499602954977502,
+      "grad_norm": 3.1475797537407226,
+      "learning_rate": 3.709542617895332e-06,
+      "loss": 0.9756,
+      "step": 7532
+    },
+    {
+      "epoch": 1.4501528021753254,
+      "grad_norm": 2.9694111380851993,
+      "learning_rate": 3.7071191697104248e-06,
+      "loss": 0.9059,
+      "step": 7533
+    },
+    {
+      "epoch": 1.4503453088529008,
+      "grad_norm": 3.277830864971584,
+      "learning_rate": 3.704696333274037e-06,
+      "loss": 0.9884,
+      "step": 7534
+    },
+    {
+      "epoch": 1.4505378155304762,
+      "grad_norm": 3.3678668580433313,
+      "learning_rate": 3.702274108821713e-06,
+      "loss": 1.0209,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "grad_norm": 3.1660710396110043,
+      "learning_rate": 3.6998524965889136e-06,
+      "loss": 0.9524,
+      "step": 7536
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 1.076,
+      "step": 7536,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.4982,
+      "step": 7536,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.5622,
+      "step": 7536,
+      "vm_loss": 0.1205
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.847,
+      "step": 7536,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.8279,
+      "step": 7536,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.7906,
+      "step": 7536,
+      "vm_loss": 0.2187
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.5357,
+      "step": 7536,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.4507303222080516,
+      "lm_loss": 0.5858,
+      "step": 7536,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 1.450922828885627,
+      "grad_norm": 3.1028320177361417,
+      "learning_rate": 3.6974314968110547e-06,
+      "loss": 0.9129,
+      "step": 7537
+    },
+    {
+      "epoch": 1.4511153355632023,
+      "grad_norm": 3.2609419564341526,
+      "learning_rate": 3.6950111097234866e-06,
+      "loss": 0.9689,
+      "step": 7538
+    },
+    {
+      "epoch": 1.4513078422407777,
+      "grad_norm": 3.289604171766757,
+      "learning_rate": 3.6925913355615075e-06,
+      "loss": 0.9961,
+      "step": 7539
+    },
+    {
+      "epoch": 1.4515003489183531,
+      "grad_norm": 3.1669610502451655,
+      "learning_rate": 3.6901721745603414e-06,
+      "loss": 0.9502,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4516928555959285,
+      "grad_norm": 3.495465152657839,
+      "learning_rate": 3.687753626955167e-06,
+      "loss": 1.0622,
+      "step": 7541
+    },
+    {
+      "epoch": 1.451885362273504,
+      "grad_norm": 3.289462938968708,
+      "learning_rate": 3.6853356929810966e-06,
+      "loss": 0.9704,
+      "step": 7542
+    },
+    {
+      "epoch": 1.4520778689510792,
+      "grad_norm": 3.260639672356218,
+      "learning_rate": 3.68291837287319e-06,
+      "loss": 0.9654,
+      "step": 7543
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "grad_norm": 3.370648975426911,
+      "learning_rate": 3.6805016668664284e-06,
+      "loss": 1.0135,
+      "step": 7544
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.5624,
+      "step": 7544,
+      "vm_loss": 0.18
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.6554,
+      "step": 7544,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.9162,
+      "step": 7544,
+      "vm_loss": 0.2429
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.8813,
+      "step": 7544,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.4676,
+      "step": 7544,
+      "vm_loss": 0.2293
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.4517,
+      "step": 7544,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.786,
+      "step": 7544,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.4522703756286546,
+      "lm_loss": 0.865,
+      "step": 7544,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.45246288230623,
+      "grad_norm": 3.207943675242393,
+      "learning_rate": 3.678085575195761e-06,
+      "loss": 1.0242,
+      "step": 7545
+    },
+    {
+      "epoch": 1.4526553889838054,
+      "grad_norm": 3.33134265411261,
+      "learning_rate": 3.6756700980960548e-06,
+      "loss": 0.9791,
+      "step": 7546
+    },
+    {
+      "epoch": 1.4528478956613808,
+      "grad_norm": 3.3400801629272605,
+      "learning_rate": 3.6732552358021258e-06,
+      "loss": 0.9816,
+      "step": 7547
+    },
+    {
+      "epoch": 1.453040402338956,
+      "grad_norm": 3.340912625850119,
+      "learning_rate": 3.6708409885487305e-06,
+      "loss": 1.0513,
+      "step": 7548
+    },
+    {
+      "epoch": 1.4532329090165315,
+      "grad_norm": 3.3629712138048617,
+      "learning_rate": 3.6684273565705686e-06,
+      "loss": 1.0565,
+      "step": 7549
+    },
+    {
+      "epoch": 1.4534254156941069,
+      "grad_norm": 3.2395413896837746,
+      "learning_rate": 3.6660143401022687e-06,
+      "loss": 1.0295,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4536179223716823,
+      "grad_norm": 3.122951720208705,
+      "learning_rate": 3.6636019393784093e-06,
+      "loss": 0.977,
+      "step": 7551
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "grad_norm": 3.194321558893384,
+      "learning_rate": 3.6611901546335093e-06,
+      "loss": 0.9901,
+      "step": 7552
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 1.2886,
+      "step": 7552,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 0.8625,
+      "step": 7552,
+      "vm_loss": 0.1085
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 0.8445,
+      "step": 7552,
+      "vm_loss": 0.137
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 0.8414,
+      "step": 7552,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 0.8963,
+      "step": 7552,
+      "vm_loss": 0.252
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 1.0389,
+      "step": 7552,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 1.2937,
+      "step": 7552,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 1.4538104290492577,
+      "lm_loss": 0.7024,
+      "step": 7552,
+      "vm_loss": 0.1378
+    },
+    {
+      "epoch": 1.454002935726833,
+      "grad_norm": 3.394555528487962,
+      "learning_rate": 3.658778986102022e-06,
+      "loss": 1.0609,
+      "step": 7553
+    },
+    {
+      "epoch": 1.4541954424044083,
+      "grad_norm": 3.2455965484785323,
+      "learning_rate": 3.6563684340183458e-06,
+      "loss": 0.9662,
+      "step": 7554
+    },
+    {
+      "epoch": 1.4543879490819838,
+      "grad_norm": 3.1950471856719442,
+      "learning_rate": 3.6539584986168163e-06,
+      "loss": 0.9818,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4545804557595592,
+      "grad_norm": 3.164321591551787,
+      "learning_rate": 3.651549180131714e-06,
+      "loss": 0.9722,
+      "step": 7556
+    },
+    {
+      "epoch": 1.4547729624371346,
+      "grad_norm": 3.1991018657563624,
+      "learning_rate": 3.6491404787972486e-06,
+      "loss": 0.9484,
+      "step": 7557
+    },
+    {
+      "epoch": 1.4549654691147098,
+      "grad_norm": 3.4178083616927935,
+      "learning_rate": 3.6467323948475786e-06,
+      "loss": 1.0199,
+      "step": 7558
+    },
+    {
+      "epoch": 1.4551579757922852,
+      "grad_norm": 3.3827364975113787,
+      "learning_rate": 3.6443249285168025e-06,
+      "loss": 1.0097,
+      "step": 7559
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "grad_norm": 3.0954711936454777,
+      "learning_rate": 3.6419180800389607e-06,
+      "loss": 0.9177,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.8161,
+      "step": 7560,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 1.4125,
+      "step": 7560,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.899,
+      "step": 7560,
+      "vm_loss": 0.2148
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.5899,
+      "step": 7560,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.8906,
+      "step": 7560,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.8449,
+      "step": 7560,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.6252,
+      "step": 7560,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 1.4553504824698607,
+      "lm_loss": 0.5667,
+      "step": 7560,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.455542989147436,
+      "grad_norm": 3.4012657303298885,
+      "learning_rate": 3.6395118496480176e-06,
+      "loss": 1.0099,
+      "step": 7561
+    },
+    {
+      "epoch": 1.4557354958250115,
+      "grad_norm": 3.1965460318289134,
+      "learning_rate": 3.637106237577903e-06,
+      "loss": 0.9137,
+      "step": 7562
+    },
+    {
+      "epoch": 1.4559280025025867,
+      "grad_norm": 3.156106085034886,
+      "learning_rate": 3.6347012440624653e-06,
+      "loss": 0.9592,
+      "step": 7563
+    },
+    {
+      "epoch": 1.4561205091801621,
+      "grad_norm": 3.1892515772761842,
+      "learning_rate": 3.6322968693355008e-06,
+      "loss": 0.9335,
+      "step": 7564
+    },
+    {
+      "epoch": 1.4563130158577375,
+      "grad_norm": 3.336148332296915,
+      "learning_rate": 3.629893113630749e-06,
+      "loss": 0.9825,
+      "step": 7565
+    },
+    {
+      "epoch": 1.456505522535313,
+      "grad_norm": 3.250439218883212,
+      "learning_rate": 3.627489977181887e-06,
+      "loss": 0.9767,
+      "step": 7566
+    },
+    {
+      "epoch": 1.4566980292128884,
+      "grad_norm": 3.1578545775510825,
+      "learning_rate": 3.625087460222524e-06,
+      "loss": 0.9337,
+      "step": 7567
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "grad_norm": 3.3216048226693635,
+      "learning_rate": 3.622685562986219e-06,
+      "loss": 0.9791,
+      "step": 7568
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.7149,
+      "step": 7568,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.874,
+      "step": 7568,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 1.0607,
+      "step": 7568,
+      "vm_loss": 0.2261
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.9421,
+      "step": 7568,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.6371,
+      "step": 7568,
+      "vm_loss": 0.1119
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.9167,
+      "step": 7568,
+      "vm_loss": 0.2672
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.7039,
+      "step": 7568,
+      "vm_loss": 0.2106
+    },
+    {
+      "epoch": 1.4568905358904636,
+      "lm_loss": 0.5965,
+      "step": 7568,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.457083042568039,
+      "grad_norm": 3.1967627215085153,
+      "learning_rate": 3.6202842857064657e-06,
+      "loss": 0.9753,
+      "step": 7569
+    },
+    {
+      "epoch": 1.4572755492456144,
+      "grad_norm": 3.0597970150074993,
+      "learning_rate": 3.6178836286167052e-06,
+      "loss": 0.9412,
+      "step": 7570
+    },
+    {
+      "epoch": 1.4574680559231898,
+      "grad_norm": 3.1929048121324164,
+      "learning_rate": 3.6154835919502996e-06,
+      "loss": 0.9467,
+      "step": 7571
+    },
+    {
+      "epoch": 1.4576605626007653,
+      "grad_norm": 3.159884695980742,
+      "learning_rate": 3.6130841759405776e-06,
+      "loss": 0.9349,
+      "step": 7572
+    },
+    {
+      "epoch": 1.4578530692783405,
+      "grad_norm": 3.3245797728228084,
+      "learning_rate": 3.6106853808207833e-06,
+      "loss": 0.9987,
+      "step": 7573
+    },
+    {
+      "epoch": 1.458045575955916,
+      "grad_norm": 3.1664880575091017,
+      "learning_rate": 3.6082872068241124e-06,
+      "loss": 0.9684,
+      "step": 7574
+    },
+    {
+      "epoch": 1.4582380826334913,
+      "grad_norm": 3.1596500132994243,
+      "learning_rate": 3.6058896541836986e-06,
+      "loss": 0.9627,
+      "step": 7575
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "grad_norm": 3.21400726079437,
+      "learning_rate": 3.6034927231326156e-06,
+      "loss": 0.9491,
+      "step": 7576
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 0.7674,
+      "step": 7576,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 0.784,
+      "step": 7576,
+      "vm_loss": 0.2127
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 1.1716,
+      "step": 7576,
+      "vm_loss": 0.0885
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 1.0539,
+      "step": 7576,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 1.1292,
+      "step": 7576,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 0.6197,
+      "step": 7576,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 0.5794,
+      "step": 7576,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 1.4584305893110667,
+      "lm_loss": 1.299,
+      "step": 7576,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 1.4586230959886421,
+      "grad_norm": 3.303567985812882,
+      "learning_rate": 3.6010964139038783e-06,
+      "loss": 0.9938,
+      "step": 7577
+    },
+    {
+      "epoch": 1.4588156026662173,
+      "grad_norm": 3.2887744292475727,
+      "learning_rate": 3.59870072673043e-06,
+      "loss": 0.9542,
+      "step": 7578
+    },
+    {
+      "epoch": 1.459008109343793,
+      "grad_norm": 3.392341811907046,
+      "learning_rate": 3.596305661845172e-06,
+      "loss": 1.0073,
+      "step": 7579
+    },
+    {
+      "epoch": 1.4592006160213682,
+      "grad_norm": 3.4059394382802135,
+      "learning_rate": 3.593911219480929e-06,
+      "loss": 1.0078,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4593931226989436,
+      "grad_norm": 3.325909839367532,
+      "learning_rate": 3.5915173998704733e-06,
+      "loss": 0.9616,
+      "step": 7581
+    },
+    {
+      "epoch": 1.459585629376519,
+      "grad_norm": 3.341954948429275,
+      "learning_rate": 3.5891242032465144e-06,
+      "loss": 0.9626,
+      "step": 7582
+    },
+    {
+      "epoch": 1.4597781360540945,
+      "grad_norm": 3.1673300061184744,
+      "learning_rate": 3.5867316298417053e-06,
+      "loss": 0.9053,
+      "step": 7583
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "grad_norm": 3.463879385629071,
+      "learning_rate": 3.5843396798886275e-06,
+      "loss": 1.0202,
+      "step": 7584
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.5899,
+      "step": 7584,
+      "vm_loss": 0.1
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.7933,
+      "step": 7584,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.5765,
+      "step": 7584,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 1.3014,
+      "step": 7584,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.5129,
+      "step": 7584,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.5223,
+      "step": 7584,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.6431,
+      "step": 7584,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.4599706427316699,
+      "lm_loss": 0.7816,
+      "step": 7584,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 1.460163149409245,
+      "grad_norm": 3.2615385854672314,
+      "learning_rate": 3.581948353619814e-06,
+      "loss": 0.9483,
+      "step": 7585
+    },
+    {
+      "epoch": 1.4603556560868205,
+      "grad_norm": 3.3058767381400593,
+      "learning_rate": 3.579557651267731e-06,
+      "loss": 0.9906,
+      "step": 7586
+    },
+    {
+      "epoch": 1.460548162764396,
+      "grad_norm": 3.2836430538495707,
+      "learning_rate": 3.5771675730647894e-06,
+      "loss": 0.9855,
+      "step": 7587
+    },
+    {
+      "epoch": 1.4607406694419713,
+      "grad_norm": 3.1257355895561467,
+      "learning_rate": 3.574778119243325e-06,
+      "loss": 0.9053,
+      "step": 7588
+    },
+    {
+      "epoch": 1.4609331761195468,
+      "grad_norm": 3.515096850020925,
+      "learning_rate": 3.572389290035636e-06,
+      "loss": 1.0862,
+      "step": 7589
+    },
+    {
+      "epoch": 1.461125682797122,
+      "grad_norm": 3.321891360331378,
+      "learning_rate": 3.5700010856739385e-06,
+      "loss": 0.9918,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4613181894746974,
+      "grad_norm": 3.29240862255625,
+      "learning_rate": 3.567613506390399e-06,
+      "loss": 0.9889,
+      "step": 7591
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "grad_norm": 3.2835252301405493,
+      "learning_rate": 3.5652265524171205e-06,
+      "loss": 0.9922,
+      "step": 7592
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.5149,
+      "step": 7592,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.9553,
+      "step": 7592,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.5854,
+      "step": 7592,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 1.1794,
+      "step": 7592,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 1.0496,
+      "step": 7592,
+      "vm_loss": 0.2439
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.7187,
+      "step": 7592,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.8927,
+      "step": 7592,
+      "vm_loss": 0.2793
+    },
+    {
+      "epoch": 1.4615106961522728,
+      "lm_loss": 0.6064,
+      "step": 7592,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.4617032028298482,
+      "grad_norm": 3.3377264045779764,
+      "learning_rate": 3.5628402239861504e-06,
+      "loss": 0.9804,
+      "step": 7593
+    },
+    {
+      "epoch": 1.4618957095074236,
+      "grad_norm": 3.222452797517878,
+      "learning_rate": 3.5604545213294616e-06,
+      "loss": 0.9852,
+      "step": 7594
+    },
+    {
+      "epoch": 1.4620882161849988,
+      "grad_norm": 3.4280855320824832,
+      "learning_rate": 3.5580694446789786e-06,
+      "loss": 1.0191,
+      "step": 7595
+    },
+    {
+      "epoch": 1.4622807228625743,
+      "grad_norm": 3.284918299654645,
+      "learning_rate": 3.5556849942665626e-06,
+      "loss": 0.9974,
+      "step": 7596
+    },
+    {
+      "epoch": 1.4624732295401497,
+      "grad_norm": 3.3171573473139335,
+      "learning_rate": 3.5533011703240127e-06,
+      "loss": 0.9669,
+      "step": 7597
+    },
+    {
+      "epoch": 1.462665736217725,
+      "grad_norm": 3.3113878570313173,
+      "learning_rate": 3.550917973083069e-06,
+      "loss": 0.9983,
+      "step": 7598
+    },
+    {
+      "epoch": 1.4628582428953005,
+      "grad_norm": 3.1279239343627494,
+      "learning_rate": 3.5485354027753992e-06,
+      "loss": 0.9523,
+      "step": 7599
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "grad_norm": 3.2349066112100706,
+      "learning_rate": 3.546153459632634e-06,
+      "loss": 0.9483,
+      "step": 7600
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.9693,
+      "step": 7600,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.8754,
+      "step": 7600,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.9397,
+      "step": 7600,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.9083,
+      "step": 7600,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.6628,
+      "step": 7600,
+      "vm_loss": 0.1603
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 1.4697,
+      "step": 7600,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.8152,
+      "step": 7600,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.4630507495728757,
+      "lm_loss": 0.9698,
+      "step": 7600,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 1.4632432562504512,
+      "grad_norm": 3.2978215667176762,
+      "learning_rate": 3.543772143886318e-06,
+      "loss": 0.9641,
+      "step": 7601
+    },
+    {
+      "epoch": 1.4634357629280266,
+      "grad_norm": 3.1678881781243673,
+      "learning_rate": 3.5413914557679486e-06,
+      "loss": 0.9476,
+      "step": 7602
+    },
+    {
+      "epoch": 1.463628269605602,
+      "grad_norm": 3.080024560280289,
+      "learning_rate": 3.539011395508959e-06,
+      "loss": 0.9382,
+      "step": 7603
+    },
+    {
+      "epoch": 1.4638207762831774,
+      "grad_norm": 3.1301046620278803,
+      "learning_rate": 3.5366319633407255e-06,
+      "loss": 0.905,
+      "step": 7604
+    },
+    {
+      "epoch": 1.4640132829607526,
+      "grad_norm": 3.2358544988563875,
+      "learning_rate": 3.5342531594945493e-06,
+      "loss": 0.9543,
+      "step": 7605
+    },
+    {
+      "epoch": 1.464205789638328,
+      "grad_norm": 3.1965287900257318,
+      "learning_rate": 3.531874984201694e-06,
+      "loss": 0.9262,
+      "step": 7606
+    },
+    {
+      "epoch": 1.4643982963159035,
+      "grad_norm": 3.278284516734381,
+      "learning_rate": 3.5294974376933377e-06,
+      "loss": 0.9521,
+      "step": 7607
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "grad_norm": 3.2343210218478275,
+      "learning_rate": 3.5271205202006163e-06,
+      "loss": 0.9504,
+      "step": 7608
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.6236,
+      "step": 7608,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.497,
+      "step": 7608,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.9825,
+      "step": 7608,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.8292,
+      "step": 7608,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 1.1448,
+      "step": 7608,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.9142,
+      "step": 7608,
+      "vm_loss": 0.1444
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 1.2976,
+      "step": 7608,
+      "vm_loss": 0.095
+    },
+    {
+      "epoch": 1.4645908029934789,
+      "lm_loss": 0.5753,
+      "step": 7608,
+      "vm_loss": 0.1146
+    },
+    {
+      "epoch": 1.4647833096710543,
+      "grad_norm": 3.3579104795701893,
+      "learning_rate": 3.5247442319545853e-06,
+      "loss": 0.9706,
+      "step": 7609
+    },
+    {
+      "epoch": 1.4649758163486295,
+      "grad_norm": 3.2936999429028813,
+      "learning_rate": 3.5223685731862657e-06,
+      "loss": 0.9457,
+      "step": 7610
+    },
+    {
+      "epoch": 1.465168323026205,
+      "grad_norm": 3.4391550431519478,
+      "learning_rate": 3.5199935441265887e-06,
+      "loss": 1.0147,
+      "step": 7611
+    },
+    {
+      "epoch": 1.4653608297037803,
+      "grad_norm": 3.344371873851089,
+      "learning_rate": 3.5176191450064445e-06,
+      "loss": 0.9753,
+      "step": 7612
+    },
+    {
+      "epoch": 1.4655533363813558,
+      "grad_norm": 3.3837372903693805,
+      "learning_rate": 3.5152453760566518e-06,
+      "loss": 0.9846,
+      "step": 7613
+    },
+    {
+      "epoch": 1.4657458430589312,
+      "grad_norm": 3.3644683039690526,
+      "learning_rate": 3.512872237507974e-06,
+      "loss": 0.9985,
+      "step": 7614
+    },
+    {
+      "epoch": 1.4659383497365064,
+      "grad_norm": 3.1706162100994115,
+      "learning_rate": 3.5104997295911115e-06,
+      "loss": 0.9482,
+      "step": 7615
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "grad_norm": 3.596283859658559,
+      "learning_rate": 3.508127852536698e-06,
+      "loss": 1.0849,
+      "step": 7616
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 1.3985,
+      "step": 7616,
+      "vm_loss": 0.1106
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.7794,
+      "step": 7616,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.6363,
+      "step": 7616,
+      "vm_loss": 0.126
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 1.1086,
+      "step": 7616,
+      "vm_loss": 0.1108
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.6449,
+      "step": 7616,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.6519,
+      "step": 7616,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.8471,
+      "step": 7616,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 1.4661308564140818,
+      "lm_loss": 0.6101,
+      "step": 7616,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 1.4663233630916572,
+      "grad_norm": 3.365089333801048,
+      "learning_rate": 3.5057566065753114e-06,
+      "loss": 1.0188,
+      "step": 7617
+    },
+    {
+      "epoch": 1.4665158697692326,
+      "grad_norm": 3.3045802593950837,
+      "learning_rate": 3.5033859919374693e-06,
+      "loss": 1.032,
+      "step": 7618
+    },
+    {
+      "epoch": 1.466708376446808,
+      "grad_norm": 3.241155900566468,
+      "learning_rate": 3.5010160088536238e-06,
+      "loss": 1.0014,
+      "step": 7619
+    },
+    {
+      "epoch": 1.4669008831243833,
+      "grad_norm": 3.1511822679533976,
+      "learning_rate": 3.4986466575541688e-06,
+      "loss": 0.9604,
+      "step": 7620
+    },
+    {
+      "epoch": 1.4670933898019587,
+      "grad_norm": 3.1535278090043253,
+      "learning_rate": 3.4962779382694387e-06,
+      "loss": 0.9499,
+      "step": 7621
+    },
+    {
+      "epoch": 1.4672858964795341,
+      "grad_norm": 3.0753821712555545,
+      "learning_rate": 3.4939098512296933e-06,
+      "loss": 0.9172,
+      "step": 7622
+    },
+    {
+      "epoch": 1.4674784031571095,
+      "grad_norm": 3.342031497366783,
+      "learning_rate": 3.491542396665155e-06,
+      "loss": 1.0126,
+      "step": 7623
+    },
+    {
+      "epoch": 1.467670909834685,
+      "grad_norm": 3.1576460273515217,
+      "learning_rate": 3.4891755748059587e-06,
+      "loss": 0.9586,
+      "step": 7624
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.8257,
+      "step": 7624,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.805,
+      "step": 7624,
+      "vm_loss": 0.2393
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.5289,
+      "step": 7624,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.9875,
+      "step": 7624,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 1.0995,
+      "step": 7624,
+      "vm_loss": 0.1553
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.4283,
+      "step": 7624,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.7987,
+      "step": 7624,
+      "vm_loss": 0.1296
+    },
+    {
+      "epoch": 1.467670909834685,
+      "lm_loss": 0.3783,
+      "step": 7624,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 1.4678634165122602,
+      "grad_norm": 3.098715100970303,
+      "learning_rate": 3.4868093858821982e-06,
+      "loss": 0.9012,
+      "step": 7625
+    },
+    {
+      "epoch": 1.4680559231898356,
+      "grad_norm": 3.340863438836121,
+      "learning_rate": 3.484443830123887e-06,
+      "loss": 1.0073,
+      "step": 7626
+    },
+    {
+      "epoch": 1.468248429867411,
+      "grad_norm": 3.255337529289207,
+      "learning_rate": 3.4820789077609997e-06,
+      "loss": 0.9847,
+      "step": 7627
+    },
+    {
+      "epoch": 1.4684409365449864,
+      "grad_norm": 3.2620871653989036,
+      "learning_rate": 3.4797146190234286e-06,
+      "loss": 1.0,
+      "step": 7628
+    },
+    {
+      "epoch": 1.4686334432225618,
+      "grad_norm": 3.32782763753346,
+      "learning_rate": 3.477350964141014e-06,
+      "loss": 0.9993,
+      "step": 7629
+    },
+    {
+      "epoch": 1.468825949900137,
+      "grad_norm": 3.453517657302616,
+      "learning_rate": 3.4749879433435364e-06,
+      "loss": 0.9681,
+      "step": 7630
+    },
+    {
+      "epoch": 1.4690184565777125,
+      "grad_norm": 3.314283208859271,
+      "learning_rate": 3.4726255568607125e-06,
+      "loss": 0.9771,
+      "step": 7631
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "grad_norm": 3.3479673159347128,
+      "learning_rate": 3.4702638049221902e-06,
+      "loss": 0.9991,
+      "step": 7632
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.6978,
+      "step": 7632,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.8667,
+      "step": 7632,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.6932,
+      "step": 7632,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.6685,
+      "step": 7632,
+      "vm_loss": 0.1283
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.9055,
+      "step": 7632,
+      "vm_loss": 0.1055
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.4773,
+      "step": 7632,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.4983,
+      "step": 7632,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 1.4692109632552879,
+      "lm_loss": 0.8448,
+      "step": 7632,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 1.4694034699328633,
+      "grad_norm": 3.399133316300144,
+      "learning_rate": 3.467902687757567e-06,
+      "loss": 0.9534,
+      "step": 7633
+    },
+    {
+      "epoch": 1.4695959766104387,
+      "grad_norm": 3.2852169695509814,
+      "learning_rate": 3.4655422055963715e-06,
+      "loss": 0.9406,
+      "step": 7634
+    },
+    {
+      "epoch": 1.469788483288014,
+      "grad_norm": 3.250399779790608,
+      "learning_rate": 3.4631823586680747e-06,
+      "loss": 0.9588,
+      "step": 7635
+    },
+    {
+      "epoch": 1.4699809899655896,
+      "grad_norm": 3.2291664040015764,
+      "learning_rate": 3.4608231472020815e-06,
+      "loss": 0.9892,
+      "step": 7636
+    },
+    {
+      "epoch": 1.4701734966431648,
+      "grad_norm": 3.2487218721689315,
+      "learning_rate": 3.4584645714277397e-06,
+      "loss": 1.028,
+      "step": 7637
+    },
+    {
+      "epoch": 1.4703660033207402,
+      "grad_norm": 3.147919480674612,
+      "learning_rate": 3.4561066315743365e-06,
+      "loss": 0.9763,
+      "step": 7638
+    },
+    {
+      "epoch": 1.4705585099983156,
+      "grad_norm": 3.148093332918957,
+      "learning_rate": 3.4537493278710853e-06,
+      "loss": 0.9365,
+      "step": 7639
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "grad_norm": 3.1778092093940082,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.9526,
+      "step": 7640
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 1.2307,
+      "step": 7640,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.7594,
+      "step": 7640,
+      "vm_loss": 0.1934
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.7765,
+      "step": 7640,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.7014,
+      "step": 7640,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.6204,
+      "step": 7640,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.6984,
+      "step": 7640,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 1.093,
+      "step": 7640,
+      "vm_loss": 0.1579
+    },
+    {
+      "epoch": 1.4707510166758908,
+      "lm_loss": 0.6794,
+      "step": 7640,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.4709435233534665,
+      "grad_norm": 3.106304167228322,
+      "learning_rate": 3.449036629831629e-06,
+      "loss": 0.9344,
+      "step": 7641
+    },
+    {
+      "epoch": 1.4711360300310417,
+      "grad_norm": 3.2077347146833985,
+      "learning_rate": 3.446681235953564e-06,
+      "loss": 0.97,
+      "step": 7642
+    },
+    {
+      "epoch": 1.471328536708617,
+      "grad_norm": 3.3466915520285894,
+      "learning_rate": 3.4443264791419164e-06,
+      "loss": 0.9722,
+      "step": 7643
+    },
+    {
+      "epoch": 1.4715210433861925,
+      "grad_norm": 3.164027558278825,
+      "learning_rate": 3.4419723596256126e-06,
+      "loss": 0.9772,
+      "step": 7644
+    },
+    {
+      "epoch": 1.471713550063768,
+      "grad_norm": 3.234882745388558,
+      "learning_rate": 3.439618877633495e-06,
+      "loss": 0.9574,
+      "step": 7645
+    },
+    {
+      "epoch": 1.4719060567413433,
+      "grad_norm": 3.133643979282079,
+      "learning_rate": 3.437266033394354e-06,
+      "loss": 0.934,
+      "step": 7646
+    },
+    {
+      "epoch": 1.4720985634189185,
+      "grad_norm": 3.540410827258322,
+      "learning_rate": 3.4349138271369163e-06,
+      "loss": 1.0199,
+      "step": 7647
+    },
+    {
+      "epoch": 1.472291070096494,
+      "grad_norm": 3.165299098636344,
+      "learning_rate": 3.43256225908985e-06,
+      "loss": 0.9181,
+      "step": 7648
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 1.2548,
+      "step": 7648,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 1.0031,
+      "step": 7648,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 1.1507,
+      "step": 7648,
+      "vm_loss": 0.0744
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 0.1779,
+      "step": 7648,
+      "vm_loss": 0.139
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 0.6426,
+      "step": 7648,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 0.5592,
+      "step": 7648,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 1.0454,
+      "step": 7648,
+      "vm_loss": 0.1086
+    },
+    {
+      "epoch": 1.472291070096494,
+      "lm_loss": 1.1815,
+      "step": 7648,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.4724835767740694,
+      "grad_norm": 3.4563873464189956,
+      "learning_rate": 3.4302113294817505e-06,
+      "loss": 1.0134,
+      "step": 7649
+    },
+    {
+      "epoch": 1.4726760834516448,
+      "grad_norm": 3.277669385941886,
+      "learning_rate": 3.4278610385411615e-06,
+      "loss": 0.9396,
+      "step": 7650
+    },
+    {
+      "epoch": 1.4728685901292202,
+      "grad_norm": 3.2130327667350516,
+      "learning_rate": 3.425511386496563e-06,
+      "loss": 0.9478,
+      "step": 7651
+    },
+    {
+      "epoch": 1.4730610968067954,
+      "grad_norm": 3.3036002545304752,
+      "learning_rate": 3.423162373576372e-06,
+      "loss": 0.9713,
+      "step": 7652
+    },
+    {
+      "epoch": 1.4732536034843708,
+      "grad_norm": 3.270411874849448,
+      "learning_rate": 3.4208140000089352e-06,
+      "loss": 0.9717,
+      "step": 7653
+    },
+    {
+      "epoch": 1.4734461101619463,
+      "grad_norm": 3.198418358837812,
+      "learning_rate": 3.4184662660225567e-06,
+      "loss": 0.9559,
+      "step": 7654
+    },
+    {
+      "epoch": 1.4736386168395217,
+      "grad_norm": 3.282354105423872,
+      "learning_rate": 3.4161191718454545e-06,
+      "loss": 0.9626,
+      "step": 7655
+    },
+    {
+      "epoch": 1.473831123517097,
+      "grad_norm": 3.3841784366276286,
+      "learning_rate": 3.4137727177058023e-06,
+      "loss": 1.0399,
+      "step": 7656
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.7541,
+      "step": 7656,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.7946,
+      "step": 7656,
+      "vm_loss": 0.1218
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.7296,
+      "step": 7656,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 1.2273,
+      "step": 7656,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.6341,
+      "step": 7656,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.5763,
+      "step": 7656,
+      "vm_loss": 0.2179
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.8286,
+      "step": 7656,
+      "vm_loss": 0.1612
+    },
+    {
+      "epoch": 1.473831123517097,
+      "lm_loss": 0.7555,
+      "step": 7656,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.4740236301946723,
+      "grad_norm": 3.1183530273437,
+      "learning_rate": 3.411426903831705e-06,
+      "loss": 0.9342,
+      "step": 7657
+    },
+    {
+      "epoch": 1.4742161368722477,
+      "grad_norm": 3.2607045404943658,
+      "learning_rate": 3.409081730451205e-06,
+      "loss": 0.9838,
+      "step": 7658
+    },
+    {
+      "epoch": 1.4744086435498232,
+      "grad_norm": 3.2481865915879067,
+      "learning_rate": 3.406737197792286e-06,
+      "loss": 0.9636,
+      "step": 7659
+    },
+    {
+      "epoch": 1.4746011502273986,
+      "grad_norm": 3.153801840715601,
+      "learning_rate": 3.4043933060828606e-06,
+      "loss": 0.9471,
+      "step": 7660
+    },
+    {
+      "epoch": 1.474793656904974,
+      "grad_norm": 3.268611465633069,
+      "learning_rate": 3.402050055550787e-06,
+      "loss": 1.0125,
+      "step": 7661
+    },
+    {
+      "epoch": 1.4749861635825492,
+      "grad_norm": 3.20372763783119,
+      "learning_rate": 3.399707446423862e-06,
+      "loss": 0.9685,
+      "step": 7662
+    },
+    {
+      "epoch": 1.4751786702601246,
+      "grad_norm": 3.199487868569656,
+      "learning_rate": 3.3973654789298162e-06,
+      "loss": 0.9174,
+      "step": 7663
+    },
+    {
+      "epoch": 1.4753711769377,
+      "grad_norm": 3.2885956979602775,
+      "learning_rate": 3.395024153296317e-06,
+      "loss": 0.9928,
+      "step": 7664
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 0.7433,
+      "step": 7664,
+      "vm_loss": 0.0994
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 0.6486,
+      "step": 7664,
+      "vm_loss": 0.3052
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 0.6649,
+      "step": 7664,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 1.3084,
+      "step": 7664,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 0.7433,
+      "step": 7664,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 1.3535,
+      "step": 7664,
+      "vm_loss": 0.1837
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 1.1201,
+      "step": 7664,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 1.4753711769377,
+      "lm_loss": 0.9644,
+      "step": 7664,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 1.4755636836152755,
+      "grad_norm": 3.2093659568400903,
+      "learning_rate": 3.392683469750976e-06,
+      "loss": 0.9625,
+      "step": 7665
+    },
+    {
+      "epoch": 1.4757561902928509,
+      "grad_norm": 3.2595373071337677,
+      "learning_rate": 3.390343428521331e-06,
+      "loss": 0.9584,
+      "step": 7666
+    },
+    {
+      "epoch": 1.475948696970426,
+      "grad_norm": 3.1292423262295874,
+      "learning_rate": 3.388004029834867e-06,
+      "loss": 0.9294,
+      "step": 7667
+    },
+    {
+      "epoch": 1.4761412036480015,
+      "grad_norm": 3.314468384780081,
+      "learning_rate": 3.3856652739190034e-06,
+      "loss": 0.9403,
+      "step": 7668
+    },
+    {
+      "epoch": 1.476333710325577,
+      "grad_norm": 3.3471634219320285,
+      "learning_rate": 3.3833271610011e-06,
+      "loss": 0.9793,
+      "step": 7669
+    },
+    {
+      "epoch": 1.4765262170031523,
+      "grad_norm": 3.319087740514296,
+      "learning_rate": 3.380989691308442e-06,
+      "loss": 0.9945,
+      "step": 7670
+    },
+    {
+      "epoch": 1.4767187236807278,
+      "grad_norm": 3.1405822665191283,
+      "learning_rate": 3.378652865068275e-06,
+      "loss": 0.9634,
+      "step": 7671
+    },
+    {
+      "epoch": 1.476911230358303,
+      "grad_norm": 3.220422842074132,
+      "learning_rate": 3.3763166825077576e-06,
+      "loss": 0.9852,
+      "step": 7672
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.7613,
+      "step": 7672,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.4764,
+      "step": 7672,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.814,
+      "step": 7672,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.6232,
+      "step": 7672,
+      "vm_loss": 0.2044
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.873,
+      "step": 7672,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.9057,
+      "step": 7672,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.7622,
+      "step": 7672,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.476911230358303,
+      "lm_loss": 0.655,
+      "step": 7672,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.4771037370358784,
+      "grad_norm": 3.2286896358689208,
+      "learning_rate": 3.3739811438540037e-06,
+      "loss": 0.9723,
+      "step": 7673
+    },
+    {
+      "epoch": 1.4772962437134538,
+      "grad_norm": 3.580271965507969,
+      "learning_rate": 3.3716462493340473e-06,
+      "loss": 1.0689,
+      "step": 7674
+    },
+    {
+      "epoch": 1.4774887503910292,
+      "grad_norm": 3.229096785538811,
+      "learning_rate": 3.3693119991748836e-06,
+      "loss": 0.9519,
+      "step": 7675
+    },
+    {
+      "epoch": 1.4776812570686046,
+      "grad_norm": 3.266786355417555,
+      "learning_rate": 3.3669783936034216e-06,
+      "loss": 0.9196,
+      "step": 7676
+    },
+    {
+      "epoch": 1.4778737637461798,
+      "grad_norm": 3.261867454546461,
+      "learning_rate": 3.3646454328465205e-06,
+      "loss": 0.9656,
+      "step": 7677
+    },
+    {
+      "epoch": 1.4780662704237553,
+      "grad_norm": 3.266164914651739,
+      "learning_rate": 3.3623131171309755e-06,
+      "loss": 0.9595,
+      "step": 7678
+    },
+    {
+      "epoch": 1.4782587771013307,
+      "grad_norm": 3.331630261545895,
+      "learning_rate": 3.3599814466835156e-06,
+      "loss": 0.9914,
+      "step": 7679
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "grad_norm": 3.140015401798985,
+      "learning_rate": 3.357650421730809e-06,
+      "loss": 0.9362,
+      "step": 7680
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.7588,
+      "step": 7680,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.4668,
+      "step": 7680,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.7358,
+      "step": 7680,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.9172,
+      "step": 7680,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.9064,
+      "step": 7680,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 1.0094,
+      "step": 7680,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 0.8003,
+      "step": 7680,
+      "vm_loss": 0.1934
+    },
+    {
+      "epoch": 1.4784512837789061,
+      "lm_loss": 1.0929,
+      "step": 7680,
+      "vm_loss": 0.1281
+    },
+    {
+      "epoch": 1.4786437904564815,
+      "grad_norm": 3.298129711406833,
+      "learning_rate": 3.3553200424994627e-06,
+      "loss": 0.984,
+      "step": 7681
+    },
+    {
+      "epoch": 1.4788362971340567,
+      "grad_norm": 3.168720645909617,
+      "learning_rate": 3.352990309216022e-06,
+      "loss": 0.9375,
+      "step": 7682
+    },
+    {
+      "epoch": 1.4790288038116322,
+      "grad_norm": 3.33303608005996,
+      "learning_rate": 3.350661222106959e-06,
+      "loss": 0.9492,
+      "step": 7683
+    },
+    {
+      "epoch": 1.4792213104892076,
+      "grad_norm": 3.3385378614676764,
+      "learning_rate": 3.3483327813986967e-06,
+      "loss": 0.9789,
+      "step": 7684
+    },
+    {
+      "epoch": 1.479413817166783,
+      "grad_norm": 3.1863423530664905,
+      "learning_rate": 3.346004987317586e-06,
+      "loss": 0.8918,
+      "step": 7685
+    },
+    {
+      "epoch": 1.4796063238443584,
+      "grad_norm": 3.2570743910180386,
+      "learning_rate": 3.343677840089924e-06,
+      "loss": 0.9321,
+      "step": 7686
+    },
+    {
+      "epoch": 1.4797988305219336,
+      "grad_norm": 3.337914097809864,
+      "learning_rate": 3.3413513399419295e-06,
+      "loss": 0.9712,
+      "step": 7687
+    },
+    {
+      "epoch": 1.479991337199509,
+      "grad_norm": 3.3116611752191747,
+      "learning_rate": 3.3390254870997796e-06,
+      "loss": 0.9216,
+      "step": 7688
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 1.1127,
+      "step": 7688,
+      "vm_loss": 0.2356
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.7878,
+      "step": 7688,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.7703,
+      "step": 7688,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.8123,
+      "step": 7688,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 1.2013,
+      "step": 7688,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.9865,
+      "step": 7688,
+      "vm_loss": 0.1517
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.8533,
+      "step": 7688,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 1.479991337199509,
+      "lm_loss": 0.4776,
+      "step": 7688,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.4801838438770845,
+      "grad_norm": 3.3403632917325496,
+      "learning_rate": 3.336700281789568e-06,
+      "loss": 0.9764,
+      "step": 7689
+    },
+    {
+      "epoch": 1.4803763505546599,
+      "grad_norm": 3.3808395436349494,
+      "learning_rate": 3.3343757242373397e-06,
+      "loss": 1.0189,
+      "step": 7690
+    },
+    {
+      "epoch": 1.4805688572322353,
+      "grad_norm": 3.1848900222907552,
+      "learning_rate": 3.332051814669064e-06,
+      "loss": 0.9232,
+      "step": 7691
+    },
+    {
+      "epoch": 1.4807613639098105,
+      "grad_norm": 3.3423341415135206,
+      "learning_rate": 3.3297285533106648e-06,
+      "loss": 0.9983,
+      "step": 7692
+    },
+    {
+      "epoch": 1.480953870587386,
+      "grad_norm": 3.1437843463376605,
+      "learning_rate": 3.3274059403879865e-06,
+      "loss": 0.95,
+      "step": 7693
+    },
+    {
+      "epoch": 1.4811463772649613,
+      "grad_norm": 3.31769356295955,
+      "learning_rate": 3.3250839761268174e-06,
+      "loss": 0.9674,
+      "step": 7694
+    },
+    {
+      "epoch": 1.4813388839425368,
+      "grad_norm": 3.428510625370614,
+      "learning_rate": 3.322762660752883e-06,
+      "loss": 0.9887,
+      "step": 7695
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "grad_norm": 3.3790686443996893,
+      "learning_rate": 3.3204419944918476e-06,
+      "loss": 0.9952,
+      "step": 7696
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.5251,
+      "step": 7696,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.8638,
+      "step": 7696,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.4394,
+      "step": 7696,
+      "vm_loss": 0.1069
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.4837,
+      "step": 7696,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.5822,
+      "step": 7696,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.5042,
+      "step": 7696,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.7243,
+      "step": 7696,
+      "vm_loss": 0.2335
+    },
+    {
+      "epoch": 1.4815313906201122,
+      "lm_loss": 0.816,
+      "step": 7696,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.4817238972976874,
+      "grad_norm": 3.0705739789682145,
+      "learning_rate": 3.3181219775693006e-06,
+      "loss": 0.9651,
+      "step": 7697
+    },
+    {
+      "epoch": 1.481916403975263,
+      "grad_norm": 3.4072719742980704,
+      "learning_rate": 3.315802610210791e-06,
+      "loss": 1.0181,
+      "step": 7698
+    },
+    {
+      "epoch": 1.4821089106528382,
+      "grad_norm": 3.2400219531711967,
+      "learning_rate": 3.313483892641779e-06,
+      "loss": 0.957,
+      "step": 7699
+    },
+    {
+      "epoch": 1.4823014173304137,
+      "grad_norm": 3.1793623976256797,
+      "learning_rate": 3.3111658250876787e-06,
+      "loss": 0.9297,
+      "step": 7700
+    },
+    {
+      "epoch": 1.482493924007989,
+      "grad_norm": 3.2544204417275795,
+      "learning_rate": 3.3088484077738357e-06,
+      "loss": 0.9594,
+      "step": 7701
+    },
+    {
+      "epoch": 1.4826864306855643,
+      "grad_norm": 3.300422945692872,
+      "learning_rate": 3.306531640925532e-06,
+      "loss": 0.9762,
+      "step": 7702
+    },
+    {
+      "epoch": 1.48287893736314,
+      "grad_norm": 3.1188481773430348,
+      "learning_rate": 3.3042155247679896e-06,
+      "loss": 0.9318,
+      "step": 7703
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "grad_norm": 3.1249433926566157,
+      "learning_rate": 3.3019000595263573e-06,
+      "loss": 0.92,
+      "step": 7704
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 1.2974,
+      "step": 7704,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.9343,
+      "step": 7704,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.6972,
+      "step": 7704,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.9308,
+      "step": 7704,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.7107,
+      "step": 7704,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.7054,
+      "step": 7704,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.8292,
+      "step": 7704,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 1.4830714440407151,
+      "lm_loss": 0.9217,
+      "step": 7704,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.4832639507182905,
+      "grad_norm": 3.428671616486981,
+      "learning_rate": 3.299585245425738e-06,
+      "loss": 1.0303,
+      "step": 7705
+    },
+    {
+      "epoch": 1.483456457395866,
+      "grad_norm": 3.280696581633513,
+      "learning_rate": 3.297271082691155e-06,
+      "loss": 0.9336,
+      "step": 7706
+    },
+    {
+      "epoch": 1.4836489640734414,
+      "grad_norm": 3.0762891421045513,
+      "learning_rate": 3.294957571547577e-06,
+      "loss": 0.8753,
+      "step": 7707
+    },
+    {
+      "epoch": 1.4838414707510168,
+      "grad_norm": 3.360564615643978,
+      "learning_rate": 3.2926447122199e-06,
+      "loss": 0.9878,
+      "step": 7708
+    },
+    {
+      "epoch": 1.484033977428592,
+      "grad_norm": 3.460083945446107,
+      "learning_rate": 3.290332504932975e-06,
+      "loss": 1.0146,
+      "step": 7709
+    },
+    {
+      "epoch": 1.4842264841061674,
+      "grad_norm": 3.1599800146266896,
+      "learning_rate": 3.288020949911569e-06,
+      "loss": 0.9325,
+      "step": 7710
+    },
+    {
+      "epoch": 1.4844189907837428,
+      "grad_norm": 3.4002943891608135,
+      "learning_rate": 3.285710047380398e-06,
+      "loss": 1.0001,
+      "step": 7711
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "grad_norm": 3.1981818334072885,
+      "learning_rate": 3.283399797564112e-06,
+      "loss": 0.9598,
+      "step": 7712
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.9238,
+      "step": 7712,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.6226,
+      "step": 7712,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.686,
+      "step": 7712,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.8612,
+      "step": 7712,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.9632,
+      "step": 7712,
+      "vm_loss": 0.2409
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.5601,
+      "step": 7712,
+      "vm_loss": 0.2169
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.7838,
+      "step": 7712,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 1.4846114974613183,
+      "lm_loss": 0.7952,
+      "step": 7712,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 1.4848040041388937,
+      "grad_norm": 3.202405742173062,
+      "learning_rate": 3.2810902006872993e-06,
+      "loss": 0.9525,
+      "step": 7713
+    },
+    {
+      "epoch": 1.4849965108164689,
+      "grad_norm": 3.259381340023161,
+      "learning_rate": 3.2787812569744716e-06,
+      "loss": 0.9509,
+      "step": 7714
+    },
+    {
+      "epoch": 1.4851890174940443,
+      "grad_norm": 3.254614066979019,
+      "learning_rate": 3.2764729666501027e-06,
+      "loss": 0.9457,
+      "step": 7715
+    },
+    {
+      "epoch": 1.4853815241716197,
+      "grad_norm": 3.350859717617728,
+      "learning_rate": 3.2741653299385767e-06,
+      "loss": 0.9589,
+      "step": 7716
+    },
+    {
+      "epoch": 1.4855740308491951,
+      "grad_norm": 3.1914184042538354,
+      "learning_rate": 3.271858347064233e-06,
+      "loss": 0.8791,
+      "step": 7717
+    },
+    {
+      "epoch": 1.4857665375267706,
+      "grad_norm": 3.2097459229258307,
+      "learning_rate": 3.2695520182513285e-06,
+      "loss": 0.9328,
+      "step": 7718
+    },
+    {
+      "epoch": 1.4859590442043458,
+      "grad_norm": 3.2882380193893472,
+      "learning_rate": 3.267246343724082e-06,
+      "loss": 0.9362,
+      "step": 7719
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "grad_norm": 3.338042557480579,
+      "learning_rate": 3.2649413237066253e-06,
+      "loss": 0.9711,
+      "step": 7720
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.6465,
+      "step": 7720,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.4273,
+      "step": 7720,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.7071,
+      "step": 7720,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.5044,
+      "step": 7720,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.8187,
+      "step": 7720,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 0.7489,
+      "step": 7720,
+      "vm_loss": 0.1748
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 1.1517,
+      "step": 7720,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 1.4861515508819212,
+      "lm_loss": 1.1211,
+      "step": 7720,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 1.4863440575594966,
+      "grad_norm": 3.327634152280964,
+      "learning_rate": 3.2626369584230387e-06,
+      "loss": 0.9911,
+      "step": 7721
+    },
+    {
+      "epoch": 1.486536564237072,
+      "grad_norm": 3.361983593182565,
+      "learning_rate": 3.260333248097335e-06,
+      "loss": 0.9757,
+      "step": 7722
+    },
+    {
+      "epoch": 1.4867290709146475,
+      "grad_norm": 3.3497177201880635,
+      "learning_rate": 3.2580301929534654e-06,
+      "loss": 0.9976,
+      "step": 7723
+    },
+    {
+      "epoch": 1.4869215775922227,
+      "grad_norm": 3.250072082007862,
+      "learning_rate": 3.25572779321532e-06,
+      "loss": 0.9274,
+      "step": 7724
+    },
+    {
+      "epoch": 1.487114084269798,
+      "grad_norm": 3.2983804550042017,
+      "learning_rate": 3.253426049106709e-06,
+      "loss": 0.9552,
+      "step": 7725
+    },
+    {
+      "epoch": 1.4873065909473735,
+      "grad_norm": 3.3550800568344843,
+      "learning_rate": 3.2511249608514083e-06,
+      "loss": 0.9635,
+      "step": 7726
+    },
+    {
+      "epoch": 1.487499097624949,
+      "grad_norm": 3.210736913768476,
+      "learning_rate": 3.2488245286731013e-06,
+      "loss": 0.9523,
+      "step": 7727
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "grad_norm": 3.282491789538895,
+      "learning_rate": 3.2465247527954226e-06,
+      "loss": 0.9549,
+      "step": 7728
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.5348,
+      "step": 7728,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.8002,
+      "step": 7728,
+      "vm_loss": 0.097
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.7222,
+      "step": 7728,
+      "vm_loss": 0.231
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.7606,
+      "step": 7728,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.7557,
+      "step": 7728,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.6734,
+      "step": 7728,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.8212,
+      "step": 7728,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.4876916043025243,
+      "lm_loss": 0.869,
+      "step": 7728,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 1.4878841109800995,
+      "grad_norm": 3.1761805424602456,
+      "learning_rate": 3.2442256334419398e-06,
+      "loss": 0.9184,
+      "step": 7729
+    },
+    {
+      "epoch": 1.488076617657675,
+      "grad_norm": 3.179179281159974,
+      "learning_rate": 3.241927170836161e-06,
+      "loss": 0.9571,
+      "step": 7730
+    },
+    {
+      "epoch": 1.4882691243352504,
+      "grad_norm": 3.1717681626726977,
+      "learning_rate": 3.2396293652015155e-06,
+      "loss": 0.9527,
+      "step": 7731
+    },
+    {
+      "epoch": 1.4884616310128258,
+      "grad_norm": 3.201785275545528,
+      "learning_rate": 3.237332216761392e-06,
+      "loss": 0.918,
+      "step": 7732
+    },
+    {
+      "epoch": 1.4886541376904012,
+      "grad_norm": 3.246982176686414,
+      "learning_rate": 3.235035725739094e-06,
+      "loss": 0.9279,
+      "step": 7733
+    },
+    {
+      "epoch": 1.4888466443679764,
+      "grad_norm": 3.2145636026407387,
+      "learning_rate": 3.2327398923578755e-06,
+      "loss": 0.9563,
+      "step": 7734
+    },
+    {
+      "epoch": 1.4890391510455518,
+      "grad_norm": 3.2755514777636847,
+      "learning_rate": 3.2304447168409115e-06,
+      "loss": 0.9207,
+      "step": 7735
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "grad_norm": 3.2709106612751224,
+      "learning_rate": 3.2281501994113363e-06,
+      "loss": 0.9221,
+      "step": 7736
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.7205,
+      "step": 7736,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.7752,
+      "step": 7736,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.6643,
+      "step": 7736,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.6819,
+      "step": 7736,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.8126,
+      "step": 7736,
+      "vm_loss": 0.1938
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 1.0848,
+      "step": 7736,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 0.677,
+      "step": 7736,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 1.4892316577231273,
+      "lm_loss": 1.0783,
+      "step": 7736,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 1.4894241644007027,
+      "grad_norm": 3.327641897128104,
+      "learning_rate": 3.225856340292195e-06,
+      "loss": 0.9833,
+      "step": 7737
+    },
+    {
+      "epoch": 1.489616671078278,
+      "grad_norm": 3.312587666576186,
+      "learning_rate": 3.2235631397064837e-06,
+      "loss": 0.9685,
+      "step": 7738
+    },
+    {
+      "epoch": 1.4898091777558533,
+      "grad_norm": 3.3874345900416296,
+      "learning_rate": 3.2212705978771318e-06,
+      "loss": 0.964,
+      "step": 7739
+    },
+    {
+      "epoch": 1.4900016844334287,
+      "grad_norm": 3.213508811423855,
+      "learning_rate": 3.218978715027006e-06,
+      "loss": 0.9759,
+      "step": 7740
+    },
+    {
+      "epoch": 1.4901941911110042,
+      "grad_norm": 3.1564147583199014,
+      "learning_rate": 3.2166874913789004e-06,
+      "loss": 0.9657,
+      "step": 7741
+    },
+    {
+      "epoch": 1.4903866977885796,
+      "grad_norm": 3.3153738830979163,
+      "learning_rate": 3.214396927155555e-06,
+      "loss": 0.9451,
+      "step": 7742
+    },
+    {
+      "epoch": 1.490579204466155,
+      "grad_norm": 3.360377429039677,
+      "learning_rate": 3.2121070225796416e-06,
+      "loss": 0.9449,
+      "step": 7743
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "grad_norm": 3.245010344793086,
+      "learning_rate": 3.209817777873767e-06,
+      "loss": 0.9357,
+      "step": 7744
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.6168,
+      "step": 7744,
+      "vm_loss": 0.1949
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.4935,
+      "step": 7744,
+      "vm_loss": 0.2001
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.7921,
+      "step": 7744,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 1.4087,
+      "step": 7744,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.7285,
+      "step": 7744,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.9514,
+      "step": 7744,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.6106,
+      "step": 7744,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 1.4907717111437302,
+      "lm_loss": 0.7972,
+      "step": 7744,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.4909642178213056,
+      "grad_norm": 3.096542702836715,
+      "learning_rate": 3.2075291932604767e-06,
+      "loss": 0.9184,
+      "step": 7745
+    },
+    {
+      "epoch": 1.491156724498881,
+      "grad_norm": 3.2724392160972537,
+      "learning_rate": 3.20524126896225e-06,
+      "loss": 0.9962,
+      "step": 7746
+    },
+    {
+      "epoch": 1.4913492311764565,
+      "grad_norm": 3.1330953220531557,
+      "learning_rate": 3.202954005201506e-06,
+      "loss": 0.9348,
+      "step": 7747
+    },
+    {
+      "epoch": 1.4915417378540319,
+      "grad_norm": 3.144468917724164,
+      "learning_rate": 3.2006674022005857e-06,
+      "loss": 0.9417,
+      "step": 7748
+    },
+    {
+      "epoch": 1.491734244531607,
+      "grad_norm": 3.4473696949701855,
+      "learning_rate": 3.1983814601817888e-06,
+      "loss": 1.0588,
+      "step": 7749
+    },
+    {
+      "epoch": 1.4919267512091825,
+      "grad_norm": 3.244718457392763,
+      "learning_rate": 3.1960961793673285e-06,
+      "loss": 0.9457,
+      "step": 7750
+    },
+    {
+      "epoch": 1.492119257886758,
+      "grad_norm": 3.190004355305222,
+      "learning_rate": 3.1938115599793708e-06,
+      "loss": 0.9411,
+      "step": 7751
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "grad_norm": 3.0831332085064496,
+      "learning_rate": 3.19152760224e-06,
+      "loss": 0.9433,
+      "step": 7752
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.9243,
+      "step": 7752,
+      "vm_loss": 0.118
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.7105,
+      "step": 7752,
+      "vm_loss": 0.165
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.7414,
+      "step": 7752,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.8962,
+      "step": 7752,
+      "vm_loss": 0.177
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.7186,
+      "step": 7752,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.8059,
+      "step": 7752,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.5081,
+      "step": 7752,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.4923117645643333,
+      "lm_loss": 0.8889,
+      "step": 7752,
+      "vm_loss": 0.0949
+    },
+    {
+      "epoch": 1.4925042712419088,
+      "grad_norm": 3.2620693381317785,
+      "learning_rate": 3.1892443063712574e-06,
+      "loss": 0.9191,
+      "step": 7753
+    },
+    {
+      "epoch": 1.492696777919484,
+      "grad_norm": 3.2102291524560664,
+      "learning_rate": 3.1869616725951012e-06,
+      "loss": 0.9172,
+      "step": 7754
+    },
+    {
+      "epoch": 1.4928892845970594,
+      "grad_norm": 3.406145037898015,
+      "learning_rate": 3.184679701133434e-06,
+      "loss": 0.9816,
+      "step": 7755
+    },
+    {
+      "epoch": 1.4930817912746348,
+      "grad_norm": 3.2968253045028515,
+      "learning_rate": 3.1823983922080926e-06,
+      "loss": 0.9911,
+      "step": 7756
+    },
+    {
+      "epoch": 1.4932742979522102,
+      "grad_norm": 3.302648899155252,
+      "learning_rate": 3.1801177460408537e-06,
+      "loss": 0.9968,
+      "step": 7757
+    },
+    {
+      "epoch": 1.4934668046297856,
+      "grad_norm": 3.3309498232761885,
+      "learning_rate": 3.1778377628534194e-06,
+      "loss": 0.9394,
+      "step": 7758
+    },
+    {
+      "epoch": 1.4936593113073608,
+      "grad_norm": 3.2072188884160258,
+      "learning_rate": 3.1755584428674345e-06,
+      "loss": 0.9078,
+      "step": 7759
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "grad_norm": 3.4244721701795333,
+      "learning_rate": 3.1732797863044795e-06,
+      "loss": 1.0008,
+      "step": 7760
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.9631,
+      "step": 7760,
+      "vm_loss": 0.1686
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.6483,
+      "step": 7760,
+      "vm_loss": 0.19
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.599,
+      "step": 7760,
+      "vm_loss": 0.2073
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.936,
+      "step": 7760,
+      "vm_loss": 0.112
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.9048,
+      "step": 7760,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 0.6556,
+      "step": 7760,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 1.3135,
+      "step": 7760,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 1.4938518179849363,
+      "lm_loss": 1.4535,
+      "step": 7760,
+      "vm_loss": 0.193
+    },
+    {
+      "epoch": 1.4940443246625117,
+      "grad_norm": 3.531376948022944,
+      "learning_rate": 3.171001793386068e-06,
+      "loss": 1.0326,
+      "step": 7761
+    },
+    {
+      "epoch": 1.4942368313400871,
+      "grad_norm": 3.2461755000387478,
+      "learning_rate": 3.1687244643336514e-06,
+      "loss": 0.9529,
+      "step": 7762
+    },
+    {
+      "epoch": 1.4944293380176625,
+      "grad_norm": 3.210999573322285,
+      "learning_rate": 3.166447799368617e-06,
+      "loss": 0.9209,
+      "step": 7763
+    },
+    {
+      "epoch": 1.4946218446952377,
+      "grad_norm": 3.3313638551805944,
+      "learning_rate": 3.16417179871228e-06,
+      "loss": 0.9788,
+      "step": 7764
+    },
+    {
+      "epoch": 1.4948143513728134,
+      "grad_norm": 3.280142236807375,
+      "learning_rate": 3.1618964625859006e-06,
+      "loss": 0.9403,
+      "step": 7765
+    },
+    {
+      "epoch": 1.4950068580503886,
+      "grad_norm": 3.3494071346407694,
+      "learning_rate": 3.1596217912106687e-06,
+      "loss": 0.969,
+      "step": 7766
+    },
+    {
+      "epoch": 1.495199364727964,
+      "grad_norm": 3.1817418460339617,
+      "learning_rate": 3.157347784807714e-06,
+      "loss": 0.9575,
+      "step": 7767
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "grad_norm": 3.2022246116984205,
+      "learning_rate": 3.1550744435980995e-06,
+      "loss": 0.9762,
+      "step": 7768
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 1.0008,
+      "step": 7768,
+      "vm_loss": 0.12
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.553,
+      "step": 7768,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.4216,
+      "step": 7768,
+      "vm_loss": 0.2159
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.5774,
+      "step": 7768,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.6663,
+      "step": 7768,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.8418,
+      "step": 7768,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.7475,
+      "step": 7768,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.4953918714055394,
+      "lm_loss": 0.701,
+      "step": 7768,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.4955843780831148,
+      "grad_norm": 3.131577978457707,
+      "learning_rate": 3.1528017678028156e-06,
+      "loss": 0.9352,
+      "step": 7769
+    },
+    {
+      "epoch": 1.4957768847606903,
+      "grad_norm": 3.110431502337773,
+      "learning_rate": 3.1505297576428074e-06,
+      "loss": 0.935,
+      "step": 7770
+    },
+    {
+      "epoch": 1.4959693914382655,
+      "grad_norm": 3.081872289025915,
+      "learning_rate": 3.148258413338935e-06,
+      "loss": 0.8718,
+      "step": 7771
+    },
+    {
+      "epoch": 1.4961618981158409,
+      "grad_norm": 3.1016206680483887,
+      "learning_rate": 3.145987735112003e-06,
+      "loss": 0.9261,
+      "step": 7772
+    },
+    {
+      "epoch": 1.4963544047934163,
+      "grad_norm": 3.230725527636669,
+      "learning_rate": 3.143717723182751e-06,
+      "loss": 0.9083,
+      "step": 7773
+    },
+    {
+      "epoch": 1.4965469114709917,
+      "grad_norm": 3.4636176251919806,
+      "learning_rate": 3.1414483777718585e-06,
+      "loss": 1.0642,
+      "step": 7774
+    },
+    {
+      "epoch": 1.4967394181485671,
+      "grad_norm": 3.3413270290674393,
+      "learning_rate": 3.139179699099926e-06,
+      "loss": 0.9393,
+      "step": 7775
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "grad_norm": 3.3708650824461834,
+      "learning_rate": 3.136911687387503e-06,
+      "loss": 0.981,
+      "step": 7776
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.6159,
+      "step": 7776,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.8462,
+      "step": 7776,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.4944,
+      "step": 7776,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 1.0182,
+      "step": 7776,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.8989,
+      "step": 7776,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.739,
+      "step": 7776,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.8847,
+      "step": 7776,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 1.4969319248261423,
+      "lm_loss": 0.7906,
+      "step": 7776,
+      "vm_loss": 0.1956
+    },
+    {
+      "epoch": 1.4971244315037178,
+      "grad_norm": 3.417335662213305,
+      "learning_rate": 3.1346443428550676e-06,
+      "loss": 0.9921,
+      "step": 7777
+    },
+    {
+      "epoch": 1.4973169381812932,
+      "grad_norm": 3.2780290555945455,
+      "learning_rate": 3.132377665723039e-06,
+      "loss": 0.9345,
+      "step": 7778
+    },
+    {
+      "epoch": 1.4975094448588686,
+      "grad_norm": 3.0854355870367742,
+      "learning_rate": 3.1301116562117572e-06,
+      "loss": 0.8491,
+      "step": 7779
+    },
+    {
+      "epoch": 1.497701951536444,
+      "grad_norm": 3.2628778850439533,
+      "learning_rate": 3.1278463145415205e-06,
+      "loss": 0.9483,
+      "step": 7780
+    },
+    {
+      "epoch": 1.4978944582140192,
+      "grad_norm": 3.3944971265147657,
+      "learning_rate": 3.12558164093254e-06,
+      "loss": 0.9707,
+      "step": 7781
+    },
+    {
+      "epoch": 1.4980869648915947,
+      "grad_norm": 3.4057424751099905,
+      "learning_rate": 3.1233176356049733e-06,
+      "loss": 0.9585,
+      "step": 7782
+    },
+    {
+      "epoch": 1.49827947156917,
+      "grad_norm": 3.3692897701341895,
+      "learning_rate": 3.1210542987789104e-06,
+      "loss": 0.9812,
+      "step": 7783
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "grad_norm": 3.360540053812142,
+      "learning_rate": 3.118791630674378e-06,
+      "loss": 0.9934,
+      "step": 7784
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.8223,
+      "step": 7784,
+      "vm_loss": 0.1722
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.6921,
+      "step": 7784,
+      "vm_loss": 0.2364
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.7079,
+      "step": 7784,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 1.224,
+      "step": 7784,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.9882,
+      "step": 7784,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.4192,
+      "step": 7784,
+      "vm_loss": 0.1352
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.5468,
+      "step": 7784,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 1.4984719782467455,
+      "lm_loss": 0.957,
+      "step": 7784,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 1.498664484924321,
+      "grad_norm": 3.2540458010806375,
+      "learning_rate": 3.116529631511338e-06,
+      "loss": 0.9763,
+      "step": 7785
+    },
+    {
+      "epoch": 1.4988569916018961,
+      "grad_norm": 3.388813406527751,
+      "learning_rate": 3.1142683015096797e-06,
+      "loss": 0.9917,
+      "step": 7786
+    },
+    {
+      "epoch": 1.4990494982794715,
+      "grad_norm": 3.188809515025512,
+      "learning_rate": 3.1120076408892374e-06,
+      "loss": 0.8953,
+      "step": 7787
+    },
+    {
+      "epoch": 1.499242004957047,
+      "grad_norm": 3.130039817311817,
+      "learning_rate": 3.109747649869774e-06,
+      "loss": 0.9345,
+      "step": 7788
+    },
+    {
+      "epoch": 1.4994345116346224,
+      "grad_norm": 3.2290389686221417,
+      "learning_rate": 3.1074883286709936e-06,
+      "loss": 0.9626,
+      "step": 7789
+    },
+    {
+      "epoch": 1.4996270183121978,
+      "grad_norm": 3.141382440953624,
+      "learning_rate": 3.105229677512527e-06,
+      "loss": 0.9117,
+      "step": 7790
+    },
+    {
+      "epoch": 1.499819524989773,
+      "grad_norm": 3.1755283220419,
+      "learning_rate": 3.102971696613949e-06,
+      "loss": 0.9062,
+      "step": 7791
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "grad_norm": 3.2476202278877357,
+      "learning_rate": 3.1007143861947575e-06,
+      "loss": 0.9752,
+      "step": 7792
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.7053,
+      "step": 7792,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.6637,
+      "step": 7792,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.8751,
+      "step": 7792,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.9144,
+      "step": 7792,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.5799,
+      "step": 7792,
+      "vm_loss": 0.131
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 1.0938,
+      "step": 7792,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 0.8213,
+      "step": 7792,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.5000120316673484,
+      "lm_loss": 1.1349,
+      "step": 7792,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.5002045383449238,
+      "grad_norm": 3.294366351781694,
+      "learning_rate": 3.098457746474396e-06,
+      "loss": 0.963,
+      "step": 7793
+    },
+    {
+      "epoch": 1.5003970450224993,
+      "grad_norm": 3.3135509686339755,
+      "learning_rate": 3.0962017776722364e-06,
+      "loss": 0.967,
+      "step": 7794
+    },
+    {
+      "epoch": 1.5005895517000747,
+      "grad_norm": 3.3842790496195034,
+      "learning_rate": 3.0939464800075947e-06,
+      "loss": 1.0102,
+      "step": 7795
+    },
+    {
+      "epoch": 1.5007820583776499,
+      "grad_norm": 3.2071537212585497,
+      "learning_rate": 3.091691853699702e-06,
+      "loss": 0.9301,
+      "step": 7796
+    },
+    {
+      "epoch": 1.5009745650552253,
+      "grad_norm": 3.399682611476646,
+      "learning_rate": 3.08943789896775e-06,
+      "loss": 1.0149,
+      "step": 7797
+    },
+    {
+      "epoch": 1.5011670717328007,
+      "grad_norm": 3.180991284415726,
+      "learning_rate": 3.087184616030844e-06,
+      "loss": 0.9317,
+      "step": 7798
+    },
+    {
+      "epoch": 1.5013595784103761,
+      "grad_norm": 3.1782859750776145,
+      "learning_rate": 3.0849320051080368e-06,
+      "loss": 0.9469,
+      "step": 7799
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "grad_norm": 3.3363292570567555,
+      "learning_rate": 3.082680066418301e-06,
+      "loss": 0.9758,
+      "step": 7800
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.5712,
+      "step": 7800,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.5198,
+      "step": 7800,
+      "vm_loss": 0.2224
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.8409,
+      "step": 7800,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.9893,
+      "step": 7800,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.9657,
+      "step": 7800,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.4379,
+      "step": 7800,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.831,
+      "step": 7800,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 1.5015520850879516,
+      "lm_loss": 0.481,
+      "step": 7800,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.5017445917655268,
+      "grad_norm": 3.2590669875017313,
+      "learning_rate": 3.0804288001805693e-06,
+      "loss": 0.9373,
+      "step": 7801
+    },
+    {
+      "epoch": 1.5019370984431022,
+      "grad_norm": 3.2236078569900646,
+      "learning_rate": 3.0781782066136812e-06,
+      "loss": 0.9446,
+      "step": 7802
+    },
+    {
+      "epoch": 1.5021296051206776,
+      "grad_norm": 3.408332658780235,
+      "learning_rate": 3.075928285936427e-06,
+      "loss": 1.0079,
+      "step": 7803
+    },
+    {
+      "epoch": 1.502322111798253,
+      "grad_norm": 3.3325751997240185,
+      "learning_rate": 3.0736790383675287e-06,
+      "loss": 0.979,
+      "step": 7804
+    },
+    {
+      "epoch": 1.5025146184758285,
+      "grad_norm": 3.112958850337292,
+      "learning_rate": 3.071430464125641e-06,
+      "loss": 0.9326,
+      "step": 7805
+    },
+    {
+      "epoch": 1.5027071251534037,
+      "grad_norm": 3.2691658090713736,
+      "learning_rate": 3.069182563429355e-06,
+      "loss": 0.9385,
+      "step": 7806
+    },
+    {
+      "epoch": 1.5028996318309793,
+      "grad_norm": 3.250751491877489,
+      "learning_rate": 3.066935336497198e-06,
+      "loss": 0.9826,
+      "step": 7807
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "grad_norm": 3.3822783336599436,
+      "learning_rate": 3.0646887835476226e-06,
+      "loss": 1.0021,
+      "step": 7808
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.6411,
+      "step": 7808,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.7659,
+      "step": 7808,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.932,
+      "step": 7808,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.951,
+      "step": 7808,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 1.1533,
+      "step": 7808,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.4582,
+      "step": 7808,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.9484,
+      "step": 7808,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.5030921385085545,
+      "lm_loss": 0.6505,
+      "step": 7808,
+      "vm_loss": 0.1667
+    },
+    {
+      "epoch": 1.50328464518613,
+      "grad_norm": 3.2930240304573206,
+      "learning_rate": 3.0624429047990255e-06,
+      "loss": 0.9666,
+      "step": 7809
+    },
+    {
+      "epoch": 1.5034771518637053,
+      "grad_norm": 3.1852727989615,
+      "learning_rate": 3.0601977004697357e-06,
+      "loss": 0.9158,
+      "step": 7810
+    },
+    {
+      "epoch": 1.5036696585412805,
+      "grad_norm": 3.205499384792309,
+      "learning_rate": 3.0579531707780162e-06,
+      "loss": 0.9618,
+      "step": 7811
+    },
+    {
+      "epoch": 1.5038621652188562,
+      "grad_norm": 3.205904308491133,
+      "learning_rate": 3.055709315942066e-06,
+      "loss": 0.8971,
+      "step": 7812
+    },
+    {
+      "epoch": 1.5040546718964314,
+      "grad_norm": 3.1524511368743093,
+      "learning_rate": 3.053466136180007e-06,
+      "loss": 0.909,
+      "step": 7813
+    },
+    {
+      "epoch": 1.5042471785740068,
+      "grad_norm": 3.2974488401174673,
+      "learning_rate": 3.0512236317099173e-06,
+      "loss": 0.9637,
+      "step": 7814
+    },
+    {
+      "epoch": 1.5044396852515822,
+      "grad_norm": 3.4551727605621174,
+      "learning_rate": 3.0489818027497895e-06,
+      "loss": 0.9725,
+      "step": 7815
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "grad_norm": 3.4765767203988043,
+      "learning_rate": 3.0467406495175633e-06,
+      "loss": 0.9499,
+      "step": 7816
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 1.5307,
+      "step": 7816,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.3988,
+      "step": 7816,
+      "vm_loss": 0.1074
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.5943,
+      "step": 7816,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.8914,
+      "step": 7816,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 1.108,
+      "step": 7816,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.87,
+      "step": 7816,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.8067,
+      "step": 7816,
+      "vm_loss": 0.1169
+    },
+    {
+      "epoch": 1.5046321919291574,
+      "lm_loss": 0.683,
+      "step": 7816,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.504824698606733,
+      "grad_norm": 3.291171560745147,
+      "learning_rate": 3.0445001722310973e-06,
+      "loss": 0.9842,
+      "step": 7817
+    },
+    {
+      "epoch": 1.5050172052843083,
+      "grad_norm": 3.2814565928466948,
+      "learning_rate": 3.0422603711082076e-06,
+      "loss": 0.9925,
+      "step": 7818
+    },
+    {
+      "epoch": 1.5052097119618837,
+      "grad_norm": 3.2003445203822074,
+      "learning_rate": 3.0400212463666224e-06,
+      "loss": 0.932,
+      "step": 7819
+    },
+    {
+      "epoch": 1.505402218639459,
+      "grad_norm": 3.279519565723928,
+      "learning_rate": 3.037782798224015e-06,
+      "loss": 0.9111,
+      "step": 7820
+    },
+    {
+      "epoch": 1.5055947253170343,
+      "grad_norm": 3.125263038178421,
+      "learning_rate": 3.0355450268979935e-06,
+      "loss": 0.9054,
+      "step": 7821
+    },
+    {
+      "epoch": 1.50578723199461,
+      "grad_norm": 3.3000042329584782,
+      "learning_rate": 3.0333079326060997e-06,
+      "loss": 0.9732,
+      "step": 7822
+    },
+    {
+      "epoch": 1.5059797386721852,
+      "grad_norm": 3.395311858369942,
+      "learning_rate": 3.0310715155657987e-06,
+      "loss": 0.9734,
+      "step": 7823
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "grad_norm": 3.1519466658749464,
+      "learning_rate": 3.028835775994512e-06,
+      "loss": 0.9265,
+      "step": 7824
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.9728,
+      "step": 7824,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.6396,
+      "step": 7824,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.8694,
+      "step": 7824,
+      "vm_loss": 0.1243
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.7613,
+      "step": 7824,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.5175,
+      "step": 7824,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.9541,
+      "step": 7824,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.6703,
+      "step": 7824,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 1.5061722453497606,
+      "lm_loss": 0.695,
+      "step": 7824,
+      "vm_loss": 0.1352
+    },
+    {
+      "epoch": 1.506364752027336,
+      "grad_norm": 3.344800185688409,
+      "learning_rate": 3.026600714109571e-06,
+      "loss": 0.9597,
+      "step": 7825
+    },
+    {
+      "epoch": 1.5065572587049112,
+      "grad_norm": 3.3540640146375025,
+      "learning_rate": 3.0243663301282567e-06,
+      "loss": 0.9829,
+      "step": 7826
+    },
+    {
+      "epoch": 1.5067497653824868,
+      "grad_norm": 3.381203114039733,
+      "learning_rate": 3.0221326242677797e-06,
+      "loss": 1.0206,
+      "step": 7827
+    },
+    {
+      "epoch": 1.506942272060062,
+      "grad_norm": 3.286401521181983,
+      "learning_rate": 3.019899596745284e-06,
+      "loss": 0.9601,
+      "step": 7828
+    },
+    {
+      "epoch": 1.5071347787376375,
+      "grad_norm": 3.362663390847099,
+      "learning_rate": 3.017667247777852e-06,
+      "loss": 1.0066,
+      "step": 7829
+    },
+    {
+      "epoch": 1.5073272854152129,
+      "grad_norm": 3.3186556926533357,
+      "learning_rate": 3.0154355775824916e-06,
+      "loss": 0.963,
+      "step": 7830
+    },
+    {
+      "epoch": 1.507519792092788,
+      "grad_norm": 3.188453438557234,
+      "learning_rate": 3.013204586376152e-06,
+      "loss": 0.9411,
+      "step": 7831
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "grad_norm": 3.26417788736505,
+      "learning_rate": 3.010974274375712e-06,
+      "loss": 0.9615,
+      "step": 7832
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 1.3322,
+      "step": 7832,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.5002,
+      "step": 7832,
+      "vm_loss": 0.2413
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.7861,
+      "step": 7832,
+      "vm_loss": 0.2121
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.5722,
+      "step": 7832,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.4425,
+      "step": 7832,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.6456,
+      "step": 7832,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.7936,
+      "step": 7832,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.5077122987703637,
+      "lm_loss": 0.9341,
+      "step": 7832,
+      "vm_loss": 0.1991
+    },
+    {
+      "epoch": 1.507904805447939,
+      "grad_norm": 3.2556036950745004,
+      "learning_rate": 3.008744641797993e-06,
+      "loss": 0.945,
+      "step": 7833
+    },
+    {
+      "epoch": 1.5080973121255143,
+      "grad_norm": 3.2045046291032966,
+      "learning_rate": 3.0065156888597313e-06,
+      "loss": 0.9807,
+      "step": 7834
+    },
+    {
+      "epoch": 1.5082898188030898,
+      "grad_norm": 3.3143383201647274,
+      "learning_rate": 3.004287415777625e-06,
+      "loss": 0.9492,
+      "step": 7835
+    },
+    {
+      "epoch": 1.508482325480665,
+      "grad_norm": 3.346507916101847,
+      "learning_rate": 3.0020598227682794e-06,
+      "loss": 0.942,
+      "step": 7836
+    },
+    {
+      "epoch": 1.5086748321582406,
+      "grad_norm": 3.012009149299936,
+      "learning_rate": 2.9998329100482493e-06,
+      "loss": 0.8855,
+      "step": 7837
+    },
+    {
+      "epoch": 1.5088673388358158,
+      "grad_norm": 3.2302572945298125,
+      "learning_rate": 2.9976066778340186e-06,
+      "loss": 0.935,
+      "step": 7838
+    },
+    {
+      "epoch": 1.5090598455133912,
+      "grad_norm": 3.0961712975697355,
+      "learning_rate": 2.9953811263420087e-06,
+      "loss": 0.8796,
+      "step": 7839
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "grad_norm": 3.2979329118484606,
+      "learning_rate": 2.993156255788565e-06,
+      "loss": 0.9715,
+      "step": 7840
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.7013,
+      "step": 7840,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.7238,
+      "step": 7840,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 1.207,
+      "step": 7840,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.6218,
+      "step": 7840,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.8391,
+      "step": 7840,
+      "vm_loss": 0.2006
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 1.0426,
+      "step": 7840,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.8236,
+      "step": 7840,
+      "vm_loss": 0.1739
+    },
+    {
+      "epoch": 1.5092523521909667,
+      "lm_loss": 0.82,
+      "step": 7840,
+      "vm_loss": 0.2145
+    },
+    {
+      "epoch": 1.5094448588685418,
+      "grad_norm": 3.298616035639577,
+      "learning_rate": 2.990932066389978e-06,
+      "loss": 0.9608,
+      "step": 7841
+    },
+    {
+      "epoch": 1.5096373655461175,
+      "grad_norm": 3.333384899053493,
+      "learning_rate": 2.9887085583624655e-06,
+      "loss": 0.9888,
+      "step": 7842
+    },
+    {
+      "epoch": 1.5098298722236927,
+      "grad_norm": 3.2292780290333267,
+      "learning_rate": 2.9864857319221852e-06,
+      "loss": 0.9637,
+      "step": 7843
+    },
+    {
+      "epoch": 1.5100223789012681,
+      "grad_norm": 3.2134135467500458,
+      "learning_rate": 2.9842635872852135e-06,
+      "loss": 0.8908,
+      "step": 7844
+    },
+    {
+      "epoch": 1.5102148855788435,
+      "grad_norm": 3.2092819128941,
+      "learning_rate": 2.9820421246675855e-06,
+      "loss": 0.9236,
+      "step": 7845
+    },
+    {
+      "epoch": 1.5104073922564187,
+      "grad_norm": 3.3036654409188912,
+      "learning_rate": 2.9798213442852466e-06,
+      "loss": 0.9492,
+      "step": 7846
+    },
+    {
+      "epoch": 1.5105998989339944,
+      "grad_norm": 3.40963758868991,
+      "learning_rate": 2.977601246354086e-06,
+      "loss": 0.9789,
+      "step": 7847
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "grad_norm": 3.149603980505544,
+      "learning_rate": 2.9753818310899276e-06,
+      "loss": 0.928,
+      "step": 7848
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.5397,
+      "step": 7848,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.6126,
+      "step": 7848,
+      "vm_loss": 0.2447
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.5625,
+      "step": 7848,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.3023,
+      "step": 7848,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.6832,
+      "step": 7848,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.6294,
+      "step": 7848,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.8626,
+      "step": 7848,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 1.5107924056115696,
+      "lm_loss": 0.7261,
+      "step": 7848,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 1.510984912289145,
+      "grad_norm": 3.175547072125256,
+      "learning_rate": 2.973163098708527e-06,
+      "loss": 0.9235,
+      "step": 7849
+    },
+    {
+      "epoch": 1.5111774189667204,
+      "grad_norm": 3.3815955939935405,
+      "learning_rate": 2.9709450494255744e-06,
+      "loss": 0.9804,
+      "step": 7850
+    },
+    {
+      "epoch": 1.5113699256442958,
+      "grad_norm": 3.038392663111802,
+      "learning_rate": 2.9687276834566856e-06,
+      "loss": 0.9041,
+      "step": 7851
+    },
+    {
+      "epoch": 1.5115624323218713,
+      "grad_norm": 3.211710795979048,
+      "learning_rate": 2.966511001017428e-06,
+      "loss": 0.9654,
+      "step": 7852
+    },
+    {
+      "epoch": 1.5117549389994465,
+      "grad_norm": 3.15260849877099,
+      "learning_rate": 2.964295002323283e-06,
+      "loss": 0.9459,
+      "step": 7853
+    },
+    {
+      "epoch": 1.5119474456770219,
+      "grad_norm": 3.250671400542431,
+      "learning_rate": 2.962079687589676e-06,
+      "loss": 0.9721,
+      "step": 7854
+    },
+    {
+      "epoch": 1.5121399523545973,
+      "grad_norm": 3.258175969691289,
+      "learning_rate": 2.9598650570319653e-06,
+      "loss": 0.9353,
+      "step": 7855
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "grad_norm": 3.4596913053078513,
+      "learning_rate": 2.9576511108654436e-06,
+      "loss": 0.9975,
+      "step": 7856
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.7401,
+      "step": 7856,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.7333,
+      "step": 7856,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.8848,
+      "step": 7856,
+      "vm_loss": 0.1338
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 1.2021,
+      "step": 7856,
+      "vm_loss": 0.1092
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.6534,
+      "step": 7856,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 1.0,
+      "step": 7856,
+      "vm_loss": 0.2015
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.5091,
+      "step": 7856,
+      "vm_loss": 0.1313
+    },
+    {
+      "epoch": 1.5123324590321727,
+      "lm_loss": 0.7526,
+      "step": 7856,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.5125249657097481,
+      "grad_norm": 3.0854046974490945,
+      "learning_rate": 2.955437849305329e-06,
+      "loss": 0.9188,
+      "step": 7857
+    },
+    {
+      "epoch": 1.5127174723873233,
+      "grad_norm": 3.2122323287307184,
+      "learning_rate": 2.953225272566782e-06,
+      "loss": 0.9503,
+      "step": 7858
+    },
+    {
+      "epoch": 1.5129099790648988,
+      "grad_norm": 3.0845922810565005,
+      "learning_rate": 2.9510133808648923e-06,
+      "loss": 0.9004,
+      "step": 7859
+    },
+    {
+      "epoch": 1.5131024857424742,
+      "grad_norm": 3.3478893482677243,
+      "learning_rate": 2.948802174414688e-06,
+      "loss": 1.0322,
+      "step": 7860
+    },
+    {
+      "epoch": 1.5132949924200496,
+      "grad_norm": 3.2317287730803086,
+      "learning_rate": 2.9465916534311177e-06,
+      "loss": 0.909,
+      "step": 7861
+    },
+    {
+      "epoch": 1.513487499097625,
+      "grad_norm": 3.153682051162729,
+      "learning_rate": 2.944381818129084e-06,
+      "loss": 0.948,
+      "step": 7862
+    },
+    {
+      "epoch": 1.5136800057752002,
+      "grad_norm": 3.2725592119711675,
+      "learning_rate": 2.9421726687234033e-06,
+      "loss": 0.97,
+      "step": 7863
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "grad_norm": 3.4323394694233187,
+      "learning_rate": 2.9399642054288336e-06,
+      "loss": 1.0502,
+      "step": 7864
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.6883,
+      "step": 7864,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.5412,
+      "step": 7864,
+      "vm_loss": 0.234
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 1.0955,
+      "step": 7864,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.775,
+      "step": 7864,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.5132,
+      "step": 7864,
+      "vm_loss": 0.098
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.6865,
+      "step": 7864,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.5881,
+      "step": 7864,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.5138725124527757,
+      "lm_loss": 0.7141,
+      "step": 7864,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 1.514065019130351,
+      "grad_norm": 3.324281687103269,
+      "learning_rate": 2.937756428460068e-06,
+      "loss": 0.9883,
+      "step": 7865
+    },
+    {
+      "epoch": 1.5142575258079265,
+      "grad_norm": 3.1941463767188365,
+      "learning_rate": 2.9355493380317334e-06,
+      "loss": 0.9133,
+      "step": 7866
+    },
+    {
+      "epoch": 1.514450032485502,
+      "grad_norm": 3.129856805838578,
+      "learning_rate": 2.933342934358381e-06,
+      "loss": 0.9373,
+      "step": 7867
+    },
+    {
+      "epoch": 1.5146425391630771,
+      "grad_norm": 3.293173417884313,
+      "learning_rate": 2.9311372176545028e-06,
+      "loss": 0.9619,
+      "step": 7868
+    },
+    {
+      "epoch": 1.5148350458406528,
+      "grad_norm": 3.1964471208157126,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.9447,
+      "step": 7869
+    },
+    {
+      "epoch": 1.515027552518228,
+      "grad_norm": 3.3227774032784643,
+      "learning_rate": 2.9267278460128045e-06,
+      "loss": 0.9729,
+      "step": 7870
+    },
+    {
+      "epoch": 1.5152200591958034,
+      "grad_norm": 3.1392236777527724,
+      "learning_rate": 2.9245241915036304e-06,
+      "loss": 0.8933,
+      "step": 7871
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "grad_norm": 3.270933592700056,
+      "learning_rate": 2.9223212248212275e-06,
+      "loss": 0.9828,
+      "step": 7872
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.372,
+      "step": 7872,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.6552,
+      "step": 7872,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.8957,
+      "step": 7872,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.3996,
+      "step": 7872,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.9829,
+      "step": 7872,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.5987,
+      "step": 7872,
+      "vm_loss": 0.2297
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 1.1111,
+      "step": 7872,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 1.5154125658733788,
+      "lm_loss": 0.8696,
+      "step": 7872,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.515605072550954,
+      "grad_norm": 3.1978790823637406,
+      "learning_rate": 2.920118946179754e-06,
+      "loss": 0.957,
+      "step": 7873
+    },
+    {
+      "epoch": 1.5157975792285296,
+      "grad_norm": 3.314784829486023,
+      "learning_rate": 2.917917355793294e-06,
+      "loss": 0.9538,
+      "step": 7874
+    },
+    {
+      "epoch": 1.5159900859061048,
+      "grad_norm": 3.1440310455661353,
+      "learning_rate": 2.915716453875873e-06,
+      "loss": 0.9397,
+      "step": 7875
+    },
+    {
+      "epoch": 1.5161825925836803,
+      "grad_norm": 3.358587416688107,
+      "learning_rate": 2.9135162406414476e-06,
+      "loss": 0.9508,
+      "step": 7876
+    },
+    {
+      "epoch": 1.5163750992612557,
+      "grad_norm": 3.5239765529201366,
+      "learning_rate": 2.911316716303908e-06,
+      "loss": 1.0264,
+      "step": 7877
+    },
+    {
+      "epoch": 1.5165676059388309,
+      "grad_norm": 3.24052667970497,
+      "learning_rate": 2.909117881077069e-06,
+      "loss": 0.9333,
+      "step": 7878
+    },
+    {
+      "epoch": 1.5167601126164065,
+      "grad_norm": 3.0790327228456564,
+      "learning_rate": 2.906919735174698e-06,
+      "loss": 0.892,
+      "step": 7879
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "grad_norm": 3.2202527269056747,
+      "learning_rate": 2.9047222788104712e-06,
+      "loss": 0.9154,
+      "step": 7880
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.7989,
+      "step": 7880,
+      "vm_loss": 0.1923
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.7383,
+      "step": 7880,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.7612,
+      "step": 7880,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.5935,
+      "step": 7880,
+      "vm_loss": 0.1424
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.7328,
+      "step": 7880,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.8368,
+      "step": 7880,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.5921,
+      "step": 7880,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.5169526192939817,
+      "lm_loss": 0.6966,
+      "step": 7880,
+      "vm_loss": 0.2142
+    },
+    {
+      "epoch": 1.5171451259715572,
+      "grad_norm": 3.212063517588743,
+      "learning_rate": 2.9025255121980155e-06,
+      "loss": 0.9165,
+      "step": 7881
+    },
+    {
+      "epoch": 1.5173376326491326,
+      "grad_norm": 3.2564129852326755,
+      "learning_rate": 2.9003294355508816e-06,
+      "loss": 0.9436,
+      "step": 7882
+    },
+    {
+      "epoch": 1.5175301393267078,
+      "grad_norm": 3.4035160443072385,
+      "learning_rate": 2.8981340490825616e-06,
+      "loss": 1.0312,
+      "step": 7883
+    },
+    {
+      "epoch": 1.5177226460042834,
+      "grad_norm": 3.306536776510789,
+      "learning_rate": 2.8959393530064696e-06,
+      "loss": 0.9688,
+      "step": 7884
+    },
+    {
+      "epoch": 1.5179151526818586,
+      "grad_norm": 3.19413145345373,
+      "learning_rate": 2.8937453475359578e-06,
+      "loss": 0.8942,
+      "step": 7885
+    },
+    {
+      "epoch": 1.518107659359434,
+      "grad_norm": 3.2393458137153597,
+      "learning_rate": 2.891552032884315e-06,
+      "loss": 0.934,
+      "step": 7886
+    },
+    {
+      "epoch": 1.5183001660370095,
+      "grad_norm": 3.0957271210821227,
+      "learning_rate": 2.8893594092647616e-06,
+      "loss": 0.8787,
+      "step": 7887
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "grad_norm": 3.203300721557911,
+      "learning_rate": 2.8871674768904377e-06,
+      "loss": 0.9614,
+      "step": 7888
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 0.7213,
+      "step": 7888,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 1.2036,
+      "step": 7888,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 0.9483,
+      "step": 7888,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 0.7984,
+      "step": 7888,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 0.9461,
+      "step": 7888,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 0.4063,
+      "step": 7888,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 1.0366,
+      "step": 7888,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 1.5184926727145847,
+      "lm_loss": 1.0646,
+      "step": 7888,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.5186851793921603,
+      "grad_norm": 3.2783994253213042,
+      "learning_rate": 2.8849762359744426e-06,
+      "loss": 0.9325,
+      "step": 7889
+    },
+    {
+      "epoch": 1.5188776860697355,
+      "grad_norm": 3.119216990453225,
+      "learning_rate": 2.882785686729781e-06,
+      "loss": 0.8994,
+      "step": 7890
+    },
+    {
+      "epoch": 1.519070192747311,
+      "grad_norm": 3.4122065764179554,
+      "learning_rate": 2.880595829369406e-06,
+      "loss": 0.9633,
+      "step": 7891
+    },
+    {
+      "epoch": 1.5192626994248863,
+      "grad_norm": 3.4339113385605677,
+      "learning_rate": 2.8784066641062013e-06,
+      "loss": 0.9425,
+      "step": 7892
+    },
+    {
+      "epoch": 1.5194552061024615,
+      "grad_norm": 3.412221430504424,
+      "learning_rate": 2.8762181911529798e-06,
+      "loss": 0.9181,
+      "step": 7893
+    },
+    {
+      "epoch": 1.5196477127800372,
+      "grad_norm": 3.2228622687664474,
+      "learning_rate": 2.8740304107224936e-06,
+      "loss": 0.8988,
+      "step": 7894
+    },
+    {
+      "epoch": 1.5198402194576124,
+      "grad_norm": 3.4439362547727788,
+      "learning_rate": 2.8718433230274124e-06,
+      "loss": 0.9846,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "grad_norm": 3.2469484901917363,
+      "learning_rate": 2.8696569282803622e-06,
+      "loss": 0.9403,
+      "step": 7896
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.7343,
+      "step": 7896,
+      "vm_loss": 0.1046
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 1.0578,
+      "step": 7896,
+      "vm_loss": 0.1227
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 1.0158,
+      "step": 7896,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.6039,
+      "step": 7896,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.8944,
+      "step": 7896,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.7894,
+      "step": 7896,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.4928,
+      "step": 7896,
+      "vm_loss": 0.1714
+    },
+    {
+      "epoch": 1.5200327261351878,
+      "lm_loss": 0.811,
+      "step": 7896,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 1.5202252328127632,
+      "grad_norm": 3.143612158537478,
+      "learning_rate": 2.8674712266938797e-06,
+      "loss": 0.9297,
+      "step": 7897
+    },
+    {
+      "epoch": 1.5204177394903384,
+      "grad_norm": 3.213090765660677,
+      "learning_rate": 2.8652862184804452e-06,
+      "loss": 0.9569,
+      "step": 7898
+    },
+    {
+      "epoch": 1.520610246167914,
+      "grad_norm": 3.1761350812790243,
+      "learning_rate": 2.863101903852471e-06,
+      "loss": 0.9358,
+      "step": 7899
+    },
+    {
+      "epoch": 1.5208027528454893,
+      "grad_norm": 3.37648226987177,
+      "learning_rate": 2.8609182830223015e-06,
+      "loss": 0.9764,
+      "step": 7900
+    },
+    {
+      "epoch": 1.5209952595230647,
+      "grad_norm": 3.287643373656489,
+      "learning_rate": 2.858735356202207e-06,
+      "loss": 0.9576,
+      "step": 7901
+    },
+    {
+      "epoch": 1.5211877662006401,
+      "grad_norm": 3.3216958851653082,
+      "learning_rate": 2.8565531236043997e-06,
+      "loss": 0.9921,
+      "step": 7902
+    },
+    {
+      "epoch": 1.5213802728782153,
+      "grad_norm": 3.210463538273395,
+      "learning_rate": 2.8543715854410203e-06,
+      "loss": 0.9243,
+      "step": 7903
+    },
+    {
+      "epoch": 1.521572779555791,
+      "grad_norm": 3.3266262878978288,
+      "learning_rate": 2.8521907419241445e-06,
+      "loss": 0.9833,
+      "step": 7904
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.8762,
+      "step": 7904,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 1.1756,
+      "step": 7904,
+      "vm_loss": 0.2393
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.5915,
+      "step": 7904,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.6991,
+      "step": 7904,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.4825,
+      "step": 7904,
+      "vm_loss": 0.1226
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.9025,
+      "step": 7904,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 0.8409,
+      "step": 7904,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.521572779555791,
+      "lm_loss": 1.1326,
+      "step": 7904,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.5217652862333662,
+      "grad_norm": 3.200664760887337,
+      "learning_rate": 2.8500105932657706e-06,
+      "loss": 0.9314,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5219577929109416,
+      "grad_norm": 3.1841842542374605,
+      "learning_rate": 2.847831139677848e-06,
+      "loss": 0.9176,
+      "step": 7906
+    },
+    {
+      "epoch": 1.522150299588517,
+      "grad_norm": 3.1638362796839776,
+      "learning_rate": 2.845652381372238e-06,
+      "loss": 0.96,
+      "step": 7907
+    },
+    {
+      "epoch": 1.5223428062660922,
+      "grad_norm": 3.3740739708409864,
+      "learning_rate": 2.8434743185607483e-06,
+      "loss": 0.9908,
+      "step": 7908
+    },
+    {
+      "epoch": 1.5225353129436678,
+      "grad_norm": 3.416488671591386,
+      "learning_rate": 2.8412969514551124e-06,
+      "loss": 1.0287,
+      "step": 7909
+    },
+    {
+      "epoch": 1.522727819621243,
+      "grad_norm": 3.195275143905709,
+      "learning_rate": 2.8391202802670025e-06,
+      "loss": 0.8638,
+      "step": 7910
+    },
+    {
+      "epoch": 1.5229203262988185,
+      "grad_norm": 3.4105116413427954,
+      "learning_rate": 2.8369443052080135e-06,
+      "loss": 0.9716,
+      "step": 7911
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "grad_norm": 3.1091468107279936,
+      "learning_rate": 2.83476902648968e-06,
+      "loss": 0.8899,
+      "step": 7912
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.7429,
+      "step": 7912,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.6468,
+      "step": 7912,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.586,
+      "step": 7912,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.57,
+      "step": 7912,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.9127,
+      "step": 7912,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.3617,
+      "step": 7912,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.7489,
+      "step": 7912,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 1.5231128329763939,
+      "lm_loss": 0.5803,
+      "step": 7912,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.523305339653969,
+      "grad_norm": 3.3998330070262215,
+      "learning_rate": 2.8325944443234674e-06,
+      "loss": 0.9365,
+      "step": 7913
+    },
+    {
+      "epoch": 1.5234978463315447,
+      "grad_norm": 3.328980404324694,
+      "learning_rate": 2.830420558920772e-06,
+      "loss": 0.9587,
+      "step": 7914
+    },
+    {
+      "epoch": 1.52369035300912,
+      "grad_norm": 3.1617280526809552,
+      "learning_rate": 2.8282473704929283e-06,
+      "loss": 0.9315,
+      "step": 7915
+    },
+    {
+      "epoch": 1.5238828596866953,
+      "grad_norm": 3.2725409445742972,
+      "learning_rate": 2.826074879251186e-06,
+      "loss": 0.9593,
+      "step": 7916
+    },
+    {
+      "epoch": 1.5240753663642708,
+      "grad_norm": 3.172251503528973,
+      "learning_rate": 2.823903085406754e-06,
+      "loss": 0.9017,
+      "step": 7917
+    },
+    {
+      "epoch": 1.5242678730418462,
+      "grad_norm": 3.1936853799073064,
+      "learning_rate": 2.8217319891707474e-06,
+      "loss": 0.9365,
+      "step": 7918
+    },
+    {
+      "epoch": 1.5244603797194216,
+      "grad_norm": 3.386063629209984,
+      "learning_rate": 2.8195615907542285e-06,
+      "loss": 0.9875,
+      "step": 7919
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "grad_norm": 3.0335218716968892,
+      "learning_rate": 2.8173918903681874e-06,
+      "loss": 0.8948,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 1.2054,
+      "step": 7920,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 0.875,
+      "step": 7920,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 0.3873,
+      "step": 7920,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 1.2146,
+      "step": 7920,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 0.9866,
+      "step": 7920,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 0.6228,
+      "step": 7920,
+      "vm_loss": 0.2196
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 1.0023,
+      "step": 7920,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.5246528863969968,
+      "lm_loss": 0.8476,
+      "step": 7920,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.5248453930745722,
+      "grad_norm": 3.4371834327541237,
+      "learning_rate": 2.8152228882235512e-06,
+      "loss": 1.0209,
+      "step": 7921
+    },
+    {
+      "epoch": 1.5250378997521477,
+      "grad_norm": 3.218700290989853,
+      "learning_rate": 2.813054584531163e-06,
+      "loss": 0.9181,
+      "step": 7922
+    },
+    {
+      "epoch": 1.525230406429723,
+      "grad_norm": 3.2618155870182584,
+      "learning_rate": 2.8108869795018233e-06,
+      "loss": 0.9354,
+      "step": 7923
+    },
+    {
+      "epoch": 1.5254229131072985,
+      "grad_norm": 3.1782015142158144,
+      "learning_rate": 2.8087200733462427e-06,
+      "loss": 0.9015,
+      "step": 7924
+    },
+    {
+      "epoch": 1.5256154197848737,
+      "grad_norm": 3.1426280150538646,
+      "learning_rate": 2.806553866275077e-06,
+      "loss": 0.8952,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5258079264624491,
+      "grad_norm": 3.2162292050820533,
+      "learning_rate": 2.8043883584988996e-06,
+      "loss": 0.9495,
+      "step": 7926
+    },
+    {
+      "epoch": 1.5260004331400245,
+      "grad_norm": 3.2438556475297,
+      "learning_rate": 2.802223550228239e-06,
+      "loss": 0.9145,
+      "step": 7927
+    },
+    {
+      "epoch": 1.5261929398176,
+      "grad_norm": 3.2794762771581363,
+      "learning_rate": 2.800059441673533e-06,
+      "loss": 0.959,
+      "step": 7928
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.5727,
+      "step": 7928,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 1.0174,
+      "step": 7928,
+      "vm_loss": 0.1412
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.6226,
+      "step": 7928,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.7331,
+      "step": 7928,
+      "vm_loss": 0.1596
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.6039,
+      "step": 7928,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.6562,
+      "step": 7928,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 0.7794,
+      "step": 7928,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 1.5261929398176,
+      "lm_loss": 1.4351,
+      "step": 7928,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.5263854464951754,
+      "grad_norm": 3.193255170579407,
+      "learning_rate": 2.797896033045163e-06,
+      "loss": 0.9079,
+      "step": 7929
+    },
+    {
+      "epoch": 1.5265779531727506,
+      "grad_norm": 3.2150010122468755,
+      "learning_rate": 2.795733324553441e-06,
+      "loss": 0.9547,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5267704598503262,
+      "grad_norm": 3.1670579134807415,
+      "learning_rate": 2.7935713164086086e-06,
+      "loss": 0.9161,
+      "step": 7931
+    },
+    {
+      "epoch": 1.5269629665279014,
+      "grad_norm": 3.488154195621208,
+      "learning_rate": 2.791410008820845e-06,
+      "loss": 1.0134,
+      "step": 7932
+    },
+    {
+      "epoch": 1.5271554732054768,
+      "grad_norm": 3.2146959501058037,
+      "learning_rate": 2.7892494020002505e-06,
+      "loss": 0.8919,
+      "step": 7933
+    },
+    {
+      "epoch": 1.5273479798830523,
+      "grad_norm": 3.169590916399676,
+      "learning_rate": 2.7870894961568662e-06,
+      "loss": 0.9452,
+      "step": 7934
+    },
+    {
+      "epoch": 1.5275404865606275,
+      "grad_norm": 3.1514412199949264,
+      "learning_rate": 2.7849302915006626e-06,
+      "loss": 0.8826,
+      "step": 7935
+    },
+    {
+      "epoch": 1.527732993238203,
+      "grad_norm": 3.2367947803501167,
+      "learning_rate": 2.782771788241543e-06,
+      "loss": 0.9734,
+      "step": 7936
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.7398,
+      "step": 7936,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.6008,
+      "step": 7936,
+      "vm_loss": 0.1049
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.4836,
+      "step": 7936,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.7368,
+      "step": 7936,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.4844,
+      "step": 7936,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.6611,
+      "step": 7936,
+      "vm_loss": 0.1039
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.7487,
+      "step": 7936,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 1.527732993238203,
+      "lm_loss": 0.5715,
+      "step": 7936,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.5279254999157783,
+      "grad_norm": 3.3783687918573957,
+      "learning_rate": 2.780613986589341e-06,
+      "loss": 0.9606,
+      "step": 7937
+    },
+    {
+      "epoch": 1.5281180065933537,
+      "grad_norm": 3.3513477808613326,
+      "learning_rate": 2.778456886753824e-06,
+      "loss": 0.9474,
+      "step": 7938
+    },
+    {
+      "epoch": 1.5283105132709291,
+      "grad_norm": 3.353058471605618,
+      "learning_rate": 2.7763004889446845e-06,
+      "loss": 0.9644,
+      "step": 7939
+    },
+    {
+      "epoch": 1.5285030199485043,
+      "grad_norm": 3.3199404914339503,
+      "learning_rate": 2.7741447933715595e-06,
+      "loss": 0.9534,
+      "step": 7940
+    },
+    {
+      "epoch": 1.52869552662608,
+      "grad_norm": 3.1643395875412033,
+      "learning_rate": 2.7719898002440037e-06,
+      "loss": 0.9612,
+      "step": 7941
+    },
+    {
+      "epoch": 1.5288880333036552,
+      "grad_norm": 3.141052928744899,
+      "learning_rate": 2.7698355097715167e-06,
+      "loss": 0.9364,
+      "step": 7942
+    },
+    {
+      "epoch": 1.5290805399812306,
+      "grad_norm": 3.3226621707174906,
+      "learning_rate": 2.767681922163512e-06,
+      "loss": 1.0054,
+      "step": 7943
+    },
+    {
+      "epoch": 1.529273046658806,
+      "grad_norm": 3.2824671922228235,
+      "learning_rate": 2.765529037629359e-06,
+      "loss": 0.9463,
+      "step": 7944
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.838,
+      "step": 7944,
+      "vm_loss": 0.1811
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.9024,
+      "step": 7944,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.7319,
+      "step": 7944,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.9065,
+      "step": 7944,
+      "vm_loss": 0.223
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.7323,
+      "step": 7944,
+      "vm_loss": 0.2251
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 1.0154,
+      "step": 7944,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.666,
+      "step": 7944,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 1.529273046658806,
+      "lm_loss": 0.9149,
+      "step": 7944,
+      "vm_loss": 0.1012
+    },
+    {
+      "epoch": 1.5294655533363812,
+      "grad_norm": 3.216402170517718,
+      "learning_rate": 2.7633768563783357e-06,
+      "loss": 0.9316,
+      "step": 7945
+    },
+    {
+      "epoch": 1.5296580600139569,
+      "grad_norm": 3.2903178911149498,
+      "learning_rate": 2.7612253786196663e-06,
+      "loss": 0.9497,
+      "step": 7946
+    },
+    {
+      "epoch": 1.529850566691532,
+      "grad_norm": 3.107944032566568,
+      "learning_rate": 2.7590746045625004e-06,
+      "loss": 0.8819,
+      "step": 7947
+    },
+    {
+      "epoch": 1.5300430733691075,
+      "grad_norm": 3.0358265909040303,
+      "learning_rate": 2.756924534415926e-06,
+      "loss": 0.8595,
+      "step": 7948
+    },
+    {
+      "epoch": 1.530235580046683,
+      "grad_norm": 3.362489582509592,
+      "learning_rate": 2.7547751683889478e-06,
+      "loss": 0.9955,
+      "step": 7949
+    },
+    {
+      "epoch": 1.5304280867242581,
+      "grad_norm": 3.368545454685749,
+      "learning_rate": 2.7526265066905177e-06,
+      "loss": 0.9856,
+      "step": 7950
+    },
+    {
+      "epoch": 1.5306205934018338,
+      "grad_norm": 3.3648455786830325,
+      "learning_rate": 2.7504785495295127e-06,
+      "loss": 0.9612,
+      "step": 7951
+    },
+    {
+      "epoch": 1.530813100079409,
+      "grad_norm": 3.0460179200882087,
+      "learning_rate": 2.7483312971147403e-06,
+      "loss": 0.8983,
+      "step": 7952
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 0.5829,
+      "step": 7952,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 0.7443,
+      "step": 7952,
+      "vm_loss": 0.2037
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 1.0858,
+      "step": 7952,
+      "vm_loss": 0.1817
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 0.7118,
+      "step": 7952,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 0.5233,
+      "step": 7952,
+      "vm_loss": 0.2523
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 0.9803,
+      "step": 7952,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 1.1859,
+      "step": 7952,
+      "vm_loss": 0.1098
+    },
+    {
+      "epoch": 1.530813100079409,
+      "lm_loss": 1.1364,
+      "step": 7952,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.5310056067569844,
+      "grad_norm": 3.3848290380901456,
+      "learning_rate": 2.746184749654942e-06,
+      "loss": 0.996,
+      "step": 7953
+    },
+    {
+      "epoch": 1.5311981134345598,
+      "grad_norm": 3.1296851778905186,
+      "learning_rate": 2.7440389073587927e-06,
+      "loss": 0.9301,
+      "step": 7954
+    },
+    {
+      "epoch": 1.531390620112135,
+      "grad_norm": 3.1792388112306456,
+      "learning_rate": 2.741893770434889e-06,
+      "loss": 0.9201,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5315831267897106,
+      "grad_norm": 3.298553622733847,
+      "learning_rate": 2.7397493390917694e-06,
+      "loss": 0.9612,
+      "step": 7956
+    },
+    {
+      "epoch": 1.5317756334672858,
+      "grad_norm": 3.2565758429365514,
+      "learning_rate": 2.737605613537899e-06,
+      "loss": 0.9204,
+      "step": 7957
+    },
+    {
+      "epoch": 1.5319681401448613,
+      "grad_norm": 3.2749339786733027,
+      "learning_rate": 2.735462593981677e-06,
+      "loss": 0.9502,
+      "step": 7958
+    },
+    {
+      "epoch": 1.5321606468224367,
+      "grad_norm": 3.0270652651930816,
+      "learning_rate": 2.733320280631434e-06,
+      "loss": 0.8678,
+      "step": 7959
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "grad_norm": 3.253472602030513,
+      "learning_rate": 2.7311786736954228e-06,
+      "loss": 0.9464,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.8057,
+      "step": 7960,
+      "vm_loss": 0.125
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.704,
+      "step": 7960,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.7755,
+      "step": 7960,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.6444,
+      "step": 7960,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.6986,
+      "step": 7960,
+      "vm_loss": 0.2247
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.708,
+      "step": 7960,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.5582,
+      "step": 7960,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 1.5323531535000119,
+      "lm_loss": 0.7389,
+      "step": 7960,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.5325456601775875,
+      "grad_norm": 3.2291358836197483,
+      "learning_rate": 2.729037773381845e-06,
+      "loss": 0.9131,
+      "step": 7961
+    },
+    {
+      "epoch": 1.5327381668551627,
+      "grad_norm": 3.1773580738369693,
+      "learning_rate": 2.726897579898815e-06,
+      "loss": 0.8895,
+      "step": 7962
+    },
+    {
+      "epoch": 1.5329306735327382,
+      "grad_norm": 3.2664355525792295,
+      "learning_rate": 2.7247580934543905e-06,
+      "loss": 0.9368,
+      "step": 7963
+    },
+    {
+      "epoch": 1.5331231802103136,
+      "grad_norm": 3.158994553941811,
+      "learning_rate": 2.722619314256557e-06,
+      "loss": 0.9145,
+      "step": 7964
+    },
+    {
+      "epoch": 1.5333156868878888,
+      "grad_norm": 3.530813063511815,
+      "learning_rate": 2.720481242513234e-06,
+      "loss": 1.0117,
+      "step": 7965
+    },
+    {
+      "epoch": 1.5335081935654644,
+      "grad_norm": 3.169532035430008,
+      "learning_rate": 2.7183438784322646e-06,
+      "loss": 0.8709,
+      "step": 7966
+    },
+    {
+      "epoch": 1.5337007002430396,
+      "grad_norm": 3.3416479133475465,
+      "learning_rate": 2.7162072222214285e-06,
+      "loss": 0.9149,
+      "step": 7967
+    },
+    {
+      "epoch": 1.533893206920615,
+      "grad_norm": 3.158020922994402,
+      "learning_rate": 2.714071274088438e-06,
+      "loss": 0.9398,
+      "step": 7968
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.6734,
+      "step": 7968,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.3745,
+      "step": 7968,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.6806,
+      "step": 7968,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.7573,
+      "step": 7968,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.6443,
+      "step": 7968,
+      "vm_loss": 0.1463
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.5062,
+      "step": 7968,
+      "vm_loss": 0.252
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.6351,
+      "step": 7968,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.533893206920615,
+      "lm_loss": 0.7953,
+      "step": 7968,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.5340857135981905,
+      "grad_norm": 3.2099254580424375,
+      "learning_rate": 2.7119360342409363e-06,
+      "loss": 0.9432,
+      "step": 7969
+    },
+    {
+      "epoch": 1.5342782202757657,
+      "grad_norm": 3.1532831296313844,
+      "learning_rate": 2.7098015028864878e-06,
+      "loss": 0.905,
+      "step": 7970
+    },
+    {
+      "epoch": 1.5344707269533413,
+      "grad_norm": 3.3746696130800324,
+      "learning_rate": 2.7076676802326084e-06,
+      "loss": 0.9853,
+      "step": 7971
+    },
+    {
+      "epoch": 1.5346632336309165,
+      "grad_norm": 3.14625616291821,
+      "learning_rate": 2.705534566486723e-06,
+      "loss": 0.913,
+      "step": 7972
+    },
+    {
+      "epoch": 1.534855740308492,
+      "grad_norm": 3.279148845372519,
+      "learning_rate": 2.7034021618562024e-06,
+      "loss": 0.9157,
+      "step": 7973
+    },
+    {
+      "epoch": 1.5350482469860673,
+      "grad_norm": 3.1076794403716974,
+      "learning_rate": 2.7012704665483413e-06,
+      "loss": 0.8862,
+      "step": 7974
+    },
+    {
+      "epoch": 1.5352407536636425,
+      "grad_norm": 3.269619903444149,
+      "learning_rate": 2.6991394807703686e-06,
+      "loss": 0.9394,
+      "step": 7975
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "grad_norm": 3.161744880009637,
+      "learning_rate": 2.697009204729447e-06,
+      "loss": 0.9244,
+      "step": 7976
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.8292,
+      "step": 7976,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 1.0268,
+      "step": 7976,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.6761,
+      "step": 7976,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.7027,
+      "step": 7976,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.6313,
+      "step": 7976,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.6142,
+      "step": 7976,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.9596,
+      "step": 7976,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.5354332603412182,
+      "lm_loss": 0.8075,
+      "step": 7976,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.5356257670187934,
+      "grad_norm": 3.231054455586062,
+      "learning_rate": 2.6948796386326615e-06,
+      "loss": 0.9193,
+      "step": 7977
+    },
+    {
+      "epoch": 1.5358182736963688,
+      "grad_norm": 3.464960503544298,
+      "learning_rate": 2.6927507826870338e-06,
+      "loss": 0.973,
+      "step": 7978
+    },
+    {
+      "epoch": 1.5360107803739442,
+      "grad_norm": 3.4564040281116917,
+      "learning_rate": 2.690622637099517e-06,
+      "loss": 0.9814,
+      "step": 7979
+    },
+    {
+      "epoch": 1.5362032870515197,
+      "grad_norm": 3.174019810427211,
+      "learning_rate": 2.688495202076994e-06,
+      "loss": 0.9006,
+      "step": 7980
+    },
+    {
+      "epoch": 1.536395793729095,
+      "grad_norm": 3.2697717538707813,
+      "learning_rate": 2.6863684778262786e-06,
+      "loss": 0.9645,
+      "step": 7981
+    },
+    {
+      "epoch": 1.5365883004066703,
+      "grad_norm": 3.2959796547954796,
+      "learning_rate": 2.68424246455412e-06,
+      "loss": 0.9716,
+      "step": 7982
+    },
+    {
+      "epoch": 1.5367808070842457,
+      "grad_norm": 3.3474441174234273,
+      "learning_rate": 2.682117162467186e-06,
+      "loss": 0.9453,
+      "step": 7983
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "grad_norm": 3.204225946582758,
+      "learning_rate": 2.6799925717720866e-06,
+      "loss": 0.9182,
+      "step": 7984
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.7671,
+      "step": 7984,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 1.0129,
+      "step": 7984,
+      "vm_loss": 0.1909
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.8394,
+      "step": 7984,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.6259,
+      "step": 7984,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.4944,
+      "step": 7984,
+      "vm_loss": 0.1261
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.7534,
+      "step": 7984,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.7852,
+      "step": 7984,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.5369733137618211,
+      "lm_loss": 0.8214,
+      "step": 7984,
+      "vm_loss": 0.0996
+    },
+    {
+      "epoch": 1.5371658204393965,
+      "grad_norm": 3.3253669885070973,
+      "learning_rate": 2.6778686926753594e-06,
+      "loss": 0.9288,
+      "step": 7985
+    },
+    {
+      "epoch": 1.537358327116972,
+      "grad_norm": 3.2155316285498348,
+      "learning_rate": 2.6757455253834763e-06,
+      "loss": 0.8838,
+      "step": 7986
+    },
+    {
+      "epoch": 1.5375508337945472,
+      "grad_norm": 3.2043528814507067,
+      "learning_rate": 2.6736230701028266e-06,
+      "loss": 0.9242,
+      "step": 7987
+    },
+    {
+      "epoch": 1.5377433404721226,
+      "grad_norm": 3.2211581773656515,
+      "learning_rate": 2.6715013270397517e-06,
+      "loss": 0.9192,
+      "step": 7988
+    },
+    {
+      "epoch": 1.537935847149698,
+      "grad_norm": 3.364400008071222,
+      "learning_rate": 2.6693802964005046e-06,
+      "loss": 0.9668,
+      "step": 7989
+    },
+    {
+      "epoch": 1.5381283538272734,
+      "grad_norm": 3.2061185646199495,
+      "learning_rate": 2.667259978391281e-06,
+      "loss": 0.9732,
+      "step": 7990
+    },
+    {
+      "epoch": 1.5383208605048488,
+      "grad_norm": 3.314719819803128,
+      "learning_rate": 2.6651403732181948e-06,
+      "loss": 0.9724,
+      "step": 7991
+    },
+    {
+      "epoch": 1.538513367182424,
+      "grad_norm": 3.3206698577924842,
+      "learning_rate": 2.66302148108731e-06,
+      "loss": 0.9471,
+      "step": 7992
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 0.9808,
+      "step": 7992,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 1.0178,
+      "step": 7992,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 1.1058,
+      "step": 7992,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 0.6054,
+      "step": 7992,
+      "vm_loss": 0.2311
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 0.4173,
+      "step": 7992,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 0.821,
+      "step": 7992,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 0.6285,
+      "step": 7992,
+      "vm_loss": 0.1254
+    },
+    {
+      "epoch": 1.538513367182424,
+      "lm_loss": 1.0019,
+      "step": 7992,
+      "vm_loss": 0.1696
+    },
+    {
+      "epoch": 1.5387058738599997,
+      "grad_norm": 3.2942341475937136,
+      "learning_rate": 2.6609033022046026e-06,
+      "loss": 0.9568,
+      "step": 7993
+    },
+    {
+      "epoch": 1.5388983805375749,
+      "grad_norm": 3.133324745317571,
+      "learning_rate": 2.6587858367759878e-06,
+      "loss": 0.8964,
+      "step": 7994
+    },
+    {
+      "epoch": 1.5390908872151503,
+      "grad_norm": 3.3536422986677663,
+      "learning_rate": 2.656669085007312e-06,
+      "loss": 0.9711,
+      "step": 7995
+    },
+    {
+      "epoch": 1.5392833938927257,
+      "grad_norm": 3.328822575143836,
+      "learning_rate": 2.654553047104349e-06,
+      "loss": 0.9891,
+      "step": 7996
+    },
+    {
+      "epoch": 1.539475900570301,
+      "grad_norm": 3.211263241867042,
+      "learning_rate": 2.652437723272806e-06,
+      "loss": 0.943,
+      "step": 7997
+    },
+    {
+      "epoch": 1.5396684072478766,
+      "grad_norm": 3.120363786721185,
+      "learning_rate": 2.6503231137183192e-06,
+      "loss": 0.8712,
+      "step": 7998
+    },
+    {
+      "epoch": 1.5398609139254518,
+      "grad_norm": 3.295522695434396,
+      "learning_rate": 2.648209218646458e-06,
+      "loss": 0.9501,
+      "step": 7999
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "grad_norm": 3.2722126523476027,
+      "learning_rate": 2.6460960382627155e-06,
+      "loss": 0.9344,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.9622,
+      "step": 8000,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 1.1346,
+      "step": 8000,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.5825,
+      "step": 8000,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.8418,
+      "step": 8000,
+      "vm_loss": 0.1129
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 1.0322,
+      "step": 8000,
+      "vm_loss": 0.2188
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.7109,
+      "step": 8000,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.9612,
+      "step": 8000,
+      "vm_loss": 0.2211
+    },
+    {
+      "epoch": 1.5400534206030272,
+      "lm_loss": 0.4593,
+      "step": 8000,
+      "vm_loss": 0.2239
+    },
+    {
+      "epoch": 1.5402459272806026,
+      "grad_norm": 3.383160649722326,
+      "learning_rate": 2.6439835727725217e-06,
+      "loss": 0.9684,
+      "step": 8001
+    },
+    {
+      "epoch": 1.5404384339581778,
+      "grad_norm": 3.187616805335515,
+      "learning_rate": 2.6418718223812357e-06,
+      "loss": 0.9313,
+      "step": 8002
+    },
+    {
+      "epoch": 1.5406309406357535,
+      "grad_norm": 3.3488764601943335,
+      "learning_rate": 2.639760787294149e-06,
+      "loss": 0.9342,
+      "step": 8003
+    },
+    {
+      "epoch": 1.5408234473133287,
+      "grad_norm": 3.421534642422455,
+      "learning_rate": 2.6376504677164727e-06,
+      "loss": 0.9464,
+      "step": 8004
+    },
+    {
+      "epoch": 1.541015953990904,
+      "grad_norm": 3.2180030301567606,
+      "learning_rate": 2.6355408638533684e-06,
+      "loss": 0.8909,
+      "step": 8005
+    },
+    {
+      "epoch": 1.5412084606684795,
+      "grad_norm": 3.225555599456573,
+      "learning_rate": 2.633431975909909e-06,
+      "loss": 0.944,
+      "step": 8006
+    },
+    {
+      "epoch": 1.5414009673460547,
+      "grad_norm": 3.1808055368756083,
+      "learning_rate": 2.6313238040911103e-06,
+      "loss": 0.9155,
+      "step": 8007
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "grad_norm": 3.223747635586637,
+      "learning_rate": 2.629216348601905e-06,
+      "loss": 0.8708,
+      "step": 8008
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.7937,
+      "step": 8008,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.9909,
+      "step": 8008,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 1.0529,
+      "step": 8008,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.4188,
+      "step": 8008,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.78,
+      "step": 8008,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.6115,
+      "step": 8008,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.7757,
+      "step": 8008,
+      "vm_loss": 0.2248
+    },
+    {
+      "epoch": 1.5415934740236303,
+      "lm_loss": 0.5175,
+      "step": 8008,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.5417859807012055,
+      "grad_norm": 3.3587088199145763,
+      "learning_rate": 2.627109609647176e-06,
+      "loss": 0.9678,
+      "step": 8009
+    },
+    {
+      "epoch": 1.541978487378781,
+      "grad_norm": 3.2493155929426507,
+      "learning_rate": 2.625003587431717e-06,
+      "loss": 0.9342,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5421709940563564,
+      "grad_norm": 3.1938556153950266,
+      "learning_rate": 2.6228982821602623e-06,
+      "loss": 0.8509,
+      "step": 8011
+    },
+    {
+      "epoch": 1.5423635007339316,
+      "grad_norm": 3.2148264232840216,
+      "learning_rate": 2.6207936940374767e-06,
+      "loss": 0.9426,
+      "step": 8012
+    },
+    {
+      "epoch": 1.5425560074115072,
+      "grad_norm": 3.3948024793351066,
+      "learning_rate": 2.6186898232679534e-06,
+      "loss": 0.9279,
+      "step": 8013
+    },
+    {
+      "epoch": 1.5427485140890824,
+      "grad_norm": 3.279163267063515,
+      "learning_rate": 2.616586670056208e-06,
+      "loss": 0.9461,
+      "step": 8014
+    },
+    {
+      "epoch": 1.5429410207666578,
+      "grad_norm": 3.2778972656379155,
+      "learning_rate": 2.6144842346067067e-06,
+      "loss": 0.948,
+      "step": 8015
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "grad_norm": 3.287355788110965,
+      "learning_rate": 2.6123825171238216e-06,
+      "loss": 0.9588,
+      "step": 8016
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 1.1373,
+      "step": 8016,
+      "vm_loss": 0.1146
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 0.8576,
+      "step": 8016,
+      "vm_loss": 0.1391
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 1.014,
+      "step": 8016,
+      "vm_loss": 0.1559
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 0.4777,
+      "step": 8016,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 0.8893,
+      "step": 8016,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 0.6298,
+      "step": 8016,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 0.8481,
+      "step": 8016,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.5431335274442333,
+      "lm_loss": 1.1882,
+      "step": 8016,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.5433260341218085,
+      "grad_norm": 3.248078114485856,
+      "learning_rate": 2.610281517811872e-06,
+      "loss": 0.9421,
+      "step": 8017
+    },
+    {
+      "epoch": 1.543518540799384,
+      "grad_norm": 3.414508479095204,
+      "learning_rate": 2.6081812368751015e-06,
+      "loss": 1.026,
+      "step": 8018
+    },
+    {
+      "epoch": 1.5437110474769593,
+      "grad_norm": 3.3334142730825502,
+      "learning_rate": 2.606081674517683e-06,
+      "loss": 0.9362,
+      "step": 8019
+    },
+    {
+      "epoch": 1.5439035541545347,
+      "grad_norm": 3.2821418238884985,
+      "learning_rate": 2.6039828309437245e-06,
+      "loss": 0.9567,
+      "step": 8020
+    },
+    {
+      "epoch": 1.5440960608321102,
+      "grad_norm": 3.2290554862601826,
+      "learning_rate": 2.6018847063572516e-06,
+      "loss": 0.9544,
+      "step": 8021
+    },
+    {
+      "epoch": 1.5442885675096854,
+      "grad_norm": 3.1612207850093474,
+      "learning_rate": 2.599787300962241e-06,
+      "loss": 0.9301,
+      "step": 8022
+    },
+    {
+      "epoch": 1.544481074187261,
+      "grad_norm": 3.18829769328562,
+      "learning_rate": 2.597690614962578e-06,
+      "loss": 0.9023,
+      "step": 8023
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "grad_norm": 3.2561905459158087,
+      "learning_rate": 2.595594648562093e-06,
+      "loss": 0.9232,
+      "step": 8024
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 0.7379,
+      "step": 8024,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 0.4718,
+      "step": 8024,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 1.0247,
+      "step": 8024,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 0.7464,
+      "step": 8024,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 0.8225,
+      "step": 8024,
+      "vm_loss": 0.2319
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 0.5027,
+      "step": 8024,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 1.1778,
+      "step": 8024,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.5446735808648362,
+      "lm_loss": 1.0253,
+      "step": 8024,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.5448660875424116,
+      "grad_norm": 3.2077979756440507,
+      "learning_rate": 2.593499401964532e-06,
+      "loss": 0.9481,
+      "step": 8025
+    },
+    {
+      "epoch": 1.545058594219987,
+      "grad_norm": 3.337386603929552,
+      "learning_rate": 2.591404875373593e-06,
+      "loss": 0.9477,
+      "step": 8026
+    },
+    {
+      "epoch": 1.5452511008975622,
+      "grad_norm": 3.3735205655263134,
+      "learning_rate": 2.58931106899288e-06,
+      "loss": 0.9571,
+      "step": 8027
+    },
+    {
+      "epoch": 1.5454436075751379,
+      "grad_norm": 3.0976357802038845,
+      "learning_rate": 2.5872179830259413e-06,
+      "loss": 0.914,
+      "step": 8028
+    },
+    {
+      "epoch": 1.545636114252713,
+      "grad_norm": 3.341977295056129,
+      "learning_rate": 2.585125617676253e-06,
+      "loss": 0.9215,
+      "step": 8029
+    },
+    {
+      "epoch": 1.5458286209302885,
+      "grad_norm": 3.2096451182066077,
+      "learning_rate": 2.583033973147221e-06,
+      "loss": 0.9102,
+      "step": 8030
+    },
+    {
+      "epoch": 1.546021127607864,
+      "grad_norm": 3.4217362382686503,
+      "learning_rate": 2.5809430496421707e-06,
+      "loss": 0.9648,
+      "step": 8031
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "grad_norm": 3.205804446345324,
+      "learning_rate": 2.57885284736438e-06,
+      "loss": 0.9243,
+      "step": 8032
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.74,
+      "step": 8032,
+      "vm_loss": 0.0934
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.4432,
+      "step": 8032,
+      "vm_loss": 0.0986
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.5572,
+      "step": 8032,
+      "vm_loss": 0.2166
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.9402,
+      "step": 8032,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.7416,
+      "step": 8032,
+      "vm_loss": 0.2138
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.7241,
+      "step": 8032,
+      "vm_loss": 0.2004
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.6982,
+      "step": 8032,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.5462136342854391,
+      "lm_loss": 0.5453,
+      "step": 8032,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 1.5464061409630148,
+      "grad_norm": 3.2343987829262457,
+      "learning_rate": 2.5767633665170343e-06,
+      "loss": 0.9199,
+      "step": 8033
+    },
+    {
+      "epoch": 1.54659864764059,
+      "grad_norm": 3.3110251423523716,
+      "learning_rate": 2.5746746073032624e-06,
+      "loss": 0.9565,
+      "step": 8034
+    },
+    {
+      "epoch": 1.5467911543181654,
+      "grad_norm": 3.3682507190932864,
+      "learning_rate": 2.572586569926111e-06,
+      "loss": 0.8937,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5469836609957408,
+      "grad_norm": 3.449785579991025,
+      "learning_rate": 2.5704992545885755e-06,
+      "loss": 1.0068,
+      "step": 8036
+    },
+    {
+      "epoch": 1.547176167673316,
+      "grad_norm": 3.315921617406086,
+      "learning_rate": 2.56841266149356e-06,
+      "loss": 0.9175,
+      "step": 8037
+    },
+    {
+      "epoch": 1.5473686743508916,
+      "grad_norm": 3.214776155482678,
+      "learning_rate": 2.5663267908439115e-06,
+      "loss": 0.8991,
+      "step": 8038
+    },
+    {
+      "epoch": 1.5475611810284668,
+      "grad_norm": 3.177546508328573,
+      "learning_rate": 2.564241642842403e-06,
+      "loss": 0.9115,
+      "step": 8039
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "grad_norm": 3.3409315281395693,
+      "learning_rate": 2.5621572176917387e-06,
+      "loss": 0.9464,
+      "step": 8040
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.7771,
+      "step": 8040,
+      "vm_loss": 0.2134
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.8359,
+      "step": 8040,
+      "vm_loss": 0.1553
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.6471,
+      "step": 8040,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.6839,
+      "step": 8040,
+      "vm_loss": 0.2457
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.5652,
+      "step": 8040,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.7827,
+      "step": 8040,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.4397,
+      "step": 8040,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.5477536877060423,
+      "lm_loss": 0.9897,
+      "step": 8040,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 1.5479461943836177,
+      "grad_norm": 3.414174748023272,
+      "learning_rate": 2.560073515594553e-06,
+      "loss": 0.997,
+      "step": 8041
+    },
+    {
+      "epoch": 1.5481387010611931,
+      "grad_norm": 3.1531722194752265,
+      "learning_rate": 2.557990536753401e-06,
+      "loss": 0.9037,
+      "step": 8042
+    },
+    {
+      "epoch": 1.5483312077387685,
+      "grad_norm": 3.057843892580745,
+      "learning_rate": 2.555908281370786e-06,
+      "loss": 0.8872,
+      "step": 8043
+    },
+    {
+      "epoch": 1.5485237144163437,
+      "grad_norm": 3.2116494676564162,
+      "learning_rate": 2.55382674964912e-06,
+      "loss": 0.8925,
+      "step": 8044
+    },
+    {
+      "epoch": 1.5487162210939192,
+      "grad_norm": 3.2344749266557287,
+      "learning_rate": 2.5517459417907575e-06,
+      "loss": 0.942,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5489087277714946,
+      "grad_norm": 3.104221482293591,
+      "learning_rate": 2.5496658579979815e-06,
+      "loss": 0.8619,
+      "step": 8046
+    },
+    {
+      "epoch": 1.54910123444907,
+      "grad_norm": 5.665101920938779,
+      "learning_rate": 2.5475864984730047e-06,
+      "loss": 0.8937,
+      "step": 8047
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "grad_norm": 3.299530216755856,
+      "learning_rate": 2.5455078634179585e-06,
+      "loss": 0.9431,
+      "step": 8048
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.792,
+      "step": 8048,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.6506,
+      "step": 8048,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.6503,
+      "step": 8048,
+      "vm_loss": 0.1017
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.7102,
+      "step": 8048,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.5696,
+      "step": 8048,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.5421,
+      "step": 8048,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 1.3349,
+      "step": 8048,
+      "vm_loss": 0.2416
+    },
+    {
+      "epoch": 1.5492937411266454,
+      "lm_loss": 0.7006,
+      "step": 8048,
+      "vm_loss": 0.0982
+    },
+    {
+      "epoch": 1.5494862478042206,
+      "grad_norm": 3.385589099523751,
+      "learning_rate": 2.543429953034925e-06,
+      "loss": 0.9719,
+      "step": 8049
+    },
+    {
+      "epoch": 1.549678754481796,
+      "grad_norm": 3.3664696626675985,
+      "learning_rate": 2.541352767525894e-06,
+      "loss": 0.9519,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5498712611593715,
+      "grad_norm": 3.3575633284852198,
+      "learning_rate": 2.5392763070928006e-06,
+      "loss": 0.9488,
+      "step": 8051
+    },
+    {
+      "epoch": 1.5500637678369469,
+      "grad_norm": 3.3275734245013973,
+      "learning_rate": 2.5372005719374957e-06,
+      "loss": 0.9372,
+      "step": 8052
+    },
+    {
+      "epoch": 1.5502562745145223,
+      "grad_norm": 3.4242098745925893,
+      "learning_rate": 2.5351255622617775e-06,
+      "loss": 0.954,
+      "step": 8053
+    },
+    {
+      "epoch": 1.5504487811920975,
+      "grad_norm": 3.14201328970626,
+      "learning_rate": 2.5330512782673554e-06,
+      "loss": 0.8365,
+      "step": 8054
+    },
+    {
+      "epoch": 1.550641287869673,
+      "grad_norm": 3.280764839740856,
+      "learning_rate": 2.5309777201558795e-06,
+      "loss": 0.9086,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "grad_norm": 3.263028325756709,
+      "learning_rate": 2.5289048881289256e-06,
+      "loss": 0.9177,
+      "step": 8056
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.6132,
+      "step": 8056,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.349,
+      "step": 8056,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.8607,
+      "step": 8056,
+      "vm_loss": 0.1151
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.4599,
+      "step": 8056,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.5041,
+      "step": 8056,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.6696,
+      "step": 8056,
+      "vm_loss": 0.1256
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.6461,
+      "step": 8056,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 1.5508337945472483,
+      "lm_loss": 0.582,
+      "step": 8056,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 1.5510263012248238,
+      "grad_norm": 3.121599770952913,
+      "learning_rate": 2.5268327823880036e-06,
+      "loss": 0.865,
+      "step": 8057
+    },
+    {
+      "epoch": 1.5512188079023992,
+      "grad_norm": 3.425979639262565,
+      "learning_rate": 2.5247614031345412e-06,
+      "loss": 0.9736,
+      "step": 8058
+    },
+    {
+      "epoch": 1.5514113145799744,
+      "grad_norm": 3.189515678924552,
+      "learning_rate": 2.522690750569906e-06,
+      "loss": 0.9005,
+      "step": 8059
+    },
+    {
+      "epoch": 1.55160382125755,
+      "grad_norm": 3.3124622292775445,
+      "learning_rate": 2.5206208248953932e-06,
+      "loss": 0.9535,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5517963279351252,
+      "grad_norm": 3.3259167107284067,
+      "learning_rate": 2.5185516263122256e-06,
+      "loss": 0.9381,
+      "step": 8061
+    },
+    {
+      "epoch": 1.5519888346127007,
+      "grad_norm": 3.2442572878533675,
+      "learning_rate": 2.5164831550215552e-06,
+      "loss": 0.9086,
+      "step": 8062
+    },
+    {
+      "epoch": 1.552181341290276,
+      "grad_norm": 3.2982612636229045,
+      "learning_rate": 2.5144154112244647e-06,
+      "loss": 0.9348,
+      "step": 8063
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "grad_norm": 3.3743817947078125,
+      "learning_rate": 2.512348395121967e-06,
+      "loss": 0.9845,
+      "step": 8064
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.8487,
+      "step": 8064,
+      "vm_loss": 0.1124
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 1.189,
+      "step": 8064,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.7536,
+      "step": 8064,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.5355,
+      "step": 8064,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.9571,
+      "step": 8064,
+      "vm_loss": 0.1878
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.5759,
+      "step": 8064,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.8864,
+      "step": 8064,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.5523738479678513,
+      "lm_loss": 0.6029,
+      "step": 8064,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 1.552566354645427,
+      "grad_norm": 3.3081428980958436,
+      "learning_rate": 2.5102821069149963e-06,
+      "loss": 0.9282,
+      "step": 8065
+    },
+    {
+      "epoch": 1.5527588613230021,
+      "grad_norm": 3.32069548727019,
+      "learning_rate": 2.5082165468044306e-06,
+      "loss": 0.9059,
+      "step": 8066
+    },
+    {
+      "epoch": 1.5529513680005775,
+      "grad_norm": 3.187461595431289,
+      "learning_rate": 2.506151714991062e-06,
+      "loss": 0.9012,
+      "step": 8067
+    },
+    {
+      "epoch": 1.553143874678153,
+      "grad_norm": 3.289097079521282,
+      "learning_rate": 2.504087611675623e-06,
+      "loss": 0.9233,
+      "step": 8068
+    },
+    {
+      "epoch": 1.5533363813557282,
+      "grad_norm": 3.430215447197617,
+      "learning_rate": 2.5020242370587643e-06,
+      "loss": 0.9548,
+      "step": 8069
+    },
+    {
+      "epoch": 1.5535288880333038,
+      "grad_norm": 3.3077061842860265,
+      "learning_rate": 2.4999615913410823e-06,
+      "loss": 0.9433,
+      "step": 8070
+    },
+    {
+      "epoch": 1.553721394710879,
+      "grad_norm": 3.28873848326804,
+      "learning_rate": 2.4978996747230834e-06,
+      "loss": 0.9317,
+      "step": 8071
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "grad_norm": 3.109929815527573,
+      "learning_rate": 2.4958384874052166e-06,
+      "loss": 0.8534,
+      "step": 8072
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.7684,
+      "step": 8072,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.905,
+      "step": 8072,
+      "vm_loss": 0.0838
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.9617,
+      "step": 8072,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.7553,
+      "step": 8072,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.7515,
+      "step": 8072,
+      "vm_loss": 0.2147
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.7598,
+      "step": 8072,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 0.7956,
+      "step": 8072,
+      "vm_loss": 0.1254
+    },
+    {
+      "epoch": 1.5539139013884544,
+      "lm_loss": 1.2563,
+      "step": 8072,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 1.5541064080660298,
+      "grad_norm": 3.2748473322767326,
+      "learning_rate": 2.4937780295878557e-06,
+      "loss": 0.9205,
+      "step": 8073
+    },
+    {
+      "epoch": 1.554298914743605,
+      "grad_norm": 3.094646244520794,
+      "learning_rate": 2.4917183014713052e-06,
+      "loss": 0.8799,
+      "step": 8074
+    },
+    {
+      "epoch": 1.5544914214211807,
+      "grad_norm": 3.248429068905733,
+      "learning_rate": 2.4896593032557916e-06,
+      "loss": 0.935,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5546839280987559,
+      "grad_norm": 3.4227510520669346,
+      "learning_rate": 2.4876010351414792e-06,
+      "loss": 0.977,
+      "step": 8076
+    },
+    {
+      "epoch": 1.5548764347763313,
+      "grad_norm": 3.238548932910585,
+      "learning_rate": 2.4855434973284585e-06,
+      "loss": 0.9099,
+      "step": 8077
+    },
+    {
+      "epoch": 1.5550689414539067,
+      "grad_norm": 3.4626305796193044,
+      "learning_rate": 2.4834866900167478e-06,
+      "loss": 1.0208,
+      "step": 8078
+    },
+    {
+      "epoch": 1.555261448131482,
+      "grad_norm": 3.186535890007029,
+      "learning_rate": 2.481430613406295e-06,
+      "loss": 0.9121,
+      "step": 8079
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "grad_norm": 3.151463090537228,
+      "learning_rate": 2.4793752676969795e-06,
+      "loss": 0.8837,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.7908,
+      "step": 8080,
+      "vm_loss": 0.2115
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 1.0147,
+      "step": 8080,
+      "vm_loss": 0.2028
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.8523,
+      "step": 8080,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.4586,
+      "step": 8080,
+      "vm_loss": 0.0933
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.5975,
+      "step": 8080,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.9358,
+      "step": 8080,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 1.3041,
+      "step": 8080,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.5554539548090576,
+      "lm_loss": 0.503,
+      "step": 8080,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 1.5556464614866328,
+      "grad_norm": 3.1700068703120152,
+      "learning_rate": 2.477320653088602e-06,
+      "loss": 0.9198,
+      "step": 8081
+    },
+    {
+      "epoch": 1.5558389681642082,
+      "grad_norm": 3.2992851740245843,
+      "learning_rate": 2.4752667697809007e-06,
+      "loss": 0.9179,
+      "step": 8082
+    },
+    {
+      "epoch": 1.5560314748417836,
+      "grad_norm": 3.1071362393579975,
+      "learning_rate": 2.473213617973538e-06,
+      "loss": 0.9027,
+      "step": 8083
+    },
+    {
+      "epoch": 1.5562239815193588,
+      "grad_norm": 3.184432023927627,
+      "learning_rate": 2.4711611978661076e-06,
+      "loss": 0.882,
+      "step": 8084
+    },
+    {
+      "epoch": 1.5564164881969345,
+      "grad_norm": 3.4120498193014677,
+      "learning_rate": 2.4691095096581343e-06,
+      "loss": 0.9697,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5566089948745097,
+      "grad_norm": 3.227166756512222,
+      "learning_rate": 2.4670585535490586e-06,
+      "loss": 0.9588,
+      "step": 8086
+    },
+    {
+      "epoch": 1.556801501552085,
+      "grad_norm": 3.2756054338265765,
+      "learning_rate": 2.465008329738272e-06,
+      "loss": 0.9371,
+      "step": 8087
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "grad_norm": 3.398754887021963,
+      "learning_rate": 2.4629588384250737e-06,
+      "loss": 0.9529,
+      "step": 8088
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.6432,
+      "step": 8088,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 1.0005,
+      "step": 8088,
+      "vm_loss": 0.1609
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.9258,
+      "step": 8088,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.8778,
+      "step": 8088,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.8446,
+      "step": 8088,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 1.1582,
+      "step": 8088,
+      "vm_loss": 0.1737
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.6223,
+      "step": 8088,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 1.5569940082296605,
+      "lm_loss": 0.5025,
+      "step": 8088,
+      "vm_loss": 0.163
+    },
+    {
+      "epoch": 1.5571865149072357,
+      "grad_norm": 3.261592689440988,
+      "learning_rate": 2.460910079808703e-06,
+      "loss": 0.9176,
+      "step": 8089
+    },
+    {
+      "epoch": 1.5573790215848113,
+      "grad_norm": 3.2467865653768526,
+      "learning_rate": 2.4588620540883257e-06,
+      "loss": 0.9096,
+      "step": 8090
+    },
+    {
+      "epoch": 1.5575715282623865,
+      "grad_norm": 3.2931729713179294,
+      "learning_rate": 2.4568147614630378e-06,
+      "loss": 0.8999,
+      "step": 8091
+    },
+    {
+      "epoch": 1.557764034939962,
+      "grad_norm": 3.2619116303949696,
+      "learning_rate": 2.4547682021318596e-06,
+      "loss": 0.9245,
+      "step": 8092
+    },
+    {
+      "epoch": 1.5579565416175374,
+      "grad_norm": 3.5874144049066996,
+      "learning_rate": 2.4527223762937422e-06,
+      "loss": 0.9962,
+      "step": 8093
+    },
+    {
+      "epoch": 1.5581490482951126,
+      "grad_norm": 3.4376427335455877,
+      "learning_rate": 2.450677284147569e-06,
+      "loss": 0.9681,
+      "step": 8094
+    },
+    {
+      "epoch": 1.5583415549726882,
+      "grad_norm": 3.1498222998796876,
+      "learning_rate": 2.44863292589215e-06,
+      "loss": 0.8592,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "grad_norm": 3.2060332358741226,
+      "learning_rate": 2.4465893017262156e-06,
+      "loss": 0.9425,
+      "step": 8096
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.5833,
+      "step": 8096,
+      "vm_loss": 0.2157
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.7539,
+      "step": 8096,
+      "vm_loss": 0.1131
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 1.1119,
+      "step": 8096,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.8628,
+      "step": 8096,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.5656,
+      "step": 8096,
+      "vm_loss": 0.121
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.5463,
+      "step": 8096,
+      "vm_loss": 0.0986
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 0.9406,
+      "step": 8096,
+      "vm_loss": 0.2596
+    },
+    {
+      "epoch": 1.5585340616502634,
+      "lm_loss": 1.078,
+      "step": 8096,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 1.5587265683278388,
+      "grad_norm": 3.1548312965062,
+      "learning_rate": 2.444546411848443e-06,
+      "loss": 0.9039,
+      "step": 8097
+    },
+    {
+      "epoch": 1.5589190750054143,
+      "grad_norm": 3.20065192290026,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.8895,
+      "step": 8098
+    },
+    {
+      "epoch": 1.5591115816829895,
+      "grad_norm": 3.384712201753457,
+      "learning_rate": 2.440462835751669e-06,
+      "loss": 0.9601,
+      "step": 8099
+    },
+    {
+      "epoch": 1.559304088360565,
+      "grad_norm": 3.24615811520348,
+      "learning_rate": 2.4384221499296466e-06,
+      "loss": 0.9596,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5594965950381403,
+      "grad_norm": 3.254830866558642,
+      "learning_rate": 2.436382199189734e-06,
+      "loss": 0.9482,
+      "step": 8101
+    },
+    {
+      "epoch": 1.5596891017157157,
+      "grad_norm": 3.200481922264514,
+      "learning_rate": 2.4343429837302367e-06,
+      "loss": 0.902,
+      "step": 8102
+    },
+    {
+      "epoch": 1.5598816083932912,
+      "grad_norm": 3.129146938125482,
+      "learning_rate": 2.432304503749394e-06,
+      "loss": 0.9116,
+      "step": 8103
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "grad_norm": 3.21760359488873,
+      "learning_rate": 2.4302667594453734e-06,
+      "loss": 0.8925,
+      "step": 8104
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.8803,
+      "step": 8104,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.9525,
+      "step": 8104,
+      "vm_loss": 0.13
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.5712,
+      "step": 8104,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.7088,
+      "step": 8104,
+      "vm_loss": 0.1951
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.6756,
+      "step": 8104,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.7027,
+      "step": 8104,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.9538,
+      "step": 8104,
+      "vm_loss": 0.1385
+    },
+    {
+      "epoch": 1.5600741150708666,
+      "lm_loss": 0.8504,
+      "step": 8104,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.560266621748442,
+      "grad_norm": 3.142369378834673,
+      "learning_rate": 2.428229751016269e-06,
+      "loss": 0.8685,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5604591284260172,
+      "grad_norm": 3.1771725856632864,
+      "learning_rate": 2.4261934786601025e-06,
+      "loss": 0.9012,
+      "step": 8106
+    },
+    {
+      "epoch": 1.5606516351035926,
+      "grad_norm": 3.100174614819856,
+      "learning_rate": 2.4241579425748284e-06,
+      "loss": 0.8673,
+      "step": 8107
+    },
+    {
+      "epoch": 1.560844141781168,
+      "grad_norm": 3.2751778244121996,
+      "learning_rate": 2.422123142958328e-06,
+      "loss": 0.9254,
+      "step": 8108
+    },
+    {
+      "epoch": 1.5610366484587435,
+      "grad_norm": 3.474570815643663,
+      "learning_rate": 2.4200890800084052e-06,
+      "loss": 0.9307,
+      "step": 8109
+    },
+    {
+      "epoch": 1.5612291551363189,
+      "grad_norm": 3.2521525690403497,
+      "learning_rate": 2.418055753922799e-06,
+      "loss": 0.9234,
+      "step": 8110
+    },
+    {
+      "epoch": 1.561421661813894,
+      "grad_norm": 3.342543265991017,
+      "learning_rate": 2.416023164899175e-06,
+      "loss": 0.9208,
+      "step": 8111
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "grad_norm": 3.312155075900073,
+      "learning_rate": 2.413991313135129e-06,
+      "loss": 0.9177,
+      "step": 8112
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.5547,
+      "step": 8112,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.566,
+      "step": 8112,
+      "vm_loss": 0.1099
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.7185,
+      "step": 8112,
+      "vm_loss": 0.1104
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.442,
+      "step": 8112,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 1.0472,
+      "step": 8112,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.9709,
+      "step": 8112,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.6095,
+      "step": 8112,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.5616141684914695,
+      "lm_loss": 0.328,
+      "step": 8112,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 1.561806675169045,
+      "grad_norm": 3.164265951924517,
+      "learning_rate": 2.411960198828175e-06,
+      "loss": 0.8689,
+      "step": 8113
+    },
+    {
+      "epoch": 1.5619991818466203,
+      "grad_norm": 3.2980978750404506,
+      "learning_rate": 2.409929822175775e-06,
+      "loss": 0.9352,
+      "step": 8114
+    },
+    {
+      "epoch": 1.5621916885241958,
+      "grad_norm": 3.195297586360202,
+      "learning_rate": 2.407900183375299e-06,
+      "loss": 0.9138,
+      "step": 8115
+    },
+    {
+      "epoch": 1.562384195201771,
+      "grad_norm": 3.3897318033117756,
+      "learning_rate": 2.4058712826240595e-06,
+      "loss": 0.9643,
+      "step": 8116
+    },
+    {
+      "epoch": 1.5625767018793464,
+      "grad_norm": 3.144066264093219,
+      "learning_rate": 2.4038431201192825e-06,
+      "loss": 0.8795,
+      "step": 8117
+    },
+    {
+      "epoch": 1.5627692085569218,
+      "grad_norm": 3.2175877368377055,
+      "learning_rate": 2.401815696058143e-06,
+      "loss": 0.9054,
+      "step": 8118
+    },
+    {
+      "epoch": 1.5629617152344972,
+      "grad_norm": 3.2806098294263677,
+      "learning_rate": 2.399789010637725e-06,
+      "loss": 0.9152,
+      "step": 8119
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "grad_norm": 3.2346207666052007,
+      "learning_rate": 2.39776306405505e-06,
+      "loss": 0.8956,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.6953,
+      "step": 8120,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.8863,
+      "step": 8120,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.8472,
+      "step": 8120,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.8541,
+      "step": 8120,
+      "vm_loss": 0.1045
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.8565,
+      "step": 8120,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.3922,
+      "step": 8120,
+      "vm_loss": 0.2042
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 0.5924,
+      "step": 8120,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.5631542219120727,
+      "lm_loss": 1.0281,
+      "step": 8120,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.5633467285896478,
+      "grad_norm": 3.1363226237620894,
+      "learning_rate": 2.3957378565070656e-06,
+      "loss": 0.8878,
+      "step": 8121
+    },
+    {
+      "epoch": 1.5635392352672235,
+      "grad_norm": 3.204317566045012,
+      "learning_rate": 2.3937133881906484e-06,
+      "loss": 0.8759,
+      "step": 8122
+    },
+    {
+      "epoch": 1.5637317419447987,
+      "grad_norm": 3.154909219088451,
+      "learning_rate": 2.3916896593026027e-06,
+      "loss": 0.8993,
+      "step": 8123
+    },
+    {
+      "epoch": 1.5639242486223741,
+      "grad_norm": 3.1829763045546478,
+      "learning_rate": 2.389666670039664e-06,
+      "loss": 0.9449,
+      "step": 8124
+    },
+    {
+      "epoch": 1.5641167552999495,
+      "grad_norm": 3.280915198095316,
+      "learning_rate": 2.3876444205984872e-06,
+      "loss": 0.917,
+      "step": 8125
+    },
+    {
+      "epoch": 1.5643092619775247,
+      "grad_norm": 3.139616102539235,
+      "learning_rate": 2.3856229111756622e-06,
+      "loss": 0.8976,
+      "step": 8126
+    },
+    {
+      "epoch": 1.5645017686551004,
+      "grad_norm": 3.2043698280066177,
+      "learning_rate": 2.383602141967708e-06,
+      "loss": 0.9345,
+      "step": 8127
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "grad_norm": 3.1811792140662,
+      "learning_rate": 2.381582113171066e-06,
+      "loss": 0.8973,
+      "step": 8128
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.6974,
+      "step": 8128,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.817,
+      "step": 8128,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.7631,
+      "step": 8128,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.5283,
+      "step": 8128,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.7091,
+      "step": 8128,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.5502,
+      "step": 8128,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 0.7149,
+      "step": 8128,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 1.5646942753326756,
+      "lm_loss": 1.6096,
+      "step": 8128,
+      "vm_loss": 0.1155
+    },
+    {
+      "epoch": 1.564886782010251,
+      "grad_norm": 3.340787377345133,
+      "learning_rate": 2.3795628249821147e-06,
+      "loss": 0.9549,
+      "step": 8129
+    },
+    {
+      "epoch": 1.5650792886878264,
+      "grad_norm": 3.2600492598972135,
+      "learning_rate": 2.3775442775971448e-06,
+      "loss": 0.936,
+      "step": 8130
+    },
+    {
+      "epoch": 1.5652717953654016,
+      "grad_norm": 3.2822791534156877,
+      "learning_rate": 2.375526471212396e-06,
+      "loss": 0.9235,
+      "step": 8131
+    },
+    {
+      "epoch": 1.5654643020429773,
+      "grad_norm": 3.282186614362665,
+      "learning_rate": 2.373509406024017e-06,
+      "loss": 0.9035,
+      "step": 8132
+    },
+    {
+      "epoch": 1.5656568087205525,
+      "grad_norm": 3.3265425431706808,
+      "learning_rate": 2.371493082228098e-06,
+      "loss": 0.9147,
+      "step": 8133
+    },
+    {
+      "epoch": 1.5658493153981279,
+      "grad_norm": 3.294296197470225,
+      "learning_rate": 2.3694775000206427e-06,
+      "loss": 0.915,
+      "step": 8134
+    },
+    {
+      "epoch": 1.5660418220757033,
+      "grad_norm": 3.2582322486076905,
+      "learning_rate": 2.3674626595976037e-06,
+      "loss": 0.8999,
+      "step": 8135
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "grad_norm": 3.3380259469162965,
+      "learning_rate": 2.3654485611548394e-06,
+      "loss": 0.913,
+      "step": 8136
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 1.05,
+      "step": 8136,
+      "vm_loss": 0.1354
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.6783,
+      "step": 8136,
+      "vm_loss": 0.0798
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.7606,
+      "step": 8136,
+      "vm_loss": 0.2127
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.8195,
+      "step": 8136,
+      "vm_loss": 0.2809
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.6626,
+      "step": 8136,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.4933,
+      "step": 8136,
+      "vm_loss": 0.0801
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.7835,
+      "step": 8136,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.5662343287532785,
+      "lm_loss": 0.38,
+      "step": 8136,
+      "vm_loss": 0.1031
+    },
+    {
+      "epoch": 1.5664268354308541,
+      "grad_norm": 3.1797893554798224,
+      "learning_rate": 2.363435204888149e-06,
+      "loss": 0.8812,
+      "step": 8137
+    },
+    {
+      "epoch": 1.5666193421084293,
+      "grad_norm": 3.2443241293356406,
+      "learning_rate": 2.3614225909932575e-06,
+      "loss": 0.9267,
+      "step": 8138
+    },
+    {
+      "epoch": 1.5668118487860048,
+      "grad_norm": 3.124435209508619,
+      "learning_rate": 2.359410719665819e-06,
+      "loss": 0.8866,
+      "step": 8139
+    },
+    {
+      "epoch": 1.5670043554635802,
+      "grad_norm": 3.2166079328069515,
+      "learning_rate": 2.357399591101404e-06,
+      "loss": 0.9022,
+      "step": 8140
+    },
+    {
+      "epoch": 1.5671968621411554,
+      "grad_norm": 3.3804288754115923,
+      "learning_rate": 2.3553892054955317e-06,
+      "loss": 0.9399,
+      "step": 8141
+    },
+    {
+      "epoch": 1.567389368818731,
+      "grad_norm": 3.2918410629600627,
+      "learning_rate": 2.3533795630436297e-06,
+      "loss": 0.9403,
+      "step": 8142
+    },
+    {
+      "epoch": 1.5675818754963062,
+      "grad_norm": 3.350468217281528,
+      "learning_rate": 2.3513706639410625e-06,
+      "loss": 0.9298,
+      "step": 8143
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "grad_norm": 3.0478209408176897,
+      "learning_rate": 2.3493625083831217e-06,
+      "loss": 0.863,
+      "step": 8144
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.8374,
+      "step": 8144,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.9028,
+      "step": 8144,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.7882,
+      "step": 8144,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.8617,
+      "step": 8144,
+      "vm_loss": 0.152
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.6157,
+      "step": 8144,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.6503,
+      "step": 8144,
+      "vm_loss": 0.1857
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.7442,
+      "step": 8144,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.5677743821738817,
+      "lm_loss": 0.4915,
+      "step": 8144,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.567966888851457,
+      "grad_norm": 3.1616191401000275,
+      "learning_rate": 2.347355096565025e-06,
+      "loss": 0.9028,
+      "step": 8145
+    },
+    {
+      "epoch": 1.5681593955290323,
+      "grad_norm": 3.3562843850140274,
+      "learning_rate": 2.345348428681922e-06,
+      "loss": 0.9469,
+      "step": 8146
+    },
+    {
+      "epoch": 1.568351902206608,
+      "grad_norm": 3.2790231434199217,
+      "learning_rate": 2.3433425049288806e-06,
+      "loss": 0.9308,
+      "step": 8147
+    },
+    {
+      "epoch": 1.5685444088841831,
+      "grad_norm": 3.275086397844839,
+      "learning_rate": 2.341337325500904e-06,
+      "loss": 0.9008,
+      "step": 8148
+    },
+    {
+      "epoch": 1.5687369155617585,
+      "grad_norm": 3.2917473904467895,
+      "learning_rate": 2.3393328905929225e-06,
+      "loss": 0.9573,
+      "step": 8149
+    },
+    {
+      "epoch": 1.568929422239334,
+      "grad_norm": 3.153765642666675,
+      "learning_rate": 2.3373292003997963e-06,
+      "loss": 0.8683,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5691219289169092,
+      "grad_norm": 3.357799039872151,
+      "learning_rate": 2.3353262551162993e-06,
+      "loss": 0.9611,
+      "step": 8151
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "grad_norm": 3.15244956783198,
+      "learning_rate": 2.3333240549371572e-06,
+      "loss": 0.8572,
+      "step": 8152
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.7314,
+      "step": 8152,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 1.1789,
+      "step": 8152,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.4106,
+      "step": 8152,
+      "vm_loss": 0.0864
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.6878,
+      "step": 8152,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.9103,
+      "step": 8152,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.3278,
+      "step": 8152,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.8873,
+      "step": 8152,
+      "vm_loss": 0.1153
+    },
+    {
+      "epoch": 1.5693144355944848,
+      "lm_loss": 0.9183,
+      "step": 8152,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.56950694227206,
+      "grad_norm": 3.3885515098720504,
+      "learning_rate": 2.3313226000569975e-06,
+      "loss": 0.9583,
+      "step": 8153
+    },
+    {
+      "epoch": 1.5696994489496354,
+      "grad_norm": 3.178625228102367,
+      "learning_rate": 2.329321890670394e-06,
+      "loss": 0.9031,
+      "step": 8154
+    },
+    {
+      "epoch": 1.5698919556272108,
+      "grad_norm": 2.999660373423965,
+      "learning_rate": 2.3273219269718384e-06,
+      "loss": 0.8263,
+      "step": 8155
+    },
+    {
+      "epoch": 1.570084462304786,
+      "grad_norm": 3.397643679766613,
+      "learning_rate": 2.325322709155757e-06,
+      "loss": 0.9731,
+      "step": 8156
+    },
+    {
+      "epoch": 1.5702769689823617,
+      "grad_norm": 3.230941346911783,
+      "learning_rate": 2.3233242374164933e-06,
+      "loss": 0.8917,
+      "step": 8157
+    },
+    {
+      "epoch": 1.5704694756599369,
+      "grad_norm": 3.264226063388696,
+      "learning_rate": 2.3213265119483265e-06,
+      "loss": 0.9428,
+      "step": 8158
+    },
+    {
+      "epoch": 1.5706619823375123,
+      "grad_norm": 3.3648367888556803,
+      "learning_rate": 2.319329532945461e-06,
+      "loss": 0.9777,
+      "step": 8159
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "grad_norm": 3.207624410673108,
+      "learning_rate": 2.3173333006020337e-06,
+      "loss": 0.8871,
+      "step": 8160
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.6826,
+      "step": 8160,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 1.0295,
+      "step": 8160,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.9451,
+      "step": 8160,
+      "vm_loss": 0.1033
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.6799,
+      "step": 8160,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.5979,
+      "step": 8160,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 1.0406,
+      "step": 8160,
+      "vm_loss": 0.1291
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.5861,
+      "step": 8160,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 1.5708544890150877,
+      "lm_loss": 0.6189,
+      "step": 8160,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.571046995692663,
+      "grad_norm": 3.1941149249792895,
+      "learning_rate": 2.315337815112093e-06,
+      "loss": 0.8958,
+      "step": 8161
+    },
+    {
+      "epoch": 1.5712395023702386,
+      "grad_norm": 3.257168598373191,
+      "learning_rate": 2.3133430766696383e-06,
+      "loss": 0.9373,
+      "step": 8162
+    },
+    {
+      "epoch": 1.5714320090478138,
+      "grad_norm": 3.3551957656097064,
+      "learning_rate": 2.311349085468575e-06,
+      "loss": 0.9508,
+      "step": 8163
+    },
+    {
+      "epoch": 1.5716245157253892,
+      "grad_norm": 3.2288193390098026,
+      "learning_rate": 2.3093558417027473e-06,
+      "loss": 0.9308,
+      "step": 8164
+    },
+    {
+      "epoch": 1.5718170224029646,
+      "grad_norm": 3.377165830861174,
+      "learning_rate": 2.3073633455659227e-06,
+      "loss": 0.9787,
+      "step": 8165
+    },
+    {
+      "epoch": 1.57200952908054,
+      "grad_norm": 3.2092360868528442,
+      "learning_rate": 2.3053715972517998e-06,
+      "loss": 0.9094,
+      "step": 8166
+    },
+    {
+      "epoch": 1.5722020357581155,
+      "grad_norm": 3.0851227318352046,
+      "learning_rate": 2.3033805969540024e-06,
+      "loss": 0.9008,
+      "step": 8167
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "grad_norm": 3.2585858081746575,
+      "learning_rate": 2.3013903448660745e-06,
+      "loss": 0.9015,
+      "step": 8168
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.6834,
+      "step": 8168,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.6106,
+      "step": 8168,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.6573,
+      "step": 8168,
+      "vm_loss": 0.2406
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.6251,
+      "step": 8168,
+      "vm_loss": 0.2209
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.7978,
+      "step": 8168,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.6573,
+      "step": 8168,
+      "vm_loss": 0.1156
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.5448,
+      "step": 8168,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 1.5723945424356907,
+      "lm_loss": 0.9146,
+      "step": 8168,
+      "vm_loss": 0.1491
+    },
+    {
+      "epoch": 1.572587049113266,
+      "grad_norm": 3.3144961416756518,
+      "learning_rate": 2.299400841181504e-06,
+      "loss": 0.9481,
+      "step": 8169
+    },
+    {
+      "epoch": 1.5727795557908415,
+      "grad_norm": 3.188380465015555,
+      "learning_rate": 2.297412086093689e-06,
+      "loss": 0.9222,
+      "step": 8170
+    },
+    {
+      "epoch": 1.572972062468417,
+      "grad_norm": 3.19698615939922,
+      "learning_rate": 2.2954240797959635e-06,
+      "loss": 0.8456,
+      "step": 8171
+    },
+    {
+      "epoch": 1.5731645691459923,
+      "grad_norm": 3.1844655735565732,
+      "learning_rate": 2.2934368224815894e-06,
+      "loss": 0.9159,
+      "step": 8172
+    },
+    {
+      "epoch": 1.5733570758235675,
+      "grad_norm": 3.2158803168224974,
+      "learning_rate": 2.2914503143437526e-06,
+      "loss": 0.9468,
+      "step": 8173
+    },
+    {
+      "epoch": 1.573549582501143,
+      "grad_norm": 3.283085889625685,
+      "learning_rate": 2.2894645555755645e-06,
+      "loss": 0.9192,
+      "step": 8174
+    },
+    {
+      "epoch": 1.5737420891787184,
+      "grad_norm": 3.2326747144532852,
+      "learning_rate": 2.287479546370068e-06,
+      "loss": 0.8958,
+      "step": 8175
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "grad_norm": 3.353028360845267,
+      "learning_rate": 2.285495286920232e-06,
+      "loss": 0.9647,
+      "step": 8176
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.8348,
+      "step": 8176,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.5109,
+      "step": 8176,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.5087,
+      "step": 8176,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.8325,
+      "step": 8176,
+      "vm_loss": 0.2366
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.6733,
+      "step": 8176,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.9432,
+      "step": 8176,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.6622,
+      "step": 8176,
+      "vm_loss": 0.194
+    },
+    {
+      "epoch": 1.5739345958562938,
+      "lm_loss": 0.8813,
+      "step": 8176,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.5741271025338692,
+      "grad_norm": 3.3527984505299386,
+      "learning_rate": 2.283511777418954e-06,
+      "loss": 0.9355,
+      "step": 8177
+    },
+    {
+      "epoch": 1.5743196092114444,
+      "grad_norm": 3.006069459448071,
+      "learning_rate": 2.281529018059049e-06,
+      "loss": 0.8631,
+      "step": 8178
+    },
+    {
+      "epoch": 1.5745121158890198,
+      "grad_norm": 3.257133935818751,
+      "learning_rate": 2.2795470090332762e-06,
+      "loss": 0.9251,
+      "step": 8179
+    },
+    {
+      "epoch": 1.5747046225665953,
+      "grad_norm": 3.173798654869818,
+      "learning_rate": 2.277565750534305e-06,
+      "loss": 0.8903,
+      "step": 8180
+    },
+    {
+      "epoch": 1.5748971292441707,
+      "grad_norm": 3.1206637015978975,
+      "learning_rate": 2.275585242754741e-06,
+      "loss": 0.8856,
+      "step": 8181
+    },
+    {
+      "epoch": 1.5750896359217461,
+      "grad_norm": 3.272775075891993,
+      "learning_rate": 2.273605485887116e-06,
+      "loss": 0.9306,
+      "step": 8182
+    },
+    {
+      "epoch": 1.5752821425993213,
+      "grad_norm": 3.3119786055097755,
+      "learning_rate": 2.2716264801238908e-06,
+      "loss": 0.9059,
+      "step": 8183
+    },
+    {
+      "epoch": 1.575474649276897,
+      "grad_norm": 3.115321483924805,
+      "learning_rate": 2.2696482256574426e-06,
+      "loss": 0.9011,
+      "step": 8184
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.8412,
+      "step": 8184,
+      "vm_loss": 0.1364
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.872,
+      "step": 8184,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.6991,
+      "step": 8184,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.8603,
+      "step": 8184,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.9292,
+      "step": 8184,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.9954,
+      "step": 8184,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.7913,
+      "step": 8184,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.575474649276897,
+      "lm_loss": 0.5971,
+      "step": 8184,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.5756671559544722,
+      "grad_norm": 3.293245656035095,
+      "learning_rate": 2.2676707226800876e-06,
+      "loss": 0.9186,
+      "step": 8185
+    },
+    {
+      "epoch": 1.5758596626320476,
+      "grad_norm": 3.2870770804201244,
+      "learning_rate": 2.265693971384063e-06,
+      "loss": 0.9065,
+      "step": 8186
+    },
+    {
+      "epoch": 1.576052169309623,
+      "grad_norm": 3.251360928941749,
+      "learning_rate": 2.2637179719615367e-06,
+      "loss": 0.9301,
+      "step": 8187
+    },
+    {
+      "epoch": 1.5762446759871982,
+      "grad_norm": 3.331542614086288,
+      "learning_rate": 2.2617427246045976e-06,
+      "loss": 0.8974,
+      "step": 8188
+    },
+    {
+      "epoch": 1.5764371826647738,
+      "grad_norm": 3.1992042923231767,
+      "learning_rate": 2.259768229505268e-06,
+      "loss": 0.9179,
+      "step": 8189
+    },
+    {
+      "epoch": 1.576629689342349,
+      "grad_norm": 3.284790450948841,
+      "learning_rate": 2.2577944868554947e-06,
+      "loss": 0.9203,
+      "step": 8190
+    },
+    {
+      "epoch": 1.5768221960199245,
+      "grad_norm": 3.291675186957191,
+      "learning_rate": 2.255821496847147e-06,
+      "loss": 0.8864,
+      "step": 8191
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "grad_norm": 3.426732793788415,
+      "learning_rate": 2.2538492596720275e-06,
+      "loss": 0.9391,
+      "step": 8192
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.6776,
+      "step": 8192,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.6365,
+      "step": 8192,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.5117,
+      "step": 8192,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.4408,
+      "step": 8192,
+      "vm_loss": 0.1412
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.6576,
+      "step": 8192,
+      "vm_loss": 0.0783
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.6194,
+      "step": 8192,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.4873,
+      "step": 8192,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 1.5770147026974999,
+      "lm_loss": 0.8917,
+      "step": 8192,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 1.577207209375075,
+      "grad_norm": 3.187442037946832,
+      "learning_rate": 2.251877775521861e-06,
+      "loss": 0.8918,
+      "step": 8193
+    },
+    {
+      "epoch": 1.5773997160526507,
+      "grad_norm": 3.3938651593298697,
+      "learning_rate": 2.2499070445883054e-06,
+      "loss": 0.9387,
+      "step": 8194
+    },
+    {
+      "epoch": 1.577592222730226,
+      "grad_norm": 3.1989251367932465,
+      "learning_rate": 2.247937067062932e-06,
+      "loss": 0.9224,
+      "step": 8195
+    },
+    {
+      "epoch": 1.5777847294078013,
+      "grad_norm": 3.1228217595534287,
+      "learning_rate": 2.245967843137259e-06,
+      "loss": 0.9039,
+      "step": 8196
+    },
+    {
+      "epoch": 1.5779772360853768,
+      "grad_norm": 3.220670253396689,
+      "learning_rate": 2.243999373002712e-06,
+      "loss": 0.8637,
+      "step": 8197
+    },
+    {
+      "epoch": 1.578169742762952,
+      "grad_norm": 3.3071434802258173,
+      "learning_rate": 2.242031656850654e-06,
+      "loss": 0.9318,
+      "step": 8198
+    },
+    {
+      "epoch": 1.5783622494405276,
+      "grad_norm": 3.192511413469724,
+      "learning_rate": 2.2400646948723724e-06,
+      "loss": 0.8818,
+      "step": 8199
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "grad_norm": 3.2523337471462885,
+      "learning_rate": 2.238098487259084e-06,
+      "loss": 0.9011,
+      "step": 8200
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.6058,
+      "step": 8200,
+      "vm_loss": 0.2292
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.7306,
+      "step": 8200,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.9154,
+      "step": 8200,
+      "vm_loss": 0.0988
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.5544,
+      "step": 8200,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.5119,
+      "step": 8200,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.9309,
+      "step": 8200,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.9151,
+      "step": 8200,
+      "vm_loss": 0.2426
+    },
+    {
+      "epoch": 1.5785547561181028,
+      "lm_loss": 0.8323,
+      "step": 8200,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.5787472627956782,
+      "grad_norm": 3.3301845399628953,
+      "learning_rate": 2.236133034201924e-06,
+      "loss": 0.8786,
+      "step": 8201
+    },
+    {
+      "epoch": 1.5789397694732537,
+      "grad_norm": 3.3761877741493973,
+      "learning_rate": 2.234168335891961e-06,
+      "loss": 0.9348,
+      "step": 8202
+    },
+    {
+      "epoch": 1.5791322761508289,
+      "grad_norm": 3.1487978734408872,
+      "learning_rate": 2.23220439252019e-06,
+      "loss": 0.8922,
+      "step": 8203
+    },
+    {
+      "epoch": 1.5793247828284045,
+      "grad_norm": 3.17940209113727,
+      "learning_rate": 2.2302412042775344e-06,
+      "loss": 0.9023,
+      "step": 8204
+    },
+    {
+      "epoch": 1.5795172895059797,
+      "grad_norm": 3.1770255131272873,
+      "learning_rate": 2.2282787713548315e-06,
+      "loss": 0.8736,
+      "step": 8205
+    },
+    {
+      "epoch": 1.5797097961835551,
+      "grad_norm": 3.2710488861028924,
+      "learning_rate": 2.226317093942868e-06,
+      "loss": 0.9217,
+      "step": 8206
+    },
+    {
+      "epoch": 1.5799023028611305,
+      "grad_norm": 3.2204827277678936,
+      "learning_rate": 2.224356172232335e-06,
+      "loss": 0.883,
+      "step": 8207
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "grad_norm": 3.2602247354373763,
+      "learning_rate": 2.2223960064138605e-06,
+      "loss": 0.9261,
+      "step": 8208
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.8811,
+      "step": 8208,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.7241,
+      "step": 8208,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 1.0384,
+      "step": 8208,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 1.0267,
+      "step": 8208,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.7457,
+      "step": 8208,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.4828,
+      "step": 8208,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.9754,
+      "step": 8208,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.5800948095387057,
+      "lm_loss": 0.6169,
+      "step": 8208,
+      "vm_loss": 0.1108
+    },
+    {
+      "epoch": 1.5802873162162814,
+      "grad_norm": 3.3094468273377937,
+      "learning_rate": 2.2204365966779983e-06,
+      "loss": 0.9531,
+      "step": 8209
+    },
+    {
+      "epoch": 1.5804798228938566,
+      "grad_norm": 3.1726877924759647,
+      "learning_rate": 2.218477943215229e-06,
+      "loss": 0.8924,
+      "step": 8210
+    },
+    {
+      "epoch": 1.580672329571432,
+      "grad_norm": 3.6417494085076387,
+      "learning_rate": 2.216520046215962e-06,
+      "loss": 1.0206,
+      "step": 8211
+    },
+    {
+      "epoch": 1.5808648362490074,
+      "grad_norm": 3.4330684466800463,
+      "learning_rate": 2.214562905870521e-06,
+      "loss": 0.9663,
+      "step": 8212
+    },
+    {
+      "epoch": 1.5810573429265826,
+      "grad_norm": 3.3981598688903825,
+      "learning_rate": 2.212606522369175e-06,
+      "loss": 0.9794,
+      "step": 8213
+    },
+    {
+      "epoch": 1.5812498496041583,
+      "grad_norm": 3.2584667017760722,
+      "learning_rate": 2.210650895902102e-06,
+      "loss": 0.8862,
+      "step": 8214
+    },
+    {
+      "epoch": 1.5814423562817335,
+      "grad_norm": 3.242158263707811,
+      "learning_rate": 2.208696026659417e-06,
+      "loss": 0.9195,
+      "step": 8215
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "grad_norm": 3.156867509445372,
+      "learning_rate": 2.2067419148311585e-06,
+      "loss": 0.9064,
+      "step": 8216
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.7287,
+      "step": 8216,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.7278,
+      "step": 8216,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.6999,
+      "step": 8216,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.7356,
+      "step": 8216,
+      "vm_loss": 0.1028
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.8408,
+      "step": 8216,
+      "vm_loss": 0.2673
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.7208,
+      "step": 8216,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.6515,
+      "step": 8216,
+      "vm_loss": 0.2032
+    },
+    {
+      "epoch": 1.5816348629593089,
+      "lm_loss": 0.4321,
+      "step": 8216,
+      "vm_loss": 0.1334
+    },
+    {
+      "epoch": 1.5818273696368843,
+      "grad_norm": 3.1852110175199226,
+      "learning_rate": 2.2047885606072926e-06,
+      "loss": 0.9176,
+      "step": 8217
+    },
+    {
+      "epoch": 1.5820198763144595,
+      "grad_norm": 3.161167760440285,
+      "learning_rate": 2.2028359641777066e-06,
+      "loss": 0.8692,
+      "step": 8218
+    },
+    {
+      "epoch": 1.5822123829920351,
+      "grad_norm": 3.2571311621277252,
+      "learning_rate": 2.20088412573222e-06,
+      "loss": 0.9363,
+      "step": 8219
+    },
+    {
+      "epoch": 1.5824048896696103,
+      "grad_norm": 3.1570445081681355,
+      "learning_rate": 2.198933045460575e-06,
+      "loss": 0.8953,
+      "step": 8220
+    },
+    {
+      "epoch": 1.5825973963471858,
+      "grad_norm": 3.1799866214407064,
+      "learning_rate": 2.1969827235524467e-06,
+      "loss": 0.8934,
+      "step": 8221
+    },
+    {
+      "epoch": 1.5827899030247612,
+      "grad_norm": 3.3450851847850807,
+      "learning_rate": 2.1950331601974207e-06,
+      "loss": 0.9384,
+      "step": 8222
+    },
+    {
+      "epoch": 1.5829824097023364,
+      "grad_norm": 3.2608110680430316,
+      "learning_rate": 2.1930843555850324e-06,
+      "loss": 0.9044,
+      "step": 8223
+    },
+    {
+      "epoch": 1.583174916379912,
+      "grad_norm": 3.22262931269552,
+      "learning_rate": 2.1911363099047222e-06,
+      "loss": 0.9165,
+      "step": 8224
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.836,
+      "step": 8224,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.881,
+      "step": 8224,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.6971,
+      "step": 8224,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 1.0072,
+      "step": 8224,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.5985,
+      "step": 8224,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.9451,
+      "step": 8224,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.9946,
+      "step": 8224,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.583174916379912,
+      "lm_loss": 0.6879,
+      "step": 8224,
+      "vm_loss": 0.1201
+    },
+    {
+      "epoch": 1.5833674230574872,
+      "grad_norm": 3.463676796008061,
+      "learning_rate": 2.189189023345866e-06,
+      "loss": 0.9423,
+      "step": 8225
+    },
+    {
+      "epoch": 1.5835599297350627,
+      "grad_norm": 3.352628672710239,
+      "learning_rate": 2.1872424960977667e-06,
+      "loss": 0.9134,
+      "step": 8226
+    },
+    {
+      "epoch": 1.583752436412638,
+      "grad_norm": 3.5250084325447077,
+      "learning_rate": 2.185296728349654e-06,
+      "loss": 0.9722,
+      "step": 8227
+    },
+    {
+      "epoch": 1.5839449430902133,
+      "grad_norm": 3.278741367798804,
+      "learning_rate": 2.183351720290676e-06,
+      "loss": 0.9162,
+      "step": 8228
+    },
+    {
+      "epoch": 1.584137449767789,
+      "grad_norm": 3.2639077693837204,
+      "learning_rate": 2.1814074721099144e-06,
+      "loss": 0.9219,
+      "step": 8229
+    },
+    {
+      "epoch": 1.5843299564453641,
+      "grad_norm": 3.17069440568728,
+      "learning_rate": 2.1794639839963762e-06,
+      "loss": 0.9212,
+      "step": 8230
+    },
+    {
+      "epoch": 1.5845224631229395,
+      "grad_norm": 3.24572136550553,
+      "learning_rate": 2.1775212561389925e-06,
+      "loss": 0.9023,
+      "step": 8231
+    },
+    {
+      "epoch": 1.584714969800515,
+      "grad_norm": 3.173234626405369,
+      "learning_rate": 2.1755792887266236e-06,
+      "loss": 0.8955,
+      "step": 8232
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 1.0978,
+      "step": 8232,
+      "vm_loss": 0.0847
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.671,
+      "step": 8232,
+      "vm_loss": 0.2121
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.5522,
+      "step": 8232,
+      "vm_loss": 0.1703
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.7667,
+      "step": 8232,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.9111,
+      "step": 8232,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.7633,
+      "step": 8232,
+      "vm_loss": 0.1308
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.4721,
+      "step": 8232,
+      "vm_loss": 0.1307
+    },
+    {
+      "epoch": 1.584714969800515,
+      "lm_loss": 0.7309,
+      "step": 8232,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 1.5849074764780904,
+      "grad_norm": 3.2109544580273286,
+      "learning_rate": 2.173638081948046e-06,
+      "loss": 0.9018,
+      "step": 8233
+    },
+    {
+      "epoch": 1.5850999831556658,
+      "grad_norm": 3.3142580995582795,
+      "learning_rate": 2.1716976359919818e-06,
+      "loss": 0.9442,
+      "step": 8234
+    },
+    {
+      "epoch": 1.585292489833241,
+      "grad_norm": 3.2627328726224367,
+      "learning_rate": 2.169757951047057e-06,
+      "loss": 0.9443,
+      "step": 8235
+    },
+    {
+      "epoch": 1.5854849965108164,
+      "grad_norm": 3.2255978387830884,
+      "learning_rate": 2.1678190273018376e-06,
+      "loss": 0.9295,
+      "step": 8236
+    },
+    {
+      "epoch": 1.5856775031883918,
+      "grad_norm": 3.278314759275568,
+      "learning_rate": 2.1658808649448105e-06,
+      "loss": 0.9043,
+      "step": 8237
+    },
+    {
+      "epoch": 1.5858700098659673,
+      "grad_norm": 3.331337344067321,
+      "learning_rate": 2.1639434641643954e-06,
+      "loss": 0.9193,
+      "step": 8238
+    },
+    {
+      "epoch": 1.5860625165435427,
+      "grad_norm": 3.371963193118649,
+      "learning_rate": 2.1620068251489214e-06,
+      "loss": 0.9522,
+      "step": 8239
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "grad_norm": 3.229596579387952,
+      "learning_rate": 2.1600709480866667e-06,
+      "loss": 0.8901,
+      "step": 8240
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 1.1651,
+      "step": 8240,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.8819,
+      "step": 8240,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.5262,
+      "step": 8240,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.7565,
+      "step": 8240,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.5928,
+      "step": 8240,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.5042,
+      "step": 8240,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.6945,
+      "step": 8240,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.5862550232211179,
+      "lm_loss": 0.626,
+      "step": 8240,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.5864475298986933,
+      "grad_norm": 3.1789923911149165,
+      "learning_rate": 2.158135833165815e-06,
+      "loss": 0.874,
+      "step": 8241
+    },
+    {
+      "epoch": 1.5866400365762687,
+      "grad_norm": 3.2201266594907314,
+      "learning_rate": 2.1562014805744913e-06,
+      "loss": 0.9344,
+      "step": 8242
+    },
+    {
+      "epoch": 1.5868325432538442,
+      "grad_norm": 3.268882139437811,
+      "learning_rate": 2.1542678905007287e-06,
+      "loss": 0.8923,
+      "step": 8243
+    },
+    {
+      "epoch": 1.5870250499314196,
+      "grad_norm": 3.1699739429129195,
+      "learning_rate": 2.152335063132508e-06,
+      "loss": 0.9255,
+      "step": 8244
+    },
+    {
+      "epoch": 1.5872175566089948,
+      "grad_norm": 3.1690268395937538,
+      "learning_rate": 2.150402998657718e-06,
+      "loss": 0.915,
+      "step": 8245
+    },
+    {
+      "epoch": 1.5874100632865704,
+      "grad_norm": 3.221311426877247,
+      "learning_rate": 2.1484716972641828e-06,
+      "loss": 0.9245,
+      "step": 8246
+    },
+    {
+      "epoch": 1.5876025699641456,
+      "grad_norm": 3.2618171724113623,
+      "learning_rate": 2.146541159139649e-06,
+      "loss": 0.9447,
+      "step": 8247
+    },
+    {
+      "epoch": 1.587795076641721,
+      "grad_norm": 3.2384727153698902,
+      "learning_rate": 2.144611384471792e-06,
+      "loss": 0.8857,
+      "step": 8248
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.3583,
+      "step": 8248,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 1.0985,
+      "step": 8248,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.7634,
+      "step": 8248,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.7449,
+      "step": 8248,
+      "vm_loss": 0.1615
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.647,
+      "step": 8248,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.852,
+      "step": 8248,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.7354,
+      "step": 8248,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.587795076641721,
+      "lm_loss": 0.5444,
+      "step": 8248,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.5879875833192965,
+      "grad_norm": 3.2719808369185825,
+      "learning_rate": 2.142682373448205e-06,
+      "loss": 0.9129,
+      "step": 8249
+    },
+    {
+      "epoch": 1.5881800899968717,
+      "grad_norm": 3.288993460020872,
+      "learning_rate": 2.1407541262564157e-06,
+      "loss": 0.8914,
+      "step": 8250
+    },
+    {
+      "epoch": 1.5883725966744473,
+      "grad_norm": 3.2399579563294845,
+      "learning_rate": 2.1388266430838757e-06,
+      "loss": 0.9467,
+      "step": 8251
+    },
+    {
+      "epoch": 1.5885651033520225,
+      "grad_norm": 3.3038289880947387,
+      "learning_rate": 2.1368999241179578e-06,
+      "loss": 0.8996,
+      "step": 8252
+    },
+    {
+      "epoch": 1.588757610029598,
+      "grad_norm": 3.153658976784306,
+      "learning_rate": 2.134973969545967e-06,
+      "loss": 0.8724,
+      "step": 8253
+    },
+    {
+      "epoch": 1.5889501167071733,
+      "grad_norm": 3.4391392516375694,
+      "learning_rate": 2.133048779555129e-06,
+      "loss": 0.9789,
+      "step": 8254
+    },
+    {
+      "epoch": 1.5891426233847485,
+      "grad_norm": 3.263121193278418,
+      "learning_rate": 2.131124354332601e-06,
+      "loss": 0.8899,
+      "step": 8255
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "grad_norm": 3.3863184790056238,
+      "learning_rate": 2.129200694065452e-06,
+      "loss": 0.9311,
+      "step": 8256
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 1.1368,
+      "step": 8256,
+      "vm_loss": 0.2064
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.8404,
+      "step": 8256,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.647,
+      "step": 8256,
+      "vm_loss": 0.126
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.9692,
+      "step": 8256,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.6332,
+      "step": 8256,
+      "vm_loss": 0.1298
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.9095,
+      "step": 8256,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.6873,
+      "step": 8256,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.5893351300623242,
+      "lm_loss": 0.7877,
+      "step": 8256,
+      "vm_loss": 0.1109
+    },
+    {
+      "epoch": 1.5895276367398994,
+      "grad_norm": 3.3680130258038123,
+      "learning_rate": 2.127277798940698e-06,
+      "loss": 0.9021,
+      "step": 8257
+    },
+    {
+      "epoch": 1.5897201434174748,
+      "grad_norm": 3.51722159129693,
+      "learning_rate": 2.1253556691452603e-06,
+      "loss": 0.9496,
+      "step": 8258
+    },
+    {
+      "epoch": 1.5899126500950502,
+      "grad_norm": 3.3200275621014845,
+      "learning_rate": 2.123434304866e-06,
+      "loss": 0.9232,
+      "step": 8259
+    },
+    {
+      "epoch": 1.5901051567726254,
+      "grad_norm": 3.5031689445659238,
+      "learning_rate": 2.121513706289692e-06,
+      "loss": 0.956,
+      "step": 8260
+    },
+    {
+      "epoch": 1.590297663450201,
+      "grad_norm": 3.0817432319221294,
+      "learning_rate": 2.1195938736030507e-06,
+      "loss": 0.8935,
+      "step": 8261
+    },
+    {
+      "epoch": 1.5904901701277763,
+      "grad_norm": 3.2296742571778934,
+      "learning_rate": 2.1176748069927035e-06,
+      "loss": 0.911,
+      "step": 8262
+    },
+    {
+      "epoch": 1.5906826768053517,
+      "grad_norm": 3.2895183400037054,
+      "learning_rate": 2.115756506645209e-06,
+      "loss": 0.9344,
+      "step": 8263
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "grad_norm": 3.326861806896037,
+      "learning_rate": 2.1138389727470508e-06,
+      "loss": 0.9498,
+      "step": 8264
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.984,
+      "step": 8264,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.5109,
+      "step": 8264,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.5482,
+      "step": 8264,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.5695,
+      "step": 8264,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.5779,
+      "step": 8264,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.7628,
+      "step": 8264,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.6919,
+      "step": 8264,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.5908751834829271,
+      "lm_loss": 0.3581,
+      "step": 8264,
+      "vm_loss": 0.1932
+    },
+    {
+      "epoch": 1.5910676901605023,
+      "grad_norm": 3.0889364039039036,
+      "learning_rate": 2.111922205484641e-06,
+      "loss": 0.8708,
+      "step": 8265
+    },
+    {
+      "epoch": 1.591260196838078,
+      "grad_norm": 3.1380712248967235,
+      "learning_rate": 2.110006205044307e-06,
+      "loss": 0.866,
+      "step": 8266
+    },
+    {
+      "epoch": 1.5914527035156532,
+      "grad_norm": 3.374327927916381,
+      "learning_rate": 2.108090971612313e-06,
+      "loss": 0.9162,
+      "step": 8267
+    },
+    {
+      "epoch": 1.5916452101932286,
+      "grad_norm": 3.2724945876137017,
+      "learning_rate": 2.106176505374844e-06,
+      "loss": 0.9164,
+      "step": 8268
+    },
+    {
+      "epoch": 1.591837716870804,
+      "grad_norm": 3.1877542397708005,
+      "learning_rate": 2.1042628065180105e-06,
+      "loss": 0.9258,
+      "step": 8269
+    },
+    {
+      "epoch": 1.5920302235483792,
+      "grad_norm": 3.4767676783801815,
+      "learning_rate": 2.102349875227847e-06,
+      "loss": 0.9931,
+      "step": 8270
+    },
+    {
+      "epoch": 1.5922227302259548,
+      "grad_norm": 3.4162759032961847,
+      "learning_rate": 2.1004377116903195e-06,
+      "loss": 0.9238,
+      "step": 8271
+    },
+    {
+      "epoch": 1.59241523690353,
+      "grad_norm": 3.174705485908778,
+      "learning_rate": 2.098526316091308e-06,
+      "loss": 0.8909,
+      "step": 8272
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 1.0846,
+      "step": 8272,
+      "vm_loss": 0.2141
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.6833,
+      "step": 8272,
+      "vm_loss": 0.1333
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.5514,
+      "step": 8272,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.6213,
+      "step": 8272,
+      "vm_loss": 0.1145
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.6342,
+      "step": 8272,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.7591,
+      "step": 8272,
+      "vm_loss": 0.105
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.553,
+      "step": 8272,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.59241523690353,
+      "lm_loss": 0.5246,
+      "step": 8272,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.5926077435811055,
+      "grad_norm": 3.209103944378494,
+      "learning_rate": 2.096615688616629e-06,
+      "loss": 0.851,
+      "step": 8273
+    },
+    {
+      "epoch": 1.5928002502586809,
+      "grad_norm": 3.1659651472179733,
+      "learning_rate": 2.094705829452018e-06,
+      "loss": 0.8988,
+      "step": 8274
+    },
+    {
+      "epoch": 1.592992756936256,
+      "grad_norm": 3.332303504014603,
+      "learning_rate": 2.0927967387831395e-06,
+      "loss": 0.9351,
+      "step": 8275
+    },
+    {
+      "epoch": 1.5931852636138317,
+      "grad_norm": 3.3133121839309543,
+      "learning_rate": 2.090888416795582e-06,
+      "loss": 0.9528,
+      "step": 8276
+    },
+    {
+      "epoch": 1.593377770291407,
+      "grad_norm": 3.1087676897121703,
+      "learning_rate": 2.088980863674853e-06,
+      "loss": 0.8593,
+      "step": 8277
+    },
+    {
+      "epoch": 1.5935702769689823,
+      "grad_norm": 3.2490769004783777,
+      "learning_rate": 2.0870740796064015e-06,
+      "loss": 0.8851,
+      "step": 8278
+    },
+    {
+      "epoch": 1.5937627836465578,
+      "grad_norm": 3.2246541185243704,
+      "learning_rate": 2.0851680647755812e-06,
+      "loss": 0.8852,
+      "step": 8279
+    },
+    {
+      "epoch": 1.593955290324133,
+      "grad_norm": 3.2637560147604745,
+      "learning_rate": 2.0832628193676864e-06,
+      "loss": 0.9056,
+      "step": 8280
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.7534,
+      "step": 8280,
+      "vm_loss": 0.2144
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.6807,
+      "step": 8280,
+      "vm_loss": 0.2277
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 1.098,
+      "step": 8280,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.726,
+      "step": 8280,
+      "vm_loss": 0.1523
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.796,
+      "step": 8280,
+      "vm_loss": 0.1941
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.283,
+      "step": 8280,
+      "vm_loss": 0.2627
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.6848,
+      "step": 8280,
+      "vm_loss": 0.1281
+    },
+    {
+      "epoch": 1.593955290324133,
+      "lm_loss": 0.6416,
+      "step": 8280,
+      "vm_loss": 0.1572
+    },
+    {
+      "epoch": 1.5941477970017086,
+      "grad_norm": 3.307097229603673,
+      "learning_rate": 2.0813583435679295e-06,
+      "loss": 0.9457,
+      "step": 8281
+    },
+    {
+      "epoch": 1.5943403036792838,
+      "grad_norm": 3.2138158606667204,
+      "learning_rate": 2.0794546375614543e-06,
+      "loss": 0.8821,
+      "step": 8282
+    },
+    {
+      "epoch": 1.5945328103568592,
+      "grad_norm": 3.353456294736932,
+      "learning_rate": 2.077551701533318e-06,
+      "loss": 0.9028,
+      "step": 8283
+    },
+    {
+      "epoch": 1.5947253170344347,
+      "grad_norm": 3.1925695940369243,
+      "learning_rate": 2.0756495356685147e-06,
+      "loss": 0.8619,
+      "step": 8284
+    },
+    {
+      "epoch": 1.5949178237120099,
+      "grad_norm": 3.2155619911772026,
+      "learning_rate": 2.0737481401519576e-06,
+      "loss": 0.9192,
+      "step": 8285
+    },
+    {
+      "epoch": 1.5951103303895855,
+      "grad_norm": 3.3890080264834808,
+      "learning_rate": 2.0718475151684913e-06,
+      "loss": 0.9258,
+      "step": 8286
+    },
+    {
+      "epoch": 1.5953028370671607,
+      "grad_norm": 3.157819241474124,
+      "learning_rate": 2.0699476609028714e-06,
+      "loss": 0.8851,
+      "step": 8287
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "grad_norm": 3.1975040811645044,
+      "learning_rate": 2.0680485775397983e-06,
+      "loss": 0.8677,
+      "step": 8288
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.7623,
+      "step": 8288,
+      "vm_loss": 0.1216
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.6548,
+      "step": 8288,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.8758,
+      "step": 8288,
+      "vm_loss": 0.1345
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.5852,
+      "step": 8288,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.9623,
+      "step": 8288,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.6871,
+      "step": 8288,
+      "vm_loss": 0.1164
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.7335,
+      "step": 8288,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.5954953437447361,
+      "lm_loss": 0.3618,
+      "step": 8288,
+      "vm_loss": 0.2121
+    },
+    {
+      "epoch": 1.5956878504223115,
+      "grad_norm": 3.178929153855023,
+      "learning_rate": 2.0661502652638822e-06,
+      "loss": 0.8982,
+      "step": 8289
+    },
+    {
+      "epoch": 1.5958803570998867,
+      "grad_norm": 3.211535634731193,
+      "learning_rate": 2.064252724259662e-06,
+      "loss": 0.941,
+      "step": 8290
+    },
+    {
+      "epoch": 1.5960728637774624,
+      "grad_norm": 3.154287789712948,
+      "learning_rate": 2.0623559547116055e-06,
+      "loss": 0.9034,
+      "step": 8291
+    },
+    {
+      "epoch": 1.5962653704550376,
+      "grad_norm": 3.1613609530987508,
+      "learning_rate": 2.060459956804103e-06,
+      "loss": 0.867,
+      "step": 8292
+    },
+    {
+      "epoch": 1.596457877132613,
+      "grad_norm": 3.2989629970352983,
+      "learning_rate": 2.0585647307214707e-06,
+      "loss": 0.9263,
+      "step": 8293
+    },
+    {
+      "epoch": 1.5966503838101884,
+      "grad_norm": 3.4278171879053065,
+      "learning_rate": 2.0566702766479454e-06,
+      "loss": 0.9745,
+      "step": 8294
+    },
+    {
+      "epoch": 1.5968428904877638,
+      "grad_norm": 3.0830439356882544,
+      "learning_rate": 2.0547765947676933e-06,
+      "loss": 0.8762,
+      "step": 8295
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "grad_norm": 3.2516764486934235,
+      "learning_rate": 2.052883685264806e-06,
+      "loss": 0.8958,
+      "step": 8296
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.9544,
+      "step": 8296,
+      "vm_loss": 0.1839
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.6604,
+      "step": 8296,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.5058,
+      "step": 8296,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.4843,
+      "step": 8296,
+      "vm_loss": 0.2179
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.5797,
+      "step": 8296,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 1.0527,
+      "step": 8296,
+      "vm_loss": 0.2119
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.4596,
+      "step": 8296,
+      "vm_loss": 0.1195
+    },
+    {
+      "epoch": 1.5970353971653393,
+      "lm_loss": 0.7037,
+      "step": 8296,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.5972279038429145,
+      "grad_norm": 3.4543530605829105,
+      "learning_rate": 2.050991548323298e-06,
+      "loss": 0.9574,
+      "step": 8297
+    },
+    {
+      "epoch": 1.5974204105204899,
+      "grad_norm": 3.3095564567194815,
+      "learning_rate": 2.0491001841271073e-06,
+      "loss": 0.9158,
+      "step": 8298
+    },
+    {
+      "epoch": 1.5976129171980653,
+      "grad_norm": 3.2974638837382204,
+      "learning_rate": 2.0472095928601045e-06,
+      "loss": 0.9056,
+      "step": 8299
+    },
+    {
+      "epoch": 1.5978054238756407,
+      "grad_norm": 3.2165264009311136,
+      "learning_rate": 2.0453197747060728e-06,
+      "loss": 0.8627,
+      "step": 8300
+    },
+    {
+      "epoch": 1.5979979305532162,
+      "grad_norm": 3.3553934836972297,
+      "learning_rate": 2.043430729848729e-06,
+      "loss": 0.9395,
+      "step": 8301
+    },
+    {
+      "epoch": 1.5981904372307913,
+      "grad_norm": 3.3075297979522795,
+      "learning_rate": 2.0415424584717113e-06,
+      "loss": 0.9105,
+      "step": 8302
+    },
+    {
+      "epoch": 1.5983829439083668,
+      "grad_norm": 3.270031232767467,
+      "learning_rate": 2.039654960758588e-06,
+      "loss": 0.9408,
+      "step": 8303
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "grad_norm": 3.2969076462985982,
+      "learning_rate": 2.0377682368928407e-06,
+      "loss": 0.8967,
+      "step": 8304
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.6728,
+      "step": 8304,
+      "vm_loss": 0.1118
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.9064,
+      "step": 8304,
+      "vm_loss": 0.1144
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.7602,
+      "step": 8304,
+      "vm_loss": 0.2321
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.8914,
+      "step": 8304,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.6725,
+      "step": 8304,
+      "vm_loss": 0.1353
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.5932,
+      "step": 8304,
+      "vm_loss": 0.1575
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 0.8768,
+      "step": 8304,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.5985754505859422,
+      "lm_loss": 1.052,
+      "step": 8304,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.5987679572635176,
+      "grad_norm": 3.184496176439511,
+      "learning_rate": 2.0358822870578922e-06,
+      "loss": 0.8853,
+      "step": 8305
+    },
+    {
+      "epoch": 1.598960463941093,
+      "grad_norm": 3.2710727618572597,
+      "learning_rate": 2.033997111437074e-06,
+      "loss": 0.9099,
+      "step": 8306
+    },
+    {
+      "epoch": 1.5991529706186682,
+      "grad_norm": 3.2971407534972306,
+      "learning_rate": 2.0321127102136528e-06,
+      "loss": 0.9123,
+      "step": 8307
+    },
+    {
+      "epoch": 1.5993454772962439,
+      "grad_norm": 3.277469259024938,
+      "learning_rate": 2.0302290835708115e-06,
+      "loss": 0.92,
+      "step": 8308
+    },
+    {
+      "epoch": 1.599537983973819,
+      "grad_norm": 3.276318087830748,
+      "learning_rate": 2.0283462316916712e-06,
+      "loss": 0.9024,
+      "step": 8309
+    },
+    {
+      "epoch": 1.5997304906513945,
+      "grad_norm": 3.3002482656381487,
+      "learning_rate": 2.026464154759261e-06,
+      "loss": 0.9017,
+      "step": 8310
+    },
+    {
+      "epoch": 1.59992299732897,
+      "grad_norm": 3.0843286695321335,
+      "learning_rate": 2.024582852956547e-06,
+      "loss": 0.8882,
+      "step": 8311
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "grad_norm": 3.2784101583412006,
+      "learning_rate": 2.022702326466416e-06,
+      "loss": 0.9017,
+      "step": 8312
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 1.2306,
+      "step": 8312,
+      "vm_loss": 0.1286
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.6293,
+      "step": 8312,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.7233,
+      "step": 8312,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.5911,
+      "step": 8312,
+      "vm_loss": 0.2123
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.7699,
+      "step": 8312,
+      "vm_loss": 0.1922
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.9302,
+      "step": 8312,
+      "vm_loss": 0.1195
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.4482,
+      "step": 8312,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 1.6001155040065451,
+      "lm_loss": 0.662,
+      "step": 8312,
+      "vm_loss": 0.1162
+    },
+    {
+      "epoch": 1.6003080106841208,
+      "grad_norm": 3.3629956586206133,
+      "learning_rate": 2.020822575471677e-06,
+      "loss": 0.951,
+      "step": 8313
+    },
+    {
+      "epoch": 1.600500517361696,
+      "grad_norm": 3.2308917588566586,
+      "learning_rate": 2.0189436001550668e-06,
+      "loss": 0.9026,
+      "step": 8314
+    },
+    {
+      "epoch": 1.6006930240392714,
+      "grad_norm": 3.1706581743766025,
+      "learning_rate": 2.0170654006992473e-06,
+      "loss": 0.926,
+      "step": 8315
+    },
+    {
+      "epoch": 1.6008855307168468,
+      "grad_norm": 3.223117217203848,
+      "learning_rate": 2.015187977286804e-06,
+      "loss": 0.9471,
+      "step": 8316
+    },
+    {
+      "epoch": 1.601078037394422,
+      "grad_norm": 3.150625318917613,
+      "learning_rate": 2.0133113301002426e-06,
+      "loss": 0.8854,
+      "step": 8317
+    },
+    {
+      "epoch": 1.6012705440719976,
+      "grad_norm": 3.1967382143421372,
+      "learning_rate": 2.0114354593220007e-06,
+      "loss": 0.8676,
+      "step": 8318
+    },
+    {
+      "epoch": 1.6014630507495728,
+      "grad_norm": 3.2693629092763588,
+      "learning_rate": 2.009560365134434e-06,
+      "loss": 0.9057,
+      "step": 8319
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "grad_norm": 3.1873409554459875,
+      "learning_rate": 2.007686047719831e-06,
+      "loss": 0.8742,
+      "step": 8320
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.347,
+      "step": 8320,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.7822,
+      "step": 8320,
+      "vm_loss": 0.1943
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.8968,
+      "step": 8320,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.4761,
+      "step": 8320,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.8188,
+      "step": 8320,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.586,
+      "step": 8320,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.7331,
+      "step": 8320,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.6016555574271483,
+      "lm_loss": 0.7576,
+      "step": 8320,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.6018480641047237,
+      "grad_norm": 3.228129237529959,
+      "learning_rate": 2.0058125072603907e-06,
+      "loss": 0.8586,
+      "step": 8321
+    },
+    {
+      "epoch": 1.602040570782299,
+      "grad_norm": 3.137441253552322,
+      "learning_rate": 2.0039397439382544e-06,
+      "loss": 0.868,
+      "step": 8322
+    },
+    {
+      "epoch": 1.6022330774598745,
+      "grad_norm": 3.3876991628001774,
+      "learning_rate": 2.002067757935473e-06,
+      "loss": 0.9317,
+      "step": 8323
+    },
+    {
+      "epoch": 1.6024255841374497,
+      "grad_norm": 3.386745238087488,
+      "learning_rate": 2.0001965494340303e-06,
+      "loss": 0.9348,
+      "step": 8324
+    },
+    {
+      "epoch": 1.6026180908150252,
+      "grad_norm": 3.175238453048385,
+      "learning_rate": 1.9983261186158265e-06,
+      "loss": 0.8481,
+      "step": 8325
+    },
+    {
+      "epoch": 1.6028105974926006,
+      "grad_norm": 3.2653911147651695,
+      "learning_rate": 1.9964564656626993e-06,
+      "loss": 0.8898,
+      "step": 8326
+    },
+    {
+      "epoch": 1.6030031041701758,
+      "grad_norm": 3.3461166423031448,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.9328,
+      "step": 8327
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "grad_norm": 3.334832116830459,
+      "learning_rate": 1.9927194940785998e-06,
+      "loss": 0.9333,
+      "step": 8328
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.6449,
+      "step": 8328,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 1.3609,
+      "step": 8328,
+      "vm_loss": 0.2348
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.5589,
+      "step": 8328,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.4405,
+      "step": 8328,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.4061,
+      "step": 8328,
+      "vm_loss": 0.2258
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 1.1004,
+      "step": 8328,
+      "vm_loss": 0.1945
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.6224,
+      "step": 8328,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.6031956108477514,
+      "lm_loss": 0.5236,
+      "step": 8328,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.6033881175253266,
+      "grad_norm": 3.29165390533546,
+      "learning_rate": 1.9908521758109114e-06,
+      "loss": 0.8882,
+      "step": 8329
+    },
+    {
+      "epoch": 1.603580624202902,
+      "grad_norm": 3.228774750231894,
+      "learning_rate": 1.9889856361348603e-06,
+      "loss": 0.9116,
+      "step": 8330
+    },
+    {
+      "epoch": 1.6037731308804775,
+      "grad_norm": 3.2595149286983705,
+      "learning_rate": 1.987119875231891e-06,
+      "loss": 0.8965,
+      "step": 8331
+    },
+    {
+      "epoch": 1.6039656375580527,
+      "grad_norm": 3.2169430903579066,
+      "learning_rate": 1.9852548932833893e-06,
+      "loss": 0.8845,
+      "step": 8332
+    },
+    {
+      "epoch": 1.6041581442356283,
+      "grad_norm": 3.3134092040973497,
+      "learning_rate": 1.983390690470649e-06,
+      "loss": 0.8902,
+      "step": 8333
+    },
+    {
+      "epoch": 1.6043506509132035,
+      "grad_norm": 3.3056309612807455,
+      "learning_rate": 1.981527266974895e-06,
+      "loss": 0.8967,
+      "step": 8334
+    },
+    {
+      "epoch": 1.604543157590779,
+      "grad_norm": 3.3964609959525447,
+      "learning_rate": 1.9796646229772775e-06,
+      "loss": 0.9804,
+      "step": 8335
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "grad_norm": 3.231816830687224,
+      "learning_rate": 1.977802758658869e-06,
+      "loss": 0.8844,
+      "step": 8336
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.4029,
+      "step": 8336,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.599,
+      "step": 8336,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.4715,
+      "step": 8336,
+      "vm_loss": 0.1185
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 1.0251,
+      "step": 8336,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.8673,
+      "step": 8336,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.4617,
+      "step": 8336,
+      "vm_loss": 0.1268
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 0.7612,
+      "step": 8336,
+      "vm_loss": 0.2686
+    },
+    {
+      "epoch": 1.6047356642683543,
+      "lm_loss": 1.037,
+      "step": 8336,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 1.6049281709459295,
+      "grad_norm": 3.1876850179137777,
+      "learning_rate": 1.975941674200668e-06,
+      "loss": 0.8957,
+      "step": 8337
+    },
+    {
+      "epoch": 1.6051206776235052,
+      "grad_norm": 3.2441251256188663,
+      "learning_rate": 1.9740813697835914e-06,
+      "loss": 0.8945,
+      "step": 8338
+    },
+    {
+      "epoch": 1.6053131843010804,
+      "grad_norm": 3.256935183580076,
+      "learning_rate": 1.9722218455884866e-06,
+      "loss": 0.9107,
+      "step": 8339
+    },
+    {
+      "epoch": 1.6055056909786558,
+      "grad_norm": 3.420023418609238,
+      "learning_rate": 1.9703631017961243e-06,
+      "loss": 0.955,
+      "step": 8340
+    },
+    {
+      "epoch": 1.6056981976562312,
+      "grad_norm": 3.12517737846043,
+      "learning_rate": 1.968505138587201e-06,
+      "loss": 0.859,
+      "step": 8341
+    },
+    {
+      "epoch": 1.6058907043338064,
+      "grad_norm": 3.1094554742084193,
+      "learning_rate": 1.9666479561423247e-06,
+      "loss": 0.8861,
+      "step": 8342
+    },
+    {
+      "epoch": 1.606083211011382,
+      "grad_norm": 3.3213613728077442,
+      "learning_rate": 1.964791554642049e-06,
+      "loss": 0.9522,
+      "step": 8343
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "grad_norm": 3.2132899690243306,
+      "learning_rate": 1.962935934266833e-06,
+      "loss": 0.8621,
+      "step": 8344
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.7518,
+      "step": 8344,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.86,
+      "step": 8344,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.7898,
+      "step": 8344,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.6418,
+      "step": 8344,
+      "vm_loss": 0.2243
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.7294,
+      "step": 8344,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.9021,
+      "step": 8344,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.9843,
+      "step": 8344,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.6062757176889573,
+      "lm_loss": 0.5909,
+      "step": 8344,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 1.6064682243665327,
+      "grad_norm": 3.226243287061956,
+      "learning_rate": 1.961081095197067e-06,
+      "loss": 0.9073,
+      "step": 8345
+    },
+    {
+      "epoch": 1.6066607310441081,
+      "grad_norm": 3.2271172538523634,
+      "learning_rate": 1.959227037613067e-06,
+      "loss": 0.8824,
+      "step": 8346
+    },
+    {
+      "epoch": 1.6068532377216833,
+      "grad_norm": 3.239449176094794,
+      "learning_rate": 1.957373761695073e-06,
+      "loss": 0.9052,
+      "step": 8347
+    },
+    {
+      "epoch": 1.607045744399259,
+      "grad_norm": 3.211515576315902,
+      "learning_rate": 1.9555212676232396e-06,
+      "loss": 0.8627,
+      "step": 8348
+    },
+    {
+      "epoch": 1.6072382510768342,
+      "grad_norm": 3.1530427504099037,
+      "learning_rate": 1.953669555577663e-06,
+      "loss": 0.8782,
+      "step": 8349
+    },
+    {
+      "epoch": 1.6074307577544096,
+      "grad_norm": 3.0249201247538653,
+      "learning_rate": 1.9518186257383453e-06,
+      "loss": 0.8474,
+      "step": 8350
+    },
+    {
+      "epoch": 1.607623264431985,
+      "grad_norm": 3.1234078016556253,
+      "learning_rate": 1.949968478285226e-06,
+      "loss": 0.8673,
+      "step": 8351
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "grad_norm": 3.257557159383585,
+      "learning_rate": 1.9481191133981557e-06,
+      "loss": 0.9131,
+      "step": 8352
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.7146,
+      "step": 8352,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.4001,
+      "step": 8352,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.9887,
+      "step": 8352,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.4478,
+      "step": 8352,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.8284,
+      "step": 8352,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.5847,
+      "step": 8352,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 1.2224,
+      "step": 8352,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.6078157711095602,
+      "lm_loss": 0.8029,
+      "step": 8352,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.6080082777871358,
+      "grad_norm": 3.3511706936007744,
+      "learning_rate": 1.9462705312569265e-06,
+      "loss": 0.9353,
+      "step": 8353
+    },
+    {
+      "epoch": 1.608200784464711,
+      "grad_norm": 3.3495207489580316,
+      "learning_rate": 1.9444227320412355e-06,
+      "loss": 0.9749,
+      "step": 8354
+    },
+    {
+      "epoch": 1.6083932911422865,
+      "grad_norm": 3.4993689114699675,
+      "learning_rate": 1.9425757159307168e-06,
+      "loss": 0.9569,
+      "step": 8355
+    },
+    {
+      "epoch": 1.6085857978198619,
+      "grad_norm": 3.283540130190931,
+      "learning_rate": 1.9407294831049226e-06,
+      "loss": 0.9036,
+      "step": 8356
+    },
+    {
+      "epoch": 1.6087783044974373,
+      "grad_norm": 3.3853501016977914,
+      "learning_rate": 1.9388840337433313e-06,
+      "loss": 0.9441,
+      "step": 8357
+    },
+    {
+      "epoch": 1.6089708111750127,
+      "grad_norm": 3.2853629931186488,
+      "learning_rate": 1.937039368025346e-06,
+      "loss": 0.8918,
+      "step": 8358
+    },
+    {
+      "epoch": 1.609163317852588,
+      "grad_norm": 3.1723898464013534,
+      "learning_rate": 1.9351954861302847e-06,
+      "loss": 0.8858,
+      "step": 8359
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "grad_norm": 3.252540386245679,
+      "learning_rate": 1.933352388237406e-06,
+      "loss": 0.888,
+      "step": 8360
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.8955,
+      "step": 8360,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.5522,
+      "step": 8360,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 1.2706,
+      "step": 8360,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.7644,
+      "step": 8360,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.4609,
+      "step": 8360,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.5524,
+      "step": 8360,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.7427,
+      "step": 8360,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 1.6093558245301633,
+      "lm_loss": 0.7653,
+      "step": 8360,
+      "vm_loss": 0.1654
+    },
+    {
+      "epoch": 1.6095483312077388,
+      "grad_norm": 3.187836269541993,
+      "learning_rate": 1.931510074525874e-06,
+      "loss": 0.8615,
+      "step": 8361
+    },
+    {
+      "epoch": 1.6097408378853142,
+      "grad_norm": 3.1978676330634546,
+      "learning_rate": 1.929668545174791e-06,
+      "loss": 0.8902,
+      "step": 8362
+    },
+    {
+      "epoch": 1.6099333445628896,
+      "grad_norm": 3.086626183351094,
+      "learning_rate": 1.9278278003631735e-06,
+      "loss": 0.8808,
+      "step": 8363
+    },
+    {
+      "epoch": 1.6101258512404648,
+      "grad_norm": 3.186101429427331,
+      "learning_rate": 1.9259878402699704e-06,
+      "loss": 0.894,
+      "step": 8364
+    },
+    {
+      "epoch": 1.6103183579180402,
+      "grad_norm": 3.098971139575795,
+      "learning_rate": 1.9241486650740406e-06,
+      "loss": 0.8355,
+      "step": 8365
+    },
+    {
+      "epoch": 1.6105108645956157,
+      "grad_norm": 3.193924022165544,
+      "learning_rate": 1.9223102749541856e-06,
+      "loss": 0.8806,
+      "step": 8366
+    },
+    {
+      "epoch": 1.610703371273191,
+      "grad_norm": 3.232335134679844,
+      "learning_rate": 1.920472670089114e-06,
+      "loss": 0.882,
+      "step": 8367
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "grad_norm": 3.4555498922303887,
+      "learning_rate": 1.9186358506574686e-06,
+      "loss": 0.9542,
+      "step": 8368
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 0.6641,
+      "step": 8368,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 0.9509,
+      "step": 8368,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 1.4275,
+      "step": 8368,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 1.1082,
+      "step": 8368,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 1.3913,
+      "step": 8368,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 1.4924,
+      "step": 8368,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 0.8781,
+      "step": 8368,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.6108958779507665,
+      "lm_loss": 0.4511,
+      "step": 8368,
+      "vm_loss": 0.1526
+    },
+    {
+      "epoch": 1.6110883846283417,
+      "grad_norm": 3.4808297222428286,
+      "learning_rate": 1.9167998168378034e-06,
+      "loss": 0.9614,
+      "step": 8369
+    },
+    {
+      "epoch": 1.6112808913059173,
+      "grad_norm": 3.322109682549969,
+      "learning_rate": 1.914964568808616e-06,
+      "loss": 0.9031,
+      "step": 8370
+    },
+    {
+      "epoch": 1.6114733979834925,
+      "grad_norm": 3.3946513457557725,
+      "learning_rate": 1.9131301067483075e-06,
+      "loss": 0.895,
+      "step": 8371
+    },
+    {
+      "epoch": 1.611665904661068,
+      "grad_norm": 3.244419771898723,
+      "learning_rate": 1.9112964308352145e-06,
+      "loss": 0.8263,
+      "step": 8372
+    },
+    {
+      "epoch": 1.6118584113386434,
+      "grad_norm": 3.273995896696473,
+      "learning_rate": 1.9094635412475927e-06,
+      "loss": 0.9115,
+      "step": 8373
+    },
+    {
+      "epoch": 1.6120509180162186,
+      "grad_norm": 3.128102081238093,
+      "learning_rate": 1.907631438163626e-06,
+      "loss": 0.8408,
+      "step": 8374
+    },
+    {
+      "epoch": 1.6122434246937942,
+      "grad_norm": 3.126744850391365,
+      "learning_rate": 1.9058001217614118e-06,
+      "loss": 0.8546,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "grad_norm": 3.2844812113609563,
+      "learning_rate": 1.9039695922189817e-06,
+      "loss": 0.9252,
+      "step": 8376
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.8637,
+      "step": 8376,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.5745,
+      "step": 8376,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.6015,
+      "step": 8376,
+      "vm_loss": 0.1822
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.5047,
+      "step": 8376,
+      "vm_loss": 0.2012
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.4382,
+      "step": 8376,
+      "vm_loss": 0.2181
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 1.0475,
+      "step": 8376,
+      "vm_loss": 0.2588
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.5956,
+      "step": 8376,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.6124359313713694,
+      "lm_loss": 0.6175,
+      "step": 8376,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.6126284380489448,
+      "grad_norm": 3.0535297509297497,
+      "learning_rate": 1.902139849714285e-06,
+      "loss": 0.8695,
+      "step": 8377
+    },
+    {
+      "epoch": 1.6128209447265203,
+      "grad_norm": 3.282675394589569,
+      "learning_rate": 1.900310894425197e-06,
+      "loss": 0.925,
+      "step": 8378
+    },
+    {
+      "epoch": 1.6130134514040955,
+      "grad_norm": 3.15395348936995,
+      "learning_rate": 1.8984827265295147e-06,
+      "loss": 0.8882,
+      "step": 8379
+    },
+    {
+      "epoch": 1.613205958081671,
+      "grad_norm": 3.2341105730541524,
+      "learning_rate": 1.8966553462049608e-06,
+      "loss": 0.8913,
+      "step": 8380
+    },
+    {
+      "epoch": 1.6133984647592463,
+      "grad_norm": 3.395267805528291,
+      "learning_rate": 1.894828753629182e-06,
+      "loss": 0.9779,
+      "step": 8381
+    },
+    {
+      "epoch": 1.6135909714368217,
+      "grad_norm": 3.2149550462320766,
+      "learning_rate": 1.8930029489797408e-06,
+      "loss": 0.8954,
+      "step": 8382
+    },
+    {
+      "epoch": 1.6137834781143972,
+      "grad_norm": 3.1458913615442303,
+      "learning_rate": 1.8911779324341317e-06,
+      "loss": 0.893,
+      "step": 8383
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "grad_norm": 3.4440901373487276,
+      "learning_rate": 1.8893537041697707e-06,
+      "loss": 0.9452,
+      "step": 8384
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.9268,
+      "step": 8384,
+      "vm_loss": 0.1072
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.726,
+      "step": 8384,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.7835,
+      "step": 8384,
+      "vm_loss": 0.2232
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.6222,
+      "step": 8384,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.3847,
+      "step": 8384,
+      "vm_loss": 0.1694
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.8808,
+      "step": 8384,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.9483,
+      "step": 8384,
+      "vm_loss": 0.2061
+    },
+    {
+      "epoch": 1.6139759847919724,
+      "lm_loss": 0.4876,
+      "step": 8384,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.614168491469548,
+      "grad_norm": 3.364944851463711,
+      "learning_rate": 1.887530264363996e-06,
+      "loss": 0.9012,
+      "step": 8385
+    },
+    {
+      "epoch": 1.6143609981471232,
+      "grad_norm": 3.216208698112367,
+      "learning_rate": 1.8857076131940643e-06,
+      "loss": 0.8319,
+      "step": 8386
+    },
+    {
+      "epoch": 1.6145535048246986,
+      "grad_norm": 3.388066159595722,
+      "learning_rate": 1.88388575083717e-06,
+      "loss": 0.9089,
+      "step": 8387
+    },
+    {
+      "epoch": 1.614746011502274,
+      "grad_norm": 3.297383004575363,
+      "learning_rate": 1.8820646774704131e-06,
+      "loss": 0.8865,
+      "step": 8388
+    },
+    {
+      "epoch": 1.6149385181798492,
+      "grad_norm": 3.3618466590428953,
+      "learning_rate": 1.880244393270828e-06,
+      "loss": 0.9095,
+      "step": 8389
+    },
+    {
+      "epoch": 1.6151310248574249,
+      "grad_norm": 3.349619236553279,
+      "learning_rate": 1.8784248984153708e-06,
+      "loss": 0.903,
+      "step": 8390
+    },
+    {
+      "epoch": 1.615323531535,
+      "grad_norm": 3.282595188101464,
+      "learning_rate": 1.876606193080921e-06,
+      "loss": 0.897,
+      "step": 8391
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "grad_norm": 3.2375819022347887,
+      "learning_rate": 1.8747882774442749e-06,
+      "loss": 0.8822,
+      "step": 8392
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.7315,
+      "step": 8392,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.5204,
+      "step": 8392,
+      "vm_loss": 0.1548
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.4955,
+      "step": 8392,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.7145,
+      "step": 8392,
+      "vm_loss": 0.0861
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.7463,
+      "step": 8392,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.8001,
+      "step": 8392,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.7204,
+      "step": 8392,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.6155160382125755,
+      "lm_loss": 0.6268,
+      "step": 8392,
+      "vm_loss": 0.248
+    },
+    {
+      "epoch": 1.615708544890151,
+      "grad_norm": 3.1966809273610783,
+      "learning_rate": 1.8729711516821603e-06,
+      "loss": 0.902,
+      "step": 8393
+    },
+    {
+      "epoch": 1.6159010515677261,
+      "grad_norm": 3.330473567364802,
+      "learning_rate": 1.8711548159712246e-06,
+      "loss": 0.9305,
+      "step": 8394
+    },
+    {
+      "epoch": 1.6160935582453018,
+      "grad_norm": 3.3675340132816243,
+      "learning_rate": 1.8693392704880409e-06,
+      "loss": 0.9552,
+      "step": 8395
+    },
+    {
+      "epoch": 1.616286064922877,
+      "grad_norm": 3.1219829488529887,
+      "learning_rate": 1.867524515409097e-06,
+      "loss": 0.8774,
+      "step": 8396
+    },
+    {
+      "epoch": 1.6164785716004524,
+      "grad_norm": 3.0600472748064145,
+      "learning_rate": 1.8657105509108208e-06,
+      "loss": 0.8205,
+      "step": 8397
+    },
+    {
+      "epoch": 1.6166710782780278,
+      "grad_norm": 3.0703600265301887,
+      "learning_rate": 1.8638973771695446e-06,
+      "loss": 0.8581,
+      "step": 8398
+    },
+    {
+      "epoch": 1.616863584955603,
+      "grad_norm": 3.1689308778930445,
+      "learning_rate": 1.8620849943615337e-06,
+      "loss": 0.903,
+      "step": 8399
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "grad_norm": 3.3200598172334077,
+      "learning_rate": 1.8602734026629753e-06,
+      "loss": 0.924,
+      "step": 8400
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.4931,
+      "step": 8400,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.6529,
+      "step": 8400,
+      "vm_loss": 0.2643
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.722,
+      "step": 8400,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.6142,
+      "step": 8400,
+      "vm_loss": 0.236
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.7254,
+      "step": 8400,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 1.0071,
+      "step": 8400,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 1.5374,
+      "step": 8400,
+      "vm_loss": 0.1457
+    },
+    {
+      "epoch": 1.6170560916331786,
+      "lm_loss": 0.5954,
+      "step": 8400,
+      "vm_loss": 0.1247
+    },
+    {
+      "epoch": 1.6172485983107538,
+      "grad_norm": 3.158453591593248,
+      "learning_rate": 1.8584626022499808e-06,
+      "loss": 0.8749,
+      "step": 8401
+    },
+    {
+      "epoch": 1.6174411049883293,
+      "grad_norm": 3.2471861623623757,
+      "learning_rate": 1.8566525932985836e-06,
+      "loss": 0.913,
+      "step": 8402
+    },
+    {
+      "epoch": 1.6176336116659047,
+      "grad_norm": 3.0919409473059174,
+      "learning_rate": 1.854843375984734e-06,
+      "loss": 0.853,
+      "step": 8403
+    },
+    {
+      "epoch": 1.61782611834348,
+      "grad_norm": 3.1755453932865247,
+      "learning_rate": 1.8530349504843193e-06,
+      "loss": 0.8857,
+      "step": 8404
+    },
+    {
+      "epoch": 1.6180186250210555,
+      "grad_norm": 3.2251113956096527,
+      "learning_rate": 1.8512273169731353e-06,
+      "loss": 0.8854,
+      "step": 8405
+    },
+    {
+      "epoch": 1.6182111316986307,
+      "grad_norm": 3.2533062859646305,
+      "learning_rate": 1.8494204756269097e-06,
+      "loss": 0.8965,
+      "step": 8406
+    },
+    {
+      "epoch": 1.6184036383762062,
+      "grad_norm": 3.2917705464218963,
+      "learning_rate": 1.8476144266212903e-06,
+      "loss": 0.8747,
+      "step": 8407
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "grad_norm": 3.155665333987961,
+      "learning_rate": 1.8458091701318504e-06,
+      "loss": 0.8611,
+      "step": 8408
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.7215,
+      "step": 8408,
+      "vm_loss": 0.2088
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.8097,
+      "step": 8408,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.6905,
+      "step": 8408,
+      "vm_loss": 0.2283
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.2811,
+      "step": 8408,
+      "vm_loss": 0.2079
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.9875,
+      "step": 8408,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.6309,
+      "step": 8408,
+      "vm_loss": 0.1961
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.5143,
+      "step": 8408,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 1.6185961450537816,
+      "lm_loss": 0.7739,
+      "step": 8408,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 1.6187886517313568,
+      "grad_norm": 3.279446183978866,
+      "learning_rate": 1.8440047063340793e-06,
+      "loss": 0.9392,
+      "step": 8409
+    },
+    {
+      "epoch": 1.6189811584089324,
+      "grad_norm": 3.3061220484966984,
+      "learning_rate": 1.8422010354033958e-06,
+      "loss": 0.8998,
+      "step": 8410
+    },
+    {
+      "epoch": 1.6191736650865076,
+      "grad_norm": 3.289514214265429,
+      "learning_rate": 1.840398157515142e-06,
+      "loss": 0.9142,
+      "step": 8411
+    },
+    {
+      "epoch": 1.619366171764083,
+      "grad_norm": 3.223875227373888,
+      "learning_rate": 1.8385960728445795e-06,
+      "loss": 0.8899,
+      "step": 8412
+    },
+    {
+      "epoch": 1.6195586784416585,
+      "grad_norm": 3.414341773206159,
+      "learning_rate": 1.8367947815668897e-06,
+      "loss": 0.9362,
+      "step": 8413
+    },
+    {
+      "epoch": 1.6197511851192337,
+      "grad_norm": 3.321590689251058,
+      "learning_rate": 1.8349942838571898e-06,
+      "loss": 0.8785,
+      "step": 8414
+    },
+    {
+      "epoch": 1.6199436917968093,
+      "grad_norm": 3.302479337004412,
+      "learning_rate": 1.8331945798905048e-06,
+      "loss": 0.9395,
+      "step": 8415
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "grad_norm": 3.2153763912663464,
+      "learning_rate": 1.831395669841789e-06,
+      "loss": 0.8722,
+      "step": 8416
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.8092,
+      "step": 8416,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.7421,
+      "step": 8416,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 1.2135,
+      "step": 8416,
+      "vm_loss": 0.1081
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.8927,
+      "step": 8416,
+      "vm_loss": 0.2702
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.4406,
+      "step": 8416,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.4905,
+      "step": 8416,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.6884,
+      "step": 8416,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.6201361984743845,
+      "lm_loss": 0.5488,
+      "step": 8416,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.62032870515196,
+      "grad_norm": 3.249231698173159,
+      "learning_rate": 1.8295975538859223e-06,
+      "loss": 0.8917,
+      "step": 8417
+    },
+    {
+      "epoch": 1.6205212118295353,
+      "grad_norm": 3.316148652215696,
+      "learning_rate": 1.827800232197705e-06,
+      "loss": 0.9102,
+      "step": 8418
+    },
+    {
+      "epoch": 1.6207137185071108,
+      "grad_norm": 3.1653376179242865,
+      "learning_rate": 1.8260037049518565e-06,
+      "loss": 0.8825,
+      "step": 8419
+    },
+    {
+      "epoch": 1.6209062251846862,
+      "grad_norm": 3.2845834132360365,
+      "learning_rate": 1.8242079723230232e-06,
+      "loss": 0.8823,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6210987318622614,
+      "grad_norm": 3.151334676801339,
+      "learning_rate": 1.8224130344857738e-06,
+      "loss": 0.865,
+      "step": 8421
+    },
+    {
+      "epoch": 1.6212912385398368,
+      "grad_norm": 3.250751693760326,
+      "learning_rate": 1.8206188916145995e-06,
+      "loss": 0.8975,
+      "step": 8422
+    },
+    {
+      "epoch": 1.6214837452174122,
+      "grad_norm": 3.3547001182476883,
+      "learning_rate": 1.818825543883914e-06,
+      "loss": 0.9069,
+      "step": 8423
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "grad_norm": 3.2711974117197395,
+      "learning_rate": 1.8170329914680529e-06,
+      "loss": 0.8804,
+      "step": 8424
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.6618,
+      "step": 8424,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.8496,
+      "step": 8424,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.6816,
+      "step": 8424,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.4988,
+      "step": 8424,
+      "vm_loss": 0.2584
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.4379,
+      "step": 8424,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.7708,
+      "step": 8424,
+      "vm_loss": 0.3248
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.5707,
+      "step": 8424,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 1.6216762518949877,
+      "lm_loss": 0.5124,
+      "step": 8424,
+      "vm_loss": 0.1298
+    },
+    {
+      "epoch": 1.621868758572563,
+      "grad_norm": 3.436717805763262,
+      "learning_rate": 1.8152412345412785e-06,
+      "loss": 0.9877,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6220612652501383,
+      "grad_norm": 3.2891167612215613,
+      "learning_rate": 1.813450273277767e-06,
+      "loss": 0.9038,
+      "step": 8426
+    },
+    {
+      "epoch": 1.6222537719277137,
+      "grad_norm": 3.0887591458038672,
+      "learning_rate": 1.811660107851626e-06,
+      "loss": 0.842,
+      "step": 8427
+    },
+    {
+      "epoch": 1.6224462786052891,
+      "grad_norm": 3.1012407371831148,
+      "learning_rate": 1.8098707384368818e-06,
+      "loss": 0.8232,
+      "step": 8428
+    },
+    {
+      "epoch": 1.6226387852828645,
+      "grad_norm": 3.1088858892259883,
+      "learning_rate": 1.8080821652074865e-06,
+      "loss": 0.8291,
+      "step": 8429
+    },
+    {
+      "epoch": 1.62283129196044,
+      "grad_norm": 3.317213780404013,
+      "learning_rate": 1.8062943883373052e-06,
+      "loss": 0.8629,
+      "step": 8430
+    },
+    {
+      "epoch": 1.6230237986380152,
+      "grad_norm": 3.402243252034283,
+      "learning_rate": 1.8045074080001423e-06,
+      "loss": 0.9125,
+      "step": 8431
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "grad_norm": 3.323981140240041,
+      "learning_rate": 1.8027212243697079e-06,
+      "loss": 0.9263,
+      "step": 8432
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.7551,
+      "step": 8432,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.5408,
+      "step": 8432,
+      "vm_loss": 0.1065
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.3836,
+      "step": 8432,
+      "vm_loss": 0.2074
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.8257,
+      "step": 8432,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.8335,
+      "step": 8432,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.7664,
+      "step": 8432,
+      "vm_loss": 0.2516
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.6236,
+      "step": 8432,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.6232163053155906,
+      "lm_loss": 0.477,
+      "step": 8432,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.623408811993166,
+      "grad_norm": 3.411989164891271,
+      "learning_rate": 1.8009358376196472e-06,
+      "loss": 0.9246,
+      "step": 8433
+    },
+    {
+      "epoch": 1.6236013186707414,
+      "grad_norm": 3.34692872799914,
+      "learning_rate": 1.799151247923514e-06,
+      "loss": 0.9258,
+      "step": 8434
+    },
+    {
+      "epoch": 1.6237938253483168,
+      "grad_norm": 3.198446431654013,
+      "learning_rate": 1.7973674554548047e-06,
+      "loss": 0.8479,
+      "step": 8435
+    },
+    {
+      "epoch": 1.623986332025892,
+      "grad_norm": 3.229715294554512,
+      "learning_rate": 1.7955844603869176e-06,
+      "loss": 0.8765,
+      "step": 8436
+    },
+    {
+      "epoch": 1.6241788387034677,
+      "grad_norm": 3.267660131337464,
+      "learning_rate": 1.7938022628931872e-06,
+      "loss": 0.9076,
+      "step": 8437
+    },
+    {
+      "epoch": 1.6243713453810429,
+      "grad_norm": 3.471213799150422,
+      "learning_rate": 1.7920208631468638e-06,
+      "loss": 0.9282,
+      "step": 8438
+    },
+    {
+      "epoch": 1.6245638520586183,
+      "grad_norm": 3.389314970537656,
+      "learning_rate": 1.7902402613211234e-06,
+      "loss": 0.9298,
+      "step": 8439
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "grad_norm": 3.3980879821869245,
+      "learning_rate": 1.7884604575890641e-06,
+      "loss": 0.9302,
+      "step": 8440
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.7974,
+      "step": 8440,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.8428,
+      "step": 8440,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.5434,
+      "step": 8440,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 1.0095,
+      "step": 8440,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.6346,
+      "step": 8440,
+      "vm_loss": 0.1204
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 1.308,
+      "step": 8440,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.856,
+      "step": 8440,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.6247563587361937,
+      "lm_loss": 0.4346,
+      "step": 8440,
+      "vm_loss": 0.1756
+    },
+    {
+      "epoch": 1.624948865413769,
+      "grad_norm": 3.233786626856265,
+      "learning_rate": 1.7866814521237064e-06,
+      "loss": 0.919,
+      "step": 8441
+    },
+    {
+      "epoch": 1.6251413720913446,
+      "grad_norm": 3.1480802322852006,
+      "learning_rate": 1.7849032450979876e-06,
+      "loss": 0.8919,
+      "step": 8442
+    },
+    {
+      "epoch": 1.6253338787689198,
+      "grad_norm": 3.3445500803941233,
+      "learning_rate": 1.783125836684776e-06,
+      "loss": 0.9082,
+      "step": 8443
+    },
+    {
+      "epoch": 1.6255263854464952,
+      "grad_norm": 3.1112929076021,
+      "learning_rate": 1.7813492270568566e-06,
+      "loss": 0.8432,
+      "step": 8444
+    },
+    {
+      "epoch": 1.6257188921240706,
+      "grad_norm": 3.2506330381964625,
+      "learning_rate": 1.7795734163869405e-06,
+      "loss": 0.8703,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6259113988016458,
+      "grad_norm": 3.3433536996165163,
+      "learning_rate": 1.7777984048476605e-06,
+      "loss": 0.908,
+      "step": 8446
+    },
+    {
+      "epoch": 1.6261039054792215,
+      "grad_norm": 3.2306418835900264,
+      "learning_rate": 1.776024192611563e-06,
+      "loss": 0.9196,
+      "step": 8447
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "grad_norm": 3.186214674865702,
+      "learning_rate": 1.7742507798511345e-06,
+      "loss": 0.8619,
+      "step": 8448
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.8864,
+      "step": 8448,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.5078,
+      "step": 8448,
+      "vm_loss": 0.1136
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.8447,
+      "step": 8448,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.7178,
+      "step": 8448,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.7485,
+      "step": 8448,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.4119,
+      "step": 8448,
+      "vm_loss": 0.1177
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.6146,
+      "step": 8448,
+      "vm_loss": 0.2764
+    },
+    {
+      "epoch": 1.6262964121567967,
+      "lm_loss": 0.9143,
+      "step": 8448,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.626488918834372,
+      "grad_norm": 3.170357594475807,
+      "learning_rate": 1.7724781667387658e-06,
+      "loss": 0.8708,
+      "step": 8449
+    },
+    {
+      "epoch": 1.6266814255119475,
+      "grad_norm": 3.309933029916249,
+      "learning_rate": 1.7707063534467816e-06,
+      "loss": 0.9212,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6268739321895227,
+      "grad_norm": 3.3356447852556803,
+      "learning_rate": 1.7689353401474197e-06,
+      "loss": 0.9163,
+      "step": 8451
+    },
+    {
+      "epoch": 1.6270664388670983,
+      "grad_norm": 3.2911631129879972,
+      "learning_rate": 1.7671651270128531e-06,
+      "loss": 0.9489,
+      "step": 8452
+    },
+    {
+      "epoch": 1.6272589455446735,
+      "grad_norm": 3.194873105069443,
+      "learning_rate": 1.7653957142151622e-06,
+      "loss": 0.8553,
+      "step": 8453
+    },
+    {
+      "epoch": 1.627451452222249,
+      "grad_norm": 3.3279564253509384,
+      "learning_rate": 1.7636271019263596e-06,
+      "loss": 0.908,
+      "step": 8454
+    },
+    {
+      "epoch": 1.6276439588998244,
+      "grad_norm": 3.1447300247087275,
+      "learning_rate": 1.7618592903183761e-06,
+      "loss": 0.863,
+      "step": 8455
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "grad_norm": 3.1006719064477886,
+      "learning_rate": 1.7600922795630693e-06,
+      "loss": 0.8685,
+      "step": 8456
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.8434,
+      "step": 8456,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.591,
+      "step": 8456,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.8466,
+      "step": 8456,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.5618,
+      "step": 8456,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.8751,
+      "step": 8456,
+      "vm_loss": 0.1363
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.8954,
+      "step": 8456,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.797,
+      "step": 8456,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.6278364655773996,
+      "lm_loss": 0.8621,
+      "step": 8456,
+      "vm_loss": 0.2175
+    },
+    {
+      "epoch": 1.6280289722549752,
+      "grad_norm": 3.2863412717117297,
+      "learning_rate": 1.7583260698322069e-06,
+      "loss": 0.8907,
+      "step": 8457
+    },
+    {
+      "epoch": 1.6282214789325504,
+      "grad_norm": 3.261475185764014,
+      "learning_rate": 1.7565606612974983e-06,
+      "loss": 0.9099,
+      "step": 8458
+    },
+    {
+      "epoch": 1.6284139856101258,
+      "grad_norm": 3.335473571539051,
+      "learning_rate": 1.7547960541305542e-06,
+      "loss": 0.9085,
+      "step": 8459
+    },
+    {
+      "epoch": 1.6286064922877013,
+      "grad_norm": 3.3044845573058304,
+      "learning_rate": 1.7530322485029206e-06,
+      "loss": 0.8913,
+      "step": 8460
+    },
+    {
+      "epoch": 1.6287989989652765,
+      "grad_norm": 3.2754737833853933,
+      "learning_rate": 1.7512692445860624e-06,
+      "loss": 0.8734,
+      "step": 8461
+    },
+    {
+      "epoch": 1.6289915056428521,
+      "grad_norm": 3.4135468627258514,
+      "learning_rate": 1.749507042551365e-06,
+      "loss": 0.9834,
+      "step": 8462
+    },
+    {
+      "epoch": 1.6291840123204273,
+      "grad_norm": 3.149913399872908,
+      "learning_rate": 1.747745642570141e-06,
+      "loss": 0.8771,
+      "step": 8463
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "grad_norm": 3.2509669728524124,
+      "learning_rate": 1.745985044813615e-06,
+      "loss": 0.882,
+      "step": 8464
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.3118,
+      "step": 8464,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.8766,
+      "step": 8464,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.5112,
+      "step": 8464,
+      "vm_loss": 0.0883
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.7389,
+      "step": 8464,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.8028,
+      "step": 8464,
+      "vm_loss": 0.1139
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.9244,
+      "step": 8464,
+      "vm_loss": 0.1226
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.7022,
+      "step": 8464,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.6293765189980027,
+      "lm_loss": 0.5815,
+      "step": 8464,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.6295690256755782,
+      "grad_norm": 3.3816111957347235,
+      "learning_rate": 1.7442252494529422e-06,
+      "loss": 0.9279,
+      "step": 8465
+    },
+    {
+      "epoch": 1.6297615323531534,
+      "grad_norm": 3.3325969122934556,
+      "learning_rate": 1.7424662566591967e-06,
+      "loss": 0.8919,
+      "step": 8466
+    },
+    {
+      "epoch": 1.629954039030729,
+      "grad_norm": 3.172126376671785,
+      "learning_rate": 1.7407080666033794e-06,
+      "loss": 0.8643,
+      "step": 8467
+    },
+    {
+      "epoch": 1.6301465457083042,
+      "grad_norm": 3.2714323919169916,
+      "learning_rate": 1.7389506794563993e-06,
+      "loss": 0.8975,
+      "step": 8468
+    },
+    {
+      "epoch": 1.6303390523858796,
+      "grad_norm": 3.1733953268311983,
+      "learning_rate": 1.7371940953891087e-06,
+      "loss": 0.8959,
+      "step": 8469
+    },
+    {
+      "epoch": 1.630531559063455,
+      "grad_norm": 3.3448553950899598,
+      "learning_rate": 1.7354383145722608e-06,
+      "loss": 0.9756,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6307240657410302,
+      "grad_norm": 3.34341426493702,
+      "learning_rate": 1.7336833371765437e-06,
+      "loss": 0.9203,
+      "step": 8471
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "grad_norm": 3.1229779210848463,
+      "learning_rate": 1.7319291633725633e-06,
+      "loss": 0.831,
+      "step": 8472
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.5602,
+      "step": 8472,
+      "vm_loss": 0.226
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.5658,
+      "step": 8472,
+      "vm_loss": 0.1131
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.8256,
+      "step": 8472,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.6543,
+      "step": 8472,
+      "vm_loss": 0.2484
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.6027,
+      "step": 8472,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 1.0274,
+      "step": 8472,
+      "vm_loss": 0.2017
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.4151,
+      "step": 8472,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 1.6309165724186059,
+      "lm_loss": 0.8648,
+      "step": 8472,
+      "vm_loss": 0.2343
+    },
+    {
+      "epoch": 1.631109079096181,
+      "grad_norm": 3.284544127801621,
+      "learning_rate": 1.7301757933308506e-06,
+      "loss": 0.9016,
+      "step": 8473
+    },
+    {
+      "epoch": 1.6313015857737565,
+      "grad_norm": 3.3163125254080814,
+      "learning_rate": 1.7284232272218503e-06,
+      "loss": 0.8838,
+      "step": 8474
+    },
+    {
+      "epoch": 1.631494092451332,
+      "grad_norm": 3.2834695956852467,
+      "learning_rate": 1.7266714652159354e-06,
+      "loss": 0.8553,
+      "step": 8475
+    },
+    {
+      "epoch": 1.6316865991289071,
+      "grad_norm": 3.3235561456293023,
+      "learning_rate": 1.7249205074834019e-06,
+      "loss": 0.9274,
+      "step": 8476
+    },
+    {
+      "epoch": 1.6318791058064828,
+      "grad_norm": 3.1648255902986917,
+      "learning_rate": 1.7231703541944666e-06,
+      "loss": 0.8467,
+      "step": 8477
+    },
+    {
+      "epoch": 1.632071612484058,
+      "grad_norm": 3.1710467917544287,
+      "learning_rate": 1.7214210055192604e-06,
+      "loss": 0.9198,
+      "step": 8478
+    },
+    {
+      "epoch": 1.6322641191616334,
+      "grad_norm": 3.045766966043062,
+      "learning_rate": 1.7196724616278504e-06,
+      "loss": 0.8223,
+      "step": 8479
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "grad_norm": 3.1697415153983646,
+      "learning_rate": 1.7179247226902107e-06,
+      "loss": 0.8947,
+      "step": 8480
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.9598,
+      "step": 8480,
+      "vm_loss": 0.0892
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.6499,
+      "step": 8480,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 1.0453,
+      "step": 8480,
+      "vm_loss": 0.2684
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.6136,
+      "step": 8480,
+      "vm_loss": 0.1085
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.8492,
+      "step": 8480,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.8723,
+      "step": 8480,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.2933,
+      "step": 8480,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 1.6324566258392088,
+      "lm_loss": 0.6891,
+      "step": 8480,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 1.6326491325167842,
+      "grad_norm": 3.3418944145619416,
+      "learning_rate": 1.7161777888762466e-06,
+      "loss": 0.907,
+      "step": 8481
+    },
+    {
+      "epoch": 1.6328416391943597,
+      "grad_norm": 3.2800683917061684,
+      "learning_rate": 1.7144316603557832e-06,
+      "loss": 0.8738,
+      "step": 8482
+    },
+    {
+      "epoch": 1.6330341458719349,
+      "grad_norm": 3.169754492854183,
+      "learning_rate": 1.7126863372985658e-06,
+      "loss": 0.8656,
+      "step": 8483
+    },
+    {
+      "epoch": 1.6332266525495103,
+      "grad_norm": 3.3629042542528773,
+      "learning_rate": 1.7109418198742644e-06,
+      "loss": 0.9378,
+      "step": 8484
+    },
+    {
+      "epoch": 1.6334191592270857,
+      "grad_norm": 3.2717371783330926,
+      "learning_rate": 1.7091981082524633e-06,
+      "loss": 0.9161,
+      "step": 8485
+    },
+    {
+      "epoch": 1.6336116659046611,
+      "grad_norm": 3.105697767564221,
+      "learning_rate": 1.7074552026026759e-06,
+      "loss": 0.8368,
+      "step": 8486
+    },
+    {
+      "epoch": 1.6338041725822365,
+      "grad_norm": 3.2565201189663266,
+      "learning_rate": 1.7057131030943364e-06,
+      "loss": 0.8834,
+      "step": 8487
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "grad_norm": 3.1515380798590265,
+      "learning_rate": 1.7039718098967983e-06,
+      "loss": 0.8381,
+      "step": 8488
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.7819,
+      "step": 8488,
+      "vm_loss": 0.2411
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.6532,
+      "step": 8488,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 1.2335,
+      "step": 8488,
+      "vm_loss": 0.2049
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.6933,
+      "step": 8488,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.4939,
+      "step": 8488,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.6833,
+      "step": 8488,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.7266,
+      "step": 8488,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.6339966792598117,
+      "lm_loss": 0.6452,
+      "step": 8488,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 1.6341891859373872,
+      "grad_norm": 3.2618775771290904,
+      "learning_rate": 1.7022313231793364e-06,
+      "loss": 0.8824,
+      "step": 8489
+    },
+    {
+      "epoch": 1.6343816926149626,
+      "grad_norm": 3.3458552636418193,
+      "learning_rate": 1.7004916431111529e-06,
+      "loss": 0.869,
+      "step": 8490
+    },
+    {
+      "epoch": 1.634574199292538,
+      "grad_norm": 3.294678228271462,
+      "learning_rate": 1.6987527698613603e-06,
+      "loss": 0.9186,
+      "step": 8491
+    },
+    {
+      "epoch": 1.6347667059701134,
+      "grad_norm": 3.1896106307745815,
+      "learning_rate": 1.6970147035990025e-06,
+      "loss": 0.8598,
+      "step": 8492
+    },
+    {
+      "epoch": 1.6349592126476886,
+      "grad_norm": 3.113586851436277,
+      "learning_rate": 1.695277444493042e-06,
+      "loss": 0.8252,
+      "step": 8493
+    },
+    {
+      "epoch": 1.635151719325264,
+      "grad_norm": 3.234654844009193,
+      "learning_rate": 1.6935409927123647e-06,
+      "loss": 0.8698,
+      "step": 8494
+    },
+    {
+      "epoch": 1.6353442260028395,
+      "grad_norm": 3.3971615641815767,
+      "learning_rate": 1.6918053484257691e-06,
+      "loss": 0.921,
+      "step": 8495
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "grad_norm": 3.186607063898726,
+      "learning_rate": 1.69007051180199e-06,
+      "loss": 0.8914,
+      "step": 8496
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.5472,
+      "step": 8496,
+      "vm_loss": 0.1213
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.8396,
+      "step": 8496,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.7031,
+      "step": 8496,
+      "vm_loss": 0.1417
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 1.0508,
+      "step": 8496,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.7819,
+      "step": 8496,
+      "vm_loss": 0.1054
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.3792,
+      "step": 8496,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.8541,
+      "step": 8496,
+      "vm_loss": 0.1048
+    },
+    {
+      "epoch": 1.6355367326804149,
+      "lm_loss": 0.6582,
+      "step": 8496,
+      "vm_loss": 0.1207
+    },
+    {
+      "epoch": 1.6357292393579903,
+      "grad_norm": 3.245727340390148,
+      "learning_rate": 1.6883364830096705e-06,
+      "loss": 0.8436,
+      "step": 8497
+    },
+    {
+      "epoch": 1.6359217460355655,
+      "grad_norm": 3.2731521555414744,
+      "learning_rate": 1.6866032622173822e-06,
+      "loss": 0.9053,
+      "step": 8498
+    },
+    {
+      "epoch": 1.6361142527131411,
+      "grad_norm": 3.370282944351226,
+      "learning_rate": 1.6848708495936161e-06,
+      "loss": 0.8897,
+      "step": 8499
+    },
+    {
+      "epoch": 1.6363067593907163,
+      "grad_norm": 3.402486474017815,
+      "learning_rate": 1.6831392453067885e-06,
+      "loss": 0.9346,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6364992660682918,
+      "grad_norm": 3.151677143009986,
+      "learning_rate": 1.6814084495252258e-06,
+      "loss": 0.8477,
+      "step": 8501
+    },
+    {
+      "epoch": 1.6366917727458672,
+      "grad_norm": 3.161219138083262,
+      "learning_rate": 1.6796784624171891e-06,
+      "loss": 0.8901,
+      "step": 8502
+    },
+    {
+      "epoch": 1.6368842794234424,
+      "grad_norm": 3.309755527022719,
+      "learning_rate": 1.6779492841508527e-06,
+      "loss": 0.9112,
+      "step": 8503
+    },
+    {
+      "epoch": 1.637076786101018,
+      "grad_norm": 3.22390679508785,
+      "learning_rate": 1.6762209148943164e-06,
+      "loss": 0.9189,
+      "step": 8504
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.7057,
+      "step": 8504,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.7496,
+      "step": 8504,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 1.1109,
+      "step": 8504,
+      "vm_loss": 0.2299
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.4952,
+      "step": 8504,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.6524,
+      "step": 8504,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.6033,
+      "step": 8504,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.9249,
+      "step": 8504,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.637076786101018,
+      "lm_loss": 0.8034,
+      "step": 8504,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.6372692927785932,
+      "grad_norm": 3.3046820577186833,
+      "learning_rate": 1.6744933548155994e-06,
+      "loss": 0.8885,
+      "step": 8505
+    },
+    {
+      "epoch": 1.6374617994561687,
+      "grad_norm": 3.0990160240948894,
+      "learning_rate": 1.672766604082643e-06,
+      "loss": 0.8427,
+      "step": 8506
+    },
+    {
+      "epoch": 1.637654306133744,
+      "grad_norm": 3.356322915855534,
+      "learning_rate": 1.6710406628633113e-06,
+      "loss": 0.9442,
+      "step": 8507
+    },
+    {
+      "epoch": 1.6378468128113193,
+      "grad_norm": 3.3079469237203347,
+      "learning_rate": 1.6693155313253828e-06,
+      "loss": 0.9227,
+      "step": 8508
+    },
+    {
+      "epoch": 1.638039319488895,
+      "grad_norm": 3.2399651576207256,
+      "learning_rate": 1.6675912096365655e-06,
+      "loss": 0.8945,
+      "step": 8509
+    },
+    {
+      "epoch": 1.6382318261664701,
+      "grad_norm": 3.2265612099292786,
+      "learning_rate": 1.6658676979644862e-06,
+      "loss": 0.9297,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6384243328440455,
+      "grad_norm": 3.35450050058546,
+      "learning_rate": 1.664144996476692e-06,
+      "loss": 0.9128,
+      "step": 8511
+    },
+    {
+      "epoch": 1.638616839521621,
+      "grad_norm": 3.237006886051614,
+      "learning_rate": 1.662423105340647e-06,
+      "loss": 0.9146,
+      "step": 8512
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.6118,
+      "step": 8512,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.4682,
+      "step": 8512,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.6492,
+      "step": 8512,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.4621,
+      "step": 8512,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.5928,
+      "step": 8512,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.4592,
+      "step": 8512,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.529,
+      "step": 8512,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 1.638616839521621,
+      "lm_loss": 0.5069,
+      "step": 8512,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 1.6388093461991962,
+      "grad_norm": 3.1268579958972498,
+      "learning_rate": 1.6607020247237504e-06,
+      "loss": 0.8296,
+      "step": 8513
+    },
+    {
+      "epoch": 1.6390018528767718,
+      "grad_norm": 3.1389705693022103,
+      "learning_rate": 1.658981754793304e-06,
+      "loss": 0.856,
+      "step": 8514
+    },
+    {
+      "epoch": 1.639194359554347,
+      "grad_norm": 3.2782678120357076,
+      "learning_rate": 1.657262295716545e-06,
+      "loss": 0.8907,
+      "step": 8515
+    },
+    {
+      "epoch": 1.6393868662319224,
+      "grad_norm": 3.2826293137258773,
+      "learning_rate": 1.6555436476606246e-06,
+      "loss": 0.9095,
+      "step": 8516
+    },
+    {
+      "epoch": 1.6395793729094978,
+      "grad_norm": 3.2464552625179204,
+      "learning_rate": 1.6538258107926219e-06,
+      "loss": 0.8571,
+      "step": 8517
+    },
+    {
+      "epoch": 1.639771879587073,
+      "grad_norm": 3.2942051842256954,
+      "learning_rate": 1.652108785279526e-06,
+      "loss": 0.8816,
+      "step": 8518
+    },
+    {
+      "epoch": 1.6399643862646487,
+      "grad_norm": 3.362549487709971,
+      "learning_rate": 1.6503925712882573e-06,
+      "loss": 0.9566,
+      "step": 8519
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "grad_norm": 3.1056394535793124,
+      "learning_rate": 1.6486771689856528e-06,
+      "loss": 0.8531,
+      "step": 8520
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.9199,
+      "step": 8520,
+      "vm_loss": 0.2464
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.7875,
+      "step": 8520,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 1.1308,
+      "step": 8520,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.5517,
+      "step": 8520,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.6735,
+      "step": 8520,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.4264,
+      "step": 8520,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.8659,
+      "step": 8520,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.6401568929422239,
+      "lm_loss": 0.3875,
+      "step": 8520,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 1.6403493996197993,
+      "grad_norm": 3.3269962025473387,
+      "learning_rate": 1.6469625785384746e-06,
+      "loss": 0.9149,
+      "step": 8521
+    },
+    {
+      "epoch": 1.6405419062973747,
+      "grad_norm": 3.3960128002494754,
+      "learning_rate": 1.645248800113396e-06,
+      "loss": 0.9088,
+      "step": 8522
+    },
+    {
+      "epoch": 1.64073441297495,
+      "grad_norm": 3.1854575576083626,
+      "learning_rate": 1.643535833877027e-06,
+      "loss": 0.8762,
+      "step": 8523
+    },
+    {
+      "epoch": 1.6409269196525256,
+      "grad_norm": 3.255071206202876,
+      "learning_rate": 1.641823679995883e-06,
+      "loss": 0.9089,
+      "step": 8524
+    },
+    {
+      "epoch": 1.6411194263301008,
+      "grad_norm": 3.2290354368499243,
+      "learning_rate": 1.6401123386364094e-06,
+      "loss": 0.8882,
+      "step": 8525
+    },
+    {
+      "epoch": 1.6413119330076762,
+      "grad_norm": 3.229150164708246,
+      "learning_rate": 1.6384018099649712e-06,
+      "loss": 0.9201,
+      "step": 8526
+    },
+    {
+      "epoch": 1.6415044396852516,
+      "grad_norm": 3.2942404252259494,
+      "learning_rate": 1.6366920941478526e-06,
+      "loss": 0.8578,
+      "step": 8527
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "grad_norm": 3.4658893470830137,
+      "learning_rate": 1.6349831913512626e-06,
+      "loss": 0.961,
+      "step": 8528
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.6479,
+      "step": 8528,
+      "vm_loss": 0.1301
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.7459,
+      "step": 8528,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.6563,
+      "step": 8528,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.4744,
+      "step": 8528,
+      "vm_loss": 0.0857
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.5223,
+      "step": 8528,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.6838,
+      "step": 8528,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.6971,
+      "step": 8528,
+      "vm_loss": 0.2503
+    },
+    {
+      "epoch": 1.6416969463628268,
+      "lm_loss": 0.5268,
+      "step": 8528,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 1.6418894530404025,
+      "grad_norm": 3.0272633848553787,
+      "learning_rate": 1.6332751017413218e-06,
+      "loss": 0.8131,
+      "step": 8529
+    },
+    {
+      "epoch": 1.6420819597179777,
+      "grad_norm": 3.1892090525502366,
+      "learning_rate": 1.6315678254840883e-06,
+      "loss": 0.8799,
+      "step": 8530
+    },
+    {
+      "epoch": 1.642274466395553,
+      "grad_norm": 3.3471680456617467,
+      "learning_rate": 1.6298613627455218e-06,
+      "loss": 0.9013,
+      "step": 8531
+    },
+    {
+      "epoch": 1.6424669730731285,
+      "grad_norm": 3.3814493772479453,
+      "learning_rate": 1.6281557136915172e-06,
+      "loss": 0.9323,
+      "step": 8532
+    },
+    {
+      "epoch": 1.6426594797507037,
+      "grad_norm": 3.4312198003595062,
+      "learning_rate": 1.6264508784878841e-06,
+      "loss": 0.9264,
+      "step": 8533
+    },
+    {
+      "epoch": 1.6428519864282793,
+      "grad_norm": 3.332540023792863,
+      "learning_rate": 1.6247468573003567e-06,
+      "loss": 0.9336,
+      "step": 8534
+    },
+    {
+      "epoch": 1.6430444931058545,
+      "grad_norm": 3.1879344360605733,
+      "learning_rate": 1.6230436502945834e-06,
+      "loss": 0.887,
+      "step": 8535
+    },
+    {
+      "epoch": 1.64323699978343,
+      "grad_norm": 3.218520511293312,
+      "learning_rate": 1.6213412576361398e-06,
+      "loss": 0.8721,
+      "step": 8536
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.5021,
+      "step": 8536,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.8111,
+      "step": 8536,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 1.0127,
+      "step": 8536,
+      "vm_loss": 0.1324
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.6422,
+      "step": 8536,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.5639,
+      "step": 8536,
+      "vm_loss": 0.2001
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.5559,
+      "step": 8536,
+      "vm_loss": 0.1259
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.8818,
+      "step": 8536,
+      "vm_loss": 0.2433
+    },
+    {
+      "epoch": 1.64323699978343,
+      "lm_loss": 0.8029,
+      "step": 8536,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 1.6434295064610054,
+      "grad_norm": 3.3612476835679668,
+      "learning_rate": 1.6196396794905212e-06,
+      "loss": 0.9,
+      "step": 8537
+    },
+    {
+      "epoch": 1.6436220131385806,
+      "grad_norm": 3.273889819108241,
+      "learning_rate": 1.617938916023143e-06,
+      "loss": 0.8946,
+      "step": 8538
+    },
+    {
+      "epoch": 1.6438145198161562,
+      "grad_norm": 3.12152354370881,
+      "learning_rate": 1.616238967399336e-06,
+      "loss": 0.8659,
+      "step": 8539
+    },
+    {
+      "epoch": 1.6440070264937314,
+      "grad_norm": 3.067912306756776,
+      "learning_rate": 1.6145398337843654e-06,
+      "loss": 0.8597,
+      "step": 8540
+    },
+    {
+      "epoch": 1.6441995331713068,
+      "grad_norm": 3.2016386983545146,
+      "learning_rate": 1.6128415153434017e-06,
+      "loss": 0.8772,
+      "step": 8541
+    },
+    {
+      "epoch": 1.6443920398488823,
+      "grad_norm": 3.273360390443935,
+      "learning_rate": 1.611144012241549e-06,
+      "loss": 0.9044,
+      "step": 8542
+    },
+    {
+      "epoch": 1.6445845465264577,
+      "grad_norm": 3.2786507347711518,
+      "learning_rate": 1.609447324643818e-06,
+      "loss": 0.912,
+      "step": 8543
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "grad_norm": 3.132915795662491,
+      "learning_rate": 1.6077514527151573e-06,
+      "loss": 0.8538,
+      "step": 8544
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.6591,
+      "step": 8544,
+      "vm_loss": 0.2326
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.79,
+      "step": 8544,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.8092,
+      "step": 8544,
+      "vm_loss": 0.1212
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.6855,
+      "step": 8544,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.4217,
+      "step": 8544,
+      "vm_loss": 0.2128
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.6045,
+      "step": 8544,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 1.1429,
+      "step": 8544,
+      "vm_loss": 0.216
+    },
+    {
+      "epoch": 1.6447770532040331,
+      "lm_loss": 0.6354,
+      "step": 8544,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.6449695598816083,
+      "grad_norm": 3.2683914646985457,
+      "learning_rate": 1.6060563966204224e-06,
+      "loss": 0.9285,
+      "step": 8545
+    },
+    {
+      "epoch": 1.6451620665591837,
+      "grad_norm": 3.3960710038787734,
+      "learning_rate": 1.604362156524395e-06,
+      "loss": 0.9454,
+      "step": 8546
+    },
+    {
+      "epoch": 1.6453545732367592,
+      "grad_norm": 3.402781970100242,
+      "learning_rate": 1.6026687325917779e-06,
+      "loss": 0.9087,
+      "step": 8547
+    },
+    {
+      "epoch": 1.6455470799143346,
+      "grad_norm": 3.213231186840626,
+      "learning_rate": 1.6009761249871936e-06,
+      "loss": 0.9269,
+      "step": 8548
+    },
+    {
+      "epoch": 1.64573958659191,
+      "grad_norm": 3.3529769349102314,
+      "learning_rate": 1.5992843338751874e-06,
+      "loss": 0.9029,
+      "step": 8549
+    },
+    {
+      "epoch": 1.6459320932694852,
+      "grad_norm": 3.2361917082805016,
+      "learning_rate": 1.5975933594202142e-06,
+      "loss": 0.8635,
+      "step": 8550
+    },
+    {
+      "epoch": 1.6461245999470606,
+      "grad_norm": 3.1983628388541985,
+      "learning_rate": 1.5959032017866705e-06,
+      "loss": 0.864,
+      "step": 8551
+    },
+    {
+      "epoch": 1.646317106624636,
+      "grad_norm": 3.127493535109191,
+      "learning_rate": 1.5942138611388536e-06,
+      "loss": 0.824,
+      "step": 8552
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.4443,
+      "step": 8552,
+      "vm_loss": 0.2146
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.6904,
+      "step": 8552,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.6321,
+      "step": 8552,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.7198,
+      "step": 8552,
+      "vm_loss": 0.08
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.7772,
+      "step": 8552,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.7045,
+      "step": 8552,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.8953,
+      "step": 8552,
+      "vm_loss": 0.2365
+    },
+    {
+      "epoch": 1.646317106624636,
+      "lm_loss": 0.7245,
+      "step": 8552,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 1.6465096133022115,
+      "grad_norm": 3.5815344476331736,
+      "learning_rate": 1.59252533764099e-06,
+      "loss": 0.9255,
+      "step": 8553
+    },
+    {
+      "epoch": 1.6467021199797869,
+      "grad_norm": 3.2068823413874283,
+      "learning_rate": 1.590837631457226e-06,
+      "loss": 0.8861,
+      "step": 8554
+    },
+    {
+      "epoch": 1.646894626657362,
+      "grad_norm": 3.2616126318494985,
+      "learning_rate": 1.5891507427516316e-06,
+      "loss": 0.8714,
+      "step": 8555
+    },
+    {
+      "epoch": 1.6470871333349375,
+      "grad_norm": 3.340829765072319,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.9373,
+      "step": 8556
+    },
+    {
+      "epoch": 1.647279640012513,
+      "grad_norm": 3.292852039147704,
+      "learning_rate": 1.5857794184308084e-06,
+      "loss": 0.91,
+      "step": 8557
+    },
+    {
+      "epoch": 1.6474721466900883,
+      "grad_norm": 3.1128553345387955,
+      "learning_rate": 1.5840949831433172e-06,
+      "loss": 0.8456,
+      "step": 8558
+    },
+    {
+      "epoch": 1.6476646533676638,
+      "grad_norm": 3.1853892489548463,
+      "learning_rate": 1.582411365989468e-06,
+      "loss": 0.8463,
+      "step": 8559
+    },
+    {
+      "epoch": 1.647857160045239,
+      "grad_norm": 3.1844989115021995,
+      "learning_rate": 1.5807285671329197e-06,
+      "loss": 0.8598,
+      "step": 8560
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.8092,
+      "step": 8560,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.5509,
+      "step": 8560,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.7338,
+      "step": 8560,
+      "vm_loss": 0.0948
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.5898,
+      "step": 8560,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.8763,
+      "step": 8560,
+      "vm_loss": 0.1927
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.8373,
+      "step": 8560,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.7365,
+      "step": 8560,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.647857160045239,
+      "lm_loss": 0.3853,
+      "step": 8560,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 1.6480496667228146,
+      "grad_norm": 3.105746661486708,
+      "learning_rate": 1.579046586737275e-06,
+      "loss": 0.8338,
+      "step": 8561
+    },
+    {
+      "epoch": 1.6482421734003898,
+      "grad_norm": 3.385922770303578,
+      "learning_rate": 1.5773654249660342e-06,
+      "loss": 0.9295,
+      "step": 8562
+    },
+    {
+      "epoch": 1.6484346800779652,
+      "grad_norm": 3.5486513855255257,
+      "learning_rate": 1.575685081982632e-06,
+      "loss": 0.9672,
+      "step": 8563
+    },
+    {
+      "epoch": 1.6486271867555407,
+      "grad_norm": 3.207737292454642,
+      "learning_rate": 1.5740055579504176e-06,
+      "loss": 0.885,
+      "step": 8564
+    },
+    {
+      "epoch": 1.6488196934331159,
+      "grad_norm": 3.2800049731018555,
+      "learning_rate": 1.5723268530326653e-06,
+      "loss": 0.8621,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6490122001106915,
+      "grad_norm": 3.2939256305741473,
+      "learning_rate": 1.5706489673925618e-06,
+      "loss": 0.8813,
+      "step": 8566
+    },
+    {
+      "epoch": 1.6492047067882667,
+      "grad_norm": 3.261284274347596,
+      "learning_rate": 1.5689719011932213e-06,
+      "loss": 0.8581,
+      "step": 8567
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "grad_norm": 3.3438703056349564,
+      "learning_rate": 1.567295654597677e-06,
+      "loss": 0.9254,
+      "step": 8568
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.8161,
+      "step": 8568,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.7037,
+      "step": 8568,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.7304,
+      "step": 8568,
+      "vm_loss": 0.2118
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.877,
+      "step": 8568,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.7825,
+      "step": 8568,
+      "vm_loss": 0.2216
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.5804,
+      "step": 8568,
+      "vm_loss": 0.116
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.3336,
+      "step": 8568,
+      "vm_loss": 0.1687
+    },
+    {
+      "epoch": 1.6493972134658421,
+      "lm_loss": 0.4479,
+      "step": 8568,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.6495897201434175,
+      "grad_norm": 3.286907855805531,
+      "learning_rate": 1.5656202277688803e-06,
+      "loss": 0.8613,
+      "step": 8569
+    },
+    {
+      "epoch": 1.6497822268209927,
+      "grad_norm": 3.5446860409728846,
+      "learning_rate": 1.5639456208697058e-06,
+      "loss": 0.9323,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6499747334985684,
+      "grad_norm": 3.469972201723012,
+      "learning_rate": 1.562271834062945e-06,
+      "loss": 0.916,
+      "step": 8571
+    },
+    {
+      "epoch": 1.6501672401761436,
+      "grad_norm": 3.431246399979431,
+      "learning_rate": 1.5605988675113137e-06,
+      "loss": 0.9067,
+      "step": 8572
+    },
+    {
+      "epoch": 1.650359746853719,
+      "grad_norm": 3.098578298550671,
+      "learning_rate": 1.5589267213774394e-06,
+      "loss": 0.8611,
+      "step": 8573
+    },
+    {
+      "epoch": 1.6505522535312944,
+      "grad_norm": 3.246010420679491,
+      "learning_rate": 1.5572553958238867e-06,
+      "loss": 0.8819,
+      "step": 8574
+    },
+    {
+      "epoch": 1.6507447602088696,
+      "grad_norm": 3.210330082708509,
+      "learning_rate": 1.55558489101312e-06,
+      "loss": 0.8519,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "grad_norm": 3.2499514432970322,
+      "learning_rate": 1.5539152071075402e-06,
+      "loss": 0.8956,
+      "step": 8576
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.6336,
+      "step": 8576,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.8766,
+      "step": 8576,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.7394,
+      "step": 8576,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.7994,
+      "step": 8576,
+      "vm_loss": 0.1476
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.6676,
+      "step": 8576,
+      "vm_loss": 0.1913
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.9927,
+      "step": 8576,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.7093,
+      "step": 8576,
+      "vm_loss": 0.1291
+    },
+    {
+      "epoch": 1.6509372668864453,
+      "lm_loss": 0.7242,
+      "step": 8576,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.6511297735640205,
+      "grad_norm": 3.309146418973245,
+      "learning_rate": 1.5522463442694546e-06,
+      "loss": 0.8736,
+      "step": 8577
+    },
+    {
+      "epoch": 1.6513222802415959,
+      "grad_norm": 3.291156938701736,
+      "learning_rate": 1.5505783026611076e-06,
+      "loss": 0.8936,
+      "step": 8578
+    },
+    {
+      "epoch": 1.6515147869191713,
+      "grad_norm": 3.2619975801347234,
+      "learning_rate": 1.5489110824446462e-06,
+      "loss": 0.8843,
+      "step": 8579
+    },
+    {
+      "epoch": 1.6517072935967465,
+      "grad_norm": 3.217711669431549,
+      "learning_rate": 1.5472446837821487e-06,
+      "loss": 0.8909,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6518998002743221,
+      "grad_norm": 3.221748673722485,
+      "learning_rate": 1.5455791068356108e-06,
+      "loss": 0.8783,
+      "step": 8581
+    },
+    {
+      "epoch": 1.6520923069518973,
+      "grad_norm": 3.0030024216598767,
+      "learning_rate": 1.5439143517669496e-06,
+      "loss": 0.8289,
+      "step": 8582
+    },
+    {
+      "epoch": 1.6522848136294728,
+      "grad_norm": 3.2378499897830086,
+      "learning_rate": 1.542250418737996e-06,
+      "loss": 0.8786,
+      "step": 8583
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "grad_norm": 3.265688236045046,
+      "learning_rate": 1.5405873079105083e-06,
+      "loss": 0.8939,
+      "step": 8584
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 1.0133,
+      "step": 8584,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.764,
+      "step": 8584,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.7781,
+      "step": 8584,
+      "vm_loss": 0.2221
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.6213,
+      "step": 8584,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 1.0419,
+      "step": 8584,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.4164,
+      "step": 8584,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.8432,
+      "step": 8584,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.6524773203070482,
+      "lm_loss": 0.5623,
+      "step": 8584,
+      "vm_loss": 0.2833
+    },
+    {
+      "epoch": 1.6526698269846234,
+      "grad_norm": 3.148808554522905,
+      "learning_rate": 1.5389250194461602e-06,
+      "loss": 0.8789,
+      "step": 8585
+    },
+    {
+      "epoch": 1.652862333662199,
+      "grad_norm": 3.2207234898934125,
+      "learning_rate": 1.5372635535065506e-06,
+      "loss": 0.8985,
+      "step": 8586
+    },
+    {
+      "epoch": 1.6530548403397742,
+      "grad_norm": 3.2975874142122885,
+      "learning_rate": 1.5356029102531933e-06,
+      "loss": 0.8988,
+      "step": 8587
+    },
+    {
+      "epoch": 1.6532473470173497,
+      "grad_norm": 3.4629366259803867,
+      "learning_rate": 1.533943089847526e-06,
+      "loss": 0.9151,
+      "step": 8588
+    },
+    {
+      "epoch": 1.653439853694925,
+      "grad_norm": 3.3645320469139035,
+      "learning_rate": 1.5322840924509009e-06,
+      "loss": 0.9093,
+      "step": 8589
+    },
+    {
+      "epoch": 1.6536323603725003,
+      "grad_norm": 3.314965778125061,
+      "learning_rate": 1.5306259182245963e-06,
+      "loss": 0.8768,
+      "step": 8590
+    },
+    {
+      "epoch": 1.653824867050076,
+      "grad_norm": 3.4655778784003086,
+      "learning_rate": 1.528968567329806e-06,
+      "loss": 0.9478,
+      "step": 8591
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "grad_norm": 3.263730130362224,
+      "learning_rate": 1.5273120399276486e-06,
+      "loss": 0.8664,
+      "step": 8592
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 1.0209,
+      "step": 8592,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.5994,
+      "step": 8592,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.5958,
+      "step": 8592,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.7338,
+      "step": 8592,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.3963,
+      "step": 8592,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.8106,
+      "step": 8592,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.8736,
+      "step": 8592,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 1.6540173737276511,
+      "lm_loss": 0.6983,
+      "step": 8592,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 1.6542098804052265,
+      "grad_norm": 3.0706368405825843,
+      "learning_rate": 1.5256563361791598e-06,
+      "loss": 0.8604,
+      "step": 8593
+    },
+    {
+      "epoch": 1.654402387082802,
+      "grad_norm": 3.2293666230213587,
+      "learning_rate": 1.5240014562452888e-06,
+      "loss": 0.8733,
+      "step": 8594
+    },
+    {
+      "epoch": 1.6545948937603772,
+      "grad_norm": 3.283028867293217,
+      "learning_rate": 1.522347400286921e-06,
+      "loss": 0.894,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6547874004379528,
+      "grad_norm": 3.2526690937293057,
+      "learning_rate": 1.5206941684648446e-06,
+      "loss": 0.8828,
+      "step": 8596
+    },
+    {
+      "epoch": 1.654979907115528,
+      "grad_norm": 3.1684084185086996,
+      "learning_rate": 1.5190417609397768e-06,
+      "loss": 0.8325,
+      "step": 8597
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 3.4292913454824787,
+      "learning_rate": 1.5173901778723532e-06,
+      "loss": 0.9443,
+      "step": 8598
+    },
+    {
+      "epoch": 1.6553649204706788,
+      "grad_norm": 3.180749309509719,
+      "learning_rate": 1.5157394194231301e-06,
+      "loss": 0.8549,
+      "step": 8599
+    },
+    {
+      "epoch": 1.655557427148254,
+      "grad_norm": 3.147228553786558,
+      "learning_rate": 1.514089485752579e-06,
+      "loss": 0.8551,
+      "step": 8600
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.7246,
+      "step": 8600,
+      "vm_loss": 0.1019
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.7496,
+      "step": 8600,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.867,
+      "step": 8600,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.7668,
+      "step": 8600,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.7195,
+      "step": 8600,
+      "vm_loss": 0.1621
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.8501,
+      "step": 8600,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.6099,
+      "step": 8600,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 1.655557427148254,
+      "lm_loss": 0.5455,
+      "step": 8600,
+      "vm_loss": 0.1664
+    },
+    {
+      "epoch": 1.6557499338258297,
+      "grad_norm": 3.292446592359703,
+      "learning_rate": 1.5124403770210972e-06,
+      "loss": 0.9009,
+      "step": 8601
+    },
+    {
+      "epoch": 1.6559424405034049,
+      "grad_norm": 3.3249381820495953,
+      "learning_rate": 1.5107920933889976e-06,
+      "loss": 0.9146,
+      "step": 8602
+    },
+    {
+      "epoch": 1.6561349471809803,
+      "grad_norm": 3.351641227618159,
+      "learning_rate": 1.5091446350165173e-06,
+      "loss": 0.865,
+      "step": 8603
+    },
+    {
+      "epoch": 1.6563274538585557,
+      "grad_norm": 3.1480824762923043,
+      "learning_rate": 1.507498002063803e-06,
+      "loss": 0.8724,
+      "step": 8604
+    },
+    {
+      "epoch": 1.656519960536131,
+      "grad_norm": 3.3068635263293973,
+      "learning_rate": 1.5058521946909399e-06,
+      "loss": 0.9284,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6567124672137066,
+      "grad_norm": 3.1165652984161345,
+      "learning_rate": 1.5042072130579122e-06,
+      "loss": 0.8535,
+      "step": 8606
+    },
+    {
+      "epoch": 1.6569049738912818,
+      "grad_norm": 3.1629556822327647,
+      "learning_rate": 1.5025630573246363e-06,
+      "loss": 0.8807,
+      "step": 8607
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "grad_norm": 3.1481275223575964,
+      "learning_rate": 1.5009197276509446e-06,
+      "loss": 0.8571,
+      "step": 8608
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.9178,
+      "step": 8608,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.7153,
+      "step": 8608,
+      "vm_loss": 0.1521
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.7857,
+      "step": 8608,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.5521,
+      "step": 8608,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.6812,
+      "step": 8608,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.7867,
+      "step": 8608,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.6027,
+      "step": 8608,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.6570974805688572,
+      "lm_loss": 0.9518,
+      "step": 8608,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 1.6572899872464326,
+      "grad_norm": 3.273485719571236,
+      "learning_rate": 1.4992772241965935e-06,
+      "loss": 0.873,
+      "step": 8609
+    },
+    {
+      "epoch": 1.657482493924008,
+      "grad_norm": 3.2619107685133697,
+      "learning_rate": 1.49763554712125e-06,
+      "loss": 0.8909,
+      "step": 8610
+    },
+    {
+      "epoch": 1.6576750006015835,
+      "grad_norm": 3.0510990190415352,
+      "learning_rate": 1.4959946965845084e-06,
+      "loss": 0.7988,
+      "step": 8611
+    },
+    {
+      "epoch": 1.6578675072791587,
+      "grad_norm": 3.3338252414019425,
+      "learning_rate": 1.4943546727458802e-06,
+      "loss": 0.9106,
+      "step": 8612
+    },
+    {
+      "epoch": 1.658060013956734,
+      "grad_norm": 3.480807468628937,
+      "learning_rate": 1.4927154757647967e-06,
+      "loss": 0.942,
+      "step": 8613
+    },
+    {
+      "epoch": 1.6582525206343095,
+      "grad_norm": 3.2978699376414076,
+      "learning_rate": 1.491077105800608e-06,
+      "loss": 0.9154,
+      "step": 8614
+    },
+    {
+      "epoch": 1.658445027311885,
+      "grad_norm": 3.2960106036443557,
+      "learning_rate": 1.4894395630125868e-06,
+      "loss": 0.8866,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "grad_norm": 3.165725497970785,
+      "learning_rate": 1.4878028475599237e-06,
+      "loss": 0.8958,
+      "step": 8616
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.7562,
+      "step": 8616,
+      "vm_loss": 0.2352
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.9552,
+      "step": 8616,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.8778,
+      "step": 8616,
+      "vm_loss": 0.213
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.5895,
+      "step": 8616,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.7849,
+      "step": 8616,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.6183,
+      "step": 8616,
+      "vm_loss": 0.0964
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.5651,
+      "step": 8616,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.6586375339894603,
+      "lm_loss": 0.7164,
+      "step": 8616,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 1.6588300406670355,
+      "grad_norm": 3.448421431615067,
+      "learning_rate": 1.4861669596017237e-06,
+      "loss": 0.9136,
+      "step": 8617
+    },
+    {
+      "epoch": 1.659022547344611,
+      "grad_norm": 3.1740875942782276,
+      "learning_rate": 1.4845318992970203e-06,
+      "loss": 0.86,
+      "step": 8618
+    },
+    {
+      "epoch": 1.6592150540221864,
+      "grad_norm": 3.3788893341495903,
+      "learning_rate": 1.48289766680476e-06,
+      "loss": 0.9262,
+      "step": 8619
+    },
+    {
+      "epoch": 1.6594075606997618,
+      "grad_norm": 3.194499805785924,
+      "learning_rate": 1.4812642622838148e-06,
+      "loss": 0.8662,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6596000673773372,
+      "grad_norm": 3.305860907010635,
+      "learning_rate": 1.4796316858929648e-06,
+      "loss": 0.8752,
+      "step": 8621
+    },
+    {
+      "epoch": 1.6597925740549124,
+      "grad_norm": 3.201176208449638,
+      "learning_rate": 1.4779999377909271e-06,
+      "loss": 0.8643,
+      "step": 8622
+    },
+    {
+      "epoch": 1.659985080732488,
+      "grad_norm": 3.316934044031229,
+      "learning_rate": 1.476369018136321e-06,
+      "loss": 0.9067,
+      "step": 8623
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "grad_norm": 3.4796896556572374,
+      "learning_rate": 1.4747389270876977e-06,
+      "loss": 0.9132,
+      "step": 8624
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 1.0447,
+      "step": 8624,
+      "vm_loss": 0.2471
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.7276,
+      "step": 8624,
+      "vm_loss": 0.196
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.6373,
+      "step": 8624,
+      "vm_loss": 0.2002
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.4689,
+      "step": 8624,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.5324,
+      "step": 8624,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.7878,
+      "step": 8624,
+      "vm_loss": 0.2042
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 1.1226,
+      "step": 8624,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.6601775874100633,
+      "lm_loss": 0.9283,
+      "step": 8624,
+      "vm_loss": 0.1915
+    },
+    {
+      "epoch": 1.6603700940876387,
+      "grad_norm": 3.2580092370535447,
+      "learning_rate": 1.4731096648035158e-06,
+      "loss": 0.9013,
+      "step": 8625
+    },
+    {
+      "epoch": 1.6605626007652141,
+      "grad_norm": 3.1987584969560903,
+      "learning_rate": 1.4714812314421711e-06,
+      "loss": 0.8889,
+      "step": 8626
+    },
+    {
+      "epoch": 1.6607551074427893,
+      "grad_norm": 3.378742204547144,
+      "learning_rate": 1.469853627161959e-06,
+      "loss": 0.9387,
+      "step": 8627
+    },
+    {
+      "epoch": 1.660947614120365,
+      "grad_norm": 3.2596257329743303,
+      "learning_rate": 1.4682268521211075e-06,
+      "loss": 0.9044,
+      "step": 8628
+    },
+    {
+      "epoch": 1.6611401207979402,
+      "grad_norm": 3.1555085556971814,
+      "learning_rate": 1.4666009064777587e-06,
+      "loss": 0.8573,
+      "step": 8629
+    },
+    {
+      "epoch": 1.6613326274755156,
+      "grad_norm": 3.11067129782355,
+      "learning_rate": 1.4649757903899752e-06,
+      "loss": 0.8309,
+      "step": 8630
+    },
+    {
+      "epoch": 1.661525134153091,
+      "grad_norm": 3.215776551744369,
+      "learning_rate": 1.463351504015741e-06,
+      "loss": 0.864,
+      "step": 8631
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "grad_norm": 3.1944682161970346,
+      "learning_rate": 1.461728047512958e-06,
+      "loss": 0.8349,
+      "step": 8632
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.8908,
+      "step": 8632,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.6637,
+      "step": 8632,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.4599,
+      "step": 8632,
+      "vm_loss": 0.117
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 1.2626,
+      "step": 8632,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.6261,
+      "step": 8632,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.719,
+      "step": 8632,
+      "vm_loss": 0.1769
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.708,
+      "step": 8632,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.6617176408306662,
+      "lm_loss": 0.6153,
+      "step": 8632,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 1.6619101475082418,
+      "grad_norm": 3.0728873791224682,
+      "learning_rate": 1.4601054210394428e-06,
+      "loss": 0.809,
+      "step": 8633
+    },
+    {
+      "epoch": 1.662102654185817,
+      "grad_norm": 3.2411696687244813,
+      "learning_rate": 1.4584836247529377e-06,
+      "loss": 0.8674,
+      "step": 8634
+    },
+    {
+      "epoch": 1.6622951608633925,
+      "grad_norm": 3.1629452086313363,
+      "learning_rate": 1.4568626588111023e-06,
+      "loss": 0.8242,
+      "step": 8635
+    },
+    {
+      "epoch": 1.6624876675409679,
+      "grad_norm": 3.26054925572343,
+      "learning_rate": 1.455242523371515e-06,
+      "loss": 0.8585,
+      "step": 8636
+    },
+    {
+      "epoch": 1.662680174218543,
+      "grad_norm": 3.275617540411821,
+      "learning_rate": 1.4536232185916765e-06,
+      "loss": 0.9255,
+      "step": 8637
+    },
+    {
+      "epoch": 1.6628726808961187,
+      "grad_norm": 3.1331767130555166,
+      "learning_rate": 1.4520047446289964e-06,
+      "loss": 0.8042,
+      "step": 8638
+    },
+    {
+      "epoch": 1.663065187573694,
+      "grad_norm": 3.1747286387788893,
+      "learning_rate": 1.45038710164082e-06,
+      "loss": 0.8556,
+      "step": 8639
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "grad_norm": 3.0959212464276904,
+      "learning_rate": 1.4487702897843969e-06,
+      "loss": 0.8385,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.6467,
+      "step": 8640,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.4168,
+      "step": 8640,
+      "vm_loss": 0.0948
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.8074,
+      "step": 8640,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.5727,
+      "step": 8640,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.9478,
+      "step": 8640,
+      "vm_loss": 0.1192
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.4803,
+      "step": 8640,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 0.6475,
+      "step": 8640,
+      "vm_loss": 0.1355
+    },
+    {
+      "epoch": 1.6632576942512693,
+      "lm_loss": 1.1566,
+      "step": 8640,
+      "vm_loss": 0.1507
+    },
+    {
+      "epoch": 1.6634502009288448,
+      "grad_norm": 3.230317078988885,
+      "learning_rate": 1.4471543092169072e-06,
+      "loss": 0.8464,
+      "step": 8641
+    },
+    {
+      "epoch": 1.66364270760642,
+      "grad_norm": 3.327929402150926,
+      "learning_rate": 1.445539160095436e-06,
+      "loss": 0.8797,
+      "step": 8642
+    },
+    {
+      "epoch": 1.6638352142839956,
+      "grad_norm": 3.273601998759998,
+      "learning_rate": 1.4439248425770058e-06,
+      "loss": 0.8864,
+      "step": 8643
+    },
+    {
+      "epoch": 1.6640277209615708,
+      "grad_norm": 3.220464535294617,
+      "learning_rate": 1.4423113568185433e-06,
+      "loss": 0.8645,
+      "step": 8644
+    },
+    {
+      "epoch": 1.6642202276391462,
+      "grad_norm": 3.2937547286579676,
+      "learning_rate": 1.4406987029769026e-06,
+      "loss": 0.8529,
+      "step": 8645
+    },
+    {
+      "epoch": 1.6644127343167217,
+      "grad_norm": 3.157043837647187,
+      "learning_rate": 1.4390868812088521e-06,
+      "loss": 0.8293,
+      "step": 8646
+    },
+    {
+      "epoch": 1.6646052409942969,
+      "grad_norm": 3.2342571082236486,
+      "learning_rate": 1.4374758916710863e-06,
+      "loss": 0.8891,
+      "step": 8647
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "grad_norm": 3.2269554137077585,
+      "learning_rate": 1.4358657345202043e-06,
+      "loss": 0.8856,
+      "step": 8648
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 1.0571,
+      "step": 8648,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.6571,
+      "step": 8648,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 1.1517,
+      "step": 8648,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.7242,
+      "step": 8648,
+      "vm_loss": 0.1091
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.9443,
+      "step": 8648,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.6744,
+      "step": 8648,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.522,
+      "step": 8648,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 1.6647977476718725,
+      "lm_loss": 0.7312,
+      "step": 8648,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.6649902543494477,
+      "grad_norm": 3.2064001453693742,
+      "learning_rate": 1.4342564099127465e-06,
+      "loss": 0.8896,
+      "step": 8649
+    },
+    {
+      "epoch": 1.6651827610270231,
+      "grad_norm": 3.1865968676083605,
+      "learning_rate": 1.43264791800515e-06,
+      "loss": 0.8844,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6653752677045985,
+      "grad_norm": 3.4245156551821254,
+      "learning_rate": 1.4310402589537852e-06,
+      "loss": 0.9231,
+      "step": 8651
+    },
+    {
+      "epoch": 1.6655677743821737,
+      "grad_norm": 3.286437050363198,
+      "learning_rate": 1.429433432914935e-06,
+      "loss": 0.8677,
+      "step": 8652
+    },
+    {
+      "epoch": 1.6657602810597494,
+      "grad_norm": 3.2948250752241552,
+      "learning_rate": 1.4278274400448056e-06,
+      "loss": 0.8979,
+      "step": 8653
+    },
+    {
+      "epoch": 1.6659527877373246,
+      "grad_norm": 3.3176248888995996,
+      "learning_rate": 1.4262222804995217e-06,
+      "loss": 0.9127,
+      "step": 8654
+    },
+    {
+      "epoch": 1.6661452944149,
+      "grad_norm": 3.396234409003037,
+      "learning_rate": 1.4246179544351202e-06,
+      "loss": 0.9161,
+      "step": 8655
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "grad_norm": 3.2404596217464636,
+      "learning_rate": 1.4230144620075648e-06,
+      "loss": 0.878,
+      "step": 8656
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.3272,
+      "step": 8656,
+      "vm_loss": 0.2388
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.9471,
+      "step": 8656,
+      "vm_loss": 0.2977
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.9748,
+      "step": 8656,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.9703,
+      "step": 8656,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.3722,
+      "step": 8656,
+      "vm_loss": 0.1466
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.764,
+      "step": 8656,
+      "vm_loss": 0.0933
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.7902,
+      "step": 8656,
+      "vm_loss": 0.1104
+    },
+    {
+      "epoch": 1.6663378010924754,
+      "lm_loss": 0.9039,
+      "step": 8656,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.6665303077700506,
+      "grad_norm": 3.312944757013938,
+      "learning_rate": 1.4214118033727343e-06,
+      "loss": 0.9283,
+      "step": 8657
+    },
+    {
+      "epoch": 1.6667228144476263,
+      "grad_norm": 3.23270419046511,
+      "learning_rate": 1.4198099786864327e-06,
+      "loss": 0.8796,
+      "step": 8658
+    },
+    {
+      "epoch": 1.6669153211252015,
+      "grad_norm": 3.1366810466161796,
+      "learning_rate": 1.418208988104368e-06,
+      "loss": 0.862,
+      "step": 8659
+    },
+    {
+      "epoch": 1.6671078278027769,
+      "grad_norm": 3.3520670955629015,
+      "learning_rate": 1.416608831782188e-06,
+      "loss": 0.8655,
+      "step": 8660
+    },
+    {
+      "epoch": 1.6673003344803523,
+      "grad_norm": 3.3538257827072964,
+      "learning_rate": 1.4150095098754402e-06,
+      "loss": 0.903,
+      "step": 8661
+    },
+    {
+      "epoch": 1.6674928411579275,
+      "grad_norm": 3.2615696767489424,
+      "learning_rate": 1.4134110225396024e-06,
+      "loss": 0.8995,
+      "step": 8662
+    },
+    {
+      "epoch": 1.6676853478355032,
+      "grad_norm": 3.263676372790851,
+      "learning_rate": 1.4118133699300685e-06,
+      "loss": 0.8819,
+      "step": 8663
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "grad_norm": 3.1008050641425493,
+      "learning_rate": 1.4102165522021516e-06,
+      "loss": 0.842,
+      "step": 8664
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.5606,
+      "step": 8664,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.7015,
+      "step": 8664,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.723,
+      "step": 8664,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.914,
+      "step": 8664,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.8645,
+      "step": 8664,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.6933,
+      "step": 8664,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.8451,
+      "step": 8664,
+      "vm_loss": 0.2329
+    },
+    {
+      "epoch": 1.6678778545130784,
+      "lm_loss": 0.6587,
+      "step": 8664,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.6680703611906538,
+      "grad_norm": 3.209617531633429,
+      "learning_rate": 1.4086205695110778e-06,
+      "loss": 0.8993,
+      "step": 8665
+    },
+    {
+      "epoch": 1.6682628678682292,
+      "grad_norm": 3.0392467582317244,
+      "learning_rate": 1.4070254220120039e-06,
+      "loss": 0.8147,
+      "step": 8666
+    },
+    {
+      "epoch": 1.6684553745458044,
+      "grad_norm": 3.2463083107447215,
+      "learning_rate": 1.4054311098599926e-06,
+      "loss": 0.8523,
+      "step": 8667
+    },
+    {
+      "epoch": 1.66864788122338,
+      "grad_norm": 3.1553068906414823,
+      "learning_rate": 1.403837633210038e-06,
+      "loss": 0.875,
+      "step": 8668
+    },
+    {
+      "epoch": 1.6688403879009552,
+      "grad_norm": 3.2210873315707955,
+      "learning_rate": 1.4022449922170368e-06,
+      "loss": 0.8686,
+      "step": 8669
+    },
+    {
+      "epoch": 1.6690328945785307,
+      "grad_norm": 3.106918134249493,
+      "learning_rate": 1.400653187035824e-06,
+      "loss": 0.858,
+      "step": 8670
+    },
+    {
+      "epoch": 1.669225401256106,
+      "grad_norm": 3.2638744093433254,
+      "learning_rate": 1.3990622178211365e-06,
+      "loss": 0.8831,
+      "step": 8671
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "grad_norm": 3.259449104393686,
+      "learning_rate": 1.3974720847276412e-06,
+      "loss": 0.9039,
+      "step": 8672
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.6322,
+      "step": 8672,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.9701,
+      "step": 8672,
+      "vm_loss": 0.1308
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.7481,
+      "step": 8672,
+      "vm_loss": 0.1516
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.7956,
+      "step": 8672,
+      "vm_loss": 0.1116
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.9709,
+      "step": 8672,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.693,
+      "step": 8672,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.445,
+      "step": 8672,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 1.6694179079336815,
+      "lm_loss": 0.4672,
+      "step": 8672,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.669610414611257,
+      "grad_norm": 3.378183535267053,
+      "learning_rate": 1.3958827879099169e-06,
+      "loss": 0.8803,
+      "step": 8673
+    },
+    {
+      "epoch": 1.6698029212888321,
+      "grad_norm": 3.3190902676777747,
+      "learning_rate": 1.3942943275224641e-06,
+      "loss": 0.8641,
+      "step": 8674
+    },
+    {
+      "epoch": 1.6699954279664075,
+      "grad_norm": 3.1162925071191263,
+      "learning_rate": 1.3927067037197039e-06,
+      "loss": 0.8261,
+      "step": 8675
+    },
+    {
+      "epoch": 1.670187934643983,
+      "grad_norm": 3.0538030158057325,
+      "learning_rate": 1.3911199166559675e-06,
+      "loss": 0.8218,
+      "step": 8676
+    },
+    {
+      "epoch": 1.6703804413215584,
+      "grad_norm": 3.415094721865597,
+      "learning_rate": 1.38953396648552e-06,
+      "loss": 0.9105,
+      "step": 8677
+    },
+    {
+      "epoch": 1.6705729479991338,
+      "grad_norm": 3.1574624157504396,
+      "learning_rate": 1.3879488533625286e-06,
+      "loss": 0.8886,
+      "step": 8678
+    },
+    {
+      "epoch": 1.670765454676709,
+      "grad_norm": 3.181947112819833,
+      "learning_rate": 1.3863645774410883e-06,
+      "loss": 0.8558,
+      "step": 8679
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "grad_norm": 3.420222345542465,
+      "learning_rate": 1.3847811388752119e-06,
+      "loss": 0.9115,
+      "step": 8680
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.5386,
+      "step": 8680,
+      "vm_loss": 0.2117
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 1.0263,
+      "step": 8680,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.7111,
+      "step": 8680,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.5011,
+      "step": 8680,
+      "vm_loss": 0.1045
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.6043,
+      "step": 8680,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.6528,
+      "step": 8680,
+      "vm_loss": 0.1638
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.8272,
+      "step": 8680,
+      "vm_loss": 0.1432
+    },
+    {
+      "epoch": 1.6709579613542844,
+      "lm_loss": 0.7701,
+      "step": 8680,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.6711504680318598,
+      "grad_norm": 3.300307039919615,
+      "learning_rate": 1.3831985378188317e-06,
+      "loss": 0.8677,
+      "step": 8681
+    },
+    {
+      "epoch": 1.6713429747094353,
+      "grad_norm": 3.2744881440019467,
+      "learning_rate": 1.3816167744257902e-06,
+      "loss": 0.8441,
+      "step": 8682
+    },
+    {
+      "epoch": 1.6715354813870107,
+      "grad_norm": 3.4880328278006796,
+      "learning_rate": 1.3800358488498645e-06,
+      "loss": 0.9261,
+      "step": 8683
+    },
+    {
+      "epoch": 1.671727988064586,
+      "grad_norm": 3.2779962678486436,
+      "learning_rate": 1.3784557612447325e-06,
+      "loss": 0.9037,
+      "step": 8684
+    },
+    {
+      "epoch": 1.6719204947421615,
+      "grad_norm": 3.195543899756157,
+      "learning_rate": 1.3768765117640059e-06,
+      "loss": 0.8676,
+      "step": 8685
+    },
+    {
+      "epoch": 1.6721130014197367,
+      "grad_norm": 3.19753293842354,
+      "learning_rate": 1.3752981005611988e-06,
+      "loss": 0.8521,
+      "step": 8686
+    },
+    {
+      "epoch": 1.6723055080973122,
+      "grad_norm": 3.304726085017399,
+      "learning_rate": 1.3737205277897637e-06,
+      "loss": 0.9105,
+      "step": 8687
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "grad_norm": 3.2585186983972743,
+      "learning_rate": 1.3721437936030524e-06,
+      "loss": 0.8776,
+      "step": 8688
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.597,
+      "step": 8688,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.7772,
+      "step": 8688,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 1.1408,
+      "step": 8688,
+      "vm_loss": 0.0958
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.5652,
+      "step": 8688,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.7737,
+      "step": 8688,
+      "vm_loss": 0.1081
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.4744,
+      "step": 8688,
+      "vm_loss": 0.0889
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.9416,
+      "step": 8688,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.6724980147748876,
+      "lm_loss": 0.551,
+      "step": 8688,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.6726905214524628,
+      "grad_norm": 3.0773575755433606,
+      "learning_rate": 1.370567898154348e-06,
+      "loss": 0.7906,
+      "step": 8689
+    },
+    {
+      "epoch": 1.6728830281300384,
+      "grad_norm": 3.211895231477191,
+      "learning_rate": 1.3689928415968456e-06,
+      "loss": 0.872,
+      "step": 8690
+    },
+    {
+      "epoch": 1.6730755348076136,
+      "grad_norm": 3.2176081339314586,
+      "learning_rate": 1.367418624083665e-06,
+      "loss": 0.8938,
+      "step": 8691
+    },
+    {
+      "epoch": 1.673268041485189,
+      "grad_norm": 3.3111613825138226,
+      "learning_rate": 1.365845245767835e-06,
+      "loss": 0.9052,
+      "step": 8692
+    },
+    {
+      "epoch": 1.6734605481627645,
+      "grad_norm": 3.1365353895482104,
+      "learning_rate": 1.3642727068023087e-06,
+      "loss": 0.8561,
+      "step": 8693
+    },
+    {
+      "epoch": 1.6736530548403397,
+      "grad_norm": 3.372813772824032,
+      "learning_rate": 1.3627010073399605e-06,
+      "loss": 0.8886,
+      "step": 8694
+    },
+    {
+      "epoch": 1.6738455615179153,
+      "grad_norm": 3.4618630067403378,
+      "learning_rate": 1.3611301475335759e-06,
+      "loss": 0.9459,
+      "step": 8695
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "grad_norm": 3.1775714126265355,
+      "learning_rate": 1.359560127535865e-06,
+      "loss": 0.8436,
+      "step": 8696
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.6031,
+      "step": 8696,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.6202,
+      "step": 8696,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.6653,
+      "step": 8696,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.6178,
+      "step": 8696,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.4381,
+      "step": 8696,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.768,
+      "step": 8696,
+      "vm_loss": 0.1153
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.6019,
+      "step": 8696,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.6740380681954905,
+      "lm_loss": 0.878,
+      "step": 8696,
+      "vm_loss": 0.154
+    },
+    {
+      "epoch": 1.674230574873066,
+      "grad_norm": 3.272170119210459,
+      "learning_rate": 1.3579909474994534e-06,
+      "loss": 0.8851,
+      "step": 8697
+    },
+    {
+      "epoch": 1.6744230815506413,
+      "grad_norm": 3.4114111981751036,
+      "learning_rate": 1.3564226075768883e-06,
+      "loss": 0.9285,
+      "step": 8698
+    },
+    {
+      "epoch": 1.6746155882282165,
+      "grad_norm": 3.337898646747744,
+      "learning_rate": 1.3548551079206262e-06,
+      "loss": 0.8961,
+      "step": 8699
+    },
+    {
+      "epoch": 1.6748080949057922,
+      "grad_norm": 3.145509714349786,
+      "learning_rate": 1.353288448683051e-06,
+      "loss": 0.8561,
+      "step": 8700
+    },
+    {
+      "epoch": 1.6750006015833674,
+      "grad_norm": 3.185125432679613,
+      "learning_rate": 1.3517226300164632e-06,
+      "loss": 0.8621,
+      "step": 8701
+    },
+    {
+      "epoch": 1.6751931082609428,
+      "grad_norm": 3.2291425655737567,
+      "learning_rate": 1.3501576520730809e-06,
+      "loss": 0.884,
+      "step": 8702
+    },
+    {
+      "epoch": 1.6753856149385182,
+      "grad_norm": 3.2291281177363027,
+      "learning_rate": 1.3485935150050345e-06,
+      "loss": 0.8311,
+      "step": 8703
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "grad_norm": 3.226636224996044,
+      "learning_rate": 1.347030218964387e-06,
+      "loss": 0.8528,
+      "step": 8704
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.7582,
+      "step": 8704,
+      "vm_loss": 0.2331
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 1.0301,
+      "step": 8704,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.5202,
+      "step": 8704,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.6198,
+      "step": 8704,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.9205,
+      "step": 8704,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.4732,
+      "step": 8704,
+      "vm_loss": 0.2216
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.529,
+      "step": 8704,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 1.6755781216160934,
+      "lm_loss": 0.4478,
+      "step": 8704,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 1.675770628293669,
+      "grad_norm": 3.1758902264920454,
+      "learning_rate": 1.3454677641031032e-06,
+      "loss": 0.864,
+      "step": 8705
+    },
+    {
+      "epoch": 1.6759631349712443,
+      "grad_norm": 3.307620336094291,
+      "learning_rate": 1.3439061505730777e-06,
+      "loss": 0.9329,
+      "step": 8706
+    },
+    {
+      "epoch": 1.6761556416488197,
+      "grad_norm": 3.28283674350169,
+      "learning_rate": 1.3423453785261164e-06,
+      "loss": 0.8889,
+      "step": 8707
+    },
+    {
+      "epoch": 1.6763481483263951,
+      "grad_norm": 3.2444700793163452,
+      "learning_rate": 1.3407854481139514e-06,
+      "loss": 0.8541,
+      "step": 8708
+    },
+    {
+      "epoch": 1.6765406550039703,
+      "grad_norm": 3.4211372911845004,
+      "learning_rate": 1.3392263594882226e-06,
+      "loss": 0.89,
+      "step": 8709
+    },
+    {
+      "epoch": 1.676733161681546,
+      "grad_norm": 3.4119764482832937,
+      "learning_rate": 1.3376681128004943e-06,
+      "loss": 0.9349,
+      "step": 8710
+    },
+    {
+      "epoch": 1.6769256683591212,
+      "grad_norm": 3.296501674875594,
+      "learning_rate": 1.3361107082022495e-06,
+      "loss": 0.9131,
+      "step": 8711
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "grad_norm": 3.20740315599562,
+      "learning_rate": 1.3345541458448896e-06,
+      "loss": 0.882,
+      "step": 8712
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.7727,
+      "step": 8712,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.8545,
+      "step": 8712,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.364,
+      "step": 8712,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.7373,
+      "step": 8712,
+      "vm_loss": 0.1492
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 1.3608,
+      "step": 8712,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.8105,
+      "step": 8712,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 0.6354,
+      "step": 8712,
+      "vm_loss": 0.1835
+    },
+    {
+      "epoch": 1.6771181750366966,
+      "lm_loss": 1.1439,
+      "step": 8712,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 1.677310681714272,
+      "grad_norm": 3.2928954222811258,
+      "learning_rate": 1.3329984258797258e-06,
+      "loss": 0.8884,
+      "step": 8713
+    },
+    {
+      "epoch": 1.6775031883918472,
+      "grad_norm": 3.0144152690407395,
+      "learning_rate": 1.3314435484580035e-06,
+      "loss": 0.7848,
+      "step": 8714
+    },
+    {
+      "epoch": 1.6776956950694228,
+      "grad_norm": 3.3837874589624732,
+      "learning_rate": 1.3298895137308687e-06,
+      "loss": 0.9118,
+      "step": 8715
+    },
+    {
+      "epoch": 1.677888201746998,
+      "grad_norm": 3.406987466021496,
+      "learning_rate": 1.3283363218493962e-06,
+      "loss": 0.9177,
+      "step": 8716
+    },
+    {
+      "epoch": 1.6780807084245735,
+      "grad_norm": 3.1088730676335286,
+      "learning_rate": 1.326783972964576e-06,
+      "loss": 0.8484,
+      "step": 8717
+    },
+    {
+      "epoch": 1.6782732151021489,
+      "grad_norm": 3.396146071827399,
+      "learning_rate": 1.3252324672273165e-06,
+      "loss": 0.9127,
+      "step": 8718
+    },
+    {
+      "epoch": 1.678465721779724,
+      "grad_norm": 3.225824077691006,
+      "learning_rate": 1.3236818047884458e-06,
+      "loss": 0.8846,
+      "step": 8719
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "grad_norm": 3.2206376541056025,
+      "learning_rate": 1.3221319857987024e-06,
+      "loss": 0.8779,
+      "step": 8720
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.9861,
+      "step": 8720,
+      "vm_loss": 0.2481
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.6711,
+      "step": 8720,
+      "vm_loss": 0.1123
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.8884,
+      "step": 8720,
+      "vm_loss": 0.1887
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.4481,
+      "step": 8720,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.6428,
+      "step": 8720,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.5598,
+      "step": 8720,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.3349,
+      "step": 8720,
+      "vm_loss": 0.1207
+    },
+    {
+      "epoch": 1.6786582284572997,
+      "lm_loss": 0.3278,
+      "step": 8720,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 1.678850735134875,
+      "grad_norm": 3.3206084775316884,
+      "learning_rate": 1.3205830104087558e-06,
+      "loss": 0.8839,
+      "step": 8721
+    },
+    {
+      "epoch": 1.6790432418124503,
+      "grad_norm": 3.409503190594079,
+      "learning_rate": 1.319034878769181e-06,
+      "loss": 0.8746,
+      "step": 8722
+    },
+    {
+      "epoch": 1.6792357484900258,
+      "grad_norm": 3.1442789119577252,
+      "learning_rate": 1.3174875910304784e-06,
+      "loss": 0.8429,
+      "step": 8723
+    },
+    {
+      "epoch": 1.679428255167601,
+      "grad_norm": 3.380796198679316,
+      "learning_rate": 1.3159411473430638e-06,
+      "loss": 0.9078,
+      "step": 8724
+    },
+    {
+      "epoch": 1.6796207618451766,
+      "grad_norm": 3.2554451948033876,
+      "learning_rate": 1.3143955478572735e-06,
+      "loss": 0.8668,
+      "step": 8725
+    },
+    {
+      "epoch": 1.6798132685227518,
+      "grad_norm": 3.237774080272247,
+      "learning_rate": 1.3128507927233536e-06,
+      "loss": 0.8673,
+      "step": 8726
+    },
+    {
+      "epoch": 1.6800057752003272,
+      "grad_norm": 3.244262203836014,
+      "learning_rate": 1.3113068820914798e-06,
+      "loss": 0.8683,
+      "step": 8727
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "grad_norm": 3.1234860378205345,
+      "learning_rate": 1.3097638161117365e-06,
+      "loss": 0.8544,
+      "step": 8728
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.7836,
+      "step": 8728,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 1.1444,
+      "step": 8728,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.7403,
+      "step": 8728,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.6521,
+      "step": 8728,
+      "vm_loss": 0.2508
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.9282,
+      "step": 8728,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.4113,
+      "step": 8728,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.8431,
+      "step": 8728,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 1.6801982818779027,
+      "lm_loss": 0.8908,
+      "step": 8728,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.6803907885554779,
+      "grad_norm": 3.4425579834010627,
+      "learning_rate": 1.3082215949341338e-06,
+      "loss": 0.9077,
+      "step": 8729
+    },
+    {
+      "epoch": 1.6805832952330535,
+      "grad_norm": 3.2688507802303937,
+      "learning_rate": 1.306680218708587e-06,
+      "loss": 0.8422,
+      "step": 8730
+    },
+    {
+      "epoch": 1.6807758019106287,
+      "grad_norm": 3.3124274672826934,
+      "learning_rate": 1.3051396875849488e-06,
+      "loss": 0.8845,
+      "step": 8731
+    },
+    {
+      "epoch": 1.6809683085882041,
+      "grad_norm": 3.3166698275243136,
+      "learning_rate": 1.303600001712969e-06,
+      "loss": 0.8506,
+      "step": 8732
+    },
+    {
+      "epoch": 1.6811608152657795,
+      "grad_norm": 3.1674969990840522,
+      "learning_rate": 1.3020611612423296e-06,
+      "loss": 0.8312,
+      "step": 8733
+    },
+    {
+      "epoch": 1.681353321943355,
+      "grad_norm": 3.195604225373869,
+      "learning_rate": 1.3005231663226237e-06,
+      "loss": 0.8694,
+      "step": 8734
+    },
+    {
+      "epoch": 1.6815458286209304,
+      "grad_norm": 3.245661432867938,
+      "learning_rate": 1.2989860171033674e-06,
+      "loss": 0.851,
+      "step": 8735
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "grad_norm": 3.224309719463014,
+      "learning_rate": 1.297449713733987e-06,
+      "loss": 0.8626,
+      "step": 8736
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.5702,
+      "step": 8736,
+      "vm_loss": 0.2285
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.394,
+      "step": 8736,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.9997,
+      "step": 8736,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.604,
+      "step": 8736,
+      "vm_loss": 0.2767
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.9586,
+      "step": 8736,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.4027,
+      "step": 8736,
+      "vm_loss": 0.1058
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.7967,
+      "step": 8736,
+      "vm_loss": 0.2396
+    },
+    {
+      "epoch": 1.6817383352985056,
+      "lm_loss": 0.9387,
+      "step": 8736,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.681930841976081,
+      "grad_norm": 3.3513537736415233,
+      "learning_rate": 1.2959142563638316e-06,
+      "loss": 0.9202,
+      "step": 8737
+    },
+    {
+      "epoch": 1.6821233486536564,
+      "grad_norm": 3.2121023934661728,
+      "learning_rate": 1.2943796451421686e-06,
+      "loss": 0.8521,
+      "step": 8738
+    },
+    {
+      "epoch": 1.6823158553312318,
+      "grad_norm": 3.1590992924814345,
+      "learning_rate": 1.292845880218182e-06,
+      "loss": 0.842,
+      "step": 8739
+    },
+    {
+      "epoch": 1.6825083620088073,
+      "grad_norm": 2.984186350413497,
+      "learning_rate": 1.2913129617409725e-06,
+      "loss": 0.8097,
+      "step": 8740
+    },
+    {
+      "epoch": 1.6827008686863825,
+      "grad_norm": 3.3496709157782623,
+      "learning_rate": 1.2897808898595599e-06,
+      "loss": 0.9128,
+      "step": 8741
+    },
+    {
+      "epoch": 1.6828933753639579,
+      "grad_norm": 3.2611090151595215,
+      "learning_rate": 1.2882496647228827e-06,
+      "loss": 0.8897,
+      "step": 8742
+    },
+    {
+      "epoch": 1.6830858820415333,
+      "grad_norm": 3.0652524913331263,
+      "learning_rate": 1.2867192864797929e-06,
+      "loss": 0.8291,
+      "step": 8743
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "grad_norm": 3.3115779452351535,
+      "learning_rate": 1.2851897552790626e-06,
+      "loss": 0.8821,
+      "step": 8744
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.5846,
+      "step": 8744,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.8508,
+      "step": 8744,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.6689,
+      "step": 8744,
+      "vm_loss": 0.1158
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 1.1186,
+      "step": 8744,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.5572,
+      "step": 8744,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 1.2138,
+      "step": 8744,
+      "vm_loss": 0.2797
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.5603,
+      "step": 8744,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.6832783887191087,
+      "lm_loss": 0.7624,
+      "step": 8744,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 1.6834708953966842,
+      "grad_norm": 3.386865354669033,
+      "learning_rate": 1.2836610712693832e-06,
+      "loss": 0.9345,
+      "step": 8745
+    },
+    {
+      "epoch": 1.6836634020742594,
+      "grad_norm": 3.4846650447672722,
+      "learning_rate": 1.2821332345993654e-06,
+      "loss": 0.8977,
+      "step": 8746
+    },
+    {
+      "epoch": 1.683855908751835,
+      "grad_norm": 3.2421638668670525,
+      "learning_rate": 1.280606245417526e-06,
+      "loss": 0.8568,
+      "step": 8747
+    },
+    {
+      "epoch": 1.6840484154294102,
+      "grad_norm": 3.404072778626749,
+      "learning_rate": 1.2790801038723178e-06,
+      "loss": 0.9046,
+      "step": 8748
+    },
+    {
+      "epoch": 1.6842409221069856,
+      "grad_norm": 3.3313672278231157,
+      "learning_rate": 1.2775548101120949e-06,
+      "loss": 0.8769,
+      "step": 8749
+    },
+    {
+      "epoch": 1.684433428784561,
+      "grad_norm": 3.272495690307263,
+      "learning_rate": 1.276030364285138e-06,
+      "loss": 0.8636,
+      "step": 8750
+    },
+    {
+      "epoch": 1.6846259354621362,
+      "grad_norm": 3.231440650854355,
+      "learning_rate": 1.2745067665396394e-06,
+      "loss": 0.8935,
+      "step": 8751
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "grad_norm": 3.2060360657411575,
+      "learning_rate": 1.2729840170237174e-06,
+      "loss": 0.844,
+      "step": 8752
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.5977,
+      "step": 8752,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.42,
+      "step": 8752,
+      "vm_loss": 0.1235
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.485,
+      "step": 8752,
+      "vm_loss": 0.1187
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.4342,
+      "step": 8752,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.6021,
+      "step": 8752,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.9447,
+      "step": 8752,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.6735,
+      "step": 8752,
+      "vm_loss": 0.2025
+    },
+    {
+      "epoch": 1.6848184421397119,
+      "lm_loss": 0.7368,
+      "step": 8752,
+      "vm_loss": 0.1675
+    },
+    {
+      "epoch": 1.685010948817287,
+      "grad_norm": 3.2971295275483072,
+      "learning_rate": 1.271462115885399e-06,
+      "loss": 0.8342,
+      "step": 8753
+    },
+    {
+      "epoch": 1.6852034554948625,
+      "grad_norm": 3.1585814923394135,
+      "learning_rate": 1.2699410632726327e-06,
+      "loss": 0.8511,
+      "step": 8754
+    },
+    {
+      "epoch": 1.685395962172438,
+      "grad_norm": 3.4391125454296194,
+      "learning_rate": 1.2684208593332858e-06,
+      "loss": 0.9064,
+      "step": 8755
+    },
+    {
+      "epoch": 1.6855884688500131,
+      "grad_norm": 3.315304705484362,
+      "learning_rate": 1.2669015042151422e-06,
+      "loss": 0.8752,
+      "step": 8756
+    },
+    {
+      "epoch": 1.6857809755275888,
+      "grad_norm": 3.2710178273322144,
+      "learning_rate": 1.2653829980658983e-06,
+      "loss": 0.8654,
+      "step": 8757
+    },
+    {
+      "epoch": 1.685973482205164,
+      "grad_norm": 3.236258565521527,
+      "learning_rate": 1.2638653410331781e-06,
+      "loss": 0.873,
+      "step": 8758
+    },
+    {
+      "epoch": 1.6861659888827394,
+      "grad_norm": 3.218784670151545,
+      "learning_rate": 1.2623485332645135e-06,
+      "loss": 0.8676,
+      "step": 8759
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "grad_norm": 3.189487807466586,
+      "learning_rate": 1.2608325749073591e-06,
+      "loss": 0.8534,
+      "step": 8760
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.5938,
+      "step": 8760,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.4621,
+      "step": 8760,
+      "vm_loss": 0.1905
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.4728,
+      "step": 8760,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.8739,
+      "step": 8760,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.9507,
+      "step": 8760,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.8914,
+      "step": 8760,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 0.4753,
+      "step": 8760,
+      "vm_loss": 0.1462
+    },
+    {
+      "epoch": 1.6863584955603148,
+      "lm_loss": 1.0074,
+      "step": 8760,
+      "vm_loss": 0.1402
+    },
+    {
+      "epoch": 1.68655100223789,
+      "grad_norm": 3.318343862451477,
+      "learning_rate": 1.2593174661090834e-06,
+      "loss": 0.8901,
+      "step": 8761
+    },
+    {
+      "epoch": 1.6867435089154657,
+      "grad_norm": 3.3305765415755295,
+      "learning_rate": 1.2578032070169777e-06,
+      "loss": 0.9042,
+      "step": 8762
+    },
+    {
+      "epoch": 1.6869360155930408,
+      "grad_norm": 3.2263822495582075,
+      "learning_rate": 1.2562897977782461e-06,
+      "loss": 0.8663,
+      "step": 8763
+    },
+    {
+      "epoch": 1.6871285222706163,
+      "grad_norm": 3.197332160763477,
+      "learning_rate": 1.2547772385400081e-06,
+      "loss": 0.8793,
+      "step": 8764
+    },
+    {
+      "epoch": 1.6873210289481917,
+      "grad_norm": 3.169701305239922,
+      "learning_rate": 1.25326552944931e-06,
+      "loss": 0.8478,
+      "step": 8765
+    },
+    {
+      "epoch": 1.687513535625767,
+      "grad_norm": 3.196920467056742,
+      "learning_rate": 1.2517546706531047e-06,
+      "loss": 0.8465,
+      "step": 8766
+    },
+    {
+      "epoch": 1.6877060423033425,
+      "grad_norm": 3.2516569158577493,
+      "learning_rate": 1.2502446622982688e-06,
+      "loss": 0.8495,
+      "step": 8767
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "grad_norm": 3.363300008182436,
+      "learning_rate": 1.2487355045315907e-06,
+      "loss": 0.8912,
+      "step": 8768
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.6801,
+      "step": 8768,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.4342,
+      "step": 8768,
+      "vm_loss": 0.2202
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.7569,
+      "step": 8768,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.4815,
+      "step": 8768,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.6115,
+      "step": 8768,
+      "vm_loss": 0.1669
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.6214,
+      "step": 8768,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 1.0029,
+      "step": 8768,
+      "vm_loss": 0.1516
+    },
+    {
+      "epoch": 1.6878985489809177,
+      "lm_loss": 0.6923,
+      "step": 8768,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.6880910556584932,
+      "grad_norm": 3.2129350435412074,
+      "learning_rate": 1.2472271974997863e-06,
+      "loss": 0.8576,
+      "step": 8769
+    },
+    {
+      "epoch": 1.6882835623360686,
+      "grad_norm": 3.222701045160172,
+      "learning_rate": 1.245719741349477e-06,
+      "loss": 0.8736,
+      "step": 8770
+    },
+    {
+      "epoch": 1.6884760690136438,
+      "grad_norm": 3.271693057210702,
+      "learning_rate": 1.2442131362272093e-06,
+      "loss": 0.8813,
+      "step": 8771
+    },
+    {
+      "epoch": 1.6886685756912194,
+      "grad_norm": 3.326589116107832,
+      "learning_rate": 1.2427073822794445e-06,
+      "loss": 0.8557,
+      "step": 8772
+    },
+    {
+      "epoch": 1.6888610823687946,
+      "grad_norm": 3.232868996477941,
+      "learning_rate": 1.2412024796525612e-06,
+      "loss": 0.8987,
+      "step": 8773
+    },
+    {
+      "epoch": 1.68905358904637,
+      "grad_norm": 3.1924700581425807,
+      "learning_rate": 1.2396984284928514e-06,
+      "loss": 0.833,
+      "step": 8774
+    },
+    {
+      "epoch": 1.6892460957239455,
+      "grad_norm": 3.25435308561204,
+      "learning_rate": 1.238195228946536e-06,
+      "loss": 0.8875,
+      "step": 8775
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "grad_norm": 3.340291984095352,
+      "learning_rate": 1.2366928811597378e-06,
+      "loss": 0.8975,
+      "step": 8776
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.772,
+      "step": 8776,
+      "vm_loss": 0.2086
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.7891,
+      "step": 8776,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.4316,
+      "step": 8776,
+      "vm_loss": 0.1874
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.8097,
+      "step": 8776,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.8689,
+      "step": 8776,
+      "vm_loss": 0.1543
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.6848,
+      "step": 8776,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.8599,
+      "step": 8776,
+      "vm_loss": 0.2112
+    },
+    {
+      "epoch": 1.6894386024015207,
+      "lm_loss": 0.6415,
+      "step": 8776,
+      "vm_loss": 0.2249
+    },
+    {
+      "epoch": 1.6896311090790963,
+      "grad_norm": 3.280167002163764,
+      "learning_rate": 1.235191385278507e-06,
+      "loss": 0.9048,
+      "step": 8777
+    },
+    {
+      "epoch": 1.6898236157566715,
+      "grad_norm": 3.2665331231164743,
+      "learning_rate": 1.2336907414488075e-06,
+      "loss": 0.8697,
+      "step": 8778
+    },
+    {
+      "epoch": 1.690016122434247,
+      "grad_norm": 3.232331044596003,
+      "learning_rate": 1.2321909498165253e-06,
+      "loss": 0.8751,
+      "step": 8779
+    },
+    {
+      "epoch": 1.6902086291118223,
+      "grad_norm": 3.3207693032688947,
+      "learning_rate": 1.2306920105274522e-06,
+      "loss": 0.8903,
+      "step": 8780
+    },
+    {
+      "epoch": 1.6904011357893975,
+      "grad_norm": 3.1370262610715454,
+      "learning_rate": 1.2291939237273088e-06,
+      "loss": 0.8484,
+      "step": 8781
+    },
+    {
+      "epoch": 1.6905936424669732,
+      "grad_norm": 3.10460582839599,
+      "learning_rate": 1.2276966895617271e-06,
+      "loss": 0.8238,
+      "step": 8782
+    },
+    {
+      "epoch": 1.6907861491445484,
+      "grad_norm": 3.2125352772695206,
+      "learning_rate": 1.2262003081762575e-06,
+      "loss": 0.8644,
+      "step": 8783
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "grad_norm": 3.4131757870152057,
+      "learning_rate": 1.2247047797163703e-06,
+      "loss": 0.9263,
+      "step": 8784
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.5714,
+      "step": 8784,
+      "vm_loss": 0.1823
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.4568,
+      "step": 8784,
+      "vm_loss": 0.1144
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.4888,
+      "step": 8784,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.6263,
+      "step": 8784,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.8641,
+      "step": 8784,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.7196,
+      "step": 8784,
+      "vm_loss": 0.1307
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.4606,
+      "step": 8784,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 1.6909786558221238,
+      "lm_loss": 0.6796,
+      "step": 8784,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 1.6911711624996992,
+      "grad_norm": 3.2619366070349267,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.8519,
+      "step": 8785
+    },
+    {
+      "epoch": 1.6913636691772744,
+      "grad_norm": 3.19861661496492,
+      "learning_rate": 1.2217162821547856e-06,
+      "loss": 0.8219,
+      "step": 8786
+    },
+    {
+      "epoch": 1.69155617585485,
+      "grad_norm": 3.215976733611148,
+      "learning_rate": 1.2202233133436103e-06,
+      "loss": 0.8398,
+      "step": 8787
+    },
+    {
+      "epoch": 1.6917486825324253,
+      "grad_norm": 3.2868616622423974,
+      "learning_rate": 1.218731198039056e-06,
+      "loss": 0.8733,
+      "step": 8788
+    },
+    {
+      "epoch": 1.6919411892100007,
+      "grad_norm": 3.296242635892787,
+      "learning_rate": 1.2172399363861731e-06,
+      "loss": 0.8692,
+      "step": 8789
+    },
+    {
+      "epoch": 1.6921336958875761,
+      "grad_norm": 3.445415280218687,
+      "learning_rate": 1.215749528529936e-06,
+      "loss": 0.906,
+      "step": 8790
+    },
+    {
+      "epoch": 1.6923262025651513,
+      "grad_norm": 3.1022956591400384,
+      "learning_rate": 1.214259974615225e-06,
+      "loss": 0.8629,
+      "step": 8791
+    },
+    {
+      "epoch": 1.692518709242727,
+      "grad_norm": 3.2987764466556206,
+      "learning_rate": 1.212771274786847e-06,
+      "loss": 0.8883,
+      "step": 8792
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.9196,
+      "step": 8792,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.7991,
+      "step": 8792,
+      "vm_loss": 0.1738
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.706,
+      "step": 8792,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.8006,
+      "step": 8792,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.3766,
+      "step": 8792,
+      "vm_loss": 0.1893
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.4099,
+      "step": 8792,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.6701,
+      "step": 8792,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.692518709242727,
+      "lm_loss": 0.7107,
+      "step": 8792,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.6927112159203022,
+      "grad_norm": 3.293036383871833,
+      "learning_rate": 1.2112834291895237e-06,
+      "loss": 0.9201,
+      "step": 8793
+    },
+    {
+      "epoch": 1.6929037225978776,
+      "grad_norm": 3.1647996403784244,
+      "learning_rate": 1.2097964379678928e-06,
+      "loss": 0.8495,
+      "step": 8794
+    },
+    {
+      "epoch": 1.693096229275453,
+      "grad_norm": 3.2883462094761686,
+      "learning_rate": 1.2083103012665032e-06,
+      "loss": 0.8603,
+      "step": 8795
+    },
+    {
+      "epoch": 1.6932887359530284,
+      "grad_norm": 3.4132015726123632,
+      "learning_rate": 1.2068250192298359e-06,
+      "loss": 0.921,
+      "step": 8796
+    },
+    {
+      "epoch": 1.6934812426306038,
+      "grad_norm": 3.1573830503695017,
+      "learning_rate": 1.2053405920022709e-06,
+      "loss": 0.8526,
+      "step": 8797
+    },
+    {
+      "epoch": 1.693673749308179,
+      "grad_norm": 3.2215020935971395,
+      "learning_rate": 1.2038570197281185e-06,
+      "loss": 0.8734,
+      "step": 8798
+    },
+    {
+      "epoch": 1.6938662559857545,
+      "grad_norm": 3.234821680935905,
+      "learning_rate": 1.2023743025515977e-06,
+      "loss": 0.8849,
+      "step": 8799
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "grad_norm": 3.3236629237547493,
+      "learning_rate": 1.200892440616851e-06,
+      "loss": 0.9253,
+      "step": 8800
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.6827,
+      "step": 8800,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.556,
+      "step": 8800,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 1.0148,
+      "step": 8800,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.6856,
+      "step": 8800,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.5674,
+      "step": 8800,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.9358,
+      "step": 8800,
+      "vm_loss": 0.2293
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.5597,
+      "step": 8800,
+      "vm_loss": 0.2256
+    },
+    {
+      "epoch": 1.6940587626633299,
+      "lm_loss": 0.4241,
+      "step": 8800,
+      "vm_loss": 0.1494
+    },
+    {
+      "epoch": 1.6942512693409053,
+      "grad_norm": 3.2657482098316617,
+      "learning_rate": 1.199411434067933e-06,
+      "loss": 0.8872,
+      "step": 8801
+    },
+    {
+      "epoch": 1.6944437760184807,
+      "grad_norm": 3.340099719276038,
+      "learning_rate": 1.1979312830488144e-06,
+      "loss": 0.9138,
+      "step": 8802
+    },
+    {
+      "epoch": 1.694636282696056,
+      "grad_norm": 3.1011893546733758,
+      "learning_rate": 1.1964519877033875e-06,
+      "loss": 0.7623,
+      "step": 8803
+    },
+    {
+      "epoch": 1.6948287893736314,
+      "grad_norm": 3.2498683166345437,
+      "learning_rate": 1.1949735481754565e-06,
+      "loss": 0.8725,
+      "step": 8804
+    },
+    {
+      "epoch": 1.6950212960512068,
+      "grad_norm": 3.2691608063931086,
+      "learning_rate": 1.1934959646087463e-06,
+      "loss": 0.8713,
+      "step": 8805
+    },
+    {
+      "epoch": 1.6952138027287822,
+      "grad_norm": 3.314856550691807,
+      "learning_rate": 1.1920192371468952e-06,
+      "loss": 0.9092,
+      "step": 8806
+    },
+    {
+      "epoch": 1.6954063094063576,
+      "grad_norm": 3.207018826587536,
+      "learning_rate": 1.1905433659334654e-06,
+      "loss": 0.8528,
+      "step": 8807
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "grad_norm": 3.3303544085614267,
+      "learning_rate": 1.1890683511119227e-06,
+      "loss": 0.8891,
+      "step": 8808
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.7064,
+      "step": 8808,
+      "vm_loss": 0.2118
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.5326,
+      "step": 8808,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.7962,
+      "step": 8808,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.6603,
+      "step": 8808,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.5298,
+      "step": 8808,
+      "vm_loss": 0.1525
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.4801,
+      "step": 8808,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.2973,
+      "step": 8808,
+      "vm_loss": 0.1648
+    },
+    {
+      "epoch": 1.6955988160839328,
+      "lm_loss": 0.5085,
+      "step": 8808,
+      "vm_loss": 0.1853
+    },
+    {
+      "epoch": 1.6957913227615082,
+      "grad_norm": 3.242280329101607,
+      "learning_rate": 1.1875941928256619e-06,
+      "loss": 0.8831,
+      "step": 8809
+    },
+    {
+      "epoch": 1.6959838294390837,
+      "grad_norm": 3.3903089925129075,
+      "learning_rate": 1.1861208912179878e-06,
+      "loss": 0.921,
+      "step": 8810
+    },
+    {
+      "epoch": 1.696176336116659,
+      "grad_norm": 3.110336189759906,
+      "learning_rate": 1.1846484464321283e-06,
+      "loss": 0.8228,
+      "step": 8811
+    },
+    {
+      "epoch": 1.6963688427942345,
+      "grad_norm": 2.9195620286093154,
+      "learning_rate": 1.183176858611218e-06,
+      "loss": 0.8101,
+      "step": 8812
+    },
+    {
+      "epoch": 1.6965613494718097,
+      "grad_norm": 3.1954545593021906,
+      "learning_rate": 1.1817061278983199e-06,
+      "loss": 0.8299,
+      "step": 8813
+    },
+    {
+      "epoch": 1.6967538561493853,
+      "grad_norm": 3.3515066839941245,
+      "learning_rate": 1.1802362544364032e-06,
+      "loss": 0.8934,
+      "step": 8814
+    },
+    {
+      "epoch": 1.6969463628269605,
+      "grad_norm": 3.2847235418717515,
+      "learning_rate": 1.1787672383683613e-06,
+      "loss": 0.8472,
+      "step": 8815
+    },
+    {
+      "epoch": 1.697138869504536,
+      "grad_norm": 3.2277639208366975,
+      "learning_rate": 1.1772990798369987e-06,
+      "loss": 0.8431,
+      "step": 8816
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.8186,
+      "step": 8816,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.9216,
+      "step": 8816,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.6937,
+      "step": 8816,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 1.1941,
+      "step": 8816,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.7434,
+      "step": 8816,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.6492,
+      "step": 8816,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.7571,
+      "step": 8816,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 1.697138869504536,
+      "lm_loss": 0.8377,
+      "step": 8816,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.6973313761821114,
+      "grad_norm": 3.2281535132777717,
+      "learning_rate": 1.1758317789850448e-06,
+      "loss": 0.8439,
+      "step": 8817
+    },
+    {
+      "epoch": 1.6975238828596866,
+      "grad_norm": 3.296686337121683,
+      "learning_rate": 1.1743653359551322e-06,
+      "loss": 0.8717,
+      "step": 8818
+    },
+    {
+      "epoch": 1.6977163895372622,
+      "grad_norm": 3.2527730765087655,
+      "learning_rate": 1.1728997508898231e-06,
+      "loss": 0.8311,
+      "step": 8819
+    },
+    {
+      "epoch": 1.6979088962148374,
+      "grad_norm": 3.36966773724426,
+      "learning_rate": 1.1714350239315897e-06,
+      "loss": 0.9131,
+      "step": 8820
+    },
+    {
+      "epoch": 1.6981014028924128,
+      "grad_norm": 3.115386333921638,
+      "learning_rate": 1.169971155222821e-06,
+      "loss": 0.807,
+      "step": 8821
+    },
+    {
+      "epoch": 1.6982939095699883,
+      "grad_norm": 3.2233436393329766,
+      "learning_rate": 1.1685081449058266e-06,
+      "loss": 0.8402,
+      "step": 8822
+    },
+    {
+      "epoch": 1.6984864162475635,
+      "grad_norm": 3.322835979842623,
+      "learning_rate": 1.1670459931228273e-06,
+      "loss": 0.9001,
+      "step": 8823
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "grad_norm": 3.2516267740552194,
+      "learning_rate": 1.165584700015967e-06,
+      "loss": 0.8987,
+      "step": 8824
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.7088,
+      "step": 8824,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.6307,
+      "step": 8824,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.6178,
+      "step": 8824,
+      "vm_loss": 0.085
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.8213,
+      "step": 8824,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.548,
+      "step": 8824,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.6688,
+      "step": 8824,
+      "vm_loss": 0.1058
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.8781,
+      "step": 8824,
+      "vm_loss": 0.1532
+    },
+    {
+      "epoch": 1.6986789229251391,
+      "lm_loss": 0.8078,
+      "step": 8824,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.6988714296027143,
+      "grad_norm": 3.4522088082983546,
+      "learning_rate": 1.1641242657272956e-06,
+      "loss": 0.8911,
+      "step": 8825
+    },
+    {
+      "epoch": 1.6990639362802897,
+      "grad_norm": 3.0845567499891,
+      "learning_rate": 1.1626646903987904e-06,
+      "loss": 0.8277,
+      "step": 8826
+    },
+    {
+      "epoch": 1.6992564429578652,
+      "grad_norm": 3.313818978770263,
+      "learning_rate": 1.1612059741723404e-06,
+      "loss": 0.8722,
+      "step": 8827
+    },
+    {
+      "epoch": 1.6994489496354404,
+      "grad_norm": 3.273247086552356,
+      "learning_rate": 1.1597481171897517e-06,
+      "loss": 0.8533,
+      "step": 8828
+    },
+    {
+      "epoch": 1.699641456313016,
+      "grad_norm": 3.2043476651241707,
+      "learning_rate": 1.1582911195927427e-06,
+      "loss": 0.8389,
+      "step": 8829
+    },
+    {
+      "epoch": 1.6998339629905912,
+      "grad_norm": 3.047389210958841,
+      "learning_rate": 1.1568349815229584e-06,
+      "loss": 0.8037,
+      "step": 8830
+    },
+    {
+      "epoch": 1.7000264696681666,
+      "grad_norm": 3.1142536885519854,
+      "learning_rate": 1.1553797031219494e-06,
+      "loss": 0.8292,
+      "step": 8831
+    },
+    {
+      "epoch": 1.700218976345742,
+      "grad_norm": 3.3186625098060487,
+      "learning_rate": 1.1539252845311889e-06,
+      "loss": 0.8555,
+      "step": 8832
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.5576,
+      "step": 8832,
+      "vm_loss": 0.1674
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.6503,
+      "step": 8832,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.8977,
+      "step": 8832,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.6176,
+      "step": 8832,
+      "vm_loss": 0.2102
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 1.0169,
+      "step": 8832,
+      "vm_loss": 0.2235
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.7094,
+      "step": 8832,
+      "vm_loss": 0.211
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.7935,
+      "step": 8832,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.700218976345742,
+      "lm_loss": 0.6207,
+      "step": 8832,
+      "vm_loss": 0.1231
+    },
+    {
+      "epoch": 1.7004114830233172,
+      "grad_norm": 3.4903967972516603,
+      "learning_rate": 1.1524717258920659e-06,
+      "loss": 0.9224,
+      "step": 8833
+    },
+    {
+      "epoch": 1.7006039897008929,
+      "grad_norm": 3.23226587650842,
+      "learning_rate": 1.1510190273458854e-06,
+      "loss": 0.8495,
+      "step": 8834
+    },
+    {
+      "epoch": 1.700796496378468,
+      "grad_norm": 3.2672053812749438,
+      "learning_rate": 1.1495671890338655e-06,
+      "loss": 0.8811,
+      "step": 8835
+    },
+    {
+      "epoch": 1.7009890030560435,
+      "grad_norm": 3.485093973705232,
+      "learning_rate": 1.1481162110971444e-06,
+      "loss": 0.9157,
+      "step": 8836
+    },
+    {
+      "epoch": 1.701181509733619,
+      "grad_norm": 2.991145489076773,
+      "learning_rate": 1.146666093676776e-06,
+      "loss": 0.8088,
+      "step": 8837
+    },
+    {
+      "epoch": 1.7013740164111941,
+      "grad_norm": 3.15955071681223,
+      "learning_rate": 1.145216836913733e-06,
+      "loss": 0.8188,
+      "step": 8838
+    },
+    {
+      "epoch": 1.7015665230887698,
+      "grad_norm": 3.3892870227751133,
+      "learning_rate": 1.1437684409488946e-06,
+      "loss": 0.8398,
+      "step": 8839
+    },
+    {
+      "epoch": 1.701759029766345,
+      "grad_norm": 3.252825081253866,
+      "learning_rate": 1.1423209059230721e-06,
+      "loss": 0.8577,
+      "step": 8840
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.8344,
+      "step": 8840,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.359,
+      "step": 8840,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.5752,
+      "step": 8840,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.4119,
+      "step": 8840,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.6752,
+      "step": 8840,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 1.4176,
+      "step": 8840,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.6824,
+      "step": 8840,
+      "vm_loss": 0.1445
+    },
+    {
+      "epoch": 1.701759029766345,
+      "lm_loss": 0.4542,
+      "step": 8840,
+      "vm_loss": 0.2145
+    },
+    {
+      "epoch": 1.7019515364439204,
+      "grad_norm": 3.1773351707510327,
+      "learning_rate": 1.1408742319769794e-06,
+      "loss": 0.8656,
+      "step": 8841
+    },
+    {
+      "epoch": 1.7021440431214958,
+      "grad_norm": 3.205217253908608,
+      "learning_rate": 1.1394284192512517e-06,
+      "loss": 0.8717,
+      "step": 8842
+    },
+    {
+      "epoch": 1.702336549799071,
+      "grad_norm": 3.2512623750535252,
+      "learning_rate": 1.137983467886442e-06,
+      "loss": 0.8895,
+      "step": 8843
+    },
+    {
+      "epoch": 1.7025290564766467,
+      "grad_norm": 3.0765128102293167,
+      "learning_rate": 1.1365393780230172e-06,
+      "loss": 0.8464,
+      "step": 8844
+    },
+    {
+      "epoch": 1.7027215631542219,
+      "grad_norm": 3.3757370732725995,
+      "learning_rate": 1.1350961498013635e-06,
+      "loss": 0.9124,
+      "step": 8845
+    },
+    {
+      "epoch": 1.7029140698317973,
+      "grad_norm": 3.44828825485161,
+      "learning_rate": 1.1336537833617777e-06,
+      "loss": 0.9366,
+      "step": 8846
+    },
+    {
+      "epoch": 1.7031065765093727,
+      "grad_norm": 3.352878235358373,
+      "learning_rate": 1.1322122788444779e-06,
+      "loss": 0.8902,
+      "step": 8847
+    },
+    {
+      "epoch": 1.703299083186948,
+      "grad_norm": 3.3688382639814862,
+      "learning_rate": 1.130771636389596e-06,
+      "loss": 0.9016,
+      "step": 8848
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.5318,
+      "step": 8848,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.6651,
+      "step": 8848,
+      "vm_loss": 0.1962
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.5197,
+      "step": 8848,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 1.1028,
+      "step": 8848,
+      "vm_loss": 0.1831
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.921,
+      "step": 8848,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.718,
+      "step": 8848,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.7463,
+      "step": 8848,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.703299083186948,
+      "lm_loss": 0.4,
+      "step": 8848,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.7034915898645235,
+      "grad_norm": 3.092255925314974,
+      "learning_rate": 1.1293318561371824e-06,
+      "loss": 0.8447,
+      "step": 8849
+    },
+    {
+      "epoch": 1.7036840965420987,
+      "grad_norm": 3.235421314293682,
+      "learning_rate": 1.1278929382272007e-06,
+      "loss": 0.8711,
+      "step": 8850
+    },
+    {
+      "epoch": 1.7038766032196742,
+      "grad_norm": 3.3291793349026904,
+      "learning_rate": 1.1264548827995347e-06,
+      "loss": 0.889,
+      "step": 8851
+    },
+    {
+      "epoch": 1.7040691098972496,
+      "grad_norm": 3.1758817065037004,
+      "learning_rate": 1.1250176899939791e-06,
+      "loss": 0.8395,
+      "step": 8852
+    },
+    {
+      "epoch": 1.7042616165748248,
+      "grad_norm": 3.1218227764345423,
+      "learning_rate": 1.1235813599502477e-06,
+      "loss": 0.8277,
+      "step": 8853
+    },
+    {
+      "epoch": 1.7044541232524004,
+      "grad_norm": 3.272083794868553,
+      "learning_rate": 1.1221458928079697e-06,
+      "loss": 0.9104,
+      "step": 8854
+    },
+    {
+      "epoch": 1.7046466299299756,
+      "grad_norm": 3.155395903635095,
+      "learning_rate": 1.120711288706695e-06,
+      "loss": 0.8427,
+      "step": 8855
+    },
+    {
+      "epoch": 1.704839136607551,
+      "grad_norm": 3.235599106719716,
+      "learning_rate": 1.1192775477858776e-06,
+      "loss": 0.9063,
+      "step": 8856
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.8269,
+      "step": 8856,
+      "vm_loss": 0.2233
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.9562,
+      "step": 8856,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.5613,
+      "step": 8856,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.8156,
+      "step": 8856,
+      "vm_loss": 0.1799
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.5601,
+      "step": 8856,
+      "vm_loss": 0.2429
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.7895,
+      "step": 8856,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.4669,
+      "step": 8856,
+      "vm_loss": 0.1186
+    },
+    {
+      "epoch": 1.704839136607551,
+      "lm_loss": 0.3733,
+      "step": 8856,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.7050316432851265,
+      "grad_norm": 3.2564330656743663,
+      "learning_rate": 1.1178446701849045e-06,
+      "loss": 0.9041,
+      "step": 8857
+    },
+    {
+      "epoch": 1.7052241499627019,
+      "grad_norm": 3.2676053615548217,
+      "learning_rate": 1.1164126560430633e-06,
+      "loss": 0.8497,
+      "step": 8858
+    },
+    {
+      "epoch": 1.7054166566402773,
+      "grad_norm": 3.209750321017128,
+      "learning_rate": 1.1149815054995682e-06,
+      "loss": 0.8197,
+      "step": 8859
+    },
+    {
+      "epoch": 1.7056091633178525,
+      "grad_norm": 3.1911299968321556,
+      "learning_rate": 1.1135512186935392e-06,
+      "loss": 0.8475,
+      "step": 8860
+    },
+    {
+      "epoch": 1.705801669995428,
+      "grad_norm": 3.3686046986211267,
+      "learning_rate": 1.1121217957640261e-06,
+      "loss": 0.8758,
+      "step": 8861
+    },
+    {
+      "epoch": 1.7059941766730033,
+      "grad_norm": 3.3407111736229838,
+      "learning_rate": 1.1106932368499824e-06,
+      "loss": 0.8809,
+      "step": 8862
+    },
+    {
+      "epoch": 1.7061866833505788,
+      "grad_norm": 3.4470173285355554,
+      "learning_rate": 1.1092655420902842e-06,
+      "loss": 0.9333,
+      "step": 8863
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "grad_norm": 3.355824641161408,
+      "learning_rate": 1.1078387116237198e-06,
+      "loss": 0.8842,
+      "step": 8864
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.5672,
+      "step": 8864,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.4775,
+      "step": 8864,
+      "vm_loss": 0.2216
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.5918,
+      "step": 8864,
+      "vm_loss": 0.1373
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 1.0284,
+      "step": 8864,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.5056,
+      "step": 8864,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.9401,
+      "step": 8864,
+      "vm_loss": 0.123
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.4532,
+      "step": 8864,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.7063791900281542,
+      "lm_loss": 0.511,
+      "step": 8864,
+      "vm_loss": 0.1145
+    },
+    {
+      "epoch": 1.7065716967057294,
+      "grad_norm": 3.4976075650895,
+      "learning_rate": 1.1064127455889974e-06,
+      "loss": 0.8965,
+      "step": 8865
+    },
+    {
+      "epoch": 1.7067642033833048,
+      "grad_norm": 3.3511588791340947,
+      "learning_rate": 1.1049876441247399e-06,
+      "loss": 0.9057,
+      "step": 8866
+    },
+    {
+      "epoch": 1.7069567100608802,
+      "grad_norm": 3.3104272371638275,
+      "learning_rate": 1.103563407369479e-06,
+      "loss": 0.8355,
+      "step": 8867
+    },
+    {
+      "epoch": 1.7071492167384557,
+      "grad_norm": 3.095068115777392,
+      "learning_rate": 1.1021400354616786e-06,
+      "loss": 0.838,
+      "step": 8868
+    },
+    {
+      "epoch": 1.707341723416031,
+      "grad_norm": 3.325558464061969,
+      "learning_rate": 1.100717528539702e-06,
+      "loss": 0.8676,
+      "step": 8869
+    },
+    {
+      "epoch": 1.7075342300936063,
+      "grad_norm": 3.079175326358344,
+      "learning_rate": 1.0992958867418358e-06,
+      "loss": 0.8183,
+      "step": 8870
+    },
+    {
+      "epoch": 1.7077267367711817,
+      "grad_norm": 3.2727756258494805,
+      "learning_rate": 1.0978751102062835e-06,
+      "loss": 0.8638,
+      "step": 8871
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "grad_norm": 3.2040327259188897,
+      "learning_rate": 1.0964551990711624e-06,
+      "loss": 0.8556,
+      "step": 8872
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.5498,
+      "step": 8872,
+      "vm_loss": 0.1406
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.5675,
+      "step": 8872,
+      "vm_loss": 0.1157
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.6545,
+      "step": 8872,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 1.3122,
+      "step": 8872,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.9058,
+      "step": 8872,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.5915,
+      "step": 8872,
+      "vm_loss": 0.1451
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 1.0227,
+      "step": 8872,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 1.7079192434487571,
+      "lm_loss": 0.7995,
+      "step": 8872,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 1.7081117501263325,
+      "grad_norm": 3.549670602581958,
+      "learning_rate": 1.0950361534745035e-06,
+      "loss": 0.8971,
+      "step": 8873
+    },
+    {
+      "epoch": 1.708304256803908,
+      "grad_norm": 3.1678794102312886,
+      "learning_rate": 1.0936179735542607e-06,
+      "loss": 0.8323,
+      "step": 8874
+    },
+    {
+      "epoch": 1.7084967634814832,
+      "grad_norm": 3.319870734717857,
+      "learning_rate": 1.0922006594482958e-06,
+      "loss": 0.868,
+      "step": 8875
+    },
+    {
+      "epoch": 1.7086892701590588,
+      "grad_norm": 3.1925521191272277,
+      "learning_rate": 1.0907842112943922e-06,
+      "loss": 0.8196,
+      "step": 8876
+    },
+    {
+      "epoch": 1.708881776836634,
+      "grad_norm": 3.2962530887831862,
+      "learning_rate": 1.089368629230242e-06,
+      "loss": 0.8774,
+      "step": 8877
+    },
+    {
+      "epoch": 1.7090742835142094,
+      "grad_norm": 3.1481665874266356,
+      "learning_rate": 1.0879539133934646e-06,
+      "loss": 0.8174,
+      "step": 8878
+    },
+    {
+      "epoch": 1.7092667901917848,
+      "grad_norm": 3.1248029117457694,
+      "learning_rate": 1.086540063921584e-06,
+      "loss": 0.83,
+      "step": 8879
+    },
+    {
+      "epoch": 1.70945929686936,
+      "grad_norm": 3.144088871676336,
+      "learning_rate": 1.0851270809520443e-06,
+      "loss": 0.8409,
+      "step": 8880
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.8083,
+      "step": 8880,
+      "vm_loss": 0.207
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.8819,
+      "step": 8880,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.7615,
+      "step": 8880,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.8085,
+      "step": 8880,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.8364,
+      "step": 8880,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 1.0129,
+      "step": 8880,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.5747,
+      "step": 8880,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.70945929686936,
+      "lm_loss": 0.5173,
+      "step": 8880,
+      "vm_loss": 0.1815
+    },
+    {
+      "epoch": 1.7096518035469357,
+      "grad_norm": 3.1494749438037895,
+      "learning_rate": 1.083714964622208e-06,
+      "loss": 0.8565,
+      "step": 8881
+    },
+    {
+      "epoch": 1.7098443102245109,
+      "grad_norm": 3.0538435932645385,
+      "learning_rate": 1.0823037150693517e-06,
+      "loss": 0.7939,
+      "step": 8882
+    },
+    {
+      "epoch": 1.7100368169020863,
+      "grad_norm": 3.3020248465267716,
+      "learning_rate": 1.0808933324306625e-06,
+      "loss": 0.9059,
+      "step": 8883
+    },
+    {
+      "epoch": 1.7102293235796617,
+      "grad_norm": 3.1571651179807114,
+      "learning_rate": 1.0794838168432498e-06,
+      "loss": 0.8262,
+      "step": 8884
+    },
+    {
+      "epoch": 1.710421830257237,
+      "grad_norm": 3.4548904385499117,
+      "learning_rate": 1.078075168444137e-06,
+      "loss": 0.9042,
+      "step": 8885
+    },
+    {
+      "epoch": 1.7106143369348126,
+      "grad_norm": 3.1062693327007698,
+      "learning_rate": 1.0766673873702637e-06,
+      "loss": 0.8193,
+      "step": 8886
+    },
+    {
+      "epoch": 1.7108068436123878,
+      "grad_norm": 3.24052405501799,
+      "learning_rate": 1.0752604737584816e-06,
+      "loss": 0.8652,
+      "step": 8887
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "grad_norm": 3.354381429865366,
+      "learning_rate": 1.0738544277455632e-06,
+      "loss": 0.8329,
+      "step": 8888
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.786,
+      "step": 8888,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.8755,
+      "step": 8888,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.4714,
+      "step": 8888,
+      "vm_loss": 0.1179
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.3446,
+      "step": 8888,
+      "vm_loss": 0.161
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 1.0874,
+      "step": 8888,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.4309,
+      "step": 8888,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.5435,
+      "step": 8888,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.7109993502899632,
+      "lm_loss": 0.7541,
+      "step": 8888,
+      "vm_loss": 0.1147
+    },
+    {
+      "epoch": 1.7111918569675386,
+      "grad_norm": 3.0629785336148614,
+      "learning_rate": 1.0724492494681948e-06,
+      "loss": 0.8017,
+      "step": 8889
+    },
+    {
+      "epoch": 1.7113843636451138,
+      "grad_norm": 3.276674143732075,
+      "learning_rate": 1.0710449390629719e-06,
+      "loss": 0.8381,
+      "step": 8890
+    },
+    {
+      "epoch": 1.7115768703226895,
+      "grad_norm": 3.3575417068128925,
+      "learning_rate": 1.0696414966664204e-06,
+      "loss": 0.8774,
+      "step": 8891
+    },
+    {
+      "epoch": 1.7117693770002647,
+      "grad_norm": 3.1326103936571785,
+      "learning_rate": 1.0682389224149648e-06,
+      "loss": 0.8079,
+      "step": 8892
+    },
+    {
+      "epoch": 1.71196188367784,
+      "grad_norm": 3.272785269420774,
+      "learning_rate": 1.0668372164449591e-06,
+      "loss": 0.8698,
+      "step": 8893
+    },
+    {
+      "epoch": 1.7121543903554155,
+      "grad_norm": 3.1438979682985413,
+      "learning_rate": 1.0654363788926614e-06,
+      "loss": 0.8231,
+      "step": 8894
+    },
+    {
+      "epoch": 1.7123468970329907,
+      "grad_norm": 3.2423727464744165,
+      "learning_rate": 1.0640364098942569e-06,
+      "loss": 0.8686,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "grad_norm": 3.412958246609101,
+      "learning_rate": 1.0626373095858367e-06,
+      "loss": 0.8951,
+      "step": 8896
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.8129,
+      "step": 8896,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.5878,
+      "step": 8896,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.5934,
+      "step": 8896,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.8676,
+      "step": 8896,
+      "vm_loss": 0.1891
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.5091,
+      "step": 8896,
+      "vm_loss": 0.1372
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.6997,
+      "step": 8896,
+      "vm_loss": 0.174
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.8642,
+      "step": 8896,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 1.7125394037105663,
+      "lm_loss": 0.6114,
+      "step": 8896,
+      "vm_loss": 0.2271
+    },
+    {
+      "epoch": 1.7127319103881415,
+      "grad_norm": 3.277949673835552,
+      "learning_rate": 1.061239078103411e-06,
+      "loss": 0.8743,
+      "step": 8897
+    },
+    {
+      "epoch": 1.712924417065717,
+      "grad_norm": 3.4287290540715865,
+      "learning_rate": 1.0598417155829078e-06,
+      "loss": 0.8971,
+      "step": 8898
+    },
+    {
+      "epoch": 1.7131169237432924,
+      "grad_norm": 3.331824573606808,
+      "learning_rate": 1.0584452221601704e-06,
+      "loss": 0.8794,
+      "step": 8899
+    },
+    {
+      "epoch": 1.7133094304208676,
+      "grad_norm": 3.205020819960327,
+      "learning_rate": 1.0570495979709505e-06,
+      "loss": 0.8341,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7135019370984432,
+      "grad_norm": 3.1647022962037012,
+      "learning_rate": 1.055654843150924e-06,
+      "loss": 0.8327,
+      "step": 8901
+    },
+    {
+      "epoch": 1.7136944437760184,
+      "grad_norm": 3.313356356073278,
+      "learning_rate": 1.0542609578356778e-06,
+      "loss": 0.8562,
+      "step": 8902
+    },
+    {
+      "epoch": 1.7138869504535938,
+      "grad_norm": 3.3872103573063823,
+      "learning_rate": 1.052867942160718e-06,
+      "loss": 0.9045,
+      "step": 8903
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "grad_norm": 3.3207980889387207,
+      "learning_rate": 1.0514757962614576e-06,
+      "loss": 0.8746,
+      "step": 8904
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.9454,
+      "step": 8904,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.417,
+      "step": 8904,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.7027,
+      "step": 8904,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.7999,
+      "step": 8904,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.4738,
+      "step": 8904,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 1.1223,
+      "step": 8904,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.8008,
+      "step": 8904,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.7140794571311693,
+      "lm_loss": 0.5958,
+      "step": 8904,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.7142719638087445,
+      "grad_norm": 3.2534060308958797,
+      "learning_rate": 1.0500845202732379e-06,
+      "loss": 0.8432,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7144644704863201,
+      "grad_norm": 3.1889812206645853,
+      "learning_rate": 1.0486941143313045e-06,
+      "loss": 0.8458,
+      "step": 8906
+    },
+    {
+      "epoch": 1.7146569771638953,
+      "grad_norm": 3.1750135545687903,
+      "learning_rate": 1.0473045785708246e-06,
+      "loss": 0.8487,
+      "step": 8907
+    },
+    {
+      "epoch": 1.7148494838414707,
+      "grad_norm": 3.1354632913089318,
+      "learning_rate": 1.045915913126877e-06,
+      "loss": 0.8337,
+      "step": 8908
+    },
+    {
+      "epoch": 1.7150419905190462,
+      "grad_norm": 3.353165955052305,
+      "learning_rate": 1.0445281181344591e-06,
+      "loss": 0.9055,
+      "step": 8909
+    },
+    {
+      "epoch": 1.7152344971966214,
+      "grad_norm": 3.416826887911343,
+      "learning_rate": 1.0431411937284842e-06,
+      "loss": 0.8593,
+      "step": 8910
+    },
+    {
+      "epoch": 1.715427003874197,
+      "grad_norm": 3.0695219998235803,
+      "learning_rate": 1.041755140043773e-06,
+      "loss": 0.8203,
+      "step": 8911
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "grad_norm": 3.1861432802485448,
+      "learning_rate": 1.040369957215076e-06,
+      "loss": 0.8685,
+      "step": 8912
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.8914,
+      "step": 8912,
+      "vm_loss": 0.133
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.742,
+      "step": 8912,
+      "vm_loss": 0.2242
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.4745,
+      "step": 8912,
+      "vm_loss": 0.1768
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.5232,
+      "step": 8912,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.5747,
+      "step": 8912,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.9237,
+      "step": 8912,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 1.1213,
+      "step": 8912,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.7156195105517722,
+      "lm_loss": 0.663,
+      "step": 8912,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 1.7158120172293476,
+      "grad_norm": 3.2214160179037585,
+      "learning_rate": 1.038985645377044e-06,
+      "loss": 0.8625,
+      "step": 8913
+    },
+    {
+      "epoch": 1.716004523906923,
+      "grad_norm": 3.2609101551257664,
+      "learning_rate": 1.037602204664252e-06,
+      "loss": 0.87,
+      "step": 8914
+    },
+    {
+      "epoch": 1.7161970305844982,
+      "grad_norm": 3.241086371573689,
+      "learning_rate": 1.0362196352111874e-06,
+      "loss": 0.8925,
+      "step": 8915
+    },
+    {
+      "epoch": 1.7163895372620739,
+      "grad_norm": 3.1531952615955987,
+      "learning_rate": 1.0348379371522577e-06,
+      "loss": 0.819,
+      "step": 8916
+    },
+    {
+      "epoch": 1.716582043939649,
+      "grad_norm": 3.134259475843129,
+      "learning_rate": 1.0334571106217762e-06,
+      "loss": 0.8379,
+      "step": 8917
+    },
+    {
+      "epoch": 1.7167745506172245,
+      "grad_norm": 3.199740697573619,
+      "learning_rate": 1.0320771557539788e-06,
+      "loss": 0.8611,
+      "step": 8918
+    },
+    {
+      "epoch": 1.7169670572948,
+      "grad_norm": 3.1992075535582165,
+      "learning_rate": 1.0306980726830151e-06,
+      "loss": 0.8327,
+      "step": 8919
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "grad_norm": 3.391728007590299,
+      "learning_rate": 1.0293198615429523e-06,
+      "loss": 0.8824,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.7107,
+      "step": 8920,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.4212,
+      "step": 8920,
+      "vm_loss": 0.167
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.737,
+      "step": 8920,
+      "vm_loss": 0.2087
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.5547,
+      "step": 8920,
+      "vm_loss": 0.1219
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.6571,
+      "step": 8920,
+      "vm_loss": 0.1929
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.5277,
+      "step": 8920,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.4847,
+      "step": 8920,
+      "vm_loss": 0.1245
+    },
+    {
+      "epoch": 1.7171595639723753,
+      "lm_loss": 0.7419,
+      "step": 8920,
+      "vm_loss": 0.1965
+    },
+    {
+      "epoch": 1.7173520706499508,
+      "grad_norm": 3.069343905186388,
+      "learning_rate": 1.027942522467762e-06,
+      "loss": 0.824,
+      "step": 8921
+    },
+    {
+      "epoch": 1.717544577327526,
+      "grad_norm": 3.2927043065668857,
+      "learning_rate": 1.0265660555913504e-06,
+      "loss": 0.8538,
+      "step": 8922
+    },
+    {
+      "epoch": 1.7177370840051014,
+      "grad_norm": 3.116163717225568,
+      "learning_rate": 1.0251904610475183e-06,
+      "loss": 0.8523,
+      "step": 8923
+    },
+    {
+      "epoch": 1.7179295906826768,
+      "grad_norm": 3.352077722368089,
+      "learning_rate": 1.0238157389699954e-06,
+      "loss": 0.8643,
+      "step": 8924
+    },
+    {
+      "epoch": 1.7181220973602522,
+      "grad_norm": 3.116826650243536,
+      "learning_rate": 1.0224418894924226e-06,
+      "loss": 0.826,
+      "step": 8925
+    },
+    {
+      "epoch": 1.7183146040378277,
+      "grad_norm": 3.250029160354693,
+      "learning_rate": 1.0210689127483552e-06,
+      "loss": 0.8379,
+      "step": 8926
+    },
+    {
+      "epoch": 1.7185071107154029,
+      "grad_norm": 3.09853203365307,
+      "learning_rate": 1.0196968088712621e-06,
+      "loss": 0.7994,
+      "step": 8927
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "grad_norm": 3.201837477430967,
+      "learning_rate": 1.0183255779945312e-06,
+      "loss": 0.837,
+      "step": 8928
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.7202,
+      "step": 8928,
+      "vm_loss": 0.1816
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.7553,
+      "step": 8928,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.9875,
+      "step": 8928,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.6147,
+      "step": 8928,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.5905,
+      "step": 8928,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.675,
+      "step": 8928,
+      "vm_loss": 0.1067
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.3813,
+      "step": 8928,
+      "vm_loss": 0.1646
+    },
+    {
+      "epoch": 1.7186996173929783,
+      "lm_loss": 0.6783,
+      "step": 8928,
+      "vm_loss": 0.2129
+    },
+    {
+      "epoch": 1.7188921240705537,
+      "grad_norm": 3.318475558073298,
+      "learning_rate": 1.0169552202514633e-06,
+      "loss": 0.8564,
+      "step": 8929
+    },
+    {
+      "epoch": 1.7190846307481291,
+      "grad_norm": 3.318063835923104,
+      "learning_rate": 1.0155857357752741e-06,
+      "loss": 0.8925,
+      "step": 8930
+    },
+    {
+      "epoch": 1.7192771374257045,
+      "grad_norm": 3.4515917711926685,
+      "learning_rate": 1.014217124699095e-06,
+      "loss": 0.9136,
+      "step": 8931
+    },
+    {
+      "epoch": 1.7194696441032797,
+      "grad_norm": 3.142491191565753,
+      "learning_rate": 1.012849387155974e-06,
+      "loss": 0.8201,
+      "step": 8932
+    },
+    {
+      "epoch": 1.7196621507808552,
+      "grad_norm": 3.2824590492340286,
+      "learning_rate": 1.011482523278874e-06,
+      "loss": 0.8733,
+      "step": 8933
+    },
+    {
+      "epoch": 1.7198546574584306,
+      "grad_norm": 3.3708114603109895,
+      "learning_rate": 1.0101165332006668e-06,
+      "loss": 0.8759,
+      "step": 8934
+    },
+    {
+      "epoch": 1.720047164136006,
+      "grad_norm": 3.32081758473011,
+      "learning_rate": 1.0087514170541479e-06,
+      "loss": 0.8303,
+      "step": 8935
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "grad_norm": 3.340233994665464,
+      "learning_rate": 1.0073871749720221e-06,
+      "loss": 0.8813,
+      "step": 8936
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.6952,
+      "step": 8936,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.8557,
+      "step": 8936,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.8264,
+      "step": 8936,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.7731,
+      "step": 8936,
+      "vm_loss": 0.1366
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.717,
+      "step": 8936,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.5549,
+      "step": 8936,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.4817,
+      "step": 8936,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.7202396708135814,
+      "lm_loss": 0.6742,
+      "step": 8936,
+      "vm_loss": 0.1407
+    },
+    {
+      "epoch": 1.7204321774911566,
+      "grad_norm": 3.2266470177293054,
+      "learning_rate": 1.0060238070869143e-06,
+      "loss": 0.8259,
+      "step": 8937
+    },
+    {
+      "epoch": 1.7206246841687323,
+      "grad_norm": 3.005420247059756,
+      "learning_rate": 1.0046613135313544e-06,
+      "loss": 0.7882,
+      "step": 8938
+    },
+    {
+      "epoch": 1.7208171908463075,
+      "grad_norm": 3.3415559983638525,
+      "learning_rate": 1.0032996944378027e-06,
+      "loss": 0.8775,
+      "step": 8939
+    },
+    {
+      "epoch": 1.7210096975238829,
+      "grad_norm": 3.44868233121355,
+      "learning_rate": 1.0019389499386212e-06,
+      "loss": 0.9046,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7212022042014583,
+      "grad_norm": 3.1713647240290435,
+      "learning_rate": 1.0005790801660942e-06,
+      "loss": 0.875,
+      "step": 8941
+    },
+    {
+      "epoch": 1.7213947108790335,
+      "grad_norm": 3.2429112801164304,
+      "learning_rate": 9.992200852524124e-07,
+      "loss": 0.8531,
+      "step": 8942
+    },
+    {
+      "epoch": 1.7215872175566092,
+      "grad_norm": 3.3411078949752975,
+      "learning_rate": 9.978619653296961e-07,
+      "loss": 0.8855,
+      "step": 8943
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "grad_norm": 3.2922186041006247,
+      "learning_rate": 9.965047205299649e-07,
+      "loss": 0.89,
+      "step": 8944
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.6766,
+      "step": 8944,
+      "vm_loss": 0.146
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.7868,
+      "step": 8944,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.4908,
+      "step": 8944,
+      "vm_loss": 0.1626
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.7814,
+      "step": 8944,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.8473,
+      "step": 8944,
+      "vm_loss": 0.1862
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.6605,
+      "step": 8944,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.5512,
+      "step": 8944,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.7217797242341843,
+      "lm_loss": 0.6782,
+      "step": 8944,
+      "vm_loss": 0.1047
+    },
+    {
+      "epoch": 1.7219722309117598,
+      "grad_norm": 3.3617177934642353,
+      "learning_rate": 9.951483509851644e-07,
+      "loss": 0.9132,
+      "step": 8945
+    },
+    {
+      "epoch": 1.7221647375893352,
+      "grad_norm": 3.324501972567766,
+      "learning_rate": 9.937928568271493e-07,
+      "loss": 0.8582,
+      "step": 8946
+    },
+    {
+      "epoch": 1.7223572442669104,
+      "grad_norm": 3.2989572301823626,
+      "learning_rate": 9.924382381876917e-07,
+      "loss": 0.8569,
+      "step": 8947
+    },
+    {
+      "epoch": 1.722549750944486,
+      "grad_norm": 3.132475058973454,
+      "learning_rate": 9.910844951984788e-07,
+      "loss": 0.836,
+      "step": 8948
+    },
+    {
+      "epoch": 1.7227422576220612,
+      "grad_norm": 3.0809040452973,
+      "learning_rate": 9.897316279911106e-07,
+      "loss": 0.8223,
+      "step": 8949
+    },
+    {
+      "epoch": 1.7229347642996367,
+      "grad_norm": 3.3126142331421957,
+      "learning_rate": 9.883796366971021e-07,
+      "loss": 0.8111,
+      "step": 8950
+    },
+    {
+      "epoch": 1.723127270977212,
+      "grad_norm": 3.450535666784199,
+      "learning_rate": 9.870285214478859e-07,
+      "loss": 0.9454,
+      "step": 8951
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "grad_norm": 3.184740417489782,
+      "learning_rate": 9.856782823748067e-07,
+      "loss": 0.8782,
+      "step": 8952
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.8355,
+      "step": 8952,
+      "vm_loss": 0.2282
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.677,
+      "step": 8952,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.3752,
+      "step": 8952,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.7342,
+      "step": 8952,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.5538,
+      "step": 8952,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.8072,
+      "step": 8952,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.8726,
+      "step": 8952,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.7233197776547873,
+      "lm_loss": 0.8837,
+      "step": 8952,
+      "vm_loss": 0.2287
+    },
+    {
+      "epoch": 1.723512284332363,
+      "grad_norm": 3.2031496758580733,
+      "learning_rate": 9.843289196091243e-07,
+      "loss": 0.8411,
+      "step": 8953
+    },
+    {
+      "epoch": 1.7237047910099381,
+      "grad_norm": 3.1436221080331204,
+      "learning_rate": 9.829804332820182e-07,
+      "loss": 0.8143,
+      "step": 8954
+    },
+    {
+      "epoch": 1.7238972976875135,
+      "grad_norm": 3.32229552729267,
+      "learning_rate": 9.81632823524571e-07,
+      "loss": 0.8764,
+      "step": 8955
+    },
+    {
+      "epoch": 1.724089804365089,
+      "grad_norm": 3.2636142067710447,
+      "learning_rate": 9.802860904677959e-07,
+      "loss": 0.8467,
+      "step": 8956
+    },
+    {
+      "epoch": 1.7242823110426642,
+      "grad_norm": 3.2950080610204027,
+      "learning_rate": 9.789402342426058e-07,
+      "loss": 0.8946,
+      "step": 8957
+    },
+    {
+      "epoch": 1.7244748177202398,
+      "grad_norm": 3.1483267203520406,
+      "learning_rate": 9.775952549798406e-07,
+      "loss": 0.8181,
+      "step": 8958
+    },
+    {
+      "epoch": 1.724667324397815,
+      "grad_norm": 3.2048427322343973,
+      "learning_rate": 9.762511528102437e-07,
+      "loss": 0.8543,
+      "step": 8959
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "grad_norm": 3.304744513621145,
+      "learning_rate": 9.749079278644846e-07,
+      "loss": 0.8927,
+      "step": 8960
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.4711,
+      "step": 8960,
+      "vm_loss": 0.2401
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.5376,
+      "step": 8960,
+      "vm_loss": 0.1088
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.8549,
+      "step": 8960,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.3762,
+      "step": 8960,
+      "vm_loss": 0.2143
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 1.2175,
+      "step": 8960,
+      "vm_loss": 0.1892
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.9253,
+      "step": 8960,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.6029,
+      "step": 8960,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.7248598310753904,
+      "lm_loss": 0.6167,
+      "step": 8960,
+      "vm_loss": 0.1544
+    },
+    {
+      "epoch": 1.7250523377529658,
+      "grad_norm": 3.334838370373949,
+      "learning_rate": 9.735655802731393e-07,
+      "loss": 0.8613,
+      "step": 8961
+    },
+    {
+      "epoch": 1.725244844430541,
+      "grad_norm": 3.234584263166714,
+      "learning_rate": 9.722241101667007e-07,
+      "loss": 0.859,
+      "step": 8962
+    },
+    {
+      "epoch": 1.7254373511081167,
+      "grad_norm": 3.3074746558211316,
+      "learning_rate": 9.70883517675577e-07,
+      "loss": 0.8494,
+      "step": 8963
+    },
+    {
+      "epoch": 1.725629857785692,
+      "grad_norm": 3.2494818069727116,
+      "learning_rate": 9.695438029300941e-07,
+      "loss": 0.8542,
+      "step": 8964
+    },
+    {
+      "epoch": 1.7258223644632673,
+      "grad_norm": 3.3043143389642595,
+      "learning_rate": 9.682049660604831e-07,
+      "loss": 0.8842,
+      "step": 8965
+    },
+    {
+      "epoch": 1.7260148711408427,
+      "grad_norm": 3.2213741608239213,
+      "learning_rate": 9.668670071969033e-07,
+      "loss": 0.8509,
+      "step": 8966
+    },
+    {
+      "epoch": 1.726207377818418,
+      "grad_norm": 3.126435078991705,
+      "learning_rate": 9.655299264694151e-07,
+      "loss": 0.825,
+      "step": 8967
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "grad_norm": 3.229934544123826,
+      "learning_rate": 9.641937240080046e-07,
+      "loss": 0.8789,
+      "step": 8968
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.4709,
+      "step": 8968,
+      "vm_loss": 0.2089
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.8347,
+      "step": 8968,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 1.0164,
+      "step": 8968,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.6292,
+      "step": 8968,
+      "vm_loss": 0.2068
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.3918,
+      "step": 8968,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.9147,
+      "step": 8968,
+      "vm_loss": 0.2303
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.6953,
+      "step": 8968,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 1.7263998844959936,
+      "lm_loss": 0.8437,
+      "step": 8968,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.7265923911735688,
+      "grad_norm": 3.2256585716227666,
+      "learning_rate": 9.628583999425644e-07,
+      "loss": 0.8918,
+      "step": 8969
+    },
+    {
+      "epoch": 1.7267848978511442,
+      "grad_norm": 3.102493749489921,
+      "learning_rate": 9.615239544029066e-07,
+      "loss": 0.8067,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7269774045287196,
+      "grad_norm": 3.2600718670955873,
+      "learning_rate": 9.60190387518759e-07,
+      "loss": 0.8748,
+      "step": 8971
+    },
+    {
+      "epoch": 1.7271699112062948,
+      "grad_norm": 3.2294348220547633,
+      "learning_rate": 9.588576994197563e-07,
+      "loss": 0.8835,
+      "step": 8972
+    },
+    {
+      "epoch": 1.7273624178838705,
+      "grad_norm": 3.274912463134301,
+      "learning_rate": 9.575258902354545e-07,
+      "loss": 0.8528,
+      "step": 8973
+    },
+    {
+      "epoch": 1.7275549245614457,
+      "grad_norm": 3.4362174095597333,
+      "learning_rate": 9.561949600953246e-07,
+      "loss": 0.874,
+      "step": 8974
+    },
+    {
+      "epoch": 1.727747431239021,
+      "grad_norm": 3.2540138174963715,
+      "learning_rate": 9.548649091287499e-07,
+      "loss": 0.8661,
+      "step": 8975
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "grad_norm": 3.3309790676960986,
+      "learning_rate": 9.535357374650234e-07,
+      "loss": 0.8776,
+      "step": 8976
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.7443,
+      "step": 8976,
+      "vm_loss": 0.1309
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.809,
+      "step": 8976,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.8912,
+      "step": 8976,
+      "vm_loss": 0.1064
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.7365,
+      "step": 8976,
+      "vm_loss": 0.1911
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.2947,
+      "step": 8976,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.4522,
+      "step": 8976,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.6013,
+      "step": 8976,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 1.7279399379165965,
+      "lm_loss": 0.9888,
+      "step": 8976,
+      "vm_loss": 0.1159
+    },
+    {
+      "epoch": 1.7281324445941717,
+      "grad_norm": 3.2086399070506224,
+      "learning_rate": 9.52207445233364e-07,
+      "loss": 0.829,
+      "step": 8977
+    },
+    {
+      "epoch": 1.7283249512717473,
+      "grad_norm": 3.2142612481278876,
+      "learning_rate": 9.508800325628953e-07,
+      "loss": 0.8203,
+      "step": 8978
+    },
+    {
+      "epoch": 1.7285174579493225,
+      "grad_norm": 3.30203227420588,
+      "learning_rate": 9.495534995826594e-07,
+      "loss": 0.8663,
+      "step": 8979
+    },
+    {
+      "epoch": 1.728709964626898,
+      "grad_norm": 3.3623202683828888,
+      "learning_rate": 9.482278464216121e-07,
+      "loss": 0.8906,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7289024713044734,
+      "grad_norm": 3.168417966935436,
+      "learning_rate": 9.469030732086259e-07,
+      "loss": 0.8296,
+      "step": 8981
+    },
+    {
+      "epoch": 1.7290949779820486,
+      "grad_norm": 3.180727006826908,
+      "learning_rate": 9.4557918007248e-07,
+      "loss": 0.8367,
+      "step": 8982
+    },
+    {
+      "epoch": 1.7292874846596242,
+      "grad_norm": 3.2421341190237003,
+      "learning_rate": 9.442561671418826e-07,
+      "loss": 0.8449,
+      "step": 8983
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "grad_norm": 3.1657655196022256,
+      "learning_rate": 9.429340345454397e-07,
+      "loss": 0.8446,
+      "step": 8984
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 1.0147,
+      "step": 8984,
+      "vm_loss": 0.1479
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.498,
+      "step": 8984,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.7227,
+      "step": 8984,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.7105,
+      "step": 8984,
+      "vm_loss": 0.2208
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.8562,
+      "step": 8984,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.7282,
+      "step": 8984,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 0.7734,
+      "step": 8984,
+      "vm_loss": 0.1846
+    },
+    {
+      "epoch": 1.7294799913371994,
+      "lm_loss": 1.0913,
+      "step": 8984,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 1.7296724980147749,
+      "grad_norm": 3.3580211341887938,
+      "learning_rate": 9.416127824116839e-07,
+      "loss": 0.8897,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7298650046923503,
+      "grad_norm": 3.219064113154529,
+      "learning_rate": 9.402924108690536e-07,
+      "loss": 0.833,
+      "step": 8986
+    },
+    {
+      "epoch": 1.7300575113699257,
+      "grad_norm": 3.331850850845125,
+      "learning_rate": 9.389729200459108e-07,
+      "loss": 0.9122,
+      "step": 8987
+    },
+    {
+      "epoch": 1.7302500180475011,
+      "grad_norm": 3.3481557477256367,
+      "learning_rate": 9.376543100705238e-07,
+      "loss": 0.9224,
+      "step": 8988
+    },
+    {
+      "epoch": 1.7304425247250763,
+      "grad_norm": 3.1726878292574665,
+      "learning_rate": 9.363365810710789e-07,
+      "loss": 0.8387,
+      "step": 8989
+    },
+    {
+      "epoch": 1.7306350314026517,
+      "grad_norm": 3.361946644076714,
+      "learning_rate": 9.350197331756761e-07,
+      "loss": 0.8998,
+      "step": 8990
+    },
+    {
+      "epoch": 1.7308275380802272,
+      "grad_norm": 3.0532075814681106,
+      "learning_rate": 9.337037665123305e-07,
+      "loss": 0.8108,
+      "step": 8991
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "grad_norm": 3.1757435609887685,
+      "learning_rate": 9.323886812089722e-07,
+      "loss": 0.8322,
+      "step": 8992
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.7484,
+      "step": 8992,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.6451,
+      "step": 8992,
+      "vm_loss": 0.1168
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.5853,
+      "step": 8992,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.7339,
+      "step": 8992,
+      "vm_loss": 0.1872
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.5924,
+      "step": 8992,
+      "vm_loss": 0.1163
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.7403,
+      "step": 8992,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.5906,
+      "step": 8992,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.7310200447578026,
+      "lm_loss": 0.6575,
+      "step": 8992,
+      "vm_loss": 0.097
+    },
+    {
+      "epoch": 1.731212551435378,
+      "grad_norm": 3.2269399865789836,
+      "learning_rate": 9.31074477393441e-07,
+      "loss": 0.8166,
+      "step": 8993
+    },
+    {
+      "epoch": 1.7314050581129532,
+      "grad_norm": 3.2261221746455986,
+      "learning_rate": 9.29761155193496e-07,
+      "loss": 0.8515,
+      "step": 8994
+    },
+    {
+      "epoch": 1.7315975647905286,
+      "grad_norm": 3.254771989080235,
+      "learning_rate": 9.284487147368093e-07,
+      "loss": 0.8676,
+      "step": 8995
+    },
+    {
+      "epoch": 1.731790071468104,
+      "grad_norm": 3.148020297369259,
+      "learning_rate": 9.271371561509657e-07,
+      "loss": 0.8014,
+      "step": 8996
+    },
+    {
+      "epoch": 1.7319825781456795,
+      "grad_norm": 3.138304011874794,
+      "learning_rate": 9.258264795634675e-07,
+      "loss": 0.8299,
+      "step": 8997
+    },
+    {
+      "epoch": 1.7321750848232549,
+      "grad_norm": 3.2520234328664914,
+      "learning_rate": 9.245166851017284e-07,
+      "loss": 0.8762,
+      "step": 8998
+    },
+    {
+      "epoch": 1.73236759150083,
+      "grad_norm": 3.2425417671841603,
+      "learning_rate": 9.232077728930744e-07,
+      "loss": 0.872,
+      "step": 8999
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "grad_norm": 3.3174267535684945,
+      "learning_rate": 9.218997430647548e-07,
+      "loss": 0.9155,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.8622,
+      "step": 9000,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.4251,
+      "step": 9000,
+      "vm_loss": 0.1478
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.8288,
+      "step": 9000,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.8252,
+      "step": 9000,
+      "vm_loss": 0.2165
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.3604,
+      "step": 9000,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.7424,
+      "step": 9000,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 1.0767,
+      "step": 9000,
+      "vm_loss": 0.0914
+    },
+    {
+      "epoch": 1.7325600981784057,
+      "lm_loss": 0.5247,
+      "step": 9000,
+      "vm_loss": 0.1524
+    },
+    {
+      "epoch": 1.732752604855981,
+      "grad_norm": 3.2549588248316135,
+      "learning_rate": 9.205925957439199e-07,
+      "loss": 0.8588,
+      "step": 9001
+    },
+    {
+      "epoch": 1.7329451115335563,
+      "grad_norm": 3.2036334965312427,
+      "learning_rate": 9.192863310576472e-07,
+      "loss": 0.8346,
+      "step": 9002
+    },
+    {
+      "epoch": 1.7331376182111318,
+      "grad_norm": 3.373110106875319,
+      "learning_rate": 9.17980949132915e-07,
+      "loss": 0.9018,
+      "step": 9003
+    },
+    {
+      "epoch": 1.733330124888707,
+      "grad_norm": 3.226449924566269,
+      "learning_rate": 9.166764500966319e-07,
+      "loss": 0.8476,
+      "step": 9004
+    },
+    {
+      "epoch": 1.7335226315662826,
+      "grad_norm": 3.2134958873577055,
+      "learning_rate": 9.153728340756052e-07,
+      "loss": 0.8562,
+      "step": 9005
+    },
+    {
+      "epoch": 1.7337151382438578,
+      "grad_norm": 3.1391893644490643,
+      "learning_rate": 9.140701011965647e-07,
+      "loss": 0.8248,
+      "step": 9006
+    },
+    {
+      "epoch": 1.7339076449214332,
+      "grad_norm": 3.2479247629223145,
+      "learning_rate": 9.127682515861546e-07,
+      "loss": 0.8871,
+      "step": 9007
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "grad_norm": 3.1520820331982593,
+      "learning_rate": 9.114672853709305e-07,
+      "loss": 0.8504,
+      "step": 9008
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.632,
+      "step": 9008,
+      "vm_loss": 0.1275
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.8274,
+      "step": 9008,
+      "vm_loss": 0.1251
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.3782,
+      "step": 9008,
+      "vm_loss": 0.1072
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.4288,
+      "step": 9008,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.6363,
+      "step": 9008,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.532,
+      "step": 9008,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.4103,
+      "step": 9008,
+      "vm_loss": 0.1106
+    },
+    {
+      "epoch": 1.7341001515990087,
+      "lm_loss": 0.5307,
+      "step": 9008,
+      "vm_loss": 0.0994
+    },
+    {
+      "epoch": 1.7342926582765839,
+      "grad_norm": 3.2181576364098854,
+      "learning_rate": 9.101672026773622e-07,
+      "loss": 0.8131,
+      "step": 9009
+    },
+    {
+      "epoch": 1.7344851649541595,
+      "grad_norm": 3.2520779694234347,
+      "learning_rate": 9.088680036318342e-07,
+      "loss": 0.885,
+      "step": 9010
+    },
+    {
+      "epoch": 1.7346776716317347,
+      "grad_norm": 3.303490851609354,
+      "learning_rate": 9.075696883606455e-07,
+      "loss": 0.9033,
+      "step": 9011
+    },
+    {
+      "epoch": 1.7348701783093101,
+      "grad_norm": 3.296322793675332,
+      "learning_rate": 9.062722569900085e-07,
+      "loss": 0.8446,
+      "step": 9012
+    },
+    {
+      "epoch": 1.7350626849868855,
+      "grad_norm": 3.446393330110615,
+      "learning_rate": 9.04975709646052e-07,
+      "loss": 0.9133,
+      "step": 9013
+    },
+    {
+      "epoch": 1.7352551916644607,
+      "grad_norm": 3.1315834919181134,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.8558,
+      "step": 9014
+    },
+    {
+      "epoch": 1.7354476983420364,
+      "grad_norm": 3.3058518717361016,
+      "learning_rate": 9.023852675422562e-07,
+      "loss": 0.8379,
+      "step": 9015
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "grad_norm": 3.2571351661639434,
+      "learning_rate": 9.010913730342397e-07,
+      "loss": 0.8438,
+      "step": 9016
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.7141,
+      "step": 9016,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.7035,
+      "step": 9016,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.4547,
+      "step": 9016,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.582,
+      "step": 9016,
+      "vm_loss": 0.0812
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.6689,
+      "step": 9016,
+      "vm_loss": 0.1729
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.8338,
+      "step": 9016,
+      "vm_loss": 0.2081
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.3363,
+      "step": 9016,
+      "vm_loss": 0.1337
+    },
+    {
+      "epoch": 1.7356402050196116,
+      "lm_loss": 0.486,
+      "step": 9016,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 1.735832711697187,
+      "grad_norm": 3.182211175190881,
+      "learning_rate": 8.99798363056551e-07,
+      "loss": 0.855,
+      "step": 9017
+    },
+    {
+      "epoch": 1.7360252183747624,
+      "grad_norm": 3.195317246702894,
+      "learning_rate": 8.985062377348863e-07,
+      "loss": 0.859,
+      "step": 9018
+    },
+    {
+      "epoch": 1.7362177250523376,
+      "grad_norm": 3.4047623851792053,
+      "learning_rate": 8.972149971948607e-07,
+      "loss": 0.9195,
+      "step": 9019
+    },
+    {
+      "epoch": 1.7364102317299133,
+      "grad_norm": 3.4958953844186182,
+      "learning_rate": 8.959246415619905e-07,
+      "loss": 0.9325,
+      "step": 9020
+    },
+    {
+      "epoch": 1.7366027384074885,
+      "grad_norm": 3.252647251365429,
+      "learning_rate": 8.946351709617251e-07,
+      "loss": 0.849,
+      "step": 9021
+    },
+    {
+      "epoch": 1.7367952450850639,
+      "grad_norm": 3.4679321251220863,
+      "learning_rate": 8.933465855194101e-07,
+      "loss": 0.904,
+      "step": 9022
+    },
+    {
+      "epoch": 1.7369877517626393,
+      "grad_norm": 3.206615775500758,
+      "learning_rate": 8.920588853603163e-07,
+      "loss": 0.8236,
+      "step": 9023
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "grad_norm": 3.290031510272475,
+      "learning_rate": 8.907720706096223e-07,
+      "loss": 0.849,
+      "step": 9024
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.9044,
+      "step": 9024,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.5596,
+      "step": 9024,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.7644,
+      "step": 9024,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.7777,
+      "step": 9024,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.5802,
+      "step": 9024,
+      "vm_loss": 0.1264
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.5636,
+      "step": 9024,
+      "vm_loss": 0.168
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.6826,
+      "step": 9024,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.7371802584402145,
+      "lm_loss": 0.5797,
+      "step": 9024,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 1.7373727651177902,
+      "grad_norm": 3.2458183846599966,
+      "learning_rate": 8.894861413924261e-07,
+      "loss": 0.8509,
+      "step": 9025
+    },
+    {
+      "epoch": 1.7375652717953654,
+      "grad_norm": 3.1980997709650154,
+      "learning_rate": 8.88201097833733e-07,
+      "loss": 0.86,
+      "step": 9026
+    },
+    {
+      "epoch": 1.7377577784729408,
+      "grad_norm": 3.2613122498285825,
+      "learning_rate": 8.869169400584665e-07,
+      "loss": 0.8707,
+      "step": 9027
+    },
+    {
+      "epoch": 1.7379502851505162,
+      "grad_norm": 3.237536655993925,
+      "learning_rate": 8.856336681914646e-07,
+      "loss": 0.8592,
+      "step": 9028
+    },
+    {
+      "epoch": 1.7381427918280914,
+      "grad_norm": 3.148861421584768,
+      "learning_rate": 8.843512823574784e-07,
+      "loss": 0.8571,
+      "step": 9029
+    },
+    {
+      "epoch": 1.738335298505667,
+      "grad_norm": 3.4609060602842887,
+      "learning_rate": 8.830697826811674e-07,
+      "loss": 0.9073,
+      "step": 9030
+    },
+    {
+      "epoch": 1.7385278051832422,
+      "grad_norm": 3.299052295075287,
+      "learning_rate": 8.817891692871161e-07,
+      "loss": 0.8489,
+      "step": 9031
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "grad_norm": 3.172806601968051,
+      "learning_rate": 8.805094422998117e-07,
+      "loss": 0.8576,
+      "step": 9032
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 0.7352,
+      "step": 9032,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 0.5268,
+      "step": 9032,
+      "vm_loss": 0.2251
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 1.0179,
+      "step": 9032,
+      "vm_loss": 0.0958
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 1.0287,
+      "step": 9032,
+      "vm_loss": 0.2244
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 0.5046,
+      "step": 9032,
+      "vm_loss": 0.1833
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 0.4661,
+      "step": 9032,
+      "vm_loss": 0.1204
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 0.8889,
+      "step": 9032,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.7387203118608177,
+      "lm_loss": 1.2154,
+      "step": 9032,
+      "vm_loss": 0.1455
+    },
+    {
+      "epoch": 1.738912818538393,
+      "grad_norm": 3.314951912785957,
+      "learning_rate": 8.792306018436625e-07,
+      "loss": 0.8696,
+      "step": 9033
+    },
+    {
+      "epoch": 1.7391053252159683,
+      "grad_norm": 3.1266469335019242,
+      "learning_rate": 8.779526480429868e-07,
+      "loss": 0.8237,
+      "step": 9034
+    },
+    {
+      "epoch": 1.739297831893544,
+      "grad_norm": 3.136114447931284,
+      "learning_rate": 8.766755810220195e-07,
+      "loss": 0.8163,
+      "step": 9035
+    },
+    {
+      "epoch": 1.7394903385711191,
+      "grad_norm": 3.269079562078465,
+      "learning_rate": 8.753994009049082e-07,
+      "loss": 0.837,
+      "step": 9036
+    },
+    {
+      "epoch": 1.7396828452486945,
+      "grad_norm": 3.160373829140482,
+      "learning_rate": 8.741241078157103e-07,
+      "loss": 0.8679,
+      "step": 9037
+    },
+    {
+      "epoch": 1.73987535192627,
+      "grad_norm": 3.437835693214052,
+      "learning_rate": 8.728497018784065e-07,
+      "loss": 0.8683,
+      "step": 9038
+    },
+    {
+      "epoch": 1.7400678586038452,
+      "grad_norm": 3.1756604690853942,
+      "learning_rate": 8.71576183216879e-07,
+      "loss": 0.8264,
+      "step": 9039
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "grad_norm": 3.3231204425868346,
+      "learning_rate": 8.703035519549341e-07,
+      "loss": 0.8767,
+      "step": 9040
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.6751,
+      "step": 9040,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.5237,
+      "step": 9040,
+      "vm_loss": 0.197
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.6484,
+      "step": 9040,
+      "vm_loss": 0.1733
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.6107,
+      "step": 9040,
+      "vm_loss": 0.1663
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.554,
+      "step": 9040,
+      "vm_loss": 0.2036
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.3648,
+      "step": 9040,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 1.052,
+      "step": 9040,
+      "vm_loss": 0.1071
+    },
+    {
+      "epoch": 1.7402603652814208,
+      "lm_loss": 0.737,
+      "step": 9040,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.740452871958996,
+      "grad_norm": 3.2486605147990266,
+      "learning_rate": 8.690318082162874e-07,
+      "loss": 0.8301,
+      "step": 9041
+    },
+    {
+      "epoch": 1.7406453786365714,
+      "grad_norm": 3.563130851788097,
+      "learning_rate": 8.677609521245689e-07,
+      "loss": 0.8896,
+      "step": 9042
+    },
+    {
+      "epoch": 1.7408378853141468,
+      "grad_norm": 3.2319477625742743,
+      "learning_rate": 8.664909838033198e-07,
+      "loss": 0.8334,
+      "step": 9043
+    },
+    {
+      "epoch": 1.741030391991722,
+      "grad_norm": 3.4206303934273006,
+      "learning_rate": 8.65221903375999e-07,
+      "loss": 0.8898,
+      "step": 9044
+    },
+    {
+      "epoch": 1.7412228986692977,
+      "grad_norm": 3.1336442818814834,
+      "learning_rate": 8.639537109659768e-07,
+      "loss": 0.8028,
+      "step": 9045
+    },
+    {
+      "epoch": 1.741415405346873,
+      "grad_norm": 3.340455889622764,
+      "learning_rate": 8.626864066965401e-07,
+      "loss": 0.8669,
+      "step": 9046
+    },
+    {
+      "epoch": 1.7416079120244483,
+      "grad_norm": 3.406345664781677,
+      "learning_rate": 8.614199906908816e-07,
+      "loss": 0.8765,
+      "step": 9047
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "grad_norm": 3.3483306032355102,
+      "learning_rate": 8.601544630721203e-07,
+      "loss": 0.8655,
+      "step": 9048
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.8541,
+      "step": 9048,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.7562,
+      "step": 9048,
+      "vm_loss": 0.1247
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.8092,
+      "step": 9048,
+      "vm_loss": 0.1198
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.322,
+      "step": 9048,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.9984,
+      "step": 9048,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.7207,
+      "step": 9048,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.3962,
+      "step": 9048,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.7418004187020237,
+      "lm_loss": 0.6018,
+      "step": 9048,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 1.7419929253795992,
+      "grad_norm": 3.1300994839524665,
+      "learning_rate": 8.588898239632759e-07,
+      "loss": 0.8061,
+      "step": 9049
+    },
+    {
+      "epoch": 1.7421854320571746,
+      "grad_norm": 3.1097956310816732,
+      "learning_rate": 8.576260734872909e-07,
+      "loss": 0.8035,
+      "step": 9050
+    },
+    {
+      "epoch": 1.7423779387347498,
+      "grad_norm": 3.2945801950378395,
+      "learning_rate": 8.56363211767014e-07,
+      "loss": 0.8551,
+      "step": 9051
+    },
+    {
+      "epoch": 1.7425704454123252,
+      "grad_norm": 3.3061136927750594,
+      "learning_rate": 8.551012389252178e-07,
+      "loss": 0.8974,
+      "step": 9052
+    },
+    {
+      "epoch": 1.7427629520899006,
+      "grad_norm": 3.1961286135381513,
+      "learning_rate": 8.538401550845755e-07,
+      "loss": 0.8356,
+      "step": 9053
+    },
+    {
+      "epoch": 1.742955458767476,
+      "grad_norm": 3.240790488252749,
+      "learning_rate": 8.525799603676843e-07,
+      "loss": 0.8297,
+      "step": 9054
+    },
+    {
+      "epoch": 1.7431479654450515,
+      "grad_norm": 3.351525773253788,
+      "learning_rate": 8.51320654897051e-07,
+      "loss": 0.8515,
+      "step": 9055
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "grad_norm": 3.2579584368283894,
+      "learning_rate": 8.500622387950963e-07,
+      "loss": 0.8392,
+      "step": 9056
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.6674,
+      "step": 9056,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.631,
+      "step": 9056,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.8712,
+      "step": 9056,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.5818,
+      "step": 9056,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.7312,
+      "step": 9056,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.9161,
+      "step": 9056,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.6469,
+      "step": 9056,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 1.7433404721226267,
+      "lm_loss": 0.9487,
+      "step": 9056,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.743532978800202,
+      "grad_norm": 3.1220706014664295,
+      "learning_rate": 8.488047121841525e-07,
+      "loss": 0.8607,
+      "step": 9057
+    },
+    {
+      "epoch": 1.7437254854777775,
+      "grad_norm": 3.2105965996865558,
+      "learning_rate": 8.475480751864695e-07,
+      "loss": 0.8236,
+      "step": 9058
+    },
+    {
+      "epoch": 1.743917992155353,
+      "grad_norm": 3.195784925865514,
+      "learning_rate": 8.462923279242097e-07,
+      "loss": 0.8522,
+      "step": 9059
+    },
+    {
+      "epoch": 1.7441104988329283,
+      "grad_norm": 3.281388049722419,
+      "learning_rate": 8.450374705194431e-07,
+      "loss": 0.8566,
+      "step": 9060
+    },
+    {
+      "epoch": 1.7443030055105035,
+      "grad_norm": 3.2844970868074728,
+      "learning_rate": 8.43783503094161e-07,
+      "loss": 0.8465,
+      "step": 9061
+    },
+    {
+      "epoch": 1.7444955121880792,
+      "grad_norm": 3.206643886525386,
+      "learning_rate": 8.425304257702649e-07,
+      "loss": 0.8393,
+      "step": 9062
+    },
+    {
+      "epoch": 1.7446880188656544,
+      "grad_norm": 3.0373121499613815,
+      "learning_rate": 8.412782386695706e-07,
+      "loss": 0.8228,
+      "step": 9063
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "grad_norm": 3.3303288410085816,
+      "learning_rate": 8.40026941913803e-07,
+      "loss": 0.8821,
+      "step": 9064
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.79,
+      "step": 9064,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.4584,
+      "step": 9064,
+      "vm_loss": 0.1273
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.7011,
+      "step": 9064,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.5816,
+      "step": 9064,
+      "vm_loss": 0.2272
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.614,
+      "step": 9064,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.4739,
+      "step": 9064,
+      "vm_loss": 0.1193
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.7292,
+      "step": 9064,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.7448805255432298,
+      "lm_loss": 0.618,
+      "step": 9064,
+      "vm_loss": 0.1809
+    },
+    {
+      "epoch": 1.7450730322208052,
+      "grad_norm": 3.252755256884406,
+      "learning_rate": 8.387765356246103e-07,
+      "loss": 0.8396,
+      "step": 9065
+    },
+    {
+      "epoch": 1.7452655388983804,
+      "grad_norm": 3.251938740101238,
+      "learning_rate": 8.37527019923543e-07,
+      "loss": 0.8439,
+      "step": 9066
+    },
+    {
+      "epoch": 1.745458045575956,
+      "grad_norm": 3.4793442412046156,
+      "learning_rate": 8.36278394932073e-07,
+      "loss": 0.8608,
+      "step": 9067
+    },
+    {
+      "epoch": 1.7456505522535313,
+      "grad_norm": 3.135055004362577,
+      "learning_rate": 8.350306607715774e-07,
+      "loss": 0.8213,
+      "step": 9068
+    },
+    {
+      "epoch": 1.7458430589311067,
+      "grad_norm": 3.3190710950868585,
+      "learning_rate": 8.337838175633595e-07,
+      "loss": 0.8737,
+      "step": 9069
+    },
+    {
+      "epoch": 1.7460355656086821,
+      "grad_norm": 3.242552754421227,
+      "learning_rate": 8.325378654286231e-07,
+      "loss": 0.835,
+      "step": 9070
+    },
+    {
+      "epoch": 1.7462280722862573,
+      "grad_norm": 3.015522336621416,
+      "learning_rate": 8.312928044884927e-07,
+      "loss": 0.8116,
+      "step": 9071
+    },
+    {
+      "epoch": 1.746420578963833,
+      "grad_norm": 3.3116097643827125,
+      "learning_rate": 8.300486348640036e-07,
+      "loss": 0.8833,
+      "step": 9072
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.6108,
+      "step": 9072,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.9349,
+      "step": 9072,
+      "vm_loss": 0.16
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.8165,
+      "step": 9072,
+      "vm_loss": 0.2001
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.7814,
+      "step": 9072,
+      "vm_loss": 0.1267
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.544,
+      "step": 9072,
+      "vm_loss": 0.1068
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.5652,
+      "step": 9072,
+      "vm_loss": 0.114
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.722,
+      "step": 9072,
+      "vm_loss": 0.0874
+    },
+    {
+      "epoch": 1.746420578963833,
+      "lm_loss": 0.7936,
+      "step": 9072,
+      "vm_loss": 0.1937
+    },
+    {
+      "epoch": 1.7466130856414082,
+      "grad_norm": 3.331721946012661,
+      "learning_rate": 8.288053566761067e-07,
+      "loss": 0.8773,
+      "step": 9073
+    },
+    {
+      "epoch": 1.7468055923189836,
+      "grad_norm": 3.1867124841193535,
+      "learning_rate": 8.275629700456588e-07,
+      "loss": 0.8513,
+      "step": 9074
+    },
+    {
+      "epoch": 1.746998098996559,
+      "grad_norm": 3.2228344828256517,
+      "learning_rate": 8.263214750934445e-07,
+      "loss": 0.8203,
+      "step": 9075
+    },
+    {
+      "epoch": 1.7471906056741342,
+      "grad_norm": 3.3618597439193683,
+      "learning_rate": 8.25080871940146e-07,
+      "loss": 0.8774,
+      "step": 9076
+    },
+    {
+      "epoch": 1.7473831123517098,
+      "grad_norm": 3.334811195752593,
+      "learning_rate": 8.238411607063678e-07,
+      "loss": 0.8588,
+      "step": 9077
+    },
+    {
+      "epoch": 1.747575619029285,
+      "grad_norm": 3.2305183102321933,
+      "learning_rate": 8.226023415126272e-07,
+      "loss": 0.8491,
+      "step": 9078
+    },
+    {
+      "epoch": 1.7477681257068605,
+      "grad_norm": 3.453126163810074,
+      "learning_rate": 8.21364414479352e-07,
+      "loss": 0.8949,
+      "step": 9079
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "grad_norm": 3.3416361802800747,
+      "learning_rate": 8.201273797268861e-07,
+      "loss": 0.8731,
+      "step": 9080
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.7205,
+      "step": 9080,
+      "vm_loss": 0.1937
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.6912,
+      "step": 9080,
+      "vm_loss": 0.1695
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.4405,
+      "step": 9080,
+      "vm_loss": 0.2039
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.6079,
+      "step": 9080,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.5921,
+      "step": 9080,
+      "vm_loss": 0.2101
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.4593,
+      "step": 9080,
+      "vm_loss": 0.1569
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.771,
+      "step": 9080,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.7479606323844359,
+      "lm_loss": 0.7845,
+      "step": 9080,
+      "vm_loss": 0.2342
+    },
+    {
+      "epoch": 1.748153139062011,
+      "grad_norm": 3.207164905035659,
+      "learning_rate": 8.188912373754809e-07,
+      "loss": 0.8605,
+      "step": 9081
+    },
+    {
+      "epoch": 1.7483456457395867,
+      "grad_norm": 3.1780833176196768,
+      "learning_rate": 8.176559875453105e-07,
+      "loss": 0.8461,
+      "step": 9082
+    },
+    {
+      "epoch": 1.748538152417162,
+      "grad_norm": 3.4602807663147326,
+      "learning_rate": 8.164216303564532e-07,
+      "loss": 0.8661,
+      "step": 9083
+    },
+    {
+      "epoch": 1.7487306590947373,
+      "grad_norm": 3.3004493534347956,
+      "learning_rate": 8.151881659289085e-07,
+      "loss": 0.84,
+      "step": 9084
+    },
+    {
+      "epoch": 1.7489231657723128,
+      "grad_norm": 3.1962456191630015,
+      "learning_rate": 8.139555943825772e-07,
+      "loss": 0.8219,
+      "step": 9085
+    },
+    {
+      "epoch": 1.749115672449888,
+      "grad_norm": 3.3780791355413933,
+      "learning_rate": 8.127239158372902e-07,
+      "loss": 0.8564,
+      "step": 9086
+    },
+    {
+      "epoch": 1.7493081791274636,
+      "grad_norm": 3.075108601875423,
+      "learning_rate": 8.11493130412776e-07,
+      "loss": 0.8363,
+      "step": 9087
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "grad_norm": 3.0302134505164555,
+      "learning_rate": 8.102632382286857e-07,
+      "loss": 0.8003,
+      "step": 9088
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.942,
+      "step": 9088,
+      "vm_loss": 0.2652
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.9533,
+      "step": 9088,
+      "vm_loss": 0.1399
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.7464,
+      "step": 9088,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.3457,
+      "step": 9088,
+      "vm_loss": 0.1025
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.5487,
+      "step": 9088,
+      "vm_loss": 0.1928
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.5846,
+      "step": 9088,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.4768,
+      "step": 9088,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.7495006858050388,
+      "lm_loss": 0.7956,
+      "step": 9088,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.7496931924826142,
+      "grad_norm": 3.3065935870386256,
+      "learning_rate": 8.090342394045792e-07,
+      "loss": 0.8706,
+      "step": 9089
+    },
+    {
+      "epoch": 1.7498856991601897,
+      "grad_norm": 3.260407904345146,
+      "learning_rate": 8.07806134059933e-07,
+      "loss": 0.8845,
+      "step": 9090
+    },
+    {
+      "epoch": 1.7500782058377649,
+      "grad_norm": 3.0455748294240657,
+      "learning_rate": 8.065789223141318e-07,
+      "loss": 0.7834,
+      "step": 9091
+    },
+    {
+      "epoch": 1.7502707125153405,
+      "grad_norm": 3.324375831743756,
+      "learning_rate": 8.053526042864768e-07,
+      "loss": 0.8782,
+      "step": 9092
+    },
+    {
+      "epoch": 1.7504632191929157,
+      "grad_norm": 3.268674086627398,
+      "learning_rate": 8.041271800961836e-07,
+      "loss": 0.8401,
+      "step": 9093
+    },
+    {
+      "epoch": 1.7506557258704911,
+      "grad_norm": 3.381288763297877,
+      "learning_rate": 8.029026498623771e-07,
+      "loss": 0.9171,
+      "step": 9094
+    },
+    {
+      "epoch": 1.7508482325480665,
+      "grad_norm": 3.2580216589811895,
+      "learning_rate": 8.016790137040987e-07,
+      "loss": 0.8288,
+      "step": 9095
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "grad_norm": 3.1325119015889236,
+      "learning_rate": 8.004562717403019e-07,
+      "loss": 0.7997,
+      "step": 9096
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.416,
+      "step": 9096,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.7128,
+      "step": 9096,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 1.0784,
+      "step": 9096,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.812,
+      "step": 9096,
+      "vm_loss": 0.141
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.7978,
+      "step": 9096,
+      "vm_loss": 0.203
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.6181,
+      "step": 9096,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.5631,
+      "step": 9096,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 1.7510407392256417,
+      "lm_loss": 0.9242,
+      "step": 9096,
+      "vm_loss": 0.1097
+    },
+    {
+      "epoch": 1.7512332459032174,
+      "grad_norm": 3.289409425976474,
+      "learning_rate": 7.992344240898519e-07,
+      "loss": 0.8586,
+      "step": 9097
+    },
+    {
+      "epoch": 1.7514257525807926,
+      "grad_norm": 3.2749794034205277,
+      "learning_rate": 7.98013470871527e-07,
+      "loss": 0.8573,
+      "step": 9098
+    },
+    {
+      "epoch": 1.751618259258368,
+      "grad_norm": 3.2429217630526708,
+      "learning_rate": 7.967934122040211e-07,
+      "loss": 0.8484,
+      "step": 9099
+    },
+    {
+      "epoch": 1.7518107659359434,
+      "grad_norm": 3.307285119440503,
+      "learning_rate": 7.955742482059381e-07,
+      "loss": 0.8769,
+      "step": 9100
+    },
+    {
+      "epoch": 1.7520032726135186,
+      "grad_norm": 3.200233189974417,
+      "learning_rate": 7.943559789957989e-07,
+      "loss": 0.8481,
+      "step": 9101
+    },
+    {
+      "epoch": 1.7521957792910943,
+      "grad_norm": 3.0708415896375816,
+      "learning_rate": 7.931386046920308e-07,
+      "loss": 0.8212,
+      "step": 9102
+    },
+    {
+      "epoch": 1.7523882859686695,
+      "grad_norm": 3.3267629971676937,
+      "learning_rate": 7.919221254129828e-07,
+      "loss": 0.8692,
+      "step": 9103
+    },
+    {
+      "epoch": 1.752580792646245,
+      "grad_norm": 3.087116651086049,
+      "learning_rate": 7.907065412769089e-07,
+      "loss": 0.7933,
+      "step": 9104
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.7309,
+      "step": 9104,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.6091,
+      "step": 9104,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.9113,
+      "step": 9104,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.7242,
+      "step": 9104,
+      "vm_loss": 0.2078
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 1.087,
+      "step": 9104,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.5255,
+      "step": 9104,
+      "vm_loss": 0.1175
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.4267,
+      "step": 9104,
+      "vm_loss": 0.1151
+    },
+    {
+      "epoch": 1.752580792646245,
+      "lm_loss": 0.4205,
+      "step": 9104,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.7527732993238203,
+      "grad_norm": 3.147319210056095,
+      "learning_rate": 7.894918524019801e-07,
+      "loss": 0.8489,
+      "step": 9105
+    },
+    {
+      "epoch": 1.7529658060013955,
+      "grad_norm": 3.2160016766072737,
+      "learning_rate": 7.8827805890628e-07,
+      "loss": 0.8693,
+      "step": 9106
+    },
+    {
+      "epoch": 1.7531583126789712,
+      "grad_norm": 3.3616927221777626,
+      "learning_rate": 7.870651609078072e-07,
+      "loss": 0.8685,
+      "step": 9107
+    },
+    {
+      "epoch": 1.7533508193565464,
+      "grad_norm": 3.189108107766812,
+      "learning_rate": 7.858531585244655e-07,
+      "loss": 0.8495,
+      "step": 9108
+    },
+    {
+      "epoch": 1.7535433260341218,
+      "grad_norm": 3.227291159284179,
+      "learning_rate": 7.846420518740816e-07,
+      "loss": 0.8476,
+      "step": 9109
+    },
+    {
+      "epoch": 1.7537358327116972,
+      "grad_norm": 3.404899730000119,
+      "learning_rate": 7.834318410743891e-07,
+      "loss": 0.881,
+      "step": 9110
+    },
+    {
+      "epoch": 1.7539283393892726,
+      "grad_norm": 3.2260613862778578,
+      "learning_rate": 7.822225262430372e-07,
+      "loss": 0.8168,
+      "step": 9111
+    },
+    {
+      "epoch": 1.754120846066848,
+      "grad_norm": 3.165097971887034,
+      "learning_rate": 7.810141074975819e-07,
+      "loss": 0.8567,
+      "step": 9112
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.7681,
+      "step": 9112,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.2385,
+      "step": 9112,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.6057,
+      "step": 9112,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.7278,
+      "step": 9112,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.9652,
+      "step": 9112,
+      "vm_loss": 0.1985
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 1.06,
+      "step": 9112,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.6473,
+      "step": 9112,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.754120846066848,
+      "lm_loss": 0.6349,
+      "step": 9112,
+      "vm_loss": 0.1726
+    },
+    {
+      "epoch": 1.7543133527444232,
+      "grad_norm": 3.1793618915258604,
+      "learning_rate": 7.798065849555037e-07,
+      "loss": 0.8427,
+      "step": 9113
+    },
+    {
+      "epoch": 1.7545058594219987,
+      "grad_norm": 3.3418824878281117,
+      "learning_rate": 7.785999587341853e-07,
+      "loss": 0.8785,
+      "step": 9114
+    },
+    {
+      "epoch": 1.754698366099574,
+      "grad_norm": 3.2938074926718293,
+      "learning_rate": 7.773942289509273e-07,
+      "loss": 0.8469,
+      "step": 9115
+    },
+    {
+      "epoch": 1.7548908727771495,
+      "grad_norm": 3.367803528827594,
+      "learning_rate": 7.761893957229405e-07,
+      "loss": 0.8718,
+      "step": 9116
+    },
+    {
+      "epoch": 1.755083379454725,
+      "grad_norm": 3.2042090946252206,
+      "learning_rate": 7.749854591673522e-07,
+      "loss": 0.8914,
+      "step": 9117
+    },
+    {
+      "epoch": 1.7552758861323001,
+      "grad_norm": 3.1210756403809903,
+      "learning_rate": 7.737824194012022e-07,
+      "loss": 0.8276,
+      "step": 9118
+    },
+    {
+      "epoch": 1.7554683928098755,
+      "grad_norm": 3.121507160293842,
+      "learning_rate": 7.725802765414359e-07,
+      "loss": 0.826,
+      "step": 9119
+    },
+    {
+      "epoch": 1.755660899487451,
+      "grad_norm": 3.2705283557820453,
+      "learning_rate": 7.713790307049207e-07,
+      "loss": 0.8562,
+      "step": 9120
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.8927,
+      "step": 9120,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.5068,
+      "step": 9120,
+      "vm_loss": 0.1747
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.4187,
+      "step": 9120,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.6097,
+      "step": 9120,
+      "vm_loss": 0.1271
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.909,
+      "step": 9120,
+      "vm_loss": 0.1497
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.39,
+      "step": 9120,
+      "vm_loss": 0.1131
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.5906,
+      "step": 9120,
+      "vm_loss": 0.1108
+    },
+    {
+      "epoch": 1.755660899487451,
+      "lm_loss": 0.4717,
+      "step": 9120,
+      "vm_loss": 0.1933
+    },
+    {
+      "epoch": 1.7558534061650264,
+      "grad_norm": 3.3482470951177246,
+      "learning_rate": 7.701786820084312e-07,
+      "loss": 0.8565,
+      "step": 9121
+    },
+    {
+      "epoch": 1.7560459128426018,
+      "grad_norm": 3.0913141418467007,
+      "learning_rate": 7.689792305686583e-07,
+      "loss": 0.82,
+      "step": 9122
+    },
+    {
+      "epoch": 1.756238419520177,
+      "grad_norm": 3.4185808812093708,
+      "learning_rate": 7.677806765022044e-07,
+      "loss": 0.8669,
+      "step": 9123
+    },
+    {
+      "epoch": 1.7564309261977527,
+      "grad_norm": 3.301512907921233,
+      "learning_rate": 7.665830199255842e-07,
+      "loss": 0.8959,
+      "step": 9124
+    },
+    {
+      "epoch": 1.7566234328753279,
+      "grad_norm": 3.3372726100674326,
+      "learning_rate": 7.653862609552221e-07,
+      "loss": 0.8926,
+      "step": 9125
+    },
+    {
+      "epoch": 1.7568159395529033,
+      "grad_norm": 3.202024410833478,
+      "learning_rate": 7.641903997074607e-07,
+      "loss": 0.8822,
+      "step": 9126
+    },
+    {
+      "epoch": 1.7570084462304787,
+      "grad_norm": 3.1979002553443014,
+      "learning_rate": 7.629954362985537e-07,
+      "loss": 0.8324,
+      "step": 9127
+    },
+    {
+      "epoch": 1.757200952908054,
+      "grad_norm": 3.176683496714559,
+      "learning_rate": 7.61801370844667e-07,
+      "loss": 0.8934,
+      "step": 9128
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.3902,
+      "step": 9128,
+      "vm_loss": 0.1881
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.4609,
+      "step": 9128,
+      "vm_loss": 0.1078
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.5858,
+      "step": 9128,
+      "vm_loss": 0.1753
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.7817,
+      "step": 9128,
+      "vm_loss": 0.1747
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.9017,
+      "step": 9128,
+      "vm_loss": 0.238
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.8686,
+      "step": 9128,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.9927,
+      "step": 9128,
+      "vm_loss": 0.1209
+    },
+    {
+      "epoch": 1.757200952908054,
+      "lm_loss": 0.4221,
+      "step": 9128,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 1.7573934595856295,
+      "grad_norm": 3.3587036790139146,
+      "learning_rate": 7.606082034618756e-07,
+      "loss": 0.8737,
+      "step": 9129
+    },
+    {
+      "epoch": 1.7575859662632047,
+      "grad_norm": 3.1150339691544455,
+      "learning_rate": 7.594159342661767e-07,
+      "loss": 0.8039,
+      "step": 9130
+    },
+    {
+      "epoch": 1.7577784729407802,
+      "grad_norm": 3.015202813482452,
+      "learning_rate": 7.582245633734675e-07,
+      "loss": 0.794,
+      "step": 9131
+    },
+    {
+      "epoch": 1.7579709796183556,
+      "grad_norm": 3.226497666892498,
+      "learning_rate": 7.570340908995688e-07,
+      "loss": 0.8246,
+      "step": 9132
+    },
+    {
+      "epoch": 1.7581634862959308,
+      "grad_norm": 3.3910305075678546,
+      "learning_rate": 7.55844516960208e-07,
+      "loss": 0.8955,
+      "step": 9133
+    },
+    {
+      "epoch": 1.7583559929735064,
+      "grad_norm": 3.2541383324943576,
+      "learning_rate": 7.546558416710292e-07,
+      "loss": 0.8643,
+      "step": 9134
+    },
+    {
+      "epoch": 1.7585484996510816,
+      "grad_norm": 3.290207772610275,
+      "learning_rate": 7.534680651475834e-07,
+      "loss": 0.8428,
+      "step": 9135
+    },
+    {
+      "epoch": 1.758741006328657,
+      "grad_norm": 3.3118627213867504,
+      "learning_rate": 7.522811875053392e-07,
+      "loss": 0.8406,
+      "step": 9136
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.7673,
+      "step": 9136,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.6391,
+      "step": 9136,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.9093,
+      "step": 9136,
+      "vm_loss": 0.2185
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.7018,
+      "step": 9136,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.2952,
+      "step": 9136,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.7054,
+      "step": 9136,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.7648,
+      "step": 9136,
+      "vm_loss": 0.1749
+    },
+    {
+      "epoch": 1.758741006328657,
+      "lm_loss": 0.73,
+      "step": 9136,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 1.7589335130062325,
+      "grad_norm": 3.1548456763327217,
+      "learning_rate": 7.510952088596768e-07,
+      "loss": 0.8672,
+      "step": 9137
+    },
+    {
+      "epoch": 1.7591260196838077,
+      "grad_norm": 3.162241128142295,
+      "learning_rate": 7.499101293258881e-07,
+      "loss": 0.8168,
+      "step": 9138
+    },
+    {
+      "epoch": 1.7593185263613833,
+      "grad_norm": 3.127715819120615,
+      "learning_rate": 7.487259490191778e-07,
+      "loss": 0.8166,
+      "step": 9139
+    },
+    {
+      "epoch": 1.7595110330389585,
+      "grad_norm": 3.265498768933632,
+      "learning_rate": 7.47542668054666e-07,
+      "loss": 0.8607,
+      "step": 9140
+    },
+    {
+      "epoch": 1.759703539716534,
+      "grad_norm": 3.4265180534507675,
+      "learning_rate": 7.463602865473774e-07,
+      "loss": 0.8835,
+      "step": 9141
+    },
+    {
+      "epoch": 1.7598960463941093,
+      "grad_norm": 3.4210355309762432,
+      "learning_rate": 7.451788046122588e-07,
+      "loss": 0.8894,
+      "step": 9142
+    },
+    {
+      "epoch": 1.7600885530716845,
+      "grad_norm": 3.2067604294346888,
+      "learning_rate": 7.439982223641629e-07,
+      "loss": 0.836,
+      "step": 9143
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "grad_norm": 3.2678172642125083,
+      "learning_rate": 7.428185399178601e-07,
+      "loss": 0.8891,
+      "step": 9144
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.5732,
+      "step": 9144,
+      "vm_loss": 0.2252
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.6637,
+      "step": 9144,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.9585,
+      "step": 9144,
+      "vm_loss": 0.1858
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.6224,
+      "step": 9144,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.6682,
+      "step": 9144,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.6115,
+      "step": 9144,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.5402,
+      "step": 9144,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.7602810597492602,
+      "lm_loss": 0.6333,
+      "step": 9144,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 1.7604735664268354,
+      "grad_norm": 3.406833009683654,
+      "learning_rate": 7.416397573880296e-07,
+      "loss": 0.9269,
+      "step": 9145
+    },
+    {
+      "epoch": 1.7606660731044108,
+      "grad_norm": 3.3343640162487835,
+      "learning_rate": 7.404618748892612e-07,
+      "loss": 0.879,
+      "step": 9146
+    },
+    {
+      "epoch": 1.7608585797819862,
+      "grad_norm": 3.4432827897339013,
+      "learning_rate": 7.392848925360651e-07,
+      "loss": 0.8682,
+      "step": 9147
+    },
+    {
+      "epoch": 1.7610510864595614,
+      "grad_norm": 3.1274069419447903,
+      "learning_rate": 7.381088104428558e-07,
+      "loss": 0.8222,
+      "step": 9148
+    },
+    {
+      "epoch": 1.761243593137137,
+      "grad_norm": 3.1699410686943543,
+      "learning_rate": 7.369336287239637e-07,
+      "loss": 0.8422,
+      "step": 9149
+    },
+    {
+      "epoch": 1.7614360998147123,
+      "grad_norm": 3.22319441201873,
+      "learning_rate": 7.35759347493632e-07,
+      "loss": 0.8326,
+      "step": 9150
+    },
+    {
+      "epoch": 1.7616286064922877,
+      "grad_norm": 3.06259210551004,
+      "learning_rate": 7.345859668660183e-07,
+      "loss": 0.7956,
+      "step": 9151
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "grad_norm": 3.1027046636356843,
+      "learning_rate": 7.334134869551857e-07,
+      "loss": 0.8484,
+      "step": 9152
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.5319,
+      "step": 9152,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.6245,
+      "step": 9152,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.6299,
+      "step": 9152,
+      "vm_loss": 0.2012
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.7788,
+      "step": 9152,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.4139,
+      "step": 9152,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.9143,
+      "step": 9152,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.7375,
+      "step": 9152,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.7618211131698631,
+      "lm_loss": 0.6463,
+      "step": 9152,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 1.7620136198474383,
+      "grad_norm": 3.1839745482976185,
+      "learning_rate": 7.322419078751164e-07,
+      "loss": 0.8494,
+      "step": 9153
+    },
+    {
+      "epoch": 1.762206126525014,
+      "grad_norm": 3.271344556112259,
+      "learning_rate": 7.310712297397038e-07,
+      "loss": 0.8708,
+      "step": 9154
+    },
+    {
+      "epoch": 1.7623986332025892,
+      "grad_norm": 3.255038698813587,
+      "learning_rate": 7.299014526627523e-07,
+      "loss": 0.8741,
+      "step": 9155
+    },
+    {
+      "epoch": 1.7625911398801646,
+      "grad_norm": 3.293389916610071,
+      "learning_rate": 7.287325767579756e-07,
+      "loss": 0.8431,
+      "step": 9156
+    },
+    {
+      "epoch": 1.76278364655774,
+      "grad_norm": 3.3721474026357,
+      "learning_rate": 7.275646021390103e-07,
+      "loss": 0.8681,
+      "step": 9157
+    },
+    {
+      "epoch": 1.7629761532353152,
+      "grad_norm": 3.1726378324181828,
+      "learning_rate": 7.263975289193937e-07,
+      "loss": 0.8291,
+      "step": 9158
+    },
+    {
+      "epoch": 1.7631686599128908,
+      "grad_norm": 3.3492187415204926,
+      "learning_rate": 7.252313572125802e-07,
+      "loss": 0.877,
+      "step": 9159
+    },
+    {
+      "epoch": 1.763361166590466,
+      "grad_norm": 3.222152139483039,
+      "learning_rate": 7.240660871319383e-07,
+      "loss": 0.8315,
+      "step": 9160
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.4679,
+      "step": 9160,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.8754,
+      "step": 9160,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.8859,
+      "step": 9160,
+      "vm_loss": 0.134
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 1.2349,
+      "step": 9160,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.5745,
+      "step": 9160,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.7343,
+      "step": 9160,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.7035,
+      "step": 9160,
+      "vm_loss": 0.3009
+    },
+    {
+      "epoch": 1.763361166590466,
+      "lm_loss": 0.5711,
+      "step": 9160,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.7635536732680415,
+      "grad_norm": 3.1990832920156067,
+      "learning_rate": 7.229017187907473e-07,
+      "loss": 0.8512,
+      "step": 9161
+    },
+    {
+      "epoch": 1.7637461799456169,
+      "grad_norm": 3.2216247539173155,
+      "learning_rate": 7.217382523021988e-07,
+      "loss": 0.8391,
+      "step": 9162
+    },
+    {
+      "epoch": 1.763938686623192,
+      "grad_norm": 3.2241713916064523,
+      "learning_rate": 7.205756877793935e-07,
+      "loss": 0.8453,
+      "step": 9163
+    },
+    {
+      "epoch": 1.7641311933007677,
+      "grad_norm": 3.5027491028638207,
+      "learning_rate": 7.19414025335351e-07,
+      "loss": 0.8726,
+      "step": 9164
+    },
+    {
+      "epoch": 1.764323699978343,
+      "grad_norm": 3.3271537162747262,
+      "learning_rate": 7.182532650829976e-07,
+      "loss": 0.8964,
+      "step": 9165
+    },
+    {
+      "epoch": 1.7645162066559184,
+      "grad_norm": 3.212824167357,
+      "learning_rate": 7.170934071351754e-07,
+      "loss": 0.805,
+      "step": 9166
+    },
+    {
+      "epoch": 1.7647087133334938,
+      "grad_norm": 3.0850974213135967,
+      "learning_rate": 7.159344516046373e-07,
+      "loss": 0.8255,
+      "step": 9167
+    },
+    {
+      "epoch": 1.764901220011069,
+      "grad_norm": 3.025587768556213,
+      "learning_rate": 7.147763986040502e-07,
+      "loss": 0.7902,
+      "step": 9168
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.6313,
+      "step": 9168,
+      "vm_loss": 0.2114
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.6482,
+      "step": 9168,
+      "vm_loss": 0.1188
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.5218,
+      "step": 9168,
+      "vm_loss": 0.1848
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.4549,
+      "step": 9168,
+      "vm_loss": 0.1235
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.5733,
+      "step": 9168,
+      "vm_loss": 0.1508
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.7217,
+      "step": 9168,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.3933,
+      "step": 9168,
+      "vm_loss": 0.1131
+    },
+    {
+      "epoch": 1.764901220011069,
+      "lm_loss": 0.9583,
+      "step": 9168,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.7650937266886446,
+      "grad_norm": 3.296325596838556,
+      "learning_rate": 7.136192482459881e-07,
+      "loss": 0.818,
+      "step": 9169
+    },
+    {
+      "epoch": 1.7652862333662198,
+      "grad_norm": 3.187321619404271,
+      "learning_rate": 7.124630006429423e-07,
+      "loss": 0.8105,
+      "step": 9170
+    },
+    {
+      "epoch": 1.7654787400437952,
+      "grad_norm": 3.128016239097787,
+      "learning_rate": 7.113076559073151e-07,
+      "loss": 0.8035,
+      "step": 9171
+    },
+    {
+      "epoch": 1.7656712467213707,
+      "grad_norm": 3.3196514589510953,
+      "learning_rate": 7.101532141514233e-07,
+      "loss": 0.9161,
+      "step": 9172
+    },
+    {
+      "epoch": 1.765863753398946,
+      "grad_norm": 3.224783099812493,
+      "learning_rate": 7.089996754874873e-07,
+      "loss": 0.8145,
+      "step": 9173
+    },
+    {
+      "epoch": 1.7660562600765215,
+      "grad_norm": 3.2808551712724916,
+      "learning_rate": 7.078470400276516e-07,
+      "loss": 0.8439,
+      "step": 9174
+    },
+    {
+      "epoch": 1.7662487667540967,
+      "grad_norm": 3.6337028868206334,
+      "learning_rate": 7.066953078839645e-07,
+      "loss": 0.9041,
+      "step": 9175
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "grad_norm": 3.3396603853352307,
+      "learning_rate": 7.055444791683907e-07,
+      "loss": 0.8918,
+      "step": 9176
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.5954,
+      "step": 9176,
+      "vm_loss": 0.1328
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.874,
+      "step": 9176,
+      "vm_loss": 0.1263
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 1.1141,
+      "step": 9176,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.8647,
+      "step": 9176,
+      "vm_loss": 0.1457
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 1.0597,
+      "step": 9176,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.6791,
+      "step": 9176,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.3902,
+      "step": 9176,
+      "vm_loss": 0.225
+    },
+    {
+      "epoch": 1.7664412734316721,
+      "lm_loss": 0.6888,
+      "step": 9176,
+      "vm_loss": 0.1828
+    },
+    {
+      "epoch": 1.7666337801092475,
+      "grad_norm": 3.5058409873416636,
+      "learning_rate": 7.04394553992801e-07,
+      "loss": 0.9071,
+      "step": 9177
+    },
+    {
+      "epoch": 1.766826286786823,
+      "grad_norm": 3.2895467015421724,
+      "learning_rate": 7.032455324689902e-07,
+      "loss": 0.8981,
+      "step": 9178
+    },
+    {
+      "epoch": 1.7670187934643984,
+      "grad_norm": 3.122765185094544,
+      "learning_rate": 7.020974147086512e-07,
+      "loss": 0.8289,
+      "step": 9179
+    },
+    {
+      "epoch": 1.7672113001419736,
+      "grad_norm": 3.2268881524928177,
+      "learning_rate": 7.009502008233993e-07,
+      "loss": 0.8441,
+      "step": 9180
+    },
+    {
+      "epoch": 1.767403806819549,
+      "grad_norm": 3.2682147495198013,
+      "learning_rate": 6.998038909247573e-07,
+      "loss": 0.8385,
+      "step": 9181
+    },
+    {
+      "epoch": 1.7675963134971244,
+      "grad_norm": 3.095162711052163,
+      "learning_rate": 6.986584851241607e-07,
+      "loss": 0.8298,
+      "step": 9182
+    },
+    {
+      "epoch": 1.7677888201746998,
+      "grad_norm": 3.1787740444095,
+      "learning_rate": 6.975139835329603e-07,
+      "loss": 0.8268,
+      "step": 9183
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "grad_norm": 3.3055546430014866,
+      "learning_rate": 6.963703862624104e-07,
+      "loss": 0.8479,
+      "step": 9184
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.6795,
+      "step": 9184,
+      "vm_loss": 0.2139
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.7216,
+      "step": 9184,
+      "vm_loss": 0.1312
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.696,
+      "step": 9184,
+      "vm_loss": 0.189
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.4869,
+      "step": 9184,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.501,
+      "step": 9184,
+      "vm_loss": 0.1836
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.7656,
+      "step": 9184,
+      "vm_loss": 0.2104
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.5313,
+      "step": 9184,
+      "vm_loss": 0.1525
+    },
+    {
+      "epoch": 1.7679813268522753,
+      "lm_loss": 0.7217,
+      "step": 9184,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 1.7681738335298505,
+      "grad_norm": 3.1302148405089247,
+      "learning_rate": 6.952276934236912e-07,
+      "loss": 0.8408,
+      "step": 9185
+    },
+    {
+      "epoch": 1.768366340207426,
+      "grad_norm": 3.1270879280449453,
+      "learning_rate": 6.940859051278814e-07,
+      "loss": 0.8286,
+      "step": 9186
+    },
+    {
+      "epoch": 1.7685588468850013,
+      "grad_norm": 3.2009342910364027,
+      "learning_rate": 6.92945021485979e-07,
+      "loss": 0.8318,
+      "step": 9187
+    },
+    {
+      "epoch": 1.7687513535625767,
+      "grad_norm": 3.306636940521181,
+      "learning_rate": 6.918050426088918e-07,
+      "loss": 0.8685,
+      "step": 9188
+    },
+    {
+      "epoch": 1.7689438602401522,
+      "grad_norm": 3.2144824607036644,
+      "learning_rate": 6.906659686074436e-07,
+      "loss": 0.8578,
+      "step": 9189
+    },
+    {
+      "epoch": 1.7691363669177274,
+      "grad_norm": 3.0542847184438764,
+      "learning_rate": 6.8952779959236e-07,
+      "loss": 0.8025,
+      "step": 9190
+    },
+    {
+      "epoch": 1.769328873595303,
+      "grad_norm": 3.2167422587656187,
+      "learning_rate": 6.883905356742948e-07,
+      "loss": 0.8599,
+      "step": 9191
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "grad_norm": 3.342015309415466,
+      "learning_rate": 6.872541769637964e-07,
+      "loss": 0.8941,
+      "step": 9192
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 1.0404,
+      "step": 9192,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.7462,
+      "step": 9192,
+      "vm_loss": 0.1279
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.5148,
+      "step": 9192,
+      "vm_loss": 0.2237
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.8642,
+      "step": 9192,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.6004,
+      "step": 9192,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.5082,
+      "step": 9192,
+      "vm_loss": 0.2059
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.5121,
+      "step": 9192,
+      "vm_loss": 0.1306
+    },
+    {
+      "epoch": 1.7695213802728782,
+      "lm_loss": 0.8031,
+      "step": 9192,
+      "vm_loss": 0.1235
+    },
+    {
+      "epoch": 1.7697138869504536,
+      "grad_norm": 3.2796719124059557,
+      "learning_rate": 6.861187235713395e-07,
+      "loss": 0.8339,
+      "step": 9193
+    },
+    {
+      "epoch": 1.769906393628029,
+      "grad_norm": 3.2239013025223056,
+      "learning_rate": 6.849841756072984e-07,
+      "loss": 0.8495,
+      "step": 9194
+    },
+    {
+      "epoch": 1.7700989003056042,
+      "grad_norm": 3.3967382774906145,
+      "learning_rate": 6.838505331819733e-07,
+      "loss": 0.8917,
+      "step": 9195
+    },
+    {
+      "epoch": 1.7702914069831799,
+      "grad_norm": 3.3696275921566605,
+      "learning_rate": 6.82717796405562e-07,
+      "loss": 0.8608,
+      "step": 9196
+    },
+    {
+      "epoch": 1.770483913660755,
+      "grad_norm": 3.334879719118258,
+      "learning_rate": 6.815859653881851e-07,
+      "loss": 0.8665,
+      "step": 9197
+    },
+    {
+      "epoch": 1.7706764203383305,
+      "grad_norm": 3.149483726817334,
+      "learning_rate": 6.804550402398679e-07,
+      "loss": 0.7964,
+      "step": 9198
+    },
+    {
+      "epoch": 1.770868927015906,
+      "grad_norm": 3.4199878991709376,
+      "learning_rate": 6.79325021070556e-07,
+      "loss": 0.8987,
+      "step": 9199
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "grad_norm": 3.3049947241360305,
+      "learning_rate": 6.781959079900958e-07,
+      "loss": 0.907,
+      "step": 9200
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.5237,
+      "step": 9200,
+      "vm_loss": 0.1248
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.6099,
+      "step": 9200,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.5679,
+      "step": 9200,
+      "vm_loss": 0.1314
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.7882,
+      "step": 9200,
+      "vm_loss": 0.1611
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.9503,
+      "step": 9200,
+      "vm_loss": 0.2522
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.5151,
+      "step": 9200,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.9043,
+      "step": 9200,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.7710614336934811,
+      "lm_loss": 0.8371,
+      "step": 9200,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 1.7712539403710568,
+      "grad_norm": 3.2678834259250555,
+      "learning_rate": 6.770677011082539e-07,
+      "loss": 0.8637,
+      "step": 9201
+    },
+    {
+      "epoch": 1.771446447048632,
+      "grad_norm": 3.1223605535717898,
+      "learning_rate": 6.75940400534707e-07,
+      "loss": 0.8432,
+      "step": 9202
+    },
+    {
+      "epoch": 1.7716389537262074,
+      "grad_norm": 3.150618501905829,
+      "learning_rate": 6.74814006379042e-07,
+      "loss": 0.8577,
+      "step": 9203
+    },
+    {
+      "epoch": 1.7718314604037828,
+      "grad_norm": 3.279169910201237,
+      "learning_rate": 6.736885187507602e-07,
+      "loss": 0.8826,
+      "step": 9204
+    },
+    {
+      "epoch": 1.772023967081358,
+      "grad_norm": 3.332192358964797,
+      "learning_rate": 6.72563937759273e-07,
+      "loss": 0.8396,
+      "step": 9205
+    },
+    {
+      "epoch": 1.7722164737589337,
+      "grad_norm": 3.291203924460957,
+      "learning_rate": 6.71440263513905e-07,
+      "loss": 0.8859,
+      "step": 9206
+    },
+    {
+      "epoch": 1.7724089804365089,
+      "grad_norm": 3.272812535001123,
+      "learning_rate": 6.703174961238867e-07,
+      "loss": 0.8438,
+      "step": 9207
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "grad_norm": 3.316516959507834,
+      "learning_rate": 6.691956356983731e-07,
+      "loss": 0.8884,
+      "step": 9208
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.4017,
+      "step": 9208,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.4183,
+      "step": 9208,
+      "vm_loss": 0.2164
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.9697,
+      "step": 9208,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.5544,
+      "step": 9208,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.8673,
+      "step": 9208,
+      "vm_loss": 0.1584
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.8152,
+      "step": 9208,
+      "vm_loss": 0.2003
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.4683,
+      "step": 9208,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.7726014871140843,
+      "lm_loss": 0.7664,
+      "step": 9208,
+      "vm_loss": 0.2055
+    },
+    {
+      "epoch": 1.7727939937916597,
+      "grad_norm": 3.14522934926373,
+      "learning_rate": 6.680746823464179e-07,
+      "loss": 0.8285,
+      "step": 9209
+    },
+    {
+      "epoch": 1.772986500469235,
+      "grad_norm": 3.136181613504236,
+      "learning_rate": 6.669546361769952e-07,
+      "loss": 0.8155,
+      "step": 9210
+    },
+    {
+      "epoch": 1.7731790071468105,
+      "grad_norm": 3.164323045322968,
+      "learning_rate": 6.658354972989833e-07,
+      "loss": 0.8205,
+      "step": 9211
+    },
+    {
+      "epoch": 1.7733715138243857,
+      "grad_norm": 3.310678509658945,
+      "learning_rate": 6.647172658211831e-07,
+      "loss": 0.9001,
+      "step": 9212
+    },
+    {
+      "epoch": 1.7735640205019612,
+      "grad_norm": 3.283355496278776,
+      "learning_rate": 6.635999418522965e-07,
+      "loss": 0.8544,
+      "step": 9213
+    },
+    {
+      "epoch": 1.7737565271795366,
+      "grad_norm": 3.2715304259435842,
+      "learning_rate": 6.624835255009432e-07,
+      "loss": 0.8594,
+      "step": 9214
+    },
+    {
+      "epoch": 1.7739490338571118,
+      "grad_norm": 3.227025813962184,
+      "learning_rate": 6.61368016875652e-07,
+      "loss": 0.8573,
+      "step": 9215
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "grad_norm": 3.380766274591073,
+      "learning_rate": 6.602534160848673e-07,
+      "loss": 0.872,
+      "step": 9216
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.5537,
+      "step": 9216,
+      "vm_loss": 0.1527
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.9421,
+      "step": 9216,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.4653,
+      "step": 9216,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.75,
+      "step": 9216,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.6928,
+      "step": 9216,
+      "vm_loss": 0.217
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 1.0293,
+      "step": 9216,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.9449,
+      "step": 9216,
+      "vm_loss": 0.1759
+    },
+    {
+      "epoch": 1.7741415405346874,
+      "lm_loss": 0.7624,
+      "step": 9216,
+      "vm_loss": 0.1071
+    },
+    {
+      "epoch": 1.7743340472122626,
+      "grad_norm": 3.378035724851111,
+      "learning_rate": 6.591397232369401e-07,
+      "loss": 0.9128,
+      "step": 9217
+    },
+    {
+      "epoch": 1.774526553889838,
+      "grad_norm": 3.1840416733891233,
+      "learning_rate": 6.58026938440135e-07,
+      "loss": 0.8548,
+      "step": 9218
+    },
+    {
+      "epoch": 1.7747190605674135,
+      "grad_norm": 3.362688991557411,
+      "learning_rate": 6.56915061802631e-07,
+      "loss": 0.8555,
+      "step": 9219
+    },
+    {
+      "epoch": 1.7749115672449887,
+      "grad_norm": 3.341396127842984,
+      "learning_rate": 6.558040934325183e-07,
+      "loss": 0.8593,
+      "step": 9220
+    },
+    {
+      "epoch": 1.7751040739225643,
+      "grad_norm": 3.189012700305559,
+      "learning_rate": 6.546940334377905e-07,
+      "loss": 0.8442,
+      "step": 9221
+    },
+    {
+      "epoch": 1.7752965806001395,
+      "grad_norm": 3.129851218420475,
+      "learning_rate": 6.535848819263679e-07,
+      "loss": 0.8633,
+      "step": 9222
+    },
+    {
+      "epoch": 1.775489087277715,
+      "grad_norm": 3.211557572174544,
+      "learning_rate": 6.524766390060677e-07,
+      "loss": 0.8528,
+      "step": 9223
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "grad_norm": 3.0312822354676623,
+      "learning_rate": 6.513693047846292e-07,
+      "loss": 0.7963,
+      "step": 9224
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.6698,
+      "step": 9224,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.5015,
+      "step": 9224,
+      "vm_loss": 0.1241
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.6723,
+      "step": 9224,
+      "vm_loss": 0.1547
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 1.2025,
+      "step": 9224,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.6682,
+      "step": 9224,
+      "vm_loss": 0.2748
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.7978,
+      "step": 9224,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.8342,
+      "step": 9224,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.7756815939552903,
+      "lm_loss": 0.6983,
+      "step": 9224,
+      "vm_loss": 0.1317
+    },
+    {
+      "epoch": 1.7758741006328655,
+      "grad_norm": 3.3300521550975035,
+      "learning_rate": 6.502628793696974e-07,
+      "loss": 0.8522,
+      "step": 9225
+    },
+    {
+      "epoch": 1.7760666073104412,
+      "grad_norm": 3.0977194561261965,
+      "learning_rate": 6.49157362868832e-07,
+      "loss": 0.7942,
+      "step": 9226
+    },
+    {
+      "epoch": 1.7762591139880164,
+      "grad_norm": 3.280793552503617,
+      "learning_rate": 6.480527553895055e-07,
+      "loss": 0.8396,
+      "step": 9227
+    },
+    {
+      "epoch": 1.7764516206655918,
+      "grad_norm": 3.143475244861368,
+      "learning_rate": 6.469490570390935e-07,
+      "loss": 0.822,
+      "step": 9228
+    },
+    {
+      "epoch": 1.7766441273431672,
+      "grad_norm": 3.176365129783468,
+      "learning_rate": 6.458462679248967e-07,
+      "loss": 0.8596,
+      "step": 9229
+    },
+    {
+      "epoch": 1.7768366340207424,
+      "grad_norm": 3.202366959012867,
+      "learning_rate": 6.44744388154116e-07,
+      "loss": 0.8279,
+      "step": 9230
+    },
+    {
+      "epoch": 1.777029140698318,
+      "grad_norm": 3.302952553004911,
+      "learning_rate": 6.436434178338691e-07,
+      "loss": 0.8249,
+      "step": 9231
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "grad_norm": 3.300311748020996,
+      "learning_rate": 6.425433570711848e-07,
+      "loss": 0.8362,
+      "step": 9232
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.6182,
+      "step": 9232,
+      "vm_loss": 0.2116
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.8127,
+      "step": 9232,
+      "vm_loss": 0.2078
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.5137,
+      "step": 9232,
+      "vm_loss": 0.1126
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.5472,
+      "step": 9232,
+      "vm_loss": 0.208
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.786,
+      "step": 9232,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.8349,
+      "step": 9232,
+      "vm_loss": 0.1728
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.65,
+      "step": 9232,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 1.7772216473758933,
+      "lm_loss": 0.8333,
+      "step": 9232,
+      "vm_loss": 0.1157
+    },
+    {
+      "epoch": 1.7774141540534687,
+      "grad_norm": 3.2229492400800837,
+      "learning_rate": 6.414442059730052e-07,
+      "loss": 0.8047,
+      "step": 9233
+    },
+    {
+      "epoch": 1.7776066607310441,
+      "grad_norm": 3.3264864929381126,
+      "learning_rate": 6.403459646461774e-07,
+      "loss": 0.8802,
+      "step": 9234
+    },
+    {
+      "epoch": 1.7777991674086195,
+      "grad_norm": 3.1954841826774105,
+      "learning_rate": 6.392486331974668e-07,
+      "loss": 0.8072,
+      "step": 9235
+    },
+    {
+      "epoch": 1.777991674086195,
+      "grad_norm": 3.128595033969252,
+      "learning_rate": 6.381522117335482e-07,
+      "loss": 0.8409,
+      "step": 9236
+    },
+    {
+      "epoch": 1.7781841807637702,
+      "grad_norm": 3.115331844638848,
+      "learning_rate": 6.370567003610095e-07,
+      "loss": 0.7891,
+      "step": 9237
+    },
+    {
+      "epoch": 1.7783766874413456,
+      "grad_norm": 3.3320659979008167,
+      "learning_rate": 6.359620991863425e-07,
+      "loss": 0.8697,
+      "step": 9238
+    },
+    {
+      "epoch": 1.778569194118921,
+      "grad_norm": 3.1837519072221894,
+      "learning_rate": 6.348684083159651e-07,
+      "loss": 0.8624,
+      "step": 9239
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "grad_norm": 3.2117770108508,
+      "learning_rate": 6.337756278561902e-07,
+      "loss": 0.8624,
+      "step": 9240
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.7106,
+      "step": 9240,
+      "vm_loss": 0.2332
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.525,
+      "step": 9240,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.4679,
+      "step": 9240,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.575,
+      "step": 9240,
+      "vm_loss": 0.1001
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.6907,
+      "step": 9240,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.7037,
+      "step": 9240,
+      "vm_loss": 0.1974
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 1.2433,
+      "step": 9240,
+      "vm_loss": 0.1215
+    },
+    {
+      "epoch": 1.7787617007964964,
+      "lm_loss": 0.4828,
+      "step": 9240,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.7789542074740718,
+      "grad_norm": 3.2706895491636736,
+      "learning_rate": 6.326837579132538e-07,
+      "loss": 0.8557,
+      "step": 9241
+    },
+    {
+      "epoch": 1.779146714151647,
+      "grad_norm": 3.2307334036245505,
+      "learning_rate": 6.315927985933001e-07,
+      "loss": 0.8041,
+      "step": 9242
+    },
+    {
+      "epoch": 1.7793392208292225,
+      "grad_norm": 3.084748508457976,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.8194,
+      "step": 9243
+    },
+    {
+      "epoch": 1.779531727506798,
+      "grad_norm": 3.079293432116286,
+      "learning_rate": 6.294136122464701e-07,
+      "loss": 0.8057,
+      "step": 9244
+    },
+    {
+      "epoch": 1.7797242341843733,
+      "grad_norm": 3.271410006086443,
+      "learning_rate": 6.283253854314376e-07,
+      "loss": 0.841,
+      "step": 9245
+    },
+    {
+      "epoch": 1.7799167408619487,
+      "grad_norm": 3.008547302609984,
+      "learning_rate": 6.272380696630765e-07,
+      "loss": 0.7755,
+      "step": 9246
+    },
+    {
+      "epoch": 1.780109247539524,
+      "grad_norm": 3.238833149115958,
+      "learning_rate": 6.261516650470878e-07,
+      "loss": 0.8393,
+      "step": 9247
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "grad_norm": 3.225089958380521,
+      "learning_rate": 6.250661716890838e-07,
+      "loss": 0.838,
+      "step": 9248
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.5288,
+      "step": 9248,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.7326,
+      "step": 9248,
+      "vm_loss": 0.2099
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.8235,
+      "step": 9248,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 1.0753,
+      "step": 9248,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.7818,
+      "step": 9248,
+      "vm_loss": 0.1624
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.466,
+      "step": 9248,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.641,
+      "step": 9248,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 1.7803017542170994,
+      "lm_loss": 0.7657,
+      "step": 9248,
+      "vm_loss": 0.1393
+    },
+    {
+      "epoch": 1.7804942608946748,
+      "grad_norm": 3.2228340982875765,
+      "learning_rate": 6.239815896945889e-07,
+      "loss": 0.8208,
+      "step": 9249
+    },
+    {
+      "epoch": 1.7806867675722502,
+      "grad_norm": 3.436525681666964,
+      "learning_rate": 6.228979191690387e-07,
+      "loss": 0.881,
+      "step": 9250
+    },
+    {
+      "epoch": 1.7808792742498256,
+      "grad_norm": 3.404937019759645,
+      "learning_rate": 6.218151602177769e-07,
+      "loss": 0.8889,
+      "step": 9251
+    },
+    {
+      "epoch": 1.7810717809274008,
+      "grad_norm": 3.245815431998424,
+      "learning_rate": 6.207333129460646e-07,
+      "loss": 0.8631,
+      "step": 9252
+    },
+    {
+      "epoch": 1.7812642876049765,
+      "grad_norm": 3.2823055949593254,
+      "learning_rate": 6.19652377459069e-07,
+      "loss": 0.8614,
+      "step": 9253
+    },
+    {
+      "epoch": 1.7814567942825517,
+      "grad_norm": 3.2936101358401655,
+      "learning_rate": 6.185723538618738e-07,
+      "loss": 0.8592,
+      "step": 9254
+    },
+    {
+      "epoch": 1.781649300960127,
+      "grad_norm": 3.382432999164009,
+      "learning_rate": 6.17493242259467e-07,
+      "loss": 0.8886,
+      "step": 9255
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "grad_norm": 3.111526722908233,
+      "learning_rate": 6.164150427567572e-07,
+      "loss": 0.7941,
+      "step": 9256
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.3625,
+      "step": 9256,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.5348,
+      "step": 9256,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.7382,
+      "step": 9256,
+      "vm_loss": 0.1522
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.8101,
+      "step": 9256,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.4439,
+      "step": 9256,
+      "vm_loss": 0.1513
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.8838,
+      "step": 9256,
+      "vm_loss": 0.1205
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.547,
+      "step": 9256,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.7818418076377025,
+      "lm_loss": 0.7169,
+      "step": 9256,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 1.7820343143152777,
+      "grad_norm": 3.404613808835494,
+      "learning_rate": 6.153377554585549e-07,
+      "loss": 0.8366,
+      "step": 9257
+    },
+    {
+      "epoch": 1.7822268209928533,
+      "grad_norm": 3.054785570234413,
+      "learning_rate": 6.142613804695885e-07,
+      "loss": 0.801,
+      "step": 9258
+    },
+    {
+      "epoch": 1.7824193276704285,
+      "grad_norm": 3.33749363987345,
+      "learning_rate": 6.131859178944922e-07,
+      "loss": 0.8981,
+      "step": 9259
+    },
+    {
+      "epoch": 1.782611834348004,
+      "grad_norm": 3.2496571440326005,
+      "learning_rate": 6.121113678378199e-07,
+      "loss": 0.8918,
+      "step": 9260
+    },
+    {
+      "epoch": 1.7828043410255794,
+      "grad_norm": 3.1280889463683104,
+      "learning_rate": 6.110377304040271e-07,
+      "loss": 0.8239,
+      "step": 9261
+    },
+    {
+      "epoch": 1.7829968477031546,
+      "grad_norm": 3.222917663309125,
+      "learning_rate": 6.09965005697487e-07,
+      "loss": 0.834,
+      "step": 9262
+    },
+    {
+      "epoch": 1.7831893543807302,
+      "grad_norm": 3.3265008147886315,
+      "learning_rate": 6.088931938224818e-07,
+      "loss": 0.8709,
+      "step": 9263
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "grad_norm": 3.1697446634346984,
+      "learning_rate": 6.078222948832058e-07,
+      "loss": 0.8191,
+      "step": 9264
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.5779,
+      "step": 9264,
+      "vm_loss": 0.1964
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.7652,
+      "step": 9264,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 1.0081,
+      "step": 9264,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.637,
+      "step": 9264,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.6511,
+      "step": 9264,
+      "vm_loss": 0.2324
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.8456,
+      "step": 9264,
+      "vm_loss": 0.1051
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 0.6702,
+      "step": 9264,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 1.7833818610583054,
+      "lm_loss": 1.1316,
+      "step": 9264,
+      "vm_loss": 0.2092
+    },
+    {
+      "epoch": 1.7835743677358809,
+      "grad_norm": 3.343411901700475,
+      "learning_rate": 6.067523089837635e-07,
+      "loss": 0.87,
+      "step": 9265
+    },
+    {
+      "epoch": 1.7837668744134563,
+      "grad_norm": 3.208265545162121,
+      "learning_rate": 6.056832362281728e-07,
+      "loss": 0.8293,
+      "step": 9266
+    },
+    {
+      "epoch": 1.7839593810910315,
+      "grad_norm": 3.2865771003767184,
+      "learning_rate": 6.046150767203585e-07,
+      "loss": 0.8311,
+      "step": 9267
+    },
+    {
+      "epoch": 1.7841518877686071,
+      "grad_norm": 3.2451078515808587,
+      "learning_rate": 6.035478305641617e-07,
+      "loss": 0.8504,
+      "step": 9268
+    },
+    {
+      "epoch": 1.7843443944461823,
+      "grad_norm": 3.239829724525075,
+      "learning_rate": 6.024814978633309e-07,
+      "loss": 0.8371,
+      "step": 9269
+    },
+    {
+      "epoch": 1.7845369011237577,
+      "grad_norm": 3.3536802046613094,
+      "learning_rate": 6.014160787215273e-07,
+      "loss": 0.8698,
+      "step": 9270
+    },
+    {
+      "epoch": 1.7847294078013332,
+      "grad_norm": 3.2248312329713555,
+      "learning_rate": 6.00351573242326e-07,
+      "loss": 0.8366,
+      "step": 9271
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "grad_norm": 3.421789033489498,
+      "learning_rate": 5.992879815292052e-07,
+      "loss": 0.9078,
+      "step": 9272
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.6616,
+      "step": 9272,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.5141,
+      "step": 9272,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.7851,
+      "step": 9272,
+      "vm_loss": 0.1931
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.5386,
+      "step": 9272,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.4687,
+      "step": 9272,
+      "vm_loss": 0.1556
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.9167,
+      "step": 9272,
+      "vm_loss": 0.1988
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.5668,
+      "step": 9272,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 1.7849219144789084,
+      "lm_loss": 0.8033,
+      "step": 9272,
+      "vm_loss": 0.1814
+    },
+    {
+      "epoch": 1.785114421156484,
+      "grad_norm": 3.1573575071062043,
+      "learning_rate": 5.982253036855656e-07,
+      "loss": 0.814,
+      "step": 9273
+    },
+    {
+      "epoch": 1.7853069278340592,
+      "grad_norm": 3.296680516675027,
+      "learning_rate": 5.971635398147091e-07,
+      "loss": 0.9145,
+      "step": 9274
+    },
+    {
+      "epoch": 1.7854994345116346,
+      "grad_norm": 3.0551457943636358,
+      "learning_rate": 5.961026900198564e-07,
+      "loss": 0.814,
+      "step": 9275
+    },
+    {
+      "epoch": 1.78569194118921,
+      "grad_norm": 3.1344385936130648,
+      "learning_rate": 5.950427544041293e-07,
+      "loss": 0.8094,
+      "step": 9276
+    },
+    {
+      "epoch": 1.7858844478667852,
+      "grad_norm": 3.5087652720375533,
+      "learning_rate": 5.939837330705733e-07,
+      "loss": 0.9412,
+      "step": 9277
+    },
+    {
+      "epoch": 1.7860769545443609,
+      "grad_norm": 3.1947165381298914,
+      "learning_rate": 5.929256261221361e-07,
+      "loss": 0.8211,
+      "step": 9278
+    },
+    {
+      "epoch": 1.786269461221936,
+      "grad_norm": 3.206555448247763,
+      "learning_rate": 5.918684336616787e-07,
+      "loss": 0.8466,
+      "step": 9279
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "grad_norm": 3.2183513926517504,
+      "learning_rate": 5.908121557919755e-07,
+      "loss": 0.824,
+      "step": 9280
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.8104,
+      "step": 9280,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.4222,
+      "step": 9280,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.9618,
+      "step": 9280,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.6091,
+      "step": 9280,
+      "vm_loss": 0.1
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.4865,
+      "step": 9280,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.5688,
+      "step": 9280,
+      "vm_loss": 0.1086
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.6236,
+      "step": 9280,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 1.7864619678995115,
+      "lm_loss": 0.9341,
+      "step": 9280,
+      "vm_loss": 0.2305
+    },
+    {
+      "epoch": 1.786654474577087,
+      "grad_norm": 3.215835781174425,
+      "learning_rate": 5.89756792615711e-07,
+      "loss": 0.8329,
+      "step": 9281
+    },
+    {
+      "epoch": 1.7868469812546621,
+      "grad_norm": 3.1170676980342606,
+      "learning_rate": 5.887023442354744e-07,
+      "loss": 0.8774,
+      "step": 9282
+    },
+    {
+      "epoch": 1.7870394879322378,
+      "grad_norm": 3.144025160494419,
+      "learning_rate": 5.876488107537792e-07,
+      "loss": 0.8312,
+      "step": 9283
+    },
+    {
+      "epoch": 1.787231994609813,
+      "grad_norm": 3.3151663693662865,
+      "learning_rate": 5.865961922730379e-07,
+      "loss": 0.8499,
+      "step": 9284
+    },
+    {
+      "epoch": 1.7874245012873884,
+      "grad_norm": 3.2966878742167944,
+      "learning_rate": 5.855444888955786e-07,
+      "loss": 0.8708,
+      "step": 9285
+    },
+    {
+      "epoch": 1.7876170079649638,
+      "grad_norm": 3.2192937950295186,
+      "learning_rate": 5.844937007236407e-07,
+      "loss": 0.8419,
+      "step": 9286
+    },
+    {
+      "epoch": 1.787809514642539,
+      "grad_norm": 3.2902553220441986,
+      "learning_rate": 5.834438278593768e-07,
+      "loss": 0.8622,
+      "step": 9287
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "grad_norm": 3.242364040778866,
+      "learning_rate": 5.823948704048443e-07,
+      "loss": 0.8617,
+      "step": 9288
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 1.1926,
+      "step": 9288,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.4136,
+      "step": 9288,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.933,
+      "step": 9288,
+      "vm_loss": 0.1863
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.8711,
+      "step": 9288,
+      "vm_loss": 0.1747
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.601,
+      "step": 9288,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.6969,
+      "step": 9288,
+      "vm_loss": 0.2151
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.7399,
+      "step": 9288,
+      "vm_loss": 0.2037
+    },
+    {
+      "epoch": 1.7880020213201147,
+      "lm_loss": 0.5307,
+      "step": 9288,
+      "vm_loss": 0.1029
+    },
+    {
+      "epoch": 1.7881945279976899,
+      "grad_norm": 3.2932063243846765,
+      "learning_rate": 5.813468284620161e-07,
+      "loss": 0.8655,
+      "step": 9289
+    },
+    {
+      "epoch": 1.7883870346752653,
+      "grad_norm": 3.3271120922929445,
+      "learning_rate": 5.802997021327761e-07,
+      "loss": 0.8659,
+      "step": 9290
+    },
+    {
+      "epoch": 1.7885795413528407,
+      "grad_norm": 3.2895643066155533,
+      "learning_rate": 5.792534915189185e-07,
+      "loss": 0.8566,
+      "step": 9291
+    },
+    {
+      "epoch": 1.788772048030416,
+      "grad_norm": 3.3427730531311877,
+      "learning_rate": 5.782081967221498e-07,
+      "loss": 0.893,
+      "step": 9292
+    },
+    {
+      "epoch": 1.7889645547079915,
+      "grad_norm": 3.32998639109651,
+      "learning_rate": 5.771638178440808e-07,
+      "loss": 0.8702,
+      "step": 9293
+    },
+    {
+      "epoch": 1.7891570613855667,
+      "grad_norm": 3.217540227518655,
+      "learning_rate": 5.761203549862448e-07,
+      "loss": 0.8015,
+      "step": 9294
+    },
+    {
+      "epoch": 1.7893495680631422,
+      "grad_norm": 3.4286631419637397,
+      "learning_rate": 5.750778082500763e-07,
+      "loss": 0.9207,
+      "step": 9295
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "grad_norm": 3.273074842613617,
+      "learning_rate": 5.740361777369241e-07,
+      "loss": 0.8824,
+      "step": 9296
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.8196,
+      "step": 9296,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.9049,
+      "step": 9296,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.6444,
+      "step": 9296,
+      "vm_loss": 0.1304
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.5757,
+      "step": 9296,
+      "vm_loss": 0.1851
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.7011,
+      "step": 9296,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.4708,
+      "step": 9296,
+      "vm_loss": 0.2166
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.7206,
+      "step": 9296,
+      "vm_loss": 0.1647
+    },
+    {
+      "epoch": 1.7895420747407176,
+      "lm_loss": 0.8499,
+      "step": 9296,
+      "vm_loss": 0.1208
+    },
+    {
+      "epoch": 1.789734581418293,
+      "grad_norm": 3.2785972409992574,
+      "learning_rate": 5.729954635480483e-07,
+      "loss": 0.8562,
+      "step": 9297
+    },
+    {
+      "epoch": 1.7899270880958684,
+      "grad_norm": 3.2755376569906614,
+      "learning_rate": 5.719556657846226e-07,
+      "loss": 0.8112,
+      "step": 9298
+    },
+    {
+      "epoch": 1.7901195947734436,
+      "grad_norm": 3.2293600704366994,
+      "learning_rate": 5.709167845477225e-07,
+      "loss": 0.8297,
+      "step": 9299
+    },
+    {
+      "epoch": 1.790312101451019,
+      "grad_norm": 3.193441751466621,
+      "learning_rate": 5.698788199383476e-07,
+      "loss": 0.8165,
+      "step": 9300
+    },
+    {
+      "epoch": 1.7905046081285945,
+      "grad_norm": 3.1143475417157154,
+      "learning_rate": 5.688417720573958e-07,
+      "loss": 0.7916,
+      "step": 9301
+    },
+    {
+      "epoch": 1.7906971148061699,
+      "grad_norm": 3.1972935909874187,
+      "learning_rate": 5.678056410056854e-07,
+      "loss": 0.852,
+      "step": 9302
+    },
+    {
+      "epoch": 1.7908896214837453,
+      "grad_norm": 3.177973552713865,
+      "learning_rate": 5.667704268839358e-07,
+      "loss": 0.8133,
+      "step": 9303
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "grad_norm": 3.2822464398528273,
+      "learning_rate": 5.65736129792791e-07,
+      "loss": 0.829,
+      "step": 9304
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.8761,
+      "step": 9304,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.8798,
+      "step": 9304,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.7196,
+      "step": 9304,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.8113,
+      "step": 9304,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.7959,
+      "step": 9304,
+      "vm_loss": 0.2356
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.7527,
+      "step": 9304,
+      "vm_loss": 0.166
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.7275,
+      "step": 9304,
+      "vm_loss": 0.1917
+    },
+    {
+      "epoch": 1.7910821281613205,
+      "lm_loss": 0.629,
+      "step": 9304,
+      "vm_loss": 0.1013
+    },
+    {
+      "epoch": 1.791274634838896,
+      "grad_norm": 3.3303301003602193,
+      "learning_rate": 5.647027498327918e-07,
+      "loss": 0.9016,
+      "step": 9305
+    },
+    {
+      "epoch": 1.7914671415164714,
+      "grad_norm": 3.1263861161841198,
+      "learning_rate": 5.636702871043986e-07,
+      "loss": 0.8131,
+      "step": 9306
+    },
+    {
+      "epoch": 1.7916596481940468,
+      "grad_norm": 3.206954243730365,
+      "learning_rate": 5.626387417079793e-07,
+      "loss": 0.8302,
+      "step": 9307
+    },
+    {
+      "epoch": 1.7918521548716222,
+      "grad_norm": 3.2779989352350056,
+      "learning_rate": 5.616081137438123e-07,
+      "loss": 0.8658,
+      "step": 9308
+    },
+    {
+      "epoch": 1.7920446615491974,
+      "grad_norm": 3.1289923026812803,
+      "learning_rate": 5.60578403312092e-07,
+      "loss": 0.7966,
+      "step": 9309
+    },
+    {
+      "epoch": 1.7922371682267728,
+      "grad_norm": 3.27722554765375,
+      "learning_rate": 5.595496105129139e-07,
+      "loss": 0.8499,
+      "step": 9310
+    },
+    {
+      "epoch": 1.7924296749043482,
+      "grad_norm": 3.4300968204868525,
+      "learning_rate": 5.585217354462924e-07,
+      "loss": 0.8888,
+      "step": 9311
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "grad_norm": 3.1934477494434863,
+      "learning_rate": 5.574947782121498e-07,
+      "loss": 0.8468,
+      "step": 9312
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.7531,
+      "step": 9312,
+      "vm_loss": 0.1311
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.6114,
+      "step": 9312,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.589,
+      "step": 9312,
+      "vm_loss": 0.1282
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.3324,
+      "step": 9312,
+      "vm_loss": 0.0991
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.9235,
+      "step": 9312,
+      "vm_loss": 0.162
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.75,
+      "step": 9312,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.5836,
+      "step": 9312,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.7926221815819237,
+      "lm_loss": 0.5484,
+      "step": 9312,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.792814688259499,
+      "grad_norm": 3.12519481115312,
+      "learning_rate": 5.564687389103207e-07,
+      "loss": 0.801,
+      "step": 9313
+    },
+    {
+      "epoch": 1.7930071949370743,
+      "grad_norm": 3.3483189553720942,
+      "learning_rate": 5.554436176405475e-07,
+      "loss": 0.9047,
+      "step": 9314
+    },
+    {
+      "epoch": 1.79319970161465,
+      "grad_norm": 3.284115775753295,
+      "learning_rate": 5.544194145024872e-07,
+      "loss": 0.8744,
+      "step": 9315
+    },
+    {
+      "epoch": 1.7933922082922251,
+      "grad_norm": 3.136461803652932,
+      "learning_rate": 5.533961295957013e-07,
+      "loss": 0.8326,
+      "step": 9316
+    },
+    {
+      "epoch": 1.7935847149698005,
+      "grad_norm": 3.372705292566454,
+      "learning_rate": 5.523737630196713e-07,
+      "loss": 0.8911,
+      "step": 9317
+    },
+    {
+      "epoch": 1.793777221647376,
+      "grad_norm": 3.3737858585394456,
+      "learning_rate": 5.513523148737809e-07,
+      "loss": 0.8638,
+      "step": 9318
+    },
+    {
+      "epoch": 1.7939697283249512,
+      "grad_norm": 3.345638230096811,
+      "learning_rate": 5.503317852573309e-07,
+      "loss": 0.8607,
+      "step": 9319
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "grad_norm": 3.4251893513047853,
+      "learning_rate": 5.493121742695251e-07,
+      "loss": 0.897,
+      "step": 9320
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.419,
+      "step": 9320,
+      "vm_loss": 0.1827
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.7904,
+      "step": 9320,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.664,
+      "step": 9320,
+      "vm_loss": 0.1323
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.919,
+      "step": 9320,
+      "vm_loss": 0.1976
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.6742,
+      "step": 9320,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.5391,
+      "step": 9320,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.7375,
+      "step": 9320,
+      "vm_loss": 0.2275
+    },
+    {
+      "epoch": 1.7941622350025268,
+      "lm_loss": 0.5958,
+      "step": 9320,
+      "vm_loss": 0.2044
+    },
+    {
+      "epoch": 1.794354741680102,
+      "grad_norm": 3.092477120183323,
+      "learning_rate": 5.482934820094876e-07,
+      "loss": 0.8398,
+      "step": 9321
+    },
+    {
+      "epoch": 1.7945472483576774,
+      "grad_norm": 3.0705010731287135,
+      "learning_rate": 5.472757085762459e-07,
+      "loss": 0.786,
+      "step": 9322
+    },
+    {
+      "epoch": 1.7947397550352528,
+      "grad_norm": 3.2965471832632542,
+      "learning_rate": 5.462588540687397e-07,
+      "loss": 0.8568,
+      "step": 9323
+    },
+    {
+      "epoch": 1.794932261712828,
+      "grad_norm": 3.184817560153111,
+      "learning_rate": 5.452429185858221e-07,
+      "loss": 0.8206,
+      "step": 9324
+    },
+    {
+      "epoch": 1.7951247683904037,
+      "grad_norm": 3.260190325186805,
+      "learning_rate": 5.442279022262564e-07,
+      "loss": 0.822,
+      "step": 9325
+    },
+    {
+      "epoch": 1.795317275067979,
+      "grad_norm": 3.2684083878219803,
+      "learning_rate": 5.432138050887115e-07,
+      "loss": 0.8149,
+      "step": 9326
+    },
+    {
+      "epoch": 1.7955097817455543,
+      "grad_norm": 3.321153270924032,
+      "learning_rate": 5.422006272717717e-07,
+      "loss": 0.8805,
+      "step": 9327
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "grad_norm": 2.979764277324341,
+      "learning_rate": 5.411883688739328e-07,
+      "loss": 0.7504,
+      "step": 9328
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.4891,
+      "step": 9328,
+      "vm_loss": 0.2018
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.5723,
+      "step": 9328,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.4893,
+      "step": 9328,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.6068,
+      "step": 9328,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.3643,
+      "step": 9328,
+      "vm_loss": 0.1396
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.5364,
+      "step": 9328,
+      "vm_loss": 0.1163
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.8323,
+      "step": 9328,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.7957022884231297,
+      "lm_loss": 0.6345,
+      "step": 9328,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.795894795100705,
+      "grad_norm": 3.3951599398942456,
+      "learning_rate": 5.401770299935971e-07,
+      "loss": 0.8778,
+      "step": 9329
+    },
+    {
+      "epoch": 1.7960873017782806,
+      "grad_norm": 3.2020205994392303,
+      "learning_rate": 5.391666107290805e-07,
+      "loss": 0.8502,
+      "step": 9330
+    },
+    {
+      "epoch": 1.7962798084558558,
+      "grad_norm": 3.1091143028082864,
+      "learning_rate": 5.381571111786099e-07,
+      "loss": 0.7952,
+      "step": 9331
+    },
+    {
+      "epoch": 1.7964723151334312,
+      "grad_norm": 3.3343196348526667,
+      "learning_rate": 5.371485314403202e-07,
+      "loss": 0.863,
+      "step": 9332
+    },
+    {
+      "epoch": 1.7966648218110066,
+      "grad_norm": 3.200568165098887,
+      "learning_rate": 5.361408716122585e-07,
+      "loss": 0.8341,
+      "step": 9333
+    },
+    {
+      "epoch": 1.7968573284885818,
+      "grad_norm": 3.0434407330028757,
+      "learning_rate": 5.351341317923808e-07,
+      "loss": 0.7836,
+      "step": 9334
+    },
+    {
+      "epoch": 1.7970498351661575,
+      "grad_norm": 3.104722092073949,
+      "learning_rate": 5.341283120785579e-07,
+      "loss": 0.8149,
+      "step": 9335
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "grad_norm": 3.3548735604886257,
+      "learning_rate": 5.331234125685669e-07,
+      "loss": 0.8617,
+      "step": 9336
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.6442,
+      "step": 9336,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.4212,
+      "step": 9336,
+      "vm_loss": 0.2187
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.3811,
+      "step": 9336,
+      "vm_loss": 0.1698
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.5379,
+      "step": 9336,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.8259,
+      "step": 9336,
+      "vm_loss": 0.1428
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.4868,
+      "step": 9336,
+      "vm_loss": 0.1211
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 1.1915,
+      "step": 9336,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.7972423418437327,
+      "lm_loss": 0.4599,
+      "step": 9336,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 1.797434848521308,
+      "grad_norm": 3.1718756111237236,
+      "learning_rate": 5.321194333600954e-07,
+      "loss": 0.8357,
+      "step": 9337
+    },
+    {
+      "epoch": 1.7976273551988835,
+      "grad_norm": 3.2873528593143133,
+      "learning_rate": 5.311163745507464e-07,
+      "loss": 0.8655,
+      "step": 9338
+    },
+    {
+      "epoch": 1.7978198618764587,
+      "grad_norm": 3.2003733281005515,
+      "learning_rate": 5.301142362380274e-07,
+      "loss": 0.8433,
+      "step": 9339
+    },
+    {
+      "epoch": 1.7980123685540343,
+      "grad_norm": 3.3807919046718626,
+      "learning_rate": 5.291130185193594e-07,
+      "loss": 0.8485,
+      "step": 9340
+    },
+    {
+      "epoch": 1.7982048752316095,
+      "grad_norm": 3.4129025834983295,
+      "learning_rate": 5.281127214920745e-07,
+      "loss": 0.8408,
+      "step": 9341
+    },
+    {
+      "epoch": 1.798397381909185,
+      "grad_norm": 3.353193767271289,
+      "learning_rate": 5.271133452534149e-07,
+      "loss": 0.8932,
+      "step": 9342
+    },
+    {
+      "epoch": 1.7985898885867604,
+      "grad_norm": 3.361393517157605,
+      "learning_rate": 5.261148899005297e-07,
+      "loss": 0.8576,
+      "step": 9343
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "grad_norm": 3.035842363424599,
+      "learning_rate": 5.251173555304845e-07,
+      "loss": 0.7771,
+      "step": 9344
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.8957,
+      "step": 9344,
+      "vm_loss": 0.1135
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.6786,
+      "step": 9344,
+      "vm_loss": 0.1625
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.7046,
+      "step": 9344,
+      "vm_loss": 0.1689
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.4745,
+      "step": 9344,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.5686,
+      "step": 9344,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.6368,
+      "step": 9344,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.5009,
+      "step": 9344,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.7987823952643356,
+      "lm_loss": 0.3397,
+      "step": 9344,
+      "vm_loss": 0.1093
+    },
+    {
+      "epoch": 1.7989749019419112,
+      "grad_norm": 3.238650458103736,
+      "learning_rate": 5.241207422402516e-07,
+      "loss": 0.7828,
+      "step": 9345
+    },
+    {
+      "epoch": 1.7991674086194864,
+      "grad_norm": 3.313445480953595,
+      "learning_rate": 5.23125050126716e-07,
+      "loss": 0.849,
+      "step": 9346
+    },
+    {
+      "epoch": 1.7993599152970619,
+      "grad_norm": 3.1550807482936607,
+      "learning_rate": 5.221302792866667e-07,
+      "loss": 0.8189,
+      "step": 9347
+    },
+    {
+      "epoch": 1.7995524219746373,
+      "grad_norm": 3.4208320095160074,
+      "learning_rate": 5.211364298168142e-07,
+      "loss": 0.8748,
+      "step": 9348
+    },
+    {
+      "epoch": 1.7997449286522125,
+      "grad_norm": 3.3066552919313557,
+      "learning_rate": 5.201435018137701e-07,
+      "loss": 0.8006,
+      "step": 9349
+    },
+    {
+      "epoch": 1.7999374353297881,
+      "grad_norm": 3.2370739221568248,
+      "learning_rate": 5.191514953740606e-07,
+      "loss": 0.8029,
+      "step": 9350
+    },
+    {
+      "epoch": 1.8001299420073633,
+      "grad_norm": 3.328435341538871,
+      "learning_rate": 5.181604105941207e-07,
+      "loss": 0.8484,
+      "step": 9351
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "grad_norm": 3.205933465828315,
+      "learning_rate": 5.171702475702967e-07,
+      "loss": 0.8239,
+      "step": 9352
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.4288,
+      "step": 9352,
+      "vm_loss": 0.1995
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.4346,
+      "step": 9352,
+      "vm_loss": 0.1706
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.5045,
+      "step": 9352,
+      "vm_loss": 0.0999
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.8446,
+      "step": 9352,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.6973,
+      "step": 9352,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.6798,
+      "step": 9352,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.9099,
+      "step": 9352,
+      "vm_loss": 0.1963
+    },
+    {
+      "epoch": 1.8003224486849387,
+      "lm_loss": 0.7159,
+      "step": 9352,
+      "vm_loss": 0.2359
+    },
+    {
+      "epoch": 1.8005149553625142,
+      "grad_norm": 3.0828030257960544,
+      "learning_rate": 5.16181006398847e-07,
+      "loss": 0.8273,
+      "step": 9353
+    },
+    {
+      "epoch": 1.8007074620400894,
+      "grad_norm": 3.3210035423085715,
+      "learning_rate": 5.15192687175935e-07,
+      "loss": 0.8551,
+      "step": 9354
+    },
+    {
+      "epoch": 1.800899968717665,
+      "grad_norm": 3.2880128154178903,
+      "learning_rate": 5.142052899976413e-07,
+      "loss": 0.8247,
+      "step": 9355
+    },
+    {
+      "epoch": 1.8010924753952402,
+      "grad_norm": 3.220242376130876,
+      "learning_rate": 5.132188149599526e-07,
+      "loss": 0.854,
+      "step": 9356
+    },
+    {
+      "epoch": 1.8012849820728156,
+      "grad_norm": 3.0951067978855797,
+      "learning_rate": 5.122332621587645e-07,
+      "loss": 0.8165,
+      "step": 9357
+    },
+    {
+      "epoch": 1.801477488750391,
+      "grad_norm": 3.389623384896271,
+      "learning_rate": 5.11248631689889e-07,
+      "loss": 0.8737,
+      "step": 9358
+    },
+    {
+      "epoch": 1.8016699954279662,
+      "grad_norm": 3.3559336549277643,
+      "learning_rate": 5.102649236490432e-07,
+      "loss": 0.85,
+      "step": 9359
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "grad_norm": 3.4458790244766506,
+      "learning_rate": 5.092821381318558e-07,
+      "loss": 0.8883,
+      "step": 9360
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.5459,
+      "step": 9360,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.7529,
+      "step": 9360,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.8306,
+      "step": 9360,
+      "vm_loss": 0.0763
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.441,
+      "step": 9360,
+      "vm_loss": 0.1237
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.833,
+      "step": 9360,
+      "vm_loss": 0.1329
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.6061,
+      "step": 9360,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.7371,
+      "step": 9360,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.8018625021055419,
+      "lm_loss": 0.7924,
+      "step": 9360,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 1.802055008783117,
+      "grad_norm": 3.2184758421167348,
+      "learning_rate": 5.083002752338651e-07,
+      "loss": 0.8368,
+      "step": 9361
+    },
+    {
+      "epoch": 1.8022475154606925,
+      "grad_norm": 3.2145464809813142,
+      "learning_rate": 5.073193350505224e-07,
+      "loss": 0.816,
+      "step": 9362
+    },
+    {
+      "epoch": 1.802440022138268,
+      "grad_norm": 3.3602811439505893,
+      "learning_rate": 5.063393176771891e-07,
+      "loss": 0.8698,
+      "step": 9363
+    },
+    {
+      "epoch": 1.8026325288158433,
+      "grad_norm": 3.2685173005618493,
+      "learning_rate": 5.053602232091304e-07,
+      "loss": 0.8383,
+      "step": 9364
+    },
+    {
+      "epoch": 1.8028250354934188,
+      "grad_norm": 3.391957322297027,
+      "learning_rate": 5.043820517415321e-07,
+      "loss": 0.8373,
+      "step": 9365
+    },
+    {
+      "epoch": 1.803017542170994,
+      "grad_norm": 3.0841682065647857,
+      "learning_rate": 5.034048033694816e-07,
+      "loss": 0.8386,
+      "step": 9366
+    },
+    {
+      "epoch": 1.8032100488485694,
+      "grad_norm": 3.1118568574819685,
+      "learning_rate": 5.024284781879829e-07,
+      "loss": 0.8053,
+      "step": 9367
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "grad_norm": 3.113797492775831,
+      "learning_rate": 5.014530762919423e-07,
+      "loss": 0.817,
+      "step": 9368
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.9392,
+      "step": 9368,
+      "vm_loss": 0.1126
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.6223,
+      "step": 9368,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.5048,
+      "step": 9368,
+      "vm_loss": 0.1284
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.4922,
+      "step": 9368,
+      "vm_loss": 0.1764
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 1.0926,
+      "step": 9368,
+      "vm_loss": 0.1826
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.5785,
+      "step": 9368,
+      "vm_loss": 0.1987
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.6226,
+      "step": 9368,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 1.8034025555261448,
+      "lm_loss": 0.7299,
+      "step": 9368,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.8035950622037202,
+      "grad_norm": 3.280044274119619,
+      "learning_rate": 5.004785977761884e-07,
+      "loss": 0.8224,
+      "step": 9369
+    },
+    {
+      "epoch": 1.8037875688812957,
+      "grad_norm": 3.195232612770297,
+      "learning_rate": 4.995050427354464e-07,
+      "loss": 0.806,
+      "step": 9370
+    },
+    {
+      "epoch": 1.8039800755588709,
+      "grad_norm": 3.426690814831063,
+      "learning_rate": 4.98532411264362e-07,
+      "loss": 0.947,
+      "step": 9371
+    },
+    {
+      "epoch": 1.8041725822364463,
+      "grad_norm": 3.3469601396181567,
+      "learning_rate": 4.975607034574859e-07,
+      "loss": 0.8482,
+      "step": 9372
+    },
+    {
+      "epoch": 1.8043650889140217,
+      "grad_norm": 3.297804878638026,
+      "learning_rate": 4.965899194092816e-07,
+      "loss": 0.8558,
+      "step": 9373
+    },
+    {
+      "epoch": 1.8045575955915971,
+      "grad_norm": 3.2765246519035367,
+      "learning_rate": 4.956200592141213e-07,
+      "loss": 0.8691,
+      "step": 9374
+    },
+    {
+      "epoch": 1.8047501022691725,
+      "grad_norm": 3.1839352497095272,
+      "learning_rate": 4.946511229662876e-07,
+      "loss": 0.8421,
+      "step": 9375
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "grad_norm": 3.3267236640530213,
+      "learning_rate": 4.936831107599749e-07,
+      "loss": 0.8471,
+      "step": 9376
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 1.0876,
+      "step": 9376,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.5483,
+      "step": 9376,
+      "vm_loss": 0.206
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.5349,
+      "step": 9376,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.9603,
+      "step": 9376,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.4511,
+      "step": 9376,
+      "vm_loss": 0.1387
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.3138,
+      "step": 9376,
+      "vm_loss": 0.1944
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.615,
+      "step": 9376,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 1.8049426089467477,
+      "lm_loss": 0.451,
+      "step": 9376,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.8051351156243234,
+      "grad_norm": 3.090837298837418,
+      "learning_rate": 4.927160226892835e-07,
+      "loss": 0.8196,
+      "step": 9377
+    },
+    {
+      "epoch": 1.8053276223018986,
+      "grad_norm": 3.1949496974142386,
+      "learning_rate": 4.917498588482295e-07,
+      "loss": 0.82,
+      "step": 9378
+    },
+    {
+      "epoch": 1.805520128979474,
+      "grad_norm": 3.193427470900945,
+      "learning_rate": 4.907846193307342e-07,
+      "loss": 0.8013,
+      "step": 9379
+    },
+    {
+      "epoch": 1.8057126356570494,
+      "grad_norm": 3.1994504460145645,
+      "learning_rate": 4.898203042306337e-07,
+      "loss": 0.843,
+      "step": 9380
+    },
+    {
+      "epoch": 1.8059051423346246,
+      "grad_norm": 3.086406355213787,
+      "learning_rate": 4.888569136416687e-07,
+      "loss": 0.8159,
+      "step": 9381
+    },
+    {
+      "epoch": 1.8060976490122003,
+      "grad_norm": 3.250327656138653,
+      "learning_rate": 4.878944476574965e-07,
+      "loss": 0.847,
+      "step": 9382
+    },
+    {
+      "epoch": 1.8062901556897755,
+      "grad_norm": 3.380748805639859,
+      "learning_rate": 4.869329063716776e-07,
+      "loss": 0.9055,
+      "step": 9383
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "grad_norm": 3.272211286399803,
+      "learning_rate": 4.859722898776898e-07,
+      "loss": 0.8352,
+      "step": 9384
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.5525,
+      "step": 9384,
+      "vm_loss": 0.085
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.5594,
+      "step": 9384,
+      "vm_loss": 0.13
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.6448,
+      "step": 9384,
+      "vm_loss": 0.1511
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.4487,
+      "step": 9384,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.6913,
+      "step": 9384,
+      "vm_loss": 0.2148
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 1.022,
+      "step": 9384,
+      "vm_loss": 0.2565
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.7156,
+      "step": 9384,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.8064826623673509,
+      "lm_loss": 0.2478,
+      "step": 9384,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.8066751690449263,
+      "grad_norm": 3.1930833416047504,
+      "learning_rate": 4.850125982689125e-07,
+      "loss": 0.8275,
+      "step": 9385
+    },
+    {
+      "epoch": 1.8068676757225015,
+      "grad_norm": 3.28107320588364,
+      "learning_rate": 4.840538316386457e-07,
+      "loss": 0.84,
+      "step": 9386
+    },
+    {
+      "epoch": 1.8070601824000772,
+      "grad_norm": 3.2106028022349378,
+      "learning_rate": 4.830959900800902e-07,
+      "loss": 0.8351,
+      "step": 9387
+    },
+    {
+      "epoch": 1.8072526890776524,
+      "grad_norm": 3.2673671297868485,
+      "learning_rate": 4.821390736863607e-07,
+      "loss": 0.8328,
+      "step": 9388
+    },
+    {
+      "epoch": 1.8074451957552278,
+      "grad_norm": 3.1604949019357265,
+      "learning_rate": 4.811830825504815e-07,
+      "loss": 0.805,
+      "step": 9389
+    },
+    {
+      "epoch": 1.8076377024328032,
+      "grad_norm": 3.215626338988843,
+      "learning_rate": 4.802280167653906e-07,
+      "loss": 0.8289,
+      "step": 9390
+    },
+    {
+      "epoch": 1.8078302091103784,
+      "grad_norm": 3.22590524519466,
+      "learning_rate": 4.792738764239269e-07,
+      "loss": 0.8681,
+      "step": 9391
+    },
+    {
+      "epoch": 1.808022715787954,
+      "grad_norm": 3.117350840472475,
+      "learning_rate": 4.783206616188496e-07,
+      "loss": 0.776,
+      "step": 9392
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.5703,
+      "step": 9392,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.5303,
+      "step": 9392,
+      "vm_loss": 0.0931
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.4471,
+      "step": 9392,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.5546,
+      "step": 9392,
+      "vm_loss": 0.1388
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.7287,
+      "step": 9392,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.6972,
+      "step": 9392,
+      "vm_loss": 0.1829
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.4452,
+      "step": 9392,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 1.808022715787954,
+      "lm_loss": 0.6672,
+      "step": 9392,
+      "vm_loss": 0.0943
+    },
+    {
+      "epoch": 1.8082152224655292,
+      "grad_norm": 3.328999970921193,
+      "learning_rate": 4.773683724428213e-07,
+      "loss": 0.8301,
+      "step": 9393
+    },
+    {
+      "epoch": 1.8084077291431047,
+      "grad_norm": 3.255319707033454,
+      "learning_rate": 4.764170089884179e-07,
+      "loss": 0.8164,
+      "step": 9394
+    },
+    {
+      "epoch": 1.80860023582068,
+      "grad_norm": 3.2216255086780863,
+      "learning_rate": 4.7546657134812213e-07,
+      "loss": 0.8269,
+      "step": 9395
+    },
+    {
+      "epoch": 1.8087927424982553,
+      "grad_norm": 3.291390470375985,
+      "learning_rate": 4.7451705961433115e-07,
+      "loss": 0.8629,
+      "step": 9396
+    },
+    {
+      "epoch": 1.808985249175831,
+      "grad_norm": 3.0083219781833357,
+      "learning_rate": 4.735684738793489e-07,
+      "loss": 0.7825,
+      "step": 9397
+    },
+    {
+      "epoch": 1.8091777558534061,
+      "grad_norm": 3.1930037153849566,
+      "learning_rate": 4.7262081423538717e-07,
+      "loss": 0.8528,
+      "step": 9398
+    },
+    {
+      "epoch": 1.8093702625309815,
+      "grad_norm": 3.2793185056427148,
+      "learning_rate": 4.7167408077457676e-07,
+      "loss": 0.8776,
+      "step": 9399
+    },
+    {
+      "epoch": 1.809562769208557,
+      "grad_norm": 3.3941752540386196,
+      "learning_rate": 4.7072827358894626e-07,
+      "loss": 0.8912,
+      "step": 9400
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.3304,
+      "step": 9400,
+      "vm_loss": 0.1011
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.8845,
+      "step": 9400,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.9758,
+      "step": 9400,
+      "vm_loss": 0.1765
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.5514,
+      "step": 9400,
+      "vm_loss": 0.1865
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.9839,
+      "step": 9400,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.7445,
+      "step": 9400,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.5048,
+      "step": 9400,
+      "vm_loss": 0.2239
+    },
+    {
+      "epoch": 1.809562769208557,
+      "lm_loss": 0.3279,
+      "step": 9400,
+      "vm_loss": 0.1502
+    },
+    {
+      "epoch": 1.8097552758861322,
+      "grad_norm": 3.2830472742796197,
+      "learning_rate": 4.6978339277044427e-07,
+      "loss": 0.8475,
+      "step": 9401
+    },
+    {
+      "epoch": 1.8099477825637078,
+      "grad_norm": 3.067944874843008,
+      "learning_rate": 4.688394384109207e-07,
+      "loss": 0.7828,
+      "step": 9402
+    },
+    {
+      "epoch": 1.810140289241283,
+      "grad_norm": 3.3273113417882887,
+      "learning_rate": 4.678964106021455e-07,
+      "loss": 0.8227,
+      "step": 9403
+    },
+    {
+      "epoch": 1.8103327959188584,
+      "grad_norm": 3.2706842106191383,
+      "learning_rate": 4.6695430943578865e-07,
+      "loss": 0.857,
+      "step": 9404
+    },
+    {
+      "epoch": 1.8105253025964338,
+      "grad_norm": 3.128044813767231,
+      "learning_rate": 4.660131350034369e-07,
+      "loss": 0.8037,
+      "step": 9405
+    },
+    {
+      "epoch": 1.810717809274009,
+      "grad_norm": 3.220675583801118,
+      "learning_rate": 4.6507288739658264e-07,
+      "loss": 0.8216,
+      "step": 9406
+    },
+    {
+      "epoch": 1.8109103159515847,
+      "grad_norm": 3.3167907751594847,
+      "learning_rate": 4.641335667066338e-07,
+      "loss": 0.8782,
+      "step": 9407
+    },
+    {
+      "epoch": 1.81110282262916,
+      "grad_norm": 3.266160835205097,
+      "learning_rate": 4.6319517302489957e-07,
+      "loss": 0.8574,
+      "step": 9408
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.4952,
+      "step": 9408,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 1.0528,
+      "step": 9408,
+      "vm_loss": 0.2168
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.4697,
+      "step": 9408,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.7696,
+      "step": 9408,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.4651,
+      "step": 9408,
+      "vm_loss": 0.1734
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.4699,
+      "step": 9408,
+      "vm_loss": 0.1992
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.4982,
+      "step": 9408,
+      "vm_loss": 0.1253
+    },
+    {
+      "epoch": 1.81110282262916,
+      "lm_loss": 0.8186,
+      "step": 9408,
+      "vm_loss": 0.1847
+    },
+    {
+      "epoch": 1.8112953293067353,
+      "grad_norm": 3.2946795730545384,
+      "learning_rate": 4.6225770644260595e-07,
+      "loss": 0.883,
+      "step": 9409
+    },
+    {
+      "epoch": 1.8114878359843107,
+      "grad_norm": 3.1772531443367757,
+      "learning_rate": 4.6132116705088657e-07,
+      "loss": 0.8252,
+      "step": 9410
+    },
+    {
+      "epoch": 1.811680342661886,
+      "grad_norm": 3.231380433652586,
+      "learning_rate": 4.603855549407854e-07,
+      "loss": 0.8472,
+      "step": 9411
+    },
+    {
+      "epoch": 1.8118728493394616,
+      "grad_norm": 3.374213779641825,
+      "learning_rate": 4.5945087020325517e-07,
+      "loss": 0.8649,
+      "step": 9412
+    },
+    {
+      "epoch": 1.8120653560170368,
+      "grad_norm": 3.2444244359321845,
+      "learning_rate": 4.58517112929161e-07,
+      "loss": 0.8248,
+      "step": 9413
+    },
+    {
+      "epoch": 1.8122578626946122,
+      "grad_norm": 3.4150854253433702,
+      "learning_rate": 4.575842832092736e-07,
+      "loss": 0.8831,
+      "step": 9414
+    },
+    {
+      "epoch": 1.8124503693721876,
+      "grad_norm": 3.2238964268883374,
+      "learning_rate": 4.566523811342771e-07,
+      "loss": 0.8321,
+      "step": 9415
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "grad_norm": 3.2174936967269288,
+      "learning_rate": 4.5572140679476463e-07,
+      "loss": 0.8352,
+      "step": 9416
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.4912,
+      "step": 9416,
+      "vm_loss": 0.1057
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.4811,
+      "step": 9416,
+      "vm_loss": 0.172
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.75,
+      "step": 9416,
+      "vm_loss": 0.2226
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.5795,
+      "step": 9416,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.4087,
+      "step": 9416,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.6448,
+      "step": 9416,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.916,
+      "step": 9416,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.8126428760497628,
+      "lm_loss": 0.7205,
+      "step": 9416,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 1.8128353827273385,
+      "grad_norm": 3.0536734160637335,
+      "learning_rate": 4.5479136028123704e-07,
+      "loss": 0.7988,
+      "step": 9417
+    },
+    {
+      "epoch": 1.8130278894049137,
+      "grad_norm": 3.2131334457063887,
+      "learning_rate": 4.5386224168411097e-07,
+      "loss": 0.8364,
+      "step": 9418
+    },
+    {
+      "epoch": 1.813220396082489,
+      "grad_norm": 3.2724589199458913,
+      "learning_rate": 4.529340510937019e-07,
+      "loss": 0.8358,
+      "step": 9419
+    },
+    {
+      "epoch": 1.8134129027600645,
+      "grad_norm": 3.353461949216105,
+      "learning_rate": 4.520067886002488e-07,
+      "loss": 0.8724,
+      "step": 9420
+    },
+    {
+      "epoch": 1.8136054094376397,
+      "grad_norm": 3.3036144466552217,
+      "learning_rate": 4.5108045429388956e-07,
+      "loss": 0.8574,
+      "step": 9421
+    },
+    {
+      "epoch": 1.8137979161152153,
+      "grad_norm": 3.2005032457145903,
+      "learning_rate": 4.501550482646755e-07,
+      "loss": 0.8792,
+      "step": 9422
+    },
+    {
+      "epoch": 1.8139904227927905,
+      "grad_norm": 3.2556618639109933,
+      "learning_rate": 4.4923057060256903e-07,
+      "loss": 0.8406,
+      "step": 9423
+    },
+    {
+      "epoch": 1.814182929470366,
+      "grad_norm": 3.2181270573242378,
+      "learning_rate": 4.4830702139744386e-07,
+      "loss": 0.8199,
+      "step": 9424
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.3965,
+      "step": 9424,
+      "vm_loss": 0.114
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.9325,
+      "step": 9424,
+      "vm_loss": 0.1475
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.5873,
+      "step": 9424,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.7009,
+      "step": 9424,
+      "vm_loss": 0.248
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.8435,
+      "step": 9424,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.5541,
+      "step": 9424,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.7144,
+      "step": 9424,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.814182929470366,
+      "lm_loss": 0.6195,
+      "step": 9424,
+      "vm_loss": 0.1474
+    },
+    {
+      "epoch": 1.8143754361479414,
+      "grad_norm": 3.340226311419131,
+      "learning_rate": 4.4738440073907484e-07,
+      "loss": 0.865,
+      "step": 9425
+    },
+    {
+      "epoch": 1.8145679428255168,
+      "grad_norm": 3.2507799288874555,
+      "learning_rate": 4.4646270871715693e-07,
+      "loss": 0.8509,
+      "step": 9426
+    },
+    {
+      "epoch": 1.8147604495030922,
+      "grad_norm": 3.300074590968059,
+      "learning_rate": 4.4554194542128837e-07,
+      "loss": 0.8662,
+      "step": 9427
+    },
+    {
+      "epoch": 1.8149529561806674,
+      "grad_norm": 3.3049519225319464,
+      "learning_rate": 4.446221109409821e-07,
+      "loss": 0.8622,
+      "step": 9428
+    },
+    {
+      "epoch": 1.8151454628582429,
+      "grad_norm": 3.2942704775606124,
+      "learning_rate": 4.43703205365652e-07,
+      "loss": 0.881,
+      "step": 9429
+    },
+    {
+      "epoch": 1.8153379695358183,
+      "grad_norm": 3.356351412684281,
+      "learning_rate": 4.427852287846346e-07,
+      "loss": 0.8809,
+      "step": 9430
+    },
+    {
+      "epoch": 1.8155304762133937,
+      "grad_norm": 3.2733802979010242,
+      "learning_rate": 4.418681812871639e-07,
+      "loss": 0.8267,
+      "step": 9431
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "grad_norm": 3.3050535855163377,
+      "learning_rate": 4.409520629623898e-07,
+      "loss": 0.8543,
+      "step": 9432
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.638,
+      "step": 9432,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 1.0204,
+      "step": 9432,
+      "vm_loss": 0.2284
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.5644,
+      "step": 9432,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.4665,
+      "step": 9432,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.4625,
+      "step": 9432,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.7166,
+      "step": 9432,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.7052,
+      "step": 9432,
+      "vm_loss": 0.1666
+    },
+    {
+      "epoch": 1.8157229828909691,
+      "lm_loss": 0.7929,
+      "step": 9432,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.8159154895685443,
+      "grad_norm": 3.222408781441603,
+      "learning_rate": 4.400368738993721e-07,
+      "loss": 0.8449,
+      "step": 9433
+    },
+    {
+      "epoch": 1.8161079962461197,
+      "grad_norm": 3.266409053698989,
+      "learning_rate": 4.391226141870786e-07,
+      "loss": 0.8457,
+      "step": 9434
+    },
+    {
+      "epoch": 1.8163005029236952,
+      "grad_norm": 3.298062890561418,
+      "learning_rate": 4.3820928391438477e-07,
+      "loss": 0.86,
+      "step": 9435
+    },
+    {
+      "epoch": 1.8164930096012706,
+      "grad_norm": 3.260219123108118,
+      "learning_rate": 4.3729688317007966e-07,
+      "loss": 0.8366,
+      "step": 9436
+    },
+    {
+      "epoch": 1.816685516278846,
+      "grad_norm": 3.3936910604462884,
+      "learning_rate": 4.363854120428601e-07,
+      "loss": 0.8892,
+      "step": 9437
+    },
+    {
+      "epoch": 1.8168780229564212,
+      "grad_norm": 3.2660571408186656,
+      "learning_rate": 4.35474870621333e-07,
+      "loss": 0.843,
+      "step": 9438
+    },
+    {
+      "epoch": 1.8170705296339968,
+      "grad_norm": 3.389533328473793,
+      "learning_rate": 4.3456525899401414e-07,
+      "loss": 0.8682,
+      "step": 9439
+    },
+    {
+      "epoch": 1.817263036311572,
+      "grad_norm": 3.067503475867899,
+      "learning_rate": 4.336565772493306e-07,
+      "loss": 0.7984,
+      "step": 9440
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.6951,
+      "step": 9440,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.6871,
+      "step": 9440,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.7169,
+      "step": 9440,
+      "vm_loss": 0.2062
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.7186,
+      "step": 9440,
+      "vm_loss": 0.1236
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.8027,
+      "step": 9440,
+      "vm_loss": 0.1118
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.904,
+      "step": 9440,
+      "vm_loss": 0.1952
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.3963,
+      "step": 9440,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 1.817263036311572,
+      "lm_loss": 0.818,
+      "step": 9440,
+      "vm_loss": 0.1539
+    },
+    {
+      "epoch": 1.8174555429891475,
+      "grad_norm": 3.331958164811603,
+      "learning_rate": 4.3274882547561734e-07,
+      "loss": 0.8464,
+      "step": 9441
+    },
+    {
+      "epoch": 1.8176480496667229,
+      "grad_norm": 3.1949212223058687,
+      "learning_rate": 4.318420037611182e-07,
+      "loss": 0.8341,
+      "step": 9442
+    },
+    {
+      "epoch": 1.817840556344298,
+      "grad_norm": 3.284468477427103,
+      "learning_rate": 4.3093611219398923e-07,
+      "loss": 0.8487,
+      "step": 9443
+    },
+    {
+      "epoch": 1.8180330630218737,
+      "grad_norm": 3.1956671405439945,
+      "learning_rate": 4.300311508622945e-07,
+      "loss": 0.8388,
+      "step": 9444
+    },
+    {
+      "epoch": 1.818225569699449,
+      "grad_norm": 3.173443300358134,
+      "learning_rate": 4.291271198540081e-07,
+      "loss": 0.8241,
+      "step": 9445
+    },
+    {
+      "epoch": 1.8184180763770244,
+      "grad_norm": 3.187716817338898,
+      "learning_rate": 4.2822401925700973e-07,
+      "loss": 0.8037,
+      "step": 9446
+    },
+    {
+      "epoch": 1.8186105830545998,
+      "grad_norm": 3.4068845747360985,
+      "learning_rate": 4.273218491590991e-07,
+      "loss": 0.8658,
+      "step": 9447
+    },
+    {
+      "epoch": 1.818803089732175,
+      "grad_norm": 3.2799445475725535,
+      "learning_rate": 4.264206096479728e-07,
+      "loss": 0.8527,
+      "step": 9448
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 1.0119,
+      "step": 9448,
+      "vm_loss": 0.209
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.4209,
+      "step": 9448,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.6887,
+      "step": 9448,
+      "vm_loss": 0.1445
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.3712,
+      "step": 9448,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.3663,
+      "step": 9448,
+      "vm_loss": 0.183
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.7808,
+      "step": 9448,
+      "vm_loss": 0.1236
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.5196,
+      "step": 9448,
+      "vm_loss": 0.2027
+    },
+    {
+      "epoch": 1.818803089732175,
+      "lm_loss": 0.3506,
+      "step": 9448,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.8189955964097506,
+      "grad_norm": 3.240357748082226,
+      "learning_rate": 4.2552030081124627e-07,
+      "loss": 0.8317,
+      "step": 9449
+    },
+    {
+      "epoch": 1.8191881030873258,
+      "grad_norm": 3.1713314013418206,
+      "learning_rate": 4.246209227364395e-07,
+      "loss": 0.8174,
+      "step": 9450
+    },
+    {
+      "epoch": 1.8193806097649012,
+      "grad_norm": 3.26215107528053,
+      "learning_rate": 4.2372247551098477e-07,
+      "loss": 0.8379,
+      "step": 9451
+    },
+    {
+      "epoch": 1.8195731164424767,
+      "grad_norm": 3.1984297910279693,
+      "learning_rate": 4.228249592222211e-07,
+      "loss": 0.7654,
+      "step": 9452
+    },
+    {
+      "epoch": 1.8197656231200519,
+      "grad_norm": 3.265262614965891,
+      "learning_rate": 4.2192837395739874e-07,
+      "loss": 0.8143,
+      "step": 9453
+    },
+    {
+      "epoch": 1.8199581297976275,
+      "grad_norm": 3.314015175924182,
+      "learning_rate": 4.210327198036779e-07,
+      "loss": 0.8942,
+      "step": 9454
+    },
+    {
+      "epoch": 1.8201506364752027,
+      "grad_norm": 3.181754635117066,
+      "learning_rate": 4.201379968481267e-07,
+      "loss": 0.8382,
+      "step": 9455
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "grad_norm": 3.271114711349186,
+      "learning_rate": 4.1924420517772567e-07,
+      "loss": 0.8319,
+      "step": 9456
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.7772,
+      "step": 9456,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.6097,
+      "step": 9456,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.568,
+      "step": 9456,
+      "vm_loss": 0.2546
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 1.1385,
+      "step": 9456,
+      "vm_loss": 0.2128
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.5089,
+      "step": 9456,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.3091,
+      "step": 9456,
+      "vm_loss": 0.1582
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.9031,
+      "step": 9456,
+      "vm_loss": 0.2212
+    },
+    {
+      "epoch": 1.8203431431527781,
+      "lm_loss": 0.4583,
+      "step": 9456,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.8205356498303535,
+      "grad_norm": 3.1912558163715032,
+      "learning_rate": 4.183513448793619e-07,
+      "loss": 0.8401,
+      "step": 9457
+    },
+    {
+      "epoch": 1.8207281565079287,
+      "grad_norm": 3.2081449124148573,
+      "learning_rate": 4.1745941603983155e-07,
+      "loss": 0.8489,
+      "step": 9458
+    },
+    {
+      "epoch": 1.8209206631855044,
+      "grad_norm": 3.3046801919679387,
+      "learning_rate": 4.1656841874584296e-07,
+      "loss": 0.8571,
+      "step": 9459
+    },
+    {
+      "epoch": 1.8211131698630796,
+      "grad_norm": 3.4789007492680213,
+      "learning_rate": 4.156783530840114e-07,
+      "loss": 0.9039,
+      "step": 9460
+    },
+    {
+      "epoch": 1.821305676540655,
+      "grad_norm": 3.199002017346473,
+      "learning_rate": 4.147892191408631e-07,
+      "loss": 0.8576,
+      "step": 9461
+    },
+    {
+      "epoch": 1.8214981832182304,
+      "grad_norm": 3.1372556741105897,
+      "learning_rate": 4.139010170028346e-07,
+      "loss": 0.8119,
+      "step": 9462
+    },
+    {
+      "epoch": 1.8216906898958056,
+      "grad_norm": 3.2893646961238825,
+      "learning_rate": 4.130137467562667e-07,
+      "loss": 0.8986,
+      "step": 9463
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "grad_norm": 3.1882090770691525,
+      "learning_rate": 4.121274084874194e-07,
+      "loss": 0.7992,
+      "step": 9464
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.6522,
+      "step": 9464,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.7543,
+      "step": 9464,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.5803,
+      "step": 9464,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.521,
+      "step": 9464,
+      "vm_loss": 0.1924
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.7159,
+      "step": 9464,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.749,
+      "step": 9464,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.9796,
+      "step": 9464,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 1.8218831965733813,
+      "lm_loss": 0.516,
+      "step": 9464,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.8220757032509565,
+      "grad_norm": 3.1491455088349714,
+      "learning_rate": 4.1124200228245037e-07,
+      "loss": 0.8443,
+      "step": 9465
+    },
+    {
+      "epoch": 1.822268209928532,
+      "grad_norm": 3.190344975667269,
+      "learning_rate": 4.1035752822743524e-07,
+      "loss": 0.8145,
+      "step": 9466
+    },
+    {
+      "epoch": 1.8224607166061073,
+      "grad_norm": 3.3399938832626135,
+      "learning_rate": 4.0947398640835636e-07,
+      "loss": 0.8749,
+      "step": 9467
+    },
+    {
+      "epoch": 1.8226532232836825,
+      "grad_norm": 3.1419700470562497,
+      "learning_rate": 4.0859137691110495e-07,
+      "loss": 0.8157,
+      "step": 9468
+    },
+    {
+      "epoch": 1.8228457299612582,
+      "grad_norm": 3.217638726886422,
+      "learning_rate": 4.0770969982148025e-07,
+      "loss": 0.8412,
+      "step": 9469
+    },
+    {
+      "epoch": 1.8230382366388334,
+      "grad_norm": 3.3210595966246945,
+      "learning_rate": 4.0682895522519474e-07,
+      "loss": 0.847,
+      "step": 9470
+    },
+    {
+      "epoch": 1.8232307433164088,
+      "grad_norm": 3.3081842533032617,
+      "learning_rate": 4.059491432078666e-07,
+      "loss": 0.8349,
+      "step": 9471
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "grad_norm": 3.217377490710455,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.8739,
+      "step": 9472
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.6206,
+      "step": 9472,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.7198,
+      "step": 9472,
+      "vm_loss": 0.1805
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.7708,
+      "step": 9472,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.4317,
+      "step": 9472,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.4208,
+      "step": 9472,
+      "vm_loss": 0.1374
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 1.276,
+      "step": 9472,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.4207,
+      "step": 9472,
+      "vm_loss": 0.1713
+    },
+    {
+      "epoch": 1.8234232499939842,
+      "lm_loss": 0.9471,
+      "step": 9472,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 1.8236157566715594,
+      "grad_norm": 3.430338053425075,
+      "learning_rate": 4.0419231725211005e-07,
+      "loss": 0.8747,
+      "step": 9473
+    },
+    {
+      "epoch": 1.823808263349135,
+      "grad_norm": 3.307244858408728,
+      "learning_rate": 4.0331530348446944e-07,
+      "loss": 0.8541,
+      "step": 9474
+    },
+    {
+      "epoch": 1.8240007700267102,
+      "grad_norm": 3.3048154437997193,
+      "learning_rate": 4.024392226373563e-07,
+      "loss": 0.8774,
+      "step": 9475
+    },
+    {
+      "epoch": 1.8241932767042857,
+      "grad_norm": 3.3255389930089536,
+      "learning_rate": 4.015640747959404e-07,
+      "loss": 0.868,
+      "step": 9476
+    },
+    {
+      "epoch": 1.824385783381861,
+      "grad_norm": 3.3005337898423606,
+      "learning_rate": 4.0068986004529577e-07,
+      "loss": 0.8345,
+      "step": 9477
+    },
+    {
+      "epoch": 1.8245782900594363,
+      "grad_norm": 3.299226665971921,
+      "learning_rate": 3.9981657847040776e-07,
+      "loss": 0.8183,
+      "step": 9478
+    },
+    {
+      "epoch": 1.824770796737012,
+      "grad_norm": 3.2109005180579313,
+      "learning_rate": 3.989442301561719e-07,
+      "loss": 0.8291,
+      "step": 9479
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "grad_norm": 3.2987902377342144,
+      "learning_rate": 3.9807281518738916e-07,
+      "loss": 0.8349,
+      "step": 9480
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.7434,
+      "step": 9480,
+      "vm_loss": 0.1807
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.4819,
+      "step": 9480,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.5675,
+      "step": 9480,
+      "vm_loss": 0.219
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.8994,
+      "step": 9480,
+      "vm_loss": 0.2045
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.7241,
+      "step": 9480,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 1.0228,
+      "step": 9480,
+      "vm_loss": 0.1341
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.8299,
+      "step": 9480,
+      "vm_loss": 0.1348
+    },
+    {
+      "epoch": 1.8249633034145871,
+      "lm_loss": 0.6174,
+      "step": 9480,
+      "vm_loss": 0.1183
+    },
+    {
+      "epoch": 1.8251558100921625,
+      "grad_norm": 3.4308905616113776,
+      "learning_rate": 3.972023336487729e-07,
+      "loss": 0.8975,
+      "step": 9481
+    },
+    {
+      "epoch": 1.825348316769738,
+      "grad_norm": 3.1579522901354156,
+      "learning_rate": 3.9633278562494657e-07,
+      "loss": 0.8092,
+      "step": 9482
+    },
+    {
+      "epoch": 1.8255408234473132,
+      "grad_norm": 3.3288312197126007,
+      "learning_rate": 3.9546417120044035e-07,
+      "loss": 0.8663,
+      "step": 9483
+    },
+    {
+      "epoch": 1.8257333301248888,
+      "grad_norm": 3.392616518180647,
+      "learning_rate": 3.9459649045969216e-07,
+      "loss": 0.8615,
+      "step": 9484
+    },
+    {
+      "epoch": 1.825925836802464,
+      "grad_norm": 3.3863379603030577,
+      "learning_rate": 3.9372974348705685e-07,
+      "loss": 0.8821,
+      "step": 9485
+    },
+    {
+      "epoch": 1.8261183434800394,
+      "grad_norm": 3.4100559828135806,
+      "learning_rate": 3.9286393036678914e-07,
+      "loss": 0.8692,
+      "step": 9486
+    },
+    {
+      "epoch": 1.8263108501576149,
+      "grad_norm": 3.0070194853357903,
+      "learning_rate": 3.919990511830585e-07,
+      "loss": 0.7841,
+      "step": 9487
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "grad_norm": 3.3824976826180415,
+      "learning_rate": 3.911351060199431e-07,
+      "loss": 0.8727,
+      "step": 9488
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.6035,
+      "step": 9488,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.4873,
+      "step": 9488,
+      "vm_loss": 0.1776
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.6396,
+      "step": 9488,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.2392,
+      "step": 9488,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.647,
+      "step": 9488,
+      "vm_loss": 0.2234
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 1.0678,
+      "step": 9488,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.7865,
+      "step": 9488,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.8265033568351903,
+      "lm_loss": 0.7229,
+      "step": 9488,
+      "vm_loss": 0.1882
+    },
+    {
+      "epoch": 1.8266958635127657,
+      "grad_norm": 3.1854970384473345,
+      "learning_rate": 3.9027209496142913e-07,
+      "loss": 0.8205,
+      "step": 9489
+    },
+    {
+      "epoch": 1.826888370190341,
+      "grad_norm": 3.136082335568287,
+      "learning_rate": 3.894100180914095e-07,
+      "loss": 0.8302,
+      "step": 9490
+    },
+    {
+      "epoch": 1.8270808768679163,
+      "grad_norm": 3.190175453799504,
+      "learning_rate": 3.885488754936939e-07,
+      "loss": 0.8387,
+      "step": 9491
+    },
+    {
+      "epoch": 1.8272733835454917,
+      "grad_norm": 3.3768010388907235,
+      "learning_rate": 3.876886672519931e-07,
+      "loss": 0.8705,
+      "step": 9492
+    },
+    {
+      "epoch": 1.8274658902230672,
+      "grad_norm": 3.2775909281947717,
+      "learning_rate": 3.868293934499312e-07,
+      "loss": 0.8396,
+      "step": 9493
+    },
+    {
+      "epoch": 1.8276583969006426,
+      "grad_norm": 3.290913518702874,
+      "learning_rate": 3.859710541710393e-07,
+      "loss": 0.8598,
+      "step": 9494
+    },
+    {
+      "epoch": 1.8278509035782178,
+      "grad_norm": 3.315770383857467,
+      "learning_rate": 3.8511364949876284e-07,
+      "loss": 0.8449,
+      "step": 9495
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "grad_norm": 3.3448790442644154,
+      "learning_rate": 3.842571795164485e-07,
+      "loss": 0.8476,
+      "step": 9496
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.6862,
+      "step": 9496,
+      "vm_loss": 0.2112
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.7716,
+      "step": 9496,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.6419,
+      "step": 9496,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.7296,
+      "step": 9496,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.8192,
+      "step": 9496,
+      "vm_loss": 0.1083
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.5988,
+      "step": 9496,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.3493,
+      "step": 9496,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.8280434102557932,
+      "lm_loss": 0.419,
+      "step": 9496,
+      "vm_loss": 0.2318
+    },
+    {
+      "epoch": 1.8282359169333686,
+      "grad_norm": 3.1472182266491475,
+      "learning_rate": 3.834016443073574e-07,
+      "loss": 0.798,
+      "step": 9497
+    },
+    {
+      "epoch": 1.828428423610944,
+      "grad_norm": 3.1008637888315205,
+      "learning_rate": 3.8254704395465856e-07,
+      "loss": 0.7781,
+      "step": 9498
+    },
+    {
+      "epoch": 1.8286209302885195,
+      "grad_norm": 3.4187812345455386,
+      "learning_rate": 3.816933785414312e-07,
+      "loss": 0.9372,
+      "step": 9499
+    },
+    {
+      "epoch": 1.8288134369660947,
+      "grad_norm": 3.172118024424408,
+      "learning_rate": 3.8084064815066104e-07,
+      "loss": 0.7911,
+      "step": 9500
+    },
+    {
+      "epoch": 1.82900594364367,
+      "grad_norm": 3.1471065054341243,
+      "learning_rate": 3.79988852865244e-07,
+      "loss": 0.8491,
+      "step": 9501
+    },
+    {
+      "epoch": 1.8291984503212455,
+      "grad_norm": 3.2790091927382186,
+      "learning_rate": 3.791379927679884e-07,
+      "loss": 0.8461,
+      "step": 9502
+    },
+    {
+      "epoch": 1.829390956998821,
+      "grad_norm": 3.140193001946722,
+      "learning_rate": 3.7828806794160456e-07,
+      "loss": 0.8233,
+      "step": 9503
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "grad_norm": 3.286493738407895,
+      "learning_rate": 3.774390784687187e-07,
+      "loss": 0.8077,
+      "step": 9504
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.7782,
+      "step": 9504,
+      "vm_loss": 0.1788
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.7257,
+      "step": 9504,
+      "vm_loss": 0.1896
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.7783,
+      "step": 9504,
+      "vm_loss": 0.0867
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.6613,
+      "step": 9504,
+      "vm_loss": 0.1642
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.8421,
+      "step": 9504,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.7399,
+      "step": 9504,
+      "vm_loss": 0.1577
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.6607,
+      "step": 9504,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.8295834636763963,
+      "lm_loss": 0.7154,
+      "step": 9504,
+      "vm_loss": 0.2408
+    },
+    {
+      "epoch": 1.8297759703539715,
+      "grad_norm": 3.1439730386916005,
+      "learning_rate": 3.765910244318638e-07,
+      "loss": 0.8058,
+      "step": 9505
+    },
+    {
+      "epoch": 1.8299684770315472,
+      "grad_norm": 3.254878105620397,
+      "learning_rate": 3.7574390591348044e-07,
+      "loss": 0.8451,
+      "step": 9506
+    },
+    {
+      "epoch": 1.8301609837091224,
+      "grad_norm": 3.231804991046132,
+      "learning_rate": 3.748977229959183e-07,
+      "loss": 0.8573,
+      "step": 9507
+    },
+    {
+      "epoch": 1.8303534903866978,
+      "grad_norm": 3.2481493719999626,
+      "learning_rate": 3.7405247576144055e-07,
+      "loss": 0.7731,
+      "step": 9508
+    },
+    {
+      "epoch": 1.8305459970642732,
+      "grad_norm": 3.3602611014099444,
+      "learning_rate": 3.732081642922125e-07,
+      "loss": 0.8738,
+      "step": 9509
+    },
+    {
+      "epoch": 1.8307385037418484,
+      "grad_norm": 3.152238527216324,
+      "learning_rate": 3.723647886703141e-07,
+      "loss": 0.7808,
+      "step": 9510
+    },
+    {
+      "epoch": 1.830931010419424,
+      "grad_norm": 3.346070852443939,
+      "learning_rate": 3.7152234897773085e-07,
+      "loss": 0.8679,
+      "step": 9511
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "grad_norm": 3.3026290344589713,
+      "learning_rate": 3.706808452963606e-07,
+      "loss": 0.8437,
+      "step": 9512
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.661,
+      "step": 9512,
+      "vm_loss": 0.2038
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.5485,
+      "step": 9512,
+      "vm_loss": 0.1976
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.7937,
+      "step": 9512,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.6154,
+      "step": 9512,
+      "vm_loss": 0.2184
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.7715,
+      "step": 9512,
+      "vm_loss": 0.1487
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.7354,
+      "step": 9512,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.5383,
+      "step": 9512,
+      "vm_loss": 0.1786
+    },
+    {
+      "epoch": 1.8311235170969993,
+      "lm_loss": 0.8047,
+      "step": 9512,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.8313160237745747,
+      "grad_norm": 3.1994216360475933,
+      "learning_rate": 3.6984027770800676e-07,
+      "loss": 0.8237,
+      "step": 9513
+    },
+    {
+      "epoch": 1.8315085304521501,
+      "grad_norm": 3.151048861314282,
+      "learning_rate": 3.690006462943829e-07,
+      "loss": 0.8214,
+      "step": 9514
+    },
+    {
+      "epoch": 1.8317010371297253,
+      "grad_norm": 3.3502839719285844,
+      "learning_rate": 3.6816195113711374e-07,
+      "loss": 0.8248,
+      "step": 9515
+    },
+    {
+      "epoch": 1.831893543807301,
+      "grad_norm": 3.4547906990984503,
+      "learning_rate": 3.6732419231773063e-07,
+      "loss": 0.8845,
+      "step": 9516
+    },
+    {
+      "epoch": 1.8320860504848762,
+      "grad_norm": 3.2973478269386427,
+      "learning_rate": 3.664873699176719e-07,
+      "loss": 0.8647,
+      "step": 9517
+    },
+    {
+      "epoch": 1.8322785571624516,
+      "grad_norm": 3.346744758591867,
+      "learning_rate": 3.656514840182912e-07,
+      "loss": 0.8477,
+      "step": 9518
+    },
+    {
+      "epoch": 1.832471063840027,
+      "grad_norm": 3.27423976087775,
+      "learning_rate": 3.648165347008448e-07,
+      "loss": 0.875,
+      "step": 9519
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "grad_norm": 3.334941521111932,
+      "learning_rate": 3.63982522046501e-07,
+      "loss": 0.839,
+      "step": 9520
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.9093,
+      "step": 9520,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 1.1019,
+      "step": 9520,
+      "vm_loss": 0.2605
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.3618,
+      "step": 9520,
+      "vm_loss": 0.1081
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.4845,
+      "step": 9520,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.6284,
+      "step": 9520,
+      "vm_loss": 0.1425
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.2545,
+      "step": 9520,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.4794,
+      "step": 9520,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 1.8326635705176022,
+      "lm_loss": 0.4807,
+      "step": 9520,
+      "vm_loss": 0.1183
+    },
+    {
+      "epoch": 1.8328560771951778,
+      "grad_norm": 3.2831563354004003,
+      "learning_rate": 3.6314944613633606e-07,
+      "loss": 0.8398,
+      "step": 9521
+    },
+    {
+      "epoch": 1.833048583872753,
+      "grad_norm": 3.1745320381986555,
+      "learning_rate": 3.623173070513375e-07,
+      "loss": 0.7842,
+      "step": 9522
+    },
+    {
+      "epoch": 1.8332410905503285,
+      "grad_norm": 3.2427427519757,
+      "learning_rate": 3.6148610487239946e-07,
+      "loss": 0.8635,
+      "step": 9523
+    },
+    {
+      "epoch": 1.8334335972279039,
+      "grad_norm": 3.199753075647246,
+      "learning_rate": 3.6065583968032283e-07,
+      "loss": 0.8115,
+      "step": 9524
+    },
+    {
+      "epoch": 1.833626103905479,
+      "grad_norm": 3.2788236138981333,
+      "learning_rate": 3.5982651155582303e-07,
+      "loss": 0.8359,
+      "step": 9525
+    },
+    {
+      "epoch": 1.8338186105830547,
+      "grad_norm": 3.294500466563377,
+      "learning_rate": 3.58998120579519e-07,
+      "loss": 0.8586,
+      "step": 9526
+    },
+    {
+      "epoch": 1.83401111726063,
+      "grad_norm": 3.298813101790542,
+      "learning_rate": 3.58170666831944e-07,
+      "loss": 0.8708,
+      "step": 9527
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "grad_norm": 3.1303530840565905,
+      "learning_rate": 3.5734415039353263e-07,
+      "loss": 0.8267,
+      "step": 9528
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.5389,
+      "step": 9528,
+      "vm_loss": 0.179
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.5756,
+      "step": 9528,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.7273,
+      "step": 9528,
+      "vm_loss": 0.2108
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.6951,
+      "step": 9528,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.5949,
+      "step": 9528,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.3952,
+      "step": 9528,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.7017,
+      "step": 9528,
+      "vm_loss": 0.1904
+    },
+    {
+      "epoch": 1.8342036239382054,
+      "lm_loss": 0.5159,
+      "step": 9528,
+      "vm_loss": 0.1982
+    },
+    {
+      "epoch": 1.8343961306157808,
+      "grad_norm": 3.2417793252633067,
+      "learning_rate": 3.565185713446373e-07,
+      "loss": 0.8656,
+      "step": 9529
+    },
+    {
+      "epoch": 1.834588637293356,
+      "grad_norm": 3.1744660241459424,
+      "learning_rate": 3.556939297655115e-07,
+      "loss": 0.8014,
+      "step": 9530
+    },
+    {
+      "epoch": 1.8347811439709316,
+      "grad_norm": 3.1962048198810504,
+      "learning_rate": 3.5487022573632343e-07,
+      "loss": 0.8138,
+      "step": 9531
+    },
+    {
+      "epoch": 1.8349736506485068,
+      "grad_norm": 3.2059522436308283,
+      "learning_rate": 3.540474593371468e-07,
+      "loss": 0.8284,
+      "step": 9532
+    },
+    {
+      "epoch": 1.8351661573260822,
+      "grad_norm": 3.255175893765566,
+      "learning_rate": 3.5322563064796535e-07,
+      "loss": 0.8372,
+      "step": 9533
+    },
+    {
+      "epoch": 1.8353586640036577,
+      "grad_norm": 3.4787265189268854,
+      "learning_rate": 3.524047397486696e-07,
+      "loss": 0.9039,
+      "step": 9534
+    },
+    {
+      "epoch": 1.8355511706812329,
+      "grad_norm": 3.1489074143203277,
+      "learning_rate": 3.515847867190625e-07,
+      "loss": 0.8423,
+      "step": 9535
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "grad_norm": 3.1705660358493244,
+      "learning_rate": 3.507657716388535e-07,
+      "loss": 0.7831,
+      "step": 9536
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.6527,
+      "step": 9536,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.8496,
+      "step": 9536,
+      "vm_loss": 0.1578
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.572,
+      "step": 9536,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.6331,
+      "step": 9536,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.4811,
+      "step": 9536,
+      "vm_loss": 0.1514
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.9438,
+      "step": 9536,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.8788,
+      "step": 9536,
+      "vm_loss": 0.2183
+    },
+    {
+      "epoch": 1.8357436773588085,
+      "lm_loss": 0.9806,
+      "step": 9536,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.8359361840363837,
+      "grad_norm": 3.299423370254257,
+      "learning_rate": 3.4994769458766233e-07,
+      "loss": 0.8541,
+      "step": 9537
+    },
+    {
+      "epoch": 1.8361286907139591,
+      "grad_norm": 3.15513076282116,
+      "learning_rate": 3.491305556450131e-07,
+      "loss": 0.8091,
+      "step": 9538
+    },
+    {
+      "epoch": 1.8363211973915345,
+      "grad_norm": 3.4448679822407615,
+      "learning_rate": 3.483143548903467e-07,
+      "loss": 0.8889,
+      "step": 9539
+    },
+    {
+      "epoch": 1.8365137040691097,
+      "grad_norm": 3.2907299642195875,
+      "learning_rate": 3.474990924030042e-07,
+      "loss": 0.8491,
+      "step": 9540
+    },
+    {
+      "epoch": 1.8367062107466854,
+      "grad_norm": 3.113449439984556,
+      "learning_rate": 3.4668476826224205e-07,
+      "loss": 0.7998,
+      "step": 9541
+    },
+    {
+      "epoch": 1.8368987174242606,
+      "grad_norm": 3.3597768355411053,
+      "learning_rate": 3.458713825472226e-07,
+      "loss": 0.8691,
+      "step": 9542
+    },
+    {
+      "epoch": 1.837091224101836,
+      "grad_norm": 3.404091040106368,
+      "learning_rate": 3.4505893533701704e-07,
+      "loss": 0.8949,
+      "step": 9543
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "grad_norm": 3.3427772751432934,
+      "learning_rate": 3.442474267106077e-07,
+      "loss": 0.8663,
+      "step": 9544
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.783,
+      "step": 9544,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.5112,
+      "step": 9544,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.7183,
+      "step": 9544,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.5819,
+      "step": 9544,
+      "vm_loss": 0.1012
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.3258,
+      "step": 9544,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.6095,
+      "step": 9544,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.5312,
+      "step": 9544,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.8372837307794114,
+      "lm_loss": 0.5948,
+      "step": 9544,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 1.8374762374569866,
+      "grad_norm": 3.3043947929699566,
+      "learning_rate": 3.434368567468782e-07,
+      "loss": 0.8461,
+      "step": 9545
+    },
+    {
+      "epoch": 1.8376687441345623,
+      "grad_norm": 3.4320712581029627,
+      "learning_rate": 3.4262722552463213e-07,
+      "loss": 0.8784,
+      "step": 9546
+    },
+    {
+      "epoch": 1.8378612508121375,
+      "grad_norm": 3.3807749870866783,
+      "learning_rate": 3.418185331225732e-07,
+      "loss": 0.8652,
+      "step": 9547
+    },
+    {
+      "epoch": 1.838053757489713,
+      "grad_norm": 3.221563034623357,
+      "learning_rate": 3.4101077961931627e-07,
+      "loss": 0.8169,
+      "step": 9548
+    },
+    {
+      "epoch": 1.8382462641672883,
+      "grad_norm": 3.255058121465018,
+      "learning_rate": 3.4020396509338636e-07,
+      "loss": 0.8196,
+      "step": 9549
+    },
+    {
+      "epoch": 1.8384387708448637,
+      "grad_norm": 3.1300023373373174,
+      "learning_rate": 3.393980896232174e-07,
+      "loss": 0.8137,
+      "step": 9550
+    },
+    {
+      "epoch": 1.8386312775224392,
+      "grad_norm": 3.219771332398671,
+      "learning_rate": 3.385931532871478e-07,
+      "loss": 0.8425,
+      "step": 9551
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "grad_norm": 3.3004988231847174,
+      "learning_rate": 3.3778915616342947e-07,
+      "loss": 0.8592,
+      "step": 9552
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.3296,
+      "step": 9552,
+      "vm_loss": 0.1052
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.3474,
+      "step": 9552,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.6391,
+      "step": 9552,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.7476,
+      "step": 9552,
+      "vm_loss": 0.1641
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 1.0186,
+      "step": 9552,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.6157,
+      "step": 9552,
+      "vm_loss": 0.1618
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.8327,
+      "step": 9552,
+      "vm_loss": 0.1796
+    },
+    {
+      "epoch": 1.8388237842000144,
+      "lm_loss": 0.894,
+      "step": 9552,
+      "vm_loss": 0.1777
+    },
+    {
+      "epoch": 1.8390162908775898,
+      "grad_norm": 3.0862680791692325,
+      "learning_rate": 3.369860983302209e-07,
+      "loss": 0.809,
+      "step": 9553
+    },
+    {
+      "epoch": 1.8392087975551652,
+      "grad_norm": 3.3058611768910366,
+      "learning_rate": 3.361839798655897e-07,
+      "loss": 0.8709,
+      "step": 9554
+    },
+    {
+      "epoch": 1.8394013042327406,
+      "grad_norm": 3.4039592810382175,
+      "learning_rate": 3.353828008475113e-07,
+      "loss": 0.8918,
+      "step": 9555
+    },
+    {
+      "epoch": 1.839593810910316,
+      "grad_norm": 3.4500258485878517,
+      "learning_rate": 3.3458256135387225e-07,
+      "loss": 0.8848,
+      "step": 9556
+    },
+    {
+      "epoch": 1.8397863175878912,
+      "grad_norm": 3.150119150357368,
+      "learning_rate": 3.337832614624648e-07,
+      "loss": 0.8514,
+      "step": 9557
+    },
+    {
+      "epoch": 1.8399788242654667,
+      "grad_norm": 3.269558438969154,
+      "learning_rate": 3.3298490125099116e-07,
+      "loss": 0.8749,
+      "step": 9558
+    },
+    {
+      "epoch": 1.840171330943042,
+      "grad_norm": 3.1418983949093544,
+      "learning_rate": 3.3218748079706153e-07,
+      "loss": 0.8228,
+      "step": 9559
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "grad_norm": 3.2807713960364704,
+      "learning_rate": 3.3139100017819835e-07,
+      "loss": 0.8698,
+      "step": 9560
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.3522,
+      "step": 9560,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.6446,
+      "step": 9560,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.5976,
+      "step": 9560,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.6688,
+      "step": 9560,
+      "vm_loss": 0.127
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.7808,
+      "step": 9560,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.587,
+      "step": 9560,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.3409,
+      "step": 9560,
+      "vm_loss": 0.1433
+    },
+    {
+      "epoch": 1.8403638376206175,
+      "lm_loss": 0.6537,
+      "step": 9560,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.840556344298193,
+      "grad_norm": 3.054919236659913,
+      "learning_rate": 3.3059545947182636e-07,
+      "loss": 0.7733,
+      "step": 9561
+    },
+    {
+      "epoch": 1.8407488509757681,
+      "grad_norm": 3.0885810012828063,
+      "learning_rate": 3.2980085875528254e-07,
+      "loss": 0.798,
+      "step": 9562
+    },
+    {
+      "epoch": 1.8409413576533435,
+      "grad_norm": 3.29829414893802,
+      "learning_rate": 3.2900719810581406e-07,
+      "loss": 0.8709,
+      "step": 9563
+    },
+    {
+      "epoch": 1.841133864330919,
+      "grad_norm": 3.135548771166346,
+      "learning_rate": 3.2821447760057467e-07,
+      "loss": 0.8009,
+      "step": 9564
+    },
+    {
+      "epoch": 1.8413263710084944,
+      "grad_norm": 3.217343032000977,
+      "learning_rate": 3.2742269731662504e-07,
+      "loss": 0.8472,
+      "step": 9565
+    },
+    {
+      "epoch": 1.8415188776860698,
+      "grad_norm": 3.317490777288784,
+      "learning_rate": 3.2663185733093796e-07,
+      "loss": 0.8525,
+      "step": 9566
+    },
+    {
+      "epoch": 1.841711384363645,
+      "grad_norm": 3.3714805204285785,
+      "learning_rate": 3.258419577203942e-07,
+      "loss": 0.8191,
+      "step": 9567
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "grad_norm": 3.2876543754563516,
+      "learning_rate": 3.250529985617801e-07,
+      "loss": 0.8525,
+      "step": 9568
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.396,
+      "step": 9568,
+      "vm_loss": 0.1422
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.4368,
+      "step": 9568,
+      "vm_loss": 0.2023
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.7416,
+      "step": 9568,
+      "vm_loss": 0.2128
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.4,
+      "step": 9568,
+      "vm_loss": 0.1088
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.6,
+      "step": 9568,
+      "vm_loss": 0.1307
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.7134,
+      "step": 9568,
+      "vm_loss": 0.2191
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.4495,
+      "step": 9568,
+      "vm_loss": 0.1208
+    },
+    {
+      "epoch": 1.8419038910412207,
+      "lm_loss": 0.659,
+      "step": 9568,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 1.8420963977187959,
+      "grad_norm": 3.161443036247088,
+      "learning_rate": 3.2426497993179207e-07,
+      "loss": 0.8085,
+      "step": 9569
+    },
+    {
+      "epoch": 1.8422889043963713,
+      "grad_norm": 3.49922044050319,
+      "learning_rate": 3.2347790190703776e-07,
+      "loss": 0.905,
+      "step": 9570
+    },
+    {
+      "epoch": 1.8424814110739467,
+      "grad_norm": 3.115539250395291,
+      "learning_rate": 3.226917645640315e-07,
+      "loss": 0.824,
+      "step": 9571
+    },
+    {
+      "epoch": 1.842673917751522,
+      "grad_norm": 3.222247052149268,
+      "learning_rate": 3.2190656797919215e-07,
+      "loss": 0.8278,
+      "step": 9572
+    },
+    {
+      "epoch": 1.8428664244290975,
+      "grad_norm": 3.2205640216521836,
+      "learning_rate": 3.2112231222885536e-07,
+      "loss": 0.8168,
+      "step": 9573
+    },
+    {
+      "epoch": 1.8430589311066727,
+      "grad_norm": 3.2583617883956872,
+      "learning_rate": 3.203389973892579e-07,
+      "loss": 0.837,
+      "step": 9574
+    },
+    {
+      "epoch": 1.8432514377842482,
+      "grad_norm": 3.2637800564939297,
+      "learning_rate": 3.195566235365499e-07,
+      "loss": 0.8551,
+      "step": 9575
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "grad_norm": 3.1884264404270524,
+      "learning_rate": 3.187751907467851e-07,
+      "loss": 0.843,
+      "step": 9576
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.9084,
+      "step": 9576,
+      "vm_loss": 0.1978
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.3514,
+      "step": 9576,
+      "vm_loss": 0.1841
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.6848,
+      "step": 9576,
+      "vm_loss": 0.1189
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.3035,
+      "step": 9576,
+      "vm_loss": 0.1619
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.7604,
+      "step": 9576,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.9533,
+      "step": 9576,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.4598,
+      "step": 9576,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.8434439444618236,
+      "lm_loss": 0.4351,
+      "step": 9576,
+      "vm_loss": 0.1999
+    },
+    {
+      "epoch": 1.8436364511393988,
+      "grad_norm": 3.3462576067718404,
+      "learning_rate": 3.179946990959326e-07,
+      "loss": 0.8614,
+      "step": 9577
+    },
+    {
+      "epoch": 1.8438289578169744,
+      "grad_norm": 3.3052975459465146,
+      "learning_rate": 3.172151486598629e-07,
+      "loss": 0.8577,
+      "step": 9578
+    },
+    {
+      "epoch": 1.8440214644945496,
+      "grad_norm": 3.1065690570671673,
+      "learning_rate": 3.1643653951436095e-07,
+      "loss": 0.8116,
+      "step": 9579
+    },
+    {
+      "epoch": 1.844213971172125,
+      "grad_norm": 3.1430785507445966,
+      "learning_rate": 3.15658871735115e-07,
+      "loss": 0.794,
+      "step": 9580
+    },
+    {
+      "epoch": 1.8444064778497005,
+      "grad_norm": 3.0770032231485067,
+      "learning_rate": 3.14882145397728e-07,
+      "loss": 0.8068,
+      "step": 9581
+    },
+    {
+      "epoch": 1.8445989845272757,
+      "grad_norm": 3.4076122832152214,
+      "learning_rate": 3.141063605777028e-07,
+      "loss": 0.8473,
+      "step": 9582
+    },
+    {
+      "epoch": 1.8447914912048513,
+      "grad_norm": 3.1578171137205207,
+      "learning_rate": 3.133315173504592e-07,
+      "loss": 0.8162,
+      "step": 9583
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "grad_norm": 3.250535141909998,
+      "learning_rate": 3.125576157913213e-07,
+      "loss": 0.848,
+      "step": 9584
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.7243,
+      "step": 9584,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.8587,
+      "step": 9584,
+      "vm_loss": 0.1866
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.4455,
+      "step": 9584,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.7672,
+      "step": 9584,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.9324,
+      "step": 9584,
+      "vm_loss": 0.2242
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.9937,
+      "step": 9584,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.8959,
+      "step": 9584,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.8449839978824265,
+      "lm_loss": 0.7313,
+      "step": 9584,
+      "vm_loss": 0.1702
+    },
+    {
+      "epoch": 1.845176504560002,
+      "grad_norm": 3.3652124468615447,
+      "learning_rate": 3.117846559755211e-07,
+      "loss": 0.8722,
+      "step": 9585
+    },
+    {
+      "epoch": 1.8453690112375774,
+      "grad_norm": 3.2385447793080644,
+      "learning_rate": 3.110126379781997e-07,
+      "loss": 0.8278,
+      "step": 9586
+    },
+    {
+      "epoch": 1.8455615179151525,
+      "grad_norm": 3.1206017394097945,
+      "learning_rate": 3.102415618744092e-07,
+      "loss": 0.7581,
+      "step": 9587
+    },
+    {
+      "epoch": 1.8457540245927282,
+      "grad_norm": 3.279019781279383,
+      "learning_rate": 3.0947142773910863e-07,
+      "loss": 0.8382,
+      "step": 9588
+    },
+    {
+      "epoch": 1.8459465312703034,
+      "grad_norm": 3.376420562140214,
+      "learning_rate": 3.087022356471603e-07,
+      "loss": 0.8692,
+      "step": 9589
+    },
+    {
+      "epoch": 1.8461390379478788,
+      "grad_norm": 3.2286511414943875,
+      "learning_rate": 3.0793398567334544e-07,
+      "loss": 0.8321,
+      "step": 9590
+    },
+    {
+      "epoch": 1.8463315446254542,
+      "grad_norm": 3.173387587854778,
+      "learning_rate": 3.071666778923432e-07,
+      "loss": 0.8377,
+      "step": 9591
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "grad_norm": 3.3325969290081763,
+      "learning_rate": 3.064003123787496e-07,
+      "loss": 0.8512,
+      "step": 9592
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.6598,
+      "step": 9592,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.7891,
+      "step": 9592,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.512,
+      "step": 9592,
+      "vm_loss": 0.2124
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 1.1378,
+      "step": 9592,
+      "vm_loss": 0.2229
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.3821,
+      "step": 9592,
+      "vm_loss": 0.1536
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.3317,
+      "step": 9592,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.9321,
+      "step": 9592,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.8465240513030294,
+      "lm_loss": 0.5165,
+      "step": 9592,
+      "vm_loss": 0.1369
+    },
+    {
+      "epoch": 1.846716557980605,
+      "grad_norm": 3.2165937163238265,
+      "learning_rate": 3.056348892070593e-07,
+      "loss": 0.8403,
+      "step": 9593
+    },
+    {
+      "epoch": 1.8469090646581803,
+      "grad_norm": 3.2142485537295427,
+      "learning_rate": 3.0487040845168847e-07,
+      "loss": 0.8321,
+      "step": 9594
+    },
+    {
+      "epoch": 1.8471015713357557,
+      "grad_norm": 3.2185217155175603,
+      "learning_rate": 3.041068701869487e-07,
+      "loss": 0.8282,
+      "step": 9595
+    },
+    {
+      "epoch": 1.8472940780133311,
+      "grad_norm": 3.2449783636220006,
+      "learning_rate": 3.033442744870685e-07,
+      "loss": 0.8567,
+      "step": 9596
+    },
+    {
+      "epoch": 1.8474865846909063,
+      "grad_norm": 3.35156198992,
+      "learning_rate": 3.025826214261807e-07,
+      "loss": 0.8838,
+      "step": 9597
+    },
+    {
+      "epoch": 1.847679091368482,
+      "grad_norm": 3.35834157574361,
+      "learning_rate": 3.0182191107833047e-07,
+      "loss": 0.8278,
+      "step": 9598
+    },
+    {
+      "epoch": 1.8478715980460572,
+      "grad_norm": 3.312555654845415,
+      "learning_rate": 3.010621435174632e-07,
+      "loss": 0.866,
+      "step": 9599
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "grad_norm": 3.0878744038169454,
+      "learning_rate": 3.0030331881744315e-07,
+      "loss": 0.7785,
+      "step": 9600
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.6234,
+      "step": 9600,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.4911,
+      "step": 9600,
+      "vm_loss": 0.1559
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.6624,
+      "step": 9600,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.9975,
+      "step": 9600,
+      "vm_loss": 0.1359
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.4601,
+      "step": 9600,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.5111,
+      "step": 9600,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.71,
+      "step": 9600,
+      "vm_loss": 0.107
+    },
+    {
+      "epoch": 1.8480641047236326,
+      "lm_loss": 0.9088,
+      "step": 9600,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.848256611401208,
+      "grad_norm": 3.1479022702775414,
+      "learning_rate": 2.9954543705203564e-07,
+      "loss": 0.8215,
+      "step": 9601
+    },
+    {
+      "epoch": 1.8484491180787832,
+      "grad_norm": 3.1686049240452983,
+      "learning_rate": 2.987884982949163e-07,
+      "loss": 0.824,
+      "step": 9602
+    },
+    {
+      "epoch": 1.8486416247563588,
+      "grad_norm": 3.309996055049136,
+      "learning_rate": 2.9803250261966956e-07,
+      "loss": 0.8661,
+      "step": 9603
+    },
+    {
+      "epoch": 1.848834131433934,
+      "grad_norm": 3.174100501190331,
+      "learning_rate": 2.9727745009978994e-07,
+      "loss": 0.8088,
+      "step": 9604
+    },
+    {
+      "epoch": 1.8490266381115095,
+      "grad_norm": 3.3527667388312383,
+      "learning_rate": 2.9652334080867427e-07,
+      "loss": 0.8395,
+      "step": 9605
+    },
+    {
+      "epoch": 1.849219144789085,
+      "grad_norm": 3.199608288721758,
+      "learning_rate": 2.95770174819634e-07,
+      "loss": 0.8338,
+      "step": 9606
+    },
+    {
+      "epoch": 1.84941165146666,
+      "grad_norm": 3.377639433104591,
+      "learning_rate": 2.95017952205886e-07,
+      "loss": 0.8332,
+      "step": 9607
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "grad_norm": 3.2989844050641555,
+      "learning_rate": 2.9426667304055525e-07,
+      "loss": 0.8392,
+      "step": 9608
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.506,
+      "step": 9608,
+      "vm_loss": 0.1554
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.4127,
+      "step": 9608,
+      "vm_loss": 0.1244
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.8654,
+      "step": 9608,
+      "vm_loss": 0.1782
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.3977,
+      "step": 9608,
+      "vm_loss": 0.129
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.4094,
+      "step": 9608,
+      "vm_loss": 0.1488
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.7141,
+      "step": 9608,
+      "vm_loss": 0.1089
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.8601,
+      "step": 9608,
+      "vm_loss": 0.1775
+    },
+    {
+      "epoch": 1.8496041581442357,
+      "lm_loss": 0.6407,
+      "step": 9608,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.849796664821811,
+      "grad_norm": 3.2639860090239,
+      "learning_rate": 2.935163373966787e-07,
+      "loss": 0.7994,
+      "step": 9609
+    },
+    {
+      "epoch": 1.8499891714993864,
+      "grad_norm": 3.160041798135915,
+      "learning_rate": 2.9276694534719375e-07,
+      "loss": 0.7838,
+      "step": 9610
+    },
+    {
+      "epoch": 1.8501816781769618,
+      "grad_norm": 3.2653342219607953,
+      "learning_rate": 2.920184969649553e-07,
+      "loss": 0.8396,
+      "step": 9611
+    },
+    {
+      "epoch": 1.8503741848545372,
+      "grad_norm": 3.2780055559578583,
+      "learning_rate": 2.912709923227197e-07,
+      "loss": 0.8416,
+      "step": 9612
+    },
+    {
+      "epoch": 1.8505666915321126,
+      "grad_norm": 3.2591498376637227,
+      "learning_rate": 2.9052443149315435e-07,
+      "loss": 0.818,
+      "step": 9613
+    },
+    {
+      "epoch": 1.8507591982096878,
+      "grad_norm": 3.235833185154272,
+      "learning_rate": 2.897788145488345e-07,
+      "loss": 0.8427,
+      "step": 9614
+    },
+    {
+      "epoch": 1.8509517048872632,
+      "grad_norm": 3.215842559161073,
+      "learning_rate": 2.890341415622444e-07,
+      "loss": 0.822,
+      "step": 9615
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "grad_norm": 3.3361864719292975,
+      "learning_rate": 2.8829041260577394e-07,
+      "loss": 0.8582,
+      "step": 9616
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.5724,
+      "step": 9616,
+      "vm_loss": 0.1128
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.5626,
+      "step": 9616,
+      "vm_loss": 0.148
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.5549,
+      "step": 9616,
+      "vm_loss": 0.1576
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.7712,
+      "step": 9616,
+      "vm_loss": 0.1754
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.7828,
+      "step": 9616,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.3769,
+      "step": 9616,
+      "vm_loss": 0.1519
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.9738,
+      "step": 9616,
+      "vm_loss": 0.1705
+    },
+    {
+      "epoch": 1.8511442115648387,
+      "lm_loss": 0.5852,
+      "step": 9616,
+      "vm_loss": 0.2192
+    },
+    {
+      "epoch": 1.851336718242414,
+      "grad_norm": 3.38528162361654,
+      "learning_rate": 2.875476277517264e-07,
+      "loss": 0.8457,
+      "step": 9617
+    },
+    {
+      "epoch": 1.8515292249199895,
+      "grad_norm": 3.1992594617820678,
+      "learning_rate": 2.868057870723073e-07,
+      "loss": 0.8045,
+      "step": 9618
+    },
+    {
+      "epoch": 1.8517217315975647,
+      "grad_norm": 3.248271856675651,
+      "learning_rate": 2.860648906396346e-07,
+      "loss": 0.8501,
+      "step": 9619
+    },
+    {
+      "epoch": 1.8519142382751401,
+      "grad_norm": 3.2620568291910153,
+      "learning_rate": 2.853249385257295e-07,
+      "loss": 0.8175,
+      "step": 9620
+    },
+    {
+      "epoch": 1.8521067449527155,
+      "grad_norm": 3.393074803490776,
+      "learning_rate": 2.8458593080253004e-07,
+      "loss": 0.859,
+      "step": 9621
+    },
+    {
+      "epoch": 1.852299251630291,
+      "grad_norm": 3.2169530497221785,
+      "learning_rate": 2.838478675418732e-07,
+      "loss": 0.8264,
+      "step": 9622
+    },
+    {
+      "epoch": 1.8524917583078664,
+      "grad_norm": 3.372702668677499,
+      "learning_rate": 2.8311074881550936e-07,
+      "loss": 0.8654,
+      "step": 9623
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "grad_norm": 3.3653128589035926,
+      "learning_rate": 2.8237457469509676e-07,
+      "loss": 0.8658,
+      "step": 9624
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.6055,
+      "step": 9624,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.524,
+      "step": 9624,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.4891,
+      "step": 9624,
+      "vm_loss": 0.1606
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.6838,
+      "step": 9624,
+      "vm_loss": 0.106
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.8178,
+      "step": 9624,
+      "vm_loss": 0.1568
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.5719,
+      "step": 9624,
+      "vm_loss": 0.1551
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 0.6926,
+      "step": 9624,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.8526842649854416,
+      "lm_loss": 1.0273,
+      "step": 9624,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.852876771663017,
+      "grad_norm": 3.3638691215876744,
+      "learning_rate": 2.816393452521993e-07,
+      "loss": 0.8338,
+      "step": 9625
+    },
+    {
+      "epoch": 1.8530692783405924,
+      "grad_norm": 3.248971959179298,
+      "learning_rate": 2.8090506055829304e-07,
+      "loss": 0.8238,
+      "step": 9626
+    },
+    {
+      "epoch": 1.8532617850181679,
+      "grad_norm": 3.1110059934914163,
+      "learning_rate": 2.8017172068475763e-07,
+      "loss": 0.7994,
+      "step": 9627
+    },
+    {
+      "epoch": 1.8534542916957433,
+      "grad_norm": 3.2279332306566166,
+      "learning_rate": 2.794393257028838e-07,
+      "loss": 0.7852,
+      "step": 9628
+    },
+    {
+      "epoch": 1.8536467983733185,
+      "grad_norm": 3.2673131854492796,
+      "learning_rate": 2.7870787568387016e-07,
+      "loss": 0.8178,
+      "step": 9629
+    },
+    {
+      "epoch": 1.8538393050508941,
+      "grad_norm": 3.436545458566446,
+      "learning_rate": 2.7797737069882203e-07,
+      "loss": 0.9113,
+      "step": 9630
+    },
+    {
+      "epoch": 1.8540318117284693,
+      "grad_norm": 3.262198868037698,
+      "learning_rate": 2.772478108187548e-07,
+      "loss": 0.8269,
+      "step": 9631
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "grad_norm": 3.2128607171330117,
+      "learning_rate": 2.765191961145908e-07,
+      "loss": 0.8536,
+      "step": 9632
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.9194,
+      "step": 9632,
+      "vm_loss": 0.2075
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.9703,
+      "step": 9632,
+      "vm_loss": 0.1937
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 1.2659,
+      "step": 9632,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 1.0282,
+      "step": 9632,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.6017,
+      "step": 9632,
+      "vm_loss": 0.1709
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.4856,
+      "step": 9632,
+      "vm_loss": 0.1676
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.6814,
+      "step": 9632,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.8542243184060447,
+      "lm_loss": 0.79,
+      "step": 9632,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.8544168250836202,
+      "grad_norm": 3.386257287798221,
+      "learning_rate": 2.757915266571598e-07,
+      "loss": 0.8797,
+      "step": 9633
+    },
+    {
+      "epoch": 1.8546093317611954,
+      "grad_norm": 3.2129110113337815,
+      "learning_rate": 2.750648025172031e-07,
+      "loss": 0.8385,
+      "step": 9634
+    },
+    {
+      "epoch": 1.854801838438771,
+      "grad_norm": 3.0132512820000197,
+      "learning_rate": 2.7433902376536426e-07,
+      "loss": 0.7625,
+      "step": 9635
+    },
+    {
+      "epoch": 1.8549943451163462,
+      "grad_norm": 3.243941489796722,
+      "learning_rate": 2.7361419047220115e-07,
+      "loss": 0.8228,
+      "step": 9636
+    },
+    {
+      "epoch": 1.8551868517939216,
+      "grad_norm": 3.266034131874567,
+      "learning_rate": 2.7289030270817305e-07,
+      "loss": 0.8425,
+      "step": 9637
+    },
+    {
+      "epoch": 1.855379358471497,
+      "grad_norm": 3.244165053334818,
+      "learning_rate": 2.72167360543657e-07,
+      "loss": 0.8207,
+      "step": 9638
+    },
+    {
+      "epoch": 1.8555718651490722,
+      "grad_norm": 3.41014535113565,
+      "learning_rate": 2.714453640489267e-07,
+      "loss": 0.8575,
+      "step": 9639
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "grad_norm": 3.262554039166125,
+      "learning_rate": 2.707243132941717e-07,
+      "loss": 0.81,
+      "step": 9640
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 1.0262,
+      "step": 9640,
+      "vm_loss": 0.1083
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.7494,
+      "step": 9640,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.4635,
+      "step": 9640,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.5387,
+      "step": 9640,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.8213,
+      "step": 9640,
+      "vm_loss": 0.1249
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.6055,
+      "step": 9640,
+      "vm_loss": 0.1259
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.4722,
+      "step": 9640,
+      "vm_loss": 0.169
+    },
+    {
+      "epoch": 1.8557643718266479,
+      "lm_loss": 0.7171,
+      "step": 9640,
+      "vm_loss": 0.144
+    },
+    {
+      "epoch": 1.855956878504223,
+      "grad_norm": 3.4041877919058945,
+      "learning_rate": 2.7000420834948693e-07,
+      "loss": 0.8514,
+      "step": 9641
+    },
+    {
+      "epoch": 1.8561493851817985,
+      "grad_norm": 3.330376841745715,
+      "learning_rate": 2.6928504928487755e-07,
+      "loss": 0.8356,
+      "step": 9642
+    },
+    {
+      "epoch": 1.856341891859374,
+      "grad_norm": 3.448141851539447,
+      "learning_rate": 2.68566836170252e-07,
+      "loss": 0.9203,
+      "step": 9643
+    },
+    {
+      "epoch": 1.8565343985369491,
+      "grad_norm": 3.2778757923510153,
+      "learning_rate": 2.6784956907543234e-07,
+      "loss": 0.8931,
+      "step": 9644
+    },
+    {
+      "epoch": 1.8567269052145248,
+      "grad_norm": 3.2501337223650655,
+      "learning_rate": 2.67133248070145e-07,
+      "loss": 0.8468,
+      "step": 9645
+    },
+    {
+      "epoch": 1.8569194118921,
+      "grad_norm": 3.2812044486300165,
+      "learning_rate": 2.664178732240252e-07,
+      "loss": 0.8609,
+      "step": 9646
+    },
+    {
+      "epoch": 1.8571119185696754,
+      "grad_norm": 3.144659049224215,
+      "learning_rate": 2.6570344460661647e-07,
+      "loss": 0.8019,
+      "step": 9647
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "grad_norm": 3.2591443051106226,
+      "learning_rate": 2.6498996228737193e-07,
+      "loss": 0.8526,
+      "step": 9648
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.7227,
+      "step": 9648,
+      "vm_loss": 0.1435
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.9284,
+      "step": 9648,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.5626,
+      "step": 9648,
+      "vm_loss": 0.1453
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.5065,
+      "step": 9648,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.7692,
+      "step": 9648,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.7949,
+      "step": 9648,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.3661,
+      "step": 9648,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 1.8573044252472508,
+      "lm_loss": 0.4883,
+      "step": 9648,
+      "vm_loss": 0.1114
+    },
+    {
+      "epoch": 1.857496931924826,
+      "grad_norm": 3.145828079323454,
+      "learning_rate": 2.642774263356507e-07,
+      "loss": 0.7879,
+      "step": 9649
+    },
+    {
+      "epoch": 1.8576894386024017,
+      "grad_norm": 3.2049003241142073,
+      "learning_rate": 2.6356583682071945e-07,
+      "loss": 0.8402,
+      "step": 9650
+    },
+    {
+      "epoch": 1.8578819452799769,
+      "grad_norm": 3.2603877132951236,
+      "learning_rate": 2.6285519381175407e-07,
+      "loss": 0.8326,
+      "step": 9651
+    },
+    {
+      "epoch": 1.8580744519575523,
+      "grad_norm": 3.342407883174142,
+      "learning_rate": 2.6214549737783813e-07,
+      "loss": 0.8562,
+      "step": 9652
+    },
+    {
+      "epoch": 1.8582669586351277,
+      "grad_norm": 3.195245475083382,
+      "learning_rate": 2.614367475879642e-07,
+      "loss": 0.8118,
+      "step": 9653
+    },
+    {
+      "epoch": 1.858459465312703,
+      "grad_norm": 3.188890736193979,
+      "learning_rate": 2.607289445110284e-07,
+      "loss": 0.8486,
+      "step": 9654
+    },
+    {
+      "epoch": 1.8586519719902785,
+      "grad_norm": 3.2726839226499327,
+      "learning_rate": 2.600220882158444e-07,
+      "loss": 0.8276,
+      "step": 9655
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "grad_norm": 3.2450376487877697,
+      "learning_rate": 2.5931617877112183e-07,
+      "loss": 0.84,
+      "step": 9656
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.7111,
+      "step": 9656,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.598,
+      "step": 9656,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.7622,
+      "step": 9656,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.6085,
+      "step": 9656,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.3055,
+      "step": 9656,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.718,
+      "step": 9656,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.5804,
+      "step": 9656,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 1.8588444786678537,
+      "lm_loss": 0.5824,
+      "step": 9656,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.8590369853454292,
+      "grad_norm": 3.3464966422192277,
+      "learning_rate": 2.586112162454868e-07,
+      "loss": 0.8333,
+      "step": 9657
+    },
+    {
+      "epoch": 1.8592294920230046,
+      "grad_norm": 3.174787677173344,
+      "learning_rate": 2.579072007074701e-07,
+      "loss": 0.817,
+      "step": 9658
+    },
+    {
+      "epoch": 1.8594219987005798,
+      "grad_norm": 3.3124093480476815,
+      "learning_rate": 2.572041322255125e-07,
+      "loss": 0.8445,
+      "step": 9659
+    },
+    {
+      "epoch": 1.8596145053781554,
+      "grad_norm": 3.2256057403923366,
+      "learning_rate": 2.5650201086795946e-07,
+      "loss": 0.8252,
+      "step": 9660
+    },
+    {
+      "epoch": 1.8598070120557306,
+      "grad_norm": 3.360017715834636,
+      "learning_rate": 2.558008367030662e-07,
+      "loss": 0.8604,
+      "step": 9661
+    },
+    {
+      "epoch": 1.859999518733306,
+      "grad_norm": 3.3032221913333895,
+      "learning_rate": 2.551006097989961e-07,
+      "loss": 0.8583,
+      "step": 9662
+    },
+    {
+      "epoch": 1.8601920254108815,
+      "grad_norm": 3.3262230806983197,
+      "learning_rate": 2.5440133022382243e-07,
+      "loss": 0.8495,
+      "step": 9663
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "grad_norm": 3.1775843456761144,
+      "learning_rate": 2.537029980455197e-07,
+      "loss": 0.8367,
+      "step": 9664
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.8675,
+      "step": 9664,
+      "vm_loss": 0.145
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.7792,
+      "step": 9664,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.5186,
+      "step": 9664,
+      "vm_loss": 0.1395
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 1.0022,
+      "step": 9664,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.9925,
+      "step": 9664,
+      "vm_loss": 0.2454
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.8654,
+      "step": 9664,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.6293,
+      "step": 9664,
+      "vm_loss": 0.1821
+    },
+    {
+      "epoch": 1.8603845320884567,
+      "lm_loss": 0.5801,
+      "step": 9664,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.8605770387660323,
+      "grad_norm": 3.47990501075481,
+      "learning_rate": 2.5300561333197803e-07,
+      "loss": 0.8945,
+      "step": 9665
+    },
+    {
+      "epoch": 1.8607695454436075,
+      "grad_norm": 3.2091242496639865,
+      "learning_rate": 2.52309176150991e-07,
+      "loss": 0.8193,
+      "step": 9666
+    },
+    {
+      "epoch": 1.860962052121183,
+      "grad_norm": 3.274251130854965,
+      "learning_rate": 2.516136865702612e-07,
+      "loss": 0.8353,
+      "step": 9667
+    },
+    {
+      "epoch": 1.8611545587987584,
+      "grad_norm": 3.140856919723914,
+      "learning_rate": 2.509191446574e-07,
+      "loss": 0.8232,
+      "step": 9668
+    },
+    {
+      "epoch": 1.8613470654763336,
+      "grad_norm": 3.1525851329813634,
+      "learning_rate": 2.5022555047992446e-07,
+      "loss": 0.8116,
+      "step": 9669
+    },
+    {
+      "epoch": 1.8615395721539092,
+      "grad_norm": 3.2726937006517747,
+      "learning_rate": 2.49532904105263e-07,
+      "loss": 0.8308,
+      "step": 9670
+    },
+    {
+      "epoch": 1.8617320788314844,
+      "grad_norm": 3.2448702586529796,
+      "learning_rate": 2.488412056007461e-07,
+      "loss": 0.8226,
+      "step": 9671
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "grad_norm": 3.334841685342883,
+      "learning_rate": 2.481504550336189e-07,
+      "loss": 0.9059,
+      "step": 9672
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.383,
+      "step": 9672,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 1.1607,
+      "step": 9672,
+      "vm_loss": 0.1458
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.5706,
+      "step": 9672,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.4752,
+      "step": 9672,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.6216,
+      "step": 9672,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.603,
+      "step": 9672,
+      "vm_loss": 0.1533
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.6829,
+      "step": 9672,
+      "vm_loss": 0.1672
+    },
+    {
+      "epoch": 1.8619245855090598,
+      "lm_loss": 0.7276,
+      "step": 9672,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 1.8621170921866352,
+      "grad_norm": 3.1890160415655116,
+      "learning_rate": 2.474606524710288e-07,
+      "loss": 0.8122,
+      "step": 9673
+    },
+    {
+      "epoch": 1.8623095988642104,
+      "grad_norm": 3.366482000308505,
+      "learning_rate": 2.467717979800355e-07,
+      "loss": 0.8477,
+      "step": 9674
+    },
+    {
+      "epoch": 1.862502105541786,
+      "grad_norm": 3.3269393040648887,
+      "learning_rate": 2.4608389162760427e-07,
+      "loss": 0.8535,
+      "step": 9675
+    },
+    {
+      "epoch": 1.8626946122193613,
+      "grad_norm": 3.227811758422907,
+      "learning_rate": 2.4539693348060835e-07,
+      "loss": 0.8404,
+      "step": 9676
+    },
+    {
+      "epoch": 1.8628871188969367,
+      "grad_norm": 3.300521378748627,
+      "learning_rate": 2.447109236058265e-07,
+      "loss": 0.8878,
+      "step": 9677
+    },
+    {
+      "epoch": 1.8630796255745121,
+      "grad_norm": 3.112021656396733,
+      "learning_rate": 2.440258620699498e-07,
+      "loss": 0.7939,
+      "step": 9678
+    },
+    {
+      "epoch": 1.8632721322520875,
+      "grad_norm": 3.1811123186668464,
+      "learning_rate": 2.4334174893957505e-07,
+      "loss": 0.8201,
+      "step": 9679
+    },
+    {
+      "epoch": 1.863464638929663,
+      "grad_norm": 3.2446747349981604,
+      "learning_rate": 2.426585842812079e-07,
+      "loss": 0.8202,
+      "step": 9680
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.5164,
+      "step": 9680,
+      "vm_loss": 0.1552
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.4945,
+      "step": 9680,
+      "vm_loss": 0.2067
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.565,
+      "step": 9680,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.7844,
+      "step": 9680,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.6199,
+      "step": 9680,
+      "vm_loss": 0.0831
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.8042,
+      "step": 9680,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.5626,
+      "step": 9680,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.863464638929663,
+      "lm_loss": 0.563,
+      "step": 9680,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.8636571456072382,
+      "grad_norm": 3.143049137460251,
+      "learning_rate": 2.4197636816125634e-07,
+      "loss": 0.8126,
+      "step": 9681
+    },
+    {
+      "epoch": 1.8638496522848136,
+      "grad_norm": 3.2426675089920205,
+      "learning_rate": 2.412951006460451e-07,
+      "loss": 0.8659,
+      "step": 9682
+    },
+    {
+      "epoch": 1.864042158962389,
+      "grad_norm": 3.2162269463474207,
+      "learning_rate": 2.406147818018001e-07,
+      "loss": 0.7919,
+      "step": 9683
+    },
+    {
+      "epoch": 1.8642346656399644,
+      "grad_norm": 3.1617350761602325,
+      "learning_rate": 2.399354116946584e-07,
+      "loss": 0.7993,
+      "step": 9684
+    },
+    {
+      "epoch": 1.8644271723175398,
+      "grad_norm": 3.1669635447357822,
+      "learning_rate": 2.392569903906605e-07,
+      "loss": 0.7747,
+      "step": 9685
+    },
+    {
+      "epoch": 1.864619678995115,
+      "grad_norm": 3.160352014176193,
+      "learning_rate": 2.385795179557604e-07,
+      "loss": 0.7938,
+      "step": 9686
+    },
+    {
+      "epoch": 1.8648121856726905,
+      "grad_norm": 3.125943571038174,
+      "learning_rate": 2.3790299445581645e-07,
+      "loss": 0.8248,
+      "step": 9687
+    },
+    {
+      "epoch": 1.865004692350266,
+      "grad_norm": 3.3510804410930666,
+      "learning_rate": 2.3722741995659382e-07,
+      "loss": 0.8523,
+      "step": 9688
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.9826,
+      "step": 9688,
+      "vm_loss": 0.2301
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.5577,
+      "step": 9688,
+      "vm_loss": 0.0892
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.6451,
+      "step": 9688,
+      "vm_loss": 0.1591
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.7339,
+      "step": 9688,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.5387,
+      "step": 9688,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 1.3206,
+      "step": 9688,
+      "vm_loss": 0.1629
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.7708,
+      "step": 9688,
+      "vm_loss": 0.122
+    },
+    {
+      "epoch": 1.865004692350266,
+      "lm_loss": 0.7254,
+      "step": 9688,
+      "vm_loss": 0.1274
+    },
+    {
+      "epoch": 1.8651971990278413,
+      "grad_norm": 3.440153434596968,
+      "learning_rate": 2.365527945237689e-07,
+      "loss": 0.8661,
+      "step": 9689
+    },
+    {
+      "epoch": 1.8653897057054167,
+      "grad_norm": 3.094718765062354,
+      "learning_rate": 2.3587911822292364e-07,
+      "loss": 0.7645,
+      "step": 9690
+    },
+    {
+      "epoch": 1.865582212382992,
+      "grad_norm": 3.418161329806052,
+      "learning_rate": 2.352063911195479e-07,
+      "loss": 0.8774,
+      "step": 9691
+    },
+    {
+      "epoch": 1.8657747190605676,
+      "grad_norm": 3.164205496457555,
+      "learning_rate": 2.3453461327903825e-07,
+      "loss": 0.8191,
+      "step": 9692
+    },
+    {
+      "epoch": 1.8659672257381428,
+      "grad_norm": 3.321684118399626,
+      "learning_rate": 2.3386378476670245e-07,
+      "loss": 0.8577,
+      "step": 9693
+    },
+    {
+      "epoch": 1.8661597324157182,
+      "grad_norm": 3.142493114974839,
+      "learning_rate": 2.3319390564775168e-07,
+      "loss": 0.7871,
+      "step": 9694
+    },
+    {
+      "epoch": 1.8663522390932936,
+      "grad_norm": 3.3976251717189068,
+      "learning_rate": 2.3252497598730606e-07,
+      "loss": 0.864,
+      "step": 9695
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "grad_norm": 3.2686074730985673,
+      "learning_rate": 2.318569958503958e-07,
+      "loss": 0.8157,
+      "step": 9696
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.6711,
+      "step": 9696,
+      "vm_loss": 0.1132
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.4747,
+      "step": 9696,
+      "vm_loss": 0.1571
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.5886,
+      "step": 9696,
+      "vm_loss": 0.1683
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.6015,
+      "step": 9696,
+      "vm_loss": 0.1717
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.798,
+      "step": 9696,
+      "vm_loss": 0.1509
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 1.0156,
+      "step": 9696,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.7591,
+      "step": 9696,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 1.8665447457708688,
+      "lm_loss": 0.7683,
+      "step": 9696,
+      "vm_loss": 0.2282
+    },
+    {
+      "epoch": 1.8667372524484445,
+      "grad_norm": 3.296592215486963,
+      "learning_rate": 2.311899653019578e-07,
+      "loss": 0.8498,
+      "step": 9697
+    },
+    {
+      "epoch": 1.8669297591260197,
+      "grad_norm": 3.354925632736047,
+      "learning_rate": 2.3052388440683248e-07,
+      "loss": 0.8834,
+      "step": 9698
+    },
+    {
+      "epoch": 1.867122265803595,
+      "grad_norm": 3.1147071902608916,
+      "learning_rate": 2.298587532297758e-07,
+      "loss": 0.7738,
+      "step": 9699
+    },
+    {
+      "epoch": 1.8673147724811705,
+      "grad_norm": 3.18120408239932,
+      "learning_rate": 2.291945718354438e-07,
+      "loss": 0.7941,
+      "step": 9700
+    },
+    {
+      "epoch": 1.8675072791587457,
+      "grad_norm": 3.267466408861681,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.8526,
+      "step": 9701
+    },
+    {
+      "epoch": 1.8676997858363213,
+      "grad_norm": 3.068626957717172,
+      "learning_rate": 2.278690586531329e-07,
+      "loss": 0.7991,
+      "step": 9702
+    },
+    {
+      "epoch": 1.8678922925138965,
+      "grad_norm": 3.1655574221918465,
+      "learning_rate": 2.2720772699401205e-07,
+      "loss": 0.8058,
+      "step": 9703
+    },
+    {
+      "epoch": 1.868084799191472,
+      "grad_norm": 3.2355871296216,
+      "learning_rate": 2.2654734537532864e-07,
+      "loss": 0.8382,
+      "step": 9704
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.598,
+      "step": 9704,
+      "vm_loss": 0.0869
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.6105,
+      "step": 9704,
+      "vm_loss": 0.205
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.6871,
+      "step": 9704,
+      "vm_loss": 0.1315
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.4307,
+      "step": 9704,
+      "vm_loss": 0.1813
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.2651,
+      "step": 9704,
+      "vm_loss": 0.2453
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.7181,
+      "step": 9704,
+      "vm_loss": 0.1574
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 1.0245,
+      "step": 9704,
+      "vm_loss": 0.1383
+    },
+    {
+      "epoch": 1.868084799191472,
+      "lm_loss": 0.6779,
+      "step": 9704,
+      "vm_loss": 0.1845
+    },
+    {
+      "epoch": 1.8682773058690474,
+      "grad_norm": 3.2927302339922853,
+      "learning_rate": 2.258879138612824e-07,
+      "loss": 0.8359,
+      "step": 9705
+    },
+    {
+      "epoch": 1.8684698125466226,
+      "grad_norm": 3.171509162301296,
+      "learning_rate": 2.2522943251597874e-07,
+      "loss": 0.8029,
+      "step": 9706
+    },
+    {
+      "epoch": 1.8686623192241982,
+      "grad_norm": 3.299729184506608,
+      "learning_rate": 2.2457190140343088e-07,
+      "loss": 0.8466,
+      "step": 9707
+    },
+    {
+      "epoch": 1.8688548259017734,
+      "grad_norm": 3.446661085823441,
+      "learning_rate": 2.2391532058755772e-07,
+      "loss": 0.8816,
+      "step": 9708
+    },
+    {
+      "epoch": 1.8690473325793489,
+      "grad_norm": 3.323875063457362,
+      "learning_rate": 2.232596901321882e-07,
+      "loss": 0.8703,
+      "step": 9709
+    },
+    {
+      "epoch": 1.8692398392569243,
+      "grad_norm": 3.095697258295933,
+      "learning_rate": 2.2260501010105795e-07,
+      "loss": 0.8163,
+      "step": 9710
+    },
+    {
+      "epoch": 1.8694323459344995,
+      "grad_norm": 3.29864542684858,
+      "learning_rate": 2.2195128055780945e-07,
+      "loss": 0.8424,
+      "step": 9711
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "grad_norm": 3.2426546367183335,
+      "learning_rate": 2.2129850156599408e-07,
+      "loss": 0.8365,
+      "step": 9712
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.2949,
+      "step": 9712,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.6369,
+      "step": 9712,
+      "vm_loss": 0.1262
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.5325,
+      "step": 9712,
+      "vm_loss": 0.1819
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.5593,
+      "step": 9712,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.5492,
+      "step": 9712,
+      "vm_loss": 0.1597
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.453,
+      "step": 9712,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.5704,
+      "step": 9712,
+      "vm_loss": 0.1371
+    },
+    {
+      "epoch": 1.8696248526120751,
+      "lm_loss": 0.3374,
+      "step": 9712,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.8698173592896503,
+      "grad_norm": 3.1229958273898153,
+      "learning_rate": 2.2064667318907108e-07,
+      "loss": 0.7664,
+      "step": 9713
+    },
+    {
+      "epoch": 1.8700098659672257,
+      "grad_norm": 3.1754028997580828,
+      "learning_rate": 2.1999579549040752e-07,
+      "loss": 0.8223,
+      "step": 9714
+    },
+    {
+      "epoch": 1.8702023726448012,
+      "grad_norm": 3.3722238714422295,
+      "learning_rate": 2.193458685332728e-07,
+      "loss": 0.8716,
+      "step": 9715
+    },
+    {
+      "epoch": 1.8703948793223764,
+      "grad_norm": 3.3717148889086683,
+      "learning_rate": 2.186968923808519e-07,
+      "loss": 0.8529,
+      "step": 9716
+    },
+    {
+      "epoch": 1.870587385999952,
+      "grad_norm": 3.181198161958038,
+      "learning_rate": 2.180488670962322e-07,
+      "loss": 0.8044,
+      "step": 9717
+    },
+    {
+      "epoch": 1.8707798926775272,
+      "grad_norm": 3.085185442809127,
+      "learning_rate": 2.1740179274241103e-07,
+      "loss": 0.7873,
+      "step": 9718
+    },
+    {
+      "epoch": 1.8709723993551026,
+      "grad_norm": 3.1285213877219498,
+      "learning_rate": 2.1675566938229032e-07,
+      "loss": 0.8136,
+      "step": 9719
+    },
+    {
+      "epoch": 1.871164906032678,
+      "grad_norm": 3.2532147430005733,
+      "learning_rate": 2.1611049707868425e-07,
+      "loss": 0.8235,
+      "step": 9720
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.591,
+      "step": 9720,
+      "vm_loss": 0.1376
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.56,
+      "step": 9720,
+      "vm_loss": 0.1869
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.9641,
+      "step": 9720,
+      "vm_loss": 0.0985
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.5021,
+      "step": 9720,
+      "vm_loss": 0.1643
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.6118,
+      "step": 9720,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.5551,
+      "step": 9720,
+      "vm_loss": 0.1958
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.3738,
+      "step": 9720,
+      "vm_loss": 0.1918
+    },
+    {
+      "epoch": 1.871164906032678,
+      "lm_loss": 0.5624,
+      "step": 9720,
+      "vm_loss": 0.1483
+    },
+    {
+      "epoch": 1.8713574127102532,
+      "grad_norm": 3.257328640580475,
+      "learning_rate": 2.1546627589430935e-07,
+      "loss": 0.8603,
+      "step": 9721
+    },
+    {
+      "epoch": 1.8715499193878289,
+      "grad_norm": 3.192045098216215,
+      "learning_rate": 2.1482300589179327e-07,
+      "loss": 0.8265,
+      "step": 9722
+    },
+    {
+      "epoch": 1.871742426065404,
+      "grad_norm": 3.194524546444081,
+      "learning_rate": 2.1418068713367046e-07,
+      "loss": 0.8358,
+      "step": 9723
+    },
+    {
+      "epoch": 1.8719349327429795,
+      "grad_norm": 3.287629593883907,
+      "learning_rate": 2.1353931968238206e-07,
+      "loss": 0.8221,
+      "step": 9724
+    },
+    {
+      "epoch": 1.872127439420555,
+      "grad_norm": 3.488361239853445,
+      "learning_rate": 2.128989036002771e-07,
+      "loss": 0.9036,
+      "step": 9725
+    },
+    {
+      "epoch": 1.8723199460981301,
+      "grad_norm": 3.132195822761052,
+      "learning_rate": 2.122594389496113e-07,
+      "loss": 0.824,
+      "step": 9726
+    },
+    {
+      "epoch": 1.8725124527757058,
+      "grad_norm": 3.292835570162771,
+      "learning_rate": 2.1162092579255055e-07,
+      "loss": 0.846,
+      "step": 9727
+    },
+    {
+      "epoch": 1.872704959453281,
+      "grad_norm": 3.262522182729641,
+      "learning_rate": 2.1098336419116628e-07,
+      "loss": 0.8523,
+      "step": 9728
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.8178,
+      "step": 9728,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.7685,
+      "step": 9728,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.4841,
+      "step": 9728,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.8488,
+      "step": 9728,
+      "vm_loss": 0.1996
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.53,
+      "step": 9728,
+      "vm_loss": 0.1616
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.6234,
+      "step": 9728,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.7864,
+      "step": 9728,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 1.872704959453281,
+      "lm_loss": 0.9771,
+      "step": 9728,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 1.8728974661308564,
+      "grad_norm": 3.1501397936146787,
+      "learning_rate": 2.1034675420743557e-07,
+      "loss": 0.8288,
+      "step": 9729
+    },
+    {
+      "epoch": 1.8730899728084318,
+      "grad_norm": 3.2668787761368185,
+      "learning_rate": 2.097110959032489e-07,
+      "loss": 0.83,
+      "step": 9730
+    },
+    {
+      "epoch": 1.873282479486007,
+      "grad_norm": 3.146049702582496,
+      "learning_rate": 2.0907638934039575e-07,
+      "loss": 0.8289,
+      "step": 9731
+    },
+    {
+      "epoch": 1.8734749861635827,
+      "grad_norm": 3.1668131620811293,
+      "learning_rate": 2.0844263458058122e-07,
+      "loss": 0.7958,
+      "step": 9732
+    },
+    {
+      "epoch": 1.8736674928411579,
+      "grad_norm": 3.3195585644546073,
+      "learning_rate": 2.0780983168541268e-07,
+      "loss": 0.8648,
+      "step": 9733
+    },
+    {
+      "epoch": 1.8738599995187333,
+      "grad_norm": 3.1890995210309288,
+      "learning_rate": 2.0717798071640761e-07,
+      "loss": 0.8216,
+      "step": 9734
+    },
+    {
+      "epoch": 1.8740525061963087,
+      "grad_norm": 3.2677739386101425,
+      "learning_rate": 2.0654708173499016e-07,
+      "loss": 0.8551,
+      "step": 9735
+    },
+    {
+      "epoch": 1.874245012873884,
+      "grad_norm": 3.200486750956276,
+      "learning_rate": 2.0591713480248909e-07,
+      "loss": 0.857,
+      "step": 9736
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.4844,
+      "step": 9736,
+      "vm_loss": 0.1321
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.9613,
+      "step": 9736,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.899,
+      "step": 9736,
+      "vm_loss": 0.1125
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.9652,
+      "step": 9736,
+      "vm_loss": 0.1921
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.5801,
+      "step": 9736,
+      "vm_loss": 0.2294
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.5664,
+      "step": 9736,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.7078,
+      "step": 9736,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 1.874245012873884,
+      "lm_loss": 0.6331,
+      "step": 9736,
+      "vm_loss": 0.1993
+    },
+    {
+      "epoch": 1.8744375195514595,
+      "grad_norm": 3.235305778019497,
+      "learning_rate": 2.0528813998014762e-07,
+      "loss": 0.8556,
+      "step": 9737
+    },
+    {
+      "epoch": 1.8746300262290347,
+      "grad_norm": 3.291903340896733,
+      "learning_rate": 2.0466009732910908e-07,
+      "loss": 0.8721,
+      "step": 9738
+    },
+    {
+      "epoch": 1.8748225329066102,
+      "grad_norm": 3.3385508919303493,
+      "learning_rate": 2.0403300691042906e-07,
+      "loss": 0.8523,
+      "step": 9739
+    },
+    {
+      "epoch": 1.8750150395841856,
+      "grad_norm": 3.309376141655951,
+      "learning_rate": 2.0340686878506655e-07,
+      "loss": 0.8545,
+      "step": 9740
+    },
+    {
+      "epoch": 1.875207546261761,
+      "grad_norm": 3.0862666470057896,
+      "learning_rate": 2.0278168301389177e-07,
+      "loss": 0.812,
+      "step": 9741
+    },
+    {
+      "epoch": 1.8754000529393364,
+      "grad_norm": 3.270874192099001,
+      "learning_rate": 2.0215744965768057e-07,
+      "loss": 0.8878,
+      "step": 9742
+    },
+    {
+      "epoch": 1.8755925596169116,
+      "grad_norm": 3.3824207127536945,
+      "learning_rate": 2.0153416877711552e-07,
+      "loss": 0.9079,
+      "step": 9743
+    },
+    {
+      "epoch": 1.875785066294487,
+      "grad_norm": 3.3554009239298974,
+      "learning_rate": 2.0091184043278923e-07,
+      "loss": 0.8626,
+      "step": 9744
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.46,
+      "step": 9744,
+      "vm_loss": 0.1994
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.701,
+      "step": 9744,
+      "vm_loss": 0.1157
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.4721,
+      "step": 9744,
+      "vm_loss": 0.1697
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.6171,
+      "step": 9744,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.5504,
+      "step": 9744,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.7514,
+      "step": 9744,
+      "vm_loss": 0.1125
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.5252,
+      "step": 9744,
+      "vm_loss": 0.136
+    },
+    {
+      "epoch": 1.875785066294487,
+      "lm_loss": 0.7013,
+      "step": 9744,
+      "vm_loss": 0.2158
+    },
+    {
+      "epoch": 1.8759775729720625,
+      "grad_norm": 3.199373082523442,
+      "learning_rate": 2.0029046468519885e-07,
+      "loss": 0.8143,
+      "step": 9745
+    },
+    {
+      "epoch": 1.876170079649638,
+      "grad_norm": 3.1163404396667014,
+      "learning_rate": 1.9967004159474946e-07,
+      "loss": 0.7806,
+      "step": 9746
+    },
+    {
+      "epoch": 1.8763625863272133,
+      "grad_norm": 3.133494924560944,
+      "learning_rate": 1.990505712217561e-07,
+      "loss": 0.8267,
+      "step": 9747
+    },
+    {
+      "epoch": 1.8765550930047885,
+      "grad_norm": 3.337141915380523,
+      "learning_rate": 1.9843205362643724e-07,
+      "loss": 0.8555,
+      "step": 9748
+    },
+    {
+      "epoch": 1.876747599682364,
+      "grad_norm": 3.4311856183189002,
+      "learning_rate": 1.978144888689204e-07,
+      "loss": 0.8794,
+      "step": 9749
+    },
+    {
+      "epoch": 1.8769401063599394,
+      "grad_norm": 3.2747941313657574,
+      "learning_rate": 1.9719787700924308e-07,
+      "loss": 0.8788,
+      "step": 9750
+    },
+    {
+      "epoch": 1.8771326130375148,
+      "grad_norm": 3.1706260957535455,
+      "learning_rate": 1.9658221810734735e-07,
+      "loss": 0.8399,
+      "step": 9751
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "grad_norm": 3.2379514835818037,
+      "learning_rate": 1.9596751222308085e-07,
+      "loss": 0.8317,
+      "step": 9752
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.8049,
+      "step": 9752,
+      "vm_loss": 0.1098
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.5397,
+      "step": 9752,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.9043,
+      "step": 9752,
+      "vm_loss": 0.132
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.899,
+      "step": 9752,
+      "vm_loss": 0.2533
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.5991,
+      "step": 9752,
+      "vm_loss": 0.1693
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.4712,
+      "step": 9752,
+      "vm_loss": 0.14
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.8153,
+      "step": 9752,
+      "vm_loss": 0.2072
+    },
+    {
+      "epoch": 1.8773251197150902,
+      "lm_loss": 0.4569,
+      "step": 9752,
+      "vm_loss": 0.1898
+    },
+    {
+      "epoch": 1.8775176263926654,
+      "grad_norm": 3.0641745742921978,
+      "learning_rate": 1.953537594162025e-07,
+      "loss": 0.783,
+      "step": 9753
+    },
+    {
+      "epoch": 1.877710133070241,
+      "grad_norm": 3.1238782542711694,
+      "learning_rate": 1.9474095974637673e-07,
+      "loss": 0.7744,
+      "step": 9754
+    },
+    {
+      "epoch": 1.8779026397478162,
+      "grad_norm": 3.305386572267922,
+      "learning_rate": 1.9412911327317595e-07,
+      "loss": 0.8296,
+      "step": 9755
+    },
+    {
+      "epoch": 1.8780951464253917,
+      "grad_norm": 3.0035863329513877,
+      "learning_rate": 1.9351822005607922e-07,
+      "loss": 0.7795,
+      "step": 9756
+    },
+    {
+      "epoch": 1.878287653102967,
+      "grad_norm": 3.319816005087962,
+      "learning_rate": 1.9290828015447237e-07,
+      "loss": 0.876,
+      "step": 9757
+    },
+    {
+      "epoch": 1.8784801597805423,
+      "grad_norm": 3.3085398684839333,
+      "learning_rate": 1.922992936276513e-07,
+      "loss": 0.8475,
+      "step": 9758
+    },
+    {
+      "epoch": 1.878672666458118,
+      "grad_norm": 3.1397651346353146,
+      "learning_rate": 1.9169126053481534e-07,
+      "loss": 0.8107,
+      "step": 9759
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "grad_norm": 3.3281318374987263,
+      "learning_rate": 1.9108418093507385e-07,
+      "loss": 0.8659,
+      "step": 9760
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.7862,
+      "step": 9760,
+      "vm_loss": 0.1715
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.7759,
+      "step": 9760,
+      "vm_loss": 0.1916
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.825,
+      "step": 9760,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.5529,
+      "step": 9760,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.6048,
+      "step": 9760,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.6929,
+      "step": 9760,
+      "vm_loss": 0.1197
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.7371,
+      "step": 9760,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.8788651731356931,
+      "lm_loss": 0.5864,
+      "step": 9760,
+      "vm_loss": 0.1467
+    },
+    {
+      "epoch": 1.8790576798132685,
+      "grad_norm": 3.2187354431328825,
+      "learning_rate": 1.9047805488744188e-07,
+      "loss": 0.8416,
+      "step": 9761
+    },
+    {
+      "epoch": 1.879250186490844,
+      "grad_norm": 3.2612582877319465,
+      "learning_rate": 1.8987288245084556e-07,
+      "loss": 0.8063,
+      "step": 9762
+    },
+    {
+      "epoch": 1.8794426931684192,
+      "grad_norm": 3.257849488628638,
+      "learning_rate": 1.8926866368411012e-07,
+      "loss": 0.846,
+      "step": 9763
+    },
+    {
+      "epoch": 1.8796351998459948,
+      "grad_norm": 3.202178638773855,
+      "learning_rate": 1.8866539864597855e-07,
+      "loss": 0.8241,
+      "step": 9764
+    },
+    {
+      "epoch": 1.87982770652357,
+      "grad_norm": 3.3844732573164262,
+      "learning_rate": 1.8806308739509394e-07,
+      "loss": 0.9206,
+      "step": 9765
+    },
+    {
+      "epoch": 1.8800202132011454,
+      "grad_norm": 3.228501560165754,
+      "learning_rate": 1.8746172999000722e-07,
+      "loss": 0.8395,
+      "step": 9766
+    },
+    {
+      "epoch": 1.8802127198787209,
+      "grad_norm": 3.3034073993584556,
+      "learning_rate": 1.8686132648918055e-07,
+      "loss": 0.8477,
+      "step": 9767
+    },
+    {
+      "epoch": 1.880405226556296,
+      "grad_norm": 3.3027977811351277,
+      "learning_rate": 1.862618769509794e-07,
+      "loss": 0.8428,
+      "step": 9768
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.9262,
+      "step": 9768,
+      "vm_loss": 0.1172
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.6772,
+      "step": 9768,
+      "vm_loss": 0.1477
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 1.2286,
+      "step": 9768,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.7982,
+      "step": 9768,
+      "vm_loss": 0.0933
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.5127,
+      "step": 9768,
+      "vm_loss": 0.2095
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.5794,
+      "step": 9768,
+      "vm_loss": 0.1438
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.5133,
+      "step": 9768,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.880405226556296,
+      "lm_loss": 0.5285,
+      "step": 9768,
+      "vm_loss": 0.1362
+    },
+    {
+      "epoch": 1.8805977332338717,
+      "grad_norm": 3.1662506759369897,
+      "learning_rate": 1.8566338143367835e-07,
+      "loss": 0.7944,
+      "step": 9769
+    },
+    {
+      "epoch": 1.880790239911447,
+      "grad_norm": 3.580192771625688,
+      "learning_rate": 1.8506583999545747e-07,
+      "loss": 0.9072,
+      "step": 9770
+    },
+    {
+      "epoch": 1.8809827465890223,
+      "grad_norm": 3.284328396357913,
+      "learning_rate": 1.8446925269440806e-07,
+      "loss": 0.8582,
+      "step": 9771
+    },
+    {
+      "epoch": 1.8811752532665977,
+      "grad_norm": 3.336742070100333,
+      "learning_rate": 1.8387361958852378e-07,
+      "loss": 0.8878,
+      "step": 9772
+    },
+    {
+      "epoch": 1.881367759944173,
+      "grad_norm": 3.2327159552030915,
+      "learning_rate": 1.8327894073570828e-07,
+      "loss": 0.8391,
+      "step": 9773
+    },
+    {
+      "epoch": 1.8815602666217486,
+      "grad_norm": 3.129188841043906,
+      "learning_rate": 1.8268521619377312e-07,
+      "loss": 0.7853,
+      "step": 9774
+    },
+    {
+      "epoch": 1.8817527732993238,
+      "grad_norm": 3.300167298964028,
+      "learning_rate": 1.820924460204354e-07,
+      "loss": 0.8463,
+      "step": 9775
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "grad_norm": 3.02643565181241,
+      "learning_rate": 1.8150063027331798e-07,
+      "loss": 0.7795,
+      "step": 9776
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.3679,
+      "step": 9776,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.658,
+      "step": 9776,
+      "vm_loss": 0.1595
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.475,
+      "step": 9776,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.6515,
+      "step": 9776,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.5932,
+      "step": 9776,
+      "vm_loss": 0.1187
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.6392,
+      "step": 9776,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.7081,
+      "step": 9776,
+      "vm_loss": 0.1482
+    },
+    {
+      "epoch": 1.8819452799768992,
+      "lm_loss": 0.7175,
+      "step": 9776,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.8821377866544746,
+      "grad_norm": 3.380567422586673,
+      "learning_rate": 1.8090976900995593e-07,
+      "loss": 0.8739,
+      "step": 9777
+    },
+    {
+      "epoch": 1.8823302933320498,
+      "grad_norm": 3.2600991120257357,
+      "learning_rate": 1.8031986228778664e-07,
+      "loss": 0.8276,
+      "step": 9778
+    },
+    {
+      "epoch": 1.8825228000096255,
+      "grad_norm": 3.2616190084311225,
+      "learning_rate": 1.7973091016415755e-07,
+      "loss": 0.8413,
+      "step": 9779
+    },
+    {
+      "epoch": 1.8827153066872007,
+      "grad_norm": 3.263726286884041,
+      "learning_rate": 1.7914291269632067e-07,
+      "loss": 0.8736,
+      "step": 9780
+    },
+    {
+      "epoch": 1.882907813364776,
+      "grad_norm": 3.146893889939857,
+      "learning_rate": 1.7855586994144025e-07,
+      "loss": 0.8242,
+      "step": 9781
+    },
+    {
+      "epoch": 1.8831003200423515,
+      "grad_norm": 3.367488898063329,
+      "learning_rate": 1.7796978195658067e-07,
+      "loss": 0.8462,
+      "step": 9782
+    },
+    {
+      "epoch": 1.8832928267199267,
+      "grad_norm": 3.132580133337081,
+      "learning_rate": 1.7738464879871963e-07,
+      "loss": 0.8128,
+      "step": 9783
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "grad_norm": 3.2646899588958855,
+      "learning_rate": 1.7680047052473837e-07,
+      "loss": 0.7963,
+      "step": 9784
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.5743,
+      "step": 9784,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.9274,
+      "step": 9784,
+      "vm_loss": 0.2082
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.5862,
+      "step": 9784,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 1.095,
+      "step": 9784,
+      "vm_loss": 0.1471
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.7025,
+      "step": 9784,
+      "vm_loss": 0.1794
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.7589,
+      "step": 9784,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.4675,
+      "step": 9784,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.8834853333975023,
+      "lm_loss": 0.7371,
+      "step": 9784,
+      "vm_loss": 0.17
+    },
+    {
+      "epoch": 1.8836778400750775,
+      "grad_norm": 3.235645300484035,
+      "learning_rate": 1.7621724719142698e-07,
+      "loss": 0.8569,
+      "step": 9785
+    },
+    {
+      "epoch": 1.883870346752653,
+      "grad_norm": 3.2626734959587886,
+      "learning_rate": 1.756349788554812e-07,
+      "loss": 0.8253,
+      "step": 9786
+    },
+    {
+      "epoch": 1.8840628534302284,
+      "grad_norm": 3.1142683895937737,
+      "learning_rate": 1.750536655735069e-07,
+      "loss": 0.786,
+      "step": 9787
+    },
+    {
+      "epoch": 1.8842553601078036,
+      "grad_norm": 3.196050184789095,
+      "learning_rate": 1.744733074020133e-07,
+      "loss": 0.8077,
+      "step": 9788
+    },
+    {
+      "epoch": 1.8844478667853792,
+      "grad_norm": 3.2693673177035056,
+      "learning_rate": 1.738939043974197e-07,
+      "loss": 0.7947,
+      "step": 9789
+    },
+    {
+      "epoch": 1.8846403734629544,
+      "grad_norm": 3.2859657650689242,
+      "learning_rate": 1.7331545661604997e-07,
+      "loss": 0.8448,
+      "step": 9790
+    },
+    {
+      "epoch": 1.8848328801405299,
+      "grad_norm": 3.267741855875794,
+      "learning_rate": 1.7273796411414023e-07,
+      "loss": 0.8529,
+      "step": 9791
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "grad_norm": 3.288271549946051,
+      "learning_rate": 1.7216142694782668e-07,
+      "loss": 0.8336,
+      "step": 9792
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.8748,
+      "step": 9792,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.4924,
+      "step": 9792,
+      "vm_loss": 0.1469
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.4441,
+      "step": 9792,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.5663,
+      "step": 9792,
+      "vm_loss": 0.0972
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 1.1189,
+      "step": 9792,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.8396,
+      "step": 9792,
+      "vm_loss": 0.1673
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.7968,
+      "step": 9792,
+      "vm_loss": 0.1779
+    },
+    {
+      "epoch": 1.8850253868181053,
+      "lm_loss": 0.9524,
+      "step": 9792,
+      "vm_loss": 0.185
+    },
+    {
+      "epoch": 1.8852178934956805,
+      "grad_norm": 3.4493391682207535,
+      "learning_rate": 1.7158584517315667e-07,
+      "loss": 0.8506,
+      "step": 9793
+    },
+    {
+      "epoch": 1.8854104001732561,
+      "grad_norm": 3.271613339158465,
+      "learning_rate": 1.710112188460844e-07,
+      "loss": 0.8623,
+      "step": 9794
+    },
+    {
+      "epoch": 1.8856029068508313,
+      "grad_norm": 3.380199588598048,
+      "learning_rate": 1.7043754802247293e-07,
+      "loss": 0.8605,
+      "step": 9795
+    },
+    {
+      "epoch": 1.8857954135284067,
+      "grad_norm": 3.228334940001412,
+      "learning_rate": 1.6986483275808762e-07,
+      "loss": 0.8377,
+      "step": 9796
+    },
+    {
+      "epoch": 1.8859879202059822,
+      "grad_norm": 3.0653181909163085,
+      "learning_rate": 1.6929307310860398e-07,
+      "loss": 0.7453,
+      "step": 9797
+    },
+    {
+      "epoch": 1.8861804268835574,
+      "grad_norm": 3.065149088907136,
+      "learning_rate": 1.6872226912960532e-07,
+      "loss": 0.7934,
+      "step": 9798
+    },
+    {
+      "epoch": 1.886372933561133,
+      "grad_norm": 3.322195596465199,
+      "learning_rate": 1.681524208765817e-07,
+      "loss": 0.8122,
+      "step": 9799
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "grad_norm": 3.2393131813235705,
+      "learning_rate": 1.6758352840492874e-07,
+      "loss": 0.8148,
+      "step": 9800
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.7451,
+      "step": 9800,
+      "vm_loss": 0.0873
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.6702,
+      "step": 9800,
+      "vm_loss": 0.1843
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 1.2447,
+      "step": 9800,
+      "vm_loss": 0.1972
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.8943,
+      "step": 9800,
+      "vm_loss": 0.159
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 1.105,
+      "step": 9800,
+      "vm_loss": 0.1512
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.4952,
+      "step": 9800,
+      "vm_loss": 0.1787
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.4821,
+      "step": 9800,
+      "vm_loss": 0.1832
+    },
+    {
+      "epoch": 1.8865654402387082,
+      "lm_loss": 0.6276,
+      "step": 9800,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.8867579469162836,
+      "grad_norm": 3.2310193654960386,
+      "learning_rate": 1.6701559176994897e-07,
+      "loss": 0.8394,
+      "step": 9801
+    },
+    {
+      "epoch": 1.886950453593859,
+      "grad_norm": 3.079022742477775,
+      "learning_rate": 1.6644861102685595e-07,
+      "loss": 0.7981,
+      "step": 9802
+    },
+    {
+      "epoch": 1.8871429602714345,
+      "grad_norm": 3.088157744894846,
+      "learning_rate": 1.6588258623076448e-07,
+      "loss": 0.7396,
+      "step": 9803
+    },
+    {
+      "epoch": 1.8873354669490099,
+      "grad_norm": 3.2448328673875704,
+      "learning_rate": 1.6531751743670053e-07,
+      "loss": 0.837,
+      "step": 9804
+    },
+    {
+      "epoch": 1.887527973626585,
+      "grad_norm": 3.445973116057066,
+      "learning_rate": 1.6475340469959689e-07,
+      "loss": 0.8919,
+      "step": 9805
+    },
+    {
+      "epoch": 1.8877204803041605,
+      "grad_norm": 3.2088208986589937,
+      "learning_rate": 1.6419024807429296e-07,
+      "loss": 0.8406,
+      "step": 9806
+    },
+    {
+      "epoch": 1.887912986981736,
+      "grad_norm": 3.3249714268679655,
+      "learning_rate": 1.6362804761553164e-07,
+      "loss": 0.8394,
+      "step": 9807
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "grad_norm": 3.1725927716841937,
+      "learning_rate": 1.630668033779692e-07,
+      "loss": 0.8087,
+      "step": 9808
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.5117,
+      "step": 9808,
+      "vm_loss": 0.1073
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.7332,
+      "step": 9808,
+      "vm_loss": 0.147
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.8572,
+      "step": 9808,
+      "vm_loss": 0.1405
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.6686,
+      "step": 9808,
+      "vm_loss": 0.261
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.7398,
+      "step": 9808,
+      "vm_loss": 0.1604
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.4486,
+      "step": 9808,
+      "vm_loss": 0.1692
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.5053,
+      "step": 9808,
+      "vm_loss": 0.15
+    },
+    {
+      "epoch": 1.8881054936593114,
+      "lm_loss": 0.6069,
+      "step": 9808,
+      "vm_loss": 0.1116
+    },
+    {
+      "epoch": 1.8882980003368868,
+      "grad_norm": 3.4642503341438844,
+      "learning_rate": 1.6250651541616426e-07,
+      "loss": 0.8963,
+      "step": 9809
+    },
+    {
+      "epoch": 1.888490507014462,
+      "grad_norm": 3.273486333381402,
+      "learning_rate": 1.619471837845854e-07,
+      "loss": 0.8556,
+      "step": 9810
+    },
+    {
+      "epoch": 1.8886830136920374,
+      "grad_norm": 3.2914875975985436,
+      "learning_rate": 1.6138880853760364e-07,
+      "loss": 0.8024,
+      "step": 9811
+    },
+    {
+      "epoch": 1.8888755203696128,
+      "grad_norm": 3.3718672402173944,
+      "learning_rate": 1.608313897295033e-07,
+      "loss": 0.8387,
+      "step": 9812
+    },
+    {
+      "epoch": 1.8890680270471882,
+      "grad_norm": 3.3052960490828283,
+      "learning_rate": 1.6027492741447214e-07,
+      "loss": 0.8813,
+      "step": 9813
+    },
+    {
+      "epoch": 1.8892605337247637,
+      "grad_norm": 3.26964543524852,
+      "learning_rate": 1.5971942164660359e-07,
+      "loss": 0.8378,
+      "step": 9814
+    },
+    {
+      "epoch": 1.8894530404023389,
+      "grad_norm": 3.181049582313773,
+      "learning_rate": 1.591648724799022e-07,
+      "loss": 0.8058,
+      "step": 9815
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "grad_norm": 3.2651242285148045,
+      "learning_rate": 1.5861127996827597e-07,
+      "loss": 0.8509,
+      "step": 9816
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.4566,
+      "step": 9816,
+      "vm_loss": 0.1377
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.9221,
+      "step": 9816,
+      "vm_loss": 0.1208
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 1.0489,
+      "step": 9816,
+      "vm_loss": 0.1557
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.757,
+      "step": 9816,
+      "vm_loss": 0.155
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.4552,
+      "step": 9816,
+      "vm_loss": 0.1598
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.8636,
+      "step": 9816,
+      "vm_loss": 0.135
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.3805,
+      "step": 9816,
+      "vm_loss": 0.1142
+    },
+    {
+      "epoch": 1.8896455470799145,
+      "lm_loss": 0.5552,
+      "step": 9816,
+      "vm_loss": 0.1545
+    },
+    {
+      "epoch": 1.8898380537574897,
+      "grad_norm": 3.1680551751191195,
+      "learning_rate": 1.5805864416554185e-07,
+      "loss": 0.7921,
+      "step": 9817
+    },
+    {
+      "epoch": 1.8900305604350651,
+      "grad_norm": 3.2000399031928617,
+      "learning_rate": 1.575069651254224e-07,
+      "loss": 0.7975,
+      "step": 9818
+    },
+    {
+      "epoch": 1.8902230671126405,
+      "grad_norm": 3.1308927697117523,
+      "learning_rate": 1.569562429015481e-07,
+      "loss": 0.7988,
+      "step": 9819
+    },
+    {
+      "epoch": 1.8904155737902157,
+      "grad_norm": 3.166668783976193,
+      "learning_rate": 1.5640647754745607e-07,
+      "loss": 0.8149,
+      "step": 9820
+    },
+    {
+      "epoch": 1.8906080804677914,
+      "grad_norm": 3.364591069347157,
+      "learning_rate": 1.558576691165914e-07,
+      "loss": 0.8624,
+      "step": 9821
+    },
+    {
+      "epoch": 1.8908005871453666,
+      "grad_norm": 3.3995215959921588,
+      "learning_rate": 1.5530981766230356e-07,
+      "loss": 0.8447,
+      "step": 9822
+    },
+    {
+      "epoch": 1.890993093822942,
+      "grad_norm": 3.2098287751045365,
+      "learning_rate": 1.5476292323785337e-07,
+      "loss": 0.8205,
+      "step": 9823
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "grad_norm": 3.112697861216151,
+      "learning_rate": 1.5421698589640267e-07,
+      "loss": 0.7992,
+      "step": 9824
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.413,
+      "step": 9824,
+      "vm_loss": 0.1581
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.6287,
+      "step": 9824,
+      "vm_loss": 0.1685
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.754,
+      "step": 9824,
+      "vm_loss": 0.1359
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.269,
+      "step": 9824,
+      "vm_loss": 0.138
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.8331,
+      "step": 9824,
+      "vm_loss": 0.1316
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.5199,
+      "step": 9824,
+      "vm_loss": 0.0919
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.5232,
+      "step": 9824,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.8911856005005174,
+      "lm_loss": 0.5469,
+      "step": 9824,
+      "vm_loss": 0.1868
+    },
+    {
+      "epoch": 1.8913781071780926,
+      "grad_norm": 3.2805933143061763,
+      "learning_rate": 1.536720056910268e-07,
+      "loss": 0.815,
+      "step": 9825
+    },
+    {
+      "epoch": 1.8915706138556683,
+      "grad_norm": 3.2215172728935597,
+      "learning_rate": 1.531279826747023e-07,
+      "loss": 0.8576,
+      "step": 9826
+    },
+    {
+      "epoch": 1.8917631205332435,
+      "grad_norm": 3.0919364531973725,
+      "learning_rate": 1.5258491690031796e-07,
+      "loss": 0.7929,
+      "step": 9827
+    },
+    {
+      "epoch": 1.891955627210819,
+      "grad_norm": 3.3238891873050016,
+      "learning_rate": 1.5204280842066266e-07,
+      "loss": 0.8445,
+      "step": 9828
+    },
+    {
+      "epoch": 1.8921481338883943,
+      "grad_norm": 3.553661107670562,
+      "learning_rate": 1.515016572884398e-07,
+      "loss": 0.9092,
+      "step": 9829
+    },
+    {
+      "epoch": 1.8923406405659695,
+      "grad_norm": 3.1649727484272643,
+      "learning_rate": 1.5096146355625396e-07,
+      "loss": 0.8489,
+      "step": 9830
+    },
+    {
+      "epoch": 1.8925331472435452,
+      "grad_norm": 3.4501169700865266,
+      "learning_rate": 1.5042222727662092e-07,
+      "loss": 0.8249,
+      "step": 9831
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "grad_norm": 3.0495338169042023,
+      "learning_rate": 1.4988394850195876e-07,
+      "loss": 0.7709,
+      "step": 9832
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.5521,
+      "step": 9832,
+      "vm_loss": 0.1605
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.6678,
+      "step": 9832,
+      "vm_loss": 0.151
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.5387,
+      "step": 9832,
+      "vm_loss": 0.1515
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.8205,
+      "step": 9832,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.8843,
+      "step": 9832,
+      "vm_loss": 0.1452
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.5839,
+      "step": 9832,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.9296,
+      "step": 9832,
+      "vm_loss": 0.1849
+    },
+    {
+      "epoch": 1.8927256539211204,
+      "lm_loss": 0.8936,
+      "step": 9832,
+      "vm_loss": 0.1942
+    },
+    {
+      "epoch": 1.8929181605986958,
+      "grad_norm": 3.3740895263904562,
+      "learning_rate": 1.4934662728459893e-07,
+      "loss": 0.8342,
+      "step": 9833
+    },
+    {
+      "epoch": 1.8931106672762712,
+      "grad_norm": 3.3515787827393297,
+      "learning_rate": 1.4881026367677188e-07,
+      "loss": 0.9085,
+      "step": 9834
+    },
+    {
+      "epoch": 1.8933031739538464,
+      "grad_norm": 3.2759730607561797,
+      "learning_rate": 1.4827485773062146e-07,
+      "loss": 0.847,
+      "step": 9835
+    },
+    {
+      "epoch": 1.893495680631422,
+      "grad_norm": 3.2303666056170166,
+      "learning_rate": 1.4774040949819378e-07,
+      "loss": 0.7955,
+      "step": 9836
+    },
+    {
+      "epoch": 1.8936881873089972,
+      "grad_norm": 3.3719887987707042,
+      "learning_rate": 1.4720691903144734e-07,
+      "loss": 0.8362,
+      "step": 9837
+    },
+    {
+      "epoch": 1.8938806939865727,
+      "grad_norm": 3.115100220963446,
+      "learning_rate": 1.4667438638224064e-07,
+      "loss": 0.8204,
+      "step": 9838
+    },
+    {
+      "epoch": 1.894073200664148,
+      "grad_norm": 3.232999195490805,
+      "learning_rate": 1.4614281160234555e-07,
+      "loss": 0.8367,
+      "step": 9839
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "grad_norm": 3.106815473788987,
+      "learning_rate": 1.4561219474343747e-07,
+      "loss": 0.7957,
+      "step": 9840
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.546,
+      "step": 9840,
+      "vm_loss": 0.1656
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.5616,
+      "step": 9840,
+      "vm_loss": 0.1984
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.7282,
+      "step": 9840,
+      "vm_loss": 0.1651
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.8378,
+      "step": 9840,
+      "vm_loss": 0.1423
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 1.0955,
+      "step": 9840,
+      "vm_loss": 0.1485
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 1.0232,
+      "step": 9840,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.4757,
+      "step": 9840,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.8942657073417233,
+      "lm_loss": 0.5586,
+      "step": 9840,
+      "vm_loss": 0.1867
+    },
+    {
+      "epoch": 1.894458214019299,
+      "grad_norm": 3.3231997355642933,
+      "learning_rate": 1.450825358570973e-07,
+      "loss": 0.8336,
+      "step": 9841
+    },
+    {
+      "epoch": 1.8946507206968741,
+      "grad_norm": 3.19496190043342,
+      "learning_rate": 1.4455383499481613e-07,
+      "loss": 0.8259,
+      "step": 9842
+    },
+    {
+      "epoch": 1.8948432273744495,
+      "grad_norm": 3.123342149495041,
+      "learning_rate": 1.4402609220799057e-07,
+      "loss": 0.8377,
+      "step": 9843
+    },
+    {
+      "epoch": 1.895035734052025,
+      "grad_norm": 3.2773964625620744,
+      "learning_rate": 1.4349930754792408e-07,
+      "loss": 0.8202,
+      "step": 9844
+    },
+    {
+      "epoch": 1.8952282407296002,
+      "grad_norm": 3.268955680949422,
+      "learning_rate": 1.4297348106582454e-07,
+      "loss": 0.8392,
+      "step": 9845
+    },
+    {
+      "epoch": 1.8954207474071758,
+      "grad_norm": 3.2694236180845584,
+      "learning_rate": 1.4244861281281218e-07,
+      "loss": 0.8205,
+      "step": 9846
+    },
+    {
+      "epoch": 1.895613254084751,
+      "grad_norm": 3.111651000924856,
+      "learning_rate": 1.4192470283990956e-07,
+      "loss": 0.8092,
+      "step": 9847
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "grad_norm": 3.24948017852796,
+      "learning_rate": 1.41401751198047e-07,
+      "loss": 0.8289,
+      "step": 9848
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.5715,
+      "step": 9848,
+      "vm_loss": 0.2043
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.7065,
+      "step": 9848,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.3542,
+      "step": 9848,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.7414,
+      "step": 9848,
+      "vm_loss": 0.173
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.3797,
+      "step": 9848,
+      "vm_loss": 0.1307
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.4394,
+      "step": 9848,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.5934,
+      "step": 9848,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.8958057607623264,
+      "lm_loss": 0.7687,
+      "step": 9848,
+      "vm_loss": 0.1566
+    },
+    {
+      "epoch": 1.8959982674399019,
+      "grad_norm": 3.0111185322965954,
+      "learning_rate": 1.4087975793806275e-07,
+      "loss": 0.7772,
+      "step": 9849
+    },
+    {
+      "epoch": 1.896190774117477,
+      "grad_norm": 3.35860871610262,
+      "learning_rate": 1.4035872311070175e-07,
+      "loss": 0.8679,
+      "step": 9850
+    },
+    {
+      "epoch": 1.8963832807950527,
+      "grad_norm": 3.324715125580184,
+      "learning_rate": 1.3983864676661352e-07,
+      "loss": 0.8518,
+      "step": 9851
+    },
+    {
+      "epoch": 1.896575787472628,
+      "grad_norm": 3.4247409431170412,
+      "learning_rate": 1.3931952895635649e-07,
+      "loss": 0.8679,
+      "step": 9852
+    },
+    {
+      "epoch": 1.8967682941502033,
+      "grad_norm": 3.2369895750285305,
+      "learning_rate": 1.3880136973039692e-07,
+      "loss": 0.8372,
+      "step": 9853
+    },
+    {
+      "epoch": 1.8969608008277787,
+      "grad_norm": 3.447736996587697,
+      "learning_rate": 1.3828416913910682e-07,
+      "loss": 0.8871,
+      "step": 9854
+    },
+    {
+      "epoch": 1.897153307505354,
+      "grad_norm": 3.0715163460263057,
+      "learning_rate": 1.3776792723276144e-07,
+      "loss": 0.8017,
+      "step": 9855
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "grad_norm": 3.3755369082685713,
+      "learning_rate": 1.372526440615507e-07,
+      "loss": 0.8642,
+      "step": 9856
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.6191,
+      "step": 9856,
+      "vm_loss": 0.1634
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.6152,
+      "step": 9856,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.7694,
+      "step": 9856,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.4861,
+      "step": 9856,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.6503,
+      "step": 9856,
+      "vm_loss": 0.1445
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.52,
+      "step": 9856,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.6167,
+      "step": 9856,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.8973458141829296,
+      "lm_loss": 0.5787,
+      "step": 9856,
+      "vm_loss": 0.1091
+    },
+    {
+      "epoch": 1.8975383208605048,
+      "grad_norm": 3.338969650245965,
+      "learning_rate": 1.3673831967556228e-07,
+      "loss": 0.8449,
+      "step": 9857
+    },
+    {
+      "epoch": 1.8977308275380802,
+      "grad_norm": 3.202734200480825,
+      "learning_rate": 1.362249541247984e-07,
+      "loss": 0.8035,
+      "step": 9858
+    },
+    {
+      "epoch": 1.8979233342156556,
+      "grad_norm": 3.3410825440864502,
+      "learning_rate": 1.3571254745916363e-07,
+      "loss": 0.8668,
+      "step": 9859
+    },
+    {
+      "epoch": 1.8981158408932308,
+      "grad_norm": 3.2925573251381555,
+      "learning_rate": 1.3520109972846918e-07,
+      "loss": 0.8554,
+      "step": 9860
+    },
+    {
+      "epoch": 1.8983083475708065,
+      "grad_norm": 3.350636137787506,
+      "learning_rate": 1.3469061098243642e-07,
+      "loss": 0.8387,
+      "step": 9861
+    },
+    {
+      "epoch": 1.8985008542483817,
+      "grad_norm": 3.189333328035987,
+      "learning_rate": 1.3418108127069008e-07,
+      "loss": 0.8125,
+      "step": 9862
+    },
+    {
+      "epoch": 1.898693360925957,
+      "grad_norm": 3.25690960268992,
+      "learning_rate": 1.3367251064276386e-07,
+      "loss": 0.8412,
+      "step": 9863
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "grad_norm": 3.1914562728521374,
+      "learning_rate": 1.3316489914809605e-07,
+      "loss": 0.7999,
+      "step": 9864
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.6434,
+      "step": 9864,
+      "vm_loss": 0.2341
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.8401,
+      "step": 9864,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.5685,
+      "step": 9864,
+      "vm_loss": 0.2014
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.4253,
+      "step": 9864,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.8736,
+      "step": 9864,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.8423,
+      "step": 9864,
+      "vm_loss": 0.1501
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.6308,
+      "step": 9864,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 1.8988858676035325,
+      "lm_loss": 0.6927,
+      "step": 9864,
+      "vm_loss": 0.1426
+    },
+    {
+      "epoch": 1.899078374281108,
+      "grad_norm": 3.37438863711428,
+      "learning_rate": 1.326582468360338e-07,
+      "loss": 0.8834,
+      "step": 9865
+    },
+    {
+      "epoch": 1.8992708809586833,
+      "grad_norm": 3.318364109074891,
+      "learning_rate": 1.3215255375582992e-07,
+      "loss": 0.8713,
+      "step": 9866
+    },
+    {
+      "epoch": 1.8994633876362585,
+      "grad_norm": 3.3323239664555646,
+      "learning_rate": 1.316478199566462e-07,
+      "loss": 0.8197,
+      "step": 9867
+    },
+    {
+      "epoch": 1.899655894313834,
+      "grad_norm": 3.2079348436948325,
+      "learning_rate": 1.3114404548754566e-07,
+      "loss": 0.7973,
+      "step": 9868
+    },
+    {
+      "epoch": 1.8998484009914094,
+      "grad_norm": 3.143200770719279,
+      "learning_rate": 1.3064123039750353e-07,
+      "loss": 0.819,
+      "step": 9869
+    },
+    {
+      "epoch": 1.9000409076689848,
+      "grad_norm": 3.328543245428068,
+      "learning_rate": 1.3013937473540073e-07,
+      "loss": 0.8418,
+      "step": 9870
+    },
+    {
+      "epoch": 1.9002334143465602,
+      "grad_norm": 3.2624666904693584,
+      "learning_rate": 1.2963847855002264e-07,
+      "loss": 0.854,
+      "step": 9871
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "grad_norm": 3.281379078986719,
+      "learning_rate": 1.291385418900626e-07,
+      "loss": 0.8596,
+      "step": 9872
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.7164,
+      "step": 9872,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.999,
+      "step": 9872,
+      "vm_loss": 0.2213
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.6343,
+      "step": 9872,
+      "vm_loss": 0.2277
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.3848,
+      "step": 9872,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.7855,
+      "step": 9872,
+      "vm_loss": 0.1141
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.515,
+      "step": 9872,
+      "vm_loss": 0.2058
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.5664,
+      "step": 9872,
+      "vm_loss": 0.1808
+    },
+    {
+      "epoch": 1.9004259210241354,
+      "lm_loss": 0.6668,
+      "step": 9872,
+      "vm_loss": 0.1505
+    },
+    {
+      "epoch": 1.9006184277017109,
+      "grad_norm": 3.2115571278196287,
+      "learning_rate": 1.2863956480412276e-07,
+      "loss": 0.8341,
+      "step": 9873
+    },
+    {
+      "epoch": 1.9008109343792863,
+      "grad_norm": 3.2056695215962416,
+      "learning_rate": 1.281415473407077e-07,
+      "loss": 0.8111,
+      "step": 9874
+    },
+    {
+      "epoch": 1.9010034410568617,
+      "grad_norm": 3.127953003319157,
+      "learning_rate": 1.2764448954823204e-07,
+      "loss": 0.785,
+      "step": 9875
+    },
+    {
+      "epoch": 1.9011959477344371,
+      "grad_norm": 3.1517299927787725,
+      "learning_rate": 1.2714839147501712e-07,
+      "loss": 0.8064,
+      "step": 9876
+    },
+    {
+      "epoch": 1.9013884544120123,
+      "grad_norm": 3.1911998113480315,
+      "learning_rate": 1.2665325316928878e-07,
+      "loss": 0.8385,
+      "step": 9877
+    },
+    {
+      "epoch": 1.9015809610895877,
+      "grad_norm": 3.3862107240064536,
+      "learning_rate": 1.2615907467918187e-07,
+      "loss": 0.9024,
+      "step": 9878
+    },
+    {
+      "epoch": 1.9017734677671632,
+      "grad_norm": 3.361385488648857,
+      "learning_rate": 1.2566585605273462e-07,
+      "loss": 0.8677,
+      "step": 9879
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "grad_norm": 3.0676184066174286,
+      "learning_rate": 1.251735973378976e-07,
+      "loss": 0.7921,
+      "step": 9880
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.7984,
+      "step": 9880,
+      "vm_loss": 0.2474
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.8176,
+      "step": 9880,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.6584,
+      "step": 9880,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.4045,
+      "step": 9880,
+      "vm_loss": 0.1529
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.6691,
+      "step": 9880,
+      "vm_loss": 0.1379
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.6673,
+      "step": 9880,
+      "vm_loss": 0.1636
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 1.2163,
+      "step": 9880,
+      "vm_loss": 0.1181
+    },
+    {
+      "epoch": 1.9019659744447386,
+      "lm_loss": 0.9016,
+      "step": 9880,
+      "vm_loss": 0.1778
+    },
+    {
+      "epoch": 1.902158481122314,
+      "grad_norm": 3.5652974434076627,
+      "learning_rate": 1.2468229858252134e-07,
+      "loss": 0.9357,
+      "step": 9881
+    },
+    {
+      "epoch": 1.9023509877998892,
+      "grad_norm": 3.342158192730118,
+      "learning_rate": 1.2419195983436881e-07,
+      "loss": 0.8361,
+      "step": 9882
+    },
+    {
+      "epoch": 1.9025434944774648,
+      "grad_norm": 3.2976363601411336,
+      "learning_rate": 1.2370258114110523e-07,
+      "loss": 0.8614,
+      "step": 9883
+    },
+    {
+      "epoch": 1.90273600115504,
+      "grad_norm": 3.323113895401307,
+      "learning_rate": 1.2321416255030694e-07,
+      "loss": 0.8343,
+      "step": 9884
+    },
+    {
+      "epoch": 1.9029285078326155,
+      "grad_norm": 3.116439863955894,
+      "learning_rate": 1.2272670410945264e-07,
+      "loss": 0.7743,
+      "step": 9885
+    },
+    {
+      "epoch": 1.903121014510191,
+      "grad_norm": 3.2476080037993564,
+      "learning_rate": 1.2224020586593e-07,
+      "loss": 0.8236,
+      "step": 9886
+    },
+    {
+      "epoch": 1.903313521187766,
+      "grad_norm": 3.246154662985664,
+      "learning_rate": 1.2175466786703227e-07,
+      "loss": 0.81,
+      "step": 9887
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "grad_norm": 3.2639114029355936,
+      "learning_rate": 1.2127009015996062e-07,
+      "loss": 0.8546,
+      "step": 9888
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.7507,
+      "step": 9888,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.7853,
+      "step": 9888,
+      "vm_loss": 0.1088
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 1.018,
+      "step": 9888,
+      "vm_loss": 0.1587
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.7663,
+      "step": 9888,
+      "vm_loss": 0.1975
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.6489,
+      "step": 9888,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.5422,
+      "step": 9888,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.7314,
+      "step": 9888,
+      "vm_loss": 0.1537
+    },
+    {
+      "epoch": 1.9035060278653417,
+      "lm_loss": 0.6153,
+      "step": 9888,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 1.903698534542917,
+      "grad_norm": 3.230043045515284,
+      "learning_rate": 1.2078647279182175e-07,
+      "loss": 0.8373,
+      "step": 9889
+    },
+    {
+      "epoch": 1.9038910412204924,
+      "grad_norm": 3.3429548658797508,
+      "learning_rate": 1.2030381580963036e-07,
+      "loss": 0.8423,
+      "step": 9890
+    },
+    {
+      "epoch": 1.9040835478980678,
+      "grad_norm": 3.3006630207801986,
+      "learning_rate": 1.1982211926030551e-07,
+      "loss": 0.8326,
+      "step": 9891
+    },
+    {
+      "epoch": 1.904276054575643,
+      "grad_norm": 3.261601858771046,
+      "learning_rate": 1.1934138319067646e-07,
+      "loss": 0.8742,
+      "step": 9892
+    },
+    {
+      "epoch": 1.9044685612532186,
+      "grad_norm": 3.193077274064226,
+      "learning_rate": 1.1886160764747356e-07,
+      "loss": 0.8441,
+      "step": 9893
+    },
+    {
+      "epoch": 1.9046610679307938,
+      "grad_norm": 3.226297739948393,
+      "learning_rate": 1.1838279267734176e-07,
+      "loss": 0.8236,
+      "step": 9894
+    },
+    {
+      "epoch": 1.9048535746083692,
+      "grad_norm": 2.893082136261642,
+      "learning_rate": 1.1790493832682271e-07,
+      "loss": 0.7491,
+      "step": 9895
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "grad_norm": 3.300736608377902,
+      "learning_rate": 1.1742804464237368e-07,
+      "loss": 0.834,
+      "step": 9896
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.8853,
+      "step": 9896,
+      "vm_loss": 0.2287
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.642,
+      "step": 9896,
+      "vm_loss": 0.1259
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.7415,
+      "step": 9896,
+      "vm_loss": 0.1772
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.6543,
+      "step": 9896,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.4442,
+      "step": 9896,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.6913,
+      "step": 9896,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.4455,
+      "step": 9896,
+      "vm_loss": 0.2599
+    },
+    {
+      "epoch": 1.9050460812859447,
+      "lm_loss": 0.8575,
+      "step": 9896,
+      "vm_loss": 0.1421
+    },
+    {
+      "epoch": 1.9052385879635199,
+      "grad_norm": 3.274786349863418,
+      "learning_rate": 1.1695211167035314e-07,
+      "loss": 0.8625,
+      "step": 9897
+    },
+    {
+      "epoch": 1.9054310946410955,
+      "grad_norm": 3.2923203019976057,
+      "learning_rate": 1.1647713945702966e-07,
+      "loss": 0.8321,
+      "step": 9898
+    },
+    {
+      "epoch": 1.9056236013186707,
+      "grad_norm": 3.034853064152262,
+      "learning_rate": 1.1600312804857405e-07,
+      "loss": 0.8006,
+      "step": 9899
+    },
+    {
+      "epoch": 1.9058161079962461,
+      "grad_norm": 3.3021380821923656,
+      "learning_rate": 1.1553007749106837e-07,
+      "loss": 0.8258,
+      "step": 9900
+    },
+    {
+      "epoch": 1.9060086146738215,
+      "grad_norm": 3.1801359586345725,
+      "learning_rate": 1.1505798783049916e-07,
+      "loss": 0.8271,
+      "step": 9901
+    },
+    {
+      "epoch": 1.9062011213513967,
+      "grad_norm": 3.239456039272056,
+      "learning_rate": 1.1458685911275746e-07,
+      "loss": 0.8313,
+      "step": 9902
+    },
+    {
+      "epoch": 1.9063936280289724,
+      "grad_norm": 3.191016302967127,
+      "learning_rate": 1.1411669138364446e-07,
+      "loss": 0.8207,
+      "step": 9903
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "grad_norm": 3.2500715005467096,
+      "learning_rate": 1.1364748468886688e-07,
+      "loss": 0.8536,
+      "step": 9904
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.7729,
+      "step": 9904,
+      "vm_loss": 0.1633
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.6795,
+      "step": 9904,
+      "vm_loss": 0.1888
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.6828,
+      "step": 9904,
+      "vm_loss": 0.1254
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.7674,
+      "step": 9904,
+      "vm_loss": 0.1731
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.752,
+      "step": 9904,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.4922,
+      "step": 9904,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 1.1116,
+      "step": 9904,
+      "vm_loss": 0.3085
+    },
+    {
+      "epoch": 1.9065861347065476,
+      "lm_loss": 0.411,
+      "step": 9904,
+      "vm_loss": 0.1588
+    },
+    {
+      "epoch": 1.906778641384123,
+      "grad_norm": 3.2610502351835127,
+      "learning_rate": 1.1317923907403827e-07,
+      "loss": 0.8658,
+      "step": 9905
+    },
+    {
+      "epoch": 1.9069711480616984,
+      "grad_norm": 3.2421997131543465,
+      "learning_rate": 1.1271195458467555e-07,
+      "loss": 0.8505,
+      "step": 9906
+    },
+    {
+      "epoch": 1.9071636547392736,
+      "grad_norm": 3.240259241230784,
+      "learning_rate": 1.122456312662068e-07,
+      "loss": 0.8247,
+      "step": 9907
+    },
+    {
+      "epoch": 1.9073561614168493,
+      "grad_norm": 3.2605844039607934,
+      "learning_rate": 1.1178026916396356e-07,
+      "loss": 0.8693,
+      "step": 9908
+    },
+    {
+      "epoch": 1.9075486680944245,
+      "grad_norm": 3.0947694256588076,
+      "learning_rate": 1.1131586832318631e-07,
+      "loss": 0.7981,
+      "step": 9909
+    },
+    {
+      "epoch": 1.907741174772,
+      "grad_norm": 3.3331988639281,
+      "learning_rate": 1.1085242878901892e-07,
+      "loss": 0.8602,
+      "step": 9910
+    },
+    {
+      "epoch": 1.9079336814495753,
+      "grad_norm": 3.2750565109747836,
+      "learning_rate": 1.1038995060651536e-07,
+      "loss": 0.832,
+      "step": 9911
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "grad_norm": 3.2519543026683,
+      "learning_rate": 1.0992843382063301e-07,
+      "loss": 0.8251,
+      "step": 9912
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.9418,
+      "step": 9912,
+      "vm_loss": 0.2026
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.711,
+      "step": 9912,
+      "vm_loss": 0.1504
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.6705,
+      "step": 9912,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.4324,
+      "step": 9912,
+      "vm_loss": 0.1742
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.6006,
+      "step": 9912,
+      "vm_loss": 0.2166
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.5787,
+      "step": 9912,
+      "vm_loss": 0.1953
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.6499,
+      "step": 9912,
+      "vm_loss": 0.1298
+    },
+    {
+      "epoch": 1.9081261881271505,
+      "lm_loss": 0.3661,
+      "step": 9912,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.9083186948047262,
+      "grad_norm": 3.3275522055311857,
+      "learning_rate": 1.0946787847623819e-07,
+      "loss": 0.8957,
+      "step": 9913
+    },
+    {
+      "epoch": 1.9085112014823014,
+      "grad_norm": 3.181604524844051,
+      "learning_rate": 1.0900828461810176e-07,
+      "loss": 0.8435,
+      "step": 9914
+    },
+    {
+      "epoch": 1.9087037081598768,
+      "grad_norm": 3.2606772503624626,
+      "learning_rate": 1.0854965229090352e-07,
+      "loss": 0.8154,
+      "step": 9915
+    },
+    {
+      "epoch": 1.9088962148374522,
+      "grad_norm": 3.1688447263236825,
+      "learning_rate": 1.0809198153922673e-07,
+      "loss": 0.8417,
+      "step": 9916
+    },
+    {
+      "epoch": 1.9090887215150274,
+      "grad_norm": 3.1259354260841006,
+      "learning_rate": 1.0763527240756578e-07,
+      "loss": 0.8015,
+      "step": 9917
+    },
+    {
+      "epoch": 1.909281228192603,
+      "grad_norm": 3.262864519739653,
+      "learning_rate": 1.0717952494031514e-07,
+      "loss": 0.825,
+      "step": 9918
+    },
+    {
+      "epoch": 1.9094737348701782,
+      "grad_norm": 3.205057023786529,
+      "learning_rate": 1.067247391817805e-07,
+      "loss": 0.8381,
+      "step": 9919
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "grad_norm": 3.4250298328493214,
+      "learning_rate": 1.0627091517617427e-07,
+      "loss": 0.867,
+      "step": 9920
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.2017,
+      "step": 9920,
+      "vm_loss": 0.1773
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.8163,
+      "step": 9920,
+      "vm_loss": 0.2009
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.4915,
+      "step": 9920,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.4239,
+      "step": 9920,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.7777,
+      "step": 9920,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.4986,
+      "step": 9920,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.4651,
+      "step": 9920,
+      "vm_loss": 0.1607
+    },
+    {
+      "epoch": 1.9096662415477537,
+      "lm_loss": 0.7211,
+      "step": 9920,
+      "vm_loss": 0.1058
+    },
+    {
+      "epoch": 1.909858748225329,
+      "grad_norm": 3.1613739876367526,
+      "learning_rate": 1.0581805296761227e-07,
+      "loss": 0.8356,
+      "step": 9921
+    },
+    {
+      "epoch": 1.9100512549029043,
+      "grad_norm": 3.3992340696033256,
+      "learning_rate": 1.0536615260011928e-07,
+      "loss": 0.8613,
+      "step": 9922
+    },
+    {
+      "epoch": 1.91024376158048,
+      "grad_norm": 3.444260482071373,
+      "learning_rate": 1.0491521411762572e-07,
+      "loss": 0.8776,
+      "step": 9923
+    },
+    {
+      "epoch": 1.9104362682580551,
+      "grad_norm": 3.336788292060354,
+      "learning_rate": 1.0446523756396764e-07,
+      "loss": 0.8368,
+      "step": 9924
+    },
+    {
+      "epoch": 1.9106287749356305,
+      "grad_norm": 3.3366431201524214,
+      "learning_rate": 1.0401622298289005e-07,
+      "loss": 0.869,
+      "step": 9925
+    },
+    {
+      "epoch": 1.910821281613206,
+      "grad_norm": 3.1982285187186585,
+      "learning_rate": 1.0356817041804246e-07,
+      "loss": 0.8512,
+      "step": 9926
+    },
+    {
+      "epoch": 1.9110137882907814,
+      "grad_norm": 3.2563313789034263,
+      "learning_rate": 1.0312107991298004e-07,
+      "loss": 0.8447,
+      "step": 9927
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "grad_norm": 3.1270523977283617,
+      "learning_rate": 1.026749515111669e-07,
+      "loss": 0.7655,
+      "step": 9928
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.4665,
+      "step": 9928,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.6368,
+      "step": 9928,
+      "vm_loss": 0.2316
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.7044,
+      "step": 9928,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.5044,
+      "step": 9928,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.4933,
+      "step": 9928,
+      "vm_loss": 0.1197
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.6454,
+      "step": 9928,
+      "vm_loss": 0.1266
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.896,
+      "step": 9928,
+      "vm_loss": 0.1498
+    },
+    {
+      "epoch": 1.9112062949683568,
+      "lm_loss": 0.7782,
+      "step": 9928,
+      "vm_loss": 0.1269
+    },
+    {
+      "epoch": 1.911398801645932,
+      "grad_norm": 3.202314516521334,
+      "learning_rate": 1.0222978525597282e-07,
+      "loss": 0.7775,
+      "step": 9929
+    },
+    {
+      "epoch": 1.9115913083235074,
+      "grad_norm": 3.2311114879379663,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.8112,
+      "step": 9930
+    },
+    {
+      "epoch": 1.9117838150010829,
+      "grad_norm": 3.2221828973454083,
+      "learning_rate": 1.0134233935844895e-07,
+      "loss": 0.8835,
+      "step": 9931
+    },
+    {
+      "epoch": 1.9119763216786583,
+      "grad_norm": 3.1304561268014393,
+      "learning_rate": 1.0090005980239236e-07,
+      "loss": 0.7846,
+      "step": 9932
+    },
+    {
+      "epoch": 1.9121688283562337,
+      "grad_norm": 3.155946356661212,
+      "learning_rate": 1.0045874256549348e-07,
+      "loss": 0.8246,
+      "step": 9933
+    },
+    {
+      "epoch": 1.912361335033809,
+      "grad_norm": 3.379190458325995,
+      "learning_rate": 1.0001838769065908e-07,
+      "loss": 0.8288,
+      "step": 9934
+    },
+    {
+      "epoch": 1.9125538417113843,
+      "grad_norm": 3.184016048369238,
+      "learning_rate": 9.957899522069491e-08,
+      "loss": 0.8187,
+      "step": 9935
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "grad_norm": 3.3063694431187196,
+      "learning_rate": 9.914056519831572e-08,
+      "loss": 0.8487,
+      "step": 9936
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.433,
+      "step": 9936,
+      "vm_loss": 0.1627
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.3984,
+      "step": 9936,
+      "vm_loss": 0.0922
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.9213,
+      "step": 9936,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.5418,
+      "step": 9936,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.8912,
+      "step": 9936,
+      "vm_loss": 0.1655
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.4956,
+      "step": 9936,
+      "vm_loss": 0.1394
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.6429,
+      "step": 9936,
+      "vm_loss": 0.1558
+    },
+    {
+      "epoch": 1.9127463483889597,
+      "lm_loss": 0.7916,
+      "step": 9936,
+      "vm_loss": 0.1593
+    },
+    {
+      "epoch": 1.9129388550665352,
+      "grad_norm": 3.038788192887153,
+      "learning_rate": 9.870309766614184e-08,
+      "loss": 0.7942,
+      "step": 9937
+    },
+    {
+      "epoch": 1.9131313617441106,
+      "grad_norm": 3.3855092457664733,
+      "learning_rate": 9.826659266670258e-08,
+      "loss": 0.8926,
+      "step": 9938
+    },
+    {
+      "epoch": 1.9133238684216858,
+      "grad_norm": 3.2337467640974724,
+      "learning_rate": 9.783105024242956e-08,
+      "loss": 0.81,
+      "step": 9939
+    },
+    {
+      "epoch": 1.9135163750992612,
+      "grad_norm": 3.3771172845783615,
+      "learning_rate": 9.739647043566447e-08,
+      "loss": 0.8637,
+      "step": 9940
+    },
+    {
+      "epoch": 1.9137088817768366,
+      "grad_norm": 3.3089180985756435,
+      "learning_rate": 9.696285328865352e-08,
+      "loss": 0.8609,
+      "step": 9941
+    },
+    {
+      "epoch": 1.913901388454412,
+      "grad_norm": 3.160928207591143,
+      "learning_rate": 9.653019884355075e-08,
+      "loss": 0.838,
+      "step": 9942
+    },
+    {
+      "epoch": 1.9140938951319875,
+      "grad_norm": 2.986331310740156,
+      "learning_rate": 9.609850714241475e-08,
+      "loss": 0.7285,
+      "step": 9943
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "grad_norm": 3.2505720078210443,
+      "learning_rate": 9.566777822721196e-08,
+      "loss": 0.8761,
+      "step": 9944
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 1.1197,
+      "step": 9944,
+      "vm_loss": 0.1493
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.4906,
+      "step": 9944,
+      "vm_loss": 0.1665
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.9033,
+      "step": 9944,
+      "vm_loss": 0.1883
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.5176,
+      "step": 9944,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.8495,
+      "step": 9944,
+      "vm_loss": 0.1564
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.8973,
+      "step": 9944,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.7788,
+      "step": 9944,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 1.9142864018095627,
+      "lm_loss": 0.6433,
+      "step": 9944,
+      "vm_loss": 0.214
+    },
+    {
+      "epoch": 1.9144789084871383,
+      "grad_norm": 3.233812671799377,
+      "learning_rate": 9.523801213981443e-08,
+      "loss": 0.8472,
+      "step": 9945
+    },
+    {
+      "epoch": 1.9146714151647135,
+      "grad_norm": 3.4125357172174526,
+      "learning_rate": 9.480920892200096e-08,
+      "loss": 0.893,
+      "step": 9946
+    },
+    {
+      "epoch": 1.914863921842289,
+      "grad_norm": 3.187284373320313,
+      "learning_rate": 9.4381368615456e-08,
+      "loss": 0.7988,
+      "step": 9947
+    },
+    {
+      "epoch": 1.9150564285198644,
+      "grad_norm": 3.19593381936061,
+      "learning_rate": 9.395449126177291e-08,
+      "loss": 0.8139,
+      "step": 9948
+    },
+    {
+      "epoch": 1.9152489351974396,
+      "grad_norm": 3.1173367344759795,
+      "learning_rate": 9.352857690244854e-08,
+      "loss": 0.8378,
+      "step": 9949
+    },
+    {
+      "epoch": 1.9154414418750152,
+      "grad_norm": 3.237628201471339,
+      "learning_rate": 9.310362557888642e-08,
+      "loss": 0.8261,
+      "step": 9950
+    },
+    {
+      "epoch": 1.9156339485525904,
+      "grad_norm": 3.383036985375902,
+      "learning_rate": 9.267963733239904e-08,
+      "loss": 0.8709,
+      "step": 9951
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "grad_norm": 3.260786093268801,
+      "learning_rate": 9.225661220420234e-08,
+      "loss": 0.8217,
+      "step": 9952
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.9081,
+      "step": 9952,
+      "vm_loss": 0.1153
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.6096,
+      "step": 9952,
+      "vm_loss": 0.2274
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.4406,
+      "step": 9952,
+      "vm_loss": 0.1576
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.4195,
+      "step": 9952,
+      "vm_loss": 0.1688
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.5355,
+      "step": 9952,
+      "vm_loss": 0.1398
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.6897,
+      "step": 9952,
+      "vm_loss": 0.1724
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 0.5694,
+      "step": 9952,
+      "vm_loss": 0.1288
+    },
+    {
+      "epoch": 1.9158264552301658,
+      "lm_loss": 1.0391,
+      "step": 9952,
+      "vm_loss": 0.1496
+    },
+    {
+      "epoch": 1.9160189619077412,
+      "grad_norm": 3.3606677888024263,
+      "learning_rate": 9.18345502354201e-08,
+      "loss": 0.866,
+      "step": 9953
+    },
+    {
+      "epoch": 1.9162114685853164,
+      "grad_norm": 3.129080059580664,
+      "learning_rate": 9.141345146708169e-08,
+      "loss": 0.8008,
+      "step": 9954
+    },
+    {
+      "epoch": 1.916403975262892,
+      "grad_norm": 3.206112737044632,
+      "learning_rate": 9.099331594012329e-08,
+      "loss": 0.8535,
+      "step": 9955
+    },
+    {
+      "epoch": 1.9165964819404673,
+      "grad_norm": 3.2603245484025107,
+      "learning_rate": 9.057414369538887e-08,
+      "loss": 0.8078,
+      "step": 9956
+    },
+    {
+      "epoch": 1.9167889886180427,
+      "grad_norm": 3.123323087210107,
+      "learning_rate": 9.015593477362472e-08,
+      "loss": 0.836,
+      "step": 9957
+    },
+    {
+      "epoch": 1.9169814952956181,
+      "grad_norm": 3.271809729413425,
+      "learning_rate": 8.973868921548834e-08,
+      "loss": 0.8384,
+      "step": 9958
+    },
+    {
+      "epoch": 1.9171740019731933,
+      "grad_norm": 3.2284625961389484,
+      "learning_rate": 8.932240706154171e-08,
+      "loss": 0.8067,
+      "step": 9959
+    },
+    {
+      "epoch": 1.917366508650769,
+      "grad_norm": 3.1097142499006507,
+      "learning_rate": 8.890708835225026e-08,
+      "loss": 0.7776,
+      "step": 9960
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.8466,
+      "step": 9960,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.672,
+      "step": 9960,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.9304,
+      "step": 9960,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.424,
+      "step": 9960,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.8495,
+      "step": 9960,
+      "vm_loss": 0.2558
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.6521,
+      "step": 9960,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.5842,
+      "step": 9960,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 1.917366508650769,
+      "lm_loss": 0.6816,
+      "step": 9960,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 1.9175590153283442,
+      "grad_norm": 3.2158704157391607,
+      "learning_rate": 8.849273312799056e-08,
+      "loss": 0.809,
+      "step": 9961
+    },
+    {
+      "epoch": 1.9177515220059196,
+      "grad_norm": 3.268655244261384,
+      "learning_rate": 8.807934142904262e-08,
+      "loss": 0.8324,
+      "step": 9962
+    },
+    {
+      "epoch": 1.917944028683495,
+      "grad_norm": 3.3504453515127945,
+      "learning_rate": 8.766691329559207e-08,
+      "loss": 0.866,
+      "step": 9963
+    },
+    {
+      "epoch": 1.9181365353610702,
+      "grad_norm": 3.208161092329222,
+      "learning_rate": 8.72554487677335e-08,
+      "loss": 0.7797,
+      "step": 9964
+    },
+    {
+      "epoch": 1.9183290420386458,
+      "grad_norm": 3.2790459550726747,
+      "learning_rate": 8.684494788546716e-08,
+      "loss": 0.8068,
+      "step": 9965
+    },
+    {
+      "epoch": 1.918521548716221,
+      "grad_norm": 3.351630112811597,
+      "learning_rate": 8.643541068869776e-08,
+      "loss": 0.8796,
+      "step": 9966
+    },
+    {
+      "epoch": 1.9187140553937965,
+      "grad_norm": 3.3083529848965245,
+      "learning_rate": 8.602683721723793e-08,
+      "loss": 0.8853,
+      "step": 9967
+    },
+    {
+      "epoch": 1.918906562071372,
+      "grad_norm": 3.0830475507712447,
+      "learning_rate": 8.561922751080697e-08,
+      "loss": 0.7518,
+      "step": 9968
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.406,
+      "step": 9968,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.8093,
+      "step": 9968,
+      "vm_loss": 0.142
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.9129,
+      "step": 9968,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.4346,
+      "step": 9968,
+      "vm_loss": 0.2875
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 1.1024,
+      "step": 9968,
+      "vm_loss": 0.2363
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.85,
+      "step": 9968,
+      "vm_loss": 0.2356
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.797,
+      "step": 9968,
+      "vm_loss": 0.1436
+    },
+    {
+      "epoch": 1.918906562071372,
+      "lm_loss": 0.5154,
+      "step": 9968,
+      "vm_loss": 0.1332
+    },
+    {
+      "epoch": 1.919099068748947,
+      "grad_norm": 3.2380418088297462,
+      "learning_rate": 8.521258160902767e-08,
+      "loss": 0.8781,
+      "step": 9969
+    },
+    {
+      "epoch": 1.9192915754265227,
+      "grad_norm": 3.070287601728227,
+      "learning_rate": 8.480689955143395e-08,
+      "loss": 0.8114,
+      "step": 9970
+    },
+    {
+      "epoch": 1.919484082104098,
+      "grad_norm": 3.30890792941629,
+      "learning_rate": 8.440218137746092e-08,
+      "loss": 0.8633,
+      "step": 9971
+    },
+    {
+      "epoch": 1.9196765887816734,
+      "grad_norm": 3.404212275537186,
+      "learning_rate": 8.399842712645378e-08,
+      "loss": 0.8702,
+      "step": 9972
+    },
+    {
+      "epoch": 1.9198690954592488,
+      "grad_norm": 3.304417692886842,
+      "learning_rate": 8.359563683766225e-08,
+      "loss": 0.8275,
+      "step": 9973
+    },
+    {
+      "epoch": 1.920061602136824,
+      "grad_norm": 3.104477219970434,
+      "learning_rate": 8.319381055024278e-08,
+      "loss": 0.8005,
+      "step": 9974
+    },
+    {
+      "epoch": 1.9202541088143996,
+      "grad_norm": 3.205647667063476,
+      "learning_rate": 8.279294830325746e-08,
+      "loss": 0.8473,
+      "step": 9975
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "grad_norm": 3.393739336683005,
+      "learning_rate": 8.239305013567734e-08,
+      "loss": 0.8685,
+      "step": 9976
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.8843,
+      "step": 9976,
+      "vm_loss": 0.1441
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.9568,
+      "step": 9976,
+      "vm_loss": 0.1766
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.4966,
+      "step": 9976,
+      "vm_loss": 0.1885
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.5389,
+      "step": 9976,
+      "vm_loss": 0.1293
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.625,
+      "step": 9976,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.6157,
+      "step": 9976,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.5504,
+      "step": 9976,
+      "vm_loss": 0.1613
+    },
+    {
+      "epoch": 1.9204466154919748,
+      "lm_loss": 0.5517,
+      "step": 9976,
+      "vm_loss": 0.1232
+    },
+    {
+      "epoch": 1.9206391221695502,
+      "grad_norm": 3.3281365535731346,
+      "learning_rate": 8.199411608637465e-08,
+      "loss": 0.8461,
+      "step": 9977
+    },
+    {
+      "epoch": 1.9208316288471257,
+      "grad_norm": 3.116934828383036,
+      "learning_rate": 8.159614619413169e-08,
+      "loss": 0.7783,
+      "step": 9978
+    },
+    {
+      "epoch": 1.9210241355247009,
+      "grad_norm": 3.2335089372004338,
+      "learning_rate": 8.119914049763755e-08,
+      "loss": 0.8323,
+      "step": 9979
+    },
+    {
+      "epoch": 1.9212166422022765,
+      "grad_norm": 3.4439366277016625,
+      "learning_rate": 8.080309903548577e-08,
+      "loss": 0.8562,
+      "step": 9980
+    },
+    {
+      "epoch": 1.9214091488798517,
+      "grad_norm": 3.280723884331258,
+      "learning_rate": 8.040802184617446e-08,
+      "loss": 0.815,
+      "step": 9981
+    },
+    {
+      "epoch": 1.9216016555574271,
+      "grad_norm": 3.2593402630784105,
+      "learning_rate": 8.0013908968114e-08,
+      "loss": 0.7906,
+      "step": 9982
+    },
+    {
+      "epoch": 1.9217941622350025,
+      "grad_norm": 3.2339832880266983,
+      "learning_rate": 7.962076043961486e-08,
+      "loss": 0.79,
+      "step": 9983
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "grad_norm": 3.103014281252802,
+      "learning_rate": 7.922857629889646e-08,
+      "loss": 0.774,
+      "step": 9984
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.5574,
+      "step": 9984,
+      "vm_loss": 0.1413
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.5323,
+      "step": 9984,
+      "vm_loss": 0.1358
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.6366,
+      "step": 9984,
+      "vm_loss": 0.1443
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.793,
+      "step": 9984,
+      "vm_loss": 0.1176
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.6337,
+      "step": 9984,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.433,
+      "step": 9984,
+      "vm_loss": 0.158
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.4286,
+      "step": 9984,
+      "vm_loss": 0.1335
+    },
+    {
+      "epoch": 1.9219866689125777,
+      "lm_loss": 0.4035,
+      "step": 9984,
+      "vm_loss": 0.186
+    },
+    {
+      "epoch": 1.9221791755901534,
+      "grad_norm": 3.2761356682684735,
+      "learning_rate": 7.883735658408386e-08,
+      "loss": 0.8048,
+      "step": 9985
+    },
+    {
+      "epoch": 1.9223716822677286,
+      "grad_norm": 3.227963267761403,
+      "learning_rate": 7.844710133320887e-08,
+      "loss": 0.8402,
+      "step": 9986
+    },
+    {
+      "epoch": 1.922564188945304,
+      "grad_norm": 3.2698343087583144,
+      "learning_rate": 7.805781058421003e-08,
+      "loss": 0.8189,
+      "step": 9987
+    },
+    {
+      "epoch": 1.9227566956228794,
+      "grad_norm": 3.1720528913860617,
+      "learning_rate": 7.766948437493039e-08,
+      "loss": 0.8152,
+      "step": 9988
+    },
+    {
+      "epoch": 1.9229492023004549,
+      "grad_norm": 3.317727886467578,
+      "learning_rate": 7.728212274312086e-08,
+      "loss": 0.8533,
+      "step": 9989
+    },
+    {
+      "epoch": 1.9231417089780303,
+      "grad_norm": 3.2837640255969647,
+      "learning_rate": 7.68957257264369e-08,
+      "loss": 0.8369,
+      "step": 9990
+    },
+    {
+      "epoch": 1.9233342156556055,
+      "grad_norm": 3.3213020320540214,
+      "learning_rate": 7.651029336244175e-08,
+      "loss": 0.8453,
+      "step": 9991
+    },
+    {
+      "epoch": 1.923526722333181,
+      "grad_norm": 3.279682932215763,
+      "learning_rate": 7.61258256886055e-08,
+      "loss": 0.8067,
+      "step": 9992
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.7507,
+      "step": 9992,
+      "vm_loss": 0.1708
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.5879,
+      "step": 9992,
+      "vm_loss": 0.1984
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.8408,
+      "step": 9992,
+      "vm_loss": 0.2065
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.9713,
+      "step": 9992,
+      "vm_loss": 0.1191
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 1.0564,
+      "step": 9992,
+      "vm_loss": 0.1442
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.848,
+      "step": 9992,
+      "vm_loss": 0.1468
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.3305,
+      "step": 9992,
+      "vm_loss": 0.1844
+    },
+    {
+      "epoch": 1.923526722333181,
+      "lm_loss": 0.4904,
+      "step": 9992,
+      "vm_loss": 0.1791
+    },
+    {
+      "epoch": 1.9237192290107563,
+      "grad_norm": 3.204432979621257,
+      "learning_rate": 7.574232274230264e-08,
+      "loss": 0.8339,
+      "step": 9993
+    },
+    {
+      "epoch": 1.9239117356883317,
+      "grad_norm": 3.1319916054962462,
+      "learning_rate": 7.535978456081339e-08,
+      "loss": 0.7901,
+      "step": 9994
+    },
+    {
+      "epoch": 1.9241042423659072,
+      "grad_norm": 3.1394839770627154,
+      "learning_rate": 7.497821118132687e-08,
+      "loss": 0.8263,
+      "step": 9995
+    },
+    {
+      "epoch": 1.9242967490434824,
+      "grad_norm": 2.9846174050187675,
+      "learning_rate": 7.459760264093563e-08,
+      "loss": 0.7968,
+      "step": 9996
+    },
+    {
+      "epoch": 1.9244892557210578,
+      "grad_norm": 3.2778192705327136,
+      "learning_rate": 7.42179589766412e-08,
+      "loss": 0.8049,
+      "step": 9997
+    },
+    {
+      "epoch": 1.9246817623986332,
+      "grad_norm": 3.1502301575505856,
+      "learning_rate": 7.383928022534847e-08,
+      "loss": 0.8312,
+      "step": 9998
+    },
+    {
+      "epoch": 1.9248742690762086,
+      "grad_norm": 3.248549460842929,
+      "learning_rate": 7.346156642387137e-08,
+      "loss": 0.8528,
+      "step": 9999
+    },
+    {
+      "epoch": 1.925066775753784,
+      "grad_norm": 3.1471421618350512,
+      "learning_rate": 7.308481760892605e-08,
+      "loss": 0.7854,
+      "step": 10000
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.6902,
+      "step": 10000,
+      "vm_loss": 0.1752
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.5175,
+      "step": 10000,
+      "vm_loss": 0.1297
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.2516,
+      "step": 10000,
+      "vm_loss": 0.1899
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.5713,
+      "step": 10000,
+      "vm_loss": 0.2327
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.6888,
+      "step": 10000,
+      "vm_loss": 0.1154
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.702,
+      "step": 10000,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.9082,
+      "step": 10000,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 1.925066775753784,
+      "lm_loss": 0.557,
+      "step": 10000,
+      "vm_loss": 0.2113
+    },
+    {
+      "epoch": 1.9252592824313592,
+      "grad_norm": 3.3684366618393975,
+      "learning_rate": 7.270903381713989e-08,
+      "loss": 0.8738,
+      "step": 10001
+    },
+    {
+      "epoch": 1.9254517891089347,
+      "grad_norm": 3.1193790826800014,
+      "learning_rate": 7.233421508504257e-08,
+      "loss": 0.7972,
+      "step": 10002
+    },
+    {
+      "epoch": 1.92564429578651,
+      "grad_norm": 3.3031903074242224,
+      "learning_rate": 7.196036144907159e-08,
+      "loss": 0.8707,
+      "step": 10003
+    },
+    {
+      "epoch": 1.9258368024640855,
+      "grad_norm": 3.3577966390052927,
+      "learning_rate": 7.1587472945569e-08,
+      "loss": 0.8781,
+      "step": 10004
+    },
+    {
+      "epoch": 1.926029309141661,
+      "grad_norm": 3.2136214711631816,
+      "learning_rate": 7.121554961078692e-08,
+      "loss": 0.8207,
+      "step": 10005
+    },
+    {
+      "epoch": 1.9262218158192361,
+      "grad_norm": 3.298741835417461,
+      "learning_rate": 7.084459148087864e-08,
+      "loss": 0.8375,
+      "step": 10006
+    },
+    {
+      "epoch": 1.9264143224968118,
+      "grad_norm": 3.179228770853609,
+      "learning_rate": 7.047459859190752e-08,
+      "loss": 0.8447,
+      "step": 10007
+    },
+    {
+      "epoch": 1.926606829174387,
+      "grad_norm": 3.1913765571956896,
+      "learning_rate": 7.01055709798415e-08,
+      "loss": 0.8008,
+      "step": 10008
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 1.1802,
+      "step": 10008,
+      "vm_loss": 0.1411
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.6101,
+      "step": 10008,
+      "vm_loss": 0.2149
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.7753,
+      "step": 10008,
+      "vm_loss": 0.1785
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.6306,
+      "step": 10008,
+      "vm_loss": 0.1789
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.3485,
+      "step": 10008,
+      "vm_loss": 0.1429
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.8707,
+      "step": 10008,
+      "vm_loss": 0.2313
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 0.4871,
+      "step": 10008,
+      "vm_loss": 0.1631
+    },
+    {
+      "epoch": 1.926606829174387,
+      "lm_loss": 1.1411,
+      "step": 10008,
+      "vm_loss": 0.1381
+    },
+    {
+      "epoch": 1.9267993358519624,
+      "grad_norm": 3.313535315065674,
+      "learning_rate": 6.973750868055406e-08,
+      "loss": 0.8562,
+      "step": 10009
+    },
+    {
+      "epoch": 1.9269918425295378,
+      "grad_norm": 3.3256902367865253,
+      "learning_rate": 6.937041172982662e-08,
+      "loss": 0.8611,
+      "step": 10010
+    },
+    {
+      "epoch": 1.927184349207113,
+      "grad_norm": 3.3460536682412987,
+      "learning_rate": 6.900428016334393e-08,
+      "loss": 0.8395,
+      "step": 10011
+    },
+    {
+      "epoch": 1.9273768558846887,
+      "grad_norm": 3.3287282106338436,
+      "learning_rate": 6.863911401669976e-08,
+      "loss": 0.8505,
+      "step": 10012
+    },
+    {
+      "epoch": 1.9275693625622639,
+      "grad_norm": 3.048516301479563,
+      "learning_rate": 6.827491332539349e-08,
+      "loss": 0.8076,
+      "step": 10013
+    },
+    {
+      "epoch": 1.9277618692398393,
+      "grad_norm": 3.20842115740196,
+      "learning_rate": 6.791167812483013e-08,
+      "loss": 0.8419,
+      "step": 10014
+    },
+    {
+      "epoch": 1.9279543759174147,
+      "grad_norm": 3.3724980902572987,
+      "learning_rate": 6.75494084503181e-08,
+      "loss": 0.8225,
+      "step": 10015
+    },
+    {
+      "epoch": 1.92814688259499,
+      "grad_norm": 3.381450113165441,
+      "learning_rate": 6.718810433707924e-08,
+      "loss": 0.8765,
+      "step": 10016
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.6617,
+      "step": 10016,
+      "vm_loss": 0.1409
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.3258,
+      "step": 10016,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.6057,
+      "step": 10016,
+      "vm_loss": 0.1154
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 1.0099,
+      "step": 10016,
+      "vm_loss": 0.1065
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.4767,
+      "step": 10016,
+      "vm_loss": 0.1367
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.5711,
+      "step": 10016,
+      "vm_loss": 0.1781
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.5324,
+      "step": 10016,
+      "vm_loss": 0.2103
+    },
+    {
+      "epoch": 1.92814688259499,
+      "lm_loss": 0.9001,
+      "step": 10016,
+      "vm_loss": 0.2371
+    },
+    {
+      "epoch": 1.9283393892725655,
+      "grad_norm": 3.153939724618915,
+      "learning_rate": 6.68277658202321e-08,
+      "loss": 0.8269,
+      "step": 10017
+    },
+    {
+      "epoch": 1.9285318959501407,
+      "grad_norm": 3.215528042978117,
+      "learning_rate": 6.646839293481088e-08,
+      "loss": 0.818,
+      "step": 10018
+    },
+    {
+      "epoch": 1.9287244026277162,
+      "grad_norm": 3.206212798912243,
+      "learning_rate": 6.610998571574656e-08,
+      "loss": 0.822,
+      "step": 10019
+    },
+    {
+      "epoch": 1.9289169093052916,
+      "grad_norm": 3.2554223949099996,
+      "learning_rate": 6.575254419788457e-08,
+      "loss": 0.8388,
+      "step": 10020
+    },
+    {
+      "epoch": 1.9291094159828668,
+      "grad_norm": 3.339778406575526,
+      "learning_rate": 6.539606841597046e-08,
+      "loss": 0.8634,
+      "step": 10021
+    },
+    {
+      "epoch": 1.9293019226604424,
+      "grad_norm": 3.329420873526788,
+      "learning_rate": 6.504055840465983e-08,
+      "loss": 0.8361,
+      "step": 10022
+    },
+    {
+      "epoch": 1.9294944293380176,
+      "grad_norm": 3.4356237032858625,
+      "learning_rate": 6.468601419851173e-08,
+      "loss": 0.8679,
+      "step": 10023
+    },
+    {
+      "epoch": 1.929686936015593,
+      "grad_norm": 3.2648005047078983,
+      "learning_rate": 6.433243583199412e-08,
+      "loss": 0.852,
+      "step": 10024
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.4727,
+      "step": 10024,
+      "vm_loss": 0.1541
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.7455,
+      "step": 10024,
+      "vm_loss": 0.1812
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.578,
+      "step": 10024,
+      "vm_loss": 0.1594
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.7262,
+      "step": 10024,
+      "vm_loss": 0.1302
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.714,
+      "step": 10024,
+      "vm_loss": 0.1318
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.4874,
+      "step": 10024,
+      "vm_loss": 0.2125
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 0.4989,
+      "step": 10024,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.929686936015593,
+      "lm_loss": 1.4171,
+      "step": 10024,
+      "vm_loss": 0.1661
+    },
+    {
+      "epoch": 1.9298794426931685,
+      "grad_norm": 3.426280175069049,
+      "learning_rate": 6.397982333947617e-08,
+      "loss": 0.8751,
+      "step": 10025
+    },
+    {
+      "epoch": 1.9300719493707437,
+      "grad_norm": 3.385736953398937,
+      "learning_rate": 6.362817675523936e-08,
+      "loss": 0.8807,
+      "step": 10026
+    },
+    {
+      "epoch": 1.9302644560483193,
+      "grad_norm": 3.178883518479652,
+      "learning_rate": 6.327749611346745e-08,
+      "loss": 0.8011,
+      "step": 10027
+    },
+    {
+      "epoch": 1.9304569627258945,
+      "grad_norm": 3.283551467295873,
+      "learning_rate": 6.292778144824984e-08,
+      "loss": 0.8467,
+      "step": 10028
+    },
+    {
+      "epoch": 1.93064946940347,
+      "grad_norm": 3.3472098100925995,
+      "learning_rate": 6.257903279358491e-08,
+      "loss": 0.8881,
+      "step": 10029
+    },
+    {
+      "epoch": 1.9308419760810454,
+      "grad_norm": 3.407160695988864,
+      "learning_rate": 6.223125018337328e-08,
+      "loss": 0.8586,
+      "step": 10030
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 3.301421813918248,
+      "learning_rate": 6.188443365142682e-08,
+      "loss": 0.8625,
+      "step": 10031
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "grad_norm": 3.0635081099564396,
+      "learning_rate": 6.153858323145856e-08,
+      "loss": 0.8045,
+      "step": 10032
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.8019,
+      "step": 10032,
+      "vm_loss": 0.1679
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.5311,
+      "step": 10032,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.5768,
+      "step": 10032,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.4148,
+      "step": 10032,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.6214,
+      "step": 10032,
+      "vm_loss": 0.1701
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.6961,
+      "step": 10032,
+      "vm_loss": 0.1757
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 1.0139,
+      "step": 10032,
+      "vm_loss": 0.1365
+    },
+    {
+      "epoch": 1.9312269894361962,
+      "lm_loss": 0.7958,
+      "step": 10032,
+      "vm_loss": 0.2085
+    },
+    {
+      "epoch": 1.9314194961137714,
+      "grad_norm": 3.2680187578225643,
+      "learning_rate": 6.119369895709048e-08,
+      "loss": 0.8561,
+      "step": 10033
+    },
+    {
+      "epoch": 1.9316120027913468,
+      "grad_norm": 3.242346144980176,
+      "learning_rate": 6.0849780861848e-08,
+      "loss": 0.8132,
+      "step": 10034
+    },
+    {
+      "epoch": 1.9318045094689222,
+      "grad_norm": 3.352650529932269,
+      "learning_rate": 6.05068289791677e-08,
+      "loss": 0.8606,
+      "step": 10035
+    },
+    {
+      "epoch": 1.9319970161464974,
+      "grad_norm": 3.140601798771611,
+      "learning_rate": 6.016484334238515e-08,
+      "loss": 0.8158,
+      "step": 10036
+    },
+    {
+      "epoch": 1.932189522824073,
+      "grad_norm": 3.251800139852715,
+      "learning_rate": 5.982382398474817e-08,
+      "loss": 0.8417,
+      "step": 10037
+    },
+    {
+      "epoch": 1.9323820295016483,
+      "grad_norm": 3.1447314784625866,
+      "learning_rate": 5.948377093940694e-08,
+      "loss": 0.7978,
+      "step": 10038
+    },
+    {
+      "epoch": 1.9325745361792237,
+      "grad_norm": 3.2401004021930686,
+      "learning_rate": 5.914468423942055e-08,
+      "loss": 0.821,
+      "step": 10039
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "grad_norm": 3.4522349113867152,
+      "learning_rate": 5.880656391775041e-08,
+      "loss": 0.9078,
+      "step": 10040
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.9056,
+      "step": 10040,
+      "vm_loss": 0.1649
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.4546,
+      "step": 10040,
+      "vm_loss": 0.2079
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 1.0492,
+      "step": 10040,
+      "vm_loss": 0.1102
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.8601,
+      "step": 10040,
+      "vm_loss": 0.1456
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.8724,
+      "step": 10040,
+      "vm_loss": 0.1795
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.7304,
+      "step": 10040,
+      "vm_loss": 0.1886
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.591,
+      "step": 10040,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.9327670428567991,
+      "lm_loss": 0.3179,
+      "step": 10040,
+      "vm_loss": 0.1658
+    },
+    {
+      "epoch": 1.9329595495343743,
+      "grad_norm": 3.212707950012238,
+      "learning_rate": 5.846941000726802e-08,
+      "loss": 0.8406,
+      "step": 10041
+    },
+    {
+      "epoch": 1.93315205621195,
+      "grad_norm": 3.36118334654075,
+      "learning_rate": 5.813322254074827e-08,
+      "loss": 0.8255,
+      "step": 10042
+    },
+    {
+      "epoch": 1.9333445628895252,
+      "grad_norm": 3.260913416622491,
+      "learning_rate": 5.779800155087167e-08,
+      "loss": 0.8723,
+      "step": 10043
+    },
+    {
+      "epoch": 1.9335370695671006,
+      "grad_norm": 3.1890492630043994,
+      "learning_rate": 5.746374707022884e-08,
+      "loss": 0.8362,
+      "step": 10044
+    },
+    {
+      "epoch": 1.933729576244676,
+      "grad_norm": 3.1999969718502035,
+      "learning_rate": 5.713045913131155e-08,
+      "loss": 0.8172,
+      "step": 10045
+    },
+    {
+      "epoch": 1.9339220829222512,
+      "grad_norm": 3.377020803701443,
+      "learning_rate": 5.679813776651943e-08,
+      "loss": 0.8583,
+      "step": 10046
+    },
+    {
+      "epoch": 1.9341145895998269,
+      "grad_norm": 3.1817266840059624,
+      "learning_rate": 5.646678300816e-08,
+      "loss": 0.8111,
+      "step": 10047
+    },
+    {
+      "epoch": 1.934307096277402,
+      "grad_norm": 3.1229156253114807,
+      "learning_rate": 5.6136394888443025e-08,
+      "loss": 0.8141,
+      "step": 10048
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.5763,
+      "step": 10048,
+      "vm_loss": 0.1681
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.6909,
+      "step": 10048,
+      "vm_loss": 0.178
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.5703,
+      "step": 10048,
+      "vm_loss": 0.1797
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.7966,
+      "step": 10048,
+      "vm_loss": 0.1305
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.8155,
+      "step": 10048,
+      "vm_loss": 0.1481
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.5978,
+      "step": 10048,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.51,
+      "step": 10048,
+      "vm_loss": 0.1138
+    },
+    {
+      "epoch": 1.934307096277402,
+      "lm_loss": 0.4611,
+      "step": 10048,
+      "vm_loss": 0.171
+    },
+    {
+      "epoch": 1.9344996029549775,
+      "grad_norm": 3.2150886586817884,
+      "learning_rate": 5.580697343948838e-08,
+      "loss": 0.8283,
+      "step": 10049
+    },
+    {
+      "epoch": 1.934692109632553,
+      "grad_norm": 3.302288258596244,
+      "learning_rate": 5.5478518693319326e-08,
+      "loss": 0.8582,
+      "step": 10050
+    },
+    {
+      "epoch": 1.934884616310128,
+      "grad_norm": 3.0802396007611677,
+      "learning_rate": 5.515103068186478e-08,
+      "loss": 0.8074,
+      "step": 10051
+    },
+    {
+      "epoch": 1.9350771229877037,
+      "grad_norm": 3.3735249507637417,
+      "learning_rate": 5.482450943696371e-08,
+      "loss": 0.8362,
+      "step": 10052
+    },
+    {
+      "epoch": 1.935269629665279,
+      "grad_norm": 3.1308826976405473,
+      "learning_rate": 5.449895499035407e-08,
+      "loss": 0.7968,
+      "step": 10053
+    },
+    {
+      "epoch": 1.9354621363428544,
+      "grad_norm": 3.0514074876814554,
+      "learning_rate": 5.4174367373688305e-08,
+      "loss": 0.7674,
+      "step": 10054
+    },
+    {
+      "epoch": 1.9356546430204298,
+      "grad_norm": 3.069136174037404,
+      "learning_rate": 5.385074661851675e-08,
+      "loss": 0.8276,
+      "step": 10055
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "grad_norm": 3.1611754259310887,
+      "learning_rate": 5.3528092756302e-08,
+      "loss": 0.8251,
+      "step": 10056
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.9981,
+      "step": 10056,
+      "vm_loss": 0.1287
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.4823,
+      "step": 10056,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.8971,
+      "step": 10056,
+      "vm_loss": 0.2144
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.6365,
+      "step": 10056,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.731,
+      "step": 10056,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.7451,
+      "step": 10056,
+      "vm_loss": 0.1431
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.8502,
+      "step": 10056,
+      "vm_loss": 0.1889
+    },
+    {
+      "epoch": 1.9358471496980052,
+      "lm_loss": 0.5152,
+      "step": 10056,
+      "vm_loss": 0.1076
+    },
+    {
+      "epoch": 1.9360396563755806,
+      "grad_norm": 3.155700554301023,
+      "learning_rate": 5.320640581840786e-08,
+      "loss": 0.8083,
+      "step": 10057
+    },
+    {
+      "epoch": 1.9362321630531558,
+      "grad_norm": 3.1966760555231475,
+      "learning_rate": 5.2885685836109316e-08,
+      "loss": 0.8,
+      "step": 10058
+    },
+    {
+      "epoch": 1.9364246697307312,
+      "grad_norm": 3.105026732066565,
+      "learning_rate": 5.2565932840582534e-08,
+      "loss": 0.7445,
+      "step": 10059
+    },
+    {
+      "epoch": 1.9366171764083067,
+      "grad_norm": 3.191592340812913,
+      "learning_rate": 5.2247146862911544e-08,
+      "loss": 0.8286,
+      "step": 10060
+    },
+    {
+      "epoch": 1.936809683085882,
+      "grad_norm": 3.400549073310816,
+      "learning_rate": 5.192932793408601e-08,
+      "loss": 0.8518,
+      "step": 10061
+    },
+    {
+      "epoch": 1.9370021897634575,
+      "grad_norm": 3.220631036128648,
+      "learning_rate": 5.1612476085004526e-08,
+      "loss": 0.8539,
+      "step": 10062
+    },
+    {
+      "epoch": 1.9371946964410327,
+      "grad_norm": 2.953337768751391,
+      "learning_rate": 5.129659134646581e-08,
+      "loss": 0.7467,
+      "step": 10063
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "grad_norm": 3.2722567806989047,
+      "learning_rate": 5.098167374917973e-08,
+      "loss": 0.8442,
+      "step": 10064
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.5045,
+      "step": 10064,
+      "vm_loss": 0.2057
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.446,
+      "step": 10064,
+      "vm_loss": 0.1937
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.8359,
+      "step": 10064,
+      "vm_loss": 0.1623
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.8131,
+      "step": 10064,
+      "vm_loss": 0.164
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.8039,
+      "step": 10064,
+      "vm_loss": 0.1031
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.4256,
+      "step": 10064,
+      "vm_loss": 0.1022
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.4718,
+      "step": 10064,
+      "vm_loss": 0.184
+    },
+    {
+      "epoch": 1.9373872031186081,
+      "lm_loss": 0.6303,
+      "step": 10064,
+      "vm_loss": 0.191
+    },
+    {
+      "epoch": 1.9375797097961835,
+      "grad_norm": 3.199055324807108,
+      "learning_rate": 5.0667723323760684e-08,
+      "loss": 0.8222,
+      "step": 10065
+    },
+    {
+      "epoch": 1.937772216473759,
+      "grad_norm": 3.323644163833223,
+      "learning_rate": 5.03547401007276e-08,
+      "loss": 0.8447,
+      "step": 10066
+    },
+    {
+      "epoch": 1.9379647231513344,
+      "grad_norm": 3.226260452132444,
+      "learning_rate": 5.004272411050726e-08,
+      "loss": 0.8227,
+      "step": 10067
+    },
+    {
+      "epoch": 1.9381572298289096,
+      "grad_norm": 3.1074136891115147,
+      "learning_rate": 4.9731675383432046e-08,
+      "loss": 0.7887,
+      "step": 10068
+    },
+    {
+      "epoch": 1.9383497365064852,
+      "grad_norm": 3.1261300044448066,
+      "learning_rate": 4.942159394973778e-08,
+      "loss": 0.7901,
+      "step": 10069
+    },
+    {
+      "epoch": 1.9385422431840604,
+      "grad_norm": 3.284298668899398,
+      "learning_rate": 4.911247983957035e-08,
+      "loss": 0.8622,
+      "step": 10070
+    },
+    {
+      "epoch": 1.9387347498616359,
+      "grad_norm": 3.317572354106562,
+      "learning_rate": 4.8804333082977937e-08,
+      "loss": 0.8437,
+      "step": 10071
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "grad_norm": 3.366810965033768,
+      "learning_rate": 4.84971537099177e-08,
+      "loss": 0.8647,
+      "step": 10072
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.2114,
+      "step": 10072,
+      "vm_loss": 0.1208
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.5062,
+      "step": 10072,
+      "vm_loss": 0.176
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.7494,
+      "step": 10072,
+      "vm_loss": 0.1437
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 1.0268,
+      "step": 10072,
+      "vm_loss": 0.1417
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.9819,
+      "step": 10072,
+      "vm_loss": 0.1343
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.6555,
+      "step": 10072,
+      "vm_loss": 0.1427
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 0.6,
+      "step": 10072,
+      "vm_loss": 0.1375
+    },
+    {
+      "epoch": 1.9389272565392113,
+      "lm_loss": 1.198,
+      "step": 10072,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.9391197632167865,
+      "grad_norm": 3.2492630354824334,
+      "learning_rate": 4.8190941750251296e-08,
+      "loss": 0.8259,
+      "step": 10073
+    },
+    {
+      "epoch": 1.9393122698943621,
+      "grad_norm": 3.1933105157475543,
+      "learning_rate": 4.788569723374603e-08,
+      "loss": 0.8248,
+      "step": 10074
+    },
+    {
+      "epoch": 1.9395047765719373,
+      "grad_norm": 3.3462045475034587,
+      "learning_rate": 4.7581420190077055e-08,
+      "loss": 0.8678,
+      "step": 10075
+    },
+    {
+      "epoch": 1.9396972832495127,
+      "grad_norm": 3.2147661053026537,
+      "learning_rate": 4.727811064882071e-08,
+      "loss": 0.8037,
+      "step": 10076
+    },
+    {
+      "epoch": 1.9398897899270882,
+      "grad_norm": 3.2360876825065965,
+      "learning_rate": 4.6975768639465626e-08,
+      "loss": 0.8703,
+      "step": 10077
+    },
+    {
+      "epoch": 1.9400822966046634,
+      "grad_norm": 3.170944778030707,
+      "learning_rate": 4.667439419140163e-08,
+      "loss": 0.817,
+      "step": 10078
+    },
+    {
+      "epoch": 1.940274803282239,
+      "grad_norm": 3.2852656658043684,
+      "learning_rate": 4.637398733392751e-08,
+      "loss": 0.8287,
+      "step": 10079
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "grad_norm": 3.3230355976955686,
+      "learning_rate": 4.6074548096244346e-08,
+      "loss": 0.8599,
+      "step": 10080
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.3998,
+      "step": 10080,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.5445,
+      "step": 10080,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.6316,
+      "step": 10080,
+      "vm_loss": 0.1699
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 1.3365,
+      "step": 10080,
+      "vm_loss": 0.2069
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.758,
+      "step": 10080,
+      "vm_loss": 0.2109
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.4581,
+      "step": 10080,
+      "vm_loss": 0.1199
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.685,
+      "step": 10080,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.9404673099598142,
+      "lm_loss": 0.941,
+      "step": 10080,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.9406598166373896,
+      "grad_norm": 3.350455898571836,
+      "learning_rate": 4.5776076507464404e-08,
+      "loss": 0.8641,
+      "step": 10081
+    },
+    {
+      "epoch": 1.940852323314965,
+      "grad_norm": 3.2479816867411246,
+      "learning_rate": 4.5478572596601154e-08,
+      "loss": 0.8272,
+      "step": 10082
+    },
+    {
+      "epoch": 1.9410448299925402,
+      "grad_norm": 3.3150998372361475,
+      "learning_rate": 4.51820363925759e-08,
+      "loss": 0.8577,
+      "step": 10083
+    },
+    {
+      "epoch": 1.9412373366701159,
+      "grad_norm": 3.254625022717895,
+      "learning_rate": 4.4886467924215584e-08,
+      "loss": 0.8337,
+      "step": 10084
+    },
+    {
+      "epoch": 1.941429843347691,
+      "grad_norm": 3.228327176167047,
+      "learning_rate": 4.459186722025499e-08,
+      "loss": 0.8394,
+      "step": 10085
+    },
+    {
+      "epoch": 1.9416223500252665,
+      "grad_norm": 3.3950981044477597,
+      "learning_rate": 4.4298234309331224e-08,
+      "loss": 0.8483,
+      "step": 10086
+    },
+    {
+      "epoch": 1.941814856702842,
+      "grad_norm": 3.3142489393923906,
+      "learning_rate": 4.400556921998922e-08,
+      "loss": 0.8412,
+      "step": 10087
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "grad_norm": 3.2547538438643744,
+      "learning_rate": 4.371387198068067e-08,
+      "loss": 0.828,
+      "step": 10088
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.7327,
+      "step": 10088,
+      "vm_loss": 0.1771
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.8187,
+      "step": 10088,
+      "vm_loss": 0.1217
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.8512,
+      "step": 10088,
+      "vm_loss": 0.1657
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.8304,
+      "step": 10088,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.8012,
+      "step": 10088,
+      "vm_loss": 0.1946
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.679,
+      "step": 10088,
+      "vm_loss": 0.143
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.3918,
+      "step": 10088,
+      "vm_loss": 0.1861
+    },
+    {
+      "epoch": 1.9420073633804171,
+      "lm_loss": 0.6492,
+      "step": 10088,
+      "vm_loss": 0.215
+    },
+    {
+      "epoch": 1.9421998700579928,
+      "grad_norm": 3.271133369606919,
+      "learning_rate": 4.342314261976177e-08,
+      "loss": 0.8464,
+      "step": 10089
+    },
+    {
+      "epoch": 1.942392376735568,
+      "grad_norm": 3.1352367547538695,
+      "learning_rate": 4.313338116549437e-08,
+      "loss": 0.8008,
+      "step": 10090
+    },
+    {
+      "epoch": 1.9425848834131434,
+      "grad_norm": 3.25802352719732,
+      "learning_rate": 4.284458764604926e-08,
+      "loss": 0.8104,
+      "step": 10091
+    },
+    {
+      "epoch": 1.9427773900907188,
+      "grad_norm": 3.2280681394410267,
+      "learning_rate": 4.255676208949844e-08,
+      "loss": 0.831,
+      "step": 10092
+    },
+    {
+      "epoch": 1.942969896768294,
+      "grad_norm": 3.558228797216585,
+      "learning_rate": 4.226990452382396e-08,
+      "loss": 0.8841,
+      "step": 10093
+    },
+    {
+      "epoch": 1.9431624034458697,
+      "grad_norm": 3.3898912323436314,
+      "learning_rate": 4.1984014976911293e-08,
+      "loss": 0.8543,
+      "step": 10094
+    },
+    {
+      "epoch": 1.9433549101234449,
+      "grad_norm": 3.4225633300778338,
+      "learning_rate": 4.1699093476551544e-08,
+      "loss": 0.8823,
+      "step": 10095
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "grad_norm": 3.159583820201095,
+      "learning_rate": 4.141514005044589e-08,
+      "loss": 0.8061,
+      "step": 10096
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.5068,
+      "step": 10096,
+      "vm_loss": 0.198
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.5668,
+      "step": 10096,
+      "vm_loss": 0.1448
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.5469,
+      "step": 10096,
+      "vm_loss": 0.1852
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.392,
+      "step": 10096,
+      "vm_loss": 0.2203
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.9583,
+      "step": 10096,
+      "vm_loss": 0.1054
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.6466,
+      "step": 10096,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.9839,
+      "step": 10096,
+      "vm_loss": 0.0768
+    },
+    {
+      "epoch": 1.9435474168010203,
+      "lm_loss": 0.3741,
+      "step": 10096,
+      "vm_loss": 0.1322
+    },
+    {
+      "epoch": 1.9437399234785957,
+      "grad_norm": 3.507125190948868,
+      "learning_rate": 4.1132154726193364e-08,
+      "loss": 0.8778,
+      "step": 10097
+    },
+    {
+      "epoch": 1.943932430156171,
+      "grad_norm": 3.1920676319393144,
+      "learning_rate": 4.085013753130862e-08,
+      "loss": 0.813,
+      "step": 10098
+    },
+    {
+      "epoch": 1.9441249368337465,
+      "grad_norm": 3.1614246271362685,
+      "learning_rate": 4.0569088493204176e-08,
+      "loss": 0.816,
+      "step": 10099
+    },
+    {
+      "epoch": 1.9443174435113217,
+      "grad_norm": 3.235445689287706,
+      "learning_rate": 4.0289007639203736e-08,
+      "loss": 0.8248,
+      "step": 10100
+    },
+    {
+      "epoch": 1.9445099501888972,
+      "grad_norm": 3.2054400367484037,
+      "learning_rate": 4.000989499653218e-08,
+      "loss": 0.8427,
+      "step": 10101
+    },
+    {
+      "epoch": 1.9447024568664726,
+      "grad_norm": 3.323331805042728,
+      "learning_rate": 3.973175059232559e-08,
+      "loss": 0.8556,
+      "step": 10102
+    },
+    {
+      "epoch": 1.9448949635440478,
+      "grad_norm": 3.2620070208671272,
+      "learning_rate": 3.945457445362122e-08,
+      "loss": 0.8402,
+      "step": 10103
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "grad_norm": 3.2572917031235606,
+      "learning_rate": 3.917836660736529e-08,
+      "loss": 0.8385,
+      "step": 10104
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.6567,
+      "step": 10104,
+      "vm_loss": 0.1573
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.4744,
+      "step": 10104,
+      "vm_loss": 0.175
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.6976,
+      "step": 10104,
+      "vm_loss": 0.1303
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.6968,
+      "step": 10104,
+      "vm_loss": 0.1418
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.6725,
+      "step": 10104,
+      "vm_loss": 0.1873
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.3953,
+      "step": 10104,
+      "vm_loss": 0.1356
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.6528,
+      "step": 10104,
+      "vm_loss": 0.1981
+    },
+    {
+      "epoch": 1.9450874702216234,
+      "lm_loss": 0.8002,
+      "step": 10104,
+      "vm_loss": 0.1677
+    },
+    {
+      "epoch": 1.9452799768991986,
+      "grad_norm": 3.263600797390648,
+      "learning_rate": 3.8903127080407446e-08,
+      "loss": 0.841,
+      "step": 10105
+    },
+    {
+      "epoch": 1.945472483576774,
+      "grad_norm": 3.2072187265620107,
+      "learning_rate": 3.8628855899506266e-08,
+      "loss": 0.8217,
+      "step": 10106
+    },
+    {
+      "epoch": 1.9456649902543495,
+      "grad_norm": 3.224795796927549,
+      "learning_rate": 3.835555309132266e-08,
+      "loss": 0.829,
+      "step": 10107
+    },
+    {
+      "epoch": 1.9458574969319247,
+      "grad_norm": 3.214690653187647,
+      "learning_rate": 3.808321868242648e-08,
+      "loss": 0.8478,
+      "step": 10108
+    },
+    {
+      "epoch": 1.9460500036095003,
+      "grad_norm": 3.0894031904245676,
+      "learning_rate": 3.7811852699291e-08,
+      "loss": 0.7829,
+      "step": 10109
+    },
+    {
+      "epoch": 1.9462425102870755,
+      "grad_norm": 3.276030170545593,
+      "learning_rate": 3.754145516829733e-08,
+      "loss": 0.8384,
+      "step": 10110
+    },
+    {
+      "epoch": 1.946435016964651,
+      "grad_norm": 3.153826866516344,
+      "learning_rate": 3.727202611573111e-08,
+      "loss": 0.8244,
+      "step": 10111
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "grad_norm": 3.069215285630629,
+      "learning_rate": 3.700356556778473e-08,
+      "loss": 0.7866,
+      "step": 10112
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.6124,
+      "step": 10112,
+      "vm_loss": 0.1741
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 1.0928,
+      "step": 10112,
+      "vm_loss": 0.1032
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.7532,
+      "step": 10112,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.4381,
+      "step": 10112,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.328,
+      "step": 10112,
+      "vm_loss": 0.1919
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.509,
+      "step": 10112,
+      "vm_loss": 0.1484
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.9468,
+      "step": 10112,
+      "vm_loss": 0.1732
+    },
+    {
+      "epoch": 1.9466275236422264,
+      "lm_loss": 0.4347,
+      "step": 10112,
+      "vm_loss": 0.1285
+    },
+    {
+      "epoch": 1.9468200303198016,
+      "grad_norm": 3.4620028156184492,
+      "learning_rate": 3.6736073550556195e-08,
+      "loss": 0.9101,
+      "step": 10113
+    },
+    {
+      "epoch": 1.9470125369973772,
+      "grad_norm": 3.3082216244669422,
+      "learning_rate": 3.6469550090048043e-08,
+      "loss": 0.8545,
+      "step": 10114
+    },
+    {
+      "epoch": 1.9472050436749524,
+      "grad_norm": 3.260157406484515,
+      "learning_rate": 3.620399521217066e-08,
+      "loss": 0.8331,
+      "step": 10115
+    },
+    {
+      "epoch": 1.9473975503525278,
+      "grad_norm": 3.389742360017076,
+      "learning_rate": 3.5939408942738955e-08,
+      "loss": 0.8773,
+      "step": 10116
+    },
+    {
+      "epoch": 1.9475900570301032,
+      "grad_norm": 3.202245488634307,
+      "learning_rate": 3.567579130747456e-08,
+      "loss": 0.8243,
+      "step": 10117
+    },
+    {
+      "epoch": 1.9477825637076787,
+      "grad_norm": 3.364834556629764,
+      "learning_rate": 3.541314233200255e-08,
+      "loss": 0.8778,
+      "step": 10118
+    },
+    {
+      "epoch": 1.947975070385254,
+      "grad_norm": 3.143481608542886,
+      "learning_rate": 3.5151462041859154e-08,
+      "loss": 0.8149,
+      "step": 10119
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "grad_norm": 3.233254059799246,
+      "learning_rate": 3.4890750462480695e-08,
+      "loss": 0.8151,
+      "step": 10120
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 1.0884,
+      "step": 10120,
+      "vm_loss": 0.2005
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.6766,
+      "step": 10120,
+      "vm_loss": 0.1226
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.7303,
+      "step": 10120,
+      "vm_loss": 0.1684
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.5767,
+      "step": 10120,
+      "vm_loss": 0.1803
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.5009,
+      "step": 10120,
+      "vm_loss": 0.224
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.4616,
+      "step": 10120,
+      "vm_loss": 0.2193
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 1.0803,
+      "step": 10120,
+      "vm_loss": 0.1534
+    },
+    {
+      "epoch": 1.9481675770628293,
+      "lm_loss": 0.9991,
+      "step": 10120,
+      "vm_loss": 0.1447
+    },
+    {
+      "epoch": 1.9483600837404047,
+      "grad_norm": 3.211727199026326,
+      "learning_rate": 3.463100761921245e-08,
+      "loss": 0.858,
+      "step": 10121
+    },
+    {
+      "epoch": 1.9485525904179801,
+      "grad_norm": 3.1663739986006947,
+      "learning_rate": 3.4372233537304236e-08,
+      "loss": 0.8063,
+      "step": 10122
+    },
+    {
+      "epoch": 1.9487450970955555,
+      "grad_norm": 3.1686360520601653,
+      "learning_rate": 3.411442824191369e-08,
+      "loss": 0.8354,
+      "step": 10123
+    },
+    {
+      "epoch": 1.948937603773131,
+      "grad_norm": 3.214496038640966,
+      "learning_rate": 3.385759175809966e-08,
+      "loss": 0.8113,
+      "step": 10124
+    },
+    {
+      "epoch": 1.9491301104507062,
+      "grad_norm": 3.241682633462666,
+      "learning_rate": 3.360172411083329e-08,
+      "loss": 0.8526,
+      "step": 10125
+    },
+    {
+      "epoch": 1.9493226171282816,
+      "grad_norm": 3.2212387952991834,
+      "learning_rate": 3.3346825324986896e-08,
+      "loss": 0.822,
+      "step": 10126
+    },
+    {
+      "epoch": 1.949515123805857,
+      "grad_norm": 3.215610096045776,
+      "learning_rate": 3.3092895425339553e-08,
+      "loss": 0.8031,
+      "step": 10127
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "grad_norm": 3.3070457664905555,
+      "learning_rate": 3.283993443657596e-08,
+      "loss": 0.8749,
+      "step": 10128
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.5524,
+      "step": 10128,
+      "vm_loss": 0.1645
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.6875,
+      "step": 10128,
+      "vm_loss": 0.2007
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.8999,
+      "step": 10128,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.4945,
+      "step": 10128,
+      "vm_loss": 0.229
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.4811,
+      "step": 10128,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.9966,
+      "step": 10128,
+      "vm_loss": 0.1257
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.6103,
+      "step": 10128,
+      "vm_loss": 0.2317
+    },
+    {
+      "epoch": 1.9497076304834324,
+      "lm_loss": 0.4787,
+      "step": 10128,
+      "vm_loss": 0.1549
+    },
+    {
+      "epoch": 1.9499001371610079,
+      "grad_norm": 3.2293690580001324,
+      "learning_rate": 3.2587942383288664e-08,
+      "loss": 0.8323,
+      "step": 10129
+    },
+    {
+      "epoch": 1.950092643838583,
+      "grad_norm": 3.326533853363896,
+      "learning_rate": 3.233691928997362e-08,
+      "loss": 0.8557,
+      "step": 10130
+    },
+    {
+      "epoch": 1.9502851505161587,
+      "grad_norm": 3.146781148874096,
+      "learning_rate": 3.208686518103465e-08,
+      "loss": 0.8251,
+      "step": 10131
+    },
+    {
+      "epoch": 1.950477657193734,
+      "grad_norm": 3.256698827587504,
+      "learning_rate": 3.183778008077787e-08,
+      "loss": 0.8736,
+      "step": 10132
+    },
+    {
+      "epoch": 1.9506701638713093,
+      "grad_norm": 3.4029389029264325,
+      "learning_rate": 3.158966401341945e-08,
+      "loss": 0.8987,
+      "step": 10133
+    },
+    {
+      "epoch": 1.9508626705488847,
+      "grad_norm": 3.3123194247520926,
+      "learning_rate": 3.1342517003079e-08,
+      "loss": 0.8339,
+      "step": 10134
+    },
+    {
+      "epoch": 1.95105517722646,
+      "grad_norm": 3.2011050721371777,
+      "learning_rate": 3.109633907378173e-08,
+      "loss": 0.8205,
+      "step": 10135
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "grad_norm": 3.187207986100193,
+      "learning_rate": 3.0851130249460737e-08,
+      "loss": 0.8746,
+      "step": 10136
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.8042,
+      "step": 10136,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.4996,
+      "step": 10136,
+      "vm_loss": 0.2241
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.3203,
+      "step": 10136,
+      "vm_loss": 0.119
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.817,
+      "step": 10136,
+      "vm_loss": 0.1825
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.8257,
+      "step": 10136,
+      "vm_loss": 0.1084
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.7018,
+      "step": 10136,
+      "vm_loss": 0.1325
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.5891,
+      "step": 10136,
+      "vm_loss": 0.1567
+    },
+    {
+      "epoch": 1.9512476839040356,
+      "lm_loss": 0.6358,
+      "step": 10136,
+      "vm_loss": 0.1761
+    },
+    {
+      "epoch": 1.9514401905816108,
+      "grad_norm": 3.128864839202728,
+      "learning_rate": 3.0606890553951384e-08,
+      "loss": 0.8242,
+      "step": 10137
+    },
+    {
+      "epoch": 1.9516326972591862,
+      "grad_norm": 3.365609096739883,
+      "learning_rate": 3.036362001099913e-08,
+      "loss": 0.8602,
+      "step": 10138
+    },
+    {
+      "epoch": 1.9518252039367616,
+      "grad_norm": 3.225548835249849,
+      "learning_rate": 3.012131864425061e-08,
+      "loss": 0.8303,
+      "step": 10139
+    },
+    {
+      "epoch": 1.9520177106143368,
+      "grad_norm": 3.3427118605209136,
+      "learning_rate": 2.987998647726365e-08,
+      "loss": 0.8636,
+      "step": 10140
+    },
+    {
+      "epoch": 1.9522102172919125,
+      "grad_norm": 3.2588357346657886,
+      "learning_rate": 2.963962353349503e-08,
+      "loss": 0.8466,
+      "step": 10141
+    },
+    {
+      "epoch": 1.9524027239694877,
+      "grad_norm": 3.3131573918408135,
+      "learning_rate": 2.9400229836314965e-08,
+      "loss": 0.8569,
+      "step": 10142
+    },
+    {
+      "epoch": 1.952595230647063,
+      "grad_norm": 3.2235368958643087,
+      "learning_rate": 2.9161805408992606e-08,
+      "loss": 0.8111,
+      "step": 10143
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "grad_norm": 3.2608283611262268,
+      "learning_rate": 2.8924350274708303e-08,
+      "loss": 0.8679,
+      "step": 10144
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.4325,
+      "step": 10144,
+      "vm_loss": 0.1834
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.4472,
+      "step": 10144,
+      "vm_loss": 0.1386
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.7697,
+      "step": 10144,
+      "vm_loss": 0.2094
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.4781,
+      "step": 10144,
+      "vm_loss": 0.1242
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.752,
+      "step": 10144,
+      "vm_loss": 0.1864
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.8934,
+      "step": 10144,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.6567,
+      "step": 10144,
+      "vm_loss": 0.1644
+    },
+    {
+      "epoch": 1.9527877373246385,
+      "lm_loss": 0.6762,
+      "step": 10144,
+      "vm_loss": 0.1592
+    },
+    {
+      "epoch": 1.9529802440022137,
+      "grad_norm": 3.206351114458451,
+      "learning_rate": 2.868786445654359e-08,
+      "loss": 0.8153,
+      "step": 10145
+    },
+    {
+      "epoch": 1.9531727506797893,
+      "grad_norm": 3.341064358288513,
+      "learning_rate": 2.845234797748897e-08,
+      "loss": 0.855,
+      "step": 10146
+    },
+    {
+      "epoch": 1.9533652573573645,
+      "grad_norm": 3.2374080580376674,
+      "learning_rate": 2.8217800860439458e-08,
+      "loss": 0.8456,
+      "step": 10147
+    },
+    {
+      "epoch": 1.95355776403494,
+      "grad_norm": 3.234795006105892,
+      "learning_rate": 2.798422312819571e-08,
+      "loss": 0.8245,
+      "step": 10148
+    },
+    {
+      "epoch": 1.9537502707125154,
+      "grad_norm": 3.2679885685405994,
+      "learning_rate": 2.775161480346511e-08,
+      "loss": 0.848,
+      "step": 10149
+    },
+    {
+      "epoch": 1.9539427773900906,
+      "grad_norm": 3.208781692450976,
+      "learning_rate": 2.7519975908859576e-08,
+      "loss": 0.8482,
+      "step": 10150
+    },
+    {
+      "epoch": 1.9541352840676662,
+      "grad_norm": 3.1018058658870147,
+      "learning_rate": 2.7289306466897758e-08,
+      "loss": 0.7718,
+      "step": 10151
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "grad_norm": 3.051579520195501,
+      "learning_rate": 2.7059606500003945e-08,
+      "loss": 0.8079,
+      "step": 10152
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.888,
+      "step": 10152,
+      "vm_loss": 0.2029
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.82,
+      "step": 10152,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.6736,
+      "step": 10152,
+      "vm_loss": 0.2013
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.772,
+      "step": 10152,
+      "vm_loss": 0.1668
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.7722,
+      "step": 10152,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.9116,
+      "step": 10152,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.6498,
+      "step": 10152,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.9543277907452414,
+      "lm_loss": 0.4299,
+      "step": 10152,
+      "vm_loss": 0.1735
+    },
+    {
+      "epoch": 1.9545202974228169,
+      "grad_norm": 3.2617761787851984,
+      "learning_rate": 2.6830876030506935e-08,
+      "loss": 0.8375,
+      "step": 10153
+    },
+    {
+      "epoch": 1.9547128041003923,
+      "grad_norm": 3.389546250088972,
+      "learning_rate": 2.6603115080643393e-08,
+      "loss": 0.8456,
+      "step": 10154
+    },
+    {
+      "epoch": 1.9549053107779675,
+      "grad_norm": 3.148277664659818,
+      "learning_rate": 2.6376323672553383e-08,
+      "loss": 0.8315,
+      "step": 10155
+    },
+    {
+      "epoch": 1.9550978174555431,
+      "grad_norm": 3.096522621588826,
+      "learning_rate": 2.6150501828284826e-08,
+      "loss": 0.8085,
+      "step": 10156
+    },
+    {
+      "epoch": 1.9552903241331183,
+      "grad_norm": 3.0705434290009443,
+      "learning_rate": 2.5925649569791266e-08,
+      "loss": 0.8087,
+      "step": 10157
+    },
+    {
+      "epoch": 1.9554828308106937,
+      "grad_norm": 3.13724463343224,
+      "learning_rate": 2.5701766918929672e-08,
+      "loss": 0.8028,
+      "step": 10158
+    },
+    {
+      "epoch": 1.9556753374882692,
+      "grad_norm": 3.207366228304054,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.8527,
+      "step": 10159
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "grad_norm": 3.400290780938795,
+      "learning_rate": 2.5256910527066136e-08,
+      "loss": 0.8451,
+      "step": 10160
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.4303,
+      "step": 10160,
+      "vm_loss": 0.1774
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.5178,
+      "step": 10160,
+      "vm_loss": 0.1101
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.5244,
+      "step": 10160,
+      "vm_loss": 0.2812
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.5143,
+      "step": 10160,
+      "vm_loss": 0.1389
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.6512,
+      "step": 10160,
+      "vm_loss": 0.1473
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.4822,
+      "step": 10160,
+      "vm_loss": 0.228
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.6638,
+      "step": 10160,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 1.9558678441658444,
+      "lm_loss": 0.8131,
+      "step": 10160,
+      "vm_loss": 0.1511
+    },
+    {
+      "epoch": 1.95606035084342,
+      "grad_norm": 3.130830682759224,
+      "learning_rate": 2.5035936829310715e-08,
+      "loss": 0.8203,
+      "step": 10161
+    },
+    {
+      "epoch": 1.9562528575209952,
+      "grad_norm": 3.2748870706060855,
+      "learning_rate": 2.4815932825679177e-08,
+      "loss": 0.8311,
+      "step": 10162
+    },
+    {
+      "epoch": 1.9564453641985706,
+      "grad_norm": 3.14441595152886,
+      "learning_rate": 2.4596898537558868e-08,
+      "loss": 0.8117,
+      "step": 10163
+    },
+    {
+      "epoch": 1.956637870876146,
+      "grad_norm": 3.2182987337532265,
+      "learning_rate": 2.4378833986242745e-08,
+      "loss": 0.8529,
+      "step": 10164
+    },
+    {
+      "epoch": 1.9568303775537212,
+      "grad_norm": 3.198950858321288,
+      "learning_rate": 2.4161739192929413e-08,
+      "loss": 0.8091,
+      "step": 10165
+    },
+    {
+      "epoch": 1.9570228842312969,
+      "grad_norm": 3.1616524660278964,
+      "learning_rate": 2.39456141787231e-08,
+      "loss": 0.8138,
+      "step": 10166
+    },
+    {
+      "epoch": 1.957215390908872,
+      "grad_norm": 3.2708745002144646,
+      "learning_rate": 2.3730458964633663e-08,
+      "loss": 0.8255,
+      "step": 10167
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "grad_norm": 3.2280621797080347,
+      "learning_rate": 2.3516273571577708e-08,
+      "loss": 0.8025,
+      "step": 10168
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.8537,
+      "step": 10168,
+      "vm_loss": 0.1948
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.7026,
+      "step": 10168,
+      "vm_loss": 0.1021
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.8448,
+      "step": 10168,
+      "vm_loss": 0.1231
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.5465,
+      "step": 10168,
+      "vm_loss": 0.1784
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.7371,
+      "step": 10168,
+      "vm_loss": 0.2035
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.6394,
+      "step": 10168,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 0.7971,
+      "step": 10168,
+      "vm_loss": 0.121
+    },
+    {
+      "epoch": 1.9574078975864475,
+      "lm_loss": 1.1849,
+      "step": 10168,
+      "vm_loss": 0.1744
+    },
+    {
+      "epoch": 1.957600404264023,
+      "grad_norm": 3.06664061265489,
+      "learning_rate": 2.3303058020376356e-08,
+      "loss": 0.8198,
+      "step": 10169
+    },
+    {
+      "epoch": 1.9577929109415981,
+      "grad_norm": 3.167473682828059,
+      "learning_rate": 2.309081233175747e-08,
+      "loss": 0.7841,
+      "step": 10170
+    },
+    {
+      "epoch": 1.9579854176191738,
+      "grad_norm": 3.498972712644818,
+      "learning_rate": 2.287953652635344e-08,
+      "loss": 0.8753,
+      "step": 10171
+    },
+    {
+      "epoch": 1.958177924296749,
+      "grad_norm": 3.2354186918751338,
+      "learning_rate": 2.2669230624702277e-08,
+      "loss": 0.8358,
+      "step": 10172
+    },
+    {
+      "epoch": 1.9583704309743244,
+      "grad_norm": 3.2327939391128457,
+      "learning_rate": 2.2459894647250956e-08,
+      "loss": 0.8227,
+      "step": 10173
+    },
+    {
+      "epoch": 1.9585629376518998,
+      "grad_norm": 3.1299842956838644,
+      "learning_rate": 2.225152861434654e-08,
+      "loss": 0.7831,
+      "step": 10174
+    },
+    {
+      "epoch": 1.958755444329475,
+      "grad_norm": 3.2964602800033216,
+      "learning_rate": 2.2044132546246156e-08,
+      "loss": 0.8349,
+      "step": 10175
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "grad_norm": 3.1965252003900066,
+      "learning_rate": 2.183770646311145e-08,
+      "loss": 0.817,
+      "step": 10176
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.5856,
+      "step": 10176,
+      "vm_loss": 0.1725
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.7886,
+      "step": 10176,
+      "vm_loss": 0.1525
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.422,
+      "step": 10176,
+      "vm_loss": 0.2509
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.5699,
+      "step": 10176,
+      "vm_loss": 0.2431
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 1.1904,
+      "step": 10176,
+      "vm_loss": 0.1755
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.7439,
+      "step": 10176,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.6648,
+      "step": 10176,
+      "vm_loss": 0.1854
+    },
+    {
+      "epoch": 1.9589479510070507,
+      "lm_loss": 0.5658,
+      "step": 10176,
+      "vm_loss": 0.1602
+    },
+    {
+      "epoch": 1.9591404576846259,
+      "grad_norm": 3.2230068015229763,
+      "learning_rate": 2.163225038501082e-08,
+      "loss": 0.8382,
+      "step": 10177
+    },
+    {
+      "epoch": 1.9593329643622013,
+      "grad_norm": 3.377070999729219,
+      "learning_rate": 2.142776433191496e-08,
+      "loss": 0.8503,
+      "step": 10178
+    },
+    {
+      "epoch": 1.9595254710397767,
+      "grad_norm": 3.2984859546716576,
+      "learning_rate": 2.1224248323703512e-08,
+      "loss": 0.8191,
+      "step": 10179
+    },
+    {
+      "epoch": 1.9597179777173521,
+      "grad_norm": 3.1991944841022906,
+      "learning_rate": 2.102170238016177e-08,
+      "loss": 0.8266,
+      "step": 10180
+    },
+    {
+      "epoch": 1.9599104843949275,
+      "grad_norm": 3.503770276607305,
+      "learning_rate": 2.0820126520978425e-08,
+      "loss": 0.8803,
+      "step": 10181
+    },
+    {
+      "epoch": 1.9601029910725027,
+      "grad_norm": 3.0893165955559643,
+      "learning_rate": 2.0619520765750022e-08,
+      "loss": 0.8125,
+      "step": 10182
+    },
+    {
+      "epoch": 1.9602954977500782,
+      "grad_norm": 3.2377109115152467,
+      "learning_rate": 2.0419885133976525e-08,
+      "loss": 0.8383,
+      "step": 10183
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "grad_norm": 3.253946682056954,
+      "learning_rate": 2.0221219645067956e-08,
+      "loss": 0.8203,
+      "step": 10184
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.5413,
+      "step": 10184,
+      "vm_loss": 0.1347
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.6911,
+      "step": 10184,
+      "vm_loss": 0.1925
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.9271,
+      "step": 10184,
+      "vm_loss": 0.1562
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.3395,
+      "step": 10184,
+      "vm_loss": 0.1589
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.842,
+      "step": 10184,
+      "vm_loss": 0.1678
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.6407,
+      "step": 10184,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.5555,
+      "step": 10184,
+      "vm_loss": 0.124
+    },
+    {
+      "epoch": 1.9604880044276536,
+      "lm_loss": 0.4456,
+      "step": 10184,
+      "vm_loss": 0.1707
+    },
+    {
+      "epoch": 1.960680511105229,
+      "grad_norm": 3.202562127832382,
+      "learning_rate": 2.0023524318334432e-08,
+      "loss": 0.7775,
+      "step": 10185
+    },
+    {
+      "epoch": 1.9608730177828044,
+      "grad_norm": 3.3294954626682336,
+      "learning_rate": 1.9826799172996125e-08,
+      "loss": 0.8881,
+      "step": 10186
+    },
+    {
+      "epoch": 1.9610655244603796,
+      "grad_norm": 3.373220143029709,
+      "learning_rate": 1.9631044228176633e-08,
+      "loss": 0.8605,
+      "step": 10187
+    },
+    {
+      "epoch": 1.961258031137955,
+      "grad_norm": 3.0916739058517955,
+      "learning_rate": 1.9436259502906284e-08,
+      "loss": 0.7834,
+      "step": 10188
+    },
+    {
+      "epoch": 1.9614505378155305,
+      "grad_norm": 3.288469126668882,
+      "learning_rate": 1.9242445016118826e-08,
+      "loss": 0.8149,
+      "step": 10189
+    },
+    {
+      "epoch": 1.961643044493106,
+      "grad_norm": 3.422558579543728,
+      "learning_rate": 1.9049600786658073e-08,
+      "loss": 0.8883,
+      "step": 10190
+    },
+    {
+      "epoch": 1.9618355511706813,
+      "grad_norm": 3.277361239622352,
+      "learning_rate": 1.885772683326903e-08,
+      "loss": 0.827,
+      "step": 10191
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "grad_norm": 3.29850755178506,
+      "learning_rate": 1.8666823174605665e-08,
+      "loss": 0.8558,
+      "step": 10192
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.8428,
+      "step": 10192,
+      "vm_loss": 0.1506
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.9339,
+      "step": 10192,
+      "vm_loss": 0.1561
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.5887,
+      "step": 10192,
+      "vm_loss": 0.1449
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.7009,
+      "step": 10192,
+      "vm_loss": 0.1745
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.8848,
+      "step": 10192,
+      "vm_loss": 0.1954
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 1.0809,
+      "step": 10192,
+      "vm_loss": 0.1045
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.6609,
+      "step": 10192,
+      "vm_loss": 0.1531
+    },
+    {
+      "epoch": 1.9620280578482565,
+      "lm_loss": 0.5582,
+      "step": 10192,
+      "vm_loss": 0.1224
+    },
+    {
+      "epoch": 1.9622205645258322,
+      "grad_norm": 3.349223870528647,
+      "learning_rate": 1.8476889829225352e-08,
+      "loss": 0.8317,
+      "step": 10193
+    },
+    {
+      "epoch": 1.9624130712034074,
+      "grad_norm": 3.18565841403218,
+      "learning_rate": 1.8287926815592217e-08,
+      "loss": 0.7681,
+      "step": 10194
+    },
+    {
+      "epoch": 1.9626055778809828,
+      "grad_norm": 3.4301533874208494,
+      "learning_rate": 1.8099934152076003e-08,
+      "loss": 0.8626,
+      "step": 10195
+    },
+    {
+      "epoch": 1.9627980845585582,
+      "grad_norm": 3.1925917490230042,
+      "learning_rate": 1.7912911856952098e-08,
+      "loss": 0.8263,
+      "step": 10196
+    },
+    {
+      "epoch": 1.9629905912361334,
+      "grad_norm": 3.4370946038941366,
+      "learning_rate": 1.7726859948401508e-08,
+      "loss": 0.8632,
+      "step": 10197
+    },
+    {
+      "epoch": 1.963183097913709,
+      "grad_norm": 3.4484958181406515,
+      "learning_rate": 1.7541778444509773e-08,
+      "loss": 0.8993,
+      "step": 10198
+    },
+    {
+      "epoch": 1.9633756045912842,
+      "grad_norm": 3.163027131757195,
+      "learning_rate": 1.735766736327138e-08,
+      "loss": 0.8003,
+      "step": 10199
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "grad_norm": 3.1013512748531964,
+      "learning_rate": 1.717452672258202e-08,
+      "loss": 0.8072,
+      "step": 10200
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.7457,
+      "step": 10200,
+      "vm_loss": 0.1586
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.2945,
+      "step": 10200,
+      "vm_loss": 0.128
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.5406,
+      "step": 10200,
+      "vm_loss": 0.1336
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.7136,
+      "step": 10200,
+      "vm_loss": 0.1555
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.456,
+      "step": 10200,
+      "vm_loss": 0.1227
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.5551,
+      "step": 10200,
+      "vm_loss": 0.1637
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.2932,
+      "step": 10200,
+      "vm_loss": 0.181
+    },
+    {
+      "epoch": 1.9635681112688597,
+      "lm_loss": 0.5556,
+      "step": 10200,
+      "vm_loss": 0.2372
+    },
+    {
+      "epoch": 1.963760617946435,
+      "grad_norm": 3.1788950564569074,
+      "learning_rate": 1.6992356540247446e-08,
+      "loss": 0.8226,
+      "step": 10201
+    },
+    {
+      "epoch": 1.9639531246240103,
+      "grad_norm": 3.1253371178934324,
+      "learning_rate": 1.6811156833974605e-08,
+      "loss": 0.7988,
+      "step": 10202
+    },
+    {
+      "epoch": 1.964145631301586,
+      "grad_norm": 3.324519195360958,
+      "learning_rate": 1.663092762137941e-08,
+      "loss": 0.8552,
+      "step": 10203
+    },
+    {
+      "epoch": 1.9643381379791611,
+      "grad_norm": 3.01185292621742,
+      "learning_rate": 1.64516689199834e-08,
+      "loss": 0.7661,
+      "step": 10204
+    },
+    {
+      "epoch": 1.9645306446567365,
+      "grad_norm": 3.19546864611251,
+      "learning_rate": 1.627338074721263e-08,
+      "loss": 0.8266,
+      "step": 10205
+    },
+    {
+      "epoch": 1.964723151334312,
+      "grad_norm": 3.2437347023354377,
+      "learning_rate": 1.6096063120396576e-08,
+      "loss": 0.8578,
+      "step": 10206
+    },
+    {
+      "epoch": 1.9649156580118872,
+      "grad_norm": 3.2013652616262984,
+      "learning_rate": 1.5919716056775892e-08,
+      "loss": 0.7826,
+      "step": 10207
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "grad_norm": 3.2535518492923075,
+      "learning_rate": 1.574433957349242e-08,
+      "loss": 0.8481,
+      "step": 10208
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.9669,
+      "step": 10208,
+      "vm_loss": 0.1234
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.8535,
+      "step": 10208,
+      "vm_loss": 0.1392
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.4212,
+      "step": 10208,
+      "vm_loss": 0.1495
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.5655,
+      "step": 10208,
+      "vm_loss": 0.1137
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.8385,
+      "step": 10208,
+      "vm_loss": 0.1871
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.5437,
+      "step": 10208,
+      "vm_loss": 0.1299
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.9348,
+      "step": 10208,
+      "vm_loss": 0.1149
+    },
+    {
+      "epoch": 1.9651081646894628,
+      "lm_loss": 0.6903,
+      "step": 10208,
+      "vm_loss": 0.1546
+    },
+    {
+      "epoch": 1.965300671367038,
+      "grad_norm": 3.2577807949095368,
+      "learning_rate": 1.556993368759585e-08,
+      "loss": 0.8217,
+      "step": 10209
+    },
+    {
+      "epoch": 1.9654931780446134,
+      "grad_norm": 3.1439382090766834,
+      "learning_rate": 1.5396498416038186e-08,
+      "loss": 0.8252,
+      "step": 10210
+    },
+    {
+      "epoch": 1.9656856847221889,
+      "grad_norm": 3.2243412725068956,
+      "learning_rate": 1.52240337756826e-08,
+      "loss": 0.8423,
+      "step": 10211
+    },
+    {
+      "epoch": 1.965878191399764,
+      "grad_norm": 3.2922122216103196,
+      "learning_rate": 1.5052539783292353e-08,
+      "loss": 0.8403,
+      "step": 10212
+    },
+    {
+      "epoch": 1.9660706980773397,
+      "grad_norm": 3.329557140736555,
+      "learning_rate": 1.4882016455540772e-08,
+      "loss": 0.8794,
+      "step": 10213
+    },
+    {
+      "epoch": 1.966263204754915,
+      "grad_norm": 3.2694591128578088,
+      "learning_rate": 1.4712463809004596e-08,
+      "loss": 0.8317,
+      "step": 10214
+    },
+    {
+      "epoch": 1.9664557114324903,
+      "grad_norm": 3.248884120966567,
+      "learning_rate": 1.4543881860165088e-08,
+      "loss": 0.8457,
+      "step": 10215
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "grad_norm": 3.164956959347906,
+      "learning_rate": 1.437627062541136e-08,
+      "loss": 0.8018,
+      "step": 10216
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.4372,
+      "step": 10216,
+      "vm_loss": 0.1404
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.7794,
+      "step": 10216,
+      "vm_loss": 0.1914
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.8622,
+      "step": 10216,
+      "vm_loss": 0.1294
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.5169,
+      "step": 10216,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.7219,
+      "step": 10216,
+      "vm_loss": 0.204
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.6495,
+      "step": 10216,
+      "vm_loss": 0.1346
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.5811,
+      "step": 10216,
+      "vm_loss": 0.1229
+    },
+    {
+      "epoch": 1.9666482181100657,
+      "lm_loss": 0.5585,
+      "step": 10216,
+      "vm_loss": 0.1382
+    },
+    {
+      "epoch": 1.966840724787641,
+      "grad_norm": 3.3482625365034675,
+      "learning_rate": 1.4209630121039264e-08,
+      "loss": 0.8257,
+      "step": 10217
+    },
+    {
+      "epoch": 1.9670332314652166,
+      "grad_norm": 3.248368575866094,
+      "learning_rate": 1.4043960363244735e-08,
+      "loss": 0.8012,
+      "step": 10218
+    },
+    {
+      "epoch": 1.9672257381427918,
+      "grad_norm": 3.343227322493252,
+      "learning_rate": 1.3879261368135999e-08,
+      "loss": 0.8644,
+      "step": 10219
+    },
+    {
+      "epoch": 1.9674182448203672,
+      "grad_norm": 3.2652494759391955,
+      "learning_rate": 1.371553315172247e-08,
+      "loss": 0.8519,
+      "step": 10220
+    },
+    {
+      "epoch": 1.9676107514979426,
+      "grad_norm": 3.1809609814871243,
+      "learning_rate": 1.355277572992142e-08,
+      "loss": 0.8303,
+      "step": 10221
+    },
+    {
+      "epoch": 1.9678032581755178,
+      "grad_norm": 3.3971726500441566,
+      "learning_rate": 1.3390989118554631e-08,
+      "loss": 0.882,
+      "step": 10222
+    },
+    {
+      "epoch": 1.9679957648530935,
+      "grad_norm": 3.3313851744587692,
+      "learning_rate": 1.3230173333348417e-08,
+      "loss": 0.8406,
+      "step": 10223
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "grad_norm": 3.3668217585292544,
+      "learning_rate": 1.3070328389939157e-08,
+      "loss": 0.8659,
+      "step": 10224
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.7855,
+      "step": 10224,
+      "vm_loss": 0.1344
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.6196,
+      "step": 10224,
+      "vm_loss": 0.156
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.6851,
+      "step": 10224,
+      "vm_loss": 0.2172
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.6548,
+      "step": 10224,
+      "vm_loss": 0.1959
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 1.0306,
+      "step": 10224,
+      "vm_loss": 0.1032
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.9352,
+      "step": 10224,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.8014,
+      "step": 10224,
+      "vm_loss": 0.2218
+    },
+    {
+      "epoch": 1.9681882715306687,
+      "lm_loss": 0.5073,
+      "step": 10224,
+      "vm_loss": 0.2306
+    },
+    {
+      "epoch": 1.968380778208244,
+      "grad_norm": 3.2626960522704174,
+      "learning_rate": 1.2911454303863314e-08,
+      "loss": 0.8367,
+      "step": 10225
+    },
+    {
+      "epoch": 1.9685732848858195,
+      "grad_norm": 3.3376822624604934,
+      "learning_rate": 1.2753551090566307e-08,
+      "loss": 0.8271,
+      "step": 10226
+    },
+    {
+      "epoch": 1.9687657915633947,
+      "grad_norm": 3.159134516745942,
+      "learning_rate": 1.2596618765398083e-08,
+      "loss": 0.8115,
+      "step": 10227
+    },
+    {
+      "epoch": 1.9689582982409704,
+      "grad_norm": 3.287706200591301,
+      "learning_rate": 1.2440657343615325e-08,
+      "loss": 0.8354,
+      "step": 10228
+    },
+    {
+      "epoch": 1.9691508049185455,
+      "grad_norm": 3.124984658151881,
+      "learning_rate": 1.228566684037813e-08,
+      "loss": 0.7755,
+      "step": 10229
+    },
+    {
+      "epoch": 1.969343311596121,
+      "grad_norm": 3.371726410472818,
+      "learning_rate": 1.2131647270754443e-08,
+      "loss": 0.8585,
+      "step": 10230
+    },
+    {
+      "epoch": 1.9695358182736964,
+      "grad_norm": 3.122454071069264,
+      "learning_rate": 1.1978598649716733e-08,
+      "loss": 0.802,
+      "step": 10231
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "grad_norm": 3.1684688914167434,
+      "learning_rate": 1.1826520992144208e-08,
+      "loss": 0.8123,
+      "step": 10232
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.8006,
+      "step": 10232,
+      "vm_loss": 0.214
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.632,
+      "step": 10232,
+      "vm_loss": 0.1461
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 1.0004,
+      "step": 10232,
+      "vm_loss": 0.1653
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.6753,
+      "step": 10232,
+      "vm_loss": 0.1727
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.7062,
+      "step": 10232,
+      "vm_loss": 0.2026
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.348,
+      "step": 10232,
+      "vm_loss": 0.157
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.4536,
+      "step": 10232,
+      "vm_loss": 0.2053
+    },
+    {
+      "epoch": 1.9697283249512716,
+      "lm_loss": 0.7613,
+      "step": 10232,
+      "vm_loss": 0.2098
+    },
+    {
+      "epoch": 1.9699208316288472,
+      "grad_norm": 3.104098898228278,
+      "learning_rate": 1.1675414312819488e-08,
+      "loss": 0.8349,
+      "step": 10233
+    },
+    {
+      "epoch": 1.9701133383064224,
+      "grad_norm": 3.3780090726454866,
+      "learning_rate": 1.1525278626431935e-08,
+      "loss": 0.858,
+      "step": 10234
+    },
+    {
+      "epoch": 1.9703058449839979,
+      "grad_norm": 3.069495782910486,
+      "learning_rate": 1.1376113947577649e-08,
+      "loss": 0.7911,
+      "step": 10235
+    },
+    {
+      "epoch": 1.9704983516615733,
+      "grad_norm": 3.2891369864121227,
+      "learning_rate": 1.1227920290757255e-08,
+      "loss": 0.8623,
+      "step": 10236
+    },
+    {
+      "epoch": 1.9706908583391485,
+      "grad_norm": 3.345186112844311,
+      "learning_rate": 1.1080697670375896e-08,
+      "loss": 0.8624,
+      "step": 10237
+    },
+    {
+      "epoch": 1.9708833650167241,
+      "grad_norm": 3.525090793930837,
+      "learning_rate": 1.0934446100746566e-08,
+      "loss": 0.9279,
+      "step": 10238
+    },
+    {
+      "epoch": 1.9710758716942993,
+      "grad_norm": 3.124773135436261,
+      "learning_rate": 1.0789165596086782e-08,
+      "loss": 0.7975,
+      "step": 10239
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "grad_norm": 3.2598522818478672,
+      "learning_rate": 1.064485617051969e-08,
+      "loss": 0.8352,
+      "step": 10240
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.3821,
+      "step": 10240,
+      "vm_loss": 0.1801
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.539,
+      "step": 10240,
+      "vm_loss": 0.1265
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.7572,
+      "step": 10240,
+      "vm_loss": 0.1955
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.5064,
+      "step": 10240,
+      "vm_loss": 0.1084
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 1.0349,
+      "step": 10240,
+      "vm_loss": 0.1238
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.4137,
+      "step": 10240,
+      "vm_loss": 0.0911
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.4583,
+      "step": 10240,
+      "vm_loss": 0.2668
+    },
+    {
+      "epoch": 1.9712683783718747,
+      "lm_loss": 0.6371,
+      "step": 10240,
+      "vm_loss": 0.2232
+    },
+    {
+      "epoch": 1.9714608850494502,
+      "grad_norm": 3.288363565488047,
+      "learning_rate": 1.050151783807296e-08,
+      "loss": 0.8548,
+      "step": 10241
+    },
+    {
+      "epoch": 1.9716533917270256,
+      "grad_norm": 3.2925694024663748,
+      "learning_rate": 1.0359150612682112e-08,
+      "loss": 0.8807,
+      "step": 10242
+    },
+    {
+      "epoch": 1.971845898404601,
+      "grad_norm": 3.16333679507627,
+      "learning_rate": 1.0217754508187183e-08,
+      "loss": 0.8169,
+      "step": 10243
+    },
+    {
+      "epoch": 1.9720384050821762,
+      "grad_norm": 3.1850822637916933,
+      "learning_rate": 1.0077329538331626e-08,
+      "loss": 0.7628,
+      "step": 10244
+    },
+    {
+      "epoch": 1.9722309117597516,
+      "grad_norm": 3.222136683736353,
+      "learning_rate": 9.937875716770074e-09,
+      "loss": 0.8248,
+      "step": 10245
+    },
+    {
+      "epoch": 1.972423418437327,
+      "grad_norm": 3.022549407975293,
+      "learning_rate": 9.799393057056128e-09,
+      "loss": 0.7663,
+      "step": 10246
+    },
+    {
+      "epoch": 1.9726159251149025,
+      "grad_norm": 3.2308311320145915,
+      "learning_rate": 9.661881572652355e-09,
+      "loss": 0.8157,
+      "step": 10247
+    },
+    {
+      "epoch": 1.972808431792478,
+      "grad_norm": 3.1239226656656505,
+      "learning_rate": 9.525341276929167e-09,
+      "loss": 0.8128,
+      "step": 10248
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.7372,
+      "step": 10248,
+      "vm_loss": 0.1565
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 1.0928,
+      "step": 10248,
+      "vm_loss": 0.1632
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 1.1286,
+      "step": 10248,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.6941,
+      "step": 10248,
+      "vm_loss": 0.1973
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.4359,
+      "step": 10248,
+      "vm_loss": 0.1419
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.3742,
+      "step": 10248,
+      "vm_loss": 0.1711
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.6855,
+      "step": 10248,
+      "vm_loss": 0.1897
+    },
+    {
+      "epoch": 1.972808431792478,
+      "lm_loss": 0.3473,
+      "step": 10248,
+      "vm_loss": 0.1486
+    },
+    {
+      "epoch": 1.973000938470053,
+      "grad_norm": 3.2558960273622652,
+      "learning_rate": 9.389772183157064e-09,
+      "loss": 0.8398,
+      "step": 10249
+    },
+    {
+      "epoch": 1.9731934451476285,
+      "grad_norm": 3.266226389932565,
+      "learning_rate": 9.255174304516611e-09,
+      "loss": 0.7842,
+      "step": 10250
+    },
+    {
+      "epoch": 1.973385951825204,
+      "grad_norm": 3.29926940947336,
+      "learning_rate": 9.121547654091789e-09,
+      "loss": 0.8475,
+      "step": 10251
+    },
+    {
+      "epoch": 1.9735784585027794,
+      "grad_norm": 3.315875973625688,
+      "learning_rate": 8.988892244874425e-09,
+      "loss": 0.8708,
+      "step": 10252
+    },
+    {
+      "epoch": 1.9737709651803548,
+      "grad_norm": 3.0834907487715,
+      "learning_rate": 8.857208089758652e-09,
+      "loss": 0.7975,
+      "step": 10253
+    },
+    {
+      "epoch": 1.97396347185793,
+      "grad_norm": 3.2983948173305473,
+      "learning_rate": 8.726495201545338e-09,
+      "loss": 0.8462,
+      "step": 10254
+    },
+    {
+      "epoch": 1.9741559785355054,
+      "grad_norm": 3.5008527061621657,
+      "learning_rate": 8.596753592943208e-09,
+      "loss": 0.8811,
+      "step": 10255
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "grad_norm": 3.2244309365477495,
+      "learning_rate": 8.467983276563285e-09,
+      "loss": 0.8253,
+      "step": 10256
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.9348,
+      "step": 10256,
+      "vm_loss": 0.1635
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.7365,
+      "step": 10256,
+      "vm_loss": 0.1465
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.5747,
+      "step": 10256,
+      "vm_loss": 0.1894
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.982,
+      "step": 10256,
+      "vm_loss": 0.1277
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.9199,
+      "step": 10256,
+      "vm_loss": 0.1351
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.7265,
+      "step": 10256,
+      "vm_loss": 0.1935
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.5097,
+      "step": 10256,
+      "vm_loss": 0.2047
+    },
+    {
+      "epoch": 1.9743484852130808,
+      "lm_loss": 0.8511,
+      "step": 10256,
+      "vm_loss": 0.1535
+    },
+    {
+      "epoch": 1.9745409918906562,
+      "grad_norm": 3.171550756440761,
+      "learning_rate": 8.340184264925555e-09,
+      "loss": 0.8221,
+      "step": 10257
+    },
+    {
+      "epoch": 1.9747334985682317,
+      "grad_norm": 3.226557052440261,
+      "learning_rate": 8.213356570452303e-09,
+      "loss": 0.8057,
+      "step": 10258
+    },
+    {
+      "epoch": 1.9749260052458069,
+      "grad_norm": 3.2514042950827373,
+      "learning_rate": 8.087500205472553e-09,
+      "loss": 0.8334,
+      "step": 10259
+    },
+    {
+      "epoch": 1.9751185119233825,
+      "grad_norm": 3.2403624841701526,
+      "learning_rate": 7.962615182222078e-09,
+      "loss": 0.8269,
+      "step": 10260
+    },
+    {
+      "epoch": 1.9753110186009577,
+      "grad_norm": 3.1835517355890492,
+      "learning_rate": 7.838701512840052e-09,
+      "loss": 0.8324,
+      "step": 10261
+    },
+    {
+      "epoch": 1.9755035252785331,
+      "grad_norm": 3.3599208788539783,
+      "learning_rate": 7.715759209373509e-09,
+      "loss": 0.8087,
+      "step": 10262
+    },
+    {
+      "epoch": 1.9756960319561085,
+      "grad_norm": 3.3042880082433514,
+      "learning_rate": 7.593788283775106e-09,
+      "loss": 0.8241,
+      "step": 10263
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "grad_norm": 3.3454027302229434,
+      "learning_rate": 7.472788747899806e-09,
+      "loss": 0.8565,
+      "step": 10264
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.7951,
+      "step": 10264,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.7826,
+      "step": 10264,
+      "vm_loss": 0.2353
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.6852,
+      "step": 10264,
+      "vm_loss": 0.1997
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.3255,
+      "step": 10264,
+      "vm_loss": 0.1166
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.756,
+      "step": 10264,
+      "vm_loss": 0.1055
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.786,
+      "step": 10264,
+      "vm_loss": 0.1751
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.7667,
+      "step": 10264,
+      "vm_loss": 0.1228
+    },
+    {
+      "epoch": 1.9758885386336837,
+      "lm_loss": 0.4622,
+      "step": 10264,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.9760810453112594,
+      "grad_norm": 3.0452741802535757,
+      "learning_rate": 7.352760613510423e-09,
+      "loss": 0.7863,
+      "step": 10265
+    },
+    {
+      "epoch": 1.9762735519888346,
+      "grad_norm": 3.202450692236302,
+      "learning_rate": 7.233703892277622e-09,
+      "loss": 0.8467,
+      "step": 10266
+    },
+    {
+      "epoch": 1.97646605866641,
+      "grad_norm": 3.1561565482033687,
+      "learning_rate": 7.115618595772145e-09,
+      "loss": 0.8188,
+      "step": 10267
+    },
+    {
+      "epoch": 1.9766585653439854,
+      "grad_norm": 3.1538533029907616,
+      "learning_rate": 6.998504735475919e-09,
+      "loss": 0.8168,
+      "step": 10268
+    },
+    {
+      "epoch": 1.9768510720215606,
+      "grad_norm": 3.3838693872137138,
+      "learning_rate": 6.882362322772063e-09,
+      "loss": 0.9067,
+      "step": 10269
+    },
+    {
+      "epoch": 1.9770435786991363,
+      "grad_norm": 3.3483323662272118,
+      "learning_rate": 6.767191368952653e-09,
+      "loss": 0.8703,
+      "step": 10270
+    },
+    {
+      "epoch": 1.9772360853767115,
+      "grad_norm": 3.1963459426587577,
+      "learning_rate": 6.652991885212068e-09,
+      "loss": 0.803,
+      "step": 10271
+    },
+    {
+      "epoch": 1.977428592054287,
+      "grad_norm": 3.1553178914155118,
+      "learning_rate": 6.539763882653649e-09,
+      "loss": 0.8075,
+      "step": 10272
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.6449,
+      "step": 10272,
+      "vm_loss": 0.1696
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.9748,
+      "step": 10272,
+      "vm_loss": 0.1286
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.3138,
+      "step": 10272,
+      "vm_loss": 0.1499
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.4256,
+      "step": 10272,
+      "vm_loss": 0.1459
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.7019,
+      "step": 10272,
+      "vm_loss": 0.2046
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.9366,
+      "step": 10272,
+      "vm_loss": 0.1876
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.5042,
+      "step": 10272,
+      "vm_loss": 0.1003
+    },
+    {
+      "epoch": 1.977428592054287,
+      "lm_loss": 0.4963,
+      "step": 10272,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.9776210987318623,
+      "grad_norm": 3.3462825031653223,
+      "learning_rate": 6.427507372283037e-09,
+      "loss": 0.8489,
+      "step": 10273
+    },
+    {
+      "epoch": 1.9778136054094375,
+      "grad_norm": 3.2860289450370157,
+      "learning_rate": 6.316222365014834e-09,
+      "loss": 0.8795,
+      "step": 10274
+    },
+    {
+      "epoch": 1.9780061120870132,
+      "grad_norm": 3.2593840235537304,
+      "learning_rate": 6.205908871665944e-09,
+      "loss": 0.8617,
+      "step": 10275
+    },
+    {
+      "epoch": 1.9781986187645884,
+      "grad_norm": 3.2826582733404597,
+      "learning_rate": 6.0965669029611205e-09,
+      "loss": 0.8416,
+      "step": 10276
+    },
+    {
+      "epoch": 1.9783911254421638,
+      "grad_norm": 3.2596252226643037,
+      "learning_rate": 5.988196469528529e-09,
+      "loss": 0.7999,
+      "step": 10277
+    },
+    {
+      "epoch": 1.9785836321197392,
+      "grad_norm": 3.076276653140899,
+      "learning_rate": 5.880797581904185e-09,
+      "loss": 0.7617,
+      "step": 10278
+    },
+    {
+      "epoch": 1.9787761387973144,
+      "grad_norm": 3.265425943029449,
+      "learning_rate": 5.774370250528627e-09,
+      "loss": 0.8102,
+      "step": 10279
+    },
+    {
+      "epoch": 1.97896864547489,
+      "grad_norm": 3.173618709191152,
+      "learning_rate": 5.668914485748022e-09,
+      "loss": 0.8411,
+      "step": 10280
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.5742,
+      "step": 10280,
+      "vm_loss": 0.1818
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.7841,
+      "step": 10280,
+      "vm_loss": 0.1628
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.5466,
+      "step": 10280,
+      "vm_loss": 0.0883
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.5518,
+      "step": 10280,
+      "vm_loss": 0.2193
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.5762,
+      "step": 10280,
+      "vm_loss": 0.1721
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.3086,
+      "step": 10280,
+      "vm_loss": 0.1401
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.5317,
+      "step": 10280,
+      "vm_loss": 0.1979
+    },
+    {
+      "epoch": 1.97896864547489,
+      "lm_loss": 0.3185,
+      "step": 10280,
+      "vm_loss": 0.1659
+    },
+    {
+      "epoch": 1.9791611521524652,
+      "grad_norm": 3.2717339503674507,
+      "learning_rate": 5.564430297813062e-09,
+      "loss": 0.8324,
+      "step": 10281
+    },
+    {
+      "epoch": 1.9793536588300407,
+      "grad_norm": 3.316101689392939,
+      "learning_rate": 5.460917696882284e-09,
+      "loss": 0.8529,
+      "step": 10282
+    },
+    {
+      "epoch": 1.979546165507616,
+      "grad_norm": 3.267550552399221,
+      "learning_rate": 5.358376693017642e-09,
+      "loss": 0.8886,
+      "step": 10283
+    },
+    {
+      "epoch": 1.9797386721851913,
+      "grad_norm": 3.2156733526081287,
+      "learning_rate": 5.256807296187827e-09,
+      "loss": 0.8322,
+      "step": 10284
+    },
+    {
+      "epoch": 1.979931178862767,
+      "grad_norm": 3.4469724245984903,
+      "learning_rate": 5.156209516266053e-09,
+      "loss": 0.8639,
+      "step": 10285
+    },
+    {
+      "epoch": 1.9801236855403421,
+      "grad_norm": 3.3279532968208114,
+      "learning_rate": 5.056583363032275e-09,
+      "loss": 0.8547,
+      "step": 10286
+    },
+    {
+      "epoch": 1.9803161922179175,
+      "grad_norm": 3.3645429520649097,
+      "learning_rate": 4.957928846170967e-09,
+      "loss": 0.8652,
+      "step": 10287
+    },
+    {
+      "epoch": 1.980508698895493,
+      "grad_norm": 3.241975506559042,
+      "learning_rate": 4.860245975273348e-09,
+      "loss": 0.8114,
+      "step": 10288
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 1.1911,
+      "step": 10288,
+      "vm_loss": 0.1036
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.7126,
+      "step": 10288,
+      "vm_loss": 0.1838
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.3572,
+      "step": 10288,
+      "vm_loss": 0.1652
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.3932,
+      "step": 10288,
+      "vm_loss": 0.1762
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.9089,
+      "step": 10288,
+      "vm_loss": 0.1599
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.5353,
+      "step": 10288,
+      "vm_loss": 0.1859
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.707,
+      "step": 10288,
+      "vm_loss": 0.199
+    },
+    {
+      "epoch": 1.980508698895493,
+      "lm_loss": 0.6611,
+      "step": 10288,
+      "vm_loss": 0.1412
+    },
+    {
+      "epoch": 1.9807012055730682,
+      "grad_norm": 3.2456576016656937,
+      "learning_rate": 4.763534759835153e-09,
+      "loss": 0.8479,
+      "step": 10289
+    },
+    {
+      "epoch": 1.9808937122506438,
+      "grad_norm": 3.290254551225311,
+      "learning_rate": 4.6677952092577525e-09,
+      "loss": 0.8887,
+      "step": 10290
+    },
+    {
+      "epoch": 1.981086218928219,
+      "grad_norm": 3.3586891231032743,
+      "learning_rate": 4.5730273328470355e-09,
+      "loss": 0.8324,
+      "step": 10291
+    },
+    {
+      "epoch": 1.9812787256057944,
+      "grad_norm": 3.5272546305254817,
+      "learning_rate": 4.4792311398178525e-09,
+      "loss": 0.8931,
+      "step": 10292
+    },
+    {
+      "epoch": 1.9814712322833699,
+      "grad_norm": 3.2049693464705302,
+      "learning_rate": 4.386406639287355e-09,
+      "loss": 0.8566,
+      "step": 10293
+    },
+    {
+      "epoch": 1.981663738960945,
+      "grad_norm": 3.0057599901725967,
+      "learning_rate": 4.2945538402794365e-09,
+      "loss": 0.7884,
+      "step": 10294
+    },
+    {
+      "epoch": 1.9818562456385207,
+      "grad_norm": 3.2166153382595417,
+      "learning_rate": 4.20367275172362e-09,
+      "loss": 0.8047,
+      "step": 10295
+    },
+    {
+      "epoch": 1.982048752316096,
+      "grad_norm": 3.2284488577300388,
+      "learning_rate": 4.11376338245395e-09,
+      "loss": 0.8246,
+      "step": 10296
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.8484,
+      "step": 10296,
+      "vm_loss": 0.1439
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.5269,
+      "step": 10296,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.8258,
+      "step": 10296,
+      "vm_loss": 0.1802
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.5883,
+      "step": 10296,
+      "vm_loss": 0.0965
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.9753,
+      "step": 10296,
+      "vm_loss": 0.1006
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.4661,
+      "step": 10296,
+      "vm_loss": 0.1736
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.7675,
+      "step": 10296,
+      "vm_loss": 0.1161
+    },
+    {
+      "epoch": 1.982048752316096,
+      "lm_loss": 0.9277,
+      "step": 10296,
+      "vm_loss": 0.115
+    },
+    {
+      "epoch": 1.9822412589936713,
+      "grad_norm": 3.2106334566274106,
+      "learning_rate": 4.024825741212324e-09,
+      "loss": 0.8116,
+      "step": 10297
+    },
+    {
+      "epoch": 1.9824337656712467,
+      "grad_norm": 3.226807181677353,
+      "learning_rate": 3.936859836641826e-09,
+      "loss": 0.8142,
+      "step": 10298
+    },
+    {
+      "epoch": 1.982626272348822,
+      "grad_norm": 3.131980566722135,
+      "learning_rate": 3.849865677296727e-09,
+      "loss": 0.803,
+      "step": 10299
+    },
+    {
+      "epoch": 1.9828187790263976,
+      "grad_norm": 3.332934969562565,
+      "learning_rate": 3.763843271631373e-09,
+      "loss": 0.8338,
+      "step": 10300
+    },
+    {
+      "epoch": 1.9830112857039728,
+      "grad_norm": 3.2837185466607246,
+      "learning_rate": 3.6787926280112963e-09,
+      "loss": 0.8536,
+      "step": 10301
+    },
+    {
+      "epoch": 1.9832037923815482,
+      "grad_norm": 3.1764902479740234,
+      "learning_rate": 3.594713754702106e-09,
+      "loss": 0.8222,
+      "step": 10302
+    },
+    {
+      "epoch": 1.9833962990591236,
+      "grad_norm": 3.155282462746852,
+      "learning_rate": 3.5116066598783747e-09,
+      "loss": 0.8304,
+      "step": 10303
+    },
+    {
+      "epoch": 1.983588805736699,
+      "grad_norm": 3.1679409558615603,
+      "learning_rate": 3.4294713516180854e-09,
+      "loss": 0.8436,
+      "step": 10304
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.5248,
+      "step": 10304,
+      "vm_loss": 0.1414
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.5858,
+      "step": 10304,
+      "vm_loss": 0.1583
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.5694,
+      "step": 10304,
+      "vm_loss": 0.1767
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.5427,
+      "step": 10304,
+      "vm_loss": 0.202
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.43,
+      "step": 10304,
+      "vm_loss": 0.1342
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.6024,
+      "step": 10304,
+      "vm_loss": 0.1446
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.6517,
+      "step": 10304,
+      "vm_loss": 0.1895
+    },
+    {
+      "epoch": 1.983588805736699,
+      "lm_loss": 0.4486,
+      "step": 10304,
+      "vm_loss": 0.1746
+    },
+    {
+      "epoch": 1.9837813124142745,
+      "grad_norm": 3.093616035277056,
+      "learning_rate": 3.348307837907072e-09,
+      "loss": 0.803,
+      "step": 10305
+    },
+    {
+      "epoch": 1.9839738190918497,
+      "grad_norm": 3.538443470696224,
+      "learning_rate": 3.2681161266356898e-09,
+      "loss": 0.9307,
+      "step": 10306
+    },
+    {
+      "epoch": 1.984166325769425,
+      "grad_norm": 3.2876091179167743,
+      "learning_rate": 3.1888962255988143e-09,
+      "loss": 0.8366,
+      "step": 10307
+    },
+    {
+      "epoch": 1.9843588324470005,
+      "grad_norm": 3.1854062953317075,
+      "learning_rate": 3.1106481424969524e-09,
+      "loss": 0.8386,
+      "step": 10308
+    },
+    {
+      "epoch": 1.984551339124576,
+      "grad_norm": 3.267671606002496,
+      "learning_rate": 3.033371884938463e-09,
+      "loss": 0.8041,
+      "step": 10309
+    },
+    {
+      "epoch": 1.9847438458021514,
+      "grad_norm": 3.408409421328903,
+      "learning_rate": 2.9570674604340046e-09,
+      "loss": 0.8645,
+      "step": 10310
+    },
+    {
+      "epoch": 1.9849363524797266,
+      "grad_norm": 3.1255932772739405,
+      "learning_rate": 2.881734876403197e-09,
+      "loss": 0.8077,
+      "step": 10311
+    },
+    {
+      "epoch": 1.985128859157302,
+      "grad_norm": 3.250726255756137,
+      "learning_rate": 2.807374140166852e-09,
+      "loss": 0.8137,
+      "step": 10312
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.9023,
+      "step": 10312,
+      "vm_loss": 0.1408
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.5499,
+      "step": 10312,
+      "vm_loss": 0.1842
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.6769,
+      "step": 10312,
+      "vm_loss": 0.1349
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 1.0874,
+      "step": 10312,
+      "vm_loss": 0.187
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.405,
+      "step": 10312,
+      "vm_loss": 0.1743
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.9951,
+      "step": 10312,
+      "vm_loss": 0.1758
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.537,
+      "step": 10312,
+      "vm_loss": 0.1585
+    },
+    {
+      "epoch": 1.985128859157302,
+      "lm_loss": 0.513,
+      "step": 10312,
+      "vm_loss": 0.1712
+    },
+    {
+      "epoch": 1.9853213658348774,
+      "grad_norm": 3.1625704609042575,
+      "learning_rate": 2.7339852589569616e-09,
+      "loss": 0.822,
+      "step": 10313
+    },
+    {
+      "epoch": 1.9855138725124528,
+      "grad_norm": 3.4382298580330244,
+      "learning_rate": 2.6615682399044883e-09,
+      "loss": 0.8786,
+      "step": 10314
+    },
+    {
+      "epoch": 1.9857063791900282,
+      "grad_norm": 3.2080342224694225,
+      "learning_rate": 2.5901230900515773e-09,
+      "loss": 0.8206,
+      "step": 10315
+    },
+    {
+      "epoch": 1.9858988858676034,
+      "grad_norm": 3.2615351810331346,
+      "learning_rate": 2.519649816342673e-09,
+      "loss": 0.8478,
+      "step": 10316
+    },
+    {
+      "epoch": 1.9860913925451789,
+      "grad_norm": 3.329770608724312,
+      "learning_rate": 2.450148425628962e-09,
+      "loss": 0.8179,
+      "step": 10317
+    },
+    {
+      "epoch": 1.9862838992227543,
+      "grad_norm": 3.316899234656234,
+      "learning_rate": 2.3816189246672616e-09,
+      "loss": 0.837,
+      "step": 10318
+    },
+    {
+      "epoch": 1.9864764059003297,
+      "grad_norm": 3.2460592904816212,
+      "learning_rate": 2.31406132011891e-09,
+      "loss": 0.8129,
+      "step": 10319
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "grad_norm": 3.3050644602280994,
+      "learning_rate": 2.247475618550876e-09,
+      "loss": 0.8333,
+      "step": 10320
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.4749,
+      "step": 10320,
+      "vm_loss": 0.1966
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 1.2338,
+      "step": 10320,
+      "vm_loss": 0.184
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.4694,
+      "step": 10320,
+      "vm_loss": 0.1417
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.7436,
+      "step": 10320,
+      "vm_loss": 0.1518
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.4122,
+      "step": 10320,
+      "vm_loss": 0.1415
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.4622,
+      "step": 10320,
+      "vm_loss": 0.2135
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.5574,
+      "step": 10320,
+      "vm_loss": 0.1617
+    },
+    {
+      "epoch": 1.9866689125779051,
+      "lm_loss": 0.6932,
+      "step": 10320,
+      "vm_loss": 0.1197
+    },
+    {
+      "epoch": 1.9868614192554803,
+      "grad_norm": 3.215475107716939,
+      "learning_rate": 2.181861826437981e-09,
+      "loss": 0.8053,
+      "step": 10321
+    },
+    {
+      "epoch": 1.987053925933056,
+      "grad_norm": 3.255744544623041,
+      "learning_rate": 2.1172199501573453e-09,
+      "loss": 0.8838,
+      "step": 10322
+    },
+    {
+      "epoch": 1.9872464326106312,
+      "grad_norm": 3.345130601404821,
+      "learning_rate": 2.0535499959917216e-09,
+      "loss": 0.8324,
+      "step": 10323
+    },
+    {
+      "epoch": 1.9874389392882066,
+      "grad_norm": 3.232323038434671,
+      "learning_rate": 1.9908519701339337e-09,
+      "loss": 0.8052,
+      "step": 10324
+    },
+    {
+      "epoch": 1.987631445965782,
+      "grad_norm": 3.2883399409649168,
+      "learning_rate": 1.929125878675775e-09,
+      "loss": 0.8242,
+      "step": 10325
+    },
+    {
+      "epoch": 1.9878239526433572,
+      "grad_norm": 3.2003250287087677,
+      "learning_rate": 1.868371727620222e-09,
+      "loss": 0.8589,
+      "step": 10326
+    },
+    {
+      "epoch": 1.9880164593209328,
+      "grad_norm": 3.1082088269988217,
+      "learning_rate": 1.8085895228725503e-09,
+      "loss": 0.8208,
+      "step": 10327
+    },
+    {
+      "epoch": 1.988208965998508,
+      "grad_norm": 3.4316102267662845,
+      "learning_rate": 1.7497792702436677e-09,
+      "loss": 0.8557,
+      "step": 10328
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.694,
+      "step": 10328,
+      "vm_loss": 0.1319
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.6742,
+      "step": 10328,
+      "vm_loss": 0.1884
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 1.1165,
+      "step": 10328,
+      "vm_loss": 0.2022
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.5041,
+      "step": 10328,
+      "vm_loss": 0.1798
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.6939,
+      "step": 10328,
+      "vm_loss": 0.1193
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.7195,
+      "step": 10328,
+      "vm_loss": 0.1222
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.942,
+      "step": 10328,
+      "vm_loss": 0.1289
+    },
+    {
+      "epoch": 1.988208965998508,
+      "lm_loss": 0.695,
+      "step": 10328,
+      "vm_loss": 0.1403
+    },
+    {
+      "epoch": 1.9884014726760835,
+      "grad_norm": 3.2722569953632914,
+      "learning_rate": 1.6919409754512228e-09,
+      "loss": 0.8024,
+      "step": 10329
+    },
+    {
+      "epoch": 1.988593979353659,
+      "grad_norm": 3.2623332848424864,
+      "learning_rate": 1.6350746441173848e-09,
+      "loss": 0.8224,
+      "step": 10330
+    },
+    {
+      "epoch": 1.988786486031234,
+      "grad_norm": 3.34550357385071,
+      "learning_rate": 1.5791802817721746e-09,
+      "loss": 0.8518,
+      "step": 10331
+    },
+    {
+      "epoch": 1.9889789927088097,
+      "grad_norm": 3.3973080147039934,
+      "learning_rate": 1.5242578938468034e-09,
+      "loss": 0.8436,
+      "step": 10332
+    },
+    {
+      "epoch": 1.989171499386385,
+      "grad_norm": 3.115532979957815,
+      "learning_rate": 1.4703074856814437e-09,
+      "loss": 0.801,
+      "step": 10333
+    },
+    {
+      "epoch": 1.9893640060639604,
+      "grad_norm": 3.3441941543828406,
+      "learning_rate": 1.4173290625207891e-09,
+      "loss": 0.8841,
+      "step": 10334
+    },
+    {
+      "epoch": 1.9895565127415358,
+      "grad_norm": 3.416234636749379,
+      "learning_rate": 1.3653226295151646e-09,
+      "loss": 0.8598,
+      "step": 10335
+    },
+    {
+      "epoch": 1.989749019419111,
+      "grad_norm": 3.207004231361259,
+      "learning_rate": 1.314288191719415e-09,
+      "loss": 0.8037,
+      "step": 10336
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.8757,
+      "step": 10336,
+      "vm_loss": 0.1926
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.5764,
+      "step": 10336,
+      "vm_loss": 0.2512
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.7766,
+      "step": 10336,
+      "vm_loss": 0.1338
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.4248,
+      "step": 10336,
+      "vm_loss": 0.2268
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.7605,
+      "step": 10336,
+      "vm_loss": 0.188
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.7669,
+      "step": 10336,
+      "vm_loss": 0.1361
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.5839,
+      "step": 10336,
+      "vm_loss": 0.1203
+    },
+    {
+      "epoch": 1.989749019419111,
+      "lm_loss": 0.427,
+      "step": 10336,
+      "vm_loss": 0.1357
+    },
+    {
+      "epoch": 1.9899415260966866,
+      "grad_norm": 3.244904998857969,
+      "learning_rate": 1.2642257540962377e-09,
+      "loss": 0.8464,
+      "step": 10337
+    },
+    {
+      "epoch": 1.9901340327742618,
+      "grad_norm": 3.0201925276628385,
+      "learning_rate": 1.21513532151063e-09,
+      "loss": 0.7615,
+      "step": 10338
+    },
+    {
+      "epoch": 1.9903265394518372,
+      "grad_norm": 3.3374943117759592,
+      "learning_rate": 1.1670168987365505e-09,
+      "loss": 0.8516,
+      "step": 10339
+    },
+    {
+      "epoch": 1.9905190461294127,
+      "grad_norm": 3.3054473535204343,
+      "learning_rate": 1.119870490450259e-09,
+      "loss": 0.8352,
+      "step": 10340
+    },
+    {
+      "epoch": 1.9907115528069879,
+      "grad_norm": 3.3707054368214338,
+      "learning_rate": 1.073696101235866e-09,
+      "loss": 0.8574,
+      "step": 10341
+    },
+    {
+      "epoch": 1.9909040594845635,
+      "grad_norm": 3.1316717434749033,
+      "learning_rate": 1.0284937355820035e-09,
+      "loss": 0.8127,
+      "step": 10342
+    },
+    {
+      "epoch": 1.9910965661621387,
+      "grad_norm": 3.1888661278329353,
+      "learning_rate": 9.842633978818238e-10,
+      "loss": 0.8332,
+      "step": 10343
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "grad_norm": 3.1075185298567742,
+      "learning_rate": 9.410050924374414e-10,
+      "loss": 0.7789,
+      "step": 10344
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.906,
+      "step": 10344,
+      "vm_loss": 0.1824
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.8038,
+      "step": 10344,
+      "vm_loss": 0.2126
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.7587,
+      "step": 10344,
+      "vm_loss": 0.1908
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.49,
+      "step": 10344,
+      "vm_loss": 0.1608
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.7159,
+      "step": 10344,
+      "vm_loss": 0.1152
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.794,
+      "step": 10344,
+      "vm_loss": 0.182
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.6915,
+      "step": 10344,
+      "vm_loss": 0.1855
+    },
+    {
+      "epoch": 1.9912890728397141,
+      "lm_loss": 0.5468,
+      "step": 10344,
+      "vm_loss": 0.149
+    },
+    {
+      "epoch": 1.9914815795172895,
+      "grad_norm": 3.2954451576893073,
+      "learning_rate": 8.987188234521604e-10,
+      "loss": 0.8623,
+      "step": 10345
+    },
+    {
+      "epoch": 1.9916740861948647,
+      "grad_norm": 3.2869728648664127,
+      "learning_rate": 8.574045950360266e-10,
+      "loss": 0.8209,
+      "step": 10346
+    },
+    {
+      "epoch": 1.9918665928724404,
+      "grad_norm": 3.239888862869303,
+      "learning_rate": 8.170624112080472e-10,
+      "loss": 0.8102,
+      "step": 10347
+    },
+    {
+      "epoch": 1.9920590995500156,
+      "grad_norm": 3.296030330432085,
+      "learning_rate": 7.776922758884198e-10,
+      "loss": 0.8358,
+      "step": 10348
+    },
+    {
+      "epoch": 1.992251606227591,
+      "grad_norm": 3.1339022347341197,
+      "learning_rate": 7.392941929040832e-10,
+      "loss": 0.8165,
+      "step": 10349
+    },
+    {
+      "epoch": 1.9924441129051664,
+      "grad_norm": 3.093440823319644,
+      "learning_rate": 7.01868165987607e-10,
+      "loss": 0.7859,
+      "step": 10350
+    },
+    {
+      "epoch": 1.9926366195827416,
+      "grad_norm": 3.231619427735801,
+      "learning_rate": 6.654141987794127e-10,
+      "loss": 0.853,
+      "step": 10351
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "grad_norm": 3.2424976031574197,
+      "learning_rate": 6.299322948211118e-10,
+      "loss": 0.8174,
+      "step": 10352
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.1905,
+      "step": 10352,
+      "vm_loss": 0.1723
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.5194,
+      "step": 10352,
+      "vm_loss": 0.1489
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.6333,
+      "step": 10352,
+      "vm_loss": 0.2156
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.6631,
+      "step": 10352,
+      "vm_loss": 0.2154
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.5846,
+      "step": 10352,
+      "vm_loss": 0.212
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 1.0106,
+      "step": 10352,
+      "vm_loss": 0.1682
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.3546,
+      "step": 10352,
+      "vm_loss": 0.2137
+    },
+    {
+      "epoch": 1.9928291262603173,
+      "lm_loss": 0.5083,
+      "step": 10352,
+      "vm_loss": 0.2589
+    },
+    {
+      "epoch": 1.9930216329378925,
+      "grad_norm": 3.1591245021862346,
+      "learning_rate": 5.95422457562167e-10,
+      "loss": 0.8265,
+      "step": 10353
+    },
+    {
+      "epoch": 1.993214139615468,
+      "grad_norm": 3.2729495364092145,
+      "learning_rate": 5.618846903576724e-10,
+      "loss": 0.8085,
+      "step": 10354
+    },
+    {
+      "epoch": 1.9934066462930433,
+      "grad_norm": 3.2095776619774266,
+      "learning_rate": 5.293189964694634e-10,
+      "loss": 0.8066,
+      "step": 10355
+    },
+    {
+      "epoch": 1.9935991529706185,
+      "grad_norm": 3.220226197953992,
+      "learning_rate": 4.977253790605652e-10,
+      "loss": 0.7952,
+      "step": 10356
+    },
+    {
+      "epoch": 1.9937916596481942,
+      "grad_norm": 3.132456075203666,
+      "learning_rate": 4.671038412051854e-10,
+      "loss": 0.8407,
+      "step": 10357
+    },
+    {
+      "epoch": 1.9939841663257694,
+      "grad_norm": 3.2305105672869043,
+      "learning_rate": 4.3745438587761145e-10,
+      "loss": 0.8255,
+      "step": 10358
+    },
+    {
+      "epoch": 1.9941766730033448,
+      "grad_norm": 3.2338921288859304,
+      "learning_rate": 4.0877701596220285e-10,
+      "loss": 0.7927,
+      "step": 10359
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "grad_norm": 3.3457167140726805,
+      "learning_rate": 3.8107173424450915e-10,
+      "loss": 0.8743,
+      "step": 10360
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.7375,
+      "step": 10360,
+      "vm_loss": 0.1969
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.7251,
+      "step": 10360,
+      "vm_loss": 0.1206
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.6318,
+      "step": 10360,
+      "vm_loss": 0.1792
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.7596,
+      "step": 10360,
+      "vm_loss": 0.1662
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.4391,
+      "step": 10360,
+      "vm_loss": 0.1912
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 1.1354,
+      "step": 10360,
+      "vm_loss": 0.192
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.8815,
+      "step": 10360,
+      "vm_loss": 0.1123
+    },
+    {
+      "epoch": 1.9943691796809202,
+      "lm_loss": 0.6504,
+      "step": 10360,
+      "vm_loss": 0.2254
+    },
+    {
+      "epoch": 1.9945616863584954,
+      "grad_norm": 3.4593600694948745,
+      "learning_rate": 3.5433854342015185e-10,
+      "loss": 0.8727,
+      "step": 10361
+    },
+    {
+      "epoch": 1.994754193036071,
+      "grad_norm": 3.1775076621605325,
+      "learning_rate": 3.285774460870528e-10,
+      "loss": 0.8433,
+      "step": 10362
+    },
+    {
+      "epoch": 1.9949466997136462,
+      "grad_norm": 3.1351824845445986,
+      "learning_rate": 3.037884447487649e-10,
+      "loss": 0.822,
+      "step": 10363
+    },
+    {
+      "epoch": 1.9951392063912217,
+      "grad_norm": 3.2514329894600706,
+      "learning_rate": 2.799715418155824e-10,
+      "loss": 0.8202,
+      "step": 10364
+    },
+    {
+      "epoch": 1.995331713068797,
+      "grad_norm": 3.3190844000511444,
+      "learning_rate": 2.571267396034305e-10,
+      "loss": 0.8306,
+      "step": 10365
+    },
+    {
+      "epoch": 1.9955242197463725,
+      "grad_norm": 3.3019548448958655,
+      "learning_rate": 2.352540403327552e-10,
+      "loss": 0.8587,
+      "step": 10366
+    },
+    {
+      "epoch": 1.995716726423948,
+      "grad_norm": 3.27279222339282,
+      "learning_rate": 2.143534461296337e-10,
+      "loss": 0.8579,
+      "step": 10367
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "grad_norm": 3.2293021714166197,
+      "learning_rate": 1.9442495902577408e-10,
+      "loss": 0.8297,
+      "step": 10368
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.4368,
+      "step": 10368,
+      "vm_loss": 0.1563
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.6769,
+      "step": 10368,
+      "vm_loss": 0.2091
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.473,
+      "step": 10368,
+      "vm_loss": 0.1902
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.7182,
+      "step": 10368,
+      "vm_loss": 0.1783
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.743,
+      "step": 10368,
+      "vm_loss": 0.2011
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.7095,
+      "step": 10368,
+      "vm_loss": 0.1719
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.8066,
+      "step": 10368,
+      "vm_loss": 0.1614
+    },
+    {
+      "epoch": 1.9959092331015231,
+      "lm_loss": 0.9259,
+      "step": 10368,
+      "vm_loss": 0.1939
+    },
+    {
+      "epoch": 1.9961017397790985,
+      "grad_norm": 3.3177833614222325,
+      "learning_rate": 1.754685809585155e-10,
+      "loss": 0.8713,
+      "step": 10369
+    },
+    {
+      "epoch": 1.996294246456674,
+      "grad_norm": 3.0334616208839367,
+      "learning_rate": 1.5748431377082817e-10,
+      "loss": 0.7873,
+      "step": 10370
+    },
+    {
+      "epoch": 1.9964867531342494,
+      "grad_norm": 3.0948606006893016,
+      "learning_rate": 1.4047215921131341e-10,
+      "loss": 0.7787,
+      "step": 10371
+    },
+    {
+      "epoch": 1.9966792598118248,
+      "grad_norm": 3.2382819962714517,
+      "learning_rate": 1.2443211893309326e-10,
+      "loss": 0.843,
+      "step": 10372
+    },
+    {
+      "epoch": 1.9968717664894,
+      "grad_norm": 3.201891769037206,
+      "learning_rate": 1.0936419449603108e-10,
+      "loss": 0.8204,
+      "step": 10373
+    },
+    {
+      "epoch": 1.9970642731669754,
+      "grad_norm": 3.2538395591913214,
+      "learning_rate": 9.526838736451105e-11,
+      "loss": 0.8629,
+      "step": 10374
+    },
+    {
+      "epoch": 1.9972567798445509,
+      "grad_norm": 3.2524755810004797,
+      "learning_rate": 8.21446989096586e-11,
+      "loss": 0.8652,
+      "step": 10375
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "grad_norm": 3.4288566241263627,
+      "learning_rate": 6.999313040600974e-11,
+      "loss": 0.8887,
+      "step": 10376
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.5095,
+      "step": 10376,
+      "vm_loss": 0.153
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.9305,
+      "step": 10376,
+      "vm_loss": 0.1054
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.9219,
+      "step": 10376,
+      "vm_loss": 0.1622
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.4402,
+      "step": 10376,
+      "vm_loss": 0.1718
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.8247,
+      "step": 10376,
+      "vm_loss": 0.1454
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.6165,
+      "step": 10376,
+      "vm_loss": 0.1528
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.7405,
+      "step": 10376,
+      "vm_loss": 0.1221
+    },
+    {
+      "epoch": 1.9974492865221263,
+      "lm_loss": 0.7824,
+      "step": 10376,
+      "vm_loss": 0.1339
+    },
+    {
+      "epoch": 1.9976417931997017,
+      "grad_norm": 3.247147272467522,
+      "learning_rate": 5.881368303595202e-11,
+      "loss": 0.8193,
+      "step": 10377
+    },
+    {
+      "epoch": 1.997834299877277,
+      "grad_norm": 3.387983436094678,
+      "learning_rate": 4.860635788528356e-11,
+      "loss": 0.8887,
+      "step": 10378
+    },
+    {
+      "epoch": 1.9980268065548523,
+      "grad_norm": 3.2076775432881415,
+      "learning_rate": 3.937115594654373e-11,
+      "loss": 0.8144,
+      "step": 10379
+    },
+    {
+      "epoch": 1.9982193132324277,
+      "grad_norm": 3.328374671669054,
+      "learning_rate": 3.1108078119013175e-11,
+      "loss": 0.835,
+      "step": 10380
+    },
+    {
+      "epoch": 1.9984118199100032,
+      "grad_norm": 3.260221242139148,
+      "learning_rate": 2.3817125204272927e-11,
+      "loss": 0.8405,
+      "step": 10381
+    },
+    {
+      "epoch": 1.9986043265875786,
+      "grad_norm": 3.2192810824075053,
+      "learning_rate": 1.749829791064528e-11,
+      "loss": 0.8321,
+      "step": 10382
+    },
+    {
+      "epoch": 1.9987968332651538,
+      "grad_norm": 3.335088050030916,
+      "learning_rate": 1.2151596854304004e-11,
+      "loss": 0.8534,
+      "step": 10383
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "grad_norm": 3.280067247254368,
+      "learning_rate": 7.777022553723258e-12,
+      "loss": 0.8263,
+      "step": 10384
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.4998,
+      "step": 10384,
+      "vm_loss": 0.1173
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.517,
+      "step": 10384,
+      "vm_loss": 0.2179
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.8516,
+      "step": 10384,
+      "vm_loss": 0.1947
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.5897,
+      "step": 10384,
+      "vm_loss": 0.1704
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.8321,
+      "step": 10384,
+      "vm_loss": 0.1231
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.5931,
+      "step": 10384,
+      "vm_loss": 0.1159
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.9918,
+      "step": 10384,
+      "vm_loss": 0.1984
+    },
+    {
+      "epoch": 1.9989893399427294,
+      "lm_loss": 0.6472,
+      "step": 10384,
+      "vm_loss": 0.1121
+    },
+    {
+      "epoch": 1.9991818466203046,
+      "grad_norm": 3.4572415083524772,
+      "learning_rate": 4.374575434118455e-12,
+      "loss": 0.8717,
+      "step": 10385
+    },
+    {
+      "epoch": 1.99937435329788,
+      "grad_norm": 3.3119665098252544,
+      "learning_rate": 1.944255827446284e-12,
+      "loss": 0.8518,
+      "step": 10386
+    },
+    {
+      "epoch": 1.9995668599754555,
+      "grad_norm": 3.1088014698276574,
+      "learning_rate": 4.860639690740243e-13,
+      "loss": 0.8285,
+      "step": 10387
+    },
+    {
+      "epoch": 1.9997593666530307,
+      "grad_norm": 3.255229368940458,
+      "learning_rate": 0.0,
+      "loss": 0.8638,
+      "step": 10388
+    },
+    {
+      "epoch": 1.9997593666530307,
+      "step": 10388,
+      "total_flos": 1.3983801185697792e+16,
+      "train_loss": 1.6275318770579388,
+      "train_runtime": 117073.7373,
+      "train_samples_per_second": 34.076,
+      "train_steps_per_second": 0.089
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 10388,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3983801185697792e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}