diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22575 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3219,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009319664492078285,
+      "grad_norm": 56.782817702956166,
+      "learning_rate": 1.5527950310559006e-07,
+      "loss": 11.0588,
+      "step": 1
+    },
+    {
+      "epoch": 0.001863932898415657,
+      "grad_norm": 54.313638739396104,
+      "learning_rate": 3.1055900621118013e-07,
+      "loss": 11.1374,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027958993476234857,
+      "grad_norm": 54.02915482135048,
+      "learning_rate": 4.6583850931677024e-07,
+      "loss": 11.1899,
+      "step": 3
+    },
+    {
+      "epoch": 0.003727865796831314,
+      "grad_norm": 56.2224030268105,
+      "learning_rate": 6.211180124223603e-07,
+      "loss": 11.0111,
+      "step": 4
+    },
+    {
+      "epoch": 0.004659832246039142,
+      "grad_norm": 55.68282767449908,
+      "learning_rate": 7.763975155279503e-07,
+      "loss": 10.97,
+      "step": 5
+    },
+    {
+      "epoch": 0.005591798695246971,
+      "grad_norm": 53.38475593505162,
+      "learning_rate": 9.316770186335405e-07,
+      "loss": 11.0165,
+      "step": 6
+    },
+    {
+      "epoch": 0.0065237651444548,
+      "grad_norm": 56.39526603365928,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 10.9046,
+      "step": 7
+    },
+    {
+      "epoch": 0.007455731593662628,
+      "grad_norm": 56.75618348055788,
+      "learning_rate": 1.2422360248447205e-06,
+      "loss": 11.0035,
+      "step": 8
+    },
+    {
+      "epoch": 0.008387698042870456,
+      "grad_norm": 58.46472303423346,
+      "learning_rate": 1.3975155279503107e-06,
+      "loss": 10.7486,
+      "step": 9
+    },
+    {
+      "epoch": 0.009319664492078284,
+      "grad_norm": 56.93782000017916,
+      "learning_rate": 1.5527950310559006e-06,
+      "loss": 10.7696,
+      "step": 10
+    },
+    {
+      "epoch": 0.010251630941286114,
+      "grad_norm": 60.06048534334742,
+      "learning_rate": 1.7080745341614908e-06,
+      "loss": 10.6603,
+      "step": 11
+    },
+    {
+      "epoch": 0.011183597390493943,
+      "grad_norm": 79.09053607138628,
+      "learning_rate": 1.863354037267081e-06,
+      "loss": 9.5003,
+      "step": 12
+    },
+    {
+      "epoch": 0.012115563839701771,
+      "grad_norm": 84.70989871084564,
+      "learning_rate": 2.018633540372671e-06,
+      "loss": 9.4497,
+      "step": 13
+    },
+    {
+      "epoch": 0.0130475302889096,
+      "grad_norm": 92.29755794929751,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 8.9454,
+      "step": 14
+    },
+    {
+      "epoch": 0.013979496738117428,
+      "grad_norm": 97.81172060488059,
+      "learning_rate": 2.329192546583851e-06,
+      "loss": 8.6873,
+      "step": 15
+    },
+    {
+      "epoch": 0.014911463187325256,
+      "grad_norm": 69.69171316190302,
+      "learning_rate": 2.484472049689441e-06,
+      "loss": 3.9914,
+      "step": 16
+    },
+    {
+      "epoch": 0.015843429636533086,
+      "grad_norm": 61.17688627396157,
+      "learning_rate": 2.639751552795031e-06,
+      "loss": 3.5445,
+      "step": 17
+    },
+    {
+      "epoch": 0.016775396085740912,
+      "grad_norm": 54.18943404505982,
+      "learning_rate": 2.7950310559006214e-06,
+      "loss": 3.3164,
+      "step": 18
+    },
+    {
+      "epoch": 0.017707362534948742,
+      "grad_norm": 41.73596926176503,
+      "learning_rate": 2.9503105590062115e-06,
+      "loss": 2.8314,
+      "step": 19
+    },
+    {
+      "epoch": 0.01863932898415657,
+      "grad_norm": 32.82249188088036,
+      "learning_rate": 3.1055900621118013e-06,
+      "loss": 2.4252,
+      "step": 20
+    },
+    {
+      "epoch": 0.0195712954333644,
+      "grad_norm": 13.460109517299115,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 1.7281,
+      "step": 21
+    },
+    {
+      "epoch": 0.02050326188257223,
+      "grad_norm": 6.115110601432725,
+      "learning_rate": 3.4161490683229816e-06,
+      "loss": 1.325,
+      "step": 22
+    },
+    {
+      "epoch": 0.021435228331780055,
+      "grad_norm": 4.827090792567416,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 1.2582,
+      "step": 23
+    },
+    {
+      "epoch": 0.022367194780987885,
+      "grad_norm": 4.064797481158908,
+      "learning_rate": 3.726708074534162e-06,
+      "loss": 1.2673,
+      "step": 24
+    },
+    {
+      "epoch": 0.023299161230195712,
+      "grad_norm": 3.350275889919475,
+      "learning_rate": 3.881987577639752e-06,
+      "loss": 1.1622,
+      "step": 25
+    },
+    {
+      "epoch": 0.024231127679403542,
+      "grad_norm": 2.5595681546646594,
+      "learning_rate": 4.037267080745342e-06,
+      "loss": 1.1147,
+      "step": 26
+    },
+    {
+      "epoch": 0.02516309412861137,
+      "grad_norm": 2.7247234151675137,
+      "learning_rate": 4.192546583850932e-06,
+      "loss": 1.0608,
+      "step": 27
+    },
+    {
+      "epoch": 0.0260950605778192,
+      "grad_norm": 1.8588536547009693,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.014,
+      "step": 28
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 1.3775842809659435,
+      "learning_rate": 4.503105590062112e-06,
+      "loss": 0.943,
+      "step": 29
+    },
+    {
+      "epoch": 0.027958993476234855,
+      "grad_norm": 1.3097870189996799,
+      "learning_rate": 4.658385093167702e-06,
+      "loss": 0.8992,
+      "step": 30
+    },
+    {
+      "epoch": 0.028890959925442685,
+      "grad_norm": 4.857292804012658,
+      "learning_rate": 4.813664596273292e-06,
+      "loss": 0.8815,
+      "step": 31
+    },
+    {
+      "epoch": 0.02982292637465051,
+      "grad_norm": 11.426041261113888,
+      "learning_rate": 4.968944099378882e-06,
+      "loss": 0.8606,
+      "step": 32
+    },
+    {
+      "epoch": 0.03075489282385834,
+      "grad_norm": 1.1876901712231354,
+      "learning_rate": 5.124223602484472e-06,
+      "loss": 0.8131,
+      "step": 33
+    },
+    {
+      "epoch": 0.03168685927306617,
+      "grad_norm": 0.9309829596450938,
+      "learning_rate": 5.279503105590062e-06,
+      "loss": 0.8069,
+      "step": 34
+    },
+    {
+      "epoch": 0.032618825722273995,
+      "grad_norm": 0.8740582428258041,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 0.8448,
+      "step": 35
+    },
+    {
+      "epoch": 0.033550792171481825,
+      "grad_norm": 0.7557767754902358,
+      "learning_rate": 5.590062111801243e-06,
+      "loss": 0.786,
+      "step": 36
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 0.6957810273744609,
+      "learning_rate": 5.745341614906832e-06,
+      "loss": 0.7813,
+      "step": 37
+    },
+    {
+      "epoch": 0.035414725069897485,
+      "grad_norm": 0.6849643983485404,
+      "learning_rate": 5.900621118012423e-06,
+      "loss": 0.7913,
+      "step": 38
+    },
+    {
+      "epoch": 0.036346691519105315,
+      "grad_norm": 0.6343582678511308,
+      "learning_rate": 6.055900621118013e-06,
+      "loss": 0.7557,
+      "step": 39
+    },
+    {
+      "epoch": 0.03727865796831314,
+      "grad_norm": 0.619685979803318,
+      "learning_rate": 6.2111801242236025e-06,
+      "loss": 0.7649,
+      "step": 40
+    },
+    {
+      "epoch": 0.03821062441752097,
+      "grad_norm": 0.623698805434108,
+      "learning_rate": 6.366459627329193e-06,
+      "loss": 0.6954,
+      "step": 41
+    },
+    {
+      "epoch": 0.0391425908667288,
+      "grad_norm": 0.6306624079643239,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.7624,
+      "step": 42
+    },
+    {
+      "epoch": 0.04007455731593663,
+      "grad_norm": 0.5616624309668669,
+      "learning_rate": 6.677018633540373e-06,
+      "loss": 0.7143,
+      "step": 43
+    },
+    {
+      "epoch": 0.04100652376514446,
+      "grad_norm": 0.48501884927815664,
+      "learning_rate": 6.832298136645963e-06,
+      "loss": 0.6711,
+      "step": 44
+    },
+    {
+      "epoch": 0.04193849021435228,
+      "grad_norm": 0.46230163216104414,
+      "learning_rate": 6.9875776397515525e-06,
+      "loss": 0.726,
+      "step": 45
+    },
+    {
+      "epoch": 0.04287045666356011,
+      "grad_norm": 0.47040311397608947,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.6891,
+      "step": 46
+    },
+    {
+      "epoch": 0.04380242311276794,
+      "grad_norm": 0.5305032792025677,
+      "learning_rate": 7.298136645962733e-06,
+      "loss": 0.7262,
+      "step": 47
+    },
+    {
+      "epoch": 0.04473438956197577,
+      "grad_norm": 0.5064161808628433,
+      "learning_rate": 7.453416149068324e-06,
+      "loss": 0.6941,
+      "step": 48
+    },
+    {
+      "epoch": 0.045666356011183594,
+      "grad_norm": 0.4382007423794495,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 0.681,
+      "step": 49
+    },
+    {
+      "epoch": 0.046598322460391424,
+      "grad_norm": 0.42890162703364376,
+      "learning_rate": 7.763975155279503e-06,
+      "loss": 0.6253,
+      "step": 50
+    },
+    {
+      "epoch": 0.047530288909599254,
+      "grad_norm": 0.41420595319380726,
+      "learning_rate": 7.919254658385093e-06,
+      "loss": 0.6241,
+      "step": 51
+    },
+    {
+      "epoch": 0.048462255358807084,
+      "grad_norm": 0.4344669766796968,
+      "learning_rate": 8.074534161490684e-06,
+      "loss": 0.644,
+      "step": 52
+    },
+    {
+      "epoch": 0.049394221808014914,
+      "grad_norm": 0.46350077606727724,
+      "learning_rate": 8.229813664596275e-06,
+      "loss": 0.6488,
+      "step": 53
+    },
+    {
+      "epoch": 0.05032618825722274,
+      "grad_norm": 0.4090538327546372,
+      "learning_rate": 8.385093167701864e-06,
+      "loss": 0.6256,
+      "step": 54
+    },
+    {
+      "epoch": 0.05125815470643057,
+      "grad_norm": 0.43774002852351135,
+      "learning_rate": 8.540372670807453e-06,
+      "loss": 0.6876,
+      "step": 55
+    },
+    {
+      "epoch": 0.0521901211556384,
+      "grad_norm": 0.35885600371264625,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.6057,
+      "step": 56
+    },
+    {
+      "epoch": 0.05312208760484623,
+      "grad_norm": 0.35053935201488784,
+      "learning_rate": 8.850931677018634e-06,
+      "loss": 0.615,
+      "step": 57
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 0.3644459343565203,
+      "learning_rate": 9.006211180124225e-06,
+      "loss": 0.6334,
+      "step": 58
+    },
+    {
+      "epoch": 0.05498602050326188,
+      "grad_norm": 0.36756477597175446,
+      "learning_rate": 9.161490683229814e-06,
+      "loss": 0.5836,
+      "step": 59
+    },
+    {
+      "epoch": 0.05591798695246971,
+      "grad_norm": 0.3130465423770163,
+      "learning_rate": 9.316770186335403e-06,
+      "loss": 0.6065,
+      "step": 60
+    },
+    {
+      "epoch": 0.05684995340167754,
+      "grad_norm": 0.3670214110186132,
+      "learning_rate": 9.472049689440994e-06,
+      "loss": 0.6214,
+      "step": 61
+    },
+    {
+      "epoch": 0.05778191985088537,
+      "grad_norm": 0.33097073639667707,
+      "learning_rate": 9.627329192546584e-06,
+      "loss": 0.6097,
+      "step": 62
+    },
+    {
+      "epoch": 0.05871388630009319,
+      "grad_norm": 0.31732488024782407,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 0.5778,
+      "step": 63
+    },
+    {
+      "epoch": 0.05964585274930102,
+      "grad_norm": 0.36289126713344994,
+      "learning_rate": 9.937888198757764e-06,
+      "loss": 0.6109,
+      "step": 64
+    },
+    {
+      "epoch": 0.06057781919850885,
+      "grad_norm": 0.3001279444105541,
+      "learning_rate": 1.0093167701863353e-05,
+      "loss": 0.5464,
+      "step": 65
+    },
+    {
+      "epoch": 0.06150978564771668,
+      "grad_norm": 0.3543269620782488,
+      "learning_rate": 1.0248447204968944e-05,
+      "loss": 0.6094,
+      "step": 66
+    },
+    {
+      "epoch": 0.06244175209692451,
+      "grad_norm": 0.31691076182907424,
+      "learning_rate": 1.0403726708074535e-05,
+      "loss": 0.6044,
+      "step": 67
+    },
+    {
+      "epoch": 0.06337371854613234,
+      "grad_norm": 0.28589438659406213,
+      "learning_rate": 1.0559006211180125e-05,
+      "loss": 0.577,
+      "step": 68
+    },
+    {
+      "epoch": 0.06430568499534017,
+      "grad_norm": 0.30819050984869734,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.5742,
+      "step": 69
+    },
+    {
+      "epoch": 0.06523765144454799,
+      "grad_norm": 0.3494484299546868,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.5809,
+      "step": 70
+    },
+    {
+      "epoch": 0.06616961789375582,
+      "grad_norm": 0.2891860244618301,
+      "learning_rate": 1.1024844720496894e-05,
+      "loss": 0.5796,
+      "step": 71
+    },
+    {
+      "epoch": 0.06710158434296365,
+      "grad_norm": 0.2854129708849679,
+      "learning_rate": 1.1180124223602485e-05,
+      "loss": 0.589,
+      "step": 72
+    },
+    {
+      "epoch": 0.06803355079217148,
+      "grad_norm": 0.2876713419094048,
+      "learning_rate": 1.1335403726708076e-05,
+      "loss": 0.5651,
+      "step": 73
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 0.2976624093926452,
+      "learning_rate": 1.1490683229813664e-05,
+      "loss": 0.5544,
+      "step": 74
+    },
+    {
+      "epoch": 0.06989748369058714,
+      "grad_norm": 0.3036505084564572,
+      "learning_rate": 1.1645962732919255e-05,
+      "loss": 0.5684,
+      "step": 75
+    },
+    {
+      "epoch": 0.07082945013979497,
+      "grad_norm": 0.28476219240931244,
+      "learning_rate": 1.1801242236024846e-05,
+      "loss": 0.5822,
+      "step": 76
+    },
+    {
+      "epoch": 0.0717614165890028,
+      "grad_norm": 0.29680579402781254,
+      "learning_rate": 1.1956521739130435e-05,
+      "loss": 0.583,
+      "step": 77
+    },
+    {
+      "epoch": 0.07269338303821063,
+      "grad_norm": 0.29560373141046103,
+      "learning_rate": 1.2111801242236026e-05,
+      "loss": 0.5419,
+      "step": 78
+    },
+    {
+      "epoch": 0.07362534948741846,
+      "grad_norm": 0.3321813454653801,
+      "learning_rate": 1.2267080745341616e-05,
+      "loss": 0.5319,
+      "step": 79
+    },
+    {
+      "epoch": 0.07455731593662628,
+      "grad_norm": 0.3078002580068892,
+      "learning_rate": 1.2422360248447205e-05,
+      "loss": 0.57,
+      "step": 80
+    },
+    {
+      "epoch": 0.0754892823858341,
+      "grad_norm": 0.27194328080533114,
+      "learning_rate": 1.2577639751552794e-05,
+      "loss": 0.5663,
+      "step": 81
+    },
+    {
+      "epoch": 0.07642124883504194,
+      "grad_norm": 0.32164899029089966,
+      "learning_rate": 1.2732919254658385e-05,
+      "loss": 0.5415,
+      "step": 82
+    },
+    {
+      "epoch": 0.07735321528424977,
+      "grad_norm": 0.3057213615086141,
+      "learning_rate": 1.2888198757763975e-05,
+      "loss": 0.5449,
+      "step": 83
+    },
+    {
+      "epoch": 0.0782851817334576,
+      "grad_norm": 0.29338478893109704,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.5546,
+      "step": 84
+    },
+    {
+      "epoch": 0.07921714818266543,
+      "grad_norm": 0.3281374446812013,
+      "learning_rate": 1.3198757763975155e-05,
+      "loss": 0.5548,
+      "step": 85
+    },
+    {
+      "epoch": 0.08014911463187326,
+      "grad_norm": 0.32045641232729927,
+      "learning_rate": 1.3354037267080746e-05,
+      "loss": 0.568,
+      "step": 86
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 0.291872640816138,
+      "learning_rate": 1.3509316770186337e-05,
+      "loss": 0.5695,
+      "step": 87
+    },
+    {
+      "epoch": 0.08201304753028892,
+      "grad_norm": 0.3037892605049127,
+      "learning_rate": 1.3664596273291926e-05,
+      "loss": 0.5751,
+      "step": 88
+    },
+    {
+      "epoch": 0.08294501397949673,
+      "grad_norm": 0.3000454307004752,
+      "learning_rate": 1.3819875776397517e-05,
+      "loss": 0.5883,
+      "step": 89
+    },
+    {
+      "epoch": 0.08387698042870456,
+      "grad_norm": 0.288973022792204,
+      "learning_rate": 1.3975155279503105e-05,
+      "loss": 0.5129,
+      "step": 90
+    },
+    {
+      "epoch": 0.08480894687791239,
+      "grad_norm": 0.2718002058856952,
+      "learning_rate": 1.4130434782608694e-05,
+      "loss": 0.5583,
+      "step": 91
+    },
+    {
+      "epoch": 0.08574091332712022,
+      "grad_norm": 0.28814297879066764,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.5443,
+      "step": 92
+    },
+    {
+      "epoch": 0.08667287977632805,
+      "grad_norm": 0.2822911269407923,
+      "learning_rate": 1.4440993788819876e-05,
+      "loss": 0.5438,
+      "step": 93
+    },
+    {
+      "epoch": 0.08760484622553588,
+      "grad_norm": 0.28086630275429925,
+      "learning_rate": 1.4596273291925466e-05,
+      "loss": 0.5407,
+      "step": 94
+    },
+    {
+      "epoch": 0.08853681267474371,
+      "grad_norm": 0.2592225499500161,
+      "learning_rate": 1.4751552795031057e-05,
+      "loss": 0.5337,
+      "step": 95
+    },
+    {
+      "epoch": 0.08946877912395154,
+      "grad_norm": 0.28618034231021516,
+      "learning_rate": 1.4906832298136648e-05,
+      "loss": 0.5574,
+      "step": 96
+    },
+    {
+      "epoch": 0.09040074557315937,
+      "grad_norm": 0.27367618714556696,
+      "learning_rate": 1.5062111801242237e-05,
+      "loss": 0.4946,
+      "step": 97
+    },
+    {
+      "epoch": 0.09133271202236719,
+      "grad_norm": 0.2762528221668063,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.5115,
+      "step": 98
+    },
+    {
+      "epoch": 0.09226467847157502,
+      "grad_norm": 0.29275445199591904,
+      "learning_rate": 1.537267080745342e-05,
+      "loss": 0.547,
+      "step": 99
+    },
+    {
+      "epoch": 0.09319664492078285,
+      "grad_norm": 0.32211651447612266,
+      "learning_rate": 1.5527950310559007e-05,
+      "loss": 0.5371,
+      "step": 100
+    },
+    {
+      "epoch": 0.09412861136999068,
+      "grad_norm": 0.28790085109170666,
+      "learning_rate": 1.5683229813664594e-05,
+      "loss": 0.5299,
+      "step": 101
+    },
+    {
+      "epoch": 0.09506057781919851,
+      "grad_norm": 0.30525820786764607,
+      "learning_rate": 1.5838509316770185e-05,
+      "loss": 0.518,
+      "step": 102
+    },
+    {
+      "epoch": 0.09599254426840634,
+      "grad_norm": 0.30374176538567466,
+      "learning_rate": 1.5993788819875776e-05,
+      "loss": 0.5396,
+      "step": 103
+    },
+    {
+      "epoch": 0.09692451071761417,
+      "grad_norm": 0.28871205333751715,
+      "learning_rate": 1.6149068322981367e-05,
+      "loss": 0.5283,
+      "step": 104
+    },
+    {
+      "epoch": 0.097856477166822,
+      "grad_norm": 0.30181784768994835,
+      "learning_rate": 1.630434782608696e-05,
+      "loss": 0.5054,
+      "step": 105
+    },
+    {
+      "epoch": 0.09878844361602983,
+      "grad_norm": 0.28079178884677175,
+      "learning_rate": 1.645962732919255e-05,
+      "loss": 0.5236,
+      "step": 106
+    },
+    {
+      "epoch": 0.09972041006523766,
+      "grad_norm": 0.2667399616888502,
+      "learning_rate": 1.6614906832298137e-05,
+      "loss": 0.5139,
+      "step": 107
+    },
+    {
+      "epoch": 0.10065237651444547,
+      "grad_norm": 0.31044901051003904,
+      "learning_rate": 1.6770186335403728e-05,
+      "loss": 0.5526,
+      "step": 108
+    },
+    {
+      "epoch": 0.1015843429636533,
+      "grad_norm": 0.3009348496814717,
+      "learning_rate": 1.6925465838509316e-05,
+      "loss": 0.5052,
+      "step": 109
+    },
+    {
+      "epoch": 0.10251630941286113,
+      "grad_norm": 0.3048614485120374,
+      "learning_rate": 1.7080745341614907e-05,
+      "loss": 0.5342,
+      "step": 110
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.2832793299192663,
+      "learning_rate": 1.7236024844720498e-05,
+      "loss": 0.5122,
+      "step": 111
+    },
+    {
+      "epoch": 0.1043802423112768,
+      "grad_norm": 0.3593405974072312,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.5613,
+      "step": 112
+    },
+    {
+      "epoch": 0.10531220876048462,
+      "grad_norm": 0.30981511770156306,
+      "learning_rate": 1.7546583850931676e-05,
+      "loss": 0.5366,
+      "step": 113
+    },
+    {
+      "epoch": 0.10624417520969245,
+      "grad_norm": 0.32219805769176935,
+      "learning_rate": 1.7701863354037267e-05,
+      "loss": 0.5024,
+      "step": 114
+    },
+    {
+      "epoch": 0.10717614165890028,
+      "grad_norm": 0.3203378154120046,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.5031,
+      "step": 115
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.3128867712448066,
+      "learning_rate": 1.801242236024845e-05,
+      "loss": 0.5086,
+      "step": 116
+    },
+    {
+      "epoch": 0.10904007455731593,
+      "grad_norm": 0.33200130361429503,
+      "learning_rate": 1.816770186335404e-05,
+      "loss": 0.5101,
+      "step": 117
+    },
+    {
+      "epoch": 0.10997204100652376,
+      "grad_norm": 0.3733354030424016,
+      "learning_rate": 1.8322981366459628e-05,
+      "loss": 0.5135,
+      "step": 118
+    },
+    {
+      "epoch": 0.11090400745573159,
+      "grad_norm": 0.4008114386689176,
+      "learning_rate": 1.8478260869565216e-05,
+      "loss": 0.5104,
+      "step": 119
+    },
+    {
+      "epoch": 0.11183597390493942,
+      "grad_norm": 0.372673697162063,
+      "learning_rate": 1.8633540372670807e-05,
+      "loss": 0.5062,
+      "step": 120
+    },
+    {
+      "epoch": 0.11276794035414725,
+      "grad_norm": 0.34201790832333295,
+      "learning_rate": 1.8788819875776398e-05,
+      "loss": 0.5278,
+      "step": 121
+    },
+    {
+      "epoch": 0.11369990680335508,
+      "grad_norm": 0.32903922248613227,
+      "learning_rate": 1.894409937888199e-05,
+      "loss": 0.5265,
+      "step": 122
+    },
+    {
+      "epoch": 0.11463187325256291,
+      "grad_norm": 0.2994641807850054,
+      "learning_rate": 1.909937888198758e-05,
+      "loss": 0.4905,
+      "step": 123
+    },
+    {
+      "epoch": 0.11556383970177074,
+      "grad_norm": 0.3741532843697418,
+      "learning_rate": 1.9254658385093167e-05,
+      "loss": 0.5039,
+      "step": 124
+    },
+    {
+      "epoch": 0.11649580615097857,
+      "grad_norm": 0.2889883308929261,
+      "learning_rate": 1.940993788819876e-05,
+      "loss": 0.4796,
+      "step": 125
+    },
+    {
+      "epoch": 0.11742777260018639,
+      "grad_norm": 0.3440267366711165,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.5169,
+      "step": 126
+    },
+    {
+      "epoch": 0.11835973904939422,
+      "grad_norm": 0.3543907558633809,
+      "learning_rate": 1.972049689440994e-05,
+      "loss": 0.5179,
+      "step": 127
+    },
+    {
+      "epoch": 0.11929170549860205,
+      "grad_norm": 0.3131085865083079,
+      "learning_rate": 1.9875776397515528e-05,
+      "loss": 0.5098,
+      "step": 128
+    },
+    {
+      "epoch": 0.12022367194780988,
+      "grad_norm": 0.3434666426488577,
+      "learning_rate": 2.003105590062112e-05,
+      "loss": 0.5005,
+      "step": 129
+    },
+    {
+      "epoch": 0.1211556383970177,
+      "grad_norm": 0.3239800968274196,
+      "learning_rate": 2.0186335403726707e-05,
+      "loss": 0.5115,
+      "step": 130
+    },
+    {
+      "epoch": 0.12208760484622554,
+      "grad_norm": 0.32328448086058864,
+      "learning_rate": 2.0341614906832298e-05,
+      "loss": 0.4933,
+      "step": 131
+    },
+    {
+      "epoch": 0.12301957129543337,
+      "grad_norm": 0.3583848599194262,
+      "learning_rate": 2.049689440993789e-05,
+      "loss": 0.5225,
+      "step": 132
+    },
+    {
+      "epoch": 0.1239515377446412,
+      "grad_norm": 0.4074774946339028,
+      "learning_rate": 2.065217391304348e-05,
+      "loss": 0.5083,
+      "step": 133
+    },
+    {
+      "epoch": 0.12488350419384903,
+      "grad_norm": 0.4004767966960763,
+      "learning_rate": 2.080745341614907e-05,
+      "loss": 0.5419,
+      "step": 134
+    },
+    {
+      "epoch": 0.12581547064305684,
+      "grad_norm": 0.41229266644243384,
+      "learning_rate": 2.096273291925466e-05,
+      "loss": 0.4863,
+      "step": 135
+    },
+    {
+      "epoch": 0.1267474370922647,
+      "grad_norm": 0.32604130375634743,
+      "learning_rate": 2.111801242236025e-05,
+      "loss": 0.4883,
+      "step": 136
+    },
+    {
+      "epoch": 0.1276794035414725,
+      "grad_norm": 0.41166893978564467,
+      "learning_rate": 2.127329192546584e-05,
+      "loss": 0.493,
+      "step": 137
+    },
+    {
+      "epoch": 0.12861136999068035,
+      "grad_norm": 0.3647351028188646,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.5063,
+      "step": 138
+    },
+    {
+      "epoch": 0.12954333643988816,
+      "grad_norm": 0.33830815032748895,
+      "learning_rate": 2.158385093167702e-05,
+      "loss": 0.4832,
+      "step": 139
+    },
+    {
+      "epoch": 0.13047530288909598,
+      "grad_norm": 0.47605037434071984,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.5044,
+      "step": 140
+    },
+    {
+      "epoch": 0.13140726933830382,
+      "grad_norm": 0.3584156056096329,
+      "learning_rate": 2.1894409937888198e-05,
+      "loss": 0.5505,
+      "step": 141
+    },
+    {
+      "epoch": 0.13233923578751164,
+      "grad_norm": 0.3873545358782271,
+      "learning_rate": 2.204968944099379e-05,
+      "loss": 0.5101,
+      "step": 142
+    },
+    {
+      "epoch": 0.13327120223671948,
+      "grad_norm": 0.37185348986080047,
+      "learning_rate": 2.220496894409938e-05,
+      "loss": 0.4883,
+      "step": 143
+    },
+    {
+      "epoch": 0.1342031686859273,
+      "grad_norm": 0.33701243081886656,
+      "learning_rate": 2.236024844720497e-05,
+      "loss": 0.5429,
+      "step": 144
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 0.36008406045114716,
+      "learning_rate": 2.2515527950310562e-05,
+      "loss": 0.518,
+      "step": 145
+    },
+    {
+      "epoch": 0.13606710158434296,
+      "grad_norm": 0.3361699856741811,
+      "learning_rate": 2.2670807453416153e-05,
+      "loss": 0.5028,
+      "step": 146
+    },
+    {
+      "epoch": 0.1369990680335508,
+      "grad_norm": 0.3365806805735381,
+      "learning_rate": 2.282608695652174e-05,
+      "loss": 0.5076,
+      "step": 147
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 0.3695301350816239,
+      "learning_rate": 2.2981366459627328e-05,
+      "loss": 0.4589,
+      "step": 148
+    },
+    {
+      "epoch": 0.13886300093196646,
+      "grad_norm": 0.3416412887901837,
+      "learning_rate": 2.313664596273292e-05,
+      "loss": 0.475,
+      "step": 149
+    },
+    {
+      "epoch": 0.13979496738117428,
+      "grad_norm": 0.4312839043669766,
+      "learning_rate": 2.329192546583851e-05,
+      "loss": 0.5347,
+      "step": 150
+    },
+    {
+      "epoch": 0.1407269338303821,
+      "grad_norm": 0.3963882098636216,
+      "learning_rate": 2.34472049689441e-05,
+      "loss": 0.525,
+      "step": 151
+    },
+    {
+      "epoch": 0.14165890027958994,
+      "grad_norm": 0.40030233235048773,
+      "learning_rate": 2.3602484472049692e-05,
+      "loss": 0.5092,
+      "step": 152
+    },
+    {
+      "epoch": 0.14259086672879775,
+      "grad_norm": 0.45224821428089196,
+      "learning_rate": 2.375776397515528e-05,
+      "loss": 0.4934,
+      "step": 153
+    },
+    {
+      "epoch": 0.1435228331780056,
+      "grad_norm": 0.4048224642380499,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 0.5079,
+      "step": 154
+    },
+    {
+      "epoch": 0.14445479962721341,
+      "grad_norm": 0.4422443131911315,
+      "learning_rate": 2.4068322981366462e-05,
+      "loss": 0.485,
+      "step": 155
+    },
+    {
+      "epoch": 0.14538676607642126,
+      "grad_norm": 0.3838612648303963,
+      "learning_rate": 2.4223602484472053e-05,
+      "loss": 0.4741,
+      "step": 156
+    },
+    {
+      "epoch": 0.14631873252562907,
+      "grad_norm": 0.42702690743636545,
+      "learning_rate": 2.437888198757764e-05,
+      "loss": 0.4746,
+      "step": 157
+    },
+    {
+      "epoch": 0.14725069897483692,
+      "grad_norm": 0.5050257313640408,
+      "learning_rate": 2.453416149068323e-05,
+      "loss": 0.5032,
+      "step": 158
+    },
+    {
+      "epoch": 0.14818266542404473,
+      "grad_norm": 0.32355621628910514,
+      "learning_rate": 2.468944099378882e-05,
+      "loss": 0.4919,
+      "step": 159
+    },
+    {
+      "epoch": 0.14911463187325255,
+      "grad_norm": 0.476745406057311,
+      "learning_rate": 2.484472049689441e-05,
+      "loss": 0.4798,
+      "step": 160
+    },
+    {
+      "epoch": 0.1500465983224604,
+      "grad_norm": 0.3982929372013835,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4802,
+      "step": 161
+    },
+    {
+      "epoch": 0.1509785647716682,
+      "grad_norm": 0.422716108126963,
+      "learning_rate": 2.515527950310559e-05,
+      "loss": 0.4871,
+      "step": 162
+    },
+    {
+      "epoch": 0.15191053122087605,
+      "grad_norm": 0.45394267501232327,
+      "learning_rate": 2.5310559006211183e-05,
+      "loss": 0.4953,
+      "step": 163
+    },
+    {
+      "epoch": 0.15284249767008387,
+      "grad_norm": 0.444703945661995,
+      "learning_rate": 2.546583850931677e-05,
+      "loss": 0.5126,
+      "step": 164
+    },
+    {
+      "epoch": 0.15377446411929171,
+      "grad_norm": 0.46144157432809024,
+      "learning_rate": 2.5621118012422362e-05,
+      "loss": 0.505,
+      "step": 165
+    },
+    {
+      "epoch": 0.15470643056849953,
+      "grad_norm": 0.3331384714302454,
+      "learning_rate": 2.577639751552795e-05,
+      "loss": 0.4894,
+      "step": 166
+    },
+    {
+      "epoch": 0.15563839701770738,
+      "grad_norm": 0.4327358532312828,
+      "learning_rate": 2.5931677018633544e-05,
+      "loss": 0.4662,
+      "step": 167
+    },
+    {
+      "epoch": 0.1565703634669152,
+      "grad_norm": 0.41507028457414397,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.4889,
+      "step": 168
+    },
+    {
+      "epoch": 0.157502329916123,
+      "grad_norm": 0.3207393755494713,
+      "learning_rate": 2.6242236024844723e-05,
+      "loss": 0.4487,
+      "step": 169
+    },
+    {
+      "epoch": 0.15843429636533085,
+      "grad_norm": 0.41266599902913287,
+      "learning_rate": 2.639751552795031e-05,
+      "loss": 0.466,
+      "step": 170
+    },
+    {
+      "epoch": 0.15936626281453867,
+      "grad_norm": 0.428215020935446,
+      "learning_rate": 2.6552795031055898e-05,
+      "loss": 0.4809,
+      "step": 171
+    },
+    {
+      "epoch": 0.1602982292637465,
+      "grad_norm": 0.3779105091640831,
+      "learning_rate": 2.6708074534161492e-05,
+      "loss": 0.475,
+      "step": 172
+    },
+    {
+      "epoch": 0.16123019571295433,
+      "grad_norm": 0.4232652928379094,
+      "learning_rate": 2.686335403726708e-05,
+      "loss": 0.5201,
+      "step": 173
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 0.3544628724558091,
+      "learning_rate": 2.7018633540372674e-05,
+      "loss": 0.467,
+      "step": 174
+    },
+    {
+      "epoch": 0.16309412861137,
+      "grad_norm": 0.45734194313541326,
+      "learning_rate": 2.7173913043478262e-05,
+      "loss": 0.4827,
+      "step": 175
+    },
+    {
+      "epoch": 0.16402609506057783,
+      "grad_norm": 0.41596237907501865,
+      "learning_rate": 2.7329192546583853e-05,
+      "loss": 0.465,
+      "step": 176
+    },
+    {
+      "epoch": 0.16495806150978565,
+      "grad_norm": 0.44585447152058544,
+      "learning_rate": 2.748447204968944e-05,
+      "loss": 0.4664,
+      "step": 177
+    },
+    {
+      "epoch": 0.16589002795899346,
+      "grad_norm": 0.44790933604646643,
+      "learning_rate": 2.7639751552795035e-05,
+      "loss": 0.4747,
+      "step": 178
+    },
+    {
+      "epoch": 0.1668219944082013,
+      "grad_norm": 0.4247432605473045,
+      "learning_rate": 2.7795031055900623e-05,
+      "loss": 0.5042,
+      "step": 179
+    },
+    {
+      "epoch": 0.16775396085740912,
+      "grad_norm": 0.4236527478706997,
+      "learning_rate": 2.795031055900621e-05,
+      "loss": 0.4792,
+      "step": 180
+    },
+    {
+      "epoch": 0.16868592730661697,
+      "grad_norm": 0.35352258563908345,
+      "learning_rate": 2.8105590062111805e-05,
+      "loss": 0.4613,
+      "step": 181
+    },
+    {
+      "epoch": 0.16961789375582478,
+      "grad_norm": 0.4774821486093568,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 0.5154,
+      "step": 182
+    },
+    {
+      "epoch": 0.17054986020503263,
+      "grad_norm": 0.3102986003717606,
+      "learning_rate": 2.8416149068322983e-05,
+      "loss": 0.4631,
+      "step": 183
+    },
+    {
+      "epoch": 0.17148182665424044,
+      "grad_norm": 0.5173261219586216,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.4778,
+      "step": 184
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 0.36342743753458534,
+      "learning_rate": 2.8726708074534165e-05,
+      "loss": 0.4869,
+      "step": 185
+    },
+    {
+      "epoch": 0.1733457595526561,
+      "grad_norm": 0.4714510651034062,
+      "learning_rate": 2.8881987577639753e-05,
+      "loss": 0.4795,
+      "step": 186
+    },
+    {
+      "epoch": 0.17427772600186392,
+      "grad_norm": 0.4709054693957223,
+      "learning_rate": 2.9037267080745344e-05,
+      "loss": 0.4828,
+      "step": 187
+    },
+    {
+      "epoch": 0.17520969245107176,
+      "grad_norm": 0.3595943408631176,
+      "learning_rate": 2.919254658385093e-05,
+      "loss": 0.4676,
+      "step": 188
+    },
+    {
+      "epoch": 0.17614165890027958,
+      "grad_norm": 0.5643318504875648,
+      "learning_rate": 2.9347826086956526e-05,
+      "loss": 0.5268,
+      "step": 189
+    },
+    {
+      "epoch": 0.17707362534948742,
+      "grad_norm": 0.38327283117564825,
+      "learning_rate": 2.9503105590062114e-05,
+      "loss": 0.4967,
+      "step": 190
+    },
+    {
+      "epoch": 0.17800559179869524,
+      "grad_norm": 0.4845666744657445,
+      "learning_rate": 2.96583850931677e-05,
+      "loss": 0.4845,
+      "step": 191
+    },
+    {
+      "epoch": 0.17893755824790308,
+      "grad_norm": 0.48009215060131666,
+      "learning_rate": 2.9813664596273296e-05,
+      "loss": 0.4545,
+      "step": 192
+    },
+    {
+      "epoch": 0.1798695246971109,
+      "grad_norm": 0.4492632010395007,
+      "learning_rate": 2.9968944099378883e-05,
+      "loss": 0.479,
+      "step": 193
+    },
+    {
+      "epoch": 0.18080149114631874,
+      "grad_norm": 0.5718318456713907,
+      "learning_rate": 3.0124223602484474e-05,
+      "loss": 0.4947,
+      "step": 194
+    },
+    {
+      "epoch": 0.18173345759552656,
+      "grad_norm": 0.44274090933929144,
+      "learning_rate": 3.0279503105590062e-05,
+      "loss": 0.4882,
+      "step": 195
+    },
+    {
+      "epoch": 0.18266542404473438,
+      "grad_norm": 0.5806933683583505,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.4826,
+      "step": 196
+    },
+    {
+      "epoch": 0.18359739049394222,
+      "grad_norm": 0.46013518255055336,
+      "learning_rate": 3.059006211180124e-05,
+      "loss": 0.498,
+      "step": 197
+    },
+    {
+      "epoch": 0.18452935694315004,
+      "grad_norm": 0.5118143895209565,
+      "learning_rate": 3.074534161490684e-05,
+      "loss": 0.4716,
+      "step": 198
+    },
+    {
+      "epoch": 0.18546132339235788,
+      "grad_norm": 0.5586275618305687,
+      "learning_rate": 3.090062111801242e-05,
+      "loss": 0.4951,
+      "step": 199
+    },
+    {
+      "epoch": 0.1863932898415657,
+      "grad_norm": 0.4743067327136167,
+      "learning_rate": 3.1055900621118014e-05,
+      "loss": 0.4614,
+      "step": 200
+    },
+    {
+      "epoch": 0.18732525629077354,
+      "grad_norm": 0.5595231439186525,
+      "learning_rate": 3.1211180124223605e-05,
+      "loss": 0.4749,
+      "step": 201
+    },
+    {
+      "epoch": 0.18825722273998136,
+      "grad_norm": 0.3797616572472874,
+      "learning_rate": 3.136645962732919e-05,
+      "loss": 0.4623,
+      "step": 202
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 0.465020284443976,
+      "learning_rate": 3.152173913043479e-05,
+      "loss": 0.4591,
+      "step": 203
+    },
+    {
+      "epoch": 0.19012115563839702,
+      "grad_norm": 0.4568130216973761,
+      "learning_rate": 3.167701863354037e-05,
+      "loss": 0.486,
+      "step": 204
+    },
+    {
+      "epoch": 0.19105312208760486,
+      "grad_norm": 0.5201510472689673,
+      "learning_rate": 3.183229813664597e-05,
+      "loss": 0.4713,
+      "step": 205
+    },
+    {
+      "epoch": 0.19198508853681268,
+      "grad_norm": 0.48509111898033974,
+      "learning_rate": 3.198757763975155e-05,
+      "loss": 0.5151,
+      "step": 206
+    },
+    {
+      "epoch": 0.1929170549860205,
+      "grad_norm": 0.48343153313312603,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.4833,
+      "step": 207
+    },
+    {
+      "epoch": 0.19384902143522834,
+      "grad_norm": 0.4183786436939928,
+      "learning_rate": 3.2298136645962735e-05,
+      "loss": 0.4759,
+      "step": 208
+    },
+    {
+      "epoch": 0.19478098788443615,
+      "grad_norm": 0.42459144146413286,
+      "learning_rate": 3.245341614906832e-05,
+      "loss": 0.4565,
+      "step": 209
+    },
+    {
+      "epoch": 0.195712954333644,
+      "grad_norm": 0.5498347809435861,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.4922,
+      "step": 210
+    },
+    {
+      "epoch": 0.1966449207828518,
+      "grad_norm": 0.4909794237208185,
+      "learning_rate": 3.27639751552795e-05,
+      "loss": 0.4923,
+      "step": 211
+    },
+    {
+      "epoch": 0.19757688723205966,
+      "grad_norm": 0.4800780680009219,
+      "learning_rate": 3.29192546583851e-05,
+      "loss": 0.486,
+      "step": 212
+    },
+    {
+      "epoch": 0.19850885368126747,
+      "grad_norm": 0.4261625277481412,
+      "learning_rate": 3.307453416149068e-05,
+      "loss": 0.4547,
+      "step": 213
+    },
+    {
+      "epoch": 0.19944082013047532,
+      "grad_norm": 0.4273368371188702,
+      "learning_rate": 3.3229813664596274e-05,
+      "loss": 0.4674,
+      "step": 214
+    },
+    {
+      "epoch": 0.20037278657968313,
+      "grad_norm": 0.396790198680162,
+      "learning_rate": 3.3385093167701865e-05,
+      "loss": 0.434,
+      "step": 215
+    },
+    {
+      "epoch": 0.20130475302889095,
+      "grad_norm": 0.38636477994735763,
+      "learning_rate": 3.3540372670807456e-05,
+      "loss": 0.4535,
+      "step": 216
+    },
+    {
+      "epoch": 0.2022367194780988,
+      "grad_norm": 0.4343668783329621,
+      "learning_rate": 3.369565217391305e-05,
+      "loss": 0.4522,
+      "step": 217
+    },
+    {
+      "epoch": 0.2031686859273066,
+      "grad_norm": 0.5437356445155337,
+      "learning_rate": 3.385093167701863e-05,
+      "loss": 0.5064,
+      "step": 218
+    },
+    {
+      "epoch": 0.20410065237651445,
+      "grad_norm": 0.46830600975937536,
+      "learning_rate": 3.400621118012422e-05,
+      "loss": 0.457,
+      "step": 219
+    },
+    {
+      "epoch": 0.20503261882572227,
+      "grad_norm": 0.4601517619102894,
+      "learning_rate": 3.4161490683229814e-05,
+      "loss": 0.4875,
+      "step": 220
+    },
+    {
+      "epoch": 0.2059645852749301,
+      "grad_norm": 0.4037905347149945,
+      "learning_rate": 3.4316770186335405e-05,
+      "loss": 0.4515,
+      "step": 221
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.48140505072267437,
+      "learning_rate": 3.4472049689440996e-05,
+      "loss": 0.4516,
+      "step": 222
+    },
+    {
+      "epoch": 0.20782851817334577,
+      "grad_norm": 0.40394980433729955,
+      "learning_rate": 3.462732919254659e-05,
+      "loss": 0.4512,
+      "step": 223
+    },
+    {
+      "epoch": 0.2087604846225536,
+      "grad_norm": 0.4547536351044821,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.4399,
+      "step": 224
+    },
+    {
+      "epoch": 0.2096924510717614,
+      "grad_norm": 0.5005104682053386,
+      "learning_rate": 3.493788819875777e-05,
+      "loss": 0.4726,
+      "step": 225
+    },
+    {
+      "epoch": 0.21062441752096925,
+      "grad_norm": 0.47044822627514055,
+      "learning_rate": 3.509316770186335e-05,
+      "loss": 0.4492,
+      "step": 226
+    },
+    {
+      "epoch": 0.21155638397017706,
+      "grad_norm": 0.5392825246580789,
+      "learning_rate": 3.524844720496895e-05,
+      "loss": 0.4833,
+      "step": 227
+    },
+    {
+      "epoch": 0.2124883504193849,
+      "grad_norm": 0.4274419065657126,
+      "learning_rate": 3.5403726708074535e-05,
+      "loss": 0.4466,
+      "step": 228
+    },
+    {
+      "epoch": 0.21342031686859272,
+      "grad_norm": 0.8886579562608834,
+      "learning_rate": 3.5559006211180126e-05,
+      "loss": 0.4866,
+      "step": 229
+    },
+    {
+      "epoch": 0.21435228331780057,
+      "grad_norm": 0.5224082962576141,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.468,
+      "step": 230
+    },
+    {
+      "epoch": 0.21528424976700838,
+      "grad_norm": 0.5646473711515231,
+      "learning_rate": 3.58695652173913e-05,
+      "loss": 0.4577,
+      "step": 231
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.47983749782808505,
+      "learning_rate": 3.60248447204969e-05,
+      "loss": 0.4788,
+      "step": 232
+    },
+    {
+      "epoch": 0.21714818266542404,
+      "grad_norm": 0.5217571142225462,
+      "learning_rate": 3.618012422360248e-05,
+      "loss": 0.4873,
+      "step": 233
+    },
+    {
+      "epoch": 0.21808014911463186,
+      "grad_norm": 0.49015785204927476,
+      "learning_rate": 3.633540372670808e-05,
+      "loss": 0.4584,
+      "step": 234
+    },
+    {
+      "epoch": 0.2190121155638397,
+      "grad_norm": 0.41949877977071165,
+      "learning_rate": 3.6490683229813665e-05,
+      "loss": 0.4409,
+      "step": 235
+    },
+    {
+      "epoch": 0.21994408201304752,
+      "grad_norm": 0.5575862848782227,
+      "learning_rate": 3.6645962732919256e-05,
+      "loss": 0.4721,
+      "step": 236
+    },
+    {
+      "epoch": 0.22087604846225536,
+      "grad_norm": 0.5533406186763868,
+      "learning_rate": 3.680124223602485e-05,
+      "loss": 0.4684,
+      "step": 237
+    },
+    {
+      "epoch": 0.22180801491146318,
+      "grad_norm": 0.6052077775010289,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 0.479,
+      "step": 238
+    },
+    {
+      "epoch": 0.22273998136067102,
+      "grad_norm": 0.4268889108178479,
+      "learning_rate": 3.711180124223603e-05,
+      "loss": 0.4361,
+      "step": 239
+    },
+    {
+      "epoch": 0.22367194780987884,
+      "grad_norm": 0.6434981263109816,
+      "learning_rate": 3.7267080745341614e-05,
+      "loss": 0.4566,
+      "step": 240
+    },
+    {
+      "epoch": 0.22460391425908668,
+      "grad_norm": 0.6605063834268231,
+      "learning_rate": 3.742236024844721e-05,
+      "loss": 0.4738,
+      "step": 241
+    },
+    {
+      "epoch": 0.2255358807082945,
+      "grad_norm": 0.4841503555788819,
+      "learning_rate": 3.7577639751552796e-05,
+      "loss": 0.4568,
+      "step": 242
+    },
+    {
+      "epoch": 0.22646784715750232,
+      "grad_norm": 0.5564237238434837,
+      "learning_rate": 3.773291925465839e-05,
+      "loss": 0.4788,
+      "step": 243
+    },
+    {
+      "epoch": 0.22739981360671016,
+      "grad_norm": 0.5625816816546112,
+      "learning_rate": 3.788819875776398e-05,
+      "loss": 0.4566,
+      "step": 244
+    },
+    {
+      "epoch": 0.22833178005591798,
+      "grad_norm": 0.571026320722015,
+      "learning_rate": 3.804347826086957e-05,
+      "loss": 0.4723,
+      "step": 245
+    },
+    {
+      "epoch": 0.22926374650512582,
+      "grad_norm": 0.4582113950751743,
+      "learning_rate": 3.819875776397516e-05,
+      "loss": 0.4469,
+      "step": 246
+    },
+    {
+      "epoch": 0.23019571295433364,
+      "grad_norm": 0.5521348013983153,
+      "learning_rate": 3.8354037267080744e-05,
+      "loss": 0.4446,
+      "step": 247
+    },
+    {
+      "epoch": 0.23112767940354148,
+      "grad_norm": 0.40752375159424387,
+      "learning_rate": 3.8509316770186335e-05,
+      "loss": 0.484,
+      "step": 248
+    },
+    {
+      "epoch": 0.2320596458527493,
+      "grad_norm": 0.4676387855249031,
+      "learning_rate": 3.8664596273291926e-05,
+      "loss": 0.4681,
+      "step": 249
+    },
+    {
+      "epoch": 0.23299161230195714,
+      "grad_norm": 0.45667671235217744,
+      "learning_rate": 3.881987577639752e-05,
+      "loss": 0.4536,
+      "step": 250
+    },
+    {
+      "epoch": 0.23392357875116496,
+      "grad_norm": 0.460915721750586,
+      "learning_rate": 3.897515527950311e-05,
+      "loss": 0.4847,
+      "step": 251
+    },
+    {
+      "epoch": 0.23485554520037277,
+      "grad_norm": 0.5004155581748244,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.4506,
+      "step": 252
+    },
+    {
+      "epoch": 0.23578751164958062,
+      "grad_norm": 0.45404356204524976,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.444,
+      "step": 253
+    },
+    {
+      "epoch": 0.23671947809878843,
+      "grad_norm": 0.6807879979625724,
+      "learning_rate": 3.944099378881988e-05,
+      "loss": 0.4861,
+      "step": 254
+    },
+    {
+      "epoch": 0.23765144454799628,
+      "grad_norm": 0.5457146601573246,
+      "learning_rate": 3.9596273291925465e-05,
+      "loss": 0.4748,
+      "step": 255
+    },
+    {
+      "epoch": 0.2385834109972041,
+      "grad_norm": 0.5212769915480384,
+      "learning_rate": 3.9751552795031056e-05,
+      "loss": 0.4483,
+      "step": 256
+    },
+    {
+      "epoch": 0.23951537744641194,
+      "grad_norm": 0.617913718729022,
+      "learning_rate": 3.990683229813665e-05,
+      "loss": 0.4605,
+      "step": 257
+    },
+    {
+      "epoch": 0.24044734389561975,
+      "grad_norm": 0.5161335724843282,
+      "learning_rate": 4.006211180124224e-05,
+      "loss": 0.4605,
+      "step": 258
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 0.5537740602972695,
+      "learning_rate": 4.021739130434783e-05,
+      "loss": 0.4488,
+      "step": 259
+    },
+    {
+      "epoch": 0.2423112767940354,
+      "grad_norm": 0.5478527134517855,
+      "learning_rate": 4.0372670807453414e-05,
+      "loss": 0.4974,
+      "step": 260
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 0.6010464556586923,
+      "learning_rate": 4.052795031055901e-05,
+      "loss": 0.4684,
+      "step": 261
+    },
+    {
+      "epoch": 0.24417520969245107,
+      "grad_norm": 0.5520960678214537,
+      "learning_rate": 4.0683229813664596e-05,
+      "loss": 0.506,
+      "step": 262
+    },
+    {
+      "epoch": 0.2451071761416589,
+      "grad_norm": 0.6322846756598782,
+      "learning_rate": 4.0838509316770193e-05,
+      "loss": 0.4622,
+      "step": 263
+    },
+    {
+      "epoch": 0.24603914259086673,
+      "grad_norm": 0.4456128378937014,
+      "learning_rate": 4.099378881987578e-05,
+      "loss": 0.421,
+      "step": 264
+    },
+    {
+      "epoch": 0.24697110904007455,
+      "grad_norm": 0.6352274559588069,
+      "learning_rate": 4.114906832298137e-05,
+      "loss": 0.4485,
+      "step": 265
+    },
+    {
+      "epoch": 0.2479030754892824,
+      "grad_norm": 0.6148018576287125,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 0.4685,
+      "step": 266
+    },
+    {
+      "epoch": 0.2488350419384902,
+      "grad_norm": 0.5021888096963858,
+      "learning_rate": 4.1459627329192544e-05,
+      "loss": 0.4704,
+      "step": 267
+    },
+    {
+      "epoch": 0.24976700838769805,
+      "grad_norm": 0.549622945783278,
+      "learning_rate": 4.161490683229814e-05,
+      "loss": 0.4675,
+      "step": 268
+    },
+    {
+      "epoch": 0.2506989748369059,
+      "grad_norm": 0.559999919557955,
+      "learning_rate": 4.1770186335403726e-05,
+      "loss": 0.4481,
+      "step": 269
+    },
+    {
+      "epoch": 0.2516309412861137,
+      "grad_norm": 0.5291209733796245,
+      "learning_rate": 4.192546583850932e-05,
+      "loss": 0.4424,
+      "step": 270
+    },
+    {
+      "epoch": 0.25256290773532153,
+      "grad_norm": 0.4756924341410583,
+      "learning_rate": 4.208074534161491e-05,
+      "loss": 0.43,
+      "step": 271
+    },
+    {
+      "epoch": 0.2534948741845294,
+      "grad_norm": 0.4739325254980473,
+      "learning_rate": 4.22360248447205e-05,
+      "loss": 0.4813,
+      "step": 272
+    },
+    {
+      "epoch": 0.25442684063373716,
+      "grad_norm": 0.511051941921151,
+      "learning_rate": 4.239130434782609e-05,
+      "loss": 0.4746,
+      "step": 273
+    },
+    {
+      "epoch": 0.255358807082945,
+      "grad_norm": 0.5673148794345746,
+      "learning_rate": 4.254658385093168e-05,
+      "loss": 0.4638,
+      "step": 274
+    },
+    {
+      "epoch": 0.25629077353215285,
+      "grad_norm": 0.48492471849583646,
+      "learning_rate": 4.270186335403727e-05,
+      "loss": 0.4768,
+      "step": 275
+    },
+    {
+      "epoch": 0.2572227399813607,
+      "grad_norm": 0.4965211334579507,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.4963,
+      "step": 276
+    },
+    {
+      "epoch": 0.2581547064305685,
+      "grad_norm": 0.491650689101607,
+      "learning_rate": 4.301242236024845e-05,
+      "loss": 0.4407,
+      "step": 277
+    },
+    {
+      "epoch": 0.2590866728797763,
+      "grad_norm": 0.47923357068990213,
+      "learning_rate": 4.316770186335404e-05,
+      "loss": 0.4465,
+      "step": 278
+    },
+    {
+      "epoch": 0.26001863932898417,
+      "grad_norm": 0.4334599818873252,
+      "learning_rate": 4.332298136645963e-05,
+      "loss": 0.4636,
+      "step": 279
+    },
+    {
+      "epoch": 0.26095060577819196,
+      "grad_norm": 0.45097293634512947,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.4389,
+      "step": 280
+    },
+    {
+      "epoch": 0.2618825722273998,
+      "grad_norm": 0.4994613757430349,
+      "learning_rate": 4.363354037267081e-05,
+      "loss": 0.4693,
+      "step": 281
+    },
+    {
+      "epoch": 0.26281453867660765,
+      "grad_norm": 0.5004320229904372,
+      "learning_rate": 4.3788819875776396e-05,
+      "loss": 0.487,
+      "step": 282
+    },
+    {
+      "epoch": 0.2637465051258155,
+      "grad_norm": 0.403092241343922,
+      "learning_rate": 4.3944099378881993e-05,
+      "loss": 0.4497,
+      "step": 283
+    },
+    {
+      "epoch": 0.2646784715750233,
+      "grad_norm": 0.47445383193393076,
+      "learning_rate": 4.409937888198758e-05,
+      "loss": 0.4468,
+      "step": 284
+    },
+    {
+      "epoch": 0.2656104380242311,
+      "grad_norm": 0.5217830218541727,
+      "learning_rate": 4.425465838509317e-05,
+      "loss": 0.4538,
+      "step": 285
+    },
+    {
+      "epoch": 0.26654240447343897,
+      "grad_norm": 0.4147434786653566,
+      "learning_rate": 4.440993788819876e-05,
+      "loss": 0.4746,
+      "step": 286
+    },
+    {
+      "epoch": 0.2674743709226468,
+      "grad_norm": 0.4742912646844642,
+      "learning_rate": 4.456521739130435e-05,
+      "loss": 0.4494,
+      "step": 287
+    },
+    {
+      "epoch": 0.2684063373718546,
+      "grad_norm": 0.5310131102369109,
+      "learning_rate": 4.472049689440994e-05,
+      "loss": 0.4658,
+      "step": 288
+    },
+    {
+      "epoch": 0.26933830382106244,
+      "grad_norm": 0.4963082648515746,
+      "learning_rate": 4.4875776397515526e-05,
+      "loss": 0.4757,
+      "step": 289
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 0.6287231280829776,
+      "learning_rate": 4.5031055900621124e-05,
+      "loss": 0.4508,
+      "step": 290
+    },
+    {
+      "epoch": 0.2712022367194781,
+      "grad_norm": 0.45327285126591543,
+      "learning_rate": 4.518633540372671e-05,
+      "loss": 0.4669,
+      "step": 291
+    },
+    {
+      "epoch": 0.2721342031686859,
+      "grad_norm": 0.6068552492786371,
+      "learning_rate": 4.5341614906832306e-05,
+      "loss": 0.4628,
+      "step": 292
+    },
+    {
+      "epoch": 0.27306616961789376,
+      "grad_norm": 0.7353709086881577,
+      "learning_rate": 4.549689440993789e-05,
+      "loss": 0.4555,
+      "step": 293
+    },
+    {
+      "epoch": 0.2739981360671016,
+      "grad_norm": 0.48968608263820923,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 0.4688,
+      "step": 294
+    },
+    {
+      "epoch": 0.2749301025163094,
+      "grad_norm": 0.46508535971363946,
+      "learning_rate": 4.580745341614907e-05,
+      "loss": 0.4823,
+      "step": 295
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 0.5189233892851299,
+      "learning_rate": 4.5962732919254656e-05,
+      "loss": 0.4435,
+      "step": 296
+    },
+    {
+      "epoch": 0.2767940354147251,
+      "grad_norm": 0.47767544980806614,
+      "learning_rate": 4.6118012422360254e-05,
+      "loss": 0.46,
+      "step": 297
+    },
+    {
+      "epoch": 0.2777260018639329,
+      "grad_norm": 0.38094558271709666,
+      "learning_rate": 4.627329192546584e-05,
+      "loss": 0.4767,
+      "step": 298
+    },
+    {
+      "epoch": 0.2786579683131407,
+      "grad_norm": 0.5155422134307653,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.458,
+      "step": 299
+    },
+    {
+      "epoch": 0.27958993476234856,
+      "grad_norm": 0.7304189879996122,
+      "learning_rate": 4.658385093167702e-05,
+      "loss": 0.5104,
+      "step": 300
+    },
+    {
+      "epoch": 0.2805219012115564,
+      "grad_norm": 0.45796577605069705,
+      "learning_rate": 4.673913043478261e-05,
+      "loss": 0.4559,
+      "step": 301
+    },
+    {
+      "epoch": 0.2814538676607642,
+      "grad_norm": 0.5519720475636191,
+      "learning_rate": 4.68944099378882e-05,
+      "loss": 0.4562,
+      "step": 302
+    },
+    {
+      "epoch": 0.28238583410997203,
+      "grad_norm": 0.45028024120092275,
+      "learning_rate": 4.7049689440993793e-05,
+      "loss": 0.4399,
+      "step": 303
+    },
+    {
+      "epoch": 0.2833178005591799,
+      "grad_norm": 0.4644433769758883,
+      "learning_rate": 4.7204968944099384e-05,
+      "loss": 0.476,
+      "step": 304
+    },
+    {
+      "epoch": 0.2842497670083877,
+      "grad_norm": 0.5027475615494298,
+      "learning_rate": 4.736024844720497e-05,
+      "loss": 0.4611,
+      "step": 305
+    },
+    {
+      "epoch": 0.2851817334575955,
+      "grad_norm": 0.46032309080526723,
+      "learning_rate": 4.751552795031056e-05,
+      "loss": 0.4361,
+      "step": 306
+    },
+    {
+      "epoch": 0.28611369990680335,
+      "grad_norm": 0.4554114823460758,
+      "learning_rate": 4.767080745341615e-05,
+      "loss": 0.4455,
+      "step": 307
+    },
+    {
+      "epoch": 0.2870456663560112,
+      "grad_norm": 0.5041397315313941,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.4443,
+      "step": 308
+    },
+    {
+      "epoch": 0.287977632805219,
+      "grad_norm": 0.48745624771090884,
+      "learning_rate": 4.798136645962733e-05,
+      "loss": 0.4481,
+      "step": 309
+    },
+    {
+      "epoch": 0.28890959925442683,
+      "grad_norm": 0.5446843697268071,
+      "learning_rate": 4.8136645962732924e-05,
+      "loss": 0.4753,
+      "step": 310
+    },
+    {
+      "epoch": 0.2898415657036347,
+      "grad_norm": 0.6695961520443128,
+      "learning_rate": 4.829192546583851e-05,
+      "loss": 0.4775,
+      "step": 311
+    },
+    {
+      "epoch": 0.2907735321528425,
+      "grad_norm": 0.5533312833129141,
+      "learning_rate": 4.8447204968944106e-05,
+      "loss": 0.4494,
+      "step": 312
+    },
+    {
+      "epoch": 0.2917054986020503,
+      "grad_norm": 0.46074215461470897,
+      "learning_rate": 4.860248447204969e-05,
+      "loss": 0.4659,
+      "step": 313
+    },
+    {
+      "epoch": 0.29263746505125815,
+      "grad_norm": 0.5851760335341725,
+      "learning_rate": 4.875776397515528e-05,
+      "loss": 0.4348,
+      "step": 314
+    },
+    {
+      "epoch": 0.293569431500466,
+      "grad_norm": 0.6469655804797101,
+      "learning_rate": 4.891304347826087e-05,
+      "loss": 0.4638,
+      "step": 315
+    },
+    {
+      "epoch": 0.29450139794967384,
+      "grad_norm": 0.5592080329365986,
+      "learning_rate": 4.906832298136646e-05,
+      "loss": 0.4311,
+      "step": 316
+    },
+    {
+      "epoch": 0.2954333643988816,
+      "grad_norm": 0.4173269297526815,
+      "learning_rate": 4.9223602484472054e-05,
+      "loss": 0.4349,
+      "step": 317
+    },
+    {
+      "epoch": 0.29636533084808947,
+      "grad_norm": 0.6059316085423471,
+      "learning_rate": 4.937888198757764e-05,
+      "loss": 0.4529,
+      "step": 318
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 0.458820327714281,
+      "learning_rate": 4.9534161490683236e-05,
+      "loss": 0.4388,
+      "step": 319
+    },
+    {
+      "epoch": 0.2982292637465051,
+      "grad_norm": 0.7478809797059247,
+      "learning_rate": 4.968944099378882e-05,
+      "loss": 0.4593,
+      "step": 320
+    },
+    {
+      "epoch": 0.29916123019571295,
+      "grad_norm": 0.6658310261740328,
+      "learning_rate": 4.984472049689442e-05,
+      "loss": 0.4313,
+      "step": 321
+    },
+    {
+      "epoch": 0.3000931966449208,
+      "grad_norm": 0.5860222669887302,
+      "learning_rate": 5e-05,
+      "loss": 0.4466,
+      "step": 322
+    },
+    {
+      "epoch": 0.30102516309412863,
+      "grad_norm": 0.742861941357761,
+      "learning_rate": 4.998274076630998e-05,
+      "loss": 0.4439,
+      "step": 323
+    },
+    {
+      "epoch": 0.3019571295433364,
+      "grad_norm": 0.4757913466729815,
+      "learning_rate": 4.996548153261996e-05,
+      "loss": 0.4437,
+      "step": 324
+    },
+    {
+      "epoch": 0.30288909599254427,
+      "grad_norm": 0.571766456801266,
+      "learning_rate": 4.994822229892993e-05,
+      "loss": 0.4405,
+      "step": 325
+    },
+    {
+      "epoch": 0.3038210624417521,
+      "grad_norm": 0.5353725332810465,
+      "learning_rate": 4.9930963065239906e-05,
+      "loss": 0.441,
+      "step": 326
+    },
+    {
+      "epoch": 0.3047530288909599,
+      "grad_norm": 0.4724784652120246,
+      "learning_rate": 4.991370383154988e-05,
+      "loss": 0.4541,
+      "step": 327
+    },
+    {
+      "epoch": 0.30568499534016774,
+      "grad_norm": 0.5488925626955962,
+      "learning_rate": 4.989644459785986e-05,
+      "loss": 0.444,
+      "step": 328
+    },
+    {
+      "epoch": 0.3066169617893756,
+      "grad_norm": 0.567341319202618,
+      "learning_rate": 4.9879185364169836e-05,
+      "loss": 0.4736,
+      "step": 329
+    },
+    {
+      "epoch": 0.30754892823858343,
+      "grad_norm": 0.4394458072561044,
+      "learning_rate": 4.986192613047981e-05,
+      "loss": 0.4457,
+      "step": 330
+    },
+    {
+      "epoch": 0.3084808946877912,
+      "grad_norm": 0.45730685941999943,
+      "learning_rate": 4.9844666896789785e-05,
+      "loss": 0.4458,
+      "step": 331
+    },
+    {
+      "epoch": 0.30941286113699906,
+      "grad_norm": 0.386841992928942,
+      "learning_rate": 4.982740766309976e-05,
+      "loss": 0.4467,
+      "step": 332
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.6073785702054916,
+      "learning_rate": 4.981014842940973e-05,
+      "loss": 0.4475,
+      "step": 333
+    },
+    {
+      "epoch": 0.31127679403541475,
+      "grad_norm": 0.5972985712873782,
+      "learning_rate": 4.979288919571971e-05,
+      "loss": 0.4547,
+      "step": 334
+    },
+    {
+      "epoch": 0.31220876048462254,
+      "grad_norm": 0.6716744951024707,
+      "learning_rate": 4.977562996202969e-05,
+      "loss": 0.4651,
+      "step": 335
+    },
+    {
+      "epoch": 0.3131407269338304,
+      "grad_norm": 0.7967607142318743,
+      "learning_rate": 4.975837072833966e-05,
+      "loss": 0.4752,
+      "step": 336
+    },
+    {
+      "epoch": 0.3140726933830382,
+      "grad_norm": 0.6478702958514692,
+      "learning_rate": 4.974111149464964e-05,
+      "loss": 0.4471,
+      "step": 337
+    },
+    {
+      "epoch": 0.315004659832246,
+      "grad_norm": 0.45051157179590146,
+      "learning_rate": 4.972385226095961e-05,
+      "loss": 0.4551,
+      "step": 338
+    },
+    {
+      "epoch": 0.31593662628145386,
+      "grad_norm": 0.7785647329095983,
+      "learning_rate": 4.970659302726959e-05,
+      "loss": 0.4315,
+      "step": 339
+    },
+    {
+      "epoch": 0.3168685927306617,
+      "grad_norm": 0.5152968357442445,
+      "learning_rate": 4.968933379357957e-05,
+      "loss": 0.4565,
+      "step": 340
+    },
+    {
+      "epoch": 0.31780055917986955,
+      "grad_norm": 0.5073564114104989,
+      "learning_rate": 4.967207455988954e-05,
+      "loss": 0.447,
+      "step": 341
+    },
+    {
+      "epoch": 0.31873252562907733,
+      "grad_norm": 0.6524407766869843,
+      "learning_rate": 4.9654815326199515e-05,
+      "loss": 0.4655,
+      "step": 342
+    },
+    {
+      "epoch": 0.3196644920782852,
+      "grad_norm": 0.39317976170574664,
+      "learning_rate": 4.9637556092509496e-05,
+      "loss": 0.4383,
+      "step": 343
+    },
+    {
+      "epoch": 0.320596458527493,
+      "grad_norm": 0.43174284257281476,
+      "learning_rate": 4.962029685881947e-05,
+      "loss": 0.4655,
+      "step": 344
+    },
+    {
+      "epoch": 0.32152842497670087,
+      "grad_norm": 0.5173274032207504,
+      "learning_rate": 4.9603037625129445e-05,
+      "loss": 0.4226,
+      "step": 345
+    },
+    {
+      "epoch": 0.32246039142590865,
+      "grad_norm": 0.3989205057345724,
+      "learning_rate": 4.9585778391439426e-05,
+      "loss": 0.4533,
+      "step": 346
+    },
+    {
+      "epoch": 0.3233923578751165,
+      "grad_norm": 0.5105406240015495,
+      "learning_rate": 4.95685191577494e-05,
+      "loss": 0.4891,
+      "step": 347
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.4981228870972702,
+      "learning_rate": 4.9551259924059375e-05,
+      "loss": 0.4635,
+      "step": 348
+    },
+    {
+      "epoch": 0.32525629077353213,
+      "grad_norm": 0.4329334116512223,
+      "learning_rate": 4.953400069036935e-05,
+      "loss": 0.4516,
+      "step": 349
+    },
+    {
+      "epoch": 0.32618825722274,
+      "grad_norm": 0.4424882171569886,
+      "learning_rate": 4.951674145667933e-05,
+      "loss": 0.4487,
+      "step": 350
+    },
+    {
+      "epoch": 0.3271202236719478,
+      "grad_norm": 0.4828692472404843,
+      "learning_rate": 4.9499482222989304e-05,
+      "loss": 0.4529,
+      "step": 351
+    },
+    {
+      "epoch": 0.32805219012115566,
+      "grad_norm": 0.3715365309718623,
+      "learning_rate": 4.948222298929928e-05,
+      "loss": 0.4441,
+      "step": 352
+    },
+    {
+      "epoch": 0.32898415657036345,
+      "grad_norm": 0.4249008119096798,
+      "learning_rate": 4.946496375560925e-05,
+      "loss": 0.436,
+      "step": 353
+    },
+    {
+      "epoch": 0.3299161230195713,
+      "grad_norm": 0.40820813996660404,
+      "learning_rate": 4.944770452191923e-05,
+      "loss": 0.4412,
+      "step": 354
+    },
+    {
+      "epoch": 0.33084808946877914,
+      "grad_norm": 0.4503474495093967,
+      "learning_rate": 4.94304452882292e-05,
+      "loss": 0.4621,
+      "step": 355
+    },
+    {
+      "epoch": 0.3317800559179869,
+      "grad_norm": 0.36039668301674344,
+      "learning_rate": 4.9413186054539176e-05,
+      "loss": 0.4245,
+      "step": 356
+    },
+    {
+      "epoch": 0.33271202236719477,
+      "grad_norm": 0.46666041779400613,
+      "learning_rate": 4.939592682084916e-05,
+      "loss": 0.4482,
+      "step": 357
+    },
+    {
+      "epoch": 0.3336439888164026,
+      "grad_norm": 0.38487843936119337,
+      "learning_rate": 4.937866758715913e-05,
+      "loss": 0.4521,
+      "step": 358
+    },
+    {
+      "epoch": 0.33457595526561046,
+      "grad_norm": 0.46962486382047963,
+      "learning_rate": 4.9361408353469105e-05,
+      "loss": 0.4712,
+      "step": 359
+    },
+    {
+      "epoch": 0.33550792171481825,
+      "grad_norm": 0.4634194481437413,
+      "learning_rate": 4.934414911977908e-05,
+      "loss": 0.4287,
+      "step": 360
+    },
+    {
+      "epoch": 0.3364398881640261,
+      "grad_norm": 0.4308211387363992,
+      "learning_rate": 4.932688988608906e-05,
+      "loss": 0.4317,
+      "step": 361
+    },
+    {
+      "epoch": 0.33737185461323393,
+      "grad_norm": 0.4912417959326823,
+      "learning_rate": 4.9309630652399035e-05,
+      "loss": 0.43,
+      "step": 362
+    },
+    {
+      "epoch": 0.3383038210624418,
+      "grad_norm": 0.39139657353331125,
+      "learning_rate": 4.929237141870901e-05,
+      "loss": 0.4402,
+      "step": 363
+    },
+    {
+      "epoch": 0.33923578751164957,
+      "grad_norm": 0.4255736583648049,
+      "learning_rate": 4.927511218501899e-05,
+      "loss": 0.4692,
+      "step": 364
+    },
+    {
+      "epoch": 0.3401677539608574,
+      "grad_norm": 0.3887387347450407,
+      "learning_rate": 4.9257852951328965e-05,
+      "loss": 0.435,
+      "step": 365
+    },
+    {
+      "epoch": 0.34109972041006525,
+      "grad_norm": 0.4687843457632725,
+      "learning_rate": 4.924059371763894e-05,
+      "loss": 0.443,
+      "step": 366
+    },
+    {
+      "epoch": 0.34203168685927304,
+      "grad_norm": 0.39327227804962384,
+      "learning_rate": 4.922333448394891e-05,
+      "loss": 0.4883,
+      "step": 367
+    },
+    {
+      "epoch": 0.3429636533084809,
+      "grad_norm": 0.6523044629586924,
+      "learning_rate": 4.9206075250258894e-05,
+      "loss": 0.4581,
+      "step": 368
+    },
+    {
+      "epoch": 0.34389561975768873,
+      "grad_norm": 0.4210406004114677,
+      "learning_rate": 4.918881601656887e-05,
+      "loss": 0.4279,
+      "step": 369
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 0.43906059445888745,
+      "learning_rate": 4.917155678287884e-05,
+      "loss": 0.4586,
+      "step": 370
+    },
+    {
+      "epoch": 0.34575955265610436,
+      "grad_norm": 0.4792828121967785,
+      "learning_rate": 4.915429754918882e-05,
+      "loss": 0.4375,
+      "step": 371
+    },
+    {
+      "epoch": 0.3466915191053122,
+      "grad_norm": 0.3907318388593238,
+      "learning_rate": 4.91370383154988e-05,
+      "loss": 0.4384,
+      "step": 372
+    },
+    {
+      "epoch": 0.34762348555452005,
+      "grad_norm": 0.5109697900521999,
+      "learning_rate": 4.911977908180877e-05,
+      "loss": 0.4479,
+      "step": 373
+    },
+    {
+      "epoch": 0.34855545200372784,
+      "grad_norm": 0.4655116768866418,
+      "learning_rate": 4.910251984811875e-05,
+      "loss": 0.4222,
+      "step": 374
+    },
+    {
+      "epoch": 0.3494874184529357,
+      "grad_norm": 0.5171739145285091,
+      "learning_rate": 4.908526061442872e-05,
+      "loss": 0.4592,
+      "step": 375
+    },
+    {
+      "epoch": 0.3504193849021435,
+      "grad_norm": 0.4620440226848061,
+      "learning_rate": 4.9068001380738695e-05,
+      "loss": 0.4457,
+      "step": 376
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 0.5739032370205556,
+      "learning_rate": 4.905074214704867e-05,
+      "loss": 0.4763,
+      "step": 377
+    },
+    {
+      "epoch": 0.35228331780055916,
+      "grad_norm": 0.42399533660851946,
+      "learning_rate": 4.9033482913358644e-05,
+      "loss": 0.4431,
+      "step": 378
+    },
+    {
+      "epoch": 0.353215284249767,
+      "grad_norm": 0.5094069108013991,
+      "learning_rate": 4.9016223679668625e-05,
+      "loss": 0.4672,
+      "step": 379
+    },
+    {
+      "epoch": 0.35414725069897485,
+      "grad_norm": 0.6695123006762566,
+      "learning_rate": 4.89989644459786e-05,
+      "loss": 0.4494,
+      "step": 380
+    },
+    {
+      "epoch": 0.3550792171481827,
+      "grad_norm": 0.48699330756644876,
+      "learning_rate": 4.8981705212288574e-05,
+      "loss": 0.4536,
+      "step": 381
+    },
+    {
+      "epoch": 0.3560111835973905,
+      "grad_norm": 0.4662099678877696,
+      "learning_rate": 4.896444597859855e-05,
+      "loss": 0.4407,
+      "step": 382
+    },
+    {
+      "epoch": 0.3569431500465983,
+      "grad_norm": 0.5302632639110546,
+      "learning_rate": 4.894718674490853e-05,
+      "loss": 0.4796,
+      "step": 383
+    },
+    {
+      "epoch": 0.35787511649580617,
+      "grad_norm": 0.38167914144438797,
+      "learning_rate": 4.89299275112185e-05,
+      "loss": 0.4638,
+      "step": 384
+    },
+    {
+      "epoch": 0.35880708294501396,
+      "grad_norm": 0.43316737588041576,
+      "learning_rate": 4.891266827752848e-05,
+      "loss": 0.4507,
+      "step": 385
+    },
+    {
+      "epoch": 0.3597390493942218,
+      "grad_norm": 0.4621337621765705,
+      "learning_rate": 4.889540904383846e-05,
+      "loss": 0.4353,
+      "step": 386
+    },
+    {
+      "epoch": 0.36067101584342964,
+      "grad_norm": 0.3519453970448515,
+      "learning_rate": 4.887814981014843e-05,
+      "loss": 0.4231,
+      "step": 387
+    },
+    {
+      "epoch": 0.3616029822926375,
+      "grad_norm": 0.4151971820857612,
+      "learning_rate": 4.886089057645841e-05,
+      "loss": 0.4477,
+      "step": 388
+    },
+    {
+      "epoch": 0.3625349487418453,
+      "grad_norm": 0.5102531155183986,
+      "learning_rate": 4.884363134276838e-05,
+      "loss": 0.4552,
+      "step": 389
+    },
+    {
+      "epoch": 0.3634669151910531,
+      "grad_norm": 0.4392258374973051,
+      "learning_rate": 4.882637210907836e-05,
+      "loss": 0.4471,
+      "step": 390
+    },
+    {
+      "epoch": 0.36439888164026096,
+      "grad_norm": 0.44108026714919335,
+      "learning_rate": 4.880911287538834e-05,
+      "loss": 0.4727,
+      "step": 391
+    },
+    {
+      "epoch": 0.36533084808946875,
+      "grad_norm": 0.48943008115652337,
+      "learning_rate": 4.879185364169831e-05,
+      "loss": 0.4082,
+      "step": 392
+    },
+    {
+      "epoch": 0.3662628145386766,
+      "grad_norm": 0.4820388794049537,
+      "learning_rate": 4.877459440800829e-05,
+      "loss": 0.4364,
+      "step": 393
+    },
+    {
+      "epoch": 0.36719478098788444,
+      "grad_norm": 0.5631374311031292,
+      "learning_rate": 4.8757335174318266e-05,
+      "loss": 0.4558,
+      "step": 394
+    },
+    {
+      "epoch": 0.3681267474370923,
+      "grad_norm": 0.5632480370503572,
+      "learning_rate": 4.874007594062824e-05,
+      "loss": 0.4675,
+      "step": 395
+    },
+    {
+      "epoch": 0.36905871388630007,
+      "grad_norm": 0.45926736004135493,
+      "learning_rate": 4.8722816706938215e-05,
+      "loss": 0.4314,
+      "step": 396
+    },
+    {
+      "epoch": 0.3699906803355079,
+      "grad_norm": 0.5059373955390541,
+      "learning_rate": 4.870555747324819e-05,
+      "loss": 0.4127,
+      "step": 397
+    },
+    {
+      "epoch": 0.37092264678471576,
+      "grad_norm": 0.49747693917439995,
+      "learning_rate": 4.8688298239558164e-05,
+      "loss": 0.4627,
+      "step": 398
+    },
+    {
+      "epoch": 0.3718546132339236,
+      "grad_norm": 0.44761949139088597,
+      "learning_rate": 4.867103900586814e-05,
+      "loss": 0.435,
+      "step": 399
+    },
+    {
+      "epoch": 0.3727865796831314,
+      "grad_norm": 0.4028833131078924,
+      "learning_rate": 4.865377977217811e-05,
+      "loss": 0.4223,
+      "step": 400
+    },
+    {
+      "epoch": 0.37371854613233924,
+      "grad_norm": 0.4366782285522345,
+      "learning_rate": 4.863652053848809e-05,
+      "loss": 0.4458,
+      "step": 401
+    },
+    {
+      "epoch": 0.3746505125815471,
+      "grad_norm": 0.5377558090555021,
+      "learning_rate": 4.861926130479807e-05,
+      "loss": 0.4768,
+      "step": 402
+    },
+    {
+      "epoch": 0.37558247903075487,
+      "grad_norm": 0.37785439974718316,
+      "learning_rate": 4.860200207110804e-05,
+      "loss": 0.4409,
+      "step": 403
+    },
+    {
+      "epoch": 0.3765144454799627,
+      "grad_norm": 0.5139325576150725,
+      "learning_rate": 4.858474283741802e-05,
+      "loss": 0.4528,
+      "step": 404
+    },
+    {
+      "epoch": 0.37744641192917056,
+      "grad_norm": 0.4566884277250674,
+      "learning_rate": 4.8567483603728e-05,
+      "loss": 0.4402,
+      "step": 405
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 0.4398791329788954,
+      "learning_rate": 4.855022437003797e-05,
+      "loss": 0.4415,
+      "step": 406
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 0.4358221261707614,
+      "learning_rate": 4.8532965136347946e-05,
+      "loss": 0.4471,
+      "step": 407
+    },
+    {
+      "epoch": 0.38024231127679403,
+      "grad_norm": 0.49146996691995987,
+      "learning_rate": 4.851570590265793e-05,
+      "loss": 0.4466,
+      "step": 408
+    },
+    {
+      "epoch": 0.3811742777260019,
+      "grad_norm": 0.4444176781372391,
+      "learning_rate": 4.84984466689679e-05,
+      "loss": 0.4533,
+      "step": 409
+    },
+    {
+      "epoch": 0.3821062441752097,
+      "grad_norm": 0.4334211028143657,
+      "learning_rate": 4.8481187435277875e-05,
+      "loss": 0.4476,
+      "step": 410
+    },
+    {
+      "epoch": 0.3830382106244175,
+      "grad_norm": 0.4379691613814686,
+      "learning_rate": 4.846392820158785e-05,
+      "loss": 0.4484,
+      "step": 411
+    },
+    {
+      "epoch": 0.38397017707362535,
+      "grad_norm": 0.49556798321467854,
+      "learning_rate": 4.844666896789783e-05,
+      "loss": 0.4651,
+      "step": 412
+    },
+    {
+      "epoch": 0.3849021435228332,
+      "grad_norm": 0.45127125112881306,
+      "learning_rate": 4.8429409734207805e-05,
+      "loss": 0.4141,
+      "step": 413
+    },
+    {
+      "epoch": 0.385834109972041,
+      "grad_norm": 0.3585652611484045,
+      "learning_rate": 4.841215050051778e-05,
+      "loss": 0.3976,
+      "step": 414
+    },
+    {
+      "epoch": 0.38676607642124883,
+      "grad_norm": 0.5492178138278585,
+      "learning_rate": 4.839489126682776e-05,
+      "loss": 0.4437,
+      "step": 415
+    },
+    {
+      "epoch": 0.38769804287045667,
+      "grad_norm": 0.41688111141257655,
+      "learning_rate": 4.8377632033137735e-05,
+      "loss": 0.4808,
+      "step": 416
+    },
+    {
+      "epoch": 0.3886300093196645,
+      "grad_norm": 0.5204144963380288,
+      "learning_rate": 4.836037279944771e-05,
+      "loss": 0.455,
+      "step": 417
+    },
+    {
+      "epoch": 0.3895619757688723,
+      "grad_norm": 0.41468555236518545,
+      "learning_rate": 4.834311356575768e-05,
+      "loss": 0.4397,
+      "step": 418
+    },
+    {
+      "epoch": 0.39049394221808015,
+      "grad_norm": 0.5045933495353204,
+      "learning_rate": 4.832585433206766e-05,
+      "loss": 0.4617,
+      "step": 419
+    },
+    {
+      "epoch": 0.391425908667288,
+      "grad_norm": 0.5210216995508107,
+      "learning_rate": 4.830859509837763e-05,
+      "loss": 0.4232,
+      "step": 420
+    },
+    {
+      "epoch": 0.3923578751164958,
+      "grad_norm": 0.4540038039492862,
+      "learning_rate": 4.8291335864687606e-05,
+      "loss": 0.4481,
+      "step": 421
+    },
+    {
+      "epoch": 0.3932898415657036,
+      "grad_norm": 0.5042006134990532,
+      "learning_rate": 4.827407663099758e-05,
+      "loss": 0.4476,
+      "step": 422
+    },
+    {
+      "epoch": 0.39422180801491147,
+      "grad_norm": 0.4142328943886276,
+      "learning_rate": 4.825681739730756e-05,
+      "loss": 0.4765,
+      "step": 423
+    },
+    {
+      "epoch": 0.3951537744641193,
+      "grad_norm": 0.6042604282210047,
+      "learning_rate": 4.8239558163617536e-05,
+      "loss": 0.4362,
+      "step": 424
+    },
+    {
+      "epoch": 0.3960857409133271,
+      "grad_norm": 0.36754402602837055,
+      "learning_rate": 4.822229892992751e-05,
+      "loss": 0.4399,
+      "step": 425
+    },
+    {
+      "epoch": 0.39701770736253494,
+      "grad_norm": 0.4629190478631869,
+      "learning_rate": 4.820503969623749e-05,
+      "loss": 0.4611,
+      "step": 426
+    },
+    {
+      "epoch": 0.3979496738117428,
+      "grad_norm": 0.41673795487213416,
+      "learning_rate": 4.8187780462547465e-05,
+      "loss": 0.442,
+      "step": 427
+    },
+    {
+      "epoch": 0.39888164026095063,
+      "grad_norm": 0.4241052978714828,
+      "learning_rate": 4.817052122885744e-05,
+      "loss": 0.4208,
+      "step": 428
+    },
+    {
+      "epoch": 0.3998136067101584,
+      "grad_norm": 0.39896875658571024,
+      "learning_rate": 4.8153261995167414e-05,
+      "loss": 0.4172,
+      "step": 429
+    },
+    {
+      "epoch": 0.40074557315936626,
+      "grad_norm": 0.444107222483263,
+      "learning_rate": 4.8136002761477395e-05,
+      "loss": 0.4422,
+      "step": 430
+    },
+    {
+      "epoch": 0.4016775396085741,
+      "grad_norm": 0.602775733307215,
+      "learning_rate": 4.811874352778737e-05,
+      "loss": 0.4487,
+      "step": 431
+    },
+    {
+      "epoch": 0.4026095060577819,
+      "grad_norm": 0.5226842925245354,
+      "learning_rate": 4.8101484294097344e-05,
+      "loss": 0.4566,
+      "step": 432
+    },
+    {
+      "epoch": 0.40354147250698974,
+      "grad_norm": 0.5503067390556056,
+      "learning_rate": 4.808422506040732e-05,
+      "loss": 0.4409,
+      "step": 433
+    },
+    {
+      "epoch": 0.4044734389561976,
+      "grad_norm": 0.4277087528787114,
+      "learning_rate": 4.80669658267173e-05,
+      "loss": 0.4546,
+      "step": 434
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 0.5224945338700132,
+      "learning_rate": 4.804970659302727e-05,
+      "loss": 0.4481,
+      "step": 435
+    },
+    {
+      "epoch": 0.4063373718546132,
+      "grad_norm": 0.45719986994881234,
+      "learning_rate": 4.803244735933725e-05,
+      "loss": 0.4538,
+      "step": 436
+    },
+    {
+      "epoch": 0.40726933830382106,
+      "grad_norm": 0.4910807809440503,
+      "learning_rate": 4.801518812564723e-05,
+      "loss": 0.454,
+      "step": 437
+    },
+    {
+      "epoch": 0.4082013047530289,
+      "grad_norm": 0.449773898401191,
+      "learning_rate": 4.79979288919572e-05,
+      "loss": 0.4385,
+      "step": 438
+    },
+    {
+      "epoch": 0.4091332712022367,
+      "grad_norm": 0.5120780680958733,
+      "learning_rate": 4.798066965826718e-05,
+      "loss": 0.4331,
+      "step": 439
+    },
+    {
+      "epoch": 0.41006523765144454,
+      "grad_norm": 0.47310815395955036,
+      "learning_rate": 4.796341042457715e-05,
+      "loss": 0.4396,
+      "step": 440
+    },
+    {
+      "epoch": 0.4109972041006524,
+      "grad_norm": 0.5812431667044037,
+      "learning_rate": 4.7946151190887126e-05,
+      "loss": 0.4396,
+      "step": 441
+    },
+    {
+      "epoch": 0.4119291705498602,
+      "grad_norm": 0.4363723301004132,
+      "learning_rate": 4.79288919571971e-05,
+      "loss": 0.4437,
+      "step": 442
+    },
+    {
+      "epoch": 0.412861136999068,
+      "grad_norm": 0.4968044762514634,
+      "learning_rate": 4.7911632723507074e-05,
+      "loss": 0.4314,
+      "step": 443
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.4876921687040951,
+      "learning_rate": 4.7894373489817056e-05,
+      "loss": 0.4224,
+      "step": 444
+    },
+    {
+      "epoch": 0.4147250698974837,
+      "grad_norm": 0.4446310728940312,
+      "learning_rate": 4.787711425612703e-05,
+      "loss": 0.455,
+      "step": 445
+    },
+    {
+      "epoch": 0.41565703634669154,
+      "grad_norm": 0.5172075684712749,
+      "learning_rate": 4.7859855022437004e-05,
+      "loss": 0.4405,
+      "step": 446
+    },
+    {
+      "epoch": 0.41658900279589933,
+      "grad_norm": 0.37750953127519166,
+      "learning_rate": 4.784259578874698e-05,
+      "loss": 0.4251,
+      "step": 447
+    },
+    {
+      "epoch": 0.4175209692451072,
+      "grad_norm": 0.4756453076342448,
+      "learning_rate": 4.782533655505696e-05,
+      "loss": 0.4275,
+      "step": 448
+    },
+    {
+      "epoch": 0.418452935694315,
+      "grad_norm": 0.42549454047898655,
+      "learning_rate": 4.7808077321366934e-05,
+      "loss": 0.4355,
+      "step": 449
+    },
+    {
+      "epoch": 0.4193849021435228,
+      "grad_norm": 0.4659771580722852,
+      "learning_rate": 4.779081808767691e-05,
+      "loss": 0.4172,
+      "step": 450
+    },
+    {
+      "epoch": 0.42031686859273065,
+      "grad_norm": 0.42958222319148,
+      "learning_rate": 4.777355885398688e-05,
+      "loss": 0.4332,
+      "step": 451
+    },
+    {
+      "epoch": 0.4212488350419385,
+      "grad_norm": 0.43254350546592,
+      "learning_rate": 4.7756299620296863e-05,
+      "loss": 0.4427,
+      "step": 452
+    },
+    {
+      "epoch": 0.42218080149114634,
+      "grad_norm": 0.43483520539124887,
+      "learning_rate": 4.773904038660684e-05,
+      "loss": 0.4436,
+      "step": 453
+    },
+    {
+      "epoch": 0.42311276794035413,
+      "grad_norm": 0.3917778298256813,
+      "learning_rate": 4.772178115291681e-05,
+      "loss": 0.436,
+      "step": 454
+    },
+    {
+      "epoch": 0.424044734389562,
+      "grad_norm": 0.49843369408793725,
+      "learning_rate": 4.770452191922679e-05,
+      "loss": 0.4246,
+      "step": 455
+    },
+    {
+      "epoch": 0.4249767008387698,
+      "grad_norm": 0.36224324206275865,
+      "learning_rate": 4.768726268553677e-05,
+      "loss": 0.4109,
+      "step": 456
+    },
+    {
+      "epoch": 0.42590866728797766,
+      "grad_norm": 0.5534460913382782,
+      "learning_rate": 4.767000345184674e-05,
+      "loss": 0.4414,
+      "step": 457
+    },
+    {
+      "epoch": 0.42684063373718545,
+      "grad_norm": 0.34782573179296483,
+      "learning_rate": 4.7652744218156716e-05,
+      "loss": 0.4419,
+      "step": 458
+    },
+    {
+      "epoch": 0.4277726001863933,
+      "grad_norm": 0.4961067198322361,
+      "learning_rate": 4.76354849844667e-05,
+      "loss": 0.4249,
+      "step": 459
+    },
+    {
+      "epoch": 0.42870456663560114,
+      "grad_norm": 0.3553668668077219,
+      "learning_rate": 4.761822575077667e-05,
+      "loss": 0.4445,
+      "step": 460
+    },
+    {
+      "epoch": 0.4296365330848089,
+      "grad_norm": 0.45548346221984887,
+      "learning_rate": 4.7600966517086646e-05,
+      "loss": 0.4333,
+      "step": 461
+    },
+    {
+      "epoch": 0.43056849953401677,
+      "grad_norm": 0.36584560527995097,
+      "learning_rate": 4.758370728339662e-05,
+      "loss": 0.4171,
+      "step": 462
+    },
+    {
+      "epoch": 0.4315004659832246,
+      "grad_norm": 0.4833336344828295,
+      "learning_rate": 4.7566448049706594e-05,
+      "loss": 0.4213,
+      "step": 463
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.3877767545141777,
+      "learning_rate": 4.754918881601657e-05,
+      "loss": 0.4083,
+      "step": 464
+    },
+    {
+      "epoch": 0.43336439888164024,
+      "grad_norm": 0.5188168403161957,
+      "learning_rate": 4.753192958232654e-05,
+      "loss": 0.4692,
+      "step": 465
+    },
+    {
+      "epoch": 0.4342963653308481,
+      "grad_norm": 0.46692053737271194,
+      "learning_rate": 4.7514670348636524e-05,
+      "loss": 0.4242,
+      "step": 466
+    },
+    {
+      "epoch": 0.43522833178005593,
+      "grad_norm": 0.5712714887817703,
+      "learning_rate": 4.74974111149465e-05,
+      "loss": 0.4648,
+      "step": 467
+    },
+    {
+      "epoch": 0.4361602982292637,
+      "grad_norm": 0.5476656412734331,
+      "learning_rate": 4.748015188125647e-05,
+      "loss": 0.4508,
+      "step": 468
+    },
+    {
+      "epoch": 0.43709226467847156,
+      "grad_norm": 0.461385002793244,
+      "learning_rate": 4.746289264756645e-05,
+      "loss": 0.4431,
+      "step": 469
+    },
+    {
+      "epoch": 0.4380242311276794,
+      "grad_norm": 0.4159422223094202,
+      "learning_rate": 4.744563341387643e-05,
+      "loss": 0.4313,
+      "step": 470
+    },
+    {
+      "epoch": 0.43895619757688725,
+      "grad_norm": 0.4338847273357992,
+      "learning_rate": 4.74283741801864e-05,
+      "loss": 0.4209,
+      "step": 471
+    },
+    {
+      "epoch": 0.43988816402609504,
+      "grad_norm": 0.4252549849344018,
+      "learning_rate": 4.7411114946496376e-05,
+      "loss": 0.4592,
+      "step": 472
+    },
+    {
+      "epoch": 0.4408201304753029,
+      "grad_norm": 0.43901773791529974,
+      "learning_rate": 4.739385571280635e-05,
+      "loss": 0.4432,
+      "step": 473
+    },
+    {
+      "epoch": 0.44175209692451073,
+      "grad_norm": 0.43691715416898136,
+      "learning_rate": 4.737659647911633e-05,
+      "loss": 0.4146,
+      "step": 474
+    },
+    {
+      "epoch": 0.4426840633737186,
+      "grad_norm": 0.4514245175382557,
+      "learning_rate": 4.7359337245426306e-05,
+      "loss": 0.4221,
+      "step": 475
+    },
+    {
+      "epoch": 0.44361602982292636,
+      "grad_norm": 0.5259586911431116,
+      "learning_rate": 4.734207801173628e-05,
+      "loss": 0.4635,
+      "step": 476
+    },
+    {
+      "epoch": 0.4445479962721342,
+      "grad_norm": 0.4326472310851479,
+      "learning_rate": 4.732481877804626e-05,
+      "loss": 0.4321,
+      "step": 477
+    },
+    {
+      "epoch": 0.44547996272134205,
+      "grad_norm": 0.5606410974659077,
+      "learning_rate": 4.7307559544356236e-05,
+      "loss": 0.457,
+      "step": 478
+    },
+    {
+      "epoch": 0.44641192917054984,
+      "grad_norm": 0.4518756055678026,
+      "learning_rate": 4.729030031066621e-05,
+      "loss": 0.4262,
+      "step": 479
+    },
+    {
+      "epoch": 0.4473438956197577,
+      "grad_norm": 0.4461931541159781,
+      "learning_rate": 4.7273041076976184e-05,
+      "loss": 0.442,
+      "step": 480
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 0.43455232236827285,
+      "learning_rate": 4.7255781843286165e-05,
+      "loss": 0.4503,
+      "step": 481
+    },
+    {
+      "epoch": 0.44920782851817337,
+      "grad_norm": 0.5194966309805924,
+      "learning_rate": 4.723852260959614e-05,
+      "loss": 0.451,
+      "step": 482
+    },
+    {
+      "epoch": 0.45013979496738116,
+      "grad_norm": 0.42882475412543336,
+      "learning_rate": 4.7221263375906114e-05,
+      "loss": 0.4359,
+      "step": 483
+    },
+    {
+      "epoch": 0.451071761416589,
+      "grad_norm": 0.46134120966802883,
+      "learning_rate": 4.720400414221609e-05,
+      "loss": 0.4107,
+      "step": 484
+    },
+    {
+      "epoch": 0.45200372786579684,
+      "grad_norm": 0.4406473670871895,
+      "learning_rate": 4.718674490852606e-05,
+      "loss": 0.4357,
+      "step": 485
+    },
+    {
+      "epoch": 0.45293569431500463,
+      "grad_norm": 0.4590689218732355,
+      "learning_rate": 4.716948567483604e-05,
+      "loss": 0.4575,
+      "step": 486
+    },
+    {
+      "epoch": 0.4538676607642125,
+      "grad_norm": 0.36556198795462674,
+      "learning_rate": 4.715222644114601e-05,
+      "loss": 0.4283,
+      "step": 487
+    },
+    {
+      "epoch": 0.4547996272134203,
+      "grad_norm": 0.5980611368746573,
+      "learning_rate": 4.713496720745599e-05,
+      "loss": 0.4356,
+      "step": 488
+    },
+    {
+      "epoch": 0.45573159366262816,
+      "grad_norm": 0.42408602752733354,
+      "learning_rate": 4.7117707973765966e-05,
+      "loss": 0.4287,
+      "step": 489
+    },
+    {
+      "epoch": 0.45666356011183595,
+      "grad_norm": 0.41661321472222657,
+      "learning_rate": 4.710044874007594e-05,
+      "loss": 0.4058,
+      "step": 490
+    },
+    {
+      "epoch": 0.4575955265610438,
+      "grad_norm": 0.4353613870417663,
+      "learning_rate": 4.7083189506385915e-05,
+      "loss": 0.4205,
+      "step": 491
+    },
+    {
+      "epoch": 0.45852749301025164,
+      "grad_norm": 0.3523332407989797,
+      "learning_rate": 4.7065930272695896e-05,
+      "loss": 0.4571,
+      "step": 492
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 0.46247247301309624,
+      "learning_rate": 4.704867103900587e-05,
+      "loss": 0.4108,
+      "step": 493
+    },
+    {
+      "epoch": 0.4603914259086673,
+      "grad_norm": 0.40407255451548374,
+      "learning_rate": 4.7031411805315845e-05,
+      "loss": 0.4317,
+      "step": 494
+    },
+    {
+      "epoch": 0.4613233923578751,
+      "grad_norm": 0.4289718593506669,
+      "learning_rate": 4.7014152571625826e-05,
+      "loss": 0.4645,
+      "step": 495
+    },
+    {
+      "epoch": 0.46225535880708296,
+      "grad_norm": 0.3911641534508682,
+      "learning_rate": 4.69968933379358e-05,
+      "loss": 0.4404,
+      "step": 496
+    },
+    {
+      "epoch": 0.46318732525629075,
+      "grad_norm": 0.46798016532445064,
+      "learning_rate": 4.6979634104245774e-05,
+      "loss": 0.4485,
+      "step": 497
+    },
+    {
+      "epoch": 0.4641192917054986,
+      "grad_norm": 0.42560365298975145,
+      "learning_rate": 4.696237487055575e-05,
+      "loss": 0.4275,
+      "step": 498
+    },
+    {
+      "epoch": 0.46505125815470644,
+      "grad_norm": 0.412101801386452,
+      "learning_rate": 4.694511563686573e-05,
+      "loss": 0.4396,
+      "step": 499
+    },
+    {
+      "epoch": 0.4659832246039143,
+      "grad_norm": 0.5171854743605055,
+      "learning_rate": 4.6927856403175704e-05,
+      "loss": 0.4434,
+      "step": 500
+    },
+    {
+      "epoch": 0.46691519105312207,
+      "grad_norm": 0.5138382488680087,
+      "learning_rate": 4.691059716948568e-05,
+      "loss": 0.4447,
+      "step": 501
+    },
+    {
+      "epoch": 0.4678471575023299,
+      "grad_norm": 0.4481053831326444,
+      "learning_rate": 4.689333793579565e-05,
+      "loss": 0.4325,
+      "step": 502
+    },
+    {
+      "epoch": 0.46877912395153776,
+      "grad_norm": 0.49387878672530655,
+      "learning_rate": 4.6876078702105634e-05,
+      "loss": 0.4128,
+      "step": 503
+    },
+    {
+      "epoch": 0.46971109040074555,
+      "grad_norm": 0.4201749615651619,
+      "learning_rate": 4.685881946841561e-05,
+      "loss": 0.4212,
+      "step": 504
+    },
+    {
+      "epoch": 0.4706430568499534,
+      "grad_norm": 0.3866506431958827,
+      "learning_rate": 4.684156023472558e-05,
+      "loss": 0.4215,
+      "step": 505
+    },
+    {
+      "epoch": 0.47157502329916123,
+      "grad_norm": 0.405874862918791,
+      "learning_rate": 4.6824301001035556e-05,
+      "loss": 0.4354,
+      "step": 506
+    },
+    {
+      "epoch": 0.4725069897483691,
+      "grad_norm": 0.4162497294071385,
+      "learning_rate": 4.680704176734553e-05,
+      "loss": 0.4425,
+      "step": 507
+    },
+    {
+      "epoch": 0.47343895619757687,
+      "grad_norm": 0.4628261489258786,
+      "learning_rate": 4.6789782533655505e-05,
+      "loss": 0.4382,
+      "step": 508
+    },
+    {
+      "epoch": 0.4743709226467847,
+      "grad_norm": 0.40008976833276944,
+      "learning_rate": 4.677252329996548e-05,
+      "loss": 0.4532,
+      "step": 509
+    },
+    {
+      "epoch": 0.47530288909599255,
+      "grad_norm": 0.47016394288172314,
+      "learning_rate": 4.675526406627546e-05,
+      "loss": 0.4053,
+      "step": 510
+    },
+    {
+      "epoch": 0.4762348555452004,
+      "grad_norm": 0.4937811918939414,
+      "learning_rate": 4.6738004832585435e-05,
+      "loss": 0.4296,
+      "step": 511
+    },
+    {
+      "epoch": 0.4771668219944082,
+      "grad_norm": 0.3809385801564945,
+      "learning_rate": 4.672074559889541e-05,
+      "loss": 0.4227,
+      "step": 512
+    },
+    {
+      "epoch": 0.47809878844361603,
+      "grad_norm": 0.6589393720277965,
+      "learning_rate": 4.670348636520538e-05,
+      "loss": 0.4553,
+      "step": 513
+    },
+    {
+      "epoch": 0.4790307548928239,
+      "grad_norm": 0.3580659578935438,
+      "learning_rate": 4.6686227131515364e-05,
+      "loss": 0.4272,
+      "step": 514
+    },
+    {
+      "epoch": 0.47996272134203166,
+      "grad_norm": 0.6194590770338666,
+      "learning_rate": 4.666896789782534e-05,
+      "loss": 0.43,
+      "step": 515
+    },
+    {
+      "epoch": 0.4808946877912395,
+      "grad_norm": 0.39394765515374286,
+      "learning_rate": 4.665170866413531e-05,
+      "loss": 0.4319,
+      "step": 516
+    },
+    {
+      "epoch": 0.48182665424044735,
+      "grad_norm": 0.5449130751173703,
+      "learning_rate": 4.6634449430445294e-05,
+      "loss": 0.466,
+      "step": 517
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 0.4570329307063504,
+      "learning_rate": 4.661719019675527e-05,
+      "loss": 0.4208,
+      "step": 518
+    },
+    {
+      "epoch": 0.483690587138863,
+      "grad_norm": 0.49400539112825037,
+      "learning_rate": 4.659993096306524e-05,
+      "loss": 0.4293,
+      "step": 519
+    },
+    {
+      "epoch": 0.4846225535880708,
+      "grad_norm": 0.5656485573763981,
+      "learning_rate": 4.658267172937522e-05,
+      "loss": 0.4357,
+      "step": 520
+    },
+    {
+      "epoch": 0.48555452003727867,
+      "grad_norm": 0.33807226940431284,
+      "learning_rate": 4.65654124956852e-05,
+      "loss": 0.4115,
+      "step": 521
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.4889135751061642,
+      "learning_rate": 4.654815326199517e-05,
+      "loss": 0.4353,
+      "step": 522
+    },
+    {
+      "epoch": 0.4874184529356943,
+      "grad_norm": 0.3573683386554365,
+      "learning_rate": 4.6530894028305146e-05,
+      "loss": 0.4337,
+      "step": 523
+    },
+    {
+      "epoch": 0.48835041938490215,
+      "grad_norm": 0.4555249726074453,
+      "learning_rate": 4.651363479461512e-05,
+      "loss": 0.4336,
+      "step": 524
+    },
+    {
+      "epoch": 0.48928238583411,
+      "grad_norm": 0.4343135259630201,
+      "learning_rate": 4.64963755609251e-05,
+      "loss": 0.4639,
+      "step": 525
+    },
+    {
+      "epoch": 0.4902143522833178,
+      "grad_norm": 0.4247850301381284,
+      "learning_rate": 4.6479116327235076e-05,
+      "loss": 0.4496,
+      "step": 526
+    },
+    {
+      "epoch": 0.4911463187325256,
+      "grad_norm": 0.49910387044192944,
+      "learning_rate": 4.646185709354505e-05,
+      "loss": 0.4246,
+      "step": 527
+    },
+    {
+      "epoch": 0.49207828518173347,
+      "grad_norm": 0.40348997102564854,
+      "learning_rate": 4.6444597859855025e-05,
+      "loss": 0.4388,
+      "step": 528
+    },
+    {
+      "epoch": 0.4930102516309413,
+      "grad_norm": 0.4514643232138925,
+      "learning_rate": 4.6427338626165e-05,
+      "loss": 0.4074,
+      "step": 529
+    },
+    {
+      "epoch": 0.4939422180801491,
+      "grad_norm": 0.49846811352051024,
+      "learning_rate": 4.641007939247497e-05,
+      "loss": 0.4418,
+      "step": 530
+    },
+    {
+      "epoch": 0.49487418452935694,
+      "grad_norm": 0.47973542408822534,
+      "learning_rate": 4.639282015878495e-05,
+      "loss": 0.4381,
+      "step": 531
+    },
+    {
+      "epoch": 0.4958061509785648,
+      "grad_norm": 0.5537079144009359,
+      "learning_rate": 4.637556092509493e-05,
+      "loss": 0.4197,
+      "step": 532
+    },
+    {
+      "epoch": 0.4967381174277726,
+      "grad_norm": 0.3455655942525359,
+      "learning_rate": 4.63583016914049e-05,
+      "loss": 0.4483,
+      "step": 533
+    },
+    {
+      "epoch": 0.4976700838769804,
+      "grad_norm": 0.4699319997660432,
+      "learning_rate": 4.634104245771488e-05,
+      "loss": 0.4449,
+      "step": 534
+    },
+    {
+      "epoch": 0.49860205032618826,
+      "grad_norm": 0.3567605389236115,
+      "learning_rate": 4.632378322402486e-05,
+      "loss": 0.4121,
+      "step": 535
+    },
+    {
+      "epoch": 0.4995340167753961,
+      "grad_norm": 0.47175326441796567,
+      "learning_rate": 4.630652399033483e-05,
+      "loss": 0.4266,
+      "step": 536
+    },
+    {
+      "epoch": 0.5004659832246039,
+      "grad_norm": 0.3822229607059117,
+      "learning_rate": 4.628926475664481e-05,
+      "loss": 0.4007,
+      "step": 537
+    },
+    {
+      "epoch": 0.5013979496738118,
+      "grad_norm": 0.41993933475023626,
+      "learning_rate": 4.627200552295478e-05,
+      "loss": 0.4348,
+      "step": 538
+    },
+    {
+      "epoch": 0.5023299161230196,
+      "grad_norm": 0.3725843974639671,
+      "learning_rate": 4.625474628926476e-05,
+      "loss": 0.437,
+      "step": 539
+    },
+    {
+      "epoch": 0.5032618825722274,
+      "grad_norm": 0.41520439822593475,
+      "learning_rate": 4.6237487055574736e-05,
+      "loss": 0.4153,
+      "step": 540
+    },
+    {
+      "epoch": 0.5041938490214353,
+      "grad_norm": 0.3496535509810763,
+      "learning_rate": 4.622022782188471e-05,
+      "loss": 0.4244,
+      "step": 541
+    },
+    {
+      "epoch": 0.5051258154706431,
+      "grad_norm": 0.4000730761208873,
+      "learning_rate": 4.6202968588194685e-05,
+      "loss": 0.4312,
+      "step": 542
+    },
+    {
+      "epoch": 0.5060577819198508,
+      "grad_norm": 0.400956510229645,
+      "learning_rate": 4.6185709354504666e-05,
+      "loss": 0.4167,
+      "step": 543
+    },
+    {
+      "epoch": 0.5069897483690587,
+      "grad_norm": 0.359552126898788,
+      "learning_rate": 4.616845012081464e-05,
+      "loss": 0.4433,
+      "step": 544
+    },
+    {
+      "epoch": 0.5079217148182665,
+      "grad_norm": 0.47560942858369654,
+      "learning_rate": 4.6151190887124615e-05,
+      "loss": 0.4293,
+      "step": 545
+    },
+    {
+      "epoch": 0.5088536812674743,
+      "grad_norm": 0.3384331918675552,
+      "learning_rate": 4.613393165343459e-05,
+      "loss": 0.4216,
+      "step": 546
+    },
+    {
+      "epoch": 0.5097856477166822,
+      "grad_norm": 0.3803029740182317,
+      "learning_rate": 4.611667241974456e-05,
+      "loss": 0.4358,
+      "step": 547
+    },
+    {
+      "epoch": 0.51071761416589,
+      "grad_norm": 0.35763836213648653,
+      "learning_rate": 4.609941318605454e-05,
+      "loss": 0.4093,
+      "step": 548
+    },
+    {
+      "epoch": 0.5116495806150979,
+      "grad_norm": 0.35135353989182894,
+      "learning_rate": 4.608215395236451e-05,
+      "loss": 0.4412,
+      "step": 549
+    },
+    {
+      "epoch": 0.5125815470643057,
+      "grad_norm": 0.40748642784629496,
+      "learning_rate": 4.606489471867449e-05,
+      "loss": 0.4388,
+      "step": 550
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 0.4077367757769607,
+      "learning_rate": 4.604763548498447e-05,
+      "loss": 0.4483,
+      "step": 551
+    },
+    {
+      "epoch": 0.5144454799627214,
+      "grad_norm": 0.4043253295138619,
+      "learning_rate": 4.603037625129444e-05,
+      "loss": 0.4118,
+      "step": 552
+    },
+    {
+      "epoch": 0.5153774464119292,
+      "grad_norm": 0.36668144319920926,
+      "learning_rate": 4.6013117017604416e-05,
+      "loss": 0.3881,
+      "step": 553
+    },
+    {
+      "epoch": 0.516309412861137,
+      "grad_norm": 0.38624531184747846,
+      "learning_rate": 4.59958577839144e-05,
+      "loss": 0.428,
+      "step": 554
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.3596466433629941,
+      "learning_rate": 4.597859855022437e-05,
+      "loss": 0.4415,
+      "step": 555
+    },
+    {
+      "epoch": 0.5181733457595527,
+      "grad_norm": 0.4473573013798326,
+      "learning_rate": 4.5961339316534345e-05,
+      "loss": 0.4805,
+      "step": 556
+    },
+    {
+      "epoch": 0.5191053122087604,
+      "grad_norm": 0.39612722381632937,
+      "learning_rate": 4.5944080082844326e-05,
+      "loss": 0.4205,
+      "step": 557
+    },
+    {
+      "epoch": 0.5200372786579683,
+      "grad_norm": 0.4461013345692974,
+      "learning_rate": 4.59268208491543e-05,
+      "loss": 0.4711,
+      "step": 558
+    },
+    {
+      "epoch": 0.5209692451071761,
+      "grad_norm": 0.3454076831898733,
+      "learning_rate": 4.5909561615464275e-05,
+      "loss": 0.4296,
+      "step": 559
+    },
+    {
+      "epoch": 0.5219012115563839,
+      "grad_norm": 0.4104050267373853,
+      "learning_rate": 4.589230238177425e-05,
+      "loss": 0.4208,
+      "step": 560
+    },
+    {
+      "epoch": 0.5228331780055918,
+      "grad_norm": 0.3885366263275777,
+      "learning_rate": 4.587504314808423e-05,
+      "loss": 0.436,
+      "step": 561
+    },
+    {
+      "epoch": 0.5237651444547996,
+      "grad_norm": 0.3725585930907195,
+      "learning_rate": 4.5857783914394205e-05,
+      "loss": 0.4319,
+      "step": 562
+    },
+    {
+      "epoch": 0.5246971109040075,
+      "grad_norm": 0.5285774461677041,
+      "learning_rate": 4.584052468070418e-05,
+      "loss": 0.4149,
+      "step": 563
+    },
+    {
+      "epoch": 0.5256290773532153,
+      "grad_norm": 0.39848303736256974,
+      "learning_rate": 4.582326544701415e-05,
+      "loss": 0.4243,
+      "step": 564
+    },
+    {
+      "epoch": 0.5265610438024231,
+      "grad_norm": 0.49009856128964185,
+      "learning_rate": 4.5806006213324134e-05,
+      "loss": 0.4023,
+      "step": 565
+    },
+    {
+      "epoch": 0.527493010251631,
+      "grad_norm": 0.516726286207246,
+      "learning_rate": 4.578874697963411e-05,
+      "loss": 0.4299,
+      "step": 566
+    },
+    {
+      "epoch": 0.5284249767008388,
+      "grad_norm": 0.41260751011984154,
+      "learning_rate": 4.577148774594408e-05,
+      "loss": 0.4479,
+      "step": 567
+    },
+    {
+      "epoch": 0.5293569431500466,
+      "grad_norm": 0.5683649094456259,
+      "learning_rate": 4.575422851225406e-05,
+      "loss": 0.3972,
+      "step": 568
+    },
+    {
+      "epoch": 0.5302889095992545,
+      "grad_norm": 0.3439947926797965,
+      "learning_rate": 4.573696927856403e-05,
+      "loss": 0.4174,
+      "step": 569
+    },
+    {
+      "epoch": 0.5312208760484622,
+      "grad_norm": 0.5521882129034127,
+      "learning_rate": 4.5719710044874006e-05,
+      "loss": 0.4174,
+      "step": 570
+    },
+    {
+      "epoch": 0.53215284249767,
+      "grad_norm": 0.45695453817927134,
+      "learning_rate": 4.570245081118398e-05,
+      "loss": 0.4045,
+      "step": 571
+    },
+    {
+      "epoch": 0.5330848089468779,
+      "grad_norm": 0.48355029949367423,
+      "learning_rate": 4.568519157749396e-05,
+      "loss": 0.439,
+      "step": 572
+    },
+    {
+      "epoch": 0.5340167753960857,
+      "grad_norm": 0.4002218724101484,
+      "learning_rate": 4.5667932343803935e-05,
+      "loss": 0.4427,
+      "step": 573
+    },
+    {
+      "epoch": 0.5349487418452936,
+      "grad_norm": 0.4398311451033267,
+      "learning_rate": 4.565067311011391e-05,
+      "loss": 0.4293,
+      "step": 574
+    },
+    {
+      "epoch": 0.5358807082945014,
+      "grad_norm": 0.41173047357400905,
+      "learning_rate": 4.563341387642389e-05,
+      "loss": 0.4244,
+      "step": 575
+    },
+    {
+      "epoch": 0.5368126747437092,
+      "grad_norm": 0.3813556268787713,
+      "learning_rate": 4.5616154642733865e-05,
+      "loss": 0.4122,
+      "step": 576
+    },
+    {
+      "epoch": 0.5377446411929171,
+      "grad_norm": 0.4446317504959361,
+      "learning_rate": 4.559889540904384e-05,
+      "loss": 0.4235,
+      "step": 577
+    },
+    {
+      "epoch": 0.5386766076421249,
+      "grad_norm": 0.37404283903197655,
+      "learning_rate": 4.5581636175353814e-05,
+      "loss": 0.4221,
+      "step": 578
+    },
+    {
+      "epoch": 0.5396085740913327,
+      "grad_norm": 0.5178603433744016,
+      "learning_rate": 4.5564376941663795e-05,
+      "loss": 0.4372,
+      "step": 579
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.38400803788816484,
+      "learning_rate": 4.554711770797377e-05,
+      "loss": 0.4456,
+      "step": 580
+    },
+    {
+      "epoch": 0.5414725069897484,
+      "grad_norm": 0.46782672701555195,
+      "learning_rate": 4.552985847428374e-05,
+      "loss": 0.436,
+      "step": 581
+    },
+    {
+      "epoch": 0.5424044734389561,
+      "grad_norm": 0.34286004074067994,
+      "learning_rate": 4.551259924059372e-05,
+      "loss": 0.4362,
+      "step": 582
+    },
+    {
+      "epoch": 0.543336439888164,
+      "grad_norm": 0.48921281046667375,
+      "learning_rate": 4.54953400069037e-05,
+      "loss": 0.397,
+      "step": 583
+    },
+    {
+      "epoch": 0.5442684063373718,
+      "grad_norm": 0.40252687961650474,
+      "learning_rate": 4.547808077321367e-05,
+      "loss": 0.437,
+      "step": 584
+    },
+    {
+      "epoch": 0.5452003727865797,
+      "grad_norm": 0.4792451494668757,
+      "learning_rate": 4.546082153952365e-05,
+      "loss": 0.4183,
+      "step": 585
+    },
+    {
+      "epoch": 0.5461323392357875,
+      "grad_norm": 0.44409660517425903,
+      "learning_rate": 4.544356230583363e-05,
+      "loss": 0.4162,
+      "step": 586
+    },
+    {
+      "epoch": 0.5470643056849953,
+      "grad_norm": 0.38342185390890976,
+      "learning_rate": 4.54263030721436e-05,
+      "loss": 0.3875,
+      "step": 587
+    },
+    {
+      "epoch": 0.5479962721342032,
+      "grad_norm": 0.47222299821120733,
+      "learning_rate": 4.540904383845358e-05,
+      "loss": 0.4134,
+      "step": 588
+    },
+    {
+      "epoch": 0.548928238583411,
+      "grad_norm": 0.40991516318870447,
+      "learning_rate": 4.539178460476355e-05,
+      "loss": 0.4684,
+      "step": 589
+    },
+    {
+      "epoch": 0.5498602050326188,
+      "grad_norm": 0.4236915699209021,
+      "learning_rate": 4.5374525371073526e-05,
+      "loss": 0.421,
+      "step": 590
+    },
+    {
+      "epoch": 0.5507921714818267,
+      "grad_norm": 0.433283728922113,
+      "learning_rate": 4.53572661373835e-05,
+      "loss": 0.4267,
+      "step": 591
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 0.420371449627224,
+      "learning_rate": 4.5340006903693474e-05,
+      "loss": 0.4116,
+      "step": 592
+    },
+    {
+      "epoch": 0.5526561043802423,
+      "grad_norm": 0.383773985173139,
+      "learning_rate": 4.532274767000345e-05,
+      "loss": 0.409,
+      "step": 593
+    },
+    {
+      "epoch": 0.5535880708294502,
+      "grad_norm": 0.34924369939738864,
+      "learning_rate": 4.530548843631343e-05,
+      "loss": 0.4395,
+      "step": 594
+    },
+    {
+      "epoch": 0.554520037278658,
+      "grad_norm": 0.3567272631453147,
+      "learning_rate": 4.5288229202623404e-05,
+      "loss": 0.4299,
+      "step": 595
+    },
+    {
+      "epoch": 0.5554520037278659,
+      "grad_norm": 0.35003964759339423,
+      "learning_rate": 4.527096996893338e-05,
+      "loss": 0.4164,
+      "step": 596
+    },
+    {
+      "epoch": 0.5563839701770736,
+      "grad_norm": 0.370231216563871,
+      "learning_rate": 4.525371073524336e-05,
+      "loss": 0.417,
+      "step": 597
+    },
+    {
+      "epoch": 0.5573159366262814,
+      "grad_norm": 0.41533339403360897,
+      "learning_rate": 4.523645150155333e-05,
+      "loss": 0.4285,
+      "step": 598
+    },
+    {
+      "epoch": 0.5582479030754893,
+      "grad_norm": 0.36147219490608906,
+      "learning_rate": 4.521919226786331e-05,
+      "loss": 0.4342,
+      "step": 599
+    },
+    {
+      "epoch": 0.5591798695246971,
+      "grad_norm": 0.4035051361166873,
+      "learning_rate": 4.520193303417328e-05,
+      "loss": 0.4224,
+      "step": 600
+    },
+    {
+      "epoch": 0.5601118359739049,
+      "grad_norm": 0.3563856781547338,
+      "learning_rate": 4.518467380048326e-05,
+      "loss": 0.4191,
+      "step": 601
+    },
+    {
+      "epoch": 0.5610438024231128,
+      "grad_norm": 0.4259734630027951,
+      "learning_rate": 4.516741456679324e-05,
+      "loss": 0.43,
+      "step": 602
+    },
+    {
+      "epoch": 0.5619757688723206,
+      "grad_norm": 0.36175135561775085,
+      "learning_rate": 4.515015533310321e-05,
+      "loss": 0.4328,
+      "step": 603
+    },
+    {
+      "epoch": 0.5629077353215284,
+      "grad_norm": 0.45140726234735357,
+      "learning_rate": 4.5132896099413186e-05,
+      "loss": 0.427,
+      "step": 604
+    },
+    {
+      "epoch": 0.5638397017707363,
+      "grad_norm": 0.38006213346358103,
+      "learning_rate": 4.511563686572317e-05,
+      "loss": 0.4032,
+      "step": 605
+    },
+    {
+      "epoch": 0.5647716682199441,
+      "grad_norm": 0.4294189457196753,
+      "learning_rate": 4.509837763203314e-05,
+      "loss": 0.4545,
+      "step": 606
+    },
+    {
+      "epoch": 0.5657036346691519,
+      "grad_norm": 0.402313252529966,
+      "learning_rate": 4.5081118398343116e-05,
+      "loss": 0.4247,
+      "step": 607
+    },
+    {
+      "epoch": 0.5666356011183598,
+      "grad_norm": 0.45540136841717044,
+      "learning_rate": 4.5063859164653097e-05,
+      "loss": 0.4475,
+      "step": 608
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 0.4314981588125385,
+      "learning_rate": 4.504659993096307e-05,
+      "loss": 0.4299,
+      "step": 609
+    },
+    {
+      "epoch": 0.5684995340167754,
+      "grad_norm": 0.43162932279779925,
+      "learning_rate": 4.5029340697273045e-05,
+      "loss": 0.4103,
+      "step": 610
+    },
+    {
+      "epoch": 0.5694315004659832,
+      "grad_norm": 0.40595725880886285,
+      "learning_rate": 4.501208146358302e-05,
+      "loss": 0.4107,
+      "step": 611
+    },
+    {
+      "epoch": 0.570363466915191,
+      "grad_norm": 0.38210778355511515,
+      "learning_rate": 4.4994822229892994e-05,
+      "loss": 0.4021,
+      "step": 612
+    },
+    {
+      "epoch": 0.5712954333643989,
+      "grad_norm": 0.3619308541461091,
+      "learning_rate": 4.497756299620297e-05,
+      "loss": 0.4425,
+      "step": 613
+    },
+    {
+      "epoch": 0.5722273998136067,
+      "grad_norm": 0.3718639886130864,
+      "learning_rate": 4.496030376251294e-05,
+      "loss": 0.4301,
+      "step": 614
+    },
+    {
+      "epoch": 0.5731593662628145,
+      "grad_norm": 0.3399591536895868,
+      "learning_rate": 4.4943044528822923e-05,
+      "loss": 0.4149,
+      "step": 615
+    },
+    {
+      "epoch": 0.5740913327120224,
+      "grad_norm": 0.3964075751386214,
+      "learning_rate": 4.49257852951329e-05,
+      "loss": 0.429,
+      "step": 616
+    },
+    {
+      "epoch": 0.5750232991612302,
+      "grad_norm": 0.3436546136448335,
+      "learning_rate": 4.490852606144287e-05,
+      "loss": 0.4076,
+      "step": 617
+    },
+    {
+      "epoch": 0.575955265610438,
+      "grad_norm": 0.49461974562935407,
+      "learning_rate": 4.4891266827752846e-05,
+      "loss": 0.4237,
+      "step": 618
+    },
+    {
+      "epoch": 0.5768872320596459,
+      "grad_norm": 0.3444615852078571,
+      "learning_rate": 4.487400759406283e-05,
+      "loss": 0.4146,
+      "step": 619
+    },
+    {
+      "epoch": 0.5778191985088537,
+      "grad_norm": 0.42703706683479153,
+      "learning_rate": 4.48567483603728e-05,
+      "loss": 0.4096,
+      "step": 620
+    },
+    {
+      "epoch": 0.5787511649580616,
+      "grad_norm": 0.4154559005111101,
+      "learning_rate": 4.4839489126682776e-05,
+      "loss": 0.4179,
+      "step": 621
+    },
+    {
+      "epoch": 0.5796831314072693,
+      "grad_norm": 0.3770163658820433,
+      "learning_rate": 4.482222989299275e-05,
+      "loss": 0.4539,
+      "step": 622
+    },
+    {
+      "epoch": 0.5806150978564771,
+      "grad_norm": 0.4426758231351413,
+      "learning_rate": 4.480497065930273e-05,
+      "loss": 0.4303,
+      "step": 623
+    },
+    {
+      "epoch": 0.581547064305685,
+      "grad_norm": 0.4408287646936583,
+      "learning_rate": 4.4787711425612706e-05,
+      "loss": 0.4401,
+      "step": 624
+    },
+    {
+      "epoch": 0.5824790307548928,
+      "grad_norm": 0.3711616320616168,
+      "learning_rate": 4.477045219192268e-05,
+      "loss": 0.452,
+      "step": 625
+    },
+    {
+      "epoch": 0.5834109972041006,
+      "grad_norm": 0.39473231897265654,
+      "learning_rate": 4.475319295823266e-05,
+      "loss": 0.4398,
+      "step": 626
+    },
+    {
+      "epoch": 0.5843429636533085,
+      "grad_norm": 0.374110685655913,
+      "learning_rate": 4.4735933724542635e-05,
+      "loss": 0.4549,
+      "step": 627
+    },
+    {
+      "epoch": 0.5852749301025163,
+      "grad_norm": 0.37787966979924636,
+      "learning_rate": 4.471867449085261e-05,
+      "loss": 0.4299,
+      "step": 628
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 0.36612034797113935,
+      "learning_rate": 4.4701415257162584e-05,
+      "loss": 0.412,
+      "step": 629
+    },
+    {
+      "epoch": 0.587138863000932,
+      "grad_norm": 0.384624517935882,
+      "learning_rate": 4.4684156023472565e-05,
+      "loss": 0.4108,
+      "step": 630
+    },
+    {
+      "epoch": 0.5880708294501398,
+      "grad_norm": 0.3687861037531805,
+      "learning_rate": 4.466689678978254e-05,
+      "loss": 0.4199,
+      "step": 631
+    },
+    {
+      "epoch": 0.5890027958993477,
+      "grad_norm": 0.47623476053224784,
+      "learning_rate": 4.4649637556092513e-05,
+      "loss": 0.4233,
+      "step": 632
+    },
+    {
+      "epoch": 0.5899347623485555,
+      "grad_norm": 0.41455951861528145,
+      "learning_rate": 4.463237832240249e-05,
+      "loss": 0.4296,
+      "step": 633
+    },
+    {
+      "epoch": 0.5908667287977633,
+      "grad_norm": 0.47018802555445816,
+      "learning_rate": 4.461511908871246e-05,
+      "loss": 0.452,
+      "step": 634
+    },
+    {
+      "epoch": 0.5917986952469712,
+      "grad_norm": 0.47406416923911754,
+      "learning_rate": 4.4597859855022436e-05,
+      "loss": 0.4446,
+      "step": 635
+    },
+    {
+      "epoch": 0.5927306616961789,
+      "grad_norm": 0.5075083122289654,
+      "learning_rate": 4.458060062133241e-05,
+      "loss": 0.4246,
+      "step": 636
+    },
+    {
+      "epoch": 0.5936626281453867,
+      "grad_norm": 0.3612321153691148,
+      "learning_rate": 4.456334138764239e-05,
+      "loss": 0.4472,
+      "step": 637
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 0.5098864113567897,
+      "learning_rate": 4.4546082153952366e-05,
+      "loss": 0.4189,
+      "step": 638
+    },
+    {
+      "epoch": 0.5955265610438024,
+      "grad_norm": 0.42402871877599874,
+      "learning_rate": 4.452882292026234e-05,
+      "loss": 0.4271,
+      "step": 639
+    },
+    {
+      "epoch": 0.5964585274930102,
+      "grad_norm": 0.5032752315526783,
+      "learning_rate": 4.4511563686572315e-05,
+      "loss": 0.4062,
+      "step": 640
+    },
+    {
+      "epoch": 0.5973904939422181,
+      "grad_norm": 0.3360296191274446,
+      "learning_rate": 4.4494304452882296e-05,
+      "loss": 0.4198,
+      "step": 641
+    },
+    {
+      "epoch": 0.5983224603914259,
+      "grad_norm": 0.3880659519689031,
+      "learning_rate": 4.447704521919227e-05,
+      "loss": 0.443,
+      "step": 642
+    },
+    {
+      "epoch": 0.5992544268406338,
+      "grad_norm": 0.3862931798670413,
+      "learning_rate": 4.4459785985502244e-05,
+      "loss": 0.4271,
+      "step": 643
+    },
+    {
+      "epoch": 0.6001863932898416,
+      "grad_norm": 0.3153120071290221,
+      "learning_rate": 4.444252675181222e-05,
+      "loss": 0.4175,
+      "step": 644
+    },
+    {
+      "epoch": 0.6011183597390494,
+      "grad_norm": 0.4239178447516735,
+      "learning_rate": 4.44252675181222e-05,
+      "loss": 0.4144,
+      "step": 645
+    },
+    {
+      "epoch": 0.6020503261882573,
+      "grad_norm": 0.40193239446332973,
+      "learning_rate": 4.4408008284432174e-05,
+      "loss": 0.4371,
+      "step": 646
+    },
+    {
+      "epoch": 0.6029822926374651,
+      "grad_norm": 0.387549663249363,
+      "learning_rate": 4.439074905074215e-05,
+      "loss": 0.4443,
+      "step": 647
+    },
+    {
+      "epoch": 0.6039142590866728,
+      "grad_norm": 0.3414180104018037,
+      "learning_rate": 4.437348981705213e-05,
+      "loss": 0.4454,
+      "step": 648
+    },
+    {
+      "epoch": 0.6048462255358807,
+      "grad_norm": 0.3354059050972693,
+      "learning_rate": 4.4356230583362103e-05,
+      "loss": 0.4151,
+      "step": 649
+    },
+    {
+      "epoch": 0.6057781919850885,
+      "grad_norm": 0.4265893143265595,
+      "learning_rate": 4.433897134967208e-05,
+      "loss": 0.424,
+      "step": 650
+    },
+    {
+      "epoch": 0.6067101584342963,
+      "grad_norm": 0.3800293337083674,
+      "learning_rate": 4.432171211598205e-05,
+      "loss": 0.4438,
+      "step": 651
+    },
+    {
+      "epoch": 0.6076421248835042,
+      "grad_norm": 0.3800957314642472,
+      "learning_rate": 4.430445288229203e-05,
+      "loss": 0.4337,
+      "step": 652
+    },
+    {
+      "epoch": 0.608574091332712,
+      "grad_norm": 0.37289060512734223,
+      "learning_rate": 4.428719364860201e-05,
+      "loss": 0.4176,
+      "step": 653
+    },
+    {
+      "epoch": 0.6095060577819198,
+      "grad_norm": 0.37534604882414874,
+      "learning_rate": 4.426993441491198e-05,
+      "loss": 0.4117,
+      "step": 654
+    },
+    {
+      "epoch": 0.6104380242311277,
+      "grad_norm": 0.4089914870843825,
+      "learning_rate": 4.4252675181221956e-05,
+      "loss": 0.4461,
+      "step": 655
+    },
+    {
+      "epoch": 0.6113699906803355,
+      "grad_norm": 0.3376088478707045,
+      "learning_rate": 4.423541594753193e-05,
+      "loss": 0.4226,
+      "step": 656
+    },
+    {
+      "epoch": 0.6123019571295434,
+      "grad_norm": 0.40436771796278753,
+      "learning_rate": 4.4218156713841905e-05,
+      "loss": 0.4288,
+      "step": 657
+    },
+    {
+      "epoch": 0.6132339235787512,
+      "grad_norm": 0.32452872858245196,
+      "learning_rate": 4.420089748015188e-05,
+      "loss": 0.3972,
+      "step": 658
+    },
+    {
+      "epoch": 0.614165890027959,
+      "grad_norm": 0.3715418431905969,
+      "learning_rate": 4.418363824646186e-05,
+      "loss": 0.3916,
+      "step": 659
+    },
+    {
+      "epoch": 0.6150978564771669,
+      "grad_norm": 0.34846232197714344,
+      "learning_rate": 4.4166379012771834e-05,
+      "loss": 0.4218,
+      "step": 660
+    },
+    {
+      "epoch": 0.6160298229263746,
+      "grad_norm": 0.40721154164947476,
+      "learning_rate": 4.414911977908181e-05,
+      "loss": 0.3973,
+      "step": 661
+    },
+    {
+      "epoch": 0.6169617893755824,
+      "grad_norm": 0.37624852446420953,
+      "learning_rate": 4.413186054539178e-05,
+      "loss": 0.4019,
+      "step": 662
+    },
+    {
+      "epoch": 0.6178937558247903,
+      "grad_norm": 0.47151020797029203,
+      "learning_rate": 4.4114601311701764e-05,
+      "loss": 0.4293,
+      "step": 663
+    },
+    {
+      "epoch": 0.6188257222739981,
+      "grad_norm": 0.4190919230666174,
+      "learning_rate": 4.409734207801174e-05,
+      "loss": 0.4212,
+      "step": 664
+    },
+    {
+      "epoch": 0.6197576887232059,
+      "grad_norm": 0.4596711910891454,
+      "learning_rate": 4.408008284432171e-05,
+      "loss": 0.4161,
+      "step": 665
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.3734958110314102,
+      "learning_rate": 4.4062823610631694e-05,
+      "loss": 0.4234,
+      "step": 666
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 0.4398863866103752,
+      "learning_rate": 4.404556437694167e-05,
+      "loss": 0.4116,
+      "step": 667
+    },
+    {
+      "epoch": 0.6225535880708295,
+      "grad_norm": 0.3572506133971031,
+      "learning_rate": 4.402830514325164e-05,
+      "loss": 0.3962,
+      "step": 668
+    },
+    {
+      "epoch": 0.6234855545200373,
+      "grad_norm": 0.45838665755343144,
+      "learning_rate": 4.4011045909561616e-05,
+      "loss": 0.4262,
+      "step": 669
+    },
+    {
+      "epoch": 0.6244175209692451,
+      "grad_norm": 0.3931181260088297,
+      "learning_rate": 4.39937866758716e-05,
+      "loss": 0.4015,
+      "step": 670
+    },
+    {
+      "epoch": 0.625349487418453,
+      "grad_norm": 0.4148633692525596,
+      "learning_rate": 4.397652744218157e-05,
+      "loss": 0.4309,
+      "step": 671
+    },
+    {
+      "epoch": 0.6262814538676608,
+      "grad_norm": 0.5091016598307129,
+      "learning_rate": 4.3959268208491546e-05,
+      "loss": 0.4366,
+      "step": 672
+    },
+    {
+      "epoch": 0.6272134203168686,
+      "grad_norm": 0.40480450902557596,
+      "learning_rate": 4.394200897480152e-05,
+      "loss": 0.4176,
+      "step": 673
+    },
+    {
+      "epoch": 0.6281453867660765,
+      "grad_norm": 0.44026491469265977,
+      "learning_rate": 4.39247497411115e-05,
+      "loss": 0.4134,
+      "step": 674
+    },
+    {
+      "epoch": 0.6290773532152842,
+      "grad_norm": 0.40811225238105475,
+      "learning_rate": 4.3907490507421476e-05,
+      "loss": 0.4028,
+      "step": 675
+    },
+    {
+      "epoch": 0.630009319664492,
+      "grad_norm": 0.405457935023565,
+      "learning_rate": 4.389023127373145e-05,
+      "loss": 0.4294,
+      "step": 676
+    },
+    {
+      "epoch": 0.6309412861136999,
+      "grad_norm": 0.46163016352945635,
+      "learning_rate": 4.3872972040041424e-05,
+      "loss": 0.4068,
+      "step": 677
+    },
+    {
+      "epoch": 0.6318732525629077,
+      "grad_norm": 0.40459637340243215,
+      "learning_rate": 4.38557128063514e-05,
+      "loss": 0.3995,
+      "step": 678
+    },
+    {
+      "epoch": 0.6328052190121156,
+      "grad_norm": 0.49159854916711204,
+      "learning_rate": 4.383845357266137e-05,
+      "loss": 0.4312,
+      "step": 679
+    },
+    {
+      "epoch": 0.6337371854613234,
+      "grad_norm": 0.49069759780539246,
+      "learning_rate": 4.382119433897135e-05,
+      "loss": 0.4508,
+      "step": 680
+    },
+    {
+      "epoch": 0.6346691519105312,
+      "grad_norm": 0.4663541745513591,
+      "learning_rate": 4.380393510528133e-05,
+      "loss": 0.4478,
+      "step": 681
+    },
+    {
+      "epoch": 0.6356011183597391,
+      "grad_norm": 0.5334307531939046,
+      "learning_rate": 4.37866758715913e-05,
+      "loss": 0.427,
+      "step": 682
+    },
+    {
+      "epoch": 0.6365330848089469,
+      "grad_norm": 0.5923647303116868,
+      "learning_rate": 4.376941663790128e-05,
+      "loss": 0.4335,
+      "step": 683
+    },
+    {
+      "epoch": 0.6374650512581547,
+      "grad_norm": 0.4097979761346584,
+      "learning_rate": 4.375215740421125e-05,
+      "loss": 0.4405,
+      "step": 684
+    },
+    {
+      "epoch": 0.6383970177073626,
+      "grad_norm": 0.5541357612861003,
+      "learning_rate": 4.373489817052123e-05,
+      "loss": 0.4165,
+      "step": 685
+    },
+    {
+      "epoch": 0.6393289841565704,
+      "grad_norm": 0.3486622875007531,
+      "learning_rate": 4.3717638936831206e-05,
+      "loss": 0.4152,
+      "step": 686
+    },
+    {
+      "epoch": 0.6402609506057781,
+      "grad_norm": 0.5167847494207316,
+      "learning_rate": 4.370037970314118e-05,
+      "loss": 0.4326,
+      "step": 687
+    },
+    {
+      "epoch": 0.641192917054986,
+      "grad_norm": 0.3868678826591993,
+      "learning_rate": 4.368312046945116e-05,
+      "loss": 0.4094,
+      "step": 688
+    },
+    {
+      "epoch": 0.6421248835041938,
+      "grad_norm": 0.5197907672434207,
+      "learning_rate": 4.3665861235761136e-05,
+      "loss": 0.4209,
+      "step": 689
+    },
+    {
+      "epoch": 0.6430568499534017,
+      "grad_norm": 0.48966550386405167,
+      "learning_rate": 4.364860200207111e-05,
+      "loss": 0.3981,
+      "step": 690
+    },
+    {
+      "epoch": 0.6439888164026095,
+      "grad_norm": 0.5283979145307307,
+      "learning_rate": 4.3631342768381085e-05,
+      "loss": 0.4298,
+      "step": 691
+    },
+    {
+      "epoch": 0.6449207828518173,
+      "grad_norm": 0.5844074225002355,
+      "learning_rate": 4.3614083534691066e-05,
+      "loss": 0.409,
+      "step": 692
+    },
+    {
+      "epoch": 0.6458527493010252,
+      "grad_norm": 0.37968180810766233,
+      "learning_rate": 4.359682430100104e-05,
+      "loss": 0.4186,
+      "step": 693
+    },
+    {
+      "epoch": 0.646784715750233,
+      "grad_norm": 0.5032846091834392,
+      "learning_rate": 4.3579565067311014e-05,
+      "loss": 0.4315,
+      "step": 694
+    },
+    {
+      "epoch": 0.6477166821994408,
+      "grad_norm": 0.3296782216896817,
+      "learning_rate": 4.356230583362099e-05,
+      "loss": 0.412,
+      "step": 695
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.5252624490150228,
+      "learning_rate": 4.354504659993097e-05,
+      "loss": 0.451,
+      "step": 696
+    },
+    {
+      "epoch": 0.6495806150978565,
+      "grad_norm": 0.4386665512997432,
+      "learning_rate": 4.3527787366240944e-05,
+      "loss": 0.4305,
+      "step": 697
+    },
+    {
+      "epoch": 0.6505125815470643,
+      "grad_norm": 0.5727280403540933,
+      "learning_rate": 4.351052813255092e-05,
+      "loss": 0.4257,
+      "step": 698
+    },
+    {
+      "epoch": 0.6514445479962722,
+      "grad_norm": 0.49141943639588453,
+      "learning_rate": 4.349326889886089e-05,
+      "loss": 0.4292,
+      "step": 699
+    },
+    {
+      "epoch": 0.65237651444548,
+      "grad_norm": 0.35395964134813224,
+      "learning_rate": 4.347600966517087e-05,
+      "loss": 0.4084,
+      "step": 700
+    },
+    {
+      "epoch": 0.6533084808946877,
+      "grad_norm": 0.4413159900321269,
+      "learning_rate": 4.345875043148084e-05,
+      "loss": 0.3929,
+      "step": 701
+    },
+    {
+      "epoch": 0.6542404473438956,
+      "grad_norm": 0.42575067926536575,
+      "learning_rate": 4.3441491197790815e-05,
+      "loss": 0.4564,
+      "step": 702
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 0.4303851452883037,
+      "learning_rate": 4.3424231964100796e-05,
+      "loss": 0.4455,
+      "step": 703
+    },
+    {
+      "epoch": 0.6561043802423113,
+      "grad_norm": 0.48750427605245117,
+      "learning_rate": 4.340697273041077e-05,
+      "loss": 0.4217,
+      "step": 704
+    },
+    {
+      "epoch": 0.6570363466915191,
+      "grad_norm": 0.36940449248771257,
+      "learning_rate": 4.3389713496720745e-05,
+      "loss": 0.3953,
+      "step": 705
+    },
+    {
+      "epoch": 0.6579683131407269,
+      "grad_norm": 0.3823465428617259,
+      "learning_rate": 4.3372454263030726e-05,
+      "loss": 0.3999,
+      "step": 706
+    },
+    {
+      "epoch": 0.6589002795899348,
+      "grad_norm": 0.4071058491151902,
+      "learning_rate": 4.33551950293407e-05,
+      "loss": 0.4401,
+      "step": 707
+    },
+    {
+      "epoch": 0.6598322460391426,
+      "grad_norm": 0.38913577516566694,
+      "learning_rate": 4.3337935795650675e-05,
+      "loss": 0.4195,
+      "step": 708
+    },
+    {
+      "epoch": 0.6607642124883504,
+      "grad_norm": 0.38841152450555777,
+      "learning_rate": 4.332067656196065e-05,
+      "loss": 0.4243,
+      "step": 709
+    },
+    {
+      "epoch": 0.6616961789375583,
+      "grad_norm": 0.4504485615934782,
+      "learning_rate": 4.330341732827063e-05,
+      "loss": 0.4307,
+      "step": 710
+    },
+    {
+      "epoch": 0.6626281453867661,
+      "grad_norm": 0.39839585076006234,
+      "learning_rate": 4.3286158094580604e-05,
+      "loss": 0.4216,
+      "step": 711
+    },
+    {
+      "epoch": 0.6635601118359739,
+      "grad_norm": 0.45473869144828954,
+      "learning_rate": 4.326889886089058e-05,
+      "loss": 0.4055,
+      "step": 712
+    },
+    {
+      "epoch": 0.6644920782851818,
+      "grad_norm": 0.37147348144640285,
+      "learning_rate": 4.325163962720055e-05,
+      "loss": 0.4117,
+      "step": 713
+    },
+    {
+      "epoch": 0.6654240447343895,
+      "grad_norm": 0.40920344207041254,
+      "learning_rate": 4.3234380393510534e-05,
+      "loss": 0.4017,
+      "step": 714
+    },
+    {
+      "epoch": 0.6663560111835974,
+      "grad_norm": 0.4131469656670445,
+      "learning_rate": 4.321712115982051e-05,
+      "loss": 0.4155,
+      "step": 715
+    },
+    {
+      "epoch": 0.6672879776328052,
+      "grad_norm": 0.33255772086083724,
+      "learning_rate": 4.319986192613048e-05,
+      "loss": 0.4148,
+      "step": 716
+    },
+    {
+      "epoch": 0.668219944082013,
+      "grad_norm": 0.4356566649359628,
+      "learning_rate": 4.3182602692440464e-05,
+      "loss": 0.4157,
+      "step": 717
+    },
+    {
+      "epoch": 0.6691519105312209,
+      "grad_norm": 0.421952814281185,
+      "learning_rate": 4.316534345875044e-05,
+      "loss": 0.4385,
+      "step": 718
+    },
+    {
+      "epoch": 0.6700838769804287,
+      "grad_norm": 0.4345247529629237,
+      "learning_rate": 4.314808422506041e-05,
+      "loss": 0.4069,
+      "step": 719
+    },
+    {
+      "epoch": 0.6710158434296365,
+      "grad_norm": 0.4093168450158888,
+      "learning_rate": 4.3130824991370387e-05,
+      "loss": 0.4329,
+      "step": 720
+    },
+    {
+      "epoch": 0.6719478098788444,
+      "grad_norm": 0.453138918519175,
+      "learning_rate": 4.311356575768036e-05,
+      "loss": 0.4416,
+      "step": 721
+    },
+    {
+      "epoch": 0.6728797763280522,
+      "grad_norm": 0.36044552842482386,
+      "learning_rate": 4.3096306523990335e-05,
+      "loss": 0.3954,
+      "step": 722
+    },
+    {
+      "epoch": 0.67381174277726,
+      "grad_norm": 0.4366086883143347,
+      "learning_rate": 4.307904729030031e-05,
+      "loss": 0.4306,
+      "step": 723
+    },
+    {
+      "epoch": 0.6747437092264679,
+      "grad_norm": 0.34997797971810735,
+      "learning_rate": 4.3061788056610284e-05,
+      "loss": 0.4227,
+      "step": 724
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 0.335869095555944,
+      "learning_rate": 4.3044528822920265e-05,
+      "loss": 0.4118,
+      "step": 725
+    },
+    {
+      "epoch": 0.6766076421248836,
+      "grad_norm": 0.4015190593241834,
+      "learning_rate": 4.302726958923024e-05,
+      "loss": 0.4167,
+      "step": 726
+    },
+    {
+      "epoch": 0.6775396085740913,
+      "grad_norm": 0.43879972167625153,
+      "learning_rate": 4.301001035554021e-05,
+      "loss": 0.4238,
+      "step": 727
+    },
+    {
+      "epoch": 0.6784715750232991,
+      "grad_norm": 0.34321500633159185,
+      "learning_rate": 4.2992751121850194e-05,
+      "loss": 0.4056,
+      "step": 728
+    },
+    {
+      "epoch": 0.679403541472507,
+      "grad_norm": 0.43553832112773333,
+      "learning_rate": 4.297549188816017e-05,
+      "loss": 0.4051,
+      "step": 729
+    },
+    {
+      "epoch": 0.6803355079217148,
+      "grad_norm": 0.44060579566703945,
+      "learning_rate": 4.295823265447014e-05,
+      "loss": 0.4259,
+      "step": 730
+    },
+    {
+      "epoch": 0.6812674743709226,
+      "grad_norm": 0.40089468475214146,
+      "learning_rate": 4.294097342078012e-05,
+      "loss": 0.4081,
+      "step": 731
+    },
+    {
+      "epoch": 0.6821994408201305,
+      "grad_norm": 0.3825317459905697,
+      "learning_rate": 4.29237141870901e-05,
+      "loss": 0.4164,
+      "step": 732
+    },
+    {
+      "epoch": 0.6831314072693383,
+      "grad_norm": 0.3873014086126696,
+      "learning_rate": 4.290645495340007e-05,
+      "loss": 0.425,
+      "step": 733
+    },
+    {
+      "epoch": 0.6840633737185461,
+      "grad_norm": 0.3722724386236482,
+      "learning_rate": 4.288919571971005e-05,
+      "loss": 0.4211,
+      "step": 734
+    },
+    {
+      "epoch": 0.684995340167754,
+      "grad_norm": 0.3734334026834433,
+      "learning_rate": 4.287193648602002e-05,
+      "loss": 0.4259,
+      "step": 735
+    },
+    {
+      "epoch": 0.6859273066169618,
+      "grad_norm": 0.36648912756124885,
+      "learning_rate": 4.285467725233e-05,
+      "loss": 0.4225,
+      "step": 736
+    },
+    {
+      "epoch": 0.6868592730661697,
+      "grad_norm": 0.39183992476871443,
+      "learning_rate": 4.2837418018639977e-05,
+      "loss": 0.423,
+      "step": 737
+    },
+    {
+      "epoch": 0.6877912395153775,
+      "grad_norm": 0.3752209069134191,
+      "learning_rate": 4.282015878494995e-05,
+      "loss": 0.3957,
+      "step": 738
+    },
+    {
+      "epoch": 0.6887232059645852,
+      "grad_norm": 0.3520982418351287,
+      "learning_rate": 4.2802899551259925e-05,
+      "loss": 0.4286,
+      "step": 739
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 0.3401875523409571,
+      "learning_rate": 4.27856403175699e-05,
+      "loss": 0.3943,
+      "step": 740
+    },
+    {
+      "epoch": 0.6905871388630009,
+      "grad_norm": 0.3310729653719841,
+      "learning_rate": 4.2768381083879874e-05,
+      "loss": 0.3997,
+      "step": 741
+    },
+    {
+      "epoch": 0.6915191053122087,
+      "grad_norm": 0.35076372509719644,
+      "learning_rate": 4.2751121850189855e-05,
+      "loss": 0.4021,
+      "step": 742
+    },
+    {
+      "epoch": 0.6924510717614166,
+      "grad_norm": 0.3560696386509464,
+      "learning_rate": 4.273386261649983e-05,
+      "loss": 0.3957,
+      "step": 743
+    },
+    {
+      "epoch": 0.6933830382106244,
+      "grad_norm": 0.3416570723254711,
+      "learning_rate": 4.27166033828098e-05,
+      "loss": 0.421,
+      "step": 744
+    },
+    {
+      "epoch": 0.6943150046598322,
+      "grad_norm": 0.37413573670752376,
+      "learning_rate": 4.269934414911978e-05,
+      "loss": 0.4248,
+      "step": 745
+    },
+    {
+      "epoch": 0.6952469711090401,
+      "grad_norm": 0.400774907092695,
+      "learning_rate": 4.268208491542976e-05,
+      "loss": 0.3894,
+      "step": 746
+    },
+    {
+      "epoch": 0.6961789375582479,
+      "grad_norm": 0.3658213562708208,
+      "learning_rate": 4.266482568173973e-05,
+      "loss": 0.4224,
+      "step": 747
+    },
+    {
+      "epoch": 0.6971109040074557,
+      "grad_norm": 0.46452600827004503,
+      "learning_rate": 4.264756644804971e-05,
+      "loss": 0.4163,
+      "step": 748
+    },
+    {
+      "epoch": 0.6980428704566636,
+      "grad_norm": 0.3775663945904702,
+      "learning_rate": 4.263030721435968e-05,
+      "loss": 0.4278,
+      "step": 749
+    },
+    {
+      "epoch": 0.6989748369058714,
+      "grad_norm": 0.4396098538887077,
+      "learning_rate": 4.261304798066966e-05,
+      "loss": 0.4116,
+      "step": 750
+    },
+    {
+      "epoch": 0.6999068033550793,
+      "grad_norm": 0.34097674895224267,
+      "learning_rate": 4.259578874697964e-05,
+      "loss": 0.4266,
+      "step": 751
+    },
+    {
+      "epoch": 0.700838769804287,
+      "grad_norm": 0.38192432209329114,
+      "learning_rate": 4.257852951328961e-05,
+      "loss": 0.4025,
+      "step": 752
+    },
+    {
+      "epoch": 0.7017707362534948,
+      "grad_norm": 0.49168138360189895,
+      "learning_rate": 4.2561270279599586e-05,
+      "loss": 0.4144,
+      "step": 753
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 0.39101345123365977,
+      "learning_rate": 4.2544011045909567e-05,
+      "loss": 0.4087,
+      "step": 754
+    },
+    {
+      "epoch": 0.7036346691519105,
+      "grad_norm": 0.3655712430370681,
+      "learning_rate": 4.252675181221954e-05,
+      "loss": 0.4269,
+      "step": 755
+    },
+    {
+      "epoch": 0.7045666356011183,
+      "grad_norm": 0.3996852244456727,
+      "learning_rate": 4.2509492578529515e-05,
+      "loss": 0.4088,
+      "step": 756
+    },
+    {
+      "epoch": 0.7054986020503262,
+      "grad_norm": 0.3872269904191273,
+      "learning_rate": 4.2492233344839496e-05,
+      "loss": 0.4172,
+      "step": 757
+    },
+    {
+      "epoch": 0.706430568499534,
+      "grad_norm": 0.43648292678565426,
+      "learning_rate": 4.247497411114947e-05,
+      "loss": 0.4368,
+      "step": 758
+    },
+    {
+      "epoch": 0.7073625349487418,
+      "grad_norm": 0.3896168955891061,
+      "learning_rate": 4.2457714877459445e-05,
+      "loss": 0.4226,
+      "step": 759
+    },
+    {
+      "epoch": 0.7082945013979497,
+      "grad_norm": 0.4311331738224165,
+      "learning_rate": 4.244045564376942e-05,
+      "loss": 0.4102,
+      "step": 760
+    },
+    {
+      "epoch": 0.7092264678471575,
+      "grad_norm": 0.4140459758539865,
+      "learning_rate": 4.242319641007939e-05,
+      "loss": 0.4108,
+      "step": 761
+    },
+    {
+      "epoch": 0.7101584342963654,
+      "grad_norm": 0.37666194793616814,
+      "learning_rate": 4.240593717638937e-05,
+      "loss": 0.4218,
+      "step": 762
+    },
+    {
+      "epoch": 0.7110904007455732,
+      "grad_norm": 0.4049479260851559,
+      "learning_rate": 4.238867794269934e-05,
+      "loss": 0.3791,
+      "step": 763
+    },
+    {
+      "epoch": 0.712022367194781,
+      "grad_norm": 0.41252641523712275,
+      "learning_rate": 4.2371418709009316e-05,
+      "loss": 0.4007,
+      "step": 764
+    },
+    {
+      "epoch": 0.7129543336439889,
+      "grad_norm": 0.37325436146909047,
+      "learning_rate": 4.23541594753193e-05,
+      "loss": 0.4024,
+      "step": 765
+    },
+    {
+      "epoch": 0.7138863000931966,
+      "grad_norm": 0.38742995681389075,
+      "learning_rate": 4.233690024162927e-05,
+      "loss": 0.4302,
+      "step": 766
+    },
+    {
+      "epoch": 0.7148182665424044,
+      "grad_norm": 0.3661125176611155,
+      "learning_rate": 4.2319641007939246e-05,
+      "loss": 0.3978,
+      "step": 767
+    },
+    {
+      "epoch": 0.7157502329916123,
+      "grad_norm": 0.43886569109245505,
+      "learning_rate": 4.230238177424923e-05,
+      "loss": 0.409,
+      "step": 768
+    },
+    {
+      "epoch": 0.7166821994408201,
+      "grad_norm": 0.39375730028985007,
+      "learning_rate": 4.22851225405592e-05,
+      "loss": 0.4196,
+      "step": 769
+    },
+    {
+      "epoch": 0.7176141658900279,
+      "grad_norm": 0.3501611315102492,
+      "learning_rate": 4.2267863306869176e-05,
+      "loss": 0.4051,
+      "step": 770
+    },
+    {
+      "epoch": 0.7185461323392358,
+      "grad_norm": 0.40051037834718073,
+      "learning_rate": 4.225060407317915e-05,
+      "loss": 0.4203,
+      "step": 771
+    },
+    {
+      "epoch": 0.7194780987884436,
+      "grad_norm": 0.4039559298487521,
+      "learning_rate": 4.223334483948913e-05,
+      "loss": 0.4419,
+      "step": 772
+    },
+    {
+      "epoch": 0.7204100652376515,
+      "grad_norm": 0.36354582454050494,
+      "learning_rate": 4.2216085605799105e-05,
+      "loss": 0.4148,
+      "step": 773
+    },
+    {
+      "epoch": 0.7213420316868593,
+      "grad_norm": 0.3755007317400082,
+      "learning_rate": 4.219882637210908e-05,
+      "loss": 0.4294,
+      "step": 774
+    },
+    {
+      "epoch": 0.7222739981360671,
+      "grad_norm": 0.38857413058960916,
+      "learning_rate": 4.2181567138419054e-05,
+      "loss": 0.4179,
+      "step": 775
+    },
+    {
+      "epoch": 0.723205964585275,
+      "grad_norm": 0.3872910979230129,
+      "learning_rate": 4.2164307904729035e-05,
+      "loss": 0.4242,
+      "step": 776
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.4078993389096558,
+      "learning_rate": 4.214704867103901e-05,
+      "loss": 0.4016,
+      "step": 777
+    },
+    {
+      "epoch": 0.7250698974836906,
+      "grad_norm": 0.4180339945434511,
+      "learning_rate": 4.2129789437348983e-05,
+      "loss": 0.4231,
+      "step": 778
+    },
+    {
+      "epoch": 0.7260018639328985,
+      "grad_norm": 0.36588314624210155,
+      "learning_rate": 4.2112530203658964e-05,
+      "loss": 0.4124,
+      "step": 779
+    },
+    {
+      "epoch": 0.7269338303821062,
+      "grad_norm": 0.3837229238059226,
+      "learning_rate": 4.209527096996894e-05,
+      "loss": 0.3857,
+      "step": 780
+    },
+    {
+      "epoch": 0.727865796831314,
+      "grad_norm": 0.3527945575871195,
+      "learning_rate": 4.207801173627891e-05,
+      "loss": 0.4169,
+      "step": 781
+    },
+    {
+      "epoch": 0.7287977632805219,
+      "grad_norm": 0.44246263028860966,
+      "learning_rate": 4.206075250258889e-05,
+      "loss": 0.4148,
+      "step": 782
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 0.40166554356035566,
+      "learning_rate": 4.204349326889886e-05,
+      "loss": 0.4162,
+      "step": 783
+    },
+    {
+      "epoch": 0.7306616961789375,
+      "grad_norm": 0.37360750202033177,
+      "learning_rate": 4.2026234035208836e-05,
+      "loss": 0.418,
+      "step": 784
+    },
+    {
+      "epoch": 0.7315936626281454,
+      "grad_norm": 0.3924208042478179,
+      "learning_rate": 4.200897480151881e-05,
+      "loss": 0.4172,
+      "step": 785
+    },
+    {
+      "epoch": 0.7325256290773532,
+      "grad_norm": 0.3866173406680904,
+      "learning_rate": 4.1991715567828785e-05,
+      "loss": 0.418,
+      "step": 786
+    },
+    {
+      "epoch": 0.7334575955265611,
+      "grad_norm": 0.40755569483722753,
+      "learning_rate": 4.1974456334138766e-05,
+      "loss": 0.403,
+      "step": 787
+    },
+    {
+      "epoch": 0.7343895619757689,
+      "grad_norm": 0.41476727512692035,
+      "learning_rate": 4.195719710044874e-05,
+      "loss": 0.4211,
+      "step": 788
+    },
+    {
+      "epoch": 0.7353215284249767,
+      "grad_norm": 0.48613047518705843,
+      "learning_rate": 4.1939937866758714e-05,
+      "loss": 0.4347,
+      "step": 789
+    },
+    {
+      "epoch": 0.7362534948741846,
+      "grad_norm": 0.4279464625766419,
+      "learning_rate": 4.1922678633068695e-05,
+      "loss": 0.4227,
+      "step": 790
+    },
+    {
+      "epoch": 0.7371854613233924,
+      "grad_norm": 0.5757496933939752,
+      "learning_rate": 4.190541939937867e-05,
+      "loss": 0.4205,
+      "step": 791
+    },
+    {
+      "epoch": 0.7381174277726001,
+      "grad_norm": 0.43776980677452537,
+      "learning_rate": 4.1888160165688644e-05,
+      "loss": 0.4317,
+      "step": 792
+    },
+    {
+      "epoch": 0.739049394221808,
+      "grad_norm": 0.44172491067551006,
+      "learning_rate": 4.187090093199862e-05,
+      "loss": 0.4203,
+      "step": 793
+    },
+    {
+      "epoch": 0.7399813606710158,
+      "grad_norm": 0.40247329545666993,
+      "learning_rate": 4.18536416983086e-05,
+      "loss": 0.3958,
+      "step": 794
+    },
+    {
+      "epoch": 0.7409133271202236,
+      "grad_norm": 0.43396874811584135,
+      "learning_rate": 4.1836382464618573e-05,
+      "loss": 0.4029,
+      "step": 795
+    },
+    {
+      "epoch": 0.7418452935694315,
+      "grad_norm": 0.3970639390146479,
+      "learning_rate": 4.181912323092855e-05,
+      "loss": 0.4037,
+      "step": 796
+    },
+    {
+      "epoch": 0.7427772600186393,
+      "grad_norm": 0.4249758755676771,
+      "learning_rate": 4.180186399723853e-05,
+      "loss": 0.4282,
+      "step": 797
+    },
+    {
+      "epoch": 0.7437092264678472,
+      "grad_norm": 0.3733510033238648,
+      "learning_rate": 4.17846047635485e-05,
+      "loss": 0.412,
+      "step": 798
+    },
+    {
+      "epoch": 0.744641192917055,
+      "grad_norm": 0.49417075173748565,
+      "learning_rate": 4.176734552985848e-05,
+      "loss": 0.4149,
+      "step": 799
+    },
+    {
+      "epoch": 0.7455731593662628,
+      "grad_norm": 0.4369941234526107,
+      "learning_rate": 4.175008629616845e-05,
+      "loss": 0.4031,
+      "step": 800
+    },
+    {
+      "epoch": 0.7465051258154707,
+      "grad_norm": 0.3964285273200971,
+      "learning_rate": 4.173282706247843e-05,
+      "loss": 0.4142,
+      "step": 801
+    },
+    {
+      "epoch": 0.7474370922646785,
+      "grad_norm": 0.46883690024021846,
+      "learning_rate": 4.171556782878841e-05,
+      "loss": 0.4076,
+      "step": 802
+    },
+    {
+      "epoch": 0.7483690587138863,
+      "grad_norm": 0.38259349349412136,
+      "learning_rate": 4.169830859509838e-05,
+      "loss": 0.4207,
+      "step": 803
+    },
+    {
+      "epoch": 0.7493010251630942,
+      "grad_norm": 0.48693831790003533,
+      "learning_rate": 4.1681049361408356e-05,
+      "loss": 0.4111,
+      "step": 804
+    },
+    {
+      "epoch": 0.750232991612302,
+      "grad_norm": 0.3755336619435926,
+      "learning_rate": 4.166379012771833e-05,
+      "loss": 0.4074,
+      "step": 805
+    },
+    {
+      "epoch": 0.7511649580615097,
+      "grad_norm": 0.4321691638326607,
+      "learning_rate": 4.1646530894028304e-05,
+      "loss": 0.4286,
+      "step": 806
+    },
+    {
+      "epoch": 0.7520969245107176,
+      "grad_norm": 0.3696936424789855,
+      "learning_rate": 4.162927166033828e-05,
+      "loss": 0.4241,
+      "step": 807
+    },
+    {
+      "epoch": 0.7530288909599254,
+      "grad_norm": 0.3202125924243289,
+      "learning_rate": 4.161201242664826e-05,
+      "loss": 0.4399,
+      "step": 808
+    },
+    {
+      "epoch": 0.7539608574091333,
+      "grad_norm": 0.4371269996216592,
+      "learning_rate": 4.1594753192958234e-05,
+      "loss": 0.4241,
+      "step": 809
+    },
+    {
+      "epoch": 0.7548928238583411,
+      "grad_norm": 0.3460145623736894,
+      "learning_rate": 4.157749395926821e-05,
+      "loss": 0.435,
+      "step": 810
+    },
+    {
+      "epoch": 0.7558247903075489,
+      "grad_norm": 0.4472767541947313,
+      "learning_rate": 4.156023472557818e-05,
+      "loss": 0.4207,
+      "step": 811
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.42968869716137265,
+      "learning_rate": 4.1542975491888163e-05,
+      "loss": 0.4036,
+      "step": 812
+    },
+    {
+      "epoch": 0.7576887232059646,
+      "grad_norm": 0.38679346997331027,
+      "learning_rate": 4.152571625819814e-05,
+      "loss": 0.4145,
+      "step": 813
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 0.43868692199992476,
+      "learning_rate": 4.150845702450811e-05,
+      "loss": 0.414,
+      "step": 814
+    },
+    {
+      "epoch": 0.7595526561043803,
+      "grad_norm": 0.3674305198270116,
+      "learning_rate": 4.1491197790818086e-05,
+      "loss": 0.3705,
+      "step": 815
+    },
+    {
+      "epoch": 0.7604846225535881,
+      "grad_norm": 0.41754251336276565,
+      "learning_rate": 4.147393855712807e-05,
+      "loss": 0.4063,
+      "step": 816
+    },
+    {
+      "epoch": 0.7614165890027959,
+      "grad_norm": 0.41680491938569875,
+      "learning_rate": 4.145667932343804e-05,
+      "loss": 0.4321,
+      "step": 817
+    },
+    {
+      "epoch": 0.7623485554520038,
+      "grad_norm": 0.4912310548737658,
+      "learning_rate": 4.1439420089748016e-05,
+      "loss": 0.4274,
+      "step": 818
+    },
+    {
+      "epoch": 0.7632805219012115,
+      "grad_norm": 0.39940698434580196,
+      "learning_rate": 4.1422160856058e-05,
+      "loss": 0.3998,
+      "step": 819
+    },
+    {
+      "epoch": 0.7642124883504194,
+      "grad_norm": 0.5378560928294738,
+      "learning_rate": 4.140490162236797e-05,
+      "loss": 0.4348,
+      "step": 820
+    },
+    {
+      "epoch": 0.7651444547996272,
+      "grad_norm": 0.33964878262496867,
+      "learning_rate": 4.1387642388677946e-05,
+      "loss": 0.3978,
+      "step": 821
+    },
+    {
+      "epoch": 0.766076421248835,
+      "grad_norm": 0.46194435534148637,
+      "learning_rate": 4.137038315498792e-05,
+      "loss": 0.4197,
+      "step": 822
+    },
+    {
+      "epoch": 0.7670083876980429,
+      "grad_norm": 0.37125635945394814,
+      "learning_rate": 4.13531239212979e-05,
+      "loss": 0.3987,
+      "step": 823
+    },
+    {
+      "epoch": 0.7679403541472507,
+      "grad_norm": 0.40365982818263857,
+      "learning_rate": 4.1335864687607875e-05,
+      "loss": 0.4272,
+      "step": 824
+    },
+    {
+      "epoch": 0.7688723205964585,
+      "grad_norm": 0.46284427430403424,
+      "learning_rate": 4.131860545391785e-05,
+      "loss": 0.4182,
+      "step": 825
+    },
+    {
+      "epoch": 0.7698042870456664,
+      "grad_norm": 0.3583859187272548,
+      "learning_rate": 4.1301346220227824e-05,
+      "loss": 0.4068,
+      "step": 826
+    },
+    {
+      "epoch": 0.7707362534948742,
+      "grad_norm": 0.3885631836934264,
+      "learning_rate": 4.12840869865378e-05,
+      "loss": 0.3888,
+      "step": 827
+    },
+    {
+      "epoch": 0.771668219944082,
+      "grad_norm": 0.3653304104883991,
+      "learning_rate": 4.126682775284777e-05,
+      "loss": 0.4209,
+      "step": 828
+    },
+    {
+      "epoch": 0.7726001863932899,
+      "grad_norm": 0.37072725900831277,
+      "learning_rate": 4.124956851915775e-05,
+      "loss": 0.4055,
+      "step": 829
+    },
+    {
+      "epoch": 0.7735321528424977,
+      "grad_norm": 0.3648840101157913,
+      "learning_rate": 4.123230928546773e-05,
+      "loss": 0.4136,
+      "step": 830
+    },
+    {
+      "epoch": 0.7744641192917054,
+      "grad_norm": 0.43885511724986,
+      "learning_rate": 4.12150500517777e-05,
+      "loss": 0.43,
+      "step": 831
+    },
+    {
+      "epoch": 0.7753960857409133,
+      "grad_norm": 0.35792802043501726,
+      "learning_rate": 4.1197790818087676e-05,
+      "loss": 0.4096,
+      "step": 832
+    },
+    {
+      "epoch": 0.7763280521901211,
+      "grad_norm": 0.3838517569513691,
+      "learning_rate": 4.118053158439765e-05,
+      "loss": 0.3927,
+      "step": 833
+    },
+    {
+      "epoch": 0.777260018639329,
+      "grad_norm": 0.5125456832943188,
+      "learning_rate": 4.116327235070763e-05,
+      "loss": 0.4129,
+      "step": 834
+    },
+    {
+      "epoch": 0.7781919850885368,
+      "grad_norm": 0.3916253775314793,
+      "learning_rate": 4.1146013117017606e-05,
+      "loss": 0.426,
+      "step": 835
+    },
+    {
+      "epoch": 0.7791239515377446,
+      "grad_norm": 0.44303231344166166,
+      "learning_rate": 4.112875388332758e-05,
+      "loss": 0.4056,
+      "step": 836
+    },
+    {
+      "epoch": 0.7800559179869525,
+      "grad_norm": 0.302421765110756,
+      "learning_rate": 4.111149464963756e-05,
+      "loss": 0.4107,
+      "step": 837
+    },
+    {
+      "epoch": 0.7809878844361603,
+      "grad_norm": 0.48641641842308314,
+      "learning_rate": 4.1094235415947536e-05,
+      "loss": 0.4209,
+      "step": 838
+    },
+    {
+      "epoch": 0.7819198508853681,
+      "grad_norm": 0.35133904627538465,
+      "learning_rate": 4.107697618225751e-05,
+      "loss": 0.4032,
+      "step": 839
+    },
+    {
+      "epoch": 0.782851817334576,
+      "grad_norm": 0.4519288010551151,
+      "learning_rate": 4.1059716948567484e-05,
+      "loss": 0.4459,
+      "step": 840
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 0.41237788772717915,
+      "learning_rate": 4.1042457714877465e-05,
+      "loss": 0.4364,
+      "step": 841
+    },
+    {
+      "epoch": 0.7847157502329916,
+      "grad_norm": 0.3759657228396316,
+      "learning_rate": 4.102519848118744e-05,
+      "loss": 0.4217,
+      "step": 842
+    },
+    {
+      "epoch": 0.7856477166821995,
+      "grad_norm": 0.5128895485996978,
+      "learning_rate": 4.1007939247497414e-05,
+      "loss": 0.4236,
+      "step": 843
+    },
+    {
+      "epoch": 0.7865796831314072,
+      "grad_norm": 0.38895290770321267,
+      "learning_rate": 4.099068001380739e-05,
+      "loss": 0.4218,
+      "step": 844
+    },
+    {
+      "epoch": 0.7875116495806151,
+      "grad_norm": 0.452564331505176,
+      "learning_rate": 4.097342078011737e-05,
+      "loss": 0.4226,
+      "step": 845
+    },
+    {
+      "epoch": 0.7884436160298229,
+      "grad_norm": 0.4508859637774081,
+      "learning_rate": 4.0956161546427344e-05,
+      "loss": 0.4319,
+      "step": 846
+    },
+    {
+      "epoch": 0.7893755824790307,
+      "grad_norm": 0.34789512507453857,
+      "learning_rate": 4.093890231273732e-05,
+      "loss": 0.4062,
+      "step": 847
+    },
+    {
+      "epoch": 0.7903075489282386,
+      "grad_norm": 0.4029367159269724,
+      "learning_rate": 4.092164307904729e-05,
+      "loss": 0.4138,
+      "step": 848
+    },
+    {
+      "epoch": 0.7912395153774464,
+      "grad_norm": 0.3788137723745779,
+      "learning_rate": 4.0904383845357266e-05,
+      "loss": 0.4,
+      "step": 849
+    },
+    {
+      "epoch": 0.7921714818266542,
+      "grad_norm": 0.4119916172627683,
+      "learning_rate": 4.088712461166724e-05,
+      "loss": 0.4233,
+      "step": 850
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 0.38821230802351425,
+      "learning_rate": 4.0869865377977215e-05,
+      "loss": 0.442,
+      "step": 851
+    },
+    {
+      "epoch": 0.7940354147250699,
+      "grad_norm": 0.43926285908763274,
+      "learning_rate": 4.0852606144287196e-05,
+      "loss": 0.4106,
+      "step": 852
+    },
+    {
+      "epoch": 0.7949673811742777,
+      "grad_norm": 0.40618335083196083,
+      "learning_rate": 4.083534691059717e-05,
+      "loss": 0.3974,
+      "step": 853
+    },
+    {
+      "epoch": 0.7958993476234856,
+      "grad_norm": 0.35405210744764076,
+      "learning_rate": 4.0818087676907145e-05,
+      "loss": 0.4089,
+      "step": 854
+    },
+    {
+      "epoch": 0.7968313140726934,
+      "grad_norm": 0.5388851154333876,
+      "learning_rate": 4.080082844321712e-05,
+      "loss": 0.4262,
+      "step": 855
+    },
+    {
+      "epoch": 0.7977632805219013,
+      "grad_norm": 0.3374603426234839,
+      "learning_rate": 4.07835692095271e-05,
+      "loss": 0.4265,
+      "step": 856
+    },
+    {
+      "epoch": 0.798695246971109,
+      "grad_norm": 0.3642161856933585,
+      "learning_rate": 4.0766309975837074e-05,
+      "loss": 0.4141,
+      "step": 857
+    },
+    {
+      "epoch": 0.7996272134203168,
+      "grad_norm": 0.40906621522068826,
+      "learning_rate": 4.074905074214705e-05,
+      "loss": 0.4055,
+      "step": 858
+    },
+    {
+      "epoch": 0.8005591798695247,
+      "grad_norm": 0.3704311510248122,
+      "learning_rate": 4.073179150845703e-05,
+      "loss": 0.4629,
+      "step": 859
+    },
+    {
+      "epoch": 0.8014911463187325,
+      "grad_norm": 0.4050795458127078,
+      "learning_rate": 4.0714532274767004e-05,
+      "loss": 0.4111,
+      "step": 860
+    },
+    {
+      "epoch": 0.8024231127679403,
+      "grad_norm": 0.34153990246842514,
+      "learning_rate": 4.069727304107698e-05,
+      "loss": 0.3924,
+      "step": 861
+    },
+    {
+      "epoch": 0.8033550792171482,
+      "grad_norm": 0.3705843815164345,
+      "learning_rate": 4.068001380738695e-05,
+      "loss": 0.429,
+      "step": 862
+    },
+    {
+      "epoch": 0.804287045666356,
+      "grad_norm": 0.3608391259983993,
+      "learning_rate": 4.0662754573696934e-05,
+      "loss": 0.4212,
+      "step": 863
+    },
+    {
+      "epoch": 0.8052190121155638,
+      "grad_norm": 0.29428830286435953,
+      "learning_rate": 4.064549534000691e-05,
+      "loss": 0.3734,
+      "step": 864
+    },
+    {
+      "epoch": 0.8061509785647717,
+      "grad_norm": 0.351999120007548,
+      "learning_rate": 4.062823610631688e-05,
+      "loss": 0.4082,
+      "step": 865
+    },
+    {
+      "epoch": 0.8070829450139795,
+      "grad_norm": 0.35137793063755185,
+      "learning_rate": 4.0610976872626856e-05,
+      "loss": 0.4148,
+      "step": 866
+    },
+    {
+      "epoch": 0.8080149114631874,
+      "grad_norm": 0.3605484527225777,
+      "learning_rate": 4.059371763893684e-05,
+      "loss": 0.4053,
+      "step": 867
+    },
+    {
+      "epoch": 0.8089468779123952,
+      "grad_norm": 0.3761538968178304,
+      "learning_rate": 4.057645840524681e-05,
+      "loss": 0.4365,
+      "step": 868
+    },
+    {
+      "epoch": 0.809878844361603,
+      "grad_norm": 0.3874817097146,
+      "learning_rate": 4.0559199171556786e-05,
+      "loss": 0.4259,
+      "step": 869
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.3215924871561967,
+      "learning_rate": 4.054193993786676e-05,
+      "loss": 0.4144,
+      "step": 870
+    },
+    {
+      "epoch": 0.8117427772600186,
+      "grad_norm": 0.3998741284875571,
+      "learning_rate": 4.0524680704176735e-05,
+      "loss": 0.4309,
+      "step": 871
+    },
+    {
+      "epoch": 0.8126747437092264,
+      "grad_norm": 0.36973540705898666,
+      "learning_rate": 4.050742147048671e-05,
+      "loss": 0.411,
+      "step": 872
+    },
+    {
+      "epoch": 0.8136067101584343,
+      "grad_norm": 0.35128115656198833,
+      "learning_rate": 4.049016223679668e-05,
+      "loss": 0.3975,
+      "step": 873
+    },
+    {
+      "epoch": 0.8145386766076421,
+      "grad_norm": 0.4110800737673518,
+      "learning_rate": 4.0472903003106664e-05,
+      "loss": 0.4204,
+      "step": 874
+    },
+    {
+      "epoch": 0.8154706430568499,
+      "grad_norm": 0.3745945601218999,
+      "learning_rate": 4.045564376941664e-05,
+      "loss": 0.4142,
+      "step": 875
+    },
+    {
+      "epoch": 0.8164026095060578,
+      "grad_norm": 0.41933245333582847,
+      "learning_rate": 4.043838453572661e-05,
+      "loss": 0.4127,
+      "step": 876
+    },
+    {
+      "epoch": 0.8173345759552656,
+      "grad_norm": 0.4466228278300321,
+      "learning_rate": 4.042112530203659e-05,
+      "loss": 0.4244,
+      "step": 877
+    },
+    {
+      "epoch": 0.8182665424044734,
+      "grad_norm": 0.38695074649626265,
+      "learning_rate": 4.040386606834657e-05,
+      "loss": 0.384,
+      "step": 878
+    },
+    {
+      "epoch": 0.8191985088536813,
+      "grad_norm": 0.4762165089875269,
+      "learning_rate": 4.038660683465654e-05,
+      "loss": 0.4128,
+      "step": 879
+    },
+    {
+      "epoch": 0.8201304753028891,
+      "grad_norm": 0.3745470166105871,
+      "learning_rate": 4.036934760096652e-05,
+      "loss": 0.3952,
+      "step": 880
+    },
+    {
+      "epoch": 0.821062441752097,
+      "grad_norm": 0.46718927482826267,
+      "learning_rate": 4.03520883672765e-05,
+      "loss": 0.4208,
+      "step": 881
+    },
+    {
+      "epoch": 0.8219944082013048,
+      "grad_norm": 0.40869049731883855,
+      "learning_rate": 4.033482913358647e-05,
+      "loss": 0.4161,
+      "step": 882
+    },
+    {
+      "epoch": 0.8229263746505125,
+      "grad_norm": 0.5108607683058191,
+      "learning_rate": 4.0317569899896447e-05,
+      "loss": 0.4302,
+      "step": 883
+    },
+    {
+      "epoch": 0.8238583410997204,
+      "grad_norm": 0.4907127070030013,
+      "learning_rate": 4.030031066620642e-05,
+      "loss": 0.4199,
+      "step": 884
+    },
+    {
+      "epoch": 0.8247903075489282,
+      "grad_norm": 0.41796862124073103,
+      "learning_rate": 4.02830514325164e-05,
+      "loss": 0.4132,
+      "step": 885
+    },
+    {
+      "epoch": 0.825722273998136,
+      "grad_norm": 0.5467368141516757,
+      "learning_rate": 4.0265792198826376e-05,
+      "loss": 0.4024,
+      "step": 886
+    },
+    {
+      "epoch": 0.8266542404473439,
+      "grad_norm": 0.3594473424437002,
+      "learning_rate": 4.024853296513635e-05,
+      "loss": 0.4037,
+      "step": 887
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.38800570130303147,
+      "learning_rate": 4.023127373144633e-05,
+      "loss": 0.4036,
+      "step": 888
+    },
+    {
+      "epoch": 0.8285181733457595,
+      "grad_norm": 0.32260245953918676,
+      "learning_rate": 4.0214014497756306e-05,
+      "loss": 0.403,
+      "step": 889
+    },
+    {
+      "epoch": 0.8294501397949674,
+      "grad_norm": 0.4178240351460631,
+      "learning_rate": 4.019675526406628e-05,
+      "loss": 0.421,
+      "step": 890
+    },
+    {
+      "epoch": 0.8303821062441752,
+      "grad_norm": 0.36645290665303787,
+      "learning_rate": 4.0179496030376254e-05,
+      "loss": 0.4264,
+      "step": 891
+    },
+    {
+      "epoch": 0.8313140726933831,
+      "grad_norm": 0.5067703900974035,
+      "learning_rate": 4.016223679668623e-05,
+      "loss": 0.4621,
+      "step": 892
+    },
+    {
+      "epoch": 0.8322460391425909,
+      "grad_norm": 0.3383962916676692,
+      "learning_rate": 4.01449775629962e-05,
+      "loss": 0.4093,
+      "step": 893
+    },
+    {
+      "epoch": 0.8331780055917987,
+      "grad_norm": 0.37938774482831944,
+      "learning_rate": 4.012771832930618e-05,
+      "loss": 0.4086,
+      "step": 894
+    },
+    {
+      "epoch": 0.8341099720410066,
+      "grad_norm": 0.3471157218191933,
+      "learning_rate": 4.011045909561615e-05,
+      "loss": 0.4244,
+      "step": 895
+    },
+    {
+      "epoch": 0.8350419384902144,
+      "grad_norm": 0.35888553621628416,
+      "learning_rate": 4.009319986192613e-05,
+      "loss": 0.4167,
+      "step": 896
+    },
+    {
+      "epoch": 0.8359739049394221,
+      "grad_norm": 0.3382897263031699,
+      "learning_rate": 4.007594062823611e-05,
+      "loss": 0.3922,
+      "step": 897
+    },
+    {
+      "epoch": 0.83690587138863,
+      "grad_norm": 0.33545209952972443,
+      "learning_rate": 4.005868139454608e-05,
+      "loss": 0.4259,
+      "step": 898
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 0.40784610513422126,
+      "learning_rate": 4.004142216085606e-05,
+      "loss": 0.4017,
+      "step": 899
+    },
+    {
+      "epoch": 0.8387698042870456,
+      "grad_norm": 0.3645000577698368,
+      "learning_rate": 4.0024162927166037e-05,
+      "loss": 0.408,
+      "step": 900
+    },
+    {
+      "epoch": 0.8397017707362535,
+      "grad_norm": 0.3564929317680359,
+      "learning_rate": 4.000690369347601e-05,
+      "loss": 0.4238,
+      "step": 901
+    },
+    {
+      "epoch": 0.8406337371854613,
+      "grad_norm": 0.3723651617370857,
+      "learning_rate": 3.9989644459785985e-05,
+      "loss": 0.4041,
+      "step": 902
+    },
+    {
+      "epoch": 0.8415657036346692,
+      "grad_norm": 0.36989374281943765,
+      "learning_rate": 3.9972385226095966e-05,
+      "loss": 0.4184,
+      "step": 903
+    },
+    {
+      "epoch": 0.842497670083877,
+      "grad_norm": 0.435803012357952,
+      "learning_rate": 3.995512599240594e-05,
+      "loss": 0.4201,
+      "step": 904
+    },
+    {
+      "epoch": 0.8434296365330848,
+      "grad_norm": 0.32436382480194553,
+      "learning_rate": 3.9937866758715915e-05,
+      "loss": 0.3857,
+      "step": 905
+    },
+    {
+      "epoch": 0.8443616029822927,
+      "grad_norm": 0.5322476047397392,
+      "learning_rate": 3.992060752502589e-05,
+      "loss": 0.4397,
+      "step": 906
+    },
+    {
+      "epoch": 0.8452935694315005,
+      "grad_norm": 0.37098832340523813,
+      "learning_rate": 3.990334829133587e-05,
+      "loss": 0.435,
+      "step": 907
+    },
+    {
+      "epoch": 0.8462255358807083,
+      "grad_norm": 0.4850951050561126,
+      "learning_rate": 3.9886089057645844e-05,
+      "loss": 0.4105,
+      "step": 908
+    },
+    {
+      "epoch": 0.8471575023299162,
+      "grad_norm": 0.3318963067438199,
+      "learning_rate": 3.986882982395582e-05,
+      "loss": 0.3919,
+      "step": 909
+    },
+    {
+      "epoch": 0.848089468779124,
+      "grad_norm": 0.5366503915720571,
+      "learning_rate": 3.98515705902658e-05,
+      "loss": 0.4057,
+      "step": 910
+    },
+    {
+      "epoch": 0.8490214352283317,
+      "grad_norm": 0.4647878695078262,
+      "learning_rate": 3.9834311356575774e-05,
+      "loss": 0.423,
+      "step": 911
+    },
+    {
+      "epoch": 0.8499534016775396,
+      "grad_norm": 0.4488248340886583,
+      "learning_rate": 3.981705212288575e-05,
+      "loss": 0.3858,
+      "step": 912
+    },
+    {
+      "epoch": 0.8508853681267474,
+      "grad_norm": 0.4611441141576221,
+      "learning_rate": 3.979979288919572e-05,
+      "loss": 0.4109,
+      "step": 913
+    },
+    {
+      "epoch": 0.8518173345759553,
+      "grad_norm": 0.3406338750405225,
+      "learning_rate": 3.97825336555057e-05,
+      "loss": 0.4016,
+      "step": 914
+    },
+    {
+      "epoch": 0.8527493010251631,
+      "grad_norm": 0.5774585753392601,
+      "learning_rate": 3.976527442181567e-05,
+      "loss": 0.4257,
+      "step": 915
+    },
+    {
+      "epoch": 0.8536812674743709,
+      "grad_norm": 0.316750449630663,
+      "learning_rate": 3.9748015188125646e-05,
+      "loss": 0.4108,
+      "step": 916
+    },
+    {
+      "epoch": 0.8546132339235788,
+      "grad_norm": 0.4560942430695926,
+      "learning_rate": 3.973075595443562e-05,
+      "loss": 0.4209,
+      "step": 917
+    },
+    {
+      "epoch": 0.8555452003727866,
+      "grad_norm": 0.3775510239873204,
+      "learning_rate": 3.97134967207456e-05,
+      "loss": 0.4047,
+      "step": 918
+    },
+    {
+      "epoch": 0.8564771668219944,
+      "grad_norm": 0.33495346797514686,
+      "learning_rate": 3.9696237487055575e-05,
+      "loss": 0.4219,
+      "step": 919
+    },
+    {
+      "epoch": 0.8574091332712023,
+      "grad_norm": 0.5182462546819264,
+      "learning_rate": 3.967897825336555e-05,
+      "loss": 0.4207,
+      "step": 920
+    },
+    {
+      "epoch": 0.8583410997204101,
+      "grad_norm": 0.3553440806947589,
+      "learning_rate": 3.966171901967553e-05,
+      "loss": 0.3867,
+      "step": 921
+    },
+    {
+      "epoch": 0.8592730661696178,
+      "grad_norm": 0.42235713711086786,
+      "learning_rate": 3.9644459785985505e-05,
+      "loss": 0.3943,
+      "step": 922
+    },
+    {
+      "epoch": 0.8602050326188257,
+      "grad_norm": 0.3836612014775603,
+      "learning_rate": 3.962720055229548e-05,
+      "loss": 0.4156,
+      "step": 923
+    },
+    {
+      "epoch": 0.8611369990680335,
+      "grad_norm": 0.35879907115564724,
+      "learning_rate": 3.9609941318605453e-05,
+      "loss": 0.4067,
+      "step": 924
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 0.3630221495586929,
+      "learning_rate": 3.9592682084915434e-05,
+      "loss": 0.4088,
+      "step": 925
+    },
+    {
+      "epoch": 0.8630009319664492,
+      "grad_norm": 0.38895960667003227,
+      "learning_rate": 3.957542285122541e-05,
+      "loss": 0.3946,
+      "step": 926
+    },
+    {
+      "epoch": 0.863932898415657,
+      "grad_norm": 0.3399338175891486,
+      "learning_rate": 3.955816361753538e-05,
+      "loss": 0.4136,
+      "step": 927
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.3906626993295965,
+      "learning_rate": 3.9540904383845364e-05,
+      "loss": 0.3962,
+      "step": 928
+    },
+    {
+      "epoch": 0.8657968313140727,
+      "grad_norm": 0.2667323583911121,
+      "learning_rate": 3.952364515015534e-05,
+      "loss": 0.3992,
+      "step": 929
+    },
+    {
+      "epoch": 0.8667287977632805,
+      "grad_norm": 0.49992625427617715,
+      "learning_rate": 3.950638591646531e-05,
+      "loss": 0.407,
+      "step": 930
+    },
+    {
+      "epoch": 0.8676607642124884,
+      "grad_norm": 0.3618113437040364,
+      "learning_rate": 3.948912668277529e-05,
+      "loss": 0.4151,
+      "step": 931
+    },
+    {
+      "epoch": 0.8685927306616962,
+      "grad_norm": 0.4182994720793084,
+      "learning_rate": 3.947186744908527e-05,
+      "loss": 0.4117,
+      "step": 932
+    },
+    {
+      "epoch": 0.869524697110904,
+      "grad_norm": 0.43057793672486866,
+      "learning_rate": 3.945460821539524e-05,
+      "loss": 0.4272,
+      "step": 933
+    },
+    {
+      "epoch": 0.8704566635601119,
+      "grad_norm": 0.35999911956401126,
+      "learning_rate": 3.943734898170522e-05,
+      "loss": 0.4027,
+      "step": 934
+    },
+    {
+      "epoch": 0.8713886300093197,
+      "grad_norm": 0.4452563700054875,
+      "learning_rate": 3.942008974801519e-05,
+      "loss": 0.4129,
+      "step": 935
+    },
+    {
+      "epoch": 0.8723205964585274,
+      "grad_norm": 0.4161773986890216,
+      "learning_rate": 3.9402830514325165e-05,
+      "loss": 0.4219,
+      "step": 936
+    },
+    {
+      "epoch": 0.8732525629077353,
+      "grad_norm": 0.3648104002127043,
+      "learning_rate": 3.938557128063514e-05,
+      "loss": 0.4027,
+      "step": 937
+    },
+    {
+      "epoch": 0.8741845293569431,
+      "grad_norm": 0.35503442649490813,
+      "learning_rate": 3.9368312046945114e-05,
+      "loss": 0.3969,
+      "step": 938
+    },
+    {
+      "epoch": 0.875116495806151,
+      "grad_norm": 0.34536972398426974,
+      "learning_rate": 3.9351052813255095e-05,
+      "loss": 0.4106,
+      "step": 939
+    },
+    {
+      "epoch": 0.8760484622553588,
+      "grad_norm": 0.31350483959808445,
+      "learning_rate": 3.933379357956507e-05,
+      "loss": 0.416,
+      "step": 940
+    },
+    {
+      "epoch": 0.8769804287045666,
+      "grad_norm": 0.38445772490054636,
+      "learning_rate": 3.9316534345875043e-05,
+      "loss": 0.3948,
+      "step": 941
+    },
+    {
+      "epoch": 0.8779123951537745,
+      "grad_norm": 0.3482762811526219,
+      "learning_rate": 3.929927511218502e-05,
+      "loss": 0.4149,
+      "step": 942
+    },
+    {
+      "epoch": 0.8788443616029823,
+      "grad_norm": 0.3569439267498268,
+      "learning_rate": 3.9282015878495e-05,
+      "loss": 0.4097,
+      "step": 943
+    },
+    {
+      "epoch": 0.8797763280521901,
+      "grad_norm": 0.39319163134588436,
+      "learning_rate": 3.926475664480497e-05,
+      "loss": 0.4368,
+      "step": 944
+    },
+    {
+      "epoch": 0.880708294501398,
+      "grad_norm": 0.42602724379833173,
+      "learning_rate": 3.924749741111495e-05,
+      "loss": 0.4022,
+      "step": 945
+    },
+    {
+      "epoch": 0.8816402609506058,
+      "grad_norm": 0.37573976049110946,
+      "learning_rate": 3.923023817742492e-05,
+      "loss": 0.4242,
+      "step": 946
+    },
+    {
+      "epoch": 0.8825722273998136,
+      "grad_norm": 0.4740900419488988,
+      "learning_rate": 3.92129789437349e-05,
+      "loss": 0.419,
+      "step": 947
+    },
+    {
+      "epoch": 0.8835041938490215,
+      "grad_norm": 0.4711946111307789,
+      "learning_rate": 3.919571971004488e-05,
+      "loss": 0.4121,
+      "step": 948
+    },
+    {
+      "epoch": 0.8844361602982292,
+      "grad_norm": 0.3692067739099133,
+      "learning_rate": 3.917846047635485e-05,
+      "loss": 0.4297,
+      "step": 949
+    },
+    {
+      "epoch": 0.8853681267474371,
+      "grad_norm": 0.5267766563474916,
+      "learning_rate": 3.916120124266483e-05,
+      "loss": 0.4123,
+      "step": 950
+    },
+    {
+      "epoch": 0.8863000931966449,
+      "grad_norm": 0.33397313294073905,
+      "learning_rate": 3.914394200897481e-05,
+      "loss": 0.4042,
+      "step": 951
+    },
+    {
+      "epoch": 0.8872320596458527,
+      "grad_norm": 0.5660062512884372,
+      "learning_rate": 3.912668277528478e-05,
+      "loss": 0.3962,
+      "step": 952
+    },
+    {
+      "epoch": 0.8881640260950606,
+      "grad_norm": 0.3813411184251707,
+      "learning_rate": 3.9109423541594755e-05,
+      "loss": 0.4421,
+      "step": 953
+    },
+    {
+      "epoch": 0.8890959925442684,
+      "grad_norm": 0.4770359984469211,
+      "learning_rate": 3.909216430790473e-05,
+      "loss": 0.4152,
+      "step": 954
+    },
+    {
+      "epoch": 0.8900279589934762,
+      "grad_norm": 0.43297497415775366,
+      "learning_rate": 3.9074905074214704e-05,
+      "loss": 0.4341,
+      "step": 955
+    },
+    {
+      "epoch": 0.8909599254426841,
+      "grad_norm": 0.3851855293521226,
+      "learning_rate": 3.905764584052468e-05,
+      "loss": 0.4179,
+      "step": 956
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 0.4258913997422698,
+      "learning_rate": 3.904038660683465e-05,
+      "loss": 0.4434,
+      "step": 957
+    },
+    {
+      "epoch": 0.8928238583410997,
+      "grad_norm": 0.3650019297499204,
+      "learning_rate": 3.9023127373144633e-05,
+      "loss": 0.4105,
+      "step": 958
+    },
+    {
+      "epoch": 0.8937558247903076,
+      "grad_norm": 0.3126128343924253,
+      "learning_rate": 3.900586813945461e-05,
+      "loss": 0.3894,
+      "step": 959
+    },
+    {
+      "epoch": 0.8946877912395154,
+      "grad_norm": 0.35374969135870443,
+      "learning_rate": 3.898860890576458e-05,
+      "loss": 0.4069,
+      "step": 960
+    },
+    {
+      "epoch": 0.8956197576887233,
+      "grad_norm": 0.3197974313536589,
+      "learning_rate": 3.897134967207456e-05,
+      "loss": 0.4214,
+      "step": 961
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 0.34527778494908495,
+      "learning_rate": 3.895409043838454e-05,
+      "loss": 0.3914,
+      "step": 962
+    },
+    {
+      "epoch": 0.8974836905871388,
+      "grad_norm": 0.35937261084652533,
+      "learning_rate": 3.893683120469451e-05,
+      "loss": 0.4109,
+      "step": 963
+    },
+    {
+      "epoch": 0.8984156570363467,
+      "grad_norm": 0.31690397909937673,
+      "learning_rate": 3.8919571971004486e-05,
+      "loss": 0.3855,
+      "step": 964
+    },
+    {
+      "epoch": 0.8993476234855545,
+      "grad_norm": 0.4201869762179012,
+      "learning_rate": 3.890231273731447e-05,
+      "loss": 0.4176,
+      "step": 965
+    },
+    {
+      "epoch": 0.9002795899347623,
+      "grad_norm": 0.3865633494960209,
+      "learning_rate": 3.888505350362444e-05,
+      "loss": 0.4044,
+      "step": 966
+    },
+    {
+      "epoch": 0.9012115563839702,
+      "grad_norm": 0.35704164121585313,
+      "learning_rate": 3.8867794269934416e-05,
+      "loss": 0.4117,
+      "step": 967
+    },
+    {
+      "epoch": 0.902143522833178,
+      "grad_norm": 0.38500949586627786,
+      "learning_rate": 3.885053503624439e-05,
+      "loss": 0.4057,
+      "step": 968
+    },
+    {
+      "epoch": 0.9030754892823858,
+      "grad_norm": 0.4245235998103044,
+      "learning_rate": 3.883327580255437e-05,
+      "loss": 0.409,
+      "step": 969
+    },
+    {
+      "epoch": 0.9040074557315937,
+      "grad_norm": 0.3077259179990962,
+      "learning_rate": 3.8816016568864345e-05,
+      "loss": 0.3849,
+      "step": 970
+    },
+    {
+      "epoch": 0.9049394221808015,
+      "grad_norm": 0.38829853412339155,
+      "learning_rate": 3.879875733517432e-05,
+      "loss": 0.4039,
+      "step": 971
+    },
+    {
+      "epoch": 0.9058713886300093,
+      "grad_norm": 0.37414651334851023,
+      "learning_rate": 3.87814981014843e-05,
+      "loss": 0.417,
+      "step": 972
+    },
+    {
+      "epoch": 0.9068033550792172,
+      "grad_norm": 0.3608810948283261,
+      "learning_rate": 3.8764238867794275e-05,
+      "loss": 0.4373,
+      "step": 973
+    },
+    {
+      "epoch": 0.907735321528425,
+      "grad_norm": 0.31604321294531046,
+      "learning_rate": 3.874697963410425e-05,
+      "loss": 0.4313,
+      "step": 974
+    },
+    {
+      "epoch": 0.9086672879776329,
+      "grad_norm": 0.3301783285534026,
+      "learning_rate": 3.8729720400414224e-05,
+      "loss": 0.4203,
+      "step": 975
+    },
+    {
+      "epoch": 0.9095992544268406,
+      "grad_norm": 0.31372727743178236,
+      "learning_rate": 3.87124611667242e-05,
+      "loss": 0.3995,
+      "step": 976
+    },
+    {
+      "epoch": 0.9105312208760484,
+      "grad_norm": 0.35601563675035824,
+      "learning_rate": 3.869520193303417e-05,
+      "loss": 0.4081,
+      "step": 977
+    },
+    {
+      "epoch": 0.9114631873252563,
+      "grad_norm": 0.29693705313948854,
+      "learning_rate": 3.8677942699344146e-05,
+      "loss": 0.4005,
+      "step": 978
+    },
+    {
+      "epoch": 0.9123951537744641,
+      "grad_norm": 0.3391655747797109,
+      "learning_rate": 3.866068346565413e-05,
+      "loss": 0.4136,
+      "step": 979
+    },
+    {
+      "epoch": 0.9133271202236719,
+      "grad_norm": 0.3835248556519902,
+      "learning_rate": 3.86434242319641e-05,
+      "loss": 0.4239,
+      "step": 980
+    },
+    {
+      "epoch": 0.9142590866728798,
+      "grad_norm": 0.38011851825396875,
+      "learning_rate": 3.8626164998274076e-05,
+      "loss": 0.4075,
+      "step": 981
+    },
+    {
+      "epoch": 0.9151910531220876,
+      "grad_norm": 0.39016155227722915,
+      "learning_rate": 3.860890576458405e-05,
+      "loss": 0.3976,
+      "step": 982
+    },
+    {
+      "epoch": 0.9161230195712954,
+      "grad_norm": 0.39012414439613646,
+      "learning_rate": 3.859164653089403e-05,
+      "loss": 0.4173,
+      "step": 983
+    },
+    {
+      "epoch": 0.9170549860205033,
+      "grad_norm": 0.28789926954994693,
+      "learning_rate": 3.8574387297204006e-05,
+      "loss": 0.4064,
+      "step": 984
+    },
+    {
+      "epoch": 0.9179869524697111,
+      "grad_norm": 0.488596184239538,
+      "learning_rate": 3.855712806351398e-05,
+      "loss": 0.4225,
+      "step": 985
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 0.4372914302977011,
+      "learning_rate": 3.8539868829823954e-05,
+      "loss": 0.3934,
+      "step": 986
+    },
+    {
+      "epoch": 0.9198508853681268,
+      "grad_norm": 0.3618161344063438,
+      "learning_rate": 3.8522609596133935e-05,
+      "loss": 0.4138,
+      "step": 987
+    },
+    {
+      "epoch": 0.9207828518173345,
+      "grad_norm": 0.4577116139850298,
+      "learning_rate": 3.850535036244391e-05,
+      "loss": 0.4286,
+      "step": 988
+    },
+    {
+      "epoch": 0.9217148182665424,
+      "grad_norm": 0.3554847083379012,
+      "learning_rate": 3.8488091128753884e-05,
+      "loss": 0.3939,
+      "step": 989
+    },
+    {
+      "epoch": 0.9226467847157502,
+      "grad_norm": 0.47371438335496424,
+      "learning_rate": 3.8470831895063865e-05,
+      "loss": 0.4095,
+      "step": 990
+    },
+    {
+      "epoch": 0.923578751164958,
+      "grad_norm": 0.3359672003684034,
+      "learning_rate": 3.845357266137384e-05,
+      "loss": 0.396,
+      "step": 991
+    },
+    {
+      "epoch": 0.9245107176141659,
+      "grad_norm": 0.441790097112171,
+      "learning_rate": 3.8436313427683814e-05,
+      "loss": 0.403,
+      "step": 992
+    },
+    {
+      "epoch": 0.9254426840633737,
+      "grad_norm": 0.3109202007577921,
+      "learning_rate": 3.841905419399379e-05,
+      "loss": 0.3779,
+      "step": 993
+    },
+    {
+      "epoch": 0.9263746505125815,
+      "grad_norm": 0.41476659408181743,
+      "learning_rate": 3.840179496030377e-05,
+      "loss": 0.4365,
+      "step": 994
+    },
+    {
+      "epoch": 0.9273066169617894,
+      "grad_norm": 0.4085388000804664,
+      "learning_rate": 3.838453572661374e-05,
+      "loss": 0.404,
+      "step": 995
+    },
+    {
+      "epoch": 0.9282385834109972,
+      "grad_norm": 0.39503788101225107,
+      "learning_rate": 3.836727649292372e-05,
+      "loss": 0.4123,
+      "step": 996
+    },
+    {
+      "epoch": 0.9291705498602051,
+      "grad_norm": 0.4054006394059083,
+      "learning_rate": 3.835001725923369e-05,
+      "loss": 0.4025,
+      "step": 997
+    },
+    {
+      "epoch": 0.9301025163094129,
+      "grad_norm": 0.3629580192062035,
+      "learning_rate": 3.8332758025543666e-05,
+      "loss": 0.4025,
+      "step": 998
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 0.38868922286888596,
+      "learning_rate": 3.831549879185364e-05,
+      "loss": 0.4106,
+      "step": 999
+    },
+    {
+      "epoch": 0.9319664492078286,
+      "grad_norm": 0.4363551198467323,
+      "learning_rate": 3.8298239558163615e-05,
+      "loss": 0.4054,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9328984156570364,
+      "grad_norm": 0.3518253528304607,
+      "learning_rate": 3.8280980324473596e-05,
+      "loss": 0.4009,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9338303821062441,
+      "grad_norm": 0.43708900884017504,
+      "learning_rate": 3.826372109078357e-05,
+      "loss": 0.4127,
+      "step": 1002
+    },
+    {
+      "epoch": 0.934762348555452,
+      "grad_norm": 0.40775783288926737,
+      "learning_rate": 3.8246461857093544e-05,
+      "loss": 0.4111,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9356943150046598,
+      "grad_norm": 0.43558596307628683,
+      "learning_rate": 3.822920262340352e-05,
+      "loss": 0.4191,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9366262814538676,
+      "grad_norm": 0.4181705787524772,
+      "learning_rate": 3.82119433897135e-05,
+      "loss": 0.4134,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9375582479030755,
+      "grad_norm": 0.3544600679426068,
+      "learning_rate": 3.8194684156023474e-05,
+      "loss": 0.4107,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9384902143522833,
+      "grad_norm": 0.4270043941077659,
+      "learning_rate": 3.817742492233345e-05,
+      "loss": 0.4043,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9394221808014911,
+      "grad_norm": 0.33377626676835936,
+      "learning_rate": 3.816016568864342e-05,
+      "loss": 0.3763,
+      "step": 1008
+    },
+    {
+      "epoch": 0.940354147250699,
+      "grad_norm": 0.4828007374776105,
+      "learning_rate": 3.8142906454953404e-05,
+      "loss": 0.4196,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9412861136999068,
+      "grad_norm": 0.3805525781440441,
+      "learning_rate": 3.812564722126338e-05,
+      "loss": 0.4105,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9422180801491147,
+      "grad_norm": 0.5949987468989795,
+      "learning_rate": 3.810838798757335e-05,
+      "loss": 0.423,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9431500465983225,
+      "grad_norm": 0.3873450762290146,
+      "learning_rate": 3.809112875388333e-05,
+      "loss": 0.4131,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9440820130475303,
+      "grad_norm": 0.4233337712549983,
+      "learning_rate": 3.807386952019331e-05,
+      "loss": 0.4039,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9450139794967382,
+      "grad_norm": 0.38423682685369304,
+      "learning_rate": 3.805661028650328e-05,
+      "loss": 0.4002,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 0.404713232141802,
+      "learning_rate": 3.8039351052813256e-05,
+      "loss": 0.4234,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9468779123951537,
+      "grad_norm": 0.43891114141243237,
+      "learning_rate": 3.802209181912324e-05,
+      "loss": 0.4049,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9478098788443616,
+      "grad_norm": 0.3273109573161215,
+      "learning_rate": 3.800483258543321e-05,
+      "loss": 0.3975,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9487418452935694,
+      "grad_norm": 0.41332463167896344,
+      "learning_rate": 3.7987573351743186e-05,
+      "loss": 0.4188,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9496738117427772,
+      "grad_norm": 0.3031487263168064,
+      "learning_rate": 3.797031411805316e-05,
+      "loss": 0.418,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9506057781919851,
+      "grad_norm": 0.41959279375390063,
+      "learning_rate": 3.7953054884363134e-05,
+      "loss": 0.3928,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9515377446411929,
+      "grad_norm": 0.3415655981472387,
+      "learning_rate": 3.793579565067311e-05,
+      "loss": 0.4085,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9524697110904008,
+      "grad_norm": 0.4256043891477172,
+      "learning_rate": 3.791853641698308e-05,
+      "loss": 0.4122,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9534016775396086,
+      "grad_norm": 0.36155367261970117,
+      "learning_rate": 3.7901277183293064e-05,
+      "loss": 0.4017,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9543336439888164,
+      "grad_norm": 0.34506961048645135,
+      "learning_rate": 3.788401794960304e-05,
+      "loss": 0.4113,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9552656104380243,
+      "grad_norm": 0.3945593443951552,
+      "learning_rate": 3.786675871591301e-05,
+      "loss": 0.4165,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9561975768872321,
+      "grad_norm": 0.32001687885668995,
+      "learning_rate": 3.784949948222299e-05,
+      "loss": 0.4112,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9571295433364398,
+      "grad_norm": 0.3734215655192567,
+      "learning_rate": 3.783224024853297e-05,
+      "loss": 0.4053,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9580615097856477,
+      "grad_norm": 0.3977729589657221,
+      "learning_rate": 3.781498101484294e-05,
+      "loss": 0.4429,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9589934762348555,
+      "grad_norm": 0.39012561445435584,
+      "learning_rate": 3.7797721781152916e-05,
+      "loss": 0.4164,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9599254426840633,
+      "grad_norm": 0.3087127179271281,
+      "learning_rate": 3.77804625474629e-05,
+      "loss": 0.3993,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9608574091332712,
+      "grad_norm": 0.41575277323376214,
+      "learning_rate": 3.776320331377287e-05,
+      "loss": 0.4302,
+      "step": 1031
+    },
+    {
+      "epoch": 0.961789375582479,
+      "grad_norm": 0.3619332478097631,
+      "learning_rate": 3.7745944080082846e-05,
+      "loss": 0.3998,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9627213420316869,
+      "grad_norm": 0.41035733004525754,
+      "learning_rate": 3.772868484639282e-05,
+      "loss": 0.4172,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9636533084808947,
+      "grad_norm": 0.29805108714096207,
+      "learning_rate": 3.77114256127028e-05,
+      "loss": 0.3863,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9645852749301025,
+      "grad_norm": 0.43348328445729384,
+      "learning_rate": 3.7694166379012776e-05,
+      "loss": 0.4044,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 0.3008289737725392,
+      "learning_rate": 3.767690714532275e-05,
+      "loss": 0.3945,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9664492078285182,
+      "grad_norm": 0.3637399853657742,
+      "learning_rate": 3.7659647911632724e-05,
+      "loss": 0.4065,
+      "step": 1037
+    },
+    {
+      "epoch": 0.967381174277726,
+      "grad_norm": 0.3891684039261977,
+      "learning_rate": 3.7642388677942705e-05,
+      "loss": 0.406,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9683131407269339,
+      "grad_norm": 0.362811441978515,
+      "learning_rate": 3.762512944425268e-05,
+      "loss": 0.4333,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9692451071761417,
+      "grad_norm": 0.379789845428218,
+      "learning_rate": 3.7607870210562654e-05,
+      "loss": 0.3849,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9701770736253494,
+      "grad_norm": 0.31387756766946295,
+      "learning_rate": 3.759061097687263e-05,
+      "loss": 0.3955,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9711090400745573,
+      "grad_norm": 0.36961853370059944,
+      "learning_rate": 3.75733517431826e-05,
+      "loss": 0.4021,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9720410065237651,
+      "grad_norm": 0.27788558138618186,
+      "learning_rate": 3.755609250949258e-05,
+      "loss": 0.3894,
+      "step": 1043
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.3757592556957609,
+      "learning_rate": 3.753883327580255e-05,
+      "loss": 0.3919,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9739049394221808,
+      "grad_norm": 0.2966683772092975,
+      "learning_rate": 3.752157404211253e-05,
+      "loss": 0.4086,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9748369058713886,
+      "grad_norm": 0.43243586503828707,
+      "learning_rate": 3.7504314808422507e-05,
+      "loss": 0.4346,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9757688723205965,
+      "grad_norm": 0.31820933593224765,
+      "learning_rate": 3.748705557473248e-05,
+      "loss": 0.4403,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9767008387698043,
+      "grad_norm": 0.47307270058147916,
+      "learning_rate": 3.7469796341042455e-05,
+      "loss": 0.4025,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9776328052190121,
+      "grad_norm": 0.38486226892681596,
+      "learning_rate": 3.7452537107352436e-05,
+      "loss": 0.4271,
+      "step": 1049
+    },
+    {
+      "epoch": 0.97856477166822,
+      "grad_norm": 0.41041703300273114,
+      "learning_rate": 3.743527787366241e-05,
+      "loss": 0.4324,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9794967381174278,
+      "grad_norm": 0.3544495917894061,
+      "learning_rate": 3.7418018639972385e-05,
+      "loss": 0.3842,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9804287045666356,
+      "grad_norm": 0.4106020191019897,
+      "learning_rate": 3.7400759406282366e-05,
+      "loss": 0.4131,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9813606710158435,
+      "grad_norm": 0.3712191747131067,
+      "learning_rate": 3.738350017259234e-05,
+      "loss": 0.4045,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9822926374650512,
+      "grad_norm": 0.41492532957323314,
+      "learning_rate": 3.7366240938902314e-05,
+      "loss": 0.3985,
+      "step": 1054
+    },
+    {
+      "epoch": 0.983224603914259,
+      "grad_norm": 0.4043505063760219,
+      "learning_rate": 3.734898170521229e-05,
+      "loss": 0.4249,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9841565703634669,
+      "grad_norm": 0.45379194890040275,
+      "learning_rate": 3.733172247152227e-05,
+      "loss": 0.4155,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9850885368126747,
+      "grad_norm": 0.4710108182066288,
+      "learning_rate": 3.7314463237832244e-05,
+      "loss": 0.4147,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9860205032618826,
+      "grad_norm": 0.32517530173545495,
+      "learning_rate": 3.729720400414222e-05,
+      "loss": 0.3969,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9869524697110904,
+      "grad_norm": 0.3854181831760744,
+      "learning_rate": 3.72799447704522e-05,
+      "loss": 0.4112,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9878844361602982,
+      "grad_norm": 0.4092510355090361,
+      "learning_rate": 3.7262685536762174e-05,
+      "loss": 0.3819,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9888164026095061,
+      "grad_norm": 0.3952251971603547,
+      "learning_rate": 3.724542630307215e-05,
+      "loss": 0.4074,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9897483690587139,
+      "grad_norm": 0.3568679536835217,
+      "learning_rate": 3.722816706938212e-05,
+      "loss": 0.3987,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9906803355079217,
+      "grad_norm": 1.0379219440117091,
+      "learning_rate": 3.7210907835692097e-05,
+      "loss": 0.422,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9916123019571296,
+      "grad_norm": 0.33957381290099775,
+      "learning_rate": 3.719364860200207e-05,
+      "loss": 0.3987,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9925442684063374,
+      "grad_norm": 0.37667561502992786,
+      "learning_rate": 3.7176389368312045e-05,
+      "loss": 0.4155,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9934762348555451,
+      "grad_norm": 0.3214909494475861,
+      "learning_rate": 3.715913013462202e-05,
+      "loss": 0.4324,
+      "step": 1066
+    },
+    {
+      "epoch": 0.994408201304753,
+      "grad_norm": 0.3447403060288721,
+      "learning_rate": 3.7141870900932e-05,
+      "loss": 0.3879,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9953401677539608,
+      "grad_norm": 0.3864185204611896,
+      "learning_rate": 3.7124611667241975e-05,
+      "loss": 0.4253,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9962721342031687,
+      "grad_norm": 0.29096153261591284,
+      "learning_rate": 3.710735243355195e-05,
+      "loss": 0.39,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9972041006523765,
+      "grad_norm": 0.42309357273640574,
+      "learning_rate": 3.709009319986193e-05,
+      "loss": 0.4085,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9981360671015843,
+      "grad_norm": 0.30446740708756725,
+      "learning_rate": 3.7072833966171904e-05,
+      "loss": 0.3901,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9990680335507922,
+      "grad_norm": 0.32451045189434685,
+      "learning_rate": 3.705557473248188e-05,
+      "loss": 0.4052,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.36784265887272843,
+      "learning_rate": 3.703831549879185e-05,
+      "loss": 0.3933,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0009319664492078,
+      "grad_norm": 0.3262647183302033,
+      "learning_rate": 3.7021056265101834e-05,
+      "loss": 0.3375,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0018639328984156,
+      "grad_norm": 0.38721983835104967,
+      "learning_rate": 3.700379703141181e-05,
+      "loss": 0.3426,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0027958993476236,
+      "grad_norm": 0.34949469148766676,
+      "learning_rate": 3.698653779772178e-05,
+      "loss": 0.3303,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0037278657968314,
+      "grad_norm": 0.3230616484940306,
+      "learning_rate": 3.696927856403176e-05,
+      "loss": 0.3624,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0046598322460392,
+      "grad_norm": 0.3319437378255196,
+      "learning_rate": 3.695201933034174e-05,
+      "loss": 0.3566,
+      "step": 1078
+    },
+    {
+      "epoch": 1.005591798695247,
+      "grad_norm": 0.33034230279187515,
+      "learning_rate": 3.693476009665171e-05,
+      "loss": 0.3433,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0065237651444547,
+      "grad_norm": 0.3285156848695929,
+      "learning_rate": 3.6917500862961687e-05,
+      "loss": 0.3558,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0074557315936625,
+      "grad_norm": 0.2973714038236142,
+      "learning_rate": 3.690024162927167e-05,
+      "loss": 0.3237,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0083876980428705,
+      "grad_norm": 0.35244823006195103,
+      "learning_rate": 3.688298239558164e-05,
+      "loss": 0.3513,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0093196644920783,
+      "grad_norm": 0.2973840415749372,
+      "learning_rate": 3.6865723161891616e-05,
+      "loss": 0.3561,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0102516309412861,
+      "grad_norm": 0.328464325353007,
+      "learning_rate": 3.684846392820159e-05,
+      "loss": 0.3583,
+      "step": 1084
+    },
+    {
+      "epoch": 1.011183597390494,
+      "grad_norm": 0.42951476359023516,
+      "learning_rate": 3.6831204694511565e-05,
+      "loss": 0.3911,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0121155638397017,
+      "grad_norm": 0.32227868588168457,
+      "learning_rate": 3.681394546082154e-05,
+      "loss": 0.3327,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0130475302889097,
+      "grad_norm": 0.4869570315054193,
+      "learning_rate": 3.6796686227131513e-05,
+      "loss": 0.3752,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0139794967381175,
+      "grad_norm": 0.41494375365376845,
+      "learning_rate": 3.677942699344149e-05,
+      "loss": 0.3748,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0149114631873253,
+      "grad_norm": 0.34592464055074684,
+      "learning_rate": 3.676216775975147e-05,
+      "loss": 0.375,
+      "step": 1089
+    },
+    {
+      "epoch": 1.015843429636533,
+      "grad_norm": 0.3807999035139463,
+      "learning_rate": 3.674490852606144e-05,
+      "loss": 0.3573,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0167753960857409,
+      "grad_norm": 0.31540064785999455,
+      "learning_rate": 3.672764929237142e-05,
+      "loss": 0.3359,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0177073625349486,
+      "grad_norm": 0.36588809024359786,
+      "learning_rate": 3.67103900586814e-05,
+      "loss": 0.3525,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0186393289841567,
+      "grad_norm": 0.36104851106235947,
+      "learning_rate": 3.669313082499137e-05,
+      "loss": 0.3425,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0195712954333644,
+      "grad_norm": 0.36660859482835156,
+      "learning_rate": 3.667587159130135e-05,
+      "loss": 0.3621,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0205032618825722,
+      "grad_norm": 0.32055013316542824,
+      "learning_rate": 3.665861235761132e-05,
+      "loss": 0.3503,
+      "step": 1095
+    },
+    {
+      "epoch": 1.02143522833178,
+      "grad_norm": 0.35200663441307123,
+      "learning_rate": 3.66413531239213e-05,
+      "loss": 0.33,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0223671947809878,
+      "grad_norm": 0.28643272669122566,
+      "learning_rate": 3.662409389023128e-05,
+      "loss": 0.3303,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0232991612301958,
+      "grad_norm": 0.3831715230980844,
+      "learning_rate": 3.660683465654125e-05,
+      "loss": 0.3532,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0242311276794036,
+      "grad_norm": 0.3379688145822085,
+      "learning_rate": 3.6589575422851225e-05,
+      "loss": 0.3374,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0251630941286114,
+      "grad_norm": 0.3688470067514589,
+      "learning_rate": 3.6572316189161206e-05,
+      "loss": 0.3477,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0260950605778192,
+      "grad_norm": 0.33293976889642984,
+      "learning_rate": 3.655505695547118e-05,
+      "loss": 0.34,
+      "step": 1101
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 0.40008597602742446,
+      "learning_rate": 3.6537797721781155e-05,
+      "loss": 0.3479,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0279589934762348,
+      "grad_norm": 0.2956818214441001,
+      "learning_rate": 3.6520538488091136e-05,
+      "loss": 0.3602,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0288909599254428,
+      "grad_norm": 0.3507888012421357,
+      "learning_rate": 3.650327925440111e-05,
+      "loss": 0.3401,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0298229263746506,
+      "grad_norm": 0.32159787130313494,
+      "learning_rate": 3.6486020020711085e-05,
+      "loss": 0.3413,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0307548928238583,
+      "grad_norm": 0.41184867605829134,
+      "learning_rate": 3.646876078702106e-05,
+      "loss": 0.3574,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0316868592730661,
+      "grad_norm": 0.3086992522845429,
+      "learning_rate": 3.645150155333103e-05,
+      "loss": 0.3336,
+      "step": 1107
+    },
+    {
+      "epoch": 1.032618825722274,
+      "grad_norm": 0.3121885034639912,
+      "learning_rate": 3.643424231964101e-05,
+      "loss": 0.3548,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0335507921714817,
+      "grad_norm": 0.29911411002430566,
+      "learning_rate": 3.641698308595098e-05,
+      "loss": 0.3534,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 0.3171804778871648,
+      "learning_rate": 3.639972385226096e-05,
+      "loss": 0.3331,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0354147250698975,
+      "grad_norm": 0.2875107038463093,
+      "learning_rate": 3.638246461857094e-05,
+      "loss": 0.3544,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0363466915191053,
+      "grad_norm": 0.3444027259425201,
+      "learning_rate": 3.636520538488091e-05,
+      "loss": 0.3245,
+      "step": 1112
+    },
+    {
+      "epoch": 1.037278657968313,
+      "grad_norm": 0.2832949167930656,
+      "learning_rate": 3.6347946151190886e-05,
+      "loss": 0.3477,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0382106244175209,
+      "grad_norm": 0.30461283898860525,
+      "learning_rate": 3.633068691750087e-05,
+      "loss": 0.3375,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0391425908667289,
+      "grad_norm": 0.32341996367144255,
+      "learning_rate": 3.631342768381084e-05,
+      "loss": 0.3692,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0400745573159367,
+      "grad_norm": 0.2961305588116086,
+      "learning_rate": 3.6296168450120815e-05,
+      "loss": 0.3298,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0410065237651445,
+      "grad_norm": 0.3570998127625885,
+      "learning_rate": 3.627890921643079e-05,
+      "loss": 0.3577,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0419384902143523,
+      "grad_norm": 0.31685693719669245,
+      "learning_rate": 3.626164998274077e-05,
+      "loss": 0.3361,
+      "step": 1118
+    },
+    {
+      "epoch": 1.04287045666356,
+      "grad_norm": 0.30330004282417494,
+      "learning_rate": 3.6244390749050745e-05,
+      "loss": 0.3568,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0438024231127678,
+      "grad_norm": 0.36647867291279845,
+      "learning_rate": 3.622713151536072e-05,
+      "loss": 0.3531,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0447343895619758,
+      "grad_norm": 0.33138800401942053,
+      "learning_rate": 3.62098722816707e-05,
+      "loss": 0.3397,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0456663560111836,
+      "grad_norm": 0.29676588369921186,
+      "learning_rate": 3.6192613047980675e-05,
+      "loss": 0.353,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0465983224603914,
+      "grad_norm": 0.3287135419049019,
+      "learning_rate": 3.617535381429065e-05,
+      "loss": 0.3422,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0475302889095992,
+      "grad_norm": 0.34415214716022335,
+      "learning_rate": 3.615809458060062e-05,
+      "loss": 0.3738,
+      "step": 1124
+    },
+    {
+      "epoch": 1.048462255358807,
+      "grad_norm": 0.3179288813997651,
+      "learning_rate": 3.6140835346910604e-05,
+      "loss": 0.3428,
+      "step": 1125
+    },
+    {
+      "epoch": 1.049394221808015,
+      "grad_norm": 0.30113716905445037,
+      "learning_rate": 3.612357611322058e-05,
+      "loss": 0.3418,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0503261882572228,
+      "grad_norm": 0.3135977871507,
+      "learning_rate": 3.610631687953055e-05,
+      "loss": 0.3613,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0512581547064306,
+      "grad_norm": 0.30017917079940726,
+      "learning_rate": 3.608905764584053e-05,
+      "loss": 0.3458,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0521901211556384,
+      "grad_norm": 0.31958854348949406,
+      "learning_rate": 3.60717984121505e-05,
+      "loss": 0.343,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0531220876048462,
+      "grad_norm": 0.3248221823323216,
+      "learning_rate": 3.6054539178460476e-05,
+      "loss": 0.3406,
+      "step": 1130
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 0.31303034303686667,
+      "learning_rate": 3.603727994477045e-05,
+      "loss": 0.3614,
+      "step": 1131
+    },
+    {
+      "epoch": 1.054986020503262,
+      "grad_norm": 0.4271794780670927,
+      "learning_rate": 3.602002071108043e-05,
+      "loss": 0.3553,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0559179869524697,
+      "grad_norm": 0.304150437692678,
+      "learning_rate": 3.6002761477390405e-05,
+      "loss": 0.3438,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0568499534016775,
+      "grad_norm": 0.38313521862200267,
+      "learning_rate": 3.598550224370038e-05,
+      "loss": 0.3786,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0577819198508853,
+      "grad_norm": 0.3670559482880488,
+      "learning_rate": 3.5968243010010354e-05,
+      "loss": 0.3771,
+      "step": 1135
+    },
+    {
+      "epoch": 1.058713886300093,
+      "grad_norm": 0.3759049041209594,
+      "learning_rate": 3.5950983776320335e-05,
+      "loss": 0.3489,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0596458527493011,
+      "grad_norm": 0.298735593872313,
+      "learning_rate": 3.593372454263031e-05,
+      "loss": 0.3705,
+      "step": 1137
+    },
+    {
+      "epoch": 1.060577819198509,
+      "grad_norm": 0.29586683097828786,
+      "learning_rate": 3.5916465308940284e-05,
+      "loss": 0.3471,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0615097856477167,
+      "grad_norm": 0.31371223542375126,
+      "learning_rate": 3.589920607525026e-05,
+      "loss": 0.3504,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0624417520969245,
+      "grad_norm": 0.33435287135306785,
+      "learning_rate": 3.588194684156024e-05,
+      "loss": 0.3538,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0633737185461323,
+      "grad_norm": 0.33346504143176964,
+      "learning_rate": 3.586468760787021e-05,
+      "loss": 0.339,
+      "step": 1141
+    },
+    {
+      "epoch": 1.06430568499534,
+      "grad_norm": 0.2897716019705582,
+      "learning_rate": 3.584742837418019e-05,
+      "loss": 0.3335,
+      "step": 1142
+    },
+    {
+      "epoch": 1.065237651444548,
+      "grad_norm": 0.37717602188430566,
+      "learning_rate": 3.583016914049017e-05,
+      "loss": 0.3565,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0661696178937559,
+      "grad_norm": 0.3070125711384053,
+      "learning_rate": 3.581290990680014e-05,
+      "loss": 0.346,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0671015843429636,
+      "grad_norm": 0.35057591296980695,
+      "learning_rate": 3.579565067311012e-05,
+      "loss": 0.3522,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0680335507921714,
+      "grad_norm": 0.3571317667219876,
+      "learning_rate": 3.577839143942009e-05,
+      "loss": 0.349,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 0.31145667448584485,
+      "learning_rate": 3.5761132205730066e-05,
+      "loss": 0.3403,
+      "step": 1147
+    },
+    {
+      "epoch": 1.0698974836905872,
+      "grad_norm": 0.2943779801174565,
+      "learning_rate": 3.574387297204005e-05,
+      "loss": 0.3374,
+      "step": 1148
+    },
+    {
+      "epoch": 1.070829450139795,
+      "grad_norm": 0.39466262333196267,
+      "learning_rate": 3.572661373835002e-05,
+      "loss": 0.3358,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0717614165890028,
+      "grad_norm": 0.3057728150261953,
+      "learning_rate": 3.5709354504659995e-05,
+      "loss": 0.3476,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0726933830382106,
+      "grad_norm": 0.3278913731915873,
+      "learning_rate": 3.569209527096997e-05,
+      "loss": 0.3456,
+      "step": 1151
+    },
+    {
+      "epoch": 1.0736253494874184,
+      "grad_norm": 0.29273861215492636,
+      "learning_rate": 3.5674836037279944e-05,
+      "loss": 0.3559,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0745573159366262,
+      "grad_norm": 0.35504034432229015,
+      "learning_rate": 3.565757680358992e-05,
+      "loss": 0.3589,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0754892823858342,
+      "grad_norm": 0.2955763507988634,
+      "learning_rate": 3.56403175698999e-05,
+      "loss": 0.3438,
+      "step": 1154
+    },
+    {
+      "epoch": 1.076421248835042,
+      "grad_norm": 0.3517965060180007,
+      "learning_rate": 3.5623058336209874e-05,
+      "loss": 0.3621,
+      "step": 1155
+    },
+    {
+      "epoch": 1.0773532152842498,
+      "grad_norm": 0.3174285208474431,
+      "learning_rate": 3.560579910251985e-05,
+      "loss": 0.3432,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0782851817334576,
+      "grad_norm": 0.38553777416258206,
+      "learning_rate": 3.558853986882982e-05,
+      "loss": 0.345,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0792171481826653,
+      "grad_norm": 0.34156039449854386,
+      "learning_rate": 3.55712806351398e-05,
+      "loss": 0.3493,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0801491146318734,
+      "grad_norm": 0.37187620027545926,
+      "learning_rate": 3.555402140144978e-05,
+      "loss": 0.3442,
+      "step": 1159
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 0.3003231772764512,
+      "learning_rate": 3.553676216775975e-05,
+      "loss": 0.3404,
+      "step": 1160
+    },
+    {
+      "epoch": 1.082013047530289,
+      "grad_norm": 0.3498381984777589,
+      "learning_rate": 3.551950293406973e-05,
+      "loss": 0.3473,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0829450139794967,
+      "grad_norm": 0.3244663165077475,
+      "learning_rate": 3.550224370037971e-05,
+      "loss": 0.3541,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0838769804287045,
+      "grad_norm": 0.3350699043985813,
+      "learning_rate": 3.548498446668968e-05,
+      "loss": 0.3893,
+      "step": 1163
+    },
+    {
+      "epoch": 1.0848089468779123,
+      "grad_norm": 0.4008957594779667,
+      "learning_rate": 3.5467725232999656e-05,
+      "loss": 0.348,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0857409133271203,
+      "grad_norm": 0.30351780833385394,
+      "learning_rate": 3.545046599930964e-05,
+      "loss": 0.3429,
+      "step": 1165
+    },
+    {
+      "epoch": 1.086672879776328,
+      "grad_norm": 0.35373664796302495,
+      "learning_rate": 3.543320676561961e-05,
+      "loss": 0.3524,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0876048462255359,
+      "grad_norm": 0.32739828279156064,
+      "learning_rate": 3.5415947531929585e-05,
+      "loss": 0.3544,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0885368126747437,
+      "grad_norm": 0.30106774570846007,
+      "learning_rate": 3.539868829823956e-05,
+      "loss": 0.3481,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0894687791239515,
+      "grad_norm": 0.3879238672544888,
+      "learning_rate": 3.5381429064549534e-05,
+      "loss": 0.3892,
+      "step": 1169
+    },
+    {
+      "epoch": 1.0904007455731595,
+      "grad_norm": 0.2821312613895853,
+      "learning_rate": 3.536416983085951e-05,
+      "loss": 0.3458,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0913327120223673,
+      "grad_norm": 0.34091351176710927,
+      "learning_rate": 3.534691059716948e-05,
+      "loss": 0.3675,
+      "step": 1171
+    },
+    {
+      "epoch": 1.092264678471575,
+      "grad_norm": 0.3284128554014004,
+      "learning_rate": 3.5329651363479464e-05,
+      "loss": 0.3431,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0931966449207828,
+      "grad_norm": 0.31225942944607477,
+      "learning_rate": 3.531239212978944e-05,
+      "loss": 0.3352,
+      "step": 1173
+    },
+    {
+      "epoch": 1.0941286113699906,
+      "grad_norm": 0.31821024077944077,
+      "learning_rate": 3.529513289609941e-05,
+      "loss": 0.3233,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0950605778191984,
+      "grad_norm": 0.3611563710125582,
+      "learning_rate": 3.5277873662409386e-05,
+      "loss": 0.353,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0959925442684064,
+      "grad_norm": 0.3356315462144002,
+      "learning_rate": 3.526061442871937e-05,
+      "loss": 0.3532,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0969245107176142,
+      "grad_norm": 0.3022977205443703,
+      "learning_rate": 3.524335519502934e-05,
+      "loss": 0.3496,
+      "step": 1177
+    },
+    {
+      "epoch": 1.097856477166822,
+      "grad_norm": 0.36095590277220235,
+      "learning_rate": 3.5226095961339316e-05,
+      "loss": 0.3535,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0987884436160298,
+      "grad_norm": 0.3192230842321107,
+      "learning_rate": 3.520883672764929e-05,
+      "loss": 0.3472,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0997204100652376,
+      "grad_norm": 0.3524730787920947,
+      "learning_rate": 3.519157749395927e-05,
+      "loss": 0.3481,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1006523765144456,
+      "grad_norm": 0.36165959274167553,
+      "learning_rate": 3.5174318260269246e-05,
+      "loss": 0.3456,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1015843429636534,
+      "grad_norm": 0.433210710962791,
+      "learning_rate": 3.515705902657922e-05,
+      "loss": 0.3501,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1025163094128612,
+      "grad_norm": 0.34551646621888593,
+      "learning_rate": 3.51397997928892e-05,
+      "loss": 0.3266,
+      "step": 1183
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 0.387234674399623,
+      "learning_rate": 3.5122540559199175e-05,
+      "loss": 0.3554,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1043802423112767,
+      "grad_norm": 0.4172735333918168,
+      "learning_rate": 3.510528132550915e-05,
+      "loss": 0.3599,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1053122087604845,
+      "grad_norm": 0.35595218466753764,
+      "learning_rate": 3.5088022091819124e-05,
+      "loss": 0.379,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1062441752096925,
+      "grad_norm": 0.4177008699836001,
+      "learning_rate": 3.5070762858129105e-05,
+      "loss": 0.3486,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1071761416589003,
+      "grad_norm": 0.349087424339184,
+      "learning_rate": 3.505350362443908e-05,
+      "loss": 0.3534,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.3321249012406462,
+      "learning_rate": 3.5036244390749054e-05,
+      "loss": 0.3291,
+      "step": 1189
+    },
+    {
+      "epoch": 1.109040074557316,
+      "grad_norm": 0.3608200032752507,
+      "learning_rate": 3.501898515705903e-05,
+      "loss": 0.3375,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1099720410065237,
+      "grad_norm": 0.3116234336948259,
+      "learning_rate": 3.5001725923369e-05,
+      "loss": 0.3518,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1109040074557317,
+      "grad_norm": 0.3553065714658423,
+      "learning_rate": 3.4984466689678977e-05,
+      "loss": 0.3583,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1118359739049395,
+      "grad_norm": 0.38199363179465806,
+      "learning_rate": 3.496720745598895e-05,
+      "loss": 0.355,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1127679403541473,
+      "grad_norm": 0.3035254602218242,
+      "learning_rate": 3.494994822229893e-05,
+      "loss": 0.3723,
+      "step": 1194
+    },
+    {
+      "epoch": 1.113699906803355,
+      "grad_norm": 0.35999463896933565,
+      "learning_rate": 3.4932688988608906e-05,
+      "loss": 0.3668,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1146318732525629,
+      "grad_norm": 0.3323932890187127,
+      "learning_rate": 3.491542975491888e-05,
+      "loss": 0.3549,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1155638397017706,
+      "grad_norm": 0.3663222769070753,
+      "learning_rate": 3.4898170521228855e-05,
+      "loss": 0.3681,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1164958061509787,
+      "grad_norm": 0.30578138515784403,
+      "learning_rate": 3.4880911287538836e-05,
+      "loss": 0.3377,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1174277726001864,
+      "grad_norm": 0.34243427780369945,
+      "learning_rate": 3.486365205384881e-05,
+      "loss": 0.3472,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1183597390493942,
+      "grad_norm": 0.3769040489422532,
+      "learning_rate": 3.4846392820158784e-05,
+      "loss": 0.3388,
+      "step": 1200
+    },
+    {
+      "epoch": 1.119291705498602,
+      "grad_norm": 0.28247657313599217,
+      "learning_rate": 3.4829133586468765e-05,
+      "loss": 0.3657,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1202236719478098,
+      "grad_norm": 0.3254297353856998,
+      "learning_rate": 3.481187435277874e-05,
+      "loss": 0.33,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1211556383970178,
+      "grad_norm": 0.33418561986561,
+      "learning_rate": 3.4794615119088714e-05,
+      "loss": 0.3409,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1220876048462256,
+      "grad_norm": 0.339226875262669,
+      "learning_rate": 3.477735588539869e-05,
+      "loss": 0.3494,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1230195712954334,
+      "grad_norm": 0.3983711230987905,
+      "learning_rate": 3.476009665170867e-05,
+      "loss": 0.3617,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1239515377446412,
+      "grad_norm": 0.34695395964526515,
+      "learning_rate": 3.4742837418018644e-05,
+      "loss": 0.3461,
+      "step": 1206
+    },
+    {
+      "epoch": 1.124883504193849,
+      "grad_norm": 0.3875058771546565,
+      "learning_rate": 3.472557818432862e-05,
+      "loss": 0.3507,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1258154706430568,
+      "grad_norm": 0.270969138581777,
+      "learning_rate": 3.470831895063859e-05,
+      "loss": 0.3426,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1267474370922648,
+      "grad_norm": 0.35742614121603966,
+      "learning_rate": 3.469105971694857e-05,
+      "loss": 0.3367,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1276794035414726,
+      "grad_norm": 0.35258924698330274,
+      "learning_rate": 3.467380048325855e-05,
+      "loss": 0.3633,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1286113699906803,
+      "grad_norm": 0.28966312926812676,
+      "learning_rate": 3.465654124956852e-05,
+      "loss": 0.3543,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1295433364398881,
+      "grad_norm": 0.3164375874284856,
+      "learning_rate": 3.4639282015878496e-05,
+      "loss": 0.3415,
+      "step": 1212
+    },
+    {
+      "epoch": 1.130475302889096,
+      "grad_norm": 0.3721347452654993,
+      "learning_rate": 3.462202278218847e-05,
+      "loss": 0.3703,
+      "step": 1213
+    },
+    {
+      "epoch": 1.131407269338304,
+      "grad_norm": 0.33945725258449966,
+      "learning_rate": 3.4604763548498445e-05,
+      "loss": 0.3575,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1323392357875117,
+      "grad_norm": 0.30647934492828494,
+      "learning_rate": 3.458750431480842e-05,
+      "loss": 0.3512,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1332712022367195,
+      "grad_norm": 0.2973242960233592,
+      "learning_rate": 3.45702450811184e-05,
+      "loss": 0.3483,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1342031686859273,
+      "grad_norm": 0.4181484947013587,
+      "learning_rate": 3.4552985847428374e-05,
+      "loss": 0.3457,
+      "step": 1217
+    },
+    {
+      "epoch": 1.135135135135135,
+      "grad_norm": 0.30642520061725437,
+      "learning_rate": 3.453572661373835e-05,
+      "loss": 0.3459,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1360671015843429,
+      "grad_norm": 0.3688742634946723,
+      "learning_rate": 3.451846738004832e-05,
+      "loss": 0.3417,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1369990680335509,
+      "grad_norm": 0.35629555820297326,
+      "learning_rate": 3.4501208146358304e-05,
+      "loss": 0.3877,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 0.3178249757910409,
+      "learning_rate": 3.448394891266828e-05,
+      "loss": 0.3643,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1388630009319665,
+      "grad_norm": 0.32562566181163977,
+      "learning_rate": 3.446668967897825e-05,
+      "loss": 0.3293,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1397949673811743,
+      "grad_norm": 0.3506617115135745,
+      "learning_rate": 3.4449430445288234e-05,
+      "loss": 0.3482,
+      "step": 1223
+    },
+    {
+      "epoch": 1.140726933830382,
+      "grad_norm": 0.3302689516686759,
+      "learning_rate": 3.443217121159821e-05,
+      "loss": 0.3467,
+      "step": 1224
+    },
+    {
+      "epoch": 1.14165890027959,
+      "grad_norm": 0.30851318301376074,
+      "learning_rate": 3.441491197790818e-05,
+      "loss": 0.3601,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1425908667287978,
+      "grad_norm": 0.3685137323980044,
+      "learning_rate": 3.4397652744218157e-05,
+      "loss": 0.3706,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1435228331780056,
+      "grad_norm": 0.29895165602947227,
+      "learning_rate": 3.438039351052814e-05,
+      "loss": 0.3443,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1444547996272134,
+      "grad_norm": 0.326309955637003,
+      "learning_rate": 3.436313427683811e-05,
+      "loss": 0.3558,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1453867660764212,
+      "grad_norm": 0.3778010366250315,
+      "learning_rate": 3.4345875043148086e-05,
+      "loss": 0.3399,
+      "step": 1229
+    },
+    {
+      "epoch": 1.146318732525629,
+      "grad_norm": 0.28345855468304537,
+      "learning_rate": 3.432861580945806e-05,
+      "loss": 0.3465,
+      "step": 1230
+    },
+    {
+      "epoch": 1.147250698974837,
+      "grad_norm": 0.29042981757251485,
+      "learning_rate": 3.431135657576804e-05,
+      "loss": 0.3711,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1481826654240448,
+      "grad_norm": 0.287784936200478,
+      "learning_rate": 3.4294097342078016e-05,
+      "loss": 0.3321,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1491146318732526,
+      "grad_norm": 0.28228794995607337,
+      "learning_rate": 3.427683810838799e-05,
+      "loss": 0.3381,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1500465983224604,
+      "grad_norm": 0.32683901816987254,
+      "learning_rate": 3.4259578874697964e-05,
+      "loss": 0.3486,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1509785647716682,
+      "grad_norm": 0.3117939615637795,
+      "learning_rate": 3.424231964100794e-05,
+      "loss": 0.3613,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1519105312208762,
+      "grad_norm": 0.2606804538081982,
+      "learning_rate": 3.422506040731791e-05,
+      "loss": 0.343,
+      "step": 1236
+    },
+    {
+      "epoch": 1.152842497670084,
+      "grad_norm": 0.2966626240174118,
+      "learning_rate": 3.420780117362789e-05,
+      "loss": 0.3442,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1537744641192917,
+      "grad_norm": 0.3052484081500861,
+      "learning_rate": 3.419054193993787e-05,
+      "loss": 0.3588,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1547064305684995,
+      "grad_norm": 0.2852423895146757,
+      "learning_rate": 3.417328270624784e-05,
+      "loss": 0.3448,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1556383970177073,
+      "grad_norm": 0.3817056431165827,
+      "learning_rate": 3.415602347255782e-05,
+      "loss": 0.3537,
+      "step": 1240
+    },
+    {
+      "epoch": 1.156570363466915,
+      "grad_norm": 0.3001873860298624,
+      "learning_rate": 3.41387642388678e-05,
+      "loss": 0.3683,
+      "step": 1241
+    },
+    {
+      "epoch": 1.157502329916123,
+      "grad_norm": 0.28847534362009664,
+      "learning_rate": 3.412150500517777e-05,
+      "loss": 0.3485,
+      "step": 1242
+    },
+    {
+      "epoch": 1.158434296365331,
+      "grad_norm": 0.3181870149025329,
+      "learning_rate": 3.410424577148775e-05,
+      "loss": 0.3556,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1593662628145387,
+      "grad_norm": 0.28736871638778216,
+      "learning_rate": 3.408698653779772e-05,
+      "loss": 0.3616,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1602982292637465,
+      "grad_norm": 0.2774127163685209,
+      "learning_rate": 3.40697273041077e-05,
+      "loss": 0.3451,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1612301957129543,
+      "grad_norm": 0.2859671189891517,
+      "learning_rate": 3.4052468070417676e-05,
+      "loss": 0.3552,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 0.3547906373451902,
+      "learning_rate": 3.403520883672765e-05,
+      "loss": 0.3713,
+      "step": 1247
+    },
+    {
+      "epoch": 1.16309412861137,
+      "grad_norm": 0.3404454477524794,
+      "learning_rate": 3.4017949603037625e-05,
+      "loss": 0.3522,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1640260950605779,
+      "grad_norm": 0.3061368934017998,
+      "learning_rate": 3.4000690369347606e-05,
+      "loss": 0.3386,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1649580615097856,
+      "grad_norm": 0.39281656858925373,
+      "learning_rate": 3.398343113565758e-05,
+      "loss": 0.3318,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1658900279589934,
+      "grad_norm": 0.29221735399825155,
+      "learning_rate": 3.3966171901967554e-05,
+      "loss": 0.33,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1668219944082012,
+      "grad_norm": 0.2904368135887982,
+      "learning_rate": 3.3948912668277536e-05,
+      "loss": 0.3218,
+      "step": 1252
+    },
+    {
+      "epoch": 1.167753960857409,
+      "grad_norm": 0.3484421348187277,
+      "learning_rate": 3.393165343458751e-05,
+      "loss": 0.3394,
+      "step": 1253
+    },
+    {
+      "epoch": 1.168685927306617,
+      "grad_norm": 0.31215586718999405,
+      "learning_rate": 3.3914394200897484e-05,
+      "loss": 0.3286,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1696178937558248,
+      "grad_norm": 0.2721408518600316,
+      "learning_rate": 3.389713496720746e-05,
+      "loss": 0.3246,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1705498602050326,
+      "grad_norm": 0.3135716517127376,
+      "learning_rate": 3.387987573351743e-05,
+      "loss": 0.3544,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1714818266542404,
+      "grad_norm": 0.27559392871376354,
+      "learning_rate": 3.386261649982741e-05,
+      "loss": 0.3451,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1724137931034484,
+      "grad_norm": 0.3065091941931709,
+      "learning_rate": 3.384535726613738e-05,
+      "loss": 0.3343,
+      "step": 1258
+    },
+    {
+      "epoch": 1.1733457595526562,
+      "grad_norm": 0.3657205838465487,
+      "learning_rate": 3.3828098032447356e-05,
+      "loss": 0.3692,
+      "step": 1259
+    },
+    {
+      "epoch": 1.174277726001864,
+      "grad_norm": 0.2740443040032835,
+      "learning_rate": 3.381083879875734e-05,
+      "loss": 0.3403,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1752096924510718,
+      "grad_norm": 0.35448607526216097,
+      "learning_rate": 3.379357956506731e-05,
+      "loss": 0.343,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1761416589002796,
+      "grad_norm": 0.31530984942073703,
+      "learning_rate": 3.3776320331377285e-05,
+      "loss": 0.3466,
+      "step": 1262
+    },
+    {
+      "epoch": 1.1770736253494873,
+      "grad_norm": 0.3542743545257663,
+      "learning_rate": 3.3759061097687266e-05,
+      "loss": 0.3328,
+      "step": 1263
+    },
+    {
+      "epoch": 1.1780055917986951,
+      "grad_norm": 0.3209515065921725,
+      "learning_rate": 3.374180186399724e-05,
+      "loss": 0.3557,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1789375582479031,
+      "grad_norm": 0.3168019501285976,
+      "learning_rate": 3.3724542630307215e-05,
+      "loss": 0.3467,
+      "step": 1265
+    },
+    {
+      "epoch": 1.179869524697111,
+      "grad_norm": 0.29663844224069214,
+      "learning_rate": 3.370728339661719e-05,
+      "loss": 0.3587,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1808014911463187,
+      "grad_norm": 0.29049234157334164,
+      "learning_rate": 3.369002416292717e-05,
+      "loss": 0.3584,
+      "step": 1267
+    },
+    {
+      "epoch": 1.1817334575955265,
+      "grad_norm": 0.30085930388126914,
+      "learning_rate": 3.3672764929237145e-05,
+      "loss": 0.3552,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1826654240447343,
+      "grad_norm": 0.2795222990352177,
+      "learning_rate": 3.365550569554712e-05,
+      "loss": 0.3636,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1835973904939423,
+      "grad_norm": 0.3207341890375812,
+      "learning_rate": 3.363824646185709e-05,
+      "loss": 0.3334,
+      "step": 1270
+    },
+    {
+      "epoch": 1.18452935694315,
+      "grad_norm": 0.3134295349312341,
+      "learning_rate": 3.3620987228167074e-05,
+      "loss": 0.3691,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1854613233923579,
+      "grad_norm": 0.3200305912793821,
+      "learning_rate": 3.360372799447705e-05,
+      "loss": 0.3525,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1863932898415657,
+      "grad_norm": 0.26740805944679197,
+      "learning_rate": 3.358646876078702e-05,
+      "loss": 0.3233,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1873252562907735,
+      "grad_norm": 0.30969274658474955,
+      "learning_rate": 3.3569209527097004e-05,
+      "loss": 0.3361,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1882572227399812,
+      "grad_norm": 0.26614232770010354,
+      "learning_rate": 3.355195029340698e-05,
+      "loss": 0.3594,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 0.3138823624679522,
+      "learning_rate": 3.353469105971695e-05,
+      "loss": 0.3479,
+      "step": 1276
+    },
+    {
+      "epoch": 1.190121155638397,
+      "grad_norm": 0.3025629127254504,
+      "learning_rate": 3.351743182602693e-05,
+      "loss": 0.3836,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1910531220876048,
+      "grad_norm": 0.33260641705685223,
+      "learning_rate": 3.35001725923369e-05,
+      "loss": 0.3794,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1919850885368126,
+      "grad_norm": 0.2978217010530524,
+      "learning_rate": 3.3482913358646875e-05,
+      "loss": 0.3472,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1929170549860204,
+      "grad_norm": 0.3217772014177255,
+      "learning_rate": 3.346565412495685e-05,
+      "loss": 0.3364,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1938490214352284,
+      "grad_norm": 0.3458494994995054,
+      "learning_rate": 3.344839489126683e-05,
+      "loss": 0.3623,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1947809878844362,
+      "grad_norm": 0.29687045381135485,
+      "learning_rate": 3.3431135657576805e-05,
+      "loss": 0.3573,
+      "step": 1282
+    },
+    {
+      "epoch": 1.195712954333644,
+      "grad_norm": 0.3398711416543203,
+      "learning_rate": 3.341387642388678e-05,
+      "loss": 0.3637,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1966449207828518,
+      "grad_norm": 0.39487233287524165,
+      "learning_rate": 3.3396617190196753e-05,
+      "loss": 0.3513,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1975768872320596,
+      "grad_norm": 0.3399680335729683,
+      "learning_rate": 3.3379357956506735e-05,
+      "loss": 0.3647,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1985088536812674,
+      "grad_norm": 0.3382989888192196,
+      "learning_rate": 3.336209872281671e-05,
+      "loss": 0.3698,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1994408201304754,
+      "grad_norm": 0.32210546924367683,
+      "learning_rate": 3.334483948912668e-05,
+      "loss": 0.3198,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2003727865796832,
+      "grad_norm": 0.3148325794292578,
+      "learning_rate": 3.332758025543666e-05,
+      "loss": 0.351,
+      "step": 1288
+    },
+    {
+      "epoch": 1.201304753028891,
+      "grad_norm": 0.3256351340521666,
+      "learning_rate": 3.331032102174664e-05,
+      "loss": 0.3315,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2022367194780987,
+      "grad_norm": 0.33411667898338715,
+      "learning_rate": 3.329306178805661e-05,
+      "loss": 0.3436,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2031686859273065,
+      "grad_norm": 0.3220499838617972,
+      "learning_rate": 3.327580255436659e-05,
+      "loss": 0.3555,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2041006523765145,
+      "grad_norm": 0.414701718055487,
+      "learning_rate": 3.325854332067657e-05,
+      "loss": 0.3216,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2050326188257223,
+      "grad_norm": 0.3254086224932852,
+      "learning_rate": 3.324128408698654e-05,
+      "loss": 0.3563,
+      "step": 1293
+    },
+    {
+      "epoch": 1.2059645852749301,
+      "grad_norm": 0.3797017901613958,
+      "learning_rate": 3.322402485329652e-05,
+      "loss": 0.3781,
+      "step": 1294
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 0.350044400260783,
+      "learning_rate": 3.320676561960649e-05,
+      "loss": 0.3284,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2078285181733457,
+      "grad_norm": 0.3183777309202059,
+      "learning_rate": 3.318950638591647e-05,
+      "loss": 0.3665,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2087604846225535,
+      "grad_norm": 0.37418021745428287,
+      "learning_rate": 3.3172247152226446e-05,
+      "loss": 0.34,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2096924510717615,
+      "grad_norm": 0.33581970304742187,
+      "learning_rate": 3.315498791853642e-05,
+      "loss": 0.3388,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2106244175209693,
+      "grad_norm": 0.34564617627631666,
+      "learning_rate": 3.3137728684846395e-05,
+      "loss": 0.3544,
+      "step": 1299
+    },
+    {
+      "epoch": 1.211556383970177,
+      "grad_norm": 0.45011555832537703,
+      "learning_rate": 3.312046945115637e-05,
+      "loss": 0.3822,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2124883504193849,
+      "grad_norm": 0.34395029568008156,
+      "learning_rate": 3.3103210217466344e-05,
+      "loss": 0.3651,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2134203168685926,
+      "grad_norm": 0.3583820993736549,
+      "learning_rate": 3.308595098377632e-05,
+      "loss": 0.3639,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2143522833178007,
+      "grad_norm": 0.3621617340275819,
+      "learning_rate": 3.30686917500863e-05,
+      "loss": 0.3517,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2152842497670084,
+      "grad_norm": 0.3343405490046225,
+      "learning_rate": 3.305143251639627e-05,
+      "loss": 0.3327,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 0.3810294020045999,
+      "learning_rate": 3.303417328270625e-05,
+      "loss": 0.3608,
+      "step": 1305
+    },
+    {
+      "epoch": 1.217148182665424,
+      "grad_norm": 0.3542066199769166,
+      "learning_rate": 3.301691404901622e-05,
+      "loss": 0.3496,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2180801491146318,
+      "grad_norm": 0.3482055686853539,
+      "learning_rate": 3.29996548153262e-05,
+      "loss": 0.3507,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2190121155638396,
+      "grad_norm": 0.3879541245467854,
+      "learning_rate": 3.298239558163618e-05,
+      "loss": 0.3374,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2199440820130476,
+      "grad_norm": 0.327346244734531,
+      "learning_rate": 3.296513634794615e-05,
+      "loss": 0.3397,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2208760484622554,
+      "grad_norm": 0.3039153182240888,
+      "learning_rate": 3.2947877114256126e-05,
+      "loss": 0.3428,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2218080149114632,
+      "grad_norm": 0.31633294608823176,
+      "learning_rate": 3.293061788056611e-05,
+      "loss": 0.3672,
+      "step": 1311
+    },
+    {
+      "epoch": 1.222739981360671,
+      "grad_norm": 0.34504626794119647,
+      "learning_rate": 3.291335864687608e-05,
+      "loss": 0.3772,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2236719478098788,
+      "grad_norm": 0.32630653335257054,
+      "learning_rate": 3.2896099413186055e-05,
+      "loss": 0.326,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2246039142590868,
+      "grad_norm": 0.29858538641723087,
+      "learning_rate": 3.2878840179496036e-05,
+      "loss": 0.3705,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2255358807082946,
+      "grad_norm": 0.320615809914272,
+      "learning_rate": 3.286158094580601e-05,
+      "loss": 0.3786,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2264678471575023,
+      "grad_norm": 0.35531709139050477,
+      "learning_rate": 3.2844321712115985e-05,
+      "loss": 0.3419,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2273998136067101,
+      "grad_norm": 0.2820094703792751,
+      "learning_rate": 3.282706247842596e-05,
+      "loss": 0.3515,
+      "step": 1317
+    },
+    {
+      "epoch": 1.228331780055918,
+      "grad_norm": 0.35737966923919395,
+      "learning_rate": 3.280980324473594e-05,
+      "loss": 0.3368,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2292637465051257,
+      "grad_norm": 0.3639905409580076,
+      "learning_rate": 3.2792544011045915e-05,
+      "loss": 0.3589,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2301957129543337,
+      "grad_norm": 0.27278281208951344,
+      "learning_rate": 3.277528477735589e-05,
+      "loss": 0.3602,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2311276794035415,
+      "grad_norm": 0.3490970039051147,
+      "learning_rate": 3.275802554366586e-05,
+      "loss": 0.374,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2320596458527493,
+      "grad_norm": 0.322216071655853,
+      "learning_rate": 3.274076630997584e-05,
+      "loss": 0.3513,
+      "step": 1322
+    },
+    {
+      "epoch": 1.232991612301957,
+      "grad_norm": 0.31000382785519837,
+      "learning_rate": 3.272350707628581e-05,
+      "loss": 0.349,
+      "step": 1323
+    },
+    {
+      "epoch": 1.2339235787511649,
+      "grad_norm": 0.34770470304631007,
+      "learning_rate": 3.2706247842595786e-05,
+      "loss": 0.3667,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2348555452003729,
+      "grad_norm": 0.3302792635452052,
+      "learning_rate": 3.268898860890577e-05,
+      "loss": 0.3435,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2357875116495807,
+      "grad_norm": 0.31044331665370356,
+      "learning_rate": 3.267172937521574e-05,
+      "loss": 0.3428,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2367194780987885,
+      "grad_norm": 0.44153248277050827,
+      "learning_rate": 3.2654470141525716e-05,
+      "loss": 0.3477,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2376514445479962,
+      "grad_norm": 0.35526246973083553,
+      "learning_rate": 3.263721090783569e-05,
+      "loss": 0.3464,
+      "step": 1328
+    },
+    {
+      "epoch": 1.238583410997204,
+      "grad_norm": 0.3073531163152437,
+      "learning_rate": 3.261995167414567e-05,
+      "loss": 0.3521,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2395153774464118,
+      "grad_norm": 0.3423810822543725,
+      "learning_rate": 3.2602692440455645e-05,
+      "loss": 0.3367,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2404473438956198,
+      "grad_norm": 0.2989013862019505,
+      "learning_rate": 3.258543320676562e-05,
+      "loss": 0.3576,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 0.2883855136641901,
+      "learning_rate": 3.25681739730756e-05,
+      "loss": 0.3324,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2423112767940354,
+      "grad_norm": 0.32797198441728515,
+      "learning_rate": 3.2550914739385575e-05,
+      "loss": 0.3388,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2432432432432432,
+      "grad_norm": 0.3028589043601128,
+      "learning_rate": 3.253365550569555e-05,
+      "loss": 0.3492,
+      "step": 1334
+    },
+    {
+      "epoch": 1.244175209692451,
+      "grad_norm": 0.3082866997739984,
+      "learning_rate": 3.2516396272005524e-05,
+      "loss": 0.3604,
+      "step": 1335
+    },
+    {
+      "epoch": 1.245107176141659,
+      "grad_norm": 0.32527772733351495,
+      "learning_rate": 3.2499137038315505e-05,
+      "loss": 0.3496,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2460391425908668,
+      "grad_norm": 0.29477693169997377,
+      "learning_rate": 3.248187780462548e-05,
+      "loss": 0.3441,
+      "step": 1337
+    },
+    {
+      "epoch": 1.2469711090400746,
+      "grad_norm": 0.37654380530926895,
+      "learning_rate": 3.246461857093545e-05,
+      "loss": 0.3348,
+      "step": 1338
+    },
+    {
+      "epoch": 1.2479030754892824,
+      "grad_norm": 0.3108536174342759,
+      "learning_rate": 3.244735933724543e-05,
+      "loss": 0.3816,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2488350419384902,
+      "grad_norm": 0.44663830747475436,
+      "learning_rate": 3.243010010355541e-05,
+      "loss": 0.3561,
+      "step": 1340
+    },
+    {
+      "epoch": 1.249767008387698,
+      "grad_norm": 0.3412572130924294,
+      "learning_rate": 3.241284086986538e-05,
+      "loss": 0.3408,
+      "step": 1341
+    },
+    {
+      "epoch": 1.250698974836906,
+      "grad_norm": 0.4778646990490752,
+      "learning_rate": 3.239558163617536e-05,
+      "loss": 0.3483,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2516309412861137,
+      "grad_norm": 0.3596025153349949,
+      "learning_rate": 3.237832240248533e-05,
+      "loss": 0.3308,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2525629077353215,
+      "grad_norm": 0.37641470094837665,
+      "learning_rate": 3.2361063168795306e-05,
+      "loss": 0.3342,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2534948741845293,
+      "grad_norm": 0.31039806440280066,
+      "learning_rate": 3.234380393510528e-05,
+      "loss": 0.3489,
+      "step": 1345
+    },
+    {
+      "epoch": 1.254426840633737,
+      "grad_norm": 0.37817675347078017,
+      "learning_rate": 3.2326544701415254e-05,
+      "loss": 0.3364,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2553588070829451,
+      "grad_norm": 0.3322230800288191,
+      "learning_rate": 3.2309285467725235e-05,
+      "loss": 0.334,
+      "step": 1347
+    },
+    {
+      "epoch": 1.256290773532153,
+      "grad_norm": 0.3024905265449303,
+      "learning_rate": 3.229202623403521e-05,
+      "loss": 0.3595,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2572227399813607,
+      "grad_norm": 0.3540229871096113,
+      "learning_rate": 3.2274767000345184e-05,
+      "loss": 0.3619,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2581547064305685,
+      "grad_norm": 0.26537517193147486,
+      "learning_rate": 3.225750776665516e-05,
+      "loss": 0.3464,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2590866728797763,
+      "grad_norm": 0.3685094246309665,
+      "learning_rate": 3.224024853296514e-05,
+      "loss": 0.3601,
+      "step": 1351
+    },
+    {
+      "epoch": 1.260018639328984,
+      "grad_norm": 0.27170623416662104,
+      "learning_rate": 3.2222989299275114e-05,
+      "loss": 0.3632,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2609506057781918,
+      "grad_norm": 0.3097454638493056,
+      "learning_rate": 3.220573006558509e-05,
+      "loss": 0.3643,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2618825722273999,
+      "grad_norm": 0.3346665038404999,
+      "learning_rate": 3.218847083189507e-05,
+      "loss": 0.3616,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2628145386766076,
+      "grad_norm": 0.3064946035929941,
+      "learning_rate": 3.217121159820504e-05,
+      "loss": 0.3398,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2637465051258154,
+      "grad_norm": 0.33968055170789135,
+      "learning_rate": 3.215395236451502e-05,
+      "loss": 0.3771,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2646784715750232,
+      "grad_norm": 0.32762285190351037,
+      "learning_rate": 3.213669313082499e-05,
+      "loss": 0.3397,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2656104380242312,
+      "grad_norm": 0.3063299530930208,
+      "learning_rate": 3.211943389713497e-05,
+      "loss": 0.3538,
+      "step": 1358
+    },
+    {
+      "epoch": 1.266542404473439,
+      "grad_norm": 0.29894595936061397,
+      "learning_rate": 3.210217466344495e-05,
+      "loss": 0.3173,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2674743709226468,
+      "grad_norm": 0.3014705677279258,
+      "learning_rate": 3.208491542975492e-05,
+      "loss": 0.3478,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2684063373718546,
+      "grad_norm": 0.34872400438707046,
+      "learning_rate": 3.2067656196064896e-05,
+      "loss": 0.3661,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2693383038210624,
+      "grad_norm": 0.34576636000876765,
+      "learning_rate": 3.205039696237487e-05,
+      "loss": 0.3567,
+      "step": 1362
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 0.2980466588776678,
+      "learning_rate": 3.2033137728684844e-05,
+      "loss": 0.3315,
+      "step": 1363
+    },
+    {
+      "epoch": 1.271202236719478,
+      "grad_norm": 0.3343104721652589,
+      "learning_rate": 3.201587849499482e-05,
+      "loss": 0.3485,
+      "step": 1364
+    },
+    {
+      "epoch": 1.272134203168686,
+      "grad_norm": 0.30123852809881396,
+      "learning_rate": 3.19986192613048e-05,
+      "loss": 0.3593,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2730661696178938,
+      "grad_norm": 0.33266900444126934,
+      "learning_rate": 3.1981360027614774e-05,
+      "loss": 0.3498,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2739981360671015,
+      "grad_norm": 0.30295451338045964,
+      "learning_rate": 3.196410079392475e-05,
+      "loss": 0.3514,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2749301025163093,
+      "grad_norm": 0.35224739882254963,
+      "learning_rate": 3.194684156023472e-05,
+      "loss": 0.3669,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 0.34863282955957775,
+      "learning_rate": 3.1929582326544704e-05,
+      "loss": 0.3481,
+      "step": 1369
+    },
+    {
+      "epoch": 1.2767940354147251,
+      "grad_norm": 0.35784288081404175,
+      "learning_rate": 3.191232309285468e-05,
+      "loss": 0.3411,
+      "step": 1370
+    },
+    {
+      "epoch": 1.277726001863933,
+      "grad_norm": 0.3134203448957504,
+      "learning_rate": 3.189506385916465e-05,
+      "loss": 0.3494,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2786579683131407,
+      "grad_norm": 0.3637126716164138,
+      "learning_rate": 3.187780462547463e-05,
+      "loss": 0.3456,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2795899347623485,
+      "grad_norm": 0.3264087842370889,
+      "learning_rate": 3.186054539178461e-05,
+      "loss": 0.3314,
+      "step": 1373
+    },
+    {
+      "epoch": 1.2805219012115563,
+      "grad_norm": 0.3475071958704886,
+      "learning_rate": 3.184328615809458e-05,
+      "loss": 0.3524,
+      "step": 1374
+    },
+    {
+      "epoch": 1.281453867660764,
+      "grad_norm": 0.34327651744095405,
+      "learning_rate": 3.1826026924404556e-05,
+      "loss": 0.3276,
+      "step": 1375
+    },
+    {
+      "epoch": 1.282385834109972,
+      "grad_norm": 0.3563499519647939,
+      "learning_rate": 3.180876769071454e-05,
+      "loss": 0.3516,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2833178005591799,
+      "grad_norm": 0.41482824117848777,
+      "learning_rate": 3.179150845702451e-05,
+      "loss": 0.3336,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2842497670083877,
+      "grad_norm": 0.31611355558665594,
+      "learning_rate": 3.1774249223334486e-05,
+      "loss": 0.365,
+      "step": 1378
+    },
+    {
+      "epoch": 1.2851817334575955,
+      "grad_norm": 0.38734841793288594,
+      "learning_rate": 3.175698998964446e-05,
+      "loss": 0.355,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2861136999068035,
+      "grad_norm": 0.33957443875266796,
+      "learning_rate": 3.173973075595444e-05,
+      "loss": 0.3749,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2870456663560113,
+      "grad_norm": 0.2971961818276081,
+      "learning_rate": 3.1722471522264415e-05,
+      "loss": 0.3526,
+      "step": 1381
+    },
+    {
+      "epoch": 1.287977632805219,
+      "grad_norm": 0.3498091703589901,
+      "learning_rate": 3.170521228857439e-05,
+      "loss": 0.3451,
+      "step": 1382
+    },
+    {
+      "epoch": 1.2889095992544268,
+      "grad_norm": 0.30840583686843875,
+      "learning_rate": 3.1687953054884364e-05,
+      "loss": 0.3697,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2898415657036346,
+      "grad_norm": 0.37623922844860247,
+      "learning_rate": 3.167069382119434e-05,
+      "loss": 0.3564,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2907735321528424,
+      "grad_norm": 0.3118551376408831,
+      "learning_rate": 3.165343458750431e-05,
+      "loss": 0.3532,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2917054986020502,
+      "grad_norm": 0.3315448041265437,
+      "learning_rate": 3.163617535381429e-05,
+      "loss": 0.3462,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2926374650512582,
+      "grad_norm": 0.3312570994091037,
+      "learning_rate": 3.161891612012427e-05,
+      "loss": 0.3428,
+      "step": 1387
+    },
+    {
+      "epoch": 1.293569431500466,
+      "grad_norm": 0.3645236817402315,
+      "learning_rate": 3.160165688643424e-05,
+      "loss": 0.364,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2945013979496738,
+      "grad_norm": 0.2971775815574898,
+      "learning_rate": 3.1584397652744217e-05,
+      "loss": 0.3433,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2954333643988816,
+      "grad_norm": 0.3397502371159504,
+      "learning_rate": 3.156713841905419e-05,
+      "loss": 0.342,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2963653308480896,
+      "grad_norm": 0.3007181534473731,
+      "learning_rate": 3.154987918536417e-05,
+      "loss": 0.3403,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 0.34043975594069736,
+      "learning_rate": 3.1532619951674146e-05,
+      "loss": 0.3488,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2982292637465052,
+      "grad_norm": 0.2958860204782596,
+      "learning_rate": 3.151536071798412e-05,
+      "loss": 0.3512,
+      "step": 1393
+    },
+    {
+      "epoch": 1.299161230195713,
+      "grad_norm": 0.34485045880953363,
+      "learning_rate": 3.14981014842941e-05,
+      "loss": 0.371,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3000931966449207,
+      "grad_norm": 0.27898320006576444,
+      "learning_rate": 3.1480842250604076e-05,
+      "loss": 0.3478,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3010251630941285,
+      "grad_norm": 0.3742623521323532,
+      "learning_rate": 3.146358301691405e-05,
+      "loss": 0.3461,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3019571295433363,
+      "grad_norm": 0.3772099241918813,
+      "learning_rate": 3.1446323783224024e-05,
+      "loss": 0.3466,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3028890959925443,
+      "grad_norm": 0.3762810573440586,
+      "learning_rate": 3.1429064549534006e-05,
+      "loss": 0.3412,
+      "step": 1398
+    },
+    {
+      "epoch": 1.303821062441752,
+      "grad_norm": 0.32802761820464094,
+      "learning_rate": 3.141180531584398e-05,
+      "loss": 0.349,
+      "step": 1399
+    },
+    {
+      "epoch": 1.30475302889096,
+      "grad_norm": 0.3761902403993999,
+      "learning_rate": 3.1394546082153954e-05,
+      "loss": 0.3473,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3056849953401677,
+      "grad_norm": 0.3328190508471384,
+      "learning_rate": 3.137728684846393e-05,
+      "loss": 0.3626,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3066169617893757,
+      "grad_norm": 0.38078722972152906,
+      "learning_rate": 3.136002761477391e-05,
+      "loss": 0.3656,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3075489282385835,
+      "grad_norm": 0.2710719066017743,
+      "learning_rate": 3.1342768381083884e-05,
+      "loss": 0.3305,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3084808946877913,
+      "grad_norm": 0.3549107381051092,
+      "learning_rate": 3.132550914739386e-05,
+      "loss": 0.3674,
+      "step": 1404
+    },
+    {
+      "epoch": 1.309412861136999,
+      "grad_norm": 0.35076354501868956,
+      "learning_rate": 3.130824991370383e-05,
+      "loss": 0.358,
+      "step": 1405
+    },
+    {
+      "epoch": 1.3103448275862069,
+      "grad_norm": 0.2898841397164584,
+      "learning_rate": 3.129099068001381e-05,
+      "loss": 0.3396,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3112767940354146,
+      "grad_norm": 0.35211619105707087,
+      "learning_rate": 3.127373144632378e-05,
+      "loss": 0.3347,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3122087604846224,
+      "grad_norm": 0.3100358103312664,
+      "learning_rate": 3.1256472212633755e-05,
+      "loss": 0.3447,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3131407269338304,
+      "grad_norm": 0.2564215605028504,
+      "learning_rate": 3.1239212978943736e-05,
+      "loss": 0.3389,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3140726933830382,
+      "grad_norm": 0.32869173196861123,
+      "learning_rate": 3.122195374525371e-05,
+      "loss": 0.3667,
+      "step": 1410
+    },
+    {
+      "epoch": 1.315004659832246,
+      "grad_norm": 0.29686748198533824,
+      "learning_rate": 3.1204694511563685e-05,
+      "loss": 0.3542,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3159366262814538,
+      "grad_norm": 0.29952101301789147,
+      "learning_rate": 3.118743527787366e-05,
+      "loss": 0.3403,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3168685927306618,
+      "grad_norm": 0.303931210395446,
+      "learning_rate": 3.117017604418364e-05,
+      "loss": 0.3436,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3178005591798696,
+      "grad_norm": 0.31136403499345644,
+      "learning_rate": 3.1152916810493615e-05,
+      "loss": 0.3709,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3187325256290774,
+      "grad_norm": 0.29072515531064475,
+      "learning_rate": 3.113565757680359e-05,
+      "loss": 0.3539,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3196644920782852,
+      "grad_norm": 0.29223049742626656,
+      "learning_rate": 3.111839834311357e-05,
+      "loss": 0.3426,
+      "step": 1416
+    },
+    {
+      "epoch": 1.320596458527493,
+      "grad_norm": 0.2793481881170594,
+      "learning_rate": 3.1101139109423544e-05,
+      "loss": 0.3233,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3215284249767008,
+      "grad_norm": 0.3034197154095062,
+      "learning_rate": 3.108387987573352e-05,
+      "loss": 0.358,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3224603914259085,
+      "grad_norm": 0.3145378725587504,
+      "learning_rate": 3.106662064204349e-05,
+      "loss": 0.33,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3233923578751166,
+      "grad_norm": 0.301008044274595,
+      "learning_rate": 3.1049361408353474e-05,
+      "loss": 0.3454,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 0.38600495552853625,
+      "learning_rate": 3.103210217466345e-05,
+      "loss": 0.3327,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3252562907735321,
+      "grad_norm": 0.2993180956379451,
+      "learning_rate": 3.101484294097342e-05,
+      "loss": 0.3423,
+      "step": 1422
+    },
+    {
+      "epoch": 1.32618825722274,
+      "grad_norm": 0.3583779911451612,
+      "learning_rate": 3.0997583707283403e-05,
+      "loss": 0.3609,
+      "step": 1423
+    },
+    {
+      "epoch": 1.327120223671948,
+      "grad_norm": 0.35830731946669375,
+      "learning_rate": 3.098032447359338e-05,
+      "loss": 0.3459,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3280521901211557,
+      "grad_norm": 0.25055558119802995,
+      "learning_rate": 3.096306523990335e-05,
+      "loss": 0.3569,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3289841565703635,
+      "grad_norm": 0.33279573855123945,
+      "learning_rate": 3.0945806006213326e-05,
+      "loss": 0.3893,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3299161230195713,
+      "grad_norm": 0.3060066484430594,
+      "learning_rate": 3.09285467725233e-05,
+      "loss": 0.3426,
+      "step": 1427
+    },
+    {
+      "epoch": 1.330848089468779,
+      "grad_norm": 0.2993077445749117,
+      "learning_rate": 3.0911287538833275e-05,
+      "loss": 0.3227,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3317800559179869,
+      "grad_norm": 0.3403364309691639,
+      "learning_rate": 3.089402830514325e-05,
+      "loss": 0.3697,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3327120223671947,
+      "grad_norm": 0.27685934738971046,
+      "learning_rate": 3.0876769071453223e-05,
+      "loss": 0.3576,
+      "step": 1430
+    },
+    {
+      "epoch": 1.3336439888164027,
+      "grad_norm": 0.31583187419427755,
+      "learning_rate": 3.0859509837763205e-05,
+      "loss": 0.3403,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3345759552656105,
+      "grad_norm": 0.33028579955463877,
+      "learning_rate": 3.084225060407318e-05,
+      "loss": 0.3791,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3355079217148182,
+      "grad_norm": 0.28220083626001413,
+      "learning_rate": 3.082499137038315e-05,
+      "loss": 0.3588,
+      "step": 1433
+    },
+    {
+      "epoch": 1.336439888164026,
+      "grad_norm": 0.2967979738612457,
+      "learning_rate": 3.0807732136693134e-05,
+      "loss": 0.3547,
+      "step": 1434
+    },
+    {
+      "epoch": 1.337371854613234,
+      "grad_norm": 0.3057704255931003,
+      "learning_rate": 3.079047290300311e-05,
+      "loss": 0.3385,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3383038210624418,
+      "grad_norm": 0.2764463321907883,
+      "learning_rate": 3.077321366931308e-05,
+      "loss": 0.3485,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3392357875116496,
+      "grad_norm": 0.29354379162980093,
+      "learning_rate": 3.075595443562306e-05,
+      "loss": 0.3452,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3401677539608574,
+      "grad_norm": 0.29222358208149346,
+      "learning_rate": 3.073869520193304e-05,
+      "loss": 0.3673,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3410997204100652,
+      "grad_norm": 0.32515621228755465,
+      "learning_rate": 3.072143596824301e-05,
+      "loss": 0.3514,
+      "step": 1439
+    },
+    {
+      "epoch": 1.342031686859273,
+      "grad_norm": 0.3070846587960357,
+      "learning_rate": 3.070417673455299e-05,
+      "loss": 0.3514,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3429636533084808,
+      "grad_norm": 0.3397085488122223,
+      "learning_rate": 3.068691750086296e-05,
+      "loss": 0.3713,
+      "step": 1441
+    },
+    {
+      "epoch": 1.3438956197576888,
+      "grad_norm": 0.2806325332439306,
+      "learning_rate": 3.066965826717294e-05,
+      "loss": 0.3419,
+      "step": 1442
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 0.29272137707130547,
+      "learning_rate": 3.0652399033482916e-05,
+      "loss": 0.3604,
+      "step": 1443
+    },
+    {
+      "epoch": 1.3457595526561044,
+      "grad_norm": 0.3279657841427319,
+      "learning_rate": 3.063513979979289e-05,
+      "loss": 0.3741,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3466915191053122,
+      "grad_norm": 0.2831140866453429,
+      "learning_rate": 3.061788056610287e-05,
+      "loss": 0.3398,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3476234855545202,
+      "grad_norm": 0.28890650539944046,
+      "learning_rate": 3.0600621332412846e-05,
+      "loss": 0.3466,
+      "step": 1446
+    },
+    {
+      "epoch": 1.348555452003728,
+      "grad_norm": 0.27826130427882595,
+      "learning_rate": 3.058336209872282e-05,
+      "loss": 0.3298,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3494874184529357,
+      "grad_norm": 0.28740677383308594,
+      "learning_rate": 3.0566102865032795e-05,
+      "loss": 0.3458,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3504193849021435,
+      "grad_norm": 0.30737752855005246,
+      "learning_rate": 3.054884363134277e-05,
+      "loss": 0.3395,
+      "step": 1449
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 0.3231346066956203,
+      "learning_rate": 3.053158439765274e-05,
+      "loss": 0.3548,
+      "step": 1450
+    },
+    {
+      "epoch": 1.352283317800559,
+      "grad_norm": 0.30597427298599167,
+      "learning_rate": 3.051432516396272e-05,
+      "loss": 0.3608,
+      "step": 1451
+    },
+    {
+      "epoch": 1.353215284249767,
+      "grad_norm": 0.37227599474332634,
+      "learning_rate": 3.0497065930272695e-05,
+      "loss": 0.3686,
+      "step": 1452
+    },
+    {
+      "epoch": 1.354147250698975,
+      "grad_norm": 0.30786019588189817,
+      "learning_rate": 3.0479806696582676e-05,
+      "loss": 0.3528,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3550792171481827,
+      "grad_norm": 0.30521235514938183,
+      "learning_rate": 3.046254746289265e-05,
+      "loss": 0.334,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3560111835973905,
+      "grad_norm": 0.2971361884816148,
+      "learning_rate": 3.0445288229202625e-05,
+      "loss": 0.338,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3569431500465983,
+      "grad_norm": 0.3416454266928099,
+      "learning_rate": 3.0428028995512602e-05,
+      "loss": 0.3507,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3578751164958063,
+      "grad_norm": 0.3080399943015756,
+      "learning_rate": 3.0410769761822577e-05,
+      "loss": 0.3706,
+      "step": 1457
+    },
+    {
+      "epoch": 1.358807082945014,
+      "grad_norm": 0.3202499394430549,
+      "learning_rate": 3.039351052813255e-05,
+      "loss": 0.3606,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3597390493942219,
+      "grad_norm": 0.3871433083820658,
+      "learning_rate": 3.0376251294442525e-05,
+      "loss": 0.372,
+      "step": 1459
+    },
+    {
+      "epoch": 1.3606710158434296,
+      "grad_norm": 0.30368610319677763,
+      "learning_rate": 3.0358992060752506e-05,
+      "loss": 0.3302,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3616029822926374,
+      "grad_norm": 0.33437310913277424,
+      "learning_rate": 3.034173282706248e-05,
+      "loss": 0.3474,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3625349487418452,
+      "grad_norm": 0.2991930396207103,
+      "learning_rate": 3.0324473593372455e-05,
+      "loss": 0.3452,
+      "step": 1462
+    },
+    {
+      "epoch": 1.363466915191053,
+      "grad_norm": 0.3750966907734571,
+      "learning_rate": 3.0307214359682433e-05,
+      "loss": 0.3611,
+      "step": 1463
+    },
+    {
+      "epoch": 1.364398881640261,
+      "grad_norm": 0.4161405880168155,
+      "learning_rate": 3.0289955125992407e-05,
+      "loss": 0.3452,
+      "step": 1464
+    },
+    {
+      "epoch": 1.3653308480894688,
+      "grad_norm": 0.30527206221788733,
+      "learning_rate": 3.027269589230238e-05,
+      "loss": 0.3556,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3662628145386766,
+      "grad_norm": 0.41369117037877606,
+      "learning_rate": 3.0255436658612356e-05,
+      "loss": 0.3574,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3671947809878844,
+      "grad_norm": 0.3649024974154033,
+      "learning_rate": 3.0238177424922337e-05,
+      "loss": 0.3331,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3681267474370924,
+      "grad_norm": 0.3291566155725294,
+      "learning_rate": 3.022091819123231e-05,
+      "loss": 0.3511,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3690587138863002,
+      "grad_norm": 0.4479024240094849,
+      "learning_rate": 3.0203658957542285e-05,
+      "loss": 0.3518,
+      "step": 1469
+    },
+    {
+      "epoch": 1.369990680335508,
+      "grad_norm": 0.314676416632654,
+      "learning_rate": 3.018639972385226e-05,
+      "loss": 0.3476,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3709226467847158,
+      "grad_norm": 0.36764438655920395,
+      "learning_rate": 3.016914049016224e-05,
+      "loss": 0.3479,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3718546132339235,
+      "grad_norm": 0.41059510438710806,
+      "learning_rate": 3.0151881256472215e-05,
+      "loss": 0.3592,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3727865796831313,
+      "grad_norm": 0.37968181632384507,
+      "learning_rate": 3.013462202278219e-05,
+      "loss": 0.3578,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3737185461323391,
+      "grad_norm": 0.3273485415959064,
+      "learning_rate": 3.0117362789092167e-05,
+      "loss": 0.3487,
+      "step": 1474
+    },
+    {
+      "epoch": 1.3746505125815471,
+      "grad_norm": 0.35498293880946613,
+      "learning_rate": 3.010010355540214e-05,
+      "loss": 0.336,
+      "step": 1475
+    },
+    {
+      "epoch": 1.375582479030755,
+      "grad_norm": 0.3093601286476094,
+      "learning_rate": 3.0082844321712115e-05,
+      "loss": 0.3342,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3765144454799627,
+      "grad_norm": 0.29172326735544046,
+      "learning_rate": 3.006558508802209e-05,
+      "loss": 0.3144,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3774464119291705,
+      "grad_norm": 0.31971717303380853,
+      "learning_rate": 3.004832585433207e-05,
+      "loss": 0.346,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 0.29041646247959213,
+      "learning_rate": 3.0031066620642045e-05,
+      "loss": 0.3343,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 0.2667442369547004,
+      "learning_rate": 3.001380738695202e-05,
+      "loss": 0.3398,
+      "step": 1480
+    },
+    {
+      "epoch": 1.380242311276794,
+      "grad_norm": 0.29544152552022207,
+      "learning_rate": 2.9996548153261994e-05,
+      "loss": 0.3451,
+      "step": 1481
+    },
+    {
+      "epoch": 1.3811742777260019,
+      "grad_norm": 0.28296923289831627,
+      "learning_rate": 2.9979288919571975e-05,
+      "loss": 0.3547,
+      "step": 1482
+    },
+    {
+      "epoch": 1.3821062441752097,
+      "grad_norm": 0.2956700879162791,
+      "learning_rate": 2.996202968588195e-05,
+      "loss": 0.3411,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3830382106244175,
+      "grad_norm": 0.31651371376087956,
+      "learning_rate": 2.9944770452191923e-05,
+      "loss": 0.3711,
+      "step": 1484
+    },
+    {
+      "epoch": 1.3839701770736252,
+      "grad_norm": 0.2693228668312376,
+      "learning_rate": 2.99275112185019e-05,
+      "loss": 0.3441,
+      "step": 1485
+    },
+    {
+      "epoch": 1.3849021435228333,
+      "grad_norm": 0.3156303382607317,
+      "learning_rate": 2.9910251984811875e-05,
+      "loss": 0.3637,
+      "step": 1486
+    },
+    {
+      "epoch": 1.385834109972041,
+      "grad_norm": 0.28857842963873437,
+      "learning_rate": 2.989299275112185e-05,
+      "loss": 0.3531,
+      "step": 1487
+    },
+    {
+      "epoch": 1.3867660764212488,
+      "grad_norm": 0.3347916934893328,
+      "learning_rate": 2.9875733517431824e-05,
+      "loss": 0.3515,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3876980428704566,
+      "grad_norm": 0.32971777239515093,
+      "learning_rate": 2.9858474283741805e-05,
+      "loss": 0.3565,
+      "step": 1489
+    },
+    {
+      "epoch": 1.3886300093196646,
+      "grad_norm": 0.3407799701575634,
+      "learning_rate": 2.984121505005178e-05,
+      "loss": 0.3567,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3895619757688724,
+      "grad_norm": 0.30522737222690566,
+      "learning_rate": 2.9823955816361753e-05,
+      "loss": 0.3662,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3904939422180802,
+      "grad_norm": 0.31900543068243703,
+      "learning_rate": 2.9806696582671728e-05,
+      "loss": 0.359,
+      "step": 1492
+    },
+    {
+      "epoch": 1.391425908667288,
+      "grad_norm": 0.35419037122407354,
+      "learning_rate": 2.978943734898171e-05,
+      "loss": 0.3645,
+      "step": 1493
+    },
+    {
+      "epoch": 1.3923578751164958,
+      "grad_norm": 0.256090213662248,
+      "learning_rate": 2.9772178115291683e-05,
+      "loss": 0.3603,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3932898415657036,
+      "grad_norm": 0.3497518829370436,
+      "learning_rate": 2.9754918881601657e-05,
+      "loss": 0.3733,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3942218080149114,
+      "grad_norm": 0.41150573247757205,
+      "learning_rate": 2.9737659647911635e-05,
+      "loss": 0.3638,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3951537744641194,
+      "grad_norm": 0.28800840924779303,
+      "learning_rate": 2.972040041422161e-05,
+      "loss": 0.369,
+      "step": 1497
+    },
+    {
+      "epoch": 1.3960857409133272,
+      "grad_norm": 0.3216767080707665,
+      "learning_rate": 2.9703141180531584e-05,
+      "loss": 0.343,
+      "step": 1498
+    },
+    {
+      "epoch": 1.397017707362535,
+      "grad_norm": 0.2855073744411169,
+      "learning_rate": 2.9685881946841558e-05,
+      "loss": 0.3482,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3979496738117427,
+      "grad_norm": 0.3245790737726752,
+      "learning_rate": 2.966862271315154e-05,
+      "loss": 0.3794,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3988816402609507,
+      "grad_norm": 0.35780968884372283,
+      "learning_rate": 2.9651363479461513e-05,
+      "loss": 0.3621,
+      "step": 1501
+    },
+    {
+      "epoch": 1.3998136067101585,
+      "grad_norm": 0.29630702184446756,
+      "learning_rate": 2.9634104245771488e-05,
+      "loss": 0.3322,
+      "step": 1502
+    },
+    {
+      "epoch": 1.4007455731593663,
+      "grad_norm": 0.9400472866184926,
+      "learning_rate": 2.9616845012081462e-05,
+      "loss": 0.3592,
+      "step": 1503
+    },
+    {
+      "epoch": 1.401677539608574,
+      "grad_norm": 0.33530952257340313,
+      "learning_rate": 2.9599585778391443e-05,
+      "loss": 0.3414,
+      "step": 1504
+    },
+    {
+      "epoch": 1.402609506057782,
+      "grad_norm": 0.33963511873772784,
+      "learning_rate": 2.9582326544701417e-05,
+      "loss": 0.3911,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4035414725069897,
+      "grad_norm": 0.2939784816954384,
+      "learning_rate": 2.956506731101139e-05,
+      "loss": 0.3544,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4044734389561975,
+      "grad_norm": 0.3395506318479654,
+      "learning_rate": 2.954780807732137e-05,
+      "loss": 0.3613,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 0.2970699351893388,
+      "learning_rate": 2.9530548843631343e-05,
+      "loss": 0.3331,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4063373718546133,
+      "grad_norm": 0.30395368195395434,
+      "learning_rate": 2.9513289609941318e-05,
+      "loss": 0.3591,
+      "step": 1509
+    },
+    {
+      "epoch": 1.407269338303821,
+      "grad_norm": 0.3135861288197787,
+      "learning_rate": 2.9496030376251292e-05,
+      "loss": 0.3401,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4082013047530288,
+      "grad_norm": 0.294058815787346,
+      "learning_rate": 2.9478771142561273e-05,
+      "loss": 0.355,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4091332712022366,
+      "grad_norm": 0.30662787512199496,
+      "learning_rate": 2.9461511908871247e-05,
+      "loss": 0.3343,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4100652376514446,
+      "grad_norm": 0.28982579523209423,
+      "learning_rate": 2.944425267518122e-05,
+      "loss": 0.3576,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4109972041006524,
+      "grad_norm": 0.30449749087111133,
+      "learning_rate": 2.9426993441491203e-05,
+      "loss": 0.3463,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4119291705498602,
+      "grad_norm": 0.2756221734404247,
+      "learning_rate": 2.9409734207801177e-05,
+      "loss": 0.3707,
+      "step": 1515
+    },
+    {
+      "epoch": 1.412861136999068,
+      "grad_norm": 0.30614390329451363,
+      "learning_rate": 2.939247497411115e-05,
+      "loss": 0.3424,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 0.31993370756923695,
+      "learning_rate": 2.9375215740421126e-05,
+      "loss": 0.3515,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4147250698974836,
+      "grad_norm": 0.3009142235571223,
+      "learning_rate": 2.9357956506731103e-05,
+      "loss": 0.3551,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4156570363466916,
+      "grad_norm": 0.2933423882492838,
+      "learning_rate": 2.9340697273041078e-05,
+      "loss": 0.3621,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4165890027958994,
+      "grad_norm": 0.3230188232256702,
+      "learning_rate": 2.9323438039351052e-05,
+      "loss": 0.3503,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4175209692451072,
+      "grad_norm": 0.31578383327524506,
+      "learning_rate": 2.9306178805661026e-05,
+      "loss": 0.3468,
+      "step": 1521
+    },
+    {
+      "epoch": 1.418452935694315,
+      "grad_norm": 0.29495257955797466,
+      "learning_rate": 2.9288919571971007e-05,
+      "loss": 0.3524,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4193849021435228,
+      "grad_norm": 0.3044939018162221,
+      "learning_rate": 2.927166033828098e-05,
+      "loss": 0.3372,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4203168685927308,
+      "grad_norm": 0.30075974051336285,
+      "learning_rate": 2.9254401104590956e-05,
+      "loss": 0.3466,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4212488350419386,
+      "grad_norm": 0.30746739833936354,
+      "learning_rate": 2.9237141870900937e-05,
+      "loss": 0.358,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4221808014911463,
+      "grad_norm": 0.3328479527262209,
+      "learning_rate": 2.921988263721091e-05,
+      "loss": 0.3447,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4231127679403541,
+      "grad_norm": 0.2853971740386399,
+      "learning_rate": 2.9202623403520885e-05,
+      "loss": 0.3436,
+      "step": 1527
+    },
+    {
+      "epoch": 1.424044734389562,
+      "grad_norm": 0.3887001744026008,
+      "learning_rate": 2.918536416983086e-05,
+      "loss": 0.3429,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4249767008387697,
+      "grad_norm": 0.3199300851117305,
+      "learning_rate": 2.9168104936140837e-05,
+      "loss": 0.3595,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4259086672879777,
+      "grad_norm": 0.3458482706309197,
+      "learning_rate": 2.9150845702450812e-05,
+      "loss": 0.3447,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4268406337371855,
+      "grad_norm": 0.3112199524925582,
+      "learning_rate": 2.9133586468760786e-05,
+      "loss": 0.3446,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4277726001863933,
+      "grad_norm": 0.2867419108229798,
+      "learning_rate": 2.911632723507076e-05,
+      "loss": 0.3496,
+      "step": 1532
+    },
+    {
+      "epoch": 1.428704566635601,
+      "grad_norm": 0.34108134040092086,
+      "learning_rate": 2.909906800138074e-05,
+      "loss": 0.3623,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4296365330848089,
+      "grad_norm": 0.3530717807294143,
+      "learning_rate": 2.9081808767690716e-05,
+      "loss": 0.3314,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4305684995340169,
+      "grad_norm": 0.3240748736241649,
+      "learning_rate": 2.906454953400069e-05,
+      "loss": 0.3726,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4315004659832247,
+      "grad_norm": 0.4050022615335281,
+      "learning_rate": 2.904729030031067e-05,
+      "loss": 0.3614,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.32062437003993355,
+      "learning_rate": 2.9030031066620645e-05,
+      "loss": 0.3516,
+      "step": 1537
+    },
+    {
+      "epoch": 1.4333643988816402,
+      "grad_norm": 0.3751994972724663,
+      "learning_rate": 2.901277183293062e-05,
+      "loss": 0.377,
+      "step": 1538
+    },
+    {
+      "epoch": 1.434296365330848,
+      "grad_norm": 0.3032408950855589,
+      "learning_rate": 2.8995512599240594e-05,
+      "loss": 0.3387,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4352283317800558,
+      "grad_norm": 0.29447034119073595,
+      "learning_rate": 2.897825336555057e-05,
+      "loss": 0.313,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4361602982292636,
+      "grad_norm": 0.36914153534960203,
+      "learning_rate": 2.8960994131860546e-05,
+      "loss": 0.3381,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4370922646784716,
+      "grad_norm": 0.3483652986208254,
+      "learning_rate": 2.894373489817052e-05,
+      "loss": 0.362,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4380242311276794,
+      "grad_norm": 0.3315851460502982,
+      "learning_rate": 2.8926475664480494e-05,
+      "loss": 0.3592,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4389561975768872,
+      "grad_norm": 0.34616071235175705,
+      "learning_rate": 2.8909216430790476e-05,
+      "loss": 0.3449,
+      "step": 1544
+    },
+    {
+      "epoch": 1.439888164026095,
+      "grad_norm": 0.41197198927686396,
+      "learning_rate": 2.889195719710045e-05,
+      "loss": 0.3391,
+      "step": 1545
+    },
+    {
+      "epoch": 1.440820130475303,
+      "grad_norm": 0.36502108557169893,
+      "learning_rate": 2.8874697963410424e-05,
+      "loss": 0.341,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4417520969245108,
+      "grad_norm": 0.36055031993429976,
+      "learning_rate": 2.8857438729720405e-05,
+      "loss": 0.346,
+      "step": 1547
+    },
+    {
+      "epoch": 1.4426840633737186,
+      "grad_norm": 0.31528118401396016,
+      "learning_rate": 2.884017949603038e-05,
+      "loss": 0.3442,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4436160298229264,
+      "grad_norm": 0.34678272981597213,
+      "learning_rate": 2.8822920262340354e-05,
+      "loss": 0.344,
+      "step": 1549
+    },
+    {
+      "epoch": 1.4445479962721341,
+      "grad_norm": 0.3225410611642017,
+      "learning_rate": 2.8805661028650328e-05,
+      "loss": 0.3491,
+      "step": 1550
+    },
+    {
+      "epoch": 1.445479962721342,
+      "grad_norm": 0.30315929689019605,
+      "learning_rate": 2.8788401794960306e-05,
+      "loss": 0.3389,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4464119291705497,
+      "grad_norm": 0.3867888247459283,
+      "learning_rate": 2.877114256127028e-05,
+      "loss": 0.3427,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4473438956197577,
+      "grad_norm": 0.28558109810720805,
+      "learning_rate": 2.8753883327580254e-05,
+      "loss": 0.3342,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 0.3841454107551035,
+      "learning_rate": 2.8736624093890235e-05,
+      "loss": 0.3392,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4492078285181733,
+      "grad_norm": 0.43247609306644097,
+      "learning_rate": 2.871936486020021e-05,
+      "loss": 0.3571,
+      "step": 1555
+    },
+    {
+      "epoch": 1.450139794967381,
+      "grad_norm": 0.4033328347684474,
+      "learning_rate": 2.8702105626510184e-05,
+      "loss": 0.346,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4510717614165891,
+      "grad_norm": 0.3848138405686901,
+      "learning_rate": 2.8684846392820158e-05,
+      "loss": 0.3399,
+      "step": 1557
+    },
+    {
+      "epoch": 1.452003727865797,
+      "grad_norm": 0.40377766457263203,
+      "learning_rate": 2.866758715913014e-05,
+      "loss": 0.3372,
+      "step": 1558
+    },
+    {
+      "epoch": 1.4529356943150047,
+      "grad_norm": 0.31557888645697824,
+      "learning_rate": 2.8650327925440114e-05,
+      "loss": 0.3408,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4538676607642125,
+      "grad_norm": 0.35685649206070336,
+      "learning_rate": 2.8633068691750088e-05,
+      "loss": 0.3608,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4547996272134203,
+      "grad_norm": 0.40728262686890676,
+      "learning_rate": 2.8615809458060062e-05,
+      "loss": 0.3414,
+      "step": 1561
+    },
+    {
+      "epoch": 1.455731593662628,
+      "grad_norm": 0.30147966324923536,
+      "learning_rate": 2.859855022437004e-05,
+      "loss": 0.3561,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4566635601118358,
+      "grad_norm": 0.3950219699198105,
+      "learning_rate": 2.8581290990680014e-05,
+      "loss": 0.3461,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4575955265610439,
+      "grad_norm": 0.3265241490482744,
+      "learning_rate": 2.856403175698999e-05,
+      "loss": 0.3537,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4585274930102516,
+      "grad_norm": 0.3927825561349364,
+      "learning_rate": 2.854677252329997e-05,
+      "loss": 0.3395,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4594594594594594,
+      "grad_norm": 0.35211722031783743,
+      "learning_rate": 2.8529513289609944e-05,
+      "loss": 0.3882,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4603914259086672,
+      "grad_norm": 0.4435421811717066,
+      "learning_rate": 2.8512254055919918e-05,
+      "loss": 0.3355,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4613233923578752,
+      "grad_norm": 0.400249761706499,
+      "learning_rate": 2.8494994822229892e-05,
+      "loss": 0.3592,
+      "step": 1568
+    },
+    {
+      "epoch": 1.462255358807083,
+      "grad_norm": 0.4070670181469492,
+      "learning_rate": 2.8477735588539873e-05,
+      "loss": 0.3555,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4631873252562908,
+      "grad_norm": 0.3640565539148266,
+      "learning_rate": 2.8460476354849848e-05,
+      "loss": 0.3342,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4641192917054986,
+      "grad_norm": 0.3873538930220281,
+      "learning_rate": 2.8443217121159822e-05,
+      "loss": 0.3598,
+      "step": 1571
+    },
+    {
+      "epoch": 1.4650512581547064,
+      "grad_norm": 0.3845003589058128,
+      "learning_rate": 2.8425957887469796e-05,
+      "loss": 0.3467,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4659832246039142,
+      "grad_norm": 0.4194713438003305,
+      "learning_rate": 2.8408698653779774e-05,
+      "loss": 0.3523,
+      "step": 1573
+    },
+    {
+      "epoch": 1.466915191053122,
+      "grad_norm": 0.33255242129434226,
+      "learning_rate": 2.8391439420089748e-05,
+      "loss": 0.3583,
+      "step": 1574
+    },
+    {
+      "epoch": 1.46784715750233,
+      "grad_norm": 0.393850526661315,
+      "learning_rate": 2.8374180186399723e-05,
+      "loss": 0.3549,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4687791239515378,
+      "grad_norm": 0.37274922344295536,
+      "learning_rate": 2.8356920952709704e-05,
+      "loss": 0.3405,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4697110904007455,
+      "grad_norm": 0.35182983343871105,
+      "learning_rate": 2.8339661719019678e-05,
+      "loss": 0.3632,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4706430568499533,
+      "grad_norm": 0.42686254562580406,
+      "learning_rate": 2.8322402485329652e-05,
+      "loss": 0.3591,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4715750232991613,
+      "grad_norm": 0.40854409552965987,
+      "learning_rate": 2.8305143251639626e-05,
+      "loss": 0.3686,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4725069897483691,
+      "grad_norm": 0.3396858158613748,
+      "learning_rate": 2.8287884017949608e-05,
+      "loss": 0.3453,
+      "step": 1580
+    },
+    {
+      "epoch": 1.473438956197577,
+      "grad_norm": 0.38180431672292914,
+      "learning_rate": 2.8270624784259582e-05,
+      "loss": 0.3496,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4743709226467847,
+      "grad_norm": 0.35042263319947375,
+      "learning_rate": 2.8253365550569556e-05,
+      "loss": 0.3343,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4753028890959925,
+      "grad_norm": 0.31758170841867445,
+      "learning_rate": 2.823610631687953e-05,
+      "loss": 0.3435,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4762348555452003,
+      "grad_norm": 0.30527370975573337,
+      "learning_rate": 2.8218847083189508e-05,
+      "loss": 0.358,
+      "step": 1584
+    },
+    {
+      "epoch": 1.477166821994408,
+      "grad_norm": 0.345531914720718,
+      "learning_rate": 2.8201587849499482e-05,
+      "loss": 0.3573,
+      "step": 1585
+    },
+    {
+      "epoch": 1.478098788443616,
+      "grad_norm": 0.31097971809542885,
+      "learning_rate": 2.8184328615809457e-05,
+      "loss": 0.3715,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4790307548928239,
+      "grad_norm": 0.30007804832852436,
+      "learning_rate": 2.8167069382119438e-05,
+      "loss": 0.3385,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4799627213420317,
+      "grad_norm": 0.2907035217427766,
+      "learning_rate": 2.8149810148429412e-05,
+      "loss": 0.3531,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4808946877912395,
+      "grad_norm": 0.3260040974917613,
+      "learning_rate": 2.8132550914739386e-05,
+      "loss": 0.3822,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4818266542404475,
+      "grad_norm": 0.3439328665602838,
+      "learning_rate": 2.811529168104936e-05,
+      "loss": 0.3471,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 0.3302175937296127,
+      "learning_rate": 2.809803244735934e-05,
+      "loss": 0.3693,
+      "step": 1591
+    },
+    {
+      "epoch": 1.483690587138863,
+      "grad_norm": 0.28054853605570906,
+      "learning_rate": 2.8080773213669316e-05,
+      "loss": 0.347,
+      "step": 1592
+    },
+    {
+      "epoch": 1.4846225535880708,
+      "grad_norm": 0.3249331879622146,
+      "learning_rate": 2.806351397997929e-05,
+      "loss": 0.3576,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4855545200372786,
+      "grad_norm": 0.2958774488500036,
+      "learning_rate": 2.8046254746289268e-05,
+      "loss": 0.3294,
+      "step": 1594
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 0.33348955123927465,
+      "learning_rate": 2.8028995512599242e-05,
+      "loss": 0.3452,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4874184529356942,
+      "grad_norm": 0.3090125387651928,
+      "learning_rate": 2.8011736278909217e-05,
+      "loss": 0.3488,
+      "step": 1596
+    },
+    {
+      "epoch": 1.4883504193849022,
+      "grad_norm": 0.3050608674271422,
+      "learning_rate": 2.799447704521919e-05,
+      "loss": 0.3436,
+      "step": 1597
+    },
+    {
+      "epoch": 1.48928238583411,
+      "grad_norm": 0.3333574644851646,
+      "learning_rate": 2.7977217811529172e-05,
+      "loss": 0.3524,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4902143522833178,
+      "grad_norm": 0.2969599234602376,
+      "learning_rate": 2.7959958577839146e-05,
+      "loss": 0.3574,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4911463187325256,
+      "grad_norm": 0.39065237702960826,
+      "learning_rate": 2.794269934414912e-05,
+      "loss": 0.3527,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4920782851817336,
+      "grad_norm": 0.3542809517208244,
+      "learning_rate": 2.7925440110459095e-05,
+      "loss": 0.3644,
+      "step": 1601
+    },
+    {
+      "epoch": 1.4930102516309414,
+      "grad_norm": 0.2987033952648609,
+      "learning_rate": 2.7908180876769076e-05,
+      "loss": 0.3582,
+      "step": 1602
+    },
+    {
+      "epoch": 1.4939422180801492,
+      "grad_norm": 0.39503480552486075,
+      "learning_rate": 2.789092164307905e-05,
+      "loss": 0.3373,
+      "step": 1603
+    },
+    {
+      "epoch": 1.494874184529357,
+      "grad_norm": 0.3219731449191371,
+      "learning_rate": 2.7873662409389024e-05,
+      "loss": 0.3692,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4958061509785647,
+      "grad_norm": 0.31883173012596916,
+      "learning_rate": 2.7856403175699002e-05,
+      "loss": 0.3478,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4967381174277725,
+      "grad_norm": 0.33604845984232357,
+      "learning_rate": 2.7839143942008976e-05,
+      "loss": 0.3367,
+      "step": 1606
+    },
+    {
+      "epoch": 1.4976700838769803,
+      "grad_norm": 0.36958082064865133,
+      "learning_rate": 2.782188470831895e-05,
+      "loss": 0.3597,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4986020503261883,
+      "grad_norm": 0.296547528386871,
+      "learning_rate": 2.7804625474628925e-05,
+      "loss": 0.3665,
+      "step": 1608
+    },
+    {
+      "epoch": 1.499534016775396,
+      "grad_norm": 0.36481076725381306,
+      "learning_rate": 2.7787366240938906e-05,
+      "loss": 0.3326,
+      "step": 1609
+    },
+    {
+      "epoch": 1.500465983224604,
+      "grad_norm": 0.3521030560699963,
+      "learning_rate": 2.777010700724888e-05,
+      "loss": 0.3565,
+      "step": 1610
+    },
+    {
+      "epoch": 1.501397949673812,
+      "grad_norm": 0.30421454062750414,
+      "learning_rate": 2.7752847773558855e-05,
+      "loss": 0.3286,
+      "step": 1611
+    },
+    {
+      "epoch": 1.5023299161230197,
+      "grad_norm": 0.3778480603352084,
+      "learning_rate": 2.773558853986883e-05,
+      "loss": 0.3598,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5032618825722275,
+      "grad_norm": 0.3471853163575119,
+      "learning_rate": 2.771832930617881e-05,
+      "loss": 0.3454,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5041938490214353,
+      "grad_norm": 0.30418831711843275,
+      "learning_rate": 2.7701070072488784e-05,
+      "loss": 0.3558,
+      "step": 1614
+    },
+    {
+      "epoch": 1.505125815470643,
+      "grad_norm": 0.3489834910562244,
+      "learning_rate": 2.768381083879876e-05,
+      "loss": 0.3377,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5060577819198508,
+      "grad_norm": 0.5666145112213601,
+      "learning_rate": 2.7666551605108736e-05,
+      "loss": 0.3495,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5069897483690586,
+      "grad_norm": 0.32754948085875496,
+      "learning_rate": 2.764929237141871e-05,
+      "loss": 0.3314,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5079217148182664,
+      "grad_norm": 0.3480575974243788,
+      "learning_rate": 2.7632033137728685e-05,
+      "loss": 0.3699,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5088536812674742,
+      "grad_norm": 0.35192961819658874,
+      "learning_rate": 2.761477390403866e-05,
+      "loss": 0.3441,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5097856477166822,
+      "grad_norm": 0.28871785694555396,
+      "learning_rate": 2.759751467034864e-05,
+      "loss": 0.3444,
+      "step": 1620
+    },
+    {
+      "epoch": 1.51071761416589,
+      "grad_norm": 0.330974576825696,
+      "learning_rate": 2.7580255436658614e-05,
+      "loss": 0.3361,
+      "step": 1621
+    },
+    {
+      "epoch": 1.511649580615098,
+      "grad_norm": 0.30722523420073333,
+      "learning_rate": 2.756299620296859e-05,
+      "loss": 0.3417,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5125815470643058,
+      "grad_norm": 0.32099156779892507,
+      "learning_rate": 2.7545736969278563e-05,
+      "loss": 0.3374,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 0.31968360284177644,
+      "learning_rate": 2.7528477735588544e-05,
+      "loss": 0.3579,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5144454799627214,
+      "grad_norm": 0.3006780381483513,
+      "learning_rate": 2.751121850189852e-05,
+      "loss": 0.3368,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5153774464119292,
+      "grad_norm": 0.27669376188331823,
+      "learning_rate": 2.7493959268208493e-05,
+      "loss": 0.3259,
+      "step": 1626
+    },
+    {
+      "epoch": 1.516309412861137,
+      "grad_norm": 0.30340696135312295,
+      "learning_rate": 2.747670003451847e-05,
+      "loss": 0.3304,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5172413793103448,
+      "grad_norm": 0.332916587682244,
+      "learning_rate": 2.7459440800828445e-05,
+      "loss": 0.3286,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5181733457595525,
+      "grad_norm": 0.25301786300902723,
+      "learning_rate": 2.744218156713842e-05,
+      "loss": 0.3347,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5191053122087603,
+      "grad_norm": 0.309788227004074,
+      "learning_rate": 2.7424922333448393e-05,
+      "loss": 0.3581,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5200372786579683,
+      "grad_norm": 0.30418102638712285,
+      "learning_rate": 2.7407663099758374e-05,
+      "loss": 0.3416,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5209692451071761,
+      "grad_norm": 0.28732045042832943,
+      "learning_rate": 2.739040386606835e-05,
+      "loss": 0.3314,
+      "step": 1632
+    },
+    {
+      "epoch": 1.521901211556384,
+      "grad_norm": 0.3738736206919394,
+      "learning_rate": 2.7373144632378323e-05,
+      "loss": 0.3679,
+      "step": 1633
+    },
+    {
+      "epoch": 1.522833178005592,
+      "grad_norm": 0.30494203569136163,
+      "learning_rate": 2.7355885398688297e-05,
+      "loss": 0.3457,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5237651444547997,
+      "grad_norm": 0.3224277283346315,
+      "learning_rate": 2.7338626164998278e-05,
+      "loss": 0.3275,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5246971109040075,
+      "grad_norm": 0.29912521533949205,
+      "learning_rate": 2.7321366931308253e-05,
+      "loss": 0.3468,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5256290773532153,
+      "grad_norm": 0.2559571701764213,
+      "learning_rate": 2.7304107697618227e-05,
+      "loss": 0.3302,
+      "step": 1637
+    },
+    {
+      "epoch": 1.526561043802423,
+      "grad_norm": 0.29458103146046855,
+      "learning_rate": 2.7286848463928204e-05,
+      "loss": 0.3403,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5274930102516309,
+      "grad_norm": 0.2918457215523616,
+      "learning_rate": 2.726958923023818e-05,
+      "loss": 0.3587,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5284249767008387,
+      "grad_norm": 0.2997158588890921,
+      "learning_rate": 2.7252329996548153e-05,
+      "loss": 0.3524,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5293569431500464,
+      "grad_norm": 0.3275325761095528,
+      "learning_rate": 2.7235070762858127e-05,
+      "loss": 0.3654,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5302889095992545,
+      "grad_norm": 0.3172953439948189,
+      "learning_rate": 2.721781152916811e-05,
+      "loss": 0.3567,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5312208760484622,
+      "grad_norm": 0.3577134740935238,
+      "learning_rate": 2.7200552295478083e-05,
+      "loss": 0.3794,
+      "step": 1643
+    },
+    {
+      "epoch": 1.53215284249767,
+      "grad_norm": 0.3178232308574679,
+      "learning_rate": 2.7183293061788057e-05,
+      "loss": 0.3584,
+      "step": 1644
+    },
+    {
+      "epoch": 1.533084808946878,
+      "grad_norm": 0.3069065748416724,
+      "learning_rate": 2.7166033828098038e-05,
+      "loss": 0.3447,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5340167753960858,
+      "grad_norm": 0.2905925863884371,
+      "learning_rate": 2.7148774594408012e-05,
+      "loss": 0.3659,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5349487418452936,
+      "grad_norm": 0.2944786661491946,
+      "learning_rate": 2.7131515360717987e-05,
+      "loss": 0.3403,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5358807082945014,
+      "grad_norm": 0.2820165868642374,
+      "learning_rate": 2.711425612702796e-05,
+      "loss": 0.352,
+      "step": 1648
+    },
+    {
+      "epoch": 1.5368126747437092,
+      "grad_norm": 0.32691935678773243,
+      "learning_rate": 2.709699689333794e-05,
+      "loss": 0.3454,
+      "step": 1649
+    },
+    {
+      "epoch": 1.537744641192917,
+      "grad_norm": 0.32507292407748006,
+      "learning_rate": 2.7079737659647913e-05,
+      "loss": 0.3459,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5386766076421248,
+      "grad_norm": 0.33689768140206644,
+      "learning_rate": 2.7062478425957887e-05,
+      "loss": 0.3665,
+      "step": 1651
+    },
+    {
+      "epoch": 1.5396085740913326,
+      "grad_norm": 0.3065322268061091,
+      "learning_rate": 2.704521919226786e-05,
+      "loss": 0.3632,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 0.35109615718540255,
+      "learning_rate": 2.7027959958577843e-05,
+      "loss": 0.3668,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5414725069897484,
+      "grad_norm": 0.3094688417182038,
+      "learning_rate": 2.7010700724887817e-05,
+      "loss": 0.3326,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5424044734389561,
+      "grad_norm": 0.3648474514573363,
+      "learning_rate": 2.699344149119779e-05,
+      "loss": 0.3549,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5433364398881642,
+      "grad_norm": 0.3669503974919837,
+      "learning_rate": 2.6976182257507772e-05,
+      "loss": 0.3452,
+      "step": 1656
+    },
+    {
+      "epoch": 1.544268406337372,
+      "grad_norm": 0.2752177444445793,
+      "learning_rate": 2.6958923023817746e-05,
+      "loss": 0.3457,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5452003727865797,
+      "grad_norm": 0.457790984212332,
+      "learning_rate": 2.694166379012772e-05,
+      "loss": 0.3506,
+      "step": 1658
+    },
+    {
+      "epoch": 1.5461323392357875,
+      "grad_norm": 0.3273064700917727,
+      "learning_rate": 2.6924404556437695e-05,
+      "loss": 0.3501,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5470643056849953,
+      "grad_norm": 0.4116808525640341,
+      "learning_rate": 2.6907145322747673e-05,
+      "loss": 0.3645,
+      "step": 1660
+    },
+    {
+      "epoch": 1.547996272134203,
+      "grad_norm": 0.39726511353130156,
+      "learning_rate": 2.6889886089057647e-05,
+      "loss": 0.3521,
+      "step": 1661
+    },
+    {
+      "epoch": 1.5489282385834109,
+      "grad_norm": 0.38219288456791045,
+      "learning_rate": 2.687262685536762e-05,
+      "loss": 0.3482,
+      "step": 1662
+    },
+    {
+      "epoch": 1.5498602050326187,
+      "grad_norm": 0.31982149134455584,
+      "learning_rate": 2.6855367621677596e-05,
+      "loss": 0.3282,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5507921714818267,
+      "grad_norm": 0.2892799278787182,
+      "learning_rate": 2.6838108387987577e-05,
+      "loss": 0.35,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 0.35092408880124437,
+      "learning_rate": 2.682084915429755e-05,
+      "loss": 0.3316,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5526561043802423,
+      "grad_norm": 0.2822144727611181,
+      "learning_rate": 2.6803589920607525e-05,
+      "loss": 0.3502,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5535880708294503,
+      "grad_norm": 0.36732802553217364,
+      "learning_rate": 2.6786330686917503e-05,
+      "loss": 0.3623,
+      "step": 1667
+    },
+    {
+      "epoch": 1.554520037278658,
+      "grad_norm": 0.3058849222641261,
+      "learning_rate": 2.6769071453227477e-05,
+      "loss": 0.3424,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5554520037278659,
+      "grad_norm": 0.3175330529270093,
+      "learning_rate": 2.675181221953745e-05,
+      "loss": 0.339,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5563839701770736,
+      "grad_norm": 0.3362358437324181,
+      "learning_rate": 2.673455298584743e-05,
+      "loss": 0.3406,
+      "step": 1670
+    },
+    {
+      "epoch": 1.5573159366262814,
+      "grad_norm": 0.26635224943076047,
+      "learning_rate": 2.6717293752157407e-05,
+      "loss": 0.3295,
+      "step": 1671
+    },
+    {
+      "epoch": 1.5582479030754892,
+      "grad_norm": 0.32789077595941024,
+      "learning_rate": 2.670003451846738e-05,
+      "loss": 0.3696,
+      "step": 1672
+    },
+    {
+      "epoch": 1.559179869524697,
+      "grad_norm": 0.3246033955902729,
+      "learning_rate": 2.6682775284777355e-05,
+      "loss": 0.3611,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5601118359739048,
+      "grad_norm": 0.2605772523135968,
+      "learning_rate": 2.666551605108733e-05,
+      "loss": 0.3429,
+      "step": 1674
+    },
+    {
+      "epoch": 1.5610438024231128,
+      "grad_norm": 0.3283991489500725,
+      "learning_rate": 2.664825681739731e-05,
+      "loss": 0.3701,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5619757688723206,
+      "grad_norm": 0.3008987950170343,
+      "learning_rate": 2.6630997583707285e-05,
+      "loss": 0.3447,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5629077353215284,
+      "grad_norm": 0.3391772780161753,
+      "learning_rate": 2.661373835001726e-05,
+      "loss": 0.3393,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5638397017707364,
+      "grad_norm": 0.27279647165722276,
+      "learning_rate": 2.6596479116327237e-05,
+      "loss": 0.3391,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5647716682199442,
+      "grad_norm": 0.3028293165968809,
+      "learning_rate": 2.657921988263721e-05,
+      "loss": 0.3359,
+      "step": 1679
+    },
+    {
+      "epoch": 1.565703634669152,
+      "grad_norm": 0.3038410397964577,
+      "learning_rate": 2.6561960648947186e-05,
+      "loss": 0.3716,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5666356011183598,
+      "grad_norm": 0.2806849011263802,
+      "learning_rate": 2.654470141525716e-05,
+      "loss": 0.3427,
+      "step": 1681
+    },
+    {
+      "epoch": 1.5675675675675675,
+      "grad_norm": 0.2712782795074391,
+      "learning_rate": 2.652744218156714e-05,
+      "loss": 0.3287,
+      "step": 1682
+    },
+    {
+      "epoch": 1.5684995340167753,
+      "grad_norm": 0.2775620083132382,
+      "learning_rate": 2.6510182947877115e-05,
+      "loss": 0.3383,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5694315004659831,
+      "grad_norm": 0.3130476497146551,
+      "learning_rate": 2.649292371418709e-05,
+      "loss": 0.356,
+      "step": 1684
+    },
+    {
+      "epoch": 1.570363466915191,
+      "grad_norm": 0.2863431627902199,
+      "learning_rate": 2.647566448049707e-05,
+      "loss": 0.3454,
+      "step": 1685
+    },
+    {
+      "epoch": 1.571295433364399,
+      "grad_norm": 0.2827974069269227,
+      "learning_rate": 2.6458405246807045e-05,
+      "loss": 0.3493,
+      "step": 1686
+    },
+    {
+      "epoch": 1.5722273998136067,
+      "grad_norm": 0.3053711923445184,
+      "learning_rate": 2.644114601311702e-05,
+      "loss": 0.3383,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5731593662628145,
+      "grad_norm": 0.2910297947103003,
+      "learning_rate": 2.6423886779426994e-05,
+      "loss": 0.3653,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5740913327120225,
+      "grad_norm": 0.29820701896785556,
+      "learning_rate": 2.640662754573697e-05,
+      "loss": 0.357,
+      "step": 1689
+    },
+    {
+      "epoch": 1.5750232991612303,
+      "grad_norm": 0.285636128965277,
+      "learning_rate": 2.6389368312046945e-05,
+      "loss": 0.3622,
+      "step": 1690
+    },
+    {
+      "epoch": 1.575955265610438,
+      "grad_norm": 0.28018037103869337,
+      "learning_rate": 2.637210907835692e-05,
+      "loss": 0.3633,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5768872320596459,
+      "grad_norm": 0.2830266714942729,
+      "learning_rate": 2.6354849844666894e-05,
+      "loss": 0.3569,
+      "step": 1692
+    },
+    {
+      "epoch": 1.5778191985088537,
+      "grad_norm": 0.3759238392427578,
+      "learning_rate": 2.6337590610976875e-05,
+      "loss": 0.3825,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5787511649580614,
+      "grad_norm": 0.26153575932568024,
+      "learning_rate": 2.632033137728685e-05,
+      "loss": 0.3344,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5796831314072692,
+      "grad_norm": 0.29923044385515174,
+      "learning_rate": 2.6303072143596824e-05,
+      "loss": 0.3551,
+      "step": 1695
+    },
+    {
+      "epoch": 1.580615097856477,
+      "grad_norm": 0.3053278497151099,
+      "learning_rate": 2.6285812909906805e-05,
+      "loss": 0.3424,
+      "step": 1696
+    },
+    {
+      "epoch": 1.581547064305685,
+      "grad_norm": 0.3070390974558346,
+      "learning_rate": 2.626855367621678e-05,
+      "loss": 0.3513,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5824790307548928,
+      "grad_norm": 0.318662754562471,
+      "learning_rate": 2.6251294442526753e-05,
+      "loss": 0.3547,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5834109972041006,
+      "grad_norm": 0.27071812882892665,
+      "learning_rate": 2.6234035208836728e-05,
+      "loss": 0.3348,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5843429636533086,
+      "grad_norm": 0.32381715261710436,
+      "learning_rate": 2.6216775975146705e-05,
+      "loss": 0.3711,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5852749301025164,
+      "grad_norm": 0.3030285092100046,
+      "learning_rate": 2.619951674145668e-05,
+      "loss": 0.3383,
+      "step": 1701
+    },
+    {
+      "epoch": 1.5862068965517242,
+      "grad_norm": 0.28740640377361865,
+      "learning_rate": 2.6182257507766654e-05,
+      "loss": 0.348,
+      "step": 1702
+    },
+    {
+      "epoch": 1.587138863000932,
+      "grad_norm": 0.3393021039333704,
+      "learning_rate": 2.6164998274076628e-05,
+      "loss": 0.3367,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5880708294501398,
+      "grad_norm": 0.3449458939752181,
+      "learning_rate": 2.614773904038661e-05,
+      "loss": 0.3704,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5890027958993476,
+      "grad_norm": 0.3000852418017459,
+      "learning_rate": 2.6130479806696584e-05,
+      "loss": 0.348,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5899347623485554,
+      "grad_norm": 0.2719011581626616,
+      "learning_rate": 2.6113220573006558e-05,
+      "loss": 0.3431,
+      "step": 1706
+    },
+    {
+      "epoch": 1.5908667287977631,
+      "grad_norm": 0.2917963931066367,
+      "learning_rate": 2.609596133931654e-05,
+      "loss": 0.3425,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5917986952469712,
+      "grad_norm": 0.3069145007751472,
+      "learning_rate": 2.6078702105626513e-05,
+      "loss": 0.3552,
+      "step": 1708
+    },
+    {
+      "epoch": 1.592730661696179,
+      "grad_norm": 0.3008160800593478,
+      "learning_rate": 2.6061442871936487e-05,
+      "loss": 0.3567,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5936626281453867,
+      "grad_norm": 0.36190068707690204,
+      "learning_rate": 2.6044183638246462e-05,
+      "loss": 0.3712,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 0.5524979636098613,
+      "learning_rate": 2.602692440455644e-05,
+      "loss": 0.3396,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5955265610438025,
+      "grad_norm": 0.3468660639005747,
+      "learning_rate": 2.6009665170866414e-05,
+      "loss": 0.3524,
+      "step": 1712
+    },
+    {
+      "epoch": 1.5964585274930103,
+      "grad_norm": 0.3048821200425553,
+      "learning_rate": 2.5992405937176388e-05,
+      "loss": 0.353,
+      "step": 1713
+    },
+    {
+      "epoch": 1.597390493942218,
+      "grad_norm": 0.2957510967419196,
+      "learning_rate": 2.5975146703486362e-05,
+      "loss": 0.3356,
+      "step": 1714
+    },
+    {
+      "epoch": 1.598322460391426,
+      "grad_norm": 0.2383605954177748,
+      "learning_rate": 2.5957887469796343e-05,
+      "loss": 0.3365,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5992544268406337,
+      "grad_norm": 0.30962704169515237,
+      "learning_rate": 2.5940628236106318e-05,
+      "loss": 0.3491,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6001863932898415,
+      "grad_norm": 0.26655826830219964,
+      "learning_rate": 2.5923369002416292e-05,
+      "loss": 0.3488,
+      "step": 1717
+    },
+    {
+      "epoch": 1.6011183597390493,
+      "grad_norm": 0.2806290112912225,
+      "learning_rate": 2.5906109768726273e-05,
+      "loss": 0.3397,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6020503261882573,
+      "grad_norm": 0.29427317314680873,
+      "learning_rate": 2.5888850535036247e-05,
+      "loss": 0.3452,
+      "step": 1719
+    },
+    {
+      "epoch": 1.602982292637465,
+      "grad_norm": 0.2880585699224808,
+      "learning_rate": 2.587159130134622e-05,
+      "loss": 0.3566,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6039142590866728,
+      "grad_norm": 0.2728536789169514,
+      "learning_rate": 2.5854332067656196e-05,
+      "loss": 0.3685,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6048462255358809,
+      "grad_norm": 0.26165390317096043,
+      "learning_rate": 2.5837072833966174e-05,
+      "loss": 0.3438,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6057781919850886,
+      "grad_norm": 0.30496382626798657,
+      "learning_rate": 2.5819813600276148e-05,
+      "loss": 0.3453,
+      "step": 1723
+    },
+    {
+      "epoch": 1.6067101584342964,
+      "grad_norm": 0.2776674849785457,
+      "learning_rate": 2.5802554366586122e-05,
+      "loss": 0.3592,
+      "step": 1724
+    },
+    {
+      "epoch": 1.6076421248835042,
+      "grad_norm": 0.27610422223488107,
+      "learning_rate": 2.5785295132896096e-05,
+      "loss": 0.3217,
+      "step": 1725
+    },
+    {
+      "epoch": 1.608574091332712,
+      "grad_norm": 0.28934440735367145,
+      "learning_rate": 2.5768035899206078e-05,
+      "loss": 0.3332,
+      "step": 1726
+    },
+    {
+      "epoch": 1.6095060577819198,
+      "grad_norm": 0.30061325953188384,
+      "learning_rate": 2.5750776665516052e-05,
+      "loss": 0.3419,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6104380242311276,
+      "grad_norm": 0.3597815881433534,
+      "learning_rate": 2.5733517431826026e-05,
+      "loss": 0.3499,
+      "step": 1728
+    },
+    {
+      "epoch": 1.6113699906803354,
+      "grad_norm": 0.27918609096892344,
+      "learning_rate": 2.5716258198136007e-05,
+      "loss": 0.3494,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6123019571295434,
+      "grad_norm": 0.3373865546808137,
+      "learning_rate": 2.569899896444598e-05,
+      "loss": 0.3345,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6132339235787512,
+      "grad_norm": 2.593049633690923,
+      "learning_rate": 2.5681739730755956e-05,
+      "loss": 0.3819,
+      "step": 1731
+    },
+    {
+      "epoch": 1.614165890027959,
+      "grad_norm": 0.30972199974578213,
+      "learning_rate": 2.566448049706593e-05,
+      "loss": 0.3774,
+      "step": 1732
+    },
+    {
+      "epoch": 1.615097856477167,
+      "grad_norm": 0.28702096429017454,
+      "learning_rate": 2.5647221263375908e-05,
+      "loss": 0.3366,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6160298229263748,
+      "grad_norm": 0.27376323758127824,
+      "learning_rate": 2.5629962029685882e-05,
+      "loss": 0.3342,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6169617893755825,
+      "grad_norm": 0.28800717700295897,
+      "learning_rate": 2.5612702795995856e-05,
+      "loss": 0.355,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6178937558247903,
+      "grad_norm": 0.25570703523245686,
+      "learning_rate": 2.5595443562305837e-05,
+      "loss": 0.3349,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6188257222739981,
+      "grad_norm": 0.3048177062545011,
+      "learning_rate": 2.557818432861581e-05,
+      "loss": 0.3476,
+      "step": 1737
+    },
+    {
+      "epoch": 1.619757688723206,
+      "grad_norm": 0.24930026352999396,
+      "learning_rate": 2.5560925094925786e-05,
+      "loss": 0.3419,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 0.294234449079981,
+      "learning_rate": 2.554366586123576e-05,
+      "loss": 0.3408,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 0.2778195515230197,
+      "learning_rate": 2.552640662754574e-05,
+      "loss": 0.3711,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6225535880708295,
+      "grad_norm": 0.27884429454643334,
+      "learning_rate": 2.5509147393855716e-05,
+      "loss": 0.355,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6234855545200373,
+      "grad_norm": 0.28739815269610747,
+      "learning_rate": 2.549188816016569e-05,
+      "loss": 0.3691,
+      "step": 1742
+    },
+    {
+      "epoch": 1.624417520969245,
+      "grad_norm": 0.24879436787744305,
+      "learning_rate": 2.5474628926475664e-05,
+      "loss": 0.3413,
+      "step": 1743
+    },
+    {
+      "epoch": 1.625349487418453,
+      "grad_norm": 0.4934801604053048,
+      "learning_rate": 2.5457369692785642e-05,
+      "loss": 0.3618,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6262814538676609,
+      "grad_norm": 0.2539905025888413,
+      "learning_rate": 2.5440110459095616e-05,
+      "loss": 0.3423,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6272134203168687,
+      "grad_norm": 0.3014719571472351,
+      "learning_rate": 2.542285122540559e-05,
+      "loss": 0.3598,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6281453867660765,
+      "grad_norm": 0.2538625771102206,
+      "learning_rate": 2.540559199171557e-05,
+      "loss": 0.3425,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6290773532152842,
+      "grad_norm": 0.31165858233152166,
+      "learning_rate": 2.5388332758025546e-05,
+      "loss": 0.3653,
+      "step": 1748
+    },
+    {
+      "epoch": 1.630009319664492,
+      "grad_norm": 0.24955393238015566,
+      "learning_rate": 2.537107352433552e-05,
+      "loss": 0.3348,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6309412861136998,
+      "grad_norm": 0.33486792055671166,
+      "learning_rate": 2.5353814290645494e-05,
+      "loss": 0.36,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6318732525629076,
+      "grad_norm": 0.30130780901618853,
+      "learning_rate": 2.5336555056955475e-05,
+      "loss": 0.3361,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6328052190121156,
+      "grad_norm": 0.34980343984037776,
+      "learning_rate": 2.531929582326545e-05,
+      "loss": 0.3487,
+      "step": 1752
+    },
+    {
+      "epoch": 1.6337371854613234,
+      "grad_norm": 0.30475946941005805,
+      "learning_rate": 2.5302036589575424e-05,
+      "loss": 0.3632,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6346691519105312,
+      "grad_norm": 0.2952855418157237,
+      "learning_rate": 2.5284777355885398e-05,
+      "loss": 0.3328,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6356011183597392,
+      "grad_norm": 0.28300671977924696,
+      "learning_rate": 2.5267518122195376e-05,
+      "loss": 0.3466,
+      "step": 1755
+    },
+    {
+      "epoch": 1.636533084808947,
+      "grad_norm": 0.27924781910711854,
+      "learning_rate": 2.525025888850535e-05,
+      "loss": 0.3568,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6374650512581548,
+      "grad_norm": 0.305446658213558,
+      "learning_rate": 2.5232999654815325e-05,
+      "loss": 0.3538,
+      "step": 1757
+    },
+    {
+      "epoch": 1.6383970177073626,
+      "grad_norm": 0.2959784892107677,
+      "learning_rate": 2.5215740421125306e-05,
+      "loss": 0.3449,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6393289841565704,
+      "grad_norm": 0.26819461370311415,
+      "learning_rate": 2.519848118743528e-05,
+      "loss": 0.3457,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6402609506057781,
+      "grad_norm": 0.36310953715573435,
+      "learning_rate": 2.5181221953745254e-05,
+      "loss": 0.3626,
+      "step": 1760
+    },
+    {
+      "epoch": 1.641192917054986,
+      "grad_norm": 0.3492440498947586,
+      "learning_rate": 2.516396272005523e-05,
+      "loss": 0.366,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6421248835041937,
+      "grad_norm": 0.3926139471466557,
+      "learning_rate": 2.514670348636521e-05,
+      "loss": 0.3378,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6430568499534017,
+      "grad_norm": 0.34122795133918643,
+      "learning_rate": 2.5129444252675184e-05,
+      "loss": 0.3585,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6439888164026095,
+      "grad_norm": 0.3305271864383135,
+      "learning_rate": 2.5112185018985158e-05,
+      "loss": 0.3638,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6449207828518173,
+      "grad_norm": 0.3002972806164204,
+      "learning_rate": 2.5094925785295132e-05,
+      "loss": 0.3471,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6458527493010253,
+      "grad_norm": 0.3088363068108621,
+      "learning_rate": 2.507766655160511e-05,
+      "loss": 0.3429,
+      "step": 1766
+    },
+    {
+      "epoch": 1.646784715750233,
+      "grad_norm": 0.26280308726154117,
+      "learning_rate": 2.5060407317915084e-05,
+      "loss": 0.3452,
+      "step": 1767
+    },
+    {
+      "epoch": 1.647716682199441,
+      "grad_norm": 0.2881365527175674,
+      "learning_rate": 2.504314808422506e-05,
+      "loss": 0.3315,
+      "step": 1768
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 0.33056603702009885,
+      "learning_rate": 2.502588885053504e-05,
+      "loss": 0.3535,
+      "step": 1769
+    },
+    {
+      "epoch": 1.6495806150978565,
+      "grad_norm": 0.29146437563666006,
+      "learning_rate": 2.5008629616845014e-05,
+      "loss": 0.3654,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6505125815470643,
+      "grad_norm": 0.3097615565275971,
+      "learning_rate": 2.499137038315499e-05,
+      "loss": 0.3319,
+      "step": 1771
+    },
+    {
+      "epoch": 1.651444547996272,
+      "grad_norm": 0.33064193437519523,
+      "learning_rate": 2.4974111149464966e-05,
+      "loss": 0.3382,
+      "step": 1772
+    },
+    {
+      "epoch": 1.6523765144454798,
+      "grad_norm": 0.30952993338219714,
+      "learning_rate": 2.495685191577494e-05,
+      "loss": 0.3627,
+      "step": 1773
+    },
+    {
+      "epoch": 1.6533084808946876,
+      "grad_norm": 0.2981801321407654,
+      "learning_rate": 2.4939592682084918e-05,
+      "loss": 0.3321,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6542404473438956,
+      "grad_norm": 0.3158083041560753,
+      "learning_rate": 2.4922333448394892e-05,
+      "loss": 0.336,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 0.38405287724160436,
+      "learning_rate": 2.4905074214704867e-05,
+      "loss": 0.3503,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6561043802423114,
+      "grad_norm": 0.31014319282578134,
+      "learning_rate": 2.4887814981014844e-05,
+      "loss": 0.3559,
+      "step": 1777
+    },
+    {
+      "epoch": 1.6570363466915192,
+      "grad_norm": 0.34625372694741274,
+      "learning_rate": 2.487055574732482e-05,
+      "loss": 0.3537,
+      "step": 1778
+    },
+    {
+      "epoch": 1.657968313140727,
+      "grad_norm": 0.3496073610250237,
+      "learning_rate": 2.4853296513634796e-05,
+      "loss": 0.3613,
+      "step": 1779
+    },
+    {
+      "epoch": 1.6589002795899348,
+      "grad_norm": 0.31723701760201517,
+      "learning_rate": 2.483603727994477e-05,
+      "loss": 0.3287,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6598322460391426,
+      "grad_norm": 0.3204326809417617,
+      "learning_rate": 2.4818778046254748e-05,
+      "loss": 0.3592,
+      "step": 1781
+    },
+    {
+      "epoch": 1.6607642124883504,
+      "grad_norm": 0.33153622953181794,
+      "learning_rate": 2.4801518812564722e-05,
+      "loss": 0.3516,
+      "step": 1782
+    },
+    {
+      "epoch": 1.6616961789375582,
+      "grad_norm": 0.28282735631590733,
+      "learning_rate": 2.47842595788747e-05,
+      "loss": 0.3374,
+      "step": 1783
+    },
+    {
+      "epoch": 1.662628145386766,
+      "grad_norm": 0.2872724955402132,
+      "learning_rate": 2.4767000345184674e-05,
+      "loss": 0.3384,
+      "step": 1784
+    },
+    {
+      "epoch": 1.6635601118359737,
+      "grad_norm": 0.29773606442166467,
+      "learning_rate": 2.4749741111494652e-05,
+      "loss": 0.35,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6644920782851818,
+      "grad_norm": 0.2740036902717188,
+      "learning_rate": 2.4732481877804626e-05,
+      "loss": 0.3417,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6654240447343895,
+      "grad_norm": 0.3458913785040286,
+      "learning_rate": 2.47152226441146e-05,
+      "loss": 0.34,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6663560111835976,
+      "grad_norm": 0.29679447083272176,
+      "learning_rate": 2.469796341042458e-05,
+      "loss": 0.3499,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6672879776328053,
+      "grad_norm": 0.25678947872474556,
+      "learning_rate": 2.4680704176734553e-05,
+      "loss": 0.3561,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6682199440820131,
+      "grad_norm": 0.28264291766175265,
+      "learning_rate": 2.466344494304453e-05,
+      "loss": 0.3468,
+      "step": 1790
+    },
+    {
+      "epoch": 1.669151910531221,
+      "grad_norm": 0.29706152559369947,
+      "learning_rate": 2.4646185709354505e-05,
+      "loss": 0.3437,
+      "step": 1791
+    },
+    {
+      "epoch": 1.6700838769804287,
+      "grad_norm": 0.26073510271850925,
+      "learning_rate": 2.4628926475664482e-05,
+      "loss": 0.3439,
+      "step": 1792
+    },
+    {
+      "epoch": 1.6710158434296365,
+      "grad_norm": 0.2861796772669478,
+      "learning_rate": 2.4611667241974457e-05,
+      "loss": 0.3462,
+      "step": 1793
+    },
+    {
+      "epoch": 1.6719478098788443,
+      "grad_norm": 0.25385987424500156,
+      "learning_rate": 2.4594408008284434e-05,
+      "loss": 0.3198,
+      "step": 1794
+    },
+    {
+      "epoch": 1.672879776328052,
+      "grad_norm": 0.28238619451690206,
+      "learning_rate": 2.457714877459441e-05,
+      "loss": 0.3426,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6738117427772599,
+      "grad_norm": 0.2736380748933511,
+      "learning_rate": 2.4559889540904386e-05,
+      "loss": 0.351,
+      "step": 1796
+    },
+    {
+      "epoch": 1.6747437092264679,
+      "grad_norm": 0.26819836846890466,
+      "learning_rate": 2.454263030721436e-05,
+      "loss": 0.3236,
+      "step": 1797
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 0.3078246113355446,
+      "learning_rate": 2.4525371073524335e-05,
+      "loss": 0.3651,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6766076421248837,
+      "grad_norm": 0.2535975866953439,
+      "learning_rate": 2.4508111839834313e-05,
+      "loss": 0.3101,
+      "step": 1799
+    },
+    {
+      "epoch": 1.6775396085740915,
+      "grad_norm": 0.2937128184153439,
+      "learning_rate": 2.4490852606144287e-05,
+      "loss": 0.3469,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6784715750232992,
+      "grad_norm": 0.2875538759557856,
+      "learning_rate": 2.4473593372454264e-05,
+      "loss": 0.3611,
+      "step": 1801
+    },
+    {
+      "epoch": 1.679403541472507,
+      "grad_norm": 0.2831978742538602,
+      "learning_rate": 2.445633413876424e-05,
+      "loss": 0.3346,
+      "step": 1802
+    },
+    {
+      "epoch": 1.6803355079217148,
+      "grad_norm": 0.32981919836353923,
+      "learning_rate": 2.4439074905074216e-05,
+      "loss": 0.369,
+      "step": 1803
+    },
+    {
+      "epoch": 1.6812674743709226,
+      "grad_norm": 0.25191815978509596,
+      "learning_rate": 2.442181567138419e-05,
+      "loss": 0.3187,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6821994408201304,
+      "grad_norm": 0.3066122259460325,
+      "learning_rate": 2.440455643769417e-05,
+      "loss": 0.3293,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6831314072693382,
+      "grad_norm": 0.26557370164473204,
+      "learning_rate": 2.4387297204004146e-05,
+      "loss": 0.3334,
+      "step": 1806
+    },
+    {
+      "epoch": 1.684063373718546,
+      "grad_norm": 0.2947813808216825,
+      "learning_rate": 2.437003797031412e-05,
+      "loss": 0.3485,
+      "step": 1807
+    },
+    {
+      "epoch": 1.684995340167754,
+      "grad_norm": 0.2949582838401032,
+      "learning_rate": 2.4352778736624095e-05,
+      "loss": 0.3691,
+      "step": 1808
+    },
+    {
+      "epoch": 1.6859273066169618,
+      "grad_norm": 0.28450764072960544,
+      "learning_rate": 2.433551950293407e-05,
+      "loss": 0.3384,
+      "step": 1809
+    },
+    {
+      "epoch": 1.6868592730661698,
+      "grad_norm": 0.2763918174215285,
+      "learning_rate": 2.4318260269244047e-05,
+      "loss": 0.3399,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6877912395153776,
+      "grad_norm": 0.23390356308905458,
+      "learning_rate": 2.430100103555402e-05,
+      "loss": 0.3557,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6887232059645854,
+      "grad_norm": 0.25403474060275477,
+      "learning_rate": 2.4283741801864e-05,
+      "loss": 0.3071,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 0.26769837958971937,
+      "learning_rate": 2.4266482568173973e-05,
+      "loss": 0.3367,
+      "step": 1813
+    },
+    {
+      "epoch": 1.690587138863001,
+      "grad_norm": 0.2703743757160016,
+      "learning_rate": 2.424922333448395e-05,
+      "loss": 0.331,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6915191053122087,
+      "grad_norm": 0.24296127590031152,
+      "learning_rate": 2.4231964100793925e-05,
+      "loss": 0.3374,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6924510717614165,
+      "grad_norm": 0.27303469326459257,
+      "learning_rate": 2.4214704867103903e-05,
+      "loss": 0.3442,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6933830382106243,
+      "grad_norm": 0.31374303377749635,
+      "learning_rate": 2.419744563341388e-05,
+      "loss": 0.349,
+      "step": 1817
+    },
+    {
+      "epoch": 1.694315004659832,
+      "grad_norm": 0.2888238188203025,
+      "learning_rate": 2.4180186399723855e-05,
+      "loss": 0.3578,
+      "step": 1818
+    },
+    {
+      "epoch": 1.69524697110904,
+      "grad_norm": 0.26046869745550644,
+      "learning_rate": 2.416292716603383e-05,
+      "loss": 0.3379,
+      "step": 1819
+    },
+    {
+      "epoch": 1.696178937558248,
+      "grad_norm": 0.2637534569873806,
+      "learning_rate": 2.4145667932343803e-05,
+      "loss": 0.335,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6971109040074557,
+      "grad_norm": 0.2778215253554295,
+      "learning_rate": 2.412840869865378e-05,
+      "loss": 0.3448,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6980428704566637,
+      "grad_norm": 0.2355521928736297,
+      "learning_rate": 2.4111149464963755e-05,
+      "loss": 0.3406,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6989748369058715,
+      "grad_norm": 0.26173934141504374,
+      "learning_rate": 2.4093890231273733e-05,
+      "loss": 0.3694,
+      "step": 1823
+    },
+    {
+      "epoch": 1.6999068033550793,
+      "grad_norm": 0.2832461007056905,
+      "learning_rate": 2.4076630997583707e-05,
+      "loss": 0.3534,
+      "step": 1824
+    },
+    {
+      "epoch": 1.700838769804287,
+      "grad_norm": 0.25327756077458224,
+      "learning_rate": 2.4059371763893685e-05,
+      "loss": 0.3405,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7017707362534948,
+      "grad_norm": 0.30631035811979196,
+      "learning_rate": 2.404211253020366e-05,
+      "loss": 0.3486,
+      "step": 1826
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 0.29191629428433746,
+      "learning_rate": 2.4024853296513637e-05,
+      "loss": 0.3813,
+      "step": 1827
+    },
+    {
+      "epoch": 1.7036346691519104,
+      "grad_norm": 0.2823502972251579,
+      "learning_rate": 2.4007594062823614e-05,
+      "loss": 0.3369,
+      "step": 1828
+    },
+    {
+      "epoch": 1.7045666356011182,
+      "grad_norm": 0.3178227608568772,
+      "learning_rate": 2.399033482913359e-05,
+      "loss": 0.3593,
+      "step": 1829
+    },
+    {
+      "epoch": 1.7054986020503262,
+      "grad_norm": 0.27290262871710785,
+      "learning_rate": 2.3973075595443563e-05,
+      "loss": 0.3467,
+      "step": 1830
+    },
+    {
+      "epoch": 1.706430568499534,
+      "grad_norm": 0.2627211881484355,
+      "learning_rate": 2.3955816361753537e-05,
+      "loss": 0.3326,
+      "step": 1831
+    },
+    {
+      "epoch": 1.7073625349487418,
+      "grad_norm": 0.29174803956978723,
+      "learning_rate": 2.3938557128063515e-05,
+      "loss": 0.3487,
+      "step": 1832
+    },
+    {
+      "epoch": 1.7082945013979498,
+      "grad_norm": 0.3112147752624366,
+      "learning_rate": 2.392129789437349e-05,
+      "loss": 0.3617,
+      "step": 1833
+    },
+    {
+      "epoch": 1.7092264678471576,
+      "grad_norm": 0.34140496241488727,
+      "learning_rate": 2.3904038660683467e-05,
+      "loss": 0.3471,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7101584342963654,
+      "grad_norm": 0.3654174279253967,
+      "learning_rate": 2.388677942699344e-05,
+      "loss": 0.3581,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7110904007455732,
+      "grad_norm": 0.29473349132855575,
+      "learning_rate": 2.386952019330342e-05,
+      "loss": 0.3534,
+      "step": 1836
+    },
+    {
+      "epoch": 1.712022367194781,
+      "grad_norm": 0.3064314801389837,
+      "learning_rate": 2.3852260959613397e-05,
+      "loss": 0.3657,
+      "step": 1837
+    },
+    {
+      "epoch": 1.7129543336439887,
+      "grad_norm": 0.3569051390748104,
+      "learning_rate": 2.383500172592337e-05,
+      "loss": 0.3101,
+      "step": 1838
+    },
+    {
+      "epoch": 1.7138863000931965,
+      "grad_norm": 0.34775462848530425,
+      "learning_rate": 2.381774249223335e-05,
+      "loss": 0.3456,
+      "step": 1839
+    },
+    {
+      "epoch": 1.7148182665424043,
+      "grad_norm": 0.2891838583846905,
+      "learning_rate": 2.3800483258543323e-05,
+      "loss": 0.3544,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7157502329916123,
+      "grad_norm": 0.34125407855714796,
+      "learning_rate": 2.3783224024853297e-05,
+      "loss": 0.3567,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7166821994408201,
+      "grad_norm": 0.3635954340817959,
+      "learning_rate": 2.376596479116327e-05,
+      "loss": 0.3532,
+      "step": 1842
+    },
+    {
+      "epoch": 1.717614165890028,
+      "grad_norm": 0.2992292111258229,
+      "learning_rate": 2.374870555747325e-05,
+      "loss": 0.3535,
+      "step": 1843
+    },
+    {
+      "epoch": 1.718546132339236,
+      "grad_norm": 0.5056911724078501,
+      "learning_rate": 2.3731446323783223e-05,
+      "loss": 0.3607,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7194780987884437,
+      "grad_norm": 0.29106843498118223,
+      "learning_rate": 2.37141870900932e-05,
+      "loss": 0.3283,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7204100652376515,
+      "grad_norm": 0.30598430654299397,
+      "learning_rate": 2.3696927856403175e-05,
+      "loss": 0.3367,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7213420316868593,
+      "grad_norm": 0.29396754929230334,
+      "learning_rate": 2.3679668622713153e-05,
+      "loss": 0.3329,
+      "step": 1847
+    },
+    {
+      "epoch": 1.722273998136067,
+      "grad_norm": 0.25998180567929846,
+      "learning_rate": 2.366240938902313e-05,
+      "loss": 0.3076,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7232059645852749,
+      "grad_norm": 0.26804879754430744,
+      "learning_rate": 2.3645150155333105e-05,
+      "loss": 0.3434,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 0.2594640272075965,
+      "learning_rate": 2.3627890921643083e-05,
+      "loss": 0.3446,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7250698974836904,
+      "grad_norm": 0.284961078223277,
+      "learning_rate": 2.3610631687953057e-05,
+      "loss": 0.3395,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7260018639328985,
+      "grad_norm": 0.3037520668886795,
+      "learning_rate": 2.359337245426303e-05,
+      "loss": 0.3496,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7269338303821062,
+      "grad_norm": 0.2582783599704508,
+      "learning_rate": 2.3576113220573005e-05,
+      "loss": 0.3395,
+      "step": 1853
+    },
+    {
+      "epoch": 1.727865796831314,
+      "grad_norm": 0.29839741395618086,
+      "learning_rate": 2.3558853986882983e-05,
+      "loss": 0.3355,
+      "step": 1854
+    },
+    {
+      "epoch": 1.728797763280522,
+      "grad_norm": 0.29964213198985584,
+      "learning_rate": 2.3541594753192957e-05,
+      "loss": 0.3512,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 0.27374013517780504,
+      "learning_rate": 2.3524335519502935e-05,
+      "loss": 0.3478,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7306616961789376,
+      "grad_norm": 0.28926723165554147,
+      "learning_rate": 2.3507076285812913e-05,
+      "loss": 0.3644,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7315936626281454,
+      "grad_norm": 0.26732952411397354,
+      "learning_rate": 2.3489817052122887e-05,
+      "loss": 0.3364,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7325256290773532,
+      "grad_norm": 0.2602736599367072,
+      "learning_rate": 2.3472557818432865e-05,
+      "loss": 0.3434,
+      "step": 1859
+    },
+    {
+      "epoch": 1.733457595526561,
+      "grad_norm": 0.29608783540651545,
+      "learning_rate": 2.345529858474284e-05,
+      "loss": 0.3265,
+      "step": 1860
+    },
+    {
+      "epoch": 1.7343895619757688,
+      "grad_norm": 0.30613738055086187,
+      "learning_rate": 2.3438039351052817e-05,
+      "loss": 0.3325,
+      "step": 1861
+    },
+    {
+      "epoch": 1.7353215284249766,
+      "grad_norm": 0.2952447184235771,
+      "learning_rate": 2.342078011736279e-05,
+      "loss": 0.3323,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7362534948741846,
+      "grad_norm": 0.34215966571055567,
+      "learning_rate": 2.3403520883672765e-05,
+      "loss": 0.3511,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7371854613233924,
+      "grad_norm": 0.25742912333076134,
+      "learning_rate": 2.338626164998274e-05,
+      "loss": 0.3571,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7381174277726001,
+      "grad_norm": 0.326520405918653,
+      "learning_rate": 2.3369002416292717e-05,
+      "loss": 0.3414,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7390493942218082,
+      "grad_norm": 0.2608096400890844,
+      "learning_rate": 2.335174318260269e-05,
+      "loss": 0.3317,
+      "step": 1866
+    },
+    {
+      "epoch": 1.739981360671016,
+      "grad_norm": 0.26102733124896427,
+      "learning_rate": 2.333448394891267e-05,
+      "loss": 0.3509,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7409133271202237,
+      "grad_norm": 0.32272168131939766,
+      "learning_rate": 2.3317224715222647e-05,
+      "loss": 0.3345,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7418452935694315,
+      "grad_norm": 0.25433261423685144,
+      "learning_rate": 2.329996548153262e-05,
+      "loss": 0.3447,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7427772600186393,
+      "grad_norm": 0.3240156203134915,
+      "learning_rate": 2.32827062478426e-05,
+      "loss": 0.3483,
+      "step": 1870
+    },
+    {
+      "epoch": 1.743709226467847,
+      "grad_norm": 0.26732084193497774,
+      "learning_rate": 2.3265447014152573e-05,
+      "loss": 0.3615,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7446411929170549,
+      "grad_norm": 0.3610580965989959,
+      "learning_rate": 2.324818778046255e-05,
+      "loss": 0.3597,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7455731593662627,
+      "grad_norm": 0.3026344571986399,
+      "learning_rate": 2.3230928546772525e-05,
+      "loss": 0.3554,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7465051258154707,
+      "grad_norm": 0.26980041003020794,
+      "learning_rate": 2.32136693130825e-05,
+      "loss": 0.3342,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7474370922646785,
+      "grad_norm": 0.26179619266996396,
+      "learning_rate": 2.3196410079392474e-05,
+      "loss": 0.3414,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7483690587138863,
+      "grad_norm": 0.2924629809437586,
+      "learning_rate": 2.317915084570245e-05,
+      "loss": 0.362,
+      "step": 1876
+    },
+    {
+      "epoch": 1.7493010251630943,
+      "grad_norm": 0.26932642770340953,
+      "learning_rate": 2.316189161201243e-05,
+      "loss": 0.3337,
+      "step": 1877
+    },
+    {
+      "epoch": 1.750232991612302,
+      "grad_norm": 0.27958932759709126,
+      "learning_rate": 2.3144632378322403e-05,
+      "loss": 0.3431,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7511649580615098,
+      "grad_norm": 0.2918913514531261,
+      "learning_rate": 2.312737314463238e-05,
+      "loss": 0.3493,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7520969245107176,
+      "grad_norm": 0.3316318590832368,
+      "learning_rate": 2.3110113910942355e-05,
+      "loss": 0.3376,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7530288909599254,
+      "grad_norm": 0.2984189523372345,
+      "learning_rate": 2.3092854677252333e-05,
+      "loss": 0.3643,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7539608574091332,
+      "grad_norm": 0.31371348155429074,
+      "learning_rate": 2.3075595443562307e-05,
+      "loss": 0.3438,
+      "step": 1882
+    },
+    {
+      "epoch": 1.754892823858341,
+      "grad_norm": 0.34014760124336646,
+      "learning_rate": 2.305833620987228e-05,
+      "loss": 0.3603,
+      "step": 1883
+    },
+    {
+      "epoch": 1.7558247903075488,
+      "grad_norm": 0.2741696008999161,
+      "learning_rate": 2.3041076976182256e-05,
+      "loss": 0.3368,
+      "step": 1884
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.29122658119249417,
+      "learning_rate": 2.3023817742492234e-05,
+      "loss": 0.3478,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7576887232059646,
+      "grad_norm": 0.345521395765362,
+      "learning_rate": 2.3006558508802208e-05,
+      "loss": 0.3631,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 0.29246302169152133,
+      "learning_rate": 2.2989299275112186e-05,
+      "loss": 0.3461,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7595526561043804,
+      "grad_norm": 0.2600564041553437,
+      "learning_rate": 2.2972040041422163e-05,
+      "loss": 0.3368,
+      "step": 1888
+    },
+    {
+      "epoch": 1.7604846225535882,
+      "grad_norm": 0.27377256278059386,
+      "learning_rate": 2.2954780807732138e-05,
+      "loss": 0.338,
+      "step": 1889
+    },
+    {
+      "epoch": 1.761416589002796,
+      "grad_norm": 0.28226526276331093,
+      "learning_rate": 2.2937521574042115e-05,
+      "loss": 0.3554,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7623485554520038,
+      "grad_norm": 0.24910840815675944,
+      "learning_rate": 2.292026234035209e-05,
+      "loss": 0.3344,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7632805219012115,
+      "grad_norm": 0.3260308497793744,
+      "learning_rate": 2.2903003106662067e-05,
+      "loss": 0.3374,
+      "step": 1892
+    },
+    {
+      "epoch": 1.7642124883504193,
+      "grad_norm": 0.30810439346895535,
+      "learning_rate": 2.288574387297204e-05,
+      "loss": 0.3434,
+      "step": 1893
+    },
+    {
+      "epoch": 1.7651444547996271,
+      "grad_norm": 0.2656624366803814,
+      "learning_rate": 2.2868484639282016e-05,
+      "loss": 0.3497,
+      "step": 1894
+    },
+    {
+      "epoch": 1.766076421248835,
+      "grad_norm": 0.3159832156073944,
+      "learning_rate": 2.285122540559199e-05,
+      "loss": 0.3639,
+      "step": 1895
+    },
+    {
+      "epoch": 1.767008387698043,
+      "grad_norm": 0.32019416815244706,
+      "learning_rate": 2.2833966171901968e-05,
+      "loss": 0.349,
+      "step": 1896
+    },
+    {
+      "epoch": 1.7679403541472507,
+      "grad_norm": 0.2713944713889817,
+      "learning_rate": 2.2816706938211945e-05,
+      "loss": 0.3355,
+      "step": 1897
+    },
+    {
+      "epoch": 1.7688723205964585,
+      "grad_norm": 0.31142049064047084,
+      "learning_rate": 2.279944770452192e-05,
+      "loss": 0.3276,
+      "step": 1898
+    },
+    {
+      "epoch": 1.7698042870456665,
+      "grad_norm": 0.26075798397494077,
+      "learning_rate": 2.2782188470831897e-05,
+      "loss": 0.3394,
+      "step": 1899
+    },
+    {
+      "epoch": 1.7707362534948743,
+      "grad_norm": 0.4409566091182875,
+      "learning_rate": 2.276492923714187e-05,
+      "loss": 0.3418,
+      "step": 1900
+    },
+    {
+      "epoch": 1.771668219944082,
+      "grad_norm": 0.3113223884838847,
+      "learning_rate": 2.274767000345185e-05,
+      "loss": 0.3277,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7726001863932899,
+      "grad_norm": 0.2975607927316072,
+      "learning_rate": 2.2730410769761824e-05,
+      "loss": 0.3331,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7735321528424977,
+      "grad_norm": 0.2397054261546087,
+      "learning_rate": 2.27131515360718e-05,
+      "loss": 0.3226,
+      "step": 1903
+    },
+    {
+      "epoch": 1.7744641192917054,
+      "grad_norm": 0.2952206060152297,
+      "learning_rate": 2.2695892302381776e-05,
+      "loss": 0.3595,
+      "step": 1904
+    },
+    {
+      "epoch": 1.7753960857409132,
+      "grad_norm": 0.28272721704386555,
+      "learning_rate": 2.267863306869175e-05,
+      "loss": 0.3451,
+      "step": 1905
+    },
+    {
+      "epoch": 1.776328052190121,
+      "grad_norm": 0.2970041336504249,
+      "learning_rate": 2.2661373835001724e-05,
+      "loss": 0.3492,
+      "step": 1906
+    },
+    {
+      "epoch": 1.777260018639329,
+      "grad_norm": 0.26476428746290787,
+      "learning_rate": 2.2644114601311702e-05,
+      "loss": 0.348,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7781919850885368,
+      "grad_norm": 0.3170240304624983,
+      "learning_rate": 2.262685536762168e-05,
+      "loss": 0.3454,
+      "step": 1908
+    },
+    {
+      "epoch": 1.7791239515377446,
+      "grad_norm": 0.332925998639465,
+      "learning_rate": 2.2609596133931654e-05,
+      "loss": 0.3556,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7800559179869526,
+      "grad_norm": 0.2653164838632094,
+      "learning_rate": 2.259233690024163e-05,
+      "loss": 0.3529,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7809878844361604,
+      "grad_norm": 0.32122420881626434,
+      "learning_rate": 2.2575077666551606e-05,
+      "loss": 0.3509,
+      "step": 1911
+    },
+    {
+      "epoch": 1.7819198508853682,
+      "grad_norm": 0.32324401981440576,
+      "learning_rate": 2.2557818432861583e-05,
+      "loss": 0.3234,
+      "step": 1912
+    },
+    {
+      "epoch": 1.782851817334576,
+      "grad_norm": 0.28522728277011367,
+      "learning_rate": 2.2540559199171558e-05,
+      "loss": 0.333,
+      "step": 1913
+    },
+    {
+      "epoch": 1.7837837837837838,
+      "grad_norm": 0.26409735472280793,
+      "learning_rate": 2.2523299965481535e-05,
+      "loss": 0.3432,
+      "step": 1914
+    },
+    {
+      "epoch": 1.7847157502329916,
+      "grad_norm": 0.29342509309687825,
+      "learning_rate": 2.250604073179151e-05,
+      "loss": 0.3165,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7856477166821993,
+      "grad_norm": 0.34746391916316877,
+      "learning_rate": 2.2488781498101484e-05,
+      "loss": 0.3486,
+      "step": 1916
+    },
+    {
+      "epoch": 1.7865796831314071,
+      "grad_norm": 0.29865942658618083,
+      "learning_rate": 2.2471522264411462e-05,
+      "loss": 0.3488,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7875116495806151,
+      "grad_norm": 0.28054372568166525,
+      "learning_rate": 2.2454263030721436e-05,
+      "loss": 0.3544,
+      "step": 1918
+    },
+    {
+      "epoch": 1.788443616029823,
+      "grad_norm": 0.3322751290621268,
+      "learning_rate": 2.2437003797031414e-05,
+      "loss": 0.3422,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7893755824790307,
+      "grad_norm": 0.32938201239340187,
+      "learning_rate": 2.2419744563341388e-05,
+      "loss": 0.3528,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7903075489282387,
+      "grad_norm": 0.270015343830145,
+      "learning_rate": 2.2402485329651366e-05,
+      "loss": 0.359,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7912395153774465,
+      "grad_norm": 0.3510456524019387,
+      "learning_rate": 2.238522609596134e-05,
+      "loss": 0.3394,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7921714818266543,
+      "grad_norm": 0.30063804836059604,
+      "learning_rate": 2.2367966862271318e-05,
+      "loss": 0.3442,
+      "step": 1923
+    },
+    {
+      "epoch": 1.793103448275862,
+      "grad_norm": 0.28259349907245734,
+      "learning_rate": 2.2350707628581292e-05,
+      "loss": 0.3363,
+      "step": 1924
+    },
+    {
+      "epoch": 1.7940354147250699,
+      "grad_norm": 0.2606383179651388,
+      "learning_rate": 2.233344839489127e-05,
+      "loss": 0.3512,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7949673811742777,
+      "grad_norm": 0.2773105934999208,
+      "learning_rate": 2.2316189161201244e-05,
+      "loss": 0.3339,
+      "step": 1926
+    },
+    {
+      "epoch": 1.7958993476234855,
+      "grad_norm": 0.2816207356314107,
+      "learning_rate": 2.2298929927511218e-05,
+      "loss": 0.3381,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7968313140726933,
+      "grad_norm": 0.2621412757615487,
+      "learning_rate": 2.2281670693821196e-05,
+      "loss": 0.3497,
+      "step": 1928
+    },
+    {
+      "epoch": 1.7977632805219013,
+      "grad_norm": 0.26434542181335646,
+      "learning_rate": 2.226441146013117e-05,
+      "loss": 0.336,
+      "step": 1929
+    },
+    {
+      "epoch": 1.798695246971109,
+      "grad_norm": 0.2927984768834388,
+      "learning_rate": 2.2247152226441148e-05,
+      "loss": 0.355,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7996272134203168,
+      "grad_norm": 0.30813739971455034,
+      "learning_rate": 2.2229892992751122e-05,
+      "loss": 0.3491,
+      "step": 1931
+    },
+    {
+      "epoch": 1.8005591798695249,
+      "grad_norm": 0.2540712434540595,
+      "learning_rate": 2.22126337590611e-05,
+      "loss": 0.3455,
+      "step": 1932
+    },
+    {
+      "epoch": 1.8014911463187326,
+      "grad_norm": 0.2867621010558032,
+      "learning_rate": 2.2195374525371074e-05,
+      "loss": 0.3533,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8024231127679404,
+      "grad_norm": 0.26102383889473113,
+      "learning_rate": 2.2178115291681052e-05,
+      "loss": 0.3437,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8033550792171482,
+      "grad_norm": 0.29147254187118155,
+      "learning_rate": 2.2160856057991026e-05,
+      "loss": 0.3453,
+      "step": 1935
+    },
+    {
+      "epoch": 1.804287045666356,
+      "grad_norm": 0.31494527040157944,
+      "learning_rate": 2.2143596824301004e-05,
+      "loss": 0.3474,
+      "step": 1936
+    },
+    {
+      "epoch": 1.8052190121155638,
+      "grad_norm": 0.2992440525794497,
+      "learning_rate": 2.2126337590610978e-05,
+      "loss": 0.3315,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8061509785647716,
+      "grad_norm": 0.27897586733476293,
+      "learning_rate": 2.2109078356920952e-05,
+      "loss": 0.3245,
+      "step": 1938
+    },
+    {
+      "epoch": 1.8070829450139794,
+      "grad_norm": 0.2703984063619173,
+      "learning_rate": 2.209181912323093e-05,
+      "loss": 0.3366,
+      "step": 1939
+    },
+    {
+      "epoch": 1.8080149114631874,
+      "grad_norm": 0.27990040237300023,
+      "learning_rate": 2.2074559889540904e-05,
+      "loss": 0.3365,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8089468779123952,
+      "grad_norm": 0.30600845226994705,
+      "learning_rate": 2.2057300655850882e-05,
+      "loss": 0.3409,
+      "step": 1941
+    },
+    {
+      "epoch": 1.809878844361603,
+      "grad_norm": 0.2716805456471926,
+      "learning_rate": 2.2040041422160856e-05,
+      "loss": 0.3541,
+      "step": 1942
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 0.2867317849971531,
+      "learning_rate": 2.2022782188470834e-05,
+      "loss": 0.3545,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8117427772600188,
+      "grad_norm": 0.2744360356925306,
+      "learning_rate": 2.2005522954780808e-05,
+      "loss": 0.3533,
+      "step": 1944
+    },
+    {
+      "epoch": 1.8126747437092265,
+      "grad_norm": 0.2616327434227867,
+      "learning_rate": 2.1988263721090786e-05,
+      "loss": 0.3577,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8136067101584343,
+      "grad_norm": 0.29637963195128186,
+      "learning_rate": 2.197100448740076e-05,
+      "loss": 0.3339,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8145386766076421,
+      "grad_norm": 0.2658073997871166,
+      "learning_rate": 2.1953745253710738e-05,
+      "loss": 0.3264,
+      "step": 1947
+    },
+    {
+      "epoch": 1.81547064305685,
+      "grad_norm": 0.2985643673509504,
+      "learning_rate": 2.1936486020020712e-05,
+      "loss": 0.3453,
+      "step": 1948
+    },
+    {
+      "epoch": 1.8164026095060577,
+      "grad_norm": 0.29318508964638024,
+      "learning_rate": 2.1919226786330686e-05,
+      "loss": 0.3334,
+      "step": 1949
+    },
+    {
+      "epoch": 1.8173345759552655,
+      "grad_norm": 0.27081857509514956,
+      "learning_rate": 2.1901967552640664e-05,
+      "loss": 0.3376,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8182665424044733,
+      "grad_norm": 0.30683408060528594,
+      "learning_rate": 2.188470831895064e-05,
+      "loss": 0.3502,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8191985088536813,
+      "grad_norm": 0.30131858772180437,
+      "learning_rate": 2.1867449085260616e-05,
+      "loss": 0.3285,
+      "step": 1952
+    },
+    {
+      "epoch": 1.820130475302889,
+      "grad_norm": 0.366576455062228,
+      "learning_rate": 2.185018985157059e-05,
+      "loss": 0.3454,
+      "step": 1953
+    },
+    {
+      "epoch": 1.821062441752097,
+      "grad_norm": 0.2708617787704819,
+      "learning_rate": 2.1832930617880568e-05,
+      "loss": 0.3455,
+      "step": 1954
+    },
+    {
+      "epoch": 1.8219944082013049,
+      "grad_norm": 0.3392725038580427,
+      "learning_rate": 2.1815671384190542e-05,
+      "loss": 0.3467,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8229263746505127,
+      "grad_norm": 0.2834021974644337,
+      "learning_rate": 2.179841215050052e-05,
+      "loss": 0.3309,
+      "step": 1956
+    },
+    {
+      "epoch": 1.8238583410997204,
+      "grad_norm": 0.3373219355304193,
+      "learning_rate": 2.1781152916810494e-05,
+      "loss": 0.3577,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8247903075489282,
+      "grad_norm": 0.2794645124602264,
+      "learning_rate": 2.1763893683120472e-05,
+      "loss": 0.3378,
+      "step": 1958
+    },
+    {
+      "epoch": 1.825722273998136,
+      "grad_norm": 0.3007525786779994,
+      "learning_rate": 2.1746634449430446e-05,
+      "loss": 0.3541,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8266542404473438,
+      "grad_norm": 0.33196315854834224,
+      "learning_rate": 2.172937521574042e-05,
+      "loss": 0.3533,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 0.263023032563813,
+      "learning_rate": 2.1712115982050398e-05,
+      "loss": 0.3478,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8285181733457594,
+      "grad_norm": 0.32533566952806203,
+      "learning_rate": 2.1694856748360373e-05,
+      "loss": 0.3529,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8294501397949674,
+      "grad_norm": 0.2791781618657612,
+      "learning_rate": 2.167759751467035e-05,
+      "loss": 0.3357,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8303821062441752,
+      "grad_norm": 0.3579433414308658,
+      "learning_rate": 2.1660338280980324e-05,
+      "loss": 0.3594,
+      "step": 1964
+    },
+    {
+      "epoch": 1.8313140726933832,
+      "grad_norm": 0.29693404197414563,
+      "learning_rate": 2.1643079047290302e-05,
+      "loss": 0.3461,
+      "step": 1965
+    },
+    {
+      "epoch": 1.832246039142591,
+      "grad_norm": 0.31987674102901575,
+      "learning_rate": 2.1625819813600276e-05,
+      "loss": 0.3467,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8331780055917988,
+      "grad_norm": 0.2975074095854175,
+      "learning_rate": 2.1608560579910254e-05,
+      "loss": 0.345,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8341099720410066,
+      "grad_norm": 0.2636079276919061,
+      "learning_rate": 2.1591301346220232e-05,
+      "loss": 0.3291,
+      "step": 1968
+    },
+    {
+      "epoch": 1.8350419384902144,
+      "grad_norm": 0.3060292134681667,
+      "learning_rate": 2.1574042112530206e-05,
+      "loss": 0.3684,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8359739049394221,
+      "grad_norm": 0.2909370419146219,
+      "learning_rate": 2.155678287884018e-05,
+      "loss": 0.3404,
+      "step": 1970
+    },
+    {
+      "epoch": 1.83690587138863,
+      "grad_norm": 0.2938983793903817,
+      "learning_rate": 2.1539523645150155e-05,
+      "loss": 0.3609,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 0.309229692202302,
+      "learning_rate": 2.1522264411460132e-05,
+      "loss": 0.3557,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8387698042870455,
+      "grad_norm": 0.3123423214993961,
+      "learning_rate": 2.1505005177770107e-05,
+      "loss": 0.3409,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8397017707362535,
+      "grad_norm": 0.3154939650701389,
+      "learning_rate": 2.1487745944080084e-05,
+      "loss": 0.3592,
+      "step": 1974
+    },
+    {
+      "epoch": 1.8406337371854613,
+      "grad_norm": 0.27716763029562624,
+      "learning_rate": 2.147048671039006e-05,
+      "loss": 0.3534,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8415657036346693,
+      "grad_norm": 0.2605064426910099,
+      "learning_rate": 2.1453227476700036e-05,
+      "loss": 0.3634,
+      "step": 1976
+    },
+    {
+      "epoch": 1.842497670083877,
+      "grad_norm": 0.2625878714883827,
+      "learning_rate": 2.143596824301001e-05,
+      "loss": 0.343,
+      "step": 1977
+    },
+    {
+      "epoch": 1.843429636533085,
+      "grad_norm": 0.2877473450304713,
+      "learning_rate": 2.1418709009319988e-05,
+      "loss": 0.3479,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8443616029822927,
+      "grad_norm": 0.24911746774211244,
+      "learning_rate": 2.1401449775629963e-05,
+      "loss": 0.3298,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8452935694315005,
+      "grad_norm": 0.28863394150484134,
+      "learning_rate": 2.1384190541939937e-05,
+      "loss": 0.3369,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8462255358807083,
+      "grad_norm": 0.3110859180889837,
+      "learning_rate": 2.1366931308249915e-05,
+      "loss": 0.3429,
+      "step": 1981
+    },
+    {
+      "epoch": 1.847157502329916,
+      "grad_norm": 0.31187855699083245,
+      "learning_rate": 2.134967207455989e-05,
+      "loss": 0.3391,
+      "step": 1982
+    },
+    {
+      "epoch": 1.8480894687791238,
+      "grad_norm": 0.2960417269689427,
+      "learning_rate": 2.1332412840869866e-05,
+      "loss": 0.3365,
+      "step": 1983
+    },
+    {
+      "epoch": 1.8490214352283316,
+      "grad_norm": 0.3110492265295226,
+      "learning_rate": 2.131515360717984e-05,
+      "loss": 0.341,
+      "step": 1984
+    },
+    {
+      "epoch": 1.8499534016775396,
+      "grad_norm": 0.2681044878531237,
+      "learning_rate": 2.129789437348982e-05,
+      "loss": 0.3412,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8508853681267474,
+      "grad_norm": 0.3173590548171478,
+      "learning_rate": 2.1280635139799793e-05,
+      "loss": 0.3674,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8518173345759554,
+      "grad_norm": 0.32916932608582566,
+      "learning_rate": 2.126337590610977e-05,
+      "loss": 0.3643,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8527493010251632,
+      "grad_norm": 0.2688170037362449,
+      "learning_rate": 2.1246116672419748e-05,
+      "loss": 0.3578,
+      "step": 1988
+    },
+    {
+      "epoch": 1.853681267474371,
+      "grad_norm": 0.3315315872118712,
+      "learning_rate": 2.1228857438729722e-05,
+      "loss": 0.3592,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8546132339235788,
+      "grad_norm": 0.268939227969717,
+      "learning_rate": 2.1211598205039697e-05,
+      "loss": 0.3451,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8555452003727866,
+      "grad_norm": 0.24915753462674506,
+      "learning_rate": 2.119433897134967e-05,
+      "loss": 0.3244,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8564771668219944,
+      "grad_norm": 0.30424989118992823,
+      "learning_rate": 2.117707973765965e-05,
+      "loss": 0.3508,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8574091332712022,
+      "grad_norm": 0.29275232358893377,
+      "learning_rate": 2.1159820503969623e-05,
+      "loss": 0.3468,
+      "step": 1993
+    },
+    {
+      "epoch": 1.85834109972041,
+      "grad_norm": 0.27684111881300066,
+      "learning_rate": 2.11425612702796e-05,
+      "loss": 0.3466,
+      "step": 1994
+    },
+    {
+      "epoch": 1.8592730661696177,
+      "grad_norm": 0.31114742175619026,
+      "learning_rate": 2.1125302036589575e-05,
+      "loss": 0.3538,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8602050326188257,
+      "grad_norm": 0.3327378093311978,
+      "learning_rate": 2.1108042802899553e-05,
+      "loss": 0.3503,
+      "step": 1996
+    },
+    {
+      "epoch": 1.8611369990680335,
+      "grad_norm": 0.3118999310257516,
+      "learning_rate": 2.1090783569209527e-05,
+      "loss": 0.3444,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 0.30088564127571665,
+      "learning_rate": 2.1073524335519505e-05,
+      "loss": 0.3139,
+      "step": 1998
+    },
+    {
+      "epoch": 1.8630009319664493,
+      "grad_norm": 0.32753618787773586,
+      "learning_rate": 2.1056265101829482e-05,
+      "loss": 0.3692,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8639328984156571,
+      "grad_norm": 0.2702428875409598,
+      "learning_rate": 2.1039005868139457e-05,
+      "loss": 0.3387,
+      "step": 2000
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 0.2825598773317577,
+      "learning_rate": 2.102174663444943e-05,
+      "loss": 0.3315,
+      "step": 2001
+    },
+    {
+      "epoch": 1.8657968313140727,
+      "grad_norm": 0.39406595161787245,
+      "learning_rate": 2.1004487400759405e-05,
+      "loss": 0.3527,
+      "step": 2002
+    },
+    {
+      "epoch": 1.8667287977632805,
+      "grad_norm": 0.33232908021436663,
+      "learning_rate": 2.0987228167069383e-05,
+      "loss": 0.3205,
+      "step": 2003
+    },
+    {
+      "epoch": 1.8676607642124883,
+      "grad_norm": 0.30130117339929735,
+      "learning_rate": 2.0969968933379357e-05,
+      "loss": 0.3336,
+      "step": 2004
+    },
+    {
+      "epoch": 1.868592730661696,
+      "grad_norm": 0.34881355447318885,
+      "learning_rate": 2.0952709699689335e-05,
+      "loss": 0.3475,
+      "step": 2005
+    },
+    {
+      "epoch": 1.8695246971109039,
+      "grad_norm": 0.36332694651704955,
+      "learning_rate": 2.093545046599931e-05,
+      "loss": 0.3543,
+      "step": 2006
+    },
+    {
+      "epoch": 1.8704566635601119,
+      "grad_norm": 0.30853574687210894,
+      "learning_rate": 2.0918191232309287e-05,
+      "loss": 0.3625,
+      "step": 2007
+    },
+    {
+      "epoch": 1.8713886300093197,
+      "grad_norm": 0.34335340541951015,
+      "learning_rate": 2.0900931998619264e-05,
+      "loss": 0.3529,
+      "step": 2008
+    },
+    {
+      "epoch": 1.8723205964585274,
+      "grad_norm": 0.33144811083534564,
+      "learning_rate": 2.088367276492924e-05,
+      "loss": 0.3511,
+      "step": 2009
+    },
+    {
+      "epoch": 1.8732525629077355,
+      "grad_norm": 0.268999172144125,
+      "learning_rate": 2.0866413531239216e-05,
+      "loss": 0.3407,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8741845293569432,
+      "grad_norm": 0.30733962134880394,
+      "learning_rate": 2.084915429754919e-05,
+      "loss": 0.3335,
+      "step": 2011
+    },
+    {
+      "epoch": 1.875116495806151,
+      "grad_norm": 0.2811554544138188,
+      "learning_rate": 2.0831895063859165e-05,
+      "loss": 0.3097,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8760484622553588,
+      "grad_norm": 0.31039277035858676,
+      "learning_rate": 2.081463583016914e-05,
+      "loss": 0.3479,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8769804287045666,
+      "grad_norm": 0.3287181398078748,
+      "learning_rate": 2.0797376596479117e-05,
+      "loss": 0.3542,
+      "step": 2014
+    },
+    {
+      "epoch": 1.8779123951537744,
+      "grad_norm": 0.3094882701594067,
+      "learning_rate": 2.078011736278909e-05,
+      "loss": 0.3464,
+      "step": 2015
+    },
+    {
+      "epoch": 1.8788443616029822,
+      "grad_norm": 0.3317833333838534,
+      "learning_rate": 2.076285812909907e-05,
+      "loss": 0.3493,
+      "step": 2016
+    },
+    {
+      "epoch": 1.87977632805219,
+      "grad_norm": 0.31168582257588034,
+      "learning_rate": 2.0745598895409043e-05,
+      "loss": 0.3201,
+      "step": 2017
+    },
+    {
+      "epoch": 1.880708294501398,
+      "grad_norm": 0.2780435160579364,
+      "learning_rate": 2.072833966171902e-05,
+      "loss": 0.3236,
+      "step": 2018
+    },
+    {
+      "epoch": 1.8816402609506058,
+      "grad_norm": 0.3301340997312112,
+      "learning_rate": 2.0711080428029e-05,
+      "loss": 0.3419,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8825722273998136,
+      "grad_norm": 0.2568167224391666,
+      "learning_rate": 2.0693821194338973e-05,
+      "loss": 0.3326,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8835041938490216,
+      "grad_norm": 0.33296219584550013,
+      "learning_rate": 2.067656196064895e-05,
+      "loss": 0.3468,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8844361602982294,
+      "grad_norm": 0.28407160832350314,
+      "learning_rate": 2.0659302726958925e-05,
+      "loss": 0.3298,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8853681267474371,
+      "grad_norm": 0.2859723203045751,
+      "learning_rate": 2.06420434932689e-05,
+      "loss": 0.3397,
+      "step": 2023
+    },
+    {
+      "epoch": 1.886300093196645,
+      "grad_norm": 0.32212132921824843,
+      "learning_rate": 2.0624784259578873e-05,
+      "loss": 0.3321,
+      "step": 2024
+    },
+    {
+      "epoch": 1.8872320596458527,
+      "grad_norm": 0.30374324691438187,
+      "learning_rate": 2.060752502588885e-05,
+      "loss": 0.3488,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8881640260950605,
+      "grad_norm": 0.29884731177008245,
+      "learning_rate": 2.0590265792198825e-05,
+      "loss": 0.3509,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8890959925442683,
+      "grad_norm": 0.31613659052722853,
+      "learning_rate": 2.0573006558508803e-05,
+      "loss": 0.3412,
+      "step": 2027
+    },
+    {
+      "epoch": 1.890027958993476,
+      "grad_norm": 0.2633094935277315,
+      "learning_rate": 2.055574732481878e-05,
+      "loss": 0.3339,
+      "step": 2028
+    },
+    {
+      "epoch": 1.890959925442684,
+      "grad_norm": 0.2930409956150796,
+      "learning_rate": 2.0538488091128755e-05,
+      "loss": 0.3542,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 0.23965490313280294,
+      "learning_rate": 2.0521228857438733e-05,
+      "loss": 0.3517,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8928238583410997,
+      "grad_norm": 0.30513395835668367,
+      "learning_rate": 2.0503969623748707e-05,
+      "loss": 0.3564,
+      "step": 2031
+    },
+    {
+      "epoch": 1.8937558247903077,
+      "grad_norm": 0.27466535476462817,
+      "learning_rate": 2.0486710390058685e-05,
+      "loss": 0.3464,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8946877912395155,
+      "grad_norm": 0.2420072940195495,
+      "learning_rate": 2.046945115636866e-05,
+      "loss": 0.34,
+      "step": 2033
+    },
+    {
+      "epoch": 1.8956197576887233,
+      "grad_norm": 0.29278560522287506,
+      "learning_rate": 2.0452191922678633e-05,
+      "loss": 0.3231,
+      "step": 2034
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 0.24800693228527723,
+      "learning_rate": 2.0434932688988608e-05,
+      "loss": 0.3315,
+      "step": 2035
+    },
+    {
+      "epoch": 1.8974836905871388,
+      "grad_norm": 0.3119059433353562,
+      "learning_rate": 2.0417673455298585e-05,
+      "loss": 0.3569,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8984156570363466,
+      "grad_norm": 0.2719152452938805,
+      "learning_rate": 2.040041422160856e-05,
+      "loss": 0.3164,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8993476234855544,
+      "grad_norm": 0.2665124264865322,
+      "learning_rate": 2.0383154987918537e-05,
+      "loss": 0.3413,
+      "step": 2038
+    },
+    {
+      "epoch": 1.9002795899347622,
+      "grad_norm": 0.3078778336043995,
+      "learning_rate": 2.0365895754228515e-05,
+      "loss": 0.3569,
+      "step": 2039
+    },
+    {
+      "epoch": 1.9012115563839702,
+      "grad_norm": 0.24743860819025804,
+      "learning_rate": 2.034863652053849e-05,
+      "loss": 0.3402,
+      "step": 2040
+    },
+    {
+      "epoch": 1.902143522833178,
+      "grad_norm": 0.2871798917020383,
+      "learning_rate": 2.0331377286848467e-05,
+      "loss": 0.3505,
+      "step": 2041
+    },
+    {
+      "epoch": 1.9030754892823858,
+      "grad_norm": 0.2942545158351549,
+      "learning_rate": 2.031411805315844e-05,
+      "loss": 0.363,
+      "step": 2042
+    },
+    {
+      "epoch": 1.9040074557315938,
+      "grad_norm": 0.293958781494719,
+      "learning_rate": 2.029685881946842e-05,
+      "loss": 0.3716,
+      "step": 2043
+    },
+    {
+      "epoch": 1.9049394221808016,
+      "grad_norm": 0.2790687700177665,
+      "learning_rate": 2.0279599585778393e-05,
+      "loss": 0.3628,
+      "step": 2044
+    },
+    {
+      "epoch": 1.9058713886300094,
+      "grad_norm": 0.23610569120878544,
+      "learning_rate": 2.0262340352088367e-05,
+      "loss": 0.3438,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9068033550792172,
+      "grad_norm": 0.3097364649081726,
+      "learning_rate": 2.024508111839834e-05,
+      "loss": 0.3405,
+      "step": 2046
+    },
+    {
+      "epoch": 1.907735321528425,
+      "grad_norm": 0.29010475737218294,
+      "learning_rate": 2.022782188470832e-05,
+      "loss": 0.3383,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9086672879776327,
+      "grad_norm": 0.29808375239113233,
+      "learning_rate": 2.0210562651018294e-05,
+      "loss": 0.3717,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9095992544268405,
+      "grad_norm": 0.2767630634173141,
+      "learning_rate": 2.019330341732827e-05,
+      "loss": 0.3337,
+      "step": 2049
+    },
+    {
+      "epoch": 1.9105312208760483,
+      "grad_norm": 0.31177964360873833,
+      "learning_rate": 2.017604418363825e-05,
+      "loss": 0.3591,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9114631873252563,
+      "grad_norm": 0.2724496816555845,
+      "learning_rate": 2.0158784949948223e-05,
+      "loss": 0.3583,
+      "step": 2051
+    },
+    {
+      "epoch": 1.9123951537744641,
+      "grad_norm": 0.30192831627092137,
+      "learning_rate": 2.01415257162582e-05,
+      "loss": 0.3274,
+      "step": 2052
+    },
+    {
+      "epoch": 1.913327120223672,
+      "grad_norm": 0.28737139184263943,
+      "learning_rate": 2.0124266482568175e-05,
+      "loss": 0.3441,
+      "step": 2053
+    },
+    {
+      "epoch": 1.91425908667288,
+      "grad_norm": 0.248380388197744,
+      "learning_rate": 2.0107007248878153e-05,
+      "loss": 0.3391,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9151910531220877,
+      "grad_norm": 0.28915904146057975,
+      "learning_rate": 2.0089748015188127e-05,
+      "loss": 0.3462,
+      "step": 2055
+    },
+    {
+      "epoch": 1.9161230195712955,
+      "grad_norm": 0.267581404208985,
+      "learning_rate": 2.00724887814981e-05,
+      "loss": 0.3327,
+      "step": 2056
+    },
+    {
+      "epoch": 1.9170549860205033,
+      "grad_norm": 0.2426233381630608,
+      "learning_rate": 2.0055229547808076e-05,
+      "loss": 0.3468,
+      "step": 2057
+    },
+    {
+      "epoch": 1.917986952469711,
+      "grad_norm": 0.28409720399546995,
+      "learning_rate": 2.0037970314118053e-05,
+      "loss": 0.3659,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 0.29761887597471737,
+      "learning_rate": 2.002071108042803e-05,
+      "loss": 0.3616,
+      "step": 2059
+    },
+    {
+      "epoch": 1.9198508853681266,
+      "grad_norm": 0.24280035302643185,
+      "learning_rate": 2.0003451846738005e-05,
+      "loss": 0.346,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9207828518173344,
+      "grad_norm": 0.24402898323221456,
+      "learning_rate": 1.9986192613047983e-05,
+      "loss": 0.3466,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9217148182665424,
+      "grad_norm": 0.2846129225662758,
+      "learning_rate": 1.9968933379357957e-05,
+      "loss": 0.3453,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9226467847157502,
+      "grad_norm": 0.2536705902414538,
+      "learning_rate": 1.9951674145667935e-05,
+      "loss": 0.3333,
+      "step": 2063
+    },
+    {
+      "epoch": 1.923578751164958,
+      "grad_norm": 0.25435116366042704,
+      "learning_rate": 1.993441491197791e-05,
+      "loss": 0.3448,
+      "step": 2064
+    },
+    {
+      "epoch": 1.924510717614166,
+      "grad_norm": 0.2898308604229828,
+      "learning_rate": 1.9917155678287887e-05,
+      "loss": 0.3303,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9254426840633738,
+      "grad_norm": 0.2705477736312823,
+      "learning_rate": 1.989989644459786e-05,
+      "loss": 0.3416,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9263746505125816,
+      "grad_norm": 0.34128111456994353,
+      "learning_rate": 1.9882637210907836e-05,
+      "loss": 0.3471,
+      "step": 2067
+    },
+    {
+      "epoch": 1.9273066169617894,
+      "grad_norm": 0.29211313959434804,
+      "learning_rate": 1.986537797721781e-05,
+      "loss": 0.3743,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9282385834109972,
+      "grad_norm": 0.24990484157151766,
+      "learning_rate": 1.9848118743527788e-05,
+      "loss": 0.328,
+      "step": 2069
+    },
+    {
+      "epoch": 1.929170549860205,
+      "grad_norm": 0.2581769330720566,
+      "learning_rate": 1.9830859509837765e-05,
+      "loss": 0.3474,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9301025163094128,
+      "grad_norm": 0.28594441457871755,
+      "learning_rate": 1.981360027614774e-05,
+      "loss": 0.3665,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 0.2746093729394535,
+      "learning_rate": 1.9796341042457717e-05,
+      "loss": 0.339,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9319664492078286,
+      "grad_norm": 0.2258568004927403,
+      "learning_rate": 1.977908180876769e-05,
+      "loss": 0.339,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9328984156570364,
+      "grad_norm": 0.2709066737208765,
+      "learning_rate": 1.976182257507767e-05,
+      "loss": 0.3496,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9338303821062441,
+      "grad_norm": 0.25903920576539485,
+      "learning_rate": 1.9744563341387643e-05,
+      "loss": 0.3509,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9347623485554521,
+      "grad_norm": 0.2696661170760633,
+      "learning_rate": 1.972730410769762e-05,
+      "loss": 0.3418,
+      "step": 2076
+    },
+    {
+      "epoch": 1.93569431500466,
+      "grad_norm": 0.27478146578685525,
+      "learning_rate": 1.9710044874007595e-05,
+      "loss": 0.3443,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9366262814538677,
+      "grad_norm": 0.31314622407178716,
+      "learning_rate": 1.969278564031757e-05,
+      "loss": 0.3431,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9375582479030755,
+      "grad_norm": 0.2682701873741881,
+      "learning_rate": 1.9675526406627547e-05,
+      "loss": 0.3509,
+      "step": 2079
+    },
+    {
+      "epoch": 1.9384902143522833,
+      "grad_norm": 0.2790402233093648,
+      "learning_rate": 1.9658267172937522e-05,
+      "loss": 0.3403,
+      "step": 2080
+    },
+    {
+      "epoch": 1.939422180801491,
+      "grad_norm": 0.2670622487572664,
+      "learning_rate": 1.96410079392475e-05,
+      "loss": 0.3185,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9403541472506989,
+      "grad_norm": 0.3170145593287183,
+      "learning_rate": 1.9623748705557474e-05,
+      "loss": 0.3439,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9412861136999067,
+      "grad_norm": 0.29200913385393595,
+      "learning_rate": 1.960648947186745e-05,
+      "loss": 0.3375,
+      "step": 2083
+    },
+    {
+      "epoch": 1.9422180801491147,
+      "grad_norm": 0.28712281272333634,
+      "learning_rate": 1.9589230238177426e-05,
+      "loss": 0.3489,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9431500465983225,
+      "grad_norm": 0.37365289989121936,
+      "learning_rate": 1.9571971004487403e-05,
+      "loss": 0.3703,
+      "step": 2085
+    },
+    {
+      "epoch": 1.9440820130475303,
+      "grad_norm": 0.32801101848194986,
+      "learning_rate": 1.9554711770797378e-05,
+      "loss": 0.3548,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9450139794967383,
+      "grad_norm": 0.3166450139011297,
+      "learning_rate": 1.9537452537107352e-05,
+      "loss": 0.3397,
+      "step": 2087
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 0.32469825933287777,
+      "learning_rate": 1.9520193303417326e-05,
+      "loss": 0.3464,
+      "step": 2088
+    },
+    {
+      "epoch": 1.9468779123951538,
+      "grad_norm": 0.35470474536258106,
+      "learning_rate": 1.9502934069727304e-05,
+      "loss": 0.346,
+      "step": 2089
+    },
+    {
+      "epoch": 1.9478098788443616,
+      "grad_norm": 0.29237553152381757,
+      "learning_rate": 1.948567483603728e-05,
+      "loss": 0.3607,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9487418452935694,
+      "grad_norm": 0.2631078881610645,
+      "learning_rate": 1.9468415602347256e-05,
+      "loss": 0.3277,
+      "step": 2091
+    },
+    {
+      "epoch": 1.9496738117427772,
+      "grad_norm": 0.3334079179167799,
+      "learning_rate": 1.9451156368657234e-05,
+      "loss": 0.3277,
+      "step": 2092
+    },
+    {
+      "epoch": 1.950605778191985,
+      "grad_norm": 0.3639337271699387,
+      "learning_rate": 1.9433897134967208e-05,
+      "loss": 0.3457,
+      "step": 2093
+    },
+    {
+      "epoch": 1.9515377446411928,
+      "grad_norm": 0.269680865357343,
+      "learning_rate": 1.9416637901277185e-05,
+      "loss": 0.3155,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9524697110904008,
+      "grad_norm": 0.32342181287509914,
+      "learning_rate": 1.939937866758716e-05,
+      "loss": 0.3436,
+      "step": 2095
+    },
+    {
+      "epoch": 1.9534016775396086,
+      "grad_norm": 0.2931661834781392,
+      "learning_rate": 1.9382119433897137e-05,
+      "loss": 0.3474,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9543336439888164,
+      "grad_norm": 0.28241913258960133,
+      "learning_rate": 1.9364860200207112e-05,
+      "loss": 0.3662,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9552656104380244,
+      "grad_norm": 0.27818269980713245,
+      "learning_rate": 1.9347600966517086e-05,
+      "loss": 0.3413,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9561975768872322,
+      "grad_norm": 0.23517815255244687,
+      "learning_rate": 1.9330341732827064e-05,
+      "loss": 0.3519,
+      "step": 2099
+    },
+    {
+      "epoch": 1.95712954333644,
+      "grad_norm": 0.28831283637015487,
+      "learning_rate": 1.9313082499137038e-05,
+      "loss": 0.3387,
+      "step": 2100
+    },
+    {
+      "epoch": 1.9580615097856477,
+      "grad_norm": 0.2566667070407751,
+      "learning_rate": 1.9295823265447016e-05,
+      "loss": 0.3301,
+      "step": 2101
+    },
+    {
+      "epoch": 1.9589934762348555,
+      "grad_norm": 0.2979428046684699,
+      "learning_rate": 1.927856403175699e-05,
+      "loss": 0.3351,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9599254426840633,
+      "grad_norm": 0.25474023564028625,
+      "learning_rate": 1.9261304798066968e-05,
+      "loss": 0.3266,
+      "step": 2103
+    },
+    {
+      "epoch": 1.9608574091332711,
+      "grad_norm": 0.2822023557274094,
+      "learning_rate": 1.9244045564376942e-05,
+      "loss": 0.3782,
+      "step": 2104
+    },
+    {
+      "epoch": 1.961789375582479,
+      "grad_norm": 0.29427832503575974,
+      "learning_rate": 1.922678633068692e-05,
+      "loss": 0.347,
+      "step": 2105
+    },
+    {
+      "epoch": 1.962721342031687,
+      "grad_norm": 0.30226932107679966,
+      "learning_rate": 1.9209527096996894e-05,
+      "loss": 0.3471,
+      "step": 2106
+    },
+    {
+      "epoch": 1.9636533084808947,
+      "grad_norm": 0.2404240355948779,
+      "learning_rate": 1.919226786330687e-05,
+      "loss": 0.3353,
+      "step": 2107
+    },
+    {
+      "epoch": 1.9645852749301025,
+      "grad_norm": 0.37068931722354953,
+      "learning_rate": 1.9175008629616846e-05,
+      "loss": 0.3722,
+      "step": 2108
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 0.2782581986178109,
+      "learning_rate": 1.915774939592682e-05,
+      "loss": 0.3417,
+      "step": 2109
+    },
+    {
+      "epoch": 1.9664492078285183,
+      "grad_norm": 0.29270554677093513,
+      "learning_rate": 1.9140490162236798e-05,
+      "loss": 0.3492,
+      "step": 2110
+    },
+    {
+      "epoch": 1.967381174277726,
+      "grad_norm": 0.28844656833074084,
+      "learning_rate": 1.9123230928546772e-05,
+      "loss": 0.371,
+      "step": 2111
+    },
+    {
+      "epoch": 1.9683131407269339,
+      "grad_norm": 0.288043369149429,
+      "learning_rate": 1.910597169485675e-05,
+      "loss": 0.3571,
+      "step": 2112
+    },
+    {
+      "epoch": 1.9692451071761417,
+      "grad_norm": 0.2768816921615798,
+      "learning_rate": 1.9088712461166724e-05,
+      "loss": 0.3463,
+      "step": 2113
+    },
+    {
+      "epoch": 1.9701770736253494,
+      "grad_norm": 0.25462458546673444,
+      "learning_rate": 1.9071453227476702e-05,
+      "loss": 0.3613,
+      "step": 2114
+    },
+    {
+      "epoch": 1.9711090400745572,
+      "grad_norm": 0.2722940384959956,
+      "learning_rate": 1.9054193993786676e-05,
+      "loss": 0.346,
+      "step": 2115
+    },
+    {
+      "epoch": 1.972041006523765,
+      "grad_norm": 0.2782200654589054,
+      "learning_rate": 1.9036934760096654e-05,
+      "loss": 0.3471,
+      "step": 2116
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 0.24689772747724417,
+      "learning_rate": 1.9019675526406628e-05,
+      "loss": 0.3532,
+      "step": 2117
+    },
+    {
+      "epoch": 1.9739049394221808,
+      "grad_norm": 0.32917936977224205,
+      "learning_rate": 1.9002416292716606e-05,
+      "loss": 0.3579,
+      "step": 2118
+    },
+    {
+      "epoch": 1.9748369058713886,
+      "grad_norm": 0.2739902655336654,
+      "learning_rate": 1.898515705902658e-05,
+      "loss": 0.3404,
+      "step": 2119
+    },
+    {
+      "epoch": 1.9757688723205966,
+      "grad_norm": 0.25487287708976514,
+      "learning_rate": 1.8967897825336554e-05,
+      "loss": 0.3229,
+      "step": 2120
+    },
+    {
+      "epoch": 1.9767008387698044,
+      "grad_norm": 0.2623480792135525,
+      "learning_rate": 1.8950638591646532e-05,
+      "loss": 0.3512,
+      "step": 2121
+    },
+    {
+      "epoch": 1.9776328052190122,
+      "grad_norm": 0.2771323672013457,
+      "learning_rate": 1.8933379357956506e-05,
+      "loss": 0.3475,
+      "step": 2122
+    },
+    {
+      "epoch": 1.97856477166822,
+      "grad_norm": 0.2902698895670476,
+      "learning_rate": 1.8916120124266484e-05,
+      "loss": 0.3484,
+      "step": 2123
+    },
+    {
+      "epoch": 1.9794967381174278,
+      "grad_norm": 0.26050617959137956,
+      "learning_rate": 1.8898860890576458e-05,
+      "loss": 0.3444,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9804287045666356,
+      "grad_norm": 0.24929504854944878,
+      "learning_rate": 1.8881601656886436e-05,
+      "loss": 0.3172,
+      "step": 2125
+    },
+    {
+      "epoch": 1.9813606710158433,
+      "grad_norm": 0.9220452323076347,
+      "learning_rate": 1.886434242319641e-05,
+      "loss": 0.3354,
+      "step": 2126
+    },
+    {
+      "epoch": 1.9822926374650511,
+      "grad_norm": 0.3502598204101326,
+      "learning_rate": 1.8847083189506388e-05,
+      "loss": 0.3514,
+      "step": 2127
+    },
+    {
+      "epoch": 1.983224603914259,
+      "grad_norm": 0.2963285364063072,
+      "learning_rate": 1.8829823955816362e-05,
+      "loss": 0.3526,
+      "step": 2128
+    },
+    {
+      "epoch": 1.984156570363467,
+      "grad_norm": 0.26473120563171737,
+      "learning_rate": 1.881256472212634e-05,
+      "loss": 0.3546,
+      "step": 2129
+    },
+    {
+      "epoch": 1.9850885368126747,
+      "grad_norm": 0.31022859287368976,
+      "learning_rate": 1.8795305488436314e-05,
+      "loss": 0.3415,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9860205032618827,
+      "grad_norm": 0.2960254482106945,
+      "learning_rate": 1.877804625474629e-05,
+      "loss": 0.3297,
+      "step": 2131
+    },
+    {
+      "epoch": 1.9869524697110905,
+      "grad_norm": 0.2732027561737865,
+      "learning_rate": 1.8760787021056266e-05,
+      "loss": 0.3419,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9878844361602983,
+      "grad_norm": 0.2681274369190567,
+      "learning_rate": 1.874352778736624e-05,
+      "loss": 0.3211,
+      "step": 2133
+    },
+    {
+      "epoch": 1.988816402609506,
+      "grad_norm": 0.3288178749563501,
+      "learning_rate": 1.8726268553676218e-05,
+      "loss": 0.3343,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9897483690587139,
+      "grad_norm": 0.3083996462770199,
+      "learning_rate": 1.8709009319986192e-05,
+      "loss": 0.3449,
+      "step": 2135
+    },
+    {
+      "epoch": 1.9906803355079217,
+      "grad_norm": 0.31431837993745637,
+      "learning_rate": 1.869175008629617e-05,
+      "loss": 0.3259,
+      "step": 2136
+    },
+    {
+      "epoch": 1.9916123019571295,
+      "grad_norm": 0.3017338603829043,
+      "learning_rate": 1.8674490852606144e-05,
+      "loss": 0.3568,
+      "step": 2137
+    },
+    {
+      "epoch": 1.9925442684063372,
+      "grad_norm": 0.2610493252939319,
+      "learning_rate": 1.8657231618916122e-05,
+      "loss": 0.3296,
+      "step": 2138
+    },
+    {
+      "epoch": 1.993476234855545,
+      "grad_norm": 0.27107304212672295,
+      "learning_rate": 1.86399723852261e-05,
+      "loss": 0.3466,
+      "step": 2139
+    },
+    {
+      "epoch": 1.994408201304753,
+      "grad_norm": 0.28127442735100555,
+      "learning_rate": 1.8622713151536074e-05,
+      "loss": 0.3489,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9953401677539608,
+      "grad_norm": 0.2504180576527886,
+      "learning_rate": 1.8605453917846048e-05,
+      "loss": 0.3427,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9962721342031688,
+      "grad_norm": 0.2835884320744997,
+      "learning_rate": 1.8588194684156023e-05,
+      "loss": 0.3333,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9972041006523766,
+      "grad_norm": 0.2853106538950936,
+      "learning_rate": 1.8570935450466e-05,
+      "loss": 0.3293,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9981360671015844,
+      "grad_norm": 0.27305216481866995,
+      "learning_rate": 1.8553676216775975e-05,
+      "loss": 0.3676,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9990680335507922,
+      "grad_norm": 0.25808595343654417,
+      "learning_rate": 1.8536416983085952e-05,
+      "loss": 0.3578,
+      "step": 2145
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.28913351561433964,
+      "learning_rate": 1.8519157749395927e-05,
+      "loss": 0.3177,
+      "step": 2146
+    },
+    {
+      "epoch": 2.000931966449208,
+      "grad_norm": 0.3311010316667172,
+      "learning_rate": 1.8501898515705904e-05,
+      "loss": 0.2777,
+      "step": 2147
+    },
+    {
+      "epoch": 2.0018639328984156,
+      "grad_norm": 0.2619828809890582,
+      "learning_rate": 1.848463928201588e-05,
+      "loss": 0.2752,
+      "step": 2148
+    },
+    {
+      "epoch": 2.0027958993476234,
+      "grad_norm": 0.26145547426041543,
+      "learning_rate": 1.8467380048325856e-05,
+      "loss": 0.2749,
+      "step": 2149
+    },
+    {
+      "epoch": 2.003727865796831,
+      "grad_norm": 0.2814247242240938,
+      "learning_rate": 1.8450120814635834e-05,
+      "loss": 0.2691,
+      "step": 2150
+    },
+    {
+      "epoch": 2.004659832246039,
+      "grad_norm": 0.29212401227423784,
+      "learning_rate": 1.8432861580945808e-05,
+      "loss": 0.2913,
+      "step": 2151
+    },
+    {
+      "epoch": 2.005591798695247,
+      "grad_norm": 0.25814085902984024,
+      "learning_rate": 1.8415602347255782e-05,
+      "loss": 0.2755,
+      "step": 2152
+    },
+    {
+      "epoch": 2.006523765144455,
+      "grad_norm": 0.3061990072141899,
+      "learning_rate": 1.8398343113565757e-05,
+      "loss": 0.2935,
+      "step": 2153
+    },
+    {
+      "epoch": 2.0074557315936628,
+      "grad_norm": 0.2881093951819176,
+      "learning_rate": 1.8381083879875734e-05,
+      "loss": 0.3001,
+      "step": 2154
+    },
+    {
+      "epoch": 2.0083876980428705,
+      "grad_norm": 0.3160753671658845,
+      "learning_rate": 1.836382464618571e-05,
+      "loss": 0.2674,
+      "step": 2155
+    },
+    {
+      "epoch": 2.0093196644920783,
+      "grad_norm": 0.23968943061589523,
+      "learning_rate": 1.8346565412495686e-05,
+      "loss": 0.2921,
+      "step": 2156
+    },
+    {
+      "epoch": 2.010251630941286,
+      "grad_norm": 0.2522162267142082,
+      "learning_rate": 1.832930617880566e-05,
+      "loss": 0.2886,
+      "step": 2157
+    },
+    {
+      "epoch": 2.011183597390494,
+      "grad_norm": 0.2648808944688652,
+      "learning_rate": 1.831204694511564e-05,
+      "loss": 0.2691,
+      "step": 2158
+    },
+    {
+      "epoch": 2.0121155638397017,
+      "grad_norm": 0.2509789096297247,
+      "learning_rate": 1.8294787711425613e-05,
+      "loss": 0.2703,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0130475302889095,
+      "grad_norm": 0.25099157478647,
+      "learning_rate": 1.827752847773559e-05,
+      "loss": 0.2736,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0139794967381173,
+      "grad_norm": 0.259634245448141,
+      "learning_rate": 1.8260269244045568e-05,
+      "loss": 0.299,
+      "step": 2161
+    },
+    {
+      "epoch": 2.014911463187325,
+      "grad_norm": 0.2422953099032814,
+      "learning_rate": 1.8243010010355542e-05,
+      "loss": 0.2651,
+      "step": 2162
+    },
+    {
+      "epoch": 2.0158434296365333,
+      "grad_norm": 0.24727658054790244,
+      "learning_rate": 1.8225750776665517e-05,
+      "loss": 0.2719,
+      "step": 2163
+    },
+    {
+      "epoch": 2.016775396085741,
+      "grad_norm": 0.2572529157422517,
+      "learning_rate": 1.820849154297549e-05,
+      "loss": 0.2791,
+      "step": 2164
+    },
+    {
+      "epoch": 2.017707362534949,
+      "grad_norm": 0.26890891579373916,
+      "learning_rate": 1.819123230928547e-05,
+      "loss": 0.2932,
+      "step": 2165
+    },
+    {
+      "epoch": 2.0186393289841567,
+      "grad_norm": 0.24282684596950346,
+      "learning_rate": 1.8173973075595443e-05,
+      "loss": 0.2704,
+      "step": 2166
+    },
+    {
+      "epoch": 2.0195712954333644,
+      "grad_norm": 0.2518918032814643,
+      "learning_rate": 1.815671384190542e-05,
+      "loss": 0.2993,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0205032618825722,
+      "grad_norm": 0.23263640728749643,
+      "learning_rate": 1.8139454608215395e-05,
+      "loss": 0.2877,
+      "step": 2168
+    },
+    {
+      "epoch": 2.02143522833178,
+      "grad_norm": 0.249673137686599,
+      "learning_rate": 1.8122195374525372e-05,
+      "loss": 0.272,
+      "step": 2169
+    },
+    {
+      "epoch": 2.022367194780988,
+      "grad_norm": 0.24965331782938574,
+      "learning_rate": 1.810493614083535e-05,
+      "loss": 0.2727,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0232991612301956,
+      "grad_norm": 0.23987357692229488,
+      "learning_rate": 1.8087676907145324e-05,
+      "loss": 0.2749,
+      "step": 2171
+    },
+    {
+      "epoch": 2.0242311276794034,
+      "grad_norm": 0.24782779329492546,
+      "learning_rate": 1.8070417673455302e-05,
+      "loss": 0.2821,
+      "step": 2172
+    },
+    {
+      "epoch": 2.025163094128611,
+      "grad_norm": 0.27740260326682137,
+      "learning_rate": 1.8053158439765276e-05,
+      "loss": 0.2847,
+      "step": 2173
+    },
+    {
+      "epoch": 2.0260950605778194,
+      "grad_norm": 0.2545515915352839,
+      "learning_rate": 1.803589920607525e-05,
+      "loss": 0.2643,
+      "step": 2174
+    },
+    {
+      "epoch": 2.027027027027027,
+      "grad_norm": 0.2429796289236236,
+      "learning_rate": 1.8018639972385225e-05,
+      "loss": 0.272,
+      "step": 2175
+    },
+    {
+      "epoch": 2.027958993476235,
+      "grad_norm": 0.23630834212285415,
+      "learning_rate": 1.8001380738695203e-05,
+      "loss": 0.2691,
+      "step": 2176
+    },
+    {
+      "epoch": 2.0288909599254428,
+      "grad_norm": 0.27312788800701276,
+      "learning_rate": 1.7984121505005177e-05,
+      "loss": 0.2815,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0298229263746506,
+      "grad_norm": 0.2492156625945811,
+      "learning_rate": 1.7966862271315155e-05,
+      "loss": 0.281,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0307548928238583,
+      "grad_norm": 0.250161413305125,
+      "learning_rate": 1.794960303762513e-05,
+      "loss": 0.2867,
+      "step": 2179
+    },
+    {
+      "epoch": 2.031686859273066,
+      "grad_norm": 0.2531096447179367,
+      "learning_rate": 1.7932343803935107e-05,
+      "loss": 0.2814,
+      "step": 2180
+    },
+    {
+      "epoch": 2.032618825722274,
+      "grad_norm": 0.261118055208972,
+      "learning_rate": 1.7915084570245084e-05,
+      "loss": 0.2847,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0335507921714817,
+      "grad_norm": 0.26478005646823544,
+      "learning_rate": 1.789782533655506e-05,
+      "loss": 0.2808,
+      "step": 2182
+    },
+    {
+      "epoch": 2.0344827586206895,
+      "grad_norm": 0.2596323899736538,
+      "learning_rate": 1.7880566102865033e-05,
+      "loss": 0.2824,
+      "step": 2183
+    },
+    {
+      "epoch": 2.0354147250698973,
+      "grad_norm": 0.2285881976432698,
+      "learning_rate": 1.786330686917501e-05,
+      "loss": 0.2748,
+      "step": 2184
+    },
+    {
+      "epoch": 2.0363466915191055,
+      "grad_norm": 0.24125406949971392,
+      "learning_rate": 1.7846047635484985e-05,
+      "loss": 0.2598,
+      "step": 2185
+    },
+    {
+      "epoch": 2.0372786579683133,
+      "grad_norm": 0.2552660193251454,
+      "learning_rate": 1.782878840179496e-05,
+      "loss": 0.2679,
+      "step": 2186
+    },
+    {
+      "epoch": 2.038210624417521,
+      "grad_norm": 0.2601764071637898,
+      "learning_rate": 1.7811529168104937e-05,
+      "loss": 0.2819,
+      "step": 2187
+    },
+    {
+      "epoch": 2.039142590866729,
+      "grad_norm": 0.24835021832381327,
+      "learning_rate": 1.779426993441491e-05,
+      "loss": 0.2852,
+      "step": 2188
+    },
+    {
+      "epoch": 2.0400745573159367,
+      "grad_norm": 0.21842024890978,
+      "learning_rate": 1.777701070072489e-05,
+      "loss": 0.266,
+      "step": 2189
+    },
+    {
+      "epoch": 2.0410065237651445,
+      "grad_norm": 0.26341365724762994,
+      "learning_rate": 1.7759751467034866e-05,
+      "loss": 0.2804,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0419384902143523,
+      "grad_norm": 0.25099142927293566,
+      "learning_rate": 1.774249223334484e-05,
+      "loss": 0.291,
+      "step": 2191
+    },
+    {
+      "epoch": 2.04287045666356,
+      "grad_norm": 0.24223903501671315,
+      "learning_rate": 1.772523299965482e-05,
+      "loss": 0.2823,
+      "step": 2192
+    },
+    {
+      "epoch": 2.043802423112768,
+      "grad_norm": 0.2440681821010892,
+      "learning_rate": 1.7707973765964793e-05,
+      "loss": 0.2756,
+      "step": 2193
+    },
+    {
+      "epoch": 2.0447343895619756,
+      "grad_norm": 0.27287453634968506,
+      "learning_rate": 1.7690714532274767e-05,
+      "loss": 0.2827,
+      "step": 2194
+    },
+    {
+      "epoch": 2.0456663560111834,
+      "grad_norm": 0.2873826575069445,
+      "learning_rate": 1.767345529858474e-05,
+      "loss": 0.3,
+      "step": 2195
+    },
+    {
+      "epoch": 2.0465983224603916,
+      "grad_norm": 0.2875584074077825,
+      "learning_rate": 1.765619606489472e-05,
+      "loss": 0.2922,
+      "step": 2196
+    },
+    {
+      "epoch": 2.0475302889095994,
+      "grad_norm": 0.22444037737052136,
+      "learning_rate": 1.7638936831204693e-05,
+      "loss": 0.2649,
+      "step": 2197
+    },
+    {
+      "epoch": 2.048462255358807,
+      "grad_norm": 0.2973738541035133,
+      "learning_rate": 1.762167759751467e-05,
+      "loss": 0.2985,
+      "step": 2198
+    },
+    {
+      "epoch": 2.049394221808015,
+      "grad_norm": 0.27286566243682103,
+      "learning_rate": 1.7604418363824645e-05,
+      "loss": 0.2925,
+      "step": 2199
+    },
+    {
+      "epoch": 2.050326188257223,
+      "grad_norm": 0.23761971818072736,
+      "learning_rate": 1.7587159130134623e-05,
+      "loss": 0.2831,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0512581547064306,
+      "grad_norm": 0.2536090204305789,
+      "learning_rate": 1.75698998964446e-05,
+      "loss": 0.2695,
+      "step": 2201
+    },
+    {
+      "epoch": 2.0521901211556384,
+      "grad_norm": 0.29503971696992537,
+      "learning_rate": 1.7552640662754575e-05,
+      "loss": 0.2876,
+      "step": 2202
+    },
+    {
+      "epoch": 2.053122087604846,
+      "grad_norm": 0.23571943386193128,
+      "learning_rate": 1.7535381429064553e-05,
+      "loss": 0.2767,
+      "step": 2203
+    },
+    {
+      "epoch": 2.054054054054054,
+      "grad_norm": 0.2548884123443062,
+      "learning_rate": 1.7518122195374527e-05,
+      "loss": 0.2614,
+      "step": 2204
+    },
+    {
+      "epoch": 2.0549860205032617,
+      "grad_norm": 0.24834426480016988,
+      "learning_rate": 1.75008629616845e-05,
+      "loss": 0.2745,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0559179869524695,
+      "grad_norm": 0.27625261448833494,
+      "learning_rate": 1.7483603727994475e-05,
+      "loss": 0.2887,
+      "step": 2206
+    },
+    {
+      "epoch": 2.0568499534016778,
+      "grad_norm": 0.23029696844100137,
+      "learning_rate": 1.7466344494304453e-05,
+      "loss": 0.2788,
+      "step": 2207
+    },
+    {
+      "epoch": 2.0577819198508855,
+      "grad_norm": 0.2230348620972378,
+      "learning_rate": 1.7449085260614427e-05,
+      "loss": 0.2777,
+      "step": 2208
+    },
+    {
+      "epoch": 2.0587138863000933,
+      "grad_norm": 0.263019227198961,
+      "learning_rate": 1.7431826026924405e-05,
+      "loss": 0.2737,
+      "step": 2209
+    },
+    {
+      "epoch": 2.059645852749301,
+      "grad_norm": 0.22992364604413906,
+      "learning_rate": 1.7414566793234383e-05,
+      "loss": 0.265,
+      "step": 2210
+    },
+    {
+      "epoch": 2.060577819198509,
+      "grad_norm": 0.2442667445233449,
+      "learning_rate": 1.7397307559544357e-05,
+      "loss": 0.2666,
+      "step": 2211
+    },
+    {
+      "epoch": 2.0615097856477167,
+      "grad_norm": 23.389771494278094,
+      "learning_rate": 1.7380048325854335e-05,
+      "loss": 0.3338,
+      "step": 2212
+    },
+    {
+      "epoch": 2.0624417520969245,
+      "grad_norm": 0.2907288053173359,
+      "learning_rate": 1.736278909216431e-05,
+      "loss": 0.2889,
+      "step": 2213
+    },
+    {
+      "epoch": 2.0633737185461323,
+      "grad_norm": 0.2528611645368693,
+      "learning_rate": 1.7345529858474287e-05,
+      "loss": 0.2669,
+      "step": 2214
+    },
+    {
+      "epoch": 2.06430568499534,
+      "grad_norm": 0.2453165357902039,
+      "learning_rate": 1.732827062478426e-05,
+      "loss": 0.2933,
+      "step": 2215
+    },
+    {
+      "epoch": 2.065237651444548,
+      "grad_norm": 0.23284925521360936,
+      "learning_rate": 1.7311011391094235e-05,
+      "loss": 0.2542,
+      "step": 2216
+    },
+    {
+      "epoch": 2.0661696178937556,
+      "grad_norm": 0.2613164096332141,
+      "learning_rate": 1.729375215740421e-05,
+      "loss": 0.2888,
+      "step": 2217
+    },
+    {
+      "epoch": 2.0671015843429634,
+      "grad_norm": 0.23706807153803972,
+      "learning_rate": 1.7276492923714187e-05,
+      "loss": 0.264,
+      "step": 2218
+    },
+    {
+      "epoch": 2.0680335507921717,
+      "grad_norm": 0.22335883647936305,
+      "learning_rate": 1.725923369002416e-05,
+      "loss": 0.2828,
+      "step": 2219
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 0.2536792394948087,
+      "learning_rate": 1.724197445633414e-05,
+      "loss": 0.275,
+      "step": 2220
+    },
+    {
+      "epoch": 2.0698974836905872,
+      "grad_norm": 0.2722798854361317,
+      "learning_rate": 1.7224715222644117e-05,
+      "loss": 0.2738,
+      "step": 2221
+    },
+    {
+      "epoch": 2.070829450139795,
+      "grad_norm": 0.21433371739852683,
+      "learning_rate": 1.720745598895409e-05,
+      "loss": 0.2733,
+      "step": 2222
+    },
+    {
+      "epoch": 2.071761416589003,
+      "grad_norm": 0.27536839497637333,
+      "learning_rate": 1.719019675526407e-05,
+      "loss": 0.2752,
+      "step": 2223
+    },
+    {
+      "epoch": 2.0726933830382106,
+      "grad_norm": 0.24731954565119865,
+      "learning_rate": 1.7172937521574043e-05,
+      "loss": 0.2692,
+      "step": 2224
+    },
+    {
+      "epoch": 2.0736253494874184,
+      "grad_norm": 0.2189265818838678,
+      "learning_rate": 1.715567828788402e-05,
+      "loss": 0.2665,
+      "step": 2225
+    },
+    {
+      "epoch": 2.074557315936626,
+      "grad_norm": 0.22712565423038547,
+      "learning_rate": 1.7138419054193995e-05,
+      "loss": 0.2621,
+      "step": 2226
+    },
+    {
+      "epoch": 2.075489282385834,
+      "grad_norm": 0.24210035664006002,
+      "learning_rate": 1.712115982050397e-05,
+      "loss": 0.2882,
+      "step": 2227
+    },
+    {
+      "epoch": 2.0764212488350418,
+      "grad_norm": 0.23434712989064307,
+      "learning_rate": 1.7103900586813944e-05,
+      "loss": 0.2684,
+      "step": 2228
+    },
+    {
+      "epoch": 2.0773532152842495,
+      "grad_norm": 0.2400226039463619,
+      "learning_rate": 1.708664135312392e-05,
+      "loss": 0.2754,
+      "step": 2229
+    },
+    {
+      "epoch": 2.0782851817334578,
+      "grad_norm": 0.22325257264033185,
+      "learning_rate": 1.70693821194339e-05,
+      "loss": 0.266,
+      "step": 2230
+    },
+    {
+      "epoch": 2.0792171481826656,
+      "grad_norm": 0.25802344091311974,
+      "learning_rate": 1.7052122885743873e-05,
+      "loss": 0.2926,
+      "step": 2231
+    },
+    {
+      "epoch": 2.0801491146318734,
+      "grad_norm": 0.22842501972702836,
+      "learning_rate": 1.703486365205385e-05,
+      "loss": 0.2645,
+      "step": 2232
+    },
+    {
+      "epoch": 2.081081081081081,
+      "grad_norm": 0.21682614908579764,
+      "learning_rate": 1.7017604418363825e-05,
+      "loss": 0.2637,
+      "step": 2233
+    },
+    {
+      "epoch": 2.082013047530289,
+      "grad_norm": 0.23384284151515589,
+      "learning_rate": 1.7000345184673803e-05,
+      "loss": 0.271,
+      "step": 2234
+    },
+    {
+      "epoch": 2.0829450139794967,
+      "grad_norm": 0.5849585932814216,
+      "learning_rate": 1.6983085950983777e-05,
+      "loss": 0.2699,
+      "step": 2235
+    },
+    {
+      "epoch": 2.0838769804287045,
+      "grad_norm": 0.23637629744504318,
+      "learning_rate": 1.6965826717293755e-05,
+      "loss": 0.2833,
+      "step": 2236
+    },
+    {
+      "epoch": 2.0848089468779123,
+      "grad_norm": 0.239898833558503,
+      "learning_rate": 1.694856748360373e-05,
+      "loss": 0.2982,
+      "step": 2237
+    },
+    {
+      "epoch": 2.08574091332712,
+      "grad_norm": 0.2501561219064165,
+      "learning_rate": 1.6931308249913704e-05,
+      "loss": 0.2862,
+      "step": 2238
+    },
+    {
+      "epoch": 2.086672879776328,
+      "grad_norm": 0.3005553707846581,
+      "learning_rate": 1.6914049016223678e-05,
+      "loss": 0.2873,
+      "step": 2239
+    },
+    {
+      "epoch": 2.0876048462255357,
+      "grad_norm": 0.2358039124368298,
+      "learning_rate": 1.6896789782533655e-05,
+      "loss": 0.273,
+      "step": 2240
+    },
+    {
+      "epoch": 2.088536812674744,
+      "grad_norm": 0.24255915650354384,
+      "learning_rate": 1.6879530548843633e-05,
+      "loss": 0.2807,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0894687791239517,
+      "grad_norm": 0.227426598289662,
+      "learning_rate": 1.6862271315153607e-05,
+      "loss": 0.2855,
+      "step": 2242
+    },
+    {
+      "epoch": 2.0904007455731595,
+      "grad_norm": 0.22353498347483355,
+      "learning_rate": 1.6845012081463585e-05,
+      "loss": 0.2717,
+      "step": 2243
+    },
+    {
+      "epoch": 2.0913327120223673,
+      "grad_norm": 1.6853089333661637,
+      "learning_rate": 1.682775284777356e-05,
+      "loss": 0.2718,
+      "step": 2244
+    },
+    {
+      "epoch": 2.092264678471575,
+      "grad_norm": 0.2391021473084039,
+      "learning_rate": 1.6810493614083537e-05,
+      "loss": 0.2772,
+      "step": 2245
+    },
+    {
+      "epoch": 2.093196644920783,
+      "grad_norm": 0.278689113170789,
+      "learning_rate": 1.679323438039351e-05,
+      "loss": 0.2873,
+      "step": 2246
+    },
+    {
+      "epoch": 2.0941286113699906,
+      "grad_norm": 0.24501525314009717,
+      "learning_rate": 1.677597514670349e-05,
+      "loss": 0.2881,
+      "step": 2247
+    },
+    {
+      "epoch": 2.0950605778191984,
+      "grad_norm": 0.22037064720008048,
+      "learning_rate": 1.6758715913013463e-05,
+      "loss": 0.2666,
+      "step": 2248
+    },
+    {
+      "epoch": 2.095992544268406,
+      "grad_norm": 0.23805864582664504,
+      "learning_rate": 1.6741456679323438e-05,
+      "loss": 0.2744,
+      "step": 2249
+    },
+    {
+      "epoch": 2.096924510717614,
+      "grad_norm": 0.22988216701474978,
+      "learning_rate": 1.6724197445633415e-05,
+      "loss": 0.2871,
+      "step": 2250
+    },
+    {
+      "epoch": 2.0978564771668218,
+      "grad_norm": 0.25275347939807763,
+      "learning_rate": 1.670693821194339e-05,
+      "loss": 0.2724,
+      "step": 2251
+    },
+    {
+      "epoch": 2.09878844361603,
+      "grad_norm": 0.24706736124340234,
+      "learning_rate": 1.6689678978253367e-05,
+      "loss": 0.2829,
+      "step": 2252
+    },
+    {
+      "epoch": 2.099720410065238,
+      "grad_norm": 0.2599446006118566,
+      "learning_rate": 1.667241974456334e-05,
+      "loss": 0.2813,
+      "step": 2253
+    },
+    {
+      "epoch": 2.1006523765144456,
+      "grad_norm": 0.2336999908600345,
+      "learning_rate": 1.665516051087332e-05,
+      "loss": 0.2914,
+      "step": 2254
+    },
+    {
+      "epoch": 2.1015843429636534,
+      "grad_norm": 0.2619691852625129,
+      "learning_rate": 1.6637901277183294e-05,
+      "loss": 0.2921,
+      "step": 2255
+    },
+    {
+      "epoch": 2.102516309412861,
+      "grad_norm": 0.232401822019891,
+      "learning_rate": 1.662064204349327e-05,
+      "loss": 0.2885,
+      "step": 2256
+    },
+    {
+      "epoch": 2.103448275862069,
+      "grad_norm": 0.229206193651887,
+      "learning_rate": 1.6603382809803246e-05,
+      "loss": 0.2687,
+      "step": 2257
+    },
+    {
+      "epoch": 2.1043802423112767,
+      "grad_norm": 0.2311827874992207,
+      "learning_rate": 1.6586123576113223e-05,
+      "loss": 0.2861,
+      "step": 2258
+    },
+    {
+      "epoch": 2.1053122087604845,
+      "grad_norm": 0.25537745431022185,
+      "learning_rate": 1.6568864342423197e-05,
+      "loss": 0.2828,
+      "step": 2259
+    },
+    {
+      "epoch": 2.1062441752096923,
+      "grad_norm": 0.22043371393126518,
+      "learning_rate": 1.6551605108733172e-05,
+      "loss": 0.2623,
+      "step": 2260
+    },
+    {
+      "epoch": 2.1071761416589,
+      "grad_norm": 0.24018452086170747,
+      "learning_rate": 1.653434587504315e-05,
+      "loss": 0.2925,
+      "step": 2261
+    },
+    {
+      "epoch": 2.108108108108108,
+      "grad_norm": 0.38004048230226206,
+      "learning_rate": 1.6517086641353124e-05,
+      "loss": 0.2845,
+      "step": 2262
+    },
+    {
+      "epoch": 2.109040074557316,
+      "grad_norm": 0.2435947179902814,
+      "learning_rate": 1.64998274076631e-05,
+      "loss": 0.2749,
+      "step": 2263
+    },
+    {
+      "epoch": 2.109972041006524,
+      "grad_norm": 0.22572858601707002,
+      "learning_rate": 1.6482568173973076e-05,
+      "loss": 0.2739,
+      "step": 2264
+    },
+    {
+      "epoch": 2.1109040074557317,
+      "grad_norm": 0.24566324600376666,
+      "learning_rate": 1.6465308940283053e-05,
+      "loss": 0.2742,
+      "step": 2265
+    },
+    {
+      "epoch": 2.1118359739049395,
+      "grad_norm": 0.22742732806609464,
+      "learning_rate": 1.6448049706593028e-05,
+      "loss": 0.2765,
+      "step": 2266
+    },
+    {
+      "epoch": 2.1127679403541473,
+      "grad_norm": 0.23185431169881648,
+      "learning_rate": 1.6430790472903005e-05,
+      "loss": 0.2787,
+      "step": 2267
+    },
+    {
+      "epoch": 2.113699906803355,
+      "grad_norm": 0.25225750346594833,
+      "learning_rate": 1.641353123921298e-05,
+      "loss": 0.2786,
+      "step": 2268
+    },
+    {
+      "epoch": 2.114631873252563,
+      "grad_norm": 0.24406222479154302,
+      "learning_rate": 1.6396272005522957e-05,
+      "loss": 0.2828,
+      "step": 2269
+    },
+    {
+      "epoch": 2.1155638397017706,
+      "grad_norm": 0.239235266670125,
+      "learning_rate": 1.637901277183293e-05,
+      "loss": 0.3015,
+      "step": 2270
+    },
+    {
+      "epoch": 2.1164958061509784,
+      "grad_norm": 0.2320437759729044,
+      "learning_rate": 1.6361753538142906e-05,
+      "loss": 0.291,
+      "step": 2271
+    },
+    {
+      "epoch": 2.117427772600186,
+      "grad_norm": 0.23479659603300385,
+      "learning_rate": 1.6344494304452884e-05,
+      "loss": 0.2757,
+      "step": 2272
+    },
+    {
+      "epoch": 2.118359739049394,
+      "grad_norm": 3.761134688750207,
+      "learning_rate": 1.6327235070762858e-05,
+      "loss": 0.2773,
+      "step": 2273
+    },
+    {
+      "epoch": 2.1192917054986022,
+      "grad_norm": 0.25112358203992674,
+      "learning_rate": 1.6309975837072836e-05,
+      "loss": 0.2836,
+      "step": 2274
+    },
+    {
+      "epoch": 2.12022367194781,
+      "grad_norm": 0.23030427907821527,
+      "learning_rate": 1.629271660338281e-05,
+      "loss": 0.2555,
+      "step": 2275
+    },
+    {
+      "epoch": 2.121155638397018,
+      "grad_norm": 0.2337557216440671,
+      "learning_rate": 1.6275457369692788e-05,
+      "loss": 0.2745,
+      "step": 2276
+    },
+    {
+      "epoch": 2.1220876048462256,
+      "grad_norm": 0.2490404375515908,
+      "learning_rate": 1.6258198136002762e-05,
+      "loss": 0.2828,
+      "step": 2277
+    },
+    {
+      "epoch": 2.1230195712954334,
+      "grad_norm": 0.30932808695666003,
+      "learning_rate": 1.624093890231274e-05,
+      "loss": 0.2878,
+      "step": 2278
+    },
+    {
+      "epoch": 2.123951537744641,
+      "grad_norm": 0.23222881577906176,
+      "learning_rate": 1.6223679668622714e-05,
+      "loss": 0.2747,
+      "step": 2279
+    },
+    {
+      "epoch": 2.124883504193849,
+      "grad_norm": 0.275537639133195,
+      "learning_rate": 1.620642043493269e-05,
+      "loss": 0.2916,
+      "step": 2280
+    },
+    {
+      "epoch": 2.1258154706430568,
+      "grad_norm": 0.21182195857486266,
+      "learning_rate": 1.6189161201242666e-05,
+      "loss": 0.2607,
+      "step": 2281
+    },
+    {
+      "epoch": 2.1267474370922645,
+      "grad_norm": 0.24928855761102417,
+      "learning_rate": 1.617190196755264e-05,
+      "loss": 0.2733,
+      "step": 2282
+    },
+    {
+      "epoch": 2.1276794035414723,
+      "grad_norm": 0.6257377732057826,
+      "learning_rate": 1.6154642733862618e-05,
+      "loss": 0.2887,
+      "step": 2283
+    },
+    {
+      "epoch": 2.12861136999068,
+      "grad_norm": 3.396838240989663,
+      "learning_rate": 1.6137383500172592e-05,
+      "loss": 0.2879,
+      "step": 2284
+    },
+    {
+      "epoch": 2.1295433364398884,
+      "grad_norm": 0.25524715004810056,
+      "learning_rate": 1.612012426648257e-05,
+      "loss": 0.2652,
+      "step": 2285
+    },
+    {
+      "epoch": 2.130475302889096,
+      "grad_norm": 1.0892487716588874,
+      "learning_rate": 1.6102865032792544e-05,
+      "loss": 0.2775,
+      "step": 2286
+    },
+    {
+      "epoch": 2.131407269338304,
+      "grad_norm": 0.22473752921107232,
+      "learning_rate": 1.608560579910252e-05,
+      "loss": 0.2903,
+      "step": 2287
+    },
+    {
+      "epoch": 2.1323392357875117,
+      "grad_norm": 0.23923116058809998,
+      "learning_rate": 1.6068346565412496e-05,
+      "loss": 0.2734,
+      "step": 2288
+    },
+    {
+      "epoch": 2.1332712022367195,
+      "grad_norm": 0.2513713792730017,
+      "learning_rate": 1.6051087331722474e-05,
+      "loss": 0.271,
+      "step": 2289
+    },
+    {
+      "epoch": 2.1342031686859273,
+      "grad_norm": 0.22776398041874324,
+      "learning_rate": 1.6033828098032448e-05,
+      "loss": 0.2728,
+      "step": 2290
+    },
+    {
+      "epoch": 2.135135135135135,
+      "grad_norm": 0.21792769994612643,
+      "learning_rate": 1.6016568864342422e-05,
+      "loss": 0.2702,
+      "step": 2291
+    },
+    {
+      "epoch": 2.136067101584343,
+      "grad_norm": 0.24449518633968034,
+      "learning_rate": 1.59993096306524e-05,
+      "loss": 0.3021,
+      "step": 2292
+    },
+    {
+      "epoch": 2.1369990680335507,
+      "grad_norm": 0.22460682181118044,
+      "learning_rate": 1.5982050396962374e-05,
+      "loss": 0.2567,
+      "step": 2293
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 0.24096868117493012,
+      "learning_rate": 1.5964791163272352e-05,
+      "loss": 0.2809,
+      "step": 2294
+    },
+    {
+      "epoch": 2.1388630009319662,
+      "grad_norm": 0.24780836136042214,
+      "learning_rate": 1.5947531929582326e-05,
+      "loss": 0.2916,
+      "step": 2295
+    },
+    {
+      "epoch": 2.1397949673811745,
+      "grad_norm": 0.24958612981498365,
+      "learning_rate": 1.5930272695892304e-05,
+      "loss": 0.2877,
+      "step": 2296
+    },
+    {
+      "epoch": 2.1407269338303823,
+      "grad_norm": 0.2526472648413272,
+      "learning_rate": 1.5913013462202278e-05,
+      "loss": 0.2801,
+      "step": 2297
+    },
+    {
+      "epoch": 2.14165890027959,
+      "grad_norm": 0.25936334508470527,
+      "learning_rate": 1.5895754228512256e-05,
+      "loss": 0.2852,
+      "step": 2298
+    },
+    {
+      "epoch": 2.142590866728798,
+      "grad_norm": 0.22259633120013783,
+      "learning_rate": 1.587849499482223e-05,
+      "loss": 0.2605,
+      "step": 2299
+    },
+    {
+      "epoch": 2.1435228331780056,
+      "grad_norm": 0.24933624273476412,
+      "learning_rate": 1.5861235761132208e-05,
+      "loss": 0.2931,
+      "step": 2300
+    },
+    {
+      "epoch": 2.1444547996272134,
+      "grad_norm": 0.24515224742195124,
+      "learning_rate": 1.5843976527442182e-05,
+      "loss": 0.2745,
+      "step": 2301
+    },
+    {
+      "epoch": 2.145386766076421,
+      "grad_norm": 0.24324305651967607,
+      "learning_rate": 1.5826717293752156e-05,
+      "loss": 0.285,
+      "step": 2302
+    },
+    {
+      "epoch": 2.146318732525629,
+      "grad_norm": 0.22035974830702829,
+      "learning_rate": 1.5809458060062134e-05,
+      "loss": 0.2767,
+      "step": 2303
+    },
+    {
+      "epoch": 2.147250698974837,
+      "grad_norm": 0.20569516803292578,
+      "learning_rate": 1.5792198826372108e-05,
+      "loss": 0.2655,
+      "step": 2304
+    },
+    {
+      "epoch": 2.1481826654240446,
+      "grad_norm": 0.2588743368908312,
+      "learning_rate": 1.5774939592682086e-05,
+      "loss": 0.2655,
+      "step": 2305
+    },
+    {
+      "epoch": 2.1491146318732524,
+      "grad_norm": 0.23617291651156924,
+      "learning_rate": 1.575768035899206e-05,
+      "loss": 0.2632,
+      "step": 2306
+    },
+    {
+      "epoch": 2.1500465983224606,
+      "grad_norm": 0.2215123129352845,
+      "learning_rate": 1.5740421125302038e-05,
+      "loss": 0.2918,
+      "step": 2307
+    },
+    {
+      "epoch": 2.1509785647716684,
+      "grad_norm": 0.2195489090954217,
+      "learning_rate": 1.5723161891612012e-05,
+      "loss": 0.2886,
+      "step": 2308
+    },
+    {
+      "epoch": 2.151910531220876,
+      "grad_norm": 0.26573414466106643,
+      "learning_rate": 1.570590265792199e-05,
+      "loss": 0.2879,
+      "step": 2309
+    },
+    {
+      "epoch": 2.152842497670084,
+      "grad_norm": 0.2522816652593159,
+      "learning_rate": 1.5688643424231964e-05,
+      "loss": 0.262,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1537744641192917,
+      "grad_norm": 0.2376557511807734,
+      "learning_rate": 1.5671384190541942e-05,
+      "loss": 0.2864,
+      "step": 2311
+    },
+    {
+      "epoch": 2.1547064305684995,
+      "grad_norm": 0.26006136091216026,
+      "learning_rate": 1.5654124956851916e-05,
+      "loss": 0.2836,
+      "step": 2312
+    },
+    {
+      "epoch": 2.1556383970177073,
+      "grad_norm": 0.2804657042496024,
+      "learning_rate": 1.563686572316189e-05,
+      "loss": 0.283,
+      "step": 2313
+    },
+    {
+      "epoch": 2.156570363466915,
+      "grad_norm": 0.2583059598879703,
+      "learning_rate": 1.5619606489471868e-05,
+      "loss": 0.2676,
+      "step": 2314
+    },
+    {
+      "epoch": 2.157502329916123,
+      "grad_norm": 0.23480838559183848,
+      "learning_rate": 1.5602347255781842e-05,
+      "loss": 0.2795,
+      "step": 2315
+    },
+    {
+      "epoch": 2.1584342963653307,
+      "grad_norm": 0.2941911776578257,
+      "learning_rate": 1.558508802209182e-05,
+      "loss": 0.2886,
+      "step": 2316
+    },
+    {
+      "epoch": 2.1593662628145385,
+      "grad_norm": 0.24923475460994537,
+      "learning_rate": 1.5567828788401794e-05,
+      "loss": 0.2641,
+      "step": 2317
+    },
+    {
+      "epoch": 2.1602982292637467,
+      "grad_norm": 0.2151698078086967,
+      "learning_rate": 1.5550569554711772e-05,
+      "loss": 0.2736,
+      "step": 2318
+    },
+    {
+      "epoch": 2.1612301957129545,
+      "grad_norm": 0.22256762028042557,
+      "learning_rate": 1.5533310321021746e-05,
+      "loss": 0.2638,
+      "step": 2319
+    },
+    {
+      "epoch": 2.1621621621621623,
+      "grad_norm": 0.22898285229786583,
+      "learning_rate": 1.5516051087331724e-05,
+      "loss": 0.2711,
+      "step": 2320
+    },
+    {
+      "epoch": 2.16309412861137,
+      "grad_norm": 0.23951327336867537,
+      "learning_rate": 1.5498791853641702e-05,
+      "loss": 0.289,
+      "step": 2321
+    },
+    {
+      "epoch": 2.164026095060578,
+      "grad_norm": 0.24512176403057032,
+      "learning_rate": 1.5481532619951676e-05,
+      "loss": 0.296,
+      "step": 2322
+    },
+    {
+      "epoch": 2.1649580615097856,
+      "grad_norm": 0.24386710323508454,
+      "learning_rate": 1.546427338626165e-05,
+      "loss": 0.2685,
+      "step": 2323
+    },
+    {
+      "epoch": 2.1658900279589934,
+      "grad_norm": 0.2523650354582828,
+      "learning_rate": 1.5447014152571625e-05,
+      "loss": 0.2679,
+      "step": 2324
+    },
+    {
+      "epoch": 2.1668219944082012,
+      "grad_norm": 0.2370561105344379,
+      "learning_rate": 1.5429754918881602e-05,
+      "loss": 0.2956,
+      "step": 2325
+    },
+    {
+      "epoch": 2.167753960857409,
+      "grad_norm": 0.2610354531829981,
+      "learning_rate": 1.5412495685191577e-05,
+      "loss": 0.2788,
+      "step": 2326
+    },
+    {
+      "epoch": 2.168685927306617,
+      "grad_norm": 0.22230294472942494,
+      "learning_rate": 1.5395236451501554e-05,
+      "loss": 0.2715,
+      "step": 2327
+    },
+    {
+      "epoch": 2.1696178937558246,
+      "grad_norm": 0.21501559364712763,
+      "learning_rate": 1.537797721781153e-05,
+      "loss": 0.2702,
+      "step": 2328
+    },
+    {
+      "epoch": 2.170549860205033,
+      "grad_norm": 0.24182125237112692,
+      "learning_rate": 1.5360717984121506e-05,
+      "loss": 0.2831,
+      "step": 2329
+    },
+    {
+      "epoch": 2.1714818266542406,
+      "grad_norm": 0.24448112172717504,
+      "learning_rate": 1.534345875043148e-05,
+      "loss": 0.2801,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 0.23274750026642557,
+      "learning_rate": 1.5326199516741458e-05,
+      "loss": 0.2735,
+      "step": 2331
+    },
+    {
+      "epoch": 2.173345759552656,
+      "grad_norm": 0.2264643154193312,
+      "learning_rate": 1.5308940283051436e-05,
+      "loss": 0.2833,
+      "step": 2332
+    },
+    {
+      "epoch": 2.174277726001864,
+      "grad_norm": 0.22933507820736052,
+      "learning_rate": 1.529168104936141e-05,
+      "loss": 0.2756,
+      "step": 2333
+    },
+    {
+      "epoch": 2.1752096924510718,
+      "grad_norm": 0.23422988443338627,
+      "learning_rate": 1.5274421815671384e-05,
+      "loss": 0.2778,
+      "step": 2334
+    },
+    {
+      "epoch": 2.1761416589002796,
+      "grad_norm": 0.2195224181441558,
+      "learning_rate": 1.525716258198136e-05,
+      "loss": 0.2694,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1770736253494873,
+      "grad_norm": 0.22787894215440938,
+      "learning_rate": 1.5239903348291338e-05,
+      "loss": 0.2574,
+      "step": 2336
+    },
+    {
+      "epoch": 2.178005591798695,
+      "grad_norm": 0.24624634413838414,
+      "learning_rate": 1.5222644114601312e-05,
+      "loss": 0.2877,
+      "step": 2337
+    },
+    {
+      "epoch": 2.178937558247903,
+      "grad_norm": 0.23085370163665211,
+      "learning_rate": 1.5205384880911288e-05,
+      "loss": 0.2723,
+      "step": 2338
+    },
+    {
+      "epoch": 2.1798695246971107,
+      "grad_norm": 0.24093009611928873,
+      "learning_rate": 1.5188125647221263e-05,
+      "loss": 0.2951,
+      "step": 2339
+    },
+    {
+      "epoch": 2.180801491146319,
+      "grad_norm": 0.26438175175205886,
+      "learning_rate": 1.517086641353124e-05,
+      "loss": 0.2651,
+      "step": 2340
+    },
+    {
+      "epoch": 2.1817334575955267,
+      "grad_norm": 0.22206208968372546,
+      "learning_rate": 1.5153607179841216e-05,
+      "loss": 0.2731,
+      "step": 2341
+    },
+    {
+      "epoch": 2.1826654240447345,
+      "grad_norm": 0.24840733896349929,
+      "learning_rate": 1.513634794615119e-05,
+      "loss": 0.2993,
+      "step": 2342
+    },
+    {
+      "epoch": 2.1835973904939423,
+      "grad_norm": 0.24484980163923553,
+      "learning_rate": 1.5119088712461168e-05,
+      "loss": 0.2828,
+      "step": 2343
+    },
+    {
+      "epoch": 2.18452935694315,
+      "grad_norm": 0.23122274433129528,
+      "learning_rate": 1.5101829478771143e-05,
+      "loss": 0.2879,
+      "step": 2344
+    },
+    {
+      "epoch": 2.185461323392358,
+      "grad_norm": 0.23346791147735674,
+      "learning_rate": 1.508457024508112e-05,
+      "loss": 0.2948,
+      "step": 2345
+    },
+    {
+      "epoch": 2.1863932898415657,
+      "grad_norm": 0.2303035996525944,
+      "learning_rate": 1.5067311011391095e-05,
+      "loss": 0.2762,
+      "step": 2346
+    },
+    {
+      "epoch": 2.1873252562907735,
+      "grad_norm": 0.23807939769632588,
+      "learning_rate": 1.505005177770107e-05,
+      "loss": 0.2788,
+      "step": 2347
+    },
+    {
+      "epoch": 2.1882572227399812,
+      "grad_norm": 0.23624751499691174,
+      "learning_rate": 1.5032792544011045e-05,
+      "loss": 0.2514,
+      "step": 2348
+    },
+    {
+      "epoch": 2.189189189189189,
+      "grad_norm": 0.2199924949039847,
+      "learning_rate": 1.5015533310321023e-05,
+      "loss": 0.2499,
+      "step": 2349
+    },
+    {
+      "epoch": 2.190121155638397,
+      "grad_norm": 0.24363428759438482,
+      "learning_rate": 1.4998274076630997e-05,
+      "loss": 0.2819,
+      "step": 2350
+    },
+    {
+      "epoch": 2.191053122087605,
+      "grad_norm": 0.25416965422067295,
+      "learning_rate": 1.4981014842940974e-05,
+      "loss": 0.2778,
+      "step": 2351
+    },
+    {
+      "epoch": 2.191985088536813,
+      "grad_norm": 0.22787195974759483,
+      "learning_rate": 1.496375560925095e-05,
+      "loss": 0.2932,
+      "step": 2352
+    },
+    {
+      "epoch": 2.1929170549860206,
+      "grad_norm": 0.22234172795987875,
+      "learning_rate": 1.4946496375560925e-05,
+      "loss": 0.2769,
+      "step": 2353
+    },
+    {
+      "epoch": 2.1938490214352284,
+      "grad_norm": 0.27848589279519276,
+      "learning_rate": 1.4929237141870902e-05,
+      "loss": 0.2867,
+      "step": 2354
+    },
+    {
+      "epoch": 2.194780987884436,
+      "grad_norm": 0.22474745437804336,
+      "learning_rate": 1.4911977908180877e-05,
+      "loss": 0.2859,
+      "step": 2355
+    },
+    {
+      "epoch": 2.195712954333644,
+      "grad_norm": 0.2355776405546879,
+      "learning_rate": 1.4894718674490854e-05,
+      "loss": 0.2959,
+      "step": 2356
+    },
+    {
+      "epoch": 2.196644920782852,
+      "grad_norm": 0.258505521559796,
+      "learning_rate": 1.4877459440800829e-05,
+      "loss": 0.283,
+      "step": 2357
+    },
+    {
+      "epoch": 2.1975768872320596,
+      "grad_norm": 0.22439743794467512,
+      "learning_rate": 1.4860200207110805e-05,
+      "loss": 0.2772,
+      "step": 2358
+    },
+    {
+      "epoch": 2.1985088536812674,
+      "grad_norm": 0.22878975639701243,
+      "learning_rate": 1.4842940973420779e-05,
+      "loss": 0.2722,
+      "step": 2359
+    },
+    {
+      "epoch": 2.199440820130475,
+      "grad_norm": 0.2258113671572503,
+      "learning_rate": 1.4825681739730757e-05,
+      "loss": 0.284,
+      "step": 2360
+    },
+    {
+      "epoch": 2.200372786579683,
+      "grad_norm": 0.21243811433778306,
+      "learning_rate": 1.4808422506040731e-05,
+      "loss": 0.2663,
+      "step": 2361
+    },
+    {
+      "epoch": 2.201304753028891,
+      "grad_norm": 0.2503197971026304,
+      "learning_rate": 1.4791163272350709e-05,
+      "loss": 0.2632,
+      "step": 2362
+    },
+    {
+      "epoch": 2.202236719478099,
+      "grad_norm": 0.23725591634752657,
+      "learning_rate": 1.4773904038660685e-05,
+      "loss": 0.266,
+      "step": 2363
+    },
+    {
+      "epoch": 2.2031686859273067,
+      "grad_norm": 0.20235250835493643,
+      "learning_rate": 1.4756644804970659e-05,
+      "loss": 0.2759,
+      "step": 2364
+    },
+    {
+      "epoch": 2.2041006523765145,
+      "grad_norm": 0.24016209814140602,
+      "learning_rate": 1.4739385571280637e-05,
+      "loss": 0.2993,
+      "step": 2365
+    },
+    {
+      "epoch": 2.2050326188257223,
+      "grad_norm": 0.2694463479476537,
+      "learning_rate": 1.472212633759061e-05,
+      "loss": 0.266,
+      "step": 2366
+    },
+    {
+      "epoch": 2.20596458527493,
+      "grad_norm": 0.2242057691735019,
+      "learning_rate": 1.4704867103900589e-05,
+      "loss": 0.2653,
+      "step": 2367
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 0.2403491414146798,
+      "learning_rate": 1.4687607870210563e-05,
+      "loss": 0.2584,
+      "step": 2368
+    },
+    {
+      "epoch": 2.2078285181733457,
+      "grad_norm": 0.21774367310450116,
+      "learning_rate": 1.4670348636520539e-05,
+      "loss": 0.2699,
+      "step": 2369
+    },
+    {
+      "epoch": 2.2087604846225535,
+      "grad_norm": 0.2196900172729427,
+      "learning_rate": 1.4653089402830513e-05,
+      "loss": 0.2754,
+      "step": 2370
+    },
+    {
+      "epoch": 2.2096924510717613,
+      "grad_norm": 0.22410497134003848,
+      "learning_rate": 1.463583016914049e-05,
+      "loss": 0.2773,
+      "step": 2371
+    },
+    {
+      "epoch": 2.210624417520969,
+      "grad_norm": 0.8430916853079145,
+      "learning_rate": 1.4618570935450468e-05,
+      "loss": 0.2785,
+      "step": 2372
+    },
+    {
+      "epoch": 2.2115563839701773,
+      "grad_norm": 0.22665500240265918,
+      "learning_rate": 1.4601311701760443e-05,
+      "loss": 0.2601,
+      "step": 2373
+    },
+    {
+      "epoch": 2.212488350419385,
+      "grad_norm": 0.2452803869039672,
+      "learning_rate": 1.4584052468070419e-05,
+      "loss": 0.2796,
+      "step": 2374
+    },
+    {
+      "epoch": 2.213420316868593,
+      "grad_norm": 0.23411562128662766,
+      "learning_rate": 1.4566793234380393e-05,
+      "loss": 0.2871,
+      "step": 2375
+    },
+    {
+      "epoch": 2.2143522833178007,
+      "grad_norm": 0.23344901626349512,
+      "learning_rate": 1.454953400069037e-05,
+      "loss": 0.2887,
+      "step": 2376
+    },
+    {
+      "epoch": 2.2152842497670084,
+      "grad_norm": 0.2225811344037117,
+      "learning_rate": 1.4532274767000345e-05,
+      "loss": 0.2731,
+      "step": 2377
+    },
+    {
+      "epoch": 2.2162162162162162,
+      "grad_norm": 0.22453602519585034,
+      "learning_rate": 1.4515015533310323e-05,
+      "loss": 0.2761,
+      "step": 2378
+    },
+    {
+      "epoch": 2.217148182665424,
+      "grad_norm": 0.259057465447909,
+      "learning_rate": 1.4497756299620297e-05,
+      "loss": 0.2982,
+      "step": 2379
+    },
+    {
+      "epoch": 2.218080149114632,
+      "grad_norm": 0.21368584860862283,
+      "learning_rate": 1.4480497065930273e-05,
+      "loss": 0.2731,
+      "step": 2380
+    },
+    {
+      "epoch": 2.2190121155638396,
+      "grad_norm": 0.20972755552449865,
+      "learning_rate": 1.4463237832240247e-05,
+      "loss": 0.2809,
+      "step": 2381
+    },
+    {
+      "epoch": 2.2199440820130474,
+      "grad_norm": 0.2210747303349958,
+      "learning_rate": 1.4445978598550225e-05,
+      "loss": 0.2671,
+      "step": 2382
+    },
+    {
+      "epoch": 2.220876048462255,
+      "grad_norm": 0.24955772759903805,
+      "learning_rate": 1.4428719364860203e-05,
+      "loss": 0.2676,
+      "step": 2383
+    },
+    {
+      "epoch": 2.2218080149114634,
+      "grad_norm": 0.23390738713371295,
+      "learning_rate": 1.4411460131170177e-05,
+      "loss": 0.2752,
+      "step": 2384
+    },
+    {
+      "epoch": 2.222739981360671,
+      "grad_norm": 0.19506075098877718,
+      "learning_rate": 1.4394200897480153e-05,
+      "loss": 0.2599,
+      "step": 2385
+    },
+    {
+      "epoch": 2.223671947809879,
+      "grad_norm": 0.2341019645123963,
+      "learning_rate": 1.4376941663790127e-05,
+      "loss": 0.2754,
+      "step": 2386
+    },
+    {
+      "epoch": 2.2246039142590868,
+      "grad_norm": 0.23922834043702754,
+      "learning_rate": 1.4359682430100105e-05,
+      "loss": 0.2903,
+      "step": 2387
+    },
+    {
+      "epoch": 2.2255358807082946,
+      "grad_norm": 0.22425772851946113,
+      "learning_rate": 1.4342423196410079e-05,
+      "loss": 0.2734,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2264678471575023,
+      "grad_norm": 0.23359623480543779,
+      "learning_rate": 1.4325163962720057e-05,
+      "loss": 0.2722,
+      "step": 2389
+    },
+    {
+      "epoch": 2.22739981360671,
+      "grad_norm": 0.2276158639116154,
+      "learning_rate": 1.4307904729030031e-05,
+      "loss": 0.2787,
+      "step": 2390
+    },
+    {
+      "epoch": 2.228331780055918,
+      "grad_norm": 0.2307470321382431,
+      "learning_rate": 1.4290645495340007e-05,
+      "loss": 0.2785,
+      "step": 2391
+    },
+    {
+      "epoch": 2.2292637465051257,
+      "grad_norm": 0.2555560749196182,
+      "learning_rate": 1.4273386261649985e-05,
+      "loss": 0.293,
+      "step": 2392
+    },
+    {
+      "epoch": 2.2301957129543335,
+      "grad_norm": 0.2488327051006588,
+      "learning_rate": 1.4256127027959959e-05,
+      "loss": 0.2918,
+      "step": 2393
+    },
+    {
+      "epoch": 2.2311276794035413,
+      "grad_norm": 0.22611721345593747,
+      "learning_rate": 1.4238867794269937e-05,
+      "loss": 0.2726,
+      "step": 2394
+    },
+    {
+      "epoch": 2.2320596458527495,
+      "grad_norm": 0.24828624559359164,
+      "learning_rate": 1.4221608560579911e-05,
+      "loss": 0.2665,
+      "step": 2395
+    },
+    {
+      "epoch": 2.2329916123019573,
+      "grad_norm": 0.24901113034561148,
+      "learning_rate": 1.4204349326889887e-05,
+      "loss": 0.2765,
+      "step": 2396
+    },
+    {
+      "epoch": 2.233923578751165,
+      "grad_norm": 0.23112058126919946,
+      "learning_rate": 1.4187090093199861e-05,
+      "loss": 0.2709,
+      "step": 2397
+    },
+    {
+      "epoch": 2.234855545200373,
+      "grad_norm": 0.2372824543097915,
+      "learning_rate": 1.4169830859509839e-05,
+      "loss": 0.2841,
+      "step": 2398
+    },
+    {
+      "epoch": 2.2357875116495807,
+      "grad_norm": 0.2414253407374196,
+      "learning_rate": 1.4152571625819813e-05,
+      "loss": 0.2716,
+      "step": 2399
+    },
+    {
+      "epoch": 2.2367194780987885,
+      "grad_norm": 0.23826732004067008,
+      "learning_rate": 1.4135312392129791e-05,
+      "loss": 0.2802,
+      "step": 2400
+    },
+    {
+      "epoch": 2.2376514445479962,
+      "grad_norm": 0.2211024867158149,
+      "learning_rate": 1.4118053158439765e-05,
+      "loss": 0.2788,
+      "step": 2401
+    },
+    {
+      "epoch": 2.238583410997204,
+      "grad_norm": 0.2196545036973834,
+      "learning_rate": 1.4100793924749741e-05,
+      "loss": 0.2524,
+      "step": 2402
+    },
+    {
+      "epoch": 2.239515377446412,
+      "grad_norm": 0.248589024861759,
+      "learning_rate": 1.4083534691059719e-05,
+      "loss": 0.2613,
+      "step": 2403
+    },
+    {
+      "epoch": 2.2404473438956196,
+      "grad_norm": 0.22455409165949683,
+      "learning_rate": 1.4066275457369693e-05,
+      "loss": 0.2666,
+      "step": 2404
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 0.20865340215004777,
+      "learning_rate": 1.404901622367967e-05,
+      "loss": 0.2664,
+      "step": 2405
+    },
+    {
+      "epoch": 2.2423112767940356,
+      "grad_norm": 0.26841271113789833,
+      "learning_rate": 1.4031756989989645e-05,
+      "loss": 0.2883,
+      "step": 2406
+    },
+    {
+      "epoch": 2.2432432432432434,
+      "grad_norm": 0.23369257856309417,
+      "learning_rate": 1.4014497756299621e-05,
+      "loss": 0.2731,
+      "step": 2407
+    },
+    {
+      "epoch": 2.244175209692451,
+      "grad_norm": 0.27987690388420716,
+      "learning_rate": 1.3997238522609595e-05,
+      "loss": 0.2944,
+      "step": 2408
+    },
+    {
+      "epoch": 2.245107176141659,
+      "grad_norm": 0.24382450477250117,
+      "learning_rate": 1.3979979288919573e-05,
+      "loss": 0.2722,
+      "step": 2409
+    },
+    {
+      "epoch": 2.246039142590867,
+      "grad_norm": 0.2675875989279932,
+      "learning_rate": 1.3962720055229547e-05,
+      "loss": 0.2925,
+      "step": 2410
+    },
+    {
+      "epoch": 2.2469711090400746,
+      "grad_norm": 0.24753731891695446,
+      "learning_rate": 1.3945460821539525e-05,
+      "loss": 0.2797,
+      "step": 2411
+    },
+    {
+      "epoch": 2.2479030754892824,
+      "grad_norm": 0.26088245066423454,
+      "learning_rate": 1.3928201587849501e-05,
+      "loss": 0.2678,
+      "step": 2412
+    },
+    {
+      "epoch": 2.24883504193849,
+      "grad_norm": 0.2557830776049621,
+      "learning_rate": 1.3910942354159475e-05,
+      "loss": 0.2793,
+      "step": 2413
+    },
+    {
+      "epoch": 2.249767008387698,
+      "grad_norm": 0.21774018135843504,
+      "learning_rate": 1.3893683120469453e-05,
+      "loss": 0.2842,
+      "step": 2414
+    },
+    {
+      "epoch": 2.2506989748369057,
+      "grad_norm": 0.2343785037458116,
+      "learning_rate": 1.3876423886779427e-05,
+      "loss": 0.2853,
+      "step": 2415
+    },
+    {
+      "epoch": 2.2516309412861135,
+      "grad_norm": 0.2808576654058452,
+      "learning_rate": 1.3859164653089405e-05,
+      "loss": 0.2763,
+      "step": 2416
+    },
+    {
+      "epoch": 2.2525629077353218,
+      "grad_norm": 0.2595856187420937,
+      "learning_rate": 1.384190541939938e-05,
+      "loss": 0.286,
+      "step": 2417
+    },
+    {
+      "epoch": 2.2534948741845295,
+      "grad_norm": 0.2331856911500192,
+      "learning_rate": 1.3824646185709355e-05,
+      "loss": 0.2919,
+      "step": 2418
+    },
+    {
+      "epoch": 2.2544268406337373,
+      "grad_norm": 0.2383187577206144,
+      "learning_rate": 1.380738695201933e-05,
+      "loss": 0.2678,
+      "step": 2419
+    },
+    {
+      "epoch": 2.255358807082945,
+      "grad_norm": 0.21594100552656548,
+      "learning_rate": 1.3790127718329307e-05,
+      "loss": 0.2649,
+      "step": 2420
+    },
+    {
+      "epoch": 2.256290773532153,
+      "grad_norm": 0.21561079101381622,
+      "learning_rate": 1.3772868484639282e-05,
+      "loss": 0.2775,
+      "step": 2421
+    },
+    {
+      "epoch": 2.2572227399813607,
+      "grad_norm": 0.23902866883739193,
+      "learning_rate": 1.375560925094926e-05,
+      "loss": 0.2832,
+      "step": 2422
+    },
+    {
+      "epoch": 2.2581547064305685,
+      "grad_norm": 0.2350789168037376,
+      "learning_rate": 1.3738350017259235e-05,
+      "loss": 0.2844,
+      "step": 2423
+    },
+    {
+      "epoch": 2.2590866728797763,
+      "grad_norm": 0.24378434132999338,
+      "learning_rate": 1.372109078356921e-05,
+      "loss": 0.2787,
+      "step": 2424
+    },
+    {
+      "epoch": 2.260018639328984,
+      "grad_norm": 0.2501467000023958,
+      "learning_rate": 1.3703831549879187e-05,
+      "loss": 0.2763,
+      "step": 2425
+    },
+    {
+      "epoch": 2.260950605778192,
+      "grad_norm": 0.2375538497748675,
+      "learning_rate": 1.3686572316189161e-05,
+      "loss": 0.2813,
+      "step": 2426
+    },
+    {
+      "epoch": 2.2618825722273996,
+      "grad_norm": 0.22293269696111717,
+      "learning_rate": 1.3669313082499139e-05,
+      "loss": 0.2688,
+      "step": 2427
+    },
+    {
+      "epoch": 2.262814538676608,
+      "grad_norm": 0.23571903311002376,
+      "learning_rate": 1.3652053848809113e-05,
+      "loss": 0.2858,
+      "step": 2428
+    },
+    {
+      "epoch": 2.2637465051258157,
+      "grad_norm": 0.21292657799763548,
+      "learning_rate": 1.363479461511909e-05,
+      "loss": 0.266,
+      "step": 2429
+    },
+    {
+      "epoch": 2.2646784715750234,
+      "grad_norm": 0.2307737399263272,
+      "learning_rate": 1.3617535381429064e-05,
+      "loss": 0.2757,
+      "step": 2430
+    },
+    {
+      "epoch": 2.2656104380242312,
+      "grad_norm": 0.20940952709364577,
+      "learning_rate": 1.3600276147739041e-05,
+      "loss": 0.2515,
+      "step": 2431
+    },
+    {
+      "epoch": 2.266542404473439,
+      "grad_norm": 0.23781606832798483,
+      "learning_rate": 1.3583016914049019e-05,
+      "loss": 0.2885,
+      "step": 2432
+    },
+    {
+      "epoch": 2.267474370922647,
+      "grad_norm": 0.23595225587950883,
+      "learning_rate": 1.3565757680358993e-05,
+      "loss": 0.2842,
+      "step": 2433
+    },
+    {
+      "epoch": 2.2684063373718546,
+      "grad_norm": 0.22265096407944948,
+      "learning_rate": 1.354849844666897e-05,
+      "loss": 0.2767,
+      "step": 2434
+    },
+    {
+      "epoch": 2.2693383038210624,
+      "grad_norm": 0.2240359507232694,
+      "learning_rate": 1.3531239212978944e-05,
+      "loss": 0.284,
+      "step": 2435
+    },
+    {
+      "epoch": 2.27027027027027,
+      "grad_norm": 0.309136468721953,
+      "learning_rate": 1.3513979979288921e-05,
+      "loss": 0.2741,
+      "step": 2436
+    },
+    {
+      "epoch": 2.271202236719478,
+      "grad_norm": 0.23270109038225362,
+      "learning_rate": 1.3496720745598896e-05,
+      "loss": 0.2808,
+      "step": 2437
+    },
+    {
+      "epoch": 2.2721342031686858,
+      "grad_norm": 0.2230656927659628,
+      "learning_rate": 1.3479461511908873e-05,
+      "loss": 0.2876,
+      "step": 2438
+    },
+    {
+      "epoch": 2.273066169617894,
+      "grad_norm": 0.2195944042720769,
+      "learning_rate": 1.3462202278218848e-05,
+      "loss": 0.2531,
+      "step": 2439
+    },
+    {
+      "epoch": 2.2739981360671018,
+      "grad_norm": 0.23005317022789085,
+      "learning_rate": 1.3444943044528824e-05,
+      "loss": 0.2831,
+      "step": 2440
+    },
+    {
+      "epoch": 2.2749301025163096,
+      "grad_norm": 0.22052812938122998,
+      "learning_rate": 1.3427683810838798e-05,
+      "loss": 0.282,
+      "step": 2441
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 0.24620282838487356,
+      "learning_rate": 1.3410424577148775e-05,
+      "loss": 0.287,
+      "step": 2442
+    },
+    {
+      "epoch": 2.276794035414725,
+      "grad_norm": 0.22082743254811746,
+      "learning_rate": 1.3393165343458751e-05,
+      "loss": 0.2919,
+      "step": 2443
+    },
+    {
+      "epoch": 2.277726001863933,
+      "grad_norm": 0.23199131140987359,
+      "learning_rate": 1.3375906109768726e-05,
+      "loss": 0.2881,
+      "step": 2444
+    },
+    {
+      "epoch": 2.2786579683131407,
+      "grad_norm": 0.23565171228817833,
+      "learning_rate": 1.3358646876078703e-05,
+      "loss": 0.2717,
+      "step": 2445
+    },
+    {
+      "epoch": 2.2795899347623485,
+      "grad_norm": 0.20430654625213066,
+      "learning_rate": 1.3341387642388678e-05,
+      "loss": 0.2663,
+      "step": 2446
+    },
+    {
+      "epoch": 2.2805219012115563,
+      "grad_norm": 0.2133911999038627,
+      "learning_rate": 1.3324128408698655e-05,
+      "loss": 0.2809,
+      "step": 2447
+    },
+    {
+      "epoch": 2.281453867660764,
+      "grad_norm": 0.23886734394361672,
+      "learning_rate": 1.330686917500863e-05,
+      "loss": 0.2642,
+      "step": 2448
+    },
+    {
+      "epoch": 2.282385834109972,
+      "grad_norm": 0.22007401247404915,
+      "learning_rate": 1.3289609941318606e-05,
+      "loss": 0.2783,
+      "step": 2449
+    },
+    {
+      "epoch": 2.28331780055918,
+      "grad_norm": 0.25403642917432423,
+      "learning_rate": 1.327235070762858e-05,
+      "loss": 0.2878,
+      "step": 2450
+    },
+    {
+      "epoch": 2.284249767008388,
+      "grad_norm": 0.21919552727718036,
+      "learning_rate": 1.3255091473938558e-05,
+      "loss": 0.2884,
+      "step": 2451
+    },
+    {
+      "epoch": 2.2851817334575957,
+      "grad_norm": 0.20481774064427652,
+      "learning_rate": 1.3237832240248535e-05,
+      "loss": 0.2818,
+      "step": 2452
+    },
+    {
+      "epoch": 2.2861136999068035,
+      "grad_norm": 0.23488876868347663,
+      "learning_rate": 1.322057300655851e-05,
+      "loss": 0.277,
+      "step": 2453
+    },
+    {
+      "epoch": 2.2870456663560113,
+      "grad_norm": 0.24760659370309548,
+      "learning_rate": 1.3203313772868486e-05,
+      "loss": 0.2706,
+      "step": 2454
+    },
+    {
+      "epoch": 2.287977632805219,
+      "grad_norm": 0.22593143838127788,
+      "learning_rate": 1.318605453917846e-05,
+      "loss": 0.2821,
+      "step": 2455
+    },
+    {
+      "epoch": 2.288909599254427,
+      "grad_norm": 0.23519814078898696,
+      "learning_rate": 1.3168795305488438e-05,
+      "loss": 0.2877,
+      "step": 2456
+    },
+    {
+      "epoch": 2.2898415657036346,
+      "grad_norm": 0.24443144798517696,
+      "learning_rate": 1.3151536071798412e-05,
+      "loss": 0.273,
+      "step": 2457
+    },
+    {
+      "epoch": 2.2907735321528424,
+      "grad_norm": 0.22038995084210433,
+      "learning_rate": 1.313427683810839e-05,
+      "loss": 0.278,
+      "step": 2458
+    },
+    {
+      "epoch": 2.29170549860205,
+      "grad_norm": 0.20808720588440974,
+      "learning_rate": 1.3117017604418364e-05,
+      "loss": 0.2686,
+      "step": 2459
+    },
+    {
+      "epoch": 2.292637465051258,
+      "grad_norm": 0.26318937673330955,
+      "learning_rate": 1.309975837072834e-05,
+      "loss": 0.2876,
+      "step": 2460
+    },
+    {
+      "epoch": 2.293569431500466,
+      "grad_norm": 0.24852824982519425,
+      "learning_rate": 1.3082499137038314e-05,
+      "loss": 0.2797,
+      "step": 2461
+    },
+    {
+      "epoch": 2.294501397949674,
+      "grad_norm": 0.22657184222289428,
+      "learning_rate": 1.3065239903348292e-05,
+      "loss": 0.2732,
+      "step": 2462
+    },
+    {
+      "epoch": 2.295433364398882,
+      "grad_norm": 0.21846753207113817,
+      "learning_rate": 1.304798066965827e-05,
+      "loss": 0.2727,
+      "step": 2463
+    },
+    {
+      "epoch": 2.2963653308480896,
+      "grad_norm": 0.25437536003145506,
+      "learning_rate": 1.3030721435968244e-05,
+      "loss": 0.2748,
+      "step": 2464
+    },
+    {
+      "epoch": 2.2972972972972974,
+      "grad_norm": 0.2503700431090204,
+      "learning_rate": 1.301346220227822e-05,
+      "loss": 0.2744,
+      "step": 2465
+    },
+    {
+      "epoch": 2.298229263746505,
+      "grad_norm": 0.220976652427566,
+      "learning_rate": 1.2996202968588194e-05,
+      "loss": 0.2754,
+      "step": 2466
+    },
+    {
+      "epoch": 2.299161230195713,
+      "grad_norm": 0.25300930482160916,
+      "learning_rate": 1.2978943734898172e-05,
+      "loss": 0.2815,
+      "step": 2467
+    },
+    {
+      "epoch": 2.3000931966449207,
+      "grad_norm": 0.27760864075637387,
+      "learning_rate": 1.2961684501208146e-05,
+      "loss": 0.275,
+      "step": 2468
+    },
+    {
+      "epoch": 2.3010251630941285,
+      "grad_norm": 0.24153061860221509,
+      "learning_rate": 1.2944425267518124e-05,
+      "loss": 0.2625,
+      "step": 2469
+    },
+    {
+      "epoch": 2.3019571295433363,
+      "grad_norm": 0.24567358638152204,
+      "learning_rate": 1.2927166033828098e-05,
+      "loss": 0.2898,
+      "step": 2470
+    },
+    {
+      "epoch": 2.302889095992544,
+      "grad_norm": 0.2468657312058213,
+      "learning_rate": 1.2909906800138074e-05,
+      "loss": 0.2907,
+      "step": 2471
+    },
+    {
+      "epoch": 2.3038210624417523,
+      "grad_norm": 0.24421350562497995,
+      "learning_rate": 1.2892647566448048e-05,
+      "loss": 0.2791,
+      "step": 2472
+    },
+    {
+      "epoch": 2.3047530288909597,
+      "grad_norm": 0.25644507778793524,
+      "learning_rate": 1.2875388332758026e-05,
+      "loss": 0.2831,
+      "step": 2473
+    },
+    {
+      "epoch": 2.305684995340168,
+      "grad_norm": 0.2198670891679119,
+      "learning_rate": 1.2858129099068004e-05,
+      "loss": 0.2681,
+      "step": 2474
+    },
+    {
+      "epoch": 2.3066169617893757,
+      "grad_norm": 0.5787286139351935,
+      "learning_rate": 1.2840869865377978e-05,
+      "loss": 0.2769,
+      "step": 2475
+    },
+    {
+      "epoch": 2.3075489282385835,
+      "grad_norm": 0.1966347855953174,
+      "learning_rate": 1.2823610631687954e-05,
+      "loss": 0.2621,
+      "step": 2476
+    },
+    {
+      "epoch": 2.3084808946877913,
+      "grad_norm": 0.2435257234341102,
+      "learning_rate": 1.2806351397997928e-05,
+      "loss": 0.2769,
+      "step": 2477
+    },
+    {
+      "epoch": 2.309412861136999,
+      "grad_norm": 0.26849506656507,
+      "learning_rate": 1.2789092164307906e-05,
+      "loss": 0.2818,
+      "step": 2478
+    },
+    {
+      "epoch": 2.310344827586207,
+      "grad_norm": 0.21842117301434574,
+      "learning_rate": 1.277183293061788e-05,
+      "loss": 0.2628,
+      "step": 2479
+    },
+    {
+      "epoch": 2.3112767940354146,
+      "grad_norm": 0.22712976622189118,
+      "learning_rate": 1.2754573696927858e-05,
+      "loss": 0.2727,
+      "step": 2480
+    },
+    {
+      "epoch": 2.3122087604846224,
+      "grad_norm": 0.26174630586870995,
+      "learning_rate": 1.2737314463237832e-05,
+      "loss": 0.2761,
+      "step": 2481
+    },
+    {
+      "epoch": 2.31314072693383,
+      "grad_norm": 0.26125267495182575,
+      "learning_rate": 1.2720055229547808e-05,
+      "loss": 0.2997,
+      "step": 2482
+    },
+    {
+      "epoch": 2.3140726933830384,
+      "grad_norm": 0.24732762807432934,
+      "learning_rate": 1.2702795995857786e-05,
+      "loss": 0.2765,
+      "step": 2483
+    },
+    {
+      "epoch": 2.315004659832246,
+      "grad_norm": 0.2302784053249447,
+      "learning_rate": 1.268553676216776e-05,
+      "loss": 0.2731,
+      "step": 2484
+    },
+    {
+      "epoch": 2.315936626281454,
+      "grad_norm": 0.23527746289952198,
+      "learning_rate": 1.2668277528477738e-05,
+      "loss": 0.2766,
+      "step": 2485
+    },
+    {
+      "epoch": 2.316868592730662,
+      "grad_norm": 0.20740713503374963,
+      "learning_rate": 1.2651018294787712e-05,
+      "loss": 0.2663,
+      "step": 2486
+    },
+    {
+      "epoch": 2.3178005591798696,
+      "grad_norm": 0.2638853706802845,
+      "learning_rate": 1.2633759061097688e-05,
+      "loss": 0.287,
+      "step": 2487
+    },
+    {
+      "epoch": 2.3187325256290774,
+      "grad_norm": 0.27404920854712517,
+      "learning_rate": 1.2616499827407662e-05,
+      "loss": 0.2795,
+      "step": 2488
+    },
+    {
+      "epoch": 2.319664492078285,
+      "grad_norm": 0.22935495850409193,
+      "learning_rate": 1.259924059371764e-05,
+      "loss": 0.2747,
+      "step": 2489
+    },
+    {
+      "epoch": 2.320596458527493,
+      "grad_norm": 0.22521331476014025,
+      "learning_rate": 1.2581981360027614e-05,
+      "loss": 0.2617,
+      "step": 2490
+    },
+    {
+      "epoch": 2.3215284249767008,
+      "grad_norm": 0.3070645717899031,
+      "learning_rate": 1.2564722126337592e-05,
+      "loss": 0.2881,
+      "step": 2491
+    },
+    {
+      "epoch": 2.3224603914259085,
+      "grad_norm": 0.2680909003133459,
+      "learning_rate": 1.2547462892647566e-05,
+      "loss": 0.2914,
+      "step": 2492
+    },
+    {
+      "epoch": 2.3233923578751163,
+      "grad_norm": 0.22759207877382054,
+      "learning_rate": 1.2530203658957542e-05,
+      "loss": 0.294,
+      "step": 2493
+    },
+    {
+      "epoch": 2.3243243243243246,
+      "grad_norm": 0.24540336984592537,
+      "learning_rate": 1.251294442526752e-05,
+      "loss": 0.2762,
+      "step": 2494
+    },
+    {
+      "epoch": 2.325256290773532,
+      "grad_norm": 0.2655357514795387,
+      "learning_rate": 1.2495685191577494e-05,
+      "loss": 0.29,
+      "step": 2495
+    },
+    {
+      "epoch": 2.32618825722274,
+      "grad_norm": 0.23311294839296678,
+      "learning_rate": 1.247842595788747e-05,
+      "loss": 0.2726,
+      "step": 2496
+    },
+    {
+      "epoch": 2.327120223671948,
+      "grad_norm": 0.22013815174637963,
+      "learning_rate": 1.2461166724197446e-05,
+      "loss": 0.2684,
+      "step": 2497
+    },
+    {
+      "epoch": 2.3280521901211557,
+      "grad_norm": 0.27829416814713853,
+      "learning_rate": 1.2443907490507422e-05,
+      "loss": 0.2914,
+      "step": 2498
+    },
+    {
+      "epoch": 2.3289841565703635,
+      "grad_norm": 0.23838235718492976,
+      "learning_rate": 1.2426648256817398e-05,
+      "loss": 0.2734,
+      "step": 2499
+    },
+    {
+      "epoch": 2.3299161230195713,
+      "grad_norm": 0.23192431396155086,
+      "learning_rate": 1.2409389023127374e-05,
+      "loss": 0.2695,
+      "step": 2500
+    },
+    {
+      "epoch": 2.330848089468779,
+      "grad_norm": 0.2609563371808852,
+      "learning_rate": 1.239212978943735e-05,
+      "loss": 0.289,
+      "step": 2501
+    },
+    {
+      "epoch": 2.331780055917987,
+      "grad_norm": 0.21977080060319532,
+      "learning_rate": 1.2374870555747326e-05,
+      "loss": 0.2655,
+      "step": 2502
+    },
+    {
+      "epoch": 2.3327120223671947,
+      "grad_norm": 0.23309150220163144,
+      "learning_rate": 1.23576113220573e-05,
+      "loss": 0.2689,
+      "step": 2503
+    },
+    {
+      "epoch": 2.3336439888164024,
+      "grad_norm": 0.24655036416917292,
+      "learning_rate": 1.2340352088367276e-05,
+      "loss": 0.2575,
+      "step": 2504
+    },
+    {
+      "epoch": 2.3345759552656107,
+      "grad_norm": 0.3073446344636903,
+      "learning_rate": 1.2323092854677252e-05,
+      "loss": 0.3008,
+      "step": 2505
+    },
+    {
+      "epoch": 2.335507921714818,
+      "grad_norm": 0.2242921544212533,
+      "learning_rate": 1.2305833620987228e-05,
+      "loss": 0.2711,
+      "step": 2506
+    },
+    {
+      "epoch": 2.3364398881640263,
+      "grad_norm": 0.26478725131600656,
+      "learning_rate": 1.2288574387297204e-05,
+      "loss": 0.2915,
+      "step": 2507
+    },
+    {
+      "epoch": 2.337371854613234,
+      "grad_norm": 0.2583207271736299,
+      "learning_rate": 1.227131515360718e-05,
+      "loss": 0.2825,
+      "step": 2508
+    },
+    {
+      "epoch": 2.338303821062442,
+      "grad_norm": 0.2322121377345241,
+      "learning_rate": 1.2254055919917156e-05,
+      "loss": 0.2953,
+      "step": 2509
+    },
+    {
+      "epoch": 2.3392357875116496,
+      "grad_norm": 0.22949591382268547,
+      "learning_rate": 1.2236796686227132e-05,
+      "loss": 0.2839,
+      "step": 2510
+    },
+    {
+      "epoch": 2.3401677539608574,
+      "grad_norm": 0.26893282281219816,
+      "learning_rate": 1.2219537452537108e-05,
+      "loss": 0.2807,
+      "step": 2511
+    },
+    {
+      "epoch": 2.341099720410065,
+      "grad_norm": 0.22416305407686726,
+      "learning_rate": 1.2202278218847084e-05,
+      "loss": 0.2725,
+      "step": 2512
+    },
+    {
+      "epoch": 2.342031686859273,
+      "grad_norm": 0.2235366241785564,
+      "learning_rate": 1.218501898515706e-05,
+      "loss": 0.2794,
+      "step": 2513
+    },
+    {
+      "epoch": 2.3429636533084808,
+      "grad_norm": 0.23641910679725903,
+      "learning_rate": 1.2167759751467034e-05,
+      "loss": 0.28,
+      "step": 2514
+    },
+    {
+      "epoch": 2.3438956197576886,
+      "grad_norm": 0.21284315703104342,
+      "learning_rate": 1.215050051777701e-05,
+      "loss": 0.294,
+      "step": 2515
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 0.6441448241674512,
+      "learning_rate": 1.2133241284086986e-05,
+      "loss": 0.2892,
+      "step": 2516
+    },
+    {
+      "epoch": 2.345759552656104,
+      "grad_norm": 0.23451849012840398,
+      "learning_rate": 1.2115982050396962e-05,
+      "loss": 0.2757,
+      "step": 2517
+    },
+    {
+      "epoch": 2.3466915191053124,
+      "grad_norm": 0.2402152503146443,
+      "learning_rate": 1.209872281670694e-05,
+      "loss": 0.2721,
+      "step": 2518
+    },
+    {
+      "epoch": 2.34762348555452,
+      "grad_norm": 0.21764572499825854,
+      "learning_rate": 1.2081463583016914e-05,
+      "loss": 0.2968,
+      "step": 2519
+    },
+    {
+      "epoch": 2.348555452003728,
+      "grad_norm": 0.21084709045533662,
+      "learning_rate": 1.206420434932689e-05,
+      "loss": 0.2806,
+      "step": 2520
+    },
+    {
+      "epoch": 2.3494874184529357,
+      "grad_norm": 0.23453367435921577,
+      "learning_rate": 1.2046945115636866e-05,
+      "loss": 0.2795,
+      "step": 2521
+    },
+    {
+      "epoch": 2.3504193849021435,
+      "grad_norm": 0.2139462777259234,
+      "learning_rate": 1.2029685881946842e-05,
+      "loss": 0.2728,
+      "step": 2522
+    },
+    {
+      "epoch": 2.3513513513513513,
+      "grad_norm": 0.2756998716650917,
+      "learning_rate": 1.2012426648256818e-05,
+      "loss": 0.277,
+      "step": 2523
+    },
+    {
+      "epoch": 2.352283317800559,
+      "grad_norm": 0.2276348035714939,
+      "learning_rate": 1.1995167414566794e-05,
+      "loss": 0.2766,
+      "step": 2524
+    },
+    {
+      "epoch": 2.353215284249767,
+      "grad_norm": 0.22961149705271375,
+      "learning_rate": 1.1977908180876769e-05,
+      "loss": 0.2943,
+      "step": 2525
+    },
+    {
+      "epoch": 2.3541472506989747,
+      "grad_norm": 0.21889014686521943,
+      "learning_rate": 1.1960648947186745e-05,
+      "loss": 0.2931,
+      "step": 2526
+    },
+    {
+      "epoch": 2.355079217148183,
+      "grad_norm": 0.22162850626759062,
+      "learning_rate": 1.194338971349672e-05,
+      "loss": 0.2945,
+      "step": 2527
+    },
+    {
+      "epoch": 2.3560111835973903,
+      "grad_norm": 0.2470847043254679,
+      "learning_rate": 1.1926130479806698e-05,
+      "loss": 0.2699,
+      "step": 2528
+    },
+    {
+      "epoch": 2.3569431500465985,
+      "grad_norm": 0.21816954775447933,
+      "learning_rate": 1.1908871246116674e-05,
+      "loss": 0.2741,
+      "step": 2529
+    },
+    {
+      "epoch": 2.3578751164958063,
+      "grad_norm": 0.24130987642953092,
+      "learning_rate": 1.1891612012426649e-05,
+      "loss": 0.2852,
+      "step": 2530
+    },
+    {
+      "epoch": 2.358807082945014,
+      "grad_norm": 0.2328269630942422,
+      "learning_rate": 1.1874352778736625e-05,
+      "loss": 0.2949,
+      "step": 2531
+    },
+    {
+      "epoch": 2.359739049394222,
+      "grad_norm": 0.22459275749151483,
+      "learning_rate": 1.18570935450466e-05,
+      "loss": 0.2847,
+      "step": 2532
+    },
+    {
+      "epoch": 2.3606710158434296,
+      "grad_norm": 0.23397003686100531,
+      "learning_rate": 1.1839834311356576e-05,
+      "loss": 0.271,
+      "step": 2533
+    },
+    {
+      "epoch": 2.3616029822926374,
+      "grad_norm": 0.22472398859757467,
+      "learning_rate": 1.1822575077666552e-05,
+      "loss": 0.2832,
+      "step": 2534
+    },
+    {
+      "epoch": 2.362534948741845,
+      "grad_norm": 0.21991819067669066,
+      "learning_rate": 1.1805315843976528e-05,
+      "loss": 0.2789,
+      "step": 2535
+    },
+    {
+      "epoch": 2.363466915191053,
+      "grad_norm": 0.24511567775651522,
+      "learning_rate": 1.1788056610286503e-05,
+      "loss": 0.2615,
+      "step": 2536
+    },
+    {
+      "epoch": 2.364398881640261,
+      "grad_norm": 0.23099249106481543,
+      "learning_rate": 1.1770797376596479e-05,
+      "loss": 0.2796,
+      "step": 2537
+    },
+    {
+      "epoch": 2.3653308480894686,
+      "grad_norm": 0.2253939191386044,
+      "learning_rate": 1.1753538142906456e-05,
+      "loss": 0.2811,
+      "step": 2538
+    },
+    {
+      "epoch": 2.3662628145386764,
+      "grad_norm": 0.20180057651028857,
+      "learning_rate": 1.1736278909216432e-05,
+      "loss": 0.2587,
+      "step": 2539
+    },
+    {
+      "epoch": 2.3671947809878846,
+      "grad_norm": 0.21921423644475382,
+      "learning_rate": 1.1719019675526408e-05,
+      "loss": 0.2642,
+      "step": 2540
+    },
+    {
+      "epoch": 2.3681267474370924,
+      "grad_norm": 0.23191072582208533,
+      "learning_rate": 1.1701760441836383e-05,
+      "loss": 0.2659,
+      "step": 2541
+    },
+    {
+      "epoch": 2.3690587138863,
+      "grad_norm": 0.22609274734373594,
+      "learning_rate": 1.1684501208146359e-05,
+      "loss": 0.2768,
+      "step": 2542
+    },
+    {
+      "epoch": 2.369990680335508,
+      "grad_norm": 0.2524450266204353,
+      "learning_rate": 1.1667241974456335e-05,
+      "loss": 0.3094,
+      "step": 2543
+    },
+    {
+      "epoch": 2.3709226467847158,
+      "grad_norm": 0.2629533648701109,
+      "learning_rate": 1.164998274076631e-05,
+      "loss": 0.292,
+      "step": 2544
+    },
+    {
+      "epoch": 2.3718546132339235,
+      "grad_norm": 0.21466925259899208,
+      "learning_rate": 1.1632723507076287e-05,
+      "loss": 0.2676,
+      "step": 2545
+    },
+    {
+      "epoch": 2.3727865796831313,
+      "grad_norm": 0.26219990517742203,
+      "learning_rate": 1.1615464273386263e-05,
+      "loss": 0.2854,
+      "step": 2546
+    },
+    {
+      "epoch": 2.373718546132339,
+      "grad_norm": 0.26561132131756854,
+      "learning_rate": 1.1598205039696237e-05,
+      "loss": 0.2729,
+      "step": 2547
+    },
+    {
+      "epoch": 2.374650512581547,
+      "grad_norm": 0.2201527889649934,
+      "learning_rate": 1.1580945806006215e-05,
+      "loss": 0.2591,
+      "step": 2548
+    },
+    {
+      "epoch": 2.3755824790307547,
+      "grad_norm": 0.24990608644973558,
+      "learning_rate": 1.156368657231619e-05,
+      "loss": 0.2687,
+      "step": 2549
+    },
+    {
+      "epoch": 2.3765144454799625,
+      "grad_norm": 0.27385851106441655,
+      "learning_rate": 1.1546427338626167e-05,
+      "loss": 0.2747,
+      "step": 2550
+    },
+    {
+      "epoch": 2.3774464119291707,
+      "grad_norm": 0.22115430877733216,
+      "learning_rate": 1.152916810493614e-05,
+      "loss": 0.2764,
+      "step": 2551
+    },
+    {
+      "epoch": 2.3783783783783785,
+      "grad_norm": 0.20083370287072283,
+      "learning_rate": 1.1511908871246117e-05,
+      "loss": 0.2634,
+      "step": 2552
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 0.2324342578930796,
+      "learning_rate": 1.1494649637556093e-05,
+      "loss": 0.2762,
+      "step": 2553
+    },
+    {
+      "epoch": 2.380242311276794,
+      "grad_norm": 0.24156062253266428,
+      "learning_rate": 1.1477390403866069e-05,
+      "loss": 0.273,
+      "step": 2554
+    },
+    {
+      "epoch": 2.381174277726002,
+      "grad_norm": 0.21758020038288803,
+      "learning_rate": 1.1460131170176045e-05,
+      "loss": 0.2712,
+      "step": 2555
+    },
+    {
+      "epoch": 2.3821062441752097,
+      "grad_norm": 0.22161937024860875,
+      "learning_rate": 1.144287193648602e-05,
+      "loss": 0.2762,
+      "step": 2556
+    },
+    {
+      "epoch": 2.3830382106244175,
+      "grad_norm": 0.23495988984099386,
+      "learning_rate": 1.1425612702795995e-05,
+      "loss": 0.2792,
+      "step": 2557
+    },
+    {
+      "epoch": 2.3839701770736252,
+      "grad_norm": 0.20966034623264168,
+      "learning_rate": 1.1408353469105973e-05,
+      "loss": 0.273,
+      "step": 2558
+    },
+    {
+      "epoch": 2.384902143522833,
+      "grad_norm": 0.23891718287903238,
+      "learning_rate": 1.1391094235415949e-05,
+      "loss": 0.2883,
+      "step": 2559
+    },
+    {
+      "epoch": 2.385834109972041,
+      "grad_norm": 0.22540951108482382,
+      "learning_rate": 1.1373835001725925e-05,
+      "loss": 0.2609,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3867660764212486,
+      "grad_norm": 0.24400622020286825,
+      "learning_rate": 1.13565757680359e-05,
+      "loss": 0.2971,
+      "step": 2561
+    },
+    {
+      "epoch": 2.387698042870457,
+      "grad_norm": 0.23848124953564856,
+      "learning_rate": 1.1339316534345875e-05,
+      "loss": 0.2689,
+      "step": 2562
+    },
+    {
+      "epoch": 2.3886300093196646,
+      "grad_norm": 0.2323843849884664,
+      "learning_rate": 1.1322057300655851e-05,
+      "loss": 0.2881,
+      "step": 2563
+    },
+    {
+      "epoch": 2.3895619757688724,
+      "grad_norm": 0.24577683550642765,
+      "learning_rate": 1.1304798066965827e-05,
+      "loss": 0.2932,
+      "step": 2564
+    },
+    {
+      "epoch": 2.39049394221808,
+      "grad_norm": 0.23922492727069689,
+      "learning_rate": 1.1287538833275803e-05,
+      "loss": 0.2933,
+      "step": 2565
+    },
+    {
+      "epoch": 2.391425908667288,
+      "grad_norm": 0.23731564838465063,
+      "learning_rate": 1.1270279599585779e-05,
+      "loss": 0.2671,
+      "step": 2566
+    },
+    {
+      "epoch": 2.392357875116496,
+      "grad_norm": 0.26358389612899974,
+      "learning_rate": 1.1253020365895755e-05,
+      "loss": 0.2824,
+      "step": 2567
+    },
+    {
+      "epoch": 2.3932898415657036,
+      "grad_norm": 0.2186116973227067,
+      "learning_rate": 1.1235761132205731e-05,
+      "loss": 0.2687,
+      "step": 2568
+    },
+    {
+      "epoch": 2.3942218080149114,
+      "grad_norm": 0.2401843727786183,
+      "learning_rate": 1.1218501898515707e-05,
+      "loss": 0.2928,
+      "step": 2569
+    },
+    {
+      "epoch": 2.395153774464119,
+      "grad_norm": 0.2483730461909779,
+      "learning_rate": 1.1201242664825683e-05,
+      "loss": 0.292,
+      "step": 2570
+    },
+    {
+      "epoch": 2.396085740913327,
+      "grad_norm": 0.2256589323192343,
+      "learning_rate": 1.1183983431135659e-05,
+      "loss": 0.2634,
+      "step": 2571
+    },
+    {
+      "epoch": 2.3970177073625347,
+      "grad_norm": 0.21656131526247724,
+      "learning_rate": 1.1166724197445635e-05,
+      "loss": 0.2927,
+      "step": 2572
+    },
+    {
+      "epoch": 2.397949673811743,
+      "grad_norm": 0.2259038522556601,
+      "learning_rate": 1.1149464963755609e-05,
+      "loss": 0.2751,
+      "step": 2573
+    },
+    {
+      "epoch": 2.3988816402609507,
+      "grad_norm": 0.25999429686292275,
+      "learning_rate": 1.1132205730065585e-05,
+      "loss": 0.2725,
+      "step": 2574
+    },
+    {
+      "epoch": 2.3998136067101585,
+      "grad_norm": 0.20599809153418228,
+      "learning_rate": 1.1114946496375561e-05,
+      "loss": 0.27,
+      "step": 2575
+    },
+    {
+      "epoch": 2.4007455731593663,
+      "grad_norm": 0.21748158808297632,
+      "learning_rate": 1.1097687262685537e-05,
+      "loss": 0.288,
+      "step": 2576
+    },
+    {
+      "epoch": 2.401677539608574,
+      "grad_norm": 0.23954952002077642,
+      "learning_rate": 1.1080428028995513e-05,
+      "loss": 0.2889,
+      "step": 2577
+    },
+    {
+      "epoch": 2.402609506057782,
+      "grad_norm": 0.21621599815452938,
+      "learning_rate": 1.1063168795305489e-05,
+      "loss": 0.2814,
+      "step": 2578
+    },
+    {
+      "epoch": 2.4035414725069897,
+      "grad_norm": 0.24092536914785856,
+      "learning_rate": 1.1045909561615465e-05,
+      "loss": 0.2781,
+      "step": 2579
+    },
+    {
+      "epoch": 2.4044734389561975,
+      "grad_norm": 0.19524569201306594,
+      "learning_rate": 1.1028650327925441e-05,
+      "loss": 0.2655,
+      "step": 2580
+    },
+    {
+      "epoch": 2.4054054054054053,
+      "grad_norm": 0.22396898405049465,
+      "learning_rate": 1.1011391094235417e-05,
+      "loss": 0.2894,
+      "step": 2581
+    },
+    {
+      "epoch": 2.406337371854613,
+      "grad_norm": 0.22353740646324502,
+      "learning_rate": 1.0994131860545393e-05,
+      "loss": 0.301,
+      "step": 2582
+    },
+    {
+      "epoch": 2.407269338303821,
+      "grad_norm": 0.23944109604276717,
+      "learning_rate": 1.0976872626855369e-05,
+      "loss": 0.2898,
+      "step": 2583
+    },
+    {
+      "epoch": 2.408201304753029,
+      "grad_norm": 0.2365843331292834,
+      "learning_rate": 1.0959613393165343e-05,
+      "loss": 0.3008,
+      "step": 2584
+    },
+    {
+      "epoch": 2.409133271202237,
+      "grad_norm": 0.2185750619087877,
+      "learning_rate": 1.094235415947532e-05,
+      "loss": 0.2774,
+      "step": 2585
+    },
+    {
+      "epoch": 2.4100652376514446,
+      "grad_norm": 0.2257896603532674,
+      "learning_rate": 1.0925094925785295e-05,
+      "loss": 0.2738,
+      "step": 2586
+    },
+    {
+      "epoch": 2.4109972041006524,
+      "grad_norm": 0.2187777644652006,
+      "learning_rate": 1.0907835692095271e-05,
+      "loss": 0.2784,
+      "step": 2587
+    },
+    {
+      "epoch": 2.4119291705498602,
+      "grad_norm": 0.19803169462871312,
+      "learning_rate": 1.0890576458405247e-05,
+      "loss": 0.2576,
+      "step": 2588
+    },
+    {
+      "epoch": 2.412861136999068,
+      "grad_norm": 0.2232843316216573,
+      "learning_rate": 1.0873317224715223e-05,
+      "loss": 0.2804,
+      "step": 2589
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 0.2888423603164031,
+      "learning_rate": 1.0856057991025199e-05,
+      "loss": 0.2754,
+      "step": 2590
+    },
+    {
+      "epoch": 2.4147250698974836,
+      "grad_norm": 0.20080571516772333,
+      "learning_rate": 1.0838798757335175e-05,
+      "loss": 0.2718,
+      "step": 2591
+    },
+    {
+      "epoch": 2.4156570363466914,
+      "grad_norm": 0.21458468659098842,
+      "learning_rate": 1.0821539523645151e-05,
+      "loss": 0.2793,
+      "step": 2592
+    },
+    {
+      "epoch": 2.416589002795899,
+      "grad_norm": 0.25540553904693664,
+      "learning_rate": 1.0804280289955127e-05,
+      "loss": 0.2827,
+      "step": 2593
+    },
+    {
+      "epoch": 2.417520969245107,
+      "grad_norm": 0.2357406278369054,
+      "learning_rate": 1.0787021056265103e-05,
+      "loss": 0.2802,
+      "step": 2594
+    },
+    {
+      "epoch": 2.418452935694315,
+      "grad_norm": 0.22005559538528127,
+      "learning_rate": 1.0769761822575077e-05,
+      "loss": 0.2785,
+      "step": 2595
+    },
+    {
+      "epoch": 2.419384902143523,
+      "grad_norm": 0.26918083691201733,
+      "learning_rate": 1.0752502588885053e-05,
+      "loss": 0.2895,
+      "step": 2596
+    },
+    {
+      "epoch": 2.4203168685927308,
+      "grad_norm": 0.21763987528726214,
+      "learning_rate": 1.073524335519503e-05,
+      "loss": 0.2829,
+      "step": 2597
+    },
+    {
+      "epoch": 2.4212488350419386,
+      "grad_norm": 0.21077755130266135,
+      "learning_rate": 1.0717984121505005e-05,
+      "loss": 0.2796,
+      "step": 2598
+    },
+    {
+      "epoch": 2.4221808014911463,
+      "grad_norm": 0.21434933521313684,
+      "learning_rate": 1.0700724887814981e-05,
+      "loss": 0.2593,
+      "step": 2599
+    },
+    {
+      "epoch": 2.423112767940354,
+      "grad_norm": 0.21920132141184157,
+      "learning_rate": 1.0683465654124957e-05,
+      "loss": 0.2731,
+      "step": 2600
+    },
+    {
+      "epoch": 2.424044734389562,
+      "grad_norm": 0.23224021707073092,
+      "learning_rate": 1.0666206420434933e-05,
+      "loss": 0.2802,
+      "step": 2601
+    },
+    {
+      "epoch": 2.4249767008387697,
+      "grad_norm": 0.2100950495465491,
+      "learning_rate": 1.064894718674491e-05,
+      "loss": 0.282,
+      "step": 2602
+    },
+    {
+      "epoch": 2.4259086672879775,
+      "grad_norm": 0.20431936579830387,
+      "learning_rate": 1.0631687953054885e-05,
+      "loss": 0.29,
+      "step": 2603
+    },
+    {
+      "epoch": 2.4268406337371853,
+      "grad_norm": 0.20619609620460352,
+      "learning_rate": 1.0614428719364861e-05,
+      "loss": 0.268,
+      "step": 2604
+    },
+    {
+      "epoch": 2.427772600186393,
+      "grad_norm": 0.2028339387893656,
+      "learning_rate": 1.0597169485674835e-05,
+      "loss": 0.2872,
+      "step": 2605
+    },
+    {
+      "epoch": 2.4287045666356013,
+      "grad_norm": 0.22660759785593776,
+      "learning_rate": 1.0579910251984811e-05,
+      "loss": 0.2825,
+      "step": 2606
+    },
+    {
+      "epoch": 2.429636533084809,
+      "grad_norm": 0.20446526845836402,
+      "learning_rate": 1.0562651018294787e-05,
+      "loss": 0.2744,
+      "step": 2607
+    },
+    {
+      "epoch": 2.430568499534017,
+      "grad_norm": 0.21770163356271194,
+      "learning_rate": 1.0545391784604763e-05,
+      "loss": 0.2789,
+      "step": 2608
+    },
+    {
+      "epoch": 2.4315004659832247,
+      "grad_norm": 0.22309385388369626,
+      "learning_rate": 1.0528132550914741e-05,
+      "loss": 0.2694,
+      "step": 2609
+    },
+    {
+      "epoch": 2.4324324324324325,
+      "grad_norm": 0.22266898725040343,
+      "learning_rate": 1.0510873317224715e-05,
+      "loss": 0.2752,
+      "step": 2610
+    },
+    {
+      "epoch": 2.4333643988816402,
+      "grad_norm": 0.22099814635381945,
+      "learning_rate": 1.0493614083534691e-05,
+      "loss": 0.2915,
+      "step": 2611
+    },
+    {
+      "epoch": 2.434296365330848,
+      "grad_norm": 0.25110590910943176,
+      "learning_rate": 1.0476354849844667e-05,
+      "loss": 0.2801,
+      "step": 2612
+    },
+    {
+      "epoch": 2.435228331780056,
+      "grad_norm": 0.19569860709968273,
+      "learning_rate": 1.0459095616154643e-05,
+      "loss": 0.2514,
+      "step": 2613
+    },
+    {
+      "epoch": 2.4361602982292636,
+      "grad_norm": 0.20791728147817656,
+      "learning_rate": 1.044183638246462e-05,
+      "loss": 0.2728,
+      "step": 2614
+    },
+    {
+      "epoch": 2.4370922646784714,
+      "grad_norm": 0.23696997531012343,
+      "learning_rate": 1.0424577148774595e-05,
+      "loss": 0.2703,
+      "step": 2615
+    },
+    {
+      "epoch": 2.438024231127679,
+      "grad_norm": 0.2057297050934854,
+      "learning_rate": 1.040731791508457e-05,
+      "loss": 0.2706,
+      "step": 2616
+    },
+    {
+      "epoch": 2.4389561975768874,
+      "grad_norm": 0.22197329295451432,
+      "learning_rate": 1.0390058681394546e-05,
+      "loss": 0.2817,
+      "step": 2617
+    },
+    {
+      "epoch": 2.439888164026095,
+      "grad_norm": 0.19953639785859545,
+      "learning_rate": 1.0372799447704522e-05,
+      "loss": 0.2654,
+      "step": 2618
+    },
+    {
+      "epoch": 2.440820130475303,
+      "grad_norm": 0.29952519044244236,
+      "learning_rate": 1.03555402140145e-05,
+      "loss": 0.2877,
+      "step": 2619
+    },
+    {
+      "epoch": 2.441752096924511,
+      "grad_norm": 0.26089356566138217,
+      "learning_rate": 1.0338280980324475e-05,
+      "loss": 0.2843,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4426840633737186,
+      "grad_norm": 0.20924755991485675,
+      "learning_rate": 1.032102174663445e-05,
+      "loss": 0.2868,
+      "step": 2621
+    },
+    {
+      "epoch": 2.4436160298229264,
+      "grad_norm": 0.22517661160426986,
+      "learning_rate": 1.0303762512944426e-05,
+      "loss": 0.2728,
+      "step": 2622
+    },
+    {
+      "epoch": 2.444547996272134,
+      "grad_norm": 0.21835793784183094,
+      "learning_rate": 1.0286503279254402e-05,
+      "loss": 0.2669,
+      "step": 2623
+    },
+    {
+      "epoch": 2.445479962721342,
+      "grad_norm": 0.23557086663402607,
+      "learning_rate": 1.0269244045564377e-05,
+      "loss": 0.272,
+      "step": 2624
+    },
+    {
+      "epoch": 2.4464119291705497,
+      "grad_norm": 0.24898302050906623,
+      "learning_rate": 1.0251984811874353e-05,
+      "loss": 0.2759,
+      "step": 2625
+    },
+    {
+      "epoch": 2.4473438956197575,
+      "grad_norm": 0.20822301120311554,
+      "learning_rate": 1.023472557818433e-05,
+      "loss": 0.272,
+      "step": 2626
+    },
+    {
+      "epoch": 2.4482758620689653,
+      "grad_norm": 0.20822735262343084,
+      "learning_rate": 1.0217466344494304e-05,
+      "loss": 0.2719,
+      "step": 2627
+    },
+    {
+      "epoch": 2.4492078285181735,
+      "grad_norm": 0.21975312023067395,
+      "learning_rate": 1.020020711080428e-05,
+      "loss": 0.2702,
+      "step": 2628
+    },
+    {
+      "epoch": 2.4501397949673813,
+      "grad_norm": 0.2223817737915316,
+      "learning_rate": 1.0182947877114257e-05,
+      "loss": 0.2781,
+      "step": 2629
+    },
+    {
+      "epoch": 2.451071761416589,
+      "grad_norm": 0.2020869607441318,
+      "learning_rate": 1.0165688643424233e-05,
+      "loss": 0.2704,
+      "step": 2630
+    },
+    {
+      "epoch": 2.452003727865797,
+      "grad_norm": 0.2001222751372582,
+      "learning_rate": 1.014842940973421e-05,
+      "loss": 0.2771,
+      "step": 2631
+    },
+    {
+      "epoch": 2.4529356943150047,
+      "grad_norm": 0.22728436725748474,
+      "learning_rate": 1.0131170176044184e-05,
+      "loss": 0.2727,
+      "step": 2632
+    },
+    {
+      "epoch": 2.4538676607642125,
+      "grad_norm": 0.22544910806917423,
+      "learning_rate": 1.011391094235416e-05,
+      "loss": 0.2716,
+      "step": 2633
+    },
+    {
+      "epoch": 2.4547996272134203,
+      "grad_norm": 0.20350370806281137,
+      "learning_rate": 1.0096651708664136e-05,
+      "loss": 0.2569,
+      "step": 2634
+    },
+    {
+      "epoch": 2.455731593662628,
+      "grad_norm": 0.24102274208100458,
+      "learning_rate": 1.0079392474974112e-05,
+      "loss": 0.2754,
+      "step": 2635
+    },
+    {
+      "epoch": 2.456663560111836,
+      "grad_norm": 0.23011887585597035,
+      "learning_rate": 1.0062133241284088e-05,
+      "loss": 0.2588,
+      "step": 2636
+    },
+    {
+      "epoch": 2.4575955265610436,
+      "grad_norm": 0.2070829184955601,
+      "learning_rate": 1.0044874007594064e-05,
+      "loss": 0.2798,
+      "step": 2637
+    },
+    {
+      "epoch": 2.4585274930102514,
+      "grad_norm": 0.2287431209956293,
+      "learning_rate": 1.0027614773904038e-05,
+      "loss": 0.2797,
+      "step": 2638
+    },
+    {
+      "epoch": 2.4594594594594597,
+      "grad_norm": 0.2370172584796964,
+      "learning_rate": 1.0010355540214016e-05,
+      "loss": 0.2754,
+      "step": 2639
+    },
+    {
+      "epoch": 2.4603914259086674,
+      "grad_norm": 0.21378358198659064,
+      "learning_rate": 9.993096306523992e-06,
+      "loss": 0.2704,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4613233923578752,
+      "grad_norm": 0.21429911886241038,
+      "learning_rate": 9.975837072833968e-06,
+      "loss": 0.2748,
+      "step": 2641
+    },
+    {
+      "epoch": 2.462255358807083,
+      "grad_norm": 0.22951129693412117,
+      "learning_rate": 9.958577839143944e-06,
+      "loss": 0.2666,
+      "step": 2642
+    },
+    {
+      "epoch": 2.463187325256291,
+      "grad_norm": 0.2061608055359151,
+      "learning_rate": 9.941318605453918e-06,
+      "loss": 0.2685,
+      "step": 2643
+    },
+    {
+      "epoch": 2.4641192917054986,
+      "grad_norm": 0.2193357755986692,
+      "learning_rate": 9.924059371763894e-06,
+      "loss": 0.2772,
+      "step": 2644
+    },
+    {
+      "epoch": 2.4650512581547064,
+      "grad_norm": 0.20375091994359532,
+      "learning_rate": 9.90680013807387e-06,
+      "loss": 0.2509,
+      "step": 2645
+    },
+    {
+      "epoch": 2.465983224603914,
+      "grad_norm": 0.24055166484722676,
+      "learning_rate": 9.889540904383846e-06,
+      "loss": 0.2879,
+      "step": 2646
+    },
+    {
+      "epoch": 2.466915191053122,
+      "grad_norm": 0.22248099309687483,
+      "learning_rate": 9.872281670693822e-06,
+      "loss": 0.2799,
+      "step": 2647
+    },
+    {
+      "epoch": 2.4678471575023297,
+      "grad_norm": 0.19816070568981456,
+      "learning_rate": 9.855022437003798e-06,
+      "loss": 0.2669,
+      "step": 2648
+    },
+    {
+      "epoch": 2.4687791239515375,
+      "grad_norm": 0.24605913719466918,
+      "learning_rate": 9.837763203313774e-06,
+      "loss": 0.2761,
+      "step": 2649
+    },
+    {
+      "epoch": 2.4697110904007458,
+      "grad_norm": 0.24397927283676732,
+      "learning_rate": 9.82050396962375e-06,
+      "loss": 0.2994,
+      "step": 2650
+    },
+    {
+      "epoch": 2.4706430568499536,
+      "grad_norm": 0.1957343970947358,
+      "learning_rate": 9.803244735933726e-06,
+      "loss": 0.2666,
+      "step": 2651
+    },
+    {
+      "epoch": 2.4715750232991613,
+      "grad_norm": 0.22211236038118418,
+      "learning_rate": 9.785985502243702e-06,
+      "loss": 0.283,
+      "step": 2652
+    },
+    {
+      "epoch": 2.472506989748369,
+      "grad_norm": 0.21437959112891306,
+      "learning_rate": 9.768726268553676e-06,
+      "loss": 0.2776,
+      "step": 2653
+    },
+    {
+      "epoch": 2.473438956197577,
+      "grad_norm": 0.2067797003529056,
+      "learning_rate": 9.751467034863652e-06,
+      "loss": 0.2722,
+      "step": 2654
+    },
+    {
+      "epoch": 2.4743709226467847,
+      "grad_norm": 0.21459789072283753,
+      "learning_rate": 9.734207801173628e-06,
+      "loss": 0.2853,
+      "step": 2655
+    },
+    {
+      "epoch": 2.4753028890959925,
+      "grad_norm": 0.2357715216576148,
+      "learning_rate": 9.716948567483604e-06,
+      "loss": 0.2826,
+      "step": 2656
+    },
+    {
+      "epoch": 2.4762348555452003,
+      "grad_norm": 0.19598206562150305,
+      "learning_rate": 9.69968933379358e-06,
+      "loss": 0.261,
+      "step": 2657
+    },
+    {
+      "epoch": 2.477166821994408,
+      "grad_norm": 0.20841017113707772,
+      "learning_rate": 9.682430100103556e-06,
+      "loss": 0.263,
+      "step": 2658
+    },
+    {
+      "epoch": 2.478098788443616,
+      "grad_norm": 0.22027861799108128,
+      "learning_rate": 9.665170866413532e-06,
+      "loss": 0.2769,
+      "step": 2659
+    },
+    {
+      "epoch": 2.4790307548928237,
+      "grad_norm": 0.20743564629476824,
+      "learning_rate": 9.647911632723508e-06,
+      "loss": 0.276,
+      "step": 2660
+    },
+    {
+      "epoch": 2.479962721342032,
+      "grad_norm": 0.20632266126761892,
+      "learning_rate": 9.630652399033484e-06,
+      "loss": 0.271,
+      "step": 2661
+    },
+    {
+      "epoch": 2.4808946877912397,
+      "grad_norm": 0.22906465099886222,
+      "learning_rate": 9.61339316534346e-06,
+      "loss": 0.2874,
+      "step": 2662
+    },
+    {
+      "epoch": 2.4818266542404475,
+      "grad_norm": 0.22140809079088197,
+      "learning_rate": 9.596133931653436e-06,
+      "loss": 0.2839,
+      "step": 2663
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 0.20375057588022757,
+      "learning_rate": 9.57887469796341e-06,
+      "loss": 0.2704,
+      "step": 2664
+    },
+    {
+      "epoch": 2.483690587138863,
+      "grad_norm": 0.2197296174853955,
+      "learning_rate": 9.561615464273386e-06,
+      "loss": 0.2769,
+      "step": 2665
+    },
+    {
+      "epoch": 2.484622553588071,
+      "grad_norm": 0.22989250445048004,
+      "learning_rate": 9.544356230583362e-06,
+      "loss": 0.2798,
+      "step": 2666
+    },
+    {
+      "epoch": 2.4855545200372786,
+      "grad_norm": 0.22966432937009226,
+      "learning_rate": 9.527096996893338e-06,
+      "loss": 0.2985,
+      "step": 2667
+    },
+    {
+      "epoch": 2.4864864864864864,
+      "grad_norm": 0.21700294708562964,
+      "learning_rate": 9.509837763203314e-06,
+      "loss": 0.2678,
+      "step": 2668
+    },
+    {
+      "epoch": 2.487418452935694,
+      "grad_norm": 0.24222085319134912,
+      "learning_rate": 9.49257852951329e-06,
+      "loss": 0.261,
+      "step": 2669
+    },
+    {
+      "epoch": 2.488350419384902,
+      "grad_norm": 0.22108230115551303,
+      "learning_rate": 9.475319295823266e-06,
+      "loss": 0.2801,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4892823858341098,
+      "grad_norm": 0.2246747099005012,
+      "learning_rate": 9.458060062133242e-06,
+      "loss": 0.2892,
+      "step": 2671
+    },
+    {
+      "epoch": 2.490214352283318,
+      "grad_norm": 0.21245110785887747,
+      "learning_rate": 9.440800828443218e-06,
+      "loss": 0.2697,
+      "step": 2672
+    },
+    {
+      "epoch": 2.491146318732526,
+      "grad_norm": 0.2509867345613903,
+      "learning_rate": 9.423541594753194e-06,
+      "loss": 0.2961,
+      "step": 2673
+    },
+    {
+      "epoch": 2.4920782851817336,
+      "grad_norm": 0.21610369801137763,
+      "learning_rate": 9.40628236106317e-06,
+      "loss": 0.2716,
+      "step": 2674
+    },
+    {
+      "epoch": 2.4930102516309414,
+      "grad_norm": 0.21592297380701186,
+      "learning_rate": 9.389023127373144e-06,
+      "loss": 0.2826,
+      "step": 2675
+    },
+    {
+      "epoch": 2.493942218080149,
+      "grad_norm": 0.2179721974187931,
+      "learning_rate": 9.37176389368312e-06,
+      "loss": 0.2731,
+      "step": 2676
+    },
+    {
+      "epoch": 2.494874184529357,
+      "grad_norm": 0.20795654085245516,
+      "learning_rate": 9.354504659993096e-06,
+      "loss": 0.2695,
+      "step": 2677
+    },
+    {
+      "epoch": 2.4958061509785647,
+      "grad_norm": 0.2306797925601343,
+      "learning_rate": 9.337245426303072e-06,
+      "loss": 0.2715,
+      "step": 2678
+    },
+    {
+      "epoch": 2.4967381174277725,
+      "grad_norm": 0.22001585966099468,
+      "learning_rate": 9.31998619261305e-06,
+      "loss": 0.2813,
+      "step": 2679
+    },
+    {
+      "epoch": 2.4976700838769803,
+      "grad_norm": 0.2243036936125682,
+      "learning_rate": 9.302726958923024e-06,
+      "loss": 0.2933,
+      "step": 2680
+    },
+    {
+      "epoch": 2.498602050326188,
+      "grad_norm": 0.2202011073588633,
+      "learning_rate": 9.285467725233e-06,
+      "loss": 0.2852,
+      "step": 2681
+    },
+    {
+      "epoch": 2.499534016775396,
+      "grad_norm": 0.2230406601854783,
+      "learning_rate": 9.268208491542976e-06,
+      "loss": 0.2681,
+      "step": 2682
+    },
+    {
+      "epoch": 2.500465983224604,
+      "grad_norm": 0.21286066654838973,
+      "learning_rate": 9.250949257852952e-06,
+      "loss": 0.2669,
+      "step": 2683
+    },
+    {
+      "epoch": 2.501397949673812,
+      "grad_norm": 0.2389168419876547,
+      "learning_rate": 9.233690024162928e-06,
+      "loss": 0.2682,
+      "step": 2684
+    },
+    {
+      "epoch": 2.5023299161230197,
+      "grad_norm": 0.2452249769123563,
+      "learning_rate": 9.216430790472904e-06,
+      "loss": 0.2903,
+      "step": 2685
+    },
+    {
+      "epoch": 2.5032618825722275,
+      "grad_norm": 0.2387382307009486,
+      "learning_rate": 9.199171556782878e-06,
+      "loss": 0.303,
+      "step": 2686
+    },
+    {
+      "epoch": 2.5041938490214353,
+      "grad_norm": 0.21340792158819152,
+      "learning_rate": 9.181912323092854e-06,
+      "loss": 0.2808,
+      "step": 2687
+    },
+    {
+      "epoch": 2.505125815470643,
+      "grad_norm": 0.22829356597508393,
+      "learning_rate": 9.16465308940283e-06,
+      "loss": 0.2767,
+      "step": 2688
+    },
+    {
+      "epoch": 2.506057781919851,
+      "grad_norm": 0.22948366315571858,
+      "learning_rate": 9.147393855712806e-06,
+      "loss": 0.2865,
+      "step": 2689
+    },
+    {
+      "epoch": 2.5069897483690586,
+      "grad_norm": 0.19139159379500573,
+      "learning_rate": 9.130134622022784e-06,
+      "loss": 0.2831,
+      "step": 2690
+    },
+    {
+      "epoch": 2.5079217148182664,
+      "grad_norm": 0.20077938758952968,
+      "learning_rate": 9.112875388332758e-06,
+      "loss": 0.2747,
+      "step": 2691
+    },
+    {
+      "epoch": 2.508853681267474,
+      "grad_norm": 0.1999183655240357,
+      "learning_rate": 9.095616154642734e-06,
+      "loss": 0.2719,
+      "step": 2692
+    },
+    {
+      "epoch": 2.509785647716682,
+      "grad_norm": 0.2180371829432636,
+      "learning_rate": 9.07835692095271e-06,
+      "loss": 0.272,
+      "step": 2693
+    },
+    {
+      "epoch": 2.5107176141658902,
+      "grad_norm": 0.23250142142199415,
+      "learning_rate": 9.061097687262686e-06,
+      "loss": 0.2789,
+      "step": 2694
+    },
+    {
+      "epoch": 2.511649580615098,
+      "grad_norm": 0.20348726356487723,
+      "learning_rate": 9.043838453572662e-06,
+      "loss": 0.2766,
+      "step": 2695
+    },
+    {
+      "epoch": 2.512581547064306,
+      "grad_norm": 0.1919358091771706,
+      "learning_rate": 9.026579219882638e-06,
+      "loss": 0.2756,
+      "step": 2696
+    },
+    {
+      "epoch": 2.5135135135135136,
+      "grad_norm": 0.24859079760420713,
+      "learning_rate": 9.009319986192612e-06,
+      "loss": 0.2782,
+      "step": 2697
+    },
+    {
+      "epoch": 2.5144454799627214,
+      "grad_norm": 0.23672544010306634,
+      "learning_rate": 8.992060752502588e-06,
+      "loss": 0.2765,
+      "step": 2698
+    },
+    {
+      "epoch": 2.515377446411929,
+      "grad_norm": 0.19330173516193494,
+      "learning_rate": 8.974801518812564e-06,
+      "loss": 0.2662,
+      "step": 2699
+    },
+    {
+      "epoch": 2.516309412861137,
+      "grad_norm": 0.20265963975260246,
+      "learning_rate": 8.957542285122542e-06,
+      "loss": 0.2678,
+      "step": 2700
+    },
+    {
+      "epoch": 2.5172413793103448,
+      "grad_norm": 0.22059980238110408,
+      "learning_rate": 8.940283051432516e-06,
+      "loss": 0.2627,
+      "step": 2701
+    },
+    {
+      "epoch": 2.5181733457595525,
+      "grad_norm": 0.23026996486530624,
+      "learning_rate": 8.923023817742492e-06,
+      "loss": 0.2912,
+      "step": 2702
+    },
+    {
+      "epoch": 2.5191053122087603,
+      "grad_norm": 0.21779852085720006,
+      "learning_rate": 8.905764584052468e-06,
+      "loss": 0.2566,
+      "step": 2703
+    },
+    {
+      "epoch": 2.520037278657968,
+      "grad_norm": 0.2058727653962132,
+      "learning_rate": 8.888505350362444e-06,
+      "loss": 0.268,
+      "step": 2704
+    },
+    {
+      "epoch": 2.5209692451071763,
+      "grad_norm": 0.22757340222781147,
+      "learning_rate": 8.87124611667242e-06,
+      "loss": 0.3034,
+      "step": 2705
+    },
+    {
+      "epoch": 2.5219012115563837,
+      "grad_norm": 0.20726195473005596,
+      "learning_rate": 8.853986882982396e-06,
+      "loss": 0.2798,
+      "step": 2706
+    },
+    {
+      "epoch": 2.522833178005592,
+      "grad_norm": 0.2115700485288949,
+      "learning_rate": 8.83672764929237e-06,
+      "loss": 0.2789,
+      "step": 2707
+    },
+    {
+      "epoch": 2.5237651444547997,
+      "grad_norm": 0.2148877710672651,
+      "learning_rate": 8.819468415602347e-06,
+      "loss": 0.2846,
+      "step": 2708
+    },
+    {
+      "epoch": 2.5246971109040075,
+      "grad_norm": 0.20912973245617314,
+      "learning_rate": 8.802209181912323e-06,
+      "loss": 0.2733,
+      "step": 2709
+    },
+    {
+      "epoch": 2.5256290773532153,
+      "grad_norm": 0.21489314208425703,
+      "learning_rate": 8.7849499482223e-06,
+      "loss": 0.277,
+      "step": 2710
+    },
+    {
+      "epoch": 2.526561043802423,
+      "grad_norm": 0.23060395519155458,
+      "learning_rate": 8.767690714532276e-06,
+      "loss": 0.2754,
+      "step": 2711
+    },
+    {
+      "epoch": 2.527493010251631,
+      "grad_norm": 0.20870136284360685,
+      "learning_rate": 8.75043148084225e-06,
+      "loss": 0.2687,
+      "step": 2712
+    },
+    {
+      "epoch": 2.5284249767008387,
+      "grad_norm": 0.19456755957397412,
+      "learning_rate": 8.733172247152227e-06,
+      "loss": 0.2746,
+      "step": 2713
+    },
+    {
+      "epoch": 2.5293569431500464,
+      "grad_norm": 0.21637797410787535,
+      "learning_rate": 8.715913013462203e-06,
+      "loss": 0.2745,
+      "step": 2714
+    },
+    {
+      "epoch": 2.5302889095992542,
+      "grad_norm": 0.20040830619444194,
+      "learning_rate": 8.698653779772179e-06,
+      "loss": 0.2841,
+      "step": 2715
+    },
+    {
+      "epoch": 2.5312208760484625,
+      "grad_norm": 0.20404396611366107,
+      "learning_rate": 8.681394546082154e-06,
+      "loss": 0.2724,
+      "step": 2716
+    },
+    {
+      "epoch": 2.53215284249767,
+      "grad_norm": 0.22846289467525688,
+      "learning_rate": 8.66413531239213e-06,
+      "loss": 0.2956,
+      "step": 2717
+    },
+    {
+      "epoch": 2.533084808946878,
+      "grad_norm": 0.21584498012273726,
+      "learning_rate": 8.646876078702105e-06,
+      "loss": 0.2924,
+      "step": 2718
+    },
+    {
+      "epoch": 2.534016775396086,
+      "grad_norm": 0.21268766948151055,
+      "learning_rate": 8.62961684501208e-06,
+      "loss": 0.2813,
+      "step": 2719
+    },
+    {
+      "epoch": 2.5349487418452936,
+      "grad_norm": 0.2215173681709518,
+      "learning_rate": 8.612357611322058e-06,
+      "loss": 0.2761,
+      "step": 2720
+    },
+    {
+      "epoch": 2.5358807082945014,
+      "grad_norm": 0.2144095447953658,
+      "learning_rate": 8.595098377632034e-06,
+      "loss": 0.2578,
+      "step": 2721
+    },
+    {
+      "epoch": 2.536812674743709,
+      "grad_norm": 0.24543319777907335,
+      "learning_rate": 8.57783914394201e-06,
+      "loss": 0.2751,
+      "step": 2722
+    },
+    {
+      "epoch": 2.537744641192917,
+      "grad_norm": 0.21150633932217341,
+      "learning_rate": 8.560579910251985e-06,
+      "loss": 0.2651,
+      "step": 2723
+    },
+    {
+      "epoch": 2.5386766076421248,
+      "grad_norm": 0.22466270263453472,
+      "learning_rate": 8.54332067656196e-06,
+      "loss": 0.2647,
+      "step": 2724
+    },
+    {
+      "epoch": 2.5396085740913326,
+      "grad_norm": 0.22459455347416504,
+      "learning_rate": 8.526061442871937e-06,
+      "loss": 0.2717,
+      "step": 2725
+    },
+    {
+      "epoch": 2.5405405405405403,
+      "grad_norm": 0.21037259192385416,
+      "learning_rate": 8.508802209181913e-06,
+      "loss": 0.2806,
+      "step": 2726
+    },
+    {
+      "epoch": 2.5414725069897486,
+      "grad_norm": 0.2195057227607386,
+      "learning_rate": 8.491542975491889e-06,
+      "loss": 0.2793,
+      "step": 2727
+    },
+    {
+      "epoch": 2.542404473438956,
+      "grad_norm": 0.20231064438924493,
+      "learning_rate": 8.474283741801865e-06,
+      "loss": 0.2653,
+      "step": 2728
+    },
+    {
+      "epoch": 2.543336439888164,
+      "grad_norm": 0.21717565771195438,
+      "learning_rate": 8.457024508111839e-06,
+      "loss": 0.2827,
+      "step": 2729
+    },
+    {
+      "epoch": 2.544268406337372,
+      "grad_norm": 0.20648730241950042,
+      "learning_rate": 8.439765274421817e-06,
+      "loss": 0.2621,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5452003727865797,
+      "grad_norm": 0.2230704149646784,
+      "learning_rate": 8.422506040731793e-06,
+      "loss": 0.2809,
+      "step": 2731
+    },
+    {
+      "epoch": 2.5461323392357875,
+      "grad_norm": 0.22458618569722102,
+      "learning_rate": 8.405246807041769e-06,
+      "loss": 0.2683,
+      "step": 2732
+    },
+    {
+      "epoch": 2.5470643056849953,
+      "grad_norm": 0.23616713672989262,
+      "learning_rate": 8.387987573351745e-06,
+      "loss": 0.2762,
+      "step": 2733
+    },
+    {
+      "epoch": 2.547996272134203,
+      "grad_norm": 0.1941190351080015,
+      "learning_rate": 8.370728339661719e-06,
+      "loss": 0.28,
+      "step": 2734
+    },
+    {
+      "epoch": 2.548928238583411,
+      "grad_norm": 0.21716570501039592,
+      "learning_rate": 8.353469105971695e-06,
+      "loss": 0.2735,
+      "step": 2735
+    },
+    {
+      "epoch": 2.5498602050326187,
+      "grad_norm": 0.23553859923711848,
+      "learning_rate": 8.33620987228167e-06,
+      "loss": 0.2614,
+      "step": 2736
+    },
+    {
+      "epoch": 2.5507921714818265,
+      "grad_norm": 0.20742690619307136,
+      "learning_rate": 8.318950638591647e-06,
+      "loss": 0.2764,
+      "step": 2737
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 0.2161292702518825,
+      "learning_rate": 8.301691404901623e-06,
+      "loss": 0.2695,
+      "step": 2738
+    },
+    {
+      "epoch": 2.552656104380242,
+      "grad_norm": 0.22085335766997993,
+      "learning_rate": 8.284432171211599e-06,
+      "loss": 0.2759,
+      "step": 2739
+    },
+    {
+      "epoch": 2.5535880708294503,
+      "grad_norm": 0.24199209939684407,
+      "learning_rate": 8.267172937521575e-06,
+      "loss": 0.2737,
+      "step": 2740
+    },
+    {
+      "epoch": 2.554520037278658,
+      "grad_norm": 0.21864911221580988,
+      "learning_rate": 8.24991370383155e-06,
+      "loss": 0.273,
+      "step": 2741
+    },
+    {
+      "epoch": 2.555452003727866,
+      "grad_norm": 0.19802545172032554,
+      "learning_rate": 8.232654470141527e-06,
+      "loss": 0.26,
+      "step": 2742
+    },
+    {
+      "epoch": 2.5563839701770736,
+      "grad_norm": 0.22764331677047953,
+      "learning_rate": 8.215395236451503e-06,
+      "loss": 0.291,
+      "step": 2743
+    },
+    {
+      "epoch": 2.5573159366262814,
+      "grad_norm": 0.23000067292702803,
+      "learning_rate": 8.198136002761479e-06,
+      "loss": 0.2802,
+      "step": 2744
+    },
+    {
+      "epoch": 2.558247903075489,
+      "grad_norm": 0.2216200126936605,
+      "learning_rate": 8.180876769071453e-06,
+      "loss": 0.2732,
+      "step": 2745
+    },
+    {
+      "epoch": 2.559179869524697,
+      "grad_norm": 0.2075502119235701,
+      "learning_rate": 8.163617535381429e-06,
+      "loss": 0.2772,
+      "step": 2746
+    },
+    {
+      "epoch": 2.560111835973905,
+      "grad_norm": 0.19257455642471277,
+      "learning_rate": 8.146358301691405e-06,
+      "loss": 0.2638,
+      "step": 2747
+    },
+    {
+      "epoch": 2.5610438024231126,
+      "grad_norm": 0.22248148750143212,
+      "learning_rate": 8.129099068001381e-06,
+      "loss": 0.2886,
+      "step": 2748
+    },
+    {
+      "epoch": 2.561975768872321,
+      "grad_norm": 0.2088187042017315,
+      "learning_rate": 8.111839834311357e-06,
+      "loss": 0.2771,
+      "step": 2749
+    },
+    {
+      "epoch": 2.562907735321528,
+      "grad_norm": 0.20324864810211624,
+      "learning_rate": 8.094580600621333e-06,
+      "loss": 0.2713,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5638397017707364,
+      "grad_norm": 0.22694885485739077,
+      "learning_rate": 8.077321366931309e-06,
+      "loss": 0.2714,
+      "step": 2751
+    },
+    {
+      "epoch": 2.564771668219944,
+      "grad_norm": 0.2144477297637796,
+      "learning_rate": 8.060062133241285e-06,
+      "loss": 0.2741,
+      "step": 2752
+    },
+    {
+      "epoch": 2.565703634669152,
+      "grad_norm": 0.21943693313146392,
+      "learning_rate": 8.04280289955126e-06,
+      "loss": 0.2833,
+      "step": 2753
+    },
+    {
+      "epoch": 2.5666356011183598,
+      "grad_norm": 0.21391232020340253,
+      "learning_rate": 8.025543665861237e-06,
+      "loss": 0.2822,
+      "step": 2754
+    },
+    {
+      "epoch": 2.5675675675675675,
+      "grad_norm": 0.24245496528481822,
+      "learning_rate": 8.008284432171211e-06,
+      "loss": 0.2714,
+      "step": 2755
+    },
+    {
+      "epoch": 2.5684995340167753,
+      "grad_norm": 0.22907378935819017,
+      "learning_rate": 7.991025198481187e-06,
+      "loss": 0.2835,
+      "step": 2756
+    },
+    {
+      "epoch": 2.569431500465983,
+      "grad_norm": 2.1136532832371993,
+      "learning_rate": 7.973765964791163e-06,
+      "loss": 0.2854,
+      "step": 2757
+    },
+    {
+      "epoch": 2.570363466915191,
+      "grad_norm": 0.21377932516276044,
+      "learning_rate": 7.956506731101139e-06,
+      "loss": 0.2776,
+      "step": 2758
+    },
+    {
+      "epoch": 2.5712954333643987,
+      "grad_norm": 0.22895134420443236,
+      "learning_rate": 7.939247497411115e-06,
+      "loss": 0.2714,
+      "step": 2759
+    },
+    {
+      "epoch": 2.572227399813607,
+      "grad_norm": 0.2450693852642098,
+      "learning_rate": 7.921988263721091e-06,
+      "loss": 0.3005,
+      "step": 2760
+    },
+    {
+      "epoch": 2.5731593662628143,
+      "grad_norm": 0.20805403529884697,
+      "learning_rate": 7.904729030031067e-06,
+      "loss": 0.2893,
+      "step": 2761
+    },
+    {
+      "epoch": 2.5740913327120225,
+      "grad_norm": 0.2107342775183357,
+      "learning_rate": 7.887469796341043e-06,
+      "loss": 0.2768,
+      "step": 2762
+    },
+    {
+      "epoch": 2.5750232991612303,
+      "grad_norm": 0.20617573703638944,
+      "learning_rate": 7.870210562651019e-06,
+      "loss": 0.2825,
+      "step": 2763
+    },
+    {
+      "epoch": 2.575955265610438,
+      "grad_norm": 0.22152743780582168,
+      "learning_rate": 7.852951328960995e-06,
+      "loss": 0.2821,
+      "step": 2764
+    },
+    {
+      "epoch": 2.576887232059646,
+      "grad_norm": 0.23452653135231494,
+      "learning_rate": 7.835692095270971e-06,
+      "loss": 0.2818,
+      "step": 2765
+    },
+    {
+      "epoch": 2.5778191985088537,
+      "grad_norm": 0.21614207308053257,
+      "learning_rate": 7.818432861580945e-06,
+      "loss": 0.2762,
+      "step": 2766
+    },
+    {
+      "epoch": 2.5787511649580614,
+      "grad_norm": 0.20705864001506574,
+      "learning_rate": 7.801173627890921e-06,
+      "loss": 0.261,
+      "step": 2767
+    },
+    {
+      "epoch": 2.5796831314072692,
+      "grad_norm": 0.20246247919828772,
+      "learning_rate": 7.783914394200897e-06,
+      "loss": 0.2691,
+      "step": 2768
+    },
+    {
+      "epoch": 2.580615097856477,
+      "grad_norm": 0.2128461839613265,
+      "learning_rate": 7.766655160510873e-06,
+      "loss": 0.2883,
+      "step": 2769
+    },
+    {
+      "epoch": 2.581547064305685,
+      "grad_norm": 0.20284152623954307,
+      "learning_rate": 7.749395926820851e-06,
+      "loss": 0.2546,
+      "step": 2770
+    },
+    {
+      "epoch": 2.582479030754893,
+      "grad_norm": 0.22886651054214505,
+      "learning_rate": 7.732136693130825e-06,
+      "loss": 0.273,
+      "step": 2771
+    },
+    {
+      "epoch": 2.5834109972041004,
+      "grad_norm": 0.20311797286533761,
+      "learning_rate": 7.714877459440801e-06,
+      "loss": 0.2659,
+      "step": 2772
+    },
+    {
+      "epoch": 2.5843429636533086,
+      "grad_norm": 0.22002971901164645,
+      "learning_rate": 7.697618225750777e-06,
+      "loss": 0.2833,
+      "step": 2773
+    },
+    {
+      "epoch": 2.5852749301025164,
+      "grad_norm": 0.2065909373952515,
+      "learning_rate": 7.680358992060753e-06,
+      "loss": 0.2673,
+      "step": 2774
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 0.23212702223927098,
+      "learning_rate": 7.663099758370729e-06,
+      "loss": 0.2781,
+      "step": 2775
+    },
+    {
+      "epoch": 2.587138863000932,
+      "grad_norm": 0.19696655632380453,
+      "learning_rate": 7.645840524680705e-06,
+      "loss": 0.2669,
+      "step": 2776
+    },
+    {
+      "epoch": 2.5880708294501398,
+      "grad_norm": 0.20909243367226235,
+      "learning_rate": 7.62858129099068e-06,
+      "loss": 0.2714,
+      "step": 2777
+    },
+    {
+      "epoch": 2.5890027958993476,
+      "grad_norm": 0.23808556053310878,
+      "learning_rate": 7.611322057300656e-06,
+      "loss": 0.2877,
+      "step": 2778
+    },
+    {
+      "epoch": 2.5899347623485554,
+      "grad_norm": 0.21947804978183177,
+      "learning_rate": 7.594062823610631e-06,
+      "loss": 0.2927,
+      "step": 2779
+    },
+    {
+      "epoch": 2.590866728797763,
+      "grad_norm": 0.21795085758970797,
+      "learning_rate": 7.576803589920608e-06,
+      "loss": 0.2914,
+      "step": 2780
+    },
+    {
+      "epoch": 2.591798695246971,
+      "grad_norm": 0.19842544338413293,
+      "learning_rate": 7.559544356230584e-06,
+      "loss": 0.2744,
+      "step": 2781
+    },
+    {
+      "epoch": 2.592730661696179,
+      "grad_norm": 0.24313270951960028,
+      "learning_rate": 7.54228512254056e-06,
+      "loss": 0.2892,
+      "step": 2782
+    },
+    {
+      "epoch": 2.5936626281453865,
+      "grad_norm": 0.20344526746096978,
+      "learning_rate": 7.525025888850535e-06,
+      "loss": 0.2739,
+      "step": 2783
+    },
+    {
+      "epoch": 2.5945945945945947,
+      "grad_norm": 0.19614945317128898,
+      "learning_rate": 7.507766655160511e-06,
+      "loss": 0.2655,
+      "step": 2784
+    },
+    {
+      "epoch": 2.5955265610438025,
+      "grad_norm": 0.19821311157970162,
+      "learning_rate": 7.490507421470487e-06,
+      "loss": 0.2748,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5964585274930103,
+      "grad_norm": 0.22235337450606354,
+      "learning_rate": 7.473248187780462e-06,
+      "loss": 0.2921,
+      "step": 2786
+    },
+    {
+      "epoch": 2.597390493942218,
+      "grad_norm": 0.21158920351882485,
+      "learning_rate": 7.455988954090438e-06,
+      "loss": 0.2773,
+      "step": 2787
+    },
+    {
+      "epoch": 2.598322460391426,
+      "grad_norm": 0.2086091007016971,
+      "learning_rate": 7.438729720400414e-06,
+      "loss": 0.2925,
+      "step": 2788
+    },
+    {
+      "epoch": 2.5992544268406337,
+      "grad_norm": 0.20099999445901862,
+      "learning_rate": 7.4214704867103895e-06,
+      "loss": 0.279,
+      "step": 2789
+    },
+    {
+      "epoch": 2.6001863932898415,
+      "grad_norm": 0.19201466888757568,
+      "learning_rate": 7.4042112530203655e-06,
+      "loss": 0.264,
+      "step": 2790
+    },
+    {
+      "epoch": 2.6011183597390493,
+      "grad_norm": 0.20919103079119547,
+      "learning_rate": 7.386952019330342e-06,
+      "loss": 0.2956,
+      "step": 2791
+    },
+    {
+      "epoch": 2.602050326188257,
+      "grad_norm": 0.20196915711068003,
+      "learning_rate": 7.369692785640318e-06,
+      "loss": 0.2763,
+      "step": 2792
+    },
+    {
+      "epoch": 2.6029822926374653,
+      "grad_norm": 0.21436724481226188,
+      "learning_rate": 7.352433551950294e-06,
+      "loss": 0.2805,
+      "step": 2793
+    },
+    {
+      "epoch": 2.6039142590866726,
+      "grad_norm": 0.20327528636980868,
+      "learning_rate": 7.335174318260269e-06,
+      "loss": 0.2769,
+      "step": 2794
+    },
+    {
+      "epoch": 2.604846225535881,
+      "grad_norm": 0.21717747626927086,
+      "learning_rate": 7.317915084570245e-06,
+      "loss": 0.2777,
+      "step": 2795
+    },
+    {
+      "epoch": 2.6057781919850886,
+      "grad_norm": 0.2054842790725039,
+      "learning_rate": 7.300655850880221e-06,
+      "loss": 0.2625,
+      "step": 2796
+    },
+    {
+      "epoch": 2.6067101584342964,
+      "grad_norm": 0.21201397324385327,
+      "learning_rate": 7.2833966171901965e-06,
+      "loss": 0.2725,
+      "step": 2797
+    },
+    {
+      "epoch": 2.607642124883504,
+      "grad_norm": 0.21874113377213975,
+      "learning_rate": 7.2661373835001725e-06,
+      "loss": 0.272,
+      "step": 2798
+    },
+    {
+      "epoch": 2.608574091332712,
+      "grad_norm": 0.2138528528470788,
+      "learning_rate": 7.2488781498101485e-06,
+      "loss": 0.2793,
+      "step": 2799
+    },
+    {
+      "epoch": 2.60950605778192,
+      "grad_norm": 0.20958851363691539,
+      "learning_rate": 7.231618916120124e-06,
+      "loss": 0.2783,
+      "step": 2800
+    },
+    {
+      "epoch": 2.6104380242311276,
+      "grad_norm": 0.23130714488928636,
+      "learning_rate": 7.214359682430101e-06,
+      "loss": 0.2822,
+      "step": 2801
+    },
+    {
+      "epoch": 2.6113699906803354,
+      "grad_norm": 0.2225916014317964,
+      "learning_rate": 7.1971004487400764e-06,
+      "loss": 0.2748,
+      "step": 2802
+    },
+    {
+      "epoch": 2.612301957129543,
+      "grad_norm": 0.22159328436065912,
+      "learning_rate": 7.179841215050052e-06,
+      "loss": 0.2872,
+      "step": 2803
+    },
+    {
+      "epoch": 2.6132339235787514,
+      "grad_norm": 0.20211096136625262,
+      "learning_rate": 7.162581981360028e-06,
+      "loss": 0.2848,
+      "step": 2804
+    },
+    {
+      "epoch": 2.6141658900279587,
+      "grad_norm": 0.19718585121531046,
+      "learning_rate": 7.1453227476700035e-06,
+      "loss": 0.2774,
+      "step": 2805
+    },
+    {
+      "epoch": 2.615097856477167,
+      "grad_norm": 0.21808364046416315,
+      "learning_rate": 7.1280635139799795e-06,
+      "loss": 0.279,
+      "step": 2806
+    },
+    {
+      "epoch": 2.6160298229263748,
+      "grad_norm": 0.19109873262860916,
+      "learning_rate": 7.1108042802899555e-06,
+      "loss": 0.2621,
+      "step": 2807
+    },
+    {
+      "epoch": 2.6169617893755825,
+      "grad_norm": 0.21450978005873053,
+      "learning_rate": 7.093545046599931e-06,
+      "loss": 0.2736,
+      "step": 2808
+    },
+    {
+      "epoch": 2.6178937558247903,
+      "grad_norm": 0.20856987939864483,
+      "learning_rate": 7.076285812909907e-06,
+      "loss": 0.2809,
+      "step": 2809
+    },
+    {
+      "epoch": 2.618825722273998,
+      "grad_norm": 0.22004692146814317,
+      "learning_rate": 7.059026579219883e-06,
+      "loss": 0.2706,
+      "step": 2810
+    },
+    {
+      "epoch": 2.619757688723206,
+      "grad_norm": 0.2106708173991987,
+      "learning_rate": 7.0417673455298594e-06,
+      "loss": 0.2805,
+      "step": 2811
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 0.19924420823426925,
+      "learning_rate": 7.024508111839835e-06,
+      "loss": 0.2777,
+      "step": 2812
+    },
+    {
+      "epoch": 2.6216216216216215,
+      "grad_norm": 0.20378844743735533,
+      "learning_rate": 7.0072488781498106e-06,
+      "loss": 0.2874,
+      "step": 2813
+    },
+    {
+      "epoch": 2.6225535880708293,
+      "grad_norm": 0.19777723935242877,
+      "learning_rate": 6.9899896444597865e-06,
+      "loss": 0.2829,
+      "step": 2814
+    },
+    {
+      "epoch": 2.6234855545200375,
+      "grad_norm": 0.20531469180385806,
+      "learning_rate": 6.9727304107697625e-06,
+      "loss": 0.2573,
+      "step": 2815
+    },
+    {
+      "epoch": 2.624417520969245,
+      "grad_norm": 0.20681309075133242,
+      "learning_rate": 6.955471177079738e-06,
+      "loss": 0.2828,
+      "step": 2816
+    },
+    {
+      "epoch": 2.625349487418453,
+      "grad_norm": 0.2248717476134633,
+      "learning_rate": 6.938211943389714e-06,
+      "loss": 0.2872,
+      "step": 2817
+    },
+    {
+      "epoch": 2.626281453867661,
+      "grad_norm": 0.21591703771709542,
+      "learning_rate": 6.92095270969969e-06,
+      "loss": 0.2888,
+      "step": 2818
+    },
+    {
+      "epoch": 2.6272134203168687,
+      "grad_norm": 0.2101188094775807,
+      "learning_rate": 6.903693476009665e-06,
+      "loss": 0.2808,
+      "step": 2819
+    },
+    {
+      "epoch": 2.6281453867660765,
+      "grad_norm": 0.23238893549927228,
+      "learning_rate": 6.886434242319641e-06,
+      "loss": 0.2847,
+      "step": 2820
+    },
+    {
+      "epoch": 2.6290773532152842,
+      "grad_norm": 0.21008885917564676,
+      "learning_rate": 6.869175008629618e-06,
+      "loss": 0.2894,
+      "step": 2821
+    },
+    {
+      "epoch": 2.630009319664492,
+      "grad_norm": 0.21455575449255052,
+      "learning_rate": 6.8519157749395936e-06,
+      "loss": 0.2989,
+      "step": 2822
+    },
+    {
+      "epoch": 2.6309412861137,
+      "grad_norm": 0.19964106060927753,
+      "learning_rate": 6.8346565412495696e-06,
+      "loss": 0.2766,
+      "step": 2823
+    },
+    {
+      "epoch": 2.6318732525629076,
+      "grad_norm": 0.2114586627954355,
+      "learning_rate": 6.817397307559545e-06,
+      "loss": 0.2756,
+      "step": 2824
+    },
+    {
+      "epoch": 2.6328052190121154,
+      "grad_norm": 0.23213506675591042,
+      "learning_rate": 6.800138073869521e-06,
+      "loss": 0.2715,
+      "step": 2825
+    },
+    {
+      "epoch": 2.6337371854613236,
+      "grad_norm": 0.19248060349979637,
+      "learning_rate": 6.782878840179497e-06,
+      "loss": 0.2707,
+      "step": 2826
+    },
+    {
+      "epoch": 2.634669151910531,
+      "grad_norm": 0.21037685755636668,
+      "learning_rate": 6.765619606489472e-06,
+      "loss": 0.2886,
+      "step": 2827
+    },
+    {
+      "epoch": 2.635601118359739,
+      "grad_norm": 0.21852831874352333,
+      "learning_rate": 6.748360372799448e-06,
+      "loss": 0.2667,
+      "step": 2828
+    },
+    {
+      "epoch": 2.636533084808947,
+      "grad_norm": 0.21685605470024993,
+      "learning_rate": 6.731101139109424e-06,
+      "loss": 0.2815,
+      "step": 2829
+    },
+    {
+      "epoch": 2.637465051258155,
+      "grad_norm": 0.20904321881015756,
+      "learning_rate": 6.713841905419399e-06,
+      "loss": 0.2749,
+      "step": 2830
+    },
+    {
+      "epoch": 2.6383970177073626,
+      "grad_norm": 0.21674312988624175,
+      "learning_rate": 6.696582671729376e-06,
+      "loss": 0.2775,
+      "step": 2831
+    },
+    {
+      "epoch": 2.6393289841565704,
+      "grad_norm": 0.2224141186552288,
+      "learning_rate": 6.679323438039352e-06,
+      "loss": 0.2681,
+      "step": 2832
+    },
+    {
+      "epoch": 2.640260950605778,
+      "grad_norm": 0.2093651933734543,
+      "learning_rate": 6.662064204349328e-06,
+      "loss": 0.282,
+      "step": 2833
+    },
+    {
+      "epoch": 2.641192917054986,
+      "grad_norm": 0.20364710488413748,
+      "learning_rate": 6.644804970659303e-06,
+      "loss": 0.2786,
+      "step": 2834
+    },
+    {
+      "epoch": 2.6421248835041937,
+      "grad_norm": 0.23861973373774792,
+      "learning_rate": 6.627545736969279e-06,
+      "loss": 0.3011,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6430568499534015,
+      "grad_norm": 0.20052711501327342,
+      "learning_rate": 6.610286503279255e-06,
+      "loss": 0.2572,
+      "step": 2836
+    },
+    {
+      "epoch": 2.6439888164026097,
+      "grad_norm": 0.21516247475785436,
+      "learning_rate": 6.59302726958923e-06,
+      "loss": 0.2833,
+      "step": 2837
+    },
+    {
+      "epoch": 2.644920782851817,
+      "grad_norm": 0.20831441314688956,
+      "learning_rate": 6.575768035899206e-06,
+      "loss": 0.2827,
+      "step": 2838
+    },
+    {
+      "epoch": 2.6458527493010253,
+      "grad_norm": 0.19502887362975393,
+      "learning_rate": 6.558508802209182e-06,
+      "loss": 0.2876,
+      "step": 2839
+    },
+    {
+      "epoch": 2.646784715750233,
+      "grad_norm": 0.20191955021802513,
+      "learning_rate": 6.541249568519157e-06,
+      "loss": 0.2742,
+      "step": 2840
+    },
+    {
+      "epoch": 2.647716682199441,
+      "grad_norm": 0.20324158893297756,
+      "learning_rate": 6.523990334829135e-06,
+      "loss": 0.2689,
+      "step": 2841
+    },
+    {
+      "epoch": 2.6486486486486487,
+      "grad_norm": 0.2045094126473417,
+      "learning_rate": 6.50673110113911e-06,
+      "loss": 0.2698,
+      "step": 2842
+    },
+    {
+      "epoch": 2.6495806150978565,
+      "grad_norm": 0.22246456913887397,
+      "learning_rate": 6.489471867449086e-06,
+      "loss": 0.2866,
+      "step": 2843
+    },
+    {
+      "epoch": 2.6505125815470643,
+      "grad_norm": 0.20446893183898557,
+      "learning_rate": 6.472212633759062e-06,
+      "loss": 0.2714,
+      "step": 2844
+    },
+    {
+      "epoch": 2.651444547996272,
+      "grad_norm": 0.2102754237358842,
+      "learning_rate": 6.454953400069037e-06,
+      "loss": 0.2649,
+      "step": 2845
+    },
+    {
+      "epoch": 2.65237651444548,
+      "grad_norm": 0.2027576221342155,
+      "learning_rate": 6.437694166379013e-06,
+      "loss": 0.2672,
+      "step": 2846
+    },
+    {
+      "epoch": 2.6533084808946876,
+      "grad_norm": 0.20351614898850073,
+      "learning_rate": 6.420434932688989e-06,
+      "loss": 0.2885,
+      "step": 2847
+    },
+    {
+      "epoch": 2.654240447343896,
+      "grad_norm": 0.19756081750992616,
+      "learning_rate": 6.403175698998964e-06,
+      "loss": 0.2765,
+      "step": 2848
+    },
+    {
+      "epoch": 2.655172413793103,
+      "grad_norm": 0.20357795076563726,
+      "learning_rate": 6.38591646530894e-06,
+      "loss": 0.2744,
+      "step": 2849
+    },
+    {
+      "epoch": 2.6561043802423114,
+      "grad_norm": 0.21554117304753453,
+      "learning_rate": 6.368657231618916e-06,
+      "loss": 0.3001,
+      "step": 2850
+    },
+    {
+      "epoch": 2.6570363466915192,
+      "grad_norm": 0.20948610694476877,
+      "learning_rate": 6.351397997928893e-06,
+      "loss": 0.2796,
+      "step": 2851
+    },
+    {
+      "epoch": 2.657968313140727,
+      "grad_norm": 0.20580536628164067,
+      "learning_rate": 6.334138764238869e-06,
+      "loss": 0.2716,
+      "step": 2852
+    },
+    {
+      "epoch": 2.658900279589935,
+      "grad_norm": 0.21110281506172893,
+      "learning_rate": 6.316879530548844e-06,
+      "loss": 0.2741,
+      "step": 2853
+    },
+    {
+      "epoch": 2.6598322460391426,
+      "grad_norm": 0.19923797148618433,
+      "learning_rate": 6.29962029685882e-06,
+      "loss": 0.2712,
+      "step": 2854
+    },
+    {
+      "epoch": 2.6607642124883504,
+      "grad_norm": 0.20482139784350567,
+      "learning_rate": 6.282361063168796e-06,
+      "loss": 0.2792,
+      "step": 2855
+    },
+    {
+      "epoch": 2.661696178937558,
+      "grad_norm": 0.19881830527007102,
+      "learning_rate": 6.265101829478771e-06,
+      "loss": 0.2791,
+      "step": 2856
+    },
+    {
+      "epoch": 2.662628145386766,
+      "grad_norm": 0.23505977958182145,
+      "learning_rate": 6.247842595788747e-06,
+      "loss": 0.2964,
+      "step": 2857
+    },
+    {
+      "epoch": 2.6635601118359737,
+      "grad_norm": 0.20629115333672896,
+      "learning_rate": 6.230583362098723e-06,
+      "loss": 0.2666,
+      "step": 2858
+    },
+    {
+      "epoch": 2.664492078285182,
+      "grad_norm": 0.21862847651004577,
+      "learning_rate": 6.213324128408699e-06,
+      "loss": 0.2769,
+      "step": 2859
+    },
+    {
+      "epoch": 2.6654240447343893,
+      "grad_norm": 0.23034684560212373,
+      "learning_rate": 6.196064894718675e-06,
+      "loss": 0.2945,
+      "step": 2860
+    },
+    {
+      "epoch": 2.6663560111835976,
+      "grad_norm": 0.20328334170477289,
+      "learning_rate": 6.17880566102865e-06,
+      "loss": 0.2663,
+      "step": 2861
+    },
+    {
+      "epoch": 2.6672879776328053,
+      "grad_norm": 0.221944528072678,
+      "learning_rate": 6.161546427338626e-06,
+      "loss": 0.3047,
+      "step": 2862
+    },
+    {
+      "epoch": 2.668219944082013,
+      "grad_norm": 0.20377686854997362,
+      "learning_rate": 6.144287193648602e-06,
+      "loss": 0.2735,
+      "step": 2863
+    },
+    {
+      "epoch": 2.669151910531221,
+      "grad_norm": 0.21075235330134176,
+      "learning_rate": 6.127027959958578e-06,
+      "loss": 0.2652,
+      "step": 2864
+    },
+    {
+      "epoch": 2.6700838769804287,
+      "grad_norm": 0.21042608526898945,
+      "learning_rate": 6.109768726268554e-06,
+      "loss": 0.2753,
+      "step": 2865
+    },
+    {
+      "epoch": 2.6710158434296365,
+      "grad_norm": 0.21178702695650492,
+      "learning_rate": 6.09250949257853e-06,
+      "loss": 0.2778,
+      "step": 2866
+    },
+    {
+      "epoch": 2.6719478098788443,
+      "grad_norm": 0.21944545357023432,
+      "learning_rate": 6.075250258888505e-06,
+      "loss": 0.2903,
+      "step": 2867
+    },
+    {
+      "epoch": 2.672879776328052,
+      "grad_norm": 0.22325553107938373,
+      "learning_rate": 6.057991025198481e-06,
+      "loss": 0.2741,
+      "step": 2868
+    },
+    {
+      "epoch": 2.67381174277726,
+      "grad_norm": 0.25379712579909414,
+      "learning_rate": 6.040731791508457e-06,
+      "loss": 0.2664,
+      "step": 2869
+    },
+    {
+      "epoch": 2.674743709226468,
+      "grad_norm": 0.2067603762513616,
+      "learning_rate": 6.023472557818433e-06,
+      "loss": 0.27,
+      "step": 2870
+    },
+    {
+      "epoch": 2.6756756756756754,
+      "grad_norm": 0.21405033289824424,
+      "learning_rate": 6.006213324128409e-06,
+      "loss": 0.2937,
+      "step": 2871
+    },
+    {
+      "epoch": 2.6766076421248837,
+      "grad_norm": 0.21622745876038438,
+      "learning_rate": 5.988954090438384e-06,
+      "loss": 0.2885,
+      "step": 2872
+    },
+    {
+      "epoch": 2.6775396085740915,
+      "grad_norm": 0.23177857759825385,
+      "learning_rate": 5.97169485674836e-06,
+      "loss": 0.271,
+      "step": 2873
+    },
+    {
+      "epoch": 2.6784715750232992,
+      "grad_norm": 0.2302656841264027,
+      "learning_rate": 5.954435623058337e-06,
+      "loss": 0.2853,
+      "step": 2874
+    },
+    {
+      "epoch": 2.679403541472507,
+      "grad_norm": 0.22504827028701332,
+      "learning_rate": 5.937176389368312e-06,
+      "loss": 0.2953,
+      "step": 2875
+    },
+    {
+      "epoch": 2.680335507921715,
+      "grad_norm": 0.23381773808511785,
+      "learning_rate": 5.919917155678288e-06,
+      "loss": 0.2934,
+      "step": 2876
+    },
+    {
+      "epoch": 2.6812674743709226,
+      "grad_norm": 0.24642167626263606,
+      "learning_rate": 5.902657921988264e-06,
+      "loss": 0.2824,
+      "step": 2877
+    },
+    {
+      "epoch": 2.6821994408201304,
+      "grad_norm": 0.23176295105561076,
+      "learning_rate": 5.885398688298239e-06,
+      "loss": 0.2904,
+      "step": 2878
+    },
+    {
+      "epoch": 2.683131407269338,
+      "grad_norm": 0.21307293684731696,
+      "learning_rate": 5.868139454608216e-06,
+      "loss": 0.2645,
+      "step": 2879
+    },
+    {
+      "epoch": 2.684063373718546,
+      "grad_norm": 0.19260889949958998,
+      "learning_rate": 5.850880220918191e-06,
+      "loss": 0.2777,
+      "step": 2880
+    },
+    {
+      "epoch": 2.684995340167754,
+      "grad_norm": 0.23178368925706822,
+      "learning_rate": 5.833620987228167e-06,
+      "loss": 0.2609,
+      "step": 2881
+    },
+    {
+      "epoch": 2.6859273066169616,
+      "grad_norm": 0.22765084189650836,
+      "learning_rate": 5.816361753538143e-06,
+      "loss": 0.2772,
+      "step": 2882
+    },
+    {
+      "epoch": 2.68685927306617,
+      "grad_norm": 0.21460313416168955,
+      "learning_rate": 5.7991025198481184e-06,
+      "loss": 0.2684,
+      "step": 2883
+    },
+    {
+      "epoch": 2.6877912395153776,
+      "grad_norm": 0.21308879946527,
+      "learning_rate": 5.781843286158095e-06,
+      "loss": 0.2793,
+      "step": 2884
+    },
+    {
+      "epoch": 2.6887232059645854,
+      "grad_norm": 0.22057049061852177,
+      "learning_rate": 5.76458405246807e-06,
+      "loss": 0.2916,
+      "step": 2885
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 0.20798313547904654,
+      "learning_rate": 5.747324818778046e-06,
+      "loss": 0.2788,
+      "step": 2886
+    },
+    {
+      "epoch": 2.690587138863001,
+      "grad_norm": 0.23594413431927727,
+      "learning_rate": 5.730065585088022e-06,
+      "loss": 0.2851,
+      "step": 2887
+    },
+    {
+      "epoch": 2.6915191053122087,
+      "grad_norm": 0.24026251918148359,
+      "learning_rate": 5.7128063513979975e-06,
+      "loss": 0.2812,
+      "step": 2888
+    },
+    {
+      "epoch": 2.6924510717614165,
+      "grad_norm": 0.22371782719972577,
+      "learning_rate": 5.695547117707974e-06,
+      "loss": 0.2759,
+      "step": 2889
+    },
+    {
+      "epoch": 2.6933830382106243,
+      "grad_norm": 0.21556502126871258,
+      "learning_rate": 5.67828788401795e-06,
+      "loss": 0.2812,
+      "step": 2890
+    },
+    {
+      "epoch": 2.694315004659832,
+      "grad_norm": 0.20652813347302015,
+      "learning_rate": 5.6610286503279255e-06,
+      "loss": 0.2815,
+      "step": 2891
+    },
+    {
+      "epoch": 2.6952469711090403,
+      "grad_norm": 0.21339807437232972,
+      "learning_rate": 5.6437694166379015e-06,
+      "loss": 0.2767,
+      "step": 2892
+    },
+    {
+      "epoch": 2.6961789375582477,
+      "grad_norm": 0.1897908728700622,
+      "learning_rate": 5.6265101829478774e-06,
+      "loss": 0.2835,
+      "step": 2893
+    },
+    {
+      "epoch": 2.697110904007456,
+      "grad_norm": 0.20659427975586753,
+      "learning_rate": 5.609250949257853e-06,
+      "loss": 0.2758,
+      "step": 2894
+    },
+    {
+      "epoch": 2.6980428704566637,
+      "grad_norm": 0.21643278470421237,
+      "learning_rate": 5.591991715567829e-06,
+      "loss": 0.2652,
+      "step": 2895
+    },
+    {
+      "epoch": 2.6989748369058715,
+      "grad_norm": 0.21001020428673337,
+      "learning_rate": 5.5747324818778045e-06,
+      "loss": 0.2642,
+      "step": 2896
+    },
+    {
+      "epoch": 2.6999068033550793,
+      "grad_norm": 0.18794513741263227,
+      "learning_rate": 5.5574732481877805e-06,
+      "loss": 0.2638,
+      "step": 2897
+    },
+    {
+      "epoch": 2.700838769804287,
+      "grad_norm": 0.20735743647845314,
+      "learning_rate": 5.5402140144977565e-06,
+      "loss": 0.2797,
+      "step": 2898
+    },
+    {
+      "epoch": 2.701770736253495,
+      "grad_norm": 0.19154345021220165,
+      "learning_rate": 5.5229547808077325e-06,
+      "loss": 0.2646,
+      "step": 2899
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 0.2366701163222664,
+      "learning_rate": 5.5056955471177085e-06,
+      "loss": 0.2882,
+      "step": 2900
+    },
+    {
+      "epoch": 2.7036346691519104,
+      "grad_norm": 0.20521583564048756,
+      "learning_rate": 5.4884363134276845e-06,
+      "loss": 0.2675,
+      "step": 2901
+    },
+    {
+      "epoch": 2.704566635601118,
+      "grad_norm": 0.1914066879272596,
+      "learning_rate": 5.47117707973766e-06,
+      "loss": 0.2693,
+      "step": 2902
+    },
+    {
+      "epoch": 2.7054986020503264,
+      "grad_norm": 0.2002086303213405,
+      "learning_rate": 5.453917846047636e-06,
+      "loss": 0.277,
+      "step": 2903
+    },
+    {
+      "epoch": 2.706430568499534,
+      "grad_norm": 0.21182082585070855,
+      "learning_rate": 5.4366586123576116e-06,
+      "loss": 0.2816,
+      "step": 2904
+    },
+    {
+      "epoch": 2.707362534948742,
+      "grad_norm": 0.2339779147902771,
+      "learning_rate": 5.4193993786675876e-06,
+      "loss": 0.2826,
+      "step": 2905
+    },
+    {
+      "epoch": 2.70829450139795,
+      "grad_norm": 0.22387323911818927,
+      "learning_rate": 5.4021401449775635e-06,
+      "loss": 0.2917,
+      "step": 2906
+    },
+    {
+      "epoch": 2.7092264678471576,
+      "grad_norm": 0.21753531669693651,
+      "learning_rate": 5.384880911287539e-06,
+      "loss": 0.2935,
+      "step": 2907
+    },
+    {
+      "epoch": 2.7101584342963654,
+      "grad_norm": 0.1924410355300424,
+      "learning_rate": 5.367621677597515e-06,
+      "loss": 0.2652,
+      "step": 2908
+    },
+    {
+      "epoch": 2.711090400745573,
+      "grad_norm": 0.19314794748989628,
+      "learning_rate": 5.350362443907491e-06,
+      "loss": 0.2771,
+      "step": 2909
+    },
+    {
+      "epoch": 2.712022367194781,
+      "grad_norm": 0.21056999821147096,
+      "learning_rate": 5.333103210217467e-06,
+      "loss": 0.288,
+      "step": 2910
+    },
+    {
+      "epoch": 2.7129543336439887,
+      "grad_norm": 0.20972949699363344,
+      "learning_rate": 5.315843976527443e-06,
+      "loss": 0.2721,
+      "step": 2911
+    },
+    {
+      "epoch": 2.7138863000931965,
+      "grad_norm": 0.22823049608817128,
+      "learning_rate": 5.298584742837418e-06,
+      "loss": 0.2754,
+      "step": 2912
+    },
+    {
+      "epoch": 2.7148182665424043,
+      "grad_norm": 0.22000574034633255,
+      "learning_rate": 5.281325509147394e-06,
+      "loss": 0.2919,
+      "step": 2913
+    },
+    {
+      "epoch": 2.7157502329916126,
+      "grad_norm": 0.19414671763325686,
+      "learning_rate": 5.2640662754573706e-06,
+      "loss": 0.2585,
+      "step": 2914
+    },
+    {
+      "epoch": 2.71668219944082,
+      "grad_norm": 0.20982886675140178,
+      "learning_rate": 5.246807041767346e-06,
+      "loss": 0.2662,
+      "step": 2915
+    },
+    {
+      "epoch": 2.717614165890028,
+      "grad_norm": 0.2371651715980282,
+      "learning_rate": 5.229547808077322e-06,
+      "loss": 0.2745,
+      "step": 2916
+    },
+    {
+      "epoch": 2.718546132339236,
+      "grad_norm": 0.18751088156795073,
+      "learning_rate": 5.212288574387298e-06,
+      "loss": 0.2524,
+      "step": 2917
+    },
+    {
+      "epoch": 2.7194780987884437,
+      "grad_norm": 0.20703126451868595,
+      "learning_rate": 5.195029340697273e-06,
+      "loss": 0.281,
+      "step": 2918
+    },
+    {
+      "epoch": 2.7204100652376515,
+      "grad_norm": 0.2125404041599179,
+      "learning_rate": 5.17777010700725e-06,
+      "loss": 0.2763,
+      "step": 2919
+    },
+    {
+      "epoch": 2.7213420316868593,
+      "grad_norm": 0.1995135400284231,
+      "learning_rate": 5.160510873317225e-06,
+      "loss": 0.2707,
+      "step": 2920
+    },
+    {
+      "epoch": 2.722273998136067,
+      "grad_norm": 0.1988272148449551,
+      "learning_rate": 5.143251639627201e-06,
+      "loss": 0.2835,
+      "step": 2921
+    },
+    {
+      "epoch": 2.723205964585275,
+      "grad_norm": 0.20695655735443008,
+      "learning_rate": 5.125992405937177e-06,
+      "loss": 0.2756,
+      "step": 2922
+    },
+    {
+      "epoch": 2.7241379310344827,
+      "grad_norm": 0.2007298294235112,
+      "learning_rate": 5.108733172247152e-06,
+      "loss": 0.2588,
+      "step": 2923
+    },
+    {
+      "epoch": 2.7250698974836904,
+      "grad_norm": 0.2257133278763949,
+      "learning_rate": 5.091473938557129e-06,
+      "loss": 0.281,
+      "step": 2924
+    },
+    {
+      "epoch": 2.7260018639328987,
+      "grad_norm": 0.20252863462270573,
+      "learning_rate": 5.074214704867105e-06,
+      "loss": 0.2738,
+      "step": 2925
+    },
+    {
+      "epoch": 2.726933830382106,
+      "grad_norm": 0.19935288982026964,
+      "learning_rate": 5.05695547117708e-06,
+      "loss": 0.2941,
+      "step": 2926
+    },
+    {
+      "epoch": 2.7278657968313142,
+      "grad_norm": 0.2115031019067344,
+      "learning_rate": 5.039696237487056e-06,
+      "loss": 0.2922,
+      "step": 2927
+    },
+    {
+      "epoch": 2.728797763280522,
+      "grad_norm": 0.2095013373483691,
+      "learning_rate": 5.022437003797032e-06,
+      "loss": 0.2762,
+      "step": 2928
+    },
+    {
+      "epoch": 2.72972972972973,
+      "grad_norm": 0.21172604443946438,
+      "learning_rate": 5.005177770107008e-06,
+      "loss": 0.278,
+      "step": 2929
+    },
+    {
+      "epoch": 2.7306616961789376,
+      "grad_norm": 0.215355203226406,
+      "learning_rate": 4.987918536416984e-06,
+      "loss": 0.2782,
+      "step": 2930
+    },
+    {
+      "epoch": 2.7315936626281454,
+      "grad_norm": 0.20825984174678588,
+      "learning_rate": 4.970659302726959e-06,
+      "loss": 0.2729,
+      "step": 2931
+    },
+    {
+      "epoch": 2.732525629077353,
+      "grad_norm": 0.2086643822311563,
+      "learning_rate": 4.953400069036935e-06,
+      "loss": 0.2777,
+      "step": 2932
+    },
+    {
+      "epoch": 2.733457595526561,
+      "grad_norm": 0.202123382135737,
+      "learning_rate": 4.936140835346911e-06,
+      "loss": 0.2752,
+      "step": 2933
+    },
+    {
+      "epoch": 2.7343895619757688,
+      "grad_norm": 0.23016795342595897,
+      "learning_rate": 4.918881601656887e-06,
+      "loss": 0.2825,
+      "step": 2934
+    },
+    {
+      "epoch": 2.7353215284249766,
+      "grad_norm": 0.28673474722428144,
+      "learning_rate": 4.901622367966863e-06,
+      "loss": 0.2616,
+      "step": 2935
+    },
+    {
+      "epoch": 2.736253494874185,
+      "grad_norm": 0.20688426871661805,
+      "learning_rate": 4.884363134276838e-06,
+      "loss": 0.2739,
+      "step": 2936
+    },
+    {
+      "epoch": 2.737185461323392,
+      "grad_norm": 0.22924650436083466,
+      "learning_rate": 4.867103900586814e-06,
+      "loss": 0.2855,
+      "step": 2937
+    },
+    {
+      "epoch": 2.7381174277726004,
+      "grad_norm": 0.21972355096170992,
+      "learning_rate": 4.84984466689679e-06,
+      "loss": 0.2797,
+      "step": 2938
+    },
+    {
+      "epoch": 2.739049394221808,
+      "grad_norm": 0.20096889359408238,
+      "learning_rate": 4.832585433206766e-06,
+      "loss": 0.2738,
+      "step": 2939
+    },
+    {
+      "epoch": 2.739981360671016,
+      "grad_norm": 0.21013739160848327,
+      "learning_rate": 4.815326199516742e-06,
+      "loss": 0.2754,
+      "step": 2940
+    },
+    {
+      "epoch": 2.7409133271202237,
+      "grad_norm": 0.21496333169651988,
+      "learning_rate": 4.798066965826718e-06,
+      "loss": 0.2819,
+      "step": 2941
+    },
+    {
+      "epoch": 2.7418452935694315,
+      "grad_norm": 0.21708865070704778,
+      "learning_rate": 4.780807732136693e-06,
+      "loss": 0.2922,
+      "step": 2942
+    },
+    {
+      "epoch": 2.7427772600186393,
+      "grad_norm": 0.20322701536167048,
+      "learning_rate": 4.763548498446669e-06,
+      "loss": 0.2661,
+      "step": 2943
+    },
+    {
+      "epoch": 2.743709226467847,
+      "grad_norm": 0.20371978682419287,
+      "learning_rate": 4.746289264756645e-06,
+      "loss": 0.2767,
+      "step": 2944
+    },
+    {
+      "epoch": 2.744641192917055,
+      "grad_norm": 0.2070390835159307,
+      "learning_rate": 4.729030031066621e-06,
+      "loss": 0.2669,
+      "step": 2945
+    },
+    {
+      "epoch": 2.7455731593662627,
+      "grad_norm": 0.19447750868954866,
+      "learning_rate": 4.711770797376597e-06,
+      "loss": 0.2833,
+      "step": 2946
+    },
+    {
+      "epoch": 2.746505125815471,
+      "grad_norm": 0.19737336102815292,
+      "learning_rate": 4.694511563686572e-06,
+      "loss": 0.2799,
+      "step": 2947
+    },
+    {
+      "epoch": 2.7474370922646782,
+      "grad_norm": 0.20656443847409453,
+      "learning_rate": 4.677252329996548e-06,
+      "loss": 0.2743,
+      "step": 2948
+    },
+    {
+      "epoch": 2.7483690587138865,
+      "grad_norm": 0.21900662114340755,
+      "learning_rate": 4.659993096306525e-06,
+      "loss": 0.2825,
+      "step": 2949
+    },
+    {
+      "epoch": 2.7493010251630943,
+      "grad_norm": 0.2103631437482066,
+      "learning_rate": 4.6427338626165e-06,
+      "loss": 0.2867,
+      "step": 2950
+    },
+    {
+      "epoch": 2.750232991612302,
+      "grad_norm": 0.19867573276254025,
+      "learning_rate": 4.625474628926476e-06,
+      "loss": 0.2716,
+      "step": 2951
+    },
+    {
+      "epoch": 2.75116495806151,
+      "grad_norm": 0.19640792850993066,
+      "learning_rate": 4.608215395236452e-06,
+      "loss": 0.2571,
+      "step": 2952
+    },
+    {
+      "epoch": 2.7520969245107176,
+      "grad_norm": 0.20316373265613533,
+      "learning_rate": 4.590956161546427e-06,
+      "loss": 0.2947,
+      "step": 2953
+    },
+    {
+      "epoch": 2.7530288909599254,
+      "grad_norm": 0.20448314040909918,
+      "learning_rate": 4.573696927856403e-06,
+      "loss": 0.2632,
+      "step": 2954
+    },
+    {
+      "epoch": 2.753960857409133,
+      "grad_norm": 0.20143478941680484,
+      "learning_rate": 4.556437694166379e-06,
+      "loss": 0.2654,
+      "step": 2955
+    },
+    {
+      "epoch": 2.754892823858341,
+      "grad_norm": 0.1951122652776208,
+      "learning_rate": 4.539178460476355e-06,
+      "loss": 0.2768,
+      "step": 2956
+    },
+    {
+      "epoch": 2.755824790307549,
+      "grad_norm": 0.20937984323000552,
+      "learning_rate": 4.521919226786331e-06,
+      "loss": 0.2728,
+      "step": 2957
+    },
+    {
+      "epoch": 2.756756756756757,
+      "grad_norm": 0.19383953967657408,
+      "learning_rate": 4.504659993096306e-06,
+      "loss": 0.2745,
+      "step": 2958
+    },
+    {
+      "epoch": 2.7576887232059644,
+      "grad_norm": 0.2062412807514913,
+      "learning_rate": 4.487400759406282e-06,
+      "loss": 0.2826,
+      "step": 2959
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 0.19383305432735873,
+      "learning_rate": 4.470141525716258e-06,
+      "loss": 0.2566,
+      "step": 2960
+    },
+    {
+      "epoch": 2.7595526561043804,
+      "grad_norm": 0.19208385622898502,
+      "learning_rate": 4.452882292026234e-06,
+      "loss": 0.2875,
+      "step": 2961
+    },
+    {
+      "epoch": 2.760484622553588,
+      "grad_norm": 0.21310028596316385,
+      "learning_rate": 4.43562305833621e-06,
+      "loss": 0.2809,
+      "step": 2962
+    },
+    {
+      "epoch": 2.761416589002796,
+      "grad_norm": 0.20756963200776127,
+      "learning_rate": 4.418363824646185e-06,
+      "loss": 0.2826,
+      "step": 2963
+    },
+    {
+      "epoch": 2.7623485554520038,
+      "grad_norm": 0.20484262127287356,
+      "learning_rate": 4.401104590956161e-06,
+      "loss": 0.2683,
+      "step": 2964
+    },
+    {
+      "epoch": 2.7632805219012115,
+      "grad_norm": 0.19948412869149035,
+      "learning_rate": 4.383845357266138e-06,
+      "loss": 0.268,
+      "step": 2965
+    },
+    {
+      "epoch": 2.7642124883504193,
+      "grad_norm": 0.20074003063945742,
+      "learning_rate": 4.366586123576113e-06,
+      "loss": 0.2917,
+      "step": 2966
+    },
+    {
+      "epoch": 2.765144454799627,
+      "grad_norm": 0.21978781543080375,
+      "learning_rate": 4.349326889886089e-06,
+      "loss": 0.2689,
+      "step": 2967
+    },
+    {
+      "epoch": 2.766076421248835,
+      "grad_norm": 0.21415622385751984,
+      "learning_rate": 4.332067656196065e-06,
+      "loss": 0.2733,
+      "step": 2968
+    },
+    {
+      "epoch": 2.767008387698043,
+      "grad_norm": 0.20619118140532008,
+      "learning_rate": 4.31480842250604e-06,
+      "loss": 0.2789,
+      "step": 2969
+    },
+    {
+      "epoch": 2.7679403541472505,
+      "grad_norm": 0.21800385570708025,
+      "learning_rate": 4.297549188816017e-06,
+      "loss": 0.2789,
+      "step": 2970
+    },
+    {
+      "epoch": 2.7688723205964587,
+      "grad_norm": 0.20331523029366677,
+      "learning_rate": 4.280289955125992e-06,
+      "loss": 0.2792,
+      "step": 2971
+    },
+    {
+      "epoch": 2.7698042870456665,
+      "grad_norm": 0.19469434775411285,
+      "learning_rate": 4.263030721435968e-06,
+      "loss": 0.276,
+      "step": 2972
+    },
+    {
+      "epoch": 2.7707362534948743,
+      "grad_norm": 0.18788640669314244,
+      "learning_rate": 4.245771487745944e-06,
+      "loss": 0.2649,
+      "step": 2973
+    },
+    {
+      "epoch": 2.771668219944082,
+      "grad_norm": 0.2181217458148137,
+      "learning_rate": 4.2285122540559194e-06,
+      "loss": 0.2803,
+      "step": 2974
+    },
+    {
+      "epoch": 2.77260018639329,
+      "grad_norm": 0.19708028078504955,
+      "learning_rate": 4.211253020365896e-06,
+      "loss": 0.2806,
+      "step": 2975
+    },
+    {
+      "epoch": 2.7735321528424977,
+      "grad_norm": 0.21873268594990125,
+      "learning_rate": 4.193993786675872e-06,
+      "loss": 0.2776,
+      "step": 2976
+    },
+    {
+      "epoch": 2.7744641192917054,
+      "grad_norm": 0.20246815423296302,
+      "learning_rate": 4.176734552985847e-06,
+      "loss": 0.2748,
+      "step": 2977
+    },
+    {
+      "epoch": 2.7753960857409132,
+      "grad_norm": 0.19227584604905182,
+      "learning_rate": 4.159475319295823e-06,
+      "loss": 0.2605,
+      "step": 2978
+    },
+    {
+      "epoch": 2.776328052190121,
+      "grad_norm": 0.19952691733105918,
+      "learning_rate": 4.142216085605799e-06,
+      "loss": 0.2784,
+      "step": 2979
+    },
+    {
+      "epoch": 2.7772600186393293,
+      "grad_norm": 0.2001879137313884,
+      "learning_rate": 4.124956851915775e-06,
+      "loss": 0.2753,
+      "step": 2980
+    },
+    {
+      "epoch": 2.7781919850885366,
+      "grad_norm": 0.19089844903418685,
+      "learning_rate": 4.107697618225751e-06,
+      "loss": 0.2634,
+      "step": 2981
+    },
+    {
+      "epoch": 2.779123951537745,
+      "grad_norm": 0.23612977335238777,
+      "learning_rate": 4.0904383845357265e-06,
+      "loss": 0.2998,
+      "step": 2982
+    },
+    {
+      "epoch": 2.7800559179869526,
+      "grad_norm": 0.2024515920266736,
+      "learning_rate": 4.0731791508457025e-06,
+      "loss": 0.2766,
+      "step": 2983
+    },
+    {
+      "epoch": 2.7809878844361604,
+      "grad_norm": 0.19621088709863435,
+      "learning_rate": 4.0559199171556784e-06,
+      "loss": 0.2858,
+      "step": 2984
+    },
+    {
+      "epoch": 2.781919850885368,
+      "grad_norm": 0.2019443445480945,
+      "learning_rate": 4.0386606834656544e-06,
+      "loss": 0.2808,
+      "step": 2985
+    },
+    {
+      "epoch": 2.782851817334576,
+      "grad_norm": 0.22044237618911713,
+      "learning_rate": 4.02140144977563e-06,
+      "loss": 0.276,
+      "step": 2986
+    },
+    {
+      "epoch": 2.7837837837837838,
+      "grad_norm": 0.21369420648490547,
+      "learning_rate": 4.0041422160856055e-06,
+      "loss": 0.2808,
+      "step": 2987
+    },
+    {
+      "epoch": 2.7847157502329916,
+      "grad_norm": 0.18844997395043311,
+      "learning_rate": 3.9868829823955815e-06,
+      "loss": 0.2669,
+      "step": 2988
+    },
+    {
+      "epoch": 2.7856477166821993,
+      "grad_norm": 0.2156699208085116,
+      "learning_rate": 3.9696237487055575e-06,
+      "loss": 0.2745,
+      "step": 2989
+    },
+    {
+      "epoch": 2.786579683131407,
+      "grad_norm": 0.1956777038216626,
+      "learning_rate": 3.9523645150155335e-06,
+      "loss": 0.2824,
+      "step": 2990
+    },
+    {
+      "epoch": 2.7875116495806154,
+      "grad_norm": 0.181168766664522,
+      "learning_rate": 3.9351052813255095e-06,
+      "loss": 0.2605,
+      "step": 2991
+    },
+    {
+      "epoch": 2.7884436160298227,
+      "grad_norm": 0.1847773775494842,
+      "learning_rate": 3.9178460476354855e-06,
+      "loss": 0.2682,
+      "step": 2992
+    },
+    {
+      "epoch": 2.789375582479031,
+      "grad_norm": 0.20233843956364075,
+      "learning_rate": 3.900586813945461e-06,
+      "loss": 0.2771,
+      "step": 2993
+    },
+    {
+      "epoch": 2.7903075489282387,
+      "grad_norm": 0.20204945409897973,
+      "learning_rate": 3.883327580255437e-06,
+      "loss": 0.278,
+      "step": 2994
+    },
+    {
+      "epoch": 2.7912395153774465,
+      "grad_norm": 0.20100104274603822,
+      "learning_rate": 3.8660683465654126e-06,
+      "loss": 0.2748,
+      "step": 2995
+    },
+    {
+      "epoch": 2.7921714818266543,
+      "grad_norm": 0.2007381347631073,
+      "learning_rate": 3.8488091128753886e-06,
+      "loss": 0.2742,
+      "step": 2996
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 0.19388789113009922,
+      "learning_rate": 3.8315498791853645e-06,
+      "loss": 0.2669,
+      "step": 2997
+    },
+    {
+      "epoch": 2.79403541472507,
+      "grad_norm": 0.2114315219470206,
+      "learning_rate": 3.81429064549534e-06,
+      "loss": 0.2754,
+      "step": 2998
+    },
+    {
+      "epoch": 2.7949673811742777,
+      "grad_norm": 0.20341889973716154,
+      "learning_rate": 3.7970314118053157e-06,
+      "loss": 0.2833,
+      "step": 2999
+    },
+    {
+      "epoch": 2.7958993476234855,
+      "grad_norm": 0.19692647787334872,
+      "learning_rate": 3.779772178115292e-06,
+      "loss": 0.2638,
+      "step": 3000
+    },
+    {
+      "epoch": 2.7968313140726933,
+      "grad_norm": 0.2032541191977724,
+      "learning_rate": 3.7625129444252676e-06,
+      "loss": 0.28,
+      "step": 3001
+    },
+    {
+      "epoch": 2.7977632805219015,
+      "grad_norm": 0.31854577858238514,
+      "learning_rate": 3.7452537107352436e-06,
+      "loss": 0.2927,
+      "step": 3002
+    },
+    {
+      "epoch": 2.798695246971109,
+      "grad_norm": 0.2002448632711451,
+      "learning_rate": 3.727994477045219e-06,
+      "loss": 0.2668,
+      "step": 3003
+    },
+    {
+      "epoch": 2.799627213420317,
+      "grad_norm": 0.21257371988699308,
+      "learning_rate": 3.7107352433551947e-06,
+      "loss": 0.2938,
+      "step": 3004
+    },
+    {
+      "epoch": 2.800559179869525,
+      "grad_norm": 0.20872409567617292,
+      "learning_rate": 3.693476009665171e-06,
+      "loss": 0.2831,
+      "step": 3005
+    },
+    {
+      "epoch": 2.8014911463187326,
+      "grad_norm": 0.200012215117581,
+      "learning_rate": 3.676216775975147e-06,
+      "loss": 0.3099,
+      "step": 3006
+    },
+    {
+      "epoch": 2.8024231127679404,
+      "grad_norm": 0.22415466560550393,
+      "learning_rate": 3.6589575422851227e-06,
+      "loss": 0.3113,
+      "step": 3007
+    },
+    {
+      "epoch": 2.803355079217148,
+      "grad_norm": 0.19948644371429816,
+      "learning_rate": 3.6416983085950983e-06,
+      "loss": 0.2688,
+      "step": 3008
+    },
+    {
+      "epoch": 2.804287045666356,
+      "grad_norm": 0.19193877315090913,
+      "learning_rate": 3.6244390749050742e-06,
+      "loss": 0.262,
+      "step": 3009
+    },
+    {
+      "epoch": 2.805219012115564,
+      "grad_norm": 0.20063048276252318,
+      "learning_rate": 3.6071798412150506e-06,
+      "loss": 0.2798,
+      "step": 3010
+    },
+    {
+      "epoch": 2.8061509785647716,
+      "grad_norm": 0.20401530109229551,
+      "learning_rate": 3.589920607525026e-06,
+      "loss": 0.2667,
+      "step": 3011
+    },
+    {
+      "epoch": 2.8070829450139794,
+      "grad_norm": 0.18652939906799426,
+      "learning_rate": 3.5726613738350018e-06,
+      "loss": 0.2769,
+      "step": 3012
+    },
+    {
+      "epoch": 2.8080149114631876,
+      "grad_norm": 0.20560051982465793,
+      "learning_rate": 3.5554021401449777e-06,
+      "loss": 0.2703,
+      "step": 3013
+    },
+    {
+      "epoch": 2.808946877912395,
+      "grad_norm": 0.20257440763904558,
+      "learning_rate": 3.5381429064549533e-06,
+      "loss": 0.274,
+      "step": 3014
+    },
+    {
+      "epoch": 2.809878844361603,
+      "grad_norm": 0.21352725580252352,
+      "learning_rate": 3.5208836727649297e-06,
+      "loss": 0.276,
+      "step": 3015
+    },
+    {
+      "epoch": 2.810810810810811,
+      "grad_norm": 0.18936218263115853,
+      "learning_rate": 3.5036244390749053e-06,
+      "loss": 0.2793,
+      "step": 3016
+    },
+    {
+      "epoch": 2.8117427772600188,
+      "grad_norm": 0.19115123776939869,
+      "learning_rate": 3.4863652053848813e-06,
+      "loss": 0.2536,
+      "step": 3017
+    },
+    {
+      "epoch": 2.8126747437092265,
+      "grad_norm": 0.20921436050788067,
+      "learning_rate": 3.469105971694857e-06,
+      "loss": 0.2816,
+      "step": 3018
+    },
+    {
+      "epoch": 2.8136067101584343,
+      "grad_norm": 0.1901612076562048,
+      "learning_rate": 3.4518467380048324e-06,
+      "loss": 0.2772,
+      "step": 3019
+    },
+    {
+      "epoch": 2.814538676607642,
+      "grad_norm": 0.21850030185802705,
+      "learning_rate": 3.434587504314809e-06,
+      "loss": 0.2785,
+      "step": 3020
+    },
+    {
+      "epoch": 2.81547064305685,
+      "grad_norm": 0.21838231920097942,
+      "learning_rate": 3.4173282706247848e-06,
+      "loss": 0.2838,
+      "step": 3021
+    },
+    {
+      "epoch": 2.8164026095060577,
+      "grad_norm": 0.20542459249738712,
+      "learning_rate": 3.4000690369347603e-06,
+      "loss": 0.291,
+      "step": 3022
+    },
+    {
+      "epoch": 2.8173345759552655,
+      "grad_norm": 0.1969002517460927,
+      "learning_rate": 3.382809803244736e-06,
+      "loss": 0.2748,
+      "step": 3023
+    },
+    {
+      "epoch": 2.8182665424044733,
+      "grad_norm": 0.19476118135666265,
+      "learning_rate": 3.365550569554712e-06,
+      "loss": 0.2799,
+      "step": 3024
+    },
+    {
+      "epoch": 2.819198508853681,
+      "grad_norm": 0.19439100154008937,
+      "learning_rate": 3.348291335864688e-06,
+      "loss": 0.282,
+      "step": 3025
+    },
+    {
+      "epoch": 2.8201304753028893,
+      "grad_norm": 0.1928490076351132,
+      "learning_rate": 3.331032102174664e-06,
+      "loss": 0.2924,
+      "step": 3026
+    },
+    {
+      "epoch": 2.821062441752097,
+      "grad_norm": 0.19542405957974024,
+      "learning_rate": 3.3137728684846394e-06,
+      "loss": 0.2818,
+      "step": 3027
+    },
+    {
+      "epoch": 2.821994408201305,
+      "grad_norm": 0.20720687527113496,
+      "learning_rate": 3.296513634794615e-06,
+      "loss": 0.3173,
+      "step": 3028
+    },
+    {
+      "epoch": 2.8229263746505127,
+      "grad_norm": 0.1969570782604848,
+      "learning_rate": 3.279254401104591e-06,
+      "loss": 0.2894,
+      "step": 3029
+    },
+    {
+      "epoch": 2.8238583410997204,
+      "grad_norm": 0.20448795308302087,
+      "learning_rate": 3.2619951674145674e-06,
+      "loss": 0.2677,
+      "step": 3030
+    },
+    {
+      "epoch": 2.8247903075489282,
+      "grad_norm": 0.2174777811153443,
+      "learning_rate": 3.244735933724543e-06,
+      "loss": 0.2889,
+      "step": 3031
+    },
+    {
+      "epoch": 2.825722273998136,
+      "grad_norm": 0.24676230625576193,
+      "learning_rate": 3.2274767000345185e-06,
+      "loss": 0.2798,
+      "step": 3032
+    },
+    {
+      "epoch": 2.826654240447344,
+      "grad_norm": 0.18746274478916045,
+      "learning_rate": 3.2102174663444945e-06,
+      "loss": 0.262,
+      "step": 3033
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 0.19428288203043748,
+      "learning_rate": 3.19295823265447e-06,
+      "loss": 0.2858,
+      "step": 3034
+    },
+    {
+      "epoch": 2.8285181733457594,
+      "grad_norm": 0.20049266618085823,
+      "learning_rate": 3.1756989989644464e-06,
+      "loss": 0.286,
+      "step": 3035
+    },
+    {
+      "epoch": 2.829450139794967,
+      "grad_norm": 0.18609336101204826,
+      "learning_rate": 3.158439765274422e-06,
+      "loss": 0.2606,
+      "step": 3036
+    },
+    {
+      "epoch": 2.8303821062441754,
+      "grad_norm": 0.19449823776814615,
+      "learning_rate": 3.141180531584398e-06,
+      "loss": 0.2837,
+      "step": 3037
+    },
+    {
+      "epoch": 2.831314072693383,
+      "grad_norm": 0.1981782256850738,
+      "learning_rate": 3.1239212978943735e-06,
+      "loss": 0.2836,
+      "step": 3038
+    },
+    {
+      "epoch": 2.832246039142591,
+      "grad_norm": 0.19034963656985357,
+      "learning_rate": 3.1066620642043495e-06,
+      "loss": 0.2851,
+      "step": 3039
+    },
+    {
+      "epoch": 2.8331780055917988,
+      "grad_norm": 0.19315419378671927,
+      "learning_rate": 3.089402830514325e-06,
+      "loss": 0.2663,
+      "step": 3040
+    },
+    {
+      "epoch": 2.8341099720410066,
+      "grad_norm": 0.19155636084855998,
+      "learning_rate": 3.072143596824301e-06,
+      "loss": 0.2653,
+      "step": 3041
+    },
+    {
+      "epoch": 2.8350419384902144,
+      "grad_norm": 0.19033274866175487,
+      "learning_rate": 3.054884363134277e-06,
+      "loss": 0.263,
+      "step": 3042
+    },
+    {
+      "epoch": 2.835973904939422,
+      "grad_norm": 0.19914301447220684,
+      "learning_rate": 3.0376251294442526e-06,
+      "loss": 0.2805,
+      "step": 3043
+    },
+    {
+      "epoch": 2.83690587138863,
+      "grad_norm": 0.18331768718125083,
+      "learning_rate": 3.0203658957542286e-06,
+      "loss": 0.2786,
+      "step": 3044
+    },
+    {
+      "epoch": 2.8378378378378377,
+      "grad_norm": 0.19760502631755147,
+      "learning_rate": 3.0031066620642046e-06,
+      "loss": 0.2866,
+      "step": 3045
+    },
+    {
+      "epoch": 2.8387698042870455,
+      "grad_norm": 0.19432888794714911,
+      "learning_rate": 2.98584742837418e-06,
+      "loss": 0.2804,
+      "step": 3046
+    },
+    {
+      "epoch": 2.8397017707362533,
+      "grad_norm": 0.19236831845839197,
+      "learning_rate": 2.968588194684156e-06,
+      "loss": 0.2752,
+      "step": 3047
+    },
+    {
+      "epoch": 2.8406337371854615,
+      "grad_norm": 0.1910742386679234,
+      "learning_rate": 2.951328960994132e-06,
+      "loss": 0.2658,
+      "step": 3048
+    },
+    {
+      "epoch": 2.8415657036346693,
+      "grad_norm": 0.21881058604205303,
+      "learning_rate": 2.934069727304108e-06,
+      "loss": 0.297,
+      "step": 3049
+    },
+    {
+      "epoch": 2.842497670083877,
+      "grad_norm": 0.20339300638225485,
+      "learning_rate": 2.9168104936140837e-06,
+      "loss": 0.2702,
+      "step": 3050
+    },
+    {
+      "epoch": 2.843429636533085,
+      "grad_norm": 0.1992683556170302,
+      "learning_rate": 2.8995512599240592e-06,
+      "loss": 0.2678,
+      "step": 3051
+    },
+    {
+      "epoch": 2.8443616029822927,
+      "grad_norm": 0.19878852815667308,
+      "learning_rate": 2.882292026234035e-06,
+      "loss": 0.2796,
+      "step": 3052
+    },
+    {
+      "epoch": 2.8452935694315005,
+      "grad_norm": 0.18545314347822645,
+      "learning_rate": 2.865032792544011e-06,
+      "loss": 0.2622,
+      "step": 3053
+    },
+    {
+      "epoch": 2.8462255358807083,
+      "grad_norm": 0.18491786250497877,
+      "learning_rate": 2.847773558853987e-06,
+      "loss": 0.2734,
+      "step": 3054
+    },
+    {
+      "epoch": 2.847157502329916,
+      "grad_norm": 0.2038511025826053,
+      "learning_rate": 2.8305143251639627e-06,
+      "loss": 0.2908,
+      "step": 3055
+    },
+    {
+      "epoch": 2.848089468779124,
+      "grad_norm": 0.20637010116367438,
+      "learning_rate": 2.8132550914739387e-06,
+      "loss": 0.2747,
+      "step": 3056
+    },
+    {
+      "epoch": 2.8490214352283316,
+      "grad_norm": 0.18852351793096087,
+      "learning_rate": 2.7959958577839147e-06,
+      "loss": 0.2726,
+      "step": 3057
+    },
+    {
+      "epoch": 2.8499534016775394,
+      "grad_norm": 0.19202813310092523,
+      "learning_rate": 2.7787366240938903e-06,
+      "loss": 0.2754,
+      "step": 3058
+    },
+    {
+      "epoch": 2.8508853681267476,
+      "grad_norm": 0.18833234610922944,
+      "learning_rate": 2.7614773904038662e-06,
+      "loss": 0.2795,
+      "step": 3059
+    },
+    {
+      "epoch": 2.8518173345759554,
+      "grad_norm": 0.1960718802950663,
+      "learning_rate": 2.7442181567138422e-06,
+      "loss": 0.2607,
+      "step": 3060
+    },
+    {
+      "epoch": 2.852749301025163,
+      "grad_norm": 0.18754349759971634,
+      "learning_rate": 2.726958923023818e-06,
+      "loss": 0.2737,
+      "step": 3061
+    },
+    {
+      "epoch": 2.853681267474371,
+      "grad_norm": 0.18732173043336564,
+      "learning_rate": 2.7096996893337938e-06,
+      "loss": 0.2708,
+      "step": 3062
+    },
+    {
+      "epoch": 2.854613233923579,
+      "grad_norm": 0.20391539866815722,
+      "learning_rate": 2.6924404556437693e-06,
+      "loss": 0.2737,
+      "step": 3063
+    },
+    {
+      "epoch": 2.8555452003727866,
+      "grad_norm": 0.1944147812396051,
+      "learning_rate": 2.6751812219537453e-06,
+      "loss": 0.2703,
+      "step": 3064
+    },
+    {
+      "epoch": 2.8564771668219944,
+      "grad_norm": 0.18250753719018248,
+      "learning_rate": 2.6579219882637213e-06,
+      "loss": 0.2739,
+      "step": 3065
+    },
+    {
+      "epoch": 2.857409133271202,
+      "grad_norm": 0.19994213870358213,
+      "learning_rate": 2.640662754573697e-06,
+      "loss": 0.2912,
+      "step": 3066
+    },
+    {
+      "epoch": 2.85834109972041,
+      "grad_norm": 0.18058177919018323,
+      "learning_rate": 2.623403520883673e-06,
+      "loss": 0.263,
+      "step": 3067
+    },
+    {
+      "epoch": 2.8592730661696177,
+      "grad_norm": 0.2050575347266619,
+      "learning_rate": 2.606144287193649e-06,
+      "loss": 0.2641,
+      "step": 3068
+    },
+    {
+      "epoch": 2.8602050326188255,
+      "grad_norm": 0.20022435417352347,
+      "learning_rate": 2.588885053503625e-06,
+      "loss": 0.2705,
+      "step": 3069
+    },
+    {
+      "epoch": 2.8611369990680338,
+      "grad_norm": 0.19070936112297893,
+      "learning_rate": 2.5716258198136004e-06,
+      "loss": 0.2728,
+      "step": 3070
+    },
+    {
+      "epoch": 2.862068965517241,
+      "grad_norm": 0.19006607981753762,
+      "learning_rate": 2.554366586123576e-06,
+      "loss": 0.277,
+      "step": 3071
+    },
+    {
+      "epoch": 2.8630009319664493,
+      "grad_norm": 0.20070814652277114,
+      "learning_rate": 2.5371073524335523e-06,
+      "loss": 0.2792,
+      "step": 3072
+    },
+    {
+      "epoch": 2.863932898415657,
+      "grad_norm": 0.18785077374913206,
+      "learning_rate": 2.519848118743528e-06,
+      "loss": 0.2728,
+      "step": 3073
+    },
+    {
+      "epoch": 2.864864864864865,
+      "grad_norm": 0.19614919144824364,
+      "learning_rate": 2.502588885053504e-06,
+      "loss": 0.2747,
+      "step": 3074
+    },
+    {
+      "epoch": 2.8657968313140727,
+      "grad_norm": 0.1889591173296693,
+      "learning_rate": 2.4853296513634795e-06,
+      "loss": 0.2759,
+      "step": 3075
+    },
+    {
+      "epoch": 2.8667287977632805,
+      "grad_norm": 0.21023443603767308,
+      "learning_rate": 2.4680704176734554e-06,
+      "loss": 0.2775,
+      "step": 3076
+    },
+    {
+      "epoch": 2.8676607642124883,
+      "grad_norm": 0.19871038682120043,
+      "learning_rate": 2.4508111839834314e-06,
+      "loss": 0.2786,
+      "step": 3077
+    },
+    {
+      "epoch": 2.868592730661696,
+      "grad_norm": 0.19392639366478343,
+      "learning_rate": 2.433551950293407e-06,
+      "loss": 0.2749,
+      "step": 3078
+    },
+    {
+      "epoch": 2.869524697110904,
+      "grad_norm": 0.21138855624086222,
+      "learning_rate": 2.416292716603383e-06,
+      "loss": 0.2931,
+      "step": 3079
+    },
+    {
+      "epoch": 2.8704566635601116,
+      "grad_norm": 0.20468378531450232,
+      "learning_rate": 2.399033482913359e-06,
+      "loss": 0.2844,
+      "step": 3080
+    },
+    {
+      "epoch": 2.87138863000932,
+      "grad_norm": 0.19714757606502553,
+      "learning_rate": 2.3817742492233345e-06,
+      "loss": 0.2835,
+      "step": 3081
+    },
+    {
+      "epoch": 2.872320596458527,
+      "grad_norm": 0.1984942419842415,
+      "learning_rate": 2.3645150155333105e-06,
+      "loss": 0.2726,
+      "step": 3082
+    },
+    {
+      "epoch": 2.8732525629077355,
+      "grad_norm": 0.18517062902215703,
+      "learning_rate": 2.347255781843286e-06,
+      "loss": 0.2636,
+      "step": 3083
+    },
+    {
+      "epoch": 2.8741845293569432,
+      "grad_norm": 0.20041065154332616,
+      "learning_rate": 2.3299965481532625e-06,
+      "loss": 0.2741,
+      "step": 3084
+    },
+    {
+      "epoch": 2.875116495806151,
+      "grad_norm": 0.18696803898184794,
+      "learning_rate": 2.312737314463238e-06,
+      "loss": 0.2622,
+      "step": 3085
+    },
+    {
+      "epoch": 2.876048462255359,
+      "grad_norm": 0.17709047806261977,
+      "learning_rate": 2.2954780807732136e-06,
+      "loss": 0.2511,
+      "step": 3086
+    },
+    {
+      "epoch": 2.8769804287045666,
+      "grad_norm": 0.19437055134866674,
+      "learning_rate": 2.2782188470831896e-06,
+      "loss": 0.287,
+      "step": 3087
+    },
+    {
+      "epoch": 2.8779123951537744,
+      "grad_norm": 0.19053712355577954,
+      "learning_rate": 2.2609596133931656e-06,
+      "loss": 0.266,
+      "step": 3088
+    },
+    {
+      "epoch": 2.878844361602982,
+      "grad_norm": 0.19610515787601318,
+      "learning_rate": 2.243700379703141e-06,
+      "loss": 0.2732,
+      "step": 3089
+    },
+    {
+      "epoch": 2.87977632805219,
+      "grad_norm": 0.19864055104226305,
+      "learning_rate": 2.226441146013117e-06,
+      "loss": 0.2796,
+      "step": 3090
+    },
+    {
+      "epoch": 2.8807082945013978,
+      "grad_norm": 0.1897603349145261,
+      "learning_rate": 2.2091819123230927e-06,
+      "loss": 0.265,
+      "step": 3091
+    },
+    {
+      "epoch": 2.881640260950606,
+      "grad_norm": 0.20021403163950588,
+      "learning_rate": 2.191922678633069e-06,
+      "loss": 0.2632,
+      "step": 3092
+    },
+    {
+      "epoch": 2.8825722273998133,
+      "grad_norm": 0.21228813219293502,
+      "learning_rate": 2.1746634449430446e-06,
+      "loss": 0.2868,
+      "step": 3093
+    },
+    {
+      "epoch": 2.8835041938490216,
+      "grad_norm": 0.3495517807416907,
+      "learning_rate": 2.15740421125302e-06,
+      "loss": 0.2909,
+      "step": 3094
+    },
+    {
+      "epoch": 2.8844361602982294,
+      "grad_norm": 0.1786527940997088,
+      "learning_rate": 2.140144977562996e-06,
+      "loss": 0.2681,
+      "step": 3095
+    },
+    {
+      "epoch": 2.885368126747437,
+      "grad_norm": 0.19556569302845087,
+      "learning_rate": 2.122885743872972e-06,
+      "loss": 0.2756,
+      "step": 3096
+    },
+    {
+      "epoch": 2.886300093196645,
+      "grad_norm": 0.19877524806729396,
+      "learning_rate": 2.105626510182948e-06,
+      "loss": 0.2817,
+      "step": 3097
+    },
+    {
+      "epoch": 2.8872320596458527,
+      "grad_norm": 0.19714461696704058,
+      "learning_rate": 2.0883672764929237e-06,
+      "loss": 0.2795,
+      "step": 3098
+    },
+    {
+      "epoch": 2.8881640260950605,
+      "grad_norm": 0.19775300650942954,
+      "learning_rate": 2.0711080428028997e-06,
+      "loss": 0.2598,
+      "step": 3099
+    },
+    {
+      "epoch": 2.8890959925442683,
+      "grad_norm": 0.1995353788570966,
+      "learning_rate": 2.0538488091128757e-06,
+      "loss": 0.267,
+      "step": 3100
+    },
+    {
+      "epoch": 2.890027958993476,
+      "grad_norm": 0.18921900576898532,
+      "learning_rate": 2.0365895754228512e-06,
+      "loss": 0.2592,
+      "step": 3101
+    },
+    {
+      "epoch": 2.890959925442684,
+      "grad_norm": 0.18470043389902377,
+      "learning_rate": 2.0193303417328272e-06,
+      "loss": 0.2721,
+      "step": 3102
+    },
+    {
+      "epoch": 2.891891891891892,
+      "grad_norm": 0.18540882492067112,
+      "learning_rate": 2.0020711080428028e-06,
+      "loss": 0.2586,
+      "step": 3103
+    },
+    {
+      "epoch": 2.8928238583410995,
+      "grad_norm": 0.20022216390680497,
+      "learning_rate": 1.9848118743527788e-06,
+      "loss": 0.2712,
+      "step": 3104
+    },
+    {
+      "epoch": 2.8937558247903077,
+      "grad_norm": 0.18928434968155539,
+      "learning_rate": 1.9675526406627547e-06,
+      "loss": 0.2876,
+      "step": 3105
+    },
+    {
+      "epoch": 2.8946877912395155,
+      "grad_norm": 0.20191915185255405,
+      "learning_rate": 1.9502934069727303e-06,
+      "loss": 0.2774,
+      "step": 3106
+    },
+    {
+      "epoch": 2.8956197576887233,
+      "grad_norm": 0.1951361819915228,
+      "learning_rate": 1.9330341732827063e-06,
+      "loss": 0.2638,
+      "step": 3107
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 0.197675901932993,
+      "learning_rate": 1.9157749395926823e-06,
+      "loss": 0.2911,
+      "step": 3108
+    },
+    {
+      "epoch": 2.897483690587139,
+      "grad_norm": 0.1928737819120238,
+      "learning_rate": 1.8985157059026578e-06,
+      "loss": 0.2916,
+      "step": 3109
+    },
+    {
+      "epoch": 2.8984156570363466,
+      "grad_norm": 0.19680649222222762,
+      "learning_rate": 1.8812564722126338e-06,
+      "loss": 0.2835,
+      "step": 3110
+    },
+    {
+      "epoch": 2.8993476234855544,
+      "grad_norm": 0.18750705457317698,
+      "learning_rate": 1.8639972385226096e-06,
+      "loss": 0.2722,
+      "step": 3111
+    },
+    {
+      "epoch": 2.900279589934762,
+      "grad_norm": 0.20714995444778417,
+      "learning_rate": 1.8467380048325856e-06,
+      "loss": 0.2662,
+      "step": 3112
+    },
+    {
+      "epoch": 2.90121155638397,
+      "grad_norm": 0.20646977392624752,
+      "learning_rate": 1.8294787711425613e-06,
+      "loss": 0.2762,
+      "step": 3113
+    },
+    {
+      "epoch": 2.9021435228331782,
+      "grad_norm": 0.18217951500379242,
+      "learning_rate": 1.8122195374525371e-06,
+      "loss": 0.2441,
+      "step": 3114
+    },
+    {
+      "epoch": 2.9030754892823856,
+      "grad_norm": 0.18474694213222892,
+      "learning_rate": 1.794960303762513e-06,
+      "loss": 0.2805,
+      "step": 3115
+    },
+    {
+      "epoch": 2.904007455731594,
+      "grad_norm": 0.18634195075203902,
+      "learning_rate": 1.7777010700724889e-06,
+      "loss": 0.2802,
+      "step": 3116
+    },
+    {
+      "epoch": 2.9049394221808016,
+      "grad_norm": 0.20142808594174186,
+      "learning_rate": 1.7604418363824649e-06,
+      "loss": 0.2751,
+      "step": 3117
+    },
+    {
+      "epoch": 2.9058713886300094,
+      "grad_norm": 0.29030328459211946,
+      "learning_rate": 1.7431826026924406e-06,
+      "loss": 0.2867,
+      "step": 3118
+    },
+    {
+      "epoch": 2.906803355079217,
+      "grad_norm": 0.2042198667506399,
+      "learning_rate": 1.7259233690024162e-06,
+      "loss": 0.2678,
+      "step": 3119
+    },
+    {
+      "epoch": 2.907735321528425,
+      "grad_norm": 0.19195977610311077,
+      "learning_rate": 1.7086641353123924e-06,
+      "loss": 0.2838,
+      "step": 3120
+    },
+    {
+      "epoch": 2.9086672879776327,
+      "grad_norm": 0.18095989480232708,
+      "learning_rate": 1.691404901622368e-06,
+      "loss": 0.2644,
+      "step": 3121
+    },
+    {
+      "epoch": 2.9095992544268405,
+      "grad_norm": 0.18557237655523964,
+      "learning_rate": 1.674145667932344e-06,
+      "loss": 0.277,
+      "step": 3122
+    },
+    {
+      "epoch": 2.9105312208760483,
+      "grad_norm": 0.18902200477423178,
+      "learning_rate": 1.6568864342423197e-06,
+      "loss": 0.2708,
+      "step": 3123
+    },
+    {
+      "epoch": 2.911463187325256,
+      "grad_norm": 0.18423597012276588,
+      "learning_rate": 1.6396272005522955e-06,
+      "loss": 0.264,
+      "step": 3124
+    },
+    {
+      "epoch": 2.9123951537744643,
+      "grad_norm": 0.1849829165312529,
+      "learning_rate": 1.6223679668622715e-06,
+      "loss": 0.2757,
+      "step": 3125
+    },
+    {
+      "epoch": 2.9133271202236717,
+      "grad_norm": 0.17510930926869414,
+      "learning_rate": 1.6051087331722472e-06,
+      "loss": 0.2569,
+      "step": 3126
+    },
+    {
+      "epoch": 2.91425908667288,
+      "grad_norm": 0.17797511673124905,
+      "learning_rate": 1.5878494994822232e-06,
+      "loss": 0.2655,
+      "step": 3127
+    },
+    {
+      "epoch": 2.9151910531220877,
+      "grad_norm": 0.18181790018158034,
+      "learning_rate": 1.570590265792199e-06,
+      "loss": 0.2643,
+      "step": 3128
+    },
+    {
+      "epoch": 2.9161230195712955,
+      "grad_norm": 0.20194301686738847,
+      "learning_rate": 1.5533310321021748e-06,
+      "loss": 0.2795,
+      "step": 3129
+    },
+    {
+      "epoch": 2.9170549860205033,
+      "grad_norm": 0.20517791324985948,
+      "learning_rate": 1.5360717984121505e-06,
+      "loss": 0.2826,
+      "step": 3130
+    },
+    {
+      "epoch": 2.917986952469711,
+      "grad_norm": 0.1935709711697975,
+      "learning_rate": 1.5188125647221263e-06,
+      "loss": 0.2556,
+      "step": 3131
+    },
+    {
+      "epoch": 2.918918918918919,
+      "grad_norm": 0.19431074628087489,
+      "learning_rate": 1.5015533310321023e-06,
+      "loss": 0.2937,
+      "step": 3132
+    },
+    {
+      "epoch": 2.9198508853681266,
+      "grad_norm": 0.19124093942323092,
+      "learning_rate": 1.484294097342078e-06,
+      "loss": 0.2653,
+      "step": 3133
+    },
+    {
+      "epoch": 2.9207828518173344,
+      "grad_norm": 0.18642793698492321,
+      "learning_rate": 1.467034863652054e-06,
+      "loss": 0.275,
+      "step": 3134
+    },
+    {
+      "epoch": 2.9217148182665422,
+      "grad_norm": 0.1920439771974822,
+      "learning_rate": 1.4497756299620296e-06,
+      "loss": 0.2773,
+      "step": 3135
+    },
+    {
+      "epoch": 2.9226467847157505,
+      "grad_norm": 0.18513239371925896,
+      "learning_rate": 1.4325163962720056e-06,
+      "loss": 0.2681,
+      "step": 3136
+    },
+    {
+      "epoch": 2.923578751164958,
+      "grad_norm": 0.17691521094286924,
+      "learning_rate": 1.4152571625819814e-06,
+      "loss": 0.2716,
+      "step": 3137
+    },
+    {
+      "epoch": 2.924510717614166,
+      "grad_norm": 0.18456562138839364,
+      "learning_rate": 1.3979979288919574e-06,
+      "loss": 0.2677,
+      "step": 3138
+    },
+    {
+      "epoch": 2.925442684063374,
+      "grad_norm": 0.29359826872342143,
+      "learning_rate": 1.3807386952019331e-06,
+      "loss": 0.2641,
+      "step": 3139
+    },
+    {
+      "epoch": 2.9263746505125816,
+      "grad_norm": 0.19254079389652903,
+      "learning_rate": 1.363479461511909e-06,
+      "loss": 0.2793,
+      "step": 3140
+    },
+    {
+      "epoch": 2.9273066169617894,
+      "grad_norm": 0.18052635821804464,
+      "learning_rate": 1.3462202278218847e-06,
+      "loss": 0.2559,
+      "step": 3141
+    },
+    {
+      "epoch": 2.928238583410997,
+      "grad_norm": 0.1873200626736531,
+      "learning_rate": 1.3289609941318607e-06,
+      "loss": 0.2701,
+      "step": 3142
+    },
+    {
+      "epoch": 2.929170549860205,
+      "grad_norm": 0.19977279554126529,
+      "learning_rate": 1.3117017604418364e-06,
+      "loss": 0.2754,
+      "step": 3143
+    },
+    {
+      "epoch": 2.9301025163094128,
+      "grad_norm": 0.19186313590757326,
+      "learning_rate": 1.2944425267518124e-06,
+      "loss": 0.2936,
+      "step": 3144
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 0.17893791121078334,
+      "learning_rate": 1.277183293061788e-06,
+      "loss": 0.2749,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9319664492078283,
+      "grad_norm": 0.19614408342124237,
+      "learning_rate": 1.259924059371764e-06,
+      "loss": 0.2876,
+      "step": 3146
+    },
+    {
+      "epoch": 2.9328984156570366,
+      "grad_norm": 0.18634100231233963,
+      "learning_rate": 1.2426648256817397e-06,
+      "loss": 0.2786,
+      "step": 3147
+    },
+    {
+      "epoch": 2.933830382106244,
+      "grad_norm": 0.1943597437986917,
+      "learning_rate": 1.2254055919917157e-06,
+      "loss": 0.2664,
+      "step": 3148
+    },
+    {
+      "epoch": 2.934762348555452,
+      "grad_norm": 0.1928919039002953,
+      "learning_rate": 1.2081463583016915e-06,
+      "loss": 0.2856,
+      "step": 3149
+    },
+    {
+      "epoch": 2.93569431500466,
+      "grad_norm": 0.18757800138510025,
+      "learning_rate": 1.1908871246116673e-06,
+      "loss": 0.2613,
+      "step": 3150
+    },
+    {
+      "epoch": 2.9366262814538677,
+      "grad_norm": 0.18387676821725837,
+      "learning_rate": 1.173627890921643e-06,
+      "loss": 0.2699,
+      "step": 3151
+    },
+    {
+      "epoch": 2.9375582479030755,
+      "grad_norm": 0.18757667376197,
+      "learning_rate": 1.156368657231619e-06,
+      "loss": 0.2776,
+      "step": 3152
+    },
+    {
+      "epoch": 2.9384902143522833,
+      "grad_norm": 0.17963861107384357,
+      "learning_rate": 1.1391094235415948e-06,
+      "loss": 0.2614,
+      "step": 3153
+    },
+    {
+      "epoch": 2.939422180801491,
+      "grad_norm": 0.19850985938548396,
+      "learning_rate": 1.1218501898515706e-06,
+      "loss": 0.2899,
+      "step": 3154
+    },
+    {
+      "epoch": 2.940354147250699,
+      "grad_norm": 0.18552889133372588,
+      "learning_rate": 1.1045909561615463e-06,
+      "loss": 0.2769,
+      "step": 3155
+    },
+    {
+      "epoch": 2.9412861136999067,
+      "grad_norm": 0.18812130326746307,
+      "learning_rate": 1.0873317224715223e-06,
+      "loss": 0.2747,
+      "step": 3156
+    },
+    {
+      "epoch": 2.9422180801491145,
+      "grad_norm": 0.19060412385975722,
+      "learning_rate": 1.070072488781498e-06,
+      "loss": 0.2892,
+      "step": 3157
+    },
+    {
+      "epoch": 2.9431500465983227,
+      "grad_norm": 0.18171513106979273,
+      "learning_rate": 1.052813255091474e-06,
+      "loss": 0.2679,
+      "step": 3158
+    },
+    {
+      "epoch": 2.94408201304753,
+      "grad_norm": 0.18321512689000957,
+      "learning_rate": 1.0355540214014498e-06,
+      "loss": 0.267,
+      "step": 3159
+    },
+    {
+      "epoch": 2.9450139794967383,
+      "grad_norm": 0.19293260140898946,
+      "learning_rate": 1.0182947877114256e-06,
+      "loss": 0.29,
+      "step": 3160
+    },
+    {
+      "epoch": 2.945945945945946,
+      "grad_norm": 0.19312332654193842,
+      "learning_rate": 1.0010355540214014e-06,
+      "loss": 0.2691,
+      "step": 3161
+    },
+    {
+      "epoch": 2.946877912395154,
+      "grad_norm": 0.18475840475999475,
+      "learning_rate": 9.837763203313774e-07,
+      "loss": 0.2685,
+      "step": 3162
+    },
+    {
+      "epoch": 2.9478098788443616,
+      "grad_norm": 0.1944014495964212,
+      "learning_rate": 9.665170866413531e-07,
+      "loss": 0.2802,
+      "step": 3163
+    },
+    {
+      "epoch": 2.9487418452935694,
+      "grad_norm": 0.1828077439824648,
+      "learning_rate": 9.492578529513289e-07,
+      "loss": 0.2602,
+      "step": 3164
+    },
+    {
+      "epoch": 2.949673811742777,
+      "grad_norm": 0.1827402151025915,
+      "learning_rate": 9.319986192613048e-07,
+      "loss": 0.2747,
+      "step": 3165
+    },
+    {
+      "epoch": 2.950605778191985,
+      "grad_norm": 0.18709539709750403,
+      "learning_rate": 9.147393855712807e-07,
+      "loss": 0.2628,
+      "step": 3166
+    },
+    {
+      "epoch": 2.951537744641193,
+      "grad_norm": 0.18816097495466125,
+      "learning_rate": 8.974801518812566e-07,
+      "loss": 0.2646,
+      "step": 3167
+    },
+    {
+      "epoch": 2.9524697110904006,
+      "grad_norm": 0.1962892785265206,
+      "learning_rate": 8.802209181912324e-07,
+      "loss": 0.2951,
+      "step": 3168
+    },
+    {
+      "epoch": 2.953401677539609,
+      "grad_norm": 0.1850207235161005,
+      "learning_rate": 8.629616845012081e-07,
+      "loss": 0.2658,
+      "step": 3169
+    },
+    {
+      "epoch": 2.954333643988816,
+      "grad_norm": 0.19836027623811306,
+      "learning_rate": 8.45702450811184e-07,
+      "loss": 0.2786,
+      "step": 3170
+    },
+    {
+      "epoch": 2.9552656104380244,
+      "grad_norm": 0.20878941288081204,
+      "learning_rate": 8.284432171211599e-07,
+      "loss": 0.2867,
+      "step": 3171
+    },
+    {
+      "epoch": 2.956197576887232,
+      "grad_norm": 0.18971214202986048,
+      "learning_rate": 8.111839834311357e-07,
+      "loss": 0.2816,
+      "step": 3172
+    },
+    {
+      "epoch": 2.95712954333644,
+      "grad_norm": 0.1979787297803025,
+      "learning_rate": 7.939247497411116e-07,
+      "loss": 0.2845,
+      "step": 3173
+    },
+    {
+      "epoch": 2.9580615097856477,
+      "grad_norm": 0.17340797274733377,
+      "learning_rate": 7.766655160510874e-07,
+      "loss": 0.2597,
+      "step": 3174
+    },
+    {
+      "epoch": 2.9589934762348555,
+      "grad_norm": 0.18565065931588529,
+      "learning_rate": 7.594062823610632e-07,
+      "loss": 0.2746,
+      "step": 3175
+    },
+    {
+      "epoch": 2.9599254426840633,
+      "grad_norm": 0.1915228444114785,
+      "learning_rate": 7.42147048671039e-07,
+      "loss": 0.2763,
+      "step": 3176
+    },
+    {
+      "epoch": 2.960857409133271,
+      "grad_norm": 0.2021614885867302,
+      "learning_rate": 7.248878149810148e-07,
+      "loss": 0.2967,
+      "step": 3177
+    },
+    {
+      "epoch": 2.961789375582479,
+      "grad_norm": 0.1816584053711945,
+      "learning_rate": 7.076285812909907e-07,
+      "loss": 0.2655,
+      "step": 3178
+    },
+    {
+      "epoch": 2.9627213420316867,
+      "grad_norm": 0.18047345543629661,
+      "learning_rate": 6.903693476009666e-07,
+      "loss": 0.276,
+      "step": 3179
+    },
+    {
+      "epoch": 2.963653308480895,
+      "grad_norm": 0.19642503447559204,
+      "learning_rate": 6.731101139109423e-07,
+      "loss": 0.2876,
+      "step": 3180
+    },
+    {
+      "epoch": 2.9645852749301023,
+      "grad_norm": 0.1894478319248315,
+      "learning_rate": 6.558508802209182e-07,
+      "loss": 0.2681,
+      "step": 3181
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 0.18227536471686712,
+      "learning_rate": 6.38591646530894e-07,
+      "loss": 0.2822,
+      "step": 3182
+    },
+    {
+      "epoch": 2.9664492078285183,
+      "grad_norm": 0.1962935483572407,
+      "learning_rate": 6.213324128408699e-07,
+      "loss": 0.2849,
+      "step": 3183
+    },
+    {
+      "epoch": 2.967381174277726,
+      "grad_norm": 0.18299938340496433,
+      "learning_rate": 6.040731791508457e-07,
+      "loss": 0.2701,
+      "step": 3184
+    },
+    {
+      "epoch": 2.968313140726934,
+      "grad_norm": 0.1852271353149681,
+      "learning_rate": 5.868139454608215e-07,
+      "loss": 0.2634,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9692451071761417,
+      "grad_norm": 0.19083397592544277,
+      "learning_rate": 5.695547117707974e-07,
+      "loss": 0.2676,
+      "step": 3186
+    },
+    {
+      "epoch": 2.9701770736253494,
+      "grad_norm": 0.19295927016314904,
+      "learning_rate": 5.522954780807732e-07,
+      "loss": 0.2705,
+      "step": 3187
+    },
+    {
+      "epoch": 2.9711090400745572,
+      "grad_norm": 0.1899430944889927,
+      "learning_rate": 5.35036244390749e-07,
+      "loss": 0.2743,
+      "step": 3188
+    },
+    {
+      "epoch": 2.972041006523765,
+      "grad_norm": 0.19525178412087957,
+      "learning_rate": 5.177770107007249e-07,
+      "loss": 0.2861,
+      "step": 3189
+    },
+    {
+      "epoch": 2.972972972972973,
+      "grad_norm": 0.18044761615669083,
+      "learning_rate": 5.005177770107007e-07,
+      "loss": 0.2692,
+      "step": 3190
+    },
+    {
+      "epoch": 2.973904939422181,
+      "grad_norm": 0.1816461640936677,
+      "learning_rate": 4.832585433206766e-07,
+      "loss": 0.2777,
+      "step": 3191
+    },
+    {
+      "epoch": 2.9748369058713884,
+      "grad_norm": 0.21404810357075688,
+      "learning_rate": 4.659993096306524e-07,
+      "loss": 0.2727,
+      "step": 3192
+    },
+    {
+      "epoch": 2.9757688723205966,
+      "grad_norm": 0.19924268950907487,
+      "learning_rate": 4.487400759406283e-07,
+      "loss": 0.2734,
+      "step": 3193
+    },
+    {
+      "epoch": 2.9767008387698044,
+      "grad_norm": 0.19332008769649855,
+      "learning_rate": 4.3148084225060405e-07,
+      "loss": 0.2725,
+      "step": 3194
+    },
+    {
+      "epoch": 2.977632805219012,
+      "grad_norm": 0.19062464793092224,
+      "learning_rate": 4.142216085605799e-07,
+      "loss": 0.2719,
+      "step": 3195
+    },
+    {
+      "epoch": 2.97856477166822,
+      "grad_norm": 0.1881773409433436,
+      "learning_rate": 3.969623748705558e-07,
+      "loss": 0.2711,
+      "step": 3196
+    },
+    {
+      "epoch": 2.9794967381174278,
+      "grad_norm": 0.18304508437338973,
+      "learning_rate": 3.797031411805316e-07,
+      "loss": 0.277,
+      "step": 3197
+    },
+    {
+      "epoch": 2.9804287045666356,
+      "grad_norm": 0.19335517644567377,
+      "learning_rate": 3.624439074905074e-07,
+      "loss": 0.291,
+      "step": 3198
+    },
+    {
+      "epoch": 2.9813606710158433,
+      "grad_norm": 0.18910604414641874,
+      "learning_rate": 3.451846738004833e-07,
+      "loss": 0.2826,
+      "step": 3199
+    },
+    {
+      "epoch": 2.982292637465051,
+      "grad_norm": 0.18641218464281595,
+      "learning_rate": 3.279254401104591e-07,
+      "loss": 0.2838,
+      "step": 3200
+    },
+    {
+      "epoch": 2.983224603914259,
+      "grad_norm": 0.19645544848890542,
+      "learning_rate": 3.1066620642043493e-07,
+      "loss": 0.2734,
+      "step": 3201
+    },
+    {
+      "epoch": 2.984156570363467,
+      "grad_norm": 0.1855772116778447,
+      "learning_rate": 2.9340697273041076e-07,
+      "loss": 0.2768,
+      "step": 3202
+    },
+    {
+      "epoch": 2.9850885368126745,
+      "grad_norm": 0.18284159758424942,
+      "learning_rate": 2.761477390403866e-07,
+      "loss": 0.2646,
+      "step": 3203
+    },
+    {
+      "epoch": 2.9860205032618827,
+      "grad_norm": 0.18483954212472017,
+      "learning_rate": 2.5888850535036246e-07,
+      "loss": 0.2854,
+      "step": 3204
+    },
+    {
+      "epoch": 2.9869524697110905,
+      "grad_norm": 0.19346195356570384,
+      "learning_rate": 2.416292716603383e-07,
+      "loss": 0.2725,
+      "step": 3205
+    },
+    {
+      "epoch": 2.9878844361602983,
+      "grad_norm": 0.18991719375444852,
+      "learning_rate": 2.2437003797031414e-07,
+      "loss": 0.2849,
+      "step": 3206
+    },
+    {
+      "epoch": 2.988816402609506,
+      "grad_norm": 0.19226419092775487,
+      "learning_rate": 2.0711080428028996e-07,
+      "loss": 0.2977,
+      "step": 3207
+    },
+    {
+      "epoch": 2.989748369058714,
+      "grad_norm": 0.1813789047818328,
+      "learning_rate": 1.898515705902658e-07,
+      "loss": 0.2554,
+      "step": 3208
+    },
+    {
+      "epoch": 2.9906803355079217,
+      "grad_norm": 0.19984677695028977,
+      "learning_rate": 1.7259233690024164e-07,
+      "loss": 0.2821,
+      "step": 3209
+    },
+    {
+      "epoch": 2.9916123019571295,
+      "grad_norm": 0.19524836862573727,
+      "learning_rate": 1.5533310321021747e-07,
+      "loss": 0.2722,
+      "step": 3210
+    },
+    {
+      "epoch": 2.9925442684063372,
+      "grad_norm": 0.33710706004468005,
+      "learning_rate": 1.380738695201933e-07,
+      "loss": 0.2675,
+      "step": 3211
+    },
+    {
+      "epoch": 2.993476234855545,
+      "grad_norm": 0.2004416480570965,
+      "learning_rate": 1.2081463583016914e-07,
+      "loss": 0.287,
+      "step": 3212
+    },
+    {
+      "epoch": 2.9944082013047533,
+      "grad_norm": 0.18458400898378163,
+      "learning_rate": 1.0355540214014498e-07,
+      "loss": 0.2785,
+      "step": 3213
+    },
+    {
+      "epoch": 2.9953401677539606,
+      "grad_norm": 0.19030848370889905,
+      "learning_rate": 8.629616845012082e-08,
+      "loss": 0.2831,
+      "step": 3214
+    },
+    {
+      "epoch": 2.996272134203169,
+      "grad_norm": 0.19013705115535734,
+      "learning_rate": 6.903693476009665e-08,
+      "loss": 0.2664,
+      "step": 3215
+    },
+    {
+      "epoch": 2.9972041006523766,
+      "grad_norm": 0.18992346843323965,
+      "learning_rate": 5.177770107007249e-08,
+      "loss": 0.2755,
+      "step": 3216
+    },
+    {
+      "epoch": 2.9981360671015844,
+      "grad_norm": 0.19422270072832717,
+      "learning_rate": 3.451846738004832e-08,
+      "loss": 0.2803,
+      "step": 3217
+    },
+    {
+      "epoch": 2.999068033550792,
+      "grad_norm": 0.18142364679184228,
+      "learning_rate": 1.725923369002416e-08,
+      "loss": 0.2678,
+      "step": 3218
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.19381177646391537,
+      "learning_rate": 0.0,
+      "loss": 0.2877,
+      "step": 3219
+    },
+    {
+      "epoch": 3.0,
+      "step": 3219,
+      "total_flos": 2.755855970469937e+18,
+      "train_loss": 0.4114767361979708,
+      "train_runtime": 186765.2867,
+      "train_samples_per_second": 0.276,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3219,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.755855970469937e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}