diff --git "a/ECD_Fintuned_phi-3-vision/trainer_state.json" "b/ECD_Fintuned_phi-3-vision/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/ECD_Fintuned_phi-3-vision/trainer_state.json"
@@ -0,0 +1,23492 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00029850746268656717,
+      "grad_norm": 14.117414815113976,
+      "learning_rate": 4.950495049504951e-08,
+      "loss": 1.4506,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005970149253731343,
+      "grad_norm": 13.931060049875214,
+      "learning_rate": 9.900990099009901e-08,
+      "loss": 1.4497,
+      "step": 2
+    },
+    {
+      "epoch": 0.0008955223880597015,
+      "grad_norm": 12.951035081832662,
+      "learning_rate": 1.4851485148514852e-07,
+      "loss": 1.4249,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011940298507462687,
+      "grad_norm": 12.506012032234118,
+      "learning_rate": 1.9801980198019803e-07,
+      "loss": 1.3703,
+      "step": 4
+    },
+    {
+      "epoch": 0.0014925373134328358,
+      "grad_norm": 12.759791737538139,
+      "learning_rate": 2.4752475247524754e-07,
+      "loss": 1.3928,
+      "step": 5
+    },
+    {
+      "epoch": 0.001791044776119403,
+      "grad_norm": 11.981120238868472,
+      "learning_rate": 2.9702970297029703e-07,
+      "loss": 1.3379,
+      "step": 6
+    },
+    {
+      "epoch": 0.00208955223880597,
+      "grad_norm": 14.232070759545241,
+      "learning_rate": 3.4653465346534657e-07,
+      "loss": 1.3504,
+      "step": 7
+    },
+    {
+      "epoch": 0.0023880597014925373,
+      "grad_norm": 14.919999725885855,
+      "learning_rate": 3.9603960396039606e-07,
+      "loss": 1.4683,
+      "step": 8
+    },
+    {
+      "epoch": 0.0026865671641791043,
+      "grad_norm": 15.081665565944352,
+      "learning_rate": 4.4554455445544555e-07,
+      "loss": 1.4529,
+      "step": 9
+    },
+    {
+      "epoch": 0.0029850746268656717,
+      "grad_norm": 14.1043643853956,
+      "learning_rate": 4.950495049504951e-07,
+      "loss": 1.4489,
+      "step": 10
+    },
+    {
+      "epoch": 0.0032835820895522386,
+      "grad_norm": 13.03984594597713,
+      "learning_rate": 5.445544554455446e-07,
+      "loss": 1.3641,
+      "step": 11
+    },
+    {
+      "epoch": 0.003582089552238806,
+      "grad_norm": 12.55643909326301,
+      "learning_rate": 5.940594059405941e-07,
+      "loss": 1.366,
+      "step": 12
+    },
+    {
+      "epoch": 0.003880597014925373,
+      "grad_norm": 13.066793846628885,
+      "learning_rate": 6.435643564356436e-07,
+      "loss": 1.3347,
+      "step": 13
+    },
+    {
+      "epoch": 0.00417910447761194,
+      "grad_norm": 14.232011881175255,
+      "learning_rate": 6.930693069306931e-07,
+      "loss": 1.3869,
+      "step": 14
+    },
+    {
+      "epoch": 0.004477611940298508,
+      "grad_norm": 13.382384812080517,
+      "learning_rate": 7.425742574257426e-07,
+      "loss": 1.2797,
+      "step": 15
+    },
+    {
+      "epoch": 0.004776119402985075,
+      "grad_norm": 14.063303937583319,
+      "learning_rate": 7.920792079207921e-07,
+      "loss": 1.2945,
+      "step": 16
+    },
+    {
+      "epoch": 0.005074626865671642,
+      "grad_norm": 12.599956283878015,
+      "learning_rate": 8.415841584158417e-07,
+      "loss": 1.2541,
+      "step": 17
+    },
+    {
+      "epoch": 0.005373134328358209,
+      "grad_norm": 15.35594033376082,
+      "learning_rate": 8.910891089108911e-07,
+      "loss": 1.4035,
+      "step": 18
+    },
+    {
+      "epoch": 0.005671641791044776,
+      "grad_norm": 13.703270603112776,
+      "learning_rate": 9.405940594059406e-07,
+      "loss": 1.3001,
+      "step": 19
+    },
+    {
+      "epoch": 0.005970149253731343,
+      "grad_norm": 11.466111679630176,
+      "learning_rate": 9.900990099009902e-07,
+      "loss": 1.1074,
+      "step": 20
+    },
+    {
+      "epoch": 0.00626865671641791,
+      "grad_norm": 12.969087798482644,
+      "learning_rate": 1.0396039603960397e-06,
+      "loss": 1.1151,
+      "step": 21
+    },
+    {
+      "epoch": 0.006567164179104477,
+      "grad_norm": 11.305356480464539,
+      "learning_rate": 1.0891089108910893e-06,
+      "loss": 1.0693,
+      "step": 22
+    },
+    {
+      "epoch": 0.006865671641791045,
+      "grad_norm": 10.571941049724712,
+      "learning_rate": 1.1386138613861388e-06,
+      "loss": 1.0079,
+      "step": 23
+    },
+    {
+      "epoch": 0.007164179104477612,
+      "grad_norm": 13.611826826791107,
+      "learning_rate": 1.1881188118811881e-06,
+      "loss": 1.1241,
+      "step": 24
+    },
+    {
+      "epoch": 0.007462686567164179,
+      "grad_norm": 10.580271079481518,
+      "learning_rate": 1.2376237623762377e-06,
+      "loss": 1.0046,
+      "step": 25
+    },
+    {
+      "epoch": 0.007761194029850746,
+      "grad_norm": 14.626205948755526,
+      "learning_rate": 1.2871287128712872e-06,
+      "loss": 1.0083,
+      "step": 26
+    },
+    {
+      "epoch": 0.008059701492537314,
+      "grad_norm": 9.659001814009544,
+      "learning_rate": 1.3366336633663367e-06,
+      "loss": 0.8455,
+      "step": 27
+    },
+    {
+      "epoch": 0.00835820895522388,
+      "grad_norm": 9.492714737777815,
+      "learning_rate": 1.3861386138613863e-06,
+      "loss": 0.744,
+      "step": 28
+    },
+    {
+      "epoch": 0.008656716417910448,
+      "grad_norm": 9.552713565555441,
+      "learning_rate": 1.4356435643564356e-06,
+      "loss": 0.7097,
+      "step": 29
+    },
+    {
+      "epoch": 0.008955223880597015,
+      "grad_norm": 9.399332350213616,
+      "learning_rate": 1.4851485148514852e-06,
+      "loss": 0.8027,
+      "step": 30
+    },
+    {
+      "epoch": 0.009253731343283582,
+      "grad_norm": 10.362788296964608,
+      "learning_rate": 1.5346534653465347e-06,
+      "loss": 0.8065,
+      "step": 31
+    },
+    {
+      "epoch": 0.00955223880597015,
+      "grad_norm": 8.883431717385413,
+      "learning_rate": 1.5841584158415842e-06,
+      "loss": 0.6679,
+      "step": 32
+    },
+    {
+      "epoch": 0.009850746268656717,
+      "grad_norm": 6.947939015528359,
+      "learning_rate": 1.6336633663366338e-06,
+      "loss": 0.6528,
+      "step": 33
+    },
+    {
+      "epoch": 0.010149253731343283,
+      "grad_norm": 5.181069033173571,
+      "learning_rate": 1.6831683168316833e-06,
+      "loss": 0.6128,
+      "step": 34
+    },
+    {
+      "epoch": 0.010447761194029851,
+      "grad_norm": 2.1652397986100635,
+      "learning_rate": 1.7326732673267326e-06,
+      "loss": 0.6552,
+      "step": 35
+    },
+    {
+      "epoch": 0.010746268656716417,
+      "grad_norm": 4.632087693327325,
+      "learning_rate": 1.7821782178217822e-06,
+      "loss": 0.6027,
+      "step": 36
+    },
+    {
+      "epoch": 0.011044776119402985,
+      "grad_norm": 7.422754544331454,
+      "learning_rate": 1.8316831683168317e-06,
+      "loss": 0.59,
+      "step": 37
+    },
+    {
+      "epoch": 0.011343283582089553,
+      "grad_norm": 7.399958472269416,
+      "learning_rate": 1.8811881188118813e-06,
+      "loss": 0.586,
+      "step": 38
+    },
+    {
+      "epoch": 0.011641791044776119,
+      "grad_norm": 3.656574253442115,
+      "learning_rate": 1.930693069306931e-06,
+      "loss": 0.5334,
+      "step": 39
+    },
+    {
+      "epoch": 0.011940298507462687,
+      "grad_norm": 2.513197596004464,
+      "learning_rate": 1.9801980198019803e-06,
+      "loss": 0.5474,
+      "step": 40
+    },
+    {
+      "epoch": 0.012238805970149255,
+      "grad_norm": 1.918336955189715,
+      "learning_rate": 2.02970297029703e-06,
+      "loss": 0.6019,
+      "step": 41
+    },
+    {
+      "epoch": 0.01253731343283582,
+      "grad_norm": 1.4284644405554494,
+      "learning_rate": 2.0792079207920794e-06,
+      "loss": 0.5576,
+      "step": 42
+    },
+    {
+      "epoch": 0.012835820895522388,
+      "grad_norm": 1.578482950062078,
+      "learning_rate": 2.1287128712871288e-06,
+      "loss": 0.5727,
+      "step": 43
+    },
+    {
+      "epoch": 0.013134328358208954,
+      "grad_norm": 1.354481047717988,
+      "learning_rate": 2.1782178217821785e-06,
+      "loss": 0.5318,
+      "step": 44
+    },
+    {
+      "epoch": 0.013432835820895522,
+      "grad_norm": 1.25497134625581,
+      "learning_rate": 2.227722772277228e-06,
+      "loss": 0.5266,
+      "step": 45
+    },
+    {
+      "epoch": 0.01373134328358209,
+      "grad_norm": 1.3741989730447257,
+      "learning_rate": 2.2772277227722776e-06,
+      "loss": 0.5577,
+      "step": 46
+    },
+    {
+      "epoch": 0.014029850746268656,
+      "grad_norm": 1.1498984206900507,
+      "learning_rate": 2.326732673267327e-06,
+      "loss": 0.4985,
+      "step": 47
+    },
+    {
+      "epoch": 0.014328358208955224,
+      "grad_norm": 1.2303306117520938,
+      "learning_rate": 2.3762376237623762e-06,
+      "loss": 0.5211,
+      "step": 48
+    },
+    {
+      "epoch": 0.014626865671641792,
+      "grad_norm": 1.11221660675085,
+      "learning_rate": 2.425742574257426e-06,
+      "loss": 0.4762,
+      "step": 49
+    },
+    {
+      "epoch": 0.014925373134328358,
+      "grad_norm": 1.1204377252400624,
+      "learning_rate": 2.4752475247524753e-06,
+      "loss": 0.5147,
+      "step": 50
+    },
+    {
+      "epoch": 0.015223880597014926,
+      "grad_norm": 1.3294182563507342,
+      "learning_rate": 2.524752475247525e-06,
+      "loss": 0.582,
+      "step": 51
+    },
+    {
+      "epoch": 0.015522388059701492,
+      "grad_norm": 1.1126190487740035,
+      "learning_rate": 2.5742574257425744e-06,
+      "loss": 0.5393,
+      "step": 52
+    },
+    {
+      "epoch": 0.01582089552238806,
+      "grad_norm": 1.0981865353995417,
+      "learning_rate": 2.623762376237624e-06,
+      "loss": 0.4749,
+      "step": 53
+    },
+    {
+      "epoch": 0.016119402985074627,
+      "grad_norm": 1.0843096961998069,
+      "learning_rate": 2.6732673267326735e-06,
+      "loss": 0.521,
+      "step": 54
+    },
+    {
+      "epoch": 0.016417910447761194,
+      "grad_norm": 1.0763196586153305,
+      "learning_rate": 2.7227722772277232e-06,
+      "loss": 0.4723,
+      "step": 55
+    },
+    {
+      "epoch": 0.01671641791044776,
+      "grad_norm": 1.0864036348823496,
+      "learning_rate": 2.7722772277227726e-06,
+      "loss": 0.5065,
+      "step": 56
+    },
+    {
+      "epoch": 0.01701492537313433,
+      "grad_norm": 1.2234419660497935,
+      "learning_rate": 2.821782178217822e-06,
+      "loss": 0.4942,
+      "step": 57
+    },
+    {
+      "epoch": 0.017313432835820895,
+      "grad_norm": 1.0591972627781436,
+      "learning_rate": 2.8712871287128712e-06,
+      "loss": 0.4778,
+      "step": 58
+    },
+    {
+      "epoch": 0.01761194029850746,
+      "grad_norm": 1.0721476207536529,
+      "learning_rate": 2.920792079207921e-06,
+      "loss": 0.5391,
+      "step": 59
+    },
+    {
+      "epoch": 0.01791044776119403,
+      "grad_norm": 1.0985626878686374,
+      "learning_rate": 2.9702970297029703e-06,
+      "loss": 0.4476,
+      "step": 60
+    },
+    {
+      "epoch": 0.018208955223880597,
+      "grad_norm": 1.1392243737050494,
+      "learning_rate": 3.01980198019802e-06,
+      "loss": 0.5497,
+      "step": 61
+    },
+    {
+      "epoch": 0.018507462686567163,
+      "grad_norm": 1.0813836369765903,
+      "learning_rate": 3.0693069306930694e-06,
+      "loss": 0.5208,
+      "step": 62
+    },
+    {
+      "epoch": 0.018805970149253733,
+      "grad_norm": 1.0889059705799737,
+      "learning_rate": 3.118811881188119e-06,
+      "loss": 0.4994,
+      "step": 63
+    },
+    {
+      "epoch": 0.0191044776119403,
+      "grad_norm": 1.0362585523500178,
+      "learning_rate": 3.1683168316831685e-06,
+      "loss": 0.4829,
+      "step": 64
+    },
+    {
+      "epoch": 0.019402985074626865,
+      "grad_norm": 1.0359519925978355,
+      "learning_rate": 3.2178217821782182e-06,
+      "loss": 0.476,
+      "step": 65
+    },
+    {
+      "epoch": 0.019701492537313434,
+      "grad_norm": 0.9185525908299487,
+      "learning_rate": 3.2673267326732676e-06,
+      "loss": 0.4103,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0418468842058957,
+      "learning_rate": 3.3168316831683173e-06,
+      "loss": 0.4146,
+      "step": 67
+    },
+    {
+      "epoch": 0.020298507462686566,
+      "grad_norm": 1.2661337234827934,
+      "learning_rate": 3.3663366336633666e-06,
+      "loss": 0.4533,
+      "step": 68
+    },
+    {
+      "epoch": 0.020597014925373136,
+      "grad_norm": 0.9917842549873999,
+      "learning_rate": 3.4158415841584164e-06,
+      "loss": 0.4651,
+      "step": 69
+    },
+    {
+      "epoch": 0.020895522388059702,
+      "grad_norm": 1.0318070695121893,
+      "learning_rate": 3.4653465346534653e-06,
+      "loss": 0.5135,
+      "step": 70
+    },
+    {
+      "epoch": 0.021194029850746268,
+      "grad_norm": 0.9942332357926064,
+      "learning_rate": 3.514851485148515e-06,
+      "loss": 0.4812,
+      "step": 71
+    },
+    {
+      "epoch": 0.021492537313432834,
+      "grad_norm": 0.9647372803003441,
+      "learning_rate": 3.5643564356435644e-06,
+      "loss": 0.5059,
+      "step": 72
+    },
+    {
+      "epoch": 0.021791044776119404,
+      "grad_norm": 0.9756894687445946,
+      "learning_rate": 3.613861386138614e-06,
+      "loss": 0.5094,
+      "step": 73
+    },
+    {
+      "epoch": 0.02208955223880597,
+      "grad_norm": 1.0710107538374112,
+      "learning_rate": 3.6633663366336635e-06,
+      "loss": 0.5479,
+      "step": 74
+    },
+    {
+      "epoch": 0.022388059701492536,
+      "grad_norm": 1.0245672263567556,
+      "learning_rate": 3.712871287128713e-06,
+      "loss": 0.4638,
+      "step": 75
+    },
+    {
+      "epoch": 0.022686567164179106,
+      "grad_norm": 0.972951009425826,
+      "learning_rate": 3.7623762376237625e-06,
+      "loss": 0.5037,
+      "step": 76
+    },
+    {
+      "epoch": 0.02298507462686567,
+      "grad_norm": 0.8913835901316376,
+      "learning_rate": 3.8118811881188123e-06,
+      "loss": 0.4415,
+      "step": 77
+    },
+    {
+      "epoch": 0.023283582089552238,
+      "grad_norm": 0.918140464986545,
+      "learning_rate": 3.861386138613862e-06,
+      "loss": 0.5016,
+      "step": 78
+    },
+    {
+      "epoch": 0.023582089552238807,
+      "grad_norm": 0.9542502229677259,
+      "learning_rate": 3.910891089108911e-06,
+      "loss": 0.4639,
+      "step": 79
+    },
+    {
+      "epoch": 0.023880597014925373,
+      "grad_norm": 0.9376495588439216,
+      "learning_rate": 3.960396039603961e-06,
+      "loss": 0.4696,
+      "step": 80
+    },
+    {
+      "epoch": 0.02417910447761194,
+      "grad_norm": 1.0118829978197192,
+      "learning_rate": 4.0099009900990104e-06,
+      "loss": 0.4393,
+      "step": 81
+    },
+    {
+      "epoch": 0.02447761194029851,
+      "grad_norm": 0.9312691594030761,
+      "learning_rate": 4.05940594059406e-06,
+      "loss": 0.4773,
+      "step": 82
+    },
+    {
+      "epoch": 0.024776119402985075,
+      "grad_norm": 0.9718484289074919,
+      "learning_rate": 4.108910891089109e-06,
+      "loss": 0.4915,
+      "step": 83
+    },
+    {
+      "epoch": 0.02507462686567164,
+      "grad_norm": 0.9125444747399516,
+      "learning_rate": 4.158415841584159e-06,
+      "loss": 0.4721,
+      "step": 84
+    },
+    {
+      "epoch": 0.025373134328358207,
+      "grad_norm": 0.8900812924955422,
+      "learning_rate": 4.207920792079208e-06,
+      "loss": 0.4223,
+      "step": 85
+    },
+    {
+      "epoch": 0.025671641791044777,
+      "grad_norm": 1.121907218319262,
+      "learning_rate": 4.2574257425742575e-06,
+      "loss": 0.427,
+      "step": 86
+    },
+    {
+      "epoch": 0.025970149253731343,
+      "grad_norm": 1.0549443213006138,
+      "learning_rate": 4.306930693069307e-06,
+      "loss": 0.4944,
+      "step": 87
+    },
+    {
+      "epoch": 0.02626865671641791,
+      "grad_norm": 0.9113639087013188,
+      "learning_rate": 4.356435643564357e-06,
+      "loss": 0.445,
+      "step": 88
+    },
+    {
+      "epoch": 0.02656716417910448,
+      "grad_norm": 0.9366021839687358,
+      "learning_rate": 4.405940594059406e-06,
+      "loss": 0.419,
+      "step": 89
+    },
+    {
+      "epoch": 0.026865671641791045,
+      "grad_norm": 1.021855680179713,
+      "learning_rate": 4.455445544554456e-06,
+      "loss": 0.5366,
+      "step": 90
+    },
+    {
+      "epoch": 0.02716417910447761,
+      "grad_norm": 1.306674789879524,
+      "learning_rate": 4.5049504950495054e-06,
+      "loss": 0.4847,
+      "step": 91
+    },
+    {
+      "epoch": 0.02746268656716418,
+      "grad_norm": 0.9466689723258144,
+      "learning_rate": 4.554455445544555e-06,
+      "loss": 0.4331,
+      "step": 92
+    },
+    {
+      "epoch": 0.027761194029850746,
+      "grad_norm": 1.0130636044595012,
+      "learning_rate": 4.603960396039605e-06,
+      "loss": 0.5162,
+      "step": 93
+    },
+    {
+      "epoch": 0.028059701492537312,
+      "grad_norm": 0.9085144452361592,
+      "learning_rate": 4.653465346534654e-06,
+      "loss": 0.4798,
+      "step": 94
+    },
+    {
+      "epoch": 0.028358208955223882,
+      "grad_norm": 0.8665298826536127,
+      "learning_rate": 4.702970297029703e-06,
+      "loss": 0.4102,
+      "step": 95
+    },
+    {
+      "epoch": 0.028656716417910448,
+      "grad_norm": 0.8922124742463072,
+      "learning_rate": 4.7524752475247525e-06,
+      "loss": 0.4531,
+      "step": 96
+    },
+    {
+      "epoch": 0.028955223880597014,
+      "grad_norm": 1.0766078947850648,
+      "learning_rate": 4.801980198019802e-06,
+      "loss": 0.4298,
+      "step": 97
+    },
+    {
+      "epoch": 0.029253731343283584,
+      "grad_norm": 0.887211185618867,
+      "learning_rate": 4.851485148514852e-06,
+      "loss": 0.4285,
+      "step": 98
+    },
+    {
+      "epoch": 0.02955223880597015,
+      "grad_norm": 0.8297755867290657,
+      "learning_rate": 4.900990099009901e-06,
+      "loss": 0.3819,
+      "step": 99
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 0.9717588300149205,
+      "learning_rate": 4.950495049504951e-06,
+      "loss": 0.5215,
+      "step": 100
+    },
+    {
+      "epoch": 0.030149253731343282,
+      "grad_norm": 0.8764330876038892,
+      "learning_rate": 5e-06,
+      "loss": 0.397,
+      "step": 101
+    },
+    {
+      "epoch": 0.03044776119402985,
+      "grad_norm": 0.9639801062167249,
+      "learning_rate": 4.99999883128047e-06,
+      "loss": 0.4042,
+      "step": 102
+    },
+    {
+      "epoch": 0.030746268656716418,
+      "grad_norm": 0.9009910276004819,
+      "learning_rate": 4.999995325122968e-06,
+      "loss": 0.4648,
+      "step": 103
+    },
+    {
+      "epoch": 0.031044776119402984,
+      "grad_norm": 0.9428548948928329,
+      "learning_rate": 4.999989481530776e-06,
+      "loss": 0.4345,
+      "step": 104
+    },
+    {
+      "epoch": 0.03134328358208955,
+      "grad_norm": 0.8987738306915273,
+      "learning_rate": 4.9999813005093556e-06,
+      "loss": 0.4682,
+      "step": 105
+    },
+    {
+      "epoch": 0.03164179104477612,
+      "grad_norm": 0.8997778964437559,
+      "learning_rate": 4.999970782066357e-06,
+      "loss": 0.4047,
+      "step": 106
+    },
+    {
+      "epoch": 0.03194029850746269,
+      "grad_norm": 0.9569315241820582,
+      "learning_rate": 4.999957926211613e-06,
+      "loss": 0.4584,
+      "step": 107
+    },
+    {
+      "epoch": 0.032238805970149255,
+      "grad_norm": 0.9078784862909539,
+      "learning_rate": 4.9999427329571445e-06,
+      "loss": 0.4117,
+      "step": 108
+    },
+    {
+      "epoch": 0.03253731343283582,
+      "grad_norm": 0.9616336053244321,
+      "learning_rate": 4.999925202317158e-06,
+      "loss": 0.4677,
+      "step": 109
+    },
+    {
+      "epoch": 0.03283582089552239,
+      "grad_norm": 0.8828990333654685,
+      "learning_rate": 4.9999053343080424e-06,
+      "loss": 0.454,
+      "step": 110
+    },
+    {
+      "epoch": 0.03313432835820895,
+      "grad_norm": 0.8971703986475054,
+      "learning_rate": 4.9998831289483745e-06,
+      "loss": 0.4401,
+      "step": 111
+    },
+    {
+      "epoch": 0.03343283582089552,
+      "grad_norm": 1.0410628753647846,
+      "learning_rate": 4.9998585862589165e-06,
+      "loss": 0.4655,
+      "step": 112
+    },
+    {
+      "epoch": 0.03373134328358209,
+      "grad_norm": 0.9301629000296456,
+      "learning_rate": 4.999831706262614e-06,
+      "loss": 0.4763,
+      "step": 113
+    },
+    {
+      "epoch": 0.03402985074626866,
+      "grad_norm": 0.8268955592230676,
+      "learning_rate": 4.999802488984598e-06,
+      "loss": 0.3696,
+      "step": 114
+    },
+    {
+      "epoch": 0.034328358208955224,
+      "grad_norm": 0.8664996618072721,
+      "learning_rate": 4.99977093445219e-06,
+      "loss": 0.3828,
+      "step": 115
+    },
+    {
+      "epoch": 0.03462686567164179,
+      "grad_norm": 1.0005293182043562,
+      "learning_rate": 4.999737042694889e-06,
+      "loss": 0.4978,
+      "step": 116
+    },
+    {
+      "epoch": 0.03492537313432836,
+      "grad_norm": 0.9140644611648783,
+      "learning_rate": 4.9997008137443845e-06,
+      "loss": 0.4464,
+      "step": 117
+    },
+    {
+      "epoch": 0.03522388059701492,
+      "grad_norm": 0.982768264944626,
+      "learning_rate": 4.99966224763455e-06,
+      "loss": 0.4059,
+      "step": 118
+    },
+    {
+      "epoch": 0.035522388059701496,
+      "grad_norm": 0.7935157638732689,
+      "learning_rate": 4.999621344401443e-06,
+      "loss": 0.3914,
+      "step": 119
+    },
+    {
+      "epoch": 0.03582089552238806,
+      "grad_norm": 0.8379591403713217,
+      "learning_rate": 4.999578104083307e-06,
+      "loss": 0.4461,
+      "step": 120
+    },
+    {
+      "epoch": 0.03611940298507463,
+      "grad_norm": 0.9772888203631277,
+      "learning_rate": 4.9995325267205715e-06,
+      "loss": 0.4251,
+      "step": 121
+    },
+    {
+      "epoch": 0.036417910447761194,
+      "grad_norm": 0.9050435971292533,
+      "learning_rate": 4.999484612355849e-06,
+      "loss": 0.4231,
+      "step": 122
+    },
+    {
+      "epoch": 0.03671641791044776,
+      "grad_norm": 0.8579300964981285,
+      "learning_rate": 4.99943436103394e-06,
+      "loss": 0.4711,
+      "step": 123
+    },
+    {
+      "epoch": 0.037014925373134326,
+      "grad_norm": 1.0200535489985578,
+      "learning_rate": 4.999381772801827e-06,
+      "loss": 0.4277,
+      "step": 124
+    },
+    {
+      "epoch": 0.03731343283582089,
+      "grad_norm": 0.9442204717239281,
+      "learning_rate": 4.999326847708678e-06,
+      "loss": 0.4448,
+      "step": 125
+    },
+    {
+      "epoch": 0.037611940298507465,
+      "grad_norm": 0.895546100855722,
+      "learning_rate": 4.999269585805848e-06,
+      "loss": 0.451,
+      "step": 126
+    },
+    {
+      "epoch": 0.03791044776119403,
+      "grad_norm": 0.9401669590071482,
+      "learning_rate": 4.999209987146876e-06,
+      "loss": 0.4118,
+      "step": 127
+    },
+    {
+      "epoch": 0.0382089552238806,
+      "grad_norm": 0.9947350109107067,
+      "learning_rate": 4.999148051787483e-06,
+      "loss": 0.4732,
+      "step": 128
+    },
+    {
+      "epoch": 0.038507462686567163,
+      "grad_norm": 0.9239692387463458,
+      "learning_rate": 4.999083779785579e-06,
+      "loss": 0.382,
+      "step": 129
+    },
+    {
+      "epoch": 0.03880597014925373,
+      "grad_norm": 0.8510728584830578,
+      "learning_rate": 4.999017171201256e-06,
+      "loss": 0.3563,
+      "step": 130
+    },
+    {
+      "epoch": 0.039104477611940296,
+      "grad_norm": 0.9327467858570172,
+      "learning_rate": 4.998948226096792e-06,
+      "loss": 0.4404,
+      "step": 131
+    },
+    {
+      "epoch": 0.03940298507462687,
+      "grad_norm": 0.9122306769001939,
+      "learning_rate": 4.998876944536648e-06,
+      "loss": 0.4465,
+      "step": 132
+    },
+    {
+      "epoch": 0.039701492537313435,
+      "grad_norm": 0.8628534940861644,
+      "learning_rate": 4.9988033265874714e-06,
+      "loss": 0.3978,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9890246363319865,
+      "learning_rate": 4.998727372318093e-06,
+      "loss": 0.4522,
+      "step": 134
+    },
+    {
+      "epoch": 0.04029850746268657,
+      "grad_norm": 0.8770139868884229,
+      "learning_rate": 4.998649081799528e-06,
+      "loss": 0.3612,
+      "step": 135
+    },
+    {
+      "epoch": 0.04059701492537313,
+      "grad_norm": 0.8875725344879101,
+      "learning_rate": 4.998568455104976e-06,
+      "loss": 0.431,
+      "step": 136
+    },
+    {
+      "epoch": 0.0408955223880597,
+      "grad_norm": 0.8689054951608837,
+      "learning_rate": 4.998485492309821e-06,
+      "loss": 0.4131,
+      "step": 137
+    },
+    {
+      "epoch": 0.04119402985074627,
+      "grad_norm": 0.8315965070349196,
+      "learning_rate": 4.998400193491632e-06,
+      "loss": 0.4025,
+      "step": 138
+    },
+    {
+      "epoch": 0.04149253731343284,
+      "grad_norm": 0.9156816933709063,
+      "learning_rate": 4.9983125587301594e-06,
+      "loss": 0.4201,
+      "step": 139
+    },
+    {
+      "epoch": 0.041791044776119404,
+      "grad_norm": 0.919113537897832,
+      "learning_rate": 4.998222588107342e-06,
+      "loss": 0.3936,
+      "step": 140
+    },
+    {
+      "epoch": 0.04208955223880597,
+      "grad_norm": 0.9372729637074827,
+      "learning_rate": 4.9981302817072984e-06,
+      "loss": 0.3789,
+      "step": 141
+    },
+    {
+      "epoch": 0.042388059701492536,
+      "grad_norm": 0.8930146991739137,
+      "learning_rate": 4.998035639616334e-06,
+      "loss": 0.4525,
+      "step": 142
+    },
+    {
+      "epoch": 0.0426865671641791,
+      "grad_norm": 0.9163631113724126,
+      "learning_rate": 4.997938661922936e-06,
+      "loss": 0.4033,
+      "step": 143
+    },
+    {
+      "epoch": 0.04298507462686567,
+      "grad_norm": 1.0610564882169262,
+      "learning_rate": 4.997839348717776e-06,
+      "loss": 0.4699,
+      "step": 144
+    },
+    {
+      "epoch": 0.04328358208955224,
+      "grad_norm": 0.9142847782134293,
+      "learning_rate": 4.99773770009371e-06,
+      "loss": 0.4134,
+      "step": 145
+    },
+    {
+      "epoch": 0.04358208955223881,
+      "grad_norm": 0.8539068181792682,
+      "learning_rate": 4.997633716145778e-06,
+      "loss": 0.3879,
+      "step": 146
+    },
+    {
+      "epoch": 0.043880597014925374,
+      "grad_norm": 0.908534305321226,
+      "learning_rate": 4.9975273969712e-06,
+      "loss": 0.4034,
+      "step": 147
+    },
+    {
+      "epoch": 0.04417910447761194,
+      "grad_norm": 0.9065531552905202,
+      "learning_rate": 4.997418742669383e-06,
+      "loss": 0.4624,
+      "step": 148
+    },
+    {
+      "epoch": 0.044477611940298506,
+      "grad_norm": 0.8900058271373764,
+      "learning_rate": 4.997307753341918e-06,
+      "loss": 0.4096,
+      "step": 149
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 0.8850620387207646,
+      "learning_rate": 4.997194429092573e-06,
+      "loss": 0.3899,
+      "step": 150
+    },
+    {
+      "epoch": 0.045074626865671645,
+      "grad_norm": 0.8701299536472148,
+      "learning_rate": 4.997078770027308e-06,
+      "loss": 0.4333,
+      "step": 151
+    },
+    {
+      "epoch": 0.04537313432835821,
+      "grad_norm": 1.020005231391521,
+      "learning_rate": 4.99696077625426e-06,
+      "loss": 0.4582,
+      "step": 152
+    },
+    {
+      "epoch": 0.04567164179104478,
+      "grad_norm": 1.042601527987692,
+      "learning_rate": 4.996840447883748e-06,
+      "loss": 0.4136,
+      "step": 153
+    },
+    {
+      "epoch": 0.04597014925373134,
+      "grad_norm": 1.0231933296827058,
+      "learning_rate": 4.996717785028278e-06,
+      "loss": 0.4359,
+      "step": 154
+    },
+    {
+      "epoch": 0.04626865671641791,
+      "grad_norm": 0.9352772150333463,
+      "learning_rate": 4.9965927878025365e-06,
+      "loss": 0.4572,
+      "step": 155
+    },
+    {
+      "epoch": 0.046567164179104475,
+      "grad_norm": 0.8932031869034676,
+      "learning_rate": 4.996465456323394e-06,
+      "loss": 0.4292,
+      "step": 156
+    },
+    {
+      "epoch": 0.04686567164179104,
+      "grad_norm": 0.9781906096105041,
+      "learning_rate": 4.9963357907099e-06,
+      "loss": 0.4399,
+      "step": 157
+    },
+    {
+      "epoch": 0.047164179104477615,
+      "grad_norm": 0.8702251247937537,
+      "learning_rate": 4.996203791083291e-06,
+      "loss": 0.3471,
+      "step": 158
+    },
+    {
+      "epoch": 0.04746268656716418,
+      "grad_norm": 0.891540485971903,
+      "learning_rate": 4.996069457566982e-06,
+      "loss": 0.3809,
+      "step": 159
+    },
+    {
+      "epoch": 0.04776119402985075,
+      "grad_norm": 0.8390421395659049,
+      "learning_rate": 4.995932790286572e-06,
+      "loss": 0.3728,
+      "step": 160
+    },
+    {
+      "epoch": 0.04805970149253731,
+      "grad_norm": 0.9759929778750225,
+      "learning_rate": 4.995793789369842e-06,
+      "loss": 0.3929,
+      "step": 161
+    },
+    {
+      "epoch": 0.04835820895522388,
+      "grad_norm": 0.9001180874449742,
+      "learning_rate": 4.9956524549467524e-06,
+      "loss": 0.4161,
+      "step": 162
+    },
+    {
+      "epoch": 0.048656716417910445,
+      "grad_norm": 1.0153327655108404,
+      "learning_rate": 4.995508787149451e-06,
+      "loss": 0.4279,
+      "step": 163
+    },
+    {
+      "epoch": 0.04895522388059702,
+      "grad_norm": 0.9940714318433066,
+      "learning_rate": 4.995362786112261e-06,
+      "loss": 0.4293,
+      "step": 164
+    },
+    {
+      "epoch": 0.049253731343283584,
+      "grad_norm": 0.8547509297200269,
+      "learning_rate": 4.99521445197169e-06,
+      "loss": 0.3878,
+      "step": 165
+    },
+    {
+      "epoch": 0.04955223880597015,
+      "grad_norm": 0.8085492708878349,
+      "learning_rate": 4.995063784866427e-06,
+      "loss": 0.3565,
+      "step": 166
+    },
+    {
+      "epoch": 0.049850746268656716,
+      "grad_norm": 0.9845859673203974,
+      "learning_rate": 4.994910784937343e-06,
+      "loss": 0.4446,
+      "step": 167
+    },
+    {
+      "epoch": 0.05014925373134328,
+      "grad_norm": 1.0461284807321518,
+      "learning_rate": 4.99475545232749e-06,
+      "loss": 0.4254,
+      "step": 168
+    },
+    {
+      "epoch": 0.05044776119402985,
+      "grad_norm": 0.9140198174691191,
+      "learning_rate": 4.994597787182097e-06,
+      "loss": 0.436,
+      "step": 169
+    },
+    {
+      "epoch": 0.050746268656716415,
+      "grad_norm": 0.9629152651211416,
+      "learning_rate": 4.99443778964858e-06,
+      "loss": 0.4742,
+      "step": 170
+    },
+    {
+      "epoch": 0.05104477611940299,
+      "grad_norm": 0.9693205236551692,
+      "learning_rate": 4.994275459876531e-06,
+      "loss": 0.4206,
+      "step": 171
+    },
+    {
+      "epoch": 0.051343283582089554,
+      "grad_norm": 0.885322451195088,
+      "learning_rate": 4.994110798017725e-06,
+      "loss": 0.3567,
+      "step": 172
+    },
+    {
+      "epoch": 0.05164179104477612,
+      "grad_norm": 0.9180041796566594,
+      "learning_rate": 4.993943804226117e-06,
+      "loss": 0.4015,
+      "step": 173
+    },
+    {
+      "epoch": 0.051940298507462686,
+      "grad_norm": 0.8915204742324517,
+      "learning_rate": 4.9937744786578425e-06,
+      "loss": 0.4378,
+      "step": 174
+    },
+    {
+      "epoch": 0.05223880597014925,
+      "grad_norm": 0.9057459288082191,
+      "learning_rate": 4.993602821471216e-06,
+      "loss": 0.3972,
+      "step": 175
+    },
+    {
+      "epoch": 0.05253731343283582,
+      "grad_norm": 0.9672531963127454,
+      "learning_rate": 4.993428832826734e-06,
+      "loss": 0.4127,
+      "step": 176
+    },
+    {
+      "epoch": 0.05283582089552239,
+      "grad_norm": 0.9087219850655597,
+      "learning_rate": 4.993252512887069e-06,
+      "loss": 0.4277,
+      "step": 177
+    },
+    {
+      "epoch": 0.05313432835820896,
+      "grad_norm": 0.8440962947146048,
+      "learning_rate": 4.993073861817078e-06,
+      "loss": 0.4241,
+      "step": 178
+    },
+    {
+      "epoch": 0.05343283582089552,
+      "grad_norm": 0.9406161881759132,
+      "learning_rate": 4.992892879783795e-06,
+      "loss": 0.4186,
+      "step": 179
+    },
+    {
+      "epoch": 0.05373134328358209,
+      "grad_norm": 0.9103960721738174,
+      "learning_rate": 4.992709566956435e-06,
+      "loss": 0.4243,
+      "step": 180
+    },
+    {
+      "epoch": 0.054029850746268655,
+      "grad_norm": 0.9301312469178198,
+      "learning_rate": 4.992523923506388e-06,
+      "loss": 0.4279,
+      "step": 181
+    },
+    {
+      "epoch": 0.05432835820895522,
+      "grad_norm": 0.9894235486712242,
+      "learning_rate": 4.992335949607229e-06,
+      "loss": 0.4456,
+      "step": 182
+    },
+    {
+      "epoch": 0.054626865671641794,
+      "grad_norm": 0.9146798098761367,
+      "learning_rate": 4.992145645434708e-06,
+      "loss": 0.3528,
+      "step": 183
+    },
+    {
+      "epoch": 0.05492537313432836,
+      "grad_norm": 0.9418084586681869,
+      "learning_rate": 4.991953011166753e-06,
+      "loss": 0.4134,
+      "step": 184
+    },
+    {
+      "epoch": 0.05522388059701493,
+      "grad_norm": 0.8443152449175418,
+      "learning_rate": 4.991758046983476e-06,
+      "loss": 0.3546,
+      "step": 185
+    },
+    {
+      "epoch": 0.05552238805970149,
+      "grad_norm": 0.8070965145975226,
+      "learning_rate": 4.991560753067161e-06,
+      "loss": 0.3852,
+      "step": 186
+    },
+    {
+      "epoch": 0.05582089552238806,
+      "grad_norm": 0.8907506048228214,
+      "learning_rate": 4.991361129602274e-06,
+      "loss": 0.4445,
+      "step": 187
+    },
+    {
+      "epoch": 0.056119402985074625,
+      "grad_norm": 0.834671843672075,
+      "learning_rate": 4.991159176775458e-06,
+      "loss": 0.3976,
+      "step": 188
+    },
+    {
+      "epoch": 0.05641791044776119,
+      "grad_norm": 0.9954785016124977,
+      "learning_rate": 4.9909548947755334e-06,
+      "loss": 0.4353,
+      "step": 189
+    },
+    {
+      "epoch": 0.056716417910447764,
+      "grad_norm": 0.7445588575209131,
+      "learning_rate": 4.990748283793499e-06,
+      "loss": 0.3131,
+      "step": 190
+    },
+    {
+      "epoch": 0.05701492537313433,
+      "grad_norm": 0.8570773114706813,
+      "learning_rate": 4.990539344022531e-06,
+      "loss": 0.3788,
+      "step": 191
+    },
+    {
+      "epoch": 0.057313432835820896,
+      "grad_norm": 1.0005429174734601,
+      "learning_rate": 4.990328075657985e-06,
+      "loss": 0.487,
+      "step": 192
+    },
+    {
+      "epoch": 0.05761194029850746,
+      "grad_norm": 0.8625742070670317,
+      "learning_rate": 4.990114478897389e-06,
+      "loss": 0.3654,
+      "step": 193
+    },
+    {
+      "epoch": 0.05791044776119403,
+      "grad_norm": 0.9721768330250884,
+      "learning_rate": 4.989898553940452e-06,
+      "loss": 0.3933,
+      "step": 194
+    },
+    {
+      "epoch": 0.058208955223880594,
+      "grad_norm": 0.9914169807245352,
+      "learning_rate": 4.989680300989058e-06,
+      "loss": 0.4732,
+      "step": 195
+    },
+    {
+      "epoch": 0.05850746268656717,
+      "grad_norm": 0.8411421119088517,
+      "learning_rate": 4.989459720247269e-06,
+      "loss": 0.4478,
+      "step": 196
+    },
+    {
+      "epoch": 0.05880597014925373,
+      "grad_norm": 0.924615837847109,
+      "learning_rate": 4.989236811921322e-06,
+      "loss": 0.4654,
+      "step": 197
+    },
+    {
+      "epoch": 0.0591044776119403,
+      "grad_norm": 0.9328373422116539,
+      "learning_rate": 4.989011576219632e-06,
+      "loss": 0.448,
+      "step": 198
+    },
+    {
+      "epoch": 0.059402985074626866,
+      "grad_norm": 0.9014890809533687,
+      "learning_rate": 4.9887840133527874e-06,
+      "loss": 0.4247,
+      "step": 199
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 0.8132765320225406,
+      "learning_rate": 4.988554123533554e-06,
+      "loss": 0.3943,
+      "step": 200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9241012173718349,
+      "learning_rate": 4.9883219069768744e-06,
+      "loss": 0.3633,
+      "step": 201
+    },
+    {
+      "epoch": 0.060298507462686564,
+      "grad_norm": 0.84126125400794,
+      "learning_rate": 4.988087363899864e-06,
+      "loss": 0.3297,
+      "step": 202
+    },
+    {
+      "epoch": 0.06059701492537314,
+      "grad_norm": 0.8347638005889463,
+      "learning_rate": 4.987850494521817e-06,
+      "loss": 0.4134,
+      "step": 203
+    },
+    {
+      "epoch": 0.0608955223880597,
+      "grad_norm": 0.8847082107220858,
+      "learning_rate": 4.987611299064197e-06,
+      "loss": 0.3754,
+      "step": 204
+    },
+    {
+      "epoch": 0.06119402985074627,
+      "grad_norm": 0.8981615926938369,
+      "learning_rate": 4.987369777750649e-06,
+      "loss": 0.407,
+      "step": 205
+    },
+    {
+      "epoch": 0.061492537313432835,
+      "grad_norm": 0.875779341191485,
+      "learning_rate": 4.9871259308069885e-06,
+      "loss": 0.4204,
+      "step": 206
+    },
+    {
+      "epoch": 0.0617910447761194,
+      "grad_norm": 0.9212258634041721,
+      "learning_rate": 4.986879758461207e-06,
+      "loss": 0.4021,
+      "step": 207
+    },
+    {
+      "epoch": 0.06208955223880597,
+      "grad_norm": 0.8857433024281737,
+      "learning_rate": 4.986631260943469e-06,
+      "loss": 0.3837,
+      "step": 208
+    },
+    {
+      "epoch": 0.06238805970149254,
+      "grad_norm": 0.967859200633979,
+      "learning_rate": 4.986380438486113e-06,
+      "loss": 0.3977,
+      "step": 209
+    },
+    {
+      "epoch": 0.0626865671641791,
+      "grad_norm": 0.925560346678998,
+      "learning_rate": 4.986127291323653e-06,
+      "loss": 0.4273,
+      "step": 210
+    },
+    {
+      "epoch": 0.06298507462686567,
+      "grad_norm": 0.863600208601708,
+      "learning_rate": 4.985871819692775e-06,
+      "loss": 0.4031,
+      "step": 211
+    },
+    {
+      "epoch": 0.06328358208955225,
+      "grad_norm": 0.9327927944187931,
+      "learning_rate": 4.985614023832339e-06,
+      "loss": 0.4167,
+      "step": 212
+    },
+    {
+      "epoch": 0.0635820895522388,
+      "grad_norm": 0.891025459096459,
+      "learning_rate": 4.985353903983377e-06,
+      "loss": 0.4188,
+      "step": 213
+    },
+    {
+      "epoch": 0.06388059701492538,
+      "grad_norm": 0.8627498456244966,
+      "learning_rate": 4.985091460389096e-06,
+      "loss": 0.3822,
+      "step": 214
+    },
+    {
+      "epoch": 0.06417910447761194,
+      "grad_norm": 0.9843279624632778,
+      "learning_rate": 4.9848266932948745e-06,
+      "loss": 0.3868,
+      "step": 215
+    },
+    {
+      "epoch": 0.06447761194029851,
+      "grad_norm": 0.9197216488420602,
+      "learning_rate": 4.984559602948261e-06,
+      "loss": 0.3866,
+      "step": 216
+    },
+    {
+      "epoch": 0.06477611940298507,
+      "grad_norm": 1.0380980582080264,
+      "learning_rate": 4.984290189598981e-06,
+      "loss": 0.4032,
+      "step": 217
+    },
+    {
+      "epoch": 0.06507462686567164,
+      "grad_norm": 0.8366437135304863,
+      "learning_rate": 4.984018453498928e-06,
+      "loss": 0.3525,
+      "step": 218
+    },
+    {
+      "epoch": 0.06537313432835822,
+      "grad_norm": 0.9649147229825452,
+      "learning_rate": 4.983744394902169e-06,
+      "loss": 0.4686,
+      "step": 219
+    },
+    {
+      "epoch": 0.06567164179104477,
+      "grad_norm": 0.8490377817349865,
+      "learning_rate": 4.983468014064942e-06,
+      "loss": 0.4265,
+      "step": 220
+    },
+    {
+      "epoch": 0.06597014925373135,
+      "grad_norm": 0.7792537616515028,
+      "learning_rate": 4.983189311245656e-06,
+      "loss": 0.3949,
+      "step": 221
+    },
+    {
+      "epoch": 0.0662686567164179,
+      "grad_norm": 0.8750639478343961,
+      "learning_rate": 4.982908286704893e-06,
+      "loss": 0.3828,
+      "step": 222
+    },
+    {
+      "epoch": 0.06656716417910448,
+      "grad_norm": 0.9455155367225592,
+      "learning_rate": 4.982624940705402e-06,
+      "loss": 0.4166,
+      "step": 223
+    },
+    {
+      "epoch": 0.06686567164179104,
+      "grad_norm": 0.9295734685485272,
+      "learning_rate": 4.982339273512106e-06,
+      "loss": 0.4153,
+      "step": 224
+    },
+    {
+      "epoch": 0.06716417910447761,
+      "grad_norm": 0.8631686207686246,
+      "learning_rate": 4.982051285392097e-06,
+      "loss": 0.3615,
+      "step": 225
+    },
+    {
+      "epoch": 0.06746268656716418,
+      "grad_norm": 0.9880414466922032,
+      "learning_rate": 4.981760976614634e-06,
+      "loss": 0.46,
+      "step": 226
+    },
+    {
+      "epoch": 0.06776119402985074,
+      "grad_norm": 0.8872311404032606,
+      "learning_rate": 4.981468347451154e-06,
+      "loss": 0.3871,
+      "step": 227
+    },
+    {
+      "epoch": 0.06805970149253732,
+      "grad_norm": 0.8943003479815301,
+      "learning_rate": 4.981173398175252e-06,
+      "loss": 0.371,
+      "step": 228
+    },
+    {
+      "epoch": 0.06835820895522388,
+      "grad_norm": 0.9525334440362165,
+      "learning_rate": 4.9808761290627035e-06,
+      "loss": 0.4159,
+      "step": 229
+    },
+    {
+      "epoch": 0.06865671641791045,
+      "grad_norm": 0.871749210147148,
+      "learning_rate": 4.9805765403914455e-06,
+      "loss": 0.3764,
+      "step": 230
+    },
+    {
+      "epoch": 0.06895522388059701,
+      "grad_norm": 0.925600078237541,
+      "learning_rate": 4.980274632441585e-06,
+      "loss": 0.3768,
+      "step": 231
+    },
+    {
+      "epoch": 0.06925373134328358,
+      "grad_norm": 0.9672291973292384,
+      "learning_rate": 4.9799704054954015e-06,
+      "loss": 0.477,
+      "step": 232
+    },
+    {
+      "epoch": 0.06955223880597015,
+      "grad_norm": 0.768340028516794,
+      "learning_rate": 4.979663859837337e-06,
+      "loss": 0.3902,
+      "step": 233
+    },
+    {
+      "epoch": 0.06985074626865671,
+      "grad_norm": 0.8740643362093679,
+      "learning_rate": 4.979354995754006e-06,
+      "loss": 0.3816,
+      "step": 234
+    },
+    {
+      "epoch": 0.07014925373134329,
+      "grad_norm": 1.0163576539062806,
+      "learning_rate": 4.979043813534189e-06,
+      "loss": 0.4156,
+      "step": 235
+    },
+    {
+      "epoch": 0.07044776119402985,
+      "grad_norm": 0.9103806310986855,
+      "learning_rate": 4.978730313468832e-06,
+      "loss": 0.4067,
+      "step": 236
+    },
+    {
+      "epoch": 0.07074626865671642,
+      "grad_norm": 0.922939829872911,
+      "learning_rate": 4.9784144958510515e-06,
+      "loss": 0.369,
+      "step": 237
+    },
+    {
+      "epoch": 0.07104477611940299,
+      "grad_norm": 1.0128842126660702,
+      "learning_rate": 4.978096360976129e-06,
+      "loss": 0.425,
+      "step": 238
+    },
+    {
+      "epoch": 0.07134328358208955,
+      "grad_norm": 0.9083583140182622,
+      "learning_rate": 4.977775909141513e-06,
+      "loss": 0.4008,
+      "step": 239
+    },
+    {
+      "epoch": 0.07164179104477612,
+      "grad_norm": 0.9204385171658356,
+      "learning_rate": 4.9774531406468164e-06,
+      "loss": 0.4098,
+      "step": 240
+    },
+    {
+      "epoch": 0.07194029850746268,
+      "grad_norm": 0.8800378945968806,
+      "learning_rate": 4.977128055793823e-06,
+      "loss": 0.4207,
+      "step": 241
+    },
+    {
+      "epoch": 0.07223880597014926,
+      "grad_norm": 0.9488253130303587,
+      "learning_rate": 4.976800654886476e-06,
+      "loss": 0.4467,
+      "step": 242
+    },
+    {
+      "epoch": 0.07253731343283581,
+      "grad_norm": 0.924076052886822,
+      "learning_rate": 4.976470938230889e-06,
+      "loss": 0.4257,
+      "step": 243
+    },
+    {
+      "epoch": 0.07283582089552239,
+      "grad_norm": 0.9270092762682594,
+      "learning_rate": 4.976138906135341e-06,
+      "loss": 0.4214,
+      "step": 244
+    },
+    {
+      "epoch": 0.07313432835820896,
+      "grad_norm": 0.897415607649877,
+      "learning_rate": 4.9758045589102696e-06,
+      "loss": 0.3953,
+      "step": 245
+    },
+    {
+      "epoch": 0.07343283582089552,
+      "grad_norm": 0.9972346268529109,
+      "learning_rate": 4.975467896868284e-06,
+      "loss": 0.4519,
+      "step": 246
+    },
+    {
+      "epoch": 0.0737313432835821,
+      "grad_norm": 0.9907973485734135,
+      "learning_rate": 4.9751289203241535e-06,
+      "loss": 0.4678,
+      "step": 247
+    },
+    {
+      "epoch": 0.07402985074626865,
+      "grad_norm": 0.8545804808558607,
+      "learning_rate": 4.974787629594815e-06,
+      "loss": 0.4088,
+      "step": 248
+    },
+    {
+      "epoch": 0.07432835820895523,
+      "grad_norm": 0.9288807835854102,
+      "learning_rate": 4.974444024999366e-06,
+      "loss": 0.4414,
+      "step": 249
+    },
+    {
+      "epoch": 0.07462686567164178,
+      "grad_norm": 0.7975187554309996,
+      "learning_rate": 4.974098106859068e-06,
+      "loss": 0.4158,
+      "step": 250
+    },
+    {
+      "epoch": 0.07492537313432836,
+      "grad_norm": 0.9898504646944991,
+      "learning_rate": 4.973749875497346e-06,
+      "loss": 0.4242,
+      "step": 251
+    },
+    {
+      "epoch": 0.07522388059701493,
+      "grad_norm": 0.8602699019118681,
+      "learning_rate": 4.973399331239789e-06,
+      "loss": 0.3789,
+      "step": 252
+    },
+    {
+      "epoch": 0.07552238805970149,
+      "grad_norm": 0.8750135247666223,
+      "learning_rate": 4.973046474414145e-06,
+      "loss": 0.3936,
+      "step": 253
+    },
+    {
+      "epoch": 0.07582089552238806,
+      "grad_norm": 0.9200590185086436,
+      "learning_rate": 4.9726913053503285e-06,
+      "loss": 0.4523,
+      "step": 254
+    },
+    {
+      "epoch": 0.07611940298507462,
+      "grad_norm": 0.9256789347650409,
+      "learning_rate": 4.972333824380414e-06,
+      "loss": 0.4328,
+      "step": 255
+    },
+    {
+      "epoch": 0.0764179104477612,
+      "grad_norm": 0.8363482584889479,
+      "learning_rate": 4.9719740318386375e-06,
+      "loss": 0.3833,
+      "step": 256
+    },
+    {
+      "epoch": 0.07671641791044777,
+      "grad_norm": 0.8386638424623721,
+      "learning_rate": 4.971611928061395e-06,
+      "loss": 0.4052,
+      "step": 257
+    },
+    {
+      "epoch": 0.07701492537313433,
+      "grad_norm": 0.7996227547339989,
+      "learning_rate": 4.9712475133872455e-06,
+      "loss": 0.3646,
+      "step": 258
+    },
+    {
+      "epoch": 0.0773134328358209,
+      "grad_norm": 0.8768980961009939,
+      "learning_rate": 4.970880788156906e-06,
+      "loss": 0.3561,
+      "step": 259
+    },
+    {
+      "epoch": 0.07761194029850746,
+      "grad_norm": 0.8551248770367569,
+      "learning_rate": 4.97051175271326e-06,
+      "loss": 0.3774,
+      "step": 260
+    },
+    {
+      "epoch": 0.07791044776119403,
+      "grad_norm": 0.8253804217485359,
+      "learning_rate": 4.970140407401343e-06,
+      "loss": 0.3577,
+      "step": 261
+    },
+    {
+      "epoch": 0.07820895522388059,
+      "grad_norm": 0.8907747036111339,
+      "learning_rate": 4.969766752568355e-06,
+      "loss": 0.3611,
+      "step": 262
+    },
+    {
+      "epoch": 0.07850746268656716,
+      "grad_norm": 0.9257974072192369,
+      "learning_rate": 4.969390788563653e-06,
+      "loss": 0.4132,
+      "step": 263
+    },
+    {
+      "epoch": 0.07880597014925374,
+      "grad_norm": 0.8701098416781633,
+      "learning_rate": 4.969012515738757e-06,
+      "loss": 0.3757,
+      "step": 264
+    },
+    {
+      "epoch": 0.0791044776119403,
+      "grad_norm": 0.8589961683118461,
+      "learning_rate": 4.9686319344473395e-06,
+      "loss": 0.3971,
+      "step": 265
+    },
+    {
+      "epoch": 0.07940298507462687,
+      "grad_norm": 0.9341286772453976,
+      "learning_rate": 4.968249045045237e-06,
+      "loss": 0.429,
+      "step": 266
+    },
+    {
+      "epoch": 0.07970149253731343,
+      "grad_norm": 0.9938003598001335,
+      "learning_rate": 4.967863847890441e-06,
+      "loss": 0.396,
+      "step": 267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9011497351262049,
+      "learning_rate": 4.9674763433431006e-06,
+      "loss": 0.4308,
+      "step": 268
+    },
+    {
+      "epoch": 0.08029850746268656,
+      "grad_norm": 0.977043100659228,
+      "learning_rate": 4.9670865317655245e-06,
+      "loss": 0.4258,
+      "step": 269
+    },
+    {
+      "epoch": 0.08059701492537313,
+      "grad_norm": 0.9059609196950806,
+      "learning_rate": 4.966694413522177e-06,
+      "loss": 0.4267,
+      "step": 270
+    },
+    {
+      "epoch": 0.0808955223880597,
+      "grad_norm": 1.0078844426437794,
+      "learning_rate": 4.966299988979678e-06,
+      "loss": 0.3847,
+      "step": 271
+    },
+    {
+      "epoch": 0.08119402985074627,
+      "grad_norm": 0.9703165705523532,
+      "learning_rate": 4.965903258506806e-06,
+      "loss": 0.4097,
+      "step": 272
+    },
+    {
+      "epoch": 0.08149253731343284,
+      "grad_norm": 0.8870735512921978,
+      "learning_rate": 4.965504222474494e-06,
+      "loss": 0.3928,
+      "step": 273
+    },
+    {
+      "epoch": 0.0817910447761194,
+      "grad_norm": 0.8739388689016976,
+      "learning_rate": 4.96510288125583e-06,
+      "loss": 0.3731,
+      "step": 274
+    },
+    {
+      "epoch": 0.08208955223880597,
+      "grad_norm": 1.0212824714114637,
+      "learning_rate": 4.9646992352260595e-06,
+      "loss": 0.3795,
+      "step": 275
+    },
+    {
+      "epoch": 0.08238805970149254,
+      "grad_norm": 0.8754845959865762,
+      "learning_rate": 4.964293284762581e-06,
+      "loss": 0.3911,
+      "step": 276
+    },
+    {
+      "epoch": 0.0826865671641791,
+      "grad_norm": 0.8698613925405617,
+      "learning_rate": 4.9638850302449485e-06,
+      "loss": 0.3532,
+      "step": 277
+    },
+    {
+      "epoch": 0.08298507462686568,
+      "grad_norm": 0.9357074185617845,
+      "learning_rate": 4.96347447205487e-06,
+      "loss": 0.4101,
+      "step": 278
+    },
+    {
+      "epoch": 0.08328358208955224,
+      "grad_norm": 0.8472354754739433,
+      "learning_rate": 4.963061610576207e-06,
+      "loss": 0.4406,
+      "step": 279
+    },
+    {
+      "epoch": 0.08358208955223881,
+      "grad_norm": 0.8564250146540285,
+      "learning_rate": 4.962646446194977e-06,
+      "loss": 0.3793,
+      "step": 280
+    },
+    {
+      "epoch": 0.08388059701492537,
+      "grad_norm": 0.8857581051447966,
+      "learning_rate": 4.962228979299345e-06,
+      "loss": 0.3822,
+      "step": 281
+    },
+    {
+      "epoch": 0.08417910447761194,
+      "grad_norm": 0.890998048990312,
+      "learning_rate": 4.961809210279634e-06,
+      "loss": 0.3944,
+      "step": 282
+    },
+    {
+      "epoch": 0.08447761194029851,
+      "grad_norm": 0.8601180120292999,
+      "learning_rate": 4.9613871395283195e-06,
+      "loss": 0.4009,
+      "step": 283
+    },
+    {
+      "epoch": 0.08477611940298507,
+      "grad_norm": 0.8277195700046859,
+      "learning_rate": 4.960962767440026e-06,
+      "loss": 0.3564,
+      "step": 284
+    },
+    {
+      "epoch": 0.08507462686567165,
+      "grad_norm": 0.8423871373655928,
+      "learning_rate": 4.96053609441153e-06,
+      "loss": 0.3747,
+      "step": 285
+    },
+    {
+      "epoch": 0.0853731343283582,
+      "grad_norm": 0.9407464456549731,
+      "learning_rate": 4.960107120841762e-06,
+      "loss": 0.3858,
+      "step": 286
+    },
+    {
+      "epoch": 0.08567164179104478,
+      "grad_norm": 0.9140065601660698,
+      "learning_rate": 4.9596758471318e-06,
+      "loss": 0.3961,
+      "step": 287
+    },
+    {
+      "epoch": 0.08597014925373134,
+      "grad_norm": 0.858838528185795,
+      "learning_rate": 4.959242273684878e-06,
+      "loss": 0.3915,
+      "step": 288
+    },
+    {
+      "epoch": 0.08626865671641791,
+      "grad_norm": 0.9199463736666107,
+      "learning_rate": 4.958806400906372e-06,
+      "loss": 0.4275,
+      "step": 289
+    },
+    {
+      "epoch": 0.08656716417910448,
+      "grad_norm": 0.9243014960259777,
+      "learning_rate": 4.958368229203816e-06,
+      "loss": 0.3851,
+      "step": 290
+    },
+    {
+      "epoch": 0.08686567164179104,
+      "grad_norm": 0.9091623645910903,
+      "learning_rate": 4.957927758986888e-06,
+      "loss": 0.3957,
+      "step": 291
+    },
+    {
+      "epoch": 0.08716417910447762,
+      "grad_norm": 0.8915385371700681,
+      "learning_rate": 4.9574849906674174e-06,
+      "loss": 0.4437,
+      "step": 292
+    },
+    {
+      "epoch": 0.08746268656716417,
+      "grad_norm": 0.8934354732288062,
+      "learning_rate": 4.957039924659382e-06,
+      "loss": 0.358,
+      "step": 293
+    },
+    {
+      "epoch": 0.08776119402985075,
+      "grad_norm": 0.9061503598549887,
+      "learning_rate": 4.956592561378907e-06,
+      "loss": 0.3692,
+      "step": 294
+    },
+    {
+      "epoch": 0.0880597014925373,
+      "grad_norm": 0.8830535399018076,
+      "learning_rate": 4.956142901244268e-06,
+      "loss": 0.4223,
+      "step": 295
+    },
+    {
+      "epoch": 0.08835820895522388,
+      "grad_norm": 0.9413406518046492,
+      "learning_rate": 4.955690944675882e-06,
+      "loss": 0.3859,
+      "step": 296
+    },
+    {
+      "epoch": 0.08865671641791045,
+      "grad_norm": 0.8947764680217752,
+      "learning_rate": 4.955236692096324e-06,
+      "loss": 0.3682,
+      "step": 297
+    },
+    {
+      "epoch": 0.08895522388059701,
+      "grad_norm": 0.8405209768536294,
+      "learning_rate": 4.954780143930303e-06,
+      "loss": 0.3826,
+      "step": 298
+    },
+    {
+      "epoch": 0.08925373134328358,
+      "grad_norm": 0.9396234783746453,
+      "learning_rate": 4.954321300604683e-06,
+      "loss": 0.4112,
+      "step": 299
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 0.9114411631804026,
+      "learning_rate": 4.953860162548472e-06,
+      "loss": 0.3612,
+      "step": 300
+    },
+    {
+      "epoch": 0.08985074626865672,
+      "grad_norm": 0.8817713479006289,
+      "learning_rate": 4.953396730192821e-06,
+      "loss": 0.4209,
+      "step": 301
+    },
+    {
+      "epoch": 0.09014925373134329,
+      "grad_norm": 0.8726071936119882,
+      "learning_rate": 4.952931003971029e-06,
+      "loss": 0.4155,
+      "step": 302
+    },
+    {
+      "epoch": 0.09044776119402985,
+      "grad_norm": 0.9048765061805111,
+      "learning_rate": 4.952462984318539e-06,
+      "loss": 0.3896,
+      "step": 303
+    },
+    {
+      "epoch": 0.09074626865671642,
+      "grad_norm": 0.9804012864737665,
+      "learning_rate": 4.9519926716729376e-06,
+      "loss": 0.4211,
+      "step": 304
+    },
+    {
+      "epoch": 0.09104477611940298,
+      "grad_norm": 0.8873017745256736,
+      "learning_rate": 4.951520066473955e-06,
+      "loss": 0.4544,
+      "step": 305
+    },
+    {
+      "epoch": 0.09134328358208955,
+      "grad_norm": 0.8062695091689972,
+      "learning_rate": 4.951045169163467e-06,
+      "loss": 0.3992,
+      "step": 306
+    },
+    {
+      "epoch": 0.09164179104477611,
+      "grad_norm": 0.8777340484612656,
+      "learning_rate": 4.950567980185489e-06,
+      "loss": 0.4128,
+      "step": 307
+    },
+    {
+      "epoch": 0.09194029850746269,
+      "grad_norm": 0.9366219952901376,
+      "learning_rate": 4.950088499986183e-06,
+      "loss": 0.3942,
+      "step": 308
+    },
+    {
+      "epoch": 0.09223880597014926,
+      "grad_norm": 0.8144148378243571,
+      "learning_rate": 4.949606729013851e-06,
+      "loss": 0.3958,
+      "step": 309
+    },
+    {
+      "epoch": 0.09253731343283582,
+      "grad_norm": 0.8968060502801483,
+      "learning_rate": 4.949122667718935e-06,
+      "loss": 0.406,
+      "step": 310
+    },
+    {
+      "epoch": 0.09283582089552239,
+      "grad_norm": 0.931747156022598,
+      "learning_rate": 4.948636316554023e-06,
+      "loss": 0.3839,
+      "step": 311
+    },
+    {
+      "epoch": 0.09313432835820895,
+      "grad_norm": 0.7978130697742478,
+      "learning_rate": 4.948147675973841e-06,
+      "loss": 0.3653,
+      "step": 312
+    },
+    {
+      "epoch": 0.09343283582089552,
+      "grad_norm": 0.7934615187323222,
+      "learning_rate": 4.947656746435255e-06,
+      "loss": 0.3813,
+      "step": 313
+    },
+    {
+      "epoch": 0.09373134328358208,
+      "grad_norm": 0.9589131627871469,
+      "learning_rate": 4.947163528397273e-06,
+      "loss": 0.469,
+      "step": 314
+    },
+    {
+      "epoch": 0.09402985074626866,
+      "grad_norm": 0.8464181634763774,
+      "learning_rate": 4.946668022321042e-06,
+      "loss": 0.3979,
+      "step": 315
+    },
+    {
+      "epoch": 0.09432835820895523,
+      "grad_norm": 0.8899369392024041,
+      "learning_rate": 4.946170228669847e-06,
+      "loss": 0.3292,
+      "step": 316
+    },
+    {
+      "epoch": 0.09462686567164179,
+      "grad_norm": 0.8553955124203195,
+      "learning_rate": 4.9456701479091155e-06,
+      "loss": 0.3925,
+      "step": 317
+    },
+    {
+      "epoch": 0.09492537313432836,
+      "grad_norm": 0.8606099277635237,
+      "learning_rate": 4.945167780506407e-06,
+      "loss": 0.433,
+      "step": 318
+    },
+    {
+      "epoch": 0.09522388059701492,
+      "grad_norm": 0.907825007701508,
+      "learning_rate": 4.944663126931426e-06,
+      "loss": 0.3645,
+      "step": 319
+    },
+    {
+      "epoch": 0.0955223880597015,
+      "grad_norm": 0.8612034752290371,
+      "learning_rate": 4.94415618765601e-06,
+      "loss": 0.3607,
+      "step": 320
+    },
+    {
+      "epoch": 0.09582089552238807,
+      "grad_norm": 0.9069258389724871,
+      "learning_rate": 4.943646963154134e-06,
+      "loss": 0.402,
+      "step": 321
+    },
+    {
+      "epoch": 0.09611940298507463,
+      "grad_norm": 0.9862506026387452,
+      "learning_rate": 4.943135453901911e-06,
+      "loss": 0.4424,
+      "step": 322
+    },
+    {
+      "epoch": 0.0964179104477612,
+      "grad_norm": 0.9122656478208079,
+      "learning_rate": 4.942621660377592e-06,
+      "loss": 0.4595,
+      "step": 323
+    },
+    {
+      "epoch": 0.09671641791044776,
+      "grad_norm": 0.9125176637122425,
+      "learning_rate": 4.942105583061558e-06,
+      "loss": 0.3824,
+      "step": 324
+    },
+    {
+      "epoch": 0.09701492537313433,
+      "grad_norm": 0.8593735642495912,
+      "learning_rate": 4.941587222436331e-06,
+      "loss": 0.3785,
+      "step": 325
+    },
+    {
+      "epoch": 0.09731343283582089,
+      "grad_norm": 0.82674079067205,
+      "learning_rate": 4.941066578986565e-06,
+      "loss": 0.3416,
+      "step": 326
+    },
+    {
+      "epoch": 0.09761194029850746,
+      "grad_norm": 0.9689144745422055,
+      "learning_rate": 4.940543653199049e-06,
+      "loss": 0.3368,
+      "step": 327
+    },
+    {
+      "epoch": 0.09791044776119404,
+      "grad_norm": 0.9277054433752907,
+      "learning_rate": 4.940018445562704e-06,
+      "loss": 0.3761,
+      "step": 328
+    },
+    {
+      "epoch": 0.0982089552238806,
+      "grad_norm": 0.7578752447120486,
+      "learning_rate": 4.939490956568589e-06,
+      "loss": 0.3509,
+      "step": 329
+    },
+    {
+      "epoch": 0.09850746268656717,
+      "grad_norm": 0.8298981698879738,
+      "learning_rate": 4.938961186709893e-06,
+      "loss": 0.3969,
+      "step": 330
+    },
+    {
+      "epoch": 0.09880597014925373,
+      "grad_norm": 0.9238719788718148,
+      "learning_rate": 4.938429136481936e-06,
+      "loss": 0.4294,
+      "step": 331
+    },
+    {
+      "epoch": 0.0991044776119403,
+      "grad_norm": 0.8589172708753758,
+      "learning_rate": 4.937894806382173e-06,
+      "loss": 0.4139,
+      "step": 332
+    },
+    {
+      "epoch": 0.09940298507462686,
+      "grad_norm": 0.8585178801953061,
+      "learning_rate": 4.937358196910191e-06,
+      "loss": 0.395,
+      "step": 333
+    },
+    {
+      "epoch": 0.09970149253731343,
+      "grad_norm": 0.8099539599389111,
+      "learning_rate": 4.936819308567705e-06,
+      "loss": 0.3669,
+      "step": 334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.839470127485901,
+      "learning_rate": 4.9362781418585635e-06,
+      "loss": 0.3418,
+      "step": 335
+    },
+    {
+      "epoch": 0.10029850746268656,
+      "grad_norm": 0.9244240560709288,
+      "learning_rate": 4.9357346972887425e-06,
+      "loss": 0.3656,
+      "step": 336
+    },
+    {
+      "epoch": 0.10059701492537314,
+      "grad_norm": 0.8543828203767136,
+      "learning_rate": 4.935188975366352e-06,
+      "loss": 0.3853,
+      "step": 337
+    },
+    {
+      "epoch": 0.1008955223880597,
+      "grad_norm": 0.9403997055915649,
+      "learning_rate": 4.934640976601627e-06,
+      "loss": 0.4266,
+      "step": 338
+    },
+    {
+      "epoch": 0.10119402985074627,
+      "grad_norm": 0.8656569961847842,
+      "learning_rate": 4.934090701506933e-06,
+      "loss": 0.3519,
+      "step": 339
+    },
+    {
+      "epoch": 0.10149253731343283,
+      "grad_norm": 1.0259990423516157,
+      "learning_rate": 4.9335381505967635e-06,
+      "loss": 0.3739,
+      "step": 340
+    },
+    {
+      "epoch": 0.1017910447761194,
+      "grad_norm": 0.9184253012427269,
+      "learning_rate": 4.932983324387742e-06,
+      "loss": 0.3891,
+      "step": 341
+    },
+    {
+      "epoch": 0.10208955223880598,
+      "grad_norm": 0.8855470935863624,
+      "learning_rate": 4.932426223398615e-06,
+      "loss": 0.4003,
+      "step": 342
+    },
+    {
+      "epoch": 0.10238805970149253,
+      "grad_norm": 0.8338863353994163,
+      "learning_rate": 4.9318668481502604e-06,
+      "loss": 0.3808,
+      "step": 343
+    },
+    {
+      "epoch": 0.10268656716417911,
+      "grad_norm": 0.8659855930783819,
+      "learning_rate": 4.93130519916568e-06,
+      "loss": 0.4153,
+      "step": 344
+    },
+    {
+      "epoch": 0.10298507462686567,
+      "grad_norm": 0.9328940913851155,
+      "learning_rate": 4.930741276970001e-06,
+      "loss": 0.4228,
+      "step": 345
+    },
+    {
+      "epoch": 0.10328358208955224,
+      "grad_norm": 0.9559469401085737,
+      "learning_rate": 4.930175082090477e-06,
+      "loss": 0.4171,
+      "step": 346
+    },
+    {
+      "epoch": 0.10358208955223881,
+      "grad_norm": 0.8361138110258766,
+      "learning_rate": 4.929606615056488e-06,
+      "loss": 0.3968,
+      "step": 347
+    },
+    {
+      "epoch": 0.10388059701492537,
+      "grad_norm": 0.8329215393697887,
+      "learning_rate": 4.929035876399535e-06,
+      "loss": 0.3942,
+      "step": 348
+    },
+    {
+      "epoch": 0.10417910447761194,
+      "grad_norm": 0.8675645842119557,
+      "learning_rate": 4.9284628666532455e-06,
+      "loss": 0.3621,
+      "step": 349
+    },
+    {
+      "epoch": 0.1044776119402985,
+      "grad_norm": 0.8507052675194776,
+      "learning_rate": 4.927887586353369e-06,
+      "loss": 0.4286,
+      "step": 350
+    },
+    {
+      "epoch": 0.10477611940298508,
+      "grad_norm": 0.837156171946447,
+      "learning_rate": 4.92731003603778e-06,
+      "loss": 0.3712,
+      "step": 351
+    },
+    {
+      "epoch": 0.10507462686567164,
+      "grad_norm": 0.9356486848372598,
+      "learning_rate": 4.926730216246472e-06,
+      "loss": 0.3913,
+      "step": 352
+    },
+    {
+      "epoch": 0.10537313432835821,
+      "grad_norm": 0.8223936751125491,
+      "learning_rate": 4.926148127521565e-06,
+      "loss": 0.3975,
+      "step": 353
+    },
+    {
+      "epoch": 0.10567164179104478,
+      "grad_norm": 0.7881227208588917,
+      "learning_rate": 4.925563770407295e-06,
+      "loss": 0.3961,
+      "step": 354
+    },
+    {
+      "epoch": 0.10597014925373134,
+      "grad_norm": 0.8453911169822759,
+      "learning_rate": 4.924977145450023e-06,
+      "loss": 0.3694,
+      "step": 355
+    },
+    {
+      "epoch": 0.10626865671641791,
+      "grad_norm": 0.8734420055725053,
+      "learning_rate": 4.924388253198229e-06,
+      "loss": 0.3865,
+      "step": 356
+    },
+    {
+      "epoch": 0.10656716417910447,
+      "grad_norm": 0.9201290119245594,
+      "learning_rate": 4.923797094202514e-06,
+      "loss": 0.4278,
+      "step": 357
+    },
+    {
+      "epoch": 0.10686567164179105,
+      "grad_norm": 0.9389093005855279,
+      "learning_rate": 4.923203669015594e-06,
+      "loss": 0.3665,
+      "step": 358
+    },
+    {
+      "epoch": 0.1071641791044776,
+      "grad_norm": 0.9042202134139647,
+      "learning_rate": 4.92260797819231e-06,
+      "loss": 0.3924,
+      "step": 359
+    },
+    {
+      "epoch": 0.10746268656716418,
+      "grad_norm": 0.8657123819229612,
+      "learning_rate": 4.922010022289618e-06,
+      "loss": 0.3972,
+      "step": 360
+    },
+    {
+      "epoch": 0.10776119402985075,
+      "grad_norm": 0.829479673932315,
+      "learning_rate": 4.921409801866591e-06,
+      "loss": 0.4135,
+      "step": 361
+    },
+    {
+      "epoch": 0.10805970149253731,
+      "grad_norm": 0.8821755114348971,
+      "learning_rate": 4.920807317484422e-06,
+      "loss": 0.3944,
+      "step": 362
+    },
+    {
+      "epoch": 0.10835820895522388,
+      "grad_norm": 0.8948274549142421,
+      "learning_rate": 4.920202569706418e-06,
+      "loss": 0.3826,
+      "step": 363
+    },
+    {
+      "epoch": 0.10865671641791044,
+      "grad_norm": 0.8652728701574908,
+      "learning_rate": 4.919595559098003e-06,
+      "loss": 0.3766,
+      "step": 364
+    },
+    {
+      "epoch": 0.10895522388059702,
+      "grad_norm": 0.9092751007745377,
+      "learning_rate": 4.9189862862267205e-06,
+      "loss": 0.4034,
+      "step": 365
+    },
+    {
+      "epoch": 0.10925373134328359,
+      "grad_norm": 0.9254208863560162,
+      "learning_rate": 4.918374751662221e-06,
+      "loss": 0.4359,
+      "step": 366
+    },
+    {
+      "epoch": 0.10955223880597015,
+      "grad_norm": 0.8168960720647488,
+      "learning_rate": 4.917760955976277e-06,
+      "loss": 0.4363,
+      "step": 367
+    },
+    {
+      "epoch": 0.10985074626865672,
+      "grad_norm": 0.830653386436508,
+      "learning_rate": 4.917144899742773e-06,
+      "loss": 0.3931,
+      "step": 368
+    },
+    {
+      "epoch": 0.11014925373134328,
+      "grad_norm": 0.840070982853219,
+      "learning_rate": 4.916526583537705e-06,
+      "loss": 0.3493,
+      "step": 369
+    },
+    {
+      "epoch": 0.11044776119402985,
+      "grad_norm": 0.979140168040529,
+      "learning_rate": 4.915906007939184e-06,
+      "loss": 0.3741,
+      "step": 370
+    },
+    {
+      "epoch": 0.11074626865671641,
+      "grad_norm": 0.9450117540952754,
+      "learning_rate": 4.915283173527434e-06,
+      "loss": 0.4575,
+      "step": 371
+    },
+    {
+      "epoch": 0.11104477611940299,
+      "grad_norm": 0.877577439309941,
+      "learning_rate": 4.9146580808847896e-06,
+      "loss": 0.4069,
+      "step": 372
+    },
+    {
+      "epoch": 0.11134328358208956,
+      "grad_norm": 0.8452549449469758,
+      "learning_rate": 4.9140307305956964e-06,
+      "loss": 0.3521,
+      "step": 373
+    },
+    {
+      "epoch": 0.11164179104477612,
+      "grad_norm": 0.8406721896093431,
+      "learning_rate": 4.913401123246713e-06,
+      "loss": 0.3441,
+      "step": 374
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 0.7975661106565912,
+      "learning_rate": 4.912769259426505e-06,
+      "loss": 0.3256,
+      "step": 375
+    },
+    {
+      "epoch": 0.11223880597014925,
+      "grad_norm": 0.8427317452222252,
+      "learning_rate": 4.912135139725851e-06,
+      "loss": 0.3885,
+      "step": 376
+    },
+    {
+      "epoch": 0.11253731343283582,
+      "grad_norm": 0.9053683742245764,
+      "learning_rate": 4.9114987647376374e-06,
+      "loss": 0.4163,
+      "step": 377
+    },
+    {
+      "epoch": 0.11283582089552238,
+      "grad_norm": 0.8630516315249676,
+      "learning_rate": 4.910860135056859e-06,
+      "loss": 0.3739,
+      "step": 378
+    },
+    {
+      "epoch": 0.11313432835820895,
+      "grad_norm": 0.9602461144012868,
+      "learning_rate": 4.91021925128062e-06,
+      "loss": 0.3943,
+      "step": 379
+    },
+    {
+      "epoch": 0.11343283582089553,
+      "grad_norm": 0.9141865424324758,
+      "learning_rate": 4.909576114008129e-06,
+      "loss": 0.4037,
+      "step": 380
+    },
+    {
+      "epoch": 0.11373134328358209,
+      "grad_norm": 0.8548880538979818,
+      "learning_rate": 4.908930723840706e-06,
+      "loss": 0.3818,
+      "step": 381
+    },
+    {
+      "epoch": 0.11402985074626866,
+      "grad_norm": 1.0328188865088241,
+      "learning_rate": 4.908283081381773e-06,
+      "loss": 0.4682,
+      "step": 382
+    },
+    {
+      "epoch": 0.11432835820895522,
+      "grad_norm": 0.9238076147148401,
+      "learning_rate": 4.907633187236861e-06,
+      "loss": 0.38,
+      "step": 383
+    },
+    {
+      "epoch": 0.11462686567164179,
+      "grad_norm": 0.8455481748001109,
+      "learning_rate": 4.906981042013605e-06,
+      "loss": 0.3761,
+      "step": 384
+    },
+    {
+      "epoch": 0.11492537313432835,
+      "grad_norm": 0.9073460304019739,
+      "learning_rate": 4.9063266463217466e-06,
+      "loss": 0.4005,
+      "step": 385
+    },
+    {
+      "epoch": 0.11522388059701492,
+      "grad_norm": 0.8931846505792465,
+      "learning_rate": 4.905670000773126e-06,
+      "loss": 0.4288,
+      "step": 386
+    },
+    {
+      "epoch": 0.1155223880597015,
+      "grad_norm": 0.851731424120729,
+      "learning_rate": 4.905011105981694e-06,
+      "loss": 0.4044,
+      "step": 387
+    },
+    {
+      "epoch": 0.11582089552238806,
+      "grad_norm": 0.9161247111493643,
+      "learning_rate": 4.9043499625635e-06,
+      "loss": 0.3995,
+      "step": 388
+    },
+    {
+      "epoch": 0.11611940298507463,
+      "grad_norm": 0.8797022877561539,
+      "learning_rate": 4.903686571136697e-06,
+      "loss": 0.3964,
+      "step": 389
+    },
+    {
+      "epoch": 0.11641791044776119,
+      "grad_norm": 0.8500731359455521,
+      "learning_rate": 4.903020932321541e-06,
+      "loss": 0.3609,
+      "step": 390
+    },
+    {
+      "epoch": 0.11671641791044776,
+      "grad_norm": 0.9945238156321736,
+      "learning_rate": 4.9023530467403856e-06,
+      "loss": 0.3801,
+      "step": 391
+    },
+    {
+      "epoch": 0.11701492537313433,
+      "grad_norm": 0.9136420198050205,
+      "learning_rate": 4.901682915017689e-06,
+      "loss": 0.3566,
+      "step": 392
+    },
+    {
+      "epoch": 0.1173134328358209,
+      "grad_norm": 0.8189636216415471,
+      "learning_rate": 4.901010537780009e-06,
+      "loss": 0.3132,
+      "step": 393
+    },
+    {
+      "epoch": 0.11761194029850747,
+      "grad_norm": 0.9844874224310621,
+      "learning_rate": 4.900335915656e-06,
+      "loss": 0.4271,
+      "step": 394
+    },
+    {
+      "epoch": 0.11791044776119403,
+      "grad_norm": 1.0487555124306993,
+      "learning_rate": 4.899659049276418e-06,
+      "loss": 0.4064,
+      "step": 395
+    },
+    {
+      "epoch": 0.1182089552238806,
+      "grad_norm": 0.7812485151038709,
+      "learning_rate": 4.898979939274118e-06,
+      "loss": 0.2882,
+      "step": 396
+    },
+    {
+      "epoch": 0.11850746268656716,
+      "grad_norm": 0.8336948833618268,
+      "learning_rate": 4.898298586284049e-06,
+      "loss": 0.3984,
+      "step": 397
+    },
+    {
+      "epoch": 0.11880597014925373,
+      "grad_norm": 0.8979301268641166,
+      "learning_rate": 4.897614990943261e-06,
+      "loss": 0.3919,
+      "step": 398
+    },
+    {
+      "epoch": 0.1191044776119403,
+      "grad_norm": 0.8427489040474724,
+      "learning_rate": 4.896929153890898e-06,
+      "loss": 0.3866,
+      "step": 399
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 0.8211103743103699,
+      "learning_rate": 4.896241075768201e-06,
+      "loss": 0.3963,
+      "step": 400
+    },
+    {
+      "epoch": 0.11970149253731344,
+      "grad_norm": 0.8445033296486258,
+      "learning_rate": 4.895550757218507e-06,
+      "loss": 0.3928,
+      "step": 401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8961945171020185,
+      "learning_rate": 4.894858198887246e-06,
+      "loss": 0.3794,
+      "step": 402
+    },
+    {
+      "epoch": 0.12029850746268657,
+      "grad_norm": 0.9818424108109866,
+      "learning_rate": 4.8941634014219454e-06,
+      "loss": 0.39,
+      "step": 403
+    },
+    {
+      "epoch": 0.12059701492537313,
+      "grad_norm": 0.8104054934733214,
+      "learning_rate": 4.8934663654722205e-06,
+      "loss": 0.2999,
+      "step": 404
+    },
+    {
+      "epoch": 0.1208955223880597,
+      "grad_norm": 0.8316791108804557,
+      "learning_rate": 4.892767091689786e-06,
+      "loss": 0.3685,
+      "step": 405
+    },
+    {
+      "epoch": 0.12119402985074627,
+      "grad_norm": 0.8310506029399883,
+      "learning_rate": 4.892065580728444e-06,
+      "loss": 0.3703,
+      "step": 406
+    },
+    {
+      "epoch": 0.12149253731343283,
+      "grad_norm": 0.7567215169549178,
+      "learning_rate": 4.8913618332440906e-06,
+      "loss": 0.3268,
+      "step": 407
+    },
+    {
+      "epoch": 0.1217910447761194,
+      "grad_norm": 0.8628920449752951,
+      "learning_rate": 4.890655849894713e-06,
+      "loss": 0.3912,
+      "step": 408
+    },
+    {
+      "epoch": 0.12208955223880597,
+      "grad_norm": 0.910090152736456,
+      "learning_rate": 4.889947631340388e-06,
+      "loss": 0.4308,
+      "step": 409
+    },
+    {
+      "epoch": 0.12238805970149254,
+      "grad_norm": 0.8591509812041254,
+      "learning_rate": 4.889237178243283e-06,
+      "loss": 0.3958,
+      "step": 410
+    },
+    {
+      "epoch": 0.12268656716417911,
+      "grad_norm": 0.8618520589363041,
+      "learning_rate": 4.888524491267653e-06,
+      "loss": 0.4158,
+      "step": 411
+    },
+    {
+      "epoch": 0.12298507462686567,
+      "grad_norm": 0.8884599985876102,
+      "learning_rate": 4.887809571079845e-06,
+      "loss": 0.3612,
+      "step": 412
+    },
+    {
+      "epoch": 0.12328358208955224,
+      "grad_norm": 0.8985780635048177,
+      "learning_rate": 4.88709241834829e-06,
+      "loss": 0.3824,
+      "step": 413
+    },
+    {
+      "epoch": 0.1235820895522388,
+      "grad_norm": 0.8673597426156886,
+      "learning_rate": 4.88637303374351e-06,
+      "loss": 0.3911,
+      "step": 414
+    },
+    {
+      "epoch": 0.12388059701492538,
+      "grad_norm": 0.8761841886398808,
+      "learning_rate": 4.885651417938112e-06,
+      "loss": 0.3596,
+      "step": 415
+    },
+    {
+      "epoch": 0.12417910447761193,
+      "grad_norm": 0.8576660751720805,
+      "learning_rate": 4.884927571606788e-06,
+      "loss": 0.4177,
+      "step": 416
+    },
+    {
+      "epoch": 0.12447761194029851,
+      "grad_norm": 0.804474354480795,
+      "learning_rate": 4.884201495426317e-06,
+      "loss": 0.3581,
+      "step": 417
+    },
+    {
+      "epoch": 0.12477611940298508,
+      "grad_norm": 0.8519755633647428,
+      "learning_rate": 4.883473190075562e-06,
+      "loss": 0.4146,
+      "step": 418
+    },
+    {
+      "epoch": 0.12507462686567164,
+      "grad_norm": 0.864412561136686,
+      "learning_rate": 4.882742656235474e-06,
+      "loss": 0.3788,
+      "step": 419
+    },
+    {
+      "epoch": 0.1253731343283582,
+      "grad_norm": 0.88627834971777,
+      "learning_rate": 4.88200989458908e-06,
+      "loss": 0.4063,
+      "step": 420
+    },
+    {
+      "epoch": 0.12567164179104479,
+      "grad_norm": 0.8104996818816044,
+      "learning_rate": 4.881274905821496e-06,
+      "loss": 0.4179,
+      "step": 421
+    },
+    {
+      "epoch": 0.12597014925373134,
+      "grad_norm": 0.7961290288194648,
+      "learning_rate": 4.88053769061992e-06,
+      "loss": 0.3743,
+      "step": 422
+    },
+    {
+      "epoch": 0.1262686567164179,
+      "grad_norm": 0.8727255096641454,
+      "learning_rate": 4.879798249673628e-06,
+      "loss": 0.4076,
+      "step": 423
+    },
+    {
+      "epoch": 0.1265671641791045,
+      "grad_norm": 0.805037045243907,
+      "learning_rate": 4.87905658367398e-06,
+      "loss": 0.3782,
+      "step": 424
+    },
+    {
+      "epoch": 0.12686567164179105,
+      "grad_norm": 0.8467074530308887,
+      "learning_rate": 4.878312693314417e-06,
+      "loss": 0.3981,
+      "step": 425
+    },
+    {
+      "epoch": 0.1271641791044776,
+      "grad_norm": 0.8591176817546119,
+      "learning_rate": 4.877566579290456e-06,
+      "loss": 0.406,
+      "step": 426
+    },
+    {
+      "epoch": 0.12746268656716417,
+      "grad_norm": 0.7675194550534261,
+      "learning_rate": 4.876818242299697e-06,
+      "loss": 0.3041,
+      "step": 427
+    },
+    {
+      "epoch": 0.12776119402985076,
+      "grad_norm": 0.974908306841169,
+      "learning_rate": 4.876067683041817e-06,
+      "loss": 0.3846,
+      "step": 428
+    },
+    {
+      "epoch": 0.12805970149253731,
+      "grad_norm": 0.9283777465319998,
+      "learning_rate": 4.875314902218569e-06,
+      "loss": 0.3893,
+      "step": 429
+    },
+    {
+      "epoch": 0.12835820895522387,
+      "grad_norm": 0.8466250701660518,
+      "learning_rate": 4.874559900533786e-06,
+      "loss": 0.3715,
+      "step": 430
+    },
+    {
+      "epoch": 0.12865671641791046,
+      "grad_norm": 0.9436206286256429,
+      "learning_rate": 4.8738026786933765e-06,
+      "loss": 0.3827,
+      "step": 431
+    },
+    {
+      "epoch": 0.12895522388059702,
+      "grad_norm": 0.8642639983887797,
+      "learning_rate": 4.8730432374053245e-06,
+      "loss": 0.4037,
+      "step": 432
+    },
+    {
+      "epoch": 0.12925373134328358,
+      "grad_norm": 0.8620504916338158,
+      "learning_rate": 4.872281577379688e-06,
+      "loss": 0.4308,
+      "step": 433
+    },
+    {
+      "epoch": 0.12955223880597014,
+      "grad_norm": 0.9813776269945593,
+      "learning_rate": 4.8715176993286e-06,
+      "loss": 0.4293,
+      "step": 434
+    },
+    {
+      "epoch": 0.12985074626865672,
+      "grad_norm": 0.8634148297447023,
+      "learning_rate": 4.8707516039662705e-06,
+      "loss": 0.3476,
+      "step": 435
+    },
+    {
+      "epoch": 0.13014925373134328,
+      "grad_norm": 0.8942167636606891,
+      "learning_rate": 4.8699832920089785e-06,
+      "loss": 0.3772,
+      "step": 436
+    },
+    {
+      "epoch": 0.13044776119402984,
+      "grad_norm": 0.8696237331815483,
+      "learning_rate": 4.869212764175076e-06,
+      "loss": 0.384,
+      "step": 437
+    },
+    {
+      "epoch": 0.13074626865671643,
+      "grad_norm": 0.8296034400995033,
+      "learning_rate": 4.8684400211849895e-06,
+      "loss": 0.3862,
+      "step": 438
+    },
+    {
+      "epoch": 0.131044776119403,
+      "grad_norm": 0.9402286904796409,
+      "learning_rate": 4.867665063761212e-06,
+      "loss": 0.3838,
+      "step": 439
+    },
+    {
+      "epoch": 0.13134328358208955,
+      "grad_norm": 0.8227141749687872,
+      "learning_rate": 4.866887892628314e-06,
+      "loss": 0.381,
+      "step": 440
+    },
+    {
+      "epoch": 0.1316417910447761,
+      "grad_norm": 0.8661103190250566,
+      "learning_rate": 4.866108508512929e-06,
+      "loss": 0.395,
+      "step": 441
+    },
+    {
+      "epoch": 0.1319402985074627,
+      "grad_norm": 0.8814800070280554,
+      "learning_rate": 4.865326912143762e-06,
+      "loss": 0.4317,
+      "step": 442
+    },
+    {
+      "epoch": 0.13223880597014925,
+      "grad_norm": 0.803476919813946,
+      "learning_rate": 4.864543104251587e-06,
+      "loss": 0.3822,
+      "step": 443
+    },
+    {
+      "epoch": 0.1325373134328358,
+      "grad_norm": 0.8519115345361534,
+      "learning_rate": 4.863757085569246e-06,
+      "loss": 0.4007,
+      "step": 444
+    },
+    {
+      "epoch": 0.1328358208955224,
+      "grad_norm": 0.8983953481126902,
+      "learning_rate": 4.862968856831646e-06,
+      "loss": 0.388,
+      "step": 445
+    },
+    {
+      "epoch": 0.13313432835820896,
+      "grad_norm": 0.7938477876103528,
+      "learning_rate": 4.862178418775763e-06,
+      "loss": 0.3488,
+      "step": 446
+    },
+    {
+      "epoch": 0.13343283582089552,
+      "grad_norm": 0.8324718218816302,
+      "learning_rate": 4.861385772140636e-06,
+      "loss": 0.3509,
+      "step": 447
+    },
+    {
+      "epoch": 0.13373134328358208,
+      "grad_norm": 0.8673532215135669,
+      "learning_rate": 4.86059091766737e-06,
+      "loss": 0.3997,
+      "step": 448
+    },
+    {
+      "epoch": 0.13402985074626866,
+      "grad_norm": 0.8720599760670272,
+      "learning_rate": 4.859793856099138e-06,
+      "loss": 0.4078,
+      "step": 449
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 0.8972789828786194,
+      "learning_rate": 4.858994588181168e-06,
+      "loss": 0.3935,
+      "step": 450
+    },
+    {
+      "epoch": 0.13462686567164178,
+      "grad_norm": 0.8581455676420515,
+      "learning_rate": 4.85819311466076e-06,
+      "loss": 0.4133,
+      "step": 451
+    },
+    {
+      "epoch": 0.13492537313432837,
+      "grad_norm": 0.8099642650298081,
+      "learning_rate": 4.857389436287271e-06,
+      "loss": 0.3684,
+      "step": 452
+    },
+    {
+      "epoch": 0.13522388059701493,
+      "grad_norm": 0.80580267367791,
+      "learning_rate": 4.85658355381212e-06,
+      "loss": 0.3736,
+      "step": 453
+    },
+    {
+      "epoch": 0.1355223880597015,
+      "grad_norm": 0.9175924370164683,
+      "learning_rate": 4.855775467988788e-06,
+      "loss": 0.3663,
+      "step": 454
+    },
+    {
+      "epoch": 0.13582089552238805,
+      "grad_norm": 0.804865363972486,
+      "learning_rate": 4.854965179572816e-06,
+      "loss": 0.3669,
+      "step": 455
+    },
+    {
+      "epoch": 0.13611940298507463,
+      "grad_norm": 0.8793877846060576,
+      "learning_rate": 4.854152689321803e-06,
+      "loss": 0.3981,
+      "step": 456
+    },
+    {
+      "epoch": 0.1364179104477612,
+      "grad_norm": 0.83704531818067,
+      "learning_rate": 4.853337997995408e-06,
+      "loss": 0.3796,
+      "step": 457
+    },
+    {
+      "epoch": 0.13671641791044775,
+      "grad_norm": 0.8197161694891909,
+      "learning_rate": 4.852521106355348e-06,
+      "loss": 0.395,
+      "step": 458
+    },
+    {
+      "epoch": 0.13701492537313434,
+      "grad_norm": 0.9341284013808222,
+      "learning_rate": 4.851702015165396e-06,
+      "loss": 0.374,
+      "step": 459
+    },
+    {
+      "epoch": 0.1373134328358209,
+      "grad_norm": 0.9043457115872434,
+      "learning_rate": 4.850880725191383e-06,
+      "loss": 0.3687,
+      "step": 460
+    },
+    {
+      "epoch": 0.13761194029850746,
+      "grad_norm": 0.8185493546695799,
+      "learning_rate": 4.850057237201194e-06,
+      "loss": 0.3623,
+      "step": 461
+    },
+    {
+      "epoch": 0.13791044776119402,
+      "grad_norm": 0.8454515033830348,
+      "learning_rate": 4.849231551964771e-06,
+      "loss": 0.345,
+      "step": 462
+    },
+    {
+      "epoch": 0.1382089552238806,
+      "grad_norm": 0.9749776836104918,
+      "learning_rate": 4.848403670254111e-06,
+      "loss": 0.4182,
+      "step": 463
+    },
+    {
+      "epoch": 0.13850746268656716,
+      "grad_norm": 0.8514730507575938,
+      "learning_rate": 4.84757359284326e-06,
+      "loss": 0.3707,
+      "step": 464
+    },
+    {
+      "epoch": 0.13880597014925372,
+      "grad_norm": 0.8959722306249783,
+      "learning_rate": 4.846741320508323e-06,
+      "loss": 0.4077,
+      "step": 465
+    },
+    {
+      "epoch": 0.1391044776119403,
+      "grad_norm": 0.7560664033693268,
+      "learning_rate": 4.8459068540274525e-06,
+      "loss": 0.3622,
+      "step": 466
+    },
+    {
+      "epoch": 0.13940298507462687,
+      "grad_norm": 0.8636823785223218,
+      "learning_rate": 4.845070194180856e-06,
+      "loss": 0.3569,
+      "step": 467
+    },
+    {
+      "epoch": 0.13970149253731343,
+      "grad_norm": 0.8161586910729753,
+      "learning_rate": 4.844231341750787e-06,
+      "loss": 0.3693,
+      "step": 468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8264271190333304,
+      "learning_rate": 4.843390297521556e-06,
+      "loss": 0.3826,
+      "step": 469
+    },
+    {
+      "epoch": 0.14029850746268657,
+      "grad_norm": 0.8329895252389455,
+      "learning_rate": 4.842547062279517e-06,
+      "loss": 0.3835,
+      "step": 470
+    },
+    {
+      "epoch": 0.14059701492537313,
+      "grad_norm": 0.8750985694869183,
+      "learning_rate": 4.841701636813074e-06,
+      "loss": 0.4365,
+      "step": 471
+    },
+    {
+      "epoch": 0.1408955223880597,
+      "grad_norm": 0.923999448212016,
+      "learning_rate": 4.84085402191268e-06,
+      "loss": 0.3355,
+      "step": 472
+    },
+    {
+      "epoch": 0.14119402985074628,
+      "grad_norm": 0.8593855485500195,
+      "learning_rate": 4.840004218370833e-06,
+      "loss": 0.3684,
+      "step": 473
+    },
+    {
+      "epoch": 0.14149253731343284,
+      "grad_norm": 0.9253486202985267,
+      "learning_rate": 4.83915222698208e-06,
+      "loss": 0.4351,
+      "step": 474
+    },
+    {
+      "epoch": 0.1417910447761194,
+      "grad_norm": 0.7704797438692463,
+      "learning_rate": 4.838298048543012e-06,
+      "loss": 0.3311,
+      "step": 475
+    },
+    {
+      "epoch": 0.14208955223880598,
+      "grad_norm": 0.84958326911992,
+      "learning_rate": 4.837441683852264e-06,
+      "loss": 0.3443,
+      "step": 476
+    },
+    {
+      "epoch": 0.14238805970149254,
+      "grad_norm": 0.9214398433351614,
+      "learning_rate": 4.8365831337105185e-06,
+      "loss": 0.3837,
+      "step": 477
+    },
+    {
+      "epoch": 0.1426865671641791,
+      "grad_norm": 0.9020552668429409,
+      "learning_rate": 4.835722398920496e-06,
+      "loss": 0.3726,
+      "step": 478
+    },
+    {
+      "epoch": 0.14298507462686566,
+      "grad_norm": 0.834953050440995,
+      "learning_rate": 4.834859480286963e-06,
+      "loss": 0.3459,
+      "step": 479
+    },
+    {
+      "epoch": 0.14328358208955225,
+      "grad_norm": 0.8287857134100892,
+      "learning_rate": 4.83399437861673e-06,
+      "loss": 0.3889,
+      "step": 480
+    },
+    {
+      "epoch": 0.1435820895522388,
+      "grad_norm": 0.8245206662676321,
+      "learning_rate": 4.833127094718643e-06,
+      "loss": 0.3667,
+      "step": 481
+    },
+    {
+      "epoch": 0.14388059701492537,
+      "grad_norm": 0.832866477275499,
+      "learning_rate": 4.832257629403592e-06,
+      "loss": 0.4004,
+      "step": 482
+    },
+    {
+      "epoch": 0.14417910447761195,
+      "grad_norm": 0.915872105196081,
+      "learning_rate": 4.8313859834845085e-06,
+      "loss": 0.4013,
+      "step": 483
+    },
+    {
+      "epoch": 0.1444776119402985,
+      "grad_norm": 0.9789956349355645,
+      "learning_rate": 4.830512157776357e-06,
+      "loss": 0.4411,
+      "step": 484
+    },
+    {
+      "epoch": 0.14477611940298507,
+      "grad_norm": 0.7690696697320152,
+      "learning_rate": 4.829636153096143e-06,
+      "loss": 0.3071,
+      "step": 485
+    },
+    {
+      "epoch": 0.14507462686567163,
+      "grad_norm": 0.9027976378763707,
+      "learning_rate": 4.828757970262913e-06,
+      "loss": 0.3413,
+      "step": 486
+    },
+    {
+      "epoch": 0.14537313432835822,
+      "grad_norm": 0.8300835301176763,
+      "learning_rate": 4.827877610097743e-06,
+      "loss": 0.3535,
+      "step": 487
+    },
+    {
+      "epoch": 0.14567164179104478,
+      "grad_norm": 0.8562294310178254,
+      "learning_rate": 4.826995073423749e-06,
+      "loss": 0.355,
+      "step": 488
+    },
+    {
+      "epoch": 0.14597014925373133,
+      "grad_norm": 0.8059680443715135,
+      "learning_rate": 4.826110361066084e-06,
+      "loss": 0.3198,
+      "step": 489
+    },
+    {
+      "epoch": 0.14626865671641792,
+      "grad_norm": 0.9642630011245373,
+      "learning_rate": 4.825223473851929e-06,
+      "loss": 0.4476,
+      "step": 490
+    },
+    {
+      "epoch": 0.14656716417910448,
+      "grad_norm": 0.925182517782558,
+      "learning_rate": 4.824334412610504e-06,
+      "loss": 0.3641,
+      "step": 491
+    },
+    {
+      "epoch": 0.14686567164179104,
+      "grad_norm": 0.8931425394201947,
+      "learning_rate": 4.823443178173058e-06,
+      "loss": 0.3963,
+      "step": 492
+    },
+    {
+      "epoch": 0.1471641791044776,
+      "grad_norm": 0.7616008631295983,
+      "learning_rate": 4.822549771372875e-06,
+      "loss": 0.3295,
+      "step": 493
+    },
+    {
+      "epoch": 0.1474626865671642,
+      "grad_norm": 0.9164279418898156,
+      "learning_rate": 4.821654193045268e-06,
+      "loss": 0.3954,
+      "step": 494
+    },
+    {
+      "epoch": 0.14776119402985075,
+      "grad_norm": 0.8242276478542254,
+      "learning_rate": 4.8207564440275816e-06,
+      "loss": 0.358,
+      "step": 495
+    },
+    {
+      "epoch": 0.1480597014925373,
+      "grad_norm": 0.8015482008189594,
+      "learning_rate": 4.819856525159187e-06,
+      "loss": 0.3724,
+      "step": 496
+    },
+    {
+      "epoch": 0.1483582089552239,
+      "grad_norm": 0.8798301252694813,
+      "learning_rate": 4.818954437281489e-06,
+      "loss": 0.3851,
+      "step": 497
+    },
+    {
+      "epoch": 0.14865671641791045,
+      "grad_norm": 0.8789684495124492,
+      "learning_rate": 4.818050181237916e-06,
+      "loss": 0.3659,
+      "step": 498
+    },
+    {
+      "epoch": 0.148955223880597,
+      "grad_norm": 0.7945015241329157,
+      "learning_rate": 4.817143757873927e-06,
+      "loss": 0.3572,
+      "step": 499
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 0.8407928444830756,
+      "learning_rate": 4.8162351680370046e-06,
+      "loss": 0.408,
+      "step": 500
+    },
+    {
+      "epoch": 0.14955223880597016,
+      "grad_norm": 0.8377445302981639,
+      "learning_rate": 4.815324412576659e-06,
+      "loss": 0.3594,
+      "step": 501
+    },
+    {
+      "epoch": 0.14985074626865671,
+      "grad_norm": 0.9360976258979777,
+      "learning_rate": 4.814411492344423e-06,
+      "loss": 0.384,
+      "step": 502
+    },
+    {
+      "epoch": 0.15014925373134327,
+      "grad_norm": 0.9337795927031363,
+      "learning_rate": 4.813496408193855e-06,
+      "loss": 0.407,
+      "step": 503
+    },
+    {
+      "epoch": 0.15044776119402986,
+      "grad_norm": 0.9407994815333237,
+      "learning_rate": 4.812579160980538e-06,
+      "loss": 0.3348,
+      "step": 504
+    },
+    {
+      "epoch": 0.15074626865671642,
+      "grad_norm": 0.9213462894334784,
+      "learning_rate": 4.8116597515620735e-06,
+      "loss": 0.3909,
+      "step": 505
+    },
+    {
+      "epoch": 0.15104477611940298,
+      "grad_norm": 0.8106559781114875,
+      "learning_rate": 4.810738180798089e-06,
+      "loss": 0.3667,
+      "step": 506
+    },
+    {
+      "epoch": 0.15134328358208957,
+      "grad_norm": 0.8651383542529051,
+      "learning_rate": 4.8098144495502295e-06,
+      "loss": 0.3689,
+      "step": 507
+    },
+    {
+      "epoch": 0.15164179104477613,
+      "grad_norm": 1.0153425024352598,
+      "learning_rate": 4.808888558682161e-06,
+      "loss": 0.3986,
+      "step": 508
+    },
+    {
+      "epoch": 0.15194029850746268,
+      "grad_norm": 0.8027998509692323,
+      "learning_rate": 4.80796050905957e-06,
+      "loss": 0.3671,
+      "step": 509
+    },
+    {
+      "epoch": 0.15223880597014924,
+      "grad_norm": 0.9698401353337921,
+      "learning_rate": 4.807030301550159e-06,
+      "loss": 0.3358,
+      "step": 510
+    },
+    {
+      "epoch": 0.15253731343283583,
+      "grad_norm": 0.8470582921938352,
+      "learning_rate": 4.806097937023652e-06,
+      "loss": 0.3692,
+      "step": 511
+    },
+    {
+      "epoch": 0.1528358208955224,
+      "grad_norm": 0.9235947242592816,
+      "learning_rate": 4.8051634163517825e-06,
+      "loss": 0.3458,
+      "step": 512
+    },
+    {
+      "epoch": 0.15313432835820895,
+      "grad_norm": 0.894327518789336,
+      "learning_rate": 4.8042267404083085e-06,
+      "loss": 0.4142,
+      "step": 513
+    },
+    {
+      "epoch": 0.15343283582089554,
+      "grad_norm": 0.9285145110926873,
+      "learning_rate": 4.803287910068997e-06,
+      "loss": 0.3749,
+      "step": 514
+    },
+    {
+      "epoch": 0.1537313432835821,
+      "grad_norm": 0.8506702933212649,
+      "learning_rate": 4.802346926211634e-06,
+      "loss": 0.3469,
+      "step": 515
+    },
+    {
+      "epoch": 0.15402985074626865,
+      "grad_norm": 0.8264106915519583,
+      "learning_rate": 4.8014037897160134e-06,
+      "loss": 0.4164,
+      "step": 516
+    },
+    {
+      "epoch": 0.1543283582089552,
+      "grad_norm": 0.855191956211793,
+      "learning_rate": 4.800458501463946e-06,
+      "loss": 0.3684,
+      "step": 517
+    },
+    {
+      "epoch": 0.1546268656716418,
+      "grad_norm": 0.8553683728972094,
+      "learning_rate": 4.7995110623392545e-06,
+      "loss": 0.3469,
+      "step": 518
+    },
+    {
+      "epoch": 0.15492537313432836,
+      "grad_norm": 0.8190621146939986,
+      "learning_rate": 4.798561473227769e-06,
+      "loss": 0.3669,
+      "step": 519
+    },
+    {
+      "epoch": 0.15522388059701492,
+      "grad_norm": 0.8060862576423745,
+      "learning_rate": 4.797609735017335e-06,
+      "loss": 0.3829,
+      "step": 520
+    },
+    {
+      "epoch": 0.1555223880597015,
+      "grad_norm": 0.8263221507606819,
+      "learning_rate": 4.796655848597803e-06,
+      "loss": 0.3787,
+      "step": 521
+    },
+    {
+      "epoch": 0.15582089552238806,
+      "grad_norm": 0.9818542527439001,
+      "learning_rate": 4.795699814861033e-06,
+      "loss": 0.3943,
+      "step": 522
+    },
+    {
+      "epoch": 0.15611940298507462,
+      "grad_norm": 0.9363390488612918,
+      "learning_rate": 4.7947416347008936e-06,
+      "loss": 0.3488,
+      "step": 523
+    },
+    {
+      "epoch": 0.15641791044776118,
+      "grad_norm": 0.8719501569051828,
+      "learning_rate": 4.793781309013261e-06,
+      "loss": 0.3037,
+      "step": 524
+    },
+    {
+      "epoch": 0.15671641791044777,
+      "grad_norm": 0.8188931174909418,
+      "learning_rate": 4.7928188386960155e-06,
+      "loss": 0.3433,
+      "step": 525
+    },
+    {
+      "epoch": 0.15701492537313433,
+      "grad_norm": 0.8576926049561543,
+      "learning_rate": 4.791854224649042e-06,
+      "loss": 0.4148,
+      "step": 526
+    },
+    {
+      "epoch": 0.1573134328358209,
+      "grad_norm": 0.9514097030300651,
+      "learning_rate": 4.7908874677742335e-06,
+      "loss": 0.4065,
+      "step": 527
+    },
+    {
+      "epoch": 0.15761194029850747,
+      "grad_norm": 0.9034660588508406,
+      "learning_rate": 4.789918568975483e-06,
+      "loss": 0.3681,
+      "step": 528
+    },
+    {
+      "epoch": 0.15791044776119403,
+      "grad_norm": 0.8634816650220751,
+      "learning_rate": 4.788947529158687e-06,
+      "loss": 0.3697,
+      "step": 529
+    },
+    {
+      "epoch": 0.1582089552238806,
+      "grad_norm": 0.9824567340256468,
+      "learning_rate": 4.787974349231745e-06,
+      "loss": 0.3842,
+      "step": 530
+    },
+    {
+      "epoch": 0.15850746268656715,
+      "grad_norm": 0.7947403079672669,
+      "learning_rate": 4.786999030104555e-06,
+      "loss": 0.3299,
+      "step": 531
+    },
+    {
+      "epoch": 0.15880597014925374,
+      "grad_norm": 0.8529781467405901,
+      "learning_rate": 4.786021572689019e-06,
+      "loss": 0.3872,
+      "step": 532
+    },
+    {
+      "epoch": 0.1591044776119403,
+      "grad_norm": 0.9209188488882705,
+      "learning_rate": 4.785041977899033e-06,
+      "loss": 0.3936,
+      "step": 533
+    },
+    {
+      "epoch": 0.15940298507462686,
+      "grad_norm": 0.885355457355636,
+      "learning_rate": 4.784060246650496e-06,
+      "loss": 0.3924,
+      "step": 534
+    },
+    {
+      "epoch": 0.15970149253731344,
+      "grad_norm": 0.9523084549565237,
+      "learning_rate": 4.783076379861304e-06,
+      "loss": 0.4318,
+      "step": 535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8863570574039847,
+      "learning_rate": 4.782090378451345e-06,
+      "loss": 0.4157,
+      "step": 536
+    },
+    {
+      "epoch": 0.16029850746268656,
+      "grad_norm": 0.8117266215970601,
+      "learning_rate": 4.781102243342508e-06,
+      "loss": 0.3463,
+      "step": 537
+    },
+    {
+      "epoch": 0.16059701492537312,
+      "grad_norm": 0.8515951237304888,
+      "learning_rate": 4.780111975458677e-06,
+      "loss": 0.4112,
+      "step": 538
+    },
+    {
+      "epoch": 0.1608955223880597,
+      "grad_norm": 0.9510326848546506,
+      "learning_rate": 4.779119575725726e-06,
+      "loss": 0.4046,
+      "step": 539
+    },
+    {
+      "epoch": 0.16119402985074627,
+      "grad_norm": 0.8449453300506576,
+      "learning_rate": 4.7781250450715245e-06,
+      "loss": 0.356,
+      "step": 540
+    },
+    {
+      "epoch": 0.16149253731343283,
+      "grad_norm": 0.931947295073069,
+      "learning_rate": 4.7771283844259365e-06,
+      "loss": 0.3974,
+      "step": 541
+    },
+    {
+      "epoch": 0.1617910447761194,
+      "grad_norm": 0.874181026365032,
+      "learning_rate": 4.776129594720813e-06,
+      "loss": 0.3629,
+      "step": 542
+    },
+    {
+      "epoch": 0.16208955223880597,
+      "grad_norm": 0.855050176983507,
+      "learning_rate": 4.77512867689e-06,
+      "loss": 0.3853,
+      "step": 543
+    },
+    {
+      "epoch": 0.16238805970149253,
+      "grad_norm": 0.8923651345310537,
+      "learning_rate": 4.77412563186933e-06,
+      "loss": 0.4007,
+      "step": 544
+    },
+    {
+      "epoch": 0.1626865671641791,
+      "grad_norm": 0.9078676220956058,
+      "learning_rate": 4.7731204605966265e-06,
+      "loss": 0.3881,
+      "step": 545
+    },
+    {
+      "epoch": 0.16298507462686568,
+      "grad_norm": 0.7538192348742706,
+      "learning_rate": 4.7721131640116996e-06,
+      "loss": 0.3036,
+      "step": 546
+    },
+    {
+      "epoch": 0.16328358208955224,
+      "grad_norm": 0.8608203405755754,
+      "learning_rate": 4.771103743056348e-06,
+      "loss": 0.3457,
+      "step": 547
+    },
+    {
+      "epoch": 0.1635820895522388,
+      "grad_norm": 0.8212962040784945,
+      "learning_rate": 4.770092198674353e-06,
+      "loss": 0.3273,
+      "step": 548
+    },
+    {
+      "epoch": 0.16388059701492538,
+      "grad_norm": 0.9028550682650915,
+      "learning_rate": 4.769078531811487e-06,
+      "loss": 0.4282,
+      "step": 549
+    },
+    {
+      "epoch": 0.16417910447761194,
+      "grad_norm": 0.8884438521513933,
+      "learning_rate": 4.768062743415502e-06,
+      "loss": 0.3837,
+      "step": 550
+    },
+    {
+      "epoch": 0.1644776119402985,
+      "grad_norm": 0.876505852629387,
+      "learning_rate": 4.767044834436137e-06,
+      "loss": 0.3951,
+      "step": 551
+    },
+    {
+      "epoch": 0.1647761194029851,
+      "grad_norm": 0.7663670395688694,
+      "learning_rate": 4.76602480582511e-06,
+      "loss": 0.3629,
+      "step": 552
+    },
+    {
+      "epoch": 0.16507462686567165,
+      "grad_norm": 0.9554436497467657,
+      "learning_rate": 4.7650026585361255e-06,
+      "loss": 0.3752,
+      "step": 553
+    },
+    {
+      "epoch": 0.1653731343283582,
+      "grad_norm": 0.8328045712770886,
+      "learning_rate": 4.7639783935248635e-06,
+      "loss": 0.363,
+      "step": 554
+    },
+    {
+      "epoch": 0.16567164179104477,
+      "grad_norm": 1.3533426086417046,
+      "learning_rate": 4.762952011748988e-06,
+      "loss": 0.293,
+      "step": 555
+    },
+    {
+      "epoch": 0.16597014925373135,
+      "grad_norm": 0.8397180497457704,
+      "learning_rate": 4.761923514168142e-06,
+      "loss": 0.3803,
+      "step": 556
+    },
+    {
+      "epoch": 0.1662686567164179,
+      "grad_norm": 0.899522921671484,
+      "learning_rate": 4.760892901743944e-06,
+      "loss": 0.4168,
+      "step": 557
+    },
+    {
+      "epoch": 0.16656716417910447,
+      "grad_norm": 0.8900270204673798,
+      "learning_rate": 4.759860175439993e-06,
+      "loss": 0.4041,
+      "step": 558
+    },
+    {
+      "epoch": 0.16686567164179106,
+      "grad_norm": 0.9386190331772859,
+      "learning_rate": 4.758825336221861e-06,
+      "loss": 0.3569,
+      "step": 559
+    },
+    {
+      "epoch": 0.16716417910447762,
+      "grad_norm": 1.4064679834783074,
+      "learning_rate": 4.7577883850570995e-06,
+      "loss": 0.4046,
+      "step": 560
+    },
+    {
+      "epoch": 0.16746268656716418,
+      "grad_norm": 0.8752446720830109,
+      "learning_rate": 4.7567493229152315e-06,
+      "loss": 0.369,
+      "step": 561
+    },
+    {
+      "epoch": 0.16776119402985074,
+      "grad_norm": 0.8921917475055471,
+      "learning_rate": 4.755708150767754e-06,
+      "loss": 0.3629,
+      "step": 562
+    },
+    {
+      "epoch": 0.16805970149253732,
+      "grad_norm": 0.9071169605857111,
+      "learning_rate": 4.754664869588139e-06,
+      "loss": 0.3196,
+      "step": 563
+    },
+    {
+      "epoch": 0.16835820895522388,
+      "grad_norm": 0.7429818510738487,
+      "learning_rate": 4.75361948035183e-06,
+      "loss": 0.322,
+      "step": 564
+    },
+    {
+      "epoch": 0.16865671641791044,
+      "grad_norm": 0.8100900713596559,
+      "learning_rate": 4.752571984036237e-06,
+      "loss": 0.3758,
+      "step": 565
+    },
+    {
+      "epoch": 0.16895522388059703,
+      "grad_norm": 0.8598955983620397,
+      "learning_rate": 4.7515223816207455e-06,
+      "loss": 0.3442,
+      "step": 566
+    },
+    {
+      "epoch": 0.1692537313432836,
+      "grad_norm": 0.8783738770144779,
+      "learning_rate": 4.750470674086709e-06,
+      "loss": 0.322,
+      "step": 567
+    },
+    {
+      "epoch": 0.16955223880597015,
+      "grad_norm": 0.9030983989078412,
+      "learning_rate": 4.749416862417448e-06,
+      "loss": 0.4017,
+      "step": 568
+    },
+    {
+      "epoch": 0.1698507462686567,
+      "grad_norm": 0.8713947394750962,
+      "learning_rate": 4.748360947598248e-06,
+      "loss": 0.3646,
+      "step": 569
+    },
+    {
+      "epoch": 0.1701492537313433,
+      "grad_norm": 0.8986926196540066,
+      "learning_rate": 4.747302930616368e-06,
+      "loss": 0.4171,
+      "step": 570
+    },
+    {
+      "epoch": 0.17044776119402985,
+      "grad_norm": 0.8073489618852552,
+      "learning_rate": 4.746242812461025e-06,
+      "loss": 0.3999,
+      "step": 571
+    },
+    {
+      "epoch": 0.1707462686567164,
+      "grad_norm": 0.9163536146278429,
+      "learning_rate": 4.7451805941234055e-06,
+      "loss": 0.4386,
+      "step": 572
+    },
+    {
+      "epoch": 0.171044776119403,
+      "grad_norm": 0.828059671354957,
+      "learning_rate": 4.744116276596656e-06,
+      "loss": 0.3435,
+      "step": 573
+    },
+    {
+      "epoch": 0.17134328358208956,
+      "grad_norm": 0.8696889215011467,
+      "learning_rate": 4.743049860875889e-06,
+      "loss": 0.3319,
+      "step": 574
+    },
+    {
+      "epoch": 0.17164179104477612,
+      "grad_norm": 0.8480598380604834,
+      "learning_rate": 4.741981347958175e-06,
+      "loss": 0.3703,
+      "step": 575
+    },
+    {
+      "epoch": 0.17194029850746267,
+      "grad_norm": 0.7909046861375709,
+      "learning_rate": 4.740910738842551e-06,
+      "loss": 0.3589,
+      "step": 576
+    },
+    {
+      "epoch": 0.17223880597014926,
+      "grad_norm": 0.8917454585581824,
+      "learning_rate": 4.739838034530008e-06,
+      "loss": 0.3876,
+      "step": 577
+    },
+    {
+      "epoch": 0.17253731343283582,
+      "grad_norm": 0.8783412360640082,
+      "learning_rate": 4.738763236023498e-06,
+      "loss": 0.3892,
+      "step": 578
+    },
+    {
+      "epoch": 0.17283582089552238,
+      "grad_norm": 0.9299982449111176,
+      "learning_rate": 4.737686344327932e-06,
+      "loss": 0.3642,
+      "step": 579
+    },
+    {
+      "epoch": 0.17313432835820897,
+      "grad_norm": 0.8702752107459392,
+      "learning_rate": 4.736607360450179e-06,
+      "loss": 0.357,
+      "step": 580
+    },
+    {
+      "epoch": 0.17343283582089553,
+      "grad_norm": 0.8641201687866182,
+      "learning_rate": 4.73552628539906e-06,
+      "loss": 0.3881,
+      "step": 581
+    },
+    {
+      "epoch": 0.17373134328358208,
+      "grad_norm": 0.7875411772022647,
+      "learning_rate": 4.734443120185357e-06,
+      "loss": 0.3886,
+      "step": 582
+    },
+    {
+      "epoch": 0.17402985074626864,
+      "grad_norm": 0.7639534640961319,
+      "learning_rate": 4.733357865821799e-06,
+      "loss": 0.3635,
+      "step": 583
+    },
+    {
+      "epoch": 0.17432835820895523,
+      "grad_norm": 0.8465197416628776,
+      "learning_rate": 4.7322705233230765e-06,
+      "loss": 0.3753,
+      "step": 584
+    },
+    {
+      "epoch": 0.1746268656716418,
+      "grad_norm": 0.8612932322468133,
+      "learning_rate": 4.731181093705825e-06,
+      "loss": 0.3584,
+      "step": 585
+    },
+    {
+      "epoch": 0.17492537313432835,
+      "grad_norm": 0.811960790500117,
+      "learning_rate": 4.730089577988637e-06,
+      "loss": 0.3794,
+      "step": 586
+    },
+    {
+      "epoch": 0.17522388059701494,
+      "grad_norm": 0.7739231568933665,
+      "learning_rate": 4.728995977192052e-06,
+      "loss": 0.3499,
+      "step": 587
+    },
+    {
+      "epoch": 0.1755223880597015,
+      "grad_norm": 0.8400654346049872,
+      "learning_rate": 4.72790029233856e-06,
+      "loss": 0.3675,
+      "step": 588
+    },
+    {
+      "epoch": 0.17582089552238805,
+      "grad_norm": 0.8830538820611432,
+      "learning_rate": 4.7268025244526e-06,
+      "loss": 0.4198,
+      "step": 589
+    },
+    {
+      "epoch": 0.1761194029850746,
+      "grad_norm": 0.8440850359948983,
+      "learning_rate": 4.725702674560558e-06,
+      "loss": 0.3436,
+      "step": 590
+    },
+    {
+      "epoch": 0.1764179104477612,
+      "grad_norm": 0.923969252299939,
+      "learning_rate": 4.724600743690766e-06,
+      "loss": 0.3726,
+      "step": 591
+    },
+    {
+      "epoch": 0.17671641791044776,
+      "grad_norm": 0.9285113562633005,
+      "learning_rate": 4.723496732873504e-06,
+      "loss": 0.3903,
+      "step": 592
+    },
+    {
+      "epoch": 0.17701492537313432,
+      "grad_norm": 0.7967222216178841,
+      "learning_rate": 4.722390643140995e-06,
+      "loss": 0.3088,
+      "step": 593
+    },
+    {
+      "epoch": 0.1773134328358209,
+      "grad_norm": 0.8941789785640152,
+      "learning_rate": 4.721282475527405e-06,
+      "loss": 0.3684,
+      "step": 594
+    },
+    {
+      "epoch": 0.17761194029850746,
+      "grad_norm": 0.8812907651438229,
+      "learning_rate": 4.720172231068845e-06,
+      "loss": 0.3697,
+      "step": 595
+    },
+    {
+      "epoch": 0.17791044776119402,
+      "grad_norm": 0.8984723387629163,
+      "learning_rate": 4.719059910803364e-06,
+      "loss": 0.3657,
+      "step": 596
+    },
+    {
+      "epoch": 0.1782089552238806,
+      "grad_norm": 0.9108229355000194,
+      "learning_rate": 4.717945515770958e-06,
+      "loss": 0.3609,
+      "step": 597
+    },
+    {
+      "epoch": 0.17850746268656717,
+      "grad_norm": 0.885813043986939,
+      "learning_rate": 4.716829047013555e-06,
+      "loss": 0.4141,
+      "step": 598
+    },
+    {
+      "epoch": 0.17880597014925373,
+      "grad_norm": 0.8127480639060701,
+      "learning_rate": 4.715710505575031e-06,
+      "loss": 0.3529,
+      "step": 599
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 0.8032237883986898,
+      "learning_rate": 4.71458989250119e-06,
+      "loss": 0.3496,
+      "step": 600
+    },
+    {
+      "epoch": 0.17940298507462688,
+      "grad_norm": 0.819201792889001,
+      "learning_rate": 4.713467208839782e-06,
+      "loss": 0.3674,
+      "step": 601
+    },
+    {
+      "epoch": 0.17970149253731343,
+      "grad_norm": 1.1406695776718605,
+      "learning_rate": 4.712342455640486e-06,
+      "loss": 0.3756,
+      "step": 602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8607621290309236,
+      "learning_rate": 4.7112156339549205e-06,
+      "loss": 0.3623,
+      "step": 603
+    },
+    {
+      "epoch": 0.18029850746268658,
+      "grad_norm": 0.8516209936354656,
+      "learning_rate": 4.710086744836635e-06,
+      "loss": 0.3934,
+      "step": 604
+    },
+    {
+      "epoch": 0.18059701492537314,
+      "grad_norm": 1.0789657160791544,
+      "learning_rate": 4.708955789341115e-06,
+      "loss": 0.3333,
+      "step": 605
+    },
+    {
+      "epoch": 0.1808955223880597,
+      "grad_norm": 0.8465101170476235,
+      "learning_rate": 4.707822768525775e-06,
+      "loss": 0.3456,
+      "step": 606
+    },
+    {
+      "epoch": 0.18119402985074626,
+      "grad_norm": 0.8021068603431047,
+      "learning_rate": 4.706687683449961e-06,
+      "loss": 0.3743,
+      "step": 607
+    },
+    {
+      "epoch": 0.18149253731343284,
+      "grad_norm": 0.9396472673097465,
+      "learning_rate": 4.705550535174952e-06,
+      "loss": 0.3594,
+      "step": 608
+    },
+    {
+      "epoch": 0.1817910447761194,
+      "grad_norm": 0.9308558717626013,
+      "learning_rate": 4.704411324763954e-06,
+      "loss": 0.3597,
+      "step": 609
+    },
+    {
+      "epoch": 0.18208955223880596,
+      "grad_norm": 0.8780878715069544,
+      "learning_rate": 4.7032700532820984e-06,
+      "loss": 0.3794,
+      "step": 610
+    },
+    {
+      "epoch": 0.18238805970149255,
+      "grad_norm": 0.8438709125133251,
+      "learning_rate": 4.702126721796448e-06,
+      "loss": 0.3534,
+      "step": 611
+    },
+    {
+      "epoch": 0.1826865671641791,
+      "grad_norm": 0.7918705122194744,
+      "learning_rate": 4.700981331375991e-06,
+      "loss": 0.3512,
+      "step": 612
+    },
+    {
+      "epoch": 0.18298507462686567,
+      "grad_norm": 0.8832849637084809,
+      "learning_rate": 4.699833883091637e-06,
+      "loss": 0.3664,
+      "step": 613
+    },
+    {
+      "epoch": 0.18328358208955223,
+      "grad_norm": 0.7766965849001322,
+      "learning_rate": 4.698684378016223e-06,
+      "loss": 0.3536,
+      "step": 614
+    },
+    {
+      "epoch": 0.18358208955223881,
+      "grad_norm": 0.8793943105130626,
+      "learning_rate": 4.69753281722451e-06,
+      "loss": 0.3859,
+      "step": 615
+    },
+    {
+      "epoch": 0.18388059701492537,
+      "grad_norm": 0.9314513871786796,
+      "learning_rate": 4.696379201793176e-06,
+      "loss": 0.367,
+      "step": 616
+    },
+    {
+      "epoch": 0.18417910447761193,
+      "grad_norm": 0.8801529809718072,
+      "learning_rate": 4.695223532800825e-06,
+      "loss": 0.4067,
+      "step": 617
+    },
+    {
+      "epoch": 0.18447761194029852,
+      "grad_norm": 0.7999007221875536,
+      "learning_rate": 4.694065811327982e-06,
+      "loss": 0.3428,
+      "step": 618
+    },
+    {
+      "epoch": 0.18477611940298508,
+      "grad_norm": 0.8859494993053622,
+      "learning_rate": 4.692906038457084e-06,
+      "loss": 0.3461,
+      "step": 619
+    },
+    {
+      "epoch": 0.18507462686567164,
+      "grad_norm": 0.7920787783745799,
+      "learning_rate": 4.6917442152724925e-06,
+      "loss": 0.3338,
+      "step": 620
+    },
+    {
+      "epoch": 0.1853731343283582,
+      "grad_norm": 0.8343099718824265,
+      "learning_rate": 4.6905803428604835e-06,
+      "loss": 0.386,
+      "step": 621
+    },
+    {
+      "epoch": 0.18567164179104478,
+      "grad_norm": 0.8227479914592881,
+      "learning_rate": 4.6894144223092496e-06,
+      "loss": 0.3514,
+      "step": 622
+    },
+    {
+      "epoch": 0.18597014925373134,
+      "grad_norm": 0.9320390654474898,
+      "learning_rate": 4.6882464547088976e-06,
+      "loss": 0.3922,
+      "step": 623
+    },
+    {
+      "epoch": 0.1862686567164179,
+      "grad_norm": 0.7694880051993461,
+      "learning_rate": 4.6870764411514495e-06,
+      "loss": 0.3571,
+      "step": 624
+    },
+    {
+      "epoch": 0.1865671641791045,
+      "grad_norm": 0.8228896792999334,
+      "learning_rate": 4.685904382730839e-06,
+      "loss": 0.3658,
+      "step": 625
+    },
+    {
+      "epoch": 0.18686567164179105,
+      "grad_norm": 0.8303039994029542,
+      "learning_rate": 4.684730280542912e-06,
+      "loss": 0.3735,
+      "step": 626
+    },
+    {
+      "epoch": 0.1871641791044776,
+      "grad_norm": 0.8427730863324351,
+      "learning_rate": 4.6835541356854255e-06,
+      "loss": 0.3918,
+      "step": 627
+    },
+    {
+      "epoch": 0.18746268656716417,
+      "grad_norm": 0.8394164814398534,
+      "learning_rate": 4.682375949258045e-06,
+      "loss": 0.3075,
+      "step": 628
+    },
+    {
+      "epoch": 0.18776119402985075,
+      "grad_norm": 0.8214529875587838,
+      "learning_rate": 4.681195722362349e-06,
+      "loss": 0.3551,
+      "step": 629
+    },
+    {
+      "epoch": 0.1880597014925373,
+      "grad_norm": 0.810136115359384,
+      "learning_rate": 4.68001345610182e-06,
+      "loss": 0.3641,
+      "step": 630
+    },
+    {
+      "epoch": 0.18835820895522387,
+      "grad_norm": 0.8803568899727253,
+      "learning_rate": 4.678829151581846e-06,
+      "loss": 0.3817,
+      "step": 631
+    },
+    {
+      "epoch": 0.18865671641791046,
+      "grad_norm": 0.7930342754419026,
+      "learning_rate": 4.677642809909725e-06,
+      "loss": 0.4024,
+      "step": 632
+    },
+    {
+      "epoch": 0.18895522388059702,
+      "grad_norm": 0.852158008550677,
+      "learning_rate": 4.6764544321946565e-06,
+      "loss": 0.4057,
+      "step": 633
+    },
+    {
+      "epoch": 0.18925373134328358,
+      "grad_norm": 0.7444755651407072,
+      "learning_rate": 4.675264019547745e-06,
+      "loss": 0.3094,
+      "step": 634
+    },
+    {
+      "epoch": 0.18955223880597014,
+      "grad_norm": 0.7810087717032494,
+      "learning_rate": 4.674071573081998e-06,
+      "loss": 0.3552,
+      "step": 635
+    },
+    {
+      "epoch": 0.18985074626865672,
+      "grad_norm": 0.8457218590063859,
+      "learning_rate": 4.672877093912323e-06,
+      "loss": 0.3554,
+      "step": 636
+    },
+    {
+      "epoch": 0.19014925373134328,
+      "grad_norm": 0.8702779931971437,
+      "learning_rate": 4.671680583155528e-06,
+      "loss": 0.4169,
+      "step": 637
+    },
+    {
+      "epoch": 0.19044776119402984,
+      "grad_norm": 0.9714348750550799,
+      "learning_rate": 4.670482041930324e-06,
+      "loss": 0.348,
+      "step": 638
+    },
+    {
+      "epoch": 0.19074626865671643,
+      "grad_norm": 0.8315101609756071,
+      "learning_rate": 4.6692814713573155e-06,
+      "loss": 0.3387,
+      "step": 639
+    },
+    {
+      "epoch": 0.191044776119403,
+      "grad_norm": 0.8737001772219148,
+      "learning_rate": 4.6680788725590086e-06,
+      "loss": 0.3723,
+      "step": 640
+    },
+    {
+      "epoch": 0.19134328358208955,
+      "grad_norm": 0.928685080600342,
+      "learning_rate": 4.6668742466598015e-06,
+      "loss": 0.3903,
+      "step": 641
+    },
+    {
+      "epoch": 0.19164179104477613,
+      "grad_norm": 0.800239252131357,
+      "learning_rate": 4.665667594785992e-06,
+      "loss": 0.3408,
+      "step": 642
+    },
+    {
+      "epoch": 0.1919402985074627,
+      "grad_norm": 0.8264024494717072,
+      "learning_rate": 4.66445891806577e-06,
+      "loss": 0.3119,
+      "step": 643
+    },
+    {
+      "epoch": 0.19223880597014925,
+      "grad_norm": 0.782897351165225,
+      "learning_rate": 4.663248217629218e-06,
+      "loss": 0.3748,
+      "step": 644
+    },
+    {
+      "epoch": 0.1925373134328358,
+      "grad_norm": 0.7565710294395597,
+      "learning_rate": 4.662035494608313e-06,
+      "loss": 0.3267,
+      "step": 645
+    },
+    {
+      "epoch": 0.1928358208955224,
+      "grad_norm": 0.8522767167672216,
+      "learning_rate": 4.660820750136918e-06,
+      "loss": 0.3771,
+      "step": 646
+    },
+    {
+      "epoch": 0.19313432835820896,
+      "grad_norm": 0.8777158029491957,
+      "learning_rate": 4.6596039853507925e-06,
+      "loss": 0.3605,
+      "step": 647
+    },
+    {
+      "epoch": 0.19343283582089552,
+      "grad_norm": 0.8449256976352308,
+      "learning_rate": 4.658385201387582e-06,
+      "loss": 0.3454,
+      "step": 648
+    },
+    {
+      "epoch": 0.1937313432835821,
+      "grad_norm": 0.9008268972405354,
+      "learning_rate": 4.657164399386818e-06,
+      "loss": 0.4138,
+      "step": 649
+    },
+    {
+      "epoch": 0.19402985074626866,
+      "grad_norm": 0.8903381842895088,
+      "learning_rate": 4.655941580489922e-06,
+      "loss": 0.3699,
+      "step": 650
+    },
+    {
+      "epoch": 0.19432835820895522,
+      "grad_norm": 0.8382545542778225,
+      "learning_rate": 4.6547167458402e-06,
+      "loss": 0.3738,
+      "step": 651
+    },
+    {
+      "epoch": 0.19462686567164178,
+      "grad_norm": 0.9109044659426834,
+      "learning_rate": 4.653489896582841e-06,
+      "loss": 0.4004,
+      "step": 652
+    },
+    {
+      "epoch": 0.19492537313432837,
+      "grad_norm": 0.9177470131519694,
+      "learning_rate": 4.65226103386492e-06,
+      "loss": 0.3654,
+      "step": 653
+    },
+    {
+      "epoch": 0.19522388059701493,
+      "grad_norm": 0.8276383969775275,
+      "learning_rate": 4.651030158835393e-06,
+      "loss": 0.3513,
+      "step": 654
+    },
+    {
+      "epoch": 0.19552238805970149,
+      "grad_norm": 0.8952139983790235,
+      "learning_rate": 4.6497972726451005e-06,
+      "loss": 0.3786,
+      "step": 655
+    },
+    {
+      "epoch": 0.19582089552238807,
+      "grad_norm": 0.7985778698618001,
+      "learning_rate": 4.648562376446759e-06,
+      "loss": 0.3638,
+      "step": 656
+    },
+    {
+      "epoch": 0.19611940298507463,
+      "grad_norm": 0.908801175999597,
+      "learning_rate": 4.6473254713949665e-06,
+      "loss": 0.397,
+      "step": 657
+    },
+    {
+      "epoch": 0.1964179104477612,
+      "grad_norm": 0.894398615813818,
+      "learning_rate": 4.6460865586462e-06,
+      "loss": 0.4207,
+      "step": 658
+    },
+    {
+      "epoch": 0.19671641791044775,
+      "grad_norm": 0.8685502281430703,
+      "learning_rate": 4.644845639358812e-06,
+      "loss": 0.3762,
+      "step": 659
+    },
+    {
+      "epoch": 0.19701492537313434,
+      "grad_norm": 0.8915809779098176,
+      "learning_rate": 4.6436027146930316e-06,
+      "loss": 0.3427,
+      "step": 660
+    },
+    {
+      "epoch": 0.1973134328358209,
+      "grad_norm": 0.8679116859759302,
+      "learning_rate": 4.642357785810964e-06,
+      "loss": 0.3944,
+      "step": 661
+    },
+    {
+      "epoch": 0.19761194029850745,
+      "grad_norm": 0.7880457160052126,
+      "learning_rate": 4.641110853876586e-06,
+      "loss": 0.3987,
+      "step": 662
+    },
+    {
+      "epoch": 0.19791044776119404,
+      "grad_norm": 0.864228831055506,
+      "learning_rate": 4.6398619200557485e-06,
+      "loss": 0.3996,
+      "step": 663
+    },
+    {
+      "epoch": 0.1982089552238806,
+      "grad_norm": 0.8022579546091344,
+      "learning_rate": 4.638610985516176e-06,
+      "loss": 0.3521,
+      "step": 664
+    },
+    {
+      "epoch": 0.19850746268656716,
+      "grad_norm": 0.7106263773090994,
+      "learning_rate": 4.6373580514274605e-06,
+      "loss": 0.3367,
+      "step": 665
+    },
+    {
+      "epoch": 0.19880597014925372,
+      "grad_norm": 0.8985506899451846,
+      "learning_rate": 4.636103118961065e-06,
+      "loss": 0.4056,
+      "step": 666
+    },
+    {
+      "epoch": 0.1991044776119403,
+      "grad_norm": 0.870760200813925,
+      "learning_rate": 4.634846189290321e-06,
+      "loss": 0.3646,
+      "step": 667
+    },
+    {
+      "epoch": 0.19940298507462687,
+      "grad_norm": 0.8551014965491996,
+      "learning_rate": 4.633587263590427e-06,
+      "loss": 0.4102,
+      "step": 668
+    },
+    {
+      "epoch": 0.19970149253731342,
+      "grad_norm": 0.8857508158230196,
+      "learning_rate": 4.632326343038448e-06,
+      "loss": 0.4184,
+      "step": 669
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7779965326593492,
+      "learning_rate": 4.631063428813314e-06,
+      "loss": 0.3521,
+      "step": 670
+    },
+    {
+      "epoch": 0.20029850746268657,
+      "grad_norm": 0.8349623707590051,
+      "learning_rate": 4.629798522095818e-06,
+      "loss": 0.381,
+      "step": 671
+    },
+    {
+      "epoch": 0.20059701492537313,
+      "grad_norm": 1.0856988327864074,
+      "learning_rate": 4.628531624068618e-06,
+      "loss": 0.3597,
+      "step": 672
+    },
+    {
+      "epoch": 0.2008955223880597,
+      "grad_norm": 0.855642649913472,
+      "learning_rate": 4.627262735916233e-06,
+      "loss": 0.3506,
+      "step": 673
+    },
+    {
+      "epoch": 0.20119402985074628,
+      "grad_norm": 0.7945501228954227,
+      "learning_rate": 4.625991858825042e-06,
+      "loss": 0.3186,
+      "step": 674
+    },
+    {
+      "epoch": 0.20149253731343283,
+      "grad_norm": 0.9656198406567287,
+      "learning_rate": 4.624718993983284e-06,
+      "loss": 0.349,
+      "step": 675
+    },
+    {
+      "epoch": 0.2017910447761194,
+      "grad_norm": 0.9620523693551688,
+      "learning_rate": 4.623444142581056e-06,
+      "loss": 0.3566,
+      "step": 676
+    },
+    {
+      "epoch": 0.20208955223880598,
+      "grad_norm": 0.7443588122091737,
+      "learning_rate": 4.622167305810315e-06,
+      "loss": 0.3195,
+      "step": 677
+    },
+    {
+      "epoch": 0.20238805970149254,
+      "grad_norm": 0.9328190975144249,
+      "learning_rate": 4.620888484864869e-06,
+      "loss": 0.3975,
+      "step": 678
+    },
+    {
+      "epoch": 0.2026865671641791,
+      "grad_norm": 0.8006835988448852,
+      "learning_rate": 4.6196076809403875e-06,
+      "loss": 0.3228,
+      "step": 679
+    },
+    {
+      "epoch": 0.20298507462686566,
+      "grad_norm": 0.9990492755928724,
+      "learning_rate": 4.618324895234391e-06,
+      "loss": 0.3728,
+      "step": 680
+    },
+    {
+      "epoch": 0.20328358208955224,
+      "grad_norm": 0.817711684027937,
+      "learning_rate": 4.61704012894625e-06,
+      "loss": 0.3876,
+      "step": 681
+    },
+    {
+      "epoch": 0.2035820895522388,
+      "grad_norm": 0.7456982963035693,
+      "learning_rate": 4.615753383277192e-06,
+      "loss": 0.334,
+      "step": 682
+    },
+    {
+      "epoch": 0.20388059701492536,
+      "grad_norm": 0.87509200555024,
+      "learning_rate": 4.614464659430292e-06,
+      "loss": 0.3982,
+      "step": 683
+    },
+    {
+      "epoch": 0.20417910447761195,
+      "grad_norm": 0.7734942878648073,
+      "learning_rate": 4.613173958610476e-06,
+      "loss": 0.3653,
+      "step": 684
+    },
+    {
+      "epoch": 0.2044776119402985,
+      "grad_norm": 1.006253095908581,
+      "learning_rate": 4.611881282024518e-06,
+      "loss": 0.4069,
+      "step": 685
+    },
+    {
+      "epoch": 0.20477611940298507,
+      "grad_norm": 0.9446767733133795,
+      "learning_rate": 4.6105866308810375e-06,
+      "loss": 0.4212,
+      "step": 686
+    },
+    {
+      "epoch": 0.20507462686567166,
+      "grad_norm": 0.9531967381845363,
+      "learning_rate": 4.609290006390503e-06,
+      "loss": 0.3525,
+      "step": 687
+    },
+    {
+      "epoch": 0.20537313432835821,
+      "grad_norm": 0.7954289882558423,
+      "learning_rate": 4.607991409765226e-06,
+      "loss": 0.3446,
+      "step": 688
+    },
+    {
+      "epoch": 0.20567164179104477,
+      "grad_norm": 0.7229225187581373,
+      "learning_rate": 4.606690842219364e-06,
+      "loss": 0.3562,
+      "step": 689
+    },
+    {
+      "epoch": 0.20597014925373133,
+      "grad_norm": 0.7579897211017783,
+      "learning_rate": 4.605388304968915e-06,
+      "loss": 0.3447,
+      "step": 690
+    },
+    {
+      "epoch": 0.20626865671641792,
+      "grad_norm": 0.9384371378838448,
+      "learning_rate": 4.604083799231719e-06,
+      "loss": 0.365,
+      "step": 691
+    },
+    {
+      "epoch": 0.20656716417910448,
+      "grad_norm": 1.1276096522058603,
+      "learning_rate": 4.602777326227459e-06,
+      "loss": 0.3468,
+      "step": 692
+    },
+    {
+      "epoch": 0.20686567164179104,
+      "grad_norm": 0.8690437426856318,
+      "learning_rate": 4.6014688871776535e-06,
+      "loss": 0.4067,
+      "step": 693
+    },
+    {
+      "epoch": 0.20716417910447762,
+      "grad_norm": 0.8340116401411932,
+      "learning_rate": 4.600158483305662e-06,
+      "loss": 0.3303,
+      "step": 694
+    },
+    {
+      "epoch": 0.20746268656716418,
+      "grad_norm": 0.9659161023612816,
+      "learning_rate": 4.59884611583668e-06,
+      "loss": 0.3845,
+      "step": 695
+    },
+    {
+      "epoch": 0.20776119402985074,
+      "grad_norm": 0.8013822632682776,
+      "learning_rate": 4.59753178599774e-06,
+      "loss": 0.343,
+      "step": 696
+    },
+    {
+      "epoch": 0.2080597014925373,
+      "grad_norm": 0.7638697285164096,
+      "learning_rate": 4.5962154950177065e-06,
+      "loss": 0.3371,
+      "step": 697
+    },
+    {
+      "epoch": 0.2083582089552239,
+      "grad_norm": 0.8584456429925702,
+      "learning_rate": 4.594897244127281e-06,
+      "loss": 0.387,
+      "step": 698
+    },
+    {
+      "epoch": 0.20865671641791045,
+      "grad_norm": 0.7652924682991881,
+      "learning_rate": 4.593577034558995e-06,
+      "loss": 0.3551,
+      "step": 699
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 0.8065003946035596,
+      "learning_rate": 4.592254867547214e-06,
+      "loss": 0.3438,
+      "step": 700
+    },
+    {
+      "epoch": 0.2092537313432836,
+      "grad_norm": 0.9004428506328457,
+      "learning_rate": 4.590930744328128e-06,
+      "loss": 0.3713,
+      "step": 701
+    },
+    {
+      "epoch": 0.20955223880597015,
+      "grad_norm": 0.8938815871012901,
+      "learning_rate": 4.5896046661397645e-06,
+      "loss": 0.3692,
+      "step": 702
+    },
+    {
+      "epoch": 0.2098507462686567,
+      "grad_norm": 0.905050953213985,
+      "learning_rate": 4.588276634221972e-06,
+      "loss": 0.3596,
+      "step": 703
+    },
+    {
+      "epoch": 0.21014925373134327,
+      "grad_norm": 0.7970736616034015,
+      "learning_rate": 4.586946649816428e-06,
+      "loss": 0.3495,
+      "step": 704
+    },
+    {
+      "epoch": 0.21044776119402986,
+      "grad_norm": 0.8579016457008204,
+      "learning_rate": 4.585614714166636e-06,
+      "loss": 0.3707,
+      "step": 705
+    },
+    {
+      "epoch": 0.21074626865671642,
+      "grad_norm": 0.9030100226974851,
+      "learning_rate": 4.584280828517924e-06,
+      "loss": 0.3828,
+      "step": 706
+    },
+    {
+      "epoch": 0.21104477611940298,
+      "grad_norm": 0.9247619159595248,
+      "learning_rate": 4.582944994117441e-06,
+      "loss": 0.3792,
+      "step": 707
+    },
+    {
+      "epoch": 0.21134328358208956,
+      "grad_norm": 0.9295226931550548,
+      "learning_rate": 4.58160721221416e-06,
+      "loss": 0.3958,
+      "step": 708
+    },
+    {
+      "epoch": 0.21164179104477612,
+      "grad_norm": 0.894595136771482,
+      "learning_rate": 4.580267484058876e-06,
+      "loss": 0.3559,
+      "step": 709
+    },
+    {
+      "epoch": 0.21194029850746268,
+      "grad_norm": 0.8774937178986986,
+      "learning_rate": 4.5789258109042e-06,
+      "loss": 0.4055,
+      "step": 710
+    },
+    {
+      "epoch": 0.21223880597014924,
+      "grad_norm": 0.8187268569344242,
+      "learning_rate": 4.577582194004565e-06,
+      "loss": 0.3495,
+      "step": 711
+    },
+    {
+      "epoch": 0.21253731343283583,
+      "grad_norm": 0.8813687056400052,
+      "learning_rate": 4.576236634616219e-06,
+      "loss": 0.3874,
+      "step": 712
+    },
+    {
+      "epoch": 0.2128358208955224,
+      "grad_norm": 0.7817355048189055,
+      "learning_rate": 4.574889133997229e-06,
+      "loss": 0.348,
+      "step": 713
+    },
+    {
+      "epoch": 0.21313432835820895,
+      "grad_norm": 0.7918141903283299,
+      "learning_rate": 4.573539693407474e-06,
+      "loss": 0.3711,
+      "step": 714
+    },
+    {
+      "epoch": 0.21343283582089553,
+      "grad_norm": 0.7682344039631848,
+      "learning_rate": 4.572188314108648e-06,
+      "loss": 0.3661,
+      "step": 715
+    },
+    {
+      "epoch": 0.2137313432835821,
+      "grad_norm": 0.7620495460154416,
+      "learning_rate": 4.570834997364258e-06,
+      "loss": 0.3651,
+      "step": 716
+    },
+    {
+      "epoch": 0.21402985074626865,
+      "grad_norm": 0.9253366569905003,
+      "learning_rate": 4.569479744439622e-06,
+      "loss": 0.3356,
+      "step": 717
+    },
+    {
+      "epoch": 0.2143283582089552,
+      "grad_norm": 0.9151048827100409,
+      "learning_rate": 4.568122556601869e-06,
+      "loss": 0.363,
+      "step": 718
+    },
+    {
+      "epoch": 0.2146268656716418,
+      "grad_norm": 0.8490011151647137,
+      "learning_rate": 4.566763435119936e-06,
+      "loss": 0.3247,
+      "step": 719
+    },
+    {
+      "epoch": 0.21492537313432836,
+      "grad_norm": 0.8786239642622696,
+      "learning_rate": 4.565402381264569e-06,
+      "loss": 0.4117,
+      "step": 720
+    },
+    {
+      "epoch": 0.21522388059701492,
+      "grad_norm": 0.8402095797172084,
+      "learning_rate": 4.564039396308319e-06,
+      "loss": 0.3549,
+      "step": 721
+    },
+    {
+      "epoch": 0.2155223880597015,
+      "grad_norm": 0.8965903934098873,
+      "learning_rate": 4.5626744815255454e-06,
+      "loss": 0.382,
+      "step": 722
+    },
+    {
+      "epoch": 0.21582089552238806,
+      "grad_norm": 0.8307615244096349,
+      "learning_rate": 4.5613076381924084e-06,
+      "loss": 0.3527,
+      "step": 723
+    },
+    {
+      "epoch": 0.21611940298507462,
+      "grad_norm": 0.8695011048660274,
+      "learning_rate": 4.559938867586874e-06,
+      "loss": 0.3689,
+      "step": 724
+    },
+    {
+      "epoch": 0.21641791044776118,
+      "grad_norm": 0.9306482302064063,
+      "learning_rate": 4.5585681709887104e-06,
+      "loss": 0.391,
+      "step": 725
+    },
+    {
+      "epoch": 0.21671641791044777,
+      "grad_norm": 0.828492781007522,
+      "learning_rate": 4.557195549679484e-06,
+      "loss": 0.3234,
+      "step": 726
+    },
+    {
+      "epoch": 0.21701492537313433,
+      "grad_norm": 0.8573294375941618,
+      "learning_rate": 4.555821004942563e-06,
+      "loss": 0.386,
+      "step": 727
+    },
+    {
+      "epoch": 0.21731343283582089,
+      "grad_norm": 0.875314700850979,
+      "learning_rate": 4.554444538063113e-06,
+      "loss": 0.3569,
+      "step": 728
+    },
+    {
+      "epoch": 0.21761194029850747,
+      "grad_norm": 0.825263881051261,
+      "learning_rate": 4.553066150328097e-06,
+      "loss": 0.3716,
+      "step": 729
+    },
+    {
+      "epoch": 0.21791044776119403,
+      "grad_norm": 0.9101676571634926,
+      "learning_rate": 4.5516858430262745e-06,
+      "loss": 0.3819,
+      "step": 730
+    },
+    {
+      "epoch": 0.2182089552238806,
+      "grad_norm": 0.7788518704841725,
+      "learning_rate": 4.550303617448198e-06,
+      "loss": 0.34,
+      "step": 731
+    },
+    {
+      "epoch": 0.21850746268656718,
+      "grad_norm": 0.7994914875640055,
+      "learning_rate": 4.548919474886217e-06,
+      "loss": 0.3317,
+      "step": 732
+    },
+    {
+      "epoch": 0.21880597014925374,
+      "grad_norm": 0.8615980705638517,
+      "learning_rate": 4.547533416634468e-06,
+      "loss": 0.3488,
+      "step": 733
+    },
+    {
+      "epoch": 0.2191044776119403,
+      "grad_norm": 0.8580797802618181,
+      "learning_rate": 4.546145443988883e-06,
+      "loss": 0.3386,
+      "step": 734
+    },
+    {
+      "epoch": 0.21940298507462686,
+      "grad_norm": 0.8023378839095241,
+      "learning_rate": 4.544755558247184e-06,
+      "loss": 0.3822,
+      "step": 735
+    },
+    {
+      "epoch": 0.21970149253731344,
+      "grad_norm": 0.8955427517527177,
+      "learning_rate": 4.543363760708878e-06,
+      "loss": 0.3656,
+      "step": 736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8836190670530261,
+      "learning_rate": 4.541970052675262e-06,
+      "loss": 0.3929,
+      "step": 737
+    },
+    {
+      "epoch": 0.22029850746268656,
+      "grad_norm": 0.9099664733412756,
+      "learning_rate": 4.540574435449421e-06,
+      "loss": 0.3515,
+      "step": 738
+    },
+    {
+      "epoch": 0.22059701492537315,
+      "grad_norm": 0.9197478364346269,
+      "learning_rate": 4.539176910336221e-06,
+      "loss": 0.4125,
+      "step": 739
+    },
+    {
+      "epoch": 0.2208955223880597,
+      "grad_norm": 0.8096717598195738,
+      "learning_rate": 4.537777478642317e-06,
+      "loss": 0.3441,
+      "step": 740
+    },
+    {
+      "epoch": 0.22119402985074627,
+      "grad_norm": 0.7424662504675141,
+      "learning_rate": 4.53637614167614e-06,
+      "loss": 0.3466,
+      "step": 741
+    },
+    {
+      "epoch": 0.22149253731343282,
+      "grad_norm": 0.8596564765091457,
+      "learning_rate": 4.534972900747907e-06,
+      "loss": 0.3607,
+      "step": 742
+    },
+    {
+      "epoch": 0.2217910447761194,
+      "grad_norm": 0.8235466276329729,
+      "learning_rate": 4.533567757169615e-06,
+      "loss": 0.3597,
+      "step": 743
+    },
+    {
+      "epoch": 0.22208955223880597,
+      "grad_norm": 1.0365662627082053,
+      "learning_rate": 4.532160712255037e-06,
+      "loss": 0.3896,
+      "step": 744
+    },
+    {
+      "epoch": 0.22238805970149253,
+      "grad_norm": 1.0076481822719652,
+      "learning_rate": 4.530751767319729e-06,
+      "loss": 0.3913,
+      "step": 745
+    },
+    {
+      "epoch": 0.22268656716417912,
+      "grad_norm": 0.8582474689745617,
+      "learning_rate": 4.529340923681016e-06,
+      "loss": 0.3245,
+      "step": 746
+    },
+    {
+      "epoch": 0.22298507462686568,
+      "grad_norm": 0.8822511043553072,
+      "learning_rate": 4.527928182658006e-06,
+      "loss": 0.3766,
+      "step": 747
+    },
+    {
+      "epoch": 0.22328358208955223,
+      "grad_norm": 0.8534191167038851,
+      "learning_rate": 4.526513545571576e-06,
+      "loss": 0.3384,
+      "step": 748
+    },
+    {
+      "epoch": 0.2235820895522388,
+      "grad_norm": 0.8670338466114618,
+      "learning_rate": 4.525097013744377e-06,
+      "loss": 0.3828,
+      "step": 749
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.794997902918133,
+      "learning_rate": 4.523678588500831e-06,
+      "loss": 0.3439,
+      "step": 750
+    },
+    {
+      "epoch": 0.22417910447761194,
+      "grad_norm": 0.8637367121235244,
+      "learning_rate": 4.522258271167134e-06,
+      "loss": 0.3854,
+      "step": 751
+    },
+    {
+      "epoch": 0.2244776119402985,
+      "grad_norm": 0.8685702777896596,
+      "learning_rate": 4.520836063071245e-06,
+      "loss": 0.3887,
+      "step": 752
+    },
+    {
+      "epoch": 0.2247761194029851,
+      "grad_norm": 0.772264036444534,
+      "learning_rate": 4.519411965542895e-06,
+      "loss": 0.3192,
+      "step": 753
+    },
+    {
+      "epoch": 0.22507462686567165,
+      "grad_norm": 0.789258733130727,
+      "learning_rate": 4.517985979913581e-06,
+      "loss": 0.3527,
+      "step": 754
+    },
+    {
+      "epoch": 0.2253731343283582,
+      "grad_norm": 0.7727803198307888,
+      "learning_rate": 4.516558107516563e-06,
+      "loss": 0.3764,
+      "step": 755
+    },
+    {
+      "epoch": 0.22567164179104476,
+      "grad_norm": 0.8985233297487031,
+      "learning_rate": 4.51512834968687e-06,
+      "loss": 0.3319,
+      "step": 756
+    },
+    {
+      "epoch": 0.22597014925373135,
+      "grad_norm": 0.7858913969509037,
+      "learning_rate": 4.513696707761287e-06,
+      "loss": 0.3422,
+      "step": 757
+    },
+    {
+      "epoch": 0.2262686567164179,
+      "grad_norm": 0.9507634234561627,
+      "learning_rate": 4.512263183078367e-06,
+      "loss": 0.3781,
+      "step": 758
+    },
+    {
+      "epoch": 0.22656716417910447,
+      "grad_norm": 0.8464279365702985,
+      "learning_rate": 4.510827776978419e-06,
+      "loss": 0.3927,
+      "step": 759
+    },
+    {
+      "epoch": 0.22686567164179106,
+      "grad_norm": 0.8120744514290998,
+      "learning_rate": 4.5093904908035145e-06,
+      "loss": 0.3431,
+      "step": 760
+    },
+    {
+      "epoch": 0.22716417910447761,
+      "grad_norm": 0.7500430603126108,
+      "learning_rate": 4.50795132589748e-06,
+      "loss": 0.2971,
+      "step": 761
+    },
+    {
+      "epoch": 0.22746268656716417,
+      "grad_norm": 1.0783281815992556,
+      "learning_rate": 4.5065102836059e-06,
+      "loss": 0.377,
+      "step": 762
+    },
+    {
+      "epoch": 0.22776119402985073,
+      "grad_norm": 0.8318583999356356,
+      "learning_rate": 4.505067365276112e-06,
+      "loss": 0.3356,
+      "step": 763
+    },
+    {
+      "epoch": 0.22805970149253732,
+      "grad_norm": 0.8126656695454364,
+      "learning_rate": 4.503622572257212e-06,
+      "loss": 0.348,
+      "step": 764
+    },
+    {
+      "epoch": 0.22835820895522388,
+      "grad_norm": 0.8301469658419047,
+      "learning_rate": 4.502175905900046e-06,
+      "loss": 0.3758,
+      "step": 765
+    },
+    {
+      "epoch": 0.22865671641791044,
+      "grad_norm": 0.8038946408180568,
+      "learning_rate": 4.50072736755721e-06,
+      "loss": 0.3598,
+      "step": 766
+    },
+    {
+      "epoch": 0.22895522388059703,
+      "grad_norm": 0.8541693014723826,
+      "learning_rate": 4.499276958583054e-06,
+      "loss": 0.3574,
+      "step": 767
+    },
+    {
+      "epoch": 0.22925373134328358,
+      "grad_norm": 0.8246286524267744,
+      "learning_rate": 4.497824680333674e-06,
+      "loss": 0.3733,
+      "step": 768
+    },
+    {
+      "epoch": 0.22955223880597014,
+      "grad_norm": 0.7782825828612279,
+      "learning_rate": 4.496370534166915e-06,
+      "loss": 0.3105,
+      "step": 769
+    },
+    {
+      "epoch": 0.2298507462686567,
+      "grad_norm": 0.8857118941633463,
+      "learning_rate": 4.494914521442368e-06,
+      "loss": 0.373,
+      "step": 770
+    },
+    {
+      "epoch": 0.2301492537313433,
+      "grad_norm": 0.8395319587604902,
+      "learning_rate": 4.49345664352137e-06,
+      "loss": 0.3468,
+      "step": 771
+    },
+    {
+      "epoch": 0.23044776119402985,
+      "grad_norm": 0.8752411500887408,
+      "learning_rate": 4.491996901766999e-06,
+      "loss": 0.3696,
+      "step": 772
+    },
+    {
+      "epoch": 0.2307462686567164,
+      "grad_norm": 0.9297321969188959,
+      "learning_rate": 4.4905352975440815e-06,
+      "loss": 0.382,
+      "step": 773
+    },
+    {
+      "epoch": 0.231044776119403,
+      "grad_norm": 0.8328696745976142,
+      "learning_rate": 4.489071832219181e-06,
+      "loss": 0.3715,
+      "step": 774
+    },
+    {
+      "epoch": 0.23134328358208955,
+      "grad_norm": 0.8741052478618722,
+      "learning_rate": 4.487606507160599e-06,
+      "loss": 0.3794,
+      "step": 775
+    },
+    {
+      "epoch": 0.2316417910447761,
+      "grad_norm": 0.8662088883531012,
+      "learning_rate": 4.486139323738382e-06,
+      "loss": 0.3659,
+      "step": 776
+    },
+    {
+      "epoch": 0.2319402985074627,
+      "grad_norm": 0.8934574672183584,
+      "learning_rate": 4.48467028332431e-06,
+      "loss": 0.3658,
+      "step": 777
+    },
+    {
+      "epoch": 0.23223880597014926,
+      "grad_norm": 1.1016352512450371,
+      "learning_rate": 4.483199387291898e-06,
+      "loss": 0.3661,
+      "step": 778
+    },
+    {
+      "epoch": 0.23253731343283582,
+      "grad_norm": 0.7828812345663573,
+      "learning_rate": 4.4817266370164e-06,
+      "loss": 0.3567,
+      "step": 779
+    },
+    {
+      "epoch": 0.23283582089552238,
+      "grad_norm": 0.8193782009262484,
+      "learning_rate": 4.480252033874801e-06,
+      "loss": 0.3539,
+      "step": 780
+    },
+    {
+      "epoch": 0.23313432835820896,
+      "grad_norm": 0.8073010853910519,
+      "learning_rate": 4.47877557924582e-06,
+      "loss": 0.3324,
+      "step": 781
+    },
+    {
+      "epoch": 0.23343283582089552,
+      "grad_norm": 0.9202437087315363,
+      "learning_rate": 4.477297274509904e-06,
+      "loss": 0.3666,
+      "step": 782
+    },
+    {
+      "epoch": 0.23373134328358208,
+      "grad_norm": 0.8160201474275349,
+      "learning_rate": 4.475817121049234e-06,
+      "loss": 0.3718,
+      "step": 783
+    },
+    {
+      "epoch": 0.23402985074626867,
+      "grad_norm": 0.8630655153632912,
+      "learning_rate": 4.474335120247716e-06,
+      "loss": 0.4114,
+      "step": 784
+    },
+    {
+      "epoch": 0.23432835820895523,
+      "grad_norm": 0.8729151161750691,
+      "learning_rate": 4.472851273490985e-06,
+      "loss": 0.3725,
+      "step": 785
+    },
+    {
+      "epoch": 0.2346268656716418,
+      "grad_norm": 0.8813347017833638,
+      "learning_rate": 4.471365582166401e-06,
+      "loss": 0.3759,
+      "step": 786
+    },
+    {
+      "epoch": 0.23492537313432835,
+      "grad_norm": 0.8663722741697655,
+      "learning_rate": 4.46987804766305e-06,
+      "loss": 0.3594,
+      "step": 787
+    },
+    {
+      "epoch": 0.23522388059701493,
+      "grad_norm": 0.7836171249800367,
+      "learning_rate": 4.46838867137174e-06,
+      "loss": 0.3079,
+      "step": 788
+    },
+    {
+      "epoch": 0.2355223880597015,
+      "grad_norm": 0.8419069340655513,
+      "learning_rate": 4.466897454685003e-06,
+      "loss": 0.3952,
+      "step": 789
+    },
+    {
+      "epoch": 0.23582089552238805,
+      "grad_norm": 0.844324374913284,
+      "learning_rate": 4.465404398997089e-06,
+      "loss": 0.3685,
+      "step": 790
+    },
+    {
+      "epoch": 0.23611940298507464,
+      "grad_norm": 0.8830236417572802,
+      "learning_rate": 4.463909505703968e-06,
+      "loss": 0.3616,
+      "step": 791
+    },
+    {
+      "epoch": 0.2364179104477612,
+      "grad_norm": 0.8444454534060524,
+      "learning_rate": 4.46241277620333e-06,
+      "loss": 0.3783,
+      "step": 792
+    },
+    {
+      "epoch": 0.23671641791044776,
+      "grad_norm": 0.7655987801662917,
+      "learning_rate": 4.460914211894579e-06,
+      "loss": 0.3183,
+      "step": 793
+    },
+    {
+      "epoch": 0.23701492537313432,
+      "grad_norm": 0.7941247862341938,
+      "learning_rate": 4.459413814178839e-06,
+      "loss": 0.366,
+      "step": 794
+    },
+    {
+      "epoch": 0.2373134328358209,
+      "grad_norm": 0.8222962634858313,
+      "learning_rate": 4.4579115844589426e-06,
+      "loss": 0.3349,
+      "step": 795
+    },
+    {
+      "epoch": 0.23761194029850746,
+      "grad_norm": 0.8913670067022349,
+      "learning_rate": 4.4564075241394386e-06,
+      "loss": 0.3692,
+      "step": 796
+    },
+    {
+      "epoch": 0.23791044776119402,
+      "grad_norm": 0.8571392122510596,
+      "learning_rate": 4.454901634626587e-06,
+      "loss": 0.3689,
+      "step": 797
+    },
+    {
+      "epoch": 0.2382089552238806,
+      "grad_norm": 0.8061419047479605,
+      "learning_rate": 4.4533939173283585e-06,
+      "loss": 0.3451,
+      "step": 798
+    },
+    {
+      "epoch": 0.23850746268656717,
+      "grad_norm": 0.83880369164832,
+      "learning_rate": 4.451884373654431e-06,
+      "loss": 0.3846,
+      "step": 799
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 0.7991787716936328,
+      "learning_rate": 4.450373005016191e-06,
+      "loss": 0.3514,
+      "step": 800
+    },
+    {
+      "epoch": 0.23910447761194029,
+      "grad_norm": 0.9096163448081129,
+      "learning_rate": 4.448859812826732e-06,
+      "loss": 0.3704,
+      "step": 801
+    },
+    {
+      "epoch": 0.23940298507462687,
+      "grad_norm": 0.8414143701711667,
+      "learning_rate": 4.447344798500852e-06,
+      "loss": 0.3674,
+      "step": 802
+    },
+    {
+      "epoch": 0.23970149253731343,
+      "grad_norm": 0.8352881489659175,
+      "learning_rate": 4.445827963455051e-06,
+      "loss": 0.3768,
+      "step": 803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8031794777452957,
+      "learning_rate": 4.444309309107535e-06,
+      "loss": 0.3423,
+      "step": 804
+    },
+    {
+      "epoch": 0.24029850746268658,
+      "grad_norm": 0.8060785944842502,
+      "learning_rate": 4.442788836878207e-06,
+      "loss": 0.3131,
+      "step": 805
+    },
+    {
+      "epoch": 0.24059701492537314,
+      "grad_norm": 0.8202351464416616,
+      "learning_rate": 4.441266548188673e-06,
+      "loss": 0.3566,
+      "step": 806
+    },
+    {
+      "epoch": 0.2408955223880597,
+      "grad_norm": 0.905814083393673,
+      "learning_rate": 4.439742444462234e-06,
+      "loss": 0.3812,
+      "step": 807
+    },
+    {
+      "epoch": 0.24119402985074626,
+      "grad_norm": 0.7839467463765396,
+      "learning_rate": 4.43821652712389e-06,
+      "loss": 0.3544,
+      "step": 808
+    },
+    {
+      "epoch": 0.24149253731343284,
+      "grad_norm": 0.9113819484994593,
+      "learning_rate": 4.436688797600338e-06,
+      "loss": 0.4163,
+      "step": 809
+    },
+    {
+      "epoch": 0.2417910447761194,
+      "grad_norm": 0.8245251973742547,
+      "learning_rate": 4.435159257319968e-06,
+      "loss": 0.349,
+      "step": 810
+    },
+    {
+      "epoch": 0.24208955223880596,
+      "grad_norm": 0.8775639237639656,
+      "learning_rate": 4.43362790771286e-06,
+      "loss": 0.4072,
+      "step": 811
+    },
+    {
+      "epoch": 0.24238805970149255,
+      "grad_norm": 0.877280517556995,
+      "learning_rate": 4.432094750210791e-06,
+      "loss": 0.3773,
+      "step": 812
+    },
+    {
+      "epoch": 0.2426865671641791,
+      "grad_norm": 0.8453157272615807,
+      "learning_rate": 4.430559786247227e-06,
+      "loss": 0.4092,
+      "step": 813
+    },
+    {
+      "epoch": 0.24298507462686567,
+      "grad_norm": 0.8066208315679838,
+      "learning_rate": 4.429023017257319e-06,
+      "loss": 0.3573,
+      "step": 814
+    },
+    {
+      "epoch": 0.24328358208955222,
+      "grad_norm": 0.8304243577622432,
+      "learning_rate": 4.42748444467791e-06,
+      "loss": 0.3086,
+      "step": 815
+    },
+    {
+      "epoch": 0.2435820895522388,
+      "grad_norm": 0.8380366969931612,
+      "learning_rate": 4.425944069947528e-06,
+      "loss": 0.3458,
+      "step": 816
+    },
+    {
+      "epoch": 0.24388059701492537,
+      "grad_norm": 0.7673079862377152,
+      "learning_rate": 4.424401894506386e-06,
+      "loss": 0.3514,
+      "step": 817
+    },
+    {
+      "epoch": 0.24417910447761193,
+      "grad_norm": 0.8657431046670215,
+      "learning_rate": 4.4228579197963795e-06,
+      "loss": 0.3648,
+      "step": 818
+    },
+    {
+      "epoch": 0.24447761194029852,
+      "grad_norm": 0.8310942286263857,
+      "learning_rate": 4.421312147261087e-06,
+      "loss": 0.2978,
+      "step": 819
+    },
+    {
+      "epoch": 0.24477611940298508,
+      "grad_norm": 0.8444445715616047,
+      "learning_rate": 4.4197645783457695e-06,
+      "loss": 0.3907,
+      "step": 820
+    },
+    {
+      "epoch": 0.24507462686567164,
+      "grad_norm": 0.9112359004688853,
+      "learning_rate": 4.418215214497366e-06,
+      "loss": 0.3833,
+      "step": 821
+    },
+    {
+      "epoch": 0.24537313432835822,
+      "grad_norm": 0.8107922970059999,
+      "learning_rate": 4.416664057164493e-06,
+      "loss": 0.3672,
+      "step": 822
+    },
+    {
+      "epoch": 0.24567164179104478,
+      "grad_norm": 0.8722125495392787,
+      "learning_rate": 4.415111107797445e-06,
+      "loss": 0.3617,
+      "step": 823
+    },
+    {
+      "epoch": 0.24597014925373134,
+      "grad_norm": 1.061183148831863,
+      "learning_rate": 4.413556367848193e-06,
+      "loss": 0.413,
+      "step": 824
+    },
+    {
+      "epoch": 0.2462686567164179,
+      "grad_norm": 0.7884747931084407,
+      "learning_rate": 4.41199983877038e-06,
+      "loss": 0.375,
+      "step": 825
+    },
+    {
+      "epoch": 0.2465671641791045,
+      "grad_norm": 0.93921749874484,
+      "learning_rate": 4.410441522019322e-06,
+      "loss": 0.3671,
+      "step": 826
+    },
+    {
+      "epoch": 0.24686567164179105,
+      "grad_norm": 0.8800331687997713,
+      "learning_rate": 4.4088814190520085e-06,
+      "loss": 0.347,
+      "step": 827
+    },
+    {
+      "epoch": 0.2471641791044776,
+      "grad_norm": 0.8608882108195471,
+      "learning_rate": 4.407319531327097e-06,
+      "loss": 0.3705,
+      "step": 828
+    },
+    {
+      "epoch": 0.2474626865671642,
+      "grad_norm": 0.8578378741311872,
+      "learning_rate": 4.405755860304915e-06,
+      "loss": 0.3574,
+      "step": 829
+    },
+    {
+      "epoch": 0.24776119402985075,
+      "grad_norm": 0.8321521972748156,
+      "learning_rate": 4.404190407447456e-06,
+      "loss": 0.3639,
+      "step": 830
+    },
+    {
+      "epoch": 0.2480597014925373,
+      "grad_norm": 0.8170637261740442,
+      "learning_rate": 4.402623174218381e-06,
+      "loss": 0.3568,
+      "step": 831
+    },
+    {
+      "epoch": 0.24835820895522387,
+      "grad_norm": 0.8552394496252084,
+      "learning_rate": 4.401054162083015e-06,
+      "loss": 0.3982,
+      "step": 832
+    },
+    {
+      "epoch": 0.24865671641791046,
+      "grad_norm": 0.8273057774794584,
+      "learning_rate": 4.399483372508345e-06,
+      "loss": 0.3223,
+      "step": 833
+    },
+    {
+      "epoch": 0.24895522388059702,
+      "grad_norm": 0.8012455720488535,
+      "learning_rate": 4.3979108069630226e-06,
+      "loss": 0.3675,
+      "step": 834
+    },
+    {
+      "epoch": 0.24925373134328357,
+      "grad_norm": 0.8788275931412666,
+      "learning_rate": 4.396336466917357e-06,
+      "loss": 0.3587,
+      "step": 835
+    },
+    {
+      "epoch": 0.24955223880597016,
+      "grad_norm": 0.8481959047937334,
+      "learning_rate": 4.394760353843318e-06,
+      "loss": 0.3916,
+      "step": 836
+    },
+    {
+      "epoch": 0.24985074626865672,
+      "grad_norm": 0.820980214165406,
+      "learning_rate": 4.393182469214533e-06,
+      "loss": 0.3731,
+      "step": 837
+    },
+    {
+      "epoch": 0.2501492537313433,
+      "grad_norm": 0.8685340899908907,
+      "learning_rate": 4.391602814506285e-06,
+      "loss": 0.3422,
+      "step": 838
+    },
+    {
+      "epoch": 0.25044776119402984,
+      "grad_norm": 0.8722156332327601,
+      "learning_rate": 4.390021391195514e-06,
+      "loss": 0.3848,
+      "step": 839
+    },
+    {
+      "epoch": 0.2507462686567164,
+      "grad_norm": 0.8131879266539699,
+      "learning_rate": 4.388438200760812e-06,
+      "loss": 0.3473,
+      "step": 840
+    },
+    {
+      "epoch": 0.251044776119403,
+      "grad_norm": 0.8738620972075055,
+      "learning_rate": 4.386853244682422e-06,
+      "loss": 0.3505,
+      "step": 841
+    },
+    {
+      "epoch": 0.25134328358208957,
+      "grad_norm": 0.8455634576936628,
+      "learning_rate": 4.385266524442241e-06,
+      "loss": 0.3656,
+      "step": 842
+    },
+    {
+      "epoch": 0.25164179104477613,
+      "grad_norm": 0.9232241608401891,
+      "learning_rate": 4.383678041523813e-06,
+      "loss": 0.3308,
+      "step": 843
+    },
+    {
+      "epoch": 0.2519402985074627,
+      "grad_norm": 0.7802438443664448,
+      "learning_rate": 4.382087797412331e-06,
+      "loss": 0.3177,
+      "step": 844
+    },
+    {
+      "epoch": 0.25223880597014925,
+      "grad_norm": 0.840623387098109,
+      "learning_rate": 4.380495793594634e-06,
+      "loss": 0.3978,
+      "step": 845
+    },
+    {
+      "epoch": 0.2525373134328358,
+      "grad_norm": 0.8598753419489917,
+      "learning_rate": 4.378902031559208e-06,
+      "loss": 0.3643,
+      "step": 846
+    },
+    {
+      "epoch": 0.25283582089552237,
+      "grad_norm": 0.8502479164780553,
+      "learning_rate": 4.377306512796179e-06,
+      "loss": 0.3116,
+      "step": 847
+    },
+    {
+      "epoch": 0.253134328358209,
+      "grad_norm": 0.8471892836766642,
+      "learning_rate": 4.375709238797322e-06,
+      "loss": 0.3459,
+      "step": 848
+    },
+    {
+      "epoch": 0.25343283582089554,
+      "grad_norm": 0.8949946135476528,
+      "learning_rate": 4.3741102110560465e-06,
+      "loss": 0.4197,
+      "step": 849
+    },
+    {
+      "epoch": 0.2537313432835821,
+      "grad_norm": 0.8390558579599957,
+      "learning_rate": 4.372509431067404e-06,
+      "loss": 0.3641,
+      "step": 850
+    },
+    {
+      "epoch": 0.25402985074626866,
+      "grad_norm": 0.8295363073039969,
+      "learning_rate": 4.370906900328087e-06,
+      "loss": 0.3085,
+      "step": 851
+    },
+    {
+      "epoch": 0.2543283582089552,
+      "grad_norm": 0.8920142421824904,
+      "learning_rate": 4.369302620336422e-06,
+      "loss": 0.3605,
+      "step": 852
+    },
+    {
+      "epoch": 0.2546268656716418,
+      "grad_norm": 0.8314417922487575,
+      "learning_rate": 4.367696592592371e-06,
+      "loss": 0.3648,
+      "step": 853
+    },
+    {
+      "epoch": 0.25492537313432834,
+      "grad_norm": 0.9416237229806054,
+      "learning_rate": 4.36608881859753e-06,
+      "loss": 0.4322,
+      "step": 854
+    },
+    {
+      "epoch": 0.25522388059701495,
+      "grad_norm": 0.8552595456528146,
+      "learning_rate": 4.364479299855131e-06,
+      "loss": 0.376,
+      "step": 855
+    },
+    {
+      "epoch": 0.2555223880597015,
+      "grad_norm": 0.7986846174676006,
+      "learning_rate": 4.362868037870033e-06,
+      "loss": 0.3129,
+      "step": 856
+    },
+    {
+      "epoch": 0.25582089552238807,
+      "grad_norm": 0.8367097908730075,
+      "learning_rate": 4.361255034148727e-06,
+      "loss": 0.34,
+      "step": 857
+    },
+    {
+      "epoch": 0.25611940298507463,
+      "grad_norm": 0.8933913196848361,
+      "learning_rate": 4.359640290199331e-06,
+      "loss": 0.3574,
+      "step": 858
+    },
+    {
+      "epoch": 0.2564179104477612,
+      "grad_norm": 0.8314203802986557,
+      "learning_rate": 4.3580238075315954e-06,
+      "loss": 0.3454,
+      "step": 859
+    },
+    {
+      "epoch": 0.25671641791044775,
+      "grad_norm": 0.8392401641828928,
+      "learning_rate": 4.356405587656886e-06,
+      "loss": 0.3616,
+      "step": 860
+    },
+    {
+      "epoch": 0.2570149253731343,
+      "grad_norm": 0.8057344142131634,
+      "learning_rate": 4.354785632088204e-06,
+      "loss": 0.3429,
+      "step": 861
+    },
+    {
+      "epoch": 0.2573134328358209,
+      "grad_norm": 0.7265185841371562,
+      "learning_rate": 4.353163942340166e-06,
+      "loss": 0.3316,
+      "step": 862
+    },
+    {
+      "epoch": 0.2576119402985075,
+      "grad_norm": 0.8861442387020552,
+      "learning_rate": 4.351540519929013e-06,
+      "loss": 0.3682,
+      "step": 863
+    },
+    {
+      "epoch": 0.25791044776119404,
+      "grad_norm": 0.8207676152529324,
+      "learning_rate": 4.349915366372605e-06,
+      "loss": 0.3465,
+      "step": 864
+    },
+    {
+      "epoch": 0.2582089552238806,
+      "grad_norm": 0.8401159847300563,
+      "learning_rate": 4.348288483190422e-06,
+      "loss": 0.3628,
+      "step": 865
+    },
+    {
+      "epoch": 0.25850746268656716,
+      "grad_norm": 0.9061691939819366,
+      "learning_rate": 4.346659871903558e-06,
+      "loss": 0.3269,
+      "step": 866
+    },
+    {
+      "epoch": 0.2588059701492537,
+      "grad_norm": 0.9554654260605913,
+      "learning_rate": 4.345029534034727e-06,
+      "loss": 0.3688,
+      "step": 867
+    },
+    {
+      "epoch": 0.2591044776119403,
+      "grad_norm": 0.8768435732446233,
+      "learning_rate": 4.343397471108254e-06,
+      "loss": 0.3467,
+      "step": 868
+    },
+    {
+      "epoch": 0.2594029850746269,
+      "grad_norm": 0.83519783168284,
+      "learning_rate": 4.341763684650078e-06,
+      "loss": 0.3664,
+      "step": 869
+    },
+    {
+      "epoch": 0.25970149253731345,
+      "grad_norm": 0.8733824091023256,
+      "learning_rate": 4.340128176187751e-06,
+      "loss": 0.3894,
+      "step": 870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7435126731069427,
+      "learning_rate": 4.338490947250431e-06,
+      "loss": 0.3265,
+      "step": 871
+    },
+    {
+      "epoch": 0.26029850746268657,
+      "grad_norm": 0.7602478532739807,
+      "learning_rate": 4.33685199936889e-06,
+      "loss": 0.3255,
+      "step": 872
+    },
+    {
+      "epoch": 0.2605970149253731,
+      "grad_norm": 0.8396951710141938,
+      "learning_rate": 4.335211334075502e-06,
+      "loss": 0.3772,
+      "step": 873
+    },
+    {
+      "epoch": 0.2608955223880597,
+      "grad_norm": 0.8535455460088422,
+      "learning_rate": 4.33356895290425e-06,
+      "loss": 0.3221,
+      "step": 874
+    },
+    {
+      "epoch": 0.26119402985074625,
+      "grad_norm": 0.934656189548925,
+      "learning_rate": 4.331924857390722e-06,
+      "loss": 0.3899,
+      "step": 875
+    },
+    {
+      "epoch": 0.26149253731343286,
+      "grad_norm": 0.7839885412025712,
+      "learning_rate": 4.330279049072103e-06,
+      "loss": 0.3278,
+      "step": 876
+    },
+    {
+      "epoch": 0.2617910447761194,
+      "grad_norm": 0.8288101118788596,
+      "learning_rate": 4.328631529487188e-06,
+      "loss": 0.359,
+      "step": 877
+    },
+    {
+      "epoch": 0.262089552238806,
+      "grad_norm": 0.7635692010308125,
+      "learning_rate": 4.326982300176366e-06,
+      "loss": 0.3315,
+      "step": 878
+    },
+    {
+      "epoch": 0.26238805970149254,
+      "grad_norm": 0.8317735162263449,
+      "learning_rate": 4.325331362681624e-06,
+      "loss": 0.3464,
+      "step": 879
+    },
+    {
+      "epoch": 0.2626865671641791,
+      "grad_norm": 0.8075219201660886,
+      "learning_rate": 4.323678718546552e-06,
+      "loss": 0.3477,
+      "step": 880
+    },
+    {
+      "epoch": 0.26298507462686566,
+      "grad_norm": 0.8112126719014313,
+      "learning_rate": 4.3220243693163305e-06,
+      "loss": 0.4148,
+      "step": 881
+    },
+    {
+      "epoch": 0.2632835820895522,
+      "grad_norm": 0.8176840947105901,
+      "learning_rate": 4.3203683165377355e-06,
+      "loss": 0.4178,
+      "step": 882
+    },
+    {
+      "epoch": 0.26358208955223883,
+      "grad_norm": 1.0009446140687355,
+      "learning_rate": 4.318710561759137e-06,
+      "loss": 0.3289,
+      "step": 883
+    },
+    {
+      "epoch": 0.2638805970149254,
+      "grad_norm": 0.7534551657111695,
+      "learning_rate": 4.317051106530492e-06,
+      "loss": 0.3402,
+      "step": 884
+    },
+    {
+      "epoch": 0.26417910447761195,
+      "grad_norm": 0.9072522401528668,
+      "learning_rate": 4.315389952403355e-06,
+      "loss": 0.3817,
+      "step": 885
+    },
+    {
+      "epoch": 0.2644776119402985,
+      "grad_norm": 0.8682117406925437,
+      "learning_rate": 4.313727100930862e-06,
+      "loss": 0.3915,
+      "step": 886
+    },
+    {
+      "epoch": 0.26477611940298507,
+      "grad_norm": 0.8515997109671464,
+      "learning_rate": 4.312062553667739e-06,
+      "loss": 0.3397,
+      "step": 887
+    },
+    {
+      "epoch": 0.2650746268656716,
+      "grad_norm": 0.837060052477179,
+      "learning_rate": 4.310396312170298e-06,
+      "loss": 0.3495,
+      "step": 888
+    },
+    {
+      "epoch": 0.2653731343283582,
+      "grad_norm": 0.960847057443914,
+      "learning_rate": 4.308728377996433e-06,
+      "loss": 0.345,
+      "step": 889
+    },
+    {
+      "epoch": 0.2656716417910448,
+      "grad_norm": 0.8594679219941264,
+      "learning_rate": 4.307058752705623e-06,
+      "loss": 0.3193,
+      "step": 890
+    },
+    {
+      "epoch": 0.26597014925373136,
+      "grad_norm": 0.8579668256137557,
+      "learning_rate": 4.3053874378589265e-06,
+      "loss": 0.3587,
+      "step": 891
+    },
+    {
+      "epoch": 0.2662686567164179,
+      "grad_norm": 0.815761166827642,
+      "learning_rate": 4.303714435018981e-06,
+      "loss": 0.392,
+      "step": 892
+    },
+    {
+      "epoch": 0.2665671641791045,
+      "grad_norm": 0.7691072442891018,
+      "learning_rate": 4.3020397457500055e-06,
+      "loss": 0.3449,
+      "step": 893
+    },
+    {
+      "epoch": 0.26686567164179104,
+      "grad_norm": 0.8546963433650799,
+      "learning_rate": 4.300363371617792e-06,
+      "loss": 0.3615,
+      "step": 894
+    },
+    {
+      "epoch": 0.2671641791044776,
+      "grad_norm": 0.8954925431678488,
+      "learning_rate": 4.29868531418971e-06,
+      "loss": 0.387,
+      "step": 895
+    },
+    {
+      "epoch": 0.26746268656716415,
+      "grad_norm": 0.8890912758541257,
+      "learning_rate": 4.297005575034701e-06,
+      "loss": 0.338,
+      "step": 896
+    },
+    {
+      "epoch": 0.26776119402985077,
+      "grad_norm": 0.9013891259842057,
+      "learning_rate": 4.295324155723283e-06,
+      "loss": 0.3627,
+      "step": 897
+    },
+    {
+      "epoch": 0.26805970149253733,
+      "grad_norm": 0.7909956097157833,
+      "learning_rate": 4.29364105782754e-06,
+      "loss": 0.3708,
+      "step": 898
+    },
+    {
+      "epoch": 0.2683582089552239,
+      "grad_norm": 0.8080884146208309,
+      "learning_rate": 4.291956282921129e-06,
+      "loss": 0.3257,
+      "step": 899
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 0.7819915704985843,
+      "learning_rate": 4.2902698325792715e-06,
+      "loss": 0.3916,
+      "step": 900
+    },
+    {
+      "epoch": 0.268955223880597,
+      "grad_norm": 0.8683299968211471,
+      "learning_rate": 4.2885817083787584e-06,
+      "loss": 0.3669,
+      "step": 901
+    },
+    {
+      "epoch": 0.26925373134328356,
+      "grad_norm": 0.8085758598925459,
+      "learning_rate": 4.286891911897944e-06,
+      "loss": 0.3187,
+      "step": 902
+    },
+    {
+      "epoch": 0.2695522388059701,
+      "grad_norm": 0.8590837328246107,
+      "learning_rate": 4.285200444716749e-06,
+      "loss": 0.3785,
+      "step": 903
+    },
+    {
+      "epoch": 0.26985074626865674,
+      "grad_norm": 0.8551915445353413,
+      "learning_rate": 4.283507308416651e-06,
+      "loss": 0.3542,
+      "step": 904
+    },
+    {
+      "epoch": 0.2701492537313433,
+      "grad_norm": 0.9079671987533409,
+      "learning_rate": 4.281812504580694e-06,
+      "loss": 0.3634,
+      "step": 905
+    },
+    {
+      "epoch": 0.27044776119402986,
+      "grad_norm": 0.8369860857576275,
+      "learning_rate": 4.280116034793477e-06,
+      "loss": 0.3708,
+      "step": 906
+    },
+    {
+      "epoch": 0.2707462686567164,
+      "grad_norm": 0.8664538534678876,
+      "learning_rate": 4.278417900641157e-06,
+      "loss": 0.4132,
+      "step": 907
+    },
+    {
+      "epoch": 0.271044776119403,
+      "grad_norm": 0.805731173102517,
+      "learning_rate": 4.2767181037114494e-06,
+      "loss": 0.3946,
+      "step": 908
+    },
+    {
+      "epoch": 0.27134328358208953,
+      "grad_norm": 0.8111842698029063,
+      "learning_rate": 4.275016645593622e-06,
+      "loss": 0.3616,
+      "step": 909
+    },
+    {
+      "epoch": 0.2716417910447761,
+      "grad_norm": 0.7534971014345517,
+      "learning_rate": 4.2733135278784975e-06,
+      "loss": 0.3284,
+      "step": 910
+    },
+    {
+      "epoch": 0.2719402985074627,
+      "grad_norm": 0.8332630746157706,
+      "learning_rate": 4.271608752158448e-06,
+      "loss": 0.3577,
+      "step": 911
+    },
+    {
+      "epoch": 0.27223880597014927,
+      "grad_norm": 0.8441430505402697,
+      "learning_rate": 4.269902320027399e-06,
+      "loss": 0.3927,
+      "step": 912
+    },
+    {
+      "epoch": 0.2725373134328358,
+      "grad_norm": 0.8088212649154801,
+      "learning_rate": 4.268194233080823e-06,
+      "loss": 0.3604,
+      "step": 913
+    },
+    {
+      "epoch": 0.2728358208955224,
+      "grad_norm": 0.8504371247895061,
+      "learning_rate": 4.266484492915738e-06,
+      "loss": 0.328,
+      "step": 914
+    },
+    {
+      "epoch": 0.27313432835820894,
+      "grad_norm": 0.8838721420845274,
+      "learning_rate": 4.264773101130711e-06,
+      "loss": 0.3491,
+      "step": 915
+    },
+    {
+      "epoch": 0.2734328358208955,
+      "grad_norm": 0.8378168662917916,
+      "learning_rate": 4.26306005932585e-06,
+      "loss": 0.349,
+      "step": 916
+    },
+    {
+      "epoch": 0.27373134328358206,
+      "grad_norm": 0.8952997918529593,
+      "learning_rate": 4.2613453691028085e-06,
+      "loss": 0.4315,
+      "step": 917
+    },
+    {
+      "epoch": 0.2740298507462687,
+      "grad_norm": 0.8262214290892325,
+      "learning_rate": 4.2596290320647795e-06,
+      "loss": 0.3247,
+      "step": 918
+    },
+    {
+      "epoch": 0.27432835820895524,
+      "grad_norm": 0.7936317313144646,
+      "learning_rate": 4.257911049816497e-06,
+      "loss": 0.3697,
+      "step": 919
+    },
+    {
+      "epoch": 0.2746268656716418,
+      "grad_norm": 0.8359230938626984,
+      "learning_rate": 4.256191423964231e-06,
+      "loss": 0.3641,
+      "step": 920
+    },
+    {
+      "epoch": 0.27492537313432835,
+      "grad_norm": 0.8439085515909566,
+      "learning_rate": 4.254470156115792e-06,
+      "loss": 0.3295,
+      "step": 921
+    },
+    {
+      "epoch": 0.2752238805970149,
+      "grad_norm": 0.9191921854597581,
+      "learning_rate": 4.252747247880521e-06,
+      "loss": 0.4339,
+      "step": 922
+    },
+    {
+      "epoch": 0.2755223880597015,
+      "grad_norm": 0.9149683459792712,
+      "learning_rate": 4.2510227008692974e-06,
+      "loss": 0.3599,
+      "step": 923
+    },
+    {
+      "epoch": 0.27582089552238803,
+      "grad_norm": 0.8114463730873631,
+      "learning_rate": 4.2492965166945295e-06,
+      "loss": 0.3911,
+      "step": 924
+    },
+    {
+      "epoch": 0.27611940298507465,
+      "grad_norm": 0.8792190926539251,
+      "learning_rate": 4.247568696970158e-06,
+      "loss": 0.3544,
+      "step": 925
+    },
+    {
+      "epoch": 0.2764179104477612,
+      "grad_norm": 0.8336709348423672,
+      "learning_rate": 4.2458392433116525e-06,
+      "loss": 0.3322,
+      "step": 926
+    },
+    {
+      "epoch": 0.27671641791044777,
+      "grad_norm": 1.020710396219329,
+      "learning_rate": 4.244108157336009e-06,
+      "loss": 0.3604,
+      "step": 927
+    },
+    {
+      "epoch": 0.2770149253731343,
+      "grad_norm": 0.8143571913624713,
+      "learning_rate": 4.2423754406617505e-06,
+      "loss": 0.3495,
+      "step": 928
+    },
+    {
+      "epoch": 0.2773134328358209,
+      "grad_norm": 0.8356882733846251,
+      "learning_rate": 4.2406410949089255e-06,
+      "loss": 0.368,
+      "step": 929
+    },
+    {
+      "epoch": 0.27761194029850744,
+      "grad_norm": 0.8645784597745637,
+      "learning_rate": 4.238905121699105e-06,
+      "loss": 0.3739,
+      "step": 930
+    },
+    {
+      "epoch": 0.27791044776119406,
+      "grad_norm": 0.8593811145607515,
+      "learning_rate": 4.237167522655382e-06,
+      "loss": 0.3174,
+      "step": 931
+    },
+    {
+      "epoch": 0.2782089552238806,
+      "grad_norm": 0.9107381802719423,
+      "learning_rate": 4.235428299402369e-06,
+      "loss": 0.4168,
+      "step": 932
+    },
+    {
+      "epoch": 0.2785074626865672,
+      "grad_norm": 0.8146960301870598,
+      "learning_rate": 4.2336874535661966e-06,
+      "loss": 0.3381,
+      "step": 933
+    },
+    {
+      "epoch": 0.27880597014925373,
+      "grad_norm": 0.7260268749376332,
+      "learning_rate": 4.231944986774513e-06,
+      "loss": 0.2668,
+      "step": 934
+    },
+    {
+      "epoch": 0.2791044776119403,
+      "grad_norm": 0.7381450828289706,
+      "learning_rate": 4.2302009006564845e-06,
+      "loss": 0.3379,
+      "step": 935
+    },
+    {
+      "epoch": 0.27940298507462685,
+      "grad_norm": 0.906697687644743,
+      "learning_rate": 4.228455196842787e-06,
+      "loss": 0.3525,
+      "step": 936
+    },
+    {
+      "epoch": 0.2797014925373134,
+      "grad_norm": 0.8801394409694048,
+      "learning_rate": 4.2267078769656115e-06,
+      "loss": 0.3684,
+      "step": 937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8229105599808183,
+      "learning_rate": 4.22495894265866e-06,
+      "loss": 0.3688,
+      "step": 938
+    },
+    {
+      "epoch": 0.2802985074626866,
+      "grad_norm": 0.8866551931642531,
+      "learning_rate": 4.223208395557142e-06,
+      "loss": 0.3939,
+      "step": 939
+    },
+    {
+      "epoch": 0.28059701492537314,
+      "grad_norm": 0.796873195210361,
+      "learning_rate": 4.221456237297779e-06,
+      "loss": 0.3246,
+      "step": 940
+    },
+    {
+      "epoch": 0.2808955223880597,
+      "grad_norm": 0.8777704725183845,
+      "learning_rate": 4.219702469518794e-06,
+      "loss": 0.3148,
+      "step": 941
+    },
+    {
+      "epoch": 0.28119402985074626,
+      "grad_norm": 0.8149960092397086,
+      "learning_rate": 4.217947093859917e-06,
+      "loss": 0.3261,
+      "step": 942
+    },
+    {
+      "epoch": 0.2814925373134328,
+      "grad_norm": 0.782356882918818,
+      "learning_rate": 4.216190111962383e-06,
+      "loss": 0.3594,
+      "step": 943
+    },
+    {
+      "epoch": 0.2817910447761194,
+      "grad_norm": 0.7838115738847105,
+      "learning_rate": 4.2144315254689265e-06,
+      "loss": 0.3658,
+      "step": 944
+    },
+    {
+      "epoch": 0.282089552238806,
+      "grad_norm": 0.861027660876004,
+      "learning_rate": 4.2126713360237835e-06,
+      "loss": 0.3216,
+      "step": 945
+    },
+    {
+      "epoch": 0.28238805970149256,
+      "grad_norm": 0.8979440300386647,
+      "learning_rate": 4.210909545272687e-06,
+      "loss": 0.347,
+      "step": 946
+    },
+    {
+      "epoch": 0.2826865671641791,
+      "grad_norm": 0.8046473678420953,
+      "learning_rate": 4.2091461548628695e-06,
+      "loss": 0.3817,
+      "step": 947
+    },
+    {
+      "epoch": 0.2829850746268657,
+      "grad_norm": 0.8999763252157099,
+      "learning_rate": 4.207381166443058e-06,
+      "loss": 0.3384,
+      "step": 948
+    },
+    {
+      "epoch": 0.28328358208955223,
+      "grad_norm": 1.0558611307630636,
+      "learning_rate": 4.205614581663472e-06,
+      "loss": 0.3764,
+      "step": 949
+    },
+    {
+      "epoch": 0.2835820895522388,
+      "grad_norm": 0.8250495089946933,
+      "learning_rate": 4.203846402175828e-06,
+      "loss": 0.3466,
+      "step": 950
+    },
+    {
+      "epoch": 0.28388059701492535,
+      "grad_norm": 0.8080126407334502,
+      "learning_rate": 4.202076629633329e-06,
+      "loss": 0.3379,
+      "step": 951
+    },
+    {
+      "epoch": 0.28417910447761197,
+      "grad_norm": 0.8189735637468861,
+      "learning_rate": 4.200305265690669e-06,
+      "loss": 0.3717,
+      "step": 952
+    },
+    {
+      "epoch": 0.2844776119402985,
+      "grad_norm": 0.8206896400863205,
+      "learning_rate": 4.198532312004031e-06,
+      "loss": 0.3424,
+      "step": 953
+    },
+    {
+      "epoch": 0.2847761194029851,
+      "grad_norm": 0.8182301432265623,
+      "learning_rate": 4.1967577702310826e-06,
+      "loss": 0.3459,
+      "step": 954
+    },
+    {
+      "epoch": 0.28507462686567164,
+      "grad_norm": 0.860760231935904,
+      "learning_rate": 4.194981642030978e-06,
+      "loss": 0.3934,
+      "step": 955
+    },
+    {
+      "epoch": 0.2853731343283582,
+      "grad_norm": 0.89467249534368,
+      "learning_rate": 4.1932039290643534e-06,
+      "loss": 0.3867,
+      "step": 956
+    },
+    {
+      "epoch": 0.28567164179104476,
+      "grad_norm": 0.9126227744137325,
+      "learning_rate": 4.191424632993326e-06,
+      "loss": 0.3452,
+      "step": 957
+    },
+    {
+      "epoch": 0.2859701492537313,
+      "grad_norm": 0.8112959766461156,
+      "learning_rate": 4.189643755481497e-06,
+      "loss": 0.2939,
+      "step": 958
+    },
+    {
+      "epoch": 0.28626865671641794,
+      "grad_norm": 0.9304054160150231,
+      "learning_rate": 4.18786129819394e-06,
+      "loss": 0.3461,
+      "step": 959
+    },
+    {
+      "epoch": 0.2865671641791045,
+      "grad_norm": 0.7922770615613615,
+      "learning_rate": 4.1860772627972125e-06,
+      "loss": 0.3435,
+      "step": 960
+    },
+    {
+      "epoch": 0.28686567164179105,
+      "grad_norm": 0.7959362445567673,
+      "learning_rate": 4.184291650959341e-06,
+      "loss": 0.3534,
+      "step": 961
+    },
+    {
+      "epoch": 0.2871641791044776,
+      "grad_norm": 0.8002993493679609,
+      "learning_rate": 4.182504464349832e-06,
+      "loss": 0.3316,
+      "step": 962
+    },
+    {
+      "epoch": 0.28746268656716417,
+      "grad_norm": 0.8167637473996379,
+      "learning_rate": 4.180715704639659e-06,
+      "loss": 0.3539,
+      "step": 963
+    },
+    {
+      "epoch": 0.28776119402985073,
+      "grad_norm": 0.802203149893345,
+      "learning_rate": 4.178925373501269e-06,
+      "loss": 0.3449,
+      "step": 964
+    },
+    {
+      "epoch": 0.2880597014925373,
+      "grad_norm": 0.8592664185496729,
+      "learning_rate": 4.17713347260858e-06,
+      "loss": 0.3828,
+      "step": 965
+    },
+    {
+      "epoch": 0.2883582089552239,
+      "grad_norm": 0.8304560933393987,
+      "learning_rate": 4.175340003636974e-06,
+      "loss": 0.3821,
+      "step": 966
+    },
+    {
+      "epoch": 0.28865671641791046,
+      "grad_norm": 0.9391386547612607,
+      "learning_rate": 4.173544968263301e-06,
+      "loss": 0.3846,
+      "step": 967
+    },
+    {
+      "epoch": 0.288955223880597,
+      "grad_norm": 0.8262795778364892,
+      "learning_rate": 4.171748368165875e-06,
+      "loss": 0.3383,
+      "step": 968
+    },
+    {
+      "epoch": 0.2892537313432836,
+      "grad_norm": 0.841522653230346,
+      "learning_rate": 4.169950205024474e-06,
+      "loss": 0.3355,
+      "step": 969
+    },
+    {
+      "epoch": 0.28955223880597014,
+      "grad_norm": 0.9913354298562964,
+      "learning_rate": 4.168150480520337e-06,
+      "loss": 0.419,
+      "step": 970
+    },
+    {
+      "epoch": 0.2898507462686567,
+      "grad_norm": 0.8071892265145763,
+      "learning_rate": 4.16634919633616e-06,
+      "loss": 0.3449,
+      "step": 971
+    },
+    {
+      "epoch": 0.29014925373134326,
+      "grad_norm": 0.8508150549887834,
+      "learning_rate": 4.164546354156104e-06,
+      "loss": 0.3672,
+      "step": 972
+    },
+    {
+      "epoch": 0.2904477611940299,
+      "grad_norm": 0.7899568543277075,
+      "learning_rate": 4.162741955665779e-06,
+      "loss": 0.3329,
+      "step": 973
+    },
+    {
+      "epoch": 0.29074626865671643,
+      "grad_norm": 0.80104254741407,
+      "learning_rate": 4.160936002552255e-06,
+      "loss": 0.3565,
+      "step": 974
+    },
+    {
+      "epoch": 0.291044776119403,
+      "grad_norm": 0.7742895075156045,
+      "learning_rate": 4.159128496504054e-06,
+      "loss": 0.3156,
+      "step": 975
+    },
+    {
+      "epoch": 0.29134328358208955,
+      "grad_norm": 0.8652205416649817,
+      "learning_rate": 4.157319439211151e-06,
+      "loss": 0.3567,
+      "step": 976
+    },
+    {
+      "epoch": 0.2916417910447761,
+      "grad_norm": 0.8552104838671654,
+      "learning_rate": 4.155508832364968e-06,
+      "loss": 0.316,
+      "step": 977
+    },
+    {
+      "epoch": 0.29194029850746267,
+      "grad_norm": 0.8757415112296788,
+      "learning_rate": 4.153696677658381e-06,
+      "loss": 0.3494,
+      "step": 978
+    },
+    {
+      "epoch": 0.29223880597014923,
+      "grad_norm": 0.8389906004653142,
+      "learning_rate": 4.151882976785709e-06,
+      "loss": 0.3291,
+      "step": 979
+    },
+    {
+      "epoch": 0.29253731343283584,
+      "grad_norm": 0.7800988745138925,
+      "learning_rate": 4.150067731442717e-06,
+      "loss": 0.3167,
+      "step": 980
+    },
+    {
+      "epoch": 0.2928358208955224,
+      "grad_norm": 0.8294199151333148,
+      "learning_rate": 4.148250943326619e-06,
+      "loss": 0.3623,
+      "step": 981
+    },
+    {
+      "epoch": 0.29313432835820896,
+      "grad_norm": 0.8578101736701281,
+      "learning_rate": 4.146432614136064e-06,
+      "loss": 0.307,
+      "step": 982
+    },
+    {
+      "epoch": 0.2934328358208955,
+      "grad_norm": 0.7856364156171273,
+      "learning_rate": 4.144612745571146e-06,
+      "loss": 0.3886,
+      "step": 983
+    },
+    {
+      "epoch": 0.2937313432835821,
+      "grad_norm": 0.8634192899478178,
+      "learning_rate": 4.1427913393333985e-06,
+      "loss": 0.3953,
+      "step": 984
+    },
+    {
+      "epoch": 0.29402985074626864,
+      "grad_norm": 0.8485105584492334,
+      "learning_rate": 4.140968397125793e-06,
+      "loss": 0.3629,
+      "step": 985
+    },
+    {
+      "epoch": 0.2943283582089552,
+      "grad_norm": 0.9742112909739822,
+      "learning_rate": 4.139143920652734e-06,
+      "loss": 0.3835,
+      "step": 986
+    },
+    {
+      "epoch": 0.2946268656716418,
+      "grad_norm": 0.9878802852692959,
+      "learning_rate": 4.137317911620063e-06,
+      "loss": 0.4003,
+      "step": 987
+    },
+    {
+      "epoch": 0.2949253731343284,
+      "grad_norm": 0.8815419197549593,
+      "learning_rate": 4.1354903717350556e-06,
+      "loss": 0.344,
+      "step": 988
+    },
+    {
+      "epoch": 0.29522388059701493,
+      "grad_norm": 0.834846080220852,
+      "learning_rate": 4.133661302706415e-06,
+      "loss": 0.3696,
+      "step": 989
+    },
+    {
+      "epoch": 0.2955223880597015,
+      "grad_norm": 0.845007866793147,
+      "learning_rate": 4.131830706244276e-06,
+      "loss": 0.3561,
+      "step": 990
+    },
+    {
+      "epoch": 0.29582089552238805,
+      "grad_norm": 0.8925684451459167,
+      "learning_rate": 4.129998584060204e-06,
+      "loss": 0.3805,
+      "step": 991
+    },
+    {
+      "epoch": 0.2961194029850746,
+      "grad_norm": 0.8371742760782053,
+      "learning_rate": 4.128164937867187e-06,
+      "loss": 0.3452,
+      "step": 992
+    },
+    {
+      "epoch": 0.29641791044776117,
+      "grad_norm": 0.8300330502109989,
+      "learning_rate": 4.12632976937964e-06,
+      "loss": 0.3614,
+      "step": 993
+    },
+    {
+      "epoch": 0.2967164179104478,
+      "grad_norm": 0.8341386461290142,
+      "learning_rate": 4.1244930803134e-06,
+      "loss": 0.3697,
+      "step": 994
+    },
+    {
+      "epoch": 0.29701492537313434,
+      "grad_norm": 0.9247230906706114,
+      "learning_rate": 4.122654872385726e-06,
+      "loss": 0.403,
+      "step": 995
+    },
+    {
+      "epoch": 0.2973134328358209,
+      "grad_norm": 0.815687144485872,
+      "learning_rate": 4.1208151473153e-06,
+      "loss": 0.3467,
+      "step": 996
+    },
+    {
+      "epoch": 0.29761194029850746,
+      "grad_norm": 0.9159383505804287,
+      "learning_rate": 4.118973906822218e-06,
+      "loss": 0.3492,
+      "step": 997
+    },
+    {
+      "epoch": 0.297910447761194,
+      "grad_norm": 0.8149117723593948,
+      "learning_rate": 4.117131152627996e-06,
+      "loss": 0.3371,
+      "step": 998
+    },
+    {
+      "epoch": 0.2982089552238806,
+      "grad_norm": 0.7227696644393177,
+      "learning_rate": 4.1152868864555626e-06,
+      "loss": 0.3277,
+      "step": 999
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.827257070561528,
+      "learning_rate": 4.113441110029265e-06,
+      "loss": 0.3227,
+      "step": 1000
+    },
+    {
+      "epoch": 0.29880597014925375,
+      "grad_norm": 0.8467214707843915,
+      "learning_rate": 4.111593825074856e-06,
+      "loss": 0.337,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2991044776119403,
+      "grad_norm": 0.8920138785669369,
+      "learning_rate": 4.109745033319504e-06,
+      "loss": 0.3982,
+      "step": 1002
+    },
+    {
+      "epoch": 0.29940298507462687,
+      "grad_norm": 0.8027122577683616,
+      "learning_rate": 4.107894736491783e-06,
+      "loss": 0.3513,
+      "step": 1003
+    },
+    {
+      "epoch": 0.29970149253731343,
+      "grad_norm": 1.2167741273718207,
+      "learning_rate": 4.106042936321676e-06,
+      "loss": 0.3689,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8627016240432611,
+      "learning_rate": 4.10418963454057e-06,
+      "loss": 0.3495,
+      "step": 1005
+    },
+    {
+      "epoch": 0.30029850746268655,
+      "grad_norm": 0.8585195931421936,
+      "learning_rate": 4.10233483288126e-06,
+      "loss": 0.317,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3005970149253731,
+      "grad_norm": 0.8429386990797031,
+      "learning_rate": 4.100478533077936e-06,
+      "loss": 0.3306,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3008955223880597,
+      "grad_norm": 0.7972123992318357,
+      "learning_rate": 4.098620736866196e-06,
+      "loss": 0.3493,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3011940298507463,
+      "grad_norm": 0.8661016284011678,
+      "learning_rate": 4.096761445983034e-06,
+      "loss": 0.3824,
+      "step": 1009
+    },
+    {
+      "epoch": 0.30149253731343284,
+      "grad_norm": 0.8703037201689672,
+      "learning_rate": 4.0949006621668405e-06,
+      "loss": 0.3902,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3017910447761194,
+      "grad_norm": 0.9376297455888316,
+      "learning_rate": 4.093038387157404e-06,
+      "loss": 0.3609,
+      "step": 1011
+    },
+    {
+      "epoch": 0.30208955223880596,
+      "grad_norm": 0.7998064580888031,
+      "learning_rate": 4.091174622695906e-06,
+      "loss": 0.3598,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3023880597014925,
+      "grad_norm": 0.8428236863313218,
+      "learning_rate": 4.089309370524921e-06,
+      "loss": 0.3676,
+      "step": 1013
+    },
+    {
+      "epoch": 0.30268656716417913,
+      "grad_norm": 0.8469734624648289,
+      "learning_rate": 4.087442632388413e-06,
+      "loss": 0.3791,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3029850746268657,
+      "grad_norm": 0.8517774956533634,
+      "learning_rate": 4.085574410031739e-06,
+      "loss": 0.3745,
+      "step": 1015
+    },
+    {
+      "epoch": 0.30328358208955225,
+      "grad_norm": 0.8281651202406228,
+      "learning_rate": 4.083704705201639e-06,
+      "loss": 0.3557,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3035820895522388,
+      "grad_norm": 0.9004766310930208,
+      "learning_rate": 4.081833519646242e-06,
+      "loss": 0.3408,
+      "step": 1017
+    },
+    {
+      "epoch": 0.30388059701492537,
+      "grad_norm": 0.7898205886854617,
+      "learning_rate": 4.0799608551150625e-06,
+      "loss": 0.37,
+      "step": 1018
+    },
+    {
+      "epoch": 0.30417910447761193,
+      "grad_norm": 0.8032659295487834,
+      "learning_rate": 4.078086713358994e-06,
+      "loss": 0.3506,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3044776119402985,
+      "grad_norm": 0.9141743806902359,
+      "learning_rate": 4.076211096130316e-06,
+      "loss": 0.3704,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3047761194029851,
+      "grad_norm": 0.8282796502709796,
+      "learning_rate": 4.074334005182682e-06,
+      "loss": 0.3571,
+      "step": 1021
+    },
+    {
+      "epoch": 0.30507462686567166,
+      "grad_norm": 0.8496527957007282,
+      "learning_rate": 4.072455442271128e-06,
+      "loss": 0.3163,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3053731343283582,
+      "grad_norm": 0.8977935551607475,
+      "learning_rate": 4.070575409152064e-06,
+      "loss": 0.4037,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3056716417910448,
+      "grad_norm": 0.8847365927783407,
+      "learning_rate": 4.068693907583276e-06,
+      "loss": 0.4017,
+      "step": 1024
+    },
+    {
+      "epoch": 0.30597014925373134,
+      "grad_norm": 0.7747661612730282,
+      "learning_rate": 4.06681093932392e-06,
+      "loss": 0.3397,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3062686567164179,
+      "grad_norm": 0.8825845190947826,
+      "learning_rate": 4.064926506134528e-06,
+      "loss": 0.3298,
+      "step": 1026
+    },
+    {
+      "epoch": 0.30656716417910446,
+      "grad_norm": 0.8592497304478822,
+      "learning_rate": 4.063040609776998e-06,
+      "loss": 0.349,
+      "step": 1027
+    },
+    {
+      "epoch": 0.30686567164179107,
+      "grad_norm": 0.8612997055386207,
+      "learning_rate": 4.0611532520145965e-06,
+      "loss": 0.3972,
+      "step": 1028
+    },
+    {
+      "epoch": 0.30716417910447763,
+      "grad_norm": 0.928710595753594,
+      "learning_rate": 4.059264434611957e-06,
+      "loss": 0.3751,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3074626865671642,
+      "grad_norm": 0.8652473522503297,
+      "learning_rate": 4.05737415933508e-06,
+      "loss": 0.3688,
+      "step": 1030
+    },
+    {
+      "epoch": 0.30776119402985075,
+      "grad_norm": 0.8978250178705632,
+      "learning_rate": 4.055482427951324e-06,
+      "loss": 0.3916,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3080597014925373,
+      "grad_norm": 0.9090489721659124,
+      "learning_rate": 4.053589242229412e-06,
+      "loss": 0.354,
+      "step": 1032
+    },
+    {
+      "epoch": 0.30835820895522387,
+      "grad_norm": 0.8801900400550225,
+      "learning_rate": 4.051694603939429e-06,
+      "loss": 0.3504,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3086567164179104,
+      "grad_norm": 0.8960014398639581,
+      "learning_rate": 4.049798514852812e-06,
+      "loss": 0.3423,
+      "step": 1034
+    },
+    {
+      "epoch": 0.30895522388059704,
+      "grad_norm": 0.798818576158037,
+      "learning_rate": 4.047900976742362e-06,
+      "loss": 0.3443,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3092537313432836,
+      "grad_norm": 0.8059330952772015,
+      "learning_rate": 4.046001991382227e-06,
+      "loss": 0.3743,
+      "step": 1036
+    },
+    {
+      "epoch": 0.30955223880597016,
+      "grad_norm": 0.8226038835739496,
+      "learning_rate": 4.044101560547915e-06,
+      "loss": 0.3515,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3098507462686567,
+      "grad_norm": 0.8230993349145885,
+      "learning_rate": 4.042199686016281e-06,
+      "loss": 0.3701,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3101492537313433,
+      "grad_norm": 0.7911941573133365,
+      "learning_rate": 4.040296369565532e-06,
+      "loss": 0.3396,
+      "step": 1039
+    },
+    {
+      "epoch": 0.31044776119402984,
+      "grad_norm": 0.8712443502975892,
+      "learning_rate": 4.038391612975222e-06,
+      "loss": 0.3927,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3107462686567164,
+      "grad_norm": 0.8233453064590726,
+      "learning_rate": 4.036485418026253e-06,
+      "loss": 0.3659,
+      "step": 1041
+    },
+    {
+      "epoch": 0.311044776119403,
+      "grad_norm": 0.8448815164890928,
+      "learning_rate": 4.034577786500869e-06,
+      "loss": 0.3512,
+      "step": 1042
+    },
+    {
+      "epoch": 0.31134328358208957,
+      "grad_norm": 0.8327563911984053,
+      "learning_rate": 4.0326687201826605e-06,
+      "loss": 0.3255,
+      "step": 1043
+    },
+    {
+      "epoch": 0.31164179104477613,
+      "grad_norm": 0.8220120310659925,
+      "learning_rate": 4.030758220856558e-06,
+      "loss": 0.3298,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3119402985074627,
+      "grad_norm": 0.8028153864586365,
+      "learning_rate": 4.02884629030883e-06,
+      "loss": 0.3678,
+      "step": 1045
+    },
+    {
+      "epoch": 0.31223880597014925,
+      "grad_norm": 0.8349137515362423,
+      "learning_rate": 4.026932930327088e-06,
+      "loss": 0.3783,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3125373134328358,
+      "grad_norm": 0.7766428030622539,
+      "learning_rate": 4.025018142700273e-06,
+      "loss": 0.2854,
+      "step": 1047
+    },
+    {
+      "epoch": 0.31283582089552237,
+      "grad_norm": 0.9182239119842932,
+      "learning_rate": 4.0231019292186685e-06,
+      "loss": 0.4211,
+      "step": 1048
+    },
+    {
+      "epoch": 0.313134328358209,
+      "grad_norm": 0.8518537676125332,
+      "learning_rate": 4.0211842916738855e-06,
+      "loss": 0.3695,
+      "step": 1049
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 0.7665539603822111,
+      "learning_rate": 4.019265231858869e-06,
+      "loss": 0.3235,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3137313432835821,
+      "grad_norm": 0.7995496595040446,
+      "learning_rate": 4.017344751567892e-06,
+      "loss": 0.3384,
+      "step": 1051
+    },
+    {
+      "epoch": 0.31402985074626866,
+      "grad_norm": 0.7860027976534756,
+      "learning_rate": 4.0154228525965585e-06,
+      "loss": 0.2767,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3143283582089552,
+      "grad_norm": 0.8580440875177763,
+      "learning_rate": 4.0134995367417965e-06,
+      "loss": 0.3525,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3146268656716418,
+      "grad_norm": 0.9336926190628849,
+      "learning_rate": 4.011574805801858e-06,
+      "loss": 0.3394,
+      "step": 1054
+    },
+    {
+      "epoch": 0.31492537313432833,
+      "grad_norm": 0.8425176180393162,
+      "learning_rate": 4.009648661576321e-06,
+      "loss": 0.3511,
+      "step": 1055
+    },
+    {
+      "epoch": 0.31522388059701495,
+      "grad_norm": 0.8276026034880272,
+      "learning_rate": 4.007721105866084e-06,
+      "loss": 0.3602,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3155223880597015,
+      "grad_norm": 0.787475800014448,
+      "learning_rate": 4.005792140473363e-06,
+      "loss": 0.3188,
+      "step": 1057
+    },
+    {
+      "epoch": 0.31582089552238807,
+      "grad_norm": 0.8880428207150297,
+      "learning_rate": 4.003861767201695e-06,
+      "loss": 0.3342,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3161194029850746,
+      "grad_norm": 0.8017106892627521,
+      "learning_rate": 4.001929987855931e-06,
+      "loss": 0.3502,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3164179104477612,
+      "grad_norm": 0.7848482690732528,
+      "learning_rate": 3.999996804242238e-06,
+      "loss": 0.3337,
+      "step": 1060
+    },
+    {
+      "epoch": 0.31671641791044775,
+      "grad_norm": 0.8099722326612471,
+      "learning_rate": 3.998062218168096e-06,
+      "loss": 0.3358,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3170149253731343,
+      "grad_norm": 0.8230124063451144,
+      "learning_rate": 3.996126231442295e-06,
+      "loss": 0.3732,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3173134328358209,
+      "grad_norm": 0.7714716419614884,
+      "learning_rate": 3.994188845874936e-06,
+      "loss": 0.3289,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3176119402985075,
+      "grad_norm": 0.7922647776658421,
+      "learning_rate": 3.992250063277427e-06,
+      "loss": 0.3154,
+      "step": 1064
+    },
+    {
+      "epoch": 0.31791044776119404,
+      "grad_norm": 0.8044498494994091,
+      "learning_rate": 3.990309885462481e-06,
+      "loss": 0.3378,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3182089552238806,
+      "grad_norm": 0.8820721478686536,
+      "learning_rate": 3.98836831424412e-06,
+      "loss": 0.3511,
+      "step": 1066
+    },
+    {
+      "epoch": 0.31850746268656716,
+      "grad_norm": 0.8572321551754879,
+      "learning_rate": 3.9864253514376634e-06,
+      "loss": 0.3646,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3188059701492537,
+      "grad_norm": 0.8802364307844034,
+      "learning_rate": 3.9844809988597355e-06,
+      "loss": 0.3448,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3191044776119403,
+      "grad_norm": 0.8291543922384458,
+      "learning_rate": 3.9825352583282585e-06,
+      "loss": 0.3219,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3194029850746269,
+      "grad_norm": 0.8993675406838366,
+      "learning_rate": 3.980588131662451e-06,
+      "loss": 0.4098,
+      "step": 1070
+    },
+    {
+      "epoch": 0.31970149253731345,
+      "grad_norm": 0.8157166363381144,
+      "learning_rate": 3.978639620682829e-06,
+      "loss": 0.3402,
+      "step": 1071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8073929451435011,
+      "learning_rate": 3.976689727211205e-06,
+      "loss": 0.3802,
+      "step": 1072
+    },
+    {
+      "epoch": 0.32029850746268657,
+      "grad_norm": 0.8488047650093633,
+      "learning_rate": 3.97473845307068e-06,
+      "loss": 0.3231,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3205970149253731,
+      "grad_norm": 0.8608697980068266,
+      "learning_rate": 3.972785800085647e-06,
+      "loss": 0.3523,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3208955223880597,
+      "grad_norm": 0.8021106185518122,
+      "learning_rate": 3.970831770081791e-06,
+      "loss": 0.3536,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32119402985074624,
+      "grad_norm": 0.8796976566955086,
+      "learning_rate": 3.968876364886082e-06,
+      "loss": 0.3275,
+      "step": 1076
+    },
+    {
+      "epoch": 0.32149253731343286,
+      "grad_norm": 0.8058440193793543,
+      "learning_rate": 3.966919586326775e-06,
+      "loss": 0.375,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3217910447761194,
+      "grad_norm": 0.8775522173367613,
+      "learning_rate": 3.964961436233412e-06,
+      "loss": 0.3588,
+      "step": 1078
+    },
+    {
+      "epoch": 0.322089552238806,
+      "grad_norm": 0.9098939150341124,
+      "learning_rate": 3.963001916436814e-06,
+      "loss": 0.3587,
+      "step": 1079
+    },
+    {
+      "epoch": 0.32238805970149254,
+      "grad_norm": 0.8473555634681235,
+      "learning_rate": 3.961041028769085e-06,
+      "loss": 0.3376,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3226865671641791,
+      "grad_norm": 0.9150706720781269,
+      "learning_rate": 3.959078775063607e-06,
+      "loss": 0.3745,
+      "step": 1081
+    },
+    {
+      "epoch": 0.32298507462686565,
+      "grad_norm": 0.9386215098514821,
+      "learning_rate": 3.95711515715504e-06,
+      "loss": 0.3778,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3232835820895522,
+      "grad_norm": 0.7483070851000795,
+      "learning_rate": 3.955150176879316e-06,
+      "loss": 0.3397,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3235820895522388,
+      "grad_norm": 0.6912249624197083,
+      "learning_rate": 3.953183836073649e-06,
+      "loss": 0.3304,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3238805970149254,
+      "grad_norm": 0.8348347584691441,
+      "learning_rate": 3.951216136576515e-06,
+      "loss": 0.3316,
+      "step": 1085
+    },
+    {
+      "epoch": 0.32417910447761195,
+      "grad_norm": 0.8690430526369205,
+      "learning_rate": 3.949247080227666e-06,
+      "loss": 0.3461,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3244776119402985,
+      "grad_norm": 0.8050979365143089,
+      "learning_rate": 3.947276668868124e-06,
+      "loss": 0.3642,
+      "step": 1087
+    },
+    {
+      "epoch": 0.32477611940298506,
+      "grad_norm": 0.7662634270514135,
+      "learning_rate": 3.945304904340174e-06,
+      "loss": 0.3445,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3250746268656716,
+      "grad_norm": 0.84107358699445,
+      "learning_rate": 3.943331788487366e-06,
+      "loss": 0.358,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3253731343283582,
+      "grad_norm": 0.824621628564205,
+      "learning_rate": 3.941357323154519e-06,
+      "loss": 0.2967,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3256716417910448,
+      "grad_norm": 0.806684118104856,
+      "learning_rate": 3.9393815101877076e-06,
+      "loss": 0.3335,
+      "step": 1091
+    },
+    {
+      "epoch": 0.32597014925373136,
+      "grad_norm": 0.8342281660142485,
+      "learning_rate": 3.937404351434269e-06,
+      "loss": 0.3748,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3262686567164179,
+      "grad_norm": 0.873757164412416,
+      "learning_rate": 3.935425848742797e-06,
+      "loss": 0.3692,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3265671641791045,
+      "grad_norm": 0.7814545073329886,
+      "learning_rate": 3.933446003963147e-06,
+      "loss": 0.3294,
+      "step": 1094
+    },
+    {
+      "epoch": 0.32686567164179103,
+      "grad_norm": 0.8482386135188483,
+      "learning_rate": 3.9314648189464226e-06,
+      "loss": 0.3259,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3271641791044776,
+      "grad_norm": 0.8036333351717462,
+      "learning_rate": 3.929482295544985e-06,
+      "loss": 0.3496,
+      "step": 1096
+    },
+    {
+      "epoch": 0.32746268656716415,
+      "grad_norm": 0.9025969332953864,
+      "learning_rate": 3.927498435612444e-06,
+      "loss": 0.3823,
+      "step": 1097
+    },
+    {
+      "epoch": 0.32776119402985077,
+      "grad_norm": 0.8330683787016878,
+      "learning_rate": 3.925513241003663e-06,
+      "loss": 0.3654,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3280597014925373,
+      "grad_norm": 0.9303082381552658,
+      "learning_rate": 3.923526713574747e-06,
+      "loss": 0.3674,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 1.181233124032353,
+      "learning_rate": 3.921538855183053e-06,
+      "loss": 0.364,
+      "step": 1100
+    },
+    {
+      "epoch": 0.32865671641791044,
+      "grad_norm": 0.8588441469224205,
+      "learning_rate": 3.919549667687178e-06,
+      "loss": 0.3635,
+      "step": 1101
+    },
+    {
+      "epoch": 0.328955223880597,
+      "grad_norm": 0.7804288737599779,
+      "learning_rate": 3.917559152946966e-06,
+      "loss": 0.3459,
+      "step": 1102
+    },
+    {
+      "epoch": 0.32925373134328356,
+      "grad_norm": 0.9027645900655169,
+      "learning_rate": 3.9155673128235e-06,
+      "loss": 0.3516,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3295522388059702,
+      "grad_norm": 0.8393324136480259,
+      "learning_rate": 3.9135741491791e-06,
+      "loss": 0.3685,
+      "step": 1104
+    },
+    {
+      "epoch": 0.32985074626865674,
+      "grad_norm": 0.7473370868680391,
+      "learning_rate": 3.9115796638773275e-06,
+      "loss": 0.2815,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3301492537313433,
+      "grad_norm": 0.8919025147790259,
+      "learning_rate": 3.9095838587829756e-06,
+      "loss": 0.3857,
+      "step": 1106
+    },
+    {
+      "epoch": 0.33044776119402985,
+      "grad_norm": 0.817241418807688,
+      "learning_rate": 3.907586735762074e-06,
+      "loss": 0.3732,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3307462686567164,
+      "grad_norm": 0.9218470406185262,
+      "learning_rate": 3.9055882966818855e-06,
+      "loss": 0.3376,
+      "step": 1108
+    },
+    {
+      "epoch": 0.331044776119403,
+      "grad_norm": 0.8177748051426704,
+      "learning_rate": 3.9035885434109014e-06,
+      "loss": 0.3184,
+      "step": 1109
+    },
+    {
+      "epoch": 0.33134328358208953,
+      "grad_norm": 0.7164318821917856,
+      "learning_rate": 3.90158747781884e-06,
+      "loss": 0.3142,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33164179104477615,
+      "grad_norm": 0.8400280769287423,
+      "learning_rate": 3.899585101776652e-06,
+      "loss": 0.302,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3319402985074627,
+      "grad_norm": 0.7773008415105561,
+      "learning_rate": 3.8975814171565075e-06,
+      "loss": 0.3799,
+      "step": 1112
+    },
+    {
+      "epoch": 0.33223880597014926,
+      "grad_norm": 0.897234720453567,
+      "learning_rate": 3.895576425831805e-06,
+      "loss": 0.3883,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3325373134328358,
+      "grad_norm": 0.8625016931833325,
+      "learning_rate": 3.893570129677161e-06,
+      "loss": 0.337,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3328358208955224,
+      "grad_norm": 0.8269153711097461,
+      "learning_rate": 3.8915625305684145e-06,
+      "loss": 0.3452,
+      "step": 1115
+    },
+    {
+      "epoch": 0.33313432835820894,
+      "grad_norm": 0.8632059531163434,
+      "learning_rate": 3.889553630382621e-06,
+      "loss": 0.3655,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3334328358208955,
+      "grad_norm": 0.7706052114027439,
+      "learning_rate": 3.8875434309980545e-06,
+      "loss": 0.3209,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3337313432835821,
+      "grad_norm": 0.8172417996393967,
+      "learning_rate": 3.8855319342942e-06,
+      "loss": 0.3588,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3340298507462687,
+      "grad_norm": 0.7457140782960014,
+      "learning_rate": 3.883519142151761e-06,
+      "loss": 0.3206,
+      "step": 1119
+    },
+    {
+      "epoch": 0.33432835820895523,
+      "grad_norm": 0.9075055402661403,
+      "learning_rate": 3.881505056452646e-06,
+      "loss": 0.388,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3346268656716418,
+      "grad_norm": 0.8633987237770478,
+      "learning_rate": 3.879489679079977e-06,
+      "loss": 0.3335,
+      "step": 1121
+    },
+    {
+      "epoch": 0.33492537313432835,
+      "grad_norm": 0.8015175637999997,
+      "learning_rate": 3.877473011918084e-06,
+      "loss": 0.3608,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3352238805970149,
+      "grad_norm": 0.9650469277739403,
+      "learning_rate": 3.8754550568525006e-06,
+      "loss": 0.3826,
+      "step": 1123
+    },
+    {
+      "epoch": 0.33552238805970147,
+      "grad_norm": 0.8531369547701335,
+      "learning_rate": 3.8734358157699666e-06,
+      "loss": 0.3907,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 0.8203181528488143,
+      "learning_rate": 3.871415290558421e-06,
+      "loss": 0.341,
+      "step": 1125
+    },
+    {
+      "epoch": 0.33611940298507464,
+      "grad_norm": 0.9506865101073372,
+      "learning_rate": 3.869393483107008e-06,
+      "loss": 0.3766,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3364179104477612,
+      "grad_norm": 0.7959518196575798,
+      "learning_rate": 3.8673703953060685e-06,
+      "loss": 0.3215,
+      "step": 1127
+    },
+    {
+      "epoch": 0.33671641791044776,
+      "grad_norm": 0.9587562044134879,
+      "learning_rate": 3.865346029047137e-06,
+      "loss": 0.425,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3370149253731343,
+      "grad_norm": 0.7933800883348958,
+      "learning_rate": 3.863320386222949e-06,
+      "loss": 0.321,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3373134328358209,
+      "grad_norm": 0.7786165000499077,
+      "learning_rate": 3.861293468727432e-06,
+      "loss": 0.3249,
+      "step": 1130
+    },
+    {
+      "epoch": 0.33761194029850744,
+      "grad_norm": 0.9211593511912554,
+      "learning_rate": 3.859265278455701e-06,
+      "loss": 0.3879,
+      "step": 1131
+    },
+    {
+      "epoch": 0.33791044776119405,
+      "grad_norm": 0.86664700650382,
+      "learning_rate": 3.8572358173040695e-06,
+      "loss": 0.3119,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3382089552238806,
+      "grad_norm": 0.8560710542669949,
+      "learning_rate": 3.85520508717003e-06,
+      "loss": 0.3335,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3385074626865672,
+      "grad_norm": 0.9378308935231966,
+      "learning_rate": 3.853173089952268e-06,
+      "loss": 0.3437,
+      "step": 1134
+    },
+    {
+      "epoch": 0.33880597014925373,
+      "grad_norm": 0.9021474935248037,
+      "learning_rate": 3.85113982755065e-06,
+      "loss": 0.3441,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3391044776119403,
+      "grad_norm": 0.8505218970552499,
+      "learning_rate": 3.849105301866227e-06,
+      "loss": 0.3595,
+      "step": 1136
+    },
+    {
+      "epoch": 0.33940298507462685,
+      "grad_norm": 1.0019110671578144,
+      "learning_rate": 3.847069514801232e-06,
+      "loss": 0.348,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3397014925373134,
+      "grad_norm": 0.896112478403269,
+      "learning_rate": 3.845032468259075e-06,
+      "loss": 0.3562,
+      "step": 1138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8272485945752788,
+      "learning_rate": 3.842994164144346e-06,
+      "loss": 0.3216,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3402985074626866,
+      "grad_norm": 0.7890421842215967,
+      "learning_rate": 3.840954604362809e-06,
+      "loss": 0.3341,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34059701492537314,
+      "grad_norm": 0.8688050179607455,
+      "learning_rate": 3.838913790821402e-06,
+      "loss": 0.3232,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3408955223880597,
+      "grad_norm": 1.0530125896366147,
+      "learning_rate": 3.8368717254282364e-06,
+      "loss": 0.389,
+      "step": 1142
+    },
+    {
+      "epoch": 0.34119402985074626,
+      "grad_norm": 0.9015300393082843,
+      "learning_rate": 3.834828410092595e-06,
+      "loss": 0.3431,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3414925373134328,
+      "grad_norm": 0.8299009904531546,
+      "learning_rate": 3.8327838467249255e-06,
+      "loss": 0.3644,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3417910447761194,
+      "grad_norm": 0.8043231321701094,
+      "learning_rate": 3.830738037236848e-06,
+      "loss": 0.3477,
+      "step": 1145
+    },
+    {
+      "epoch": 0.342089552238806,
+      "grad_norm": 0.753965995052167,
+      "learning_rate": 3.82869098354114e-06,
+      "loss": 0.3473,
+      "step": 1146
+    },
+    {
+      "epoch": 0.34238805970149255,
+      "grad_norm": 0.7796417555071672,
+      "learning_rate": 3.826642687551751e-06,
+      "loss": 0.3402,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3426865671641791,
+      "grad_norm": 0.8401115556379568,
+      "learning_rate": 3.824593151183785e-06,
+      "loss": 0.3521,
+      "step": 1148
+    },
+    {
+      "epoch": 0.34298507462686567,
+      "grad_norm": 0.8843481013387703,
+      "learning_rate": 3.82254237635351e-06,
+      "loss": 0.3826,
+      "step": 1149
+    },
+    {
+      "epoch": 0.34328358208955223,
+      "grad_norm": 0.8732317874736554,
+      "learning_rate": 3.82049036497835e-06,
+      "loss": 0.3005,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3435820895522388,
+      "grad_norm": 0.8189473191628437,
+      "learning_rate": 3.8184371189768855e-06,
+      "loss": 0.3464,
+      "step": 1151
+    },
+    {
+      "epoch": 0.34388059701492535,
+      "grad_norm": 1.107894296285839,
+      "learning_rate": 3.816382640268852e-06,
+      "loss": 0.3991,
+      "step": 1152
+    },
+    {
+      "epoch": 0.34417910447761196,
+      "grad_norm": 0.8200875925736727,
+      "learning_rate": 3.8143269307751373e-06,
+      "loss": 0.3489,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3444776119402985,
+      "grad_norm": 0.9460453425867347,
+      "learning_rate": 3.8122699924177786e-06,
+      "loss": 0.3836,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3447761194029851,
+      "grad_norm": 0.772503033014387,
+      "learning_rate": 3.8102118271199638e-06,
+      "loss": 0.3285,
+      "step": 1155
+    },
+    {
+      "epoch": 0.34507462686567164,
+      "grad_norm": 0.8815073767971605,
+      "learning_rate": 3.8081524368060273e-06,
+      "loss": 0.3988,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3453731343283582,
+      "grad_norm": 0.8885895081542389,
+      "learning_rate": 3.806091823401448e-06,
+      "loss": 0.3049,
+      "step": 1157
+    },
+    {
+      "epoch": 0.34567164179104476,
+      "grad_norm": 0.895020437102532,
+      "learning_rate": 3.80402998883285e-06,
+      "loss": 0.4085,
+      "step": 1158
+    },
+    {
+      "epoch": 0.3459701492537313,
+      "grad_norm": 0.9171169870830792,
+      "learning_rate": 3.8019669350279985e-06,
+      "loss": 0.3629,
+      "step": 1159
+    },
+    {
+      "epoch": 0.34626865671641793,
+      "grad_norm": 0.8081376377551374,
+      "learning_rate": 3.7999026639157983e-06,
+      "loss": 0.3319,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3465671641791045,
+      "grad_norm": 0.8402653067564256,
+      "learning_rate": 3.797837177426292e-06,
+      "loss": 0.3931,
+      "step": 1161
+    },
+    {
+      "epoch": 0.34686567164179105,
+      "grad_norm": 0.837482981106821,
+      "learning_rate": 3.79577047749066e-06,
+      "loss": 0.3157,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3471641791044776,
+      "grad_norm": 0.7786230841490198,
+      "learning_rate": 3.793702566041216e-06,
+      "loss": 0.3587,
+      "step": 1163
+    },
+    {
+      "epoch": 0.34746268656716417,
+      "grad_norm": 0.8720932882663056,
+      "learning_rate": 3.7916334450114073e-06,
+      "loss": 0.3514,
+      "step": 1164
+    },
+    {
+      "epoch": 0.34776119402985073,
+      "grad_norm": 0.8359944457321459,
+      "learning_rate": 3.7895631163358106e-06,
+      "loss": 0.345,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3480597014925373,
+      "grad_norm": 0.8019229336175024,
+      "learning_rate": 3.787491581950133e-06,
+      "loss": 0.3482,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3483582089552239,
+      "grad_norm": 0.8820518367174046,
+      "learning_rate": 3.7854188437912097e-06,
+      "loss": 0.3733,
+      "step": 1167
+    },
+    {
+      "epoch": 0.34865671641791046,
+      "grad_norm": 0.8403047783123814,
+      "learning_rate": 3.783344903796999e-06,
+      "loss": 0.3402,
+      "step": 1168
+    },
+    {
+      "epoch": 0.348955223880597,
+      "grad_norm": 0.7415680720483135,
+      "learning_rate": 3.7812697639065843e-06,
+      "loss": 0.3043,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3492537313432836,
+      "grad_norm": 0.743125129225478,
+      "learning_rate": 3.779193426060172e-06,
+      "loss": 0.322,
+      "step": 1170
+    },
+    {
+      "epoch": 0.34955223880597014,
+      "grad_norm": 0.8684344515884909,
+      "learning_rate": 3.7771158921990865e-06,
+      "loss": 0.4043,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3498507462686567,
+      "grad_norm": 0.8825012733228704,
+      "learning_rate": 3.7750371642657722e-06,
+      "loss": 0.3566,
+      "step": 1172
+    },
+    {
+      "epoch": 0.35014925373134326,
+      "grad_norm": 0.8095109148739124,
+      "learning_rate": 3.7729572442037877e-06,
+      "loss": 0.3221,
+      "step": 1173
+    },
+    {
+      "epoch": 0.35044776119402987,
+      "grad_norm": 0.8568019262787748,
+      "learning_rate": 3.7708761339578082e-06,
+      "loss": 0.3464,
+      "step": 1174
+    },
+    {
+      "epoch": 0.35074626865671643,
+      "grad_norm": 0.8431627633291477,
+      "learning_rate": 3.768793835473622e-06,
+      "loss": 0.3341,
+      "step": 1175
+    },
+    {
+      "epoch": 0.351044776119403,
+      "grad_norm": 0.8852857190511301,
+      "learning_rate": 3.766710350698125e-06,
+      "loss": 0.4152,
+      "step": 1176
+    },
+    {
+      "epoch": 0.35134328358208955,
+      "grad_norm": 0.85299125587319,
+      "learning_rate": 3.764625681579327e-06,
+      "loss": 0.3475,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3516417910447761,
+      "grad_norm": 0.9703649850814429,
+      "learning_rate": 3.762539830066343e-06,
+      "loss": 0.3788,
+      "step": 1178
+    },
+    {
+      "epoch": 0.35194029850746267,
+      "grad_norm": 0.8714521980948533,
+      "learning_rate": 3.760452798109391e-06,
+      "loss": 0.388,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3522388059701492,
+      "grad_norm": 0.7822951433929752,
+      "learning_rate": 3.758364587659796e-06,
+      "loss": 0.3682,
+      "step": 1180
+    },
+    {
+      "epoch": 0.35253731343283584,
+      "grad_norm": 0.785928197254092,
+      "learning_rate": 3.756275200669986e-06,
+      "loss": 0.3522,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3528358208955224,
+      "grad_norm": 0.8654717177672067,
+      "learning_rate": 3.754184639093484e-06,
+      "loss": 0.3752,
+      "step": 1182
+    },
+    {
+      "epoch": 0.35313432835820896,
+      "grad_norm": 0.8035897529969991,
+      "learning_rate": 3.7520929048849154e-06,
+      "loss": 0.3194,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3534328358208955,
+      "grad_norm": 0.8793098053746985,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3511,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3537313432835821,
+      "grad_norm": 0.9325267349782886,
+      "learning_rate": 3.747905926395554e-06,
+      "loss": 0.3763,
+      "step": 1185
+    },
+    {
+      "epoch": 0.35402985074626864,
+      "grad_norm": 0.948428300474812,
+      "learning_rate": 3.745810686029485e-06,
+      "loss": 0.3896,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3543283582089552,
+      "grad_norm": 0.7393311008440809,
+      "learning_rate": 3.743714280860791e-06,
+      "loss": 0.3037,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3546268656716418,
+      "grad_norm": 0.8041532895009262,
+      "learning_rate": 3.7416167128495596e-06,
+      "loss": 0.3509,
+      "step": 1188
+    },
+    {
+      "epoch": 0.35492537313432837,
+      "grad_norm": 0.7550348359422494,
+      "learning_rate": 3.7395179839569656e-06,
+      "loss": 0.3718,
+      "step": 1189
+    },
+    {
+      "epoch": 0.35522388059701493,
+      "grad_norm": 0.8634479052711203,
+      "learning_rate": 3.7374180961452704e-06,
+      "loss": 0.355,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3555223880597015,
+      "grad_norm": 0.7576581623194972,
+      "learning_rate": 3.735317051377818e-06,
+      "loss": 0.3351,
+      "step": 1191
+    },
+    {
+      "epoch": 0.35582089552238805,
+      "grad_norm": 0.8046261548713685,
+      "learning_rate": 3.7332148516190327e-06,
+      "loss": 0.358,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3561194029850746,
+      "grad_norm": 0.8406094996665042,
+      "learning_rate": 3.731111498834421e-06,
+      "loss": 0.4038,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3564179104477612,
+      "grad_norm": 0.7827745650131553,
+      "learning_rate": 3.7290069949905665e-06,
+      "loss": 0.317,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3567164179104478,
+      "grad_norm": 0.8857669556458677,
+      "learning_rate": 3.7269013420551286e-06,
+      "loss": 0.3502,
+      "step": 1195
+    },
+    {
+      "epoch": 0.35701492537313434,
+      "grad_norm": 0.9284361811395723,
+      "learning_rate": 3.7247945419968416e-06,
+      "loss": 0.3782,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3573134328358209,
+      "grad_norm": 0.8341033368677571,
+      "learning_rate": 3.722686596785513e-06,
+      "loss": 0.3861,
+      "step": 1197
+    },
+    {
+      "epoch": 0.35761194029850746,
+      "grad_norm": 0.9798340588994884,
+      "learning_rate": 3.720577508392018e-06,
+      "loss": 0.3544,
+      "step": 1198
+    },
+    {
+      "epoch": 0.357910447761194,
+      "grad_norm": 0.831648222529357,
+      "learning_rate": 3.7184672787883058e-06,
+      "loss": 0.3828,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.8402477136138807,
+      "learning_rate": 3.7163559099473874e-06,
+      "loss": 0.339,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3585074626865672,
+      "grad_norm": 0.7486776097613386,
+      "learning_rate": 3.7142434038433415e-06,
+      "loss": 0.3296,
+      "step": 1201
+    },
+    {
+      "epoch": 0.35880597014925375,
+      "grad_norm": 0.8479281742963486,
+      "learning_rate": 3.7121297624513108e-06,
+      "loss": 0.3301,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3591044776119403,
+      "grad_norm": 0.9005107528987859,
+      "learning_rate": 3.710014987747498e-06,
+      "loss": 0.4088,
+      "step": 1203
+    },
+    {
+      "epoch": 0.35940298507462687,
+      "grad_norm": 0.8059051563499257,
+      "learning_rate": 3.707899081709165e-06,
+      "loss": 0.3517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3597014925373134,
+      "grad_norm": 0.8552050817452692,
+      "learning_rate": 3.705782046314634e-06,
+      "loss": 0.3445,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9019044436471652,
+      "learning_rate": 3.7036638835432803e-06,
+      "loss": 0.3489,
+      "step": 1206
+    },
+    {
+      "epoch": 0.36029850746268655,
+      "grad_norm": 0.843202812648593,
+      "learning_rate": 3.7015445953755347e-06,
+      "loss": 0.3417,
+      "step": 1207
+    },
+    {
+      "epoch": 0.36059701492537316,
+      "grad_norm": 0.7443346275252014,
+      "learning_rate": 3.6994241837928803e-06,
+      "loss": 0.3425,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3608955223880597,
+      "grad_norm": 0.8737220707930936,
+      "learning_rate": 3.6973026507778497e-06,
+      "loss": 0.3388,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3611940298507463,
+      "grad_norm": 0.8319570014436627,
+      "learning_rate": 3.695179998314026e-06,
+      "loss": 0.3783,
+      "step": 1210
+    },
+    {
+      "epoch": 0.36149253731343284,
+      "grad_norm": 1.0254313443105356,
+      "learning_rate": 3.6930562283860356e-06,
+      "loss": 0.3262,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3617910447761194,
+      "grad_norm": 0.8868819070775161,
+      "learning_rate": 3.690931342979552e-06,
+      "loss": 0.3693,
+      "step": 1212
+    },
+    {
+      "epoch": 0.36208955223880596,
+      "grad_norm": 0.8512110198615549,
+      "learning_rate": 3.6888053440812928e-06,
+      "loss": 0.3534,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3623880597014925,
+      "grad_norm": 0.9606477539687734,
+      "learning_rate": 3.6866782336790137e-06,
+      "loss": 0.335,
+      "step": 1214
+    },
+    {
+      "epoch": 0.36268656716417913,
+      "grad_norm": 0.8299706662665464,
+      "learning_rate": 3.684550013761511e-06,
+      "loss": 0.3069,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3629850746268657,
+      "grad_norm": 0.8583028517456109,
+      "learning_rate": 3.6824206863186195e-06,
+      "loss": 0.3534,
+      "step": 1216
+    },
+    {
+      "epoch": 0.36328358208955225,
+      "grad_norm": 0.795689265721912,
+      "learning_rate": 3.680290253341207e-06,
+      "loss": 0.3217,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3635820895522388,
+      "grad_norm": 0.8155242986949494,
+      "learning_rate": 3.6781587168211785e-06,
+      "loss": 0.3139,
+      "step": 1218
+    },
+    {
+      "epoch": 0.36388059701492537,
+      "grad_norm": 0.9284470749372372,
+      "learning_rate": 3.676026078751466e-06,
+      "loss": 0.3994,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3641791044776119,
+      "grad_norm": 0.7846735459117187,
+      "learning_rate": 3.673892341126036e-06,
+      "loss": 0.42,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3644776119402985,
+      "grad_norm": 0.7394437630997585,
+      "learning_rate": 3.6717575059398818e-06,
+      "loss": 0.3425,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3647761194029851,
+      "grad_norm": 0.8410128872441591,
+      "learning_rate": 3.66962157518902e-06,
+      "loss": 0.309,
+      "step": 1222
+    },
+    {
+      "epoch": 0.36507462686567166,
+      "grad_norm": 0.7127446629430065,
+      "learning_rate": 3.6674845508704954e-06,
+      "loss": 0.3246,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3653731343283582,
+      "grad_norm": 0.8303226055174142,
+      "learning_rate": 3.665346434982373e-06,
+      "loss": 0.3568,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3656716417910448,
+      "grad_norm": 0.8759974317217685,
+      "learning_rate": 3.6632072295237385e-06,
+      "loss": 0.3725,
+      "step": 1225
+    },
+    {
+      "epoch": 0.36597014925373134,
+      "grad_norm": 0.9100058785675172,
+      "learning_rate": 3.6610669364946993e-06,
+      "loss": 0.3733,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3662686567164179,
+      "grad_norm": 0.8872689350119385,
+      "learning_rate": 3.6589255578963744e-06,
+      "loss": 0.3834,
+      "step": 1227
+    },
+    {
+      "epoch": 0.36656716417910445,
+      "grad_norm": 0.859809747889946,
+      "learning_rate": 3.656783095730902e-06,
+      "loss": 0.3207,
+      "step": 1228
+    },
+    {
+      "epoch": 0.36686567164179107,
+      "grad_norm": 0.743367008555944,
+      "learning_rate": 3.6546395520014324e-06,
+      "loss": 0.3311,
+      "step": 1229
+    },
+    {
+      "epoch": 0.36716417910447763,
+      "grad_norm": 0.8284032271379078,
+      "learning_rate": 3.6524949287121247e-06,
+      "loss": 0.3271,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3674626865671642,
+      "grad_norm": 0.9040431058647576,
+      "learning_rate": 3.650349227868151e-06,
+      "loss": 0.3805,
+      "step": 1231
+    },
+    {
+      "epoch": 0.36776119402985075,
+      "grad_norm": 0.788723325583723,
+      "learning_rate": 3.6482024514756905e-06,
+      "loss": 0.3369,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3680597014925373,
+      "grad_norm": 0.8292299897136874,
+      "learning_rate": 3.646054601541924e-06,
+      "loss": 0.3526,
+      "step": 1233
+    },
+    {
+      "epoch": 0.36835820895522386,
+      "grad_norm": 0.7780280218539215,
+      "learning_rate": 3.6439056800750406e-06,
+      "loss": 0.3404,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3686567164179104,
+      "grad_norm": 0.8640955670674403,
+      "learning_rate": 3.641755689084229e-06,
+      "loss": 0.359,
+      "step": 1235
+    },
+    {
+      "epoch": 0.36895522388059704,
+      "grad_norm": 0.9296153051752855,
+      "learning_rate": 3.6396046305796783e-06,
+      "loss": 0.3492,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3692537313432836,
+      "grad_norm": 0.8539806741089728,
+      "learning_rate": 3.6374525065725763e-06,
+      "loss": 0.3732,
+      "step": 1237
+    },
+    {
+      "epoch": 0.36955223880597016,
+      "grad_norm": 0.8791568970858674,
+      "learning_rate": 3.635299319075106e-06,
+      "loss": 0.3471,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3698507462686567,
+      "grad_norm": 0.8410283775191358,
+      "learning_rate": 3.6331450701004444e-06,
+      "loss": 0.3714,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3701492537313433,
+      "grad_norm": 0.8135682137125765,
+      "learning_rate": 3.6309897616627644e-06,
+      "loss": 0.3211,
+      "step": 1240
+    },
+    {
+      "epoch": 0.37044776119402983,
+      "grad_norm": 1.0986621561108818,
+      "learning_rate": 3.628833395777224e-06,
+      "loss": 0.3671,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3707462686567164,
+      "grad_norm": 0.8336925612419427,
+      "learning_rate": 3.626675974459974e-06,
+      "loss": 0.3488,
+      "step": 1242
+    },
+    {
+      "epoch": 0.371044776119403,
+      "grad_norm": 0.7967364289439731,
+      "learning_rate": 3.624517499728151e-06,
+      "loss": 0.3496,
+      "step": 1243
+    },
+    {
+      "epoch": 0.37134328358208957,
+      "grad_norm": 0.8235464715233171,
+      "learning_rate": 3.622357973599875e-06,
+      "loss": 0.3435,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3716417910447761,
+      "grad_norm": 0.7670146106690454,
+      "learning_rate": 3.6201973980942525e-06,
+      "loss": 0.3054,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3719402985074627,
+      "grad_norm": 0.8127305346359484,
+      "learning_rate": 3.618035775231367e-06,
+      "loss": 0.337,
+      "step": 1246
+    },
+    {
+      "epoch": 0.37223880597014924,
+      "grad_norm": 0.8739704252887466,
+      "learning_rate": 3.6158731070322833e-06,
+      "loss": 0.3904,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3725373134328358,
+      "grad_norm": 0.8589201227163017,
+      "learning_rate": 3.613709395519045e-06,
+      "loss": 0.3251,
+      "step": 1248
+    },
+    {
+      "epoch": 0.37283582089552236,
+      "grad_norm": 1.0287866995944308,
+      "learning_rate": 3.611544642714668e-06,
+      "loss": 0.342,
+      "step": 1249
+    },
+    {
+      "epoch": 0.373134328358209,
+      "grad_norm": 0.784684328253123,
+      "learning_rate": 3.609378850643144e-06,
+      "loss": 0.329,
+      "step": 1250
+    },
+    {
+      "epoch": 0.37343283582089554,
+      "grad_norm": 0.9957577337951871,
+      "learning_rate": 3.607212021329436e-06,
+      "loss": 0.3318,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3737313432835821,
+      "grad_norm": 0.8067111270985596,
+      "learning_rate": 3.6050441567994766e-06,
+      "loss": 0.291,
+      "step": 1252
+    },
+    {
+      "epoch": 0.37402985074626866,
+      "grad_norm": 0.814611603363173,
+      "learning_rate": 3.6028752590801653e-06,
+      "loss": 0.3724,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3743283582089552,
+      "grad_norm": 0.8212971184460834,
+      "learning_rate": 3.60070533019937e-06,
+      "loss": 0.3413,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3746268656716418,
+      "grad_norm": 0.8492409029801361,
+      "learning_rate": 3.5985343721859205e-06,
+      "loss": 0.4186,
+      "step": 1255
+    },
+    {
+      "epoch": 0.37492537313432833,
+      "grad_norm": 0.8293206499269064,
+      "learning_rate": 3.59636238706961e-06,
+      "loss": 0.3273,
+      "step": 1256
+    },
+    {
+      "epoch": 0.37522388059701495,
+      "grad_norm": 0.7439436947261291,
+      "learning_rate": 3.5941893768811915e-06,
+      "loss": 0.3171,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3755223880597015,
+      "grad_norm": 0.8417772206584229,
+      "learning_rate": 3.5920153436523762e-06,
+      "loss": 0.3727,
+      "step": 1258
+    },
+    {
+      "epoch": 0.37582089552238807,
+      "grad_norm": 0.7863852090186266,
+      "learning_rate": 3.589840289415833e-06,
+      "loss": 0.3473,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3761194029850746,
+      "grad_norm": 1.2126822142651668,
+      "learning_rate": 3.5876642162051833e-06,
+      "loss": 0.3347,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3764179104477612,
+      "grad_norm": 0.8029602835003353,
+      "learning_rate": 3.585487126055004e-06,
+      "loss": 0.3459,
+      "step": 1261
+    },
+    {
+      "epoch": 0.37671641791044774,
+      "grad_norm": 0.9206998569587098,
+      "learning_rate": 3.5833090210008204e-06,
+      "loss": 0.3634,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3770149253731343,
+      "grad_norm": 0.8641555534787082,
+      "learning_rate": 3.5811299030791074e-06,
+      "loss": 0.361,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3773134328358209,
+      "grad_norm": 0.7872304084887846,
+      "learning_rate": 3.578949774327288e-06,
+      "loss": 0.3268,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3776119402985075,
+      "grad_norm": 0.8680512566886363,
+      "learning_rate": 3.5767686367837295e-06,
+      "loss": 0.3436,
+      "step": 1265
+    },
+    {
+      "epoch": 0.37791044776119403,
+      "grad_norm": 0.826653018279734,
+      "learning_rate": 3.5745864924877412e-06,
+      "loss": 0.3587,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3782089552238806,
+      "grad_norm": 0.8881579540530804,
+      "learning_rate": 3.572403343479576e-06,
+      "loss": 0.3683,
+      "step": 1267
+    },
+    {
+      "epoch": 0.37850746268656715,
+      "grad_norm": 0.811695971069644,
+      "learning_rate": 3.570219191800424e-06,
+      "loss": 0.3417,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3788059701492537,
+      "grad_norm": 0.7952960741104704,
+      "learning_rate": 3.5680340394924144e-06,
+      "loss": 0.366,
+      "step": 1269
+    },
+    {
+      "epoch": 0.37910447761194027,
+      "grad_norm": 0.812111528734765,
+      "learning_rate": 3.565847888598612e-06,
+      "loss": 0.3031,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3794029850746269,
+      "grad_norm": 0.8238499604306999,
+      "learning_rate": 3.5636607411630133e-06,
+      "loss": 0.3639,
+      "step": 1271
+    },
+    {
+      "epoch": 0.37970149253731345,
+      "grad_norm": 1.2980492858910848,
+      "learning_rate": 3.5614725992305487e-06,
+      "loss": 0.3809,
+      "step": 1272
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7995909221221797,
+      "learning_rate": 3.5592834648470763e-06,
+      "loss": 0.3372,
+      "step": 1273
+    },
+    {
+      "epoch": 0.38029850746268656,
+      "grad_norm": 0.8051516225042442,
+      "learning_rate": 3.557093340059385e-06,
+      "loss": 0.3444,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3805970149253731,
+      "grad_norm": 0.7977178460505244,
+      "learning_rate": 3.5549022269151876e-06,
+      "loss": 0.3087,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3808955223880597,
+      "grad_norm": 0.8617753129014368,
+      "learning_rate": 3.552710127463121e-06,
+      "loss": 0.399,
+      "step": 1276
+    },
+    {
+      "epoch": 0.38119402985074624,
+      "grad_norm": 0.8085602277043268,
+      "learning_rate": 3.550517043752745e-06,
+      "loss": 0.3762,
+      "step": 1277
+    },
+    {
+      "epoch": 0.38149253731343286,
+      "grad_norm": 0.8125284069150092,
+      "learning_rate": 3.5483229778345403e-06,
+      "loss": 0.3532,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3817910447761194,
+      "grad_norm": 0.7339040692073175,
+      "learning_rate": 3.546127931759903e-06,
+      "loss": 0.3073,
+      "step": 1279
+    },
+    {
+      "epoch": 0.382089552238806,
+      "grad_norm": 0.7778739951423395,
+      "learning_rate": 3.5439319075811496e-06,
+      "loss": 0.3276,
+      "step": 1280
+    },
+    {
+      "epoch": 0.38238805970149253,
+      "grad_norm": 0.7892393982929026,
+      "learning_rate": 3.5417349073515085e-06,
+      "loss": 0.3358,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3826865671641791,
+      "grad_norm": 0.8241466550279929,
+      "learning_rate": 3.5395369331251205e-06,
+      "loss": 0.3427,
+      "step": 1282
+    },
+    {
+      "epoch": 0.38298507462686565,
+      "grad_norm": 0.8892819954376281,
+      "learning_rate": 3.53733798695704e-06,
+      "loss": 0.3166,
+      "step": 1283
+    },
+    {
+      "epoch": 0.38328358208955227,
+      "grad_norm": 0.8269062884932116,
+      "learning_rate": 3.5351380709032265e-06,
+      "loss": 0.3722,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3835820895522388,
+      "grad_norm": 0.905649324964948,
+      "learning_rate": 3.5329371870205477e-06,
+      "loss": 0.3142,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3838805970149254,
+      "grad_norm": 0.8106636299372714,
+      "learning_rate": 3.5307353373667772e-06,
+      "loss": 0.3387,
+      "step": 1286
+    },
+    {
+      "epoch": 0.38417910447761194,
+      "grad_norm": 0.811430184665767,
+      "learning_rate": 3.528532524000591e-06,
+      "loss": 0.3286,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3844776119402985,
+      "grad_norm": 0.8218271629980521,
+      "learning_rate": 3.5263287489815643e-06,
+      "loss": 0.3835,
+      "step": 1288
+    },
+    {
+      "epoch": 0.38477611940298506,
+      "grad_norm": 0.816710352160504,
+      "learning_rate": 3.524124014370175e-06,
+      "loss": 0.3582,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3850746268656716,
+      "grad_norm": 0.7305897895271865,
+      "learning_rate": 3.5219183222277954e-06,
+      "loss": 0.3257,
+      "step": 1290
+    },
+    {
+      "epoch": 0.38537313432835824,
+      "grad_norm": 0.7546275525313664,
+      "learning_rate": 3.519711674616694e-06,
+      "loss": 0.3344,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3856716417910448,
+      "grad_norm": 0.8631937455194517,
+      "learning_rate": 3.517504073600031e-06,
+      "loss": 0.351,
+      "step": 1292
+    },
+    {
+      "epoch": 0.38597014925373135,
+      "grad_norm": 0.8204252773546967,
+      "learning_rate": 3.5152955212418616e-06,
+      "loss": 0.2986,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3862686567164179,
+      "grad_norm": 0.8154202257697561,
+      "learning_rate": 3.5130860196071283e-06,
+      "loss": 0.3418,
+      "step": 1294
+    },
+    {
+      "epoch": 0.38656716417910447,
+      "grad_norm": 0.8526837643437616,
+      "learning_rate": 3.51087557076166e-06,
+      "loss": 0.3821,
+      "step": 1295
+    },
+    {
+      "epoch": 0.38686567164179103,
+      "grad_norm": 0.8428109150390479,
+      "learning_rate": 3.508664176772173e-06,
+      "loss": 0.3234,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3871641791044776,
+      "grad_norm": 0.8008585623503791,
+      "learning_rate": 3.506451839706268e-06,
+      "loss": 0.3399,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3874626865671642,
+      "grad_norm": 0.8269646097065997,
+      "learning_rate": 3.5042385616324243e-06,
+      "loss": 0.3505,
+      "step": 1298
+    },
+    {
+      "epoch": 0.38776119402985076,
+      "grad_norm": 0.7972302337986306,
+      "learning_rate": 3.5020243446200034e-06,
+      "loss": 0.3302,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 0.8833042412930123,
+      "learning_rate": 3.4998091907392463e-06,
+      "loss": 0.3793,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3883582089552239,
+      "grad_norm": 0.9850943543964298,
+      "learning_rate": 3.497593102061264e-06,
+      "loss": 0.3532,
+      "step": 1301
+    },
+    {
+      "epoch": 0.38865671641791044,
+      "grad_norm": 0.9676051683553516,
+      "learning_rate": 3.4953760806580494e-06,
+      "loss": 0.3635,
+      "step": 1302
+    },
+    {
+      "epoch": 0.388955223880597,
+      "grad_norm": 0.8090901217456243,
+      "learning_rate": 3.4931581286024607e-06,
+      "loss": 0.2871,
+      "step": 1303
+    },
+    {
+      "epoch": 0.38925373134328356,
+      "grad_norm": 0.8348404220329789,
+      "learning_rate": 3.4909392479682303e-06,
+      "loss": 0.3315,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3895522388059702,
+      "grad_norm": 0.7991008815236104,
+      "learning_rate": 3.488719440829958e-06,
+      "loss": 0.36,
+      "step": 1305
+    },
+    {
+      "epoch": 0.38985074626865673,
+      "grad_norm": 0.8438228763160631,
+      "learning_rate": 3.4864987092631074e-06,
+      "loss": 0.3869,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3901492537313433,
+      "grad_norm": 0.8489110544841167,
+      "learning_rate": 3.4842770553440103e-06,
+      "loss": 0.3675,
+      "step": 1307
+    },
+    {
+      "epoch": 0.39044776119402985,
+      "grad_norm": 0.7917206514304985,
+      "learning_rate": 3.4820544811498584e-06,
+      "loss": 0.3385,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3907462686567164,
+      "grad_norm": 1.0183165833490218,
+      "learning_rate": 3.479830988758704e-06,
+      "loss": 0.351,
+      "step": 1309
+    },
+    {
+      "epoch": 0.39104477611940297,
+      "grad_norm": 0.7666673934748648,
+      "learning_rate": 3.4776065802494585e-06,
+      "loss": 0.3123,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39134328358208953,
+      "grad_norm": 0.8481681061835175,
+      "learning_rate": 3.47538125770189e-06,
+      "loss": 0.3304,
+      "step": 1311
+    },
+    {
+      "epoch": 0.39164179104477614,
+      "grad_norm": 0.8626463063388562,
+      "learning_rate": 3.4731550231966193e-06,
+      "loss": 0.3429,
+      "step": 1312
+    },
+    {
+      "epoch": 0.3919402985074627,
+      "grad_norm": 0.7994976728058045,
+      "learning_rate": 3.470927878815124e-06,
+      "loss": 0.3368,
+      "step": 1313
+    },
+    {
+      "epoch": 0.39223880597014926,
+      "grad_norm": 0.8542479136676837,
+      "learning_rate": 3.4686998266397275e-06,
+      "loss": 0.3284,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3925373134328358,
+      "grad_norm": 0.8028452068951644,
+      "learning_rate": 3.466470868753606e-06,
+      "loss": 0.3699,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3928358208955224,
+      "grad_norm": 0.8192750229513925,
+      "learning_rate": 3.4642410072407797e-06,
+      "loss": 0.3374,
+      "step": 1316
+    },
+    {
+      "epoch": 0.39313432835820894,
+      "grad_norm": 0.7911937120174724,
+      "learning_rate": 3.4620102441861147e-06,
+      "loss": 0.3068,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3934328358208955,
+      "grad_norm": 0.7856414622293901,
+      "learning_rate": 3.4597785816753203e-06,
+      "loss": 0.3815,
+      "step": 1318
+    },
+    {
+      "epoch": 0.3937313432835821,
+      "grad_norm": 0.8719965350651078,
+      "learning_rate": 3.4575460217949475e-06,
+      "loss": 0.3249,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3940298507462687,
+      "grad_norm": 0.8038659900828976,
+      "learning_rate": 3.4553125666323828e-06,
+      "loss": 0.3609,
+      "step": 1320
+    },
+    {
+      "epoch": 0.39432835820895523,
+      "grad_norm": 0.8437738135633144,
+      "learning_rate": 3.453078218275856e-06,
+      "loss": 0.3547,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3946268656716418,
+      "grad_norm": 0.8893980819638573,
+      "learning_rate": 3.4508429788144255e-06,
+      "loss": 0.3835,
+      "step": 1322
+    },
+    {
+      "epoch": 0.39492537313432835,
+      "grad_norm": 0.7966994623889064,
+      "learning_rate": 3.4486068503379864e-06,
+      "loss": 0.3278,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3952238805970149,
+      "grad_norm": 0.8837643559806236,
+      "learning_rate": 3.4463698349372655e-06,
+      "loss": 0.4,
+      "step": 1324
+    },
+    {
+      "epoch": 0.39552238805970147,
+      "grad_norm": 0.8420254919427764,
+      "learning_rate": 3.444131934703816e-06,
+      "loss": 0.3145,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3958208955223881,
+      "grad_norm": 0.8412477035174881,
+      "learning_rate": 3.4418931517300207e-06,
+      "loss": 0.3318,
+      "step": 1326
+    },
+    {
+      "epoch": 0.39611940298507464,
+      "grad_norm": 0.7378862322828338,
+      "learning_rate": 3.4396534881090882e-06,
+      "loss": 0.3295,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3964179104477612,
+      "grad_norm": 0.814528536394005,
+      "learning_rate": 3.437412945935047e-06,
+      "loss": 0.3715,
+      "step": 1328
+    },
+    {
+      "epoch": 0.39671641791044776,
+      "grad_norm": 0.8472682635407821,
+      "learning_rate": 3.435171527302752e-06,
+      "loss": 0.3495,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3970149253731343,
+      "grad_norm": 0.7719473237614821,
+      "learning_rate": 3.4329292343078733e-06,
+      "loss": 0.3388,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3973134328358209,
+      "grad_norm": 0.8467959201969155,
+      "learning_rate": 3.430686069046901e-06,
+      "loss": 0.3357,
+      "step": 1331
+    },
+    {
+      "epoch": 0.39761194029850744,
+      "grad_norm": 0.8099659593829108,
+      "learning_rate": 3.4284420336171393e-06,
+      "loss": 0.3495,
+      "step": 1332
+    },
+    {
+      "epoch": 0.39791044776119405,
+      "grad_norm": 0.8144878750567746,
+      "learning_rate": 3.426197130116707e-06,
+      "loss": 0.3371,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3982089552238806,
+      "grad_norm": 0.7434099123184534,
+      "learning_rate": 3.423951360644534e-06,
+      "loss": 0.2841,
+      "step": 1334
+    },
+    {
+      "epoch": 0.39850746268656717,
+      "grad_norm": 0.8453291988577794,
+      "learning_rate": 3.4217047273003605e-06,
+      "loss": 0.3219,
+      "step": 1335
+    },
+    {
+      "epoch": 0.39880597014925373,
+      "grad_norm": 0.9447504107148086,
+      "learning_rate": 3.4194572321847336e-06,
+      "loss": 0.3831,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3991044776119403,
+      "grad_norm": 0.8778360194938137,
+      "learning_rate": 3.417208877399006e-06,
+      "loss": 0.363,
+      "step": 1337
+    },
+    {
+      "epoch": 0.39940298507462685,
+      "grad_norm": 0.858110018415135,
+      "learning_rate": 3.4149596650453354e-06,
+      "loss": 0.345,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3997014925373134,
+      "grad_norm": 0.8216385808533025,
+      "learning_rate": 3.4127095972266795e-06,
+      "loss": 0.4099,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8344309431851922,
+      "learning_rate": 3.4104586760467984e-06,
+      "loss": 0.3873,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4002985074626866,
+      "grad_norm": 0.8372552783746257,
+      "learning_rate": 3.408206903610247e-06,
+      "loss": 0.3612,
+      "step": 1341
+    },
+    {
+      "epoch": 0.40059701492537314,
+      "grad_norm": 0.8372600966982839,
+      "learning_rate": 3.4059542820223782e-06,
+      "loss": 0.3692,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4008955223880597,
+      "grad_norm": 0.9587410316998514,
+      "learning_rate": 3.4037008133893395e-06,
+      "loss": 0.3882,
+      "step": 1343
+    },
+    {
+      "epoch": 0.40119402985074626,
+      "grad_norm": 0.7735563287004625,
+      "learning_rate": 3.4014464998180673e-06,
+      "loss": 0.3486,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4014925373134328,
+      "grad_norm": 0.7836991921229477,
+      "learning_rate": 3.3991913434162905e-06,
+      "loss": 0.3254,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4017910447761194,
+      "grad_norm": 0.7627893713273461,
+      "learning_rate": 3.396935346292526e-06,
+      "loss": 0.3461,
+      "step": 1346
+    },
+    {
+      "epoch": 0.402089552238806,
+      "grad_norm": 0.8945538621615577,
+      "learning_rate": 3.3946785105560742e-06,
+      "loss": 0.312,
+      "step": 1347
+    },
+    {
+      "epoch": 0.40238805970149255,
+      "grad_norm": 0.8301080381657066,
+      "learning_rate": 3.3924208383170244e-06,
+      "loss": 0.384,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4026865671641791,
+      "grad_norm": 0.8441180205827015,
+      "learning_rate": 3.3901623316862424e-06,
+      "loss": 0.3237,
+      "step": 1349
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 0.917883619027885,
+      "learning_rate": 3.3879029927753782e-06,
+      "loss": 0.3451,
+      "step": 1350
+    },
+    {
+      "epoch": 0.40328358208955223,
+      "grad_norm": 0.7676285335284508,
+      "learning_rate": 3.3856428236968593e-06,
+      "loss": 0.3348,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4035820895522388,
+      "grad_norm": 0.7694667977006615,
+      "learning_rate": 3.3833818265638868e-06,
+      "loss": 0.3202,
+      "step": 1352
+    },
+    {
+      "epoch": 0.40388059701492535,
+      "grad_norm": 0.7859618711908494,
+      "learning_rate": 3.3811200034904392e-06,
+      "loss": 0.3447,
+      "step": 1353
+    },
+    {
+      "epoch": 0.40417910447761196,
+      "grad_norm": 0.7857504939863141,
+      "learning_rate": 3.3788573565912665e-06,
+      "loss": 0.2722,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4044776119402985,
+      "grad_norm": 0.782826471359172,
+      "learning_rate": 3.3765938879818865e-06,
+      "loss": 0.3318,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4047761194029851,
+      "grad_norm": 0.855706197844801,
+      "learning_rate": 3.3743295997785884e-06,
+      "loss": 0.3204,
+      "step": 1356
+    },
+    {
+      "epoch": 0.40507462686567164,
+      "grad_norm": 0.8343684499136247,
+      "learning_rate": 3.372064494098427e-06,
+      "loss": 0.3173,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4053731343283582,
+      "grad_norm": 0.8369262504159325,
+      "learning_rate": 3.3697985730592187e-06,
+      "loss": 0.3339,
+      "step": 1358
+    },
+    {
+      "epoch": 0.40567164179104476,
+      "grad_norm": 0.7799099543460535,
+      "learning_rate": 3.3675318387795473e-06,
+      "loss": 0.3423,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4059701492537313,
+      "grad_norm": 0.8648556337576385,
+      "learning_rate": 3.3652642933787526e-06,
+      "loss": 0.3493,
+      "step": 1360
+    },
+    {
+      "epoch": 0.40626865671641793,
+      "grad_norm": 0.8204663128828968,
+      "learning_rate": 3.362995938976934e-06,
+      "loss": 0.3317,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4065671641791045,
+      "grad_norm": 0.792102559533383,
+      "learning_rate": 3.3607267776949485e-06,
+      "loss": 0.3639,
+      "step": 1362
+    },
+    {
+      "epoch": 0.40686567164179105,
+      "grad_norm": 0.8308207742246098,
+      "learning_rate": 3.358456811654406e-06,
+      "loss": 0.3679,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4071641791044776,
+      "grad_norm": 0.8656195971316198,
+      "learning_rate": 3.35618604297767e-06,
+      "loss": 0.3797,
+      "step": 1364
+    },
+    {
+      "epoch": 0.40746268656716417,
+      "grad_norm": 0.8006170230930989,
+      "learning_rate": 3.3539144737878525e-06,
+      "loss": 0.3471,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4077611940298507,
+      "grad_norm": 0.8502026092595737,
+      "learning_rate": 3.351642106208816e-06,
+      "loss": 0.3814,
+      "step": 1366
+    },
+    {
+      "epoch": 0.40805970149253734,
+      "grad_norm": 0.7775186619424316,
+      "learning_rate": 3.3493689423651697e-06,
+      "loss": 0.3086,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4083582089552239,
+      "grad_norm": 0.798430336568009,
+      "learning_rate": 3.3470949843822657e-06,
+      "loss": 0.3406,
+      "step": 1368
+    },
+    {
+      "epoch": 0.40865671641791046,
+      "grad_norm": 0.8053325058264433,
+      "learning_rate": 3.3448202343861992e-06,
+      "loss": 0.3458,
+      "step": 1369
+    },
+    {
+      "epoch": 0.408955223880597,
+      "grad_norm": 0.8343334939840257,
+      "learning_rate": 3.3425446945038058e-06,
+      "loss": 0.337,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4092537313432836,
+      "grad_norm": 0.8191029497223187,
+      "learning_rate": 3.34026836686266e-06,
+      "loss": 0.3625,
+      "step": 1371
+    },
+    {
+      "epoch": 0.40955223880597014,
+      "grad_norm": 0.8618543242740297,
+      "learning_rate": 3.337991253591073e-06,
+      "loss": 0.3382,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4098507462686567,
+      "grad_norm": 0.7548489326509085,
+      "learning_rate": 3.3357133568180905e-06,
+      "loss": 0.3296,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4101492537313433,
+      "grad_norm": 0.842491815374688,
+      "learning_rate": 3.3334346786734894e-06,
+      "loss": 0.3493,
+      "step": 1374
+    },
+    {
+      "epoch": 0.41044776119402987,
+      "grad_norm": 0.8011968378789714,
+      "learning_rate": 3.3311552212877787e-06,
+      "loss": 0.3359,
+      "step": 1375
+    },
+    {
+      "epoch": 0.41074626865671643,
+      "grad_norm": 0.8683971005840104,
+      "learning_rate": 3.3288749867921953e-06,
+      "loss": 0.3896,
+      "step": 1376
+    },
+    {
+      "epoch": 0.411044776119403,
+      "grad_norm": 0.8853597587848415,
+      "learning_rate": 3.3265939773187026e-06,
+      "loss": 0.31,
+      "step": 1377
+    },
+    {
+      "epoch": 0.41134328358208955,
+      "grad_norm": 0.8397333555979059,
+      "learning_rate": 3.3243121949999906e-06,
+      "loss": 0.3478,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4116417910447761,
+      "grad_norm": 0.7964931329993589,
+      "learning_rate": 3.3220296419694686e-06,
+      "loss": 0.2821,
+      "step": 1379
+    },
+    {
+      "epoch": 0.41194029850746267,
+      "grad_norm": 0.8116358539123523,
+      "learning_rate": 3.319746320361268e-06,
+      "loss": 0.3519,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4122388059701493,
+      "grad_norm": 1.417004497957898,
+      "learning_rate": 3.3174622323102396e-06,
+      "loss": 0.3152,
+      "step": 1381
+    },
+    {
+      "epoch": 0.41253731343283584,
+      "grad_norm": 0.7731890086167517,
+      "learning_rate": 3.3151773799519492e-06,
+      "loss": 0.3873,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4128358208955224,
+      "grad_norm": 0.7984276296697358,
+      "learning_rate": 3.3128917654226794e-06,
+      "loss": 0.3897,
+      "step": 1383
+    },
+    {
+      "epoch": 0.41313432835820896,
+      "grad_norm": 0.8349251849409903,
+      "learning_rate": 3.310605390859422e-06,
+      "loss": 0.3516,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4134328358208955,
+      "grad_norm": 0.8765686889024715,
+      "learning_rate": 3.3083182583998835e-06,
+      "loss": 0.3757,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4137313432835821,
+      "grad_norm": 0.9054669051528854,
+      "learning_rate": 3.3060303701824763e-06,
+      "loss": 0.3599,
+      "step": 1386
+    },
+    {
+      "epoch": 0.41402985074626864,
+      "grad_norm": 0.7717625673330878,
+      "learning_rate": 3.303741728346319e-06,
+      "loss": 0.3221,
+      "step": 1387
+    },
+    {
+      "epoch": 0.41432835820895525,
+      "grad_norm": 0.7734383019569172,
+      "learning_rate": 3.301452335031238e-06,
+      "loss": 0.3417,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4146268656716418,
+      "grad_norm": 0.9879867361611887,
+      "learning_rate": 3.299162192377759e-06,
+      "loss": 0.4098,
+      "step": 1389
+    },
+    {
+      "epoch": 0.41492537313432837,
+      "grad_norm": 0.8150464123862557,
+      "learning_rate": 3.2968713025271095e-06,
+      "loss": 0.3322,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4152238805970149,
+      "grad_norm": 0.8573334079386694,
+      "learning_rate": 3.2945796676212155e-06,
+      "loss": 0.2965,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4155223880597015,
+      "grad_norm": 0.9215312294103535,
+      "learning_rate": 3.2922872898027007e-06,
+      "loss": 0.3696,
+      "step": 1392
+    },
+    {
+      "epoch": 0.41582089552238805,
+      "grad_norm": 0.8431748611402596,
+      "learning_rate": 3.289994171214882e-06,
+      "loss": 0.3596,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4161194029850746,
+      "grad_norm": 0.9100190938713786,
+      "learning_rate": 3.287700314001769e-06,
+      "loss": 0.3549,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4164179104477612,
+      "grad_norm": 0.7828037598477029,
+      "learning_rate": 3.2854057203080624e-06,
+      "loss": 0.3741,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4167164179104478,
+      "grad_norm": 0.8491175801452148,
+      "learning_rate": 3.283110392279152e-06,
+      "loss": 0.3451,
+      "step": 1396
+    },
+    {
+      "epoch": 0.41701492537313434,
+      "grad_norm": 0.8220914334757844,
+      "learning_rate": 3.2808143320611137e-06,
+      "loss": 0.349,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4173134328358209,
+      "grad_norm": 0.7625729166024875,
+      "learning_rate": 3.2785175418007066e-06,
+      "loss": 0.3334,
+      "step": 1398
+    },
+    {
+      "epoch": 0.41761194029850746,
+      "grad_norm": 0.7835308374457285,
+      "learning_rate": 3.276220023645374e-06,
+      "loss": 0.3221,
+      "step": 1399
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 0.8199114774599374,
+      "learning_rate": 3.2739217797432405e-06,
+      "loss": 0.3238,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4182089552238806,
+      "grad_norm": 0.7932003246008847,
+      "learning_rate": 3.2716228122431072e-06,
+      "loss": 0.3066,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4185074626865672,
+      "grad_norm": 0.8150947385287399,
+      "learning_rate": 3.2693231232944527e-06,
+      "loss": 0.3184,
+      "step": 1402
+    },
+    {
+      "epoch": 0.41880597014925375,
+      "grad_norm": 0.9738953531287554,
+      "learning_rate": 3.2670227150474298e-06,
+      "loss": 0.3689,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4191044776119403,
+      "grad_norm": 0.9195470631640194,
+      "learning_rate": 3.2647215896528643e-06,
+      "loss": 0.3668,
+      "step": 1404
+    },
+    {
+      "epoch": 0.41940298507462687,
+      "grad_norm": 0.8600925044550167,
+      "learning_rate": 3.262419749262254e-06,
+      "loss": 0.3313,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4197014925373134,
+      "grad_norm": 0.7587232892677159,
+      "learning_rate": 3.260117196027761e-06,
+      "loss": 0.3819,
+      "step": 1406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.822014542644426,
+      "learning_rate": 3.2578139321022175e-06,
+      "loss": 0.335,
+      "step": 1407
+    },
+    {
+      "epoch": 0.42029850746268654,
+      "grad_norm": 0.9420250400800545,
+      "learning_rate": 3.25550995963912e-06,
+      "loss": 0.3181,
+      "step": 1408
+    },
+    {
+      "epoch": 0.42059701492537316,
+      "grad_norm": 0.7516372540343567,
+      "learning_rate": 3.253205280792625e-06,
+      "loss": 0.2851,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4208955223880597,
+      "grad_norm": 0.7948134058388104,
+      "learning_rate": 3.250899897717552e-06,
+      "loss": 0.3454,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4211940298507463,
+      "grad_norm": 0.7240159536362353,
+      "learning_rate": 3.248593812569379e-06,
+      "loss": 0.3136,
+      "step": 1411
+    },
+    {
+      "epoch": 0.42149253731343284,
+      "grad_norm": 0.759842065598249,
+      "learning_rate": 3.246287027504237e-06,
+      "loss": 0.3287,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4217910447761194,
+      "grad_norm": 0.790615746686793,
+      "learning_rate": 3.2439795446789152e-06,
+      "loss": 0.3589,
+      "step": 1413
+    },
+    {
+      "epoch": 0.42208955223880595,
+      "grad_norm": 0.8340970220469927,
+      "learning_rate": 3.241671366250854e-06,
+      "loss": 0.3562,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4223880597014925,
+      "grad_norm": 0.8939868975796444,
+      "learning_rate": 3.2393624943781426e-06,
+      "loss": 0.3438,
+      "step": 1415
+    },
+    {
+      "epoch": 0.42268656716417913,
+      "grad_norm": 0.8450509789557636,
+      "learning_rate": 3.2370529312195225e-06,
+      "loss": 0.3233,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4229850746268657,
+      "grad_norm": 0.8150254784866996,
+      "learning_rate": 3.2347426789343766e-06,
+      "loss": 0.3523,
+      "step": 1417
+    },
+    {
+      "epoch": 0.42328358208955225,
+      "grad_norm": 1.0717236306406788,
+      "learning_rate": 3.2324317396827355e-06,
+      "loss": 0.3495,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4235820895522388,
+      "grad_norm": 0.7615376345185246,
+      "learning_rate": 3.2301201156252704e-06,
+      "loss": 0.3615,
+      "step": 1419
+    },
+    {
+      "epoch": 0.42388059701492536,
+      "grad_norm": 0.8626725672132948,
+      "learning_rate": 3.2278078089232945e-06,
+      "loss": 0.359,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4241791044776119,
+      "grad_norm": 0.7597752656447071,
+      "learning_rate": 3.2254948217387576e-06,
+      "loss": 0.3437,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4244776119402985,
+      "grad_norm": 0.8020767792341673,
+      "learning_rate": 3.223181156234246e-06,
+      "loss": 0.3578,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4247761194029851,
+      "grad_norm": 0.7733888823557834,
+      "learning_rate": 3.2208668145729806e-06,
+      "loss": 0.3439,
+      "step": 1423
+    },
+    {
+      "epoch": 0.42507462686567166,
+      "grad_norm": 0.8786679302972779,
+      "learning_rate": 3.2185517989188154e-06,
+      "loss": 0.3349,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4253731343283582,
+      "grad_norm": 0.7638976687742035,
+      "learning_rate": 3.216236111436233e-06,
+      "loss": 0.3331,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4256716417910448,
+      "grad_norm": 0.7840962222569484,
+      "learning_rate": 3.213919754290343e-06,
+      "loss": 0.354,
+      "step": 1426
+    },
+    {
+      "epoch": 0.42597014925373133,
+      "grad_norm": 0.8613687089983829,
+      "learning_rate": 3.2116027296468866e-06,
+      "loss": 0.3771,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4262686567164179,
+      "grad_norm": 0.7666875547356334,
+      "learning_rate": 3.2092850396722227e-06,
+      "loss": 0.3428,
+      "step": 1428
+    },
+    {
+      "epoch": 0.42656716417910445,
+      "grad_norm": 0.8148557834884871,
+      "learning_rate": 3.2069666865333356e-06,
+      "loss": 0.3541,
+      "step": 1429
+    },
+    {
+      "epoch": 0.42686567164179107,
+      "grad_norm": 0.7914353848681119,
+      "learning_rate": 3.204647672397829e-06,
+      "loss": 0.3338,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4271641791044776,
+      "grad_norm": 0.8843259464989137,
+      "learning_rate": 3.2023279994339242e-06,
+      "loss": 0.3511,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4274626865671642,
+      "grad_norm": 0.788441005552561,
+      "learning_rate": 3.2000076698104585e-06,
+      "loss": 0.359,
+      "step": 1432
+    },
+    {
+      "epoch": 0.42776119402985074,
+      "grad_norm": 0.8244525477083475,
+      "learning_rate": 3.197686685696885e-06,
+      "loss": 0.3729,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4280597014925373,
+      "grad_norm": 0.8468029355697649,
+      "learning_rate": 3.1953650492632664e-06,
+      "loss": 0.3583,
+      "step": 1434
+    },
+    {
+      "epoch": 0.42835820895522386,
+      "grad_norm": 0.7871095643951108,
+      "learning_rate": 3.193042762680277e-06,
+      "loss": 0.3595,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4286567164179104,
+      "grad_norm": 0.8360573541165367,
+      "learning_rate": 3.1907198281191963e-06,
+      "loss": 0.3374,
+      "step": 1436
+    },
+    {
+      "epoch": 0.42895522388059704,
+      "grad_norm": 0.8275698735469698,
+      "learning_rate": 3.1883962477519136e-06,
+      "loss": 0.3247,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4292537313432836,
+      "grad_norm": 0.7431117381085794,
+      "learning_rate": 3.1860720237509186e-06,
+      "loss": 0.3186,
+      "step": 1438
+    },
+    {
+      "epoch": 0.42955223880597015,
+      "grad_norm": 0.8236400271008097,
+      "learning_rate": 3.1837471582893044e-06,
+      "loss": 0.4015,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4298507462686567,
+      "grad_norm": 0.7284531636328365,
+      "learning_rate": 3.181421653540764e-06,
+      "loss": 0.3023,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4301492537313433,
+      "grad_norm": 0.8203967087943976,
+      "learning_rate": 3.1790955116795865e-06,
+      "loss": 0.3109,
+      "step": 1441
+    },
+    {
+      "epoch": 0.43044776119402983,
+      "grad_norm": 0.8560428452075539,
+      "learning_rate": 3.176768734880658e-06,
+      "loss": 0.3423,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4307462686567164,
+      "grad_norm": 0.9450551948814531,
+      "learning_rate": 3.17444132531946e-06,
+      "loss": 0.3602,
+      "step": 1443
+    },
+    {
+      "epoch": 0.431044776119403,
+      "grad_norm": 0.7426532649143752,
+      "learning_rate": 3.1721132851720615e-06,
+      "loss": 0.3561,
+      "step": 1444
+    },
+    {
+      "epoch": 0.43134328358208956,
+      "grad_norm": 0.8374537611379493,
+      "learning_rate": 3.169784616615125e-06,
+      "loss": 0.3457,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4316417910447761,
+      "grad_norm": 0.7637827169896023,
+      "learning_rate": 3.1674553218258976e-06,
+      "loss": 0.2929,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4319402985074627,
+      "grad_norm": 0.8669603581918557,
+      "learning_rate": 3.1651254029822126e-06,
+      "loss": 0.3499,
+      "step": 1447
+    },
+    {
+      "epoch": 0.43223880597014924,
+      "grad_norm": 0.8055692147761853,
+      "learning_rate": 3.1627948622624894e-06,
+      "loss": 0.3469,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4325373134328358,
+      "grad_norm": 1.081434396687064,
+      "learning_rate": 3.160463701845725e-06,
+      "loss": 0.4056,
+      "step": 1449
+    },
+    {
+      "epoch": 0.43283582089552236,
+      "grad_norm": 0.866468982049259,
+      "learning_rate": 3.1581319239114983e-06,
+      "loss": 0.3664,
+      "step": 1450
+    },
+    {
+      "epoch": 0.433134328358209,
+      "grad_norm": 0.7980338117580451,
+      "learning_rate": 3.1557995306399657e-06,
+      "loss": 0.3377,
+      "step": 1451
+    },
+    {
+      "epoch": 0.43343283582089553,
+      "grad_norm": 0.9044155994386817,
+      "learning_rate": 3.1534665242118557e-06,
+      "loss": 0.3559,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4337313432835821,
+      "grad_norm": 0.8274492131309824,
+      "learning_rate": 3.151132906808474e-06,
+      "loss": 0.2981,
+      "step": 1453
+    },
+    {
+      "epoch": 0.43402985074626865,
+      "grad_norm": 1.1457765974493999,
+      "learning_rate": 3.1487986806116964e-06,
+      "loss": 0.3406,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4343283582089552,
+      "grad_norm": 0.7341840039215745,
+      "learning_rate": 3.1464638478039665e-06,
+      "loss": 0.3597,
+      "step": 1455
+    },
+    {
+      "epoch": 0.43462686567164177,
+      "grad_norm": 0.8382632632518243,
+      "learning_rate": 3.1441284105682973e-06,
+      "loss": 0.3517,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4349253731343284,
+      "grad_norm": 0.8205128080490478,
+      "learning_rate": 3.1417923710882643e-06,
+      "loss": 0.3274,
+      "step": 1457
+    },
+    {
+      "epoch": 0.43522388059701494,
+      "grad_norm": 0.8607995458918247,
+      "learning_rate": 3.1394557315480077e-06,
+      "loss": 0.3733,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4355223880597015,
+      "grad_norm": 0.9177467111453076,
+      "learning_rate": 3.13711849413223e-06,
+      "loss": 0.3382,
+      "step": 1459
+    },
+    {
+      "epoch": 0.43582089552238806,
+      "grad_norm": 0.8007964740073961,
+      "learning_rate": 3.1347806610261886e-06,
+      "loss": 0.3501,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4361194029850746,
+      "grad_norm": 0.8794813487482264,
+      "learning_rate": 3.1324422344157026e-06,
+      "loss": 0.3731,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4364179104477612,
+      "grad_norm": 0.9082175046205051,
+      "learning_rate": 3.1301032164871436e-06,
+      "loss": 0.4047,
+      "step": 1462
+    },
+    {
+      "epoch": 0.43671641791044774,
+      "grad_norm": 0.9375256531298286,
+      "learning_rate": 3.1277636094274357e-06,
+      "loss": 0.3549,
+      "step": 1463
+    },
+    {
+      "epoch": 0.43701492537313436,
+      "grad_norm": 0.7978433496058229,
+      "learning_rate": 3.1254234154240544e-06,
+      "loss": 0.3325,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4373134328358209,
+      "grad_norm": 0.8332407184088709,
+      "learning_rate": 3.1230826366650245e-06,
+      "loss": 0.347,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4376119402985075,
+      "grad_norm": 0.7369892800300631,
+      "learning_rate": 3.1207412753389173e-06,
+      "loss": 0.3096,
+      "step": 1466
+    },
+    {
+      "epoch": 0.43791044776119403,
+      "grad_norm": 0.912860064618065,
+      "learning_rate": 3.118399333634848e-06,
+      "loss": 0.4124,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4382089552238806,
+      "grad_norm": 0.9428428169193727,
+      "learning_rate": 3.1160568137424757e-06,
+      "loss": 0.3071,
+      "step": 1468
+    },
+    {
+      "epoch": 0.43850746268656715,
+      "grad_norm": 0.8484227296987369,
+      "learning_rate": 3.1137137178519983e-06,
+      "loss": 0.3957,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4388059701492537,
+      "grad_norm": 0.8297150076443598,
+      "learning_rate": 3.1113700481541547e-06,
+      "loss": 0.3014,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4391044776119403,
+      "grad_norm": 0.8049525013596943,
+      "learning_rate": 3.1090258068402173e-06,
+      "loss": 0.3452,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4394029850746269,
+      "grad_norm": 0.871830475989278,
+      "learning_rate": 3.1066809961019954e-06,
+      "loss": 0.3659,
+      "step": 1472
+    },
+    {
+      "epoch": 0.43970149253731344,
+      "grad_norm": 0.8080522829751793,
+      "learning_rate": 3.1043356181318313e-06,
+      "loss": 0.3318,
+      "step": 1473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8144887418555589,
+      "learning_rate": 3.101989675122594e-06,
+      "loss": 0.3259,
+      "step": 1474
+    },
+    {
+      "epoch": 0.44029850746268656,
+      "grad_norm": 0.8505193832128495,
+      "learning_rate": 3.099643169267685e-06,
+      "loss": 0.3377,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4405970149253731,
+      "grad_norm": 0.8155829812521912,
+      "learning_rate": 3.097296102761028e-06,
+      "loss": 0.3389,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4408955223880597,
+      "grad_norm": 0.786054888865815,
+      "learning_rate": 3.0949484777970747e-06,
+      "loss": 0.3208,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4411940298507463,
+      "grad_norm": 0.7510442793736704,
+      "learning_rate": 3.0926002965707965e-06,
+      "loss": 0.3468,
+      "step": 1478
+    },
+    {
+      "epoch": 0.44149253731343285,
+      "grad_norm": 0.760538965279755,
+      "learning_rate": 3.090251561277685e-06,
+      "loss": 0.3432,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4417910447761194,
+      "grad_norm": 0.838884239640953,
+      "learning_rate": 3.0879022741137515e-06,
+      "loss": 0.3718,
+      "step": 1480
+    },
+    {
+      "epoch": 0.44208955223880597,
+      "grad_norm": 0.7876393667924823,
+      "learning_rate": 3.085552437275522e-06,
+      "loss": 0.3722,
+      "step": 1481
+    },
+    {
+      "epoch": 0.44238805970149253,
+      "grad_norm": 0.8465102759230003,
+      "learning_rate": 3.0832020529600367e-06,
+      "loss": 0.3485,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4426865671641791,
+      "grad_norm": 0.8627284225172639,
+      "learning_rate": 3.0808511233648466e-06,
+      "loss": 0.3689,
+      "step": 1483
+    },
+    {
+      "epoch": 0.44298507462686565,
+      "grad_norm": 0.7790417749973847,
+      "learning_rate": 3.0784996506880157e-06,
+      "loss": 0.313,
+      "step": 1484
+    },
+    {
+      "epoch": 0.44328358208955226,
+      "grad_norm": 0.7197974569889182,
+      "learning_rate": 3.076147637128111e-06,
+      "loss": 0.3109,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4435820895522388,
+      "grad_norm": 0.9066043015509634,
+      "learning_rate": 3.0737950848842097e-06,
+      "loss": 0.33,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4438805970149254,
+      "grad_norm": 0.7323591878235486,
+      "learning_rate": 3.0714419961558907e-06,
+      "loss": 0.3428,
+      "step": 1487
+    },
+    {
+      "epoch": 0.44417910447761194,
+      "grad_norm": 0.816034591616698,
+      "learning_rate": 3.069088373143234e-06,
+      "loss": 0.3388,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4444776119402985,
+      "grad_norm": 0.8142476229964963,
+      "learning_rate": 3.06673421804682e-06,
+      "loss": 0.29,
+      "step": 1489
+    },
+    {
+      "epoch": 0.44477611940298506,
+      "grad_norm": 0.7722627326209343,
+      "learning_rate": 3.064379533067726e-06,
+      "loss": 0.303,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4450746268656716,
+      "grad_norm": 0.8826032609144622,
+      "learning_rate": 3.062024320407525e-06,
+      "loss": 0.3546,
+      "step": 1491
+    },
+    {
+      "epoch": 0.44537313432835823,
+      "grad_norm": 0.7817296477576442,
+      "learning_rate": 3.059668582268285e-06,
+      "loss": 0.35,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4456716417910448,
+      "grad_norm": 0.7259562656822198,
+      "learning_rate": 3.0573123208525613e-06,
+      "loss": 0.3178,
+      "step": 1493
+    },
+    {
+      "epoch": 0.44597014925373135,
+      "grad_norm": 0.755357913464589,
+      "learning_rate": 3.0549555383634032e-06,
+      "loss": 0.354,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4462686567164179,
+      "grad_norm": 0.792525116985471,
+      "learning_rate": 3.052598237004343e-06,
+      "loss": 0.345,
+      "step": 1495
+    },
+    {
+      "epoch": 0.44656716417910447,
+      "grad_norm": 0.8285212042311892,
+      "learning_rate": 3.0502404189794012e-06,
+      "loss": 0.3618,
+      "step": 1496
+    },
+    {
+      "epoch": 0.44686567164179103,
+      "grad_norm": 0.8922327205331874,
+      "learning_rate": 3.0478820864930796e-06,
+      "loss": 0.3616,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4471641791044776,
+      "grad_norm": 0.8584872212973166,
+      "learning_rate": 3.0455232417503617e-06,
+      "loss": 0.3367,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4474626865671642,
+      "grad_norm": 0.7992271644708301,
+      "learning_rate": 3.0431638869567097e-06,
+      "loss": 0.3214,
+      "step": 1499
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.7540776769854936,
+      "learning_rate": 3.0408040243180638e-06,
+      "loss": 0.3285,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4480597014925373,
+      "grad_norm": 0.8633455449268056,
+      "learning_rate": 3.0384436560408363e-06,
+      "loss": 0.3401,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4483582089552239,
+      "grad_norm": 0.9222419115283121,
+      "learning_rate": 3.0360827843319156e-06,
+      "loss": 0.3838,
+      "step": 1502
+    },
+    {
+      "epoch": 0.44865671641791044,
+      "grad_norm": 0.8368812623261956,
+      "learning_rate": 3.033721411398659e-06,
+      "loss": 0.3591,
+      "step": 1503
+    },
+    {
+      "epoch": 0.448955223880597,
+      "grad_norm": 0.8378315376207885,
+      "learning_rate": 3.0313595394488917e-06,
+      "loss": 0.3567,
+      "step": 1504
+    },
+    {
+      "epoch": 0.44925373134328356,
+      "grad_norm": 0.7763761976990962,
+      "learning_rate": 3.0289971706909064e-06,
+      "loss": 0.3387,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4495522388059702,
+      "grad_norm": 0.7607283367215002,
+      "learning_rate": 3.026634307333462e-06,
+      "loss": 0.3541,
+      "step": 1506
+    },
+    {
+      "epoch": 0.44985074626865673,
+      "grad_norm": 0.8127172993351834,
+      "learning_rate": 3.024270951585776e-06,
+      "loss": 0.3546,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4501492537313433,
+      "grad_norm": 0.8648174995674591,
+      "learning_rate": 3.02190710565753e-06,
+      "loss": 0.3627,
+      "step": 1508
+    },
+    {
+      "epoch": 0.45044776119402985,
+      "grad_norm": 0.8480105965484397,
+      "learning_rate": 3.019542771758861e-06,
+      "loss": 0.3599,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4507462686567164,
+      "grad_norm": 0.7968525603371781,
+      "learning_rate": 3.0171779521003647e-06,
+      "loss": 0.3186,
+      "step": 1510
+    },
+    {
+      "epoch": 0.45104477611940297,
+      "grad_norm": 0.9005901290048014,
+      "learning_rate": 3.0148126488930896e-06,
+      "loss": 0.3793,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4513432835820895,
+      "grad_norm": 0.8428645379248311,
+      "learning_rate": 3.012446864348536e-06,
+      "loss": 0.3971,
+      "step": 1512
+    },
+    {
+      "epoch": 0.45164179104477614,
+      "grad_norm": 0.8480294223291627,
+      "learning_rate": 3.010080600678656e-06,
+      "loss": 0.3705,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4519402985074627,
+      "grad_norm": 0.8472349283442467,
+      "learning_rate": 3.0077138600958468e-06,
+      "loss": 0.3656,
+      "step": 1514
+    },
+    {
+      "epoch": 0.45223880597014926,
+      "grad_norm": 0.9103614344387116,
+      "learning_rate": 3.0053466448129535e-06,
+      "loss": 0.3812,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4525373134328358,
+      "grad_norm": 0.8231552945682205,
+      "learning_rate": 3.0029789570432665e-06,
+      "loss": 0.348,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4528358208955224,
+      "grad_norm": 0.9675315430191836,
+      "learning_rate": 3.000610799000514e-06,
+      "loss": 0.3675,
+      "step": 1517
+    },
+    {
+      "epoch": 0.45313432835820894,
+      "grad_norm": 0.8070529408063196,
+      "learning_rate": 2.9982421728988663e-06,
+      "loss": 0.341,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4534328358208955,
+      "grad_norm": 0.8536363490650818,
+      "learning_rate": 2.9958730809529326e-06,
+      "loss": 0.3794,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4537313432835821,
+      "grad_norm": 0.8122339636569735,
+      "learning_rate": 2.9935035253777544e-06,
+      "loss": 0.3582,
+      "step": 1520
+    },
+    {
+      "epoch": 0.45402985074626867,
+      "grad_norm": 0.7993919608319511,
+      "learning_rate": 2.9911335083888093e-06,
+      "loss": 0.344,
+      "step": 1521
+    },
+    {
+      "epoch": 0.45432835820895523,
+      "grad_norm": 0.8228888153873775,
+      "learning_rate": 2.988763032202006e-06,
+      "loss": 0.3235,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4546268656716418,
+      "grad_norm": 0.8201968407407756,
+      "learning_rate": 2.9863920990336803e-06,
+      "loss": 0.3707,
+      "step": 1523
+    },
+    {
+      "epoch": 0.45492537313432835,
+      "grad_norm": 0.8436534725037953,
+      "learning_rate": 2.9840207111005987e-06,
+      "loss": 0.3886,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4552238805970149,
+      "grad_norm": 0.7968744092690195,
+      "learning_rate": 2.9816488706199498e-06,
+      "loss": 0.3498,
+      "step": 1525
+    },
+    {
+      "epoch": 0.45552238805970147,
+      "grad_norm": 0.8197065867829275,
+      "learning_rate": 2.9792765798093466e-06,
+      "loss": 0.3676,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4558208955223881,
+      "grad_norm": 0.8141683013758337,
+      "learning_rate": 2.9769038408868246e-06,
+      "loss": 0.3175,
+      "step": 1527
+    },
+    {
+      "epoch": 0.45611940298507464,
+      "grad_norm": 0.8682309359789031,
+      "learning_rate": 2.9745306560708343e-06,
+      "loss": 0.3582,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4564179104477612,
+      "grad_norm": 0.8430286979266656,
+      "learning_rate": 2.9721570275802487e-06,
+      "loss": 0.3643,
+      "step": 1529
+    },
+    {
+      "epoch": 0.45671641791044776,
+      "grad_norm": 0.8327189426692697,
+      "learning_rate": 2.969782957634351e-06,
+      "loss": 0.3869,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4570149253731343,
+      "grad_norm": 0.8303352738239019,
+      "learning_rate": 2.967408448452838e-06,
+      "loss": 0.3396,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4573134328358209,
+      "grad_norm": 0.8521303530894394,
+      "learning_rate": 2.9650335022558202e-06,
+      "loss": 0.3593,
+      "step": 1532
+    },
+    {
+      "epoch": 0.45761194029850744,
+      "grad_norm": 0.7973923282393381,
+      "learning_rate": 2.962658121263812e-06,
+      "loss": 0.3826,
+      "step": 1533
+    },
+    {
+      "epoch": 0.45791044776119405,
+      "grad_norm": 0.8307181709707767,
+      "learning_rate": 2.9602823076977376e-06,
+      "loss": 0.385,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4582089552238806,
+      "grad_norm": 0.7681378144807851,
+      "learning_rate": 2.9579060637789257e-06,
+      "loss": 0.3345,
+      "step": 1535
+    },
+    {
+      "epoch": 0.45850746268656717,
+      "grad_norm": 0.8862805383356972,
+      "learning_rate": 2.955529391729105e-06,
+      "loss": 0.3342,
+      "step": 1536
+    },
+    {
+      "epoch": 0.45880597014925373,
+      "grad_norm": 0.7914026956159501,
+      "learning_rate": 2.9531522937704065e-06,
+      "loss": 0.3161,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4591044776119403,
+      "grad_norm": 0.815340219137627,
+      "learning_rate": 2.9507747721253598e-06,
+      "loss": 0.3161,
+      "step": 1538
+    },
+    {
+      "epoch": 0.45940298507462685,
+      "grad_norm": 0.8486280883954298,
+      "learning_rate": 2.948396829016888e-06,
+      "loss": 0.3425,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4597014925373134,
+      "grad_norm": 0.8507231569423965,
+      "learning_rate": 2.9460184666683112e-06,
+      "loss": 0.3601,
+      "step": 1540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8316847618482142,
+      "learning_rate": 2.9436396873033396e-06,
+      "loss": 0.3566,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4602985074626866,
+      "grad_norm": 0.7838154911787112,
+      "learning_rate": 2.941260493146074e-06,
+      "loss": 0.3075,
+      "step": 1542
+    },
+    {
+      "epoch": 0.46059701492537314,
+      "grad_norm": 0.8919729652488453,
+      "learning_rate": 2.938880886421004e-06,
+      "loss": 0.3628,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4608955223880597,
+      "grad_norm": 0.8463176490111552,
+      "learning_rate": 2.9365008693530017e-06,
+      "loss": 0.3916,
+      "step": 1544
+    },
+    {
+      "epoch": 0.46119402985074626,
+      "grad_norm": 0.8415000303394304,
+      "learning_rate": 2.9341204441673267e-06,
+      "loss": 0.3378,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4614925373134328,
+      "grad_norm": 0.9092886626244101,
+      "learning_rate": 2.931739613089618e-06,
+      "loss": 0.3549,
+      "step": 1546
+    },
+    {
+      "epoch": 0.46179104477611943,
+      "grad_norm": 0.8845564844751359,
+      "learning_rate": 2.929358378345894e-06,
+      "loss": 0.3646,
+      "step": 1547
+    },
+    {
+      "epoch": 0.462089552238806,
+      "grad_norm": 0.7119788198918313,
+      "learning_rate": 2.9269767421625535e-06,
+      "loss": 0.2987,
+      "step": 1548
+    },
+    {
+      "epoch": 0.46238805970149255,
+      "grad_norm": 0.812817673515338,
+      "learning_rate": 2.9245947067663653e-06,
+      "loss": 0.3138,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4626865671641791,
+      "grad_norm": 0.8782182840009076,
+      "learning_rate": 2.922212274384476e-06,
+      "loss": 0.3709,
+      "step": 1550
+    },
+    {
+      "epoch": 0.46298507462686567,
+      "grad_norm": 0.7972696289980897,
+      "learning_rate": 2.9198294472444022e-06,
+      "loss": 0.3182,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4632835820895522,
+      "grad_norm": 0.8368020184190745,
+      "learning_rate": 2.9174462275740286e-06,
+      "loss": 0.4036,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4635820895522388,
+      "grad_norm": 0.8084845969111005,
+      "learning_rate": 2.9150626176016065e-06,
+      "loss": 0.3457,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4638805970149254,
+      "grad_norm": 0.9420253305451717,
+      "learning_rate": 2.9126786195557554e-06,
+      "loss": 0.3444,
+      "step": 1554
+    },
+    {
+      "epoch": 0.46417910447761196,
+      "grad_norm": 0.8848388474337436,
+      "learning_rate": 2.910294235665453e-06,
+      "loss": 0.3494,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4644776119402985,
+      "grad_norm": 0.8223510020899589,
+      "learning_rate": 2.9079094681600416e-06,
+      "loss": 0.3289,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4647761194029851,
+      "grad_norm": 0.7905450158929105,
+      "learning_rate": 2.9055243192692207e-06,
+      "loss": 0.3338,
+      "step": 1557
+    },
+    {
+      "epoch": 0.46507462686567164,
+      "grad_norm": 0.8490757892161332,
+      "learning_rate": 2.9031387912230454e-06,
+      "loss": 0.378,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4653731343283582,
+      "grad_norm": 0.7906965441882455,
+      "learning_rate": 2.900752886251927e-06,
+      "loss": 0.3149,
+      "step": 1559
+    },
+    {
+      "epoch": 0.46567164179104475,
+      "grad_norm": 0.7577279466640751,
+      "learning_rate": 2.898366606586628e-06,
+      "loss": 0.2942,
+      "step": 1560
+    },
+    {
+      "epoch": 0.46597014925373137,
+      "grad_norm": 0.8589842568909583,
+      "learning_rate": 2.895979954458263e-06,
+      "loss": 0.366,
+      "step": 1561
+    },
+    {
+      "epoch": 0.46626865671641793,
+      "grad_norm": 0.756467710812116,
+      "learning_rate": 2.893592932098292e-06,
+      "loss": 0.3039,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4665671641791045,
+      "grad_norm": 0.749749204455234,
+      "learning_rate": 2.891205541738523e-06,
+      "loss": 0.3461,
+      "step": 1563
+    },
+    {
+      "epoch": 0.46686567164179105,
+      "grad_norm": 0.7995654210279144,
+      "learning_rate": 2.8888177856111082e-06,
+      "loss": 0.2976,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4671641791044776,
+      "grad_norm": 0.9693221870057586,
+      "learning_rate": 2.8864296659485413e-06,
+      "loss": 0.3557,
+      "step": 1565
+    },
+    {
+      "epoch": 0.46746268656716417,
+      "grad_norm": 0.7943690346767394,
+      "learning_rate": 2.8840411849836565e-06,
+      "loss": 0.3598,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4677611940298507,
+      "grad_norm": 0.8435450993312467,
+      "learning_rate": 2.881652344949625e-06,
+      "loss": 0.4217,
+      "step": 1567
+    },
+    {
+      "epoch": 0.46805970149253734,
+      "grad_norm": 0.7333154530834609,
+      "learning_rate": 2.8792631480799526e-06,
+      "loss": 0.2883,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4683582089552239,
+      "grad_norm": 0.7562359363934774,
+      "learning_rate": 2.8768735966084817e-06,
+      "loss": 0.3055,
+      "step": 1569
+    },
+    {
+      "epoch": 0.46865671641791046,
+      "grad_norm": 0.7398426343466973,
+      "learning_rate": 2.874483692769385e-06,
+      "loss": 0.3478,
+      "step": 1570
+    },
+    {
+      "epoch": 0.468955223880597,
+      "grad_norm": 0.8230158538149063,
+      "learning_rate": 2.8720934387971627e-06,
+      "loss": 0.3278,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4692537313432836,
+      "grad_norm": 1.3545109216720357,
+      "learning_rate": 2.869702836926645e-06,
+      "loss": 0.2997,
+      "step": 1572
+    },
+    {
+      "epoch": 0.46955223880597013,
+      "grad_norm": 0.926104402556891,
+      "learning_rate": 2.8673118893929876e-06,
+      "loss": 0.3141,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4698507462686567,
+      "grad_norm": 0.7738598668200026,
+      "learning_rate": 2.864920598431665e-06,
+      "loss": 0.3212,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4701492537313433,
+      "grad_norm": 0.7844426674437163,
+      "learning_rate": 2.862528966278479e-06,
+      "loss": 0.3281,
+      "step": 1575
+    },
+    {
+      "epoch": 0.47044776119402987,
+      "grad_norm": 0.8678599800587856,
+      "learning_rate": 2.8601369951695463e-06,
+      "loss": 0.3721,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4707462686567164,
+      "grad_norm": 0.863778592486986,
+      "learning_rate": 2.8577446873413007e-06,
+      "loss": 0.3629,
+      "step": 1577
+    },
+    {
+      "epoch": 0.471044776119403,
+      "grad_norm": 0.9033383927815183,
+      "learning_rate": 2.855352045030493e-06,
+      "loss": 0.3508,
+      "step": 1578
+    },
+    {
+      "epoch": 0.47134328358208955,
+      "grad_norm": 0.812991807684829,
+      "learning_rate": 2.8529590704741843e-06,
+      "loss": 0.384,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4716417910447761,
+      "grad_norm": 0.8187956043141631,
+      "learning_rate": 2.8505657659097486e-06,
+      "loss": 0.3707,
+      "step": 1580
+    },
+    {
+      "epoch": 0.47194029850746266,
+      "grad_norm": 0.9346578308192153,
+      "learning_rate": 2.8481721335748674e-06,
+      "loss": 0.4094,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4722388059701493,
+      "grad_norm": 0.7880484291645268,
+      "learning_rate": 2.845778175707527e-06,
+      "loss": 0.3299,
+      "step": 1582
+    },
+    {
+      "epoch": 0.47253731343283584,
+      "grad_norm": 0.811099440312082,
+      "learning_rate": 2.8433838945460207e-06,
+      "loss": 0.33,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4728358208955224,
+      "grad_norm": 0.7026137359784628,
+      "learning_rate": 2.8409892923289432e-06,
+      "loss": 0.3127,
+      "step": 1584
+    },
+    {
+      "epoch": 0.47313432835820896,
+      "grad_norm": 0.8034762357536271,
+      "learning_rate": 2.838594371295189e-06,
+      "loss": 0.3265,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4734328358208955,
+      "grad_norm": 0.8485262509519692,
+      "learning_rate": 2.8361991336839513e-06,
+      "loss": 0.381,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4737313432835821,
+      "grad_norm": 0.9427056505224326,
+      "learning_rate": 2.833803581734718e-06,
+      "loss": 0.4068,
+      "step": 1587
+    },
+    {
+      "epoch": 0.47402985074626863,
+      "grad_norm": 0.8021996861465295,
+      "learning_rate": 2.8314077176872724e-06,
+      "loss": 0.335,
+      "step": 1588
+    },
+    {
+      "epoch": 0.47432835820895525,
+      "grad_norm": 0.7919679011390528,
+      "learning_rate": 2.8290115437816894e-06,
+      "loss": 0.3561,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4746268656716418,
+      "grad_norm": 0.7810376577747256,
+      "learning_rate": 2.8266150622583315e-06,
+      "loss": 0.3501,
+      "step": 1590
+    },
+    {
+      "epoch": 0.47492537313432837,
+      "grad_norm": 0.9272988287616046,
+      "learning_rate": 2.8242182753578523e-06,
+      "loss": 0.3466,
+      "step": 1591
+    },
+    {
+      "epoch": 0.4752238805970149,
+      "grad_norm": 0.8467896274245239,
+      "learning_rate": 2.8218211853211893e-06,
+      "loss": 0.357,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4755223880597015,
+      "grad_norm": 0.9156541553176574,
+      "learning_rate": 2.819423794389561e-06,
+      "loss": 0.3529,
+      "step": 1593
+    },
+    {
+      "epoch": 0.47582089552238804,
+      "grad_norm": 0.7927076329626589,
+      "learning_rate": 2.817026104804471e-06,
+      "loss": 0.3155,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4761194029850746,
+      "grad_norm": 0.767743906513666,
+      "learning_rate": 2.8146281188077017e-06,
+      "loss": 0.3155,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4764179104477612,
+      "grad_norm": 0.8361708178822149,
+      "learning_rate": 2.8122298386413094e-06,
+      "loss": 0.361,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4767164179104478,
+      "grad_norm": 0.8455590406670277,
+      "learning_rate": 2.8098312665476283e-06,
+      "loss": 0.3462,
+      "step": 1597
+    },
+    {
+      "epoch": 0.47701492537313434,
+      "grad_norm": 0.8438149986929196,
+      "learning_rate": 2.8074324047692662e-06,
+      "loss": 0.3597,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4773134328358209,
+      "grad_norm": 0.8286441056776019,
+      "learning_rate": 2.8050332555490987e-06,
+      "loss": 0.3648,
+      "step": 1599
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.7719571836786463,
+      "learning_rate": 2.8026338211302735e-06,
+      "loss": 0.3333,
+      "step": 1600
+    },
+    {
+      "epoch": 0.477910447761194,
+      "grad_norm": 0.8233026839737,
+      "learning_rate": 2.800234103756201e-06,
+      "loss": 0.3337,
+      "step": 1601
+    },
+    {
+      "epoch": 0.47820895522388057,
+      "grad_norm": 0.7688811413487633,
+      "learning_rate": 2.7978341056705592e-06,
+      "loss": 0.3075,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4785074626865672,
+      "grad_norm": 0.9247126459680931,
+      "learning_rate": 2.7954338291172892e-06,
+      "loss": 0.3842,
+      "step": 1603
+    },
+    {
+      "epoch": 0.47880597014925375,
+      "grad_norm": 0.8669810824469362,
+      "learning_rate": 2.79303327634059e-06,
+      "loss": 0.3619,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4791044776119403,
+      "grad_norm": 0.824675144757377,
+      "learning_rate": 2.7906324495849206e-06,
+      "loss": 0.3491,
+      "step": 1605
+    },
+    {
+      "epoch": 0.47940298507462686,
+      "grad_norm": 0.7606474249944587,
+      "learning_rate": 2.788231351094995e-06,
+      "loss": 0.3409,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4797014925373134,
+      "grad_norm": 0.8588453625011868,
+      "learning_rate": 2.785829983115781e-06,
+      "loss": 0.3826,
+      "step": 1607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8069935126683544,
+      "learning_rate": 2.7834283478925007e-06,
+      "loss": 0.3424,
+      "step": 1608
+    },
+    {
+      "epoch": 0.48029850746268654,
+      "grad_norm": 0.8572397041684174,
+      "learning_rate": 2.7810264476706227e-06,
+      "loss": 0.3258,
+      "step": 1609
+    },
+    {
+      "epoch": 0.48059701492537316,
+      "grad_norm": 0.850023545997115,
+      "learning_rate": 2.778624284695867e-06,
+      "loss": 0.3235,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4808955223880597,
+      "grad_norm": 0.8475591867049286,
+      "learning_rate": 2.7762218612141966e-06,
+      "loss": 0.3337,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4811940298507463,
+      "grad_norm": 0.8536920148847397,
+      "learning_rate": 2.7738191794718183e-06,
+      "loss": 0.307,
+      "step": 1612
+    },
+    {
+      "epoch": 0.48149253731343283,
+      "grad_norm": 0.8281648081371542,
+      "learning_rate": 2.771416241715182e-06,
+      "loss": 0.3839,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4817910447761194,
+      "grad_norm": 0.7677447755495962,
+      "learning_rate": 2.7690130501909756e-06,
+      "loss": 0.3291,
+      "step": 1614
+    },
+    {
+      "epoch": 0.48208955223880595,
+      "grad_norm": 3.3342305797562215,
+      "learning_rate": 2.766609607146124e-06,
+      "loss": 0.3939,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4823880597014925,
+      "grad_norm": 0.7437237275043507,
+      "learning_rate": 2.7642059148277894e-06,
+      "loss": 0.3558,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4826865671641791,
+      "grad_norm": 0.8563459859815404,
+      "learning_rate": 2.761801975483363e-06,
+      "loss": 0.3396,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4829850746268657,
+      "grad_norm": 0.8531096858219094,
+      "learning_rate": 2.7593977913604717e-06,
+      "loss": 0.3741,
+      "step": 1618
+    },
+    {
+      "epoch": 0.48328358208955224,
+      "grad_norm": 0.8588995102283145,
+      "learning_rate": 2.7569933647069685e-06,
+      "loss": 0.3189,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4835820895522388,
+      "grad_norm": 0.8321789763467867,
+      "learning_rate": 2.754588697770933e-06,
+      "loss": 0.3456,
+      "step": 1620
+    },
+    {
+      "epoch": 0.48388059701492536,
+      "grad_norm": 0.8853198602133011,
+      "learning_rate": 2.752183792800671e-06,
+      "loss": 0.3932,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4841791044776119,
+      "grad_norm": 0.8434334328362673,
+      "learning_rate": 2.7497786520447093e-06,
+      "loss": 0.4004,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4844776119402985,
+      "grad_norm": 0.8669235672763962,
+      "learning_rate": 2.7473732777517965e-06,
+      "loss": 0.3448,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4847761194029851,
+      "grad_norm": 0.8146513938626443,
+      "learning_rate": 2.7449676721708995e-06,
+      "loss": 0.3658,
+      "step": 1624
+    },
+    {
+      "epoch": 0.48507462686567165,
+      "grad_norm": 0.7682677410362674,
+      "learning_rate": 2.7425618375511992e-06,
+      "loss": 0.3429,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4853731343283582,
+      "grad_norm": 0.8659290979909456,
+      "learning_rate": 2.7401557761420933e-06,
+      "loss": 0.3989,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4856716417910448,
+      "grad_norm": 0.9005007791678644,
+      "learning_rate": 2.737749490193191e-06,
+      "loss": 0.4066,
+      "step": 1627
+    },
+    {
+      "epoch": 0.48597014925373133,
+      "grad_norm": 0.8269529526190118,
+      "learning_rate": 2.7353429819543104e-06,
+      "loss": 0.3104,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4862686567164179,
+      "grad_norm": 0.7655526786732026,
+      "learning_rate": 2.7329362536754777e-06,
+      "loss": 0.3489,
+      "step": 1629
+    },
+    {
+      "epoch": 0.48656716417910445,
+      "grad_norm": 0.8373568352027102,
+      "learning_rate": 2.7305293076069263e-06,
+      "loss": 0.3264,
+      "step": 1630
+    },
+    {
+      "epoch": 0.48686567164179106,
+      "grad_norm": 0.7899465343213389,
+      "learning_rate": 2.728122145999091e-06,
+      "loss": 0.3057,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4871641791044776,
+      "grad_norm": 2.475275899634775,
+      "learning_rate": 2.72571477110261e-06,
+      "loss": 0.3467,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4874626865671642,
+      "grad_norm": 0.8557171237524993,
+      "learning_rate": 2.72330718516832e-06,
+      "loss": 0.3587,
+      "step": 1633
+    },
+    {
+      "epoch": 0.48776119402985074,
+      "grad_norm": 0.7280642925743758,
+      "learning_rate": 2.7208993904472543e-06,
+      "loss": 0.3306,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4880597014925373,
+      "grad_norm": 0.7753145285378411,
+      "learning_rate": 2.7184913891906433e-06,
+      "loss": 0.3074,
+      "step": 1635
+    },
+    {
+      "epoch": 0.48835820895522386,
+      "grad_norm": 0.7395330717244848,
+      "learning_rate": 2.716083183649909e-06,
+      "loss": 0.3426,
+      "step": 1636
+    },
+    {
+      "epoch": 0.4886567164179105,
+      "grad_norm": 0.8002310765854909,
+      "learning_rate": 2.7136747760766653e-06,
+      "loss": 0.3534,
+      "step": 1637
+    },
+    {
+      "epoch": 0.48895522388059703,
+      "grad_norm": 0.8324334035465984,
+      "learning_rate": 2.7112661687227142e-06,
+      "loss": 0.3355,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4892537313432836,
+      "grad_norm": 0.7743794963426636,
+      "learning_rate": 2.708857363840045e-06,
+      "loss": 0.3329,
+      "step": 1639
+    },
+    {
+      "epoch": 0.48955223880597015,
+      "grad_norm": 0.8228579838838092,
+      "learning_rate": 2.7064483636808314e-06,
+      "loss": 0.3755,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4898507462686567,
+      "grad_norm": 0.7759076884150069,
+      "learning_rate": 2.7040391704974293e-06,
+      "loss": 0.3534,
+      "step": 1641
+    },
+    {
+      "epoch": 0.49014925373134327,
+      "grad_norm": 0.8250088563687552,
+      "learning_rate": 2.7016297865423767e-06,
+      "loss": 0.3562,
+      "step": 1642
+    },
+    {
+      "epoch": 0.49044776119402983,
+      "grad_norm": 0.8627019755137874,
+      "learning_rate": 2.699220214068389e-06,
+      "loss": 0.292,
+      "step": 1643
+    },
+    {
+      "epoch": 0.49074626865671644,
+      "grad_norm": 0.8814707303291824,
+      "learning_rate": 2.696810455328357e-06,
+      "loss": 0.3616,
+      "step": 1644
+    },
+    {
+      "epoch": 0.491044776119403,
+      "grad_norm": 0.8932767174110039,
+      "learning_rate": 2.694400512575346e-06,
+      "loss": 0.3596,
+      "step": 1645
+    },
+    {
+      "epoch": 0.49134328358208956,
+      "grad_norm": 0.8118919974953459,
+      "learning_rate": 2.6919903880625954e-06,
+      "loss": 0.3343,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4916417910447761,
+      "grad_norm": 0.8648850492388545,
+      "learning_rate": 2.6895800840435106e-06,
+      "loss": 0.335,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4919402985074627,
+      "grad_norm": 0.8549039551168018,
+      "learning_rate": 2.687169602771668e-06,
+      "loss": 0.3723,
+      "step": 1648
+    },
+    {
+      "epoch": 0.49223880597014924,
+      "grad_norm": 0.7944249187287936,
+      "learning_rate": 2.68475894650081e-06,
+      "loss": 0.3085,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 0.7927257364542748,
+      "learning_rate": 2.6823481174848405e-06,
+      "loss": 0.3133,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4928358208955224,
+      "grad_norm": 0.7886213022780711,
+      "learning_rate": 2.679937117977825e-06,
+      "loss": 0.3822,
+      "step": 1651
+    },
+    {
+      "epoch": 0.493134328358209,
+      "grad_norm": 0.8280071240535853,
+      "learning_rate": 2.6775259502339913e-06,
+      "loss": 0.3573,
+      "step": 1652
+    },
+    {
+      "epoch": 0.49343283582089553,
+      "grad_norm": 0.8418482347983869,
+      "learning_rate": 2.67511461650772e-06,
+      "loss": 0.3835,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4937313432835821,
+      "grad_norm": 0.8387035834884874,
+      "learning_rate": 2.672703119053552e-06,
+      "loss": 0.3586,
+      "step": 1654
+    },
+    {
+      "epoch": 0.49402985074626865,
+      "grad_norm": 0.7816419105725156,
+      "learning_rate": 2.670291460126177e-06,
+      "loss": 0.3339,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4943283582089552,
+      "grad_norm": 0.8001221218291323,
+      "learning_rate": 2.667879641980437e-06,
+      "loss": 0.3204,
+      "step": 1656
+    },
+    {
+      "epoch": 0.49462686567164177,
+      "grad_norm": 0.7879284043181888,
+      "learning_rate": 2.6654676668713245e-06,
+      "loss": 0.3224,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4949253731343284,
+      "grad_norm": 0.801234203959891,
+      "learning_rate": 2.6630555370539763e-06,
+      "loss": 0.3327,
+      "step": 1658
+    },
+    {
+      "epoch": 0.49522388059701494,
+      "grad_norm": 0.7997197712338734,
+      "learning_rate": 2.6606432547836757e-06,
+      "loss": 0.3375,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4955223880597015,
+      "grad_norm": 0.7623217742754592,
+      "learning_rate": 2.658230822315847e-06,
+      "loss": 0.3187,
+      "step": 1660
+    },
+    {
+      "epoch": 0.49582089552238806,
+      "grad_norm": 0.7962363516446156,
+      "learning_rate": 2.655818241906057e-06,
+      "loss": 0.3493,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4961194029850746,
+      "grad_norm": 0.9034198102424835,
+      "learning_rate": 2.653405515810009e-06,
+      "loss": 0.3753,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4964179104477612,
+      "grad_norm": 0.7925807514012592,
+      "learning_rate": 2.650992646283542e-06,
+      "loss": 0.3077,
+      "step": 1663
+    },
+    {
+      "epoch": 0.49671641791044774,
+      "grad_norm": 0.7247430409943646,
+      "learning_rate": 2.648579635582632e-06,
+      "loss": 0.3098,
+      "step": 1664
+    },
+    {
+      "epoch": 0.49701492537313435,
+      "grad_norm": 0.6769452012660345,
+      "learning_rate": 2.6461664859633844e-06,
+      "loss": 0.2895,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4973134328358209,
+      "grad_norm": 0.806049068867384,
+      "learning_rate": 2.6437531996820353e-06,
+      "loss": 0.3263,
+      "step": 1666
+    },
+    {
+      "epoch": 0.49761194029850747,
+      "grad_norm": 0.8550960252280426,
+      "learning_rate": 2.641339778994948e-06,
+      "loss": 0.37,
+      "step": 1667
+    },
+    {
+      "epoch": 0.49791044776119403,
+      "grad_norm": 0.8478170482025786,
+      "learning_rate": 2.6389262261586127e-06,
+      "loss": 0.3892,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4982089552238806,
+      "grad_norm": 0.8345120172992568,
+      "learning_rate": 2.636512543429642e-06,
+      "loss": 0.348,
+      "step": 1669
+    },
+    {
+      "epoch": 0.49850746268656715,
+      "grad_norm": 0.8028565137146465,
+      "learning_rate": 2.634098733064771e-06,
+      "loss": 0.3574,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4988059701492537,
+      "grad_norm": 0.7451403703442233,
+      "learning_rate": 2.6316847973208535e-06,
+      "loss": 0.2899,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4991044776119403,
+      "grad_norm": 0.8352856655412605,
+      "learning_rate": 2.6292707384548604e-06,
+      "loss": 0.3597,
+      "step": 1672
+    },
+    {
+      "epoch": 0.4994029850746269,
+      "grad_norm": 0.8545513078252729,
+      "learning_rate": 2.6268565587238777e-06,
+      "loss": 0.3325,
+      "step": 1673
+    },
+    {
+      "epoch": 0.49970149253731344,
+      "grad_norm": 0.8282524133497315,
+      "learning_rate": 2.6244422603851046e-06,
+      "loss": 0.3492,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7991480826878556,
+      "learning_rate": 2.622027845695851e-06,
+      "loss": 0.3506,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5002985074626866,
+      "grad_norm": 0.7582015801891304,
+      "learning_rate": 2.6196133169135368e-06,
+      "loss": 0.3293,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5005970149253731,
+      "grad_norm": 0.9612014821814638,
+      "learning_rate": 2.6171986762956856e-06,
+      "loss": 0.3996,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5008955223880597,
+      "grad_norm": 0.7898075941393438,
+      "learning_rate": 2.61478392609993e-06,
+      "loss": 0.3312,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5011940298507462,
+      "grad_norm": 0.8015043125699043,
+      "learning_rate": 2.612369068584001e-06,
+      "loss": 0.3096,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5014925373134328,
+      "grad_norm": 0.8389408382259228,
+      "learning_rate": 2.6099541060057316e-06,
+      "loss": 0.3338,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5017910447761194,
+      "grad_norm": 0.8096612317183629,
+      "learning_rate": 2.607539040623054e-06,
+      "loss": 0.3828,
+      "step": 1681
+    },
+    {
+      "epoch": 0.502089552238806,
+      "grad_norm": 0.8259878912971927,
+      "learning_rate": 2.6051238746939934e-06,
+      "loss": 0.3031,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5023880597014926,
+      "grad_norm": 0.8631597448564671,
+      "learning_rate": 2.602708610476673e-06,
+      "loss": 0.3664,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5026865671641791,
+      "grad_norm": 0.7716339492340298,
+      "learning_rate": 2.600293250229306e-06,
+      "loss": 0.338,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5029850746268657,
+      "grad_norm": 0.899796414901459,
+      "learning_rate": 2.597877796210194e-06,
+      "loss": 0.3815,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5032835820895523,
+      "grad_norm": 0.9820931245782578,
+      "learning_rate": 2.5954622506777285e-06,
+      "loss": 0.3691,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5035820895522388,
+      "grad_norm": 0.7494455392178265,
+      "learning_rate": 2.5930466158903856e-06,
+      "loss": 0.3215,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5038805970149254,
+      "grad_norm": 0.8086992100975138,
+      "learning_rate": 2.5906308941067243e-06,
+      "loss": 0.3297,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5041791044776119,
+      "grad_norm": 0.9419257958465664,
+      "learning_rate": 2.588215087585387e-06,
+      "loss": 0.3312,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5044776119402985,
+      "grad_norm": 0.7786799854356972,
+      "learning_rate": 2.5857991985850924e-06,
+      "loss": 0.3236,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5047761194029851,
+      "grad_norm": 0.8680079607603017,
+      "learning_rate": 2.583383229364639e-06,
+      "loss": 0.3484,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5050746268656716,
+      "grad_norm": 0.867654415324763,
+      "learning_rate": 2.580967182182898e-06,
+      "loss": 0.3494,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5053731343283582,
+      "grad_norm": 0.7824582583766272,
+      "learning_rate": 2.5785510592988156e-06,
+      "loss": 0.3345,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5056716417910447,
+      "grad_norm": 0.9778668777323667,
+      "learning_rate": 2.5761348629714073e-06,
+      "loss": 0.3447,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5059701492537313,
+      "grad_norm": 0.879824672451696,
+      "learning_rate": 2.5737185954597583e-06,
+      "loss": 0.3486,
+      "step": 1695
+    },
+    {
+      "epoch": 0.506268656716418,
+      "grad_norm": 0.7272504403686626,
+      "learning_rate": 2.571302259023019e-06,
+      "loss": 0.3284,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5065671641791045,
+      "grad_norm": 0.8269875685389965,
+      "learning_rate": 2.5688858559204056e-06,
+      "loss": 0.3189,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5068656716417911,
+      "grad_norm": 0.7107142314587015,
+      "learning_rate": 2.5664693884111958e-06,
+      "loss": 0.3069,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5071641791044776,
+      "grad_norm": 0.8300310961227638,
+      "learning_rate": 2.564052858754728e-06,
+      "loss": 0.2923,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5074626865671642,
+      "grad_norm": 0.8721700820774255,
+      "learning_rate": 2.561636269210399e-06,
+      "loss": 0.3377,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5077611940298508,
+      "grad_norm": 0.7550427576298773,
+      "learning_rate": 2.55921962203766e-06,
+      "loss": 0.3222,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5080597014925373,
+      "grad_norm": 0.7968625493472227,
+      "learning_rate": 2.5568029194960186e-06,
+      "loss": 0.3291,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5083582089552239,
+      "grad_norm": 0.7511883213331606,
+      "learning_rate": 2.554386163845032e-06,
+      "loss": 0.3107,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5086567164179104,
+      "grad_norm": 0.752772967275015,
+      "learning_rate": 2.551969357344308e-06,
+      "loss": 0.3579,
+      "step": 1704
+    },
+    {
+      "epoch": 0.508955223880597,
+      "grad_norm": 0.7805447167279405,
+      "learning_rate": 2.5495525022535013e-06,
+      "loss": 0.3091,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5092537313432836,
+      "grad_norm": 0.7465985414472608,
+      "learning_rate": 2.547135600832313e-06,
+      "loss": 0.3116,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5095522388059701,
+      "grad_norm": 0.7452102417332194,
+      "learning_rate": 2.544718655340486e-06,
+      "loss": 0.3141,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5098507462686567,
+      "grad_norm": 0.8594711328205182,
+      "learning_rate": 2.5423016680378076e-06,
+      "loss": 0.3254,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5101492537313432,
+      "grad_norm": 0.8724045713981662,
+      "learning_rate": 2.5398846411840998e-06,
+      "loss": 0.3306,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5104477611940299,
+      "grad_norm": 0.8250540195559437,
+      "learning_rate": 2.5374675770392247e-06,
+      "loss": 0.2953,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5107462686567165,
+      "grad_norm": 0.7579796546729729,
+      "learning_rate": 2.5350504778630795e-06,
+      "loss": 0.3514,
+      "step": 1711
+    },
+    {
+      "epoch": 0.511044776119403,
+      "grad_norm": 0.7968045617982358,
+      "learning_rate": 2.5326333459155904e-06,
+      "loss": 0.3443,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5113432835820896,
+      "grad_norm": 0.7725314533480072,
+      "learning_rate": 2.530216183456719e-06,
+      "loss": 0.3345,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5116417910447761,
+      "grad_norm": 0.8181569525639177,
+      "learning_rate": 2.527798992746453e-06,
+      "loss": 0.3061,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5119402985074627,
+      "grad_norm": 0.8292279194686958,
+      "learning_rate": 2.525381776044806e-06,
+      "loss": 0.3494,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5122388059701493,
+      "grad_norm": 0.9155665246514125,
+      "learning_rate": 2.5229645356118166e-06,
+      "loss": 0.3805,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5125373134328358,
+      "grad_norm": 0.8426243739571815,
+      "learning_rate": 2.520547273707546e-06,
+      "loss": 0.3978,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5128358208955224,
+      "grad_norm": 0.7578253079200755,
+      "learning_rate": 2.5181299925920756e-06,
+      "loss": 0.3387,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5131343283582089,
+      "grad_norm": 0.7152262315331874,
+      "learning_rate": 2.5157126945255043e-06,
+      "loss": 0.2707,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5134328358208955,
+      "grad_norm": 0.7803958098892375,
+      "learning_rate": 2.5132953817679466e-06,
+      "loss": 0.3587,
+      "step": 1720
+    },
+    {
+      "epoch": 0.513731343283582,
+      "grad_norm": 0.7380296998637139,
+      "learning_rate": 2.510878056579531e-06,
+      "loss": 0.3272,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5140298507462686,
+      "grad_norm": 0.8282169867776825,
+      "learning_rate": 2.5084607212203983e-06,
+      "loss": 0.3618,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5143283582089552,
+      "grad_norm": 0.8758117401270619,
+      "learning_rate": 2.5060433779506966e-06,
+      "loss": 0.3096,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5146268656716418,
+      "grad_norm": 0.9027136861131375,
+      "learning_rate": 2.5036260290305837e-06,
+      "loss": 0.3365,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5149253731343284,
+      "grad_norm": 0.8734789095232062,
+      "learning_rate": 2.501208676720223e-06,
+      "loss": 0.3564,
+      "step": 1725
+    },
+    {
+      "epoch": 0.515223880597015,
+      "grad_norm": 0.9223885876305432,
+      "learning_rate": 2.498791323279778e-06,
+      "loss": 0.361,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5155223880597015,
+      "grad_norm": 0.7412513939737528,
+      "learning_rate": 2.496373970969417e-06,
+      "loss": 0.2946,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5158208955223881,
+      "grad_norm": 0.8125704355183377,
+      "learning_rate": 2.4939566220493043e-06,
+      "loss": 0.3597,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5161194029850746,
+      "grad_norm": 0.875058571982458,
+      "learning_rate": 2.491539278779603e-06,
+      "loss": 0.3321,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5164179104477612,
+      "grad_norm": 1.4129333220086666,
+      "learning_rate": 2.48912194342047e-06,
+      "loss": 0.3021,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5167164179104478,
+      "grad_norm": 0.822837577529842,
+      "learning_rate": 2.486704618232054e-06,
+      "loss": 0.2875,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5170149253731343,
+      "grad_norm": 0.8508719610068957,
+      "learning_rate": 2.484287305474496e-06,
+      "loss": 0.332,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5173134328358209,
+      "grad_norm": 0.8371266747144144,
+      "learning_rate": 2.481870007407925e-06,
+      "loss": 0.3505,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5176119402985074,
+      "grad_norm": 0.7674347672744616,
+      "learning_rate": 2.4794527262924546e-06,
+      "loss": 0.3378,
+      "step": 1734
+    },
+    {
+      "epoch": 0.517910447761194,
+      "grad_norm": 0.9297458178829942,
+      "learning_rate": 2.4770354643881843e-06,
+      "loss": 0.3841,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5182089552238806,
+      "grad_norm": 0.9461877714567667,
+      "learning_rate": 2.474618223955196e-06,
+      "loss": 0.3962,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5185074626865671,
+      "grad_norm": 0.9240923700130257,
+      "learning_rate": 2.4722010072535485e-06,
+      "loss": 0.3352,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5188059701492538,
+      "grad_norm": 0.7914811230349772,
+      "learning_rate": 2.4697838165432816e-06,
+      "loss": 0.351,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5191044776119403,
+      "grad_norm": 0.9433458310818481,
+      "learning_rate": 2.4673666540844105e-06,
+      "loss": 0.3725,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5194029850746269,
+      "grad_norm": 0.8700776482081186,
+      "learning_rate": 2.4649495221369218e-06,
+      "loss": 0.3117,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5197014925373135,
+      "grad_norm": 0.7373330115096584,
+      "learning_rate": 2.462532422960776e-06,
+      "loss": 0.3273,
+      "step": 1741
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7741460492182761,
+      "learning_rate": 2.460115358815901e-06,
+      "loss": 0.3185,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5202985074626866,
+      "grad_norm": 0.9184552422451486,
+      "learning_rate": 2.457698331962193e-06,
+      "loss": 0.3499,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5205970149253731,
+      "grad_norm": 0.7416538846201887,
+      "learning_rate": 2.4552813446595148e-06,
+      "loss": 0.3278,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5208955223880597,
+      "grad_norm": 0.8354557623430109,
+      "learning_rate": 2.4528643991676875e-06,
+      "loss": 0.3341,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5211940298507463,
+      "grad_norm": 0.8538898399358935,
+      "learning_rate": 2.450447497746499e-06,
+      "loss": 0.3757,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5214925373134328,
+      "grad_norm": 0.7395052286353708,
+      "learning_rate": 2.4480306426556925e-06,
+      "loss": 0.2997,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5217910447761194,
+      "grad_norm": 0.7583313377731395,
+      "learning_rate": 2.4456138361549678e-06,
+      "loss": 0.3343,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5220895522388059,
+      "grad_norm": 0.8107731144138431,
+      "learning_rate": 2.4431970805039814e-06,
+      "loss": 0.2842,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5223880597014925,
+      "grad_norm": 0.7658894882988232,
+      "learning_rate": 2.4407803779623394e-06,
+      "loss": 0.3124,
+      "step": 1750
+    },
+    {
+      "epoch": 0.522686567164179,
+      "grad_norm": 0.8719707167891656,
+      "learning_rate": 2.4383637307896017e-06,
+      "loss": 0.2977,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5229850746268657,
+      "grad_norm": 0.8510394601889362,
+      "learning_rate": 2.4359471412452724e-06,
+      "loss": 0.3698,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5232835820895523,
+      "grad_norm": 0.9805015870419773,
+      "learning_rate": 2.4335306115888046e-06,
+      "loss": 0.3421,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5235820895522388,
+      "grad_norm": 0.884081775452208,
+      "learning_rate": 2.4311141440795956e-06,
+      "loss": 0.3362,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5238805970149254,
+      "grad_norm": 0.8500753318797264,
+      "learning_rate": 2.4286977409769818e-06,
+      "loss": 0.3223,
+      "step": 1755
+    },
+    {
+      "epoch": 0.524179104477612,
+      "grad_norm": 0.7775675727762501,
+      "learning_rate": 2.4262814045402425e-06,
+      "loss": 0.3587,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5244776119402985,
+      "grad_norm": 0.8771202940506111,
+      "learning_rate": 2.4238651370285936e-06,
+      "loss": 0.3609,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5247761194029851,
+      "grad_norm": 0.9343708174797646,
+      "learning_rate": 2.421448940701185e-06,
+      "loss": 0.3788,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5250746268656716,
+      "grad_norm": 0.7830270720430653,
+      "learning_rate": 2.4190328178171026e-06,
+      "loss": 0.3327,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5253731343283582,
+      "grad_norm": 0.7893636098117658,
+      "learning_rate": 2.4166167706353623e-06,
+      "loss": 0.3195,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5256716417910448,
+      "grad_norm": 0.8111132911057031,
+      "learning_rate": 2.414200801414908e-06,
+      "loss": 0.3735,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5259701492537313,
+      "grad_norm": 0.9047377513665926,
+      "learning_rate": 2.4117849124146136e-06,
+      "loss": 0.3782,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5262686567164179,
+      "grad_norm": 0.8119207582887555,
+      "learning_rate": 2.4093691058932765e-06,
+      "loss": 0.3387,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5265671641791044,
+      "grad_norm": 0.7829803719766413,
+      "learning_rate": 2.4069533841096153e-06,
+      "loss": 0.3503,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5268656716417911,
+      "grad_norm": 0.8065607770759078,
+      "learning_rate": 2.404537749322273e-06,
+      "loss": 0.3407,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5271641791044777,
+      "grad_norm": 0.8371021912735477,
+      "learning_rate": 2.4021222037898066e-06,
+      "loss": 0.3698,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5274626865671642,
+      "grad_norm": 0.8296224322060857,
+      "learning_rate": 2.399706749770695e-06,
+      "loss": 0.2996,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5277611940298508,
+      "grad_norm": 0.7980329495819588,
+      "learning_rate": 2.3972913895233278e-06,
+      "loss": 0.3164,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5280597014925373,
+      "grad_norm": 0.84030721922861,
+      "learning_rate": 2.394876125306007e-06,
+      "loss": 0.3384,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5283582089552239,
+      "grad_norm": 0.8654667250374329,
+      "learning_rate": 2.392460959376947e-06,
+      "loss": 0.3168,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5286567164179105,
+      "grad_norm": 0.8688067439209197,
+      "learning_rate": 2.3900458939942696e-06,
+      "loss": 0.3503,
+      "step": 1771
+    },
+    {
+      "epoch": 0.528955223880597,
+      "grad_norm": 0.7877176392427366,
+      "learning_rate": 2.387630931416e-06,
+      "loss": 0.3048,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5292537313432836,
+      "grad_norm": 0.8446237767654405,
+      "learning_rate": 2.3852160739000706e-06,
+      "loss": 0.3626,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5295522388059701,
+      "grad_norm": 0.8025405273097861,
+      "learning_rate": 2.382801323704315e-06,
+      "loss": 0.3284,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5298507462686567,
+      "grad_norm": 0.9052209154464622,
+      "learning_rate": 2.380386683086465e-06,
+      "loss": 0.3916,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5301492537313433,
+      "grad_norm": 0.83787822521044,
+      "learning_rate": 2.3779721543041504e-06,
+      "loss": 0.3503,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5304477611940298,
+      "grad_norm": 0.8029033496086618,
+      "learning_rate": 2.3755577396148967e-06,
+      "loss": 0.3198,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5307462686567164,
+      "grad_norm": 0.9561232832711954,
+      "learning_rate": 2.3731434412761236e-06,
+      "loss": 0.3653,
+      "step": 1778
+    },
+    {
+      "epoch": 0.531044776119403,
+      "grad_norm": 0.8072801892942726,
+      "learning_rate": 2.370729261545141e-06,
+      "loss": 0.34,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5313432835820896,
+      "grad_norm": 0.7980807938842572,
+      "learning_rate": 2.3683152026791473e-06,
+      "loss": 0.2973,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5316417910447762,
+      "grad_norm": 0.8316738799919338,
+      "learning_rate": 2.365901266935229e-06,
+      "loss": 0.3422,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5319402985074627,
+      "grad_norm": 0.7980662515304342,
+      "learning_rate": 2.363487456570359e-06,
+      "loss": 0.3575,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5322388059701493,
+      "grad_norm": 0.833051789155456,
+      "learning_rate": 2.3610737738413872e-06,
+      "loss": 0.3391,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5325373134328358,
+      "grad_norm": 0.8007945992425828,
+      "learning_rate": 2.3586602210050525e-06,
+      "loss": 0.2989,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5328358208955224,
+      "grad_norm": 0.9029970072294182,
+      "learning_rate": 2.356246800317965e-06,
+      "loss": 0.3445,
+      "step": 1785
+    },
+    {
+      "epoch": 0.533134328358209,
+      "grad_norm": 0.990141115180327,
+      "learning_rate": 2.353833514036616e-06,
+      "loss": 0.3393,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5334328358208955,
+      "grad_norm": 0.8954234000787314,
+      "learning_rate": 2.351420364417368e-06,
+      "loss": 0.3891,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5337313432835821,
+      "grad_norm": 0.8292622437441727,
+      "learning_rate": 2.349007353716458e-06,
+      "loss": 0.344,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5340298507462686,
+      "grad_norm": 0.8448034395608284,
+      "learning_rate": 2.3465944841899916e-06,
+      "loss": 0.3248,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5343283582089552,
+      "grad_norm": 0.9091150167264239,
+      "learning_rate": 2.344181758093943e-06,
+      "loss": 0.3164,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5346268656716417,
+      "grad_norm": 0.77933196910969,
+      "learning_rate": 2.3417691776841532e-06,
+      "loss": 0.3361,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5349253731343283,
+      "grad_norm": 0.8157989161197443,
+      "learning_rate": 2.339356745216325e-06,
+      "loss": 0.2778,
+      "step": 1792
+    },
+    {
+      "epoch": 0.535223880597015,
+      "grad_norm": 0.7979239591724641,
+      "learning_rate": 2.336944462946024e-06,
+      "loss": 0.3471,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5355223880597015,
+      "grad_norm": 0.8373875356209881,
+      "learning_rate": 2.3345323331286763e-06,
+      "loss": 0.3221,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5358208955223881,
+      "grad_norm": 0.8199518664561318,
+      "learning_rate": 2.3321203580195635e-06,
+      "loss": 0.332,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5361194029850747,
+      "grad_norm": 0.8762948103305159,
+      "learning_rate": 2.329708539873824e-06,
+      "loss": 0.3909,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5364179104477612,
+      "grad_norm": 0.8353475092926272,
+      "learning_rate": 2.3272968809464486e-06,
+      "loss": 0.3553,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5367164179104478,
+      "grad_norm": 0.8759847723242117,
+      "learning_rate": 2.3248853834922803e-06,
+      "loss": 0.3686,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5370149253731343,
+      "grad_norm": 0.8615201742620705,
+      "learning_rate": 2.3224740497660096e-06,
+      "loss": 0.3754,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 0.9099685739043476,
+      "learning_rate": 2.3200628820221756e-06,
+      "loss": 0.3581,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5376119402985075,
+      "grad_norm": 0.777596772205429,
+      "learning_rate": 2.3176518825151608e-06,
+      "loss": 0.3378,
+      "step": 1801
+    },
+    {
+      "epoch": 0.537910447761194,
+      "grad_norm": 0.8450144255131784,
+      "learning_rate": 2.315241053499191e-06,
+      "loss": 0.3387,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5382089552238806,
+      "grad_norm": 0.9315730643786867,
+      "learning_rate": 2.3128303972283327e-06,
+      "loss": 0.3867,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5385074626865671,
+      "grad_norm": 0.8448720469427933,
+      "learning_rate": 2.3104199159564902e-06,
+      "loss": 0.3539,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5388059701492537,
+      "grad_norm": 0.8164611944170581,
+      "learning_rate": 2.308009611937406e-06,
+      "loss": 0.3366,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5391044776119402,
+      "grad_norm": 0.8705803506185734,
+      "learning_rate": 2.3055994874246544e-06,
+      "loss": 0.3427,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5394029850746269,
+      "grad_norm": 0.8111592075727269,
+      "learning_rate": 2.3031895446716438e-06,
+      "loss": 0.3498,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5397014925373135,
+      "grad_norm": 0.82443436935079,
+      "learning_rate": 2.300779785931611e-06,
+      "loss": 0.3402,
+      "step": 1808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7893728042942114,
+      "learning_rate": 2.2983702134576237e-06,
+      "loss": 0.3534,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5402985074626866,
+      "grad_norm": 0.7788581120840403,
+      "learning_rate": 2.295960829502571e-06,
+      "loss": 0.3632,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5405970149253732,
+      "grad_norm": 0.7144098288588688,
+      "learning_rate": 2.2935516363191695e-06,
+      "loss": 0.3292,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5408955223880597,
+      "grad_norm": 0.8718613502814399,
+      "learning_rate": 2.2911426361599563e-06,
+      "loss": 0.3914,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5411940298507463,
+      "grad_norm": 0.8140978897388683,
+      "learning_rate": 2.288733831277287e-06,
+      "loss": 0.3285,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5414925373134328,
+      "grad_norm": 0.9065681350482651,
+      "learning_rate": 2.286325223923336e-06,
+      "loss": 0.3734,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5417910447761194,
+      "grad_norm": 0.8578142995410984,
+      "learning_rate": 2.283916816350092e-06,
+      "loss": 0.3417,
+      "step": 1815
+    },
+    {
+      "epoch": 0.542089552238806,
+      "grad_norm": 0.8951438939421517,
+      "learning_rate": 2.2815086108093575e-06,
+      "loss": 0.3678,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5423880597014925,
+      "grad_norm": 0.7710865449797061,
+      "learning_rate": 2.279100609552747e-06,
+      "loss": 0.3212,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5426865671641791,
+      "grad_norm": 0.8413189744305332,
+      "learning_rate": 2.2766928148316815e-06,
+      "loss": 0.3546,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5429850746268656,
+      "grad_norm": 0.7918974236582986,
+      "learning_rate": 2.274285228897391e-06,
+      "loss": 0.3157,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5432835820895522,
+      "grad_norm": 0.8039993848081551,
+      "learning_rate": 2.271877854000909e-06,
+      "loss": 0.3115,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5435820895522389,
+      "grad_norm": 0.8401858367747947,
+      "learning_rate": 2.2694706923930737e-06,
+      "loss": 0.3542,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5438805970149254,
+      "grad_norm": 0.7815175224613609,
+      "learning_rate": 2.2670637463245223e-06,
+      "loss": 0.3417,
+      "step": 1822
+    },
+    {
+      "epoch": 0.544179104477612,
+      "grad_norm": 0.8583325595180437,
+      "learning_rate": 2.2646570180456896e-06,
+      "loss": 0.3806,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5444776119402985,
+      "grad_norm": 0.8765871329748808,
+      "learning_rate": 2.2622505098068098e-06,
+      "loss": 0.3702,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5447761194029851,
+      "grad_norm": 0.7542775615348616,
+      "learning_rate": 2.2598442238579067e-06,
+      "loss": 0.3352,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5450746268656717,
+      "grad_norm": 0.8460036442491983,
+      "learning_rate": 2.2574381624488008e-06,
+      "loss": 0.3481,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5453731343283582,
+      "grad_norm": 0.8043856469909816,
+      "learning_rate": 2.2550323278291013e-06,
+      "loss": 0.3685,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5456716417910448,
+      "grad_norm": 0.8739643120061339,
+      "learning_rate": 2.2526267222482035e-06,
+      "loss": 0.346,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5459701492537313,
+      "grad_norm": 0.9831654255410001,
+      "learning_rate": 2.250221347955291e-06,
+      "loss": 0.3252,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5462686567164179,
+      "grad_norm": 0.8015068502149022,
+      "learning_rate": 2.2478162071993296e-06,
+      "loss": 0.357,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5465671641791044,
+      "grad_norm": 0.7756446807152939,
+      "learning_rate": 2.2454113022290676e-06,
+      "loss": 0.3478,
+      "step": 1831
+    },
+    {
+      "epoch": 0.546865671641791,
+      "grad_norm": 0.9555139693052305,
+      "learning_rate": 2.2430066352930323e-06,
+      "loss": 0.3294,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5471641791044776,
+      "grad_norm": 0.8021249504547362,
+      "learning_rate": 2.240602208639529e-06,
+      "loss": 0.3343,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5474626865671641,
+      "grad_norm": 0.7978682944455774,
+      "learning_rate": 2.238198024516637e-06,
+      "loss": 0.3285,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5477611940298508,
+      "grad_norm": 0.8124634613968112,
+      "learning_rate": 2.2357940851722114e-06,
+      "loss": 0.336,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5480597014925374,
+      "grad_norm": 0.8713005908230844,
+      "learning_rate": 2.2333903928538765e-06,
+      "loss": 0.3149,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5483582089552239,
+      "grad_norm": 0.9595459470624672,
+      "learning_rate": 2.230986949809025e-06,
+      "loss": 0.3468,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5486567164179105,
+      "grad_norm": 0.8161907559262533,
+      "learning_rate": 2.2285837582848185e-06,
+      "loss": 0.3801,
+      "step": 1838
+    },
+    {
+      "epoch": 0.548955223880597,
+      "grad_norm": 0.8893397151532759,
+      "learning_rate": 2.226180820528182e-06,
+      "loss": 0.3721,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5492537313432836,
+      "grad_norm": 0.7700788867504089,
+      "learning_rate": 2.223778138785804e-06,
+      "loss": 0.3014,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5495522388059702,
+      "grad_norm": 0.871007482090454,
+      "learning_rate": 2.2213757153041337e-06,
+      "loss": 0.3615,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5498507462686567,
+      "grad_norm": 0.8633799398623063,
+      "learning_rate": 2.2189735523293777e-06,
+      "loss": 0.361,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5501492537313433,
+      "grad_norm": 0.8645468422430543,
+      "learning_rate": 2.2165716521075e-06,
+      "loss": 0.3533,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5504477611940298,
+      "grad_norm": 0.8540892598497185,
+      "learning_rate": 2.21417001688422e-06,
+      "loss": 0.3592,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5507462686567164,
+      "grad_norm": 0.8694634042599341,
+      "learning_rate": 2.211768648905006e-06,
+      "loss": 0.352,
+      "step": 1845
+    },
+    {
+      "epoch": 0.551044776119403,
+      "grad_norm": 0.7911027777874611,
+      "learning_rate": 2.20936755041508e-06,
+      "loss": 0.3108,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5513432835820895,
+      "grad_norm": 0.8066430351634646,
+      "learning_rate": 2.2069667236594104e-06,
+      "loss": 0.2852,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5516417910447761,
+      "grad_norm": 0.8185535177696429,
+      "learning_rate": 2.204566170882711e-06,
+      "loss": 0.34,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5519402985074627,
+      "grad_norm": 0.7993975027083716,
+      "learning_rate": 2.202165894329441e-06,
+      "loss": 0.3781,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5522388059701493,
+      "grad_norm": 0.849262989720656,
+      "learning_rate": 2.1997658962438003e-06,
+      "loss": 0.3165,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5525373134328359,
+      "grad_norm": 0.7278759041881264,
+      "learning_rate": 2.197366178869728e-06,
+      "loss": 0.2826,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5528358208955224,
+      "grad_norm": 0.828521762307412,
+      "learning_rate": 2.1949667444509025e-06,
+      "loss": 0.3265,
+      "step": 1852
+    },
+    {
+      "epoch": 0.553134328358209,
+      "grad_norm": 0.7855801046085621,
+      "learning_rate": 2.192567595230735e-06,
+      "loss": 0.319,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5534328358208955,
+      "grad_norm": 0.7381825111391765,
+      "learning_rate": 2.190168733452372e-06,
+      "loss": 0.3203,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5537313432835821,
+      "grad_norm": 0.8446345881014914,
+      "learning_rate": 2.187770161358692e-06,
+      "loss": 0.3271,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5540298507462686,
+      "grad_norm": 0.725953827886251,
+      "learning_rate": 2.1853718811922996e-06,
+      "loss": 0.3016,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5543283582089552,
+      "grad_norm": 0.8971268582384204,
+      "learning_rate": 2.1829738951955295e-06,
+      "loss": 0.3595,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5546268656716418,
+      "grad_norm": 0.8332215625700863,
+      "learning_rate": 2.180576205610439e-06,
+      "loss": 0.2938,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5549253731343283,
+      "grad_norm": 0.8275970156258079,
+      "learning_rate": 2.178178814678812e-06,
+      "loss": 0.3862,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5552238805970149,
+      "grad_norm": 0.86905690858234,
+      "learning_rate": 2.1757817246421477e-06,
+      "loss": 0.3728,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5555223880597014,
+      "grad_norm": 0.7889094120217989,
+      "learning_rate": 2.173384937741668e-06,
+      "loss": 0.307,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5558208955223881,
+      "grad_norm": 0.9468308573254213,
+      "learning_rate": 2.1709884562183114e-06,
+      "loss": 0.3373,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5561194029850747,
+      "grad_norm": 0.8358643562239164,
+      "learning_rate": 2.1685922823127276e-06,
+      "loss": 0.3595,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5564179104477612,
+      "grad_norm": 0.8195541395992418,
+      "learning_rate": 2.166196418265282e-06,
+      "loss": 0.3175,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5567164179104478,
+      "grad_norm": 0.8872518656245127,
+      "learning_rate": 2.163800866316049e-06,
+      "loss": 0.3329,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5570149253731344,
+      "grad_norm": 0.8974786463656703,
+      "learning_rate": 2.161405628704811e-06,
+      "loss": 0.3594,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5573134328358209,
+      "grad_norm": 0.9090184559871817,
+      "learning_rate": 2.1590107076710567e-06,
+      "loss": 0.341,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5576119402985075,
+      "grad_norm": 0.795923872852375,
+      "learning_rate": 2.1566161054539797e-06,
+      "loss": 0.3174,
+      "step": 1868
+    },
+    {
+      "epoch": 0.557910447761194,
+      "grad_norm": 0.836212534397218,
+      "learning_rate": 2.1542218242924737e-06,
+      "loss": 0.3334,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5582089552238806,
+      "grad_norm": 0.9386342551389201,
+      "learning_rate": 2.1518278664251334e-06,
+      "loss": 0.3427,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5585074626865671,
+      "grad_norm": 0.8270322091053605,
+      "learning_rate": 2.1494342340902518e-06,
+      "loss": 0.3581,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5588059701492537,
+      "grad_norm": 0.8355544694865434,
+      "learning_rate": 2.147040929525816e-06,
+      "loss": 0.3277,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5591044776119403,
+      "grad_norm": 0.7760309647668585,
+      "learning_rate": 2.1446479549695083e-06,
+      "loss": 0.3003,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5594029850746268,
+      "grad_norm": 0.8285302940910255,
+      "learning_rate": 2.1422553126587e-06,
+      "loss": 0.3398,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 0.8028111781327222,
+      "learning_rate": 2.139863004830455e-06,
+      "loss": 0.3068,
+      "step": 1875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9241018551228057,
+      "learning_rate": 2.1374710337215223e-06,
+      "loss": 0.3117,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5602985074626866,
+      "grad_norm": 0.8859437856608664,
+      "learning_rate": 2.1350794015683356e-06,
+      "loss": 0.3366,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5605970149253732,
+      "grad_norm": 0.8066823845490054,
+      "learning_rate": 2.1326881106070137e-06,
+      "loss": 0.3239,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5608955223880597,
+      "grad_norm": 0.8280625069457178,
+      "learning_rate": 2.1302971630733553e-06,
+      "loss": 0.3256,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5611940298507463,
+      "grad_norm": 0.7654383674574374,
+      "learning_rate": 2.127906561202838e-06,
+      "loss": 0.3152,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5614925373134328,
+      "grad_norm": 0.8019819178042276,
+      "learning_rate": 2.125516307230616e-06,
+      "loss": 0.3496,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5617910447761194,
+      "grad_norm": 0.8424062174944292,
+      "learning_rate": 2.1231264033915188e-06,
+      "loss": 0.3457,
+      "step": 1882
+    },
+    {
+      "epoch": 0.562089552238806,
+      "grad_norm": 1.0405265753175954,
+      "learning_rate": 2.1207368519200483e-06,
+      "loss": 0.3606,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5623880597014925,
+      "grad_norm": 0.8012391807155793,
+      "learning_rate": 2.1183476550503763e-06,
+      "loss": 0.3336,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5626865671641791,
+      "grad_norm": 0.8303562700296768,
+      "learning_rate": 2.1159588150163447e-06,
+      "loss": 0.3525,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5629850746268656,
+      "grad_norm": 0.8141402829710194,
+      "learning_rate": 2.113570334051459e-06,
+      "loss": 0.3743,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5632835820895522,
+      "grad_norm": 0.7825708947108343,
+      "learning_rate": 2.111182214388893e-06,
+      "loss": 0.2933,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5635820895522388,
+      "grad_norm": 0.8309709447567482,
+      "learning_rate": 2.108794458261478e-06,
+      "loss": 0.3238,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5638805970149253,
+      "grad_norm": 0.8224296447050079,
+      "learning_rate": 2.106407067901709e-06,
+      "loss": 0.3258,
+      "step": 1889
+    },
+    {
+      "epoch": 0.564179104477612,
+      "grad_norm": 0.9018945970876204,
+      "learning_rate": 2.104020045541739e-06,
+      "loss": 0.3827,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5644776119402986,
+      "grad_norm": 0.8513253106321182,
+      "learning_rate": 2.1016333934133727e-06,
+      "loss": 0.3399,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5647761194029851,
+      "grad_norm": 0.9164770773291085,
+      "learning_rate": 2.099247113748074e-06,
+      "loss": 0.3529,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5650746268656717,
+      "grad_norm": 0.8334813106827352,
+      "learning_rate": 2.096861208776956e-06,
+      "loss": 0.342,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5653731343283582,
+      "grad_norm": 0.79013138961232,
+      "learning_rate": 2.0944756807307805e-06,
+      "loss": 0.3404,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5656716417910448,
+      "grad_norm": 0.8495550812479902,
+      "learning_rate": 2.092090531839959e-06,
+      "loss": 0.3307,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5659701492537313,
+      "grad_norm": 0.8325862898480291,
+      "learning_rate": 2.089705764334547e-06,
+      "loss": 0.3557,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5662686567164179,
+      "grad_norm": 0.8222800051067589,
+      "learning_rate": 2.0873213804442454e-06,
+      "loss": 0.3185,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5665671641791045,
+      "grad_norm": 0.7723465046550894,
+      "learning_rate": 2.0849373823983935e-06,
+      "loss": 0.3464,
+      "step": 1898
+    },
+    {
+      "epoch": 0.566865671641791,
+      "grad_norm": 0.8615414329369864,
+      "learning_rate": 2.082553772425972e-06,
+      "loss": 0.3505,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5671641791044776,
+      "grad_norm": 0.8521312443511204,
+      "learning_rate": 2.080170552755598e-06,
+      "loss": 0.338,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5674626865671641,
+      "grad_norm": 0.8669205691719605,
+      "learning_rate": 2.077787725615524e-06,
+      "loss": 0.3645,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5677611940298507,
+      "grad_norm": 0.750771605956206,
+      "learning_rate": 2.075405293233635e-06,
+      "loss": 0.2969,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5680597014925373,
+      "grad_norm": 0.770946338972137,
+      "learning_rate": 2.073023257837448e-06,
+      "loss": 0.3189,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5683582089552239,
+      "grad_norm": 0.8472428925223515,
+      "learning_rate": 2.070641621654106e-06,
+      "loss": 0.3146,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5686567164179105,
+      "grad_norm": 0.8076339095091497,
+      "learning_rate": 2.068260386910383e-06,
+      "loss": 0.3259,
+      "step": 1905
+    },
+    {
+      "epoch": 0.568955223880597,
+      "grad_norm": 0.7546439694354958,
+      "learning_rate": 2.0658795558326745e-06,
+      "loss": 0.3046,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5692537313432836,
+      "grad_norm": 0.7863545601633702,
+      "learning_rate": 2.063499130646999e-06,
+      "loss": 0.3119,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5695522388059702,
+      "grad_norm": 0.9196917659773726,
+      "learning_rate": 2.0611191135789972e-06,
+      "loss": 0.3762,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5698507462686567,
+      "grad_norm": 0.8725237416065241,
+      "learning_rate": 2.0587395068539268e-06,
+      "loss": 0.3656,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5701492537313433,
+      "grad_norm": 0.8403813292122189,
+      "learning_rate": 2.056360312696661e-06,
+      "loss": 0.3555,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5704477611940298,
+      "grad_norm": 0.739162644070513,
+      "learning_rate": 2.0539815333316896e-06,
+      "loss": 0.3363,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5707462686567164,
+      "grad_norm": 0.8072527202747288,
+      "learning_rate": 2.0516031709831124e-06,
+      "loss": 0.3338,
+      "step": 1912
+    },
+    {
+      "epoch": 0.571044776119403,
+      "grad_norm": 0.7551658882134142,
+      "learning_rate": 2.0492252278746406e-06,
+      "loss": 0.3133,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5713432835820895,
+      "grad_norm": 0.8610162991085163,
+      "learning_rate": 2.046847706229594e-06,
+      "loss": 0.3595,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5716417910447761,
+      "grad_norm": 0.8314973416295284,
+      "learning_rate": 2.0444706082708957e-06,
+      "loss": 0.3151,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5719402985074626,
+      "grad_norm": 0.7775346420782155,
+      "learning_rate": 2.042093936221075e-06,
+      "loss": 0.328,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5722388059701492,
+      "grad_norm": 1.4529917092934526,
+      "learning_rate": 2.039717692302263e-06,
+      "loss": 0.3158,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5725373134328359,
+      "grad_norm": 0.8832461020890808,
+      "learning_rate": 2.0373418787361886e-06,
+      "loss": 0.3249,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5728358208955224,
+      "grad_norm": 0.8529873184087406,
+      "learning_rate": 2.0349664977441806e-06,
+      "loss": 0.3627,
+      "step": 1919
+    },
+    {
+      "epoch": 0.573134328358209,
+      "grad_norm": 0.8801116310966653,
+      "learning_rate": 2.0325915515471627e-06,
+      "loss": 0.3887,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5734328358208955,
+      "grad_norm": 0.8146237688887484,
+      "learning_rate": 2.03021704236565e-06,
+      "loss": 0.3567,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5737313432835821,
+      "grad_norm": 0.8299437970321636,
+      "learning_rate": 2.0278429724197517e-06,
+      "loss": 0.3358,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5740298507462687,
+      "grad_norm": 0.8343117163911652,
+      "learning_rate": 2.0254693439291665e-06,
+      "loss": 0.3466,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5743283582089552,
+      "grad_norm": 0.7664529851147533,
+      "learning_rate": 2.023096159113177e-06,
+      "loss": 0.3201,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5746268656716418,
+      "grad_norm": 0.8295280556984195,
+      "learning_rate": 2.0207234201906546e-06,
+      "loss": 0.3285,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5749253731343283,
+      "grad_norm": 0.9191354365414987,
+      "learning_rate": 2.018351129380052e-06,
+      "loss": 0.3826,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5752238805970149,
+      "grad_norm": 0.7452870603167776,
+      "learning_rate": 2.0159792888994025e-06,
+      "loss": 0.3166,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5755223880597015,
+      "grad_norm": 1.0403191478557996,
+      "learning_rate": 2.0136079009663205e-06,
+      "loss": 0.3165,
+      "step": 1928
+    },
+    {
+      "epoch": 0.575820895522388,
+      "grad_norm": 0.8016626447497367,
+      "learning_rate": 2.0112369677979955e-06,
+      "loss": 0.3246,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5761194029850746,
+      "grad_norm": 0.8057426738001215,
+      "learning_rate": 2.008866491611191e-06,
+      "loss": 0.2706,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5764179104477611,
+      "grad_norm": 0.8464060075703335,
+      "learning_rate": 2.0064964746222464e-06,
+      "loss": 0.3619,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5767164179104478,
+      "grad_norm": 0.7572279430026017,
+      "learning_rate": 2.0041269190470687e-06,
+      "loss": 0.3071,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5770149253731344,
+      "grad_norm": 0.8070777794735167,
+      "learning_rate": 2.001757827101134e-06,
+      "loss": 0.3298,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5773134328358209,
+      "grad_norm": 0.7406332845203685,
+      "learning_rate": 1.9993892009994862e-06,
+      "loss": 0.31,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5776119402985075,
+      "grad_norm": 0.8057738674065215,
+      "learning_rate": 1.9970210429567343e-06,
+      "loss": 0.3689,
+      "step": 1935
+    },
+    {
+      "epoch": 0.577910447761194,
+      "grad_norm": 0.8334636514302126,
+      "learning_rate": 1.9946533551870465e-06,
+      "loss": 0.3531,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5782089552238806,
+      "grad_norm": 0.7674289139447094,
+      "learning_rate": 1.9922861399041537e-06,
+      "loss": 0.357,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5785074626865672,
+      "grad_norm": 0.8048568434131651,
+      "learning_rate": 1.989919399321345e-06,
+      "loss": 0.346,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5788059701492537,
+      "grad_norm": 0.7475874633539165,
+      "learning_rate": 1.9875531356514642e-06,
+      "loss": 0.3308,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5791044776119403,
+      "grad_norm": 1.1995168376736667,
+      "learning_rate": 1.9851873511069104e-06,
+      "loss": 0.3112,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5794029850746268,
+      "grad_norm": 0.8706989388369915,
+      "learning_rate": 1.9828220478996357e-06,
+      "loss": 0.3703,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5797014925373134,
+      "grad_norm": 0.8260670136022402,
+      "learning_rate": 1.980457228241139e-06,
+      "loss": 0.317,
+      "step": 1942
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7553465208304225,
+      "learning_rate": 1.9780928943424703e-06,
+      "loss": 0.3407,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5802985074626865,
+      "grad_norm": 0.9544720767792053,
+      "learning_rate": 1.9757290484142244e-06,
+      "loss": 0.3739,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5805970149253732,
+      "grad_norm": 0.8720188388239823,
+      "learning_rate": 1.9733656926665388e-06,
+      "loss": 0.3476,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5808955223880597,
+      "grad_norm": 0.8709469192102466,
+      "learning_rate": 1.971002829309094e-06,
+      "loss": 0.3701,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5811940298507463,
+      "grad_norm": 0.8604365263309919,
+      "learning_rate": 1.968640460551109e-06,
+      "loss": 0.3698,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5814925373134329,
+      "grad_norm": 0.7533527794831715,
+      "learning_rate": 1.9662785886013417e-06,
+      "loss": 0.3349,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5817910447761194,
+      "grad_norm": 0.8247273401204854,
+      "learning_rate": 1.9639172156680848e-06,
+      "loss": 0.3361,
+      "step": 1949
+    },
+    {
+      "epoch": 0.582089552238806,
+      "grad_norm": 0.8120190282994632,
+      "learning_rate": 1.961556343959164e-06,
+      "loss": 0.2857,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5823880597014925,
+      "grad_norm": 0.862894378293909,
+      "learning_rate": 1.9591959756819366e-06,
+      "loss": 0.3927,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5826865671641791,
+      "grad_norm": 0.8218779267658843,
+      "learning_rate": 1.9568361130432907e-06,
+      "loss": 0.3287,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5829850746268657,
+      "grad_norm": 0.8143989499819814,
+      "learning_rate": 1.954476758249639e-06,
+      "loss": 0.3454,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5832835820895522,
+      "grad_norm": 0.8454839729330965,
+      "learning_rate": 1.9521179135069213e-06,
+      "loss": 0.3276,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5835820895522388,
+      "grad_norm": 0.750200670423591,
+      "learning_rate": 1.9497595810206e-06,
+      "loss": 0.3304,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5838805970149253,
+      "grad_norm": 0.8045091719653029,
+      "learning_rate": 1.9474017629956576e-06,
+      "loss": 0.3525,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5841791044776119,
+      "grad_norm": 0.904970348742468,
+      "learning_rate": 1.9450444616365976e-06,
+      "loss": 0.3555,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5844776119402985,
+      "grad_norm": 0.7892323857546758,
+      "learning_rate": 1.9426876791474396e-06,
+      "loss": 0.3307,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5847761194029851,
+      "grad_norm": 0.8188095376158899,
+      "learning_rate": 1.940331417731716e-06,
+      "loss": 0.3035,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5850746268656717,
+      "grad_norm": 0.7550954391161062,
+      "learning_rate": 1.9379756795924757e-06,
+      "loss": 0.309,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5853731343283582,
+      "grad_norm": 0.785376246418163,
+      "learning_rate": 1.935620466932275e-06,
+      "loss": 0.3116,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5856716417910448,
+      "grad_norm": 0.8351399359015248,
+      "learning_rate": 1.933265781953181e-06,
+      "loss": 0.3944,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5859701492537314,
+      "grad_norm": 0.8086095361360239,
+      "learning_rate": 1.9309116268567675e-06,
+      "loss": 0.3319,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5862686567164179,
+      "grad_norm": 0.953784526278762,
+      "learning_rate": 1.9285580038441105e-06,
+      "loss": 0.3373,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5865671641791045,
+      "grad_norm": 0.909438958477141,
+      "learning_rate": 1.926204915115791e-06,
+      "loss": 0.3498,
+      "step": 1965
+    },
+    {
+      "epoch": 0.586865671641791,
+      "grad_norm": 0.8033996032006552,
+      "learning_rate": 1.92385236287189e-06,
+      "loss": 0.3342,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5871641791044776,
+      "grad_norm": 0.8135989857406488,
+      "learning_rate": 1.921500349311986e-06,
+      "loss": 0.3831,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5874626865671642,
+      "grad_norm": 0.850789249214276,
+      "learning_rate": 1.919148876635154e-06,
+      "loss": 0.2999,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5877611940298507,
+      "grad_norm": 0.7746152215096338,
+      "learning_rate": 1.916797947039965e-06,
+      "loss": 0.3327,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5880597014925373,
+      "grad_norm": 0.9288841282660558,
+      "learning_rate": 1.914447562724479e-06,
+      "loss": 0.3201,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5883582089552238,
+      "grad_norm": 0.7498990167094447,
+      "learning_rate": 1.9120977258862493e-06,
+      "loss": 0.3018,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5886567164179104,
+      "grad_norm": 0.7828793001916144,
+      "learning_rate": 1.909748438722315e-06,
+      "loss": 0.3376,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5889552238805971,
+      "grad_norm": 0.9745524576411368,
+      "learning_rate": 1.9073997034292043e-06,
+      "loss": 0.4012,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5892537313432836,
+      "grad_norm": 0.7597359048329623,
+      "learning_rate": 1.905051522202926e-06,
+      "loss": 0.2966,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5895522388059702,
+      "grad_norm": 0.7985139531152796,
+      "learning_rate": 1.902703897238972e-06,
+      "loss": 0.3545,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5898507462686567,
+      "grad_norm": 0.7828851032723473,
+      "learning_rate": 1.9003568307323156e-06,
+      "loss": 0.3402,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5901492537313433,
+      "grad_norm": 0.8844102544530947,
+      "learning_rate": 1.898010324877406e-06,
+      "loss": 0.3512,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5904477611940299,
+      "grad_norm": 1.2103129479742514,
+      "learning_rate": 1.8956643818681685e-06,
+      "loss": 0.3557,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5907462686567164,
+      "grad_norm": 0.8338907004205794,
+      "learning_rate": 1.8933190038980044e-06,
+      "loss": 0.2936,
+      "step": 1979
+    },
+    {
+      "epoch": 0.591044776119403,
+      "grad_norm": 0.798234969689793,
+      "learning_rate": 1.890974193159783e-06,
+      "loss": 0.3208,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5913432835820895,
+      "grad_norm": 0.8265334530994278,
+      "learning_rate": 1.888629951845846e-06,
+      "loss": 0.3321,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5916417910447761,
+      "grad_norm": 0.8484611458052064,
+      "learning_rate": 1.8862862821480023e-06,
+      "loss": 0.3302,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5919402985074627,
+      "grad_norm": 0.7732292982110145,
+      "learning_rate": 1.8839431862575252e-06,
+      "loss": 0.3178,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5922388059701492,
+      "grad_norm": 0.7854227204822434,
+      "learning_rate": 1.881600666365153e-06,
+      "loss": 0.2743,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5925373134328358,
+      "grad_norm": 0.8822786181065627,
+      "learning_rate": 1.8792587246610833e-06,
+      "loss": 0.2931,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5928358208955223,
+      "grad_norm": 0.8413866496121205,
+      "learning_rate": 1.8769173633349757e-06,
+      "loss": 0.3055,
+      "step": 1986
+    },
+    {
+      "epoch": 0.593134328358209,
+      "grad_norm": 0.8929662942782096,
+      "learning_rate": 1.8745765845759466e-06,
+      "loss": 0.3441,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5934328358208956,
+      "grad_norm": 0.8084666675397074,
+      "learning_rate": 1.8722363905725654e-06,
+      "loss": 0.3335,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5937313432835821,
+      "grad_norm": 0.8138764969858261,
+      "learning_rate": 1.8698967835128572e-06,
+      "loss": 0.3642,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5940298507462687,
+      "grad_norm": 0.7905280819462355,
+      "learning_rate": 1.867557765584298e-06,
+      "loss": 0.3279,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5943283582089552,
+      "grad_norm": 0.7863823629467317,
+      "learning_rate": 1.8652193389738122e-06,
+      "loss": 0.3364,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5946268656716418,
+      "grad_norm": 0.7267169725040129,
+      "learning_rate": 1.8628815058677712e-06,
+      "loss": 0.3171,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5949253731343284,
+      "grad_norm": 0.8524291293019064,
+      "learning_rate": 1.8605442684519932e-06,
+      "loss": 0.3593,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5952238805970149,
+      "grad_norm": 0.7776370242867892,
+      "learning_rate": 1.8582076289117367e-06,
+      "loss": 0.3241,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5955223880597015,
+      "grad_norm": 0.899435349098279,
+      "learning_rate": 1.8558715894317033e-06,
+      "loss": 0.3719,
+      "step": 1995
+    },
+    {
+      "epoch": 0.595820895522388,
+      "grad_norm": 0.8848148638993476,
+      "learning_rate": 1.8535361521960341e-06,
+      "loss": 0.3868,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5961194029850746,
+      "grad_norm": 0.8154092596098508,
+      "learning_rate": 1.8512013193883044e-06,
+      "loss": 0.3443,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5964179104477612,
+      "grad_norm": 0.8146716902809428,
+      "learning_rate": 1.8488670931915268e-06,
+      "loss": 0.3366,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5967164179104477,
+      "grad_norm": 0.8576493081535792,
+      "learning_rate": 1.8465334757881453e-06,
+      "loss": 0.3671,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.8128896259402698,
+      "learning_rate": 1.8442004693600358e-06,
+      "loss": 0.3748,
+      "step": 2000
+    },
+    {
+      "epoch": 0.597313432835821,
+      "grad_norm": 0.7918103129525407,
+      "learning_rate": 1.8418680760885028e-06,
+      "loss": 0.3413,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5976119402985075,
+      "grad_norm": 0.8813005048512403,
+      "learning_rate": 1.8395362981542757e-06,
+      "loss": 0.3706,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5979104477611941,
+      "grad_norm": 0.823115743992545,
+      "learning_rate": 1.8372051377375117e-06,
+      "loss": 0.3721,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5982089552238806,
+      "grad_norm": 0.7894206010095408,
+      "learning_rate": 1.8348745970177884e-06,
+      "loss": 0.3229,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5985074626865672,
+      "grad_norm": 1.3615178436392212,
+      "learning_rate": 1.8325446781741041e-06,
+      "loss": 0.3459,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5988059701492537,
+      "grad_norm": 0.7843156247200797,
+      "learning_rate": 1.8302153833848762e-06,
+      "loss": 0.3221,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5991044776119403,
+      "grad_norm": 1.075019211892318,
+      "learning_rate": 1.8278867148279395e-06,
+      "loss": 0.3769,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5994029850746269,
+      "grad_norm": 0.8427937686071887,
+      "learning_rate": 1.8255586746805412e-06,
+      "loss": 0.3626,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5997014925373134,
+      "grad_norm": 1.0019280320812132,
+      "learning_rate": 1.8232312651193418e-06,
+      "loss": 0.319,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.752076030590557,
+      "learning_rate": 1.8209044883204141e-06,
+      "loss": 0.3165,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6002985074626865,
+      "grad_norm": 0.7911868281154993,
+      "learning_rate": 1.8185783464592372e-06,
+      "loss": 0.3463,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6005970149253731,
+      "grad_norm": 0.8613079162533939,
+      "learning_rate": 1.8162528417106962e-06,
+      "loss": 0.3539,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6008955223880597,
+      "grad_norm": 0.7566901309370632,
+      "learning_rate": 1.8139279762490816e-06,
+      "loss": 0.3156,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6011940298507462,
+      "grad_norm": 0.8761370837990474,
+      "learning_rate": 1.8116037522480866e-06,
+      "loss": 0.3873,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6014925373134329,
+      "grad_norm": 0.9028602770938112,
+      "learning_rate": 1.8092801718808035e-06,
+      "loss": 0.3325,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6017910447761194,
+      "grad_norm": 0.8144205633300299,
+      "learning_rate": 1.8069572373197234e-06,
+      "loss": 0.3361,
+      "step": 2016
+    },
+    {
+      "epoch": 0.602089552238806,
+      "grad_norm": 0.8544221934437295,
+      "learning_rate": 1.8046349507367336e-06,
+      "loss": 0.3611,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6023880597014926,
+      "grad_norm": 0.808086145145757,
+      "learning_rate": 1.802313314303115e-06,
+      "loss": 0.3479,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6026865671641791,
+      "grad_norm": 0.9626316828813825,
+      "learning_rate": 1.7999923301895419e-06,
+      "loss": 0.357,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6029850746268657,
+      "grad_norm": 0.8065948651275312,
+      "learning_rate": 1.797672000566077e-06,
+      "loss": 0.3632,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6032835820895522,
+      "grad_norm": 0.8600310373365637,
+      "learning_rate": 1.795352327602172e-06,
+      "loss": 0.3467,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6035820895522388,
+      "grad_norm": 0.7840180123347607,
+      "learning_rate": 1.7930333134666655e-06,
+      "loss": 0.3145,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6038805970149254,
+      "grad_norm": 0.8760399951287269,
+      "learning_rate": 1.790714960327778e-06,
+      "loss": 0.3523,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6041791044776119,
+      "grad_norm": 0.9401970743207244,
+      "learning_rate": 1.7883972703531138e-06,
+      "loss": 0.3923,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6044776119402985,
+      "grad_norm": 0.7714575169408807,
+      "learning_rate": 1.786080245709657e-06,
+      "loss": 0.331,
+      "step": 2025
+    },
+    {
+      "epoch": 0.604776119402985,
+      "grad_norm": 0.8115570578510135,
+      "learning_rate": 1.7837638885637682e-06,
+      "loss": 0.2821,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6050746268656716,
+      "grad_norm": 0.8360919692031952,
+      "learning_rate": 1.7814482010811852e-06,
+      "loss": 0.3322,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6053731343283583,
+      "grad_norm": 0.8034337429029215,
+      "learning_rate": 1.77913318542702e-06,
+      "loss": 0.3147,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6056716417910448,
+      "grad_norm": 0.7672266818780934,
+      "learning_rate": 1.776818843765755e-06,
+      "loss": 0.3489,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6059701492537314,
+      "grad_norm": 0.7896411247534145,
+      "learning_rate": 1.7745051782612432e-06,
+      "loss": 0.3341,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6062686567164179,
+      "grad_norm": 0.9298746237897938,
+      "learning_rate": 1.7721921910767066e-06,
+      "loss": 0.3759,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6065671641791045,
+      "grad_norm": 0.7832828837477588,
+      "learning_rate": 1.76987988437473e-06,
+      "loss": 0.3277,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6068656716417911,
+      "grad_norm": 0.8044375141975681,
+      "learning_rate": 1.7675682603172656e-06,
+      "loss": 0.3269,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6071641791044776,
+      "grad_norm": 0.8011482806477036,
+      "learning_rate": 1.7652573210656242e-06,
+      "loss": 0.3285,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6074626865671642,
+      "grad_norm": 0.8705610431211035,
+      "learning_rate": 1.7629470687804783e-06,
+      "loss": 0.3102,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6077611940298507,
+      "grad_norm": 0.8594206025341075,
+      "learning_rate": 1.7606375056218578e-06,
+      "loss": 0.3346,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6080597014925373,
+      "grad_norm": 0.8509811944236282,
+      "learning_rate": 1.7583286337491472e-06,
+      "loss": 0.327,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6083582089552239,
+      "grad_norm": 0.7545342074644574,
+      "learning_rate": 1.7560204553210858e-06,
+      "loss": 0.3468,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6086567164179104,
+      "grad_norm": 0.8580794849406649,
+      "learning_rate": 1.7537129724957644e-06,
+      "loss": 0.3539,
+      "step": 2039
+    },
+    {
+      "epoch": 0.608955223880597,
+      "grad_norm": 0.9223321696220602,
+      "learning_rate": 1.7514061874306226e-06,
+      "loss": 0.3834,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6092537313432835,
+      "grad_norm": 0.7309630284623936,
+      "learning_rate": 1.7491001022824483e-06,
+      "loss": 0.3494,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6095522388059702,
+      "grad_norm": 0.9098414650575413,
+      "learning_rate": 1.7467947192073759e-06,
+      "loss": 0.3578,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6098507462686568,
+      "grad_norm": 0.8443943503187832,
+      "learning_rate": 1.744490040360881e-06,
+      "loss": 0.3264,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6101492537313433,
+      "grad_norm": 0.8020741133467957,
+      "learning_rate": 1.7421860678977831e-06,
+      "loss": 0.327,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6104477611940299,
+      "grad_norm": 0.8065034628311121,
+      "learning_rate": 1.7398828039722403e-06,
+      "loss": 0.333,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6107462686567164,
+      "grad_norm": 0.932836415200145,
+      "learning_rate": 1.7375802507377475e-06,
+      "loss": 0.3337,
+      "step": 2046
+    },
+    {
+      "epoch": 0.611044776119403,
+      "grad_norm": 0.8386207975303591,
+      "learning_rate": 1.7352784103471355e-06,
+      "loss": 0.3384,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6113432835820896,
+      "grad_norm": 0.9979807581300122,
+      "learning_rate": 1.7329772849525707e-06,
+      "loss": 0.3463,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6116417910447761,
+      "grad_norm": 0.9144481465630234,
+      "learning_rate": 1.7306768767055481e-06,
+      "loss": 0.3171,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6119402985074627,
+      "grad_norm": 0.9179258800006835,
+      "learning_rate": 1.7283771877568934e-06,
+      "loss": 0.2897,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6122388059701492,
+      "grad_norm": 0.9088510671380466,
+      "learning_rate": 1.7260782202567595e-06,
+      "loss": 0.337,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6125373134328358,
+      "grad_norm": 0.7587527894434154,
+      "learning_rate": 1.723779976354626e-06,
+      "loss": 0.3155,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6128358208955224,
+      "grad_norm": 0.8273268196048457,
+      "learning_rate": 1.721482458199294e-06,
+      "loss": 0.3487,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6131343283582089,
+      "grad_norm": 0.875544314571328,
+      "learning_rate": 1.7191856679388869e-06,
+      "loss": 0.3419,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6134328358208955,
+      "grad_norm": 0.8395751763266083,
+      "learning_rate": 1.7168896077208482e-06,
+      "loss": 0.3419,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6137313432835821,
+      "grad_norm": 0.8125721253715453,
+      "learning_rate": 1.714594279691938e-06,
+      "loss": 0.3201,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6140298507462687,
+      "grad_norm": 0.91793798749558,
+      "learning_rate": 1.7122996859982318e-06,
+      "loss": 0.324,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6143283582089553,
+      "grad_norm": 0.8806848474353035,
+      "learning_rate": 1.710005828785119e-06,
+      "loss": 0.3682,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6146268656716418,
+      "grad_norm": 0.7958603606959694,
+      "learning_rate": 1.7077127101973e-06,
+      "loss": 0.3475,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6149253731343284,
+      "grad_norm": 0.8477880797334701,
+      "learning_rate": 1.7054203323787854e-06,
+      "loss": 0.3667,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6152238805970149,
+      "grad_norm": 0.8049693453434607,
+      "learning_rate": 1.7031286974728916e-06,
+      "loss": 0.3672,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6155223880597015,
+      "grad_norm": 0.8445392848976708,
+      "learning_rate": 1.7008378076222417e-06,
+      "loss": 0.3454,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6158208955223881,
+      "grad_norm": 0.8355246174084053,
+      "learning_rate": 1.698547664968763e-06,
+      "loss": 0.3214,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6161194029850746,
+      "grad_norm": 0.8464507961501435,
+      "learning_rate": 1.6962582716536813e-06,
+      "loss": 0.3299,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6164179104477612,
+      "grad_norm": 0.9279680601344305,
+      "learning_rate": 1.6939696298175245e-06,
+      "loss": 0.3676,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6167164179104477,
+      "grad_norm": 0.9394401389859991,
+      "learning_rate": 1.6916817416001175e-06,
+      "loss": 0.3252,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6170149253731343,
+      "grad_norm": 0.8255999859088444,
+      "learning_rate": 1.6893946091405784e-06,
+      "loss": 0.3492,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6173134328358209,
+      "grad_norm": 0.8393914123113455,
+      "learning_rate": 1.6871082345773215e-06,
+      "loss": 0.326,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6176119402985074,
+      "grad_norm": 0.7540269358482177,
+      "learning_rate": 1.6848226200480514e-06,
+      "loss": 0.3164,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6179104477611941,
+      "grad_norm": 0.8407057610257092,
+      "learning_rate": 1.6825377676897608e-06,
+      "loss": 0.3426,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6182089552238806,
+      "grad_norm": 0.9601578918299282,
+      "learning_rate": 1.6802536796387328e-06,
+      "loss": 0.4112,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6185074626865672,
+      "grad_norm": 0.8027639919807739,
+      "learning_rate": 1.6779703580305323e-06,
+      "loss": 0.3493,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6188059701492538,
+      "grad_norm": 0.8391959617782264,
+      "learning_rate": 1.6756878050000098e-06,
+      "loss": 0.3591,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6191044776119403,
+      "grad_norm": 0.7871628742456087,
+      "learning_rate": 1.6734060226812976e-06,
+      "loss": 0.3414,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6194029850746269,
+      "grad_norm": 0.8096322967159341,
+      "learning_rate": 1.6711250132078055e-06,
+      "loss": 0.3215,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6197014925373134,
+      "grad_norm": 0.8138284215248026,
+      "learning_rate": 1.6688447787122226e-06,
+      "loss": 0.3173,
+      "step": 2076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8329557552536543,
+      "learning_rate": 1.666565321326512e-06,
+      "loss": 0.301,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6202985074626866,
+      "grad_norm": 0.8459261392156018,
+      "learning_rate": 1.6642866431819107e-06,
+      "loss": 0.3869,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6205970149253731,
+      "grad_norm": 0.7415893886114104,
+      "learning_rate": 1.6620087464089275e-06,
+      "loss": 0.3512,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6208955223880597,
+      "grad_norm": 0.8509533039291013,
+      "learning_rate": 1.659731633137341e-06,
+      "loss": 0.356,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6211940298507462,
+      "grad_norm": 0.7465210952477122,
+      "learning_rate": 1.657455305496195e-06,
+      "loss": 0.3329,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6214925373134328,
+      "grad_norm": 0.863374891709103,
+      "learning_rate": 1.6551797656138018e-06,
+      "loss": 0.3296,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6217910447761194,
+      "grad_norm": 0.8457477256461274,
+      "learning_rate": 1.6529050156177356e-06,
+      "loss": 0.3386,
+      "step": 2083
+    },
+    {
+      "epoch": 0.622089552238806,
+      "grad_norm": 0.8417932226084862,
+      "learning_rate": 1.650631057634831e-06,
+      "loss": 0.3294,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6223880597014926,
+      "grad_norm": 0.7940353343813964,
+      "learning_rate": 1.6483578937911836e-06,
+      "loss": 0.3573,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6226865671641791,
+      "grad_norm": 0.8984545759070848,
+      "learning_rate": 1.6460855262121479e-06,
+      "loss": 0.335,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6229850746268657,
+      "grad_norm": 0.9689021624736345,
+      "learning_rate": 1.6438139570223311e-06,
+      "loss": 0.3637,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6232835820895523,
+      "grad_norm": 0.7948610205026269,
+      "learning_rate": 1.641543188345594e-06,
+      "loss": 0.318,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6235820895522388,
+      "grad_norm": 0.8388231085500125,
+      "learning_rate": 1.6392732223050515e-06,
+      "loss": 0.3366,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6238805970149254,
+      "grad_norm": 0.9034038641733735,
+      "learning_rate": 1.6370040610230662e-06,
+      "loss": 0.3811,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6241791044776119,
+      "grad_norm": 0.8056850552409819,
+      "learning_rate": 1.6347357066212478e-06,
+      "loss": 0.3246,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6244776119402985,
+      "grad_norm": 0.8924171728248095,
+      "learning_rate": 1.6324681612204527e-06,
+      "loss": 0.3099,
+      "step": 2092
+    },
+    {
+      "epoch": 0.624776119402985,
+      "grad_norm": 0.8889195100531483,
+      "learning_rate": 1.6302014269407812e-06,
+      "loss": 0.3916,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6250746268656716,
+      "grad_norm": 0.802744288589964,
+      "learning_rate": 1.6279355059015739e-06,
+      "loss": 0.3216,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6253731343283582,
+      "grad_norm": 0.8244538019031434,
+      "learning_rate": 1.6256704002214124e-06,
+      "loss": 0.3436,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6256716417910447,
+      "grad_norm": 0.8806302935844983,
+      "learning_rate": 1.6234061120181144e-06,
+      "loss": 0.3234,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6259701492537313,
+      "grad_norm": 0.7542539098364519,
+      "learning_rate": 1.6211426434087347e-06,
+      "loss": 0.3112,
+      "step": 2097
+    },
+    {
+      "epoch": 0.626268656716418,
+      "grad_norm": 0.8779997865446432,
+      "learning_rate": 1.6188799965095614e-06,
+      "loss": 0.3213,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6265671641791045,
+      "grad_norm": 0.8512972841056398,
+      "learning_rate": 1.6166181734361136e-06,
+      "loss": 0.3806,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 0.780603070362957,
+      "learning_rate": 1.6143571763031413e-06,
+      "loss": 0.3381,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6271641791044776,
+      "grad_norm": 0.8792815515680216,
+      "learning_rate": 1.6120970072246222e-06,
+      "loss": 0.3795,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6274626865671642,
+      "grad_norm": 0.8439949848861296,
+      "learning_rate": 1.6098376683137582e-06,
+      "loss": 0.2953,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6277611940298508,
+      "grad_norm": 0.9313525702028004,
+      "learning_rate": 1.6075791616829764e-06,
+      "loss": 0.3475,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6280597014925373,
+      "grad_norm": 0.7518808217696689,
+      "learning_rate": 1.6053214894439262e-06,
+      "loss": 0.3147,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6283582089552239,
+      "grad_norm": 0.7506483709151667,
+      "learning_rate": 1.6030646537074751e-06,
+      "loss": 0.3167,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6286567164179104,
+      "grad_norm": 0.8052425362278406,
+      "learning_rate": 1.6008086565837106e-06,
+      "loss": 0.2981,
+      "step": 2106
+    },
+    {
+      "epoch": 0.628955223880597,
+      "grad_norm": 0.8210002191055628,
+      "learning_rate": 1.5985535001819335e-06,
+      "loss": 0.3501,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6292537313432836,
+      "grad_norm": 0.8941400826776075,
+      "learning_rate": 1.596299186610661e-06,
+      "loss": 0.3456,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6295522388059701,
+      "grad_norm": 1.0083217537161828,
+      "learning_rate": 1.594045717977622e-06,
+      "loss": 0.3841,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6298507462686567,
+      "grad_norm": 1.078514382968689,
+      "learning_rate": 1.5917930963897534e-06,
+      "loss": 0.3614,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6301492537313432,
+      "grad_norm": 0.7917547916905323,
+      "learning_rate": 1.5895413239532022e-06,
+      "loss": 0.3412,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6304477611940299,
+      "grad_norm": 0.7691216432708953,
+      "learning_rate": 1.5872904027733211e-06,
+      "loss": 0.3178,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6307462686567165,
+      "grad_norm": 0.77802800862977,
+      "learning_rate": 1.5850403349546655e-06,
+      "loss": 0.309,
+      "step": 2113
+    },
+    {
+      "epoch": 0.631044776119403,
+      "grad_norm": 0.8255197315967899,
+      "learning_rate": 1.5827911226009945e-06,
+      "loss": 0.3138,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6313432835820896,
+      "grad_norm": 0.8386811603427105,
+      "learning_rate": 1.5805427678152677e-06,
+      "loss": 0.2891,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6316417910447761,
+      "grad_norm": 0.8153875110996863,
+      "learning_rate": 1.5782952726996403e-06,
+      "loss": 0.2995,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6319402985074627,
+      "grad_norm": 0.8409552206683609,
+      "learning_rate": 1.5760486393554667e-06,
+      "loss": 0.3386,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6322388059701493,
+      "grad_norm": 0.8250066653051528,
+      "learning_rate": 1.5738028698832942e-06,
+      "loss": 0.3827,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6325373134328358,
+      "grad_norm": 0.8610197982443525,
+      "learning_rate": 1.5715579663828618e-06,
+      "loss": 0.3365,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6328358208955224,
+      "grad_norm": 0.9413453392741797,
+      "learning_rate": 1.5693139309531006e-06,
+      "loss": 0.3692,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6331343283582089,
+      "grad_norm": 0.818537165590243,
+      "learning_rate": 1.567070765692128e-06,
+      "loss": 0.3237,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6334328358208955,
+      "grad_norm": 0.8210467581037414,
+      "learning_rate": 1.5648284726972491e-06,
+      "loss": 0.3422,
+      "step": 2122
+    },
+    {
+      "epoch": 0.633731343283582,
+      "grad_norm": 0.8300643168242267,
+      "learning_rate": 1.562587054064953e-06,
+      "loss": 0.3112,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6340298507462686,
+      "grad_norm": 0.8063304103666424,
+      "learning_rate": 1.5603465118909122e-06,
+      "loss": 0.3277,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6343283582089553,
+      "grad_norm": 0.7841383009639938,
+      "learning_rate": 1.5581068482699797e-06,
+      "loss": 0.3132,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6346268656716418,
+      "grad_norm": 0.7663654715100157,
+      "learning_rate": 1.5558680652961844e-06,
+      "loss": 0.3549,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6349253731343284,
+      "grad_norm": 0.716788049651756,
+      "learning_rate": 1.553630165062735e-06,
+      "loss": 0.3293,
+      "step": 2127
+    },
+    {
+      "epoch": 0.635223880597015,
+      "grad_norm": 0.8376097154978661,
+      "learning_rate": 1.5513931496620138e-06,
+      "loss": 0.348,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6355223880597015,
+      "grad_norm": 0.7844875450357711,
+      "learning_rate": 1.549157021185575e-06,
+      "loss": 0.3238,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6358208955223881,
+      "grad_norm": 0.7845014380845483,
+      "learning_rate": 1.546921781724145e-06,
+      "loss": 0.3272,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6361194029850746,
+      "grad_norm": 0.8084890497779735,
+      "learning_rate": 1.544687433367617e-06,
+      "loss": 0.3503,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6364179104477612,
+      "grad_norm": 1.155872428085641,
+      "learning_rate": 1.5424539782050535e-06,
+      "loss": 0.3463,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6367164179104478,
+      "grad_norm": 0.8237151552290364,
+      "learning_rate": 1.5402214183246805e-06,
+      "loss": 0.3751,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6370149253731343,
+      "grad_norm": 0.8146345074095411,
+      "learning_rate": 1.5379897558138861e-06,
+      "loss": 0.3444,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6373134328358209,
+      "grad_norm": 0.7508094621607114,
+      "learning_rate": 1.5357589927592211e-06,
+      "loss": 0.2862,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6376119402985074,
+      "grad_norm": 0.7803467497000177,
+      "learning_rate": 1.5335291312463952e-06,
+      "loss": 0.2941,
+      "step": 2136
+    },
+    {
+      "epoch": 0.637910447761194,
+      "grad_norm": 0.7746140356066684,
+      "learning_rate": 1.531300173360273e-06,
+      "loss": 0.338,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6382089552238805,
+      "grad_norm": 0.7466760142466666,
+      "learning_rate": 1.5290721211848767e-06,
+      "loss": 0.2746,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6385074626865672,
+      "grad_norm": 0.8289320396750041,
+      "learning_rate": 1.5268449768033811e-06,
+      "loss": 0.3639,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6388059701492538,
+      "grad_norm": 0.7786271838202594,
+      "learning_rate": 1.5246187422981113e-06,
+      "loss": 0.3195,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6391044776119403,
+      "grad_norm": 0.7961531887157887,
+      "learning_rate": 1.522393419750542e-06,
+      "loss": 0.3472,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6394029850746269,
+      "grad_norm": 0.7553655687048026,
+      "learning_rate": 1.520169011241297e-06,
+      "loss": 0.3181,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6397014925373135,
+      "grad_norm": 0.8791350108013017,
+      "learning_rate": 1.5179455188501425e-06,
+      "loss": 0.3573,
+      "step": 2143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.791573138903334,
+      "learning_rate": 1.5157229446559903e-06,
+      "loss": 0.3146,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6402985074626866,
+      "grad_norm": 0.7982451287466451,
+      "learning_rate": 1.5135012907368928e-06,
+      "loss": 0.3238,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6405970149253731,
+      "grad_norm": 0.9547736379566139,
+      "learning_rate": 1.5112805591700426e-06,
+      "loss": 0.3464,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6408955223880597,
+      "grad_norm": 0.9389426478984015,
+      "learning_rate": 1.50906075203177e-06,
+      "loss": 0.3831,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6411940298507462,
+      "grad_norm": 0.7559187426695012,
+      "learning_rate": 1.5068418713975397e-06,
+      "loss": 0.2961,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6414925373134328,
+      "grad_norm": 0.8009593103452896,
+      "learning_rate": 1.5046239193419514e-06,
+      "loss": 0.3345,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6417910447761194,
+      "grad_norm": 0.8808029291340134,
+      "learning_rate": 1.5024068979387365e-06,
+      "loss": 0.3622,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6420895522388059,
+      "grad_norm": 0.8048602418552092,
+      "learning_rate": 1.5001908092607553e-06,
+      "loss": 0.3272,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6423880597014925,
+      "grad_norm": 0.886841416239724,
+      "learning_rate": 1.497975655379997e-06,
+      "loss": 0.3637,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6426865671641792,
+      "grad_norm": 0.8501760322071722,
+      "learning_rate": 1.495761438367577e-06,
+      "loss": 0.3315,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6429850746268657,
+      "grad_norm": 0.8499429690154281,
+      "learning_rate": 1.4935481602937334e-06,
+      "loss": 0.3519,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6432835820895523,
+      "grad_norm": 0.8164800240232427,
+      "learning_rate": 1.4913358232278274e-06,
+      "loss": 0.327,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6435820895522388,
+      "grad_norm": 0.8053874302696963,
+      "learning_rate": 1.4891244292383412e-06,
+      "loss": 0.3362,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6438805970149254,
+      "grad_norm": 0.7798998762739349,
+      "learning_rate": 1.4869139803928727e-06,
+      "loss": 0.3538,
+      "step": 2157
+    },
+    {
+      "epoch": 0.644179104477612,
+      "grad_norm": 0.925044008666226,
+      "learning_rate": 1.4847044787581395e-06,
+      "loss": 0.3266,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6444776119402985,
+      "grad_norm": 0.8652309852841852,
+      "learning_rate": 1.48249592639997e-06,
+      "loss": 0.3687,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6447761194029851,
+      "grad_norm": 0.7856735870745902,
+      "learning_rate": 1.4802883253833073e-06,
+      "loss": 0.2956,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6450746268656716,
+      "grad_norm": 0.7855897694933878,
+      "learning_rate": 1.4780816777722052e-06,
+      "loss": 0.3162,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6453731343283582,
+      "grad_norm": 0.808546264261996,
+      "learning_rate": 1.4758759856298248e-06,
+      "loss": 0.3095,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6456716417910447,
+      "grad_norm": 0.8076290541474285,
+      "learning_rate": 1.4736712510184359e-06,
+      "loss": 0.3276,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6459701492537313,
+      "grad_norm": 0.7968186141075612,
+      "learning_rate": 1.4714674759994096e-06,
+      "loss": 0.2801,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6462686567164179,
+      "grad_norm": 0.7847934498100252,
+      "learning_rate": 1.4692646626332228e-06,
+      "loss": 0.3368,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6465671641791044,
+      "grad_norm": 0.9749102135271982,
+      "learning_rate": 1.4670628129794527e-06,
+      "loss": 0.3638,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6468656716417911,
+      "grad_norm": 0.7325101572568912,
+      "learning_rate": 1.4648619290967742e-06,
+      "loss": 0.3243,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6471641791044777,
+      "grad_norm": 0.9104372675901333,
+      "learning_rate": 1.4626620130429608e-06,
+      "loss": 0.3549,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6474626865671642,
+      "grad_norm": 0.7969438524329776,
+      "learning_rate": 1.4604630668748795e-06,
+      "loss": 0.3465,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6477611940298508,
+      "grad_norm": 0.8828900184898577,
+      "learning_rate": 1.4582650926484921e-06,
+      "loss": 0.3556,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6480597014925373,
+      "grad_norm": 0.798311888333371,
+      "learning_rate": 1.4560680924188508e-06,
+      "loss": 0.3184,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6483582089552239,
+      "grad_norm": 0.860792245744944,
+      "learning_rate": 1.453872068240097e-06,
+      "loss": 0.3363,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6486567164179105,
+      "grad_norm": 0.7649776213408986,
+      "learning_rate": 1.4516770221654614e-06,
+      "loss": 0.3355,
+      "step": 2173
+    },
+    {
+      "epoch": 0.648955223880597,
+      "grad_norm": 0.7942691044713359,
+      "learning_rate": 1.4494829562472558e-06,
+      "loss": 0.3341,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6492537313432836,
+      "grad_norm": 0.8621884015676841,
+      "learning_rate": 1.4472898725368795e-06,
+      "loss": 0.3343,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6495522388059701,
+      "grad_norm": 0.7924705469899432,
+      "learning_rate": 1.4450977730848132e-06,
+      "loss": 0.3275,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6498507462686567,
+      "grad_norm": 1.2686960712118278,
+      "learning_rate": 1.4429066599406152e-06,
+      "loss": 0.3576,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6501492537313432,
+      "grad_norm": 0.7074769773056945,
+      "learning_rate": 1.4407165351529246e-06,
+      "loss": 0.2896,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6504477611940298,
+      "grad_norm": 0.8333612223546273,
+      "learning_rate": 1.4385274007694527e-06,
+      "loss": 0.3181,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6507462686567164,
+      "grad_norm": 0.8657347511763104,
+      "learning_rate": 1.4363392588369876e-06,
+      "loss": 0.3312,
+      "step": 2180
+    },
+    {
+      "epoch": 0.651044776119403,
+      "grad_norm": 0.7696448268786413,
+      "learning_rate": 1.4341521114013888e-06,
+      "loss": 0.3079,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6513432835820896,
+      "grad_norm": 0.8733955547423778,
+      "learning_rate": 1.4319659605075855e-06,
+      "loss": 0.3489,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6516417910447762,
+      "grad_norm": 1.079444796750812,
+      "learning_rate": 1.429780808199576e-06,
+      "loss": 0.3833,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6519402985074627,
+      "grad_norm": 0.7722744529314012,
+      "learning_rate": 1.4275966565204251e-06,
+      "loss": 0.3301,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6522388059701493,
+      "grad_norm": 0.8366589681087087,
+      "learning_rate": 1.4254135075122596e-06,
+      "loss": 0.2948,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6525373134328358,
+      "grad_norm": 0.8263300107210038,
+      "learning_rate": 1.4232313632162714e-06,
+      "loss": 0.3113,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6528358208955224,
+      "grad_norm": 0.7782108665859974,
+      "learning_rate": 1.421050225672712e-06,
+      "loss": 0.3098,
+      "step": 2187
+    },
+    {
+      "epoch": 0.653134328358209,
+      "grad_norm": 0.9032704619470634,
+      "learning_rate": 1.418870096920893e-06,
+      "loss": 0.3312,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6534328358208955,
+      "grad_norm": 0.7838544001315462,
+      "learning_rate": 1.4166909789991813e-06,
+      "loss": 0.3091,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6537313432835821,
+      "grad_norm": 0.8249256923116727,
+      "learning_rate": 1.4145128739449974e-06,
+      "loss": 0.3657,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6540298507462686,
+      "grad_norm": 0.7937594433877714,
+      "learning_rate": 1.4123357837948177e-06,
+      "loss": 0.3459,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6543283582089552,
+      "grad_norm": 1.0209366045959114,
+      "learning_rate": 1.410159710584168e-06,
+      "loss": 0.3305,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6546268656716417,
+      "grad_norm": 0.8416577328440887,
+      "learning_rate": 1.4079846563476246e-06,
+      "loss": 0.3428,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6549253731343283,
+      "grad_norm": 0.8381158795554763,
+      "learning_rate": 1.4058106231188093e-06,
+      "loss": 0.3278,
+      "step": 2194
+    },
+    {
+      "epoch": 0.655223880597015,
+      "grad_norm": 0.7994100548783899,
+      "learning_rate": 1.4036376129303914e-06,
+      "loss": 0.3187,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6555223880597015,
+      "grad_norm": 0.8441055293209245,
+      "learning_rate": 1.4014656278140806e-06,
+      "loss": 0.3585,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6558208955223881,
+      "grad_norm": 0.8224053402613939,
+      "learning_rate": 1.399294669800631e-06,
+      "loss": 0.3106,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6561194029850747,
+      "grad_norm": 0.893366800253749,
+      "learning_rate": 1.3971247409198347e-06,
+      "loss": 0.3897,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6564179104477612,
+      "grad_norm": 0.8805627796048343,
+      "learning_rate": 1.3949558432005245e-06,
+      "loss": 0.347,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6567164179104478,
+      "grad_norm": 0.764024546070756,
+      "learning_rate": 1.3927879786705645e-06,
+      "loss": 0.3213,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6570149253731343,
+      "grad_norm": 0.722603696748916,
+      "learning_rate": 1.3906211493568563e-06,
+      "loss": 0.3137,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6573134328358209,
+      "grad_norm": 0.9269224842693098,
+      "learning_rate": 1.3884553572853324e-06,
+      "loss": 0.3678,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6576119402985074,
+      "grad_norm": 0.7631786113345267,
+      "learning_rate": 1.3862906044809554e-06,
+      "loss": 0.314,
+      "step": 2203
+    },
+    {
+      "epoch": 0.657910447761194,
+      "grad_norm": 0.7773514605842012,
+      "learning_rate": 1.3841268929677165e-06,
+      "loss": 0.3133,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6582089552238806,
+      "grad_norm": 0.7946457192152244,
+      "learning_rate": 1.381964224768634e-06,
+      "loss": 0.3181,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6585074626865671,
+      "grad_norm": 0.8529990432133921,
+      "learning_rate": 1.3798026019057483e-06,
+      "loss": 0.3426,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6588059701492537,
+      "grad_norm": 0.9216515652907142,
+      "learning_rate": 1.377642026400125e-06,
+      "loss": 0.3643,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6591044776119404,
+      "grad_norm": 0.832800795535474,
+      "learning_rate": 1.3754825002718498e-06,
+      "loss": 0.3629,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6594029850746269,
+      "grad_norm": 0.990654044032596,
+      "learning_rate": 1.3733240255400263e-06,
+      "loss": 0.3506,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6597014925373135,
+      "grad_norm": 0.6854970183511663,
+      "learning_rate": 1.3711666042227772e-06,
+      "loss": 0.3319,
+      "step": 2210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8428108856258283,
+      "learning_rate": 1.3690102383372369e-06,
+      "loss": 0.2742,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6602985074626866,
+      "grad_norm": 0.8741591762317693,
+      "learning_rate": 1.3668549298995558e-06,
+      "loss": 0.3687,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6605970149253731,
+      "grad_norm": 0.7955836464506277,
+      "learning_rate": 1.3647006809248947e-06,
+      "loss": 0.3032,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6608955223880597,
+      "grad_norm": 0.8711695616206405,
+      "learning_rate": 1.3625474934274241e-06,
+      "loss": 0.381,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6611940298507463,
+      "grad_norm": 0.8401790125361631,
+      "learning_rate": 1.360395369420322e-06,
+      "loss": 0.3229,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6614925373134328,
+      "grad_norm": 0.8676846365068158,
+      "learning_rate": 1.3582443109157722e-06,
+      "loss": 0.3615,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6617910447761194,
+      "grad_norm": 0.7929648763121615,
+      "learning_rate": 1.3560943199249605e-06,
+      "loss": 0.3588,
+      "step": 2217
+    },
+    {
+      "epoch": 0.662089552238806,
+      "grad_norm": 0.9511947801173409,
+      "learning_rate": 1.3539453984580767e-06,
+      "loss": 0.3105,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6623880597014925,
+      "grad_norm": 0.7404356758215146,
+      "learning_rate": 1.3517975485243103e-06,
+      "loss": 0.2941,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6626865671641791,
+      "grad_norm": 1.0227800450584272,
+      "learning_rate": 1.3496507721318486e-06,
+      "loss": 0.3296,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6629850746268656,
+      "grad_norm": 0.7690958302654547,
+      "learning_rate": 1.3475050712878755e-06,
+      "loss": 0.321,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6632835820895523,
+      "grad_norm": 0.8740712288587946,
+      "learning_rate": 1.345360447998569e-06,
+      "loss": 0.3596,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6635820895522389,
+      "grad_norm": 0.8474627076368212,
+      "learning_rate": 1.3432169042690988e-06,
+      "loss": 0.3634,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6638805970149254,
+      "grad_norm": 0.9253686814199646,
+      "learning_rate": 1.3410744421036262e-06,
+      "loss": 0.2892,
+      "step": 2224
+    },
+    {
+      "epoch": 0.664179104477612,
+      "grad_norm": 0.8348049414436733,
+      "learning_rate": 1.3389330635053013e-06,
+      "loss": 0.3515,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6644776119402985,
+      "grad_norm": 0.822846204285828,
+      "learning_rate": 1.3367927704762613e-06,
+      "loss": 0.3425,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6647761194029851,
+      "grad_norm": 0.8678200777615288,
+      "learning_rate": 1.3346535650176284e-06,
+      "loss": 0.3978,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6650746268656716,
+      "grad_norm": 0.7358960864836035,
+      "learning_rate": 1.3325154491295062e-06,
+      "loss": 0.2916,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6653731343283582,
+      "grad_norm": 0.7423135388554165,
+      "learning_rate": 1.330378424810981e-06,
+      "loss": 0.3068,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6656716417910448,
+      "grad_norm": 0.8230255273207777,
+      "learning_rate": 1.3282424940601197e-06,
+      "loss": 0.355,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6659701492537313,
+      "grad_norm": 0.8023837250650214,
+      "learning_rate": 1.326107658873964e-06,
+      "loss": 0.2787,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6662686567164179,
+      "grad_norm": 0.7899498370106252,
+      "learning_rate": 1.3239739212485342e-06,
+      "loss": 0.2977,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6665671641791044,
+      "grad_norm": 0.9616260736960074,
+      "learning_rate": 1.3218412831788232e-06,
+      "loss": 0.3403,
+      "step": 2233
+    },
+    {
+      "epoch": 0.666865671641791,
+      "grad_norm": 0.9674895085633961,
+      "learning_rate": 1.3197097466587939e-06,
+      "loss": 0.3681,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6671641791044776,
+      "grad_norm": 0.808185104408685,
+      "learning_rate": 1.317579313681382e-06,
+      "loss": 0.3109,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6674626865671642,
+      "grad_norm": 0.8737274124495528,
+      "learning_rate": 1.315449986238489e-06,
+      "loss": 0.3471,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6677611940298508,
+      "grad_norm": 0.8084788833507394,
+      "learning_rate": 1.3133217663209873e-06,
+      "loss": 0.3272,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6680597014925374,
+      "grad_norm": 0.8888189135315439,
+      "learning_rate": 1.3111946559187078e-06,
+      "loss": 0.359,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6683582089552239,
+      "grad_norm": 0.8035833032123232,
+      "learning_rate": 1.309068657020448e-06,
+      "loss": 0.3285,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6686567164179105,
+      "grad_norm": 0.9254549420427741,
+      "learning_rate": 1.3069437716139648e-06,
+      "loss": 0.377,
+      "step": 2240
+    },
+    {
+      "epoch": 0.668955223880597,
+      "grad_norm": 0.768533993055729,
+      "learning_rate": 1.3048200016859743e-06,
+      "loss": 0.3516,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6692537313432836,
+      "grad_norm": 0.7917607581116863,
+      "learning_rate": 1.3026973492221501e-06,
+      "loss": 0.3074,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6695522388059701,
+      "grad_norm": 0.8087854690730707,
+      "learning_rate": 1.3005758162071206e-06,
+      "loss": 0.3164,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6698507462686567,
+      "grad_norm": 0.8650182535397217,
+      "learning_rate": 1.298455404624466e-06,
+      "loss": 0.3296,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6701492537313433,
+      "grad_norm": 0.7997154190905688,
+      "learning_rate": 1.2963361164567207e-06,
+      "loss": 0.3546,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6704477611940298,
+      "grad_norm": 0.9966416873938866,
+      "learning_rate": 1.2942179536853666e-06,
+      "loss": 0.3685,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6707462686567164,
+      "grad_norm": 0.8148532817098773,
+      "learning_rate": 1.2921009182908351e-06,
+      "loss": 0.3242,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6710447761194029,
+      "grad_norm": 0.75826379276025,
+      "learning_rate": 1.2899850122525037e-06,
+      "loss": 0.2655,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6713432835820895,
+      "grad_norm": 0.9350771859645716,
+      "learning_rate": 1.2878702375486905e-06,
+      "loss": 0.3477,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 0.7895404077047398,
+      "learning_rate": 1.2857565961566593e-06,
+      "loss": 0.3619,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6719402985074627,
+      "grad_norm": 0.956687016013122,
+      "learning_rate": 1.2836440900526136e-06,
+      "loss": 0.3528,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6722388059701493,
+      "grad_norm": 0.8027731716484646,
+      "learning_rate": 1.281532721211695e-06,
+      "loss": 0.3426,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6725373134328358,
+      "grad_norm": 0.8683386280279184,
+      "learning_rate": 1.2794224916079818e-06,
+      "loss": 0.3624,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6728358208955224,
+      "grad_norm": 0.8021378966828665,
+      "learning_rate": 1.2773134032144885e-06,
+      "loss": 0.3344,
+      "step": 2254
+    },
+    {
+      "epoch": 0.673134328358209,
+      "grad_norm": 0.822448080875513,
+      "learning_rate": 1.2752054580031592e-06,
+      "loss": 0.3012,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6734328358208955,
+      "grad_norm": 0.8066506836401879,
+      "learning_rate": 1.2730986579448719e-06,
+      "loss": 0.3137,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6737313432835821,
+      "grad_norm": 0.7452823368578445,
+      "learning_rate": 1.2709930050094337e-06,
+      "loss": 0.3389,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6740298507462686,
+      "grad_norm": 0.7510246741662416,
+      "learning_rate": 1.268888501165579e-06,
+      "loss": 0.2977,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6743283582089552,
+      "grad_norm": 0.8248092475825504,
+      "learning_rate": 1.2667851483809673e-06,
+      "loss": 0.3085,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6746268656716418,
+      "grad_norm": 0.8232298732269487,
+      "learning_rate": 1.264682948622183e-06,
+      "loss": 0.3586,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6749253731343283,
+      "grad_norm": 0.87600455902864,
+      "learning_rate": 1.2625819038547302e-06,
+      "loss": 0.3549,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6752238805970149,
+      "grad_norm": 0.8569153532990559,
+      "learning_rate": 1.2604820160430348e-06,
+      "loss": 0.3223,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6755223880597014,
+      "grad_norm": 0.8313060298524098,
+      "learning_rate": 1.2583832871504415e-06,
+      "loss": 0.3437,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6758208955223881,
+      "grad_norm": 0.8731999146655798,
+      "learning_rate": 1.25628571913921e-06,
+      "loss": 0.3928,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6761194029850747,
+      "grad_norm": 0.830291452853221,
+      "learning_rate": 1.2541893139705162e-06,
+      "loss": 0.3391,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6764179104477612,
+      "grad_norm": 0.7748207330310743,
+      "learning_rate": 1.2520940736044468e-06,
+      "loss": 0.3191,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6767164179104478,
+      "grad_norm": 0.7873823715976541,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.3354,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6770149253731343,
+      "grad_norm": 0.8238073008584107,
+      "learning_rate": 1.2479070951150857e-06,
+      "loss": 0.3722,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6773134328358209,
+      "grad_norm": 0.7684716974580467,
+      "learning_rate": 1.245815360906517e-06,
+      "loss": 0.3427,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6776119402985075,
+      "grad_norm": 0.9061590212050163,
+      "learning_rate": 1.2437247993300147e-06,
+      "loss": 0.3099,
+      "step": 2270
+    },
+    {
+      "epoch": 0.677910447761194,
+      "grad_norm": 0.8775458433192322,
+      "learning_rate": 1.2416354123402047e-06,
+      "loss": 0.3393,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6782089552238806,
+      "grad_norm": 0.8109025654110851,
+      "learning_rate": 1.2395472018906102e-06,
+      "loss": 0.3608,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6785074626865671,
+      "grad_norm": 0.8511617196960201,
+      "learning_rate": 1.2374601699336586e-06,
+      "loss": 0.3599,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6788059701492537,
+      "grad_norm": 1.2567441350562456,
+      "learning_rate": 1.2353743184206724e-06,
+      "loss": 0.3313,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6791044776119403,
+      "grad_norm": 0.7862840705237462,
+      "learning_rate": 1.2332896493018753e-06,
+      "loss": 0.323,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6794029850746268,
+      "grad_norm": 0.7520760436868094,
+      "learning_rate": 1.231206164526379e-06,
+      "loss": 0.3234,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6797014925373134,
+      "grad_norm": 0.7862063406522266,
+      "learning_rate": 1.229123866042192e-06,
+      "loss": 0.3108,
+      "step": 2277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7931130898996858,
+      "learning_rate": 1.2270427557962125e-06,
+      "loss": 0.2967,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6802985074626866,
+      "grad_norm": 0.8831726582888332,
+      "learning_rate": 1.2249628357342284e-06,
+      "loss": 0.3366,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6805970149253732,
+      "grad_norm": 0.8362589161672496,
+      "learning_rate": 1.2228841078009133e-06,
+      "loss": 0.32,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6808955223880597,
+      "grad_norm": 0.7825482021832821,
+      "learning_rate": 1.2208065739398286e-06,
+      "loss": 0.3236,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6811940298507463,
+      "grad_norm": 0.7726512675817263,
+      "learning_rate": 1.2187302360934162e-06,
+      "loss": 0.3659,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6814925373134328,
+      "grad_norm": 0.8087551514263144,
+      "learning_rate": 1.216655096203002e-06,
+      "loss": 0.3497,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6817910447761194,
+      "grad_norm": 0.8321435088465757,
+      "learning_rate": 1.214581156208791e-06,
+      "loss": 0.3367,
+      "step": 2284
+    },
+    {
+      "epoch": 0.682089552238806,
+      "grad_norm": 0.8973773653830277,
+      "learning_rate": 1.2125084180498672e-06,
+      "loss": 0.3681,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6823880597014925,
+      "grad_norm": 0.7645793362757674,
+      "learning_rate": 1.2104368836641909e-06,
+      "loss": 0.3385,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6826865671641791,
+      "grad_norm": 0.7860470543681153,
+      "learning_rate": 1.208366554988594e-06,
+      "loss": 0.3102,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6829850746268656,
+      "grad_norm": 0.7711760963359957,
+      "learning_rate": 1.2062974339587844e-06,
+      "loss": 0.3456,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6832835820895522,
+      "grad_norm": 0.8535820450749667,
+      "learning_rate": 1.2042295225093404e-06,
+      "loss": 0.3325,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6835820895522388,
+      "grad_norm": 0.849860239170701,
+      "learning_rate": 1.202162822573708e-06,
+      "loss": 0.3324,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6838805970149253,
+      "grad_norm": 1.3404957810983344,
+      "learning_rate": 1.2000973360842019e-06,
+      "loss": 0.3833,
+      "step": 2291
+    },
+    {
+      "epoch": 0.684179104477612,
+      "grad_norm": 0.819656752872545,
+      "learning_rate": 1.1980330649720024e-06,
+      "loss": 0.3346,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6844776119402985,
+      "grad_norm": 0.9829794779330354,
+      "learning_rate": 1.1959700111671508e-06,
+      "loss": 0.3462,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6847761194029851,
+      "grad_norm": 0.7859290591207186,
+      "learning_rate": 1.1939081765985528e-06,
+      "loss": 0.291,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6850746268656717,
+      "grad_norm": 0.810647132569844,
+      "learning_rate": 1.1918475631939735e-06,
+      "loss": 0.3525,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6853731343283582,
+      "grad_norm": 0.8879767152876183,
+      "learning_rate": 1.1897881728800364e-06,
+      "loss": 0.3398,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6856716417910448,
+      "grad_norm": 0.8081639689735962,
+      "learning_rate": 1.1877300075822223e-06,
+      "loss": 0.3221,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6859701492537313,
+      "grad_norm": 0.8175265918913039,
+      "learning_rate": 1.1856730692248635e-06,
+      "loss": 0.3699,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6862686567164179,
+      "grad_norm": 0.7821789581867407,
+      "learning_rate": 1.1836173597311484e-06,
+      "loss": 0.316,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6865671641791045,
+      "grad_norm": 0.7996767148379933,
+      "learning_rate": 1.1815628810231147e-06,
+      "loss": 0.3795,
+      "step": 2300
+    },
+    {
+      "epoch": 0.686865671641791,
+      "grad_norm": 0.8382268679790396,
+      "learning_rate": 1.1795096350216506e-06,
+      "loss": 0.3339,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6871641791044776,
+      "grad_norm": 0.8667351459692307,
+      "learning_rate": 1.1774576236464906e-06,
+      "loss": 0.3781,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6874626865671641,
+      "grad_norm": 0.7891693977056329,
+      "learning_rate": 1.1754068488162165e-06,
+      "loss": 0.2936,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6877611940298507,
+      "grad_norm": 0.7802694773636687,
+      "learning_rate": 1.1733573124482505e-06,
+      "loss": 0.3155,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6880597014925374,
+      "grad_norm": 0.7825736393246602,
+      "learning_rate": 1.1713090164588608e-06,
+      "loss": 0.3262,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6883582089552239,
+      "grad_norm": 0.7824674762625962,
+      "learning_rate": 1.1692619627631539e-06,
+      "loss": 0.3152,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6886567164179105,
+      "grad_norm": 0.7859681853056969,
+      "learning_rate": 1.1672161532750749e-06,
+      "loss": 0.2974,
+      "step": 2307
+    },
+    {
+      "epoch": 0.688955223880597,
+      "grad_norm": 0.7718704768344236,
+      "learning_rate": 1.1651715899074057e-06,
+      "loss": 0.3303,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6892537313432836,
+      "grad_norm": 1.0009242400338725,
+      "learning_rate": 1.1631282745717646e-06,
+      "loss": 0.3184,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6895522388059702,
+      "grad_norm": 0.7836521432415025,
+      "learning_rate": 1.1610862091785993e-06,
+      "loss": 0.3612,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6898507462686567,
+      "grad_norm": 0.8471632552947357,
+      "learning_rate": 1.1590453956371925e-06,
+      "loss": 0.2925,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6901492537313433,
+      "grad_norm": 0.7698745533567938,
+      "learning_rate": 1.157005835855654e-06,
+      "loss": 0.3735,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6904477611940298,
+      "grad_norm": 0.7805344892196837,
+      "learning_rate": 1.1549675317409254e-06,
+      "loss": 0.3117,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6907462686567164,
+      "grad_norm": 0.8027470646073804,
+      "learning_rate": 1.1529304851987685e-06,
+      "loss": 0.3219,
+      "step": 2314
+    },
+    {
+      "epoch": 0.691044776119403,
+      "grad_norm": 0.8673613890056963,
+      "learning_rate": 1.1508946981337729e-06,
+      "loss": 0.3615,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6913432835820895,
+      "grad_norm": 0.7571748507097132,
+      "learning_rate": 1.1488601724493504e-06,
+      "loss": 0.3159,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6916417910447761,
+      "grad_norm": 0.8645844258196447,
+      "learning_rate": 1.1468269100477322e-06,
+      "loss": 0.3551,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6919402985074626,
+      "grad_norm": 0.7659538316934659,
+      "learning_rate": 1.1447949128299695e-06,
+      "loss": 0.34,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6922388059701493,
+      "grad_norm": 0.9851936264003477,
+      "learning_rate": 1.1427641826959313e-06,
+      "loss": 0.3599,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6925373134328359,
+      "grad_norm": 0.8517192723805389,
+      "learning_rate": 1.1407347215442985e-06,
+      "loss": 0.3335,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6928358208955224,
+      "grad_norm": 0.8252636410104004,
+      "learning_rate": 1.138706531272569e-06,
+      "loss": 0.3648,
+      "step": 2321
+    },
+    {
+      "epoch": 0.693134328358209,
+      "grad_norm": 0.7905456222424458,
+      "learning_rate": 1.1366796137770512e-06,
+      "loss": 0.3108,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6934328358208955,
+      "grad_norm": 0.8262859024812593,
+      "learning_rate": 1.1346539709528634e-06,
+      "loss": 0.3211,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6937313432835821,
+      "grad_norm": 0.7940913246786866,
+      "learning_rate": 1.1326296046939334e-06,
+      "loss": 0.3616,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6940298507462687,
+      "grad_norm": 0.9195285544655638,
+      "learning_rate": 1.1306065168929925e-06,
+      "loss": 0.3563,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6943283582089552,
+      "grad_norm": 0.7768973786157898,
+      "learning_rate": 1.1285847094415792e-06,
+      "loss": 0.3046,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6946268656716418,
+      "grad_norm": 0.8076162534794478,
+      "learning_rate": 1.126564184230034e-06,
+      "loss": 0.3552,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6949253731343283,
+      "grad_norm": 0.7569456200327547,
+      "learning_rate": 1.1245449431474994e-06,
+      "loss": 0.3296,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6952238805970149,
+      "grad_norm": 0.8337061109406402,
+      "learning_rate": 1.1225269880819158e-06,
+      "loss": 0.2923,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6955223880597015,
+      "grad_norm": 0.8392372442467751,
+      "learning_rate": 1.1205103209200238e-06,
+      "loss": 0.3642,
+      "step": 2330
+    },
+    {
+      "epoch": 0.695820895522388,
+      "grad_norm": 0.8358260600131869,
+      "learning_rate": 1.1184949435473556e-06,
+      "loss": 0.328,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6961194029850746,
+      "grad_norm": 0.8295407000320718,
+      "learning_rate": 1.1164808578482405e-06,
+      "loss": 0.3583,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6964179104477612,
+      "grad_norm": 0.8276689046645763,
+      "learning_rate": 1.1144680657058005e-06,
+      "loss": 0.3463,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6967164179104478,
+      "grad_norm": 0.8590053159382256,
+      "learning_rate": 1.1124565690019465e-06,
+      "loss": 0.3184,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6970149253731344,
+      "grad_norm": 0.8071370342595704,
+      "learning_rate": 1.1104463696173798e-06,
+      "loss": 0.3232,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6973134328358209,
+      "grad_norm": 0.8642648814478965,
+      "learning_rate": 1.1084374694315863e-06,
+      "loss": 0.3465,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6976119402985075,
+      "grad_norm": 0.7645757881216615,
+      "learning_rate": 1.1064298703228397e-06,
+      "loss": 0.329,
+      "step": 2337
+    },
+    {
+      "epoch": 0.697910447761194,
+      "grad_norm": 0.8218872469842694,
+      "learning_rate": 1.1044235741681957e-06,
+      "loss": 0.3571,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6982089552238806,
+      "grad_norm": 0.8082592854417281,
+      "learning_rate": 1.102418582843493e-06,
+      "loss": 0.3386,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6985074626865672,
+      "grad_norm": 0.8551551134548198,
+      "learning_rate": 1.100414898223349e-06,
+      "loss": 0.3187,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6988059701492537,
+      "grad_norm": 0.8454776918886312,
+      "learning_rate": 1.0984125221811611e-06,
+      "loss": 0.3226,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6991044776119403,
+      "grad_norm": 0.8274576186330233,
+      "learning_rate": 1.0964114565891005e-06,
+      "loss": 0.339,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6994029850746268,
+      "grad_norm": 0.8865291981715888,
+      "learning_rate": 1.0944117033181151e-06,
+      "loss": 0.2815,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6997014925373134,
+      "grad_norm": 0.7691276300346361,
+      "learning_rate": 1.0924132642379262e-06,
+      "loss": 0.2907,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8838758440348157,
+      "learning_rate": 1.090416141217025e-06,
+      "loss": 0.3687,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7002985074626865,
+      "grad_norm": 0.834499070283479,
+      "learning_rate": 1.0884203361226733e-06,
+      "loss": 0.3357,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7005970149253732,
+      "grad_norm": 0.7837295011852984,
+      "learning_rate": 1.0864258508209008e-06,
+      "loss": 0.3141,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7008955223880597,
+      "grad_norm": 0.9621383837777848,
+      "learning_rate": 1.0844326871765012e-06,
+      "loss": 0.3344,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7011940298507463,
+      "grad_norm": 0.836293896828257,
+      "learning_rate": 1.0824408470530334e-06,
+      "loss": 0.3519,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7014925373134329,
+      "grad_norm": 0.8713650513051289,
+      "learning_rate": 1.0804503323128214e-06,
+      "loss": 0.3365,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7017910447761194,
+      "grad_norm": 0.7793599987932461,
+      "learning_rate": 1.0784611448169482e-06,
+      "loss": 0.3324,
+      "step": 2351
+    },
+    {
+      "epoch": 0.702089552238806,
+      "grad_norm": 0.747029931462924,
+      "learning_rate": 1.0764732864252538e-06,
+      "loss": 0.3049,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7023880597014925,
+      "grad_norm": 0.7301826389896193,
+      "learning_rate": 1.0744867589963378e-06,
+      "loss": 0.2968,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7026865671641791,
+      "grad_norm": 0.80685723875685,
+      "learning_rate": 1.0725015643875553e-06,
+      "loss": 0.3443,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7029850746268657,
+      "grad_norm": 0.8130732259035608,
+      "learning_rate": 1.0705177044550147e-06,
+      "loss": 0.3303,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7032835820895522,
+      "grad_norm": 1.0494784933389165,
+      "learning_rate": 1.0685351810535779e-06,
+      "loss": 0.3631,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7035820895522388,
+      "grad_norm": 0.771095705431318,
+      "learning_rate": 1.0665539960368536e-06,
+      "loss": 0.3097,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7038805970149253,
+      "grad_norm": 0.8622991762796726,
+      "learning_rate": 1.0645741512572031e-06,
+      "loss": 0.3797,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7041791044776119,
+      "grad_norm": 0.8468783834550917,
+      "learning_rate": 1.0625956485657321e-06,
+      "loss": 0.313,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7044776119402985,
+      "grad_norm": 0.7455704245680197,
+      "learning_rate": 1.060618489812293e-06,
+      "loss": 0.3154,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7047761194029851,
+      "grad_norm": 0.8264996079174645,
+      "learning_rate": 1.0586426768454813e-06,
+      "loss": 0.3475,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7050746268656717,
+      "grad_norm": 0.8225831760639863,
+      "learning_rate": 1.0566682115126345e-06,
+      "loss": 0.3354,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7053731343283582,
+      "grad_norm": 0.7582974133085499,
+      "learning_rate": 1.0546950956598276e-06,
+      "loss": 0.3143,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7056716417910448,
+      "grad_norm": 0.837445402977908,
+      "learning_rate": 1.0527233311318768e-06,
+      "loss": 0.3432,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7059701492537314,
+      "grad_norm": 0.8064450995316124,
+      "learning_rate": 1.050752919772334e-06,
+      "loss": 0.3427,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7062686567164179,
+      "grad_norm": 0.7708046546693463,
+      "learning_rate": 1.048783863423486e-06,
+      "loss": 0.2782,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7065671641791045,
+      "grad_norm": 0.7890086390145694,
+      "learning_rate": 1.0468161639263518e-06,
+      "loss": 0.3246,
+      "step": 2367
+    },
+    {
+      "epoch": 0.706865671641791,
+      "grad_norm": 0.9613407733550523,
+      "learning_rate": 1.0448498231206843e-06,
+      "loss": 0.3788,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7071641791044776,
+      "grad_norm": 0.8377614962573282,
+      "learning_rate": 1.0428848428449618e-06,
+      "loss": 0.2917,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7074626865671642,
+      "grad_norm": 0.7644477209892271,
+      "learning_rate": 1.040921224936394e-06,
+      "loss": 0.2539,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7077611940298507,
+      "grad_norm": 0.8073027138319966,
+      "learning_rate": 1.0389589712309156e-06,
+      "loss": 0.3853,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7080597014925373,
+      "grad_norm": 0.773541394858571,
+      "learning_rate": 1.0369980835631862e-06,
+      "loss": 0.3109,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7083582089552238,
+      "grad_norm": 0.8939609435492758,
+      "learning_rate": 1.035038563766589e-06,
+      "loss": 0.3331,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7086567164179104,
+      "grad_norm": 0.8792285135356904,
+      "learning_rate": 1.0330804136732253e-06,
+      "loss": 0.3031,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7089552238805971,
+      "grad_norm": 0.8249719781900057,
+      "learning_rate": 1.0311236351139186e-06,
+      "loss": 0.3463,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7092537313432836,
+      "grad_norm": 0.9446425866318262,
+      "learning_rate": 1.0291682299182092e-06,
+      "loss": 0.3444,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7095522388059702,
+      "grad_norm": 0.8352690148905147,
+      "learning_rate": 1.027214199914353e-06,
+      "loss": 0.3358,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7098507462686567,
+      "grad_norm": 0.7396486359215371,
+      "learning_rate": 1.025261546929321e-06,
+      "loss": 0.3504,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7101492537313433,
+      "grad_norm": 0.8423507780215546,
+      "learning_rate": 1.0233102727887967e-06,
+      "loss": 0.3328,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7104477611940299,
+      "grad_norm": 0.8857400481832675,
+      "learning_rate": 1.0213603793171717e-06,
+      "loss": 0.3581,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7107462686567164,
+      "grad_norm": 0.8154084588304341,
+      "learning_rate": 1.0194118683375502e-06,
+      "loss": 0.3725,
+      "step": 2381
+    },
+    {
+      "epoch": 0.711044776119403,
+      "grad_norm": 0.9479114104669211,
+      "learning_rate": 1.0174647416717428e-06,
+      "loss": 0.3963,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7113432835820895,
+      "grad_norm": 0.8293843508242847,
+      "learning_rate": 1.015519001140265e-06,
+      "loss": 0.3394,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7116417910447761,
+      "grad_norm": 0.8038453888278364,
+      "learning_rate": 1.0135746485623376e-06,
+      "loss": 0.2989,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7119402985074627,
+      "grad_norm": 0.8752892342124312,
+      "learning_rate": 1.0116316857558814e-06,
+      "loss": 0.3621,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7122388059701492,
+      "grad_norm": 0.7967390929681123,
+      "learning_rate": 1.0096901145375197e-06,
+      "loss": 0.3461,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7125373134328358,
+      "grad_norm": 0.8551632337211533,
+      "learning_rate": 1.0077499367225737e-06,
+      "loss": 0.3295,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7128358208955224,
+      "grad_norm": 0.8773095549519137,
+      "learning_rate": 1.0058111541250642e-06,
+      "loss": 0.3562,
+      "step": 2388
+    },
+    {
+      "epoch": 0.713134328358209,
+      "grad_norm": 0.8061003756386974,
+      "learning_rate": 1.0038737685577057e-06,
+      "loss": 0.3063,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7134328358208956,
+      "grad_norm": 0.842089773836329,
+      "learning_rate": 1.0019377818319045e-06,
+      "loss": 0.3467,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7137313432835821,
+      "grad_norm": 0.7906352842135439,
+      "learning_rate": 1.0000031957577618e-06,
+      "loss": 0.3419,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7140298507462687,
+      "grad_norm": 0.9218014121141632,
+      "learning_rate": 9.98070012144069e-07,
+      "loss": 0.3658,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7143283582089552,
+      "grad_norm": 0.7939940692921382,
+      "learning_rate": 9.96138232798305e-07,
+      "loss": 0.3208,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7146268656716418,
+      "grad_norm": 0.8770021483327558,
+      "learning_rate": 9.942078595266374e-07,
+      "loss": 0.3714,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7149253731343284,
+      "grad_norm": 0.8900060492450977,
+      "learning_rate": 9.922788941339169e-07,
+      "loss": 0.3235,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7152238805970149,
+      "grad_norm": 0.7812682445205359,
+      "learning_rate": 9.90351338423679e-07,
+      "loss": 0.3309,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7155223880597015,
+      "grad_norm": 1.0022155072876207,
+      "learning_rate": 9.884251941981424e-07,
+      "loss": 0.3809,
+      "step": 2397
+    },
+    {
+      "epoch": 0.715820895522388,
+      "grad_norm": 0.917525828896109,
+      "learning_rate": 9.865004632582045e-07,
+      "loss": 0.361,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7161194029850746,
+      "grad_norm": 0.7859374963278128,
+      "learning_rate": 9.845771474034419e-07,
+      "loss": 0.355,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.8984576640616055,
+      "learning_rate": 9.826552484321086e-07,
+      "loss": 0.3332,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7167164179104477,
+      "grad_norm": 0.8103441506377737,
+      "learning_rate": 9.80734768141132e-07,
+      "loss": 0.3105,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7170149253731344,
+      "grad_norm": 0.7674707580678842,
+      "learning_rate": 9.788157083261152e-07,
+      "loss": 0.2509,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7173134328358209,
+      "grad_norm": 0.86655253489264,
+      "learning_rate": 9.768980707813319e-07,
+      "loss": 0.4078,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7176119402985075,
+      "grad_norm": 0.8359834932221715,
+      "learning_rate": 9.74981857299727e-07,
+      "loss": 0.3915,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7179104477611941,
+      "grad_norm": 0.8302643419290284,
+      "learning_rate": 9.730670696729128e-07,
+      "loss": 0.3383,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7182089552238806,
+      "grad_norm": 0.8047361286111895,
+      "learning_rate": 9.711537096911704e-07,
+      "loss": 0.3249,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7185074626865672,
+      "grad_norm": 0.7966244710402332,
+      "learning_rate": 9.692417791434431e-07,
+      "loss": 0.284,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7188059701492537,
+      "grad_norm": 0.8341828877431555,
+      "learning_rate": 9.673312798173399e-07,
+      "loss": 0.321,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7191044776119403,
+      "grad_norm": 0.8810165039262209,
+      "learning_rate": 9.654222134991312e-07,
+      "loss": 0.3862,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7194029850746269,
+      "grad_norm": 0.861373072096721,
+      "learning_rate": 9.635145819737476e-07,
+      "loss": 0.336,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7197014925373134,
+      "grad_norm": 0.8448013124798284,
+      "learning_rate": 9.616083870247785e-07,
+      "loss": 0.3962,
+      "step": 2411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8451750195129675,
+      "learning_rate": 9.597036304344688e-07,
+      "loss": 0.3554,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7202985074626865,
+      "grad_norm": 0.7751019468536666,
+      "learning_rate": 9.578003139837196e-07,
+      "loss": 0.305,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7205970149253731,
+      "grad_norm": 0.8478856215524544,
+      "learning_rate": 9.558984394520856e-07,
+      "loss": 0.3419,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7208955223880597,
+      "grad_norm": 0.7578568314911519,
+      "learning_rate": 9.539980086177734e-07,
+      "loss": 0.2998,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7211940298507463,
+      "grad_norm": 0.8308842159873571,
+      "learning_rate": 9.520990232576391e-07,
+      "loss": 0.3567,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7214925373134329,
+      "grad_norm": 0.9597039646097656,
+      "learning_rate": 9.502014851471888e-07,
+      "loss": 0.3741,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7217910447761194,
+      "grad_norm": 0.8391053251773948,
+      "learning_rate": 9.483053960605726e-07,
+      "loss": 0.3388,
+      "step": 2418
+    },
+    {
+      "epoch": 0.722089552238806,
+      "grad_norm": 0.7561556408738367,
+      "learning_rate": 9.464107577705887e-07,
+      "loss": 0.3408,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7223880597014926,
+      "grad_norm": 0.9032217114431608,
+      "learning_rate": 9.445175720486771e-07,
+      "loss": 0.3679,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7226865671641791,
+      "grad_norm": 0.909162048526229,
+      "learning_rate": 9.426258406649211e-07,
+      "loss": 0.3379,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7229850746268657,
+      "grad_norm": 0.7673893074765138,
+      "learning_rate": 9.407355653880437e-07,
+      "loss": 0.2601,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7232835820895522,
+      "grad_norm": 0.8015859918855246,
+      "learning_rate": 9.388467479854046e-07,
+      "loss": 0.3201,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7235820895522388,
+      "grad_norm": 0.7805401715822302,
+      "learning_rate": 9.369593902230032e-07,
+      "loss": 0.3044,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7238805970149254,
+      "grad_norm": 0.7760863233512528,
+      "learning_rate": 9.350734938654715e-07,
+      "loss": 0.3316,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7241791044776119,
+      "grad_norm": 0.8337237951101096,
+      "learning_rate": 9.331890606760791e-07,
+      "loss": 0.3232,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7244776119402985,
+      "grad_norm": 0.7698626803368143,
+      "learning_rate": 9.313060924167247e-07,
+      "loss": 0.3024,
+      "step": 2427
+    },
+    {
+      "epoch": 0.724776119402985,
+      "grad_norm": 0.7772472006066783,
+      "learning_rate": 9.29424590847936e-07,
+      "loss": 0.3098,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7250746268656716,
+      "grad_norm": 0.8379061596545783,
+      "learning_rate": 9.275445577288722e-07,
+      "loss": 0.3738,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7253731343283583,
+      "grad_norm": 0.89865650654176,
+      "learning_rate": 9.256659948173181e-07,
+      "loss": 0.3338,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7256716417910448,
+      "grad_norm": 0.8370452187741632,
+      "learning_rate": 9.237889038696843e-07,
+      "loss": 0.3382,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7259701492537314,
+      "grad_norm": 0.8134596631444544,
+      "learning_rate": 9.219132866410063e-07,
+      "loss": 0.3064,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7262686567164179,
+      "grad_norm": 0.8264430928633055,
+      "learning_rate": 9.200391448849383e-07,
+      "loss": 0.3432,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7265671641791045,
+      "grad_norm": 0.7728324764571474,
+      "learning_rate": 9.181664803537585e-07,
+      "loss": 0.3278,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7268656716417911,
+      "grad_norm": 0.8542443384866948,
+      "learning_rate": 9.162952947983619e-07,
+      "loss": 0.311,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7271641791044776,
+      "grad_norm": 0.7451019337472867,
+      "learning_rate": 9.144255899682622e-07,
+      "loss": 0.3326,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7274626865671642,
+      "grad_norm": 0.7771194432519887,
+      "learning_rate": 9.125573676115873e-07,
+      "loss": 0.3467,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7277611940298507,
+      "grad_norm": 0.7687912387800152,
+      "learning_rate": 9.106906294750806e-07,
+      "loss": 0.3147,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7280597014925373,
+      "grad_norm": 0.9034996997981363,
+      "learning_rate": 9.088253773040947e-07,
+      "loss": 0.3687,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7283582089552239,
+      "grad_norm": 0.8836325866150969,
+      "learning_rate": 9.069616128425964e-07,
+      "loss": 0.4087,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7286567164179104,
+      "grad_norm": 1.0041210920794168,
+      "learning_rate": 9.050993378331599e-07,
+      "loss": 0.3454,
+      "step": 2441
+    },
+    {
+      "epoch": 0.728955223880597,
+      "grad_norm": 0.9216219623020782,
+      "learning_rate": 9.032385540169664e-07,
+      "loss": 0.3622,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7292537313432835,
+      "grad_norm": 0.8079424743719126,
+      "learning_rate": 9.013792631338048e-07,
+      "loss": 0.3418,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7295522388059702,
+      "grad_norm": 0.9118676627334383,
+      "learning_rate": 8.995214669220648e-07,
+      "loss": 0.3449,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7298507462686568,
+      "grad_norm": 0.8611393356418761,
+      "learning_rate": 8.976651671187417e-07,
+      "loss": 0.3299,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7301492537313433,
+      "grad_norm": 0.7639011408029112,
+      "learning_rate": 8.958103654594302e-07,
+      "loss": 0.3003,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7304477611940299,
+      "grad_norm": 0.8195346005096708,
+      "learning_rate": 8.939570636783249e-07,
+      "loss": 0.3394,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7307462686567164,
+      "grad_norm": 0.8782576394827525,
+      "learning_rate": 8.921052635082175e-07,
+      "loss": 0.361,
+      "step": 2448
+    },
+    {
+      "epoch": 0.731044776119403,
+      "grad_norm": 0.8671344977010738,
+      "learning_rate": 8.902549666804971e-07,
+      "loss": 0.3828,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7313432835820896,
+      "grad_norm": 0.809160024689468,
+      "learning_rate": 8.884061749251446e-07,
+      "loss": 0.3113,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7316417910447761,
+      "grad_norm": 0.8158507141596787,
+      "learning_rate": 8.865588899707358e-07,
+      "loss": 0.3481,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7319402985074627,
+      "grad_norm": 0.8308589930715008,
+      "learning_rate": 8.847131135444373e-07,
+      "loss": 0.3448,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7322388059701492,
+      "grad_norm": 0.8737351072965946,
+      "learning_rate": 8.828688473720051e-07,
+      "loss": 0.3881,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7325373134328358,
+      "grad_norm": 0.8195810108023454,
+      "learning_rate": 8.810260931777828e-07,
+      "loss": 0.3467,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7328358208955223,
+      "grad_norm": 0.8343708212051258,
+      "learning_rate": 8.791848526847016e-07,
+      "loss": 0.3586,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7331343283582089,
+      "grad_norm": 0.8073884493919178,
+      "learning_rate": 8.77345127614275e-07,
+      "loss": 0.3176,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7334328358208955,
+      "grad_norm": 0.9085802883997077,
+      "learning_rate": 8.755069196866014e-07,
+      "loss": 0.3577,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7337313432835821,
+      "grad_norm": 0.7503848280448058,
+      "learning_rate": 8.736702306203612e-07,
+      "loss": 0.2852,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7340298507462687,
+      "grad_norm": 1.180773273009247,
+      "learning_rate": 8.718350621328137e-07,
+      "loss": 0.353,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7343283582089553,
+      "grad_norm": 0.8460574776608701,
+      "learning_rate": 8.700014159397971e-07,
+      "loss": 0.3534,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7346268656716418,
+      "grad_norm": 0.8636506104198776,
+      "learning_rate": 8.681692937557246e-07,
+      "loss": 0.2733,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7349253731343284,
+      "grad_norm": 0.7554477293718002,
+      "learning_rate": 8.663386972935864e-07,
+      "loss": 0.3375,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7352238805970149,
+      "grad_norm": 0.8170180017573404,
+      "learning_rate": 8.645096282649448e-07,
+      "loss": 0.3501,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7355223880597015,
+      "grad_norm": 0.785002214591176,
+      "learning_rate": 8.626820883799364e-07,
+      "loss": 0.322,
+      "step": 2464
+    },
+    {
+      "epoch": 0.735820895522388,
+      "grad_norm": 0.8124518915482682,
+      "learning_rate": 8.608560793472667e-07,
+      "loss": 0.3488,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7361194029850746,
+      "grad_norm": 0.8819651422697621,
+      "learning_rate": 8.590316028742079e-07,
+      "loss": 0.3871,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7364179104477612,
+      "grad_norm": 0.8051048323090729,
+      "learning_rate": 8.572086606666016e-07,
+      "loss": 0.3391,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7367164179104477,
+      "grad_norm": 0.8716400545844462,
+      "learning_rate": 8.553872544288544e-07,
+      "loss": 0.3561,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7370149253731343,
+      "grad_norm": 1.0077273262938058,
+      "learning_rate": 8.535673858639368e-07,
+      "loss": 0.3269,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7373134328358208,
+      "grad_norm": 0.7495287643901639,
+      "learning_rate": 8.517490566733827e-07,
+      "loss": 0.3547,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7376119402985075,
+      "grad_norm": 0.94770061816034,
+      "learning_rate": 8.499322685572834e-07,
+      "loss": 0.3732,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7379104477611941,
+      "grad_norm": 0.8036303894234751,
+      "learning_rate": 8.481170232142923e-07,
+      "loss": 0.3387,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7382089552238806,
+      "grad_norm": 0.770395344765945,
+      "learning_rate": 8.463033223416201e-07,
+      "loss": 0.2754,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7385074626865672,
+      "grad_norm": 0.7821239157671365,
+      "learning_rate": 8.444911676350326e-07,
+      "loss": 0.3406,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7388059701492538,
+      "grad_norm": 0.8604005758402877,
+      "learning_rate": 8.426805607888502e-07,
+      "loss": 0.3882,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7391044776119403,
+      "grad_norm": 0.7436455606846663,
+      "learning_rate": 8.408715034959469e-07,
+      "loss": 0.3026,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7394029850746269,
+      "grad_norm": 0.8531423147657465,
+      "learning_rate": 8.390639974477461e-07,
+      "loss": 0.3147,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7397014925373134,
+      "grad_norm": 0.897354179300323,
+      "learning_rate": 8.372580443342218e-07,
+      "loss": 0.3369,
+      "step": 2478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7937336535275226,
+      "learning_rate": 8.354536458438969e-07,
+      "loss": 0.3417,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7402985074626866,
+      "grad_norm": 0.7948747004341821,
+      "learning_rate": 8.3365080366384e-07,
+      "loss": 0.3138,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7405970149253731,
+      "grad_norm": 0.8198556735193836,
+      "learning_rate": 8.31849519479665e-07,
+      "loss": 0.3394,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7408955223880597,
+      "grad_norm": 0.8619919834482459,
+      "learning_rate": 8.300497949755271e-07,
+      "loss": 0.312,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7411940298507462,
+      "grad_norm": 0.8257206030147372,
+      "learning_rate": 8.282516318341258e-07,
+      "loss": 0.3253,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7414925373134328,
+      "grad_norm": 0.8361638399301501,
+      "learning_rate": 8.264550317366998e-07,
+      "loss": 0.3585,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7417910447761195,
+      "grad_norm": 0.9277677786456653,
+      "learning_rate": 8.246599963630266e-07,
+      "loss": 0.3364,
+      "step": 2485
+    },
+    {
+      "epoch": 0.742089552238806,
+      "grad_norm": 0.8008796354592284,
+      "learning_rate": 8.228665273914202e-07,
+      "loss": 0.305,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7423880597014926,
+      "grad_norm": 0.7154547857109149,
+      "learning_rate": 8.210746264987315e-07,
+      "loss": 0.3003,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7426865671641791,
+      "grad_norm": 0.8086001172238579,
+      "learning_rate": 8.192842953603422e-07,
+      "loss": 0.3445,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7429850746268657,
+      "grad_norm": 0.809703550422652,
+      "learning_rate": 8.174955356501693e-07,
+      "loss": 0.3528,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7432835820895523,
+      "grad_norm": 0.7940020900981949,
+      "learning_rate": 8.157083490406593e-07,
+      "loss": 0.3609,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7435820895522388,
+      "grad_norm": 0.8667872980953691,
+      "learning_rate": 8.139227372027883e-07,
+      "loss": 0.3538,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7438805970149254,
+      "grad_norm": 0.8022738790006476,
+      "learning_rate": 8.121387018060601e-07,
+      "loss": 0.3551,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7441791044776119,
+      "grad_norm": 0.7944094322718988,
+      "learning_rate": 8.103562445185045e-07,
+      "loss": 0.2928,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7444776119402985,
+      "grad_norm": 0.9134473219974198,
+      "learning_rate": 8.085753670066746e-07,
+      "loss": 0.3283,
+      "step": 2494
+    },
+    {
+      "epoch": 0.744776119402985,
+      "grad_norm": 0.9171291476302468,
+      "learning_rate": 8.067960709356479e-07,
+      "loss": 0.3432,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7450746268656716,
+      "grad_norm": 0.7630031740731616,
+      "learning_rate": 8.05018357969023e-07,
+      "loss": 0.2707,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7453731343283582,
+      "grad_norm": 0.8408211927826157,
+      "learning_rate": 8.03242229768918e-07,
+      "loss": 0.3425,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7456716417910447,
+      "grad_norm": 0.8479202019499819,
+      "learning_rate": 8.014676879959704e-07,
+      "loss": 0.3288,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7459701492537314,
+      "grad_norm": 0.8959963249910727,
+      "learning_rate": 7.996947343093323e-07,
+      "loss": 0.3825,
+      "step": 2499
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 0.8469480512025105,
+      "learning_rate": 7.979233703666714e-07,
+      "loss": 0.3364,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7465671641791045,
+      "grad_norm": 0.8363118813056599,
+      "learning_rate": 7.961535978241722e-07,
+      "loss": 0.3655,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7468656716417911,
+      "grad_norm": 0.7993313849201408,
+      "learning_rate": 7.943854183365277e-07,
+      "loss": 0.3081,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7471641791044776,
+      "grad_norm": 0.8028836757795254,
+      "learning_rate": 7.926188335569432e-07,
+      "loss": 0.3151,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7474626865671642,
+      "grad_norm": 0.8193299901791269,
+      "learning_rate": 7.908538451371311e-07,
+      "loss": 0.3327,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7477611940298508,
+      "grad_norm": 0.812821295525394,
+      "learning_rate": 7.890904547273134e-07,
+      "loss": 0.3028,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7480597014925373,
+      "grad_norm": 0.7946631103592113,
+      "learning_rate": 7.873286639762171e-07,
+      "loss": 0.3305,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7483582089552239,
+      "grad_norm": 0.8168449751348733,
+      "learning_rate": 7.855684745310732e-07,
+      "loss": 0.311,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7486567164179104,
+      "grad_norm": 0.7577415278692218,
+      "learning_rate": 7.838098880376175e-07,
+      "loss": 0.3005,
+      "step": 2508
+    },
+    {
+      "epoch": 0.748955223880597,
+      "grad_norm": 1.007066007674342,
+      "learning_rate": 7.820529061400833e-07,
+      "loss": 0.3422,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7492537313432835,
+      "grad_norm": 0.8285843650225135,
+      "learning_rate": 7.802975304812068e-07,
+      "loss": 0.3649,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7495522388059701,
+      "grad_norm": 0.8802781806958206,
+      "learning_rate": 7.785437627022216e-07,
+      "loss": 0.3442,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7498507462686567,
+      "grad_norm": 0.8872697995230727,
+      "learning_rate": 7.767916044428578e-07,
+      "loss": 0.3535,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7501492537313433,
+      "grad_norm": 0.7493499394945767,
+      "learning_rate": 7.750410573413403e-07,
+      "loss": 0.3,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7504477611940299,
+      "grad_norm": 0.8420235208681436,
+      "learning_rate": 7.732921230343893e-07,
+      "loss": 0.3361,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7507462686567165,
+      "grad_norm": 0.9016435397252184,
+      "learning_rate": 7.715448031572137e-07,
+      "loss": 0.3665,
+      "step": 2515
+    },
+    {
+      "epoch": 0.751044776119403,
+      "grad_norm": 0.8254349682410594,
+      "learning_rate": 7.697990993435162e-07,
+      "loss": 0.3658,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7513432835820896,
+      "grad_norm": 0.8283799880314672,
+      "learning_rate": 7.680550132254869e-07,
+      "loss": 0.2997,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7516417910447761,
+      "grad_norm": 0.7889635309394748,
+      "learning_rate": 7.663125464338042e-07,
+      "loss": 0.3662,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7519402985074627,
+      "grad_norm": 0.7589268133008595,
+      "learning_rate": 7.645717005976328e-07,
+      "loss": 0.2914,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7522388059701492,
+      "grad_norm": 0.8796818140938326,
+      "learning_rate": 7.628324773446192e-07,
+      "loss": 0.3653,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7525373134328358,
+      "grad_norm": 0.8567321353644352,
+      "learning_rate": 7.610948783008956e-07,
+      "loss": 0.3185,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7528358208955224,
+      "grad_norm": 0.8340526715199753,
+      "learning_rate": 7.593589050910752e-07,
+      "loss": 0.3609,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7531343283582089,
+      "grad_norm": 0.8400627991409522,
+      "learning_rate": 7.576245593382503e-07,
+      "loss": 0.3362,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7534328358208955,
+      "grad_norm": 0.8335760592291928,
+      "learning_rate": 7.558918426639919e-07,
+      "loss": 0.3326,
+      "step": 2524
+    },
+    {
+      "epoch": 0.753731343283582,
+      "grad_norm": 0.7330595579589215,
+      "learning_rate": 7.541607566883486e-07,
+      "loss": 0.316,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7540298507462686,
+      "grad_norm": 0.8319155733303641,
+      "learning_rate": 7.524313030298422e-07,
+      "loss": 0.3825,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7543283582089553,
+      "grad_norm": 0.8537564302430024,
+      "learning_rate": 7.507034833054705e-07,
+      "loss": 0.3675,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7546268656716418,
+      "grad_norm": 0.8541046358406873,
+      "learning_rate": 7.489772991307026e-07,
+      "loss": 0.3463,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7549253731343284,
+      "grad_norm": 0.8333446760301982,
+      "learning_rate": 7.472527521194789e-07,
+      "loss": 0.3248,
+      "step": 2529
+    },
+    {
+      "epoch": 0.755223880597015,
+      "grad_norm": 0.918746792475965,
+      "learning_rate": 7.455298438842096e-07,
+      "loss": 0.3618,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7555223880597015,
+      "grad_norm": 0.7709947075005573,
+      "learning_rate": 7.4380857603577e-07,
+      "loss": 0.323,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7558208955223881,
+      "grad_norm": 0.8113705340822994,
+      "learning_rate": 7.420889501835046e-07,
+      "loss": 0.3353,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7561194029850746,
+      "grad_norm": 0.8316485419308024,
+      "learning_rate": 7.403709679352216e-07,
+      "loss": 0.3197,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7564179104477612,
+      "grad_norm": 0.9244834653645814,
+      "learning_rate": 7.386546308971926e-07,
+      "loss": 0.3625,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7567164179104477,
+      "grad_norm": 0.7698158757397285,
+      "learning_rate": 7.369399406741509e-07,
+      "loss": 0.3401,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7570149253731343,
+      "grad_norm": 0.7936072499011064,
+      "learning_rate": 7.35226898869291e-07,
+      "loss": 0.3294,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7573134328358209,
+      "grad_norm": 0.8096055523595336,
+      "learning_rate": 7.335155070842631e-07,
+      "loss": 0.3354,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7576119402985074,
+      "grad_norm": 0.8150395641292778,
+      "learning_rate": 7.318057669191775e-07,
+      "loss": 0.3378,
+      "step": 2538
+    },
+    {
+      "epoch": 0.757910447761194,
+      "grad_norm": 0.8749495077069125,
+      "learning_rate": 7.300976799726007e-07,
+      "loss": 0.3548,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7582089552238805,
+      "grad_norm": 0.7864759817642071,
+      "learning_rate": 7.283912478415522e-07,
+      "loss": 0.298,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7585074626865672,
+      "grad_norm": 0.85761429358801,
+      "learning_rate": 7.266864721215033e-07,
+      "loss": 0.3667,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7588059701492538,
+      "grad_norm": 0.7614755089756268,
+      "learning_rate": 7.249833544063786e-07,
+      "loss": 0.3075,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7591044776119403,
+      "grad_norm": 0.7998879965445289,
+      "learning_rate": 7.232818962885513e-07,
+      "loss": 0.3155,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7594029850746269,
+      "grad_norm": 1.1882550211054863,
+      "learning_rate": 7.215820993588435e-07,
+      "loss": 0.3549,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7597014925373134,
+      "grad_norm": 0.6960475242707338,
+      "learning_rate": 7.198839652065234e-07,
+      "loss": 0.2709,
+      "step": 2545
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8929595250452405,
+      "learning_rate": 7.181874954193066e-07,
+      "loss": 0.4011,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7602985074626866,
+      "grad_norm": 0.8085099466146569,
+      "learning_rate": 7.164926915833489e-07,
+      "loss": 0.3258,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7605970149253731,
+      "grad_norm": 0.7526251937377817,
+      "learning_rate": 7.147995552832518e-07,
+      "loss": 0.3357,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7608955223880597,
+      "grad_norm": 0.7452474358495229,
+      "learning_rate": 7.13108088102056e-07,
+      "loss": 0.2711,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7611940298507462,
+      "grad_norm": 0.8572099572681284,
+      "learning_rate": 7.114182916212423e-07,
+      "loss": 0.3695,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7614925373134328,
+      "grad_norm": 0.9598211721736167,
+      "learning_rate": 7.097301674207291e-07,
+      "loss": 0.3611,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7617910447761194,
+      "grad_norm": 0.9085391291886787,
+      "learning_rate": 7.080437170788723e-07,
+      "loss": 0.3824,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7620895522388059,
+      "grad_norm": 0.8177813221411165,
+      "learning_rate": 7.063589421724601e-07,
+      "loss": 0.2772,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7623880597014925,
+      "grad_norm": 0.8628641410267316,
+      "learning_rate": 7.046758442767171e-07,
+      "loss": 0.334,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7626865671641792,
+      "grad_norm": 0.928648183310679,
+      "learning_rate": 7.029944249652987e-07,
+      "loss": 0.3221,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7629850746268657,
+      "grad_norm": 0.9255552592094868,
+      "learning_rate": 7.013146858102906e-07,
+      "loss": 0.3226,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7632835820895523,
+      "grad_norm": 0.7920749204965727,
+      "learning_rate": 6.996366283822093e-07,
+      "loss": 0.2868,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7635820895522388,
+      "grad_norm": 0.9935006013461068,
+      "learning_rate": 6.979602542499955e-07,
+      "loss": 0.313,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7638805970149254,
+      "grad_norm": 0.790352931784802,
+      "learning_rate": 6.962855649810193e-07,
+      "loss": 0.3091,
+      "step": 2559
+    },
+    {
+      "epoch": 0.764179104477612,
+      "grad_norm": 0.784302622452047,
+      "learning_rate": 6.946125621410746e-07,
+      "loss": 0.3021,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7644776119402985,
+      "grad_norm": 0.8521244035570953,
+      "learning_rate": 6.929412472943775e-07,
+      "loss": 0.3065,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7647761194029851,
+      "grad_norm": 0.7990696640196155,
+      "learning_rate": 6.91271622003567e-07,
+      "loss": 0.317,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7650746268656716,
+      "grad_norm": 0.7301484428928479,
+      "learning_rate": 6.896036878297033e-07,
+      "loss": 0.3071,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7653731343283582,
+      "grad_norm": 0.7370476408839566,
+      "learning_rate": 6.879374463322619e-07,
+      "loss": 0.2726,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7656716417910447,
+      "grad_norm": 0.9000167299852614,
+      "learning_rate": 6.862728990691392e-07,
+      "loss": 0.3357,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7659701492537313,
+      "grad_norm": 0.7870895798188804,
+      "learning_rate": 6.846100475966461e-07,
+      "loss": 0.2879,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7662686567164179,
+      "grad_norm": 0.8360237218908565,
+      "learning_rate": 6.829488934695083e-07,
+      "loss": 0.3215,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7665671641791045,
+      "grad_norm": 0.9359634021391391,
+      "learning_rate": 6.812894382408652e-07,
+      "loss": 0.3455,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7668656716417911,
+      "grad_norm": 0.930052847096152,
+      "learning_rate": 6.796316834622654e-07,
+      "loss": 0.3582,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7671641791044777,
+      "grad_norm": 0.9038130785381302,
+      "learning_rate": 6.779756306836702e-07,
+      "loss": 0.2985,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7674626865671642,
+      "grad_norm": 0.7826417580282887,
+      "learning_rate": 6.763212814534484e-07,
+      "loss": 0.3482,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7677611940298508,
+      "grad_norm": 0.8486630906094814,
+      "learning_rate": 6.746686373183761e-07,
+      "loss": 0.3688,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7680597014925373,
+      "grad_norm": 0.8735312621938363,
+      "learning_rate": 6.730176998236355e-07,
+      "loss": 0.3815,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7683582089552239,
+      "grad_norm": 0.8940128219367052,
+      "learning_rate": 6.713684705128135e-07,
+      "loss": 0.3194,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7686567164179104,
+      "grad_norm": 0.9408242887386759,
+      "learning_rate": 6.697209509278979e-07,
+      "loss": 0.3658,
+      "step": 2575
+    },
+    {
+      "epoch": 0.768955223880597,
+      "grad_norm": 0.8959030487388594,
+      "learning_rate": 6.680751426092791e-07,
+      "loss": 0.3391,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7692537313432836,
+      "grad_norm": 0.7803592674313101,
+      "learning_rate": 6.664310470957497e-07,
+      "loss": 0.2969,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7695522388059701,
+      "grad_norm": 0.7505213649281478,
+      "learning_rate": 6.647886659244987e-07,
+      "loss": 0.2958,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7698507462686567,
+      "grad_norm": 0.8011860941500494,
+      "learning_rate": 6.631480006311108e-07,
+      "loss": 0.311,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7701492537313432,
+      "grad_norm": 0.7936319294693487,
+      "learning_rate": 6.61509052749569e-07,
+      "loss": 0.3458,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7704477611940298,
+      "grad_norm": 0.7316957671250423,
+      "learning_rate": 6.598718238122497e-07,
+      "loss": 0.2991,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7707462686567165,
+      "grad_norm": 0.8104928093316218,
+      "learning_rate": 6.582363153499221e-07,
+      "loss": 0.3437,
+      "step": 2582
+    },
+    {
+      "epoch": 0.771044776119403,
+      "grad_norm": 0.9010645711228973,
+      "learning_rate": 6.566025288917463e-07,
+      "loss": 0.3652,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7713432835820896,
+      "grad_norm": 0.8089700577841448,
+      "learning_rate": 6.549704659652742e-07,
+      "loss": 0.371,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7716417910447761,
+      "grad_norm": 0.8261177060625268,
+      "learning_rate": 6.533401280964427e-07,
+      "loss": 0.3337,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7719402985074627,
+      "grad_norm": 0.8072522990077107,
+      "learning_rate": 6.517115168095792e-07,
+      "loss": 0.3177,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7722388059701493,
+      "grad_norm": 0.8196927890724248,
+      "learning_rate": 6.500846336273953e-07,
+      "loss": 0.3239,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7725373134328358,
+      "grad_norm": 0.8447623565955257,
+      "learning_rate": 6.484594800709873e-07,
+      "loss": 0.3301,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7728358208955224,
+      "grad_norm": 0.8469818517147698,
+      "learning_rate": 6.468360576598339e-07,
+      "loss": 0.3407,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7731343283582089,
+      "grad_norm": 0.8200398390703326,
+      "learning_rate": 6.452143679117965e-07,
+      "loss": 0.3304,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7734328358208955,
+      "grad_norm": 0.8316560060664724,
+      "learning_rate": 6.435944123431137e-07,
+      "loss": 0.3667,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7737313432835821,
+      "grad_norm": 0.8479185212991167,
+      "learning_rate": 6.419761924684056e-07,
+      "loss": 0.3306,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7740298507462686,
+      "grad_norm": 0.7788855711723567,
+      "learning_rate": 6.403597098006684e-07,
+      "loss": 0.3336,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7743283582089552,
+      "grad_norm": 0.8580372589130707,
+      "learning_rate": 6.387449658512735e-07,
+      "loss": 0.3446,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7746268656716417,
+      "grad_norm": 0.8545734419513868,
+      "learning_rate": 6.371319621299679e-07,
+      "loss": 0.338,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7749253731343284,
+      "grad_norm": 1.3812292290362789,
+      "learning_rate": 6.355207001448696e-07,
+      "loss": 0.3346,
+      "step": 2596
+    },
+    {
+      "epoch": 0.775223880597015,
+      "grad_norm": 0.853363224946842,
+      "learning_rate": 6.339111814024701e-07,
+      "loss": 0.391,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7755223880597015,
+      "grad_norm": 0.8265934162348483,
+      "learning_rate": 6.323034074076298e-07,
+      "loss": 0.3339,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7758208955223881,
+      "grad_norm": 0.9064944014108518,
+      "learning_rate": 6.306973796635785e-07,
+      "loss": 0.3589,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7761194029850746,
+      "grad_norm": 0.9677195208881169,
+      "learning_rate": 6.29093099671913e-07,
+      "loss": 0.317,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7764179104477612,
+      "grad_norm": 0.8311565173033396,
+      "learning_rate": 6.274905689325966e-07,
+      "loss": 0.3472,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7767164179104478,
+      "grad_norm": 0.7950764731124217,
+      "learning_rate": 6.258897889439549e-07,
+      "loss": 0.3109,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7770149253731343,
+      "grad_norm": 0.9141762608429268,
+      "learning_rate": 6.242907612026792e-07,
+      "loss": 0.3974,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7773134328358209,
+      "grad_norm": 0.8803598215750571,
+      "learning_rate": 6.22693487203821e-07,
+      "loss": 0.3331,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7776119402985074,
+      "grad_norm": 0.7668038317932448,
+      "learning_rate": 6.210979684407931e-07,
+      "loss": 0.3298,
+      "step": 2605
+    },
+    {
+      "epoch": 0.777910447761194,
+      "grad_norm": 0.8448858692096748,
+      "learning_rate": 6.195042064053672e-07,
+      "loss": 0.3259,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7782089552238806,
+      "grad_norm": 0.8115377621130089,
+      "learning_rate": 6.179122025876702e-07,
+      "loss": 0.3155,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7785074626865671,
+      "grad_norm": 0.7350727849438435,
+      "learning_rate": 6.163219584761879e-07,
+      "loss": 0.3448,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7788059701492537,
+      "grad_norm": 0.803225857620502,
+      "learning_rate": 6.147334755577597e-07,
+      "loss": 0.3125,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7791044776119403,
+      "grad_norm": 0.7835583970542247,
+      "learning_rate": 6.131467553175785e-07,
+      "loss": 0.295,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7794029850746269,
+      "grad_norm": 0.802514614490111,
+      "learning_rate": 6.11561799239189e-07,
+      "loss": 0.3624,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7797014925373135,
+      "grad_norm": 0.8264140923298133,
+      "learning_rate": 6.099786088044871e-07,
+      "loss": 0.3453,
+      "step": 2612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7730123427638769,
+      "learning_rate": 6.083971854937157e-07,
+      "loss": 0.3126,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7802985074626866,
+      "grad_norm": 0.876817858400939,
+      "learning_rate": 6.068175307854676e-07,
+      "loss": 0.3374,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7805970149253731,
+      "grad_norm": 0.7780503068811441,
+      "learning_rate": 6.052396461566823e-07,
+      "loss": 0.402,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7808955223880597,
+      "grad_norm": 0.769230896426925,
+      "learning_rate": 6.036635330826437e-07,
+      "loss": 0.3172,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7811940298507463,
+      "grad_norm": 0.9137700677139416,
+      "learning_rate": 6.02089193036978e-07,
+      "loss": 0.3312,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7814925373134328,
+      "grad_norm": 0.7641819962198311,
+      "learning_rate": 6.00516627491655e-07,
+      "loss": 0.3646,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7817910447761194,
+      "grad_norm": 0.7985862077567278,
+      "learning_rate": 5.989458379169852e-07,
+      "loss": 0.3615,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7820895522388059,
+      "grad_norm": 0.8430048020986005,
+      "learning_rate": 5.973768257816187e-07,
+      "loss": 0.3587,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7823880597014925,
+      "grad_norm": 0.8377553518393402,
+      "learning_rate": 5.958095925525437e-07,
+      "loss": 0.2958,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7826865671641791,
+      "grad_norm": 0.822488397832201,
+      "learning_rate": 5.942441396950857e-07,
+      "loss": 0.3351,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7829850746268656,
+      "grad_norm": 0.8772777506321962,
+      "learning_rate": 5.926804686729034e-07,
+      "loss": 0.3915,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7832835820895523,
+      "grad_norm": 0.8598726121904382,
+      "learning_rate": 5.911185809479919e-07,
+      "loss": 0.2941,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 0.767393624496139,
+      "learning_rate": 5.895584779806782e-07,
+      "loss": 0.3313,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7838805970149254,
+      "grad_norm": 0.8365941078767859,
+      "learning_rate": 5.880001612296208e-07,
+      "loss": 0.3248,
+      "step": 2626
+    },
+    {
+      "epoch": 0.784179104477612,
+      "grad_norm": 0.8689097485635112,
+      "learning_rate": 5.864436321518077e-07,
+      "loss": 0.3254,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7844776119402985,
+      "grad_norm": 0.8501156388344782,
+      "learning_rate": 5.848888922025553e-07,
+      "loss": 0.356,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7847761194029851,
+      "grad_norm": 0.871278978235916,
+      "learning_rate": 5.833359428355076e-07,
+      "loss": 0.3179,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7850746268656716,
+      "grad_norm": 0.8117067102631867,
+      "learning_rate": 5.817847855026346e-07,
+      "loss": 0.3056,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7853731343283582,
+      "grad_norm": 0.7831623164451796,
+      "learning_rate": 5.802354216542305e-07,
+      "loss": 0.2978,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7856716417910448,
+      "grad_norm": 0.8648788615413797,
+      "learning_rate": 5.78687852738913e-07,
+      "loss": 0.3269,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7859701492537313,
+      "grad_norm": 1.2129702699211065,
+      "learning_rate": 5.771420802036218e-07,
+      "loss": 0.3086,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7862686567164179,
+      "grad_norm": 0.8425751147312559,
+      "learning_rate": 5.755981054936152e-07,
+      "loss": 0.3467,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7865671641791044,
+      "grad_norm": 0.8226182315548135,
+      "learning_rate": 5.740559300524726e-07,
+      "loss": 0.3224,
+      "step": 2635
+    },
+    {
+      "epoch": 0.786865671641791,
+      "grad_norm": 0.7695238455194613,
+      "learning_rate": 5.725155553220904e-07,
+      "loss": 0.282,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7871641791044776,
+      "grad_norm": 0.8454114923108801,
+      "learning_rate": 5.709769827426814e-07,
+      "loss": 0.3547,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7874626865671642,
+      "grad_norm": 0.9164178244850171,
+      "learning_rate": 5.694402137527735e-07,
+      "loss": 0.3695,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7877611940298508,
+      "grad_norm": 0.9339118467727153,
+      "learning_rate": 5.679052497892093e-07,
+      "loss": 0.3797,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7880597014925373,
+      "grad_norm": 0.8732595479379647,
+      "learning_rate": 5.663720922871408e-07,
+      "loss": 0.3355,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7883582089552239,
+      "grad_norm": 0.7644896912195902,
+      "learning_rate": 5.648407426800337e-07,
+      "loss": 0.3427,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7886567164179105,
+      "grad_norm": 0.9237943784299637,
+      "learning_rate": 5.633112023996626e-07,
+      "loss": 0.3387,
+      "step": 2642
+    },
+    {
+      "epoch": 0.788955223880597,
+      "grad_norm": 0.8362466946619596,
+      "learning_rate": 5.617834728761104e-07,
+      "loss": 0.329,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7892537313432836,
+      "grad_norm": 0.819307805738088,
+      "learning_rate": 5.602575555377676e-07,
+      "loss": 0.3064,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7895522388059701,
+      "grad_norm": 0.7447822449758292,
+      "learning_rate": 5.587334518113285e-07,
+      "loss": 0.3393,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7898507462686567,
+      "grad_norm": 0.8444167330983842,
+      "learning_rate": 5.572111631217936e-07,
+      "loss": 0.339,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7901492537313433,
+      "grad_norm": 0.8006438974002873,
+      "learning_rate": 5.556906908924656e-07,
+      "loss": 0.3163,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7904477611940298,
+      "grad_norm": 0.9751532556450391,
+      "learning_rate": 5.541720365449488e-07,
+      "loss": 0.3634,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7907462686567164,
+      "grad_norm": 0.8268096722848733,
+      "learning_rate": 5.526552014991485e-07,
+      "loss": 0.3187,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7910447761194029,
+      "grad_norm": 0.7212108804420516,
+      "learning_rate": 5.511401871732685e-07,
+      "loss": 0.3009,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7913432835820896,
+      "grad_norm": 0.8746502710892506,
+      "learning_rate": 5.496269949838096e-07,
+      "loss": 0.3687,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7916417910447762,
+      "grad_norm": 0.8435667329953185,
+      "learning_rate": 5.481156263455692e-07,
+      "loss": 0.3187,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7919402985074627,
+      "grad_norm": 0.7718206010241203,
+      "learning_rate": 5.466060826716416e-07,
+      "loss": 0.2981,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7922388059701493,
+      "grad_norm": 0.8420876178922836,
+      "learning_rate": 5.450983653734135e-07,
+      "loss": 0.3497,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7925373134328358,
+      "grad_norm": 0.7794372062492068,
+      "learning_rate": 5.435924758605621e-07,
+      "loss": 0.3162,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7928358208955224,
+      "grad_norm": 0.7218324952169056,
+      "learning_rate": 5.420884155410583e-07,
+      "loss": 0.2859,
+      "step": 2656
+    },
+    {
+      "epoch": 0.793134328358209,
+      "grad_norm": 0.8935853893312067,
+      "learning_rate": 5.405861858211617e-07,
+      "loss": 0.3275,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7934328358208955,
+      "grad_norm": 0.8885132950616169,
+      "learning_rate": 5.390857881054206e-07,
+      "loss": 0.335,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7937313432835821,
+      "grad_norm": 0.7811013082337122,
+      "learning_rate": 5.375872237966706e-07,
+      "loss": 0.3193,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7940298507462686,
+      "grad_norm": 0.761806698796138,
+      "learning_rate": 5.360904942960327e-07,
+      "loss": 0.3307,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7943283582089552,
+      "grad_norm": 0.8600929269448834,
+      "learning_rate": 5.345956010029118e-07,
+      "loss": 0.3242,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7946268656716418,
+      "grad_norm": 0.8628730775985842,
+      "learning_rate": 5.331025453149971e-07,
+      "loss": 0.2922,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7949253731343283,
+      "grad_norm": 0.8558985245441267,
+      "learning_rate": 5.316113286282595e-07,
+      "loss": 0.3258,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7952238805970149,
+      "grad_norm": 0.8797671914656098,
+      "learning_rate": 5.301219523369499e-07,
+      "loss": 0.3694,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7955223880597015,
+      "grad_norm": 0.7806683157905394,
+      "learning_rate": 5.286344178336e-07,
+      "loss": 0.3144,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7958208955223881,
+      "grad_norm": 0.81264272258197,
+      "learning_rate": 5.271487265090163e-07,
+      "loss": 0.3667,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7961194029850747,
+      "grad_norm": 0.8601858762639831,
+      "learning_rate": 5.256648797522851e-07,
+      "loss": 0.337,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7964179104477612,
+      "grad_norm": 0.8271856093309297,
+      "learning_rate": 5.241828789507669e-07,
+      "loss": 0.2935,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7967164179104478,
+      "grad_norm": 0.859387934817203,
+      "learning_rate": 5.227027254900963e-07,
+      "loss": 0.3188,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7970149253731343,
+      "grad_norm": 0.9900633077884397,
+      "learning_rate": 5.212244207541806e-07,
+      "loss": 0.4112,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7973134328358209,
+      "grad_norm": 0.7598238168460656,
+      "learning_rate": 5.197479661251994e-07,
+      "loss": 0.2827,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7976119402985075,
+      "grad_norm": 0.8105463195253384,
+      "learning_rate": 5.182733629836006e-07,
+      "loss": 0.3,
+      "step": 2672
+    },
+    {
+      "epoch": 0.797910447761194,
+      "grad_norm": 0.8470648410134773,
+      "learning_rate": 5.168006127081027e-07,
+      "loss": 0.348,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7982089552238806,
+      "grad_norm": 0.8543097630359251,
+      "learning_rate": 5.153297166756913e-07,
+      "loss": 0.3301,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7985074626865671,
+      "grad_norm": 0.9187030166832789,
+      "learning_rate": 5.138606762616186e-07,
+      "loss": 0.3212,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7988059701492537,
+      "grad_norm": 0.746950665181267,
+      "learning_rate": 5.123934928394011e-07,
+      "loss": 0.2924,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7991044776119403,
+      "grad_norm": 0.8206743880698498,
+      "learning_rate": 5.109281677808207e-07,
+      "loss": 0.3299,
+      "step": 2677
+    },
+    {
+      "epoch": 0.7994029850746268,
+      "grad_norm": 0.8977645162297897,
+      "learning_rate": 5.094647024559188e-07,
+      "loss": 0.363,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7997014925373135,
+      "grad_norm": 0.7603505513127645,
+      "learning_rate": 5.080030982330008e-07,
+      "loss": 0.2685,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7885960133919522,
+      "learning_rate": 5.065433564786312e-07,
+      "loss": 0.3079,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8002985074626866,
+      "grad_norm": 0.801692605365031,
+      "learning_rate": 5.050854785576326e-07,
+      "loss": 0.3456,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8005970149253732,
+      "grad_norm": 0.7935079560838534,
+      "learning_rate": 5.036294658330862e-07,
+      "loss": 0.359,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8008955223880597,
+      "grad_norm": 0.7511048696552357,
+      "learning_rate": 5.021753196663268e-07,
+      "loss": 0.3289,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8011940298507463,
+      "grad_norm": 0.9217091503137018,
+      "learning_rate": 5.007230414169464e-07,
+      "loss": 0.3606,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8014925373134328,
+      "grad_norm": 0.8424874150110574,
+      "learning_rate": 4.992726324427901e-07,
+      "loss": 0.3526,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8017910447761194,
+      "grad_norm": 0.7801243835807536,
+      "learning_rate": 4.978240940999546e-07,
+      "loss": 0.2924,
+      "step": 2686
+    },
+    {
+      "epoch": 0.802089552238806,
+      "grad_norm": 0.7866205363691048,
+      "learning_rate": 4.96377427742788e-07,
+      "loss": 0.3286,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8023880597014925,
+      "grad_norm": 0.8260723183869202,
+      "learning_rate": 4.949326347238887e-07,
+      "loss": 0.3512,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8026865671641791,
+      "grad_norm": 0.9243904923211944,
+      "learning_rate": 4.934897163941008e-07,
+      "loss": 0.3652,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8029850746268656,
+      "grad_norm": 0.7325605543620152,
+      "learning_rate": 4.920486741025202e-07,
+      "loss": 0.3073,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8032835820895522,
+      "grad_norm": 0.825997047722358,
+      "learning_rate": 4.906095091964855e-07,
+      "loss": 0.3904,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8035820895522388,
+      "grad_norm": 0.9371944525084498,
+      "learning_rate": 4.891722230215812e-07,
+      "loss": 0.3535,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8038805970149254,
+      "grad_norm": 0.8203179911092189,
+      "learning_rate": 4.87736816921634e-07,
+      "loss": 0.3135,
+      "step": 2693
+    },
+    {
+      "epoch": 0.804179104477612,
+      "grad_norm": 0.9856622184168011,
+      "learning_rate": 4.863032922387137e-07,
+      "loss": 0.3265,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8044776119402985,
+      "grad_norm": 0.8181896117890224,
+      "learning_rate": 4.848716503131312e-07,
+      "loss": 0.3039,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8047761194029851,
+      "grad_norm": 0.8035708164782546,
+      "learning_rate": 4.834418924834372e-07,
+      "loss": 0.3458,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8050746268656717,
+      "grad_norm": 0.8265207452128133,
+      "learning_rate": 4.820140200864198e-07,
+      "loss": 0.3108,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8053731343283582,
+      "grad_norm": 0.8263073899283685,
+      "learning_rate": 4.805880344571057e-07,
+      "loss": 0.336,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8056716417910448,
+      "grad_norm": 0.7679933299366023,
+      "learning_rate": 4.791639369287557e-07,
+      "loss": 0.3183,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 0.8067035384863501,
+      "learning_rate": 4.777417288328667e-07,
+      "loss": 0.3481,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8062686567164179,
+      "grad_norm": 0.8326924635445278,
+      "learning_rate": 4.763214114991685e-07,
+      "loss": 0.3581,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8065671641791045,
+      "grad_norm": 0.8792635453313903,
+      "learning_rate": 4.749029862556234e-07,
+      "loss": 0.3514,
+      "step": 2702
+    },
+    {
+      "epoch": 0.806865671641791,
+      "grad_norm": 0.8231309020118249,
+      "learning_rate": 4.7348645442842486e-07,
+      "loss": 0.3245,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8071641791044776,
+      "grad_norm": 0.8220357789415259,
+      "learning_rate": 4.720718173419947e-07,
+      "loss": 0.3068,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8074626865671641,
+      "grad_norm": 0.8172111167713195,
+      "learning_rate": 4.706590763189842e-07,
+      "loss": 0.311,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8077611940298507,
+      "grad_norm": 0.8197511984015036,
+      "learning_rate": 4.692482326802722e-07,
+      "loss": 0.3214,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8080597014925374,
+      "grad_norm": 0.795191617688341,
+      "learning_rate": 4.678392877449631e-07,
+      "loss": 0.343,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8083582089552239,
+      "grad_norm": 0.9261906113677095,
+      "learning_rate": 4.6643224283038606e-07,
+      "loss": 0.3497,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8086567164179105,
+      "grad_norm": 0.8073388835958137,
+      "learning_rate": 4.6502709925209415e-07,
+      "loss": 0.3566,
+      "step": 2709
+    },
+    {
+      "epoch": 0.808955223880597,
+      "grad_norm": 0.8727196491973467,
+      "learning_rate": 4.636238583238611e-07,
+      "loss": 0.3442,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8092537313432836,
+      "grad_norm": 0.7474594969428882,
+      "learning_rate": 4.622225213576842e-07,
+      "loss": 0.3188,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8095522388059702,
+      "grad_norm": 0.8418571780348719,
+      "learning_rate": 4.608230896637786e-07,
+      "loss": 0.3451,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8098507462686567,
+      "grad_norm": 0.8706546788357333,
+      "learning_rate": 4.5942556455057907e-07,
+      "loss": 0.328,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8101492537313433,
+      "grad_norm": 0.82036774081707,
+      "learning_rate": 4.580299473247385e-07,
+      "loss": 0.3743,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8104477611940298,
+      "grad_norm": 0.8382687141631925,
+      "learning_rate": 4.566362392911233e-07,
+      "loss": 0.3347,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8107462686567164,
+      "grad_norm": 0.8501600750136287,
+      "learning_rate": 4.5524444175281764e-07,
+      "loss": 0.3122,
+      "step": 2716
+    },
+    {
+      "epoch": 0.811044776119403,
+      "grad_norm": 0.7513505089871687,
+      "learning_rate": 4.5385455601111764e-07,
+      "loss": 0.2863,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8113432835820895,
+      "grad_norm": 0.8351608395297877,
+      "learning_rate": 4.5246658336553285e-07,
+      "loss": 0.3418,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8116417910447761,
+      "grad_norm": 0.8672168533410403,
+      "learning_rate": 4.51080525113784e-07,
+      "loss": 0.3373,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8119402985074626,
+      "grad_norm": 0.891251070972832,
+      "learning_rate": 4.496963825518025e-07,
+      "loss": 0.3924,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8122388059701493,
+      "grad_norm": 1.0096715094848157,
+      "learning_rate": 4.483141569737262e-07,
+      "loss": 0.3393,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8125373134328359,
+      "grad_norm": 0.7978505191567731,
+      "learning_rate": 4.4693384967190336e-07,
+      "loss": 0.3533,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8128358208955224,
+      "grad_norm": 0.8736364444396616,
+      "learning_rate": 4.455554619368874e-07,
+      "loss": 0.3452,
+      "step": 2723
+    },
+    {
+      "epoch": 0.813134328358209,
+      "grad_norm": 0.868351671110689,
+      "learning_rate": 4.441789950574374e-07,
+      "loss": 0.3279,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8134328358208955,
+      "grad_norm": 0.8426500515372894,
+      "learning_rate": 4.4280445032051644e-07,
+      "loss": 0.3321,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8137313432835821,
+      "grad_norm": 0.921670497533751,
+      "learning_rate": 4.4143182901129065e-07,
+      "loss": 0.3849,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8140298507462687,
+      "grad_norm": 0.8427699059645096,
+      "learning_rate": 4.40061132413126e-07,
+      "loss": 0.3491,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8143283582089552,
+      "grad_norm": 1.0244827972367638,
+      "learning_rate": 4.386923618075919e-07,
+      "loss": 0.3091,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8146268656716418,
+      "grad_norm": 0.8797755605970998,
+      "learning_rate": 4.3732551847445514e-07,
+      "loss": 0.3726,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8149253731343283,
+      "grad_norm": 0.8300697363391679,
+      "learning_rate": 4.3596060369168166e-07,
+      "loss": 0.339,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8152238805970149,
+      "grad_norm": 0.8232225916495611,
+      "learning_rate": 4.3459761873543173e-07,
+      "loss": 0.3237,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8155223880597015,
+      "grad_norm": 0.7786052553153007,
+      "learning_rate": 4.3323656488006433e-07,
+      "loss": 0.3131,
+      "step": 2732
+    },
+    {
+      "epoch": 0.815820895522388,
+      "grad_norm": 0.8470582237532246,
+      "learning_rate": 4.3187744339813096e-07,
+      "loss": 0.3337,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8161194029850747,
+      "grad_norm": 0.8893530166853234,
+      "learning_rate": 4.30520255560378e-07,
+      "loss": 0.3582,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8164179104477612,
+      "grad_norm": 0.8553834397028571,
+      "learning_rate": 4.291650026357419e-07,
+      "loss": 0.3852,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8167164179104478,
+      "grad_norm": 0.8952589728174519,
+      "learning_rate": 4.278116858913525e-07,
+      "loss": 0.401,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8170149253731344,
+      "grad_norm": 0.7976255563281726,
+      "learning_rate": 4.2646030659252657e-07,
+      "loss": 0.326,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8173134328358209,
+      "grad_norm": 0.9629301647838667,
+      "learning_rate": 4.2511086600277143e-07,
+      "loss": 0.3878,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8176119402985075,
+      "grad_norm": 0.8360663192438232,
+      "learning_rate": 4.2376336538378084e-07,
+      "loss": 0.3345,
+      "step": 2739
+    },
+    {
+      "epoch": 0.817910447761194,
+      "grad_norm": 0.8380674045458814,
+      "learning_rate": 4.2241780599543563e-07,
+      "loss": 0.3498,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8182089552238806,
+      "grad_norm": 0.8738374475513894,
+      "learning_rate": 4.210741890958009e-07,
+      "loss": 0.3488,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8185074626865672,
+      "grad_norm": 1.27003433660867,
+      "learning_rate": 4.19732515941125e-07,
+      "loss": 0.3179,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8188059701492537,
+      "grad_norm": 0.744478080565413,
+      "learning_rate": 4.1839278778584014e-07,
+      "loss": 0.3175,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8191044776119403,
+      "grad_norm": 0.7803913692575267,
+      "learning_rate": 4.1705500588255956e-07,
+      "loss": 0.2942,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8194029850746268,
+      "grad_norm": 0.7539919885280165,
+      "learning_rate": 4.157191714820766e-07,
+      "loss": 0.2942,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8197014925373134,
+      "grad_norm": 0.8191488598187618,
+      "learning_rate": 4.1438528583336384e-07,
+      "loss": 0.3379,
+      "step": 2746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8663924389620965,
+      "learning_rate": 4.1305335018357235e-07,
+      "loss": 0.3193,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8202985074626866,
+      "grad_norm": 0.8843844251532348,
+      "learning_rate": 4.117233657780287e-07,
+      "loss": 0.3684,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8205970149253732,
+      "grad_norm": 0.7260699768191158,
+      "learning_rate": 4.1039533386023595e-07,
+      "loss": 0.2681,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8208955223880597,
+      "grad_norm": 0.8255147907973385,
+      "learning_rate": 4.09069255671872e-07,
+      "loss": 0.3105,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8211940298507463,
+      "grad_norm": 0.7640188918178487,
+      "learning_rate": 4.077451324527873e-07,
+      "loss": 0.2879,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8214925373134329,
+      "grad_norm": 0.8228944124806514,
+      "learning_rate": 4.06422965441006e-07,
+      "loss": 0.3215,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8217910447761194,
+      "grad_norm": 0.8372918580266345,
+      "learning_rate": 4.0510275587271995e-07,
+      "loss": 0.3465,
+      "step": 2753
+    },
+    {
+      "epoch": 0.822089552238806,
+      "grad_norm": 0.7687323322051585,
+      "learning_rate": 4.0378450498229417e-07,
+      "loss": 0.3526,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8223880597014925,
+      "grad_norm": 0.7942261241475315,
+      "learning_rate": 4.0246821400226095e-07,
+      "loss": 0.3391,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8226865671641791,
+      "grad_norm": 0.9083289569523623,
+      "learning_rate": 4.0115388416332e-07,
+      "loss": 0.3058,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8229850746268657,
+      "grad_norm": 0.7389692614308948,
+      "learning_rate": 3.998415166943381e-07,
+      "loss": 0.2869,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8232835820895522,
+      "grad_norm": 0.7891964084273392,
+      "learning_rate": 3.9853111282234736e-07,
+      "loss": 0.3293,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8235820895522388,
+      "grad_norm": 0.8794771704418654,
+      "learning_rate": 3.972226737725421e-07,
+      "loss": 0.3271,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8238805970149253,
+      "grad_norm": 0.7902448566807511,
+      "learning_rate": 3.9591620076828127e-07,
+      "loss": 0.2814,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8241791044776119,
+      "grad_norm": 0.7750307089786643,
+      "learning_rate": 3.94611695031086e-07,
+      "loss": 0.3438,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8244776119402986,
+      "grad_norm": 0.9123133939275277,
+      "learning_rate": 3.9330915778063666e-07,
+      "loss": 0.3011,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8247761194029851,
+      "grad_norm": 0.8857984121410764,
+      "learning_rate": 3.920085902347745e-07,
+      "loss": 0.3484,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8250746268656717,
+      "grad_norm": 0.849801189339493,
+      "learning_rate": 3.9070999360949824e-07,
+      "loss": 0.3515,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8253731343283582,
+      "grad_norm": 0.8945130952730969,
+      "learning_rate": 3.8941336911896306e-07,
+      "loss": 0.372,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8256716417910448,
+      "grad_norm": 0.8619546550134426,
+      "learning_rate": 3.881187179754828e-07,
+      "loss": 0.3446,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8259701492537314,
+      "grad_norm": 0.910387966103673,
+      "learning_rate": 3.868260413895239e-07,
+      "loss": 0.3719,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8262686567164179,
+      "grad_norm": 0.7564349469839857,
+      "learning_rate": 3.855353405697082e-07,
+      "loss": 0.337,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8265671641791045,
+      "grad_norm": 0.8389296921319255,
+      "learning_rate": 3.842466167228082e-07,
+      "loss": 0.3324,
+      "step": 2769
+    },
+    {
+      "epoch": 0.826865671641791,
+      "grad_norm": 0.7638208844437337,
+      "learning_rate": 3.829598710537502e-07,
+      "loss": 0.2851,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8271641791044776,
+      "grad_norm": 0.7857817130843364,
+      "learning_rate": 3.816751047656098e-07,
+      "loss": 0.33,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8274626865671642,
+      "grad_norm": 0.8418912981277041,
+      "learning_rate": 3.8039231905961204e-07,
+      "loss": 0.2974,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8277611940298507,
+      "grad_norm": 0.9404182488278726,
+      "learning_rate": 3.791115151351313e-07,
+      "loss": 0.3785,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8280597014925373,
+      "grad_norm": 0.8318189034227368,
+      "learning_rate": 3.778326941896862e-07,
+      "loss": 0.353,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8283582089552238,
+      "grad_norm": 0.8300336051160621,
+      "learning_rate": 3.7655585741894454e-07,
+      "loss": 0.3494,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8286567164179105,
+      "grad_norm": 0.7380131940769754,
+      "learning_rate": 3.7528100601671694e-07,
+      "loss": 0.2907,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8289552238805971,
+      "grad_norm": 0.792304457008959,
+      "learning_rate": 3.740081411749588e-07,
+      "loss": 0.3718,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8292537313432836,
+      "grad_norm": 0.8319630537903212,
+      "learning_rate": 3.7273726408376734e-07,
+      "loss": 0.3514,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8295522388059702,
+      "grad_norm": 0.8174896078993995,
+      "learning_rate": 3.714683759313825e-07,
+      "loss": 0.3168,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8298507462686567,
+      "grad_norm": 0.8243230824612728,
+      "learning_rate": 3.7020147790418266e-07,
+      "loss": 0.3156,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8301492537313433,
+      "grad_norm": 0.8235097405737742,
+      "learning_rate": 3.689365711866869e-07,
+      "loss": 0.3215,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8304477611940299,
+      "grad_norm": 0.972966084001807,
+      "learning_rate": 3.676736569615524e-07,
+      "loss": 0.3932,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8307462686567164,
+      "grad_norm": 0.8685667990613507,
+      "learning_rate": 3.664127364095732e-07,
+      "loss": 0.3553,
+      "step": 2783
+    },
+    {
+      "epoch": 0.831044776119403,
+      "grad_norm": 0.9074240672477265,
+      "learning_rate": 3.6515381070967916e-07,
+      "loss": 0.3711,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8313432835820895,
+      "grad_norm": 0.9317748916806734,
+      "learning_rate": 3.6389688103893565e-07,
+      "loss": 0.3269,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8316417910447761,
+      "grad_norm": 0.7730908407067022,
+      "learning_rate": 3.626419485725402e-07,
+      "loss": 0.291,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8319402985074626,
+      "grad_norm": 0.8220545588467743,
+      "learning_rate": 3.6138901448382475e-07,
+      "loss": 0.2818,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8322388059701492,
+      "grad_norm": 0.7753288586198012,
+      "learning_rate": 3.601380799442519e-07,
+      "loss": 0.3117,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8325373134328358,
+      "grad_norm": 0.7319571164319162,
+      "learning_rate": 3.5888914612341506e-07,
+      "loss": 0.2653,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8328358208955224,
+      "grad_norm": 0.9086402808294651,
+      "learning_rate": 3.576422141890376e-07,
+      "loss": 0.3587,
+      "step": 2790
+    },
+    {
+      "epoch": 0.833134328358209,
+      "grad_norm": 0.9224261714666895,
+      "learning_rate": 3.5639728530696944e-07,
+      "loss": 0.3608,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8334328358208956,
+      "grad_norm": 0.8510496483807699,
+      "learning_rate": 3.551543606411889e-07,
+      "loss": 0.3436,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8337313432835821,
+      "grad_norm": 0.8921671526472281,
+      "learning_rate": 3.5391344135380065e-07,
+      "loss": 0.332,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8340298507462687,
+      "grad_norm": 0.8000190887650788,
+      "learning_rate": 3.526745286050334e-07,
+      "loss": 0.3302,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8343283582089552,
+      "grad_norm": 0.7752036186758762,
+      "learning_rate": 3.514376235532413e-07,
+      "loss": 0.3355,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8346268656716418,
+      "grad_norm": 0.789049365859591,
+      "learning_rate": 3.5020272735490023e-07,
+      "loss": 0.3275,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8349253731343284,
+      "grad_norm": 0.827050594275851,
+      "learning_rate": 3.4896984116460697e-07,
+      "loss": 0.3277,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8352238805970149,
+      "grad_norm": 0.9597439299150683,
+      "learning_rate": 3.477389661350811e-07,
+      "loss": 0.3553,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8355223880597015,
+      "grad_norm": 0.8061366601518952,
+      "learning_rate": 3.465101034171603e-07,
+      "loss": 0.3236,
+      "step": 2799
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 0.8294064364634925,
+      "learning_rate": 3.452832541598014e-07,
+      "loss": 0.3105,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8361194029850746,
+      "grad_norm": 0.7827890707493881,
+      "learning_rate": 3.4405841951007907e-07,
+      "loss": 0.3295,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8364179104477611,
+      "grad_norm": 0.8274777597695597,
+      "learning_rate": 3.4283560061318276e-07,
+      "loss": 0.3279,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8367164179104477,
+      "grad_norm": 0.8368484624668843,
+      "learning_rate": 3.416147986124185e-07,
+      "loss": 0.3126,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8370149253731344,
+      "grad_norm": 0.8310718332591835,
+      "learning_rate": 3.403960146492072e-07,
+      "loss": 0.307,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8373134328358209,
+      "grad_norm": 1.1731444253779697,
+      "learning_rate": 3.391792498630819e-07,
+      "loss": 0.3181,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8376119402985075,
+      "grad_norm": 0.8303714814188278,
+      "learning_rate": 3.3796450539168855e-07,
+      "loss": 0.3548,
+      "step": 2806
+    },
+    {
+      "epoch": 0.837910447761194,
+      "grad_norm": 0.8551149029659069,
+      "learning_rate": 3.367517823707822e-07,
+      "loss": 0.3245,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8382089552238806,
+      "grad_norm": 0.8090639602963838,
+      "learning_rate": 3.355410819342303e-07,
+      "loss": 0.3548,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8385074626865672,
+      "grad_norm": 0.8944815883645018,
+      "learning_rate": 3.343324052140079e-07,
+      "loss": 0.3413,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8388059701492537,
+      "grad_norm": 0.832569540017415,
+      "learning_rate": 3.331257533401985e-07,
+      "loss": 0.3283,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8391044776119403,
+      "grad_norm": 0.9058511577320896,
+      "learning_rate": 3.3192112744099255e-07,
+      "loss": 0.3237,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8394029850746269,
+      "grad_norm": 0.8466505427268364,
+      "learning_rate": 3.3071852864268476e-07,
+      "loss": 0.3318,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8397014925373134,
+      "grad_norm": 0.8482105495452263,
+      "learning_rate": 3.2951795806967667e-07,
+      "loss": 0.3832,
+      "step": 2813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.760237333698982,
+      "learning_rate": 3.28319416844472e-07,
+      "loss": 0.3267,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8402985074626865,
+      "grad_norm": 0.907957579652845,
+      "learning_rate": 3.2712290608767796e-07,
+      "loss": 0.353,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8405970149253731,
+      "grad_norm": 0.8309270569530571,
+      "learning_rate": 3.259284269180027e-07,
+      "loss": 0.3212,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8408955223880596,
+      "grad_norm": 0.8751671504075764,
+      "learning_rate": 3.247359804522557e-07,
+      "loss": 0.3945,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8411940298507463,
+      "grad_norm": 0.7771991272718066,
+      "learning_rate": 3.2354556780534423e-07,
+      "loss": 0.3495,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8414925373134329,
+      "grad_norm": 0.7300330257017509,
+      "learning_rate": 3.223571900902758e-07,
+      "loss": 0.2848,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8417910447761194,
+      "grad_norm": 0.7636406065115092,
+      "learning_rate": 3.2117084841815445e-07,
+      "loss": 0.3463,
+      "step": 2820
+    },
+    {
+      "epoch": 0.842089552238806,
+      "grad_norm": 0.7679199725342977,
+      "learning_rate": 3.199865438981808e-07,
+      "loss": 0.3386,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8423880597014926,
+      "grad_norm": 0.8654021210207075,
+      "learning_rate": 3.18804277637651e-07,
+      "loss": 0.374,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8426865671641791,
+      "grad_norm": 0.8588936766950285,
+      "learning_rate": 3.1762405074195505e-07,
+      "loss": 0.2811,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8429850746268657,
+      "grad_norm": 0.8219003046328764,
+      "learning_rate": 3.164458643145757e-07,
+      "loss": 0.3322,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8432835820895522,
+      "grad_norm": 0.8476446887601962,
+      "learning_rate": 3.152697194570892e-07,
+      "loss": 0.3511,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8435820895522388,
+      "grad_norm": 0.8506563729141355,
+      "learning_rate": 3.14095617269162e-07,
+      "loss": 0.3435,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8438805970149253,
+      "grad_norm": 0.8442762907020646,
+      "learning_rate": 3.12923558848551e-07,
+      "loss": 0.2946,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8441791044776119,
+      "grad_norm": 0.9559566043579282,
+      "learning_rate": 3.11753545291103e-07,
+      "loss": 0.3085,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8444776119402985,
+      "grad_norm": 0.8597722096339706,
+      "learning_rate": 3.1058557769075127e-07,
+      "loss": 0.3444,
+      "step": 2829
+    },
+    {
+      "epoch": 0.844776119402985,
+      "grad_norm": 0.7840152793786416,
+      "learning_rate": 3.0941965713951723e-07,
+      "loss": 0.3346,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8450746268656717,
+      "grad_norm": 0.7867462780601826,
+      "learning_rate": 3.0825578472750806e-07,
+      "loss": 0.3301,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8453731343283583,
+      "grad_norm": 1.183250521423748,
+      "learning_rate": 3.070939615429167e-07,
+      "loss": 0.3585,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8456716417910448,
+      "grad_norm": 0.8624837995021237,
+      "learning_rate": 3.0593418867201877e-07,
+      "loss": 0.3036,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8459701492537314,
+      "grad_norm": 0.7736676319547169,
+      "learning_rate": 3.047764671991749e-07,
+      "loss": 0.3162,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8462686567164179,
+      "grad_norm": 0.8290268285441122,
+      "learning_rate": 3.0362079820682485e-07,
+      "loss": 0.3256,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8465671641791045,
+      "grad_norm": 0.7956305622703035,
+      "learning_rate": 3.0246718277549157e-07,
+      "loss": 0.2927,
+      "step": 2836
+    },
+    {
+      "epoch": 0.846865671641791,
+      "grad_norm": 0.9449677011825268,
+      "learning_rate": 3.0131562198377763e-07,
+      "loss": 0.3221,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8471641791044776,
+      "grad_norm": 0.8544674441401762,
+      "learning_rate": 3.001661169083639e-07,
+      "loss": 0.3511,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8474626865671642,
+      "grad_norm": 0.8378591210472813,
+      "learning_rate": 2.990186686240104e-07,
+      "loss": 0.3299,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8477611940298507,
+      "grad_norm": 0.7415214017957397,
+      "learning_rate": 2.978732782035518e-07,
+      "loss": 0.335,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8480597014925373,
+      "grad_norm": 0.9365276617601792,
+      "learning_rate": 2.967299467179019e-07,
+      "loss": 0.3189,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8483582089552238,
+      "grad_norm": 0.803685125509278,
+      "learning_rate": 2.955886752360468e-07,
+      "loss": 0.3054,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8486567164179104,
+      "grad_norm": 0.8054257795996826,
+      "learning_rate": 2.944494648250476e-07,
+      "loss": 0.3736,
+      "step": 2843
+    },
+    {
+      "epoch": 0.848955223880597,
+      "grad_norm": 0.8427531576716745,
+      "learning_rate": 2.93312316550039e-07,
+      "loss": 0.3018,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8492537313432836,
+      "grad_norm": 0.7542191685088523,
+      "learning_rate": 2.9217723147422603e-07,
+      "loss": 0.3425,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8495522388059702,
+      "grad_norm": 0.8846788028740314,
+      "learning_rate": 2.9104421065888566e-07,
+      "loss": 0.3495,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8498507462686568,
+      "grad_norm": 0.7448582027758741,
+      "learning_rate": 2.8991325516336516e-07,
+      "loss": 0.2709,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8501492537313433,
+      "grad_norm": 0.9689479036479011,
+      "learning_rate": 2.887843660450798e-07,
+      "loss": 0.3394,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8504477611940299,
+      "grad_norm": 0.9730142586691932,
+      "learning_rate": 2.8765754435951446e-07,
+      "loss": 0.3772,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8507462686567164,
+      "grad_norm": 0.8711119065481331,
+      "learning_rate": 2.865327911602189e-07,
+      "loss": 0.3281,
+      "step": 2850
+    },
+    {
+      "epoch": 0.851044776119403,
+      "grad_norm": 0.8353842456994517,
+      "learning_rate": 2.8541010749881e-07,
+      "loss": 0.3442,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8513432835820895,
+      "grad_norm": 0.8230749752409625,
+      "learning_rate": 2.8428949442496996e-07,
+      "loss": 0.328,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8516417910447761,
+      "grad_norm": 0.8209618759870918,
+      "learning_rate": 2.831709529864446e-07,
+      "loss": 0.3234,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8519402985074627,
+      "grad_norm": 0.8756915629518247,
+      "learning_rate": 2.820544842290429e-07,
+      "loss": 0.3458,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8522388059701492,
+      "grad_norm": 0.8506109532491893,
+      "learning_rate": 2.809400891966363e-07,
+      "loss": 0.3495,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8525373134328358,
+      "grad_norm": 0.7609796908968595,
+      "learning_rate": 2.798277689311563e-07,
+      "loss": 0.3166,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8528358208955223,
+      "grad_norm": 0.9292199709147685,
+      "learning_rate": 2.7871752447259564e-07,
+      "loss": 0.3637,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8531343283582089,
+      "grad_norm": 0.8140854628051475,
+      "learning_rate": 2.7760935685900576e-07,
+      "loss": 0.3422,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8534328358208956,
+      "grad_norm": 0.8210792005498305,
+      "learning_rate": 2.765032671264961e-07,
+      "loss": 0.2837,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8537313432835821,
+      "grad_norm": 0.7602160506021929,
+      "learning_rate": 2.7539925630923473e-07,
+      "loss": 0.3127,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8540298507462687,
+      "grad_norm": 0.8787744611254319,
+      "learning_rate": 2.7429732543944323e-07,
+      "loss": 0.3677,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8543283582089553,
+      "grad_norm": 0.7774552964487964,
+      "learning_rate": 2.7319747554740096e-07,
+      "loss": 0.3148,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8546268656716418,
+      "grad_norm": 0.8133612375263352,
+      "learning_rate": 2.720997076614407e-07,
+      "loss": 0.301,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8549253731343284,
+      "grad_norm": 0.8837935762329368,
+      "learning_rate": 2.710040228079486e-07,
+      "loss": 0.3679,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8552238805970149,
+      "grad_norm": 0.8348061759356361,
+      "learning_rate": 2.6991042201136327e-07,
+      "loss": 0.351,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8555223880597015,
+      "grad_norm": 0.9256883374968553,
+      "learning_rate": 2.688189062941754e-07,
+      "loss": 0.3454,
+      "step": 2866
+    },
+    {
+      "epoch": 0.855820895522388,
+      "grad_norm": 0.8766267825494243,
+      "learning_rate": 2.677294766769245e-07,
+      "loss": 0.3868,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8561194029850746,
+      "grad_norm": 0.7835550371050818,
+      "learning_rate": 2.6664213417820104e-07,
+      "loss": 0.299,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8564179104477612,
+      "grad_norm": 0.758646833090557,
+      "learning_rate": 2.655568798146443e-07,
+      "loss": 0.3072,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8567164179104477,
+      "grad_norm": 0.7818620602758218,
+      "learning_rate": 2.644737146009402e-07,
+      "loss": 0.3201,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8570149253731343,
+      "grad_norm": 0.7465671035572722,
+      "learning_rate": 2.633926395498218e-07,
+      "loss": 0.3185,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8573134328358208,
+      "grad_norm": 0.8201353059035935,
+      "learning_rate": 2.6231365567206844e-07,
+      "loss": 0.3589,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8576119402985075,
+      "grad_norm": 0.8180993739540929,
+      "learning_rate": 2.6123676397650314e-07,
+      "loss": 0.3187,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8579104477611941,
+      "grad_norm": 0.8056216721587721,
+      "learning_rate": 2.601619654699933e-07,
+      "loss": 0.3326,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8582089552238806,
+      "grad_norm": 0.786235331989252,
+      "learning_rate": 2.5908926115744997e-07,
+      "loss": 0.2917,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8585074626865672,
+      "grad_norm": 0.7791737700058262,
+      "learning_rate": 2.5801865204182486e-07,
+      "loss": 0.3158,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8588059701492538,
+      "grad_norm": 0.8728995256661342,
+      "learning_rate": 2.569501391241122e-07,
+      "loss": 0.387,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8591044776119403,
+      "grad_norm": 0.76411187451701,
+      "learning_rate": 2.5588372340334427e-07,
+      "loss": 0.3338,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8594029850746269,
+      "grad_norm": 0.7959879543698507,
+      "learning_rate": 2.548194058765949e-07,
+      "loss": 0.3169,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8597014925373134,
+      "grad_norm": 0.7821916530349274,
+      "learning_rate": 2.5375718753897493e-07,
+      "loss": 0.3115,
+      "step": 2880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.910323877482503,
+      "learning_rate": 2.5269706938363196e-07,
+      "loss": 0.3825,
+      "step": 2881
+    },
+    {
+      "epoch": 0.8602985074626865,
+      "grad_norm": 0.8333574082581855,
+      "learning_rate": 2.5163905240175175e-07,
+      "loss": 0.3386,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8605970149253731,
+      "grad_norm": 1.4726326278298185,
+      "learning_rate": 2.505831375825532e-07,
+      "loss": 0.3895,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8608955223880597,
+      "grad_norm": 0.7646029930002108,
+      "learning_rate": 2.495293259132914e-07,
+      "loss": 0.3438,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8611940298507462,
+      "grad_norm": 0.917822050579998,
+      "learning_rate": 2.484776183792545e-07,
+      "loss": 0.4223,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8614925373134328,
+      "grad_norm": 0.9260960233947864,
+      "learning_rate": 2.474280159637635e-07,
+      "loss": 0.34,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8617910447761195,
+      "grad_norm": 0.843848719502158,
+      "learning_rate": 2.463805196481714e-07,
+      "loss": 0.3272,
+      "step": 2887
+    },
+    {
+      "epoch": 0.862089552238806,
+      "grad_norm": 0.8251578240160266,
+      "learning_rate": 2.453351304118609e-07,
+      "loss": 0.3677,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8623880597014926,
+      "grad_norm": 0.7232437460080843,
+      "learning_rate": 2.442918492322463e-07,
+      "loss": 0.3187,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8626865671641791,
+      "grad_norm": 0.7702961173547361,
+      "learning_rate": 2.4325067708476924e-07,
+      "loss": 0.3192,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8629850746268657,
+      "grad_norm": 0.8797092318207074,
+      "learning_rate": 2.42211614942901e-07,
+      "loss": 0.3281,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8632835820895522,
+      "grad_norm": 0.7723014760959183,
+      "learning_rate": 2.4117466377813927e-07,
+      "loss": 0.3015,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8635820895522388,
+      "grad_norm": 0.770812872138971,
+      "learning_rate": 2.4013982456000813e-07,
+      "loss": 0.3174,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8638805970149254,
+      "grad_norm": 0.8891261100131612,
+      "learning_rate": 2.3910709825605645e-07,
+      "loss": 0.3458,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8641791044776119,
+      "grad_norm": 0.8361856897100667,
+      "learning_rate": 2.380764858318585e-07,
+      "loss": 0.3498,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8644776119402985,
+      "grad_norm": 0.9085874630351144,
+      "learning_rate": 2.370479882510121e-07,
+      "loss": 0.3246,
+      "step": 2896
+    },
+    {
+      "epoch": 0.864776119402985,
+      "grad_norm": 0.8197815633910296,
+      "learning_rate": 2.3602160647513693e-07,
+      "loss": 0.3716,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8650746268656716,
+      "grad_norm": 0.8746527165561627,
+      "learning_rate": 2.3499734146387565e-07,
+      "loss": 0.3132,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8653731343283582,
+      "grad_norm": 0.8083218694058425,
+      "learning_rate": 2.3397519417489022e-07,
+      "loss": 0.3384,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8656716417910447,
+      "grad_norm": 0.8505934710512888,
+      "learning_rate": 2.3295516556386372e-07,
+      "loss": 0.3654,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8659701492537314,
+      "grad_norm": 0.8163302330969157,
+      "learning_rate": 2.3193725658449794e-07,
+      "loss": 0.3231,
+      "step": 2901
+    },
+    {
+      "epoch": 0.866268656716418,
+      "grad_norm": 0.9049008261636325,
+      "learning_rate": 2.3092146818851357e-07,
+      "loss": 0.3785,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8665671641791045,
+      "grad_norm": 0.8815041271123635,
+      "learning_rate": 2.2990780132564729e-07,
+      "loss": 0.3432,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8668656716417911,
+      "grad_norm": 0.8388624020002282,
+      "learning_rate": 2.288962569436537e-07,
+      "loss": 0.3187,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8671641791044776,
+      "grad_norm": 0.9394777411904285,
+      "learning_rate": 2.2788683598830101e-07,
+      "loss": 0.3456,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8674626865671642,
+      "grad_norm": 0.8140678456634438,
+      "learning_rate": 2.2687953940337403e-07,
+      "loss": 0.3278,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8677611940298507,
+      "grad_norm": 0.8236736666005042,
+      "learning_rate": 2.258743681306702e-07,
+      "loss": 0.3195,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8680597014925373,
+      "grad_norm": 0.8327417961490085,
+      "learning_rate": 2.2487132311000055e-07,
+      "loss": 0.3545,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8683582089552239,
+      "grad_norm": 0.8708133440716065,
+      "learning_rate": 2.2387040527918708e-07,
+      "loss": 0.3167,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8686567164179104,
+      "grad_norm": 0.812026737884432,
+      "learning_rate": 2.2287161557406455e-07,
+      "loss": 0.2901,
+      "step": 2910
+    },
+    {
+      "epoch": 0.868955223880597,
+      "grad_norm": 0.7460255803585844,
+      "learning_rate": 2.21874954928476e-07,
+      "loss": 0.2604,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8692537313432835,
+      "grad_norm": 0.8937718738194345,
+      "learning_rate": 2.2088042427427515e-07,
+      "loss": 0.361,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8695522388059701,
+      "grad_norm": 0.8470926320063893,
+      "learning_rate": 2.19888024541324e-07,
+      "loss": 0.3481,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8698507462686568,
+      "grad_norm": 0.7648679073540151,
+      "learning_rate": 2.188977566574921e-07,
+      "loss": 0.2797,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8701492537313433,
+      "grad_norm": 0.7915971704662402,
+      "learning_rate": 2.179096215486562e-07,
+      "loss": 0.3656,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8704477611940299,
+      "grad_norm": 0.7609586429057615,
+      "learning_rate": 2.1692362013869705e-07,
+      "loss": 0.3232,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8707462686567164,
+      "grad_norm": 0.9187944488879568,
+      "learning_rate": 2.1593975334950363e-07,
+      "loss": 0.3782,
+      "step": 2917
+    },
+    {
+      "epoch": 0.871044776119403,
+      "grad_norm": 0.8742675214436783,
+      "learning_rate": 2.14958022100967e-07,
+      "loss": 0.35,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8713432835820896,
+      "grad_norm": 0.773745759889458,
+      "learning_rate": 2.139784273109813e-07,
+      "loss": 0.3365,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8716417910447761,
+      "grad_norm": 0.9230337430192534,
+      "learning_rate": 2.1300096989544494e-07,
+      "loss": 0.3368,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8719402985074627,
+      "grad_norm": 0.8006731655410944,
+      "learning_rate": 2.1202565076825554e-07,
+      "loss": 0.3213,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8722388059701492,
+      "grad_norm": 0.7820026265974506,
+      "learning_rate": 2.1105247084131308e-07,
+      "loss": 0.3169,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8725373134328358,
+      "grad_norm": 0.7990682212516456,
+      "learning_rate": 2.100814310245175e-07,
+      "loss": 0.3447,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8728358208955224,
+      "grad_norm": 0.793417044723113,
+      "learning_rate": 2.0911253222576673e-07,
+      "loss": 0.3729,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8731343283582089,
+      "grad_norm": 0.8157859780494326,
+      "learning_rate": 2.081457753509586e-07,
+      "loss": 0.2663,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8734328358208955,
+      "grad_norm": 0.7620840057646765,
+      "learning_rate": 2.0718116130398592e-07,
+      "loss": 0.3219,
+      "step": 2926
+    },
+    {
+      "epoch": 0.873731343283582,
+      "grad_norm": 0.8189546759348862,
+      "learning_rate": 2.0621869098673974e-07,
+      "loss": 0.3365,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8740298507462687,
+      "grad_norm": 0.8239722079023885,
+      "learning_rate": 2.0525836529910665e-07,
+      "loss": 0.3365,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8743283582089553,
+      "grad_norm": 0.7714328059715659,
+      "learning_rate": 2.0430018513896754e-07,
+      "loss": 0.3434,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8746268656716418,
+      "grad_norm": 0.8713435989825455,
+      "learning_rate": 2.033441514021975e-07,
+      "loss": 0.3619,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8749253731343284,
+      "grad_norm": 0.8802611699490916,
+      "learning_rate": 2.0239026498266535e-07,
+      "loss": 0.3763,
+      "step": 2931
+    },
+    {
+      "epoch": 0.875223880597015,
+      "grad_norm": 0.8223587805396835,
+      "learning_rate": 2.0143852677223074e-07,
+      "loss": 0.3458,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8755223880597015,
+      "grad_norm": 0.82268739377152,
+      "learning_rate": 2.0048893766074608e-07,
+      "loss": 0.3284,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8758208955223881,
+      "grad_norm": 0.8682099708563312,
+      "learning_rate": 1.9954149853605386e-07,
+      "loss": 0.3522,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8761194029850746,
+      "grad_norm": 0.8184563001143988,
+      "learning_rate": 1.9859621028398697e-07,
+      "loss": 0.326,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8764179104477612,
+      "grad_norm": 0.8396543309711917,
+      "learning_rate": 1.9765307378836695e-07,
+      "loss": 0.3512,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8767164179104477,
+      "grad_norm": 0.7884392531195238,
+      "learning_rate": 1.9671208993100292e-07,
+      "loss": 0.3335,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8770149253731343,
+      "grad_norm": 0.8535929495828234,
+      "learning_rate": 1.957732595916917e-07,
+      "loss": 0.3245,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8773134328358209,
+      "grad_norm": 0.8257674644312324,
+      "learning_rate": 1.9483658364821744e-07,
+      "loss": 0.3041,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8776119402985074,
+      "grad_norm": 0.8124568964960845,
+      "learning_rate": 1.9390206297634912e-07,
+      "loss": 0.3304,
+      "step": 2940
+    },
+    {
+      "epoch": 0.877910447761194,
+      "grad_norm": 0.8215547653162424,
+      "learning_rate": 1.9296969844984054e-07,
+      "loss": 0.2944,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8782089552238806,
+      "grad_norm": 0.8357215000420828,
+      "learning_rate": 1.9203949094043024e-07,
+      "loss": 0.3408,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8785074626865672,
+      "grad_norm": 0.8231402612725451,
+      "learning_rate": 1.9111144131783914e-07,
+      "loss": 0.3437,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8788059701492538,
+      "grad_norm": 0.9024692930586209,
+      "learning_rate": 1.90185550449771e-07,
+      "loss": 0.3731,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8791044776119403,
+      "grad_norm": 0.7126389443706246,
+      "learning_rate": 1.892618192019116e-07,
+      "loss": 0.2756,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8794029850746269,
+      "grad_norm": 0.7750142540785868,
+      "learning_rate": 1.883402484379268e-07,
+      "loss": 0.3293,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8797014925373134,
+      "grad_norm": 0.8251890473704655,
+      "learning_rate": 1.8742083901946317e-07,
+      "loss": 0.2914,
+      "step": 2947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8619570811698483,
+      "learning_rate": 1.8650359180614557e-07,
+      "loss": 0.3061,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8802985074626866,
+      "grad_norm": 0.805795293872737,
+      "learning_rate": 1.85588507655578e-07,
+      "loss": 0.3264,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8805970149253731,
+      "grad_norm": 0.8331813857436107,
+      "learning_rate": 1.8467558742334219e-07,
+      "loss": 0.3111,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8808955223880597,
+      "grad_norm": 0.7786687931228663,
+      "learning_rate": 1.837648319629956e-07,
+      "loss": 0.2909,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8811940298507462,
+      "grad_norm": 0.7905242232392001,
+      "learning_rate": 1.8285624212607322e-07,
+      "loss": 0.3318,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8814925373134328,
+      "grad_norm": 0.7851585515172644,
+      "learning_rate": 1.819498187620841e-07,
+      "loss": 0.2932,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8817910447761194,
+      "grad_norm": 0.8344992885181066,
+      "learning_rate": 1.810455627185112e-07,
+      "loss": 0.3474,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8820895522388059,
+      "grad_norm": 0.8778172382134446,
+      "learning_rate": 1.801434748408129e-07,
+      "loss": 0.3121,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8823880597014926,
+      "grad_norm": 0.8166282816674805,
+      "learning_rate": 1.7924355597241927e-07,
+      "loss": 0.3558,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8826865671641791,
+      "grad_norm": 0.8719815044140756,
+      "learning_rate": 1.7834580695473254e-07,
+      "loss": 0.38,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8829850746268657,
+      "grad_norm": 0.8076792947849909,
+      "learning_rate": 1.774502286271254e-07,
+      "loss": 0.3016,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8832835820895523,
+      "grad_norm": 0.9801413967360424,
+      "learning_rate": 1.7655682182694228e-07,
+      "loss": 0.4087,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8835820895522388,
+      "grad_norm": 0.8289310466533708,
+      "learning_rate": 1.756655873894969e-07,
+      "loss": 0.3196,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8838805970149254,
+      "grad_norm": 0.770440394202249,
+      "learning_rate": 1.7477652614807134e-07,
+      "loss": 0.3295,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8841791044776119,
+      "grad_norm": 0.8473369272270318,
+      "learning_rate": 1.7388963893391676e-07,
+      "loss": 0.3282,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8844776119402985,
+      "grad_norm": 0.8319003049685355,
+      "learning_rate": 1.7300492657625094e-07,
+      "loss": 0.3384,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8847761194029851,
+      "grad_norm": 0.69805804384644,
+      "learning_rate": 1.7212238990225756e-07,
+      "loss": 0.3107,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8850746268656716,
+      "grad_norm": 0.8316282683239232,
+      "learning_rate": 1.7124202973708788e-07,
+      "loss": 0.3394,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8853731343283582,
+      "grad_norm": 0.8189231259197391,
+      "learning_rate": 1.7036384690385681e-07,
+      "loss": 0.3359,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8856716417910447,
+      "grad_norm": 0.8816856984730904,
+      "learning_rate": 1.6948784222364372e-07,
+      "loss": 0.3283,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8859701492537313,
+      "grad_norm": 0.9024059566893298,
+      "learning_rate": 1.6861401651549203e-07,
+      "loss": 0.3404,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8862686567164179,
+      "grad_norm": 0.7660680841699447,
+      "learning_rate": 1.6774237059640764e-07,
+      "loss": 0.3183,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8865671641791045,
+      "grad_norm": 0.9030202060790797,
+      "learning_rate": 1.6687290528135725e-07,
+      "loss": 0.3394,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8868656716417911,
+      "grad_norm": 0.8416102003078251,
+      "learning_rate": 1.660056213832706e-07,
+      "loss": 0.3156,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8871641791044776,
+      "grad_norm": 0.8428340924275619,
+      "learning_rate": 1.651405197130368e-07,
+      "loss": 0.343,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8874626865671642,
+      "grad_norm": 0.8500621805800072,
+      "learning_rate": 1.642776010795047e-07,
+      "loss": 0.3628,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8877611940298508,
+      "grad_norm": 0.99516910680903,
+      "learning_rate": 1.634168662894825e-07,
+      "loss": 0.3386,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8880597014925373,
+      "grad_norm": 0.9229080750641229,
+      "learning_rate": 1.6255831614773594e-07,
+      "loss": 0.3529,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8883582089552239,
+      "grad_norm": 0.791162434060006,
+      "learning_rate": 1.6170195145698842e-07,
+      "loss": 0.3473,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8886567164179104,
+      "grad_norm": 0.7220798985091988,
+      "learning_rate": 1.6084777301792031e-07,
+      "loss": 0.3026,
+      "step": 2977
+    },
+    {
+      "epoch": 0.888955223880597,
+      "grad_norm": 0.747801405803127,
+      "learning_rate": 1.5999578162916723e-07,
+      "loss": 0.2975,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8892537313432836,
+      "grad_norm": 0.8265984529095876,
+      "learning_rate": 1.5914597808732085e-07,
+      "loss": 0.3365,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8895522388059701,
+      "grad_norm": 0.7476801517306761,
+      "learning_rate": 1.58298363186927e-07,
+      "loss": 0.3497,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8898507462686567,
+      "grad_norm": 0.9086508062842112,
+      "learning_rate": 1.5745293772048393e-07,
+      "loss": 0.3154,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8901492537313432,
+      "grad_norm": 0.8224473282900949,
+      "learning_rate": 1.5660970247844437e-07,
+      "loss": 0.354,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8904477611940298,
+      "grad_norm": 0.7683788999407549,
+      "learning_rate": 1.5576865824921295e-07,
+      "loss": 0.3067,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8907462686567165,
+      "grad_norm": 1.013004681398411,
+      "learning_rate": 1.5492980581914535e-07,
+      "loss": 0.3294,
+      "step": 2984
+    },
+    {
+      "epoch": 0.891044776119403,
+      "grad_norm": 0.8640995726797704,
+      "learning_rate": 1.5409314597254864e-07,
+      "loss": 0.3699,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8913432835820896,
+      "grad_norm": 0.8465105815609653,
+      "learning_rate": 1.5325867949167823e-07,
+      "loss": 0.3255,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8916417910447761,
+      "grad_norm": 0.9416694811441639,
+      "learning_rate": 1.5242640715674079e-07,
+      "loss": 0.3372,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8919402985074627,
+      "grad_norm": 0.8527515210916904,
+      "learning_rate": 1.5159632974589028e-07,
+      "loss": 0.3654,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8922388059701493,
+      "grad_norm": 0.8989050252519671,
+      "learning_rate": 1.507684480352292e-07,
+      "loss": 0.3533,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8925373134328358,
+      "grad_norm": 0.8129306542683746,
+      "learning_rate": 1.4994276279880648e-07,
+      "loss": 0.3405,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8928358208955224,
+      "grad_norm": 0.8832284244419019,
+      "learning_rate": 1.4911927480861843e-07,
+      "loss": 0.3266,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8931343283582089,
+      "grad_norm": 0.9068919741311265,
+      "learning_rate": 1.4829798483460471e-07,
+      "loss": 0.3655,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8934328358208955,
+      "grad_norm": 0.7024871886169809,
+      "learning_rate": 1.474788936446525e-07,
+      "loss": 0.3323,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8937313432835821,
+      "grad_norm": 0.7474825332275427,
+      "learning_rate": 1.4666200200459224e-07,
+      "loss": 0.3027,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8940298507462686,
+      "grad_norm": 0.8040131965806334,
+      "learning_rate": 1.458473106781977e-07,
+      "loss": 0.3541,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8943283582089552,
+      "grad_norm": 0.7470197078575609,
+      "learning_rate": 1.450348204271848e-07,
+      "loss": 0.3267,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8946268656716417,
+      "grad_norm": 0.800564447211567,
+      "learning_rate": 1.4422453201121234e-07,
+      "loss": 0.365,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8949253731343284,
+      "grad_norm": 0.7546300002347592,
+      "learning_rate": 1.4341644618788037e-07,
+      "loss": 0.2846,
+      "step": 2998
+    },
+    {
+      "epoch": 0.895223880597015,
+      "grad_norm": 0.7717377048978583,
+      "learning_rate": 1.4261056371272953e-07,
+      "loss": 0.3157,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.9958856245080583,
+      "learning_rate": 1.4180688533924014e-07,
+      "loss": 0.3462,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8958208955223881,
+      "grad_norm": 0.7740493958337736,
+      "learning_rate": 1.4100541181883225e-07,
+      "loss": 0.2892,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8961194029850746,
+      "grad_norm": 0.8871065072732562,
+      "learning_rate": 1.402061439008634e-07,
+      "loss": 0.3561,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8964179104477612,
+      "grad_norm": 0.7521482462008626,
+      "learning_rate": 1.394090823326297e-07,
+      "loss": 0.2957,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8967164179104478,
+      "grad_norm": 0.8691973266040441,
+      "learning_rate": 1.386142278593647e-07,
+      "loss": 0.3191,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8970149253731343,
+      "grad_norm": 0.8681542943381358,
+      "learning_rate": 1.3782158122423783e-07,
+      "loss": 0.362,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8973134328358209,
+      "grad_norm": 0.8478073154828707,
+      "learning_rate": 1.3703114316835436e-07,
+      "loss": 0.3601,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8976119402985074,
+      "grad_norm": 0.7832272639251721,
+      "learning_rate": 1.3624291443075505e-07,
+      "loss": 0.3219,
+      "step": 3007
+    },
+    {
+      "epoch": 0.897910447761194,
+      "grad_norm": 0.8259684062539756,
+      "learning_rate": 1.3545689574841341e-07,
+      "loss": 0.3591,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8982089552238806,
+      "grad_norm": 0.8048015763714349,
+      "learning_rate": 1.3467308785623856e-07,
+      "loss": 0.3752,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8985074626865671,
+      "grad_norm": 0.8513994404977496,
+      "learning_rate": 1.3389149148707176e-07,
+      "loss": 0.3726,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8988059701492538,
+      "grad_norm": 0.8112900887766165,
+      "learning_rate": 1.3311210737168624e-07,
+      "loss": 0.3092,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8991044776119403,
+      "grad_norm": 0.8531040207429486,
+      "learning_rate": 1.3233493623878796e-07,
+      "loss": 0.3495,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8994029850746269,
+      "grad_norm": 0.872868595281458,
+      "learning_rate": 1.3155997881501181e-07,
+      "loss": 0.3465,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8997014925373135,
+      "grad_norm": 0.8868213740208793,
+      "learning_rate": 1.307872358249246e-07,
+      "loss": 0.3229,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8061758493940485,
+      "learning_rate": 1.300167079910225e-07,
+      "loss": 0.3782,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9002985074626866,
+      "grad_norm": 0.7826850147527407,
+      "learning_rate": 1.2924839603372986e-07,
+      "loss": 0.3566,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9005970149253731,
+      "grad_norm": 0.7652865811941827,
+      "learning_rate": 1.2848230067139977e-07,
+      "loss": 0.2958,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9008955223880597,
+      "grad_norm": 0.8684758983402867,
+      "learning_rate": 1.2771842262031293e-07,
+      "loss": 0.3499,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9011940298507463,
+      "grad_norm": 0.8121980361863542,
+      "learning_rate": 1.2695676259467632e-07,
+      "loss": 0.342,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9014925373134328,
+      "grad_norm": 0.8695703424315466,
+      "learning_rate": 1.2619732130662365e-07,
+      "loss": 0.3466,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9017910447761194,
+      "grad_norm": 0.7408411983724599,
+      "learning_rate": 1.2544009946621417e-07,
+      "loss": 0.3154,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9020895522388059,
+      "grad_norm": 0.7520252535004338,
+      "learning_rate": 1.246850977814315e-07,
+      "loss": 0.2926,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9023880597014925,
+      "grad_norm": 0.8015680226963483,
+      "learning_rate": 1.2393231695818435e-07,
+      "loss": 0.3424,
+      "step": 3023
+    },
+    {
+      "epoch": 0.902686567164179,
+      "grad_norm": 0.8844284820852588,
+      "learning_rate": 1.2318175770030388e-07,
+      "loss": 0.3086,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9029850746268657,
+      "grad_norm": 1.4074341950451352,
+      "learning_rate": 1.2243342070954485e-07,
+      "loss": 0.3021,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9032835820895523,
+      "grad_norm": 0.8769608874212167,
+      "learning_rate": 1.216873066855845e-07,
+      "loss": 0.3686,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9035820895522388,
+      "grad_norm": 0.8331928786433949,
+      "learning_rate": 1.2094341632602063e-07,
+      "loss": 0.3711,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9038805970149254,
+      "grad_norm": 0.8114919419794893,
+      "learning_rate": 1.2020175032637278e-07,
+      "loss": 0.3104,
+      "step": 3028
+    },
+    {
+      "epoch": 0.904179104477612,
+      "grad_norm": 0.8382734489680684,
+      "learning_rate": 1.194623093800809e-07,
+      "loss": 0.3222,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9044776119402985,
+      "grad_norm": 0.8090241891495422,
+      "learning_rate": 1.1872509417850425e-07,
+      "loss": 0.3221,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9047761194029851,
+      "grad_norm": 0.7762081906772144,
+      "learning_rate": 1.179901054109206e-07,
+      "loss": 0.3045,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9050746268656716,
+      "grad_norm": 0.7687394313447948,
+      "learning_rate": 1.1725734376452691e-07,
+      "loss": 0.3097,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9053731343283582,
+      "grad_norm": 0.8494786804724461,
+      "learning_rate": 1.1652680992443765e-07,
+      "loss": 0.2838,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9056716417910448,
+      "grad_norm": 0.7426720969321979,
+      "learning_rate": 1.1579850457368342e-07,
+      "loss": 0.3068,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9059701492537313,
+      "grad_norm": 0.8542802364049378,
+      "learning_rate": 1.1507242839321281e-07,
+      "loss": 0.3151,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9062686567164179,
+      "grad_norm": 0.8047828696688691,
+      "learning_rate": 1.1434858206188864e-07,
+      "loss": 0.3468,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9065671641791044,
+      "grad_norm": 0.8784472135696588,
+      "learning_rate": 1.1362696625648983e-07,
+      "loss": 0.3373,
+      "step": 3037
+    },
+    {
+      "epoch": 0.906865671641791,
+      "grad_norm": 0.8176344259816754,
+      "learning_rate": 1.1290758165171e-07,
+      "loss": 0.3417,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9071641791044777,
+      "grad_norm": 0.84514932245467,
+      "learning_rate": 1.1219042892015586e-07,
+      "loss": 0.3071,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9074626865671642,
+      "grad_norm": 0.7963261643887735,
+      "learning_rate": 1.1147550873234769e-07,
+      "loss": 0.2971,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9077611940298508,
+      "grad_norm": 0.8505558295228394,
+      "learning_rate": 1.1076282175671832e-07,
+      "loss": 0.2992,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9080597014925373,
+      "grad_norm": 0.9346232263581281,
+      "learning_rate": 1.1005236865961277e-07,
+      "loss": 0.3754,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9083582089552239,
+      "grad_norm": 0.7821481899052896,
+      "learning_rate": 1.0934415010528748e-07,
+      "loss": 0.3066,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9086567164179105,
+      "grad_norm": 0.8906980599877051,
+      "learning_rate": 1.0863816675590999e-07,
+      "loss": 0.3545,
+      "step": 3044
+    },
+    {
+      "epoch": 0.908955223880597,
+      "grad_norm": 0.8482058328893826,
+      "learning_rate": 1.0793441927155673e-07,
+      "loss": 0.3525,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9092537313432836,
+      "grad_norm": 0.8148838793718065,
+      "learning_rate": 1.0723290831021471e-07,
+      "loss": 0.3553,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9095522388059701,
+      "grad_norm": 0.8334685767932932,
+      "learning_rate": 1.0653363452777953e-07,
+      "loss": 0.3473,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9098507462686567,
+      "grad_norm": 0.7622666260376155,
+      "learning_rate": 1.0583659857805545e-07,
+      "loss": 0.3193,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9101492537313433,
+      "grad_norm": 0.8736868598440297,
+      "learning_rate": 1.0514180111275391e-07,
+      "loss": 0.3266,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9104477611940298,
+      "grad_norm": 0.8589697295961596,
+      "learning_rate": 1.044492427814936e-07,
+      "loss": 0.3075,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9107462686567164,
+      "grad_norm": 0.8373246294908004,
+      "learning_rate": 1.0375892423179962e-07,
+      "loss": 0.2927,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9110447761194029,
+      "grad_norm": 0.9038840022385726,
+      "learning_rate": 1.030708461091029e-07,
+      "loss": 0.3729,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9113432835820896,
+      "grad_norm": 0.846932324486075,
+      "learning_rate": 1.0238500905673992e-07,
+      "loss": 0.3424,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9116417910447762,
+      "grad_norm": 0.7863656327081202,
+      "learning_rate": 1.0170141371595138e-07,
+      "loss": 0.3041,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9119402985074627,
+      "grad_norm": 0.861611603226045,
+      "learning_rate": 1.0102006072588239e-07,
+      "loss": 0.303,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9122388059701493,
+      "grad_norm": 0.8354159151319559,
+      "learning_rate": 1.0034095072358196e-07,
+      "loss": 0.2905,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9125373134328358,
+      "grad_norm": 0.9076705935835198,
+      "learning_rate": 9.966408434400026e-08,
+      "loss": 0.3596,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9128358208955224,
+      "grad_norm": 0.7940993218471586,
+      "learning_rate": 9.898946221999162e-08,
+      "loss": 0.344,
+      "step": 3058
+    },
+    {
+      "epoch": 0.913134328358209,
+      "grad_norm": 0.8683284494278827,
+      "learning_rate": 9.83170849823109e-08,
+      "loss": 0.2837,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9134328358208955,
+      "grad_norm": 0.8679841336149956,
+      "learning_rate": 9.764695325961471e-08,
+      "loss": 0.3426,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9137313432835821,
+      "grad_norm": 0.9136469022117164,
+      "learning_rate": 9.697906767845988e-08,
+      "loss": 0.3216,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9140298507462686,
+      "grad_norm": 0.9280938115066125,
+      "learning_rate": 9.631342886330302e-08,
+      "loss": 0.3441,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9143283582089552,
+      "grad_norm": 0.7989785840668151,
+      "learning_rate": 9.565003743650019e-08,
+      "loss": 0.3037,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9146268656716418,
+      "grad_norm": 0.7840797063644848,
+      "learning_rate": 9.498889401830636e-08,
+      "loss": 0.3593,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9149253731343283,
+      "grad_norm": 0.8742322626184073,
+      "learning_rate": 9.432999922687397e-08,
+      "loss": 0.3853,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9152238805970149,
+      "grad_norm": 0.8309689917110207,
+      "learning_rate": 9.367335367825442e-08,
+      "loss": 0.3548,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9155223880597015,
+      "grad_norm": 0.7682603888770035,
+      "learning_rate": 9.301895798639465e-08,
+      "loss": 0.3458,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9158208955223881,
+      "grad_norm": 0.8406021828495258,
+      "learning_rate": 9.236681276313914e-08,
+      "loss": 0.3439,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9161194029850747,
+      "grad_norm": 0.7676640227562936,
+      "learning_rate": 9.171691861822735e-08,
+      "loss": 0.3377,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9164179104477612,
+      "grad_norm": 0.8241123436966133,
+      "learning_rate": 9.106927615929462e-08,
+      "loss": 0.3461,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9167164179104478,
+      "grad_norm": 0.7865175510079805,
+      "learning_rate": 9.042388599187158e-08,
+      "loss": 0.3356,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9170149253731343,
+      "grad_norm": 0.8416832778666247,
+      "learning_rate": 8.978074871938109e-08,
+      "loss": 0.3225,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9173134328358209,
+      "grad_norm": 0.925094121401764,
+      "learning_rate": 8.91398649431413e-08,
+      "loss": 0.35,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9176119402985075,
+      "grad_norm": 0.9245170617415597,
+      "learning_rate": 8.850123526236292e-08,
+      "loss": 0.3473,
+      "step": 3074
+    },
+    {
+      "epoch": 0.917910447761194,
+      "grad_norm": 0.8830284323162437,
+      "learning_rate": 8.786486027414942e-08,
+      "loss": 0.3263,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9182089552238806,
+      "grad_norm": 0.8322762071581037,
+      "learning_rate": 8.723074057349568e-08,
+      "loss": 0.3267,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9185074626865671,
+      "grad_norm": 0.8218742132897513,
+      "learning_rate": 8.659887675328826e-08,
+      "loss": 0.3525,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9188059701492537,
+      "grad_norm": 0.7524701289147285,
+      "learning_rate": 8.596926940430406e-08,
+      "loss": 0.347,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9191044776119403,
+      "grad_norm": 0.7973411802668999,
+      "learning_rate": 8.534191911521106e-08,
+      "loss": 0.3203,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9194029850746268,
+      "grad_norm": 0.755081736754136,
+      "learning_rate": 8.471682647256619e-08,
+      "loss": 0.3012,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9197014925373135,
+      "grad_norm": 0.7506268914890571,
+      "learning_rate": 8.409399206081609e-08,
+      "loss": 0.3365,
+      "step": 3081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8943820675775366,
+      "learning_rate": 8.347341646229578e-08,
+      "loss": 0.3694,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9202985074626866,
+      "grad_norm": 0.8565711585170784,
+      "learning_rate": 8.285510025722781e-08,
+      "loss": 0.3066,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9205970149253732,
+      "grad_norm": 0.6915296836970478,
+      "learning_rate": 8.223904402372334e-08,
+      "loss": 0.2471,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9208955223880597,
+      "grad_norm": 0.7910425039296988,
+      "learning_rate": 8.16252483377794e-08,
+      "loss": 0.3063,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9211940298507463,
+      "grad_norm": 0.8626761328335215,
+      "learning_rate": 8.101371377328055e-08,
+      "loss": 0.323,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9214925373134328,
+      "grad_norm": 0.811154258383053,
+      "learning_rate": 8.040444090199634e-08,
+      "loss": 0.3412,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9217910447761194,
+      "grad_norm": 0.8740316342417893,
+      "learning_rate": 7.97974302935825e-08,
+      "loss": 0.3647,
+      "step": 3088
+    },
+    {
+      "epoch": 0.922089552238806,
+      "grad_norm": 0.9132701160611257,
+      "learning_rate": 7.919268251557838e-08,
+      "loss": 0.3722,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9223880597014925,
+      "grad_norm": 0.7691408535525245,
+      "learning_rate": 7.85901981334089e-08,
+      "loss": 0.3096,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9226865671641791,
+      "grad_norm": 0.839417162083958,
+      "learning_rate": 7.798997771038236e-08,
+      "loss": 0.3487,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9229850746268656,
+      "grad_norm": 0.869748021900408,
+      "learning_rate": 7.739202180769013e-08,
+      "loss": 0.3477,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9232835820895522,
+      "grad_norm": 0.8159308852427606,
+      "learning_rate": 7.679633098440609e-08,
+      "loss": 0.3281,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9235820895522389,
+      "grad_norm": 0.9129695074897101,
+      "learning_rate": 7.620290579748723e-08,
+      "loss": 0.356,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9238805970149254,
+      "grad_norm": 0.8713069828640914,
+      "learning_rate": 7.561174680177114e-08,
+      "loss": 0.329,
+      "step": 3095
+    },
+    {
+      "epoch": 0.924179104477612,
+      "grad_norm": 0.7559944876673397,
+      "learning_rate": 7.502285454997732e-08,
+      "loss": 0.3056,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9244776119402985,
+      "grad_norm": 0.829846374274675,
+      "learning_rate": 7.443622959270535e-08,
+      "loss": 0.3374,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9247761194029851,
+      "grad_norm": 0.9373023004429427,
+      "learning_rate": 7.385187247843567e-08,
+      "loss": 0.3328,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9250746268656717,
+      "grad_norm": 0.7512508155243499,
+      "learning_rate": 7.32697837535279e-08,
+      "loss": 0.3161,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9253731343283582,
+      "grad_norm": 0.7853324032451395,
+      "learning_rate": 7.268996396222056e-08,
+      "loss": 0.3507,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9256716417910448,
+      "grad_norm": 0.8502143296601243,
+      "learning_rate": 7.211241364663113e-08,
+      "loss": 0.3854,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9259701492537313,
+      "grad_norm": 0.8250790098331866,
+      "learning_rate": 7.153713334675516e-08,
+      "loss": 0.3361,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9262686567164179,
+      "grad_norm": 0.8122977611367114,
+      "learning_rate": 7.096412360046545e-08,
+      "loss": 0.3371,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9265671641791045,
+      "grad_norm": 0.8020325644972325,
+      "learning_rate": 7.039338494351261e-08,
+      "loss": 0.327,
+      "step": 3104
+    },
+    {
+      "epoch": 0.926865671641791,
+      "grad_norm": 0.8098687914753534,
+      "learning_rate": 6.982491790952284e-08,
+      "loss": 0.3228,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9271641791044776,
+      "grad_norm": 0.8343057527737529,
+      "learning_rate": 6.925872302999931e-08,
+      "loss": 0.3383,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9274626865671641,
+      "grad_norm": 0.8582956934198305,
+      "learning_rate": 6.86948008343205e-08,
+      "loss": 0.3705,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9277611940298508,
+      "grad_norm": 0.8051927725802557,
+      "learning_rate": 6.813315184973968e-08,
+      "loss": 0.3325,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9280597014925374,
+      "grad_norm": 0.7892911817116857,
+      "learning_rate": 6.757377660138508e-08,
+      "loss": 0.3413,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9283582089552239,
+      "grad_norm": 0.7927728140846071,
+      "learning_rate": 6.701667561225894e-08,
+      "loss": 0.3533,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9286567164179105,
+      "grad_norm": 0.8148550066294056,
+      "learning_rate": 6.64618494032368e-08,
+      "loss": 0.3803,
+      "step": 3111
+    },
+    {
+      "epoch": 0.928955223880597,
+      "grad_norm": 0.8284362735918354,
+      "learning_rate": 6.590929849306788e-08,
+      "loss": 0.3344,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9292537313432836,
+      "grad_norm": 0.7868083459028574,
+      "learning_rate": 6.535902339837392e-08,
+      "loss": 0.3141,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9295522388059702,
+      "grad_norm": 0.8202287813945078,
+      "learning_rate": 6.481102463364864e-08,
+      "loss": 0.3267,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9298507462686567,
+      "grad_norm": 1.011406528561342,
+      "learning_rate": 6.426530271125775e-08,
+      "loss": 0.3625,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9301492537313433,
+      "grad_norm": 0.9568538573975606,
+      "learning_rate": 6.372185814143756e-08,
+      "loss": 0.3092,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9304477611940298,
+      "grad_norm": 0.8120426908903172,
+      "learning_rate": 6.31806914322955e-08,
+      "loss": 0.2998,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9307462686567164,
+      "grad_norm": 0.8154259717600727,
+      "learning_rate": 6.264180308980933e-08,
+      "loss": 0.3441,
+      "step": 3118
+    },
+    {
+      "epoch": 0.931044776119403,
+      "grad_norm": 0.8366039008724436,
+      "learning_rate": 6.210519361782685e-08,
+      "loss": 0.3123,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9313432835820895,
+      "grad_norm": 0.7249537665139266,
+      "learning_rate": 6.157086351806451e-08,
+      "loss": 0.3187,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9316417910447761,
+      "grad_norm": 0.834336635014129,
+      "learning_rate": 6.103881329010797e-08,
+      "loss": 0.3238,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9319402985074627,
+      "grad_norm": 0.921548871205445,
+      "learning_rate": 6.050904343141095e-08,
+      "loss": 0.3376,
+      "step": 3122
+    },
+    {
+      "epoch": 0.9322388059701493,
+      "grad_norm": 0.8938173280654804,
+      "learning_rate": 5.998155443729586e-08,
+      "loss": 0.3725,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9325373134328359,
+      "grad_norm": 0.8368698873876483,
+      "learning_rate": 5.9456346800951805e-08,
+      "loss": 0.352,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9328358208955224,
+      "grad_norm": 0.8916547607557948,
+      "learning_rate": 5.8933421013435135e-08,
+      "loss": 0.3883,
+      "step": 3125
+    },
+    {
+      "epoch": 0.933134328358209,
+      "grad_norm": 0.8029972819508524,
+      "learning_rate": 5.841277756366892e-08,
+      "loss": 0.273,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9334328358208955,
+      "grad_norm": 0.9090008716050973,
+      "learning_rate": 5.7894416938441834e-08,
+      "loss": 0.3846,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9337313432835821,
+      "grad_norm": 0.737087695091503,
+      "learning_rate": 5.737833962240841e-08,
+      "loss": 0.3197,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9340298507462687,
+      "grad_norm": 0.9304110314164876,
+      "learning_rate": 5.68645460980885e-08,
+      "loss": 0.3527,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9343283582089552,
+      "grad_norm": 0.8089432663681636,
+      "learning_rate": 5.635303684586646e-08,
+      "loss": 0.3151,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9346268656716418,
+      "grad_norm": 1.165243151408962,
+      "learning_rate": 5.584381234399111e-08,
+      "loss": 0.3267,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9349253731343283,
+      "grad_norm": 0.7348025145722581,
+      "learning_rate": 5.533687306857466e-08,
+      "loss": 0.2696,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9352238805970149,
+      "grad_norm": 0.8262555992907743,
+      "learning_rate": 5.483221949359324e-08,
+      "loss": 0.3516,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9355223880597014,
+      "grad_norm": 0.9156843006759944,
+      "learning_rate": 5.432985209088526e-08,
+      "loss": 0.3481,
+      "step": 3134
+    },
+    {
+      "epoch": 0.935820895522388,
+      "grad_norm": 0.8572239132276424,
+      "learning_rate": 5.3829771330152495e-08,
+      "loss": 0.3254,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9361194029850747,
+      "grad_norm": 1.028962020667184,
+      "learning_rate": 5.3331977678958145e-08,
+      "loss": 0.3123,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9364179104477612,
+      "grad_norm": 0.83452568472891,
+      "learning_rate": 5.2836471602727144e-08,
+      "loss": 0.3421,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9367164179104478,
+      "grad_norm": 0.8286325370096002,
+      "learning_rate": 5.234325356474529e-08,
+      "loss": 0.3499,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9370149253731344,
+      "grad_norm": 0.8160036150223631,
+      "learning_rate": 5.185232402615953e-08,
+      "loss": 0.3105,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9373134328358209,
+      "grad_norm": 0.8066012270688009,
+      "learning_rate": 5.1363683445977144e-08,
+      "loss": 0.3075,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9376119402985075,
+      "grad_norm": 0.8033304417193656,
+      "learning_rate": 5.087733228106517e-08,
+      "loss": 0.3218,
+      "step": 3141
+    },
+    {
+      "epoch": 0.937910447761194,
+      "grad_norm": 0.8465912538140584,
+      "learning_rate": 5.0393270986150155e-08,
+      "loss": 0.3174,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9382089552238806,
+      "grad_norm": 0.8023677069921564,
+      "learning_rate": 4.991150001381756e-08,
+      "loss": 0.3562,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9385074626865672,
+      "grad_norm": 0.9290258509536511,
+      "learning_rate": 4.9432019814511235e-08,
+      "loss": 0.3718,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9388059701492537,
+      "grad_norm": 0.8889616159939564,
+      "learning_rate": 4.8954830836533963e-08,
+      "loss": 0.3495,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9391044776119403,
+      "grad_norm": 0.811509055722367,
+      "learning_rate": 4.847993352604524e-08,
+      "loss": 0.3429,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9394029850746268,
+      "grad_norm": 0.8199197753898051,
+      "learning_rate": 4.800732832706323e-08,
+      "loss": 0.3656,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9397014925373134,
+      "grad_norm": 0.9063245467477791,
+      "learning_rate": 4.753701568146168e-08,
+      "loss": 0.3317,
+      "step": 3148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8342353851057329,
+      "learning_rate": 4.706899602897136e-08,
+      "loss": 0.3312,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9402985074626866,
+      "grad_norm": 0.8720693495116465,
+      "learning_rate": 4.6603269807179716e-08,
+      "loss": 0.3593,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9405970149253732,
+      "grad_norm": 0.8586192607934766,
+      "learning_rate": 4.6139837451529004e-08,
+      "loss": 0.2832,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9408955223880597,
+      "grad_norm": 0.9695827592543284,
+      "learning_rate": 4.5678699395317326e-08,
+      "loss": 0.409,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9411940298507463,
+      "grad_norm": 0.8419990980437638,
+      "learning_rate": 4.5219856069697866e-08,
+      "loss": 0.3268,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9414925373134329,
+      "grad_norm": 0.8452212009450679,
+      "learning_rate": 4.476330790367717e-08,
+      "loss": 0.3211,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9417910447761194,
+      "grad_norm": 0.8498008661362966,
+      "learning_rate": 4.4309055324117386e-08,
+      "loss": 0.3485,
+      "step": 3155
+    },
+    {
+      "epoch": 0.942089552238806,
+      "grad_norm": 0.8431790042502505,
+      "learning_rate": 4.385709875573324e-08,
+      "loss": 0.3482,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9423880597014925,
+      "grad_norm": 0.8448462756069092,
+      "learning_rate": 4.340743862109309e-08,
+      "loss": 0.3571,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9426865671641791,
+      "grad_norm": 0.8098626117139088,
+      "learning_rate": 4.296007534061869e-08,
+      "loss": 0.3451,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9429850746268656,
+      "grad_norm": 0.9323258425698503,
+      "learning_rate": 4.2515009332582954e-08,
+      "loss": 0.3325,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9432835820895522,
+      "grad_norm": 0.8628485400460011,
+      "learning_rate": 4.207224101311247e-08,
+      "loss": 0.3334,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9435820895522388,
+      "grad_norm": 0.8956281853483311,
+      "learning_rate": 4.163177079618441e-08,
+      "loss": 0.3362,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9438805970149253,
+      "grad_norm": 0.8074849186203403,
+      "learning_rate": 4.1193599093627964e-08,
+      "loss": 0.3114,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9441791044776119,
+      "grad_norm": 0.8490755318936722,
+      "learning_rate": 4.0757726315122646e-08,
+      "loss": 0.3558,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9444776119402986,
+      "grad_norm": 0.7610965070669408,
+      "learning_rate": 4.032415286819941e-08,
+      "loss": 0.3158,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9447761194029851,
+      "grad_norm": 0.7655473588220373,
+      "learning_rate": 3.989287915823842e-08,
+      "loss": 0.3121,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9450746268656717,
+      "grad_norm": 0.8444397049637858,
+      "learning_rate": 3.9463905588470186e-08,
+      "loss": 0.3323,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9453731343283582,
+      "grad_norm": 0.8236973140905199,
+      "learning_rate": 3.9037232559974714e-08,
+      "loss": 0.3348,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9456716417910448,
+      "grad_norm": 0.8003724680619415,
+      "learning_rate": 3.861286047168067e-08,
+      "loss": 0.3795,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9459701492537314,
+      "grad_norm": 1.0280919880161592,
+      "learning_rate": 3.8190789720365665e-08,
+      "loss": 0.3209,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9462686567164179,
+      "grad_norm": 0.8000546455088566,
+      "learning_rate": 3.777102070065569e-08,
+      "loss": 0.3505,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9465671641791045,
+      "grad_norm": 0.8935839440883604,
+      "learning_rate": 3.735355380502431e-08,
+      "loss": 0.3476,
+      "step": 3171
+    },
+    {
+      "epoch": 0.946865671641791,
+      "grad_norm": 0.8848630298774833,
+      "learning_rate": 3.693838942379291e-08,
+      "loss": 0.3168,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9471641791044776,
+      "grad_norm": 0.7685629417514018,
+      "learning_rate": 3.6525527945130424e-08,
+      "loss": 0.3267,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9474626865671641,
+      "grad_norm": 0.7874491430144344,
+      "learning_rate": 3.611496975505169e-08,
+      "loss": 0.3076,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9477611940298507,
+      "grad_norm": 0.8513167764285567,
+      "learning_rate": 3.5706715237419366e-08,
+      "loss": 0.3351,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9480597014925373,
+      "grad_norm": 0.7679536427243635,
+      "learning_rate": 3.5300764773940896e-08,
+      "loss": 0.333,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9483582089552239,
+      "grad_norm": 0.7672420889067244,
+      "learning_rate": 3.4897118744170175e-08,
+      "loss": 0.3435,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9486567164179105,
+      "grad_norm": 0.872282837076114,
+      "learning_rate": 3.4495777525506703e-08,
+      "loss": 0.3501,
+      "step": 3178
+    },
+    {
+      "epoch": 0.948955223880597,
+      "grad_norm": 0.8460843443255879,
+      "learning_rate": 3.4096741493194196e-08,
+      "loss": 0.3564,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9492537313432836,
+      "grad_norm": 0.8100002833851134,
+      "learning_rate": 3.3700011020322e-08,
+      "loss": 0.3143,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9495522388059702,
+      "grad_norm": 0.7984025736476822,
+      "learning_rate": 3.330558647782312e-08,
+      "loss": 0.3666,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9498507462686567,
+      "grad_norm": 0.754287300991587,
+      "learning_rate": 3.291346823447533e-08,
+      "loss": 0.3031,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9501492537313433,
+      "grad_norm": 0.7752734270024597,
+      "learning_rate": 3.252365665689955e-08,
+      "loss": 0.3004,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9504477611940298,
+      "grad_norm": 0.8191799374185267,
+      "learning_rate": 3.213615210955978e-08,
+      "loss": 0.3527,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9507462686567164,
+      "grad_norm": 0.907730673037401,
+      "learning_rate": 3.1750954954763716e-08,
+      "loss": 0.3272,
+      "step": 3185
+    },
+    {
+      "epoch": 0.951044776119403,
+      "grad_norm": 0.767757547991333,
+      "learning_rate": 3.136806555266103e-08,
+      "loss": 0.278,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9513432835820895,
+      "grad_norm": 0.8243908687352361,
+      "learning_rate": 3.098748426124398e-08,
+      "loss": 0.3542,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9516417910447761,
+      "grad_norm": 0.712848097621701,
+      "learning_rate": 3.0609211436347095e-08,
+      "loss": 0.2936,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9519402985074626,
+      "grad_norm": 0.8024929014526022,
+      "learning_rate": 3.02332474316458e-08,
+      "loss": 0.325,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9522388059701492,
+      "grad_norm": 0.8494487804866518,
+      "learning_rate": 2.985959259865778e-08,
+      "loss": 0.3904,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9525373134328359,
+      "grad_norm": 0.7778696644607079,
+      "learning_rate": 2.9488247286740546e-08,
+      "loss": 0.3079,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9528358208955224,
+      "grad_norm": 0.7349126770422241,
+      "learning_rate": 2.9119211843093574e-08,
+      "loss": 0.2997,
+      "step": 3192
+    },
+    {
+      "epoch": 0.953134328358209,
+      "grad_norm": 0.7872392631708472,
+      "learning_rate": 2.8752486612755593e-08,
+      "loss": 0.2926,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9534328358208956,
+      "grad_norm": 0.8767962603475438,
+      "learning_rate": 2.8388071938605655e-08,
+      "loss": 0.3712,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9537313432835821,
+      "grad_norm": 0.779215026325899,
+      "learning_rate": 2.802596816136316e-08,
+      "loss": 0.3147,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9540298507462687,
+      "grad_norm": 0.7598752044921813,
+      "learning_rate": 2.766617561958618e-08,
+      "loss": 0.3079,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9543283582089552,
+      "grad_norm": 0.7165696161507437,
+      "learning_rate": 2.7308694649671453e-08,
+      "loss": 0.3027,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9546268656716418,
+      "grad_norm": 0.8812874250973933,
+      "learning_rate": 2.6953525585855233e-08,
+      "loss": 0.3666,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9549253731343283,
+      "grad_norm": 0.876566087660891,
+      "learning_rate": 2.660066876021189e-08,
+      "loss": 0.3457,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.7854233313802679,
+      "learning_rate": 2.625012450265446e-08,
+      "loss": 0.3486,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9555223880597015,
+      "grad_norm": 0.8325761566945635,
+      "learning_rate": 2.5901893140932444e-08,
+      "loss": 0.3254,
+      "step": 3201
+    },
+    {
+      "epoch": 0.955820895522388,
+      "grad_norm": 0.8166375313979258,
+      "learning_rate": 2.555597500063456e-08,
+      "loss": 0.3446,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9561194029850746,
+      "grad_norm": 0.9210937286901646,
+      "learning_rate": 2.521237040518515e-08,
+      "loss": 0.3436,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9564179104477611,
+      "grad_norm": 0.7559699274090332,
+      "learning_rate": 2.4871079675846398e-08,
+      "loss": 0.3649,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9567164179104478,
+      "grad_norm": 0.7400183635230114,
+      "learning_rate": 2.4532103131716668e-08,
+      "loss": 0.2962,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9570149253731344,
+      "grad_norm": 0.7802880622650665,
+      "learning_rate": 2.419544108973104e-08,
+      "loss": 0.2953,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9573134328358209,
+      "grad_norm": 0.8494660495105342,
+      "learning_rate": 2.3861093864660233e-08,
+      "loss": 0.3352,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9576119402985075,
+      "grad_norm": 0.8112621260443245,
+      "learning_rate": 2.3529061769110573e-08,
+      "loss": 0.3118,
+      "step": 3208
+    },
+    {
+      "epoch": 0.957910447761194,
+      "grad_norm": 0.7589488656121378,
+      "learning_rate": 2.3199345113524007e-08,
+      "loss": 0.2918,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9582089552238806,
+      "grad_norm": 0.8313292700601174,
+      "learning_rate": 2.287194420617783e-08,
+      "loss": 0.3353,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9585074626865672,
+      "grad_norm": 0.871232307176741,
+      "learning_rate": 2.254685935318357e-08,
+      "loss": 0.273,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9588059701492537,
+      "grad_norm": 0.8200691382729434,
+      "learning_rate": 2.222409085848809e-08,
+      "loss": 0.3358,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9591044776119403,
+      "grad_norm": 0.8045701864458568,
+      "learning_rate": 2.1903639023871658e-08,
+      "loss": 0.3378,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9594029850746268,
+      "grad_norm": 0.8197181943630443,
+      "learning_rate": 2.1585504148949056e-08,
+      "loss": 0.3174,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9597014925373134,
+      "grad_norm": 0.8494644190786257,
+      "learning_rate": 2.1269686531168456e-08,
+      "loss": 0.3223,
+      "step": 3215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9396349631933137,
+      "learning_rate": 2.095618646581199e-08,
+      "loss": 0.3686,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9602985074626865,
+      "grad_norm": 0.8323583157384636,
+      "learning_rate": 2.064500424599436e-08,
+      "loss": 0.3196,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9605970149253731,
+      "grad_norm": 0.8892166342620139,
+      "learning_rate": 2.0336140162663386e-08,
+      "loss": 0.3216,
+      "step": 3218
+    },
+    {
+      "epoch": 0.9608955223880598,
+      "grad_norm": 0.8483612538763416,
+      "learning_rate": 2.002959450459918e-08,
+      "loss": 0.3309,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9611940298507463,
+      "grad_norm": 0.7559863326126044,
+      "learning_rate": 1.9725367558415253e-08,
+      "loss": 0.3091,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9614925373134329,
+      "grad_norm": 0.8020646850199934,
+      "learning_rate": 1.9423459608555462e-08,
+      "loss": 0.3299,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9617910447761194,
+      "grad_norm": 0.7418019278634108,
+      "learning_rate": 1.912387093729706e-08,
+      "loss": 0.302,
+      "step": 3222
+    },
+    {
+      "epoch": 0.962089552238806,
+      "grad_norm": 0.7754195649660184,
+      "learning_rate": 1.8826601824747936e-08,
+      "loss": 0.3138,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9623880597014925,
+      "grad_norm": 0.9516965112975473,
+      "learning_rate": 1.8531652548847146e-08,
+      "loss": 0.384,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9626865671641791,
+      "grad_norm": 0.8417351207690223,
+      "learning_rate": 1.8239023385365484e-08,
+      "loss": 0.3457,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9629850746268657,
+      "grad_norm": 0.7831713431001753,
+      "learning_rate": 1.7948714607903816e-08,
+      "loss": 0.3093,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9632835820895522,
+      "grad_norm": 0.7856443433041905,
+      "learning_rate": 1.7660726487894188e-08,
+      "loss": 0.314,
+      "step": 3227
+    },
+    {
+      "epoch": 0.9635820895522388,
+      "grad_norm": 0.872583774357775,
+      "learning_rate": 1.7375059294598152e-08,
+      "loss": 0.3313,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9638805970149253,
+      "grad_norm": 0.7486216858015543,
+      "learning_rate": 1.7091713295107337e-08,
+      "loss": 0.2724,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9641791044776119,
+      "grad_norm": 0.8129958391856543,
+      "learning_rate": 1.6810688754343717e-08,
+      "loss": 0.3504,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9644776119402985,
+      "grad_norm": 0.7739654221427428,
+      "learning_rate": 1.6531985935058504e-08,
+      "loss": 0.2831,
+      "step": 3231
+    },
+    {
+      "epoch": 0.964776119402985,
+      "grad_norm": 0.8564741384154801,
+      "learning_rate": 1.6255605097831584e-08,
+      "loss": 0.3567,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9650746268656717,
+      "grad_norm": 0.8973102182259116,
+      "learning_rate": 1.598154650107264e-08,
+      "loss": 0.3538,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9653731343283583,
+      "grad_norm": 0.8793936915110651,
+      "learning_rate": 1.570981040101949e-08,
+      "loss": 0.3381,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9656716417910448,
+      "grad_norm": 0.7753860087917981,
+      "learning_rate": 1.5440397051739163e-08,
+      "loss": 0.2941,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9659701492537314,
+      "grad_norm": 0.7683611135195979,
+      "learning_rate": 1.517330670512629e-08,
+      "loss": 0.3361,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9662686567164179,
+      "grad_norm": 0.8618392461978379,
+      "learning_rate": 1.4908539610903882e-08,
+      "loss": 0.3968,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9665671641791045,
+      "grad_norm": 0.7885885536521883,
+      "learning_rate": 1.4646096016622813e-08,
+      "loss": 0.3317,
+      "step": 3238
+    },
+    {
+      "epoch": 0.966865671641791,
+      "grad_norm": 0.8240073075438972,
+      "learning_rate": 1.4385976167661241e-08,
+      "loss": 0.3054,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9671641791044776,
+      "grad_norm": 0.8544446967641939,
+      "learning_rate": 1.412818030722546e-08,
+      "loss": 0.3274,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9674626865671642,
+      "grad_norm": 0.7981604304770294,
+      "learning_rate": 1.387270867634738e-08,
+      "loss": 0.3452,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9677611940298507,
+      "grad_norm": 0.7709030094314585,
+      "learning_rate": 1.3619561513887603e-08,
+      "loss": 0.3228,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9680597014925373,
+      "grad_norm": 0.9060585553579815,
+      "learning_rate": 1.33687390565318e-08,
+      "loss": 0.3067,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9683582089552238,
+      "grad_norm": 0.819120949141105,
+      "learning_rate": 1.3120241538793487e-08,
+      "loss": 0.3123,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9686567164179104,
+      "grad_norm": 0.7972272943948284,
+      "learning_rate": 1.287406919301154e-08,
+      "loss": 0.3486,
+      "step": 3245
+    },
+    {
+      "epoch": 0.968955223880597,
+      "grad_norm": 0.8855995326432275,
+      "learning_rate": 1.2630222249351287e-08,
+      "loss": 0.3682,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9692537313432836,
+      "grad_norm": 0.7894355860076914,
+      "learning_rate": 1.2388700935803133e-08,
+      "loss": 0.3357,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9695522388059702,
+      "grad_norm": 0.8366342328352925,
+      "learning_rate": 1.214950547818422e-08,
+      "loss": 0.35,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9698507462686567,
+      "grad_norm": 0.8043634556674311,
+      "learning_rate": 1.191263610013621e-08,
+      "loss": 0.3296,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9701492537313433,
+      "grad_norm": 0.8860564254201281,
+      "learning_rate": 1.1678093023126392e-08,
+      "loss": 0.3182,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9704477611940299,
+      "grad_norm": 0.8007581402509059,
+      "learning_rate": 1.144587646644657e-08,
+      "loss": 0.3505,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9707462686567164,
+      "grad_norm": 0.7966262785299418,
+      "learning_rate": 1.121598664721335e-08,
+      "loss": 0.34,
+      "step": 3252
+    },
+    {
+      "epoch": 0.971044776119403,
+      "grad_norm": 0.7323979717681582,
+      "learning_rate": 1.0988423780368685e-08,
+      "loss": 0.2943,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9713432835820895,
+      "grad_norm": 0.8685782481705663,
+      "learning_rate": 1.0763188078678211e-08,
+      "loss": 0.2686,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9716417910447761,
+      "grad_norm": 0.8163158923186468,
+      "learning_rate": 1.0540279752731252e-08,
+      "loss": 0.3165,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9719402985074627,
+      "grad_norm": 0.8202125520717879,
+      "learning_rate": 1.0319699010942207e-08,
+      "loss": 0.2752,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9722388059701492,
+      "grad_norm": 0.7860376878250228,
+      "learning_rate": 1.0101446059548604e-08,
+      "loss": 0.3302,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9725373134328358,
+      "grad_norm": 0.7983641641459549,
+      "learning_rate": 9.88552110261165e-09,
+      "loss": 0.3623,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9728358208955223,
+      "grad_norm": 0.8275216318089765,
+      "learning_rate": 9.671924342015692e-09,
+      "loss": 0.354,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9731343283582089,
+      "grad_norm": 0.8231617072873623,
+      "learning_rate": 9.460655977468757e-09,
+      "loss": 0.2916,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9734328358208956,
+      "grad_norm": 0.8054748071183675,
+      "learning_rate": 9.251716206501449e-09,
+      "loss": 0.3232,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9737313432835821,
+      "grad_norm": 0.9502388363988287,
+      "learning_rate": 9.045105224467221e-09,
+      "loss": 0.374,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9740298507462687,
+      "grad_norm": 0.8180489799155873,
+      "learning_rate": 8.84082322454266e-09,
+      "loss": 0.2988,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9743283582089552,
+      "grad_norm": 0.8508694679094144,
+      "learning_rate": 8.638870397726374e-09,
+      "loss": 0.352,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9746268656716418,
+      "grad_norm": 0.8289100432687119,
+      "learning_rate": 8.439246932839262e-09,
+      "loss": 0.3237,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9749253731343284,
+      "grad_norm": 0.8121348616630165,
+      "learning_rate": 8.241953016524251e-09,
+      "loss": 0.3371,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9752238805970149,
+      "grad_norm": 0.8052003128568289,
+      "learning_rate": 8.04698883324656e-09,
+      "loss": 0.3187,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9755223880597015,
+      "grad_norm": 0.7856612139320543,
+      "learning_rate": 7.854354565292877e-09,
+      "loss": 0.3022,
+      "step": 3268
+    },
+    {
+      "epoch": 0.975820895522388,
+      "grad_norm": 0.8114505322634475,
+      "learning_rate": 7.66405039277135e-09,
+      "loss": 0.3531,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9761194029850746,
+      "grad_norm": 0.9170043571131018,
+      "learning_rate": 7.47607649361215e-09,
+      "loss": 0.3955,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9764179104477612,
+      "grad_norm": 0.8428130140130116,
+      "learning_rate": 7.290433043565803e-09,
+      "loss": 0.3136,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9767164179104477,
+      "grad_norm": 0.843369641143616,
+      "learning_rate": 7.107120216205132e-09,
+      "loss": 0.3433,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9770149253731343,
+      "grad_norm": 0.7547310709601116,
+      "learning_rate": 6.926138182922204e-09,
+      "loss": 0.2943,
+      "step": 3273
+    },
+    {
+      "epoch": 0.977313432835821,
+      "grad_norm": 0.7834878975003178,
+      "learning_rate": 6.747487112931661e-09,
+      "loss": 0.3006,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9776119402985075,
+      "grad_norm": 0.8222964720647756,
+      "learning_rate": 6.57116717326739e-09,
+      "loss": 0.3323,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9779104477611941,
+      "grad_norm": 0.7184401155481505,
+      "learning_rate": 6.397178528784464e-09,
+      "loss": 0.2594,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9782089552238806,
+      "grad_norm": 0.8094077697755245,
+      "learning_rate": 6.225521342158036e-09,
+      "loss": 0.3232,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9785074626865672,
+      "grad_norm": 0.8180072345165923,
+      "learning_rate": 6.056195773883056e-09,
+      "loss": 0.3605,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9788059701492537,
+      "grad_norm": 0.8328869774640257,
+      "learning_rate": 5.889201982275383e-09,
+      "loss": 0.3574,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9791044776119403,
+      "grad_norm": 0.80665947348926,
+      "learning_rate": 5.724540123469569e-09,
+      "loss": 0.3129,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9794029850746269,
+      "grad_norm": 0.803178162202097,
+      "learning_rate": 5.562210351420794e-09,
+      "loss": 0.3278,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9797014925373134,
+      "grad_norm": 0.7597600431927148,
+      "learning_rate": 5.402212817903207e-09,
+      "loss": 0.2825,
+      "step": 3282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7663037088333483,
+      "learning_rate": 5.244547672510758e-09,
+      "loss": 0.2932,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9802985074626865,
+      "grad_norm": 0.9425580206593992,
+      "learning_rate": 5.0892150626566384e-09,
+      "loss": 0.3402,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9805970149253731,
+      "grad_norm": 0.8617742011068167,
+      "learning_rate": 4.93621513357273e-09,
+      "loss": 0.3574,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9808955223880597,
+      "grad_norm": 0.9143536666483425,
+      "learning_rate": 4.785548028310438e-09,
+      "loss": 0.3447,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9811940298507462,
+      "grad_norm": 0.7753440496822905,
+      "learning_rate": 4.637213887739856e-09,
+      "loss": 0.3166,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9814925373134329,
+      "grad_norm": 0.870713877541679,
+      "learning_rate": 4.4912128505497644e-09,
+      "loss": 0.3141,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9817910447761194,
+      "grad_norm": 0.9813206925694251,
+      "learning_rate": 4.347545053247637e-09,
+      "loss": 0.3278,
+      "step": 3289
+    },
+    {
+      "epoch": 0.982089552238806,
+      "grad_norm": 0.8586592952573566,
+      "learning_rate": 4.20621063015908e-09,
+      "loss": 0.3509,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9823880597014926,
+      "grad_norm": 0.7818349485955345,
+      "learning_rate": 4.067209713428388e-09,
+      "loss": 0.3258,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9826865671641791,
+      "grad_norm": 0.7927890018750996,
+      "learning_rate": 3.930542433018547e-09,
+      "loss": 0.3466,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9829850746268657,
+      "grad_norm": 0.7627515359114002,
+      "learning_rate": 3.796208916709565e-09,
+      "loss": 0.3308,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9832835820895522,
+      "grad_norm": 0.8615140995217084,
+      "learning_rate": 3.66420929010014e-09,
+      "loss": 0.3603,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9835820895522388,
+      "grad_norm": 0.8056178237621043,
+      "learning_rate": 3.5345436766065498e-09,
+      "loss": 0.3186,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9838805970149254,
+      "grad_norm": 0.8976274390176755,
+      "learning_rate": 3.407212197463483e-09,
+      "loss": 0.3176,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9841791044776119,
+      "grad_norm": 0.896342713527404,
+      "learning_rate": 3.282214971722375e-09,
+      "loss": 0.345,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9844776119402985,
+      "grad_norm": 0.8134845252531054,
+      "learning_rate": 3.159552116252795e-09,
+      "loss": 0.3436,
+      "step": 3298
+    },
+    {
+      "epoch": 0.984776119402985,
+      "grad_norm": 0.7899298838638901,
+      "learning_rate": 3.0392237457413377e-09,
+      "loss": 0.3398,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 0.8684483689426634,
+      "learning_rate": 2.9212299726921746e-09,
+      "loss": 0.3693,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9853731343283582,
+      "grad_norm": 0.7802224704645674,
+      "learning_rate": 2.80557090742678e-09,
+      "loss": 0.3245,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9856716417910448,
+      "grad_norm": 0.759169607116071,
+      "learning_rate": 2.6922466580830975e-09,
+      "loss": 0.2923,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9859701492537314,
+      "grad_norm": 0.8262473666639696,
+      "learning_rate": 2.5812573306169263e-09,
+      "loss": 0.3262,
+      "step": 3303
+    },
+    {
+      "epoch": 0.986268656716418,
+      "grad_norm": 0.8618008393149142,
+      "learning_rate": 2.4726030288005356e-09,
+      "loss": 0.3574,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9865671641791045,
+      "grad_norm": 0.7676567792662928,
+      "learning_rate": 2.3662838542229392e-09,
+      "loss": 0.2872,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9868656716417911,
+      "grad_norm": 0.7963744228618038,
+      "learning_rate": 2.2622999062899e-09,
+      "loss": 0.3347,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9871641791044776,
+      "grad_norm": 0.862404131304968,
+      "learning_rate": 2.160651282224202e-09,
+      "loss": 0.3633,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9874626865671642,
+      "grad_norm": 0.8788985953060575,
+      "learning_rate": 2.0613380770645452e-09,
+      "loss": 0.3345,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9877611940298507,
+      "grad_norm": 0.8092901595234544,
+      "learning_rate": 1.9643603836666527e-09,
+      "loss": 0.364,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9880597014925373,
+      "grad_norm": 0.8618273355933869,
+      "learning_rate": 1.869718292701883e-09,
+      "loss": 0.3583,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9883582089552239,
+      "grad_norm": 0.7424770027809615,
+      "learning_rate": 1.7774118926586204e-09,
+      "loss": 0.3108,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9886567164179104,
+      "grad_norm": 0.8322502568478948,
+      "learning_rate": 1.6874412698408837e-09,
+      "loss": 0.3482,
+      "step": 3312
+    },
+    {
+      "epoch": 0.988955223880597,
+      "grad_norm": 0.767877374434005,
+      "learning_rate": 1.599806508368884e-09,
+      "loss": 0.2953,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9892537313432835,
+      "grad_norm": 0.7552281111497002,
+      "learning_rate": 1.5145076901795785e-09,
+      "loss": 0.3104,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9895522388059701,
+      "grad_norm": 0.8290814127495036,
+      "learning_rate": 1.431544895024728e-09,
+      "loss": 0.3276,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9898507462686568,
+      "grad_norm": 0.9192486683083075,
+      "learning_rate": 1.3509182004725618e-09,
+      "loss": 0.3404,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9901492537313433,
+      "grad_norm": 0.8117587672326521,
+      "learning_rate": 1.2726276819075012e-09,
+      "loss": 0.3382,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9904477611940299,
+      "grad_norm": 0.8043944237738767,
+      "learning_rate": 1.1966734125287704e-09,
+      "loss": 0.3145,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9907462686567164,
+      "grad_norm": 0.7769831757123559,
+      "learning_rate": 1.1230554633523406e-09,
+      "loss": 0.3126,
+      "step": 3319
+    },
+    {
+      "epoch": 0.991044776119403,
+      "grad_norm": 0.8223550217599328,
+      "learning_rate": 1.0517739032084307e-09,
+      "loss": 0.3554,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9913432835820896,
+      "grad_norm": 0.7735518380562161,
+      "learning_rate": 9.828287987442842e-10,
+      "loss": 0.3166,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9916417910447761,
+      "grad_norm": 0.8254997943791005,
+      "learning_rate": 9.162202144213927e-10,
+      "loss": 0.353,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9919402985074627,
+      "grad_norm": 0.813538581959752,
+      "learning_rate": 8.519482125171618e-10,
+      "loss": 0.3559,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9922388059701492,
+      "grad_norm": 0.7599430996322734,
+      "learning_rate": 7.900128531249107e-10,
+      "loss": 0.3131,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9925373134328358,
+      "grad_norm": 0.8113270433764241,
+      "learning_rate": 7.304141941522069e-10,
+      "loss": 0.3595,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9928358208955224,
+      "grad_norm": 0.9811336819537293,
+      "learning_rate": 6.731522913222544e-10,
+      "loss": 0.3199,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9931343283582089,
+      "grad_norm": 0.8454076686303549,
+      "learning_rate": 6.18227198173893e-10,
+      "loss": 0.3376,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9934328358208955,
+      "grad_norm": 0.8629013881810315,
+      "learning_rate": 5.656389660604888e-10,
+      "loss": 0.3391,
+      "step": 3328
+    },
+    {
+      "epoch": 0.993731343283582,
+      "grad_norm": 0.9074954180694659,
+      "learning_rate": 5.153876441510441e-10,
+      "loss": 0.3412,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9940298507462687,
+      "grad_norm": 0.8314510642754133,
+      "learning_rate": 4.674732794288095e-10,
+      "loss": 0.3593,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9943283582089553,
+      "grad_norm": 0.9116668780870614,
+      "learning_rate": 4.2189591669322684e-10,
+      "loss": 0.3986,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9946268656716418,
+      "grad_norm": 0.7948460052793381,
+      "learning_rate": 3.786555985574314e-10,
+      "loss": 0.2573,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9949253731343284,
+      "grad_norm": 0.7312524937623769,
+      "learning_rate": 3.3775236545019464e-10,
+      "loss": 0.298,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9952238805970149,
+      "grad_norm": 0.7302545942476094,
+      "learning_rate": 2.9918625561536907e-10,
+      "loss": 0.3069,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9955223880597015,
+      "grad_norm": 0.8697911226163374,
+      "learning_rate": 2.6295730511105564e-10,
+      "loss": 0.3235,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9958208955223881,
+      "grad_norm": 0.8879989503033112,
+      "learning_rate": 2.2906554781043645e-10,
+      "loss": 0.3554,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9961194029850746,
+      "grad_norm": 0.7574075071972748,
+      "learning_rate": 1.9751101540149697e-10,
+      "loss": 0.3168,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9964179104477612,
+      "grad_norm": 0.9752022971438031,
+      "learning_rate": 1.6829373738702636e-10,
+      "loss": 0.3387,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9967164179104477,
+      "grad_norm": 0.7698823309368659,
+      "learning_rate": 1.4141374108433968e-10,
+      "loss": 0.2791,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9970149253731343,
+      "grad_norm": 0.8921339982339014,
+      "learning_rate": 1.1687105162583311e-10,
+      "loss": 0.3688,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9973134328358209,
+      "grad_norm": 0.7650117442391594,
+      "learning_rate": 9.466569195787367e-11,
+      "loss": 0.3043,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9976119402985074,
+      "grad_norm": 0.8092924984819057,
+      "learning_rate": 7.479768284246458e-11,
+      "loss": 0.3063,
+      "step": 3342
+    },
+    {
+      "epoch": 0.997910447761194,
+      "grad_norm": 0.7743808795054051,
+      "learning_rate": 5.726704285530238e-11,
+      "loss": 0.3209,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9982089552238806,
+      "grad_norm": 0.8125424691553139,
+      "learning_rate": 4.207378838744225e-11,
+      "loss": 0.3303,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9985074626865672,
+      "grad_norm": 0.814816088361442,
+      "learning_rate": 2.9217933643910236e-11,
+      "loss": 0.33,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9988059701492538,
+      "grad_norm": 0.7687063639245223,
+      "learning_rate": 1.8699490644813467e-11,
+      "loss": 0.2951,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9991044776119403,
+      "grad_norm": 0.8274962606742341,
+      "learning_rate": 1.05184692245075e-11,
+      "loss": 0.2992,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9994029850746269,
+      "grad_norm": 0.7641695113854454,
+      "learning_rate": 4.6748770321514145e-12,
+      "loss": 0.3547,
+      "step": 3348
+    },
+    {
+      "epoch": 0.9997014925373134,
+      "grad_norm": 0.7988942617182022,
+      "learning_rate": 1.1687195311527176e-12,
+      "loss": 0.3496,
+      "step": 3349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8568561243747301,
+      "learning_rate": 0.0,
+      "loss": 0.3795,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0,
+      "step": 3350,
+      "total_flos": 2762833588060160.0,
+      "train_loss": 0.3609271458858874,
+      "train_runtime": 54753.0786,
+      "train_samples_per_second": 5.873,
+      "train_steps_per_second": 0.061
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3350,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2762833588060160.0,
+  "train_batch_size": 24,
+  "trial_name": null,
+  "trial_params": null
+}