diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20750 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2956,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006765899864682003,
+      "grad_norm": 16.730741267512215,
+      "learning_rate": 3.3783783783783786e-08,
+      "loss": 1.4003,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013531799729364006,
+      "grad_norm": 15.571454259355779,
+      "learning_rate": 6.756756756756757e-08,
+      "loss": 1.3802,
+      "step": 2
+    },
+    {
+      "epoch": 0.0020297699594046007,
+      "grad_norm": 15.861673808084099,
+      "learning_rate": 1.0135135135135137e-07,
+      "loss": 1.373,
+      "step": 3
+    },
+    {
+      "epoch": 0.0027063599458728013,
+      "grad_norm": 16.519050340423366,
+      "learning_rate": 1.3513513513513515e-07,
+      "loss": 1.3622,
+      "step": 4
+    },
+    {
+      "epoch": 0.0033829499323410014,
+      "grad_norm": 16.322036539691197,
+      "learning_rate": 1.6891891891891894e-07,
+      "loss": 1.3839,
+      "step": 5
+    },
+    {
+      "epoch": 0.0040595399188092015,
+      "grad_norm": 15.806865323362244,
+      "learning_rate": 2.0270270270270273e-07,
+      "loss": 1.3616,
+      "step": 6
+    },
+    {
+      "epoch": 0.004736129905277402,
+      "grad_norm": 16.30647607824906,
+      "learning_rate": 2.3648648648648652e-07,
+      "loss": 1.3834,
+      "step": 7
+    },
+    {
+      "epoch": 0.005412719891745603,
+      "grad_norm": 16.097760757273917,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 1.3828,
+      "step": 8
+    },
+    {
+      "epoch": 0.006089309878213802,
+      "grad_norm": 15.790648487246665,
+      "learning_rate": 3.040540540540541e-07,
+      "loss": 1.3705,
+      "step": 9
+    },
+    {
+      "epoch": 0.006765899864682003,
+      "grad_norm": 15.618496912944824,
+      "learning_rate": 3.378378378378379e-07,
+      "loss": 1.3526,
+      "step": 10
+    },
+    {
+      "epoch": 0.007442489851150203,
+      "grad_norm": 15.440313477614094,
+      "learning_rate": 3.716216216216217e-07,
+      "loss": 1.3662,
+      "step": 11
+    },
+    {
+      "epoch": 0.008119079837618403,
+      "grad_norm": 15.328802293841417,
+      "learning_rate": 4.0540540540540546e-07,
+      "loss": 1.3611,
+      "step": 12
+    },
+    {
+      "epoch": 0.008795669824086604,
+      "grad_norm": 13.482186578449982,
+      "learning_rate": 4.3918918918918923e-07,
+      "loss": 1.3257,
+      "step": 13
+    },
+    {
+      "epoch": 0.009472259810554804,
+      "grad_norm": 12.78547332777934,
+      "learning_rate": 4.7297297297297305e-07,
+      "loss": 1.2768,
+      "step": 14
+    },
+    {
+      "epoch": 0.010148849797023005,
+      "grad_norm": 11.736419832719868,
+      "learning_rate": 5.067567567567568e-07,
+      "loss": 1.2301,
+      "step": 15
+    },
+    {
+      "epoch": 0.010825439783491205,
+      "grad_norm": 12.36350937570391,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 1.252,
+      "step": 16
+    },
+    {
+      "epoch": 0.011502029769959404,
+      "grad_norm": 8.137523845755103,
+      "learning_rate": 5.743243243243245e-07,
+      "loss": 1.1234,
+      "step": 17
+    },
+    {
+      "epoch": 0.012178619756427604,
+      "grad_norm": 7.925612172145099,
+      "learning_rate": 6.081081081081082e-07,
+      "loss": 1.1318,
+      "step": 18
+    },
+    {
+      "epoch": 0.012855209742895805,
+      "grad_norm": 8.065725866048334,
+      "learning_rate": 6.418918918918919e-07,
+      "loss": 1.1282,
+      "step": 19
+    },
+    {
+      "epoch": 0.013531799729364006,
+      "grad_norm": 7.783987576844463,
+      "learning_rate": 6.756756756756758e-07,
+      "loss": 1.1031,
+      "step": 20
+    },
+    {
+      "epoch": 0.014208389715832206,
+      "grad_norm": 7.537433144057233,
+      "learning_rate": 7.094594594594595e-07,
+      "loss": 1.0974,
+      "step": 21
+    },
+    {
+      "epoch": 0.014884979702300407,
+      "grad_norm": 7.33918326680637,
+      "learning_rate": 7.432432432432434e-07,
+      "loss": 1.0281,
+      "step": 22
+    },
+    {
+      "epoch": 0.015561569688768605,
+      "grad_norm": 6.847447951454805,
+      "learning_rate": 7.770270270270271e-07,
+      "loss": 0.8803,
+      "step": 23
+    },
+    {
+      "epoch": 0.016238159675236806,
+      "grad_norm": 7.107323623205301,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.8526,
+      "step": 24
+    },
+    {
+      "epoch": 0.016914749661705007,
+      "grad_norm": 6.984389688094926,
+      "learning_rate": 8.445945945945947e-07,
+      "loss": 0.8347,
+      "step": 25
+    },
+    {
+      "epoch": 0.017591339648173207,
+      "grad_norm": 6.957244760107825,
+      "learning_rate": 8.783783783783785e-07,
+      "loss": 0.8158,
+      "step": 26
+    },
+    {
+      "epoch": 0.018267929634641408,
+      "grad_norm": 5.951868854528439,
+      "learning_rate": 9.121621621621622e-07,
+      "loss": 0.7551,
+      "step": 27
+    },
+    {
+      "epoch": 0.018944519621109608,
+      "grad_norm": 5.576563392855383,
+      "learning_rate": 9.459459459459461e-07,
+      "loss": 0.7502,
+      "step": 28
+    },
+    {
+      "epoch": 0.01962110960757781,
+      "grad_norm": 5.333821136741542,
+      "learning_rate": 9.797297297297298e-07,
+      "loss": 0.7318,
+      "step": 29
+    },
+    {
+      "epoch": 0.02029769959404601,
+      "grad_norm": 4.71953267641669,
+      "learning_rate": 1.0135135135135136e-06,
+      "loss": 0.7009,
+      "step": 30
+    },
+    {
+      "epoch": 0.02097428958051421,
+      "grad_norm": 6.210600970087328,
+      "learning_rate": 1.0472972972972973e-06,
+      "loss": 0.6156,
+      "step": 31
+    },
+    {
+      "epoch": 0.02165087956698241,
+      "grad_norm": 2.882402036869178,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.5813,
+      "step": 32
+    },
+    {
+      "epoch": 0.022327469553450607,
+      "grad_norm": 2.7328332516592364,
+      "learning_rate": 1.114864864864865e-06,
+      "loss": 0.5673,
+      "step": 33
+    },
+    {
+      "epoch": 0.023004059539918808,
+      "grad_norm": 2.290783908283054,
+      "learning_rate": 1.148648648648649e-06,
+      "loss": 0.5578,
+      "step": 34
+    },
+    {
+      "epoch": 0.02368064952638701,
+      "grad_norm": 1.8020363472220706,
+      "learning_rate": 1.1824324324324326e-06,
+      "loss": 0.5145,
+      "step": 35
+    },
+    {
+      "epoch": 0.02435723951285521,
+      "grad_norm": 1.6372723330361931,
+      "learning_rate": 1.2162162162162164e-06,
+      "loss": 0.5459,
+      "step": 36
+    },
+    {
+      "epoch": 0.02503382949932341,
+      "grad_norm": 1.634384920098834,
+      "learning_rate": 1.25e-06,
+      "loss": 0.5434,
+      "step": 37
+    },
+    {
+      "epoch": 0.02571041948579161,
+      "grad_norm": 1.4803481693799974,
+      "learning_rate": 1.2837837837837838e-06,
+      "loss": 0.5155,
+      "step": 38
+    },
+    {
+      "epoch": 0.02638700947225981,
+      "grad_norm": 1.410478596337927,
+      "learning_rate": 1.3175675675675676e-06,
+      "loss": 0.5011,
+      "step": 39
+    },
+    {
+      "epoch": 0.02706359945872801,
+      "grad_norm": 1.3427462135117854,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.4879,
+      "step": 40
+    },
+    {
+      "epoch": 0.02774018944519621,
+      "grad_norm": 1.2931079526982363,
+      "learning_rate": 1.3851351351351352e-06,
+      "loss": 0.4814,
+      "step": 41
+    },
+    {
+      "epoch": 0.028416779431664412,
+      "grad_norm": 1.2980357899645447,
+      "learning_rate": 1.418918918918919e-06,
+      "loss": 0.4813,
+      "step": 42
+    },
+    {
+      "epoch": 0.029093369418132613,
+      "grad_norm": 1.3249799973103618,
+      "learning_rate": 1.4527027027027027e-06,
+      "loss": 0.4524,
+      "step": 43
+    },
+    {
+      "epoch": 0.029769959404600813,
+      "grad_norm": 1.4109631750201308,
+      "learning_rate": 1.4864864864864868e-06,
+      "loss": 0.4715,
+      "step": 44
+    },
+    {
+      "epoch": 0.030446549391069014,
+      "grad_norm": 1.3124047019000802,
+      "learning_rate": 1.5202702702702704e-06,
+      "loss": 0.4718,
+      "step": 45
+    },
+    {
+      "epoch": 0.03112313937753721,
+      "grad_norm": 1.3577219505300853,
+      "learning_rate": 1.5540540540540541e-06,
+      "loss": 0.4666,
+      "step": 46
+    },
+    {
+      "epoch": 0.031799729364005415,
+      "grad_norm": 1.206525378928271,
+      "learning_rate": 1.5878378378378378e-06,
+      "loss": 0.4581,
+      "step": 47
+    },
+    {
+      "epoch": 0.03247631935047361,
+      "grad_norm": 1.1358040114562087,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.4633,
+      "step": 48
+    },
+    {
+      "epoch": 0.033152909336941816,
+      "grad_norm": 1.1352051258810618,
+      "learning_rate": 1.6554054054054055e-06,
+      "loss": 0.4453,
+      "step": 49
+    },
+    {
+      "epoch": 0.03382949932341001,
+      "grad_norm": 1.1547392514255554,
+      "learning_rate": 1.6891891891891894e-06,
+      "loss": 0.4449,
+      "step": 50
+    },
+    {
+      "epoch": 0.03450608930987822,
+      "grad_norm": 1.138578275755696,
+      "learning_rate": 1.722972972972973e-06,
+      "loss": 0.4289,
+      "step": 51
+    },
+    {
+      "epoch": 0.035182679296346414,
+      "grad_norm": 1.1111420898608284,
+      "learning_rate": 1.756756756756757e-06,
+      "loss": 0.4399,
+      "step": 52
+    },
+    {
+      "epoch": 0.03585926928281461,
+      "grad_norm": 1.055663222633348,
+      "learning_rate": 1.7905405405405408e-06,
+      "loss": 0.4384,
+      "step": 53
+    },
+    {
+      "epoch": 0.036535859269282815,
+      "grad_norm": 1.055347179354741,
+      "learning_rate": 1.8243243243243245e-06,
+      "loss": 0.4332,
+      "step": 54
+    },
+    {
+      "epoch": 0.03721244925575101,
+      "grad_norm": 1.052291843131986,
+      "learning_rate": 1.8581081081081081e-06,
+      "loss": 0.4302,
+      "step": 55
+    },
+    {
+      "epoch": 0.037889039242219216,
+      "grad_norm": 1.015375417637575,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.4375,
+      "step": 56
+    },
+    {
+      "epoch": 0.03856562922868741,
+      "grad_norm": 1.0371652541062233,
+      "learning_rate": 1.925675675675676e-06,
+      "loss": 0.4035,
+      "step": 57
+    },
+    {
+      "epoch": 0.03924221921515562,
+      "grad_norm": 1.0641509539695226,
+      "learning_rate": 1.9594594594594595e-06,
+      "loss": 0.4223,
+      "step": 58
+    },
+    {
+      "epoch": 0.039918809201623814,
+      "grad_norm": 0.9914265277383638,
+      "learning_rate": 1.9932432432432434e-06,
+      "loss": 0.399,
+      "step": 59
+    },
+    {
+      "epoch": 0.04059539918809202,
+      "grad_norm": 1.0263164791533792,
+      "learning_rate": 2.0270270270270273e-06,
+      "loss": 0.4187,
+      "step": 60
+    },
+    {
+      "epoch": 0.041271989174560215,
+      "grad_norm": 0.9947505937515272,
+      "learning_rate": 2.060810810810811e-06,
+      "loss": 0.4158,
+      "step": 61
+    },
+    {
+      "epoch": 0.04194857916102842,
+      "grad_norm": 0.9909395863726204,
+      "learning_rate": 2.0945945945945946e-06,
+      "loss": 0.4304,
+      "step": 62
+    },
+    {
+      "epoch": 0.04262516914749662,
+      "grad_norm": 0.9292999289285129,
+      "learning_rate": 2.1283783783783785e-06,
+      "loss": 0.4169,
+      "step": 63
+    },
+    {
+      "epoch": 0.04330175913396482,
+      "grad_norm": 1.0314768531808776,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.4216,
+      "step": 64
+    },
+    {
+      "epoch": 0.04397834912043302,
+      "grad_norm": 1.0309902965388054,
+      "learning_rate": 2.195945945945946e-06,
+      "loss": 0.4218,
+      "step": 65
+    },
+    {
+      "epoch": 0.044654939106901215,
+      "grad_norm": 0.9852987927365103,
+      "learning_rate": 2.22972972972973e-06,
+      "loss": 0.4033,
+      "step": 66
+    },
+    {
+      "epoch": 0.04533152909336942,
+      "grad_norm": 0.9434944427346157,
+      "learning_rate": 2.2635135135135135e-06,
+      "loss": 0.4116,
+      "step": 67
+    },
+    {
+      "epoch": 0.046008119079837616,
+      "grad_norm": 0.9982076756005573,
+      "learning_rate": 2.297297297297298e-06,
+      "loss": 0.4097,
+      "step": 68
+    },
+    {
+      "epoch": 0.04668470906630582,
+      "grad_norm": 1.0226952189078897,
+      "learning_rate": 2.3310810810810813e-06,
+      "loss": 0.4002,
+      "step": 69
+    },
+    {
+      "epoch": 0.04736129905277402,
+      "grad_norm": 0.9773100130798488,
+      "learning_rate": 2.364864864864865e-06,
+      "loss": 0.3915,
+      "step": 70
+    },
+    {
+      "epoch": 0.04803788903924222,
+      "grad_norm": 0.9762812638918047,
+      "learning_rate": 2.3986486486486486e-06,
+      "loss": 0.3919,
+      "step": 71
+    },
+    {
+      "epoch": 0.04871447902571042,
+      "grad_norm": 1.0020207991206935,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.3921,
+      "step": 72
+    },
+    {
+      "epoch": 0.04939106901217862,
+      "grad_norm": 1.0120884462936703,
+      "learning_rate": 2.4662162162162163e-06,
+      "loss": 0.3776,
+      "step": 73
+    },
+    {
+      "epoch": 0.05006765899864682,
+      "grad_norm": 1.0516468695441452,
+      "learning_rate": 2.5e-06,
+      "loss": 0.3859,
+      "step": 74
+    },
+    {
+      "epoch": 0.05074424898511502,
+      "grad_norm": 0.9780870933091569,
+      "learning_rate": 2.533783783783784e-06,
+      "loss": 0.3894,
+      "step": 75
+    },
+    {
+      "epoch": 0.05142083897158322,
+      "grad_norm": 1.0311856688315713,
+      "learning_rate": 2.5675675675675675e-06,
+      "loss": 0.3883,
+      "step": 76
+    },
+    {
+      "epoch": 0.052097428958051424,
+      "grad_norm": 0.9496324692105611,
+      "learning_rate": 2.601351351351352e-06,
+      "loss": 0.386,
+      "step": 77
+    },
+    {
+      "epoch": 0.05277401894451962,
+      "grad_norm": 0.9697738269386031,
+      "learning_rate": 2.6351351351351353e-06,
+      "loss": 0.3793,
+      "step": 78
+    },
+    {
+      "epoch": 0.05345060893098782,
+      "grad_norm": 0.9965712520544989,
+      "learning_rate": 2.668918918918919e-06,
+      "loss": 0.388,
+      "step": 79
+    },
+    {
+      "epoch": 0.05412719891745602,
+      "grad_norm": 0.9418392636899636,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.3792,
+      "step": 80
+    },
+    {
+      "epoch": 0.05480378890392422,
+      "grad_norm": 0.9231211200890874,
+      "learning_rate": 2.7364864864864865e-06,
+      "loss": 0.3686,
+      "step": 81
+    },
+    {
+      "epoch": 0.05548037889039242,
+      "grad_norm": 1.062806789745046,
+      "learning_rate": 2.7702702702702703e-06,
+      "loss": 0.3827,
+      "step": 82
+    },
+    {
+      "epoch": 0.05615696887686062,
+      "grad_norm": 0.9986462714361596,
+      "learning_rate": 2.8040540540540546e-06,
+      "loss": 0.3888,
+      "step": 83
+    },
+    {
+      "epoch": 0.056833558863328824,
+      "grad_norm": 1.056713018386747,
+      "learning_rate": 2.837837837837838e-06,
+      "loss": 0.382,
+      "step": 84
+    },
+    {
+      "epoch": 0.05751014884979702,
+      "grad_norm": 1.0007102796338614,
+      "learning_rate": 2.871621621621622e-06,
+      "loss": 0.3968,
+      "step": 85
+    },
+    {
+      "epoch": 0.058186738836265225,
+      "grad_norm": 1.03144449589489,
+      "learning_rate": 2.9054054054054054e-06,
+      "loss": 0.3669,
+      "step": 86
+    },
+    {
+      "epoch": 0.05886332882273342,
+      "grad_norm": 1.041159509051703,
+      "learning_rate": 2.9391891891891893e-06,
+      "loss": 0.3747,
+      "step": 87
+    },
+    {
+      "epoch": 0.05953991880920163,
+      "grad_norm": 1.0378568292691996,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.3717,
+      "step": 88
+    },
+    {
+      "epoch": 0.060216508795669824,
+      "grad_norm": 1.1645083070325495,
+      "learning_rate": 3.006756756756757e-06,
+      "loss": 0.3864,
+      "step": 89
+    },
+    {
+      "epoch": 0.06089309878213803,
+      "grad_norm": 1.0664838033380974,
+      "learning_rate": 3.040540540540541e-06,
+      "loss": 0.3837,
+      "step": 90
+    },
+    {
+      "epoch": 0.061569688768606225,
+      "grad_norm": 0.9795777593045356,
+      "learning_rate": 3.0743243243243248e-06,
+      "loss": 0.3733,
+      "step": 91
+    },
+    {
+      "epoch": 0.06224627875507442,
+      "grad_norm": 1.1173424271948955,
+      "learning_rate": 3.1081081081081082e-06,
+      "loss": 0.3748,
+      "step": 92
+    },
+    {
+      "epoch": 0.06292286874154263,
+      "grad_norm": 1.114397692571731,
+      "learning_rate": 3.141891891891892e-06,
+      "loss": 0.3563,
+      "step": 93
+    },
+    {
+      "epoch": 0.06359945872801083,
+      "grad_norm": 0.9878127612356723,
+      "learning_rate": 3.1756756756756755e-06,
+      "loss": 0.3696,
+      "step": 94
+    },
+    {
+      "epoch": 0.06427604871447902,
+      "grad_norm": 0.9738688151710514,
+      "learning_rate": 3.20945945945946e-06,
+      "loss": 0.3724,
+      "step": 95
+    },
+    {
+      "epoch": 0.06495263870094722,
+      "grad_norm": 1.0801619847575732,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.3704,
+      "step": 96
+    },
+    {
+      "epoch": 0.06562922868741543,
+      "grad_norm": 1.1355268850212117,
+      "learning_rate": 3.277027027027027e-06,
+      "loss": 0.3719,
+      "step": 97
+    },
+    {
+      "epoch": 0.06630581867388363,
+      "grad_norm": 0.9166917516935064,
+      "learning_rate": 3.310810810810811e-06,
+      "loss": 0.3656,
+      "step": 98
+    },
+    {
+      "epoch": 0.06698240866035182,
+      "grad_norm": 1.0313368398797989,
+      "learning_rate": 3.3445945945945953e-06,
+      "loss": 0.3621,
+      "step": 99
+    },
+    {
+      "epoch": 0.06765899864682003,
+      "grad_norm": 0.9836880800425399,
+      "learning_rate": 3.3783783783783788e-06,
+      "loss": 0.3537,
+      "step": 100
+    },
+    {
+      "epoch": 0.06833558863328823,
+      "grad_norm": 0.9263217520072473,
+      "learning_rate": 3.4121621621621626e-06,
+      "loss": 0.3641,
+      "step": 101
+    },
+    {
+      "epoch": 0.06901217861975643,
+      "grad_norm": 1.0112742073005092,
+      "learning_rate": 3.445945945945946e-06,
+      "loss": 0.36,
+      "step": 102
+    },
+    {
+      "epoch": 0.06968876860622462,
+      "grad_norm": 0.9916126686386084,
+      "learning_rate": 3.47972972972973e-06,
+      "loss": 0.3814,
+      "step": 103
+    },
+    {
+      "epoch": 0.07036535859269283,
+      "grad_norm": 0.9847415548825377,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.3698,
+      "step": 104
+    },
+    {
+      "epoch": 0.07104194857916103,
+      "grad_norm": 0.9588024088872578,
+      "learning_rate": 3.5472972972972973e-06,
+      "loss": 0.3625,
+      "step": 105
+    },
+    {
+      "epoch": 0.07171853856562922,
+      "grad_norm": 0.9831235499155904,
+      "learning_rate": 3.5810810810810816e-06,
+      "loss": 0.3705,
+      "step": 106
+    },
+    {
+      "epoch": 0.07239512855209743,
+      "grad_norm": 0.9861768633987433,
+      "learning_rate": 3.6148648648648655e-06,
+      "loss": 0.3647,
+      "step": 107
+    },
+    {
+      "epoch": 0.07307171853856563,
+      "grad_norm": 1.0606338063172631,
+      "learning_rate": 3.648648648648649e-06,
+      "loss": 0.3703,
+      "step": 108
+    },
+    {
+      "epoch": 0.07374830852503383,
+      "grad_norm": 0.9881508203610974,
+      "learning_rate": 3.6824324324324328e-06,
+      "loss": 0.3729,
+      "step": 109
+    },
+    {
+      "epoch": 0.07442489851150202,
+      "grad_norm": 1.0001913986166444,
+      "learning_rate": 3.7162162162162162e-06,
+      "loss": 0.3703,
+      "step": 110
+    },
+    {
+      "epoch": 0.07510148849797023,
+      "grad_norm": 1.0114756809458896,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3723,
+      "step": 111
+    },
+    {
+      "epoch": 0.07577807848443843,
+      "grad_norm": 0.9512222873658841,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.3582,
+      "step": 112
+    },
+    {
+      "epoch": 0.07645466847090664,
+      "grad_norm": 0.9453179055223212,
+      "learning_rate": 3.817567567567567e-06,
+      "loss": 0.3565,
+      "step": 113
+    },
+    {
+      "epoch": 0.07713125845737483,
+      "grad_norm": 0.9524715891096117,
+      "learning_rate": 3.851351351351352e-06,
+      "loss": 0.3675,
+      "step": 114
+    },
+    {
+      "epoch": 0.07780784844384303,
+      "grad_norm": 0.9999912241113618,
+      "learning_rate": 3.885135135135135e-06,
+      "loss": 0.3639,
+      "step": 115
+    },
+    {
+      "epoch": 0.07848443843031123,
+      "grad_norm": 1.055972560726516,
+      "learning_rate": 3.918918918918919e-06,
+      "loss": 0.3539,
+      "step": 116
+    },
+    {
+      "epoch": 0.07916102841677942,
+      "grad_norm": 0.9391582163968005,
+      "learning_rate": 3.952702702702703e-06,
+      "loss": 0.3508,
+      "step": 117
+    },
+    {
+      "epoch": 0.07983761840324763,
+      "grad_norm": 0.94833910465346,
+      "learning_rate": 3.986486486486487e-06,
+      "loss": 0.3632,
+      "step": 118
+    },
+    {
+      "epoch": 0.08051420838971583,
+      "grad_norm": 1.070789034259915,
+      "learning_rate": 4.020270270270271e-06,
+      "loss": 0.3639,
+      "step": 119
+    },
+    {
+      "epoch": 0.08119079837618404,
+      "grad_norm": 1.0133398094963086,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.3645,
+      "step": 120
+    },
+    {
+      "epoch": 0.08186738836265223,
+      "grad_norm": 0.8886040431328776,
+      "learning_rate": 4.087837837837838e-06,
+      "loss": 0.3615,
+      "step": 121
+    },
+    {
+      "epoch": 0.08254397834912043,
+      "grad_norm": 0.9757625925379442,
+      "learning_rate": 4.121621621621622e-06,
+      "loss": 0.3566,
+      "step": 122
+    },
+    {
+      "epoch": 0.08322056833558863,
+      "grad_norm": 1.0128075128638858,
+      "learning_rate": 4.155405405405405e-06,
+      "loss": 0.3583,
+      "step": 123
+    },
+    {
+      "epoch": 0.08389715832205684,
+      "grad_norm": 0.9049852654356164,
+      "learning_rate": 4.189189189189189e-06,
+      "loss": 0.3676,
+      "step": 124
+    },
+    {
+      "epoch": 0.08457374830852503,
+      "grad_norm": 1.0131521395391876,
+      "learning_rate": 4.222972972972974e-06,
+      "loss": 0.3577,
+      "step": 125
+    },
+    {
+      "epoch": 0.08525033829499323,
+      "grad_norm": 1.0789716766915414,
+      "learning_rate": 4.256756756756757e-06,
+      "loss": 0.3532,
+      "step": 126
+    },
+    {
+      "epoch": 0.08592692828146144,
+      "grad_norm": 0.9783591849570962,
+      "learning_rate": 4.290540540540541e-06,
+      "loss": 0.3536,
+      "step": 127
+    },
+    {
+      "epoch": 0.08660351826792964,
+      "grad_norm": 0.9421245409069976,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.3403,
+      "step": 128
+    },
+    {
+      "epoch": 0.08728010825439783,
+      "grad_norm": 0.9271592880849898,
+      "learning_rate": 4.3581081081081085e-06,
+      "loss": 0.3541,
+      "step": 129
+    },
+    {
+      "epoch": 0.08795669824086604,
+      "grad_norm": 1.059729011063821,
+      "learning_rate": 4.391891891891892e-06,
+      "loss": 0.3577,
+      "step": 130
+    },
+    {
+      "epoch": 0.08863328822733424,
+      "grad_norm": 1.039889679856772,
+      "learning_rate": 4.4256756756756754e-06,
+      "loss": 0.3637,
+      "step": 131
+    },
+    {
+      "epoch": 0.08930987821380243,
+      "grad_norm": 0.9583751612541778,
+      "learning_rate": 4.45945945945946e-06,
+      "loss": 0.3545,
+      "step": 132
+    },
+    {
+      "epoch": 0.08998646820027063,
+      "grad_norm": 0.9019133632068663,
+      "learning_rate": 4.493243243243244e-06,
+      "loss": 0.3696,
+      "step": 133
+    },
+    {
+      "epoch": 0.09066305818673884,
+      "grad_norm": 1.0882899845215799,
+      "learning_rate": 4.527027027027027e-06,
+      "loss": 0.3542,
+      "step": 134
+    },
+    {
+      "epoch": 0.09133964817320704,
+      "grad_norm": 0.972920748411265,
+      "learning_rate": 4.560810810810811e-06,
+      "loss": 0.3563,
+      "step": 135
+    },
+    {
+      "epoch": 0.09201623815967523,
+      "grad_norm": 0.9607177937348906,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.3693,
+      "step": 136
+    },
+    {
+      "epoch": 0.09269282814614344,
+      "grad_norm": 0.9425453677839248,
+      "learning_rate": 4.628378378378379e-06,
+      "loss": 0.3381,
+      "step": 137
+    },
+    {
+      "epoch": 0.09336941813261164,
+      "grad_norm": 0.954177334088261,
+      "learning_rate": 4.6621621621621625e-06,
+      "loss": 0.3626,
+      "step": 138
+    },
+    {
+      "epoch": 0.09404600811907984,
+      "grad_norm": 0.994579898932673,
+      "learning_rate": 4.695945945945946e-06,
+      "loss": 0.356,
+      "step": 139
+    },
+    {
+      "epoch": 0.09472259810554803,
+      "grad_norm": 0.9843907088893953,
+      "learning_rate": 4.72972972972973e-06,
+      "loss": 0.3525,
+      "step": 140
+    },
+    {
+      "epoch": 0.09539918809201624,
+      "grad_norm": 0.8681485240229405,
+      "learning_rate": 4.763513513513514e-06,
+      "loss": 0.351,
+      "step": 141
+    },
+    {
+      "epoch": 0.09607577807848444,
+      "grad_norm": 0.9595662426177981,
+      "learning_rate": 4.797297297297297e-06,
+      "loss": 0.3419,
+      "step": 142
+    },
+    {
+      "epoch": 0.09675236806495263,
+      "grad_norm": 0.9239545208320784,
+      "learning_rate": 4.831081081081082e-06,
+      "loss": 0.3536,
+      "step": 143
+    },
+    {
+      "epoch": 0.09742895805142084,
+      "grad_norm": 0.89666580058855,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.3526,
+      "step": 144
+    },
+    {
+      "epoch": 0.09810554803788904,
+      "grad_norm": 0.9673666901174596,
+      "learning_rate": 4.898648648648649e-06,
+      "loss": 0.3506,
+      "step": 145
+    },
+    {
+      "epoch": 0.09878213802435724,
+      "grad_norm": 0.8817116329539537,
+      "learning_rate": 4.932432432432433e-06,
+      "loss": 0.3428,
+      "step": 146
+    },
+    {
+      "epoch": 0.09945872801082543,
+      "grad_norm": 0.9867492064635516,
+      "learning_rate": 4.9662162162162165e-06,
+      "loss": 0.3484,
+      "step": 147
+    },
+    {
+      "epoch": 0.10013531799729364,
+      "grad_norm": 0.9043665990251071,
+      "learning_rate": 5e-06,
+      "loss": 0.3492,
+      "step": 148
+    },
+    {
+      "epoch": 0.10081190798376184,
+      "grad_norm": 0.9011500354117904,
+      "learning_rate": 5.033783783783784e-06,
+      "loss": 0.3436,
+      "step": 149
+    },
+    {
+      "epoch": 0.10148849797023005,
+      "grad_norm": 0.953843668390464,
+      "learning_rate": 5.067567567567568e-06,
+      "loss": 0.3529,
+      "step": 150
+    },
+    {
+      "epoch": 0.10216508795669824,
+      "grad_norm": 0.9215559520914532,
+      "learning_rate": 5.101351351351351e-06,
+      "loss": 0.3538,
+      "step": 151
+    },
+    {
+      "epoch": 0.10284167794316644,
+      "grad_norm": 0.9357092491232858,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.3584,
+      "step": 152
+    },
+    {
+      "epoch": 0.10351826792963464,
+      "grad_norm": 0.9365767128996711,
+      "learning_rate": 5.168918918918919e-06,
+      "loss": 0.3546,
+      "step": 153
+    },
+    {
+      "epoch": 0.10419485791610285,
+      "grad_norm": 0.9624474687891851,
+      "learning_rate": 5.202702702702704e-06,
+      "loss": 0.3603,
+      "step": 154
+    },
+    {
+      "epoch": 0.10487144790257104,
+      "grad_norm": 0.9441923631377324,
+      "learning_rate": 5.2364864864864875e-06,
+      "loss": 0.3621,
+      "step": 155
+    },
+    {
+      "epoch": 0.10554803788903924,
+      "grad_norm": 0.9768875901794846,
+      "learning_rate": 5.2702702702702705e-06,
+      "loss": 0.3336,
+      "step": 156
+    },
+    {
+      "epoch": 0.10622462787550745,
+      "grad_norm": 0.9672310364129455,
+      "learning_rate": 5.304054054054054e-06,
+      "loss": 0.3577,
+      "step": 157
+    },
+    {
+      "epoch": 0.10690121786197564,
+      "grad_norm": 0.8938714866222908,
+      "learning_rate": 5.337837837837838e-06,
+      "loss": 0.3517,
+      "step": 158
+    },
+    {
+      "epoch": 0.10757780784844384,
+      "grad_norm": 0.9461498992246354,
+      "learning_rate": 5.371621621621622e-06,
+      "loss": 0.3733,
+      "step": 159
+    },
+    {
+      "epoch": 0.10825439783491204,
+      "grad_norm": 0.9420642339852112,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.3404,
+      "step": 160
+    },
+    {
+      "epoch": 0.10893098782138025,
+      "grad_norm": 0.8571660933345047,
+      "learning_rate": 5.43918918918919e-06,
+      "loss": 0.3301,
+      "step": 161
+    },
+    {
+      "epoch": 0.10960757780784844,
+      "grad_norm": 0.8739947191097851,
+      "learning_rate": 5.472972972972973e-06,
+      "loss": 0.3509,
+      "step": 162
+    },
+    {
+      "epoch": 0.11028416779431664,
+      "grad_norm": 1.0068377235851085,
+      "learning_rate": 5.506756756756757e-06,
+      "loss": 0.3455,
+      "step": 163
+    },
+    {
+      "epoch": 0.11096075778078485,
+      "grad_norm": 0.8502231162724772,
+      "learning_rate": 5.540540540540541e-06,
+      "loss": 0.3459,
+      "step": 164
+    },
+    {
+      "epoch": 0.11163734776725305,
+      "grad_norm": 0.9982633482472686,
+      "learning_rate": 5.574324324324325e-06,
+      "loss": 0.3499,
+      "step": 165
+    },
+    {
+      "epoch": 0.11231393775372124,
+      "grad_norm": 0.9658330618822643,
+      "learning_rate": 5.608108108108109e-06,
+      "loss": 0.3388,
+      "step": 166
+    },
+    {
+      "epoch": 0.11299052774018944,
+      "grad_norm": 0.9006851866188571,
+      "learning_rate": 5.641891891891892e-06,
+      "loss": 0.3483,
+      "step": 167
+    },
+    {
+      "epoch": 0.11366711772665765,
+      "grad_norm": 0.9352204994688179,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.3449,
+      "step": 168
+    },
+    {
+      "epoch": 0.11434370771312584,
+      "grad_norm": 0.8870096765403227,
+      "learning_rate": 5.70945945945946e-06,
+      "loss": 0.3522,
+      "step": 169
+    },
+    {
+      "epoch": 0.11502029769959404,
+      "grad_norm": 0.8969307960911421,
+      "learning_rate": 5.743243243243244e-06,
+      "loss": 0.3432,
+      "step": 170
+    },
+    {
+      "epoch": 0.11569688768606225,
+      "grad_norm": 0.908483782850771,
+      "learning_rate": 5.777027027027028e-06,
+      "loss": 0.3419,
+      "step": 171
+    },
+    {
+      "epoch": 0.11637347767253045,
+      "grad_norm": 0.8764836095650671,
+      "learning_rate": 5.810810810810811e-06,
+      "loss": 0.3549,
+      "step": 172
+    },
+    {
+      "epoch": 0.11705006765899864,
+      "grad_norm": 0.8693655023211483,
+      "learning_rate": 5.844594594594595e-06,
+      "loss": 0.3436,
+      "step": 173
+    },
+    {
+      "epoch": 0.11772665764546685,
+      "grad_norm": 0.8629649880466117,
+      "learning_rate": 5.8783783783783786e-06,
+      "loss": 0.3377,
+      "step": 174
+    },
+    {
+      "epoch": 0.11840324763193505,
+      "grad_norm": 0.9655818319073995,
+      "learning_rate": 5.912162162162162e-06,
+      "loss": 0.3473,
+      "step": 175
+    },
+    {
+      "epoch": 0.11907983761840325,
+      "grad_norm": 0.8388724068663871,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.3387,
+      "step": 176
+    },
+    {
+      "epoch": 0.11975642760487144,
+      "grad_norm": 0.8921658350034631,
+      "learning_rate": 5.979729729729731e-06,
+      "loss": 0.338,
+      "step": 177
+    },
+    {
+      "epoch": 0.12043301759133965,
+      "grad_norm": 0.8501350702200059,
+      "learning_rate": 6.013513513513514e-06,
+      "loss": 0.3365,
+      "step": 178
+    },
+    {
+      "epoch": 0.12110960757780785,
+      "grad_norm": 0.8460122502599188,
+      "learning_rate": 6.047297297297298e-06,
+      "loss": 0.3402,
+      "step": 179
+    },
+    {
+      "epoch": 0.12178619756427606,
+      "grad_norm": 0.8973361672929452,
+      "learning_rate": 6.081081081081082e-06,
+      "loss": 0.3442,
+      "step": 180
+    },
+    {
+      "epoch": 0.12246278755074425,
+      "grad_norm": 0.9069042614342414,
+      "learning_rate": 6.114864864864866e-06,
+      "loss": 0.3362,
+      "step": 181
+    },
+    {
+      "epoch": 0.12313937753721245,
+      "grad_norm": 0.8860906783387148,
+      "learning_rate": 6.1486486486486495e-06,
+      "loss": 0.3299,
+      "step": 182
+    },
+    {
+      "epoch": 0.12381596752368065,
+      "grad_norm": 0.8551730686715848,
+      "learning_rate": 6.1824324324324326e-06,
+      "loss": 0.348,
+      "step": 183
+    },
+    {
+      "epoch": 0.12449255751014884,
+      "grad_norm": 0.8766985699493755,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.3375,
+      "step": 184
+    },
+    {
+      "epoch": 0.12516914749661706,
+      "grad_norm": 0.9225543493262915,
+      "learning_rate": 6.25e-06,
+      "loss": 0.3424,
+      "step": 185
+    },
+    {
+      "epoch": 0.12584573748308525,
+      "grad_norm": 0.8471558820346015,
+      "learning_rate": 6.283783783783784e-06,
+      "loss": 0.3279,
+      "step": 186
+    },
+    {
+      "epoch": 0.12652232746955344,
+      "grad_norm": 0.8503227272610423,
+      "learning_rate": 6.317567567567569e-06,
+      "loss": 0.326,
+      "step": 187
+    },
+    {
+      "epoch": 0.12719891745602166,
+      "grad_norm": 0.9135946214244975,
+      "learning_rate": 6.351351351351351e-06,
+      "loss": 0.3441,
+      "step": 188
+    },
+    {
+      "epoch": 0.12787550744248985,
+      "grad_norm": 0.8912738787342082,
+      "learning_rate": 6.385135135135136e-06,
+      "loss": 0.3515,
+      "step": 189
+    },
+    {
+      "epoch": 0.12855209742895804,
+      "grad_norm": 0.8712120577876538,
+      "learning_rate": 6.41891891891892e-06,
+      "loss": 0.3346,
+      "step": 190
+    },
+    {
+      "epoch": 0.12922868741542626,
+      "grad_norm": 0.9287050994691649,
+      "learning_rate": 6.4527027027027035e-06,
+      "loss": 0.3459,
+      "step": 191
+    },
+    {
+      "epoch": 0.12990527740189445,
+      "grad_norm": 0.8581490251456293,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.3486,
+      "step": 192
+    },
+    {
+      "epoch": 0.13058186738836267,
+      "grad_norm": 0.8942155499722226,
+      "learning_rate": 6.5202702702702704e-06,
+      "loss": 0.3413,
+      "step": 193
+    },
+    {
+      "epoch": 0.13125845737483086,
+      "grad_norm": 0.9054037429104935,
+      "learning_rate": 6.554054054054054e-06,
+      "loss": 0.3413,
+      "step": 194
+    },
+    {
+      "epoch": 0.13193504736129905,
+      "grad_norm": 0.8685048265527847,
+      "learning_rate": 6.587837837837838e-06,
+      "loss": 0.3432,
+      "step": 195
+    },
+    {
+      "epoch": 0.13261163734776726,
+      "grad_norm": 0.8595915682513737,
+      "learning_rate": 6.621621621621622e-06,
+      "loss": 0.3332,
+      "step": 196
+    },
+    {
+      "epoch": 0.13328822733423545,
+      "grad_norm": 0.8661792550561265,
+      "learning_rate": 6.655405405405406e-06,
+      "loss": 0.3296,
+      "step": 197
+    },
+    {
+      "epoch": 0.13396481732070364,
+      "grad_norm": 0.8648875656286432,
+      "learning_rate": 6.689189189189191e-06,
+      "loss": 0.3311,
+      "step": 198
+    },
+    {
+      "epoch": 0.13464140730717186,
+      "grad_norm": 0.8898885781197285,
+      "learning_rate": 6.722972972972973e-06,
+      "loss": 0.3441,
+      "step": 199
+    },
+    {
+      "epoch": 0.13531799729364005,
+      "grad_norm": 1.0064185808659396,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.3398,
+      "step": 200
+    },
+    {
+      "epoch": 0.13599458728010824,
+      "grad_norm": 0.8488676927743474,
+      "learning_rate": 6.790540540540541e-06,
+      "loss": 0.3348,
+      "step": 201
+    },
+    {
+      "epoch": 0.13667117726657646,
+      "grad_norm": 0.8338753305142843,
+      "learning_rate": 6.824324324324325e-06,
+      "loss": 0.342,
+      "step": 202
+    },
+    {
+      "epoch": 0.13734776725304465,
+      "grad_norm": 0.9451790114855093,
+      "learning_rate": 6.858108108108109e-06,
+      "loss": 0.3461,
+      "step": 203
+    },
+    {
+      "epoch": 0.13802435723951287,
+      "grad_norm": 0.8906201870575157,
+      "learning_rate": 6.891891891891892e-06,
+      "loss": 0.3452,
+      "step": 204
+    },
+    {
+      "epoch": 0.13870094722598106,
+      "grad_norm": 0.8530465629443367,
+      "learning_rate": 6.925675675675676e-06,
+      "loss": 0.3472,
+      "step": 205
+    },
+    {
+      "epoch": 0.13937753721244925,
+      "grad_norm": 0.8685095309624206,
+      "learning_rate": 6.95945945945946e-06,
+      "loss": 0.3357,
+      "step": 206
+    },
+    {
+      "epoch": 0.14005412719891747,
+      "grad_norm": 0.8641588636876336,
+      "learning_rate": 6.993243243243244e-06,
+      "loss": 0.3432,
+      "step": 207
+    },
+    {
+      "epoch": 0.14073071718538566,
+      "grad_norm": 0.8135358923354462,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.3366,
+      "step": 208
+    },
+    {
+      "epoch": 0.14140730717185385,
+      "grad_norm": 0.7944919501238901,
+      "learning_rate": 7.060810810810811e-06,
+      "loss": 0.3345,
+      "step": 209
+    },
+    {
+      "epoch": 0.14208389715832206,
+      "grad_norm": 0.8818931083400321,
+      "learning_rate": 7.0945945945945946e-06,
+      "loss": 0.3351,
+      "step": 210
+    },
+    {
+      "epoch": 0.14276048714479025,
+      "grad_norm": 0.832227859323065,
+      "learning_rate": 7.128378378378379e-06,
+      "loss": 0.3394,
+      "step": 211
+    },
+    {
+      "epoch": 0.14343707713125844,
+      "grad_norm": 0.8374932477499295,
+      "learning_rate": 7.162162162162163e-06,
+      "loss": 0.3372,
+      "step": 212
+    },
+    {
+      "epoch": 0.14411366711772666,
+      "grad_norm": 0.801713296556622,
+      "learning_rate": 7.195945945945947e-06,
+      "loss": 0.3453,
+      "step": 213
+    },
+    {
+      "epoch": 0.14479025710419485,
+      "grad_norm": 0.8425078994820807,
+      "learning_rate": 7.229729729729731e-06,
+      "loss": 0.3334,
+      "step": 214
+    },
+    {
+      "epoch": 0.14546684709066307,
+      "grad_norm": 0.8165957140324827,
+      "learning_rate": 7.263513513513514e-06,
+      "loss": 0.3383,
+      "step": 215
+    },
+    {
+      "epoch": 0.14614343707713126,
+      "grad_norm": 0.8269447244193878,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.3337,
+      "step": 216
+    },
+    {
+      "epoch": 0.14682002706359945,
+      "grad_norm": 0.7785212763654489,
+      "learning_rate": 7.331081081081082e-06,
+      "loss": 0.3313,
+      "step": 217
+    },
+    {
+      "epoch": 0.14749661705006767,
+      "grad_norm": 0.8360693119035515,
+      "learning_rate": 7.3648648648648655e-06,
+      "loss": 0.3494,
+      "step": 218
+    },
+    {
+      "epoch": 0.14817320703653586,
+      "grad_norm": 0.8552586338707748,
+      "learning_rate": 7.398648648648649e-06,
+      "loss": 0.3445,
+      "step": 219
+    },
+    {
+      "epoch": 0.14884979702300405,
+      "grad_norm": 0.780737840620897,
+      "learning_rate": 7.4324324324324324e-06,
+      "loss": 0.3372,
+      "step": 220
+    },
+    {
+      "epoch": 0.14952638700947227,
+      "grad_norm": 0.8183492714062136,
+      "learning_rate": 7.466216216216216e-06,
+      "loss": 0.336,
+      "step": 221
+    },
+    {
+      "epoch": 0.15020297699594046,
+      "grad_norm": 0.8450841667172513,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3277,
+      "step": 222
+    },
+    {
+      "epoch": 0.15087956698240865,
+      "grad_norm": 0.7980075139308351,
+      "learning_rate": 7.533783783783785e-06,
+      "loss": 0.3392,
+      "step": 223
+    },
+    {
+      "epoch": 0.15155615696887687,
+      "grad_norm": 0.8555238793947134,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.3324,
+      "step": 224
+    },
+    {
+      "epoch": 0.15223274695534506,
+      "grad_norm": 0.8544886059464901,
+      "learning_rate": 7.601351351351352e-06,
+      "loss": 0.3492,
+      "step": 225
+    },
+    {
+      "epoch": 0.15290933694181327,
+      "grad_norm": 0.8629137329616998,
+      "learning_rate": 7.635135135135135e-06,
+      "loss": 0.3414,
+      "step": 226
+    },
+    {
+      "epoch": 0.15358592692828146,
+      "grad_norm": 0.887225279695075,
+      "learning_rate": 7.668918918918919e-06,
+      "loss": 0.3428,
+      "step": 227
+    },
+    {
+      "epoch": 0.15426251691474965,
+      "grad_norm": 0.8309492854826752,
+      "learning_rate": 7.702702702702704e-06,
+      "loss": 0.3433,
+      "step": 228
+    },
+    {
+      "epoch": 0.15493910690121787,
+      "grad_norm": 0.8162196223267019,
+      "learning_rate": 7.736486486486488e-06,
+      "loss": 0.3199,
+      "step": 229
+    },
+    {
+      "epoch": 0.15561569688768606,
+      "grad_norm": 0.8161488289421107,
+      "learning_rate": 7.77027027027027e-06,
+      "loss": 0.3141,
+      "step": 230
+    },
+    {
+      "epoch": 0.15629228687415425,
+      "grad_norm": 0.8625073153106998,
+      "learning_rate": 7.804054054054054e-06,
+      "loss": 0.3406,
+      "step": 231
+    },
+    {
+      "epoch": 0.15696887686062247,
+      "grad_norm": 0.7696156006298618,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.3231,
+      "step": 232
+    },
+    {
+      "epoch": 0.15764546684709066,
+      "grad_norm": 0.8223138333117583,
+      "learning_rate": 7.871621621621622e-06,
+      "loss": 0.3424,
+      "step": 233
+    },
+    {
+      "epoch": 0.15832205683355885,
+      "grad_norm": 0.8458985905034395,
+      "learning_rate": 7.905405405405406e-06,
+      "loss": 0.3296,
+      "step": 234
+    },
+    {
+      "epoch": 0.15899864682002707,
+      "grad_norm": 0.8054757643753689,
+      "learning_rate": 7.93918918918919e-06,
+      "loss": 0.3267,
+      "step": 235
+    },
+    {
+      "epoch": 0.15967523680649526,
+      "grad_norm": 0.8649006176454097,
+      "learning_rate": 7.972972972972974e-06,
+      "loss": 0.3386,
+      "step": 236
+    },
+    {
+      "epoch": 0.16035182679296348,
+      "grad_norm": 0.8531529543945617,
+      "learning_rate": 8.006756756756757e-06,
+      "loss": 0.3215,
+      "step": 237
+    },
+    {
+      "epoch": 0.16102841677943167,
+      "grad_norm": 0.8000272587465453,
+      "learning_rate": 8.040540540540541e-06,
+      "loss": 0.3337,
+      "step": 238
+    },
+    {
+      "epoch": 0.16170500676589986,
+      "grad_norm": 0.8976529685095862,
+      "learning_rate": 8.074324324324325e-06,
+      "loss": 0.342,
+      "step": 239
+    },
+    {
+      "epoch": 0.16238159675236807,
+      "grad_norm": 0.808196429166087,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.3312,
+      "step": 240
+    },
+    {
+      "epoch": 0.16305818673883626,
+      "grad_norm": 0.8099645164859992,
+      "learning_rate": 8.141891891891893e-06,
+      "loss": 0.3437,
+      "step": 241
+    },
+    {
+      "epoch": 0.16373477672530445,
+      "grad_norm": 0.848085315064657,
+      "learning_rate": 8.175675675675677e-06,
+      "loss": 0.3384,
+      "step": 242
+    },
+    {
+      "epoch": 0.16441136671177267,
+      "grad_norm": 0.8010656933028365,
+      "learning_rate": 8.20945945945946e-06,
+      "loss": 0.3378,
+      "step": 243
+    },
+    {
+      "epoch": 0.16508795669824086,
+      "grad_norm": 0.7730580274807233,
+      "learning_rate": 8.243243243243245e-06,
+      "loss": 0.3262,
+      "step": 244
+    },
+    {
+      "epoch": 0.16576454668470908,
+      "grad_norm": 0.8327533897123965,
+      "learning_rate": 8.277027027027028e-06,
+      "loss": 0.3241,
+      "step": 245
+    },
+    {
+      "epoch": 0.16644113667117727,
+      "grad_norm": 0.8015396933833453,
+      "learning_rate": 8.31081081081081e-06,
+      "loss": 0.3394,
+      "step": 246
+    },
+    {
+      "epoch": 0.16711772665764546,
+      "grad_norm": 0.8001509187042862,
+      "learning_rate": 8.344594594594594e-06,
+      "loss": 0.3304,
+      "step": 247
+    },
+    {
+      "epoch": 0.16779431664411368,
+      "grad_norm": 0.8313438783713415,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.3294,
+      "step": 248
+    },
+    {
+      "epoch": 0.16847090663058187,
+      "grad_norm": 0.8118320962346404,
+      "learning_rate": 8.412162162162162e-06,
+      "loss": 0.3242,
+      "step": 249
+    },
+    {
+      "epoch": 0.16914749661705006,
+      "grad_norm": 0.7969297983964378,
+      "learning_rate": 8.445945945945948e-06,
+      "loss": 0.3237,
+      "step": 250
+    },
+    {
+      "epoch": 0.16982408660351828,
+      "grad_norm": 0.9167940232320994,
+      "learning_rate": 8.479729729729732e-06,
+      "loss": 0.3306,
+      "step": 251
+    },
+    {
+      "epoch": 0.17050067658998647,
+      "grad_norm": 0.8178030936476689,
+      "learning_rate": 8.513513513513514e-06,
+      "loss": 0.3234,
+      "step": 252
+    },
+    {
+      "epoch": 0.17117726657645466,
+      "grad_norm": 0.723320427907144,
+      "learning_rate": 8.547297297297298e-06,
+      "loss": 0.3313,
+      "step": 253
+    },
+    {
+      "epoch": 0.17185385656292287,
+      "grad_norm": 0.8490630570763466,
+      "learning_rate": 8.581081081081082e-06,
+      "loss": 0.326,
+      "step": 254
+    },
+    {
+      "epoch": 0.17253044654939106,
+      "grad_norm": 0.8050936764006791,
+      "learning_rate": 8.614864864864865e-06,
+      "loss": 0.3304,
+      "step": 255
+    },
+    {
+      "epoch": 0.17320703653585928,
+      "grad_norm": 0.8021316627878387,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.3205,
+      "step": 256
+    },
+    {
+      "epoch": 0.17388362652232747,
+      "grad_norm": 0.854957295850016,
+      "learning_rate": 8.682432432432433e-06,
+      "loss": 0.3236,
+      "step": 257
+    },
+    {
+      "epoch": 0.17456021650879566,
+      "grad_norm": 0.7872059217232962,
+      "learning_rate": 8.716216216216217e-06,
+      "loss": 0.3281,
+      "step": 258
+    },
+    {
+      "epoch": 0.17523680649526388,
+      "grad_norm": 0.7948236920064615,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.3325,
+      "step": 259
+    },
+    {
+      "epoch": 0.17591339648173207,
+      "grad_norm": 0.8333016404188799,
+      "learning_rate": 8.783783783783785e-06,
+      "loss": 0.333,
+      "step": 260
+    },
+    {
+      "epoch": 0.17658998646820026,
+      "grad_norm": 0.8137808811423364,
+      "learning_rate": 8.817567567567569e-06,
+      "loss": 0.3268,
+      "step": 261
+    },
+    {
+      "epoch": 0.17726657645466848,
+      "grad_norm": 0.7731565925256881,
+      "learning_rate": 8.851351351351351e-06,
+      "loss": 0.3293,
+      "step": 262
+    },
+    {
+      "epoch": 0.17794316644113667,
+      "grad_norm": 0.7683919633669537,
+      "learning_rate": 8.885135135135136e-06,
+      "loss": 0.3299,
+      "step": 263
+    },
+    {
+      "epoch": 0.17861975642760486,
+      "grad_norm": 0.819555331626936,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.3355,
+      "step": 264
+    },
+    {
+      "epoch": 0.17929634641407308,
+      "grad_norm": 0.7781306429348449,
+      "learning_rate": 8.952702702702704e-06,
+      "loss": 0.3233,
+      "step": 265
+    },
+    {
+      "epoch": 0.17997293640054127,
+      "grad_norm": 0.785509377820432,
+      "learning_rate": 8.986486486486488e-06,
+      "loss": 0.3291,
+      "step": 266
+    },
+    {
+      "epoch": 0.18064952638700948,
+      "grad_norm": 0.8208983205271441,
+      "learning_rate": 9.02027027027027e-06,
+      "loss": 0.3333,
+      "step": 267
+    },
+    {
+      "epoch": 0.18132611637347767,
+      "grad_norm": 0.7931088156225514,
+      "learning_rate": 9.054054054054054e-06,
+      "loss": 0.3186,
+      "step": 268
+    },
+    {
+      "epoch": 0.18200270635994586,
+      "grad_norm": 0.8035106213295651,
+      "learning_rate": 9.087837837837838e-06,
+      "loss": 0.3166,
+      "step": 269
+    },
+    {
+      "epoch": 0.18267929634641408,
+      "grad_norm": 0.7741955901504421,
+      "learning_rate": 9.121621621621622e-06,
+      "loss": 0.3264,
+      "step": 270
+    },
+    {
+      "epoch": 0.18335588633288227,
+      "grad_norm": 0.8180600177812944,
+      "learning_rate": 9.155405405405406e-06,
+      "loss": 0.3339,
+      "step": 271
+    },
+    {
+      "epoch": 0.18403247631935046,
+      "grad_norm": 0.866459485540038,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.3244,
+      "step": 272
+    },
+    {
+      "epoch": 0.18470906630581868,
+      "grad_norm": 0.7823472378067216,
+      "learning_rate": 9.222972972972973e-06,
+      "loss": 0.3291,
+      "step": 273
+    },
+    {
+      "epoch": 0.18538565629228687,
+      "grad_norm": 0.8610321534441738,
+      "learning_rate": 9.256756756756757e-06,
+      "loss": 0.338,
+      "step": 274
+    },
+    {
+      "epoch": 0.18606224627875506,
+      "grad_norm": 0.7618907443885038,
+      "learning_rate": 9.290540540540541e-06,
+      "loss": 0.3281,
+      "step": 275
+    },
+    {
+      "epoch": 0.18673883626522328,
+      "grad_norm": 0.8049045352742612,
+      "learning_rate": 9.324324324324325e-06,
+      "loss": 0.322,
+      "step": 276
+    },
+    {
+      "epoch": 0.18741542625169147,
+      "grad_norm": 0.7861965444748493,
+      "learning_rate": 9.358108108108109e-06,
+      "loss": 0.3177,
+      "step": 277
+    },
+    {
+      "epoch": 0.1880920162381597,
+      "grad_norm": 0.7897762321100874,
+      "learning_rate": 9.391891891891893e-06,
+      "loss": 0.332,
+      "step": 278
+    },
+    {
+      "epoch": 0.18876860622462788,
+      "grad_norm": 0.8788310882786632,
+      "learning_rate": 9.425675675675677e-06,
+      "loss": 0.3313,
+      "step": 279
+    },
+    {
+      "epoch": 0.18944519621109607,
+      "grad_norm": 0.7811693504857151,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.3385,
+      "step": 280
+    },
+    {
+      "epoch": 0.19012178619756429,
+      "grad_norm": 0.7754862014396956,
+      "learning_rate": 9.493243243243244e-06,
+      "loss": 0.3332,
+      "step": 281
+    },
+    {
+      "epoch": 0.19079837618403248,
+      "grad_norm": 0.8122520146710277,
+      "learning_rate": 9.527027027027028e-06,
+      "loss": 0.335,
+      "step": 282
+    },
+    {
+      "epoch": 0.19147496617050067,
+      "grad_norm": 0.7298813992473239,
+      "learning_rate": 9.56081081081081e-06,
+      "loss": 0.3109,
+      "step": 283
+    },
+    {
+      "epoch": 0.19215155615696888,
+      "grad_norm": 0.7221936539490259,
+      "learning_rate": 9.594594594594594e-06,
+      "loss": 0.3307,
+      "step": 284
+    },
+    {
+      "epoch": 0.19282814614343707,
+      "grad_norm": 0.8342843395499431,
+      "learning_rate": 9.62837837837838e-06,
+      "loss": 0.3293,
+      "step": 285
+    },
+    {
+      "epoch": 0.19350473612990526,
+      "grad_norm": 0.7777084901763289,
+      "learning_rate": 9.662162162162164e-06,
+      "loss": 0.3208,
+      "step": 286
+    },
+    {
+      "epoch": 0.19418132611637348,
+      "grad_norm": 0.8172639830470911,
+      "learning_rate": 9.695945945945948e-06,
+      "loss": 0.3268,
+      "step": 287
+    },
+    {
+      "epoch": 0.19485791610284167,
+      "grad_norm": 0.7674474584858759,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.3282,
+      "step": 288
+    },
+    {
+      "epoch": 0.1955345060893099,
+      "grad_norm": 0.7770709011335428,
+      "learning_rate": 9.763513513513514e-06,
+      "loss": 0.3294,
+      "step": 289
+    },
+    {
+      "epoch": 0.19621109607577808,
+      "grad_norm": 0.7678884697339083,
+      "learning_rate": 9.797297297297298e-06,
+      "loss": 0.3169,
+      "step": 290
+    },
+    {
+      "epoch": 0.19688768606224627,
+      "grad_norm": 0.7865828177253228,
+      "learning_rate": 9.831081081081081e-06,
+      "loss": 0.3331,
+      "step": 291
+    },
+    {
+      "epoch": 0.1975642760487145,
+      "grad_norm": 0.8080930078633354,
+      "learning_rate": 9.864864864864865e-06,
+      "loss": 0.3272,
+      "step": 292
+    },
+    {
+      "epoch": 0.19824086603518268,
+      "grad_norm": 0.759717542045495,
+      "learning_rate": 9.89864864864865e-06,
+      "loss": 0.3324,
+      "step": 293
+    },
+    {
+      "epoch": 0.19891745602165087,
+      "grad_norm": 0.7549920836500931,
+      "learning_rate": 9.932432432432433e-06,
+      "loss": 0.3367,
+      "step": 294
+    },
+    {
+      "epoch": 0.19959404600811909,
+      "grad_norm": 0.7719081399128706,
+      "learning_rate": 9.966216216216217e-06,
+      "loss": 0.3229,
+      "step": 295
+    },
+    {
+      "epoch": 0.20027063599458728,
+      "grad_norm": 0.7882836370096206,
+      "learning_rate": 1e-05,
+      "loss": 0.3342,
+      "step": 296
+    },
+    {
+      "epoch": 0.2009472259810555,
+      "grad_norm": 0.7467690177833016,
+      "learning_rate": 9.999996512803418e-06,
+      "loss": 0.3174,
+      "step": 297
+    },
+    {
+      "epoch": 0.20162381596752368,
+      "grad_norm": 0.7765581370169501,
+      "learning_rate": 9.999986051218538e-06,
+      "loss": 0.3301,
+      "step": 298
+    },
+    {
+      "epoch": 0.20230040595399187,
+      "grad_norm": 0.7459278607295443,
+      "learning_rate": 9.99996861525995e-06,
+      "loss": 0.3324,
+      "step": 299
+    },
+    {
+      "epoch": 0.2029769959404601,
+      "grad_norm": 0.7034749747661629,
+      "learning_rate": 9.999944204951974e-06,
+      "loss": 0.3326,
+      "step": 300
+    },
+    {
+      "epoch": 0.20365358592692828,
+      "grad_norm": 0.7202900537200706,
+      "learning_rate": 9.999912820328665e-06,
+      "loss": 0.308,
+      "step": 301
+    },
+    {
+      "epoch": 0.20433017591339647,
+      "grad_norm": 0.7716372864187919,
+      "learning_rate": 9.999874461433796e-06,
+      "loss": 0.3357,
+      "step": 302
+    },
+    {
+      "epoch": 0.2050067658998647,
+      "grad_norm": 0.9647240258696954,
+      "learning_rate": 9.999829128320873e-06,
+      "loss": 0.3342,
+      "step": 303
+    },
+    {
+      "epoch": 0.20568335588633288,
+      "grad_norm": 25.635252553252815,
+      "learning_rate": 9.999776821053134e-06,
+      "loss": 0.3318,
+      "step": 304
+    },
+    {
+      "epoch": 0.20635994587280107,
+      "grad_norm": 6.444807365789432,
+      "learning_rate": 9.999717539703537e-06,
+      "loss": 0.3343,
+      "step": 305
+    },
+    {
+      "epoch": 0.2070365358592693,
+      "grad_norm": 1.0667176755110874,
+      "learning_rate": 9.999651284354774e-06,
+      "loss": 0.3346,
+      "step": 306
+    },
+    {
+      "epoch": 0.20771312584573748,
+      "grad_norm": 1.0462130038433695,
+      "learning_rate": 9.999578055099264e-06,
+      "loss": 0.3256,
+      "step": 307
+    },
+    {
+      "epoch": 0.2083897158322057,
+      "grad_norm": 0.8588850970989159,
+      "learning_rate": 9.999497852039152e-06,
+      "loss": 0.3288,
+      "step": 308
+    },
+    {
+      "epoch": 0.2090663058186739,
+      "grad_norm": 0.890093795694103,
+      "learning_rate": 9.999410675286313e-06,
+      "loss": 0.3167,
+      "step": 309
+    },
+    {
+      "epoch": 0.20974289580514208,
+      "grad_norm": 0.804218556080729,
+      "learning_rate": 9.999316524962347e-06,
+      "loss": 0.3297,
+      "step": 310
+    },
+    {
+      "epoch": 0.2104194857916103,
+      "grad_norm": 0.8156612055895035,
+      "learning_rate": 9.999215401198579e-06,
+      "loss": 0.3296,
+      "step": 311
+    },
+    {
+      "epoch": 0.21109607577807848,
+      "grad_norm": 0.8205392275301405,
+      "learning_rate": 9.999107304136068e-06,
+      "loss": 0.3282,
+      "step": 312
+    },
+    {
+      "epoch": 0.21177266576454667,
+      "grad_norm": 0.8047843674789104,
+      "learning_rate": 9.998992233925597e-06,
+      "loss": 0.3348,
+      "step": 313
+    },
+    {
+      "epoch": 0.2124492557510149,
+      "grad_norm": 0.7869652823594478,
+      "learning_rate": 9.998870190727674e-06,
+      "loss": 0.3311,
+      "step": 314
+    },
+    {
+      "epoch": 0.21312584573748308,
+      "grad_norm": 0.8054610741588345,
+      "learning_rate": 9.998741174712534e-06,
+      "loss": 0.3324,
+      "step": 315
+    },
+    {
+      "epoch": 0.21380243572395127,
+      "grad_norm": 0.7901959146376306,
+      "learning_rate": 9.998605186060138e-06,
+      "loss": 0.3473,
+      "step": 316
+    },
+    {
+      "epoch": 0.2144790257104195,
+      "grad_norm": 0.6940088237673759,
+      "learning_rate": 9.998462224960176e-06,
+      "loss": 0.3247,
+      "step": 317
+    },
+    {
+      "epoch": 0.21515561569688768,
+      "grad_norm": 0.7628690891181793,
+      "learning_rate": 9.998312291612056e-06,
+      "loss": 0.3308,
+      "step": 318
+    },
+    {
+      "epoch": 0.2158322056833559,
+      "grad_norm": 0.7907871892000858,
+      "learning_rate": 9.998155386224924e-06,
+      "loss": 0.3249,
+      "step": 319
+    },
+    {
+      "epoch": 0.2165087956698241,
+      "grad_norm": 0.7399292833476737,
+      "learning_rate": 9.99799150901764e-06,
+      "loss": 0.3047,
+      "step": 320
+    },
+    {
+      "epoch": 0.21718538565629228,
+      "grad_norm": 0.7459946190156378,
+      "learning_rate": 9.997820660218793e-06,
+      "loss": 0.3153,
+      "step": 321
+    },
+    {
+      "epoch": 0.2178619756427605,
+      "grad_norm": 0.7890710078242472,
+      "learning_rate": 9.997642840066696e-06,
+      "loss": 0.327,
+      "step": 322
+    },
+    {
+      "epoch": 0.2185385656292287,
+      "grad_norm": 0.7407355909745483,
+      "learning_rate": 9.99745804880939e-06,
+      "loss": 0.3328,
+      "step": 323
+    },
+    {
+      "epoch": 0.21921515561569688,
+      "grad_norm": 0.7415916080169501,
+      "learning_rate": 9.99726628670463e-06,
+      "loss": 0.3268,
+      "step": 324
+    },
+    {
+      "epoch": 0.2198917456021651,
+      "grad_norm": 0.7651442891896504,
+      "learning_rate": 9.997067554019908e-06,
+      "loss": 0.3109,
+      "step": 325
+    },
+    {
+      "epoch": 0.22056833558863329,
+      "grad_norm": 0.735294877142389,
+      "learning_rate": 9.996861851032426e-06,
+      "loss": 0.3218,
+      "step": 326
+    },
+    {
+      "epoch": 0.22124492557510148,
+      "grad_norm": 0.7307546439145841,
+      "learning_rate": 9.996649178029118e-06,
+      "loss": 0.3285,
+      "step": 327
+    },
+    {
+      "epoch": 0.2219215155615697,
+      "grad_norm": 0.7328840513592791,
+      "learning_rate": 9.996429535306638e-06,
+      "loss": 0.318,
+      "step": 328
+    },
+    {
+      "epoch": 0.22259810554803788,
+      "grad_norm": 0.7069847553846568,
+      "learning_rate": 9.996202923171356e-06,
+      "loss": 0.3206,
+      "step": 329
+    },
+    {
+      "epoch": 0.2232746955345061,
+      "grad_norm": 0.7520320253222692,
+      "learning_rate": 9.995969341939373e-06,
+      "loss": 0.3238,
+      "step": 330
+    },
+    {
+      "epoch": 0.2239512855209743,
+      "grad_norm": 0.7109457294179562,
+      "learning_rate": 9.995728791936505e-06,
+      "loss": 0.3152,
+      "step": 331
+    },
+    {
+      "epoch": 0.22462787550744248,
+      "grad_norm": 0.6957311508123549,
+      "learning_rate": 9.995481273498291e-06,
+      "loss": 0.3432,
+      "step": 332
+    },
+    {
+      "epoch": 0.2253044654939107,
+      "grad_norm": 0.7189386129204676,
+      "learning_rate": 9.995226786969988e-06,
+      "loss": 0.3198,
+      "step": 333
+    },
+    {
+      "epoch": 0.2259810554803789,
+      "grad_norm": 0.7582421324570706,
+      "learning_rate": 9.994965332706574e-06,
+      "loss": 0.3164,
+      "step": 334
+    },
+    {
+      "epoch": 0.22665764546684708,
+      "grad_norm": 0.7335925129229004,
+      "learning_rate": 9.994696911072745e-06,
+      "loss": 0.3285,
+      "step": 335
+    },
+    {
+      "epoch": 0.2273342354533153,
+      "grad_norm": 0.722230869179683,
+      "learning_rate": 9.99442152244292e-06,
+      "loss": 0.3278,
+      "step": 336
+    },
+    {
+      "epoch": 0.2280108254397835,
+      "grad_norm": 0.6921768544402361,
+      "learning_rate": 9.994139167201229e-06,
+      "loss": 0.3139,
+      "step": 337
+    },
+    {
+      "epoch": 0.22868741542625168,
+      "grad_norm": 0.7155919270725442,
+      "learning_rate": 9.993849845741525e-06,
+      "loss": 0.3275,
+      "step": 338
+    },
+    {
+      "epoch": 0.2293640054127199,
+      "grad_norm": 0.7016116281069982,
+      "learning_rate": 9.993553558467376e-06,
+      "loss": 0.3153,
+      "step": 339
+    },
+    {
+      "epoch": 0.23004059539918809,
+      "grad_norm": 0.7586629837817358,
+      "learning_rate": 9.993250305792067e-06,
+      "loss": 0.3328,
+      "step": 340
+    },
+    {
+      "epoch": 0.2307171853856563,
+      "grad_norm": 0.7275380930246326,
+      "learning_rate": 9.992940088138598e-06,
+      "loss": 0.3339,
+      "step": 341
+    },
+    {
+      "epoch": 0.2313937753721245,
+      "grad_norm": 0.7120875041750986,
+      "learning_rate": 9.992622905939686e-06,
+      "loss": 0.3133,
+      "step": 342
+    },
+    {
+      "epoch": 0.23207036535859268,
+      "grad_norm": 0.7411887530647576,
+      "learning_rate": 9.99229875963776e-06,
+      "loss": 0.3369,
+      "step": 343
+    },
+    {
+      "epoch": 0.2327469553450609,
+      "grad_norm": 0.7715039273023475,
+      "learning_rate": 9.991967649684967e-06,
+      "loss": 0.3223,
+      "step": 344
+    },
+    {
+      "epoch": 0.2334235453315291,
+      "grad_norm": 0.7024817330857115,
+      "learning_rate": 9.991629576543164e-06,
+      "loss": 0.3259,
+      "step": 345
+    },
+    {
+      "epoch": 0.23410013531799728,
+      "grad_norm": 0.72440507380587,
+      "learning_rate": 9.991284540683922e-06,
+      "loss": 0.3332,
+      "step": 346
+    },
+    {
+      "epoch": 0.2347767253044655,
+      "grad_norm": 0.7496790046680725,
+      "learning_rate": 9.990932542588524e-06,
+      "loss": 0.3235,
+      "step": 347
+    },
+    {
+      "epoch": 0.2354533152909337,
+      "grad_norm": 0.7282899676532798,
+      "learning_rate": 9.990573582747965e-06,
+      "loss": 0.3238,
+      "step": 348
+    },
+    {
+      "epoch": 0.2361299052774019,
+      "grad_norm": 0.6917063483038498,
+      "learning_rate": 9.99020766166295e-06,
+      "loss": 0.3338,
+      "step": 349
+    },
+    {
+      "epoch": 0.2368064952638701,
+      "grad_norm": 0.7511376052974671,
+      "learning_rate": 9.989834779843895e-06,
+      "loss": 0.3217,
+      "step": 350
+    },
+    {
+      "epoch": 0.2374830852503383,
+      "grad_norm": 0.7260244750224937,
+      "learning_rate": 9.989454937810926e-06,
+      "loss": 0.3272,
+      "step": 351
+    },
+    {
+      "epoch": 0.2381596752368065,
+      "grad_norm": 0.6892138858970657,
+      "learning_rate": 9.989068136093873e-06,
+      "loss": 0.3176,
+      "step": 352
+    },
+    {
+      "epoch": 0.2388362652232747,
+      "grad_norm": 0.705313812505697,
+      "learning_rate": 9.98867437523228e-06,
+      "loss": 0.3268,
+      "step": 353
+    },
+    {
+      "epoch": 0.2395128552097429,
+      "grad_norm": 0.7421506698657179,
+      "learning_rate": 9.988273655775398e-06,
+      "loss": 0.3297,
+      "step": 354
+    },
+    {
+      "epoch": 0.2401894451962111,
+      "grad_norm": 0.6942955697743944,
+      "learning_rate": 9.987865978282177e-06,
+      "loss": 0.3144,
+      "step": 355
+    },
+    {
+      "epoch": 0.2408660351826793,
+      "grad_norm": 0.711084605255173,
+      "learning_rate": 9.98745134332128e-06,
+      "loss": 0.3287,
+      "step": 356
+    },
+    {
+      "epoch": 0.24154262516914748,
+      "grad_norm": 0.7418772516476061,
+      "learning_rate": 9.987029751471072e-06,
+      "loss": 0.3171,
+      "step": 357
+    },
+    {
+      "epoch": 0.2422192151556157,
+      "grad_norm": 0.7422162227173037,
+      "learning_rate": 9.986601203319623e-06,
+      "loss": 0.3254,
+      "step": 358
+    },
+    {
+      "epoch": 0.2428958051420839,
+      "grad_norm": 0.7445627452809618,
+      "learning_rate": 9.986165699464706e-06,
+      "loss": 0.3418,
+      "step": 359
+    },
+    {
+      "epoch": 0.2435723951285521,
+      "grad_norm": 0.7300828760676705,
+      "learning_rate": 9.985723240513795e-06,
+      "loss": 0.3371,
+      "step": 360
+    },
+    {
+      "epoch": 0.2442489851150203,
+      "grad_norm": 0.6823845374996537,
+      "learning_rate": 9.985273827084068e-06,
+      "loss": 0.3293,
+      "step": 361
+    },
+    {
+      "epoch": 0.2449255751014885,
+      "grad_norm": 0.723048002129932,
+      "learning_rate": 9.9848174598024e-06,
+      "loss": 0.3171,
+      "step": 362
+    },
+    {
+      "epoch": 0.2456021650879567,
+      "grad_norm": 0.7189844634292714,
+      "learning_rate": 9.98435413930537e-06,
+      "loss": 0.318,
+      "step": 363
+    },
+    {
+      "epoch": 0.2462787550744249,
+      "grad_norm": 0.7069237821827394,
+      "learning_rate": 9.983883866239253e-06,
+      "loss": 0.3147,
+      "step": 364
+    },
+    {
+      "epoch": 0.2469553450608931,
+      "grad_norm": 0.7113777473123136,
+      "learning_rate": 9.983406641260023e-06,
+      "loss": 0.3326,
+      "step": 365
+    },
+    {
+      "epoch": 0.2476319350473613,
+      "grad_norm": 0.6992974839392972,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.3115,
+      "step": 366
+    },
+    {
+      "epoch": 0.2483085250338295,
+      "grad_norm": 0.7527107717720851,
+      "learning_rate": 9.982431338234602e-06,
+      "loss": 0.3256,
+      "step": 367
+    },
+    {
+      "epoch": 0.2489851150202977,
+      "grad_norm": 0.74304279619225,
+      "learning_rate": 9.981933261548841e-06,
+      "loss": 0.3292,
+      "step": 368
+    },
+    {
+      "epoch": 0.2496617050067659,
+      "grad_norm": 0.7101368716732751,
+      "learning_rate": 9.981428235670825e-06,
+      "loss": 0.3133,
+      "step": 369
+    },
+    {
+      "epoch": 0.2503382949932341,
+      "grad_norm": 0.681924721787861,
+      "learning_rate": 9.980916261305002e-06,
+      "loss": 0.3096,
+      "step": 370
+    },
+    {
+      "epoch": 0.2510148849797023,
+      "grad_norm": 0.6966233432435366,
+      "learning_rate": 9.980397339165514e-06,
+      "loss": 0.342,
+      "step": 371
+    },
+    {
+      "epoch": 0.2516914749661705,
+      "grad_norm": 0.7111788880077314,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.3283,
+      "step": 372
+    },
+    {
+      "epoch": 0.2523680649526387,
+      "grad_norm": 0.6621387582938183,
+      "learning_rate": 9.97933865447057e-06,
+      "loss": 0.3265,
+      "step": 373
+    },
+    {
+      "epoch": 0.2530446549391069,
+      "grad_norm": 0.6470924071582645,
+      "learning_rate": 9.97879889339185e-06,
+      "loss": 0.3289,
+      "step": 374
+    },
+    {
+      "epoch": 0.25372124492557513,
+      "grad_norm": 0.6851135223072917,
+      "learning_rate": 9.978252187492936e-06,
+      "loss": 0.3099,
+      "step": 375
+    },
+    {
+      "epoch": 0.2543978349120433,
+      "grad_norm": 0.6987762338079896,
+      "learning_rate": 9.97769853753642e-06,
+      "loss": 0.328,
+      "step": 376
+    },
+    {
+      "epoch": 0.2550744248985115,
+      "grad_norm": 0.7192546923305027,
+      "learning_rate": 9.97713794429457e-06,
+      "loss": 0.3162,
+      "step": 377
+    },
+    {
+      "epoch": 0.2557510148849797,
+      "grad_norm": 0.6714300111389941,
+      "learning_rate": 9.97657040854935e-06,
+      "loss": 0.3236,
+      "step": 378
+    },
+    {
+      "epoch": 0.2564276048714479,
+      "grad_norm": 0.6729206417173349,
+      "learning_rate": 9.975995931092404e-06,
+      "loss": 0.3153,
+      "step": 379
+    },
+    {
+      "epoch": 0.2571041948579161,
+      "grad_norm": 0.7530023117360258,
+      "learning_rate": 9.975414512725058e-06,
+      "loss": 0.3283,
+      "step": 380
+    },
+    {
+      "epoch": 0.2577807848443843,
+      "grad_norm": 0.7187498671286208,
+      "learning_rate": 9.974826154258318e-06,
+      "loss": 0.322,
+      "step": 381
+    },
+    {
+      "epoch": 0.2584573748308525,
+      "grad_norm": 0.6553822163418016,
+      "learning_rate": 9.974230856512874e-06,
+      "loss": 0.306,
+      "step": 382
+    },
+    {
+      "epoch": 0.2591339648173207,
+      "grad_norm": 0.7672133045119465,
+      "learning_rate": 9.973628620319094e-06,
+      "loss": 0.3338,
+      "step": 383
+    },
+    {
+      "epoch": 0.2598105548037889,
+      "grad_norm": 0.744754949666086,
+      "learning_rate": 9.973019446517023e-06,
+      "loss": 0.3275,
+      "step": 384
+    },
+    {
+      "epoch": 0.2604871447902571,
+      "grad_norm": 0.7703938156711007,
+      "learning_rate": 9.972403335956387e-06,
+      "loss": 0.3222,
+      "step": 385
+    },
+    {
+      "epoch": 0.26116373477672533,
+      "grad_norm": 0.7510295605284385,
+      "learning_rate": 9.971780289496585e-06,
+      "loss": 0.3261,
+      "step": 386
+    },
+    {
+      "epoch": 0.2618403247631935,
+      "grad_norm": 0.6603337182591643,
+      "learning_rate": 9.971150308006689e-06,
+      "loss": 0.3167,
+      "step": 387
+    },
+    {
+      "epoch": 0.2625169147496617,
+      "grad_norm": 0.7232758127259956,
+      "learning_rate": 9.970513392365449e-06,
+      "loss": 0.3221,
+      "step": 388
+    },
+    {
+      "epoch": 0.2631935047361299,
+      "grad_norm": 0.6950854352159602,
+      "learning_rate": 9.969869543461283e-06,
+      "loss": 0.3196,
+      "step": 389
+    },
+    {
+      "epoch": 0.2638700947225981,
+      "grad_norm": 0.7031267503331153,
+      "learning_rate": 9.969218762192286e-06,
+      "loss": 0.3329,
+      "step": 390
+    },
+    {
+      "epoch": 0.2645466847090663,
+      "grad_norm": 0.7209635777660744,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.3243,
+      "step": 391
+    },
+    {
+      "epoch": 0.2652232746955345,
+      "grad_norm": 0.7614265079968179,
+      "learning_rate": 9.967896406200498e-06,
+      "loss": 0.3252,
+      "step": 392
+    },
+    {
+      "epoch": 0.2658998646820027,
+      "grad_norm": 0.6953159265418555,
+      "learning_rate": 9.967224833322237e-06,
+      "loss": 0.3278,
+      "step": 393
+    },
+    {
+      "epoch": 0.2665764546684709,
+      "grad_norm": 0.6730219997567399,
+      "learning_rate": 9.966546331768192e-06,
+      "loss": 0.3145,
+      "step": 394
+    },
+    {
+      "epoch": 0.2672530446549391,
+      "grad_norm": 0.6959150306925767,
+      "learning_rate": 9.96586090248479e-06,
+      "loss": 0.3053,
+      "step": 395
+    },
+    {
+      "epoch": 0.2679296346414073,
+      "grad_norm": 0.697315117361425,
+      "learning_rate": 9.965168546428122e-06,
+      "loss": 0.3133,
+      "step": 396
+    },
+    {
+      "epoch": 0.26860622462787553,
+      "grad_norm": 0.6754309224218298,
+      "learning_rate": 9.964469264563939e-06,
+      "loss": 0.3244,
+      "step": 397
+    },
+    {
+      "epoch": 0.2692828146143437,
+      "grad_norm": 0.674570828717732,
+      "learning_rate": 9.963763057867658e-06,
+      "loss": 0.3081,
+      "step": 398
+    },
+    {
+      "epoch": 0.2699594046008119,
+      "grad_norm": 0.6220348056577781,
+      "learning_rate": 9.963049927324349e-06,
+      "loss": 0.3171,
+      "step": 399
+    },
+    {
+      "epoch": 0.2706359945872801,
+      "grad_norm": 0.6512864734731331,
+      "learning_rate": 9.962329873928743e-06,
+      "loss": 0.3043,
+      "step": 400
+    },
+    {
+      "epoch": 0.2713125845737483,
+      "grad_norm": 0.6714841719046208,
+      "learning_rate": 9.961602898685225e-06,
+      "loss": 0.3193,
+      "step": 401
+    },
+    {
+      "epoch": 0.2719891745602165,
+      "grad_norm": 0.6684337463677027,
+      "learning_rate": 9.960869002607843e-06,
+      "loss": 0.3265,
+      "step": 402
+    },
+    {
+      "epoch": 0.27266576454668473,
+      "grad_norm": 0.684676365005976,
+      "learning_rate": 9.960128186720288e-06,
+      "loss": 0.3144,
+      "step": 403
+    },
+    {
+      "epoch": 0.2733423545331529,
+      "grad_norm": 0.7015877158944209,
+      "learning_rate": 9.959380452055909e-06,
+      "loss": 0.3346,
+      "step": 404
+    },
+    {
+      "epoch": 0.2740189445196211,
+      "grad_norm": 0.6579912796903548,
+      "learning_rate": 9.958625799657707e-06,
+      "loss": 0.3105,
+      "step": 405
+    },
+    {
+      "epoch": 0.2746955345060893,
+      "grad_norm": 0.7024510331880638,
+      "learning_rate": 9.95786423057833e-06,
+      "loss": 0.3222,
+      "step": 406
+    },
+    {
+      "epoch": 0.2753721244925575,
+      "grad_norm": 0.6549827726998128,
+      "learning_rate": 9.95709574588007e-06,
+      "loss": 0.3235,
+      "step": 407
+    },
+    {
+      "epoch": 0.27604871447902574,
+      "grad_norm": 0.6690954433838975,
+      "learning_rate": 9.956320346634877e-06,
+      "loss": 0.3189,
+      "step": 408
+    },
+    {
+      "epoch": 0.2767253044654939,
+      "grad_norm": 0.6541006411259631,
+      "learning_rate": 9.955538033924334e-06,
+      "loss": 0.3089,
+      "step": 409
+    },
+    {
+      "epoch": 0.2774018944519621,
+      "grad_norm": 0.7110822662058611,
+      "learning_rate": 9.954748808839675e-06,
+      "loss": 0.3189,
+      "step": 410
+    },
+    {
+      "epoch": 0.2780784844384303,
+      "grad_norm": 0.6807201705024405,
+      "learning_rate": 9.95395267248177e-06,
+      "loss": 0.3102,
+      "step": 411
+    },
+    {
+      "epoch": 0.2787550744248985,
+      "grad_norm": 0.6634764039010798,
+      "learning_rate": 9.953149625961136e-06,
+      "loss": 0.3289,
+      "step": 412
+    },
+    {
+      "epoch": 0.2794316644113667,
+      "grad_norm": 0.7093419654198937,
+      "learning_rate": 9.952339670397924e-06,
+      "loss": 0.3218,
+      "step": 413
+    },
+    {
+      "epoch": 0.28010825439783493,
+      "grad_norm": 0.6488917571249994,
+      "learning_rate": 9.951522806921922e-06,
+      "loss": 0.3205,
+      "step": 414
+    },
+    {
+      "epoch": 0.2807848443843031,
+      "grad_norm": 0.6603105222613779,
+      "learning_rate": 9.95069903667256e-06,
+      "loss": 0.3224,
+      "step": 415
+    },
+    {
+      "epoch": 0.2814614343707713,
+      "grad_norm": 0.7691701702118307,
+      "learning_rate": 9.949868360798893e-06,
+      "loss": 0.3143,
+      "step": 416
+    },
+    {
+      "epoch": 0.2821380243572395,
+      "grad_norm": 0.7226424225320364,
+      "learning_rate": 9.949030780459615e-06,
+      "loss": 0.3293,
+      "step": 417
+    },
+    {
+      "epoch": 0.2828146143437077,
+      "grad_norm": 0.6492183077783911,
+      "learning_rate": 9.948186296823048e-06,
+      "loss": 0.3072,
+      "step": 418
+    },
+    {
+      "epoch": 0.28349120433017594,
+      "grad_norm": 0.8274401098788938,
+      "learning_rate": 9.947334911067144e-06,
+      "loss": 0.3074,
+      "step": 419
+    },
+    {
+      "epoch": 0.28416779431664413,
+      "grad_norm": 0.6777348774997134,
+      "learning_rate": 9.946476624379485e-06,
+      "loss": 0.3092,
+      "step": 420
+    },
+    {
+      "epoch": 0.2848443843031123,
+      "grad_norm": 0.7412658130282925,
+      "learning_rate": 9.945611437957274e-06,
+      "loss": 0.3261,
+      "step": 421
+    },
+    {
+      "epoch": 0.2855209742895805,
+      "grad_norm": 0.6892513805623189,
+      "learning_rate": 9.944739353007344e-06,
+      "loss": 0.3138,
+      "step": 422
+    },
+    {
+      "epoch": 0.2861975642760487,
+      "grad_norm": 0.6685193371942978,
+      "learning_rate": 9.943860370746144e-06,
+      "loss": 0.3166,
+      "step": 423
+    },
+    {
+      "epoch": 0.2868741542625169,
+      "grad_norm": 0.7066798755292086,
+      "learning_rate": 9.942974492399751e-06,
+      "loss": 0.3198,
+      "step": 424
+    },
+    {
+      "epoch": 0.28755074424898514,
+      "grad_norm": 0.640811479423961,
+      "learning_rate": 9.942081719203855e-06,
+      "loss": 0.3084,
+      "step": 425
+    },
+    {
+      "epoch": 0.2882273342354533,
+      "grad_norm": 0.6755129067822345,
+      "learning_rate": 9.941182052403768e-06,
+      "loss": 0.3035,
+      "step": 426
+    },
+    {
+      "epoch": 0.2889039242219215,
+      "grad_norm": 0.6939772201149468,
+      "learning_rate": 9.940275493254416e-06,
+      "loss": 0.3175,
+      "step": 427
+    },
+    {
+      "epoch": 0.2895805142083897,
+      "grad_norm": 0.6428680576137114,
+      "learning_rate": 9.93936204302034e-06,
+      "loss": 0.3142,
+      "step": 428
+    },
+    {
+      "epoch": 0.2902571041948579,
+      "grad_norm": 0.713877964705485,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.3166,
+      "step": 429
+    },
+    {
+      "epoch": 0.29093369418132614,
+      "grad_norm": 0.6832619265290361,
+      "learning_rate": 9.937514474404229e-06,
+      "loss": 0.3223,
+      "step": 430
+    },
+    {
+      "epoch": 0.29161028416779433,
+      "grad_norm": 0.6444180002860292,
+      "learning_rate": 9.936580358599327e-06,
+      "loss": 0.3087,
+      "step": 431
+    },
+    {
+      "epoch": 0.2922868741542625,
+      "grad_norm": 0.63103813405727,
+      "learning_rate": 9.935639356863966e-06,
+      "loss": 0.3164,
+      "step": 432
+    },
+    {
+      "epoch": 0.2929634641407307,
+      "grad_norm": 0.6665414453574531,
+      "learning_rate": 9.934691470510728e-06,
+      "loss": 0.3296,
+      "step": 433
+    },
+    {
+      "epoch": 0.2936400541271989,
+      "grad_norm": 0.699950846560052,
+      "learning_rate": 9.933736700861798e-06,
+      "loss": 0.3233,
+      "step": 434
+    },
+    {
+      "epoch": 0.2943166441136671,
+      "grad_norm": 0.6438949495537493,
+      "learning_rate": 9.932775049248965e-06,
+      "loss": 0.3189,
+      "step": 435
+    },
+    {
+      "epoch": 0.29499323410013534,
+      "grad_norm": 0.6732695729186846,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.3076,
+      "step": 436
+    },
+    {
+      "epoch": 0.2956698240866035,
+      "grad_norm": 0.6742980459445292,
+      "learning_rate": 9.93083110550673e-06,
+      "loss": 0.3114,
+      "step": 437
+    },
+    {
+      "epoch": 0.2963464140730717,
+      "grad_norm": 0.6432210697057751,
+      "learning_rate": 9.929848816088898e-06,
+      "loss": 0.3196,
+      "step": 438
+    },
+    {
+      "epoch": 0.2970230040595399,
+      "grad_norm": 0.6607136716698079,
+      "learning_rate": 9.92885965013029e-06,
+      "loss": 0.3169,
+      "step": 439
+    },
+    {
+      "epoch": 0.2976995940460081,
+      "grad_norm": 0.6468156714868835,
+      "learning_rate": 9.92786360901067e-06,
+      "loss": 0.3196,
+      "step": 440
+    },
+    {
+      "epoch": 0.29837618403247634,
+      "grad_norm": 0.6502510319302798,
+      "learning_rate": 9.9268606941194e-06,
+      "loss": 0.3176,
+      "step": 441
+    },
+    {
+      "epoch": 0.29905277401894453,
+      "grad_norm": 0.6675648325789031,
+      "learning_rate": 9.925850906855419e-06,
+      "loss": 0.3176,
+      "step": 442
+    },
+    {
+      "epoch": 0.2997293640054127,
+      "grad_norm": 0.6606113899406185,
+      "learning_rate": 9.92483424862726e-06,
+      "loss": 0.3063,
+      "step": 443
+    },
+    {
+      "epoch": 0.3004059539918809,
+      "grad_norm": 0.6598257928008405,
+      "learning_rate": 9.923810720853038e-06,
+      "loss": 0.3125,
+      "step": 444
+    },
+    {
+      "epoch": 0.3010825439783491,
+      "grad_norm": 0.6241620066060988,
+      "learning_rate": 9.92278032496045e-06,
+      "loss": 0.3175,
+      "step": 445
+    },
+    {
+      "epoch": 0.3017591339648173,
+      "grad_norm": 0.6587577554390699,
+      "learning_rate": 9.921743062386773e-06,
+      "loss": 0.3202,
+      "step": 446
+    },
+    {
+      "epoch": 0.30243572395128554,
+      "grad_norm": 0.6461322892612362,
+      "learning_rate": 9.920698934578862e-06,
+      "loss": 0.3218,
+      "step": 447
+    },
+    {
+      "epoch": 0.30311231393775373,
+      "grad_norm": 0.6567905944143188,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.3265,
+      "step": 448
+    },
+    {
+      "epoch": 0.3037889039242219,
+      "grad_norm": 0.6734684840026418,
+      "learning_rate": 9.91859008909564e-06,
+      "loss": 0.3103,
+      "step": 449
+    },
+    {
+      "epoch": 0.3044654939106901,
+      "grad_norm": 0.6501469770602251,
+      "learning_rate": 9.917525374361913e-06,
+      "loss": 0.3228,
+      "step": 450
+    },
+    {
+      "epoch": 0.3051420838971583,
+      "grad_norm": 0.6342524924935761,
+      "learning_rate": 9.916453800277115e-06,
+      "loss": 0.3106,
+      "step": 451
+    },
+    {
+      "epoch": 0.30581867388362655,
+      "grad_norm": 0.6539923892015517,
+      "learning_rate": 9.915375368335962e-06,
+      "loss": 0.3113,
+      "step": 452
+    },
+    {
+      "epoch": 0.30649526387009474,
+      "grad_norm": 0.6490882101473643,
+      "learning_rate": 9.914290080042735e-06,
+      "loss": 0.3135,
+      "step": 453
+    },
+    {
+      "epoch": 0.3071718538565629,
+      "grad_norm": 0.6676368896210113,
+      "learning_rate": 9.91319793691128e-06,
+      "loss": 0.3108,
+      "step": 454
+    },
+    {
+      "epoch": 0.3078484438430311,
+      "grad_norm": 0.6186900116530775,
+      "learning_rate": 9.912098940465007e-06,
+      "loss": 0.3046,
+      "step": 455
+    },
+    {
+      "epoch": 0.3085250338294993,
+      "grad_norm": 0.6736008238581569,
+      "learning_rate": 9.910993092236878e-06,
+      "loss": 0.3226,
+      "step": 456
+    },
+    {
+      "epoch": 0.3092016238159675,
+      "grad_norm": 0.6717238942800557,
+      "learning_rate": 9.90988039376942e-06,
+      "loss": 0.3159,
+      "step": 457
+    },
+    {
+      "epoch": 0.30987821380243574,
+      "grad_norm": 0.6352848151589852,
+      "learning_rate": 9.90876084661471e-06,
+      "loss": 0.3156,
+      "step": 458
+    },
+    {
+      "epoch": 0.31055480378890393,
+      "grad_norm": 0.6812280411751457,
+      "learning_rate": 9.907634452334382e-06,
+      "loss": 0.3138,
+      "step": 459
+    },
+    {
+      "epoch": 0.3112313937753721,
+      "grad_norm": 0.6833702960798785,
+      "learning_rate": 9.90650121249962e-06,
+      "loss": 0.304,
+      "step": 460
+    },
+    {
+      "epoch": 0.3119079837618403,
+      "grad_norm": 0.6561957038192524,
+      "learning_rate": 9.905361128691156e-06,
+      "loss": 0.3038,
+      "step": 461
+    },
+    {
+      "epoch": 0.3125845737483085,
+      "grad_norm": 0.6596600009882967,
+      "learning_rate": 9.904214202499266e-06,
+      "loss": 0.3198,
+      "step": 462
+    },
+    {
+      "epoch": 0.31326116373477675,
+      "grad_norm": 0.6257171151002868,
+      "learning_rate": 9.903060435523775e-06,
+      "loss": 0.3108,
+      "step": 463
+    },
+    {
+      "epoch": 0.31393775372124494,
+      "grad_norm": 0.6216290252208717,
+      "learning_rate": 9.901899829374048e-06,
+      "loss": 0.3152,
+      "step": 464
+    },
+    {
+      "epoch": 0.31461434370771313,
+      "grad_norm": 0.6246865629632653,
+      "learning_rate": 9.900732385668989e-06,
+      "loss": 0.3114,
+      "step": 465
+    },
+    {
+      "epoch": 0.3152909336941813,
+      "grad_norm": 0.6942870668356571,
+      "learning_rate": 9.899558106037039e-06,
+      "loss": 0.3097,
+      "step": 466
+    },
+    {
+      "epoch": 0.3159675236806495,
+      "grad_norm": 0.5867768565703964,
+      "learning_rate": 9.898376992116179e-06,
+      "loss": 0.3086,
+      "step": 467
+    },
+    {
+      "epoch": 0.3166441136671177,
+      "grad_norm": 0.6542854443440924,
+      "learning_rate": 9.897189045553917e-06,
+      "loss": 0.3255,
+      "step": 468
+    },
+    {
+      "epoch": 0.31732070365358594,
+      "grad_norm": 0.7366122755751578,
+      "learning_rate": 9.895994268007294e-06,
+      "loss": 0.3078,
+      "step": 469
+    },
+    {
+      "epoch": 0.31799729364005414,
+      "grad_norm": 0.6402704595240659,
+      "learning_rate": 9.89479266114288e-06,
+      "loss": 0.3149,
+      "step": 470
+    },
+    {
+      "epoch": 0.3186738836265223,
+      "grad_norm": 0.6319421336346136,
+      "learning_rate": 9.893584226636773e-06,
+      "loss": 0.3033,
+      "step": 471
+    },
+    {
+      "epoch": 0.3193504736129905,
+      "grad_norm": 0.6774141803903538,
+      "learning_rate": 9.89236896617459e-06,
+      "loss": 0.3193,
+      "step": 472
+    },
+    {
+      "epoch": 0.3200270635994587,
+      "grad_norm": 0.6486621280180213,
+      "learning_rate": 9.891146881451472e-06,
+      "loss": 0.3107,
+      "step": 473
+    },
+    {
+      "epoch": 0.32070365358592695,
+      "grad_norm": 0.6434572731190457,
+      "learning_rate": 9.88991797417208e-06,
+      "loss": 0.3157,
+      "step": 474
+    },
+    {
+      "epoch": 0.32138024357239514,
+      "grad_norm": 0.6772837960063797,
+      "learning_rate": 9.88868224605059e-06,
+      "loss": 0.3081,
+      "step": 475
+    },
+    {
+      "epoch": 0.32205683355886333,
+      "grad_norm": 0.6291887884357894,
+      "learning_rate": 9.887439698810694e-06,
+      "loss": 0.3115,
+      "step": 476
+    },
+    {
+      "epoch": 0.3227334235453315,
+      "grad_norm": 0.659242104067569,
+      "learning_rate": 9.886190334185592e-06,
+      "loss": 0.3161,
+      "step": 477
+    },
+    {
+      "epoch": 0.3234100135317997,
+      "grad_norm": 0.6823443141262217,
+      "learning_rate": 9.884934153917998e-06,
+      "loss": 0.309,
+      "step": 478
+    },
+    {
+      "epoch": 0.32408660351826796,
+      "grad_norm": 0.6157891747237658,
+      "learning_rate": 9.883671159760128e-06,
+      "loss": 0.3156,
+      "step": 479
+    },
+    {
+      "epoch": 0.32476319350473615,
+      "grad_norm": 0.6508924177734028,
+      "learning_rate": 9.882401353473711e-06,
+      "loss": 0.3356,
+      "step": 480
+    },
+    {
+      "epoch": 0.32543978349120434,
+      "grad_norm": 0.6797198524360808,
+      "learning_rate": 9.881124736829968e-06,
+      "loss": 0.3109,
+      "step": 481
+    },
+    {
+      "epoch": 0.3261163734776725,
+      "grad_norm": 0.6251355373308037,
+      "learning_rate": 9.879841311609625e-06,
+      "loss": 0.3016,
+      "step": 482
+    },
+    {
+      "epoch": 0.3267929634641407,
+      "grad_norm": 0.6288682162807113,
+      "learning_rate": 9.878551079602908e-06,
+      "loss": 0.3162,
+      "step": 483
+    },
+    {
+      "epoch": 0.3274695534506089,
+      "grad_norm": 0.6252111749955878,
+      "learning_rate": 9.877254042609529e-06,
+      "loss": 0.3019,
+      "step": 484
+    },
+    {
+      "epoch": 0.32814614343707715,
+      "grad_norm": 0.6444375743267886,
+      "learning_rate": 9.8759502024387e-06,
+      "loss": 0.3115,
+      "step": 485
+    },
+    {
+      "epoch": 0.32882273342354534,
+      "grad_norm": 0.6183799521688081,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.3098,
+      "step": 486
+    },
+    {
+      "epoch": 0.32949932341001353,
+      "grad_norm": 0.6421633666735025,
+      "learning_rate": 9.873322119848973e-06,
+      "loss": 0.3156,
+      "step": 487
+    },
+    {
+      "epoch": 0.3301759133964817,
+      "grad_norm": 0.6346251243688272,
+      "learning_rate": 9.87199788109593e-06,
+      "loss": 0.3076,
+      "step": 488
+    },
+    {
+      "epoch": 0.3308525033829499,
+      "grad_norm": 0.6371289349499826,
+      "learning_rate": 9.870666846497145e-06,
+      "loss": 0.314,
+      "step": 489
+    },
+    {
+      "epoch": 0.33152909336941816,
+      "grad_norm": 0.6181985542947318,
+      "learning_rate": 9.869329017909248e-06,
+      "loss": 0.3094,
+      "step": 490
+    },
+    {
+      "epoch": 0.33220568335588635,
+      "grad_norm": 0.6884667078329131,
+      "learning_rate": 9.867984397198349e-06,
+      "loss": 0.316,
+      "step": 491
+    },
+    {
+      "epoch": 0.33288227334235454,
+      "grad_norm": 0.6022162541045866,
+      "learning_rate": 9.86663298624003e-06,
+      "loss": 0.3099,
+      "step": 492
+    },
+    {
+      "epoch": 0.33355886332882273,
+      "grad_norm": 0.6370187752401233,
+      "learning_rate": 9.865274786919345e-06,
+      "loss": 0.3197,
+      "step": 493
+    },
+    {
+      "epoch": 0.3342354533152909,
+      "grad_norm": 0.7006203414556026,
+      "learning_rate": 9.863909801130816e-06,
+      "loss": 0.3212,
+      "step": 494
+    },
+    {
+      "epoch": 0.3349120433017591,
+      "grad_norm": 0.6337150332215001,
+      "learning_rate": 9.862538030778434e-06,
+      "loss": 0.3111,
+      "step": 495
+    },
+    {
+      "epoch": 0.33558863328822736,
+      "grad_norm": 0.657219345154801,
+      "learning_rate": 9.861159477775653e-06,
+      "loss": 0.3148,
+      "step": 496
+    },
+    {
+      "epoch": 0.33626522327469555,
+      "grad_norm": 0.6751056842805132,
+      "learning_rate": 9.859774144045387e-06,
+      "loss": 0.3074,
+      "step": 497
+    },
+    {
+      "epoch": 0.33694181326116374,
+      "grad_norm": 0.658712543289006,
+      "learning_rate": 9.858382031520005e-06,
+      "loss": 0.3183,
+      "step": 498
+    },
+    {
+      "epoch": 0.3376184032476319,
+      "grad_norm": 0.6451305989963522,
+      "learning_rate": 9.856983142141338e-06,
+      "loss": 0.3083,
+      "step": 499
+    },
+    {
+      "epoch": 0.3382949932341001,
+      "grad_norm": 0.6759282211832467,
+      "learning_rate": 9.855577477860669e-06,
+      "loss": 0.3123,
+      "step": 500
+    },
+    {
+      "epoch": 0.33897158322056836,
+      "grad_norm": 0.6869600248493148,
+      "learning_rate": 9.854165040638724e-06,
+      "loss": 0.3155,
+      "step": 501
+    },
+    {
+      "epoch": 0.33964817320703655,
+      "grad_norm": 0.6600973677103916,
+      "learning_rate": 9.852745832445684e-06,
+      "loss": 0.3111,
+      "step": 502
+    },
+    {
+      "epoch": 0.34032476319350474,
+      "grad_norm": 0.6166334793569577,
+      "learning_rate": 9.851319855261173e-06,
+      "loss": 0.3052,
+      "step": 503
+    },
+    {
+      "epoch": 0.34100135317997293,
+      "grad_norm": 0.5852079040600452,
+      "learning_rate": 9.849887111074256e-06,
+      "loss": 0.3159,
+      "step": 504
+    },
+    {
+      "epoch": 0.3416779431664411,
+      "grad_norm": 0.7000454930815192,
+      "learning_rate": 9.848447601883436e-06,
+      "loss": 0.32,
+      "step": 505
+    },
+    {
+      "epoch": 0.3423545331529093,
+      "grad_norm": 0.6125113286170768,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 0.3003,
+      "step": 506
+    },
+    {
+      "epoch": 0.34303112313937756,
+      "grad_norm": 0.5948356370447204,
+      "learning_rate": 9.845548296531283e-06,
+      "loss": 0.3028,
+      "step": 507
+    },
+    {
+      "epoch": 0.34370771312584575,
+      "grad_norm": 0.6210089438469427,
+      "learning_rate": 9.84408850441413e-06,
+      "loss": 0.3162,
+      "step": 508
+    },
+    {
+      "epoch": 0.34438430311231394,
+      "grad_norm": 0.6625234157100467,
+      "learning_rate": 9.842621955381428e-06,
+      "loss": 0.3173,
+      "step": 509
+    },
+    {
+      "epoch": 0.34506089309878213,
+      "grad_norm": 0.6363586626570846,
+      "learning_rate": 9.841148651478833e-06,
+      "loss": 0.311,
+      "step": 510
+    },
+    {
+      "epoch": 0.3457374830852503,
+      "grad_norm": 0.6241785640667638,
+      "learning_rate": 9.839668594761427e-06,
+      "loss": 0.3128,
+      "step": 511
+    },
+    {
+      "epoch": 0.34641407307171856,
+      "grad_norm": 0.6679986038833047,
+      "learning_rate": 9.838181787293707e-06,
+      "loss": 0.3178,
+      "step": 512
+    },
+    {
+      "epoch": 0.34709066305818675,
+      "grad_norm": 0.5868086805741166,
+      "learning_rate": 9.836688231149593e-06,
+      "loss": 0.3063,
+      "step": 513
+    },
+    {
+      "epoch": 0.34776725304465494,
+      "grad_norm": 0.6027079466193433,
+      "learning_rate": 9.835187928412412e-06,
+      "loss": 0.3072,
+      "step": 514
+    },
+    {
+      "epoch": 0.34844384303112313,
+      "grad_norm": 0.6566033096204844,
+      "learning_rate": 9.833680881174901e-06,
+      "loss": 0.3137,
+      "step": 515
+    },
+    {
+      "epoch": 0.3491204330175913,
+      "grad_norm": 0.6044115515089298,
+      "learning_rate": 9.832167091539215e-06,
+      "loss": 0.3069,
+      "step": 516
+    },
+    {
+      "epoch": 0.3497970230040595,
+      "grad_norm": 0.6898301597033742,
+      "learning_rate": 9.830646561616901e-06,
+      "loss": 0.3218,
+      "step": 517
+    },
+    {
+      "epoch": 0.35047361299052776,
+      "grad_norm": 0.6051340764756622,
+      "learning_rate": 9.829119293528916e-06,
+      "loss": 0.318,
+      "step": 518
+    },
+    {
+      "epoch": 0.35115020297699595,
+      "grad_norm": 0.6268521261157523,
+      "learning_rate": 9.827585289405614e-06,
+      "loss": 0.3229,
+      "step": 519
+    },
+    {
+      "epoch": 0.35182679296346414,
+      "grad_norm": 0.6778669902931774,
+      "learning_rate": 9.826044551386743e-06,
+      "loss": 0.3181,
+      "step": 520
+    },
+    {
+      "epoch": 0.35250338294993233,
+      "grad_norm": 0.6633381674463105,
+      "learning_rate": 9.824497081621449e-06,
+      "loss": 0.3132,
+      "step": 521
+    },
+    {
+      "epoch": 0.3531799729364005,
+      "grad_norm": 0.6225136554147355,
+      "learning_rate": 9.822942882268261e-06,
+      "loss": 0.298,
+      "step": 522
+    },
+    {
+      "epoch": 0.35385656292286877,
+      "grad_norm": 0.6549761249556743,
+      "learning_rate": 9.8213819554951e-06,
+      "loss": 0.3105,
+      "step": 523
+    },
+    {
+      "epoch": 0.35453315290933696,
+      "grad_norm": 0.6519050244250838,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.3075,
+      "step": 524
+    },
+    {
+      "epoch": 0.35520974289580515,
+      "grad_norm": 0.6195346991931288,
+      "learning_rate": 9.818239928407449e-06,
+      "loss": 0.3123,
+      "step": 525
+    },
+    {
+      "epoch": 0.35588633288227334,
+      "grad_norm": 0.6658331498522633,
+      "learning_rate": 9.816658832475709e-06,
+      "loss": 0.3128,
+      "step": 526
+    },
+    {
+      "epoch": 0.3565629228687415,
+      "grad_norm": 0.6862245623917979,
+      "learning_rate": 9.81507101788948e-06,
+      "loss": 0.3165,
+      "step": 527
+    },
+    {
+      "epoch": 0.3572395128552097,
+      "grad_norm": 0.6039080662100713,
+      "learning_rate": 9.813476486863575e-06,
+      "loss": 0.3139,
+      "step": 528
+    },
+    {
+      "epoch": 0.35791610284167796,
+      "grad_norm": 0.6674175212607096,
+      "learning_rate": 9.811875241622169e-06,
+      "loss": 0.3227,
+      "step": 529
+    },
+    {
+      "epoch": 0.35859269282814615,
+      "grad_norm": 0.6407346548266009,
+      "learning_rate": 9.810267284398805e-06,
+      "loss": 0.3093,
+      "step": 530
+    },
+    {
+      "epoch": 0.35926928281461434,
+      "grad_norm": 0.6418881407130957,
+      "learning_rate": 9.808652617436386e-06,
+      "loss": 0.3124,
+      "step": 531
+    },
+    {
+      "epoch": 0.35994587280108253,
+      "grad_norm": 0.6380660624596824,
+      "learning_rate": 9.807031242987182e-06,
+      "loss": 0.3127,
+      "step": 532
+    },
+    {
+      "epoch": 0.3606224627875507,
+      "grad_norm": 0.6904681195957602,
+      "learning_rate": 9.805403163312808e-06,
+      "loss": 0.3121,
+      "step": 533
+    },
+    {
+      "epoch": 0.36129905277401897,
+      "grad_norm": 0.6136811585154887,
+      "learning_rate": 9.803768380684242e-06,
+      "loss": 0.2907,
+      "step": 534
+    },
+    {
+      "epoch": 0.36197564276048716,
+      "grad_norm": 0.6242675944732607,
+      "learning_rate": 9.802126897381804e-06,
+      "loss": 0.32,
+      "step": 535
+    },
+    {
+      "epoch": 0.36265223274695535,
+      "grad_norm": 0.6496421010453656,
+      "learning_rate": 9.800478715695165e-06,
+      "loss": 0.3214,
+      "step": 536
+    },
+    {
+      "epoch": 0.36332882273342354,
+      "grad_norm": 0.6512157323003713,
+      "learning_rate": 9.798823837923336e-06,
+      "loss": 0.3054,
+      "step": 537
+    },
+    {
+      "epoch": 0.36400541271989173,
+      "grad_norm": 0.6468424645456742,
+      "learning_rate": 9.797162266374677e-06,
+      "loss": 0.3009,
+      "step": 538
+    },
+    {
+      "epoch": 0.3646820027063599,
+      "grad_norm": 0.5971815068996954,
+      "learning_rate": 9.795494003366873e-06,
+      "loss": 0.3046,
+      "step": 539
+    },
+    {
+      "epoch": 0.36535859269282817,
+      "grad_norm": 0.6434777914383665,
+      "learning_rate": 9.79381905122695e-06,
+      "loss": 0.3139,
+      "step": 540
+    },
+    {
+      "epoch": 0.36603518267929636,
+      "grad_norm": 0.6038755901007039,
+      "learning_rate": 9.792137412291265e-06,
+      "loss": 0.3037,
+      "step": 541
+    },
+    {
+      "epoch": 0.36671177266576455,
+      "grad_norm": 0.6220840387622848,
+      "learning_rate": 9.790449088905496e-06,
+      "loss": 0.3143,
+      "step": 542
+    },
+    {
+      "epoch": 0.36738836265223274,
+      "grad_norm": 0.5900814016961722,
+      "learning_rate": 9.788754083424654e-06,
+      "loss": 0.3085,
+      "step": 543
+    },
+    {
+      "epoch": 0.3680649526387009,
+      "grad_norm": 0.6083537877358284,
+      "learning_rate": 9.787052398213062e-06,
+      "loss": 0.2986,
+      "step": 544
+    },
+    {
+      "epoch": 0.36874154262516917,
+      "grad_norm": 0.6269891617882138,
+      "learning_rate": 9.785344035644366e-06,
+      "loss": 0.2939,
+      "step": 545
+    },
+    {
+      "epoch": 0.36941813261163736,
+      "grad_norm": 0.6052773251121962,
+      "learning_rate": 9.783628998101525e-06,
+      "loss": 0.3151,
+      "step": 546
+    },
+    {
+      "epoch": 0.37009472259810555,
+      "grad_norm": 0.5968545213184993,
+      "learning_rate": 9.781907287976807e-06,
+      "loss": 0.3052,
+      "step": 547
+    },
+    {
+      "epoch": 0.37077131258457374,
+      "grad_norm": 0.615927841972164,
+      "learning_rate": 9.780178907671788e-06,
+      "loss": 0.3044,
+      "step": 548
+    },
+    {
+      "epoch": 0.37144790257104193,
+      "grad_norm": 0.6310503123256876,
+      "learning_rate": 9.778443859597353e-06,
+      "loss": 0.31,
+      "step": 549
+    },
+    {
+      "epoch": 0.3721244925575101,
+      "grad_norm": 0.6344520395801881,
+      "learning_rate": 9.776702146173678e-06,
+      "loss": 0.3147,
+      "step": 550
+    },
+    {
+      "epoch": 0.37280108254397837,
+      "grad_norm": 0.6084496528543195,
+      "learning_rate": 9.774953769830245e-06,
+      "loss": 0.3166,
+      "step": 551
+    },
+    {
+      "epoch": 0.37347767253044656,
+      "grad_norm": 0.6273051288181115,
+      "learning_rate": 9.773198733005827e-06,
+      "loss": 0.3095,
+      "step": 552
+    },
+    {
+      "epoch": 0.37415426251691475,
+      "grad_norm": 0.6340854691366966,
+      "learning_rate": 9.771437038148486e-06,
+      "loss": 0.3126,
+      "step": 553
+    },
+    {
+      "epoch": 0.37483085250338294,
+      "grad_norm": 0.655940210504197,
+      "learning_rate": 9.769668687715572e-06,
+      "loss": 0.3044,
+      "step": 554
+    },
+    {
+      "epoch": 0.37550744248985113,
+      "grad_norm": 0.6352877518670023,
+      "learning_rate": 9.767893684173722e-06,
+      "loss": 0.3066,
+      "step": 555
+    },
+    {
+      "epoch": 0.3761840324763194,
+      "grad_norm": 0.6129833591357998,
+      "learning_rate": 9.766112029998847e-06,
+      "loss": 0.3089,
+      "step": 556
+    },
+    {
+      "epoch": 0.37686062246278756,
+      "grad_norm": 0.6493120407527091,
+      "learning_rate": 9.764323727676142e-06,
+      "loss": 0.313,
+      "step": 557
+    },
+    {
+      "epoch": 0.37753721244925575,
+      "grad_norm": 0.632749899157854,
+      "learning_rate": 9.762528779700067e-06,
+      "loss": 0.2954,
+      "step": 558
+    },
+    {
+      "epoch": 0.37821380243572394,
+      "grad_norm": 0.6127040488316374,
+      "learning_rate": 9.76072718857436e-06,
+      "loss": 0.3096,
+      "step": 559
+    },
+    {
+      "epoch": 0.37889039242219213,
+      "grad_norm": 0.6752081465461858,
+      "learning_rate": 9.758918956812024e-06,
+      "loss": 0.3127,
+      "step": 560
+    },
+    {
+      "epoch": 0.3795669824086603,
+      "grad_norm": 0.6474832902819029,
+      "learning_rate": 9.757104086935319e-06,
+      "loss": 0.2991,
+      "step": 561
+    },
+    {
+      "epoch": 0.38024357239512857,
+      "grad_norm": 0.6456328232880437,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.3078,
+      "step": 562
+    },
+    {
+      "epoch": 0.38092016238159676,
+      "grad_norm": 0.6707478091603916,
+      "learning_rate": 9.753454442974154e-06,
+      "loss": 0.3106,
+      "step": 563
+    },
+    {
+      "epoch": 0.38159675236806495,
+      "grad_norm": 0.6367273623754901,
+      "learning_rate": 9.751619673980503e-06,
+      "loss": 0.3046,
+      "step": 564
+    },
+    {
+      "epoch": 0.38227334235453314,
+      "grad_norm": 0.6363032300839486,
+      "learning_rate": 9.7497782770541e-06,
+      "loss": 0.3069,
+      "step": 565
+    },
+    {
+      "epoch": 0.38294993234100133,
+      "grad_norm": 0.6547507146290692,
+      "learning_rate": 9.747930254763467e-06,
+      "loss": 0.3141,
+      "step": 566
+    },
+    {
+      "epoch": 0.3836265223274696,
+      "grad_norm": 0.6342331997018092,
+      "learning_rate": 9.746075609686372e-06,
+      "loss": 0.2985,
+      "step": 567
+    },
+    {
+      "epoch": 0.38430311231393777,
+      "grad_norm": 0.6520669995141555,
+      "learning_rate": 9.74421434440982e-06,
+      "loss": 0.2994,
+      "step": 568
+    },
+    {
+      "epoch": 0.38497970230040596,
+      "grad_norm": 0.6472644577510097,
+      "learning_rate": 9.742346461530048e-06,
+      "loss": 0.3092,
+      "step": 569
+    },
+    {
+      "epoch": 0.38565629228687415,
+      "grad_norm": 0.7071062226125303,
+      "learning_rate": 9.74047196365253e-06,
+      "loss": 0.3109,
+      "step": 570
+    },
+    {
+      "epoch": 0.38633288227334234,
+      "grad_norm": 0.6430810999891908,
+      "learning_rate": 9.738590853391959e-06,
+      "loss": 0.3118,
+      "step": 571
+    },
+    {
+      "epoch": 0.3870094722598105,
+      "grad_norm": 0.6815208944881339,
+      "learning_rate": 9.736703133372259e-06,
+      "loss": 0.3136,
+      "step": 572
+    },
+    {
+      "epoch": 0.3876860622462788,
+      "grad_norm": 0.6186013286076679,
+      "learning_rate": 9.734808806226568e-06,
+      "loss": 0.3238,
+      "step": 573
+    },
+    {
+      "epoch": 0.38836265223274696,
+      "grad_norm": 0.6725503014051902,
+      "learning_rate": 9.732907874597241e-06,
+      "loss": 0.305,
+      "step": 574
+    },
+    {
+      "epoch": 0.38903924221921515,
+      "grad_norm": 0.6099574781862094,
+      "learning_rate": 9.731000341135852e-06,
+      "loss": 0.3099,
+      "step": 575
+    },
+    {
+      "epoch": 0.38971583220568334,
+      "grad_norm": 0.647874206563038,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.3187,
+      "step": 576
+    },
+    {
+      "epoch": 0.39039242219215153,
+      "grad_norm": 0.6008367006435882,
+      "learning_rate": 9.727165479369192e-06,
+      "loss": 0.3006,
+      "step": 577
+    },
+    {
+      "epoch": 0.3910690121786198,
+      "grad_norm": 0.5893239853616612,
+      "learning_rate": 9.725238156413089e-06,
+      "loss": 0.3032,
+      "step": 578
+    },
+    {
+      "epoch": 0.39174560216508797,
+      "grad_norm": 0.6139538557664592,
+      "learning_rate": 9.723304242323248e-06,
+      "loss": 0.3045,
+      "step": 579
+    },
+    {
+      "epoch": 0.39242219215155616,
+      "grad_norm": 0.6573402513622209,
+      "learning_rate": 9.721363739797243e-06,
+      "loss": 0.3072,
+      "step": 580
+    },
+    {
+      "epoch": 0.39309878213802435,
+      "grad_norm": 0.58189515660078,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.3045,
+      "step": 581
+    },
+    {
+      "epoch": 0.39377537212449254,
+      "grad_norm": 0.6400910133347768,
+      "learning_rate": 9.717462980272989e-06,
+      "loss": 0.305,
+      "step": 582
+    },
+    {
+      "epoch": 0.3944519621109608,
+      "grad_norm": 0.5905776069865982,
+      "learning_rate": 9.715502728715827e-06,
+      "loss": 0.315,
+      "step": 583
+    },
+    {
+      "epoch": 0.395128552097429,
+      "grad_norm": 0.6259305498122035,
+      "learning_rate": 9.713535899604667e-06,
+      "loss": 0.307,
+      "step": 584
+    },
+    {
+      "epoch": 0.39580514208389717,
+      "grad_norm": 0.5837036854769092,
+      "learning_rate": 9.711562495682996e-06,
+      "loss": 0.3172,
+      "step": 585
+    },
+    {
+      "epoch": 0.39648173207036536,
+      "grad_norm": 0.661981551118665,
+      "learning_rate": 9.70958251970347e-06,
+      "loss": 0.3059,
+      "step": 586
+    },
+    {
+      "epoch": 0.39715832205683355,
+      "grad_norm": 0.613989382268929,
+      "learning_rate": 9.707595974427922e-06,
+      "loss": 0.2997,
+      "step": 587
+    },
+    {
+      "epoch": 0.39783491204330174,
+      "grad_norm": 0.573742207269287,
+      "learning_rate": 9.705602862627335e-06,
+      "loss": 0.3006,
+      "step": 588
+    },
+    {
+      "epoch": 0.39851150202977,
+      "grad_norm": 0.6377589058562296,
+      "learning_rate": 9.703603187081863e-06,
+      "loss": 0.309,
+      "step": 589
+    },
+    {
+      "epoch": 0.39918809201623817,
+      "grad_norm": 0.6334997323645234,
+      "learning_rate": 9.701596950580807e-06,
+      "loss": 0.3003,
+      "step": 590
+    },
+    {
+      "epoch": 0.39986468200270636,
+      "grad_norm": 0.5992496437745587,
+      "learning_rate": 9.699584155922625e-06,
+      "loss": 0.3006,
+      "step": 591
+    },
+    {
+      "epoch": 0.40054127198917455,
+      "grad_norm": 0.603328702525179,
+      "learning_rate": 9.697564805914922e-06,
+      "loss": 0.2919,
+      "step": 592
+    },
+    {
+      "epoch": 0.40121786197564274,
+      "grad_norm": 0.5779454754309823,
+      "learning_rate": 9.695538903374444e-06,
+      "loss": 0.3019,
+      "step": 593
+    },
+    {
+      "epoch": 0.401894451962111,
+      "grad_norm": 0.5769909505530895,
+      "learning_rate": 9.693506451127082e-06,
+      "loss": 0.3182,
+      "step": 594
+    },
+    {
+      "epoch": 0.4025710419485792,
+      "grad_norm": 0.6123606624158388,
+      "learning_rate": 9.691467452007857e-06,
+      "loss": 0.311,
+      "step": 595
+    },
+    {
+      "epoch": 0.40324763193504737,
+      "grad_norm": 0.5848918642972003,
+      "learning_rate": 9.689421908860928e-06,
+      "loss": 0.3079,
+      "step": 596
+    },
+    {
+      "epoch": 0.40392422192151556,
+      "grad_norm": 0.6120924606651011,
+      "learning_rate": 9.687369824539577e-06,
+      "loss": 0.3051,
+      "step": 597
+    },
+    {
+      "epoch": 0.40460081190798375,
+      "grad_norm": 0.5630854672059997,
+      "learning_rate": 9.685311201906216e-06,
+      "loss": 0.3068,
+      "step": 598
+    },
+    {
+      "epoch": 0.40527740189445194,
+      "grad_norm": 0.5769061949026719,
+      "learning_rate": 9.683246043832371e-06,
+      "loss": 0.3048,
+      "step": 599
+    },
+    {
+      "epoch": 0.4059539918809202,
+      "grad_norm": 0.6402260128939908,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.3148,
+      "step": 600
+    },
+    {
+      "epoch": 0.4066305818673884,
+      "grad_norm": 0.6192990080910205,
+      "learning_rate": 9.679096132894922e-06,
+      "loss": 0.3154,
+      "step": 601
+    },
+    {
+      "epoch": 0.40730717185385656,
+      "grad_norm": 0.5806726309395249,
+      "learning_rate": 9.67701138581994e-06,
+      "loss": 0.3136,
+      "step": 602
+    },
+    {
+      "epoch": 0.40798376184032475,
+      "grad_norm": 0.6208589590957669,
+      "learning_rate": 9.674920114881712e-06,
+      "loss": 0.3169,
+      "step": 603
+    },
+    {
+      "epoch": 0.40866035182679294,
+      "grad_norm": 0.6070969921893409,
+      "learning_rate": 9.672822322997305e-06,
+      "loss": 0.3086,
+      "step": 604
+    },
+    {
+      "epoch": 0.4093369418132612,
+      "grad_norm": 0.6383939306805547,
+      "learning_rate": 9.670718013092886e-06,
+      "loss": 0.2964,
+      "step": 605
+    },
+    {
+      "epoch": 0.4100135317997294,
+      "grad_norm": 0.5942821448884972,
+      "learning_rate": 9.668607188103708e-06,
+      "loss": 0.3073,
+      "step": 606
+    },
+    {
+      "epoch": 0.41069012178619757,
+      "grad_norm": 0.5994955099984979,
+      "learning_rate": 9.666489850974123e-06,
+      "loss": 0.3078,
+      "step": 607
+    },
+    {
+      "epoch": 0.41136671177266576,
+      "grad_norm": 0.6375550172759893,
+      "learning_rate": 9.664366004657553e-06,
+      "loss": 0.3015,
+      "step": 608
+    },
+    {
+      "epoch": 0.41204330175913395,
+      "grad_norm": 0.6152297527709155,
+      "learning_rate": 9.662235652116507e-06,
+      "loss": 0.3023,
+      "step": 609
+    },
+    {
+      "epoch": 0.41271989174560214,
+      "grad_norm": 0.6115779493822586,
+      "learning_rate": 9.66009879632257e-06,
+      "loss": 0.3045,
+      "step": 610
+    },
+    {
+      "epoch": 0.4133964817320704,
+      "grad_norm": 0.5942035936323302,
+      "learning_rate": 9.657955440256396e-06,
+      "loss": 0.2904,
+      "step": 611
+    },
+    {
+      "epoch": 0.4140730717185386,
+      "grad_norm": 0.6313440210780771,
+      "learning_rate": 9.655805586907705e-06,
+      "loss": 0.3176,
+      "step": 612
+    },
+    {
+      "epoch": 0.41474966170500677,
+      "grad_norm": 0.6025134848734571,
+      "learning_rate": 9.653649239275284e-06,
+      "loss": 0.2934,
+      "step": 613
+    },
+    {
+      "epoch": 0.41542625169147496,
+      "grad_norm": 0.6062263616899043,
+      "learning_rate": 9.651486400366972e-06,
+      "loss": 0.3069,
+      "step": 614
+    },
+    {
+      "epoch": 0.41610284167794315,
+      "grad_norm": 0.6156611773769662,
+      "learning_rate": 9.649317073199673e-06,
+      "loss": 0.313,
+      "step": 615
+    },
+    {
+      "epoch": 0.4167794316644114,
+      "grad_norm": 0.6150008945539498,
+      "learning_rate": 9.64714126079933e-06,
+      "loss": 0.3032,
+      "step": 616
+    },
+    {
+      "epoch": 0.4174560216508796,
+      "grad_norm": 0.614535991516847,
+      "learning_rate": 9.644958966200939e-06,
+      "loss": 0.3033,
+      "step": 617
+    },
+    {
+      "epoch": 0.4181326116373478,
+      "grad_norm": 0.6276597531579805,
+      "learning_rate": 9.642770192448537e-06,
+      "loss": 0.3085,
+      "step": 618
+    },
+    {
+      "epoch": 0.41880920162381596,
+      "grad_norm": 0.5992263796408506,
+      "learning_rate": 9.640574942595195e-06,
+      "loss": 0.3042,
+      "step": 619
+    },
+    {
+      "epoch": 0.41948579161028415,
+      "grad_norm": 0.6404735555409361,
+      "learning_rate": 9.638373219703023e-06,
+      "loss": 0.2985,
+      "step": 620
+    },
+    {
+      "epoch": 0.42016238159675234,
+      "grad_norm": 0.6437145845667945,
+      "learning_rate": 9.636165026843155e-06,
+      "loss": 0.3032,
+      "step": 621
+    },
+    {
+      "epoch": 0.4208389715832206,
+      "grad_norm": 0.667044925072939,
+      "learning_rate": 9.633950367095758e-06,
+      "loss": 0.2927,
+      "step": 622
+    },
+    {
+      "epoch": 0.4215155615696888,
+      "grad_norm": 0.6148296208489038,
+      "learning_rate": 9.631729243550003e-06,
+      "loss": 0.2994,
+      "step": 623
+    },
+    {
+      "epoch": 0.42219215155615697,
+      "grad_norm": 0.5880583740825737,
+      "learning_rate": 9.629501659304096e-06,
+      "loss": 0.3108,
+      "step": 624
+    },
+    {
+      "epoch": 0.42286874154262516,
+      "grad_norm": 0.6328070993428794,
+      "learning_rate": 9.627267617465243e-06,
+      "loss": 0.3004,
+      "step": 625
+    },
+    {
+      "epoch": 0.42354533152909335,
+      "grad_norm": 0.636496170723556,
+      "learning_rate": 9.625027121149665e-06,
+      "loss": 0.2989,
+      "step": 626
+    },
+    {
+      "epoch": 0.4242219215155616,
+      "grad_norm": 0.5989739906097346,
+      "learning_rate": 9.622780173482577e-06,
+      "loss": 0.295,
+      "step": 627
+    },
+    {
+      "epoch": 0.4248985115020298,
+      "grad_norm": 0.6223715209420049,
+      "learning_rate": 9.620526777598202e-06,
+      "loss": 0.3182,
+      "step": 628
+    },
+    {
+      "epoch": 0.425575101488498,
+      "grad_norm": 0.6466302258065751,
+      "learning_rate": 9.618266936639752e-06,
+      "loss": 0.3086,
+      "step": 629
+    },
+    {
+      "epoch": 0.42625169147496617,
+      "grad_norm": 0.6400127959126503,
+      "learning_rate": 9.616000653759435e-06,
+      "loss": 0.2974,
+      "step": 630
+    },
+    {
+      "epoch": 0.42692828146143436,
+      "grad_norm": 0.5932643325133256,
+      "learning_rate": 9.613727932118435e-06,
+      "loss": 0.3111,
+      "step": 631
+    },
+    {
+      "epoch": 0.42760487144790255,
+      "grad_norm": 0.6018233942666171,
+      "learning_rate": 9.611448774886925e-06,
+      "loss": 0.2974,
+      "step": 632
+    },
+    {
+      "epoch": 0.4282814614343708,
+      "grad_norm": 0.5913481965331164,
+      "learning_rate": 9.609163185244052e-06,
+      "loss": 0.2951,
+      "step": 633
+    },
+    {
+      "epoch": 0.428958051420839,
+      "grad_norm": 0.5945891923550551,
+      "learning_rate": 9.606871166377939e-06,
+      "loss": 0.307,
+      "step": 634
+    },
+    {
+      "epoch": 0.42963464140730717,
+      "grad_norm": 0.6365656041598183,
+      "learning_rate": 9.604572721485673e-06,
+      "loss": 0.3109,
+      "step": 635
+    },
+    {
+      "epoch": 0.43031123139377536,
+      "grad_norm": 0.5747548348745365,
+      "learning_rate": 9.602267853773301e-06,
+      "loss": 0.3117,
+      "step": 636
+    },
+    {
+      "epoch": 0.43098782138024355,
+      "grad_norm": 0.6017169745543313,
+      "learning_rate": 9.59995656645584e-06,
+      "loss": 0.3068,
+      "step": 637
+    },
+    {
+      "epoch": 0.4316644113667118,
+      "grad_norm": 0.614303984443308,
+      "learning_rate": 9.597638862757255e-06,
+      "loss": 0.2957,
+      "step": 638
+    },
+    {
+      "epoch": 0.43234100135318,
+      "grad_norm": 0.6616252525266761,
+      "learning_rate": 9.595314745910455e-06,
+      "loss": 0.3066,
+      "step": 639
+    },
+    {
+      "epoch": 0.4330175913396482,
+      "grad_norm": 0.5738506341417521,
+      "learning_rate": 9.59298421915731e-06,
+      "loss": 0.3102,
+      "step": 640
+    },
+    {
+      "epoch": 0.43369418132611637,
+      "grad_norm": 0.6273385609700511,
+      "learning_rate": 9.590647285748614e-06,
+      "loss": 0.3226,
+      "step": 641
+    },
+    {
+      "epoch": 0.43437077131258456,
+      "grad_norm": 0.6714531835760587,
+      "learning_rate": 9.588303948944109e-06,
+      "loss": 0.3018,
+      "step": 642
+    },
+    {
+      "epoch": 0.43504736129905275,
+      "grad_norm": 0.6457313486565687,
+      "learning_rate": 9.585954212012465e-06,
+      "loss": 0.3034,
+      "step": 643
+    },
+    {
+      "epoch": 0.435723951285521,
+      "grad_norm": 0.6404823891763967,
+      "learning_rate": 9.58359807823128e-06,
+      "loss": 0.3131,
+      "step": 644
+    },
+    {
+      "epoch": 0.4364005412719892,
+      "grad_norm": 0.6181210015959896,
+      "learning_rate": 9.581235550887074e-06,
+      "loss": 0.3005,
+      "step": 645
+    },
+    {
+      "epoch": 0.4370771312584574,
+      "grad_norm": 0.6651691552956707,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 0.3018,
+      "step": 646
+    },
+    {
+      "epoch": 0.43775372124492556,
+      "grad_norm": 0.6060967258642173,
+      "learning_rate": 9.576491328700272e-06,
+      "loss": 0.3011,
+      "step": 647
+    },
+    {
+      "epoch": 0.43843031123139375,
+      "grad_norm": 0.6532876185696262,
+      "learning_rate": 9.57410964047529e-06,
+      "loss": 0.3126,
+      "step": 648
+    },
+    {
+      "epoch": 0.439106901217862,
+      "grad_norm": 0.6224172548164867,
+      "learning_rate": 9.571721571922505e-06,
+      "loss": 0.3011,
+      "step": 649
+    },
+    {
+      "epoch": 0.4397834912043302,
+      "grad_norm": 0.5937312594941452,
+      "learning_rate": 9.569327126372985e-06,
+      "loss": 0.2817,
+      "step": 650
+    },
+    {
+      "epoch": 0.4404600811907984,
+      "grad_norm": 0.6106497493885047,
+      "learning_rate": 9.56692630716669e-06,
+      "loss": 0.3082,
+      "step": 651
+    },
+    {
+      "epoch": 0.44113667117726657,
+      "grad_norm": 0.5603490682252269,
+      "learning_rate": 9.564519117652473e-06,
+      "loss": 0.3024,
+      "step": 652
+    },
+    {
+      "epoch": 0.44181326116373476,
+      "grad_norm": 0.590182887843202,
+      "learning_rate": 9.562105561188069e-06,
+      "loss": 0.3018,
+      "step": 653
+    },
+    {
+      "epoch": 0.44248985115020295,
+      "grad_norm": 0.5888604509673957,
+      "learning_rate": 9.559685641140098e-06,
+      "loss": 0.2985,
+      "step": 654
+    },
+    {
+      "epoch": 0.4431664411366712,
+      "grad_norm": 0.6213706602738642,
+      "learning_rate": 9.557259360884054e-06,
+      "loss": 0.3075,
+      "step": 655
+    },
+    {
+      "epoch": 0.4438430311231394,
+      "grad_norm": 0.5651441777203571,
+      "learning_rate": 9.554826723804304e-06,
+      "loss": 0.3015,
+      "step": 656
+    },
+    {
+      "epoch": 0.4445196211096076,
+      "grad_norm": 0.5527950463599898,
+      "learning_rate": 9.552387733294081e-06,
+      "loss": 0.3019,
+      "step": 657
+    },
+    {
+      "epoch": 0.44519621109607577,
+      "grad_norm": 0.5974888668532754,
+      "learning_rate": 9.54994239275548e-06,
+      "loss": 0.3112,
+      "step": 658
+    },
+    {
+      "epoch": 0.44587280108254396,
+      "grad_norm": 0.5976497336639846,
+      "learning_rate": 9.547490705599456e-06,
+      "loss": 0.3024,
+      "step": 659
+    },
+    {
+      "epoch": 0.4465493910690122,
+      "grad_norm": 0.6041417861067779,
+      "learning_rate": 9.545032675245814e-06,
+      "loss": 0.3231,
+      "step": 660
+    },
+    {
+      "epoch": 0.4472259810554804,
+      "grad_norm": 0.5901869905353382,
+      "learning_rate": 9.542568305123207e-06,
+      "loss": 0.3051,
+      "step": 661
+    },
+    {
+      "epoch": 0.4479025710419486,
+      "grad_norm": 0.5802069028004831,
+      "learning_rate": 9.540097598669135e-06,
+      "loss": 0.3088,
+      "step": 662
+    },
+    {
+      "epoch": 0.4485791610284168,
+      "grad_norm": 0.5603292281273322,
+      "learning_rate": 9.537620559329931e-06,
+      "loss": 0.3094,
+      "step": 663
+    },
+    {
+      "epoch": 0.44925575101488496,
+      "grad_norm": 0.5836902713404347,
+      "learning_rate": 9.535137190560765e-06,
+      "loss": 0.2955,
+      "step": 664
+    },
+    {
+      "epoch": 0.44993234100135315,
+      "grad_norm": 0.5788042071472832,
+      "learning_rate": 9.532647495825636e-06,
+      "loss": 0.3055,
+      "step": 665
+    },
+    {
+      "epoch": 0.4506089309878214,
+      "grad_norm": 0.5779954766153762,
+      "learning_rate": 9.530151478597366e-06,
+      "loss": 0.3008,
+      "step": 666
+    },
+    {
+      "epoch": 0.4512855209742896,
+      "grad_norm": 0.5999038142246513,
+      "learning_rate": 9.527649142357596e-06,
+      "loss": 0.3049,
+      "step": 667
+    },
+    {
+      "epoch": 0.4519621109607578,
+      "grad_norm": 0.5891185260176424,
+      "learning_rate": 9.525140490596778e-06,
+      "loss": 0.3081,
+      "step": 668
+    },
+    {
+      "epoch": 0.45263870094722597,
+      "grad_norm": 0.5818519014184372,
+      "learning_rate": 9.52262552681418e-06,
+      "loss": 0.3136,
+      "step": 669
+    },
+    {
+      "epoch": 0.45331529093369416,
+      "grad_norm": 0.6023458758686873,
+      "learning_rate": 9.520104254517873e-06,
+      "loss": 0.2948,
+      "step": 670
+    },
+    {
+      "epoch": 0.4539918809201624,
+      "grad_norm": 0.5838005373239196,
+      "learning_rate": 9.517576677224723e-06,
+      "loss": 0.2994,
+      "step": 671
+    },
+    {
+      "epoch": 0.4546684709066306,
+      "grad_norm": 0.5633728326204963,
+      "learning_rate": 9.515042798460393e-06,
+      "loss": 0.3183,
+      "step": 672
+    },
+    {
+      "epoch": 0.4553450608930988,
+      "grad_norm": 0.5782090307017909,
+      "learning_rate": 9.512502621759339e-06,
+      "loss": 0.31,
+      "step": 673
+    },
+    {
+      "epoch": 0.456021650879567,
+      "grad_norm": 0.5825149223196275,
+      "learning_rate": 9.509956150664796e-06,
+      "loss": 0.2996,
+      "step": 674
+    },
+    {
+      "epoch": 0.45669824086603517,
+      "grad_norm": 0.6432575412420874,
+      "learning_rate": 9.507403388728784e-06,
+      "loss": 0.304,
+      "step": 675
+    },
+    {
+      "epoch": 0.45737483085250336,
+      "grad_norm": 0.5621282477092442,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.2939,
+      "step": 676
+    },
+    {
+      "epoch": 0.4580514208389716,
+      "grad_norm": 0.5607346354089252,
+      "learning_rate": 9.502279006584297e-06,
+      "loss": 0.3065,
+      "step": 677
+    },
+    {
+      "epoch": 0.4587280108254398,
+      "grad_norm": 0.5401951211870069,
+      "learning_rate": 9.49970739352371e-06,
+      "loss": 0.2914,
+      "step": 678
+    },
+    {
+      "epoch": 0.459404600811908,
+      "grad_norm": 0.6291769750957792,
+      "learning_rate": 9.497129503917429e-06,
+      "loss": 0.3065,
+      "step": 679
+    },
+    {
+      "epoch": 0.46008119079837617,
+      "grad_norm": 0.6177125833731143,
+      "learning_rate": 9.494545341361291e-06,
+      "loss": 0.2936,
+      "step": 680
+    },
+    {
+      "epoch": 0.46075778078484436,
+      "grad_norm": 0.5844309782003548,
+      "learning_rate": 9.491954909459895e-06,
+      "loss": 0.3132,
+      "step": 681
+    },
+    {
+      "epoch": 0.4614343707713126,
+      "grad_norm": 0.5554676207720802,
+      "learning_rate": 9.489358211826577e-06,
+      "loss": 0.3118,
+      "step": 682
+    },
+    {
+      "epoch": 0.4621109607577808,
+      "grad_norm": 0.5860693594762715,
+      "learning_rate": 9.486755252083413e-06,
+      "loss": 0.3069,
+      "step": 683
+    },
+    {
+      "epoch": 0.462787550744249,
+      "grad_norm": 0.5728427387987377,
+      "learning_rate": 9.484146033861216e-06,
+      "loss": 0.2948,
+      "step": 684
+    },
+    {
+      "epoch": 0.4634641407307172,
+      "grad_norm": 0.5622035636303329,
+      "learning_rate": 9.481530560799534e-06,
+      "loss": 0.3013,
+      "step": 685
+    },
+    {
+      "epoch": 0.46414073071718537,
+      "grad_norm": 0.5671719303961282,
+      "learning_rate": 9.478908836546629e-06,
+      "loss": 0.3034,
+      "step": 686
+    },
+    {
+      "epoch": 0.4648173207036536,
+      "grad_norm": 0.5695815189984385,
+      "learning_rate": 9.47628086475949e-06,
+      "loss": 0.3047,
+      "step": 687
+    },
+    {
+      "epoch": 0.4654939106901218,
+      "grad_norm": 0.5666997429765435,
+      "learning_rate": 9.473646649103819e-06,
+      "loss": 0.2972,
+      "step": 688
+    },
+    {
+      "epoch": 0.46617050067659,
+      "grad_norm": 0.5626743462336447,
+      "learning_rate": 9.471006193254026e-06,
+      "loss": 0.3065,
+      "step": 689
+    },
+    {
+      "epoch": 0.4668470906630582,
+      "grad_norm": 0.5961537473140269,
+      "learning_rate": 9.468359500893227e-06,
+      "loss": 0.318,
+      "step": 690
+    },
+    {
+      "epoch": 0.4675236806495264,
+      "grad_norm": 0.5523433869930681,
+      "learning_rate": 9.465706575713235e-06,
+      "loss": 0.3023,
+      "step": 691
+    },
+    {
+      "epoch": 0.46820027063599456,
+      "grad_norm": 0.5635279108490268,
+      "learning_rate": 9.463047421414564e-06,
+      "loss": 0.2919,
+      "step": 692
+    },
+    {
+      "epoch": 0.4688768606224628,
+      "grad_norm": 0.5696567860248648,
+      "learning_rate": 9.460382041706406e-06,
+      "loss": 0.3098,
+      "step": 693
+    },
+    {
+      "epoch": 0.469553450608931,
+      "grad_norm": 0.5806106475386249,
+      "learning_rate": 9.457710440306645e-06,
+      "loss": 0.3001,
+      "step": 694
+    },
+    {
+      "epoch": 0.4702300405953992,
+      "grad_norm": 0.5990942110482262,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.3044,
+      "step": 695
+    },
+    {
+      "epoch": 0.4709066305818674,
+      "grad_norm": 0.5952880875638943,
+      "learning_rate": 9.452348587347224e-06,
+      "loss": 0.297,
+      "step": 696
+    },
+    {
+      "epoch": 0.47158322056833557,
+      "grad_norm": 0.577647515019447,
+      "learning_rate": 9.449658343266697e-06,
+      "loss": 0.2891,
+      "step": 697
+    },
+    {
+      "epoch": 0.4722598105548038,
+      "grad_norm": 0.6336618538196243,
+      "learning_rate": 9.446961892452824e-06,
+      "loss": 0.3077,
+      "step": 698
+    },
+    {
+      "epoch": 0.472936400541272,
+      "grad_norm": 0.5984991255901817,
+      "learning_rate": 9.444259238666827e-06,
+      "loss": 0.3027,
+      "step": 699
+    },
+    {
+      "epoch": 0.4736129905277402,
+      "grad_norm": 0.5763000256588728,
+      "learning_rate": 9.44155038567858e-06,
+      "loss": 0.3131,
+      "step": 700
+    },
+    {
+      "epoch": 0.4742895805142084,
+      "grad_norm": 0.6362514805847823,
+      "learning_rate": 9.438835337266603e-06,
+      "loss": 0.3167,
+      "step": 701
+    },
+    {
+      "epoch": 0.4749661705006766,
+      "grad_norm": 0.5864927769077318,
+      "learning_rate": 9.43611409721806e-06,
+      "loss": 0.3135,
+      "step": 702
+    },
+    {
+      "epoch": 0.47564276048714477,
+      "grad_norm": 0.5935598578107959,
+      "learning_rate": 9.43338666932875e-06,
+      "loss": 0.3033,
+      "step": 703
+    },
+    {
+      "epoch": 0.476319350473613,
+      "grad_norm": 0.5774320122481774,
+      "learning_rate": 9.430653057403105e-06,
+      "loss": 0.3014,
+      "step": 704
+    },
+    {
+      "epoch": 0.4769959404600812,
+      "grad_norm": 0.6226603645346016,
+      "learning_rate": 9.42791326525418e-06,
+      "loss": 0.3137,
+      "step": 705
+    },
+    {
+      "epoch": 0.4776725304465494,
+      "grad_norm": 0.597999143596453,
+      "learning_rate": 9.425167296703655e-06,
+      "loss": 0.2967,
+      "step": 706
+    },
+    {
+      "epoch": 0.4783491204330176,
+      "grad_norm": 0.5691406676074343,
+      "learning_rate": 9.422415155581822e-06,
+      "loss": 0.3059,
+      "step": 707
+    },
+    {
+      "epoch": 0.4790257104194858,
+      "grad_norm": 0.5867821041576133,
+      "learning_rate": 9.419656845727582e-06,
+      "loss": 0.3014,
+      "step": 708
+    },
+    {
+      "epoch": 0.479702300405954,
+      "grad_norm": 0.576552313884612,
+      "learning_rate": 9.416892370988445e-06,
+      "loss": 0.3032,
+      "step": 709
+    },
+    {
+      "epoch": 0.4803788903924222,
+      "grad_norm": 0.5968997781127228,
+      "learning_rate": 9.414121735220513e-06,
+      "loss": 0.2975,
+      "step": 710
+    },
+    {
+      "epoch": 0.4810554803788904,
+      "grad_norm": 0.6332750102172536,
+      "learning_rate": 9.411344942288493e-06,
+      "loss": 0.2992,
+      "step": 711
+    },
+    {
+      "epoch": 0.4817320703653586,
+      "grad_norm": 0.5670879015697637,
+      "learning_rate": 9.408561996065672e-06,
+      "loss": 0.2957,
+      "step": 712
+    },
+    {
+      "epoch": 0.4824086603518268,
+      "grad_norm": 0.5548400801051895,
+      "learning_rate": 9.405772900433922e-06,
+      "loss": 0.2928,
+      "step": 713
+    },
+    {
+      "epoch": 0.48308525033829497,
+      "grad_norm": 0.5517876636249128,
+      "learning_rate": 9.40297765928369e-06,
+      "loss": 0.3021,
+      "step": 714
+    },
+    {
+      "epoch": 0.4837618403247632,
+      "grad_norm": 0.5778716958415082,
+      "learning_rate": 9.400176276514002e-06,
+      "loss": 0.3076,
+      "step": 715
+    },
+    {
+      "epoch": 0.4844384303112314,
+      "grad_norm": 0.5684720065937993,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.3081,
+      "step": 716
+    },
+    {
+      "epoch": 0.4851150202976996,
+      "grad_norm": 0.5821870498506955,
+      "learning_rate": 9.394555101755173e-06,
+      "loss": 0.3082,
+      "step": 717
+    },
+    {
+      "epoch": 0.4857916102841678,
+      "grad_norm": 0.5508732557622712,
+      "learning_rate": 9.391735317606885e-06,
+      "loss": 0.3058,
+      "step": 718
+    },
+    {
+      "epoch": 0.486468200270636,
+      "grad_norm": 0.570571997112907,
+      "learning_rate": 9.388909407520843e-06,
+      "loss": 0.3089,
+      "step": 719
+    },
+    {
+      "epoch": 0.4871447902571042,
+      "grad_norm": 0.5803559614467765,
+      "learning_rate": 9.386077375438848e-06,
+      "loss": 0.2957,
+      "step": 720
+    },
+    {
+      "epoch": 0.4878213802435724,
+      "grad_norm": 0.5518050666730866,
+      "learning_rate": 9.38323922531124e-06,
+      "loss": 0.2915,
+      "step": 721
+    },
+    {
+      "epoch": 0.4884979702300406,
+      "grad_norm": 0.5642088812749738,
+      "learning_rate": 9.380394961096895e-06,
+      "loss": 0.2988,
+      "step": 722
+    },
+    {
+      "epoch": 0.4891745602165088,
+      "grad_norm": 0.5831941689431623,
+      "learning_rate": 9.377544586763216e-06,
+      "loss": 0.2847,
+      "step": 723
+    },
+    {
+      "epoch": 0.489851150202977,
+      "grad_norm": 0.5743739052810604,
+      "learning_rate": 9.374688106286127e-06,
+      "loss": 0.3043,
+      "step": 724
+    },
+    {
+      "epoch": 0.49052774018944517,
+      "grad_norm": 0.5554345694557313,
+      "learning_rate": 9.371825523650075e-06,
+      "loss": 0.2903,
+      "step": 725
+    },
+    {
+      "epoch": 0.4912043301759134,
+      "grad_norm": 0.584314042506332,
+      "learning_rate": 9.368956842848014e-06,
+      "loss": 0.3045,
+      "step": 726
+    },
+    {
+      "epoch": 0.4918809201623816,
+      "grad_norm": 0.5363674116973478,
+      "learning_rate": 9.366082067881406e-06,
+      "loss": 0.299,
+      "step": 727
+    },
+    {
+      "epoch": 0.4925575101488498,
+      "grad_norm": 0.5809483035562704,
+      "learning_rate": 9.363201202760212e-06,
+      "loss": 0.3148,
+      "step": 728
+    },
+    {
+      "epoch": 0.493234100135318,
+      "grad_norm": 0.554608296618633,
+      "learning_rate": 9.36031425150289e-06,
+      "loss": 0.3087,
+      "step": 729
+    },
+    {
+      "epoch": 0.4939106901217862,
+      "grad_norm": 0.5850022151393639,
+      "learning_rate": 9.357421218136387e-06,
+      "loss": 0.3027,
+      "step": 730
+    },
+    {
+      "epoch": 0.4945872801082544,
+      "grad_norm": 0.5611020296498642,
+      "learning_rate": 9.354522106696133e-06,
+      "loss": 0.3039,
+      "step": 731
+    },
+    {
+      "epoch": 0.4952638700947226,
+      "grad_norm": 0.540593085853543,
+      "learning_rate": 9.351616921226036e-06,
+      "loss": 0.3029,
+      "step": 732
+    },
+    {
+      "epoch": 0.4959404600811908,
+      "grad_norm": 0.5595080561332507,
+      "learning_rate": 9.348705665778479e-06,
+      "loss": 0.2962,
+      "step": 733
+    },
+    {
+      "epoch": 0.496617050067659,
+      "grad_norm": 0.5229361800320294,
+      "learning_rate": 9.345788344414306e-06,
+      "loss": 0.2876,
+      "step": 734
+    },
+    {
+      "epoch": 0.4972936400541272,
+      "grad_norm": 0.570940663700859,
+      "learning_rate": 9.34286496120283e-06,
+      "loss": 0.3034,
+      "step": 735
+    },
+    {
+      "epoch": 0.4979702300405954,
+      "grad_norm": 0.55005659106939,
+      "learning_rate": 9.339935520221816e-06,
+      "loss": 0.298,
+      "step": 736
+    },
+    {
+      "epoch": 0.4986468200270636,
+      "grad_norm": 0.5943457503245411,
+      "learning_rate": 9.337000025557477e-06,
+      "loss": 0.304,
+      "step": 737
+    },
+    {
+      "epoch": 0.4993234100135318,
+      "grad_norm": 0.5787640041690718,
+      "learning_rate": 9.33405848130447e-06,
+      "loss": 0.2962,
+      "step": 738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5481869873493219,
+      "learning_rate": 9.331110891565898e-06,
+      "loss": 0.2995,
+      "step": 739
+    },
+    {
+      "epoch": 0.5006765899864682,
+      "grad_norm": 0.5971582580763398,
+      "learning_rate": 9.328157260453286e-06,
+      "loss": 0.3026,
+      "step": 740
+    },
+    {
+      "epoch": 0.5013531799729364,
+      "grad_norm": 0.6030895703718813,
+      "learning_rate": 9.32519759208659e-06,
+      "loss": 0.3177,
+      "step": 741
+    },
+    {
+      "epoch": 0.5020297699594046,
+      "grad_norm": 0.5732906503233692,
+      "learning_rate": 9.322231890594193e-06,
+      "loss": 0.2916,
+      "step": 742
+    },
+    {
+      "epoch": 0.5027063599458728,
+      "grad_norm": 0.5635363669867718,
+      "learning_rate": 9.319260160112884e-06,
+      "loss": 0.3047,
+      "step": 743
+    },
+    {
+      "epoch": 0.503382949932341,
+      "grad_norm": 0.5565821024838211,
+      "learning_rate": 9.31628240478787e-06,
+      "loss": 0.3005,
+      "step": 744
+    },
+    {
+      "epoch": 0.5040595399188093,
+      "grad_norm": 0.5165607557582695,
+      "learning_rate": 9.313298628772756e-06,
+      "loss": 0.2976,
+      "step": 745
+    },
+    {
+      "epoch": 0.5047361299052774,
+      "grad_norm": 0.5772867732674161,
+      "learning_rate": 9.310308836229548e-06,
+      "loss": 0.3041,
+      "step": 746
+    },
+    {
+      "epoch": 0.5054127198917456,
+      "grad_norm": 0.5880939570042524,
+      "learning_rate": 9.307313031328644e-06,
+      "loss": 0.3112,
+      "step": 747
+    },
+    {
+      "epoch": 0.5060893098782138,
+      "grad_norm": 0.5479667295010011,
+      "learning_rate": 9.304311218248828e-06,
+      "loss": 0.2881,
+      "step": 748
+    },
+    {
+      "epoch": 0.506765899864682,
+      "grad_norm": 0.5398695105966872,
+      "learning_rate": 9.301303401177266e-06,
+      "loss": 0.2995,
+      "step": 749
+    },
+    {
+      "epoch": 0.5074424898511503,
+      "grad_norm": 0.5570098007134875,
+      "learning_rate": 9.298289584309496e-06,
+      "loss": 0.2988,
+      "step": 750
+    },
+    {
+      "epoch": 0.5081190798376184,
+      "grad_norm": 0.5911603398735041,
+      "learning_rate": 9.295269771849426e-06,
+      "loss": 0.3106,
+      "step": 751
+    },
+    {
+      "epoch": 0.5087956698240866,
+      "grad_norm": 0.5489027508097252,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.2841,
+      "step": 752
+    },
+    {
+      "epoch": 0.5094722598105548,
+      "grad_norm": 0.5672620138822246,
+      "learning_rate": 9.289212177009839e-06,
+      "loss": 0.3022,
+      "step": 753
+    },
+    {
+      "epoch": 0.510148849797023,
+      "grad_norm": 0.5649460402479857,
+      "learning_rate": 9.286174403079928e-06,
+      "loss": 0.2993,
+      "step": 754
+    },
+    {
+      "epoch": 0.5108254397834912,
+      "grad_norm": 0.5528223972596815,
+      "learning_rate": 9.283130650456929e-06,
+      "loss": 0.3053,
+      "step": 755
+    },
+    {
+      "epoch": 0.5115020297699594,
+      "grad_norm": 0.5474812711054658,
+      "learning_rate": 9.280080923386501e-06,
+      "loss": 0.2907,
+      "step": 756
+    },
+    {
+      "epoch": 0.5121786197564276,
+      "grad_norm": 0.6183920117502267,
+      "learning_rate": 9.27702522612265e-06,
+      "loss": 0.3162,
+      "step": 757
+    },
+    {
+      "epoch": 0.5128552097428958,
+      "grad_norm": 0.5376407595845764,
+      "learning_rate": 9.273963562927695e-06,
+      "loss": 0.2948,
+      "step": 758
+    },
+    {
+      "epoch": 0.513531799729364,
+      "grad_norm": 0.5595303673378601,
+      "learning_rate": 9.270895938072293e-06,
+      "loss": 0.308,
+      "step": 759
+    },
+    {
+      "epoch": 0.5142083897158322,
+      "grad_norm": 0.5497287092038473,
+      "learning_rate": 9.267822355835402e-06,
+      "loss": 0.3009,
+      "step": 760
+    },
+    {
+      "epoch": 0.5148849797023004,
+      "grad_norm": 0.5578727398149145,
+      "learning_rate": 9.2647428205043e-06,
+      "loss": 0.3078,
+      "step": 761
+    },
+    {
+      "epoch": 0.5155615696887687,
+      "grad_norm": 0.5422517252375881,
+      "learning_rate": 9.261657336374561e-06,
+      "loss": 0.3074,
+      "step": 762
+    },
+    {
+      "epoch": 0.5162381596752368,
+      "grad_norm": 0.5635519154411008,
+      "learning_rate": 9.258565907750065e-06,
+      "loss": 0.2879,
+      "step": 763
+    },
+    {
+      "epoch": 0.516914749661705,
+      "grad_norm": 0.5571622189281922,
+      "learning_rate": 9.25546853894298e-06,
+      "loss": 0.3047,
+      "step": 764
+    },
+    {
+      "epoch": 0.5175913396481732,
+      "grad_norm": 0.5704023498932288,
+      "learning_rate": 9.252365234273754e-06,
+      "loss": 0.3034,
+      "step": 765
+    },
+    {
+      "epoch": 0.5182679296346414,
+      "grad_norm": 0.5623491152101785,
+      "learning_rate": 9.249255998071127e-06,
+      "loss": 0.302,
+      "step": 766
+    },
+    {
+      "epoch": 0.5189445196211097,
+      "grad_norm": 0.5809468671534228,
+      "learning_rate": 9.246140834672103e-06,
+      "loss": 0.3043,
+      "step": 767
+    },
+    {
+      "epoch": 0.5196211096075778,
+      "grad_norm": 0.5617160794203792,
+      "learning_rate": 9.243019748421956e-06,
+      "loss": 0.3047,
+      "step": 768
+    },
+    {
+      "epoch": 0.520297699594046,
+      "grad_norm": 0.5991321479782231,
+      "learning_rate": 9.239892743674224e-06,
+      "loss": 0.3061,
+      "step": 769
+    },
+    {
+      "epoch": 0.5209742895805142,
+      "grad_norm": 0.5996902389783957,
+      "learning_rate": 9.236759824790698e-06,
+      "loss": 0.304,
+      "step": 770
+    },
+    {
+      "epoch": 0.5216508795669824,
+      "grad_norm": 0.5413489349467447,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.3014,
+      "step": 771
+    },
+    {
+      "epoch": 0.5223274695534507,
+      "grad_norm": 0.588701909036404,
+      "learning_rate": 9.230476262104678e-06,
+      "loss": 0.3143,
+      "step": 772
+    },
+    {
+      "epoch": 0.5230040595399188,
+      "grad_norm": 0.5528438577558575,
+      "learning_rate": 9.227325627066989e-06,
+      "loss": 0.2961,
+      "step": 773
+    },
+    {
+      "epoch": 0.523680649526387,
+      "grad_norm": 0.5420308380517652,
+      "learning_rate": 9.22416909542311e-06,
+      "loss": 0.3036,
+      "step": 774
+    },
+    {
+      "epoch": 0.5243572395128552,
+      "grad_norm": 0.5592212673426991,
+      "learning_rate": 9.22100667157602e-06,
+      "loss": 0.3117,
+      "step": 775
+    },
+    {
+      "epoch": 0.5250338294993234,
+      "grad_norm": 0.5723557855060515,
+      "learning_rate": 9.217838359936914e-06,
+      "loss": 0.2967,
+      "step": 776
+    },
+    {
+      "epoch": 0.5257104194857916,
+      "grad_norm": 0.5560898205095529,
+      "learning_rate": 9.214664164925204e-06,
+      "loss": 0.3049,
+      "step": 777
+    },
+    {
+      "epoch": 0.5263870094722598,
+      "grad_norm": 0.5270386621848234,
+      "learning_rate": 9.211484090968505e-06,
+      "loss": 0.3047,
+      "step": 778
+    },
+    {
+      "epoch": 0.527063599458728,
+      "grad_norm": 0.5560744392173804,
+      "learning_rate": 9.208298142502637e-06,
+      "loss": 0.3031,
+      "step": 779
+    },
+    {
+      "epoch": 0.5277401894451962,
+      "grad_norm": 0.5777722504723846,
+      "learning_rate": 9.205106323971607e-06,
+      "loss": 0.3105,
+      "step": 780
+    },
+    {
+      "epoch": 0.5284167794316644,
+      "grad_norm": 0.5460783241187096,
+      "learning_rate": 9.201908639827619e-06,
+      "loss": 0.2958,
+      "step": 781
+    },
+    {
+      "epoch": 0.5290933694181326,
+      "grad_norm": 0.5563471964684251,
+      "learning_rate": 9.198705094531053e-06,
+      "loss": 0.3061,
+      "step": 782
+    },
+    {
+      "epoch": 0.5297699594046008,
+      "grad_norm": 0.5354561245195842,
+      "learning_rate": 9.195495692550465e-06,
+      "loss": 0.2901,
+      "step": 783
+    },
+    {
+      "epoch": 0.530446549391069,
+      "grad_norm": 0.5911763498547373,
+      "learning_rate": 9.192280438362581e-06,
+      "loss": 0.2996,
+      "step": 784
+    },
+    {
+      "epoch": 0.5311231393775372,
+      "grad_norm": 0.5485197472153303,
+      "learning_rate": 9.189059336452292e-06,
+      "loss": 0.3011,
+      "step": 785
+    },
+    {
+      "epoch": 0.5317997293640054,
+      "grad_norm": 0.5437810111182592,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.2957,
+      "step": 786
+    },
+    {
+      "epoch": 0.5324763193504736,
+      "grad_norm": 0.5714887484046028,
+      "learning_rate": 9.182599607444833e-06,
+      "loss": 0.306,
+      "step": 787
+    },
+    {
+      "epoch": 0.5331529093369418,
+      "grad_norm": 0.5639806834347181,
+      "learning_rate": 9.179360989358199e-06,
+      "loss": 0.2895,
+      "step": 788
+    },
+    {
+      "epoch": 0.5338294993234101,
+      "grad_norm": 0.5641414860374772,
+      "learning_rate": 9.17611654157022e-06,
+      "loss": 0.3151,
+      "step": 789
+    },
+    {
+      "epoch": 0.5345060893098782,
+      "grad_norm": 0.5457962148744243,
+      "learning_rate": 9.172866268606514e-06,
+      "loss": 0.2817,
+      "step": 790
+    },
+    {
+      "epoch": 0.5351826792963464,
+      "grad_norm": 0.5542792445008103,
+      "learning_rate": 9.169610175000812e-06,
+      "loss": 0.3023,
+      "step": 791
+    },
+    {
+      "epoch": 0.5358592692828146,
+      "grad_norm": 0.5547740239276749,
+      "learning_rate": 9.166348265294968e-06,
+      "loss": 0.2977,
+      "step": 792
+    },
+    {
+      "epoch": 0.5365358592692828,
+      "grad_norm": 0.5277485637050191,
+      "learning_rate": 9.163080544038953e-06,
+      "loss": 0.2991,
+      "step": 793
+    },
+    {
+      "epoch": 0.5372124492557511,
+      "grad_norm": 0.5880123570210535,
+      "learning_rate": 9.15980701579084e-06,
+      "loss": 0.2964,
+      "step": 794
+    },
+    {
+      "epoch": 0.5378890392422192,
+      "grad_norm": 0.5790189468999908,
+      "learning_rate": 9.156527685116806e-06,
+      "loss": 0.2992,
+      "step": 795
+    },
+    {
+      "epoch": 0.5385656292286874,
+      "grad_norm": 0.5407021497742042,
+      "learning_rate": 9.153242556591115e-06,
+      "loss": 0.2942,
+      "step": 796
+    },
+    {
+      "epoch": 0.5392422192151556,
+      "grad_norm": 0.5912848776401304,
+      "learning_rate": 9.149951634796128e-06,
+      "loss": 0.3087,
+      "step": 797
+    },
+    {
+      "epoch": 0.5399188092016238,
+      "grad_norm": 0.5819941538897025,
+      "learning_rate": 9.146654924322277e-06,
+      "loss": 0.2943,
+      "step": 798
+    },
+    {
+      "epoch": 0.540595399188092,
+      "grad_norm": 0.543412380697745,
+      "learning_rate": 9.143352429768075e-06,
+      "loss": 0.295,
+      "step": 799
+    },
+    {
+      "epoch": 0.5412719891745602,
+      "grad_norm": 0.5860569872825235,
+      "learning_rate": 9.140044155740102e-06,
+      "loss": 0.2962,
+      "step": 800
+    },
+    {
+      "epoch": 0.5419485791610285,
+      "grad_norm": 0.6004777553899262,
+      "learning_rate": 9.136730106852995e-06,
+      "loss": 0.2962,
+      "step": 801
+    },
+    {
+      "epoch": 0.5426251691474966,
+      "grad_norm": 0.5908162952133265,
+      "learning_rate": 9.133410287729454e-06,
+      "loss": 0.3053,
+      "step": 802
+    },
+    {
+      "epoch": 0.5433017591339648,
+      "grad_norm": 0.5671439136754883,
+      "learning_rate": 9.130084703000223e-06,
+      "loss": 0.302,
+      "step": 803
+    },
+    {
+      "epoch": 0.543978349120433,
+      "grad_norm": 0.5790862133795486,
+      "learning_rate": 9.126753357304088e-06,
+      "loss": 0.3033,
+      "step": 804
+    },
+    {
+      "epoch": 0.5446549391069012,
+      "grad_norm": 0.6006372901185502,
+      "learning_rate": 9.123416255287872e-06,
+      "loss": 0.2976,
+      "step": 805
+    },
+    {
+      "epoch": 0.5453315290933695,
+      "grad_norm": 0.5433073221131227,
+      "learning_rate": 9.120073401606427e-06,
+      "loss": 0.2989,
+      "step": 806
+    },
+    {
+      "epoch": 0.5460081190798376,
+      "grad_norm": 0.6338134689464464,
+      "learning_rate": 9.11672480092263e-06,
+      "loss": 0.2982,
+      "step": 807
+    },
+    {
+      "epoch": 0.5466847090663058,
+      "grad_norm": 0.6105445042972905,
+      "learning_rate": 9.11337045790737e-06,
+      "loss": 0.3021,
+      "step": 808
+    },
+    {
+      "epoch": 0.547361299052774,
+      "grad_norm": 0.5521255140268764,
+      "learning_rate": 9.110010377239552e-06,
+      "loss": 0.3001,
+      "step": 809
+    },
+    {
+      "epoch": 0.5480378890392422,
+      "grad_norm": 0.5654046914862141,
+      "learning_rate": 9.106644563606076e-06,
+      "loss": 0.2949,
+      "step": 810
+    },
+    {
+      "epoch": 0.5487144790257105,
+      "grad_norm": 0.5784350313690156,
+      "learning_rate": 9.103273021701846e-06,
+      "loss": 0.2986,
+      "step": 811
+    },
+    {
+      "epoch": 0.5493910690121786,
+      "grad_norm": 0.5396891965473214,
+      "learning_rate": 9.099895756229754e-06,
+      "loss": 0.2976,
+      "step": 812
+    },
+    {
+      "epoch": 0.5500676589986468,
+      "grad_norm": 0.5457991342153601,
+      "learning_rate": 9.096512771900675e-06,
+      "loss": 0.2949,
+      "step": 813
+    },
+    {
+      "epoch": 0.550744248985115,
+      "grad_norm": 0.5431328226879331,
+      "learning_rate": 9.093124073433464e-06,
+      "loss": 0.2906,
+      "step": 814
+    },
+    {
+      "epoch": 0.5514208389715832,
+      "grad_norm": 0.561021031908741,
+      "learning_rate": 9.089729665554939e-06,
+      "loss": 0.2981,
+      "step": 815
+    },
+    {
+      "epoch": 0.5520974289580515,
+      "grad_norm": 0.5463397263216477,
+      "learning_rate": 9.08632955299989e-06,
+      "loss": 0.3048,
+      "step": 816
+    },
+    {
+      "epoch": 0.5527740189445196,
+      "grad_norm": 0.5553814415702706,
+      "learning_rate": 9.082923740511063e-06,
+      "loss": 0.3022,
+      "step": 817
+    },
+    {
+      "epoch": 0.5534506089309879,
+      "grad_norm": 0.5587705848639869,
+      "learning_rate": 9.07951223283915e-06,
+      "loss": 0.2994,
+      "step": 818
+    },
+    {
+      "epoch": 0.554127198917456,
+      "grad_norm": 0.5663529073623439,
+      "learning_rate": 9.076095034742792e-06,
+      "loss": 0.3039,
+      "step": 819
+    },
+    {
+      "epoch": 0.5548037889039242,
+      "grad_norm": 0.5524199841236451,
+      "learning_rate": 9.072672150988563e-06,
+      "loss": 0.3062,
+      "step": 820
+    },
+    {
+      "epoch": 0.5554803788903924,
+      "grad_norm": 0.5522549690106191,
+      "learning_rate": 9.069243586350976e-06,
+      "loss": 0.2995,
+      "step": 821
+    },
+    {
+      "epoch": 0.5561569688768606,
+      "grad_norm": 0.5611626764212988,
+      "learning_rate": 9.065809345612458e-06,
+      "loss": 0.3066,
+      "step": 822
+    },
+    {
+      "epoch": 0.5568335588633289,
+      "grad_norm": 0.5446624645155115,
+      "learning_rate": 9.062369433563357e-06,
+      "loss": 0.2987,
+      "step": 823
+    },
+    {
+      "epoch": 0.557510148849797,
+      "grad_norm": 0.5630674098240749,
+      "learning_rate": 9.058923855001935e-06,
+      "loss": 0.2981,
+      "step": 824
+    },
+    {
+      "epoch": 0.5581867388362652,
+      "grad_norm": 0.5649415848867698,
+      "learning_rate": 9.055472614734358e-06,
+      "loss": 0.3095,
+      "step": 825
+    },
+    {
+      "epoch": 0.5588633288227334,
+      "grad_norm": 0.5275767032673048,
+      "learning_rate": 9.052015717574683e-06,
+      "loss": 0.2959,
+      "step": 826
+    },
+    {
+      "epoch": 0.5595399188092016,
+      "grad_norm": 0.5498945057567547,
+      "learning_rate": 9.048553168344864e-06,
+      "loss": 0.299,
+      "step": 827
+    },
+    {
+      "epoch": 0.5602165087956699,
+      "grad_norm": 0.5618002519716603,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.3001,
+      "step": 828
+    },
+    {
+      "epoch": 0.560893098782138,
+      "grad_norm": 0.56163736064354,
+      "learning_rate": 9.041611133002016e-06,
+      "loss": 0.3032,
+      "step": 829
+    },
+    {
+      "epoch": 0.5615696887686062,
+      "grad_norm": 0.5486554244532377,
+      "learning_rate": 9.038131656572284e-06,
+      "loss": 0.2973,
+      "step": 830
+    },
+    {
+      "epoch": 0.5622462787550744,
+      "grad_norm": 0.5429179967186969,
+      "learning_rate": 9.034646547438987e-06,
+      "loss": 0.2991,
+      "step": 831
+    },
+    {
+      "epoch": 0.5629228687415426,
+      "grad_norm": 0.5895171759332049,
+      "learning_rate": 9.03115581046343e-06,
+      "loss": 0.2999,
+      "step": 832
+    },
+    {
+      "epoch": 0.5635994587280109,
+      "grad_norm": 0.5817757193043149,
+      "learning_rate": 9.027659450514769e-06,
+      "loss": 0.2969,
+      "step": 833
+    },
+    {
+      "epoch": 0.564276048714479,
+      "grad_norm": 0.5380392039980519,
+      "learning_rate": 9.02415747247e-06,
+      "loss": 0.2946,
+      "step": 834
+    },
+    {
+      "epoch": 0.5649526387009473,
+      "grad_norm": 0.5856419612561949,
+      "learning_rate": 9.02064988121396e-06,
+      "loss": 0.3028,
+      "step": 835
+    },
+    {
+      "epoch": 0.5656292286874154,
+      "grad_norm": 0.5608080664875977,
+      "learning_rate": 9.017136681639307e-06,
+      "loss": 0.2968,
+      "step": 836
+    },
+    {
+      "epoch": 0.5663058186738836,
+      "grad_norm": 0.5845406825290534,
+      "learning_rate": 9.013617878646534e-06,
+      "loss": 0.3036,
+      "step": 837
+    },
+    {
+      "epoch": 0.5669824086603519,
+      "grad_norm": 0.5551103104411049,
+      "learning_rate": 9.010093477143942e-06,
+      "loss": 0.3017,
+      "step": 838
+    },
+    {
+      "epoch": 0.56765899864682,
+      "grad_norm": 0.5427911738378117,
+      "learning_rate": 9.006563482047643e-06,
+      "loss": 0.2928,
+      "step": 839
+    },
+    {
+      "epoch": 0.5683355886332883,
+      "grad_norm": 0.529058669319154,
+      "learning_rate": 9.003027898281551e-06,
+      "loss": 0.2949,
+      "step": 840
+    },
+    {
+      "epoch": 0.5690121786197564,
+      "grad_norm": 0.5654318627787605,
+      "learning_rate": 8.99948673077738e-06,
+      "loss": 0.3038,
+      "step": 841
+    },
+    {
+      "epoch": 0.5696887686062246,
+      "grad_norm": 0.5392295698871644,
+      "learning_rate": 8.995939984474624e-06,
+      "loss": 0.3005,
+      "step": 842
+    },
+    {
+      "epoch": 0.5703653585926928,
+      "grad_norm": 0.6127866528020111,
+      "learning_rate": 8.992387664320568e-06,
+      "loss": 0.3006,
+      "step": 843
+    },
+    {
+      "epoch": 0.571041948579161,
+      "grad_norm": 0.5403707397324433,
+      "learning_rate": 8.988829775270265e-06,
+      "loss": 0.2945,
+      "step": 844
+    },
+    {
+      "epoch": 0.5717185385656293,
+      "grad_norm": 0.5413747118717694,
+      "learning_rate": 8.985266322286538e-06,
+      "loss": 0.2986,
+      "step": 845
+    },
+    {
+      "epoch": 0.5723951285520974,
+      "grad_norm": 0.5533030512572205,
+      "learning_rate": 8.981697310339972e-06,
+      "loss": 0.3105,
+      "step": 846
+    },
+    {
+      "epoch": 0.5730717185385656,
+      "grad_norm": 0.5431968020135576,
+      "learning_rate": 8.978122744408905e-06,
+      "loss": 0.2998,
+      "step": 847
+    },
+    {
+      "epoch": 0.5737483085250338,
+      "grad_norm": 0.5848290922884716,
+      "learning_rate": 8.974542629479426e-06,
+      "loss": 0.2975,
+      "step": 848
+    },
+    {
+      "epoch": 0.574424898511502,
+      "grad_norm": 0.5668526190559471,
+      "learning_rate": 8.970956970545356e-06,
+      "loss": 0.3145,
+      "step": 849
+    },
+    {
+      "epoch": 0.5751014884979703,
+      "grad_norm": 0.5368939834179557,
+      "learning_rate": 8.967365772608258e-06,
+      "loss": 0.2941,
+      "step": 850
+    },
+    {
+      "epoch": 0.5757780784844384,
+      "grad_norm": 0.6144919772009699,
+      "learning_rate": 8.963769040677413e-06,
+      "loss": 0.3001,
+      "step": 851
+    },
+    {
+      "epoch": 0.5764546684709067,
+      "grad_norm": 0.5752483339450997,
+      "learning_rate": 8.96016677976983e-06,
+      "loss": 0.2967,
+      "step": 852
+    },
+    {
+      "epoch": 0.5771312584573748,
+      "grad_norm": 0.5512547646595551,
+      "learning_rate": 8.956558994910224e-06,
+      "loss": 0.2932,
+      "step": 853
+    },
+    {
+      "epoch": 0.577807848443843,
+      "grad_norm": 0.5964056761099592,
+      "learning_rate": 8.952945691131016e-06,
+      "loss": 0.3093,
+      "step": 854
+    },
+    {
+      "epoch": 0.5784844384303113,
+      "grad_norm": 0.5918235712116676,
+      "learning_rate": 8.949326873472327e-06,
+      "loss": 0.3116,
+      "step": 855
+    },
+    {
+      "epoch": 0.5791610284167794,
+      "grad_norm": 0.5836715203980507,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.2971,
+      "step": 856
+    },
+    {
+      "epoch": 0.5798376184032477,
+      "grad_norm": 0.5675143953149822,
+      "learning_rate": 8.942072716715436e-06,
+      "loss": 0.2939,
+      "step": 857
+    },
+    {
+      "epoch": 0.5805142083897158,
+      "grad_norm": 0.5655057389878644,
+      "learning_rate": 8.938437387735903e-06,
+      "loss": 0.3055,
+      "step": 858
+    },
+    {
+      "epoch": 0.581190798376184,
+      "grad_norm": 0.5613407351180515,
+      "learning_rate": 8.934796565114211e-06,
+      "loss": 0.3044,
+      "step": 859
+    },
+    {
+      "epoch": 0.5818673883626523,
+      "grad_norm": 0.5724212550581983,
+      "learning_rate": 8.931150253928866e-06,
+      "loss": 0.2908,
+      "step": 860
+    },
+    {
+      "epoch": 0.5825439783491204,
+      "grad_norm": 0.5781157536366013,
+      "learning_rate": 8.92749845926603e-06,
+      "loss": 0.2941,
+      "step": 861
+    },
+    {
+      "epoch": 0.5832205683355887,
+      "grad_norm": 0.550870043050043,
+      "learning_rate": 8.923841186219512e-06,
+      "loss": 0.3036,
+      "step": 862
+    },
+    {
+      "epoch": 0.5838971583220568,
+      "grad_norm": 0.5623359285175002,
+      "learning_rate": 8.920178439890765e-06,
+      "loss": 0.2938,
+      "step": 863
+    },
+    {
+      "epoch": 0.584573748308525,
+      "grad_norm": 0.6001398849789111,
+      "learning_rate": 8.916510225388878e-06,
+      "loss": 0.2953,
+      "step": 864
+    },
+    {
+      "epoch": 0.5852503382949933,
+      "grad_norm": 0.5632970494570869,
+      "learning_rate": 8.912836547830562e-06,
+      "loss": 0.2837,
+      "step": 865
+    },
+    {
+      "epoch": 0.5859269282814614,
+      "grad_norm": 0.5562124388230466,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.2921,
+      "step": 866
+    },
+    {
+      "epoch": 0.5866035182679297,
+      "grad_norm": 0.6335149656334864,
+      "learning_rate": 8.905472824049592e-06,
+      "loss": 0.3025,
+      "step": 867
+    },
+    {
+      "epoch": 0.5872801082543978,
+      "grad_norm": 0.5630222300460045,
+      "learning_rate": 8.901782788098442e-06,
+      "loss": 0.3015,
+      "step": 868
+    },
+    {
+      "epoch": 0.587956698240866,
+      "grad_norm": 0.5663950523188785,
+      "learning_rate": 8.898087309633851e-06,
+      "loss": 0.306,
+      "step": 869
+    },
+    {
+      "epoch": 0.5886332882273342,
+      "grad_norm": 0.5469823040129249,
+      "learning_rate": 8.894386393810563e-06,
+      "loss": 0.2985,
+      "step": 870
+    },
+    {
+      "epoch": 0.5893098782138024,
+      "grad_norm": 0.5832101052154601,
+      "learning_rate": 8.890680045790907e-06,
+      "loss": 0.2926,
+      "step": 871
+    },
+    {
+      "epoch": 0.5899864682002707,
+      "grad_norm": 0.5608779015264297,
+      "learning_rate": 8.886968270744789e-06,
+      "loss": 0.2956,
+      "step": 872
+    },
+    {
+      "epoch": 0.5906630581867388,
+      "grad_norm": 0.5553667848489872,
+      "learning_rate": 8.883251073849682e-06,
+      "loss": 0.3068,
+      "step": 873
+    },
+    {
+      "epoch": 0.591339648173207,
+      "grad_norm": 0.5433155878022492,
+      "learning_rate": 8.879528460290628e-06,
+      "loss": 0.3002,
+      "step": 874
+    },
+    {
+      "epoch": 0.5920162381596752,
+      "grad_norm": 0.5371904233962611,
+      "learning_rate": 8.875800435260218e-06,
+      "loss": 0.2863,
+      "step": 875
+    },
+    {
+      "epoch": 0.5926928281461434,
+      "grad_norm": 0.549511401213718,
+      "learning_rate": 8.872067003958597e-06,
+      "loss": 0.2996,
+      "step": 876
+    },
+    {
+      "epoch": 0.5933694181326117,
+      "grad_norm": 0.5506848891885107,
+      "learning_rate": 8.868328171593448e-06,
+      "loss": 0.2895,
+      "step": 877
+    },
+    {
+      "epoch": 0.5940460081190798,
+      "grad_norm": 0.5890834398975353,
+      "learning_rate": 8.864583943379987e-06,
+      "loss": 0.3112,
+      "step": 878
+    },
+    {
+      "epoch": 0.5947225981055481,
+      "grad_norm": 0.5721074155459656,
+      "learning_rate": 8.860834324540961e-06,
+      "loss": 0.2925,
+      "step": 879
+    },
+    {
+      "epoch": 0.5953991880920162,
+      "grad_norm": 0.5815263624516371,
+      "learning_rate": 8.85707932030663e-06,
+      "loss": 0.2922,
+      "step": 880
+    },
+    {
+      "epoch": 0.5960757780784844,
+      "grad_norm": 0.5629479816147226,
+      "learning_rate": 8.85331893591477e-06,
+      "loss": 0.3133,
+      "step": 881
+    },
+    {
+      "epoch": 0.5967523680649527,
+      "grad_norm": 0.5502266278493668,
+      "learning_rate": 8.849553176610661e-06,
+      "loss": 0.2971,
+      "step": 882
+    },
+    {
+      "epoch": 0.5974289580514208,
+      "grad_norm": 0.5969888282672955,
+      "learning_rate": 8.84578204764708e-06,
+      "loss": 0.2975,
+      "step": 883
+    },
+    {
+      "epoch": 0.5981055480378891,
+      "grad_norm": 0.541669977840359,
+      "learning_rate": 8.842005554284296e-06,
+      "loss": 0.2915,
+      "step": 884
+    },
+    {
+      "epoch": 0.5987821380243572,
+      "grad_norm": 0.5608832867074686,
+      "learning_rate": 8.838223701790057e-06,
+      "loss": 0.2996,
+      "step": 885
+    },
+    {
+      "epoch": 0.5994587280108254,
+      "grad_norm": 0.5846875342757033,
+      "learning_rate": 8.834436495439588e-06,
+      "loss": 0.2959,
+      "step": 886
+    },
+    {
+      "epoch": 0.6001353179972937,
+      "grad_norm": 0.549068486973641,
+      "learning_rate": 8.830643940515582e-06,
+      "loss": 0.2926,
+      "step": 887
+    },
+    {
+      "epoch": 0.6008119079837618,
+      "grad_norm": 0.5616363734685056,
+      "learning_rate": 8.826846042308195e-06,
+      "loss": 0.2928,
+      "step": 888
+    },
+    {
+      "epoch": 0.6014884979702301,
+      "grad_norm": 0.6154200576420003,
+      "learning_rate": 8.823042806115034e-06,
+      "loss": 0.2946,
+      "step": 889
+    },
+    {
+      "epoch": 0.6021650879566982,
+      "grad_norm": 0.5540532197422215,
+      "learning_rate": 8.819234237241148e-06,
+      "loss": 0.2914,
+      "step": 890
+    },
+    {
+      "epoch": 0.6028416779431665,
+      "grad_norm": 0.6076486499428294,
+      "learning_rate": 8.815420340999034e-06,
+      "loss": 0.3064,
+      "step": 891
+    },
+    {
+      "epoch": 0.6035182679296346,
+      "grad_norm": 0.5783488972052218,
+      "learning_rate": 8.81160112270861e-06,
+      "loss": 0.2926,
+      "step": 892
+    },
+    {
+      "epoch": 0.6041948579161028,
+      "grad_norm": 0.5996767957880053,
+      "learning_rate": 8.807776587697224e-06,
+      "loss": 0.294,
+      "step": 893
+    },
+    {
+      "epoch": 0.6048714479025711,
+      "grad_norm": 0.5370209758250004,
+      "learning_rate": 8.803946741299635e-06,
+      "loss": 0.2891,
+      "step": 894
+    },
+    {
+      "epoch": 0.6055480378890392,
+      "grad_norm": 0.610173278494983,
+      "learning_rate": 8.800111588858017e-06,
+      "loss": 0.2992,
+      "step": 895
+    },
+    {
+      "epoch": 0.6062246278755075,
+      "grad_norm": 0.6311058835560682,
+      "learning_rate": 8.796271135721944e-06,
+      "loss": 0.3007,
+      "step": 896
+    },
+    {
+      "epoch": 0.6069012178619756,
+      "grad_norm": 0.5830614682948251,
+      "learning_rate": 8.792425387248379e-06,
+      "loss": 0.3044,
+      "step": 897
+    },
+    {
+      "epoch": 0.6075778078484438,
+      "grad_norm": 0.614590152772781,
+      "learning_rate": 8.788574348801676e-06,
+      "loss": 0.3087,
+      "step": 898
+    },
+    {
+      "epoch": 0.6082543978349121,
+      "grad_norm": 0.6331056836499985,
+      "learning_rate": 8.784718025753562e-06,
+      "loss": 0.2948,
+      "step": 899
+    },
+    {
+      "epoch": 0.6089309878213802,
+      "grad_norm": 0.5934321954393875,
+      "learning_rate": 8.780856423483145e-06,
+      "loss": 0.3066,
+      "step": 900
+    },
+    {
+      "epoch": 0.6096075778078485,
+      "grad_norm": 0.6066227474665069,
+      "learning_rate": 8.77698954737689e-06,
+      "loss": 0.2965,
+      "step": 901
+    },
+    {
+      "epoch": 0.6102841677943166,
+      "grad_norm": 0.6098147123507792,
+      "learning_rate": 8.773117402828618e-06,
+      "loss": 0.3009,
+      "step": 902
+    },
+    {
+      "epoch": 0.6109607577807848,
+      "grad_norm": 0.5674155281311741,
+      "learning_rate": 8.769239995239502e-06,
+      "loss": 0.2976,
+      "step": 903
+    },
+    {
+      "epoch": 0.6116373477672531,
+      "grad_norm": 0.5840434383796752,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.2977,
+      "step": 904
+    },
+    {
+      "epoch": 0.6123139377537212,
+      "grad_norm": 0.5455609803124205,
+      "learning_rate": 8.761469412580126e-06,
+      "loss": 0.2853,
+      "step": 905
+    },
+    {
+      "epoch": 0.6129905277401895,
+      "grad_norm": 0.583139855868769,
+      "learning_rate": 8.757576248348883e-06,
+      "loss": 0.2963,
+      "step": 906
+    },
+    {
+      "epoch": 0.6136671177266576,
+      "grad_norm": 0.5655128743583371,
+      "learning_rate": 8.75367784275482e-06,
+      "loss": 0.2854,
+      "step": 907
+    },
+    {
+      "epoch": 0.6143437077131259,
+      "grad_norm": 0.5889356519708898,
+      "learning_rate": 8.74977420123574e-06,
+      "loss": 0.2914,
+      "step": 908
+    },
+    {
+      "epoch": 0.6150202976995941,
+      "grad_norm": 0.6380317067863059,
+      "learning_rate": 8.745865329236751e-06,
+      "loss": 0.3066,
+      "step": 909
+    },
+    {
+      "epoch": 0.6156968876860622,
+      "grad_norm": 0.5127644227319094,
+      "learning_rate": 8.741951232210254e-06,
+      "loss": 0.2904,
+      "step": 910
+    },
+    {
+      "epoch": 0.6163734776725305,
+      "grad_norm": 0.5158887082037273,
+      "learning_rate": 8.738031915615934e-06,
+      "loss": 0.2943,
+      "step": 911
+    },
+    {
+      "epoch": 0.6170500676589986,
+      "grad_norm": 0.5696676121204085,
+      "learning_rate": 8.734107384920771e-06,
+      "loss": 0.2992,
+      "step": 912
+    },
+    {
+      "epoch": 0.6177266576454669,
+      "grad_norm": 0.5556950593854647,
+      "learning_rate": 8.730177645599002e-06,
+      "loss": 0.2862,
+      "step": 913
+    },
+    {
+      "epoch": 0.618403247631935,
+      "grad_norm": 0.5354586414312311,
+      "learning_rate": 8.726242703132139e-06,
+      "loss": 0.3054,
+      "step": 914
+    },
+    {
+      "epoch": 0.6190798376184032,
+      "grad_norm": 0.5432946313182098,
+      "learning_rate": 8.72230256300895e-06,
+      "loss": 0.2988,
+      "step": 915
+    },
+    {
+      "epoch": 0.6197564276048715,
+      "grad_norm": 0.6281871375978919,
+      "learning_rate": 8.71835723072545e-06,
+      "loss": 0.3051,
+      "step": 916
+    },
+    {
+      "epoch": 0.6204330175913396,
+      "grad_norm": 0.5715784704005251,
+      "learning_rate": 8.7144067117849e-06,
+      "loss": 0.307,
+      "step": 917
+    },
+    {
+      "epoch": 0.6211096075778079,
+      "grad_norm": 0.5246534299864413,
+      "learning_rate": 8.710451011697794e-06,
+      "loss": 0.297,
+      "step": 918
+    },
+    {
+      "epoch": 0.621786197564276,
+      "grad_norm": 0.5903708105203975,
+      "learning_rate": 8.706490135981856e-06,
+      "loss": 0.3001,
+      "step": 919
+    },
+    {
+      "epoch": 0.6224627875507442,
+      "grad_norm": 0.5330170633632928,
+      "learning_rate": 8.702524090162023e-06,
+      "loss": 0.2977,
+      "step": 920
+    },
+    {
+      "epoch": 0.6231393775372125,
+      "grad_norm": 0.5325353830995546,
+      "learning_rate": 8.69855287977045e-06,
+      "loss": 0.2829,
+      "step": 921
+    },
+    {
+      "epoch": 0.6238159675236806,
+      "grad_norm": 0.5149976876175524,
+      "learning_rate": 8.694576510346493e-06,
+      "loss": 0.2894,
+      "step": 922
+    },
+    {
+      "epoch": 0.6244925575101489,
+      "grad_norm": 0.529587785260569,
+      "learning_rate": 8.690594987436705e-06,
+      "loss": 0.3048,
+      "step": 923
+    },
+    {
+      "epoch": 0.625169147496617,
+      "grad_norm": 0.543922999123769,
+      "learning_rate": 8.686608316594826e-06,
+      "loss": 0.3095,
+      "step": 924
+    },
+    {
+      "epoch": 0.6258457374830853,
+      "grad_norm": 0.5294163175558032,
+      "learning_rate": 8.68261650338178e-06,
+      "loss": 0.2939,
+      "step": 925
+    },
+    {
+      "epoch": 0.6265223274695535,
+      "grad_norm": 0.5407162539832638,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 0.3013,
+      "step": 926
+    },
+    {
+      "epoch": 0.6271989174560216,
+      "grad_norm": 0.5152658943251931,
+      "learning_rate": 8.674617472121726e-06,
+      "loss": 0.29,
+      "step": 927
+    },
+    {
+      "epoch": 0.6278755074424899,
+      "grad_norm": 0.5250839975619317,
+      "learning_rate": 8.670610265232398e-06,
+      "loss": 0.2912,
+      "step": 928
+    },
+    {
+      "epoch": 0.628552097428958,
+      "grad_norm": 0.5402857727809378,
+      "learning_rate": 8.666597938287242e-06,
+      "loss": 0.2908,
+      "step": 929
+    },
+    {
+      "epoch": 0.6292286874154263,
+      "grad_norm": 0.5425918452516918,
+      "learning_rate": 8.662580496882967e-06,
+      "loss": 0.2891,
+      "step": 930
+    },
+    {
+      "epoch": 0.6299052774018945,
+      "grad_norm": 0.5737760971714359,
+      "learning_rate": 8.658557946623416e-06,
+      "loss": 0.2975,
+      "step": 931
+    },
+    {
+      "epoch": 0.6305818673883626,
+      "grad_norm": 0.5279936681314105,
+      "learning_rate": 8.654530293119558e-06,
+      "loss": 0.2979,
+      "step": 932
+    },
+    {
+      "epoch": 0.6312584573748309,
+      "grad_norm": 0.548338976453665,
+      "learning_rate": 8.650497541989483e-06,
+      "loss": 0.2978,
+      "step": 933
+    },
+    {
+      "epoch": 0.631935047361299,
+      "grad_norm": 0.5618814566010114,
+      "learning_rate": 8.646459698858386e-06,
+      "loss": 0.3057,
+      "step": 934
+    },
+    {
+      "epoch": 0.6326116373477673,
+      "grad_norm": 0.5730552212773551,
+      "learning_rate": 8.64241676935857e-06,
+      "loss": 0.2971,
+      "step": 935
+    },
+    {
+      "epoch": 0.6332882273342354,
+      "grad_norm": 0.5380633791694546,
+      "learning_rate": 8.638368759129433e-06,
+      "loss": 0.2863,
+      "step": 936
+    },
+    {
+      "epoch": 0.6339648173207036,
+      "grad_norm": 0.5274462818779537,
+      "learning_rate": 8.634315673817453e-06,
+      "loss": 0.2878,
+      "step": 937
+    },
+    {
+      "epoch": 0.6346414073071719,
+      "grad_norm": 0.5366968240366689,
+      "learning_rate": 8.630257519076196e-06,
+      "loss": 0.2899,
+      "step": 938
+    },
+    {
+      "epoch": 0.63531799729364,
+      "grad_norm": 0.5154732587354501,
+      "learning_rate": 8.626194300566295e-06,
+      "loss": 0.3012,
+      "step": 939
+    },
+    {
+      "epoch": 0.6359945872801083,
+      "grad_norm": 0.5797730010571516,
+      "learning_rate": 8.622126023955446e-06,
+      "loss": 0.3205,
+      "step": 940
+    },
+    {
+      "epoch": 0.6366711772665764,
+      "grad_norm": 0.5295098767241785,
+      "learning_rate": 8.6180526949184e-06,
+      "loss": 0.299,
+      "step": 941
+    },
+    {
+      "epoch": 0.6373477672530447,
+      "grad_norm": 0.5264280635334435,
+      "learning_rate": 8.613974319136959e-06,
+      "loss": 0.2988,
+      "step": 942
+    },
+    {
+      "epoch": 0.6380243572395129,
+      "grad_norm": 0.5544848502742725,
+      "learning_rate": 8.60989090229996e-06,
+      "loss": 0.3094,
+      "step": 943
+    },
+    {
+      "epoch": 0.638700947225981,
+      "grad_norm": 0.5230611221981714,
+      "learning_rate": 8.605802450103276e-06,
+      "loss": 0.2888,
+      "step": 944
+    },
+    {
+      "epoch": 0.6393775372124493,
+      "grad_norm": 0.5164077330353328,
+      "learning_rate": 8.601708968249797e-06,
+      "loss": 0.2815,
+      "step": 945
+    },
+    {
+      "epoch": 0.6400541271989174,
+      "grad_norm": 0.5179266808820038,
+      "learning_rate": 8.597610462449441e-06,
+      "loss": 0.2995,
+      "step": 946
+    },
+    {
+      "epoch": 0.6407307171853857,
+      "grad_norm": 0.5494991898356408,
+      "learning_rate": 8.59350693841912e-06,
+      "loss": 0.2934,
+      "step": 947
+    },
+    {
+      "epoch": 0.6414073071718539,
+      "grad_norm": 0.5232006422161414,
+      "learning_rate": 8.589398401882755e-06,
+      "loss": 0.2955,
+      "step": 948
+    },
+    {
+      "epoch": 0.642083897158322,
+      "grad_norm": 0.5254649383149836,
+      "learning_rate": 8.585284858571254e-06,
+      "loss": 0.2974,
+      "step": 949
+    },
+    {
+      "epoch": 0.6427604871447903,
+      "grad_norm": 0.5184845001426555,
+      "learning_rate": 8.581166314222512e-06,
+      "loss": 0.3023,
+      "step": 950
+    },
+    {
+      "epoch": 0.6434370771312584,
+      "grad_norm": 0.5374625988916433,
+      "learning_rate": 8.577042774581397e-06,
+      "loss": 0.3,
+      "step": 951
+    },
+    {
+      "epoch": 0.6441136671177267,
+      "grad_norm": 0.6209827432043309,
+      "learning_rate": 8.572914245399748e-06,
+      "loss": 0.3006,
+      "step": 952
+    },
+    {
+      "epoch": 0.6447902571041949,
+      "grad_norm": 0.525076676495374,
+      "learning_rate": 8.568780732436361e-06,
+      "loss": 0.2992,
+      "step": 953
+    },
+    {
+      "epoch": 0.645466847090663,
+      "grad_norm": 0.5249064143812362,
+      "learning_rate": 8.564642241456986e-06,
+      "loss": 0.2844,
+      "step": 954
+    },
+    {
+      "epoch": 0.6461434370771313,
+      "grad_norm": 0.5340259784921277,
+      "learning_rate": 8.560498778234316e-06,
+      "loss": 0.2922,
+      "step": 955
+    },
+    {
+      "epoch": 0.6468200270635994,
+      "grad_norm": 0.5313735589045534,
+      "learning_rate": 8.556350348547978e-06,
+      "loss": 0.2911,
+      "step": 956
+    },
+    {
+      "epoch": 0.6474966170500677,
+      "grad_norm": 0.5678821426820465,
+      "learning_rate": 8.552196958184526e-06,
+      "loss": 0.3153,
+      "step": 957
+    },
+    {
+      "epoch": 0.6481732070365359,
+      "grad_norm": 0.5368873625000248,
+      "learning_rate": 8.54803861293744e-06,
+      "loss": 0.2987,
+      "step": 958
+    },
+    {
+      "epoch": 0.648849797023004,
+      "grad_norm": 0.5562363825493358,
+      "learning_rate": 8.543875318607102e-06,
+      "loss": 0.2952,
+      "step": 959
+    },
+    {
+      "epoch": 0.6495263870094723,
+      "grad_norm": 0.5440556472241357,
+      "learning_rate": 8.539707081000808e-06,
+      "loss": 0.3025,
+      "step": 960
+    },
+    {
+      "epoch": 0.6502029769959404,
+      "grad_norm": 0.5116926049397693,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.2875,
+      "step": 961
+    },
+    {
+      "epoch": 0.6508795669824087,
+      "grad_norm": 0.5396565454823112,
+      "learning_rate": 8.531355799223968e-06,
+      "loss": 0.2984,
+      "step": 962
+    },
+    {
+      "epoch": 0.6515561569688768,
+      "grad_norm": 0.5252804709441166,
+      "learning_rate": 8.52717276670245e-06,
+      "loss": 0.2897,
+      "step": 963
+    },
+    {
+      "epoch": 0.652232746955345,
+      "grad_norm": 0.5269714055301262,
+      "learning_rate": 8.522984814203006e-06,
+      "loss": 0.2975,
+      "step": 964
+    },
+    {
+      "epoch": 0.6529093369418133,
+      "grad_norm": 0.5574648316291245,
+      "learning_rate": 8.51879194756732e-06,
+      "loss": 0.3016,
+      "step": 965
+    },
+    {
+      "epoch": 0.6535859269282814,
+      "grad_norm": 0.539040678558386,
+      "learning_rate": 8.514594172643934e-06,
+      "loss": 0.3097,
+      "step": 966
+    },
+    {
+      "epoch": 0.6542625169147497,
+      "grad_norm": 0.5316611020889722,
+      "learning_rate": 8.510391495288234e-06,
+      "loss": 0.3054,
+      "step": 967
+    },
+    {
+      "epoch": 0.6549391069012178,
+      "grad_norm": 0.5386033463955999,
+      "learning_rate": 8.506183921362443e-06,
+      "loss": 0.2959,
+      "step": 968
+    },
+    {
+      "epoch": 0.6556156968876861,
+      "grad_norm": 0.5156482433367408,
+      "learning_rate": 8.50197145673562e-06,
+      "loss": 0.2944,
+      "step": 969
+    },
+    {
+      "epoch": 0.6562922868741543,
+      "grad_norm": 0.522401333444416,
+      "learning_rate": 8.497754107283637e-06,
+      "loss": 0.2788,
+      "step": 970
+    },
+    {
+      "epoch": 0.6569688768606224,
+      "grad_norm": 0.5098290725412649,
+      "learning_rate": 8.49353187888919e-06,
+      "loss": 0.2991,
+      "step": 971
+    },
+    {
+      "epoch": 0.6576454668470907,
+      "grad_norm": 0.5416123728552092,
+      "learning_rate": 8.489304777441772e-06,
+      "loss": 0.3012,
+      "step": 972
+    },
+    {
+      "epoch": 0.6583220568335588,
+      "grad_norm": 0.5423262190753975,
+      "learning_rate": 8.485072808837673e-06,
+      "loss": 0.3016,
+      "step": 973
+    },
+    {
+      "epoch": 0.6589986468200271,
+      "grad_norm": 0.5171719531820111,
+      "learning_rate": 8.480835978979983e-06,
+      "loss": 0.293,
+      "step": 974
+    },
+    {
+      "epoch": 0.6596752368064953,
+      "grad_norm": 0.5857469820727066,
+      "learning_rate": 8.476594293778561e-06,
+      "loss": 0.2992,
+      "step": 975
+    },
+    {
+      "epoch": 0.6603518267929634,
+      "grad_norm": 0.5259210406243775,
+      "learning_rate": 8.472347759150044e-06,
+      "loss": 0.2813,
+      "step": 976
+    },
+    {
+      "epoch": 0.6610284167794317,
+      "grad_norm": 0.5331196330280126,
+      "learning_rate": 8.468096381017832e-06,
+      "loss": 0.3019,
+      "step": 977
+    },
+    {
+      "epoch": 0.6617050067658998,
+      "grad_norm": 0.5372689843755181,
+      "learning_rate": 8.463840165312083e-06,
+      "loss": 0.3134,
+      "step": 978
+    },
+    {
+      "epoch": 0.6623815967523681,
+      "grad_norm": 0.560476251828771,
+      "learning_rate": 8.4595791179697e-06,
+      "loss": 0.3039,
+      "step": 979
+    },
+    {
+      "epoch": 0.6630581867388363,
+      "grad_norm": 0.5391329312969535,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.2938,
+      "step": 980
+    },
+    {
+      "epoch": 0.6637347767253045,
+      "grad_norm": 0.5439991373773334,
+      "learning_rate": 8.451042552156335e-06,
+      "loss": 0.3048,
+      "step": 981
+    },
+    {
+      "epoch": 0.6644113667117727,
+      "grad_norm": 0.5416045149888572,
+      "learning_rate": 8.446767045592829e-06,
+      "loss": 0.2759,
+      "step": 982
+    },
+    {
+      "epoch": 0.6650879566982408,
+      "grad_norm": 0.5615542248779019,
+      "learning_rate": 8.442486731207622e-06,
+      "loss": 0.2773,
+      "step": 983
+    },
+    {
+      "epoch": 0.6657645466847091,
+      "grad_norm": 0.5576099393303231,
+      "learning_rate": 8.438201614971227e-06,
+      "loss": 0.3029,
+      "step": 984
+    },
+    {
+      "epoch": 0.6664411366711772,
+      "grad_norm": 0.5389154467680758,
+      "learning_rate": 8.433911702860864e-06,
+      "loss": 0.29,
+      "step": 985
+    },
+    {
+      "epoch": 0.6671177266576455,
+      "grad_norm": 0.5574919249468511,
+      "learning_rate": 8.429617000860441e-06,
+      "loss": 0.2974,
+      "step": 986
+    },
+    {
+      "epoch": 0.6677943166441137,
+      "grad_norm": 0.5796540276401466,
+      "learning_rate": 8.425317514960547e-06,
+      "loss": 0.2963,
+      "step": 987
+    },
+    {
+      "epoch": 0.6684709066305818,
+      "grad_norm": 0.5351848293985333,
+      "learning_rate": 8.421013251158437e-06,
+      "loss": 0.2976,
+      "step": 988
+    },
+    {
+      "epoch": 0.6691474966170501,
+      "grad_norm": 0.5428000510582028,
+      "learning_rate": 8.416704215458042e-06,
+      "loss": 0.3035,
+      "step": 989
+    },
+    {
+      "epoch": 0.6698240866035182,
+      "grad_norm": 0.5600948288055342,
+      "learning_rate": 8.412390413869944e-06,
+      "loss": 0.2966,
+      "step": 990
+    },
+    {
+      "epoch": 0.6705006765899865,
+      "grad_norm": 0.5680544938659107,
+      "learning_rate": 8.40807185241137e-06,
+      "loss": 0.2963,
+      "step": 991
+    },
+    {
+      "epoch": 0.6711772665764547,
+      "grad_norm": 0.532289680386378,
+      "learning_rate": 8.40374853710619e-06,
+      "loss": 0.293,
+      "step": 992
+    },
+    {
+      "epoch": 0.6718538565629228,
+      "grad_norm": 0.5713592461221124,
+      "learning_rate": 8.399420473984905e-06,
+      "loss": 0.2907,
+      "step": 993
+    },
+    {
+      "epoch": 0.6725304465493911,
+      "grad_norm": 0.551934210226489,
+      "learning_rate": 8.395087669084638e-06,
+      "loss": 0.2981,
+      "step": 994
+    },
+    {
+      "epoch": 0.6732070365358592,
+      "grad_norm": 0.5149809941864839,
+      "learning_rate": 8.390750128449124e-06,
+      "loss": 0.2918,
+      "step": 995
+    },
+    {
+      "epoch": 0.6738836265223275,
+      "grad_norm": 0.5447166121443857,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.2988,
+      "step": 996
+    },
+    {
+      "epoch": 0.6745602165087957,
+      "grad_norm": 0.5435214989244523,
+      "learning_rate": 8.382060864180326e-06,
+      "loss": 0.2911,
+      "step": 997
+    },
+    {
+      "epoch": 0.6752368064952639,
+      "grad_norm": 0.5448303377284845,
+      "learning_rate": 8.377709152667513e-06,
+      "loss": 0.2996,
+      "step": 998
+    },
+    {
+      "epoch": 0.6759133964817321,
+      "grad_norm": 0.5660733275463123,
+      "learning_rate": 8.373352729660373e-06,
+      "loss": 0.3105,
+      "step": 999
+    },
+    {
+      "epoch": 0.6765899864682002,
+      "grad_norm": 0.5340001421569021,
+      "learning_rate": 8.36899160123559e-06,
+      "loss": 0.2942,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6772665764546685,
+      "grad_norm": 0.5521444282348594,
+      "learning_rate": 8.36462577347641e-06,
+      "loss": 0.293,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6779431664411367,
+      "grad_norm": 0.5488152587777889,
+      "learning_rate": 8.36025525247263e-06,
+      "loss": 0.3056,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6786197564276049,
+      "grad_norm": 0.533612040544712,
+      "learning_rate": 8.355880044320599e-06,
+      "loss": 0.2953,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6792963464140731,
+      "grad_norm": 0.5279988037569934,
+      "learning_rate": 8.3515001551232e-06,
+      "loss": 0.2923,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6799729364005412,
+      "grad_norm": 0.5250469116467068,
+      "learning_rate": 8.347115590989845e-06,
+      "loss": 0.2933,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6806495263870095,
+      "grad_norm": 0.5710056400701043,
+      "learning_rate": 8.342726358036473e-06,
+      "loss": 0.298,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6813261163734776,
+      "grad_norm": 0.543967476530912,
+      "learning_rate": 8.338332462385528e-06,
+      "loss": 0.3052,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6820027063599459,
+      "grad_norm": 0.5509519454658223,
+      "learning_rate": 8.333933910165964e-06,
+      "loss": 0.3006,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6826792963464141,
+      "grad_norm": 0.5547968289049332,
+      "learning_rate": 8.329530707513223e-06,
+      "loss": 0.3119,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6833558863328822,
+      "grad_norm": 0.5192124680028315,
+      "learning_rate": 8.325122860569241e-06,
+      "loss": 0.2843,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6840324763193505,
+      "grad_norm": 0.5171186595967231,
+      "learning_rate": 8.320710375482432e-06,
+      "loss": 0.2874,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6847090663058186,
+      "grad_norm": 0.5513822447250257,
+      "learning_rate": 8.316293258407673e-06,
+      "loss": 0.292,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6853856562922869,
+      "grad_norm": 0.5511355785881631,
+      "learning_rate": 8.31187151550631e-06,
+      "loss": 0.3028,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6860622462787551,
+      "grad_norm": 0.5076273833065873,
+      "learning_rate": 8.307445152946133e-06,
+      "loss": 0.2968,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6867388362652233,
+      "grad_norm": 0.5089404514862232,
+      "learning_rate": 8.303014176901384e-06,
+      "loss": 0.2787,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6874154262516915,
+      "grad_norm": 0.5421874526537435,
+      "learning_rate": 8.298578593552737e-06,
+      "loss": 0.2919,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6880920162381596,
+      "grad_norm": 0.5428987435946161,
+      "learning_rate": 8.29413840908729e-06,
+      "loss": 0.299,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6887686062246279,
+      "grad_norm": 0.5224815815848058,
+      "learning_rate": 8.289693629698564e-06,
+      "loss": 0.2993,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6894451962110961,
+      "grad_norm": 0.5243043811755902,
+      "learning_rate": 8.285244261586485e-06,
+      "loss": 0.2897,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6901217861975643,
+      "grad_norm": 0.5253772365927546,
+      "learning_rate": 8.280790310957382e-06,
+      "loss": 0.2842,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6907983761840325,
+      "grad_norm": 0.6206955120452043,
+      "learning_rate": 8.276331784023976e-06,
+      "loss": 0.3126,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6914749661705006,
+      "grad_norm": 0.5480447453219145,
+      "learning_rate": 8.271868687005371e-06,
+      "loss": 0.2927,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6921515561569689,
+      "grad_norm": 0.5335892150757401,
+      "learning_rate": 8.267401026127046e-06,
+      "loss": 0.2982,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6928281461434371,
+      "grad_norm": 0.5399353291463875,
+      "learning_rate": 8.262928807620843e-06,
+      "loss": 0.2816,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6935047361299053,
+      "grad_norm": 0.5406026353468125,
+      "learning_rate": 8.258452037724968e-06,
+      "loss": 0.2927,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6941813261163735,
+      "grad_norm": 0.5130967639832842,
+      "learning_rate": 8.253970722683968e-06,
+      "loss": 0.2902,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6948579161028416,
+      "grad_norm": 0.5559955022016767,
+      "learning_rate": 8.249484868748737e-06,
+      "loss": 0.2925,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6955345060893099,
+      "grad_norm": 0.509800894039279,
+      "learning_rate": 8.244994482176495e-06,
+      "loss": 0.2789,
+      "step": 1028
+    },
+    {
+      "epoch": 0.696211096075778,
+      "grad_norm": 0.5212861389529568,
+      "learning_rate": 8.240499569230787e-06,
+      "loss": 0.2938,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6968876860622463,
+      "grad_norm": 0.5204687393689884,
+      "learning_rate": 8.236000136181468e-06,
+      "loss": 0.3085,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6975642760487145,
+      "grad_norm": 0.533802315522298,
+      "learning_rate": 8.231496189304704e-06,
+      "loss": 0.2987,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6982408660351827,
+      "grad_norm": 0.5261387497361262,
+      "learning_rate": 8.226987734882956e-06,
+      "loss": 0.2962,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6989174560216509,
+      "grad_norm": 0.5249246727129384,
+      "learning_rate": 8.222474779204968e-06,
+      "loss": 0.2975,
+      "step": 1033
+    },
+    {
+      "epoch": 0.699594046008119,
+      "grad_norm": 0.5248481902810384,
+      "learning_rate": 8.217957328565765e-06,
+      "loss": 0.2928,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7002706359945873,
+      "grad_norm": 0.5166426592235464,
+      "learning_rate": 8.213435389266641e-06,
+      "loss": 0.294,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7009472259810555,
+      "grad_norm": 0.5279929934121549,
+      "learning_rate": 8.208908967615159e-06,
+      "loss": 0.2975,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7016238159675237,
+      "grad_norm": 0.5518023009235741,
+      "learning_rate": 8.204378069925121e-06,
+      "loss": 0.2878,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7023004059539919,
+      "grad_norm": 0.5267589619726432,
+      "learning_rate": 8.199842702516584e-06,
+      "loss": 0.3034,
+      "step": 1038
+    },
+    {
+      "epoch": 0.70297699594046,
+      "grad_norm": 0.5313608544924104,
+      "learning_rate": 8.195302871715831e-06,
+      "loss": 0.2946,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7036535859269283,
+      "grad_norm": 0.5392427329442301,
+      "learning_rate": 8.190758583855379e-06,
+      "loss": 0.2959,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7043301759133965,
+      "grad_norm": 0.5639766168560917,
+      "learning_rate": 8.186209845273954e-06,
+      "loss": 0.2903,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7050067658998647,
+      "grad_norm": 0.5549981994386375,
+      "learning_rate": 8.181656662316498e-06,
+      "loss": 0.2995,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7056833558863329,
+      "grad_norm": 0.5112585363358638,
+      "learning_rate": 8.177099041334146e-06,
+      "loss": 0.2913,
+      "step": 1043
+    },
+    {
+      "epoch": 0.706359945872801,
+      "grad_norm": 0.5234437867013138,
+      "learning_rate": 8.172536988684227e-06,
+      "loss": 0.2968,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7070365358592693,
+      "grad_norm": 0.5458434788550469,
+      "learning_rate": 8.167970510730254e-06,
+      "loss": 0.2926,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7077131258457375,
+      "grad_norm": 0.5494487509202765,
+      "learning_rate": 8.163399613841903e-06,
+      "loss": 0.2887,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7083897158322057,
+      "grad_norm": 0.5230076996486837,
+      "learning_rate": 8.158824304395026e-06,
+      "loss": 0.2945,
+      "step": 1047
+    },
+    {
+      "epoch": 0.7090663058186739,
+      "grad_norm": 0.526115670768814,
+      "learning_rate": 8.15424458877162e-06,
+      "loss": 0.2757,
+      "step": 1048
+    },
+    {
+      "epoch": 0.709742895805142,
+      "grad_norm": 0.5296250588452163,
+      "learning_rate": 8.149660473359836e-06,
+      "loss": 0.2893,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7104194857916103,
+      "grad_norm": 0.5569970915884735,
+      "learning_rate": 8.145071964553956e-06,
+      "loss": 0.2992,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7110960757780784,
+      "grad_norm": 0.5172263232911845,
+      "learning_rate": 8.140479068754396e-06,
+      "loss": 0.2978,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7117726657645467,
+      "grad_norm": 0.5725194825986271,
+      "learning_rate": 8.135881792367686e-06,
+      "loss": 0.3079,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7124492557510149,
+      "grad_norm": 0.5452487794197558,
+      "learning_rate": 8.131280141806468e-06,
+      "loss": 0.2972,
+      "step": 1053
+    },
+    {
+      "epoch": 0.713125845737483,
+      "grad_norm": 0.539055022522886,
+      "learning_rate": 8.12667412348949e-06,
+      "loss": 0.3082,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7138024357239513,
+      "grad_norm": 0.582421856325855,
+      "learning_rate": 8.122063743841581e-06,
+      "loss": 0.305,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7144790257104194,
+      "grad_norm": 0.5004060240835847,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.2871,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7151556156968877,
+      "grad_norm": 0.540721311651463,
+      "learning_rate": 8.112829926282743e-06,
+      "loss": 0.3023,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7158322056833559,
+      "grad_norm": 0.5440597734388322,
+      "learning_rate": 8.108206501251868e-06,
+      "loss": 0.2903,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7165087956698241,
+      "grad_norm": 0.5436975557375594,
+      "learning_rate": 8.103578740650157e-06,
+      "loss": 0.2894,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7171853856562923,
+      "grad_norm": 0.4934490186369908,
+      "learning_rate": 8.098946650932776e-06,
+      "loss": 0.2812,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7178619756427604,
+      "grad_norm": 0.523668331431008,
+      "learning_rate": 8.094310238560926e-06,
+      "loss": 0.3005,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7185385656292287,
+      "grad_norm": 0.5362773676503786,
+      "learning_rate": 8.089669510001843e-06,
+      "loss": 0.2965,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7192151556156969,
+      "grad_norm": 0.5485126648187933,
+      "learning_rate": 8.085024471728778e-06,
+      "loss": 0.2984,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7198917456021651,
+      "grad_norm": 0.5310407272780361,
+      "learning_rate": 8.080375130220995e-06,
+      "loss": 0.3035,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7205683355886333,
+      "grad_norm": 0.5367435691275575,
+      "learning_rate": 8.075721491963761e-06,
+      "loss": 0.2893,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7212449255751014,
+      "grad_norm": 0.5443316214667175,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 0.3002,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7219215155615697,
+      "grad_norm": 0.4883922857700187,
+      "learning_rate": 8.066401351171973e-06,
+      "loss": 0.2974,
+      "step": 1067
+    },
+    {
+      "epoch": 0.7225981055480379,
+      "grad_norm": 0.5472130101752405,
+      "learning_rate": 8.061734861637883e-06,
+      "loss": 0.3097,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7232746955345061,
+      "grad_norm": 0.5435270179476893,
+      "learning_rate": 8.057064101355255e-06,
+      "loss": 0.2928,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7239512855209743,
+      "grad_norm": 0.5252100617794201,
+      "learning_rate": 8.052389076839233e-06,
+      "loss": 0.2904,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7246278755074425,
+      "grad_norm": 0.5347972110536291,
+      "learning_rate": 8.047709794610907e-06,
+      "loss": 0.302,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7253044654939107,
+      "grad_norm": 0.533262795852714,
+      "learning_rate": 8.043026261197312e-06,
+      "loss": 0.2961,
+      "step": 1072
+    },
+    {
+      "epoch": 0.725981055480379,
+      "grad_norm": 0.5152705644073176,
+      "learning_rate": 8.038338483131408e-06,
+      "loss": 0.303,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7266576454668471,
+      "grad_norm": 0.5725509774533468,
+      "learning_rate": 8.033646466952072e-06,
+      "loss": 0.3033,
+      "step": 1074
+    },
+    {
+      "epoch": 0.7273342354533153,
+      "grad_norm": 0.5344593632114969,
+      "learning_rate": 8.0289502192041e-06,
+      "loss": 0.2934,
+      "step": 1075
+    },
+    {
+      "epoch": 0.7280108254397835,
+      "grad_norm": 0.5191924107952616,
+      "learning_rate": 8.024249746438189e-06,
+      "loss": 0.2924,
+      "step": 1076
+    },
+    {
+      "epoch": 0.7286874154262517,
+      "grad_norm": 0.5644834015318716,
+      "learning_rate": 8.019545055210924e-06,
+      "loss": 0.2904,
+      "step": 1077
+    },
+    {
+      "epoch": 0.7293640054127198,
+      "grad_norm": 0.5649435067358575,
+      "learning_rate": 8.014836152084784e-06,
+      "loss": 0.2886,
+      "step": 1078
+    },
+    {
+      "epoch": 0.7300405953991881,
+      "grad_norm": 0.5339926077812096,
+      "learning_rate": 8.010123043628113e-06,
+      "loss": 0.2923,
+      "step": 1079
+    },
+    {
+      "epoch": 0.7307171853856563,
+      "grad_norm": 0.5491575552060038,
+      "learning_rate": 8.005405736415127e-06,
+      "loss": 0.2824,
+      "step": 1080
+    },
+    {
+      "epoch": 0.7313937753721245,
+      "grad_norm": 0.5523354344710706,
+      "learning_rate": 8.000684237025894e-06,
+      "loss": 0.2921,
+      "step": 1081
+    },
+    {
+      "epoch": 0.7320703653585927,
+      "grad_norm": 0.5120799255972635,
+      "learning_rate": 7.995958552046338e-06,
+      "loss": 0.3027,
+      "step": 1082
+    },
+    {
+      "epoch": 0.7327469553450608,
+      "grad_norm": 0.5546552336495316,
+      "learning_rate": 7.991228688068213e-06,
+      "loss": 0.3052,
+      "step": 1083
+    },
+    {
+      "epoch": 0.7334235453315291,
+      "grad_norm": 0.5467080532374307,
+      "learning_rate": 7.986494651689104e-06,
+      "loss": 0.3043,
+      "step": 1084
+    },
+    {
+      "epoch": 0.7341001353179973,
+      "grad_norm": 0.5114169433258975,
+      "learning_rate": 7.98175644951242e-06,
+      "loss": 0.2947,
+      "step": 1085
+    },
+    {
+      "epoch": 0.7347767253044655,
+      "grad_norm": 0.5530938884406995,
+      "learning_rate": 7.977014088147375e-06,
+      "loss": 0.2904,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7354533152909337,
+      "grad_norm": 0.540446351738585,
+      "learning_rate": 7.972267574208991e-06,
+      "loss": 0.2971,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7361299052774019,
+      "grad_norm": 0.5505516501140981,
+      "learning_rate": 7.967516914318075e-06,
+      "loss": 0.2984,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7368064952638701,
+      "grad_norm": 0.5212713570424431,
+      "learning_rate": 7.962762115101224e-06,
+      "loss": 0.301,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7374830852503383,
+      "grad_norm": 0.5076388059912484,
+      "learning_rate": 7.958003183190804e-06,
+      "loss": 0.2809,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7381596752368065,
+      "grad_norm": 0.5580023702238869,
+      "learning_rate": 7.953240125224948e-06,
+      "loss": 0.2989,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7388362652232747,
+      "grad_norm": 0.5557481325271645,
+      "learning_rate": 7.948472947847546e-06,
+      "loss": 0.3045,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7395128552097429,
+      "grad_norm": 0.5705871502987903,
+      "learning_rate": 7.943701657708227e-06,
+      "loss": 0.3002,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7401894451962111,
+      "grad_norm": 0.5518188323100711,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.2955,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7408660351826793,
+      "grad_norm": 0.5208006275660501,
+      "learning_rate": 7.934146765771061e-06,
+      "loss": 0.2886,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7415426251691475,
+      "grad_norm": 0.5259356524892155,
+      "learning_rate": 7.929363177301124e-06,
+      "loss": 0.2933,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7422192151556157,
+      "grad_norm": 0.5562685133080384,
+      "learning_rate": 7.924575502725086e-06,
+      "loss": 0.2807,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7428958051420839,
+      "grad_norm": 0.5381743944612812,
+      "learning_rate": 7.919783748721169e-06,
+      "loss": 0.3037,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7435723951285521,
+      "grad_norm": 0.5301053872762932,
+      "learning_rate": 7.914987921973287e-06,
+      "loss": 0.3005,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7442489851150202,
+      "grad_norm": 0.547734865478386,
+      "learning_rate": 7.910188029171039e-06,
+      "loss": 0.2878,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7449255751014885,
+      "grad_norm": 0.5352793524715399,
+      "learning_rate": 7.905384077009693e-06,
+      "loss": 0.2935,
+      "step": 1101
+    },
+    {
+      "epoch": 0.7456021650879567,
+      "grad_norm": 0.5287116209096365,
+      "learning_rate": 7.900576072190177e-06,
+      "loss": 0.2925,
+      "step": 1102
+    },
+    {
+      "epoch": 0.7462787550744249,
+      "grad_norm": 0.5347888359231753,
+      "learning_rate": 7.895764021419074e-06,
+      "loss": 0.2861,
+      "step": 1103
+    },
+    {
+      "epoch": 0.7469553450608931,
+      "grad_norm": 0.5047590506312314,
+      "learning_rate": 7.890947931408614e-06,
+      "loss": 0.2899,
+      "step": 1104
+    },
+    {
+      "epoch": 0.7476319350473613,
+      "grad_norm": 0.5078279751277103,
+      "learning_rate": 7.886127808876655e-06,
+      "loss": 0.2978,
+      "step": 1105
+    },
+    {
+      "epoch": 0.7483085250338295,
+      "grad_norm": 0.5256404323897009,
+      "learning_rate": 7.881303660546684e-06,
+      "loss": 0.2833,
+      "step": 1106
+    },
+    {
+      "epoch": 0.7489851150202977,
+      "grad_norm": 0.5315817933518733,
+      "learning_rate": 7.876475493147805e-06,
+      "loss": 0.3065,
+      "step": 1107
+    },
+    {
+      "epoch": 0.7496617050067659,
+      "grad_norm": 0.5640363044343041,
+      "learning_rate": 7.871643313414718e-06,
+      "loss": 0.2973,
+      "step": 1108
+    },
+    {
+      "epoch": 0.7503382949932341,
+      "grad_norm": 0.5333395271036443,
+      "learning_rate": 7.866807128087736e-06,
+      "loss": 0.2899,
+      "step": 1109
+    },
+    {
+      "epoch": 0.7510148849797023,
+      "grad_norm": 0.5371560751979332,
+      "learning_rate": 7.861966943912746e-06,
+      "loss": 0.2879,
+      "step": 1110
+    },
+    {
+      "epoch": 0.7516914749661705,
+      "grad_norm": 0.5405355512479946,
+      "learning_rate": 7.857122767641218e-06,
+      "loss": 0.2962,
+      "step": 1111
+    },
+    {
+      "epoch": 0.7523680649526387,
+      "grad_norm": 0.5548844929082567,
+      "learning_rate": 7.852274606030191e-06,
+      "loss": 0.2959,
+      "step": 1112
+    },
+    {
+      "epoch": 0.7530446549391069,
+      "grad_norm": 0.5440059215691581,
+      "learning_rate": 7.84742246584226e-06,
+      "loss": 0.3062,
+      "step": 1113
+    },
+    {
+      "epoch": 0.7537212449255751,
+      "grad_norm": 0.5434463015721115,
+      "learning_rate": 7.842566353845575e-06,
+      "loss": 0.2932,
+      "step": 1114
+    },
+    {
+      "epoch": 0.7543978349120433,
+      "grad_norm": 0.5789816889015749,
+      "learning_rate": 7.837706276813819e-06,
+      "loss": 0.2915,
+      "step": 1115
+    },
+    {
+      "epoch": 0.7550744248985115,
+      "grad_norm": 0.5733457402745237,
+      "learning_rate": 7.832842241526212e-06,
+      "loss": 0.2995,
+      "step": 1116
+    },
+    {
+      "epoch": 0.7557510148849798,
+      "grad_norm": 0.5358341768446587,
+      "learning_rate": 7.827974254767493e-06,
+      "loss": 0.2884,
+      "step": 1117
+    },
+    {
+      "epoch": 0.7564276048714479,
+      "grad_norm": 0.5504485430255516,
+      "learning_rate": 7.82310232332791e-06,
+      "loss": 0.3012,
+      "step": 1118
+    },
+    {
+      "epoch": 0.7571041948579161,
+      "grad_norm": 0.5458930400213989,
+      "learning_rate": 7.81822645400322e-06,
+      "loss": 0.2871,
+      "step": 1119
+    },
+    {
+      "epoch": 0.7577807848443843,
+      "grad_norm": 0.5367455440374713,
+      "learning_rate": 7.813346653594667e-06,
+      "loss": 0.2907,
+      "step": 1120
+    },
+    {
+      "epoch": 0.7584573748308525,
+      "grad_norm": 0.5066684588232969,
+      "learning_rate": 7.808462928908982e-06,
+      "loss": 0.2803,
+      "step": 1121
+    },
+    {
+      "epoch": 0.7591339648173207,
+      "grad_norm": 0.49062857501734064,
+      "learning_rate": 7.803575286758365e-06,
+      "loss": 0.2767,
+      "step": 1122
+    },
+    {
+      "epoch": 0.7598105548037889,
+      "grad_norm": 0.5022734915337267,
+      "learning_rate": 7.798683733960483e-06,
+      "loss": 0.2806,
+      "step": 1123
+    },
+    {
+      "epoch": 0.7604871447902571,
+      "grad_norm": 0.5337046517800563,
+      "learning_rate": 7.793788277338464e-06,
+      "loss": 0.2985,
+      "step": 1124
+    },
+    {
+      "epoch": 0.7611637347767253,
+      "grad_norm": 0.5174763857311346,
+      "learning_rate": 7.788888923720872e-06,
+      "loss": 0.3028,
+      "step": 1125
+    },
+    {
+      "epoch": 0.7618403247631935,
+      "grad_norm": 0.5320571852959397,
+      "learning_rate": 7.78398567994171e-06,
+      "loss": 0.2949,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7625169147496617,
+      "grad_norm": 0.5282129352595681,
+      "learning_rate": 7.77907855284041e-06,
+      "loss": 0.2926,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7631935047361299,
+      "grad_norm": 0.4827338220966895,
+      "learning_rate": 7.774167549261817e-06,
+      "loss": 0.2746,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7638700947225981,
+      "grad_norm": 0.5167683418544465,
+      "learning_rate": 7.769252676056186e-06,
+      "loss": 0.296,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7645466847090663,
+      "grad_norm": 0.5514778596653243,
+      "learning_rate": 7.764333940079169e-06,
+      "loss": 0.2928,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7652232746955345,
+      "grad_norm": 0.5156760034436229,
+      "learning_rate": 7.759411348191806e-06,
+      "loss": 0.2901,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7658998646820027,
+      "grad_norm": 0.541711479217456,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.2981,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7665764546684709,
+      "grad_norm": 0.5159824945179852,
+      "learning_rate": 7.74955462415708e-06,
+      "loss": 0.2891,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7672530446549392,
+      "grad_norm": 0.5104872979200005,
+      "learning_rate": 7.744620505758652e-06,
+      "loss": 0.2828,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7679296346414073,
+      "grad_norm": 0.5153821998867437,
+      "learning_rate": 7.739682558947725e-06,
+      "loss": 0.2841,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7686062246278755,
+      "grad_norm": 0.5021721357135888,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.2877,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7692828146143437,
+      "grad_norm": 0.5082659176315707,
+      "learning_rate": 7.729795207645051e-06,
+      "loss": 0.2869,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7699594046008119,
+      "grad_norm": 0.5644631129686093,
+      "learning_rate": 7.724845816944962e-06,
+      "loss": 0.2797,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7706359945872802,
+      "grad_norm": 0.542377641138436,
+      "learning_rate": 7.719892625415662e-06,
+      "loss": 0.2868,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7713125845737483,
+      "grad_norm": 0.5554511433804457,
+      "learning_rate": 7.714935639966257e-06,
+      "loss": 0.3034,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7719891745602165,
+      "grad_norm": 0.5661215277303339,
+      "learning_rate": 7.709974867511139e-06,
+      "loss": 0.2964,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7726657645466847,
+      "grad_norm": 0.5325063138986836,
+      "learning_rate": 7.705010314969983e-06,
+      "loss": 0.2934,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7733423545331529,
+      "grad_norm": 0.5170239627683275,
+      "learning_rate": 7.700041989267738e-06,
+      "loss": 0.2891,
+      "step": 1143
+    },
+    {
+      "epoch": 0.774018944519621,
+      "grad_norm": 0.5100883589851496,
+      "learning_rate": 7.695069897334613e-06,
+      "loss": 0.2947,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7746955345060893,
+      "grad_norm": 0.532933844771571,
+      "learning_rate": 7.690094046106073e-06,
+      "loss": 0.2708,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7753721244925575,
+      "grad_norm": 0.5615399859733609,
+      "learning_rate": 7.685114442522831e-06,
+      "loss": 0.2994,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7760487144790257,
+      "grad_norm": 0.48736958054904844,
+      "learning_rate": 7.680131093530826e-06,
+      "loss": 0.2898,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7767253044654939,
+      "grad_norm": 0.5197490489916861,
+      "learning_rate": 7.675144006081225e-06,
+      "loss": 0.3024,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7774018944519621,
+      "grad_norm": 0.5310227219134894,
+      "learning_rate": 7.670153187130411e-06,
+      "loss": 0.2908,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7780784844384303,
+      "grad_norm": 0.5101929105114068,
+      "learning_rate": 7.66515864363997e-06,
+      "loss": 0.2872,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7787550744248986,
+      "grad_norm": 0.5104166315712599,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.283,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7794316644113667,
+      "grad_norm": 0.5150006461668013,
+      "learning_rate": 7.65515841091252e-06,
+      "loss": 0.3013,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7801082543978349,
+      "grad_norm": 0.5152604207871935,
+      "learning_rate": 7.650152735624623e-06,
+      "loss": 0.2928,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7807848443843031,
+      "grad_norm": 0.5653871568162103,
+      "learning_rate": 7.645143363695302e-06,
+      "loss": 0.2962,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7814614343707713,
+      "grad_norm": 0.4846930781223465,
+      "learning_rate": 7.640130302112021e-06,
+      "loss": 0.2878,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7821380243572396,
+      "grad_norm": 0.5083944910239082,
+      "learning_rate": 7.635113557867395e-06,
+      "loss": 0.2772,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7828146143437077,
+      "grad_norm": 0.5171998509773861,
+      "learning_rate": 7.63009313795917e-06,
+      "loss": 0.2891,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7834912043301759,
+      "grad_norm": 0.527273562153083,
+      "learning_rate": 7.625069049390228e-06,
+      "loss": 0.2985,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7841677943166441,
+      "grad_norm": 0.5007616020807756,
+      "learning_rate": 7.6200412991685594e-06,
+      "loss": 0.2836,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7848443843031123,
+      "grad_norm": 0.5100125322664146,
+      "learning_rate": 7.615009894307263e-06,
+      "loss": 0.2994,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7855209742895806,
+      "grad_norm": 0.5283593015841002,
+      "learning_rate": 7.609974841824543e-06,
+      "loss": 0.2895,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7861975642760487,
+      "grad_norm": 0.5020718131409405,
+      "learning_rate": 7.604936148743682e-06,
+      "loss": 0.291,
+      "step": 1162
+    },
+    {
+      "epoch": 0.786874154262517,
+      "grad_norm": 0.5221560502797526,
+      "learning_rate": 7.59989382209305e-06,
+      "loss": 0.2938,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7875507442489851,
+      "grad_norm": 0.5163226365784069,
+      "learning_rate": 7.594847868906076e-06,
+      "loss": 0.2851,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7882273342354533,
+      "grad_norm": 0.516485275317283,
+      "learning_rate": 7.5897982962212555e-06,
+      "loss": 0.2852,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7889039242219216,
+      "grad_norm": 0.5177324467918335,
+      "learning_rate": 7.584745111082128e-06,
+      "loss": 0.2929,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7895805142083897,
+      "grad_norm": 0.4956460660733004,
+      "learning_rate": 7.579688320537275e-06,
+      "loss": 0.2878,
+      "step": 1167
+    },
+    {
+      "epoch": 0.790257104194858,
+      "grad_norm": 0.511657882364339,
+      "learning_rate": 7.574627931640304e-06,
+      "loss": 0.2832,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7909336941813261,
+      "grad_norm": 0.5082570171029187,
+      "learning_rate": 7.569563951449843e-06,
+      "loss": 0.3041,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7916102841677943,
+      "grad_norm": 0.4911783590110438,
+      "learning_rate": 7.564496387029532e-06,
+      "loss": 0.2923,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7922868741542625,
+      "grad_norm": 0.5132214545312058,
+      "learning_rate": 7.559425245448006e-06,
+      "loss": 0.2881,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7929634641407307,
+      "grad_norm": 0.5264696720321423,
+      "learning_rate": 7.5543505337788934e-06,
+      "loss": 0.2952,
+      "step": 1172
+    },
+    {
+      "epoch": 0.793640054127199,
+      "grad_norm": 0.4992700644476591,
+      "learning_rate": 7.5492722591008e-06,
+      "loss": 0.2942,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7943166441136671,
+      "grad_norm": 0.49137414253277817,
+      "learning_rate": 7.544190428497304e-06,
+      "loss": 0.2943,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7949932341001353,
+      "grad_norm": 0.5279869258872314,
+      "learning_rate": 7.539105049056943e-06,
+      "loss": 0.3016,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7956698240866035,
+      "grad_norm": 0.49257298956271767,
+      "learning_rate": 7.5340161278732e-06,
+      "loss": 0.279,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7963464140730717,
+      "grad_norm": 0.5075326417080966,
+      "learning_rate": 7.528923672044507e-06,
+      "loss": 0.2858,
+      "step": 1177
+    },
+    {
+      "epoch": 0.79702300405954,
+      "grad_norm": 0.5057429807682727,
+      "learning_rate": 7.52382768867422e-06,
+      "loss": 0.2935,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7976995940460081,
+      "grad_norm": 0.5503856758407208,
+      "learning_rate": 7.518728184870616e-06,
+      "loss": 0.2962,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7983761840324763,
+      "grad_norm": 0.5041797636978502,
+      "learning_rate": 7.5136251677468856e-06,
+      "loss": 0.2939,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7990527740189445,
+      "grad_norm": 0.5115254721004205,
+      "learning_rate": 7.508518644421119e-06,
+      "loss": 0.2824,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7997293640054127,
+      "grad_norm": 0.5196716966896301,
+      "learning_rate": 7.5034086220162945e-06,
+      "loss": 0.2823,
+      "step": 1182
+    },
+    {
+      "epoch": 0.800405953991881,
+      "grad_norm": 0.552815855381993,
+      "learning_rate": 7.498295107660276e-06,
+      "loss": 0.2789,
+      "step": 1183
+    },
+    {
+      "epoch": 0.8010825439783491,
+      "grad_norm": 0.5035747925646005,
+      "learning_rate": 7.4931781084857915e-06,
+      "loss": 0.2798,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8017591339648173,
+      "grad_norm": 0.521414687490373,
+      "learning_rate": 7.488057631630438e-06,
+      "loss": 0.2964,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8024357239512855,
+      "grad_norm": 0.5060885023197377,
+      "learning_rate": 7.482933684236654e-06,
+      "loss": 0.2974,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8031123139377537,
+      "grad_norm": 0.5614209951312902,
+      "learning_rate": 7.4778062734517294e-06,
+      "loss": 0.2852,
+      "step": 1187
+    },
+    {
+      "epoch": 0.803788903924222,
+      "grad_norm": 0.5247824495892759,
+      "learning_rate": 7.4726754064277775e-06,
+      "loss": 0.2963,
+      "step": 1188
+    },
+    {
+      "epoch": 0.8044654939106901,
+      "grad_norm": 0.5302776872643215,
+      "learning_rate": 7.467541090321735e-06,
+      "loss": 0.2973,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8051420838971584,
+      "grad_norm": 0.5438393785369889,
+      "learning_rate": 7.462403332295351e-06,
+      "loss": 0.2995,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8058186738836265,
+      "grad_norm": 0.5120083170797697,
+      "learning_rate": 7.457262139515172e-06,
+      "loss": 0.2863,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8064952638700947,
+      "grad_norm": 0.4970137132511043,
+      "learning_rate": 7.452117519152542e-06,
+      "loss": 0.2899,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8071718538565629,
+      "grad_norm": 0.5346934411971949,
+      "learning_rate": 7.446969478383577e-06,
+      "loss": 0.2962,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8078484438430311,
+      "grad_norm": 0.5204148558722844,
+      "learning_rate": 7.441818024389173e-06,
+      "loss": 0.2895,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8085250338294994,
+      "grad_norm": 0.5105879609070342,
+      "learning_rate": 7.436663164354982e-06,
+      "loss": 0.2842,
+      "step": 1195
+    },
+    {
+      "epoch": 0.8092016238159675,
+      "grad_norm": 0.5384659059121913,
+      "learning_rate": 7.431504905471407e-06,
+      "loss": 0.2884,
+      "step": 1196
+    },
+    {
+      "epoch": 0.8098782138024357,
+      "grad_norm": 0.5477081150417581,
+      "learning_rate": 7.4263432549335946e-06,
+      "loss": 0.2971,
+      "step": 1197
+    },
+    {
+      "epoch": 0.8105548037889039,
+      "grad_norm": 0.5026707540611514,
+      "learning_rate": 7.4211782199414204e-06,
+      "loss": 0.2875,
+      "step": 1198
+    },
+    {
+      "epoch": 0.8112313937753721,
+      "grad_norm": 0.48973646449444286,
+      "learning_rate": 7.416009807699481e-06,
+      "loss": 0.2962,
+      "step": 1199
+    },
+    {
+      "epoch": 0.8119079837618404,
+      "grad_norm": 0.5344108916019269,
+      "learning_rate": 7.410838025417083e-06,
+      "loss": 0.2856,
+      "step": 1200
+    },
+    {
+      "epoch": 0.8125845737483085,
+      "grad_norm": 0.5257347484096067,
+      "learning_rate": 7.405662880308239e-06,
+      "loss": 0.2862,
+      "step": 1201
+    },
+    {
+      "epoch": 0.8132611637347767,
+      "grad_norm": 0.5404903308208926,
+      "learning_rate": 7.400484379591644e-06,
+      "loss": 0.2873,
+      "step": 1202
+    },
+    {
+      "epoch": 0.8139377537212449,
+      "grad_norm": 0.5187618591450285,
+      "learning_rate": 7.395302530490679e-06,
+      "loss": 0.2927,
+      "step": 1203
+    },
+    {
+      "epoch": 0.8146143437077131,
+      "grad_norm": 0.5122267224864078,
+      "learning_rate": 7.390117340233396e-06,
+      "loss": 0.2888,
+      "step": 1204
+    },
+    {
+      "epoch": 0.8152909336941814,
+      "grad_norm": 0.49683197670608065,
+      "learning_rate": 7.384928816052506e-06,
+      "loss": 0.2911,
+      "step": 1205
+    },
+    {
+      "epoch": 0.8159675236806495,
+      "grad_norm": 0.4977612591399441,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.2897,
+      "step": 1206
+    },
+    {
+      "epoch": 0.8166441136671178,
+      "grad_norm": 0.5289592395056446,
+      "learning_rate": 7.374541794873987e-06,
+      "loss": 0.294,
+      "step": 1207
+    },
+    {
+      "epoch": 0.8173207036535859,
+      "grad_norm": 0.533321598866801,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.2976,
+      "step": 1208
+    },
+    {
+      "epoch": 0.8179972936400541,
+      "grad_norm": 0.5148821973181089,
+      "learning_rate": 7.364141524909641e-06,
+      "loss": 0.3028,
+      "step": 1209
+    },
+    {
+      "epoch": 0.8186738836265224,
+      "grad_norm": 0.5187783246036881,
+      "learning_rate": 7.358936439763789e-06,
+      "loss": 0.294,
+      "step": 1210
+    },
+    {
+      "epoch": 0.8193504736129905,
+      "grad_norm": 0.5220413061472483,
+      "learning_rate": 7.353728064187901e-06,
+      "loss": 0.2793,
+      "step": 1211
+    },
+    {
+      "epoch": 0.8200270635994588,
+      "grad_norm": 0.49274578176068373,
+      "learning_rate": 7.348516405447031e-06,
+      "loss": 0.297,
+      "step": 1212
+    },
+    {
+      "epoch": 0.8207036535859269,
+      "grad_norm": 0.5154631863709815,
+      "learning_rate": 7.343301470810809e-06,
+      "loss": 0.289,
+      "step": 1213
+    },
+    {
+      "epoch": 0.8213802435723951,
+      "grad_norm": 0.5098075632198067,
+      "learning_rate": 7.338083267553433e-06,
+      "loss": 0.2962,
+      "step": 1214
+    },
+    {
+      "epoch": 0.8220568335588633,
+      "grad_norm": 0.4930337652918319,
+      "learning_rate": 7.332861802953666e-06,
+      "loss": 0.2951,
+      "step": 1215
+    },
+    {
+      "epoch": 0.8227334235453315,
+      "grad_norm": 0.5062139004310491,
+      "learning_rate": 7.327637084294818e-06,
+      "loss": 0.2894,
+      "step": 1216
+    },
+    {
+      "epoch": 0.8234100135317998,
+      "grad_norm": 0.49357924554126703,
+      "learning_rate": 7.322409118864733e-06,
+      "loss": 0.2748,
+      "step": 1217
+    },
+    {
+      "epoch": 0.8240866035182679,
+      "grad_norm": 0.52528962434527,
+      "learning_rate": 7.317177913955795e-06,
+      "loss": 0.3001,
+      "step": 1218
+    },
+    {
+      "epoch": 0.8247631935047361,
+      "grad_norm": 0.4856818168748149,
+      "learning_rate": 7.311943476864895e-06,
+      "loss": 0.2807,
+      "step": 1219
+    },
+    {
+      "epoch": 0.8254397834912043,
+      "grad_norm": 0.5289904876094443,
+      "learning_rate": 7.30670581489344e-06,
+      "loss": 0.3021,
+      "step": 1220
+    },
+    {
+      "epoch": 0.8261163734776725,
+      "grad_norm": 0.5351263566606644,
+      "learning_rate": 7.301464935347331e-06,
+      "loss": 0.2886,
+      "step": 1221
+    },
+    {
+      "epoch": 0.8267929634641408,
+      "grad_norm": 0.5409883765385054,
+      "learning_rate": 7.29622084553696e-06,
+      "loss": 0.2848,
+      "step": 1222
+    },
+    {
+      "epoch": 0.8274695534506089,
+      "grad_norm": 0.5397510511680542,
+      "learning_rate": 7.290973552777197e-06,
+      "loss": 0.2952,
+      "step": 1223
+    },
+    {
+      "epoch": 0.8281461434370772,
+      "grad_norm": 0.49851980408663693,
+      "learning_rate": 7.285723064387373e-06,
+      "loss": 0.2945,
+      "step": 1224
+    },
+    {
+      "epoch": 0.8288227334235453,
+      "grad_norm": 0.509754258690803,
+      "learning_rate": 7.28046938769129e-06,
+      "loss": 0.2872,
+      "step": 1225
+    },
+    {
+      "epoch": 0.8294993234100135,
+      "grad_norm": 0.49340929759523594,
+      "learning_rate": 7.2752125300171835e-06,
+      "loss": 0.2862,
+      "step": 1226
+    },
+    {
+      "epoch": 0.8301759133964818,
+      "grad_norm": 0.5246663866735056,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.2875,
+      "step": 1227
+    },
+    {
+      "epoch": 0.8308525033829499,
+      "grad_norm": 0.49936809538468435,
+      "learning_rate": 7.264689301070048e-06,
+      "loss": 0.2847,
+      "step": 1228
+    },
+    {
+      "epoch": 0.8315290933694182,
+      "grad_norm": 0.5042806044962409,
+      "learning_rate": 7.259422944475645e-06,
+      "loss": 0.2889,
+      "step": 1229
+    },
+    {
+      "epoch": 0.8322056833558863,
+      "grad_norm": 0.5225278093781809,
+      "learning_rate": 7.254153436260456e-06,
+      "loss": 0.2948,
+      "step": 1230
+    },
+    {
+      "epoch": 0.8328822733423545,
+      "grad_norm": 0.5038493621519162,
+      "learning_rate": 7.248880783774801e-06,
+      "loss": 0.2871,
+      "step": 1231
+    },
+    {
+      "epoch": 0.8335588633288228,
+      "grad_norm": 0.5099080936688112,
+      "learning_rate": 7.2436049943733955e-06,
+      "loss": 0.2964,
+      "step": 1232
+    },
+    {
+      "epoch": 0.8342354533152909,
+      "grad_norm": 0.4974928059829499,
+      "learning_rate": 7.238326075415323e-06,
+      "loss": 0.2876,
+      "step": 1233
+    },
+    {
+      "epoch": 0.8349120433017592,
+      "grad_norm": 0.5249895712863232,
+      "learning_rate": 7.233044034264034e-06,
+      "loss": 0.2874,
+      "step": 1234
+    },
+    {
+      "epoch": 0.8355886332882273,
+      "grad_norm": 0.494882477078539,
+      "learning_rate": 7.2277588782873345e-06,
+      "loss": 0.2859,
+      "step": 1235
+    },
+    {
+      "epoch": 0.8362652232746955,
+      "grad_norm": 0.5198069091353015,
+      "learning_rate": 7.22247061485738e-06,
+      "loss": 0.2876,
+      "step": 1236
+    },
+    {
+      "epoch": 0.8369418132611637,
+      "grad_norm": 0.5038480790884718,
+      "learning_rate": 7.21717925135065e-06,
+      "loss": 0.2802,
+      "step": 1237
+    },
+    {
+      "epoch": 0.8376184032476319,
+      "grad_norm": 0.5015497291531253,
+      "learning_rate": 7.211884795147958e-06,
+      "loss": 0.2893,
+      "step": 1238
+    },
+    {
+      "epoch": 0.8382949932341002,
+      "grad_norm": 0.5108063634218935,
+      "learning_rate": 7.206587253634427e-06,
+      "loss": 0.2848,
+      "step": 1239
+    },
+    {
+      "epoch": 0.8389715832205683,
+      "grad_norm": 0.5363080994847788,
+      "learning_rate": 7.201286634199484e-06,
+      "loss": 0.303,
+      "step": 1240
+    },
+    {
+      "epoch": 0.8396481732070366,
+      "grad_norm": 0.5183761723692146,
+      "learning_rate": 7.195982944236853e-06,
+      "loss": 0.296,
+      "step": 1241
+    },
+    {
+      "epoch": 0.8403247631935047,
+      "grad_norm": 0.5012235980013421,
+      "learning_rate": 7.190676191144532e-06,
+      "loss": 0.2868,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8410013531799729,
+      "grad_norm": 0.49488671362899184,
+      "learning_rate": 7.185366382324804e-06,
+      "loss": 0.2982,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8416779431664412,
+      "grad_norm": 0.530108496689666,
+      "learning_rate": 7.180053525184202e-06,
+      "loss": 0.2965,
+      "step": 1244
+    },
+    {
+      "epoch": 0.8423545331529093,
+      "grad_norm": 0.499685324748837,
+      "learning_rate": 7.1747376271335215e-06,
+      "loss": 0.2881,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8430311231393776,
+      "grad_norm": 0.5118052241031257,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.2989,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8437077131258457,
+      "grad_norm": 0.5088857889927775,
+      "learning_rate": 7.1640967379662775e-06,
+      "loss": 0.286,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8443843031123139,
+      "grad_norm": 0.5047732071873224,
+      "learning_rate": 7.158771761692464e-06,
+      "loss": 0.287,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8450608930987822,
+      "grad_norm": 0.5057955553659181,
+      "learning_rate": 7.153443774194049e-06,
+      "loss": 0.2867,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8457374830852503,
+      "grad_norm": 0.5090977300877949,
+      "learning_rate": 7.148112782902927e-06,
+      "loss": 0.2879,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8464140730717186,
+      "grad_norm": 0.501082349049618,
+      "learning_rate": 7.14277879525518e-06,
+      "loss": 0.2924,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8470906630581867,
+      "grad_norm": 0.5074656989507028,
+      "learning_rate": 7.137441818691081e-06,
+      "loss": 0.295,
+      "step": 1252
+    },
+    {
+      "epoch": 0.847767253044655,
+      "grad_norm": 0.5063998022960192,
+      "learning_rate": 7.1321018606550574e-06,
+      "loss": 0.2857,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8484438430311232,
+      "grad_norm": 0.5131695092597027,
+      "learning_rate": 7.1267589285957075e-06,
+      "loss": 0.2979,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8491204330175913,
+      "grad_norm": 0.5151293296394088,
+      "learning_rate": 7.121413029965769e-06,
+      "loss": 0.2846,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8497970230040596,
+      "grad_norm": 0.4817374266578085,
+      "learning_rate": 7.1160641722221255e-06,
+      "loss": 0.2663,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8504736129905277,
+      "grad_norm": 0.4981903503020046,
+      "learning_rate": 7.110712362825783e-06,
+      "loss": 0.278,
+      "step": 1257
+    },
+    {
+      "epoch": 0.851150202976996,
+      "grad_norm": 0.5429825410547419,
+      "learning_rate": 7.105357609241863e-06,
+      "loss": 0.2919,
+      "step": 1258
+    },
+    {
+      "epoch": 0.8518267929634641,
+      "grad_norm": 0.5056899511252598,
+      "learning_rate": 7.0999999189396015e-06,
+      "loss": 0.2775,
+      "step": 1259
+    },
+    {
+      "epoch": 0.8525033829499323,
+      "grad_norm": 0.5177568809944318,
+      "learning_rate": 7.094639299392324e-06,
+      "loss": 0.2854,
+      "step": 1260
+    },
+    {
+      "epoch": 0.8531799729364006,
+      "grad_norm": 0.5120926696227671,
+      "learning_rate": 7.0892757580774455e-06,
+      "loss": 0.285,
+      "step": 1261
+    },
+    {
+      "epoch": 0.8538565629228687,
+      "grad_norm": 0.4877370991590401,
+      "learning_rate": 7.083909302476453e-06,
+      "loss": 0.2976,
+      "step": 1262
+    },
+    {
+      "epoch": 0.854533152909337,
+      "grad_norm": 0.5107253523769092,
+      "learning_rate": 7.078539940074902e-06,
+      "loss": 0.3055,
+      "step": 1263
+    },
+    {
+      "epoch": 0.8552097428958051,
+      "grad_norm": 0.5235336876197503,
+      "learning_rate": 7.0731676783624015e-06,
+      "loss": 0.3045,
+      "step": 1264
+    },
+    {
+      "epoch": 0.8558863328822733,
+      "grad_norm": 0.5184055764815209,
+      "learning_rate": 7.067792524832604e-06,
+      "loss": 0.2965,
+      "step": 1265
+    },
+    {
+      "epoch": 0.8565629228687416,
+      "grad_norm": 0.4820286023761079,
+      "learning_rate": 7.062414486983197e-06,
+      "loss": 0.2854,
+      "step": 1266
+    },
+    {
+      "epoch": 0.8572395128552097,
+      "grad_norm": 0.49504218324028065,
+      "learning_rate": 7.057033572315891e-06,
+      "loss": 0.2923,
+      "step": 1267
+    },
+    {
+      "epoch": 0.857916102841678,
+      "grad_norm": 0.5230913549656236,
+      "learning_rate": 7.051649788336405e-06,
+      "loss": 0.2972,
+      "step": 1268
+    },
+    {
+      "epoch": 0.8585926928281461,
+      "grad_norm": 0.5091362175477512,
+      "learning_rate": 7.04626314255447e-06,
+      "loss": 0.2886,
+      "step": 1269
+    },
+    {
+      "epoch": 0.8592692828146143,
+      "grad_norm": 0.5212931497824757,
+      "learning_rate": 7.040873642483801e-06,
+      "loss": 0.306,
+      "step": 1270
+    },
+    {
+      "epoch": 0.8599458728010826,
+      "grad_norm": 0.5011410072797908,
+      "learning_rate": 7.035481295642096e-06,
+      "loss": 0.287,
+      "step": 1271
+    },
+    {
+      "epoch": 0.8606224627875507,
+      "grad_norm": 0.540353736608924,
+      "learning_rate": 7.030086109551023e-06,
+      "loss": 0.2858,
+      "step": 1272
+    },
+    {
+      "epoch": 0.861299052774019,
+      "grad_norm": 0.5298712746286626,
+      "learning_rate": 7.024688091736214e-06,
+      "loss": 0.2914,
+      "step": 1273
+    },
+    {
+      "epoch": 0.8619756427604871,
+      "grad_norm": 0.5390605325457568,
+      "learning_rate": 7.019287249727248e-06,
+      "loss": 0.2884,
+      "step": 1274
+    },
+    {
+      "epoch": 0.8626522327469553,
+      "grad_norm": 0.5072165885925364,
+      "learning_rate": 7.013883591057642e-06,
+      "loss": 0.2811,
+      "step": 1275
+    },
+    {
+      "epoch": 0.8633288227334236,
+      "grad_norm": 0.49736214799630707,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.2901,
+      "step": 1276
+    },
+    {
+      "epoch": 0.8640054127198917,
+      "grad_norm": 0.5297546644112249,
+      "learning_rate": 7.00306785389023e-06,
+      "loss": 0.3049,
+      "step": 1277
+    },
+    {
+      "epoch": 0.86468200270636,
+      "grad_norm": 0.5088607433699045,
+      "learning_rate": 6.997655790479062e-06,
+      "loss": 0.2926,
+      "step": 1278
+    },
+    {
+      "epoch": 0.8653585926928281,
+      "grad_norm": 0.4856890572738931,
+      "learning_rate": 6.992240940580517e-06,
+      "loss": 0.2911,
+      "step": 1279
+    },
+    {
+      "epoch": 0.8660351826792964,
+      "grad_norm": 0.49205945096555903,
+      "learning_rate": 6.986823311747652e-06,
+      "loss": 0.293,
+      "step": 1280
+    },
+    {
+      "epoch": 0.8667117726657646,
+      "grad_norm": 0.5031738794249846,
+      "learning_rate": 6.981402911537405e-06,
+      "loss": 0.2885,
+      "step": 1281
+    },
+    {
+      "epoch": 0.8673883626522327,
+      "grad_norm": 0.4937744745156734,
+      "learning_rate": 6.97597974751057e-06,
+      "loss": 0.2976,
+      "step": 1282
+    },
+    {
+      "epoch": 0.868064952638701,
+      "grad_norm": 0.5171874968837452,
+      "learning_rate": 6.970553827231809e-06,
+      "loss": 0.2755,
+      "step": 1283
+    },
+    {
+      "epoch": 0.8687415426251691,
+      "grad_norm": 0.4888082358116817,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.2808,
+      "step": 1284
+    },
+    {
+      "epoch": 0.8694181326116374,
+      "grad_norm": 0.4898872134619127,
+      "learning_rate": 6.959693748196335e-06,
+      "loss": 0.2938,
+      "step": 1285
+    },
+    {
+      "epoch": 0.8700947225981055,
+      "grad_norm": 0.5067308516850026,
+      "learning_rate": 6.954259604588114e-06,
+      "loss": 0.2952,
+      "step": 1286
+    },
+    {
+      "epoch": 0.8707713125845737,
+      "grad_norm": 0.4901636868531918,
+      "learning_rate": 6.94882273502493e-06,
+      "loss": 0.271,
+      "step": 1287
+    },
+    {
+      "epoch": 0.871447902571042,
+      "grad_norm": 0.5147931027398728,
+      "learning_rate": 6.943383147090552e-06,
+      "loss": 0.2982,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8721244925575101,
+      "grad_norm": 0.5224507776329675,
+      "learning_rate": 6.937940848372546e-06,
+      "loss": 0.2943,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8728010825439784,
+      "grad_norm": 0.49836528865005403,
+      "learning_rate": 6.932495846462262e-06,
+      "loss": 0.2929,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8734776725304465,
+      "grad_norm": 0.5047546991534498,
+      "learning_rate": 6.9270481489548125e-06,
+      "loss": 0.284,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8741542625169147,
+      "grad_norm": 0.5142329454898764,
+      "learning_rate": 6.921597763449075e-06,
+      "loss": 0.2718,
+      "step": 1292
+    },
+    {
+      "epoch": 0.874830852503383,
+      "grad_norm": 0.5278713233970972,
+      "learning_rate": 6.9161446975476775e-06,
+      "loss": 0.2979,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8755074424898511,
+      "grad_norm": 0.5136695458200962,
+      "learning_rate": 6.9106889588569845e-06,
+      "loss": 0.2877,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8761840324763194,
+      "grad_norm": 0.5160883071815348,
+      "learning_rate": 6.905230554987088e-06,
+      "loss": 0.2914,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8768606224627875,
+      "grad_norm": 0.5158935268734325,
+      "learning_rate": 6.8997694935518e-06,
+      "loss": 0.2869,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8775372124492558,
+      "grad_norm": 0.4806965517852258,
+      "learning_rate": 6.8943057821686386e-06,
+      "loss": 0.2834,
+      "step": 1297
+    },
+    {
+      "epoch": 0.878213802435724,
+      "grad_norm": 0.5297237103976068,
+      "learning_rate": 6.888839428458819e-06,
+      "loss": 0.304,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8788903924221921,
+      "grad_norm": 0.47893432820045173,
+      "learning_rate": 6.883370440047238e-06,
+      "loss": 0.2952,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8795669824086604,
+      "grad_norm": 0.4835669687548482,
+      "learning_rate": 6.877898824562472e-06,
+      "loss": 0.2735,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8802435723951285,
+      "grad_norm": 0.5223596709849508,
+      "learning_rate": 6.8724245896367636e-06,
+      "loss": 0.2965,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8809201623815968,
+      "grad_norm": 0.5141300561680373,
+      "learning_rate": 6.8669477429060026e-06,
+      "loss": 0.2768,
+      "step": 1302
+    },
+    {
+      "epoch": 0.881596752368065,
+      "grad_norm": 0.4951032347933015,
+      "learning_rate": 6.8614682920097265e-06,
+      "loss": 0.2806,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8822733423545331,
+      "grad_norm": 0.5214575251525285,
+      "learning_rate": 6.855986244591104e-06,
+      "loss": 0.3032,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8829499323410014,
+      "grad_norm": 0.5156359552723714,
+      "learning_rate": 6.850501608296928e-06,
+      "loss": 0.2886,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8836265223274695,
+      "grad_norm": 0.48358077759781726,
+      "learning_rate": 6.845014390777595e-06,
+      "loss": 0.2819,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8843031123139378,
+      "grad_norm": 0.4865200684596849,
+      "learning_rate": 6.839524599687115e-06,
+      "loss": 0.2858,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8849797023004059,
+      "grad_norm": 0.4866188832561783,
+      "learning_rate": 6.834032242683075e-06,
+      "loss": 0.2845,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8856562922868741,
+      "grad_norm": 0.5295075103258354,
+      "learning_rate": 6.8285373274266474e-06,
+      "loss": 0.2897,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8863328822733424,
+      "grad_norm": 0.5415319850894017,
+      "learning_rate": 6.823039861582574e-06,
+      "loss": 0.301,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8870094722598105,
+      "grad_norm": 0.5071470223646076,
+      "learning_rate": 6.817539852819149e-06,
+      "loss": 0.294,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8876860622462788,
+      "grad_norm": 0.5002449744781545,
+      "learning_rate": 6.8120373088082215e-06,
+      "loss": 0.2939,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8883626522327469,
+      "grad_norm": 0.5366527509961133,
+      "learning_rate": 6.806532237225168e-06,
+      "loss": 0.284,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8890392422192152,
+      "grad_norm": 0.5204012973689583,
+      "learning_rate": 6.801024645748899e-06,
+      "loss": 0.2897,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8897158322056834,
+      "grad_norm": 0.5030863220749361,
+      "learning_rate": 6.795514542061834e-06,
+      "loss": 0.2844,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8903924221921515,
+      "grad_norm": 0.5327191920822224,
+      "learning_rate": 6.7900019338499005e-06,
+      "loss": 0.2858,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8910690121786198,
+      "grad_norm": 0.5270592249473931,
+      "learning_rate": 6.7844868288025155e-06,
+      "loss": 0.2837,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8917456021650879,
+      "grad_norm": 0.509392974791801,
+      "learning_rate": 6.778969234612583e-06,
+      "loss": 0.2866,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8924221921515562,
+      "grad_norm": 0.5053445739792091,
+      "learning_rate": 6.773449158976477e-06,
+      "loss": 0.2827,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8930987821380244,
+      "grad_norm": 0.49440090473035586,
+      "learning_rate": 6.767926609594032e-06,
+      "loss": 0.2788,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8937753721244925,
+      "grad_norm": 0.47323986988036804,
+      "learning_rate": 6.762401594168537e-06,
+      "loss": 0.2879,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8944519621109608,
+      "grad_norm": 0.5021874917521402,
+      "learning_rate": 6.7568741204067145e-06,
+      "loss": 0.2926,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8951285520974289,
+      "grad_norm": 0.49905761375698354,
+      "learning_rate": 6.751344196018721e-06,
+      "loss": 0.2801,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8958051420838972,
+      "grad_norm": 0.5146797296386512,
+      "learning_rate": 6.74581182871813e-06,
+      "loss": 0.2888,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8964817320703654,
+      "grad_norm": 0.535471717499363,
+      "learning_rate": 6.7402770262219234e-06,
+      "loss": 0.2949,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8971583220568335,
+      "grad_norm": 0.4969063027367392,
+      "learning_rate": 6.734739796250477e-06,
+      "loss": 0.2933,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8978349120433018,
+      "grad_norm": 0.5113497859425116,
+      "learning_rate": 6.729200146527554e-06,
+      "loss": 0.2917,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8985115020297699,
+      "grad_norm": 0.5104617295130721,
+      "learning_rate": 6.723658084780297e-06,
+      "loss": 0.2843,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8991880920162382,
+      "grad_norm": 0.47458879263020254,
+      "learning_rate": 6.718113618739206e-06,
+      "loss": 0.2922,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8998646820027063,
+      "grad_norm": 0.4966493613519298,
+      "learning_rate": 6.712566756138142e-06,
+      "loss": 0.2913,
+      "step": 1330
+    },
+    {
+      "epoch": 0.9005412719891746,
+      "grad_norm": 0.5122126851253148,
+      "learning_rate": 6.707017504714299e-06,
+      "loss": 0.2992,
+      "step": 1331
+    },
+    {
+      "epoch": 0.9012178619756428,
+      "grad_norm": 0.5219241034911285,
+      "learning_rate": 6.701465872208216e-06,
+      "loss": 0.3036,
+      "step": 1332
+    },
+    {
+      "epoch": 0.9018944519621109,
+      "grad_norm": 0.5129960648412032,
+      "learning_rate": 6.695911866363742e-06,
+      "loss": 0.2779,
+      "step": 1333
+    },
+    {
+      "epoch": 0.9025710419485792,
+      "grad_norm": 0.49913413466961615,
+      "learning_rate": 6.690355494928043e-06,
+      "loss": 0.2972,
+      "step": 1334
+    },
+    {
+      "epoch": 0.9032476319350473,
+      "grad_norm": 0.48775119445080545,
+      "learning_rate": 6.684796765651583e-06,
+      "loss": 0.2887,
+      "step": 1335
+    },
+    {
+      "epoch": 0.9039242219215156,
+      "grad_norm": 0.4795024787282662,
+      "learning_rate": 6.6792356862881144e-06,
+      "loss": 0.2773,
+      "step": 1336
+    },
+    {
+      "epoch": 0.9046008119079838,
+      "grad_norm": 0.49584734463377966,
+      "learning_rate": 6.673672264594665e-06,
+      "loss": 0.2872,
+      "step": 1337
+    },
+    {
+      "epoch": 0.9052774018944519,
+      "grad_norm": 0.47326903400727577,
+      "learning_rate": 6.668106508331539e-06,
+      "loss": 0.2816,
+      "step": 1338
+    },
+    {
+      "epoch": 0.9059539918809202,
+      "grad_norm": 0.5170197819917796,
+      "learning_rate": 6.662538425262285e-06,
+      "loss": 0.293,
+      "step": 1339
+    },
+    {
+      "epoch": 0.9066305818673883,
+      "grad_norm": 0.4890499217287929,
+      "learning_rate": 6.656968023153706e-06,
+      "loss": 0.2859,
+      "step": 1340
+    },
+    {
+      "epoch": 0.9073071718538566,
+      "grad_norm": 0.4657129206075007,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.2632,
+      "step": 1341
+    },
+    {
+      "epoch": 0.9079837618403248,
+      "grad_norm": 0.4965575720645063,
+      "learning_rate": 6.6458202929019345e-06,
+      "loss": 0.2884,
+      "step": 1342
+    },
+    {
+      "epoch": 0.908660351826793,
+      "grad_norm": 0.4960997967045921,
+      "learning_rate": 6.640242980308474e-06,
+      "loss": 0.2831,
+      "step": 1343
+    },
+    {
+      "epoch": 0.9093369418132612,
+      "grad_norm": 0.5054773165685083,
+      "learning_rate": 6.634663379775126e-06,
+      "loss": 0.2959,
+      "step": 1344
+    },
+    {
+      "epoch": 0.9100135317997293,
+      "grad_norm": 0.5130719898977725,
+      "learning_rate": 6.62908149908476e-06,
+      "loss": 0.2866,
+      "step": 1345
+    },
+    {
+      "epoch": 0.9106901217861976,
+      "grad_norm": 0.5134125274403117,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.2814,
+      "step": 1346
+    },
+    {
+      "epoch": 0.9113667117726658,
+      "grad_norm": 0.4751872236229903,
+      "learning_rate": 6.617910928380319e-06,
+      "loss": 0.282,
+      "step": 1347
+    },
+    {
+      "epoch": 0.912043301759134,
+      "grad_norm": 0.5042894350235606,
+      "learning_rate": 6.612322253947836e-06,
+      "loss": 0.2824,
+      "step": 1348
+    },
+    {
+      "epoch": 0.9127198917456022,
+      "grad_norm": 0.4859740253310109,
+      "learning_rate": 6.606731330521491e-06,
+      "loss": 0.2888,
+      "step": 1349
+    },
+    {
+      "epoch": 0.9133964817320703,
+      "grad_norm": 0.4992213598896632,
+      "learning_rate": 6.601138165899945e-06,
+      "loss": 0.2845,
+      "step": 1350
+    },
+    {
+      "epoch": 0.9140730717185386,
+      "grad_norm": 0.5209044284282254,
+      "learning_rate": 6.595542767884984e-06,
+      "loss": 0.295,
+      "step": 1351
+    },
+    {
+      "epoch": 0.9147496617050067,
+      "grad_norm": 0.5004433848648135,
+      "learning_rate": 6.589945144281508e-06,
+      "loss": 0.2832,
+      "step": 1352
+    },
+    {
+      "epoch": 0.915426251691475,
+      "grad_norm": 0.5121052820792209,
+      "learning_rate": 6.584345302897522e-06,
+      "loss": 0.3016,
+      "step": 1353
+    },
+    {
+      "epoch": 0.9161028416779432,
+      "grad_norm": 0.48790157985065885,
+      "learning_rate": 6.578743251544128e-06,
+      "loss": 0.2903,
+      "step": 1354
+    },
+    {
+      "epoch": 0.9167794316644113,
+      "grad_norm": 0.498946717727405,
+      "learning_rate": 6.573138998035504e-06,
+      "loss": 0.2941,
+      "step": 1355
+    },
+    {
+      "epoch": 0.9174560216508796,
+      "grad_norm": 0.49344360007461124,
+      "learning_rate": 6.567532550188908e-06,
+      "loss": 0.2913,
+      "step": 1356
+    },
+    {
+      "epoch": 0.9181326116373477,
+      "grad_norm": 0.501723247838122,
+      "learning_rate": 6.561923915824649e-06,
+      "loss": 0.3009,
+      "step": 1357
+    },
+    {
+      "epoch": 0.918809201623816,
+      "grad_norm": 0.5111923341740495,
+      "learning_rate": 6.556313102766094e-06,
+      "loss": 0.287,
+      "step": 1358
+    },
+    {
+      "epoch": 0.9194857916102842,
+      "grad_norm": 0.4902140172915019,
+      "learning_rate": 6.550700118839647e-06,
+      "loss": 0.2919,
+      "step": 1359
+    },
+    {
+      "epoch": 0.9201623815967523,
+      "grad_norm": 0.5488548835648784,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2786,
+      "step": 1360
+    },
+    {
+      "epoch": 0.9208389715832206,
+      "grad_norm": 0.4939014982132556,
+      "learning_rate": 6.539467669703816e-06,
+      "loss": 0.2852,
+      "step": 1361
+    },
+    {
+      "epoch": 0.9215155615696887,
+      "grad_norm": 0.49062505168779685,
+      "learning_rate": 6.533848220162336e-06,
+      "loss": 0.285,
+      "step": 1362
+    },
+    {
+      "epoch": 0.922192151556157,
+      "grad_norm": 0.48575570322264827,
+      "learning_rate": 6.528226631088748e-06,
+      "loss": 0.287,
+      "step": 1363
+    },
+    {
+      "epoch": 0.9228687415426252,
+      "grad_norm": 0.5048940471592656,
+      "learning_rate": 6.5226029103244846e-06,
+      "loss": 0.2765,
+      "step": 1364
+    },
+    {
+      "epoch": 0.9235453315290933,
+      "grad_norm": 0.4979502997719499,
+      "learning_rate": 6.516977065713958e-06,
+      "loss": 0.2826,
+      "step": 1365
+    },
+    {
+      "epoch": 0.9242219215155616,
+      "grad_norm": 0.5226392387845722,
+      "learning_rate": 6.511349105104534e-06,
+      "loss": 0.2845,
+      "step": 1366
+    },
+    {
+      "epoch": 0.9248985115020297,
+      "grad_norm": 0.4776295122479072,
+      "learning_rate": 6.505719036346538e-06,
+      "loss": 0.2797,
+      "step": 1367
+    },
+    {
+      "epoch": 0.925575101488498,
+      "grad_norm": 0.4989314624922866,
+      "learning_rate": 6.500086867293231e-06,
+      "loss": 0.2743,
+      "step": 1368
+    },
+    {
+      "epoch": 0.9262516914749662,
+      "grad_norm": 0.5321032604201313,
+      "learning_rate": 6.4944526058008054e-06,
+      "loss": 0.2978,
+      "step": 1369
+    },
+    {
+      "epoch": 0.9269282814614344,
+      "grad_norm": 0.5034236285368575,
+      "learning_rate": 6.488816259728372e-06,
+      "loss": 0.2754,
+      "step": 1370
+    },
+    {
+      "epoch": 0.9276048714479026,
+      "grad_norm": 0.5008701614227415,
+      "learning_rate": 6.48317783693795e-06,
+      "loss": 0.2815,
+      "step": 1371
+    },
+    {
+      "epoch": 0.9282814614343707,
+      "grad_norm": 0.485094208855566,
+      "learning_rate": 6.477537345294455e-06,
+      "loss": 0.2778,
+      "step": 1372
+    },
+    {
+      "epoch": 0.928958051420839,
+      "grad_norm": 0.5059017917206042,
+      "learning_rate": 6.471894792665687e-06,
+      "loss": 0.2827,
+      "step": 1373
+    },
+    {
+      "epoch": 0.9296346414073072,
+      "grad_norm": 0.48462937150474145,
+      "learning_rate": 6.466250186922325e-06,
+      "loss": 0.2912,
+      "step": 1374
+    },
+    {
+      "epoch": 0.9303112313937754,
+      "grad_norm": 0.5217985256568758,
+      "learning_rate": 6.460603535937905e-06,
+      "loss": 0.2859,
+      "step": 1375
+    },
+    {
+      "epoch": 0.9309878213802436,
+      "grad_norm": 0.4981879096001895,
+      "learning_rate": 6.454954847588824e-06,
+      "loss": 0.2767,
+      "step": 1376
+    },
+    {
+      "epoch": 0.9316644113667117,
+      "grad_norm": 0.5034555305477874,
+      "learning_rate": 6.449304129754313e-06,
+      "loss": 0.2779,
+      "step": 1377
+    },
+    {
+      "epoch": 0.93234100135318,
+      "grad_norm": 0.5282972585918071,
+      "learning_rate": 6.443651390316438e-06,
+      "loss": 0.2834,
+      "step": 1378
+    },
+    {
+      "epoch": 0.9330175913396481,
+      "grad_norm": 0.5154985232149497,
+      "learning_rate": 6.437996637160086e-06,
+      "loss": 0.291,
+      "step": 1379
+    },
+    {
+      "epoch": 0.9336941813261164,
+      "grad_norm": 0.4585181031405808,
+      "learning_rate": 6.4323398781729525e-06,
+      "loss": 0.2749,
+      "step": 1380
+    },
+    {
+      "epoch": 0.9343707713125846,
+      "grad_norm": 0.4832200155107634,
+      "learning_rate": 6.426681121245527e-06,
+      "loss": 0.2791,
+      "step": 1381
+    },
+    {
+      "epoch": 0.9350473612990527,
+      "grad_norm": 0.503695189371936,
+      "learning_rate": 6.42102037427109e-06,
+      "loss": 0.2915,
+      "step": 1382
+    },
+    {
+      "epoch": 0.935723951285521,
+      "grad_norm": 0.5129635263157485,
+      "learning_rate": 6.4153576451456965e-06,
+      "loss": 0.2938,
+      "step": 1383
+    },
+    {
+      "epoch": 0.9364005412719891,
+      "grad_norm": 0.4965362624727899,
+      "learning_rate": 6.409692941768166e-06,
+      "loss": 0.2825,
+      "step": 1384
+    },
+    {
+      "epoch": 0.9370771312584574,
+      "grad_norm": 0.48960594065585,
+      "learning_rate": 6.404026272040071e-06,
+      "loss": 0.2802,
+      "step": 1385
+    },
+    {
+      "epoch": 0.9377537212449256,
+      "grad_norm": 0.5124083372518266,
+      "learning_rate": 6.398357643865731e-06,
+      "loss": 0.2891,
+      "step": 1386
+    },
+    {
+      "epoch": 0.9384303112313938,
+      "grad_norm": 0.5385933172854248,
+      "learning_rate": 6.392687065152192e-06,
+      "loss": 0.2804,
+      "step": 1387
+    },
+    {
+      "epoch": 0.939106901217862,
+      "grad_norm": 0.5150324621992184,
+      "learning_rate": 6.387014543809224e-06,
+      "loss": 0.2878,
+      "step": 1388
+    },
+    {
+      "epoch": 0.9397834912043301,
+      "grad_norm": 0.5156336591313562,
+      "learning_rate": 6.381340087749304e-06,
+      "loss": 0.2924,
+      "step": 1389
+    },
+    {
+      "epoch": 0.9404600811907984,
+      "grad_norm": 0.4961834938395387,
+      "learning_rate": 6.375663704887614e-06,
+      "loss": 0.2842,
+      "step": 1390
+    },
+    {
+      "epoch": 0.9411366711772666,
+      "grad_norm": 0.5698520264909013,
+      "learning_rate": 6.369985403142014e-06,
+      "loss": 0.2964,
+      "step": 1391
+    },
+    {
+      "epoch": 0.9418132611637348,
+      "grad_norm": 0.4688278133931245,
+      "learning_rate": 6.364305190433049e-06,
+      "loss": 0.2914,
+      "step": 1392
+    },
+    {
+      "epoch": 0.942489851150203,
+      "grad_norm": 0.4655033056633345,
+      "learning_rate": 6.358623074683924e-06,
+      "loss": 0.287,
+      "step": 1393
+    },
+    {
+      "epoch": 0.9431664411366711,
+      "grad_norm": 0.46657356470058453,
+      "learning_rate": 6.3529390638205036e-06,
+      "loss": 0.2836,
+      "step": 1394
+    },
+    {
+      "epoch": 0.9438430311231394,
+      "grad_norm": 0.5201375077134995,
+      "learning_rate": 6.34725316577129e-06,
+      "loss": 0.2855,
+      "step": 1395
+    },
+    {
+      "epoch": 0.9445196211096076,
+      "grad_norm": 0.5004868173788369,
+      "learning_rate": 6.341565388467425e-06,
+      "loss": 0.2855,
+      "step": 1396
+    },
+    {
+      "epoch": 0.9451962110960758,
+      "grad_norm": 0.49523439509474737,
+      "learning_rate": 6.335875739842663e-06,
+      "loss": 0.2874,
+      "step": 1397
+    },
+    {
+      "epoch": 0.945872801082544,
+      "grad_norm": 0.48405302799164013,
+      "learning_rate": 6.330184227833376e-06,
+      "loss": 0.2917,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9465493910690121,
+      "grad_norm": 0.5097288535296727,
+      "learning_rate": 6.324490860378533e-06,
+      "loss": 0.2815,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9472259810554804,
+      "grad_norm": 0.49333606409524233,
+      "learning_rate": 6.3187956454196885e-06,
+      "loss": 0.2846,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9479025710419485,
+      "grad_norm": 0.4563276440204552,
+      "learning_rate": 6.313098590900978e-06,
+      "loss": 0.277,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9485791610284168,
+      "grad_norm": 0.46594948299831873,
+      "learning_rate": 6.3073997047691e-06,
+      "loss": 0.2762,
+      "step": 1402
+    },
+    {
+      "epoch": 0.949255751014885,
+      "grad_norm": 0.4896678008496494,
+      "learning_rate": 6.301698994973308e-06,
+      "loss": 0.297,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9499323410013532,
+      "grad_norm": 0.5777164672002452,
+      "learning_rate": 6.295996469465404e-06,
+      "loss": 0.2809,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9506089309878214,
+      "grad_norm": 0.4915240018900208,
+      "learning_rate": 6.290292136199716e-06,
+      "loss": 0.274,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9512855209742895,
+      "grad_norm": 0.5071609586177751,
+      "learning_rate": 6.284586003133096e-06,
+      "loss": 0.2898,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9519621109607578,
+      "grad_norm": 0.48552437605872617,
+      "learning_rate": 6.2788780782249106e-06,
+      "loss": 0.2852,
+      "step": 1407
+    },
+    {
+      "epoch": 0.952638700947226,
+      "grad_norm": 0.4894377198225572,
+      "learning_rate": 6.2731683694370185e-06,
+      "loss": 0.2827,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9533152909336942,
+      "grad_norm": 0.5053661170935958,
+      "learning_rate": 6.26745688473377e-06,
+      "loss": 0.2942,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9539918809201624,
+      "grad_norm": 0.4830839927939992,
+      "learning_rate": 6.261743632081998e-06,
+      "loss": 0.2888,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9546684709066305,
+      "grad_norm": 0.5006258007660077,
+      "learning_rate": 6.256028619450993e-06,
+      "loss": 0.2899,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9553450608930988,
+      "grad_norm": 0.5002741961207722,
+      "learning_rate": 6.250311854812504e-06,
+      "loss": 0.2938,
+      "step": 1412
+    },
+    {
+      "epoch": 0.956021650879567,
+      "grad_norm": 0.47082050797639413,
+      "learning_rate": 6.244593346140725e-06,
+      "loss": 0.2745,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9566982408660352,
+      "grad_norm": 0.4698729962799257,
+      "learning_rate": 6.238873101412282e-06,
+      "loss": 0.2811,
+      "step": 1414
+    },
+    {
+      "epoch": 0.9573748308525034,
+      "grad_norm": 0.48322151144040165,
+      "learning_rate": 6.2331511286062186e-06,
+      "loss": 0.2889,
+      "step": 1415
+    },
+    {
+      "epoch": 0.9580514208389715,
+      "grad_norm": 0.48533463103319363,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.2809,
+      "step": 1416
+    },
+    {
+      "epoch": 0.9587280108254398,
+      "grad_norm": 0.47988836459167555,
+      "learning_rate": 6.2217020306894705e-06,
+      "loss": 0.2806,
+      "step": 1417
+    },
+    {
+      "epoch": 0.959404600811908,
+      "grad_norm": 0.54357915142844,
+      "learning_rate": 6.215974921548888e-06,
+      "loss": 0.2851,
+      "step": 1418
+    },
+    {
+      "epoch": 0.9600811907983762,
+      "grad_norm": 0.4982383889320541,
+      "learning_rate": 6.210246116270867e-06,
+      "loss": 0.3009,
+      "step": 1419
+    },
+    {
+      "epoch": 0.9607577807848444,
+      "grad_norm": 0.48871251377732633,
+      "learning_rate": 6.204515622846399e-06,
+      "loss": 0.2815,
+      "step": 1420
+    },
+    {
+      "epoch": 0.9614343707713126,
+      "grad_norm": 0.5020478105255516,
+      "learning_rate": 6.198783449268827e-06,
+      "loss": 0.2867,
+      "step": 1421
+    },
+    {
+      "epoch": 0.9621109607577808,
+      "grad_norm": 0.5094123355110645,
+      "learning_rate": 6.193049603533835e-06,
+      "loss": 0.2934,
+      "step": 1422
+    },
+    {
+      "epoch": 0.9627875507442489,
+      "grad_norm": 0.4778648168962283,
+      "learning_rate": 6.187314093639444e-06,
+      "loss": 0.2797,
+      "step": 1423
+    },
+    {
+      "epoch": 0.9634641407307172,
+      "grad_norm": 0.4962998059268294,
+      "learning_rate": 6.181576927585993e-06,
+      "loss": 0.2814,
+      "step": 1424
+    },
+    {
+      "epoch": 0.9641407307171854,
+      "grad_norm": 0.47753673089481996,
+      "learning_rate": 6.175838113376133e-06,
+      "loss": 0.2812,
+      "step": 1425
+    },
+    {
+      "epoch": 0.9648173207036536,
+      "grad_norm": 0.47572131755673425,
+      "learning_rate": 6.170097659014812e-06,
+      "loss": 0.287,
+      "step": 1426
+    },
+    {
+      "epoch": 0.9654939106901218,
+      "grad_norm": 0.49693692464535344,
+      "learning_rate": 6.164355572509271e-06,
+      "loss": 0.3004,
+      "step": 1427
+    },
+    {
+      "epoch": 0.9661705006765899,
+      "grad_norm": 0.48440602769136804,
+      "learning_rate": 6.158611861869018e-06,
+      "loss": 0.2901,
+      "step": 1428
+    },
+    {
+      "epoch": 0.9668470906630582,
+      "grad_norm": 0.4507345986285861,
+      "learning_rate": 6.152866535105837e-06,
+      "loss": 0.2766,
+      "step": 1429
+    },
+    {
+      "epoch": 0.9675236806495264,
+      "grad_norm": 0.4808543040196935,
+      "learning_rate": 6.147119600233758e-06,
+      "loss": 0.295,
+      "step": 1430
+    },
+    {
+      "epoch": 0.9682002706359946,
+      "grad_norm": 0.46896469935854346,
+      "learning_rate": 6.141371065269061e-06,
+      "loss": 0.2877,
+      "step": 1431
+    },
+    {
+      "epoch": 0.9688768606224628,
+      "grad_norm": 0.5325000078330345,
+      "learning_rate": 6.135620938230254e-06,
+      "loss": 0.2933,
+      "step": 1432
+    },
+    {
+      "epoch": 0.969553450608931,
+      "grad_norm": 0.4652290144981375,
+      "learning_rate": 6.129869227138062e-06,
+      "loss": 0.2838,
+      "step": 1433
+    },
+    {
+      "epoch": 0.9702300405953992,
+      "grad_norm": 0.4687569920985748,
+      "learning_rate": 6.1241159400154306e-06,
+      "loss": 0.2779,
+      "step": 1434
+    },
+    {
+      "epoch": 0.9709066305818674,
+      "grad_norm": 0.5079379501894264,
+      "learning_rate": 6.118361084887492e-06,
+      "loss": 0.2925,
+      "step": 1435
+    },
+    {
+      "epoch": 0.9715832205683356,
+      "grad_norm": 0.5089281816752588,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.2838,
+      "step": 1436
+    },
+    {
+      "epoch": 0.9722598105548038,
+      "grad_norm": 0.5189985233088312,
+      "learning_rate": 6.106846702727173e-06,
+      "loss": 0.2793,
+      "step": 1437
+    },
+    {
+      "epoch": 0.972936400541272,
+      "grad_norm": 0.5012139300105123,
+      "learning_rate": 6.101087191755958e-06,
+      "loss": 0.2869,
+      "step": 1438
+    },
+    {
+      "epoch": 0.9736129905277402,
+      "grad_norm": 0.46478240900277157,
+      "learning_rate": 6.095326144901746e-06,
+      "loss": 0.2847,
+      "step": 1439
+    },
+    {
+      "epoch": 0.9742895805142084,
+      "grad_norm": 0.48713954143626703,
+      "learning_rate": 6.0895635702004985e-06,
+      "loss": 0.2873,
+      "step": 1440
+    },
+    {
+      "epoch": 0.9749661705006766,
+      "grad_norm": 0.489701089930619,
+      "learning_rate": 6.08379947569031e-06,
+      "loss": 0.2915,
+      "step": 1441
+    },
+    {
+      "epoch": 0.9756427604871448,
+      "grad_norm": 0.4908854187072097,
+      "learning_rate": 6.078033869411389e-06,
+      "loss": 0.2843,
+      "step": 1442
+    },
+    {
+      "epoch": 0.976319350473613,
+      "grad_norm": 0.4905225178630297,
+      "learning_rate": 6.072266759406057e-06,
+      "loss": 0.2835,
+      "step": 1443
+    },
+    {
+      "epoch": 0.9769959404600812,
+      "grad_norm": 0.48741848155098166,
+      "learning_rate": 6.066498153718735e-06,
+      "loss": 0.2843,
+      "step": 1444
+    },
+    {
+      "epoch": 0.9776725304465493,
+      "grad_norm": 0.5003209594673714,
+      "learning_rate": 6.060728060395927e-06,
+      "loss": 0.2806,
+      "step": 1445
+    },
+    {
+      "epoch": 0.9783491204330176,
+      "grad_norm": 0.5047963564545831,
+      "learning_rate": 6.054956487486212e-06,
+      "loss": 0.2886,
+      "step": 1446
+    },
+    {
+      "epoch": 0.9790257104194858,
+      "grad_norm": 0.501922286236791,
+      "learning_rate": 6.049183443040233e-06,
+      "loss": 0.2906,
+      "step": 1447
+    },
+    {
+      "epoch": 0.979702300405954,
+      "grad_norm": 0.4895979181016119,
+      "learning_rate": 6.043408935110688e-06,
+      "loss": 0.2985,
+      "step": 1448
+    },
+    {
+      "epoch": 0.9803788903924222,
+      "grad_norm": 0.4716902352862774,
+      "learning_rate": 6.037632971752315e-06,
+      "loss": 0.2773,
+      "step": 1449
+    },
+    {
+      "epoch": 0.9810554803788903,
+      "grad_norm": 0.47384416221236436,
+      "learning_rate": 6.0318555610218796e-06,
+      "loss": 0.2805,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9817320703653586,
+      "grad_norm": 0.5100330733116825,
+      "learning_rate": 6.026076710978172e-06,
+      "loss": 0.2981,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9824086603518268,
+      "grad_norm": 0.47483228559685126,
+      "learning_rate": 6.020296429681985e-06,
+      "loss": 0.2737,
+      "step": 1452
+    },
+    {
+      "epoch": 0.983085250338295,
+      "grad_norm": 0.4833900302686072,
+      "learning_rate": 6.014514725196108e-06,
+      "loss": 0.2873,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9837618403247632,
+      "grad_norm": 0.4893880963652393,
+      "learning_rate": 6.0087316055853175e-06,
+      "loss": 0.2959,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9844384303112313,
+      "grad_norm": 0.45476804016719646,
+      "learning_rate": 6.002947078916365e-06,
+      "loss": 0.2788,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9851150202976996,
+      "grad_norm": 0.4756318766606038,
+      "learning_rate": 5.997161153257963e-06,
+      "loss": 0.279,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9857916102841678,
+      "grad_norm": 0.47327318254883216,
+      "learning_rate": 5.9913738366807725e-06,
+      "loss": 0.2767,
+      "step": 1457
+    },
+    {
+      "epoch": 0.986468200270636,
+      "grad_norm": 0.5162317675367587,
+      "learning_rate": 5.985585137257401e-06,
+      "loss": 0.2964,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9871447902571042,
+      "grad_norm": 0.4920483276003604,
+      "learning_rate": 5.9797950630623834e-06,
+      "loss": 0.2829,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9878213802435724,
+      "grad_norm": 0.5021171236650912,
+      "learning_rate": 5.974003622172167e-06,
+      "loss": 0.2859,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9884979702300406,
+      "grad_norm": 0.49806185523003293,
+      "learning_rate": 5.9682108226651084e-06,
+      "loss": 0.2779,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9891745602165088,
+      "grad_norm": 0.486364859569392,
+      "learning_rate": 5.962416672621461e-06,
+      "loss": 0.2834,
+      "step": 1462
+    },
+    {
+      "epoch": 0.989851150202977,
+      "grad_norm": 0.5121618665881045,
+      "learning_rate": 5.956621180123362e-06,
+      "loss": 0.2884,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9905277401894452,
+      "grad_norm": 0.5146630787237592,
+      "learning_rate": 5.950824353254818e-06,
+      "loss": 0.2869,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9912043301759134,
+      "grad_norm": 0.4621659014716319,
+      "learning_rate": 5.945026200101702e-06,
+      "loss": 0.2658,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9918809201623816,
+      "grad_norm": 0.488169200757033,
+      "learning_rate": 5.939226728751733e-06,
+      "loss": 0.282,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9925575101488497,
+      "grad_norm": 0.47877225286649994,
+      "learning_rate": 5.933425947294466e-06,
+      "loss": 0.2764,
+      "step": 1467
+    },
+    {
+      "epoch": 0.993234100135318,
+      "grad_norm": 0.5067601350664532,
+      "learning_rate": 5.92762386382129e-06,
+      "loss": 0.2906,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9939106901217862,
+      "grad_norm": 0.5073601187310524,
+      "learning_rate": 5.921820486425409e-06,
+      "loss": 0.2844,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9945872801082544,
+      "grad_norm": 0.4985497346176548,
+      "learning_rate": 5.916015823201827e-06,
+      "loss": 0.3034,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9952638700947226,
+      "grad_norm": 0.5062140105484784,
+      "learning_rate": 5.910209882247346e-06,
+      "loss": 0.2802,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9959404600811907,
+      "grad_norm": 0.4909288164680948,
+      "learning_rate": 5.904402671660551e-06,
+      "loss": 0.2907,
+      "step": 1472
+    },
+    {
+      "epoch": 0.996617050067659,
+      "grad_norm": 0.49319087765656283,
+      "learning_rate": 5.8985941995417915e-06,
+      "loss": 0.2752,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9972936400541272,
+      "grad_norm": 0.5106637738748453,
+      "learning_rate": 5.892784473993184e-06,
+      "loss": 0.2986,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9979702300405954,
+      "grad_norm": 0.5030797409068497,
+      "learning_rate": 5.88697350311859e-06,
+      "loss": 0.3007,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9986468200270636,
+      "grad_norm": 0.4875775378395366,
+      "learning_rate": 5.88116129502361e-06,
+      "loss": 0.2834,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9993234100135318,
+      "grad_norm": 0.5202309618381707,
+      "learning_rate": 5.875347857815565e-06,
+      "loss": 0.2939,
+      "step": 1477
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5075404498605539,
+      "learning_rate": 5.869533199603498e-06,
+      "loss": 0.2944,
+      "step": 1478
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.28650158643722534,
+      "eval_runtime": 99.1623,
+      "eval_samples_per_second": 100.391,
+      "eval_steps_per_second": 0.787,
+      "step": 1478
+    },
+    {
+      "epoch": 1.0006765899864682,
+      "grad_norm": 0.51060726546731,
+      "learning_rate": 5.863717328498153e-06,
+      "loss": 0.2639,
+      "step": 1479
+    },
+    {
+      "epoch": 1.0013531799729365,
+      "grad_norm": 0.48522294338409805,
+      "learning_rate": 5.857900252611959e-06,
+      "loss": 0.2603,
+      "step": 1480
+    },
+    {
+      "epoch": 1.0020297699594045,
+      "grad_norm": 0.5154858500458256,
+      "learning_rate": 5.8520819800590345e-06,
+      "loss": 0.2632,
+      "step": 1481
+    },
+    {
+      "epoch": 1.0027063599458728,
+      "grad_norm": 0.4828291391642033,
+      "learning_rate": 5.846262518955163e-06,
+      "loss": 0.2549,
+      "step": 1482
+    },
+    {
+      "epoch": 1.003382949932341,
+      "grad_norm": 0.50546825334058,
+      "learning_rate": 5.840441877417785e-06,
+      "loss": 0.2726,
+      "step": 1483
+    },
+    {
+      "epoch": 1.0040595399188093,
+      "grad_norm": 0.4916555316201046,
+      "learning_rate": 5.83462006356599e-06,
+      "loss": 0.2538,
+      "step": 1484
+    },
+    {
+      "epoch": 1.0047361299052775,
+      "grad_norm": 0.5141388495342926,
+      "learning_rate": 5.828797085520504e-06,
+      "loss": 0.2611,
+      "step": 1485
+    },
+    {
+      "epoch": 1.0054127198917455,
+      "grad_norm": 0.5242912062651525,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.2566,
+      "step": 1486
+    },
+    {
+      "epoch": 1.0060893098782138,
+      "grad_norm": 0.5387226211959644,
+      "learning_rate": 5.8171476693394525e-06,
+      "loss": 0.2596,
+      "step": 1487
+    },
+    {
+      "epoch": 1.006765899864682,
+      "grad_norm": 0.5412146539976774,
+      "learning_rate": 5.811321247453409e-06,
+      "loss": 0.2653,
+      "step": 1488
+    },
+    {
+      "epoch": 1.0074424898511503,
+      "grad_norm": 0.5019158727405902,
+      "learning_rate": 5.805493693872696e-06,
+      "loss": 0.2634,
+      "step": 1489
+    },
+    {
+      "epoch": 1.0081190798376185,
+      "grad_norm": 0.4853047824574284,
+      "learning_rate": 5.799665016726039e-06,
+      "loss": 0.2519,
+      "step": 1490
+    },
+    {
+      "epoch": 1.0087956698240865,
+      "grad_norm": 0.5047083151623284,
+      "learning_rate": 5.793835224143737e-06,
+      "loss": 0.2602,
+      "step": 1491
+    },
+    {
+      "epoch": 1.0094722598105548,
+      "grad_norm": 0.48805511994232514,
+      "learning_rate": 5.788004324257643e-06,
+      "loss": 0.2526,
+      "step": 1492
+    },
+    {
+      "epoch": 1.010148849797023,
+      "grad_norm": 0.5044020250711329,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.2575,
+      "step": 1493
+    },
+    {
+      "epoch": 1.0108254397834913,
+      "grad_norm": 0.5032388454171268,
+      "learning_rate": 5.776339235109203e-06,
+      "loss": 0.2644,
+      "step": 1494
+    },
+    {
+      "epoch": 1.0115020297699595,
+      "grad_norm": 0.49612499503343577,
+      "learning_rate": 5.77050506211824e-06,
+      "loss": 0.2566,
+      "step": 1495
+    },
+    {
+      "epoch": 1.0121786197564275,
+      "grad_norm": 0.47980075503312547,
+      "learning_rate": 5.764669814366231e-06,
+      "loss": 0.2586,
+      "step": 1496
+    },
+    {
+      "epoch": 1.0128552097428958,
+      "grad_norm": 0.5335770655702309,
+      "learning_rate": 5.758833499992636e-06,
+      "loss": 0.2592,
+      "step": 1497
+    },
+    {
+      "epoch": 1.013531799729364,
+      "grad_norm": 0.4924318763387056,
+      "learning_rate": 5.752996127138404e-06,
+      "loss": 0.2543,
+      "step": 1498
+    },
+    {
+      "epoch": 1.0142083897158323,
+      "grad_norm": 0.49295916614633073,
+      "learning_rate": 5.747157703945966e-06,
+      "loss": 0.2546,
+      "step": 1499
+    },
+    {
+      "epoch": 1.0148849797023005,
+      "grad_norm": 0.5431753908980437,
+      "learning_rate": 5.74131823855921e-06,
+      "loss": 0.2648,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0155615696887685,
+      "grad_norm": 0.5205248997837939,
+      "learning_rate": 5.735477739123484e-06,
+      "loss": 0.26,
+      "step": 1501
+    },
+    {
+      "epoch": 1.0162381596752368,
+      "grad_norm": 0.5149647433345754,
+      "learning_rate": 5.729636213785574e-06,
+      "loss": 0.2658,
+      "step": 1502
+    },
+    {
+      "epoch": 1.016914749661705,
+      "grad_norm": 0.5242359347209952,
+      "learning_rate": 5.723793670693702e-06,
+      "loss": 0.2669,
+      "step": 1503
+    },
+    {
+      "epoch": 1.0175913396481733,
+      "grad_norm": 0.5164447432719136,
+      "learning_rate": 5.717950117997502e-06,
+      "loss": 0.2585,
+      "step": 1504
+    },
+    {
+      "epoch": 1.0182679296346413,
+      "grad_norm": 0.5119456301958378,
+      "learning_rate": 5.712105563848024e-06,
+      "loss": 0.2502,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0189445196211095,
+      "grad_norm": 0.5155585284736887,
+      "learning_rate": 5.70626001639771e-06,
+      "loss": 0.2478,
+      "step": 1506
+    },
+    {
+      "epoch": 1.0196211096075778,
+      "grad_norm": 0.48340067489982996,
+      "learning_rate": 5.70041348380039e-06,
+      "loss": 0.2537,
+      "step": 1507
+    },
+    {
+      "epoch": 1.020297699594046,
+      "grad_norm": 0.49942543196765843,
+      "learning_rate": 5.694565974211267e-06,
+      "loss": 0.2629,
+      "step": 1508
+    },
+    {
+      "epoch": 1.0209742895805143,
+      "grad_norm": 0.513073985033831,
+      "learning_rate": 5.688717495786909e-06,
+      "loss": 0.2595,
+      "step": 1509
+    },
+    {
+      "epoch": 1.0216508795669823,
+      "grad_norm": 0.4992543614191154,
+      "learning_rate": 5.6828680566852314e-06,
+      "loss": 0.2468,
+      "step": 1510
+    },
+    {
+      "epoch": 1.0223274695534506,
+      "grad_norm": 0.5126555263427828,
+      "learning_rate": 5.677017665065492e-06,
+      "loss": 0.2578,
+      "step": 1511
+    },
+    {
+      "epoch": 1.0230040595399188,
+      "grad_norm": 0.5022441434155804,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.2549,
+      "step": 1512
+    },
+    {
+      "epoch": 1.023680649526387,
+      "grad_norm": 0.5172756673917124,
+      "learning_rate": 5.665314056915493e-06,
+      "loss": 0.2659,
+      "step": 1513
+    },
+    {
+      "epoch": 1.0243572395128553,
+      "grad_norm": 0.4946482421316732,
+      "learning_rate": 5.659460856710346e-06,
+      "loss": 0.26,
+      "step": 1514
+    },
+    {
+      "epoch": 1.0250338294993233,
+      "grad_norm": 0.4941783154281281,
+      "learning_rate": 5.653606736637339e-06,
+      "loss": 0.2656,
+      "step": 1515
+    },
+    {
+      "epoch": 1.0257104194857916,
+      "grad_norm": 0.49595803188583426,
+      "learning_rate": 5.647751704862263e-06,
+      "loss": 0.2714,
+      "step": 1516
+    },
+    {
+      "epoch": 1.0263870094722598,
+      "grad_norm": 0.5039013270971785,
+      "learning_rate": 5.6418957695521735e-06,
+      "loss": 0.2624,
+      "step": 1517
+    },
+    {
+      "epoch": 1.027063599458728,
+      "grad_norm": 0.4979034679542835,
+      "learning_rate": 5.636038938875391e-06,
+      "loss": 0.2586,
+      "step": 1518
+    },
+    {
+      "epoch": 1.0277401894451963,
+      "grad_norm": 0.506775534092181,
+      "learning_rate": 5.630181221001483e-06,
+      "loss": 0.2645,
+      "step": 1519
+    },
+    {
+      "epoch": 1.0284167794316643,
+      "grad_norm": 0.4876570587576805,
+      "learning_rate": 5.624322624101255e-06,
+      "loss": 0.2446,
+      "step": 1520
+    },
+    {
+      "epoch": 1.0290933694181326,
+      "grad_norm": 0.5054996893122956,
+      "learning_rate": 5.61846315634674e-06,
+      "loss": 0.2647,
+      "step": 1521
+    },
+    {
+      "epoch": 1.0297699594046008,
+      "grad_norm": 0.5166764485800169,
+      "learning_rate": 5.612602825911179e-06,
+      "loss": 0.2658,
+      "step": 1522
+    },
+    {
+      "epoch": 1.030446549391069,
+      "grad_norm": 0.4950730006235408,
+      "learning_rate": 5.60674164096903e-06,
+      "loss": 0.2503,
+      "step": 1523
+    },
+    {
+      "epoch": 1.0311231393775373,
+      "grad_norm": 0.4941202954321686,
+      "learning_rate": 5.600879609695929e-06,
+      "loss": 0.2602,
+      "step": 1524
+    },
+    {
+      "epoch": 1.0317997293640053,
+      "grad_norm": 0.4752359619182735,
+      "learning_rate": 5.595016740268699e-06,
+      "loss": 0.2604,
+      "step": 1525
+    },
+    {
+      "epoch": 1.0324763193504736,
+      "grad_norm": 0.49217190742791445,
+      "learning_rate": 5.589153040865333e-06,
+      "loss": 0.2636,
+      "step": 1526
+    },
+    {
+      "epoch": 1.0331529093369418,
+      "grad_norm": 0.4974017237174421,
+      "learning_rate": 5.5832885196649776e-06,
+      "loss": 0.2569,
+      "step": 1527
+    },
+    {
+      "epoch": 1.03382949932341,
+      "grad_norm": 0.526869222995278,
+      "learning_rate": 5.577423184847932e-06,
+      "loss": 0.2621,
+      "step": 1528
+    },
+    {
+      "epoch": 1.0345060893098783,
+      "grad_norm": 0.5117282694347058,
+      "learning_rate": 5.57155704459562e-06,
+      "loss": 0.2535,
+      "step": 1529
+    },
+    {
+      "epoch": 1.0351826792963463,
+      "grad_norm": 0.4935521187189948,
+      "learning_rate": 5.565690107090603e-06,
+      "loss": 0.254,
+      "step": 1530
+    },
+    {
+      "epoch": 1.0358592692828146,
+      "grad_norm": 0.4855595897767847,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.2487,
+      "step": 1531
+    },
+    {
+      "epoch": 1.0365358592692828,
+      "grad_norm": 0.5226442518282415,
+      "learning_rate": 5.553953873058201e-06,
+      "loss": 0.2505,
+      "step": 1532
+    },
+    {
+      "epoch": 1.037212449255751,
+      "grad_norm": 0.4776851712654877,
+      "learning_rate": 5.5480845929014416e-06,
+      "loss": 0.2607,
+      "step": 1533
+    },
+    {
+      "epoch": 1.0378890392422193,
+      "grad_norm": 0.4922431397509027,
+      "learning_rate": 5.542214548233195e-06,
+      "loss": 0.2561,
+      "step": 1534
+    },
+    {
+      "epoch": 1.0385656292286873,
+      "grad_norm": 0.5445167838326881,
+      "learning_rate": 5.53634374724146e-06,
+      "loss": 0.2576,
+      "step": 1535
+    },
+    {
+      "epoch": 1.0392422192151556,
+      "grad_norm": 0.5106127217808536,
+      "learning_rate": 5.530472198115291e-06,
+      "loss": 0.2556,
+      "step": 1536
+    },
+    {
+      "epoch": 1.0399188092016238,
+      "grad_norm": 0.5332918538257682,
+      "learning_rate": 5.5245999090447886e-06,
+      "loss": 0.2722,
+      "step": 1537
+    },
+    {
+      "epoch": 1.040595399188092,
+      "grad_norm": 0.4935446837315038,
+      "learning_rate": 5.518726888221082e-06,
+      "loss": 0.2512,
+      "step": 1538
+    },
+    {
+      "epoch": 1.0412719891745603,
+      "grad_norm": 0.4815788911576198,
+      "learning_rate": 5.512853143836323e-06,
+      "loss": 0.2513,
+      "step": 1539
+    },
+    {
+      "epoch": 1.0419485791610283,
+      "grad_norm": 0.5128306260573813,
+      "learning_rate": 5.506978684083672e-06,
+      "loss": 0.269,
+      "step": 1540
+    },
+    {
+      "epoch": 1.0426251691474966,
+      "grad_norm": 0.4852052086999163,
+      "learning_rate": 5.501103517157288e-06,
+      "loss": 0.2645,
+      "step": 1541
+    },
+    {
+      "epoch": 1.0433017591339648,
+      "grad_norm": 0.4950112301498589,
+      "learning_rate": 5.495227651252315e-06,
+      "loss": 0.2467,
+      "step": 1542
+    },
+    {
+      "epoch": 1.043978349120433,
+      "grad_norm": 0.49361920207276094,
+      "learning_rate": 5.489351094564872e-06,
+      "loss": 0.2488,
+      "step": 1543
+    },
+    {
+      "epoch": 1.044654939106901,
+      "grad_norm": 0.5079642107266851,
+      "learning_rate": 5.483473855292043e-06,
+      "loss": 0.2656,
+      "step": 1544
+    },
+    {
+      "epoch": 1.0453315290933693,
+      "grad_norm": 0.48569966777422047,
+      "learning_rate": 5.477595941631866e-06,
+      "loss": 0.2576,
+      "step": 1545
+    },
+    {
+      "epoch": 1.0460081190798376,
+      "grad_norm": 0.489943514669429,
+      "learning_rate": 5.471717361783312e-06,
+      "loss": 0.264,
+      "step": 1546
+    },
+    {
+      "epoch": 1.0466847090663058,
+      "grad_norm": 0.5072415575087923,
+      "learning_rate": 5.46583812394629e-06,
+      "loss": 0.2557,
+      "step": 1547
+    },
+    {
+      "epoch": 1.047361299052774,
+      "grad_norm": 0.5220424683135019,
+      "learning_rate": 5.459958236321625e-06,
+      "loss": 0.2553,
+      "step": 1548
+    },
+    {
+      "epoch": 1.048037889039242,
+      "grad_norm": 0.5057274446760508,
+      "learning_rate": 5.4540777071110415e-06,
+      "loss": 0.2472,
+      "step": 1549
+    },
+    {
+      "epoch": 1.0487144790257104,
+      "grad_norm": 0.5248045917622753,
+      "learning_rate": 5.448196544517168e-06,
+      "loss": 0.2571,
+      "step": 1550
+    },
+    {
+      "epoch": 1.0493910690121786,
+      "grad_norm": 0.517828608833534,
+      "learning_rate": 5.442314756743511e-06,
+      "loss": 0.2655,
+      "step": 1551
+    },
+    {
+      "epoch": 1.0500676589986468,
+      "grad_norm": 0.4936326374250009,
+      "learning_rate": 5.436432351994452e-06,
+      "loss": 0.2629,
+      "step": 1552
+    },
+    {
+      "epoch": 1.050744248985115,
+      "grad_norm": 0.49537290657450295,
+      "learning_rate": 5.430549338475229e-06,
+      "loss": 0.257,
+      "step": 1553
+    },
+    {
+      "epoch": 1.0514208389715831,
+      "grad_norm": 0.5113633937635248,
+      "learning_rate": 5.4246657243919345e-06,
+      "loss": 0.2622,
+      "step": 1554
+    },
+    {
+      "epoch": 1.0520974289580514,
+      "grad_norm": 0.48807365074940584,
+      "learning_rate": 5.418781517951494e-06,
+      "loss": 0.2517,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0527740189445196,
+      "grad_norm": 0.4991642721551504,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.2589,
+      "step": 1556
+    },
+    {
+      "epoch": 1.0534506089309879,
+      "grad_norm": 0.48571810504654506,
+      "learning_rate": 5.4070113608310086e-06,
+      "loss": 0.2503,
+      "step": 1557
+    },
+    {
+      "epoch": 1.054127198917456,
+      "grad_norm": 0.5462498313651222,
+      "learning_rate": 5.401125426568904e-06,
+      "loss": 0.2728,
+      "step": 1558
+    },
+    {
+      "epoch": 1.0548037889039241,
+      "grad_norm": 0.5166844383366058,
+      "learning_rate": 5.395238932785514e-06,
+      "loss": 0.2571,
+      "step": 1559
+    },
+    {
+      "epoch": 1.0554803788903924,
+      "grad_norm": 0.5136902543997381,
+      "learning_rate": 5.3893518876917795e-06,
+      "loss": 0.2566,
+      "step": 1560
+    },
+    {
+      "epoch": 1.0561569688768606,
+      "grad_norm": 0.48248903353537337,
+      "learning_rate": 5.383464299499419e-06,
+      "loss": 0.2652,
+      "step": 1561
+    },
+    {
+      "epoch": 1.0568335588633289,
+      "grad_norm": 0.5159525294705649,
+      "learning_rate": 5.377576176420899e-06,
+      "loss": 0.2648,
+      "step": 1562
+    },
+    {
+      "epoch": 1.057510148849797,
+      "grad_norm": 0.5323753007462892,
+      "learning_rate": 5.371687526669439e-06,
+      "loss": 0.2684,
+      "step": 1563
+    },
+    {
+      "epoch": 1.0581867388362651,
+      "grad_norm": 0.49882119379381873,
+      "learning_rate": 5.365798358458989e-06,
+      "loss": 0.2604,
+      "step": 1564
+    },
+    {
+      "epoch": 1.0588633288227334,
+      "grad_norm": 0.5257763324503214,
+      "learning_rate": 5.359908680004227e-06,
+      "loss": 0.2593,
+      "step": 1565
+    },
+    {
+      "epoch": 1.0595399188092016,
+      "grad_norm": 0.4811128135568375,
+      "learning_rate": 5.354018499520536e-06,
+      "loss": 0.255,
+      "step": 1566
+    },
+    {
+      "epoch": 1.0602165087956699,
+      "grad_norm": 0.511984295544459,
+      "learning_rate": 5.348127825224003e-06,
+      "loss": 0.2639,
+      "step": 1567
+    },
+    {
+      "epoch": 1.060893098782138,
+      "grad_norm": 0.4987479120590218,
+      "learning_rate": 5.342236665331407e-06,
+      "loss": 0.2557,
+      "step": 1568
+    },
+    {
+      "epoch": 1.0615696887686061,
+      "grad_norm": 0.5027987514460075,
+      "learning_rate": 5.336345028060199e-06,
+      "loss": 0.2619,
+      "step": 1569
+    },
+    {
+      "epoch": 1.0622462787550744,
+      "grad_norm": 0.500482845347824,
+      "learning_rate": 5.3304529216284974e-06,
+      "loss": 0.2607,
+      "step": 1570
+    },
+    {
+      "epoch": 1.0629228687415426,
+      "grad_norm": 0.5235287050141275,
+      "learning_rate": 5.324560354255077e-06,
+      "loss": 0.2444,
+      "step": 1571
+    },
+    {
+      "epoch": 1.0635994587280109,
+      "grad_norm": 0.5265144686110922,
+      "learning_rate": 5.318667334159354e-06,
+      "loss": 0.2638,
+      "step": 1572
+    },
+    {
+      "epoch": 1.0642760487144791,
+      "grad_norm": 0.49728830025308013,
+      "learning_rate": 5.312773869561374e-06,
+      "loss": 0.2518,
+      "step": 1573
+    },
+    {
+      "epoch": 1.0649526387009471,
+      "grad_norm": 0.5164136248620481,
+      "learning_rate": 5.306879968681808e-06,
+      "loss": 0.2652,
+      "step": 1574
+    },
+    {
+      "epoch": 1.0656292286874154,
+      "grad_norm": 0.4881346147075505,
+      "learning_rate": 5.30098563974193e-06,
+      "loss": 0.2672,
+      "step": 1575
+    },
+    {
+      "epoch": 1.0663058186738836,
+      "grad_norm": 0.49607844475379165,
+      "learning_rate": 5.2950908909636144e-06,
+      "loss": 0.2539,
+      "step": 1576
+    },
+    {
+      "epoch": 1.0669824086603519,
+      "grad_norm": 0.5202292754545554,
+      "learning_rate": 5.289195730569321e-06,
+      "loss": 0.2743,
+      "step": 1577
+    },
+    {
+      "epoch": 1.0676589986468201,
+      "grad_norm": 0.5025779083659317,
+      "learning_rate": 5.2833001667820815e-06,
+      "loss": 0.2529,
+      "step": 1578
+    },
+    {
+      "epoch": 1.0683355886332881,
+      "grad_norm": 0.5367108483524559,
+      "learning_rate": 5.277404207825494e-06,
+      "loss": 0.2688,
+      "step": 1579
+    },
+    {
+      "epoch": 1.0690121786197564,
+      "grad_norm": 0.5221053603742699,
+      "learning_rate": 5.271507861923701e-06,
+      "loss": 0.2583,
+      "step": 1580
+    },
+    {
+      "epoch": 1.0696887686062246,
+      "grad_norm": 0.4823238951736119,
+      "learning_rate": 5.265611137301397e-06,
+      "loss": 0.2519,
+      "step": 1581
+    },
+    {
+      "epoch": 1.0703653585926929,
+      "grad_norm": 0.49875806307140685,
+      "learning_rate": 5.2597140421837915e-06,
+      "loss": 0.2624,
+      "step": 1582
+    },
+    {
+      "epoch": 1.0710419485791611,
+      "grad_norm": 0.5246836920487586,
+      "learning_rate": 5.253816584796617e-06,
+      "loss": 0.2617,
+      "step": 1583
+    },
+    {
+      "epoch": 1.0717185385656292,
+      "grad_norm": 0.5091984503420885,
+      "learning_rate": 5.247918773366112e-06,
+      "loss": 0.2643,
+      "step": 1584
+    },
+    {
+      "epoch": 1.0723951285520974,
+      "grad_norm": 0.5009162294615511,
+      "learning_rate": 5.242020616119008e-06,
+      "loss": 0.2575,
+      "step": 1585
+    },
+    {
+      "epoch": 1.0730717185385656,
+      "grad_norm": 0.5319307314870665,
+      "learning_rate": 5.2361221212825175e-06,
+      "loss": 0.2518,
+      "step": 1586
+    },
+    {
+      "epoch": 1.073748308525034,
+      "grad_norm": 0.5196594569440378,
+      "learning_rate": 5.230223297084325e-06,
+      "loss": 0.2597,
+      "step": 1587
+    },
+    {
+      "epoch": 1.0744248985115021,
+      "grad_norm": 0.5361380932828659,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.2597,
+      "step": 1588
+    },
+    {
+      "epoch": 1.0751014884979702,
+      "grad_norm": 0.5101496341285419,
+      "learning_rate": 5.218424693515859e-06,
+      "loss": 0.2684,
+      "step": 1589
+    },
+    {
+      "epoch": 1.0757780784844384,
+      "grad_norm": 0.5042041274901425,
+      "learning_rate": 5.212524930603205e-06,
+      "loss": 0.2491,
+      "step": 1590
+    },
+    {
+      "epoch": 1.0764546684709067,
+      "grad_norm": 0.5257168720289142,
+      "learning_rate": 5.206624871244066e-06,
+      "loss": 0.2758,
+      "step": 1591
+    },
+    {
+      "epoch": 1.077131258457375,
+      "grad_norm": 0.5162961963173214,
+      "learning_rate": 5.200724523668311e-06,
+      "loss": 0.2518,
+      "step": 1592
+    },
+    {
+      "epoch": 1.0778078484438431,
+      "grad_norm": 0.5041686271022378,
+      "learning_rate": 5.1948238961062045e-06,
+      "loss": 0.264,
+      "step": 1593
+    },
+    {
+      "epoch": 1.0784844384303112,
+      "grad_norm": 0.4830813023191563,
+      "learning_rate": 5.188922996788409e-06,
+      "loss": 0.2589,
+      "step": 1594
+    },
+    {
+      "epoch": 1.0791610284167794,
+      "grad_norm": 0.492969460596816,
+      "learning_rate": 5.183021833945962e-06,
+      "loss": 0.2552,
+      "step": 1595
+    },
+    {
+      "epoch": 1.0798376184032477,
+      "grad_norm": 0.525475830252623,
+      "learning_rate": 5.177120415810271e-06,
+      "loss": 0.257,
+      "step": 1596
+    },
+    {
+      "epoch": 1.080514208389716,
+      "grad_norm": 0.5008818924332257,
+      "learning_rate": 5.171218750613094e-06,
+      "loss": 0.2484,
+      "step": 1597
+    },
+    {
+      "epoch": 1.0811907983761841,
+      "grad_norm": 0.5238617061167149,
+      "learning_rate": 5.165316846586541e-06,
+      "loss": 0.2588,
+      "step": 1598
+    },
+    {
+      "epoch": 1.0818673883626522,
+      "grad_norm": 0.5129188163490401,
+      "learning_rate": 5.1594147119630525e-06,
+      "loss": 0.2665,
+      "step": 1599
+    },
+    {
+      "epoch": 1.0825439783491204,
+      "grad_norm": 0.5119751808391425,
+      "learning_rate": 5.153512354975388e-06,
+      "loss": 0.2566,
+      "step": 1600
+    },
+    {
+      "epoch": 1.0832205683355887,
+      "grad_norm": 0.5282705216257856,
+      "learning_rate": 5.147609783856619e-06,
+      "loss": 0.2613,
+      "step": 1601
+    },
+    {
+      "epoch": 1.083897158322057,
+      "grad_norm": 0.5338352343703437,
+      "learning_rate": 5.1417070068401165e-06,
+      "loss": 0.2627,
+      "step": 1602
+    },
+    {
+      "epoch": 1.084573748308525,
+      "grad_norm": 0.5343772558049744,
+      "learning_rate": 5.135804032159539e-06,
+      "loss": 0.2666,
+      "step": 1603
+    },
+    {
+      "epoch": 1.0852503382949932,
+      "grad_norm": 0.49784013300670565,
+      "learning_rate": 5.129900868048817e-06,
+      "loss": 0.2583,
+      "step": 1604
+    },
+    {
+      "epoch": 1.0859269282814614,
+      "grad_norm": 0.5021375106927548,
+      "learning_rate": 5.123997522742151e-06,
+      "loss": 0.2557,
+      "step": 1605
+    },
+    {
+      "epoch": 1.0866035182679297,
+      "grad_norm": 0.49916252150695184,
+      "learning_rate": 5.11809400447399e-06,
+      "loss": 0.2481,
+      "step": 1606
+    },
+    {
+      "epoch": 1.087280108254398,
+      "grad_norm": 0.5138141173363374,
+      "learning_rate": 5.112190321479026e-06,
+      "loss": 0.2655,
+      "step": 1607
+    },
+    {
+      "epoch": 1.087956698240866,
+      "grad_norm": 0.5213063131633964,
+      "learning_rate": 5.106286481992179e-06,
+      "loss": 0.25,
+      "step": 1608
+    },
+    {
+      "epoch": 1.0886332882273342,
+      "grad_norm": 0.5056192920763816,
+      "learning_rate": 5.1003824942485895e-06,
+      "loss": 0.2549,
+      "step": 1609
+    },
+    {
+      "epoch": 1.0893098782138024,
+      "grad_norm": 0.5190571077885133,
+      "learning_rate": 5.094478366483604e-06,
+      "loss": 0.2493,
+      "step": 1610
+    },
+    {
+      "epoch": 1.0899864682002707,
+      "grad_norm": 0.5037805929868497,
+      "learning_rate": 5.088574106932762e-06,
+      "loss": 0.2577,
+      "step": 1611
+    },
+    {
+      "epoch": 1.090663058186739,
+      "grad_norm": 0.506274552964166,
+      "learning_rate": 5.082669723831793e-06,
+      "loss": 0.2567,
+      "step": 1612
+    },
+    {
+      "epoch": 1.091339648173207,
+      "grad_norm": 0.516016693554664,
+      "learning_rate": 5.076765225416593e-06,
+      "loss": 0.2565,
+      "step": 1613
+    },
+    {
+      "epoch": 1.0920162381596752,
+      "grad_norm": 0.5013854774918234,
+      "learning_rate": 5.070860619923218e-06,
+      "loss": 0.2551,
+      "step": 1614
+    },
+    {
+      "epoch": 1.0926928281461434,
+      "grad_norm": 0.5007715179214884,
+      "learning_rate": 5.064955915587879e-06,
+      "loss": 0.259,
+      "step": 1615
+    },
+    {
+      "epoch": 1.0933694181326117,
+      "grad_norm": 0.5230351491638601,
+      "learning_rate": 5.059051120646924e-06,
+      "loss": 0.253,
+      "step": 1616
+    },
+    {
+      "epoch": 1.09404600811908,
+      "grad_norm": 0.5367102240519886,
+      "learning_rate": 5.053146243336819e-06,
+      "loss": 0.2516,
+      "step": 1617
+    },
+    {
+      "epoch": 1.094722598105548,
+      "grad_norm": 0.5053387492003762,
+      "learning_rate": 5.047241291894156e-06,
+      "loss": 0.2517,
+      "step": 1618
+    },
+    {
+      "epoch": 1.0953991880920162,
+      "grad_norm": 0.5175354310876213,
+      "learning_rate": 5.041336274555625e-06,
+      "loss": 0.2506,
+      "step": 1619
+    },
+    {
+      "epoch": 1.0960757780784844,
+      "grad_norm": 0.5197197282354055,
+      "learning_rate": 5.035431199558008e-06,
+      "loss": 0.2576,
+      "step": 1620
+    },
+    {
+      "epoch": 1.0967523680649527,
+      "grad_norm": 0.5092972433142781,
+      "learning_rate": 5.029526075138167e-06,
+      "loss": 0.2615,
+      "step": 1621
+    },
+    {
+      "epoch": 1.097428958051421,
+      "grad_norm": 0.5138426895321271,
+      "learning_rate": 5.0236209095330344e-06,
+      "loss": 0.2571,
+      "step": 1622
+    },
+    {
+      "epoch": 1.098105548037889,
+      "grad_norm": 0.48186497091711855,
+      "learning_rate": 5.017715710979602e-06,
+      "loss": 0.2468,
+      "step": 1623
+    },
+    {
+      "epoch": 1.0987821380243572,
+      "grad_norm": 0.5422823951364704,
+      "learning_rate": 5.011810487714901e-06,
+      "loss": 0.2573,
+      "step": 1624
+    },
+    {
+      "epoch": 1.0994587280108254,
+      "grad_norm": 0.525556017847293,
+      "learning_rate": 5.005905247976003e-06,
+      "loss": 0.2609,
+      "step": 1625
+    },
+    {
+      "epoch": 1.1001353179972937,
+      "grad_norm": 0.498331772890971,
+      "learning_rate": 5e-06,
+      "loss": 0.2541,
+      "step": 1626
+    },
+    {
+      "epoch": 1.100811907983762,
+      "grad_norm": 0.4871827771017889,
+      "learning_rate": 4.994094752024e-06,
+      "loss": 0.2473,
+      "step": 1627
+    },
+    {
+      "epoch": 1.10148849797023,
+      "grad_norm": 0.5047499090495516,
+      "learning_rate": 4.988189512285101e-06,
+      "loss": 0.2469,
+      "step": 1628
+    },
+    {
+      "epoch": 1.1021650879566982,
+      "grad_norm": 0.4918338461626717,
+      "learning_rate": 4.9822842890204e-06,
+      "loss": 0.2615,
+      "step": 1629
+    },
+    {
+      "epoch": 1.1028416779431665,
+      "grad_norm": 0.4924626017815475,
+      "learning_rate": 4.976379090466966e-06,
+      "loss": 0.2451,
+      "step": 1630
+    },
+    {
+      "epoch": 1.1035182679296347,
+      "grad_norm": 0.5040254358106789,
+      "learning_rate": 4.970473924861835e-06,
+      "loss": 0.2547,
+      "step": 1631
+    },
+    {
+      "epoch": 1.104194857916103,
+      "grad_norm": 0.49785508219043667,
+      "learning_rate": 4.964568800441993e-06,
+      "loss": 0.2526,
+      "step": 1632
+    },
+    {
+      "epoch": 1.104871447902571,
+      "grad_norm": 0.4970096806208725,
+      "learning_rate": 4.958663725444376e-06,
+      "loss": 0.2478,
+      "step": 1633
+    },
+    {
+      "epoch": 1.1055480378890392,
+      "grad_norm": 0.5243443363832974,
+      "learning_rate": 4.952758708105845e-06,
+      "loss": 0.2584,
+      "step": 1634
+    },
+    {
+      "epoch": 1.1062246278755075,
+      "grad_norm": 0.5000974690062139,
+      "learning_rate": 4.946853756663182e-06,
+      "loss": 0.2547,
+      "step": 1635
+    },
+    {
+      "epoch": 1.1069012178619757,
+      "grad_norm": 0.4921802989150207,
+      "learning_rate": 4.940948879353078e-06,
+      "loss": 0.2493,
+      "step": 1636
+    },
+    {
+      "epoch": 1.1075778078484437,
+      "grad_norm": 0.5024267393433528,
+      "learning_rate": 4.935044084412122e-06,
+      "loss": 0.2547,
+      "step": 1637
+    },
+    {
+      "epoch": 1.108254397834912,
+      "grad_norm": 0.5146934464685443,
+      "learning_rate": 4.929139380076784e-06,
+      "loss": 0.2588,
+      "step": 1638
+    },
+    {
+      "epoch": 1.1089309878213802,
+      "grad_norm": 0.48303675835309196,
+      "learning_rate": 4.923234774583408e-06,
+      "loss": 0.2468,
+      "step": 1639
+    },
+    {
+      "epoch": 1.1096075778078485,
+      "grad_norm": 0.5126981553724491,
+      "learning_rate": 4.917330276168208e-06,
+      "loss": 0.2618,
+      "step": 1640
+    },
+    {
+      "epoch": 1.1102841677943167,
+      "grad_norm": 0.5121169268856428,
+      "learning_rate": 4.911425893067239e-06,
+      "loss": 0.2654,
+      "step": 1641
+    },
+    {
+      "epoch": 1.1109607577807847,
+      "grad_norm": 0.5328612770462663,
+      "learning_rate": 4.905521633516399e-06,
+      "loss": 0.2666,
+      "step": 1642
+    },
+    {
+      "epoch": 1.111637347767253,
+      "grad_norm": 0.4801032047820691,
+      "learning_rate": 4.899617505751411e-06,
+      "loss": 0.2567,
+      "step": 1643
+    },
+    {
+      "epoch": 1.1123139377537212,
+      "grad_norm": 0.5005058767597181,
+      "learning_rate": 4.8937135180078236e-06,
+      "loss": 0.265,
+      "step": 1644
+    },
+    {
+      "epoch": 1.1129905277401895,
+      "grad_norm": 0.5304173040230161,
+      "learning_rate": 4.887809678520976e-06,
+      "loss": 0.2635,
+      "step": 1645
+    },
+    {
+      "epoch": 1.1136671177266577,
+      "grad_norm": 0.5320477689677321,
+      "learning_rate": 4.8819059955260105e-06,
+      "loss": 0.261,
+      "step": 1646
+    },
+    {
+      "epoch": 1.1143437077131257,
+      "grad_norm": 0.4837319363076158,
+      "learning_rate": 4.87600247725785e-06,
+      "loss": 0.2543,
+      "step": 1647
+    },
+    {
+      "epoch": 1.115020297699594,
+      "grad_norm": 0.5007477878096721,
+      "learning_rate": 4.870099131951185e-06,
+      "loss": 0.2663,
+      "step": 1648
+    },
+    {
+      "epoch": 1.1156968876860622,
+      "grad_norm": 0.4968784277869084,
+      "learning_rate": 4.864195967840464e-06,
+      "loss": 0.2416,
+      "step": 1649
+    },
+    {
+      "epoch": 1.1163734776725305,
+      "grad_norm": 0.5415854955440947,
+      "learning_rate": 4.858292993159884e-06,
+      "loss": 0.2675,
+      "step": 1650
+    },
+    {
+      "epoch": 1.1170500676589987,
+      "grad_norm": 0.46425990435250053,
+      "learning_rate": 4.852390216143383e-06,
+      "loss": 0.2466,
+      "step": 1651
+    },
+    {
+      "epoch": 1.1177266576454667,
+      "grad_norm": 0.5050676184901075,
+      "learning_rate": 4.846487645024614e-06,
+      "loss": 0.262,
+      "step": 1652
+    },
+    {
+      "epoch": 1.118403247631935,
+      "grad_norm": 0.5137162989912462,
+      "learning_rate": 4.84058528803695e-06,
+      "loss": 0.2549,
+      "step": 1653
+    },
+    {
+      "epoch": 1.1190798376184032,
+      "grad_norm": 0.5216078741120943,
+      "learning_rate": 4.8346831534134595e-06,
+      "loss": 0.2615,
+      "step": 1654
+    },
+    {
+      "epoch": 1.1197564276048715,
+      "grad_norm": 0.5044046958575569,
+      "learning_rate": 4.828781249386907e-06,
+      "loss": 0.255,
+      "step": 1655
+    },
+    {
+      "epoch": 1.1204330175913397,
+      "grad_norm": 0.5318282308039745,
+      "learning_rate": 4.822879584189732e-06,
+      "loss": 0.26,
+      "step": 1656
+    },
+    {
+      "epoch": 1.1211096075778078,
+      "grad_norm": 0.5121782156524974,
+      "learning_rate": 4.816978166054039e-06,
+      "loss": 0.2596,
+      "step": 1657
+    },
+    {
+      "epoch": 1.121786197564276,
+      "grad_norm": 0.5502817125469918,
+      "learning_rate": 4.811077003211592e-06,
+      "loss": 0.2606,
+      "step": 1658
+    },
+    {
+      "epoch": 1.1224627875507442,
+      "grad_norm": 0.5064862994359235,
+      "learning_rate": 4.805176103893796e-06,
+      "loss": 0.2605,
+      "step": 1659
+    },
+    {
+      "epoch": 1.1231393775372125,
+      "grad_norm": 0.5433040737733111,
+      "learning_rate": 4.799275476331692e-06,
+      "loss": 0.2572,
+      "step": 1660
+    },
+    {
+      "epoch": 1.1238159675236807,
+      "grad_norm": 0.5133486138969722,
+      "learning_rate": 4.793375128755934e-06,
+      "loss": 0.2597,
+      "step": 1661
+    },
+    {
+      "epoch": 1.1244925575101488,
+      "grad_norm": 0.5148453705274467,
+      "learning_rate": 4.787475069396796e-06,
+      "loss": 0.2602,
+      "step": 1662
+    },
+    {
+      "epoch": 1.125169147496617,
+      "grad_norm": 0.5125707132728878,
+      "learning_rate": 4.781575306484142e-06,
+      "loss": 0.254,
+      "step": 1663
+    },
+    {
+      "epoch": 1.1258457374830853,
+      "grad_norm": 0.5135944266008912,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.2492,
+      "step": 1664
+    },
+    {
+      "epoch": 1.1265223274695535,
+      "grad_norm": 0.49664005461150373,
+      "learning_rate": 4.769776702915676e-06,
+      "loss": 0.2622,
+      "step": 1665
+    },
+    {
+      "epoch": 1.1271989174560217,
+      "grad_norm": 0.4931405459471713,
+      "learning_rate": 4.763877878717484e-06,
+      "loss": 0.2502,
+      "step": 1666
+    },
+    {
+      "epoch": 1.1278755074424898,
+      "grad_norm": 0.5450956355509277,
+      "learning_rate": 4.757979383880994e-06,
+      "loss": 0.2686,
+      "step": 1667
+    },
+    {
+      "epoch": 1.128552097428958,
+      "grad_norm": 0.5033780926327562,
+      "learning_rate": 4.752081226633888e-06,
+      "loss": 0.2487,
+      "step": 1668
+    },
+    {
+      "epoch": 1.1292286874154263,
+      "grad_norm": 0.5670653911972409,
+      "learning_rate": 4.7461834152033845e-06,
+      "loss": 0.2643,
+      "step": 1669
+    },
+    {
+      "epoch": 1.1299052774018945,
+      "grad_norm": 0.5013656916034278,
+      "learning_rate": 4.740285957816211e-06,
+      "loss": 0.2642,
+      "step": 1670
+    },
+    {
+      "epoch": 1.1305818673883627,
+      "grad_norm": 0.5401201149979813,
+      "learning_rate": 4.734388862698605e-06,
+      "loss": 0.2691,
+      "step": 1671
+    },
+    {
+      "epoch": 1.1312584573748308,
+      "grad_norm": 0.5091597125402244,
+      "learning_rate": 4.728492138076299e-06,
+      "loss": 0.2543,
+      "step": 1672
+    },
+    {
+      "epoch": 1.131935047361299,
+      "grad_norm": 0.5161314084342067,
+      "learning_rate": 4.722595792174508e-06,
+      "loss": 0.2522,
+      "step": 1673
+    },
+    {
+      "epoch": 1.1326116373477673,
+      "grad_norm": 0.491697330010215,
+      "learning_rate": 4.71669983321792e-06,
+      "loss": 0.2486,
+      "step": 1674
+    },
+    {
+      "epoch": 1.1332882273342355,
+      "grad_norm": 0.48283529684877546,
+      "learning_rate": 4.710804269430681e-06,
+      "loss": 0.2508,
+      "step": 1675
+    },
+    {
+      "epoch": 1.1339648173207038,
+      "grad_norm": 0.5384262487910895,
+      "learning_rate": 4.704909109036387e-06,
+      "loss": 0.2624,
+      "step": 1676
+    },
+    {
+      "epoch": 1.1346414073071718,
+      "grad_norm": 0.5166882826104905,
+      "learning_rate": 4.699014360258071e-06,
+      "loss": 0.262,
+      "step": 1677
+    },
+    {
+      "epoch": 1.13531799729364,
+      "grad_norm": 0.5053442873745394,
+      "learning_rate": 4.693120031318194e-06,
+      "loss": 0.2558,
+      "step": 1678
+    },
+    {
+      "epoch": 1.1359945872801083,
+      "grad_norm": 0.5465522197847549,
+      "learning_rate": 4.687226130438626e-06,
+      "loss": 0.2635,
+      "step": 1679
+    },
+    {
+      "epoch": 1.1366711772665765,
+      "grad_norm": 0.5310651911941321,
+      "learning_rate": 4.681332665840647e-06,
+      "loss": 0.2568,
+      "step": 1680
+    },
+    {
+      "epoch": 1.1373477672530448,
+      "grad_norm": 0.4868333796252307,
+      "learning_rate": 4.675439645744924e-06,
+      "loss": 0.238,
+      "step": 1681
+    },
+    {
+      "epoch": 1.1380243572395128,
+      "grad_norm": 0.5594297195059,
+      "learning_rate": 4.669547078371503e-06,
+      "loss": 0.2628,
+      "step": 1682
+    },
+    {
+      "epoch": 1.138700947225981,
+      "grad_norm": 0.5625241083232347,
+      "learning_rate": 4.663654971939802e-06,
+      "loss": 0.2608,
+      "step": 1683
+    },
+    {
+      "epoch": 1.1393775372124493,
+      "grad_norm": 0.5111030458466095,
+      "learning_rate": 4.657763334668594e-06,
+      "loss": 0.26,
+      "step": 1684
+    },
+    {
+      "epoch": 1.1400541271989175,
+      "grad_norm": 0.502389550756731,
+      "learning_rate": 4.651872174775999e-06,
+      "loss": 0.2583,
+      "step": 1685
+    },
+    {
+      "epoch": 1.1407307171853858,
+      "grad_norm": 0.5129871071683002,
+      "learning_rate": 4.645981500479466e-06,
+      "loss": 0.2581,
+      "step": 1686
+    },
+    {
+      "epoch": 1.1414073071718538,
+      "grad_norm": 0.5310953077232031,
+      "learning_rate": 4.640091319995775e-06,
+      "loss": 0.2559,
+      "step": 1687
+    },
+    {
+      "epoch": 1.142083897158322,
+      "grad_norm": 0.5427289536038403,
+      "learning_rate": 4.634201641541013e-06,
+      "loss": 0.2558,
+      "step": 1688
+    },
+    {
+      "epoch": 1.1427604871447903,
+      "grad_norm": 0.496594037125896,
+      "learning_rate": 4.628312473330563e-06,
+      "loss": 0.2549,
+      "step": 1689
+    },
+    {
+      "epoch": 1.1434370771312585,
+      "grad_norm": 0.4800932875635843,
+      "learning_rate": 4.622423823579102e-06,
+      "loss": 0.2471,
+      "step": 1690
+    },
+    {
+      "epoch": 1.1441136671177268,
+      "grad_norm": 0.5052849646497213,
+      "learning_rate": 4.616535700500583e-06,
+      "loss": 0.2471,
+      "step": 1691
+    },
+    {
+      "epoch": 1.1447902571041948,
+      "grad_norm": 0.5373631514041572,
+      "learning_rate": 4.610648112308221e-06,
+      "loss": 0.2525,
+      "step": 1692
+    },
+    {
+      "epoch": 1.145466847090663,
+      "grad_norm": 0.5629742787069011,
+      "learning_rate": 4.604761067214488e-06,
+      "loss": 0.2631,
+      "step": 1693
+    },
+    {
+      "epoch": 1.1461434370771313,
+      "grad_norm": 0.5300140616046646,
+      "learning_rate": 4.598874573431097e-06,
+      "loss": 0.262,
+      "step": 1694
+    },
+    {
+      "epoch": 1.1468200270635995,
+      "grad_norm": 0.5026420969609792,
+      "learning_rate": 4.592988639168993e-06,
+      "loss": 0.2553,
+      "step": 1695
+    },
+    {
+      "epoch": 1.1474966170500678,
+      "grad_norm": 0.5064328694755703,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.2576,
+      "step": 1696
+    },
+    {
+      "epoch": 1.1481732070365358,
+      "grad_norm": 0.491911982827002,
+      "learning_rate": 4.581218482048506e-06,
+      "loss": 0.2601,
+      "step": 1697
+    },
+    {
+      "epoch": 1.148849797023004,
+      "grad_norm": 0.512298805342294,
+      "learning_rate": 4.575334275608067e-06,
+      "loss": 0.257,
+      "step": 1698
+    },
+    {
+      "epoch": 1.1495263870094723,
+      "grad_norm": 0.5156140481480181,
+      "learning_rate": 4.569450661524773e-06,
+      "loss": 0.2647,
+      "step": 1699
+    },
+    {
+      "epoch": 1.1502029769959405,
+      "grad_norm": 0.5067396307682654,
+      "learning_rate": 4.563567648005551e-06,
+      "loss": 0.2553,
+      "step": 1700
+    },
+    {
+      "epoch": 1.1508795669824086,
+      "grad_norm": 0.48961621441238384,
+      "learning_rate": 4.5576852432564896e-06,
+      "loss": 0.2549,
+      "step": 1701
+    },
+    {
+      "epoch": 1.1515561569688768,
+      "grad_norm": 0.5004838812612049,
+      "learning_rate": 4.551803455482833e-06,
+      "loss": 0.2482,
+      "step": 1702
+    },
+    {
+      "epoch": 1.152232746955345,
+      "grad_norm": 0.49201406680517323,
+      "learning_rate": 4.545922292888959e-06,
+      "loss": 0.2443,
+      "step": 1703
+    },
+    {
+      "epoch": 1.1529093369418133,
+      "grad_norm": 0.5031835232424022,
+      "learning_rate": 4.540041763678377e-06,
+      "loss": 0.2644,
+      "step": 1704
+    },
+    {
+      "epoch": 1.1535859269282815,
+      "grad_norm": 0.5118679962949959,
+      "learning_rate": 4.5341618760537094e-06,
+      "loss": 0.2615,
+      "step": 1705
+    },
+    {
+      "epoch": 1.1542625169147496,
+      "grad_norm": 0.5593693433792524,
+      "learning_rate": 4.528282638216689e-06,
+      "loss": 0.2637,
+      "step": 1706
+    },
+    {
+      "epoch": 1.1549391069012178,
+      "grad_norm": 0.5041826486207662,
+      "learning_rate": 4.522404058368137e-06,
+      "loss": 0.2581,
+      "step": 1707
+    },
+    {
+      "epoch": 1.155615696887686,
+      "grad_norm": 0.4979293341167642,
+      "learning_rate": 4.516526144707957e-06,
+      "loss": 0.252,
+      "step": 1708
+    },
+    {
+      "epoch": 1.1562922868741543,
+      "grad_norm": 0.4957451868058958,
+      "learning_rate": 4.51064890543513e-06,
+      "loss": 0.2411,
+      "step": 1709
+    },
+    {
+      "epoch": 1.1569688768606226,
+      "grad_norm": 0.5508387162548588,
+      "learning_rate": 4.504772348747687e-06,
+      "loss": 0.2477,
+      "step": 1710
+    },
+    {
+      "epoch": 1.1576454668470906,
+      "grad_norm": 0.5276727213578922,
+      "learning_rate": 4.498896482842715e-06,
+      "loss": 0.2641,
+      "step": 1711
+    },
+    {
+      "epoch": 1.1583220568335588,
+      "grad_norm": 0.5015035470097373,
+      "learning_rate": 4.493021315916328e-06,
+      "loss": 0.2589,
+      "step": 1712
+    },
+    {
+      "epoch": 1.158998646820027,
+      "grad_norm": 0.5280603506427982,
+      "learning_rate": 4.4871468561636784e-06,
+      "loss": 0.2593,
+      "step": 1713
+    },
+    {
+      "epoch": 1.1596752368064953,
+      "grad_norm": 0.5064385857095365,
+      "learning_rate": 4.481273111778919e-06,
+      "loss": 0.259,
+      "step": 1714
+    },
+    {
+      "epoch": 1.1603518267929636,
+      "grad_norm": 0.48254018741750543,
+      "learning_rate": 4.475400090955214e-06,
+      "loss": 0.2446,
+      "step": 1715
+    },
+    {
+      "epoch": 1.1610284167794316,
+      "grad_norm": 0.49358753832257013,
+      "learning_rate": 4.46952780188471e-06,
+      "loss": 0.2538,
+      "step": 1716
+    },
+    {
+      "epoch": 1.1617050067658998,
+      "grad_norm": 0.5077380971843242,
+      "learning_rate": 4.463656252758543e-06,
+      "loss": 0.2614,
+      "step": 1717
+    },
+    {
+      "epoch": 1.162381596752368,
+      "grad_norm": 0.5204132303062271,
+      "learning_rate": 4.457785451766808e-06,
+      "loss": 0.268,
+      "step": 1718
+    },
+    {
+      "epoch": 1.1630581867388363,
+      "grad_norm": 0.5126255197171146,
+      "learning_rate": 4.451915407098559e-06,
+      "loss": 0.2554,
+      "step": 1719
+    },
+    {
+      "epoch": 1.1637347767253043,
+      "grad_norm": 0.5121255622811841,
+      "learning_rate": 4.446046126941801e-06,
+      "loss": 0.2632,
+      "step": 1720
+    },
+    {
+      "epoch": 1.1644113667117726,
+      "grad_norm": 0.5176060735034389,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.2573,
+      "step": 1721
+    },
+    {
+      "epoch": 1.1650879566982408,
+      "grad_norm": 0.48961903454565553,
+      "learning_rate": 4.4343098929094e-06,
+      "loss": 0.2516,
+      "step": 1722
+    },
+    {
+      "epoch": 1.165764546684709,
+      "grad_norm": 0.4933211249391572,
+      "learning_rate": 4.42844295540438e-06,
+      "loss": 0.2485,
+      "step": 1723
+    },
+    {
+      "epoch": 1.1664411366711773,
+      "grad_norm": 0.49841500042151204,
+      "learning_rate": 4.42257681515207e-06,
+      "loss": 0.2541,
+      "step": 1724
+    },
+    {
+      "epoch": 1.1671177266576453,
+      "grad_norm": 0.5000158318218074,
+      "learning_rate": 4.416711480335023e-06,
+      "loss": 0.2629,
+      "step": 1725
+    },
+    {
+      "epoch": 1.1677943166441136,
+      "grad_norm": 0.49689680978946166,
+      "learning_rate": 4.410846959134667e-06,
+      "loss": 0.2592,
+      "step": 1726
+    },
+    {
+      "epoch": 1.1684709066305818,
+      "grad_norm": 0.5030709756474844,
+      "learning_rate": 4.404983259731302e-06,
+      "loss": 0.2655,
+      "step": 1727
+    },
+    {
+      "epoch": 1.16914749661705,
+      "grad_norm": 0.5020415238850887,
+      "learning_rate": 4.399120390304072e-06,
+      "loss": 0.2636,
+      "step": 1728
+    },
+    {
+      "epoch": 1.1698240866035183,
+      "grad_norm": 0.5206551523437551,
+      "learning_rate": 4.393258359030972e-06,
+      "loss": 0.2586,
+      "step": 1729
+    },
+    {
+      "epoch": 1.1705006765899864,
+      "grad_norm": 0.5069818093046231,
+      "learning_rate": 4.3873971740888205e-06,
+      "loss": 0.2531,
+      "step": 1730
+    },
+    {
+      "epoch": 1.1711772665764546,
+      "grad_norm": 0.4977689551514409,
+      "learning_rate": 4.381536843653262e-06,
+      "loss": 0.2546,
+      "step": 1731
+    },
+    {
+      "epoch": 1.1718538565629228,
+      "grad_norm": 0.5041599589994815,
+      "learning_rate": 4.375677375898746e-06,
+      "loss": 0.2579,
+      "step": 1732
+    },
+    {
+      "epoch": 1.172530446549391,
+      "grad_norm": 0.5303067044908718,
+      "learning_rate": 4.36981877899852e-06,
+      "loss": 0.2627,
+      "step": 1733
+    },
+    {
+      "epoch": 1.1732070365358593,
+      "grad_norm": 0.5153798937581642,
+      "learning_rate": 4.3639610611246106e-06,
+      "loss": 0.2614,
+      "step": 1734
+    },
+    {
+      "epoch": 1.1738836265223274,
+      "grad_norm": 0.5052900516951899,
+      "learning_rate": 4.358104230447827e-06,
+      "loss": 0.2521,
+      "step": 1735
+    },
+    {
+      "epoch": 1.1745602165087956,
+      "grad_norm": 0.539582229740494,
+      "learning_rate": 4.352248295137739e-06,
+      "loss": 0.2582,
+      "step": 1736
+    },
+    {
+      "epoch": 1.1752368064952639,
+      "grad_norm": 0.5106670667328143,
+      "learning_rate": 4.346393263362662e-06,
+      "loss": 0.2595,
+      "step": 1737
+    },
+    {
+      "epoch": 1.175913396481732,
+      "grad_norm": 0.5117018364081974,
+      "learning_rate": 4.340539143289655e-06,
+      "loss": 0.2624,
+      "step": 1738
+    },
+    {
+      "epoch": 1.1765899864682003,
+      "grad_norm": 0.5184436220665856,
+      "learning_rate": 4.334685943084509e-06,
+      "loss": 0.2579,
+      "step": 1739
+    },
+    {
+      "epoch": 1.1772665764546684,
+      "grad_norm": 0.5014616172037716,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.2632,
+      "step": 1740
+    },
+    {
+      "epoch": 1.1779431664411366,
+      "grad_norm": 0.5027436328325222,
+      "learning_rate": 4.322982334934509e-06,
+      "loss": 0.2572,
+      "step": 1741
+    },
+    {
+      "epoch": 1.1786197564276049,
+      "grad_norm": 0.5194848141058733,
+      "learning_rate": 4.31713194331477e-06,
+      "loss": 0.2583,
+      "step": 1742
+    },
+    {
+      "epoch": 1.179296346414073,
+      "grad_norm": 0.5202865628309163,
+      "learning_rate": 4.311282504213094e-06,
+      "loss": 0.2534,
+      "step": 1743
+    },
+    {
+      "epoch": 1.1799729364005414,
+      "grad_norm": 0.5951786186308918,
+      "learning_rate": 4.305434025788735e-06,
+      "loss": 0.2718,
+      "step": 1744
+    },
+    {
+      "epoch": 1.1806495263870094,
+      "grad_norm": 0.5137257450379361,
+      "learning_rate": 4.299586516199611e-06,
+      "loss": 0.2722,
+      "step": 1745
+    },
+    {
+      "epoch": 1.1813261163734776,
+      "grad_norm": 0.4972419315370277,
+      "learning_rate": 4.293739983602292e-06,
+      "loss": 0.2537,
+      "step": 1746
+    },
+    {
+      "epoch": 1.1820027063599459,
+      "grad_norm": 0.4984357483688694,
+      "learning_rate": 4.2878944361519795e-06,
+      "loss": 0.2581,
+      "step": 1747
+    },
+    {
+      "epoch": 1.182679296346414,
+      "grad_norm": 0.5183405126881376,
+      "learning_rate": 4.282049882002499e-06,
+      "loss": 0.2604,
+      "step": 1748
+    },
+    {
+      "epoch": 1.1833558863328824,
+      "grad_norm": 0.4937717895841812,
+      "learning_rate": 4.2762063293063e-06,
+      "loss": 0.2588,
+      "step": 1749
+    },
+    {
+      "epoch": 1.1840324763193504,
+      "grad_norm": 0.5061098543165853,
+      "learning_rate": 4.270363786214427e-06,
+      "loss": 0.2595,
+      "step": 1750
+    },
+    {
+      "epoch": 1.1847090663058186,
+      "grad_norm": 0.5038214128078949,
+      "learning_rate": 4.264522260876518e-06,
+      "loss": 0.264,
+      "step": 1751
+    },
+    {
+      "epoch": 1.1853856562922869,
+      "grad_norm": 0.5060322734397749,
+      "learning_rate": 4.25868176144079e-06,
+      "loss": 0.2634,
+      "step": 1752
+    },
+    {
+      "epoch": 1.1860622462787551,
+      "grad_norm": 0.536693140826359,
+      "learning_rate": 4.252842296054036e-06,
+      "loss": 0.2603,
+      "step": 1753
+    },
+    {
+      "epoch": 1.1867388362652234,
+      "grad_norm": 0.5043023768346515,
+      "learning_rate": 4.247003872861598e-06,
+      "loss": 0.2505,
+      "step": 1754
+    },
+    {
+      "epoch": 1.1874154262516914,
+      "grad_norm": 0.5236988732914363,
+      "learning_rate": 4.241166500007368e-06,
+      "loss": 0.2565,
+      "step": 1755
+    },
+    {
+      "epoch": 1.1880920162381596,
+      "grad_norm": 0.49642741527693146,
+      "learning_rate": 4.23533018563377e-06,
+      "loss": 0.26,
+      "step": 1756
+    },
+    {
+      "epoch": 1.1887686062246279,
+      "grad_norm": 0.5135614107721375,
+      "learning_rate": 4.229494937881761e-06,
+      "loss": 0.2448,
+      "step": 1757
+    },
+    {
+      "epoch": 1.1894451962110961,
+      "grad_norm": 0.4770296726551818,
+      "learning_rate": 4.223660764890799e-06,
+      "loss": 0.2426,
+      "step": 1758
+    },
+    {
+      "epoch": 1.1901217861975644,
+      "grad_norm": 0.5143384903089994,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.2517,
+      "step": 1759
+    },
+    {
+      "epoch": 1.1907983761840324,
+      "grad_norm": 0.523608913134806,
+      "learning_rate": 4.211995675742358e-06,
+      "loss": 0.2565,
+      "step": 1760
+    },
+    {
+      "epoch": 1.1914749661705006,
+      "grad_norm": 0.5235621490095477,
+      "learning_rate": 4.206164775856265e-06,
+      "loss": 0.2657,
+      "step": 1761
+    },
+    {
+      "epoch": 1.1921515561569689,
+      "grad_norm": 0.5125923913622754,
+      "learning_rate": 4.2003349832739624e-06,
+      "loss": 0.2699,
+      "step": 1762
+    },
+    {
+      "epoch": 1.1928281461434371,
+      "grad_norm": 0.47536666712531794,
+      "learning_rate": 4.194506306127305e-06,
+      "loss": 0.2536,
+      "step": 1763
+    },
+    {
+      "epoch": 1.1935047361299054,
+      "grad_norm": 0.5303543924754485,
+      "learning_rate": 4.1886787525465914e-06,
+      "loss": 0.2569,
+      "step": 1764
+    },
+    {
+      "epoch": 1.1941813261163734,
+      "grad_norm": 0.5203958487068109,
+      "learning_rate": 4.182852330660549e-06,
+      "loss": 0.245,
+      "step": 1765
+    },
+    {
+      "epoch": 1.1948579161028416,
+      "grad_norm": 0.5018838005571208,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.2602,
+      "step": 1766
+    },
+    {
+      "epoch": 1.19553450608931,
+      "grad_norm": 0.47239113938911464,
+      "learning_rate": 4.171202914479498e-06,
+      "loss": 0.2515,
+      "step": 1767
+    },
+    {
+      "epoch": 1.1962110960757781,
+      "grad_norm": 0.5155080653958877,
+      "learning_rate": 4.165379936434011e-06,
+      "loss": 0.2695,
+      "step": 1768
+    },
+    {
+      "epoch": 1.1968876860622464,
+      "grad_norm": 0.5078251428692967,
+      "learning_rate": 4.159558122582217e-06,
+      "loss": 0.2474,
+      "step": 1769
+    },
+    {
+      "epoch": 1.1975642760487144,
+      "grad_norm": 0.49723873506874966,
+      "learning_rate": 4.153737481044838e-06,
+      "loss": 0.2529,
+      "step": 1770
+    },
+    {
+      "epoch": 1.1982408660351827,
+      "grad_norm": 0.5130883374232702,
+      "learning_rate": 4.147918019940967e-06,
+      "loss": 0.2672,
+      "step": 1771
+    },
+    {
+      "epoch": 1.198917456021651,
+      "grad_norm": 0.5217621421134566,
+      "learning_rate": 4.142099747388042e-06,
+      "loss": 0.2647,
+      "step": 1772
+    },
+    {
+      "epoch": 1.1995940460081191,
+      "grad_norm": 0.510081113862726,
+      "learning_rate": 4.13628267150185e-06,
+      "loss": 0.2592,
+      "step": 1773
+    },
+    {
+      "epoch": 1.2002706359945874,
+      "grad_norm": 0.5093825665450684,
+      "learning_rate": 4.1304668003965016e-06,
+      "loss": 0.2547,
+      "step": 1774
+    },
+    {
+      "epoch": 1.2009472259810554,
+      "grad_norm": 0.521897642509576,
+      "learning_rate": 4.124652142184436e-06,
+      "loss": 0.2651,
+      "step": 1775
+    },
+    {
+      "epoch": 1.2016238159675237,
+      "grad_norm": 0.5144735599042707,
+      "learning_rate": 4.118838704976392e-06,
+      "loss": 0.2544,
+      "step": 1776
+    },
+    {
+      "epoch": 1.202300405953992,
+      "grad_norm": 0.5204216586326839,
+      "learning_rate": 4.1130264968814104e-06,
+      "loss": 0.2666,
+      "step": 1777
+    },
+    {
+      "epoch": 1.2029769959404601,
+      "grad_norm": 0.4888175902165075,
+      "learning_rate": 4.107215526006818e-06,
+      "loss": 0.2591,
+      "step": 1778
+    },
+    {
+      "epoch": 1.2036535859269284,
+      "grad_norm": 0.5329646472873127,
+      "learning_rate": 4.10140580045821e-06,
+      "loss": 0.266,
+      "step": 1779
+    },
+    {
+      "epoch": 1.2043301759133964,
+      "grad_norm": 0.5064352871756844,
+      "learning_rate": 4.0955973283394525e-06,
+      "loss": 0.268,
+      "step": 1780
+    },
+    {
+      "epoch": 1.2050067658998647,
+      "grad_norm": 0.49035833189125577,
+      "learning_rate": 4.089790117752655e-06,
+      "loss": 0.2555,
+      "step": 1781
+    },
+    {
+      "epoch": 1.205683355886333,
+      "grad_norm": 0.507704851622085,
+      "learning_rate": 4.083984176798175e-06,
+      "loss": 0.255,
+      "step": 1782
+    },
+    {
+      "epoch": 1.2063599458728012,
+      "grad_norm": 0.5337200501381372,
+      "learning_rate": 4.078179513574593e-06,
+      "loss": 0.2608,
+      "step": 1783
+    },
+    {
+      "epoch": 1.2070365358592694,
+      "grad_norm": 0.47533825488045384,
+      "learning_rate": 4.072376136178712e-06,
+      "loss": 0.2485,
+      "step": 1784
+    },
+    {
+      "epoch": 1.2077131258457374,
+      "grad_norm": 0.5313790264797449,
+      "learning_rate": 4.066574052705535e-06,
+      "loss": 0.2603,
+      "step": 1785
+    },
+    {
+      "epoch": 1.2083897158322057,
+      "grad_norm": 0.507060234081431,
+      "learning_rate": 4.06077327124827e-06,
+      "loss": 0.2464,
+      "step": 1786
+    },
+    {
+      "epoch": 1.209066305818674,
+      "grad_norm": 0.5170909943216785,
+      "learning_rate": 4.054973799898299e-06,
+      "loss": 0.2588,
+      "step": 1787
+    },
+    {
+      "epoch": 1.2097428958051422,
+      "grad_norm": 0.4925085806795408,
+      "learning_rate": 4.049175646745182e-06,
+      "loss": 0.2505,
+      "step": 1788
+    },
+    {
+      "epoch": 1.2104194857916104,
+      "grad_norm": 0.5076099145761229,
+      "learning_rate": 4.04337881987664e-06,
+      "loss": 0.2566,
+      "step": 1789
+    },
+    {
+      "epoch": 1.2110960757780784,
+      "grad_norm": 0.5162697173615535,
+      "learning_rate": 4.03758332737854e-06,
+      "loss": 0.2581,
+      "step": 1790
+    },
+    {
+      "epoch": 1.2117726657645467,
+      "grad_norm": 0.5176931287296272,
+      "learning_rate": 4.031789177334895e-06,
+      "loss": 0.2556,
+      "step": 1791
+    },
+    {
+      "epoch": 1.212449255751015,
+      "grad_norm": 0.48662804925224434,
+      "learning_rate": 4.025996377827836e-06,
+      "loss": 0.254,
+      "step": 1792
+    },
+    {
+      "epoch": 1.2131258457374832,
+      "grad_norm": 0.49338456705212524,
+      "learning_rate": 4.020204936937618e-06,
+      "loss": 0.2595,
+      "step": 1793
+    },
+    {
+      "epoch": 1.2138024357239512,
+      "grad_norm": 0.512715944424132,
+      "learning_rate": 4.0144148627426e-06,
+      "loss": 0.2596,
+      "step": 1794
+    },
+    {
+      "epoch": 1.2144790257104194,
+      "grad_norm": 0.507344081870421,
+      "learning_rate": 4.008626163319227e-06,
+      "loss": 0.2674,
+      "step": 1795
+    },
+    {
+      "epoch": 1.2151556156968877,
+      "grad_norm": 0.49551201924704097,
+      "learning_rate": 4.002838846742039e-06,
+      "loss": 0.2508,
+      "step": 1796
+    },
+    {
+      "epoch": 1.215832205683356,
+      "grad_norm": 0.5245847550091759,
+      "learning_rate": 3.997052921083637e-06,
+      "loss": 0.2512,
+      "step": 1797
+    },
+    {
+      "epoch": 1.2165087956698242,
+      "grad_norm": 0.510423178609213,
+      "learning_rate": 3.991268394414685e-06,
+      "loss": 0.2493,
+      "step": 1798
+    },
+    {
+      "epoch": 1.2171853856562922,
+      "grad_norm": 0.504299980936725,
+      "learning_rate": 3.985485274803894e-06,
+      "loss": 0.2522,
+      "step": 1799
+    },
+    {
+      "epoch": 1.2178619756427604,
+      "grad_norm": 0.5250488107385729,
+      "learning_rate": 3.979703570318017e-06,
+      "loss": 0.2664,
+      "step": 1800
+    },
+    {
+      "epoch": 1.2185385656292287,
+      "grad_norm": 0.5244183885911718,
+      "learning_rate": 3.973923289021829e-06,
+      "loss": 0.2698,
+      "step": 1801
+    },
+    {
+      "epoch": 1.219215155615697,
+      "grad_norm": 0.5190699739699128,
+      "learning_rate": 3.968144438978121e-06,
+      "loss": 0.256,
+      "step": 1802
+    },
+    {
+      "epoch": 1.2198917456021652,
+      "grad_norm": 0.537855540238927,
+      "learning_rate": 3.962367028247686e-06,
+      "loss": 0.2656,
+      "step": 1803
+    },
+    {
+      "epoch": 1.2205683355886332,
+      "grad_norm": 0.5137983371052295,
+      "learning_rate": 3.956591064889313e-06,
+      "loss": 0.2584,
+      "step": 1804
+    },
+    {
+      "epoch": 1.2212449255751014,
+      "grad_norm": 0.5110270015624713,
+      "learning_rate": 3.950816556959769e-06,
+      "loss": 0.2649,
+      "step": 1805
+    },
+    {
+      "epoch": 1.2219215155615697,
+      "grad_norm": 0.4846873231718402,
+      "learning_rate": 3.94504351251379e-06,
+      "loss": 0.2619,
+      "step": 1806
+    },
+    {
+      "epoch": 1.222598105548038,
+      "grad_norm": 0.506098866624769,
+      "learning_rate": 3.939271939604075e-06,
+      "loss": 0.2599,
+      "step": 1807
+    },
+    {
+      "epoch": 1.2232746955345062,
+      "grad_norm": 0.5272175154017164,
+      "learning_rate": 3.9335018462812664e-06,
+      "loss": 0.2599,
+      "step": 1808
+    },
+    {
+      "epoch": 1.2239512855209742,
+      "grad_norm": 0.49929013845779774,
+      "learning_rate": 3.9277332405939446e-06,
+      "loss": 0.248,
+      "step": 1809
+    },
+    {
+      "epoch": 1.2246278755074425,
+      "grad_norm": 0.5436022357422624,
+      "learning_rate": 3.921966130588612e-06,
+      "loss": 0.2647,
+      "step": 1810
+    },
+    {
+      "epoch": 1.2253044654939107,
+      "grad_norm": 0.5371500005499671,
+      "learning_rate": 3.916200524309693e-06,
+      "loss": 0.2524,
+      "step": 1811
+    },
+    {
+      "epoch": 1.225981055480379,
+      "grad_norm": 0.49704361901067057,
+      "learning_rate": 3.910436429799503e-06,
+      "loss": 0.255,
+      "step": 1812
+    },
+    {
+      "epoch": 1.226657645466847,
+      "grad_norm": 0.5088644935733247,
+      "learning_rate": 3.904673855098256e-06,
+      "loss": 0.25,
+      "step": 1813
+    },
+    {
+      "epoch": 1.2273342354533152,
+      "grad_norm": 0.5444718951751712,
+      "learning_rate": 3.898912808244043e-06,
+      "loss": 0.2568,
+      "step": 1814
+    },
+    {
+      "epoch": 1.2280108254397835,
+      "grad_norm": 0.5148433432491903,
+      "learning_rate": 3.893153297272829e-06,
+      "loss": 0.2564,
+      "step": 1815
+    },
+    {
+      "epoch": 1.2286874154262517,
+      "grad_norm": 0.517224945211232,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.2521,
+      "step": 1816
+    },
+    {
+      "epoch": 1.22936400541272,
+      "grad_norm": 0.4996902485431754,
+      "learning_rate": 3.881638915112508e-06,
+      "loss": 0.2513,
+      "step": 1817
+    },
+    {
+      "epoch": 1.230040595399188,
+      "grad_norm": 0.5105241785191917,
+      "learning_rate": 3.875884059984571e-06,
+      "loss": 0.2624,
+      "step": 1818
+    },
+    {
+      "epoch": 1.2307171853856562,
+      "grad_norm": 0.5100054866361706,
+      "learning_rate": 3.870130772861939e-06,
+      "loss": 0.2509,
+      "step": 1819
+    },
+    {
+      "epoch": 1.2313937753721245,
+      "grad_norm": 0.5058884704451506,
+      "learning_rate": 3.864379061769749e-06,
+      "loss": 0.257,
+      "step": 1820
+    },
+    {
+      "epoch": 1.2320703653585927,
+      "grad_norm": 0.5300597007473856,
+      "learning_rate": 3.858628934730939e-06,
+      "loss": 0.2648,
+      "step": 1821
+    },
+    {
+      "epoch": 1.232746955345061,
+      "grad_norm": 0.4957082627888165,
+      "learning_rate": 3.852880399766243e-06,
+      "loss": 0.2546,
+      "step": 1822
+    },
+    {
+      "epoch": 1.233423545331529,
+      "grad_norm": 0.5171564094810134,
+      "learning_rate": 3.847133464894165e-06,
+      "loss": 0.2588,
+      "step": 1823
+    },
+    {
+      "epoch": 1.2341001353179972,
+      "grad_norm": 0.5224357374762719,
+      "learning_rate": 3.8413881381309845e-06,
+      "loss": 0.2514,
+      "step": 1824
+    },
+    {
+      "epoch": 1.2347767253044655,
+      "grad_norm": 0.4901736206262608,
+      "learning_rate": 3.835644427490732e-06,
+      "loss": 0.2492,
+      "step": 1825
+    },
+    {
+      "epoch": 1.2354533152909337,
+      "grad_norm": 0.5583590174880381,
+      "learning_rate": 3.829902340985189e-06,
+      "loss": 0.2606,
+      "step": 1826
+    },
+    {
+      "epoch": 1.236129905277402,
+      "grad_norm": 0.5100767767537174,
+      "learning_rate": 3.824161886623869e-06,
+      "loss": 0.2626,
+      "step": 1827
+    },
+    {
+      "epoch": 1.23680649526387,
+      "grad_norm": 0.4999214242215081,
+      "learning_rate": 3.818423072414007e-06,
+      "loss": 0.2616,
+      "step": 1828
+    },
+    {
+      "epoch": 1.2374830852503382,
+      "grad_norm": 0.471141037799205,
+      "learning_rate": 3.8126859063605576e-06,
+      "loss": 0.2408,
+      "step": 1829
+    },
+    {
+      "epoch": 1.2381596752368065,
+      "grad_norm": 0.4914768382314962,
+      "learning_rate": 3.8069503964661656e-06,
+      "loss": 0.2539,
+      "step": 1830
+    },
+    {
+      "epoch": 1.2388362652232747,
+      "grad_norm": 0.5320700974920602,
+      "learning_rate": 3.8012165507311756e-06,
+      "loss": 0.2601,
+      "step": 1831
+    },
+    {
+      "epoch": 1.239512855209743,
+      "grad_norm": 0.5124310997201873,
+      "learning_rate": 3.795484377153601e-06,
+      "loss": 0.2579,
+      "step": 1832
+    },
+    {
+      "epoch": 1.240189445196211,
+      "grad_norm": 0.5116516268883082,
+      "learning_rate": 3.7897538837291347e-06,
+      "loss": 0.2522,
+      "step": 1833
+    },
+    {
+      "epoch": 1.2408660351826792,
+      "grad_norm": 0.5064810073228588,
+      "learning_rate": 3.7840250784511147e-06,
+      "loss": 0.2586,
+      "step": 1834
+    },
+    {
+      "epoch": 1.2415426251691475,
+      "grad_norm": 0.49594297691047745,
+      "learning_rate": 3.778297969310529e-06,
+      "loss": 0.2536,
+      "step": 1835
+    },
+    {
+      "epoch": 1.2422192151556157,
+      "grad_norm": 0.5079098542254323,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.2587,
+      "step": 1836
+    },
+    {
+      "epoch": 1.242895805142084,
+      "grad_norm": 0.5248540629145394,
+      "learning_rate": 3.7668488713937822e-06,
+      "loss": 0.2586,
+      "step": 1837
+    },
+    {
+      "epoch": 1.243572395128552,
+      "grad_norm": 0.5189067079687815,
+      "learning_rate": 3.7611268985877213e-06,
+      "loss": 0.265,
+      "step": 1838
+    },
+    {
+      "epoch": 1.2442489851150202,
+      "grad_norm": 0.5125063277270316,
+      "learning_rate": 3.755406653859276e-06,
+      "loss": 0.2518,
+      "step": 1839
+    },
+    {
+      "epoch": 1.2449255751014885,
+      "grad_norm": 0.4811430711403407,
+      "learning_rate": 3.749688145187497e-06,
+      "loss": 0.2499,
+      "step": 1840
+    },
+    {
+      "epoch": 1.2456021650879567,
+      "grad_norm": 0.4682922314156268,
+      "learning_rate": 3.7439713805490087e-06,
+      "loss": 0.245,
+      "step": 1841
+    },
+    {
+      "epoch": 1.246278755074425,
+      "grad_norm": 0.5113652337103233,
+      "learning_rate": 3.738256367918004e-06,
+      "loss": 0.2602,
+      "step": 1842
+    },
+    {
+      "epoch": 1.246955345060893,
+      "grad_norm": 0.5125284756097966,
+      "learning_rate": 3.73254311526623e-06,
+      "loss": 0.2482,
+      "step": 1843
+    },
+    {
+      "epoch": 1.2476319350473613,
+      "grad_norm": 0.5087220786579383,
+      "learning_rate": 3.7268316305629836e-06,
+      "loss": 0.2511,
+      "step": 1844
+    },
+    {
+      "epoch": 1.2483085250338295,
+      "grad_norm": 0.5001820378871923,
+      "learning_rate": 3.721121921775092e-06,
+      "loss": 0.2439,
+      "step": 1845
+    },
+    {
+      "epoch": 1.2489851150202977,
+      "grad_norm": 0.49514836565132114,
+      "learning_rate": 3.7154139968669043e-06,
+      "loss": 0.256,
+      "step": 1846
+    },
+    {
+      "epoch": 1.249661705006766,
+      "grad_norm": 0.4956651066298383,
+      "learning_rate": 3.709707863800286e-06,
+      "loss": 0.2586,
+      "step": 1847
+    },
+    {
+      "epoch": 1.250338294993234,
+      "grad_norm": 0.5052638874406074,
+      "learning_rate": 3.704003530534597e-06,
+      "loss": 0.2535,
+      "step": 1848
+    },
+    {
+      "epoch": 1.2510148849797023,
+      "grad_norm": 0.5100026455657584,
+      "learning_rate": 3.6983010050266934e-06,
+      "loss": 0.2536,
+      "step": 1849
+    },
+    {
+      "epoch": 1.2516914749661705,
+      "grad_norm": 0.4965340709135591,
+      "learning_rate": 3.6926002952309015e-06,
+      "loss": 0.2586,
+      "step": 1850
+    },
+    {
+      "epoch": 1.2523680649526387,
+      "grad_norm": 0.5037214523257226,
+      "learning_rate": 3.686901409099023e-06,
+      "loss": 0.2626,
+      "step": 1851
+    },
+    {
+      "epoch": 1.253044654939107,
+      "grad_norm": 0.49586156695860967,
+      "learning_rate": 3.681204354580313e-06,
+      "loss": 0.2533,
+      "step": 1852
+    },
+    {
+      "epoch": 1.253721244925575,
+      "grad_norm": 0.5290914935382336,
+      "learning_rate": 3.67550913962147e-06,
+      "loss": 0.2521,
+      "step": 1853
+    },
+    {
+      "epoch": 1.2543978349120433,
+      "grad_norm": 0.5013357161747359,
+      "learning_rate": 3.669815772166625e-06,
+      "loss": 0.2597,
+      "step": 1854
+    },
+    {
+      "epoch": 1.2550744248985115,
+      "grad_norm": 0.5255567047370956,
+      "learning_rate": 3.6641242601573378e-06,
+      "loss": 0.2583,
+      "step": 1855
+    },
+    {
+      "epoch": 1.2557510148849798,
+      "grad_norm": 0.5080531283553643,
+      "learning_rate": 3.658434611532578e-06,
+      "loss": 0.2515,
+      "step": 1856
+    },
+    {
+      "epoch": 1.256427604871448,
+      "grad_norm": 0.519173363481449,
+      "learning_rate": 3.6527468342287104e-06,
+      "loss": 0.2461,
+      "step": 1857
+    },
+    {
+      "epoch": 1.257104194857916,
+      "grad_norm": 0.5314077276241241,
+      "learning_rate": 3.6470609361794972e-06,
+      "loss": 0.263,
+      "step": 1858
+    },
+    {
+      "epoch": 1.2577807848443843,
+      "grad_norm": 0.510935369032836,
+      "learning_rate": 3.6413769253160776e-06,
+      "loss": 0.2513,
+      "step": 1859
+    },
+    {
+      "epoch": 1.2584573748308525,
+      "grad_norm": 0.5197182724852086,
+      "learning_rate": 3.635694809566954e-06,
+      "loss": 0.2599,
+      "step": 1860
+    },
+    {
+      "epoch": 1.2591339648173208,
+      "grad_norm": 0.5057190002702041,
+      "learning_rate": 3.6300145968579876e-06,
+      "loss": 0.2632,
+      "step": 1861
+    },
+    {
+      "epoch": 1.259810554803789,
+      "grad_norm": 0.5274864813833785,
+      "learning_rate": 3.624336295112388e-06,
+      "loss": 0.2515,
+      "step": 1862
+    },
+    {
+      "epoch": 1.260487144790257,
+      "grad_norm": 0.517826961027545,
+      "learning_rate": 3.6186599122506972e-06,
+      "loss": 0.2424,
+      "step": 1863
+    },
+    {
+      "epoch": 1.2611637347767253,
+      "grad_norm": 0.5432389636554358,
+      "learning_rate": 3.6129854561907786e-06,
+      "loss": 0.2601,
+      "step": 1864
+    },
+    {
+      "epoch": 1.2618403247631935,
+      "grad_norm": 0.5236779915192686,
+      "learning_rate": 3.6073129348478087e-06,
+      "loss": 0.2619,
+      "step": 1865
+    },
+    {
+      "epoch": 1.2625169147496618,
+      "grad_norm": 0.47780949888442537,
+      "learning_rate": 3.6016423561342707e-06,
+      "loss": 0.2514,
+      "step": 1866
+    },
+    {
+      "epoch": 1.26319350473613,
+      "grad_norm": 0.49118278704650886,
+      "learning_rate": 3.5959737279599306e-06,
+      "loss": 0.2467,
+      "step": 1867
+    },
+    {
+      "epoch": 1.263870094722598,
+      "grad_norm": 0.5089370678545724,
+      "learning_rate": 3.5903070582318356e-06,
+      "loss": 0.2642,
+      "step": 1868
+    },
+    {
+      "epoch": 1.2645466847090663,
+      "grad_norm": 0.5032559037737842,
+      "learning_rate": 3.584642354854305e-06,
+      "loss": 0.2491,
+      "step": 1869
+    },
+    {
+      "epoch": 1.2652232746955345,
+      "grad_norm": 0.5219844862234095,
+      "learning_rate": 3.5789796257289117e-06,
+      "loss": 0.2516,
+      "step": 1870
+    },
+    {
+      "epoch": 1.2658998646820028,
+      "grad_norm": 0.5056609517960806,
+      "learning_rate": 3.573318878754475e-06,
+      "loss": 0.2547,
+      "step": 1871
+    },
+    {
+      "epoch": 1.266576454668471,
+      "grad_norm": 0.48633569855961434,
+      "learning_rate": 3.567660121827048e-06,
+      "loss": 0.2488,
+      "step": 1872
+    },
+    {
+      "epoch": 1.267253044654939,
+      "grad_norm": 0.511245040498462,
+      "learning_rate": 3.562003362839914e-06,
+      "loss": 0.2532,
+      "step": 1873
+    },
+    {
+      "epoch": 1.2679296346414073,
+      "grad_norm": 0.5179498691412971,
+      "learning_rate": 3.5563486096835643e-06,
+      "loss": 0.2545,
+      "step": 1874
+    },
+    {
+      "epoch": 1.2686062246278755,
+      "grad_norm": 0.5106844901130884,
+      "learning_rate": 3.550695870245689e-06,
+      "loss": 0.2526,
+      "step": 1875
+    },
+    {
+      "epoch": 1.2692828146143438,
+      "grad_norm": 0.5074586430691448,
+      "learning_rate": 3.545045152411178e-06,
+      "loss": 0.2686,
+      "step": 1876
+    },
+    {
+      "epoch": 1.269959404600812,
+      "grad_norm": 0.499836714673921,
+      "learning_rate": 3.5393964640620967e-06,
+      "loss": 0.2594,
+      "step": 1877
+    },
+    {
+      "epoch": 1.27063599458728,
+      "grad_norm": 0.5289866669517533,
+      "learning_rate": 3.533749813077677e-06,
+      "loss": 0.266,
+      "step": 1878
+    },
+    {
+      "epoch": 1.2713125845737483,
+      "grad_norm": 0.5022609279265148,
+      "learning_rate": 3.5281052073343126e-06,
+      "loss": 0.2511,
+      "step": 1879
+    },
+    {
+      "epoch": 1.2719891745602165,
+      "grad_norm": 0.49523365349264414,
+      "learning_rate": 3.5224626547055463e-06,
+      "loss": 0.2496,
+      "step": 1880
+    },
+    {
+      "epoch": 1.2726657645466848,
+      "grad_norm": 0.48999965555512254,
+      "learning_rate": 3.516822163062052e-06,
+      "loss": 0.2531,
+      "step": 1881
+    },
+    {
+      "epoch": 1.273342354533153,
+      "grad_norm": 0.5180077013883351,
+      "learning_rate": 3.5111837402716297e-06,
+      "loss": 0.2524,
+      "step": 1882
+    },
+    {
+      "epoch": 1.274018944519621,
+      "grad_norm": 0.5122027227263035,
+      "learning_rate": 3.505547394199196e-06,
+      "loss": 0.2605,
+      "step": 1883
+    },
+    {
+      "epoch": 1.2746955345060893,
+      "grad_norm": 0.4890603054537694,
+      "learning_rate": 3.499913132706771e-06,
+      "loss": 0.2516,
+      "step": 1884
+    },
+    {
+      "epoch": 1.2753721244925575,
+      "grad_norm": 0.49271484389787423,
+      "learning_rate": 3.4942809636534637e-06,
+      "loss": 0.2508,
+      "step": 1885
+    },
+    {
+      "epoch": 1.2760487144790258,
+      "grad_norm": 0.5003212098974307,
+      "learning_rate": 3.4886508948954656e-06,
+      "loss": 0.2537,
+      "step": 1886
+    },
+    {
+      "epoch": 1.276725304465494,
+      "grad_norm": 0.50051452890456,
+      "learning_rate": 3.4830229342860434e-06,
+      "loss": 0.2546,
+      "step": 1887
+    },
+    {
+      "epoch": 1.277401894451962,
+      "grad_norm": 0.4883659321953847,
+      "learning_rate": 3.4773970896755167e-06,
+      "loss": 0.2413,
+      "step": 1888
+    },
+    {
+      "epoch": 1.2780784844384303,
+      "grad_norm": 0.4961302596873515,
+      "learning_rate": 3.4717733689112545e-06,
+      "loss": 0.2545,
+      "step": 1889
+    },
+    {
+      "epoch": 1.2787550744248986,
+      "grad_norm": 0.5136695579416148,
+      "learning_rate": 3.466151779837665e-06,
+      "loss": 0.2577,
+      "step": 1890
+    },
+    {
+      "epoch": 1.2794316644113666,
+      "grad_norm": 0.49946046852024917,
+      "learning_rate": 3.4605323302961857e-06,
+      "loss": 0.258,
+      "step": 1891
+    },
+    {
+      "epoch": 1.280108254397835,
+      "grad_norm": 0.5049757277720484,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.262,
+      "step": 1892
+    },
+    {
+      "epoch": 1.280784844384303,
+      "grad_norm": 0.5102676098749941,
+      "learning_rate": 3.4492998811603557e-06,
+      "loss": 0.2574,
+      "step": 1893
+    },
+    {
+      "epoch": 1.2814614343707713,
+      "grad_norm": 0.4781011341123066,
+      "learning_rate": 3.4436868972339073e-06,
+      "loss": 0.2468,
+      "step": 1894
+    },
+    {
+      "epoch": 1.2821380243572396,
+      "grad_norm": 0.5013206387466711,
+      "learning_rate": 3.4380760841753534e-06,
+      "loss": 0.2516,
+      "step": 1895
+    },
+    {
+      "epoch": 1.2828146143437076,
+      "grad_norm": 0.47000062253253333,
+      "learning_rate": 3.4324674498110956e-06,
+      "loss": 0.2503,
+      "step": 1896
+    },
+    {
+      "epoch": 1.283491204330176,
+      "grad_norm": 0.5100314032117335,
+      "learning_rate": 3.426861001964496e-06,
+      "loss": 0.256,
+      "step": 1897
+    },
+    {
+      "epoch": 1.284167794316644,
+      "grad_norm": 0.4915782546339698,
+      "learning_rate": 3.4212567484558735e-06,
+      "loss": 0.252,
+      "step": 1898
+    },
+    {
+      "epoch": 1.2848443843031123,
+      "grad_norm": 0.48403896933803947,
+      "learning_rate": 3.4156546971024783e-06,
+      "loss": 0.2571,
+      "step": 1899
+    },
+    {
+      "epoch": 1.2855209742895806,
+      "grad_norm": 0.49555112454337596,
+      "learning_rate": 3.4100548557184944e-06,
+      "loss": 0.2515,
+      "step": 1900
+    },
+    {
+      "epoch": 1.2861975642760486,
+      "grad_norm": 0.481872309029285,
+      "learning_rate": 3.404457232115017e-06,
+      "loss": 0.2521,
+      "step": 1901
+    },
+    {
+      "epoch": 1.2868741542625168,
+      "grad_norm": 0.49390471491581833,
+      "learning_rate": 3.3988618341000566e-06,
+      "loss": 0.2526,
+      "step": 1902
+    },
+    {
+      "epoch": 1.287550744248985,
+      "grad_norm": 0.5094180662362029,
+      "learning_rate": 3.3932686694785104e-06,
+      "loss": 0.2598,
+      "step": 1903
+    },
+    {
+      "epoch": 1.2882273342354533,
+      "grad_norm": 0.48249772472030156,
+      "learning_rate": 3.3876777460521647e-06,
+      "loss": 0.2443,
+      "step": 1904
+    },
+    {
+      "epoch": 1.2889039242219216,
+      "grad_norm": 0.5266137229858894,
+      "learning_rate": 3.382089071619683e-06,
+      "loss": 0.2549,
+      "step": 1905
+    },
+    {
+      "epoch": 1.2895805142083896,
+      "grad_norm": 0.5261446187996301,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 0.2556,
+      "step": 1906
+    },
+    {
+      "epoch": 1.2902571041948578,
+      "grad_norm": 0.49649304442086756,
+      "learning_rate": 3.3709185009152424e-06,
+      "loss": 0.2541,
+      "step": 1907
+    },
+    {
+      "epoch": 1.290933694181326,
+      "grad_norm": 0.49783896208812384,
+      "learning_rate": 3.3653366202248738e-06,
+      "loss": 0.2456,
+      "step": 1908
+    },
+    {
+      "epoch": 1.2916102841677943,
+      "grad_norm": 0.5134276909682299,
+      "learning_rate": 3.3597570196915275e-06,
+      "loss": 0.2515,
+      "step": 1909
+    },
+    {
+      "epoch": 1.2922868741542626,
+      "grad_norm": 0.48464252037946703,
+      "learning_rate": 3.3541797070980663e-06,
+      "loss": 0.2541,
+      "step": 1910
+    },
+    {
+      "epoch": 1.2929634641407306,
+      "grad_norm": 0.47948584532400657,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 0.2456,
+      "step": 1911
+    },
+    {
+      "epoch": 1.2936400541271988,
+      "grad_norm": 0.5026635191376861,
+      "learning_rate": 3.3430319768462956e-06,
+      "loss": 0.2556,
+      "step": 1912
+    },
+    {
+      "epoch": 1.294316644113667,
+      "grad_norm": 0.5125079474629946,
+      "learning_rate": 3.3374615747377165e-06,
+      "loss": 0.2588,
+      "step": 1913
+    },
+    {
+      "epoch": 1.2949932341001353,
+      "grad_norm": 0.519774797740816,
+      "learning_rate": 3.331893491668464e-06,
+      "loss": 0.2556,
+      "step": 1914
+    },
+    {
+      "epoch": 1.2956698240866036,
+      "grad_norm": 0.5249809865535788,
+      "learning_rate": 3.3263277354053346e-06,
+      "loss": 0.2659,
+      "step": 1915
+    },
+    {
+      "epoch": 1.2963464140730716,
+      "grad_norm": 0.5111909182366542,
+      "learning_rate": 3.3207643137118872e-06,
+      "loss": 0.2518,
+      "step": 1916
+    },
+    {
+      "epoch": 1.2970230040595399,
+      "grad_norm": 0.5115331267898925,
+      "learning_rate": 3.3152032343484187e-06,
+      "loss": 0.2676,
+      "step": 1917
+    },
+    {
+      "epoch": 1.297699594046008,
+      "grad_norm": 0.5072597638616144,
+      "learning_rate": 3.309644505071959e-06,
+      "loss": 0.2593,
+      "step": 1918
+    },
+    {
+      "epoch": 1.2983761840324763,
+      "grad_norm": 0.5256051448644492,
+      "learning_rate": 3.3040881336362584e-06,
+      "loss": 0.2506,
+      "step": 1919
+    },
+    {
+      "epoch": 1.2990527740189446,
+      "grad_norm": 0.5467417724333338,
+      "learning_rate": 3.298534127791785e-06,
+      "loss": 0.2603,
+      "step": 1920
+    },
+    {
+      "epoch": 1.2997293640054126,
+      "grad_norm": 0.5001281538896736,
+      "learning_rate": 3.2929824952857014e-06,
+      "loss": 0.2527,
+      "step": 1921
+    },
+    {
+      "epoch": 1.3004059539918809,
+      "grad_norm": 0.498130126416776,
+      "learning_rate": 3.2874332438618607e-06,
+      "loss": 0.2601,
+      "step": 1922
+    },
+    {
+      "epoch": 1.301082543978349,
+      "grad_norm": 0.518506492470925,
+      "learning_rate": 3.2818863812607937e-06,
+      "loss": 0.2559,
+      "step": 1923
+    },
+    {
+      "epoch": 1.3017591339648173,
+      "grad_norm": 0.5075533581843087,
+      "learning_rate": 3.276341915219704e-06,
+      "loss": 0.2551,
+      "step": 1924
+    },
+    {
+      "epoch": 1.3024357239512856,
+      "grad_norm": 0.5177892931316658,
+      "learning_rate": 3.2707998534724473e-06,
+      "loss": 0.2564,
+      "step": 1925
+    },
+    {
+      "epoch": 1.3031123139377536,
+      "grad_norm": 0.5182854437172116,
+      "learning_rate": 3.2652602037495247e-06,
+      "loss": 0.2548,
+      "step": 1926
+    },
+    {
+      "epoch": 1.3037889039242219,
+      "grad_norm": 0.5151527970863518,
+      "learning_rate": 3.259722973778078e-06,
+      "loss": 0.2571,
+      "step": 1927
+    },
+    {
+      "epoch": 1.30446549391069,
+      "grad_norm": 0.5155426535812653,
+      "learning_rate": 3.254188171281871e-06,
+      "loss": 0.2489,
+      "step": 1928
+    },
+    {
+      "epoch": 1.3051420838971584,
+      "grad_norm": 0.4986708790829614,
+      "learning_rate": 3.2486558039812803e-06,
+      "loss": 0.2525,
+      "step": 1929
+    },
+    {
+      "epoch": 1.3058186738836266,
+      "grad_norm": 0.5180282900673404,
+      "learning_rate": 3.2431258795932863e-06,
+      "loss": 0.2547,
+      "step": 1930
+    },
+    {
+      "epoch": 1.3064952638700946,
+      "grad_norm": 0.520834366420758,
+      "learning_rate": 3.2375984058314647e-06,
+      "loss": 0.2544,
+      "step": 1931
+    },
+    {
+      "epoch": 1.3071718538565629,
+      "grad_norm": 0.5121617516871944,
+      "learning_rate": 3.232073390405969e-06,
+      "loss": 0.2486,
+      "step": 1932
+    },
+    {
+      "epoch": 1.3078484438430311,
+      "grad_norm": 0.5300477483497522,
+      "learning_rate": 3.2265508410235247e-06,
+      "loss": 0.2498,
+      "step": 1933
+    },
+    {
+      "epoch": 1.3085250338294994,
+      "grad_norm": 0.5175006613283677,
+      "learning_rate": 3.2210307653874175e-06,
+      "loss": 0.2592,
+      "step": 1934
+    },
+    {
+      "epoch": 1.3092016238159676,
+      "grad_norm": 0.503315925383524,
+      "learning_rate": 3.215513171197486e-06,
+      "loss": 0.2447,
+      "step": 1935
+    },
+    {
+      "epoch": 1.3098782138024356,
+      "grad_norm": 0.4964304374405437,
+      "learning_rate": 3.2099980661501016e-06,
+      "loss": 0.2555,
+      "step": 1936
+    },
+    {
+      "epoch": 1.3105548037889039,
+      "grad_norm": 0.5075654333702144,
+      "learning_rate": 3.2044854579381657e-06,
+      "loss": 0.2523,
+      "step": 1937
+    },
+    {
+      "epoch": 1.3112313937753721,
+      "grad_norm": 0.5342417776286681,
+      "learning_rate": 3.1989753542511016e-06,
+      "loss": 0.2513,
+      "step": 1938
+    },
+    {
+      "epoch": 1.3119079837618404,
+      "grad_norm": 0.49380597904301043,
+      "learning_rate": 3.1934677627748333e-06,
+      "loss": 0.2407,
+      "step": 1939
+    },
+    {
+      "epoch": 1.3125845737483086,
+      "grad_norm": 0.4961794273769398,
+      "learning_rate": 3.1879626911917806e-06,
+      "loss": 0.2558,
+      "step": 1940
+    },
+    {
+      "epoch": 1.3132611637347766,
+      "grad_norm": 0.49818025780293207,
+      "learning_rate": 3.1824601471808504e-06,
+      "loss": 0.2538,
+      "step": 1941
+    },
+    {
+      "epoch": 1.3139377537212449,
+      "grad_norm": 0.5220475038680646,
+      "learning_rate": 3.1769601384174274e-06,
+      "loss": 0.2668,
+      "step": 1942
+    },
+    {
+      "epoch": 1.3146143437077131,
+      "grad_norm": 0.5259803096159635,
+      "learning_rate": 3.1714626725733534e-06,
+      "loss": 0.2559,
+      "step": 1943
+    },
+    {
+      "epoch": 1.3152909336941814,
+      "grad_norm": 0.4970554953468933,
+      "learning_rate": 3.165967757316925e-06,
+      "loss": 0.2557,
+      "step": 1944
+    },
+    {
+      "epoch": 1.3159675236806496,
+      "grad_norm": 0.5088587261617094,
+      "learning_rate": 3.1604754003128857e-06,
+      "loss": 0.2647,
+      "step": 1945
+    },
+    {
+      "epoch": 1.3166441136671176,
+      "grad_norm": 0.5093845727514164,
+      "learning_rate": 3.154985609222405e-06,
+      "loss": 0.2501,
+      "step": 1946
+    },
+    {
+      "epoch": 1.317320703653586,
+      "grad_norm": 0.5182330330698063,
+      "learning_rate": 3.149498391703075e-06,
+      "loss": 0.2625,
+      "step": 1947
+    },
+    {
+      "epoch": 1.3179972936400541,
+      "grad_norm": 0.5046117446605803,
+      "learning_rate": 3.1440137554088957e-06,
+      "loss": 0.2554,
+      "step": 1948
+    },
+    {
+      "epoch": 1.3186738836265224,
+      "grad_norm": 0.5100145321625769,
+      "learning_rate": 3.1385317079902743e-06,
+      "loss": 0.2543,
+      "step": 1949
+    },
+    {
+      "epoch": 1.3193504736129906,
+      "grad_norm": 0.5010385627102185,
+      "learning_rate": 3.1330522570939987e-06,
+      "loss": 0.254,
+      "step": 1950
+    },
+    {
+      "epoch": 1.3200270635994586,
+      "grad_norm": 0.532185148190828,
+      "learning_rate": 3.1275754103632385e-06,
+      "loss": 0.2681,
+      "step": 1951
+    },
+    {
+      "epoch": 1.320703653585927,
+      "grad_norm": 0.5071842598191438,
+      "learning_rate": 3.1221011754375275e-06,
+      "loss": 0.2572,
+      "step": 1952
+    },
+    {
+      "epoch": 1.3213802435723951,
+      "grad_norm": 0.49101665614133394,
+      "learning_rate": 3.116629559952764e-06,
+      "loss": 0.2551,
+      "step": 1953
+    },
+    {
+      "epoch": 1.3220568335588634,
+      "grad_norm": 0.5214842198056288,
+      "learning_rate": 3.111160571541183e-06,
+      "loss": 0.2541,
+      "step": 1954
+    },
+    {
+      "epoch": 1.3227334235453316,
+      "grad_norm": 0.5110105276797563,
+      "learning_rate": 3.105694217831361e-06,
+      "loss": 0.2478,
+      "step": 1955
+    },
+    {
+      "epoch": 1.3234100135317997,
+      "grad_norm": 0.4887583648340985,
+      "learning_rate": 3.1002305064482006e-06,
+      "loss": 0.261,
+      "step": 1956
+    },
+    {
+      "epoch": 1.324086603518268,
+      "grad_norm": 0.5173150545543822,
+      "learning_rate": 3.094769445012913e-06,
+      "loss": 0.2645,
+      "step": 1957
+    },
+    {
+      "epoch": 1.3247631935047361,
+      "grad_norm": 0.4977474209796807,
+      "learning_rate": 3.089311041143017e-06,
+      "loss": 0.2625,
+      "step": 1958
+    },
+    {
+      "epoch": 1.3254397834912044,
+      "grad_norm": 0.5144197651977577,
+      "learning_rate": 3.083855302452323e-06,
+      "loss": 0.2574,
+      "step": 1959
+    },
+    {
+      "epoch": 1.3261163734776726,
+      "grad_norm": 0.5010978648179522,
+      "learning_rate": 3.078402236550926e-06,
+      "loss": 0.254,
+      "step": 1960
+    },
+    {
+      "epoch": 1.3267929634641407,
+      "grad_norm": 0.4883841474531989,
+      "learning_rate": 3.0729518510451888e-06,
+      "loss": 0.2486,
+      "step": 1961
+    },
+    {
+      "epoch": 1.327469553450609,
+      "grad_norm": 0.5197575130501223,
+      "learning_rate": 3.06750415353774e-06,
+      "loss": 0.2547,
+      "step": 1962
+    },
+    {
+      "epoch": 1.3281461434370772,
+      "grad_norm": 0.5009565275217402,
+      "learning_rate": 3.0620591516274532e-06,
+      "loss": 0.253,
+      "step": 1963
+    },
+    {
+      "epoch": 1.3288227334235454,
+      "grad_norm": 0.5095763090065845,
+      "learning_rate": 3.0566168529094485e-06,
+      "loss": 0.2633,
+      "step": 1964
+    },
+    {
+      "epoch": 1.3294993234100136,
+      "grad_norm": 0.4848432820196966,
+      "learning_rate": 3.0511772649750728e-06,
+      "loss": 0.2446,
+      "step": 1965
+    },
+    {
+      "epoch": 1.3301759133964817,
+      "grad_norm": 0.5042196637080786,
+      "learning_rate": 3.045740395411886e-06,
+      "loss": 0.257,
+      "step": 1966
+    },
+    {
+      "epoch": 1.33085250338295,
+      "grad_norm": 0.49909213390098744,
+      "learning_rate": 3.0403062518036664e-06,
+      "loss": 0.249,
+      "step": 1967
+    },
+    {
+      "epoch": 1.3315290933694182,
+      "grad_norm": 0.5095065838614894,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.25,
+      "step": 1968
+    },
+    {
+      "epoch": 1.3322056833558864,
+      "grad_norm": 0.5070173218543467,
+      "learning_rate": 3.0294461727681936e-06,
+      "loss": 0.2487,
+      "step": 1969
+    },
+    {
+      "epoch": 1.3328822733423547,
+      "grad_norm": 0.48482280778081605,
+      "learning_rate": 3.0240202524894304e-06,
+      "loss": 0.2462,
+      "step": 1970
+    },
+    {
+      "epoch": 1.3335588633288227,
+      "grad_norm": 0.5274238941365568,
+      "learning_rate": 3.018597088462597e-06,
+      "loss": 0.2535,
+      "step": 1971
+    },
+    {
+      "epoch": 1.334235453315291,
+      "grad_norm": 0.5231684922752187,
+      "learning_rate": 3.013176688252349e-06,
+      "loss": 0.2594,
+      "step": 1972
+    },
+    {
+      "epoch": 1.3349120433017592,
+      "grad_norm": 0.5111278536181871,
+      "learning_rate": 3.007759059419485e-06,
+      "loss": 0.2579,
+      "step": 1973
+    },
+    {
+      "epoch": 1.3355886332882274,
+      "grad_norm": 0.5155269197531276,
+      "learning_rate": 3.0023442095209386e-06,
+      "loss": 0.2632,
+      "step": 1974
+    },
+    {
+      "epoch": 1.3362652232746957,
+      "grad_norm": 0.5058223937349804,
+      "learning_rate": 2.9969321461097712e-06,
+      "loss": 0.2593,
+      "step": 1975
+    },
+    {
+      "epoch": 1.3369418132611637,
+      "grad_norm": 0.5086239447233707,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.2556,
+      "step": 1976
+    },
+    {
+      "epoch": 1.337618403247632,
+      "grad_norm": 0.48576309545320173,
+      "learning_rate": 2.9861164089423584e-06,
+      "loss": 0.2502,
+      "step": 1977
+    },
+    {
+      "epoch": 1.3382949932341002,
+      "grad_norm": 0.47001502979043447,
+      "learning_rate": 2.980712750272754e-06,
+      "loss": 0.2523,
+      "step": 1978
+    },
+    {
+      "epoch": 1.3389715832205684,
+      "grad_norm": 0.48305758991018916,
+      "learning_rate": 2.975311908263788e-06,
+      "loss": 0.2581,
+      "step": 1979
+    },
+    {
+      "epoch": 1.3396481732070367,
+      "grad_norm": 0.4875766921794759,
+      "learning_rate": 2.96991389044898e-06,
+      "loss": 0.2515,
+      "step": 1980
+    },
+    {
+      "epoch": 1.3403247631935047,
+      "grad_norm": 0.5157520208091969,
+      "learning_rate": 2.964518704357906e-06,
+      "loss": 0.26,
+      "step": 1981
+    },
+    {
+      "epoch": 1.341001353179973,
+      "grad_norm": 0.5013795790090423,
+      "learning_rate": 2.9591263575162e-06,
+      "loss": 0.2568,
+      "step": 1982
+    },
+    {
+      "epoch": 1.3416779431664412,
+      "grad_norm": 0.4983491290270124,
+      "learning_rate": 2.9537368574455304e-06,
+      "loss": 0.2486,
+      "step": 1983
+    },
+    {
+      "epoch": 1.3423545331529092,
+      "grad_norm": 0.500266895102474,
+      "learning_rate": 2.9483502116635943e-06,
+      "loss": 0.2521,
+      "step": 1984
+    },
+    {
+      "epoch": 1.3430311231393777,
+      "grad_norm": 0.48898696895201516,
+      "learning_rate": 2.9429664276841107e-06,
+      "loss": 0.2521,
+      "step": 1985
+    },
+    {
+      "epoch": 1.3437077131258457,
+      "grad_norm": 0.5180995641795345,
+      "learning_rate": 2.9375855130168046e-06,
+      "loss": 0.2637,
+      "step": 1986
+    },
+    {
+      "epoch": 1.344384303112314,
+      "grad_norm": 0.5185064874997003,
+      "learning_rate": 2.932207475167398e-06,
+      "loss": 0.2639,
+      "step": 1987
+    },
+    {
+      "epoch": 1.3450608930987822,
+      "grad_norm": 0.5119322519623422,
+      "learning_rate": 2.9268323216375997e-06,
+      "loss": 0.2424,
+      "step": 1988
+    },
+    {
+      "epoch": 1.3457374830852502,
+      "grad_norm": 0.5186008511583682,
+      "learning_rate": 2.9214600599250993e-06,
+      "loss": 0.2643,
+      "step": 1989
+    },
+    {
+      "epoch": 1.3464140730717187,
+      "grad_norm": 0.5354174490942603,
+      "learning_rate": 2.9160906975235493e-06,
+      "loss": 0.2614,
+      "step": 1990
+    },
+    {
+      "epoch": 1.3470906630581867,
+      "grad_norm": 0.5215512741730531,
+      "learning_rate": 2.910724241922558e-06,
+      "loss": 0.2508,
+      "step": 1991
+    },
+    {
+      "epoch": 1.347767253044655,
+      "grad_norm": 0.501630462934942,
+      "learning_rate": 2.9053607006076766e-06,
+      "loss": 0.2593,
+      "step": 1992
+    },
+    {
+      "epoch": 1.3484438430311232,
+      "grad_norm": 0.4948497644936073,
+      "learning_rate": 2.9000000810603985e-06,
+      "loss": 0.2458,
+      "step": 1993
+    },
+    {
+      "epoch": 1.3491204330175912,
+      "grad_norm": 0.5000568412085239,
+      "learning_rate": 2.8946423907581377e-06,
+      "loss": 0.2428,
+      "step": 1994
+    },
+    {
+      "epoch": 1.3497970230040595,
+      "grad_norm": 0.5193154063975651,
+      "learning_rate": 2.8892876371742185e-06,
+      "loss": 0.2485,
+      "step": 1995
+    },
+    {
+      "epoch": 1.3504736129905277,
+      "grad_norm": 0.5176501681006508,
+      "learning_rate": 2.8839358277778758e-06,
+      "loss": 0.2536,
+      "step": 1996
+    },
+    {
+      "epoch": 1.351150202976996,
+      "grad_norm": 0.5436825232569681,
+      "learning_rate": 2.8785869700342322e-06,
+      "loss": 0.2596,
+      "step": 1997
+    },
+    {
+      "epoch": 1.3518267929634642,
+      "grad_norm": 0.5013337624470315,
+      "learning_rate": 2.873241071404296e-06,
+      "loss": 0.2577,
+      "step": 1998
+    },
+    {
+      "epoch": 1.3525033829499322,
+      "grad_norm": 0.5140326302126593,
+      "learning_rate": 2.8678981393449434e-06,
+      "loss": 0.258,
+      "step": 1999
+    },
+    {
+      "epoch": 1.3531799729364005,
+      "grad_norm": 0.5349817517822526,
+      "learning_rate": 2.86255818130892e-06,
+      "loss": 0.2695,
+      "step": 2000
+    },
+    {
+      "epoch": 1.3538565629228687,
+      "grad_norm": 0.49603638073949063,
+      "learning_rate": 2.8572212047448196e-06,
+      "loss": 0.2555,
+      "step": 2001
+    },
+    {
+      "epoch": 1.354533152909337,
+      "grad_norm": 0.49607972580200277,
+      "learning_rate": 2.8518872170970758e-06,
+      "loss": 0.2546,
+      "step": 2002
+    },
+    {
+      "epoch": 1.3552097428958052,
+      "grad_norm": 0.49407173375072727,
+      "learning_rate": 2.846556225805951e-06,
+      "loss": 0.2432,
+      "step": 2003
+    },
+    {
+      "epoch": 1.3558863328822732,
+      "grad_norm": 0.514175600808731,
+      "learning_rate": 2.8412282383075362e-06,
+      "loss": 0.2549,
+      "step": 2004
+    },
+    {
+      "epoch": 1.3565629228687415,
+      "grad_norm": 0.5166580678089843,
+      "learning_rate": 2.8359032620337246e-06,
+      "loss": 0.2675,
+      "step": 2005
+    },
+    {
+      "epoch": 1.3572395128552097,
+      "grad_norm": 0.5160549628904729,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.2545,
+      "step": 2006
+    },
+    {
+      "epoch": 1.357916102841678,
+      "grad_norm": 0.5044475092605399,
+      "learning_rate": 2.8252623728664797e-06,
+      "loss": 0.2576,
+      "step": 2007
+    },
+    {
+      "epoch": 1.3585926928281462,
+      "grad_norm": 0.525076595251361,
+      "learning_rate": 2.8199464748157983e-06,
+      "loss": 0.2579,
+      "step": 2008
+    },
+    {
+      "epoch": 1.3592692828146142,
+      "grad_norm": 0.49130479244810854,
+      "learning_rate": 2.8146336176751988e-06,
+      "loss": 0.2367,
+      "step": 2009
+    },
+    {
+      "epoch": 1.3599458728010825,
+      "grad_norm": 0.5060705012775366,
+      "learning_rate": 2.8093238088554676e-06,
+      "loss": 0.2536,
+      "step": 2010
+    },
+    {
+      "epoch": 1.3606224627875507,
+      "grad_norm": 0.4905287489619616,
+      "learning_rate": 2.804017055763149e-06,
+      "loss": 0.2496,
+      "step": 2011
+    },
+    {
+      "epoch": 1.361299052774019,
+      "grad_norm": 0.4926140919094704,
+      "learning_rate": 2.7987133658005174e-06,
+      "loss": 0.2567,
+      "step": 2012
+    },
+    {
+      "epoch": 1.3619756427604872,
+      "grad_norm": 0.5132651481960547,
+      "learning_rate": 2.793412746365574e-06,
+      "loss": 0.2582,
+      "step": 2013
+    },
+    {
+      "epoch": 1.3626522327469552,
+      "grad_norm": 0.5091355954278813,
+      "learning_rate": 2.788115204852042e-06,
+      "loss": 0.2508,
+      "step": 2014
+    },
+    {
+      "epoch": 1.3633288227334235,
+      "grad_norm": 0.49522033333591525,
+      "learning_rate": 2.7828207486493513e-06,
+      "loss": 0.248,
+      "step": 2015
+    },
+    {
+      "epoch": 1.3640054127198917,
+      "grad_norm": 0.5073054703597139,
+      "learning_rate": 2.7775293851426233e-06,
+      "loss": 0.2518,
+      "step": 2016
+    },
+    {
+      "epoch": 1.36468200270636,
+      "grad_norm": 0.509091500934284,
+      "learning_rate": 2.7722411217126655e-06,
+      "loss": 0.265,
+      "step": 2017
+    },
+    {
+      "epoch": 1.3653585926928282,
+      "grad_norm": 0.5212604737000412,
+      "learning_rate": 2.766955965735968e-06,
+      "loss": 0.2648,
+      "step": 2018
+    },
+    {
+      "epoch": 1.3660351826792962,
+      "grad_norm": 0.502252312969035,
+      "learning_rate": 2.76167392458468e-06,
+      "loss": 0.2591,
+      "step": 2019
+    },
+    {
+      "epoch": 1.3667117726657645,
+      "grad_norm": 0.4950579699867225,
+      "learning_rate": 2.7563950056266053e-06,
+      "loss": 0.2588,
+      "step": 2020
+    },
+    {
+      "epoch": 1.3673883626522327,
+      "grad_norm": 0.49687741183609724,
+      "learning_rate": 2.751119216225198e-06,
+      "loss": 0.2465,
+      "step": 2021
+    },
+    {
+      "epoch": 1.368064952638701,
+      "grad_norm": 0.4932368847918004,
+      "learning_rate": 2.745846563739546e-06,
+      "loss": 0.2488,
+      "step": 2022
+    },
+    {
+      "epoch": 1.3687415426251692,
+      "grad_norm": 0.4847708695047045,
+      "learning_rate": 2.740577055524356e-06,
+      "loss": 0.2525,
+      "step": 2023
+    },
+    {
+      "epoch": 1.3694181326116373,
+      "grad_norm": 0.49274202332041905,
+      "learning_rate": 2.7353106989299528e-06,
+      "loss": 0.2532,
+      "step": 2024
+    },
+    {
+      "epoch": 1.3700947225981055,
+      "grad_norm": 0.4943690606666166,
+      "learning_rate": 2.7300475013022666e-06,
+      "loss": 0.2586,
+      "step": 2025
+    },
+    {
+      "epoch": 1.3707713125845737,
+      "grad_norm": 0.48729291185206897,
+      "learning_rate": 2.7247874699828186e-06,
+      "loss": 0.2562,
+      "step": 2026
+    },
+    {
+      "epoch": 1.371447902571042,
+      "grad_norm": 0.5048120720987606,
+      "learning_rate": 2.719530612308712e-06,
+      "loss": 0.2535,
+      "step": 2027
+    },
+    {
+      "epoch": 1.3721244925575102,
+      "grad_norm": 0.5044584942573134,
+      "learning_rate": 2.7142769356126258e-06,
+      "loss": 0.2551,
+      "step": 2028
+    },
+    {
+      "epoch": 1.3728010825439783,
+      "grad_norm": 0.48838185060292105,
+      "learning_rate": 2.709026447222806e-06,
+      "loss": 0.2554,
+      "step": 2029
+    },
+    {
+      "epoch": 1.3734776725304465,
+      "grad_norm": 0.5075848872412796,
+      "learning_rate": 2.7037791544630414e-06,
+      "loss": 0.2587,
+      "step": 2030
+    },
+    {
+      "epoch": 1.3741542625169147,
+      "grad_norm": 0.5234498369903007,
+      "learning_rate": 2.6985350646526713e-06,
+      "loss": 0.2648,
+      "step": 2031
+    },
+    {
+      "epoch": 1.374830852503383,
+      "grad_norm": 0.4907858487362599,
+      "learning_rate": 2.693294185106562e-06,
+      "loss": 0.2562,
+      "step": 2032
+    },
+    {
+      "epoch": 1.3755074424898512,
+      "grad_norm": 0.5193092310404074,
+      "learning_rate": 2.688056523135107e-06,
+      "loss": 0.2599,
+      "step": 2033
+    },
+    {
+      "epoch": 1.3761840324763193,
+      "grad_norm": 0.5099334366934776,
+      "learning_rate": 2.682822086044206e-06,
+      "loss": 0.2565,
+      "step": 2034
+    },
+    {
+      "epoch": 1.3768606224627875,
+      "grad_norm": 0.5069020164970048,
+      "learning_rate": 2.677590881135266e-06,
+      "loss": 0.2645,
+      "step": 2035
+    },
+    {
+      "epoch": 1.3775372124492558,
+      "grad_norm": 0.48591778194320484,
+      "learning_rate": 2.6723629157051844e-06,
+      "loss": 0.2425,
+      "step": 2036
+    },
+    {
+      "epoch": 1.378213802435724,
+      "grad_norm": 0.5069404719526072,
+      "learning_rate": 2.6671381970463363e-06,
+      "loss": 0.2499,
+      "step": 2037
+    },
+    {
+      "epoch": 1.3788903924221922,
+      "grad_norm": 0.5013635612916136,
+      "learning_rate": 2.66191673244657e-06,
+      "loss": 0.2495,
+      "step": 2038
+    },
+    {
+      "epoch": 1.3795669824086603,
+      "grad_norm": 0.5011368647321768,
+      "learning_rate": 2.6566985291891932e-06,
+      "loss": 0.2526,
+      "step": 2039
+    },
+    {
+      "epoch": 1.3802435723951285,
+      "grad_norm": 0.514250896899684,
+      "learning_rate": 2.6514835945529706e-06,
+      "loss": 0.2561,
+      "step": 2040
+    },
+    {
+      "epoch": 1.3809201623815968,
+      "grad_norm": 0.5137284012093906,
+      "learning_rate": 2.6462719358120983e-06,
+      "loss": 0.2652,
+      "step": 2041
+    },
+    {
+      "epoch": 1.381596752368065,
+      "grad_norm": 0.4865256441410564,
+      "learning_rate": 2.641063560236212e-06,
+      "loss": 0.2369,
+      "step": 2042
+    },
+    {
+      "epoch": 1.3822733423545333,
+      "grad_norm": 0.5039863147550203,
+      "learning_rate": 2.63585847509036e-06,
+      "loss": 0.2553,
+      "step": 2043
+    },
+    {
+      "epoch": 1.3829499323410013,
+      "grad_norm": 0.512331013630883,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.2573,
+      "step": 2044
+    },
+    {
+      "epoch": 1.3836265223274695,
+      "grad_norm": 0.519018512138971,
+      "learning_rate": 2.625458205126015e-06,
+      "loss": 0.2511,
+      "step": 2045
+    },
+    {
+      "epoch": 1.3843031123139378,
+      "grad_norm": 0.5216974056049689,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.2561,
+      "step": 2046
+    },
+    {
+      "epoch": 1.384979702300406,
+      "grad_norm": 0.4998362820104067,
+      "learning_rate": 2.615071183947496e-06,
+      "loss": 0.2596,
+      "step": 2047
+    },
+    {
+      "epoch": 1.3856562922868743,
+      "grad_norm": 0.5004370634069251,
+      "learning_rate": 2.609882659766605e-06,
+      "loss": 0.2522,
+      "step": 2048
+    },
+    {
+      "epoch": 1.3863328822733423,
+      "grad_norm": 0.515680711308714,
+      "learning_rate": 2.604697469509323e-06,
+      "loss": 0.2505,
+      "step": 2049
+    },
+    {
+      "epoch": 1.3870094722598105,
+      "grad_norm": 0.516997741841569,
+      "learning_rate": 2.5995156204083573e-06,
+      "loss": 0.2538,
+      "step": 2050
+    },
+    {
+      "epoch": 1.3876860622462788,
+      "grad_norm": 0.48500741921186413,
+      "learning_rate": 2.5943371196917633e-06,
+      "loss": 0.2454,
+      "step": 2051
+    },
+    {
+      "epoch": 1.388362652232747,
+      "grad_norm": 0.49920482006330713,
+      "learning_rate": 2.5891619745829184e-06,
+      "loss": 0.2588,
+      "step": 2052
+    },
+    {
+      "epoch": 1.3890392422192153,
+      "grad_norm": 0.5157840309727497,
+      "learning_rate": 2.5839901923005207e-06,
+      "loss": 0.264,
+      "step": 2053
+    },
+    {
+      "epoch": 1.3897158322056833,
+      "grad_norm": 0.5165966742662046,
+      "learning_rate": 2.5788217800585812e-06,
+      "loss": 0.2557,
+      "step": 2054
+    },
+    {
+      "epoch": 1.3903924221921515,
+      "grad_norm": 0.4994417593810354,
+      "learning_rate": 2.573656745066406e-06,
+      "loss": 0.2395,
+      "step": 2055
+    },
+    {
+      "epoch": 1.3910690121786198,
+      "grad_norm": 0.4792485992476495,
+      "learning_rate": 2.5684950945285937e-06,
+      "loss": 0.2481,
+      "step": 2056
+    },
+    {
+      "epoch": 1.391745602165088,
+      "grad_norm": 0.4967825895383718,
+      "learning_rate": 2.5633368356450185e-06,
+      "loss": 0.252,
+      "step": 2057
+    },
+    {
+      "epoch": 1.3924221921515563,
+      "grad_norm": 0.5096621332107023,
+      "learning_rate": 2.558181975610827e-06,
+      "loss": 0.2603,
+      "step": 2058
+    },
+    {
+      "epoch": 1.3930987821380243,
+      "grad_norm": 0.5132641302041446,
+      "learning_rate": 2.553030521616424e-06,
+      "loss": 0.2578,
+      "step": 2059
+    },
+    {
+      "epoch": 1.3937753721244925,
+      "grad_norm": 0.4979381100910691,
+      "learning_rate": 2.5478824808474613e-06,
+      "loss": 0.2509,
+      "step": 2060
+    },
+    {
+      "epoch": 1.3944519621109608,
+      "grad_norm": 0.5101070245358168,
+      "learning_rate": 2.5427378604848285e-06,
+      "loss": 0.2622,
+      "step": 2061
+    },
+    {
+      "epoch": 1.395128552097429,
+      "grad_norm": 0.5131540379928476,
+      "learning_rate": 2.53759666770465e-06,
+      "loss": 0.272,
+      "step": 2062
+    },
+    {
+      "epoch": 1.3958051420838973,
+      "grad_norm": 0.4807353924466645,
+      "learning_rate": 2.532458909678266e-06,
+      "loss": 0.2444,
+      "step": 2063
+    },
+    {
+      "epoch": 1.3964817320703653,
+      "grad_norm": 0.5199212555163054,
+      "learning_rate": 2.527324593572223e-06,
+      "loss": 0.2678,
+      "step": 2064
+    },
+    {
+      "epoch": 1.3971583220568335,
+      "grad_norm": 0.5123408504176789,
+      "learning_rate": 2.5221937265482722e-06,
+      "loss": 0.2482,
+      "step": 2065
+    },
+    {
+      "epoch": 1.3978349120433018,
+      "grad_norm": 0.4913761581166269,
+      "learning_rate": 2.517066315763348e-06,
+      "loss": 0.2542,
+      "step": 2066
+    },
+    {
+      "epoch": 1.39851150202977,
+      "grad_norm": 0.508752934746744,
+      "learning_rate": 2.511942368369566e-06,
+      "loss": 0.2547,
+      "step": 2067
+    },
+    {
+      "epoch": 1.3991880920162383,
+      "grad_norm": 0.4977032787499384,
+      "learning_rate": 2.5068218915142093e-06,
+      "loss": 0.2525,
+      "step": 2068
+    },
+    {
+      "epoch": 1.3998646820027063,
+      "grad_norm": 0.49725024967814885,
+      "learning_rate": 2.5017048923397247e-06,
+      "loss": 0.2538,
+      "step": 2069
+    },
+    {
+      "epoch": 1.4005412719891746,
+      "grad_norm": 0.4881807628340449,
+      "learning_rate": 2.496591377983706e-06,
+      "loss": 0.2464,
+      "step": 2070
+    },
+    {
+      "epoch": 1.4012178619756428,
+      "grad_norm": 0.4807678148660376,
+      "learning_rate": 2.4914813555788827e-06,
+      "loss": 0.2419,
+      "step": 2071
+    },
+    {
+      "epoch": 1.401894451962111,
+      "grad_norm": 0.4902462956159903,
+      "learning_rate": 2.4863748322531144e-06,
+      "loss": 0.2586,
+      "step": 2072
+    },
+    {
+      "epoch": 1.4025710419485793,
+      "grad_norm": 0.5070043137710929,
+      "learning_rate": 2.4812718151293858e-06,
+      "loss": 0.2575,
+      "step": 2073
+    },
+    {
+      "epoch": 1.4032476319350473,
+      "grad_norm": 0.4938516826245986,
+      "learning_rate": 2.476172311325783e-06,
+      "loss": 0.2481,
+      "step": 2074
+    },
+    {
+      "epoch": 1.4039242219215156,
+      "grad_norm": 0.5029404898197912,
+      "learning_rate": 2.4710763279554944e-06,
+      "loss": 0.2535,
+      "step": 2075
+    },
+    {
+      "epoch": 1.4046008119079838,
+      "grad_norm": 0.495767428855396,
+      "learning_rate": 2.4659838721268005e-06,
+      "loss": 0.2578,
+      "step": 2076
+    },
+    {
+      "epoch": 1.4052774018944518,
+      "grad_norm": 0.4732310288313788,
+      "learning_rate": 2.460894950943059e-06,
+      "loss": 0.2338,
+      "step": 2077
+    },
+    {
+      "epoch": 1.4059539918809203,
+      "grad_norm": 0.4919509446770192,
+      "learning_rate": 2.4558095715026975e-06,
+      "loss": 0.2477,
+      "step": 2078
+    },
+    {
+      "epoch": 1.4066305818673883,
+      "grad_norm": 0.48389505482764844,
+      "learning_rate": 2.4507277408992004e-06,
+      "loss": 0.2384,
+      "step": 2079
+    },
+    {
+      "epoch": 1.4073071718538566,
+      "grad_norm": 0.5138272961113601,
+      "learning_rate": 2.4456494662211082e-06,
+      "loss": 0.2536,
+      "step": 2080
+    },
+    {
+      "epoch": 1.4079837618403248,
+      "grad_norm": 0.5043124455235967,
+      "learning_rate": 2.4405747545519966e-06,
+      "loss": 0.2635,
+      "step": 2081
+    },
+    {
+      "epoch": 1.4086603518267928,
+      "grad_norm": 0.4980430979141146,
+      "learning_rate": 2.43550361297047e-06,
+      "loss": 0.2479,
+      "step": 2082
+    },
+    {
+      "epoch": 1.4093369418132613,
+      "grad_norm": 0.5280517849106742,
+      "learning_rate": 2.430436048550157e-06,
+      "loss": 0.2589,
+      "step": 2083
+    },
+    {
+      "epoch": 1.4100135317997293,
+      "grad_norm": 0.5013862912140888,
+      "learning_rate": 2.4253720683596976e-06,
+      "loss": 0.2518,
+      "step": 2084
+    },
+    {
+      "epoch": 1.4106901217861976,
+      "grad_norm": 0.48551465689576634,
+      "learning_rate": 2.420311679462727e-06,
+      "loss": 0.2524,
+      "step": 2085
+    },
+    {
+      "epoch": 1.4113667117726658,
+      "grad_norm": 0.5008834668788076,
+      "learning_rate": 2.4152548889178722e-06,
+      "loss": 0.2562,
+      "step": 2086
+    },
+    {
+      "epoch": 1.4120433017591338,
+      "grad_norm": 0.49943442341617217,
+      "learning_rate": 2.4102017037787457e-06,
+      "loss": 0.2557,
+      "step": 2087
+    },
+    {
+      "epoch": 1.412719891745602,
+      "grad_norm": 0.4932617377757013,
+      "learning_rate": 2.4051521310939258e-06,
+      "loss": 0.2465,
+      "step": 2088
+    },
+    {
+      "epoch": 1.4133964817320703,
+      "grad_norm": 0.49536993936383444,
+      "learning_rate": 2.4001061779069516e-06,
+      "loss": 0.2445,
+      "step": 2089
+    },
+    {
+      "epoch": 1.4140730717185386,
+      "grad_norm": 0.48730133573483647,
+      "learning_rate": 2.3950638512563173e-06,
+      "loss": 0.2512,
+      "step": 2090
+    },
+    {
+      "epoch": 1.4147496617050068,
+      "grad_norm": 0.5097280555007337,
+      "learning_rate": 2.390025158175458e-06,
+      "loss": 0.2595,
+      "step": 2091
+    },
+    {
+      "epoch": 1.4154262516914748,
+      "grad_norm": 0.4819118500024701,
+      "learning_rate": 2.3849901056927383e-06,
+      "loss": 0.2525,
+      "step": 2092
+    },
+    {
+      "epoch": 1.416102841677943,
+      "grad_norm": 0.5237148701744913,
+      "learning_rate": 2.3799587008314422e-06,
+      "loss": 0.2452,
+      "step": 2093
+    },
+    {
+      "epoch": 1.4167794316644113,
+      "grad_norm": 0.4785528188252626,
+      "learning_rate": 2.374930950609773e-06,
+      "loss": 0.2436,
+      "step": 2094
+    },
+    {
+      "epoch": 1.4174560216508796,
+      "grad_norm": 0.4733064843989778,
+      "learning_rate": 2.3699068620408305e-06,
+      "loss": 0.2462,
+      "step": 2095
+    },
+    {
+      "epoch": 1.4181326116373478,
+      "grad_norm": 0.4922571799516969,
+      "learning_rate": 2.364886442132606e-06,
+      "loss": 0.246,
+      "step": 2096
+    },
+    {
+      "epoch": 1.4188092016238159,
+      "grad_norm": 0.528825978048522,
+      "learning_rate": 2.3598696978879787e-06,
+      "loss": 0.26,
+      "step": 2097
+    },
+    {
+      "epoch": 1.419485791610284,
+      "grad_norm": 0.4979481004692495,
+      "learning_rate": 2.3548566363046993e-06,
+      "loss": 0.2476,
+      "step": 2098
+    },
+    {
+      "epoch": 1.4201623815967523,
+      "grad_norm": 0.505264264423603,
+      "learning_rate": 2.349847264375378e-06,
+      "loss": 0.2543,
+      "step": 2099
+    },
+    {
+      "epoch": 1.4208389715832206,
+      "grad_norm": 0.4772269358178425,
+      "learning_rate": 2.344841589087482e-06,
+      "loss": 0.244,
+      "step": 2100
+    },
+    {
+      "epoch": 1.4215155615696888,
+      "grad_norm": 0.5054856224255921,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.265,
+      "step": 2101
+    },
+    {
+      "epoch": 1.4221921515561569,
+      "grad_norm": 0.5279798968804738,
+      "learning_rate": 2.3348413563600324e-06,
+      "loss": 0.2595,
+      "step": 2102
+    },
+    {
+      "epoch": 1.422868741542625,
+      "grad_norm": 0.5184521712015991,
+      "learning_rate": 2.3298468128695904e-06,
+      "loss": 0.2609,
+      "step": 2103
+    },
+    {
+      "epoch": 1.4235453315290933,
+      "grad_norm": 0.5068710880928842,
+      "learning_rate": 2.324855993918775e-06,
+      "loss": 0.2558,
+      "step": 2104
+    },
+    {
+      "epoch": 1.4242219215155616,
+      "grad_norm": 0.503323495137593,
+      "learning_rate": 2.3198689064691754e-06,
+      "loss": 0.2515,
+      "step": 2105
+    },
+    {
+      "epoch": 1.4248985115020298,
+      "grad_norm": 0.49873043194306393,
+      "learning_rate": 2.3148855574771706e-06,
+      "loss": 0.2443,
+      "step": 2106
+    },
+    {
+      "epoch": 1.4255751014884979,
+      "grad_norm": 0.4977659865127549,
+      "learning_rate": 2.3099059538939285e-06,
+      "loss": 0.2406,
+      "step": 2107
+    },
+    {
+      "epoch": 1.426251691474966,
+      "grad_norm": 0.507585656247614,
+      "learning_rate": 2.304930102665389e-06,
+      "loss": 0.2418,
+      "step": 2108
+    },
+    {
+      "epoch": 1.4269282814614344,
+      "grad_norm": 0.5187093492230298,
+      "learning_rate": 2.2999580107322654e-06,
+      "loss": 0.2546,
+      "step": 2109
+    },
+    {
+      "epoch": 1.4276048714479026,
+      "grad_norm": 0.509492325678473,
+      "learning_rate": 2.2949896850300186e-06,
+      "loss": 0.2529,
+      "step": 2110
+    },
+    {
+      "epoch": 1.4282814614343708,
+      "grad_norm": 0.4873159680522553,
+      "learning_rate": 2.2900251324888627e-06,
+      "loss": 0.2467,
+      "step": 2111
+    },
+    {
+      "epoch": 1.4289580514208389,
+      "grad_norm": 0.5122553581485823,
+      "learning_rate": 2.2850643600337435e-06,
+      "loss": 0.2633,
+      "step": 2112
+    },
+    {
+      "epoch": 1.4296346414073071,
+      "grad_norm": 0.48414967289101296,
+      "learning_rate": 2.2801073745843395e-06,
+      "loss": 0.2443,
+      "step": 2113
+    },
+    {
+      "epoch": 1.4303112313937754,
+      "grad_norm": 0.4730668372549669,
+      "learning_rate": 2.2751541830550417e-06,
+      "loss": 0.2422,
+      "step": 2114
+    },
+    {
+      "epoch": 1.4309878213802436,
+      "grad_norm": 0.48610088138201274,
+      "learning_rate": 2.27020479235495e-06,
+      "loss": 0.2493,
+      "step": 2115
+    },
+    {
+      "epoch": 1.4316644113667119,
+      "grad_norm": 0.48888135821677,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.2406,
+      "step": 2116
+    },
+    {
+      "epoch": 1.4323410013531799,
+      "grad_norm": 0.5085609168436923,
+      "learning_rate": 2.2603174410522766e-06,
+      "loss": 0.2524,
+      "step": 2117
+    },
+    {
+      "epoch": 1.4330175913396481,
+      "grad_norm": 0.5174089391420021,
+      "learning_rate": 2.2553794942413506e-06,
+      "loss": 0.247,
+      "step": 2118
+    },
+    {
+      "epoch": 1.4336941813261164,
+      "grad_norm": 0.5065506379761218,
+      "learning_rate": 2.2504453758429206e-06,
+      "loss": 0.2527,
+      "step": 2119
+    },
+    {
+      "epoch": 1.4343707713125846,
+      "grad_norm": 0.504192252428722,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.2509,
+      "step": 2120
+    },
+    {
+      "epoch": 1.4350473612990529,
+      "grad_norm": 0.5115858143560573,
+      "learning_rate": 2.2405886518081967e-06,
+      "loss": 0.2524,
+      "step": 2121
+    },
+    {
+      "epoch": 1.4357239512855209,
+      "grad_norm": 0.48254121946613265,
+      "learning_rate": 2.2356660599208335e-06,
+      "loss": 0.2467,
+      "step": 2122
+    },
+    {
+      "epoch": 1.4364005412719891,
+      "grad_norm": 0.5709723463116735,
+      "learning_rate": 2.2307473239438153e-06,
+      "loss": 0.248,
+      "step": 2123
+    },
+    {
+      "epoch": 1.4370771312584574,
+      "grad_norm": 0.5210333690657513,
+      "learning_rate": 2.2258324507381834e-06,
+      "loss": 0.2565,
+      "step": 2124
+    },
+    {
+      "epoch": 1.4377537212449256,
+      "grad_norm": 0.5045068086521571,
+      "learning_rate": 2.2209214471595918e-06,
+      "loss": 0.251,
+      "step": 2125
+    },
+    {
+      "epoch": 1.4384303112313939,
+      "grad_norm": 0.5004464618883946,
+      "learning_rate": 2.2160143200582906e-06,
+      "loss": 0.2429,
+      "step": 2126
+    },
+    {
+      "epoch": 1.439106901217862,
+      "grad_norm": 0.5121667863976368,
+      "learning_rate": 2.21111107627913e-06,
+      "loss": 0.2563,
+      "step": 2127
+    },
+    {
+      "epoch": 1.4397834912043301,
+      "grad_norm": 0.5051440167222413,
+      "learning_rate": 2.2062117226615375e-06,
+      "loss": 0.2524,
+      "step": 2128
+    },
+    {
+      "epoch": 1.4404600811907984,
+      "grad_norm": 0.5169400158362245,
+      "learning_rate": 2.2013162660395182e-06,
+      "loss": 0.2476,
+      "step": 2129
+    },
+    {
+      "epoch": 1.4411366711772666,
+      "grad_norm": 0.4955721208185394,
+      "learning_rate": 2.1964247132416373e-06,
+      "loss": 0.2541,
+      "step": 2130
+    },
+    {
+      "epoch": 1.4418132611637349,
+      "grad_norm": 0.5129792819137408,
+      "learning_rate": 2.1915370710910188e-06,
+      "loss": 0.2504,
+      "step": 2131
+    },
+    {
+      "epoch": 1.442489851150203,
+      "grad_norm": 0.4877467820665676,
+      "learning_rate": 2.186653346405333e-06,
+      "loss": 0.251,
+      "step": 2132
+    },
+    {
+      "epoch": 1.4431664411366711,
+      "grad_norm": 0.49965702043583643,
+      "learning_rate": 2.1817735459967787e-06,
+      "loss": 0.2535,
+      "step": 2133
+    },
+    {
+      "epoch": 1.4438430311231394,
+      "grad_norm": 0.49452129236533326,
+      "learning_rate": 2.1768976766720896e-06,
+      "loss": 0.2528,
+      "step": 2134
+    },
+    {
+      "epoch": 1.4445196211096076,
+      "grad_norm": 0.4987343190753758,
+      "learning_rate": 2.1720257452325087e-06,
+      "loss": 0.2492,
+      "step": 2135
+    },
+    {
+      "epoch": 1.4451962110960759,
+      "grad_norm": 0.4900407892537662,
+      "learning_rate": 2.16715775847379e-06,
+      "loss": 0.2488,
+      "step": 2136
+    },
+    {
+      "epoch": 1.445872801082544,
+      "grad_norm": 0.4907109717293334,
+      "learning_rate": 2.1622937231861823e-06,
+      "loss": 0.2624,
+      "step": 2137
+    },
+    {
+      "epoch": 1.4465493910690121,
+      "grad_norm": 0.49541248481914985,
+      "learning_rate": 2.157433646154426e-06,
+      "loss": 0.2527,
+      "step": 2138
+    },
+    {
+      "epoch": 1.4472259810554804,
+      "grad_norm": 0.5098391069446023,
+      "learning_rate": 2.1525775341577404e-06,
+      "loss": 0.2558,
+      "step": 2139
+    },
+    {
+      "epoch": 1.4479025710419486,
+      "grad_norm": 0.4992562465677881,
+      "learning_rate": 2.147725393969811e-06,
+      "loss": 0.2492,
+      "step": 2140
+    },
+    {
+      "epoch": 1.4485791610284169,
+      "grad_norm": 0.49524082729748903,
+      "learning_rate": 2.1428772323587827e-06,
+      "loss": 0.2487,
+      "step": 2141
+    },
+    {
+      "epoch": 1.449255751014885,
+      "grad_norm": 0.5099189461102216,
+      "learning_rate": 2.138033056087256e-06,
+      "loss": 0.2616,
+      "step": 2142
+    },
+    {
+      "epoch": 1.4499323410013532,
+      "grad_norm": 0.5006816223711742,
+      "learning_rate": 2.133192871912266e-06,
+      "loss": 0.2561,
+      "step": 2143
+    },
+    {
+      "epoch": 1.4506089309878214,
+      "grad_norm": 0.5266994551366269,
+      "learning_rate": 2.1283566865852824e-06,
+      "loss": 0.2713,
+      "step": 2144
+    },
+    {
+      "epoch": 1.4512855209742896,
+      "grad_norm": 0.5845905374514943,
+      "learning_rate": 2.123524506852197e-06,
+      "loss": 0.2585,
+      "step": 2145
+    },
+    {
+      "epoch": 1.451962110960758,
+      "grad_norm": 0.5071587692765479,
+      "learning_rate": 2.1186963394533165e-06,
+      "loss": 0.2539,
+      "step": 2146
+    },
+    {
+      "epoch": 1.452638700947226,
+      "grad_norm": 0.49553987399424726,
+      "learning_rate": 2.113872191123347e-06,
+      "loss": 0.2597,
+      "step": 2147
+    },
+    {
+      "epoch": 1.4533152909336942,
+      "grad_norm": 0.5127330537694187,
+      "learning_rate": 2.1090520685913874e-06,
+      "loss": 0.2659,
+      "step": 2148
+    },
+    {
+      "epoch": 1.4539918809201624,
+      "grad_norm": 0.5069581082949981,
+      "learning_rate": 2.1042359785809274e-06,
+      "loss": 0.26,
+      "step": 2149
+    },
+    {
+      "epoch": 1.4546684709066307,
+      "grad_norm": 0.4982795573504502,
+      "learning_rate": 2.099423927809826e-06,
+      "loss": 0.2553,
+      "step": 2150
+    },
+    {
+      "epoch": 1.455345060893099,
+      "grad_norm": 0.5204800399598737,
+      "learning_rate": 2.094615922990309e-06,
+      "loss": 0.2614,
+      "step": 2151
+    },
+    {
+      "epoch": 1.456021650879567,
+      "grad_norm": 0.48594355769122666,
+      "learning_rate": 2.089811970828961e-06,
+      "loss": 0.249,
+      "step": 2152
+    },
+    {
+      "epoch": 1.4566982408660352,
+      "grad_norm": 0.4935631135699461,
+      "learning_rate": 2.085012078026714e-06,
+      "loss": 0.2583,
+      "step": 2153
+    },
+    {
+      "epoch": 1.4573748308525034,
+      "grad_norm": 0.4793324116777188,
+      "learning_rate": 2.0802162512788337e-06,
+      "loss": 0.2421,
+      "step": 2154
+    },
+    {
+      "epoch": 1.4580514208389717,
+      "grad_norm": 0.49699732419713716,
+      "learning_rate": 2.0754244972749154e-06,
+      "loss": 0.2466,
+      "step": 2155
+    },
+    {
+      "epoch": 1.45872801082544,
+      "grad_norm": 0.4985143959781602,
+      "learning_rate": 2.0706368226988772e-06,
+      "loss": 0.2418,
+      "step": 2156
+    },
+    {
+      "epoch": 1.459404600811908,
+      "grad_norm": 0.5481360128916896,
+      "learning_rate": 2.0658532342289424e-06,
+      "loss": 0.2612,
+      "step": 2157
+    },
+    {
+      "epoch": 1.4600811907983762,
+      "grad_norm": 0.5022532862855017,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.2454,
+      "step": 2158
+    },
+    {
+      "epoch": 1.4607577807848444,
+      "grad_norm": 0.5217477942515937,
+      "learning_rate": 2.0562983422917726e-06,
+      "loss": 0.2528,
+      "step": 2159
+    },
+    {
+      "epoch": 1.4614343707713127,
+      "grad_norm": 0.4971512201107231,
+      "learning_rate": 2.0515270521524562e-06,
+      "loss": 0.2444,
+      "step": 2160
+    },
+    {
+      "epoch": 1.462110960757781,
+      "grad_norm": 0.5235150877140263,
+      "learning_rate": 2.0467598747750533e-06,
+      "loss": 0.2592,
+      "step": 2161
+    },
+    {
+      "epoch": 1.462787550744249,
+      "grad_norm": 0.503997146745054,
+      "learning_rate": 2.041996816809197e-06,
+      "loss": 0.2578,
+      "step": 2162
+    },
+    {
+      "epoch": 1.4634641407307172,
+      "grad_norm": 0.525952452696541,
+      "learning_rate": 2.0372378848987777e-06,
+      "loss": 0.2571,
+      "step": 2163
+    },
+    {
+      "epoch": 1.4641407307171854,
+      "grad_norm": 0.5032033539961679,
+      "learning_rate": 2.032483085681927e-06,
+      "loss": 0.2425,
+      "step": 2164
+    },
+    {
+      "epoch": 1.4648173207036537,
+      "grad_norm": 0.5123494424185453,
+      "learning_rate": 2.027732425791011e-06,
+      "loss": 0.2465,
+      "step": 2165
+    },
+    {
+      "epoch": 1.465493910690122,
+      "grad_norm": 0.502047997800729,
+      "learning_rate": 2.0229859118526244e-06,
+      "loss": 0.2534,
+      "step": 2166
+    },
+    {
+      "epoch": 1.46617050067659,
+      "grad_norm": 0.5013988339777017,
+      "learning_rate": 2.018243550487581e-06,
+      "loss": 0.2501,
+      "step": 2167
+    },
+    {
+      "epoch": 1.4668470906630582,
+      "grad_norm": 0.4951761594502989,
+      "learning_rate": 2.0135053483108973e-06,
+      "loss": 0.2534,
+      "step": 2168
+    },
+    {
+      "epoch": 1.4675236806495264,
+      "grad_norm": 0.4668955330246146,
+      "learning_rate": 2.0087713119317896e-06,
+      "loss": 0.2381,
+      "step": 2169
+    },
+    {
+      "epoch": 1.4682002706359945,
+      "grad_norm": 0.5139878331259563,
+      "learning_rate": 2.004041447953663e-06,
+      "loss": 0.2583,
+      "step": 2170
+    },
+    {
+      "epoch": 1.468876860622463,
+      "grad_norm": 0.5217267634996896,
+      "learning_rate": 1.999315762974107e-06,
+      "loss": 0.262,
+      "step": 2171
+    },
+    {
+      "epoch": 1.469553450608931,
+      "grad_norm": 0.5090737167712815,
+      "learning_rate": 1.9945942635848745e-06,
+      "loss": 0.2496,
+      "step": 2172
+    },
+    {
+      "epoch": 1.4702300405953992,
+      "grad_norm": 0.532289259465562,
+      "learning_rate": 1.9898769563718873e-06,
+      "loss": 0.2547,
+      "step": 2173
+    },
+    {
+      "epoch": 1.4709066305818674,
+      "grad_norm": 0.4946456162196744,
+      "learning_rate": 1.985163847915217e-06,
+      "loss": 0.2572,
+      "step": 2174
+    },
+    {
+      "epoch": 1.4715832205683355,
+      "grad_norm": 0.49203767949173915,
+      "learning_rate": 1.9804549447890764e-06,
+      "loss": 0.2546,
+      "step": 2175
+    },
+    {
+      "epoch": 1.472259810554804,
+      "grad_norm": 0.5038903923342116,
+      "learning_rate": 1.9757502535618137e-06,
+      "loss": 0.2479,
+      "step": 2176
+    },
+    {
+      "epoch": 1.472936400541272,
+      "grad_norm": 0.5182291579632705,
+      "learning_rate": 1.971049780795901e-06,
+      "loss": 0.2577,
+      "step": 2177
+    },
+    {
+      "epoch": 1.4736129905277402,
+      "grad_norm": 0.5202108930486209,
+      "learning_rate": 1.9663535330479305e-06,
+      "loss": 0.2414,
+      "step": 2178
+    },
+    {
+      "epoch": 1.4742895805142084,
+      "grad_norm": 0.5287580993550303,
+      "learning_rate": 1.961661516868594e-06,
+      "loss": 0.2644,
+      "step": 2179
+    },
+    {
+      "epoch": 1.4749661705006765,
+      "grad_norm": 0.5153506772999866,
+      "learning_rate": 1.956973738802689e-06,
+      "loss": 0.2445,
+      "step": 2180
+    },
+    {
+      "epoch": 1.4756427604871447,
+      "grad_norm": 0.5409280970115099,
+      "learning_rate": 1.9522902053890925e-06,
+      "loss": 0.258,
+      "step": 2181
+    },
+    {
+      "epoch": 1.476319350473613,
+      "grad_norm": 0.5024900537651106,
+      "learning_rate": 1.9476109231607687e-06,
+      "loss": 0.2486,
+      "step": 2182
+    },
+    {
+      "epoch": 1.4769959404600812,
+      "grad_norm": 0.4890235828833275,
+      "learning_rate": 1.9429358986447476e-06,
+      "loss": 0.2461,
+      "step": 2183
+    },
+    {
+      "epoch": 1.4776725304465494,
+      "grad_norm": 0.5049333592598533,
+      "learning_rate": 1.938265138362118e-06,
+      "loss": 0.2511,
+      "step": 2184
+    },
+    {
+      "epoch": 1.4783491204330175,
+      "grad_norm": 0.5073314097023166,
+      "learning_rate": 1.9335986488280266e-06,
+      "loss": 0.2477,
+      "step": 2185
+    },
+    {
+      "epoch": 1.4790257104194857,
+      "grad_norm": 0.4845425565075815,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.2552,
+      "step": 2186
+    },
+    {
+      "epoch": 1.479702300405954,
+      "grad_norm": 0.5126629672203393,
+      "learning_rate": 1.9242785080362398e-06,
+      "loss": 0.2458,
+      "step": 2187
+    },
+    {
+      "epoch": 1.4803788903924222,
+      "grad_norm": 0.4891308629037526,
+      "learning_rate": 1.9196248697790066e-06,
+      "loss": 0.244,
+      "step": 2188
+    },
+    {
+      "epoch": 1.4810554803788905,
+      "grad_norm": 0.5168022420421734,
+      "learning_rate": 1.9149755282712247e-06,
+      "loss": 0.2441,
+      "step": 2189
+    },
+    {
+      "epoch": 1.4817320703653585,
+      "grad_norm": 0.48659671516972475,
+      "learning_rate": 1.9103304899981603e-06,
+      "loss": 0.2492,
+      "step": 2190
+    },
+    {
+      "epoch": 1.4824086603518267,
+      "grad_norm": 0.5016883221126635,
+      "learning_rate": 1.905689761439075e-06,
+      "loss": 0.2411,
+      "step": 2191
+    },
+    {
+      "epoch": 1.483085250338295,
+      "grad_norm": 0.5122248101017031,
+      "learning_rate": 1.901053349067225e-06,
+      "loss": 0.271,
+      "step": 2192
+    },
+    {
+      "epoch": 1.4837618403247632,
+      "grad_norm": 0.49844087681545296,
+      "learning_rate": 1.8964212593498444e-06,
+      "loss": 0.2578,
+      "step": 2193
+    },
+    {
+      "epoch": 1.4844384303112315,
+      "grad_norm": 0.5155195089788851,
+      "learning_rate": 1.891793498748134e-06,
+      "loss": 0.2602,
+      "step": 2194
+    },
+    {
+      "epoch": 1.4851150202976995,
+      "grad_norm": 0.49740451071521874,
+      "learning_rate": 1.8871700737172572e-06,
+      "loss": 0.2447,
+      "step": 2195
+    },
+    {
+      "epoch": 1.4857916102841677,
+      "grad_norm": 0.4836623699420873,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.248,
+      "step": 2196
+    },
+    {
+      "epoch": 1.486468200270636,
+      "grad_norm": 0.5064840001071086,
+      "learning_rate": 1.8779362561584209e-06,
+      "loss": 0.2497,
+      "step": 2197
+    },
+    {
+      "epoch": 1.4871447902571042,
+      "grad_norm": 0.5065890296247518,
+      "learning_rate": 1.8733258765105129e-06,
+      "loss": 0.2623,
+      "step": 2198
+    },
+    {
+      "epoch": 1.4878213802435725,
+      "grad_norm": 0.5181408069465013,
+      "learning_rate": 1.8687198581935318e-06,
+      "loss": 0.2729,
+      "step": 2199
+    },
+    {
+      "epoch": 1.4884979702300405,
+      "grad_norm": 0.5146307012504393,
+      "learning_rate": 1.864118207632315e-06,
+      "loss": 0.2506,
+      "step": 2200
+    },
+    {
+      "epoch": 1.4891745602165087,
+      "grad_norm": 0.5079529589067043,
+      "learning_rate": 1.8595209312456052e-06,
+      "loss": 0.2684,
+      "step": 2201
+    },
+    {
+      "epoch": 1.489851150202977,
+      "grad_norm": 0.4992902343651994,
+      "learning_rate": 1.8549280354460437e-06,
+      "loss": 0.2543,
+      "step": 2202
+    },
+    {
+      "epoch": 1.4905277401894452,
+      "grad_norm": 0.4907393173250745,
+      "learning_rate": 1.8503395266401652e-06,
+      "loss": 0.2514,
+      "step": 2203
+    },
+    {
+      "epoch": 1.4912043301759135,
+      "grad_norm": 0.5004570415735573,
+      "learning_rate": 1.845755411228382e-06,
+      "loss": 0.249,
+      "step": 2204
+    },
+    {
+      "epoch": 1.4918809201623815,
+      "grad_norm": 0.5177338790980202,
+      "learning_rate": 1.8411756956049758e-06,
+      "loss": 0.2625,
+      "step": 2205
+    },
+    {
+      "epoch": 1.4925575101488497,
+      "grad_norm": 0.49034220898338254,
+      "learning_rate": 1.8366003861580966e-06,
+      "loss": 0.2421,
+      "step": 2206
+    },
+    {
+      "epoch": 1.493234100135318,
+      "grad_norm": 0.47707123785749406,
+      "learning_rate": 1.8320294892697477e-06,
+      "loss": 0.2441,
+      "step": 2207
+    },
+    {
+      "epoch": 1.4939106901217862,
+      "grad_norm": 0.4887188329272896,
+      "learning_rate": 1.8274630113157727e-06,
+      "loss": 0.2423,
+      "step": 2208
+    },
+    {
+      "epoch": 1.4945872801082545,
+      "grad_norm": 0.5094788232800171,
+      "learning_rate": 1.8229009586658552e-06,
+      "loss": 0.2528,
+      "step": 2209
+    },
+    {
+      "epoch": 1.4952638700947225,
+      "grad_norm": 0.5301942609229953,
+      "learning_rate": 1.818343337683503e-06,
+      "loss": 0.2644,
+      "step": 2210
+    },
+    {
+      "epoch": 1.4959404600811907,
+      "grad_norm": 0.5051470315552674,
+      "learning_rate": 1.8137901547260472e-06,
+      "loss": 0.2549,
+      "step": 2211
+    },
+    {
+      "epoch": 1.496617050067659,
+      "grad_norm": 0.4863470153196226,
+      "learning_rate": 1.8092414161446225e-06,
+      "loss": 0.2548,
+      "step": 2212
+    },
+    {
+      "epoch": 1.4972936400541272,
+      "grad_norm": 0.5101771418035616,
+      "learning_rate": 1.8046971282841686e-06,
+      "loss": 0.2494,
+      "step": 2213
+    },
+    {
+      "epoch": 1.4979702300405955,
+      "grad_norm": 0.506658497542102,
+      "learning_rate": 1.8001572974834169e-06,
+      "loss": 0.2558,
+      "step": 2214
+    },
+    {
+      "epoch": 1.4986468200270635,
+      "grad_norm": 0.497120110991891,
+      "learning_rate": 1.7956219300748796e-06,
+      "loss": 0.2465,
+      "step": 2215
+    },
+    {
+      "epoch": 1.4993234100135318,
+      "grad_norm": 0.5371600455989929,
+      "learning_rate": 1.7910910323848435e-06,
+      "loss": 0.2509,
+      "step": 2216
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4978761500823446,
+      "learning_rate": 1.7865646107333585e-06,
+      "loss": 0.2495,
+      "step": 2217
+    },
+    {
+      "epoch": 1.5006765899864682,
+      "grad_norm": 0.48304875949671816,
+      "learning_rate": 1.7820426714342375e-06,
+      "loss": 0.2431,
+      "step": 2218
+    },
+    {
+      "epoch": 1.5013531799729365,
+      "grad_norm": 0.538759445416668,
+      "learning_rate": 1.777525220795034e-06,
+      "loss": 0.2587,
+      "step": 2219
+    },
+    {
+      "epoch": 1.5020297699594045,
+      "grad_norm": 0.5022381164535056,
+      "learning_rate": 1.7730122651170457e-06,
+      "loss": 0.2529,
+      "step": 2220
+    },
+    {
+      "epoch": 1.5027063599458728,
+      "grad_norm": 0.47069737729056504,
+      "learning_rate": 1.7685038106952952e-06,
+      "loss": 0.2385,
+      "step": 2221
+    },
+    {
+      "epoch": 1.503382949932341,
+      "grad_norm": 0.48885546174632694,
+      "learning_rate": 1.763999863818533e-06,
+      "loss": 0.2443,
+      "step": 2222
+    },
+    {
+      "epoch": 1.5040595399188093,
+      "grad_norm": 0.5111337540112224,
+      "learning_rate": 1.759500430769216e-06,
+      "loss": 0.2553,
+      "step": 2223
+    },
+    {
+      "epoch": 1.5047361299052775,
+      "grad_norm": 0.4831420596818458,
+      "learning_rate": 1.755005517823506e-06,
+      "loss": 0.2474,
+      "step": 2224
+    },
+    {
+      "epoch": 1.5054127198917455,
+      "grad_norm": 0.4979862505690422,
+      "learning_rate": 1.7505151312512642e-06,
+      "loss": 0.246,
+      "step": 2225
+    },
+    {
+      "epoch": 1.5060893098782138,
+      "grad_norm": 0.5030029798673477,
+      "learning_rate": 1.7460292773160315e-06,
+      "loss": 0.2486,
+      "step": 2226
+    },
+    {
+      "epoch": 1.506765899864682,
+      "grad_norm": 0.5056416467897953,
+      "learning_rate": 1.7415479622750336e-06,
+      "loss": 0.2549,
+      "step": 2227
+    },
+    {
+      "epoch": 1.5074424898511503,
+      "grad_norm": 0.5153726543397203,
+      "learning_rate": 1.7370711923791567e-06,
+      "loss": 0.2585,
+      "step": 2228
+    },
+    {
+      "epoch": 1.5081190798376185,
+      "grad_norm": 0.5084531165890156,
+      "learning_rate": 1.7325989738729555e-06,
+      "loss": 0.2591,
+      "step": 2229
+    },
+    {
+      "epoch": 1.5087956698240865,
+      "grad_norm": 0.5045544775048464,
+      "learning_rate": 1.7281313129946302e-06,
+      "loss": 0.2518,
+      "step": 2230
+    },
+    {
+      "epoch": 1.5094722598105548,
+      "grad_norm": 0.4829349857244556,
+      "learning_rate": 1.723668215976026e-06,
+      "loss": 0.2506,
+      "step": 2231
+    },
+    {
+      "epoch": 1.510148849797023,
+      "grad_norm": 0.4908122899421733,
+      "learning_rate": 1.7192096890426192e-06,
+      "loss": 0.2518,
+      "step": 2232
+    },
+    {
+      "epoch": 1.510825439783491,
+      "grad_norm": 0.520890119837811,
+      "learning_rate": 1.7147557384135156e-06,
+      "loss": 0.2699,
+      "step": 2233
+    },
+    {
+      "epoch": 1.5115020297699595,
+      "grad_norm": 0.5100057115172657,
+      "learning_rate": 1.7103063703014372e-06,
+      "loss": 0.2422,
+      "step": 2234
+    },
+    {
+      "epoch": 1.5121786197564275,
+      "grad_norm": 0.489933170700976,
+      "learning_rate": 1.7058615909127102e-06,
+      "loss": 0.2483,
+      "step": 2235
+    },
+    {
+      "epoch": 1.5128552097428958,
+      "grad_norm": 0.49964097014985515,
+      "learning_rate": 1.7014214064472646e-06,
+      "loss": 0.2438,
+      "step": 2236
+    },
+    {
+      "epoch": 1.513531799729364,
+      "grad_norm": 0.518908070485271,
+      "learning_rate": 1.6969858230986174e-06,
+      "loss": 0.2553,
+      "step": 2237
+    },
+    {
+      "epoch": 1.514208389715832,
+      "grad_norm": 0.49839750202927946,
+      "learning_rate": 1.6925548470538695e-06,
+      "loss": 0.2616,
+      "step": 2238
+    },
+    {
+      "epoch": 1.5148849797023005,
+      "grad_norm": 0.49940082577477213,
+      "learning_rate": 1.6881284844936923e-06,
+      "loss": 0.2452,
+      "step": 2239
+    },
+    {
+      "epoch": 1.5155615696887685,
+      "grad_norm": 0.5207769564384882,
+      "learning_rate": 1.683706741592327e-06,
+      "loss": 0.2555,
+      "step": 2240
+    },
+    {
+      "epoch": 1.5162381596752368,
+      "grad_norm": 0.529780185578341,
+      "learning_rate": 1.6792896245175693e-06,
+      "loss": 0.2626,
+      "step": 2241
+    },
+    {
+      "epoch": 1.516914749661705,
+      "grad_norm": 0.5157578014441285,
+      "learning_rate": 1.6748771394307584e-06,
+      "loss": 0.2569,
+      "step": 2242
+    },
+    {
+      "epoch": 1.517591339648173,
+      "grad_norm": 0.48633322108769683,
+      "learning_rate": 1.6704692924867783e-06,
+      "loss": 0.2482,
+      "step": 2243
+    },
+    {
+      "epoch": 1.5182679296346415,
+      "grad_norm": 0.5135354314583718,
+      "learning_rate": 1.6660660898340392e-06,
+      "loss": 0.2556,
+      "step": 2244
+    },
+    {
+      "epoch": 1.5189445196211095,
+      "grad_norm": 0.48227244222831817,
+      "learning_rate": 1.6616675376144741e-06,
+      "loss": 0.2515,
+      "step": 2245
+    },
+    {
+      "epoch": 1.5196211096075778,
+      "grad_norm": 0.4954547739409768,
+      "learning_rate": 1.6572736419635288e-06,
+      "loss": 0.2549,
+      "step": 2246
+    },
+    {
+      "epoch": 1.520297699594046,
+      "grad_norm": 0.4916277866857519,
+      "learning_rate": 1.6528844090101548e-06,
+      "loss": 0.2406,
+      "step": 2247
+    },
+    {
+      "epoch": 1.520974289580514,
+      "grad_norm": 0.5302259016237963,
+      "learning_rate": 1.648499844876802e-06,
+      "loss": 0.2562,
+      "step": 2248
+    },
+    {
+      "epoch": 1.5216508795669825,
+      "grad_norm": 0.5154912123882129,
+      "learning_rate": 1.6441199556794036e-06,
+      "loss": 0.2533,
+      "step": 2249
+    },
+    {
+      "epoch": 1.5223274695534506,
+      "grad_norm": 0.48118305719574817,
+      "learning_rate": 1.639744747527371e-06,
+      "loss": 0.2355,
+      "step": 2250
+    },
+    {
+      "epoch": 1.5230040595399188,
+      "grad_norm": 0.49611352782876417,
+      "learning_rate": 1.6353742265235923e-06,
+      "loss": 0.2472,
+      "step": 2251
+    },
+    {
+      "epoch": 1.523680649526387,
+      "grad_norm": 0.48324739510390763,
+      "learning_rate": 1.631008398764412e-06,
+      "loss": 0.2438,
+      "step": 2252
+    },
+    {
+      "epoch": 1.524357239512855,
+      "grad_norm": 0.5095770051755408,
+      "learning_rate": 1.6266472703396286e-06,
+      "loss": 0.2566,
+      "step": 2253
+    },
+    {
+      "epoch": 1.5250338294993235,
+      "grad_norm": 0.5116206131848741,
+      "learning_rate": 1.6222908473324889e-06,
+      "loss": 0.2618,
+      "step": 2254
+    },
+    {
+      "epoch": 1.5257104194857916,
+      "grad_norm": 0.515743142235952,
+      "learning_rate": 1.617939135819675e-06,
+      "loss": 0.2639,
+      "step": 2255
+    },
+    {
+      "epoch": 1.5263870094722598,
+      "grad_norm": 0.5362045442174803,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.2603,
+      "step": 2256
+    },
+    {
+      "epoch": 1.527063599458728,
+      "grad_norm": 0.49777481823140085,
+      "learning_rate": 1.6092498715508776e-06,
+      "loss": 0.2464,
+      "step": 2257
+    },
+    {
+      "epoch": 1.527740189445196,
+      "grad_norm": 0.5143201468260413,
+      "learning_rate": 1.604912330915364e-06,
+      "loss": 0.2593,
+      "step": 2258
+    },
+    {
+      "epoch": 1.5284167794316645,
+      "grad_norm": 0.496119108090543,
+      "learning_rate": 1.6005795260150963e-06,
+      "loss": 0.2454,
+      "step": 2259
+    },
+    {
+      "epoch": 1.5290933694181326,
+      "grad_norm": 0.49146253664142636,
+      "learning_rate": 1.5962514628938103e-06,
+      "loss": 0.2568,
+      "step": 2260
+    },
+    {
+      "epoch": 1.5297699594046008,
+      "grad_norm": 0.48113139584234954,
+      "learning_rate": 1.59192814758863e-06,
+      "loss": 0.2419,
+      "step": 2261
+    },
+    {
+      "epoch": 1.530446549391069,
+      "grad_norm": 0.49368983278849476,
+      "learning_rate": 1.5876095861300567e-06,
+      "loss": 0.2491,
+      "step": 2262
+    },
+    {
+      "epoch": 1.531123139377537,
+      "grad_norm": 0.49811782852443287,
+      "learning_rate": 1.5832957845419583e-06,
+      "loss": 0.2496,
+      "step": 2263
+    },
+    {
+      "epoch": 1.5317997293640055,
+      "grad_norm": 0.4925244724245926,
+      "learning_rate": 1.5789867488415633e-06,
+      "loss": 0.251,
+      "step": 2264
+    },
+    {
+      "epoch": 1.5324763193504736,
+      "grad_norm": 0.48633080202066975,
+      "learning_rate": 1.5746824850394559e-06,
+      "loss": 0.2399,
+      "step": 2265
+    },
+    {
+      "epoch": 1.5331529093369418,
+      "grad_norm": 0.509685213589456,
+      "learning_rate": 1.5703829991395602e-06,
+      "loss": 0.2502,
+      "step": 2266
+    },
+    {
+      "epoch": 1.53382949932341,
+      "grad_norm": 0.4985575908026002,
+      "learning_rate": 1.5660882971391367e-06,
+      "loss": 0.2447,
+      "step": 2267
+    },
+    {
+      "epoch": 1.534506089309878,
+      "grad_norm": 0.5065433209098394,
+      "learning_rate": 1.5617983850287737e-06,
+      "loss": 0.2591,
+      "step": 2268
+    },
+    {
+      "epoch": 1.5351826792963466,
+      "grad_norm": 0.48244600317726066,
+      "learning_rate": 1.55751326879238e-06,
+      "loss": 0.2425,
+      "step": 2269
+    },
+    {
+      "epoch": 1.5358592692828146,
+      "grad_norm": 0.5076282483132682,
+      "learning_rate": 1.5532329544071712e-06,
+      "loss": 0.2499,
+      "step": 2270
+    },
+    {
+      "epoch": 1.5365358592692828,
+      "grad_norm": 0.49204960547426035,
+      "learning_rate": 1.5489574478436664e-06,
+      "loss": 0.2443,
+      "step": 2271
+    },
+    {
+      "epoch": 1.537212449255751,
+      "grad_norm": 0.5207033739993037,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.2445,
+      "step": 2272
+    },
+    {
+      "epoch": 1.537889039242219,
+      "grad_norm": 0.5036719148355483,
+      "learning_rate": 1.5404208820303041e-06,
+      "loss": 0.2471,
+      "step": 2273
+    },
+    {
+      "epoch": 1.5385656292286876,
+      "grad_norm": 0.5137284678580989,
+      "learning_rate": 1.5361598346879193e-06,
+      "loss": 0.2552,
+      "step": 2274
+    },
+    {
+      "epoch": 1.5392422192151556,
+      "grad_norm": 0.5065100494359286,
+      "learning_rate": 1.5319036189821678e-06,
+      "loss": 0.242,
+      "step": 2275
+    },
+    {
+      "epoch": 1.5399188092016238,
+      "grad_norm": 0.5530591336603423,
+      "learning_rate": 1.5276522408499567e-06,
+      "loss": 0.2519,
+      "step": 2276
+    },
+    {
+      "epoch": 1.540595399188092,
+      "grad_norm": 0.515399999811938,
+      "learning_rate": 1.5234057062214403e-06,
+      "loss": 0.2524,
+      "step": 2277
+    },
+    {
+      "epoch": 1.54127198917456,
+      "grad_norm": 0.49706303964285503,
+      "learning_rate": 1.5191640210200186e-06,
+      "loss": 0.2484,
+      "step": 2278
+    },
+    {
+      "epoch": 1.5419485791610286,
+      "grad_norm": 0.5040314240257913,
+      "learning_rate": 1.514927191162327e-06,
+      "loss": 0.2631,
+      "step": 2279
+    },
+    {
+      "epoch": 1.5426251691474966,
+      "grad_norm": 0.5087276939619023,
+      "learning_rate": 1.5106952225582312e-06,
+      "loss": 0.25,
+      "step": 2280
+    },
+    {
+      "epoch": 1.5433017591339648,
+      "grad_norm": 0.49488526035134994,
+      "learning_rate": 1.5064681211108112e-06,
+      "loss": 0.242,
+      "step": 2281
+    },
+    {
+      "epoch": 1.543978349120433,
+      "grad_norm": 0.5192600328379066,
+      "learning_rate": 1.5022458927163618e-06,
+      "loss": 0.2511,
+      "step": 2282
+    },
+    {
+      "epoch": 1.544654939106901,
+      "grad_norm": 0.540131556502969,
+      "learning_rate": 1.498028543264381e-06,
+      "loss": 0.2515,
+      "step": 2283
+    },
+    {
+      "epoch": 1.5453315290933696,
+      "grad_norm": 0.48940012995836046,
+      "learning_rate": 1.4938160786375571e-06,
+      "loss": 0.2418,
+      "step": 2284
+    },
+    {
+      "epoch": 1.5460081190798376,
+      "grad_norm": 0.4969496268285726,
+      "learning_rate": 1.4896085047117682e-06,
+      "loss": 0.2503,
+      "step": 2285
+    },
+    {
+      "epoch": 1.5466847090663058,
+      "grad_norm": 0.4919168609771574,
+      "learning_rate": 1.4854058273560667e-06,
+      "loss": 0.2463,
+      "step": 2286
+    },
+    {
+      "epoch": 1.547361299052774,
+      "grad_norm": 0.5095318070080931,
+      "learning_rate": 1.4812080524326817e-06,
+      "loss": 0.2449,
+      "step": 2287
+    },
+    {
+      "epoch": 1.548037889039242,
+      "grad_norm": 0.5092506070645164,
+      "learning_rate": 1.477015185796995e-06,
+      "loss": 0.2536,
+      "step": 2288
+    },
+    {
+      "epoch": 1.5487144790257106,
+      "grad_norm": 0.513321624520788,
+      "learning_rate": 1.4728272332975507e-06,
+      "loss": 0.2486,
+      "step": 2289
+    },
+    {
+      "epoch": 1.5493910690121786,
+      "grad_norm": 0.4994426230499301,
+      "learning_rate": 1.4686442007760315e-06,
+      "loss": 0.2467,
+      "step": 2290
+    },
+    {
+      "epoch": 1.5500676589986468,
+      "grad_norm": 0.48326191970897137,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.2528,
+      "step": 2291
+    },
+    {
+      "epoch": 1.550744248985115,
+      "grad_norm": 0.4863685048092452,
+      "learning_rate": 1.460292918999195e-06,
+      "loss": 0.2488,
+      "step": 2292
+    },
+    {
+      "epoch": 1.5514208389715831,
+      "grad_norm": 0.5129250668809845,
+      "learning_rate": 1.4561246813928985e-06,
+      "loss": 0.2531,
+      "step": 2293
+    },
+    {
+      "epoch": 1.5520974289580516,
+      "grad_norm": 0.5156650140413478,
+      "learning_rate": 1.4519613870625632e-06,
+      "loss": 0.2579,
+      "step": 2294
+    },
+    {
+      "epoch": 1.5527740189445196,
+      "grad_norm": 0.5053508121089274,
+      "learning_rate": 1.4478030418154753e-06,
+      "loss": 0.2542,
+      "step": 2295
+    },
+    {
+      "epoch": 1.5534506089309879,
+      "grad_norm": 0.4770399557807679,
+      "learning_rate": 1.4436496514520253e-06,
+      "loss": 0.252,
+      "step": 2296
+    },
+    {
+      "epoch": 1.554127198917456,
+      "grad_norm": 0.5087901541666479,
+      "learning_rate": 1.4395012217656856e-06,
+      "loss": 0.2523,
+      "step": 2297
+    },
+    {
+      "epoch": 1.5548037889039241,
+      "grad_norm": 0.4794938052772905,
+      "learning_rate": 1.4353577585430152e-06,
+      "loss": 0.2503,
+      "step": 2298
+    },
+    {
+      "epoch": 1.5554803788903924,
+      "grad_norm": 0.4994055699024229,
+      "learning_rate": 1.4312192675636406e-06,
+      "loss": 0.2443,
+      "step": 2299
+    },
+    {
+      "epoch": 1.5561569688768606,
+      "grad_norm": 0.5086687913509113,
+      "learning_rate": 1.4270857546002548e-06,
+      "loss": 0.252,
+      "step": 2300
+    },
+    {
+      "epoch": 1.5568335588633289,
+      "grad_norm": 0.5149626555561578,
+      "learning_rate": 1.4229572254186047e-06,
+      "loss": 0.2477,
+      "step": 2301
+    },
+    {
+      "epoch": 1.557510148849797,
+      "grad_norm": 0.4595522623291559,
+      "learning_rate": 1.4188336857774892e-06,
+      "loss": 0.2459,
+      "step": 2302
+    },
+    {
+      "epoch": 1.5581867388362651,
+      "grad_norm": 0.4890133471232923,
+      "learning_rate": 1.4147151414287475e-06,
+      "loss": 0.2464,
+      "step": 2303
+    },
+    {
+      "epoch": 1.5588633288227334,
+      "grad_norm": 0.4885307127938241,
+      "learning_rate": 1.410601598117246e-06,
+      "loss": 0.2528,
+      "step": 2304
+    },
+    {
+      "epoch": 1.5595399188092016,
+      "grad_norm": 0.49945376015906306,
+      "learning_rate": 1.406493061580881e-06,
+      "loss": 0.2471,
+      "step": 2305
+    },
+    {
+      "epoch": 1.5602165087956699,
+      "grad_norm": 0.5021121483605259,
+      "learning_rate": 1.4023895375505608e-06,
+      "loss": 0.2456,
+      "step": 2306
+    },
+    {
+      "epoch": 1.560893098782138,
+      "grad_norm": 0.5051262239875945,
+      "learning_rate": 1.3982910317502042e-06,
+      "loss": 0.2593,
+      "step": 2307
+    },
+    {
+      "epoch": 1.5615696887686061,
+      "grad_norm": 0.49976004489416337,
+      "learning_rate": 1.3941975498967265e-06,
+      "loss": 0.2579,
+      "step": 2308
+    },
+    {
+      "epoch": 1.5622462787550744,
+      "grad_norm": 0.5091309088295877,
+      "learning_rate": 1.3901090977000409e-06,
+      "loss": 0.2487,
+      "step": 2309
+    },
+    {
+      "epoch": 1.5629228687415426,
+      "grad_norm": 0.49353558668292125,
+      "learning_rate": 1.3860256808630429e-06,
+      "loss": 0.2415,
+      "step": 2310
+    },
+    {
+      "epoch": 1.5635994587280109,
+      "grad_norm": 0.486910519051264,
+      "learning_rate": 1.3819473050816002e-06,
+      "loss": 0.2474,
+      "step": 2311
+    },
+    {
+      "epoch": 1.5642760487144791,
+      "grad_norm": 0.5062883607533053,
+      "learning_rate": 1.3778739760445552e-06,
+      "loss": 0.2495,
+      "step": 2312
+    },
+    {
+      "epoch": 1.5649526387009471,
+      "grad_norm": 0.4961406956516028,
+      "learning_rate": 1.3738056994337068e-06,
+      "loss": 0.2444,
+      "step": 2313
+    },
+    {
+      "epoch": 1.5656292286874154,
+      "grad_norm": 0.5034306630112082,
+      "learning_rate": 1.3697424809238058e-06,
+      "loss": 0.2561,
+      "step": 2314
+    },
+    {
+      "epoch": 1.5663058186738836,
+      "grad_norm": 0.5002454084476611,
+      "learning_rate": 1.365684326182548e-06,
+      "loss": 0.2379,
+      "step": 2315
+    },
+    {
+      "epoch": 1.5669824086603519,
+      "grad_norm": 0.5401755246246636,
+      "learning_rate": 1.361631240870569e-06,
+      "loss": 0.2596,
+      "step": 2316
+    },
+    {
+      "epoch": 1.5676589986468201,
+      "grad_norm": 0.5037057384170571,
+      "learning_rate": 1.357583230641431e-06,
+      "loss": 0.254,
+      "step": 2317
+    },
+    {
+      "epoch": 1.5683355886332881,
+      "grad_norm": 0.4800788828077926,
+      "learning_rate": 1.3535403011416159e-06,
+      "loss": 0.2432,
+      "step": 2318
+    },
+    {
+      "epoch": 1.5690121786197564,
+      "grad_norm": 0.49550376949144637,
+      "learning_rate": 1.3495024580105193e-06,
+      "loss": 0.2532,
+      "step": 2319
+    },
+    {
+      "epoch": 1.5696887686062246,
+      "grad_norm": 0.49833595353267474,
+      "learning_rate": 1.3454697068804434e-06,
+      "loss": 0.25,
+      "step": 2320
+    },
+    {
+      "epoch": 1.5703653585926927,
+      "grad_norm": 0.5119887986703091,
+      "learning_rate": 1.341442053376587e-06,
+      "loss": 0.2581,
+      "step": 2321
+    },
+    {
+      "epoch": 1.5710419485791611,
+      "grad_norm": 0.5228915662366979,
+      "learning_rate": 1.337419503117035e-06,
+      "loss": 0.256,
+      "step": 2322
+    },
+    {
+      "epoch": 1.5717185385656292,
+      "grad_norm": 0.4875404031073257,
+      "learning_rate": 1.3334020617127585e-06,
+      "loss": 0.2457,
+      "step": 2323
+    },
+    {
+      "epoch": 1.5723951285520974,
+      "grad_norm": 0.5110919098813891,
+      "learning_rate": 1.3293897347676032e-06,
+      "loss": 0.2432,
+      "step": 2324
+    },
+    {
+      "epoch": 1.5730717185385656,
+      "grad_norm": 0.5143359790761378,
+      "learning_rate": 1.3253825278782756e-06,
+      "loss": 0.2511,
+      "step": 2325
+    },
+    {
+      "epoch": 1.5737483085250337,
+      "grad_norm": 0.5083210388274393,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.2346,
+      "step": 2326
+    },
+    {
+      "epoch": 1.5744248985115021,
+      "grad_norm": 0.5232895366732199,
+      "learning_rate": 1.3173834966182226e-06,
+      "loss": 0.2502,
+      "step": 2327
+    },
+    {
+      "epoch": 1.5751014884979702,
+      "grad_norm": 0.49700734337916885,
+      "learning_rate": 1.313391683405177e-06,
+      "loss": 0.248,
+      "step": 2328
+    },
+    {
+      "epoch": 1.5757780784844384,
+      "grad_norm": 0.5185083510367375,
+      "learning_rate": 1.3094050125632973e-06,
+      "loss": 0.2494,
+      "step": 2329
+    },
+    {
+      "epoch": 1.5764546684709067,
+      "grad_norm": 0.5121054836662874,
+      "learning_rate": 1.305423489653508e-06,
+      "loss": 0.2516,
+      "step": 2330
+    },
+    {
+      "epoch": 1.5771312584573747,
+      "grad_norm": 0.4826278144703736,
+      "learning_rate": 1.3014471202295514e-06,
+      "loss": 0.2532,
+      "step": 2331
+    },
+    {
+      "epoch": 1.5778078484438431,
+      "grad_norm": 0.4937382766324471,
+      "learning_rate": 1.297475909837979e-06,
+      "loss": 0.2503,
+      "step": 2332
+    },
+    {
+      "epoch": 1.5784844384303112,
+      "grad_norm": 0.49915281933442934,
+      "learning_rate": 1.293509864018146e-06,
+      "loss": 0.2483,
+      "step": 2333
+    },
+    {
+      "epoch": 1.5791610284167794,
+      "grad_norm": 0.4978540270134428,
+      "learning_rate": 1.289548988302207e-06,
+      "loss": 0.251,
+      "step": 2334
+    },
+    {
+      "epoch": 1.5798376184032477,
+      "grad_norm": 0.5093440626299053,
+      "learning_rate": 1.2855932882151024e-06,
+      "loss": 0.258,
+      "step": 2335
+    },
+    {
+      "epoch": 1.5805142083897157,
+      "grad_norm": 0.49652803418860036,
+      "learning_rate": 1.281642769274552e-06,
+      "loss": 0.2523,
+      "step": 2336
+    },
+    {
+      "epoch": 1.5811907983761841,
+      "grad_norm": 0.49268464687412544,
+      "learning_rate": 1.2776974369910517e-06,
+      "loss": 0.2517,
+      "step": 2337
+    },
+    {
+      "epoch": 1.5818673883626522,
+      "grad_norm": 0.4943781305556024,
+      "learning_rate": 1.2737572968678624e-06,
+      "loss": 0.2569,
+      "step": 2338
+    },
+    {
+      "epoch": 1.5825439783491204,
+      "grad_norm": 0.4878796455710039,
+      "learning_rate": 1.2698223544010003e-06,
+      "loss": 0.2477,
+      "step": 2339
+    },
+    {
+      "epoch": 1.5832205683355887,
+      "grad_norm": 0.5102175444926299,
+      "learning_rate": 1.2658926150792321e-06,
+      "loss": 0.2553,
+      "step": 2340
+    },
+    {
+      "epoch": 1.5838971583220567,
+      "grad_norm": 0.5201162687091095,
+      "learning_rate": 1.261968084384066e-06,
+      "loss": 0.2631,
+      "step": 2341
+    },
+    {
+      "epoch": 1.5845737483085252,
+      "grad_norm": 0.5082152270344162,
+      "learning_rate": 1.2580487677897496e-06,
+      "loss": 0.2429,
+      "step": 2342
+    },
+    {
+      "epoch": 1.5852503382949932,
+      "grad_norm": 0.5057432079699692,
+      "learning_rate": 1.2541346707632495e-06,
+      "loss": 0.2422,
+      "step": 2343
+    },
+    {
+      "epoch": 1.5859269282814614,
+      "grad_norm": 0.4801073129734185,
+      "learning_rate": 1.250225798764259e-06,
+      "loss": 0.2476,
+      "step": 2344
+    },
+    {
+      "epoch": 1.5866035182679297,
+      "grad_norm": 0.49729434075629914,
+      "learning_rate": 1.2463221572451805e-06,
+      "loss": 0.2462,
+      "step": 2345
+    },
+    {
+      "epoch": 1.5872801082543977,
+      "grad_norm": 0.5092566694891173,
+      "learning_rate": 1.242423751651119e-06,
+      "loss": 0.2575,
+      "step": 2346
+    },
+    {
+      "epoch": 1.5879566982408662,
+      "grad_norm": 0.48393258184928617,
+      "learning_rate": 1.2385305874198778e-06,
+      "loss": 0.244,
+      "step": 2347
+    },
+    {
+      "epoch": 1.5886332882273342,
+      "grad_norm": 0.5019597414425726,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.2484,
+      "step": 2348
+    },
+    {
+      "epoch": 1.5893098782138024,
+      "grad_norm": 0.5102861554816435,
+      "learning_rate": 1.2307600047604995e-06,
+      "loss": 0.2546,
+      "step": 2349
+    },
+    {
+      "epoch": 1.5899864682002707,
+      "grad_norm": 0.4988204171131413,
+      "learning_rate": 1.2268825971713833e-06,
+      "loss": 0.2413,
+      "step": 2350
+    },
+    {
+      "epoch": 1.5906630581867387,
+      "grad_norm": 0.5173250161233183,
+      "learning_rate": 1.2230104526231107e-06,
+      "loss": 0.2543,
+      "step": 2351
+    },
+    {
+      "epoch": 1.5913396481732072,
+      "grad_norm": 0.5195203030377427,
+      "learning_rate": 1.2191435765168557e-06,
+      "loss": 0.2472,
+      "step": 2352
+    },
+    {
+      "epoch": 1.5920162381596752,
+      "grad_norm": 0.49045964406440823,
+      "learning_rate": 1.2152819742464394e-06,
+      "loss": 0.2479,
+      "step": 2353
+    },
+    {
+      "epoch": 1.5926928281461434,
+      "grad_norm": 0.4992711926537573,
+      "learning_rate": 1.2114256511983274e-06,
+      "loss": 0.2503,
+      "step": 2354
+    },
+    {
+      "epoch": 1.5933694181326117,
+      "grad_norm": 0.49167964381533064,
+      "learning_rate": 1.2075746127516224e-06,
+      "loss": 0.2527,
+      "step": 2355
+    },
+    {
+      "epoch": 1.5940460081190797,
+      "grad_norm": 0.48743136911149176,
+      "learning_rate": 1.2037288642780575e-06,
+      "loss": 0.2538,
+      "step": 2356
+    },
+    {
+      "epoch": 1.5947225981055482,
+      "grad_norm": 0.481325186427767,
+      "learning_rate": 1.1998884111419828e-06,
+      "loss": 0.2487,
+      "step": 2357
+    },
+    {
+      "epoch": 1.5953991880920162,
+      "grad_norm": 0.5176632073288059,
+      "learning_rate": 1.1960532587003666e-06,
+      "loss": 0.2614,
+      "step": 2358
+    },
+    {
+      "epoch": 1.5960757780784844,
+      "grad_norm": 0.5015022536054188,
+      "learning_rate": 1.192223412302778e-06,
+      "loss": 0.2507,
+      "step": 2359
+    },
+    {
+      "epoch": 1.5967523680649527,
+      "grad_norm": 0.5164667561851997,
+      "learning_rate": 1.1883988772913924e-06,
+      "loss": 0.2535,
+      "step": 2360
+    },
+    {
+      "epoch": 1.5974289580514207,
+      "grad_norm": 0.4879423401168632,
+      "learning_rate": 1.1845796590009684e-06,
+      "loss": 0.2435,
+      "step": 2361
+    },
+    {
+      "epoch": 1.5981055480378892,
+      "grad_norm": 0.49947035219966457,
+      "learning_rate": 1.180765762758852e-06,
+      "loss": 0.2456,
+      "step": 2362
+    },
+    {
+      "epoch": 1.5987821380243572,
+      "grad_norm": 0.4987191209959,
+      "learning_rate": 1.1769571938849683e-06,
+      "loss": 0.2531,
+      "step": 2363
+    },
+    {
+      "epoch": 1.5994587280108254,
+      "grad_norm": 0.4903192326370862,
+      "learning_rate": 1.173153957691805e-06,
+      "loss": 0.2482,
+      "step": 2364
+    },
+    {
+      "epoch": 1.6001353179972937,
+      "grad_norm": 0.5003189391052887,
+      "learning_rate": 1.1693560594844183e-06,
+      "loss": 0.2543,
+      "step": 2365
+    },
+    {
+      "epoch": 1.6008119079837617,
+      "grad_norm": 0.48690873688597364,
+      "learning_rate": 1.165563504560413e-06,
+      "loss": 0.2471,
+      "step": 2366
+    },
+    {
+      "epoch": 1.6014884979702302,
+      "grad_norm": 0.5136188912595475,
+      "learning_rate": 1.1617762982099446e-06,
+      "loss": 0.255,
+      "step": 2367
+    },
+    {
+      "epoch": 1.6021650879566982,
+      "grad_norm": 0.505801439270711,
+      "learning_rate": 1.157994445715706e-06,
+      "loss": 0.2497,
+      "step": 2368
+    },
+    {
+      "epoch": 1.6028416779431665,
+      "grad_norm": 0.5111741679227052,
+      "learning_rate": 1.1542179523529218e-06,
+      "loss": 0.257,
+      "step": 2369
+    },
+    {
+      "epoch": 1.6035182679296347,
+      "grad_norm": 0.48940448385602414,
+      "learning_rate": 1.1504468233893408e-06,
+      "loss": 0.2432,
+      "step": 2370
+    },
+    {
+      "epoch": 1.6041948579161027,
+      "grad_norm": 0.4867619950943536,
+      "learning_rate": 1.146681064085231e-06,
+      "loss": 0.2435,
+      "step": 2371
+    },
+    {
+      "epoch": 1.6048714479025712,
+      "grad_norm": 0.49639618053509976,
+      "learning_rate": 1.1429206796933717e-06,
+      "loss": 0.2451,
+      "step": 2372
+    },
+    {
+      "epoch": 1.6055480378890392,
+      "grad_norm": 0.47386305266006007,
+      "learning_rate": 1.1391656754590397e-06,
+      "loss": 0.2403,
+      "step": 2373
+    },
+    {
+      "epoch": 1.6062246278755075,
+      "grad_norm": 0.4948633331427711,
+      "learning_rate": 1.1354160566200128e-06,
+      "loss": 0.2492,
+      "step": 2374
+    },
+    {
+      "epoch": 1.6069012178619757,
+      "grad_norm": 0.47397596832375877,
+      "learning_rate": 1.1316718284065536e-06,
+      "loss": 0.2449,
+      "step": 2375
+    },
+    {
+      "epoch": 1.6075778078484437,
+      "grad_norm": 0.5196050030140132,
+      "learning_rate": 1.1279329960414047e-06,
+      "loss": 0.2524,
+      "step": 2376
+    },
+    {
+      "epoch": 1.6082543978349122,
+      "grad_norm": 0.5010242752946684,
+      "learning_rate": 1.124199564739783e-06,
+      "loss": 0.251,
+      "step": 2377
+    },
+    {
+      "epoch": 1.6089309878213802,
+      "grad_norm": 0.4883832602239757,
+      "learning_rate": 1.1204715397093735e-06,
+      "loss": 0.2453,
+      "step": 2378
+    },
+    {
+      "epoch": 1.6096075778078485,
+      "grad_norm": 0.5123340053173335,
+      "learning_rate": 1.1167489261503194e-06,
+      "loss": 0.2428,
+      "step": 2379
+    },
+    {
+      "epoch": 1.6102841677943167,
+      "grad_norm": 0.4695714605402934,
+      "learning_rate": 1.113031729255214e-06,
+      "loss": 0.2376,
+      "step": 2380
+    },
+    {
+      "epoch": 1.6109607577807847,
+      "grad_norm": 0.5059061172740084,
+      "learning_rate": 1.1093199542090944e-06,
+      "loss": 0.2503,
+      "step": 2381
+    },
+    {
+      "epoch": 1.6116373477672532,
+      "grad_norm": 0.49414152805142053,
+      "learning_rate": 1.1056136061894386e-06,
+      "loss": 0.2449,
+      "step": 2382
+    },
+    {
+      "epoch": 1.6123139377537212,
+      "grad_norm": 0.4780129319472459,
+      "learning_rate": 1.1019126903661514e-06,
+      "loss": 0.2398,
+      "step": 2383
+    },
+    {
+      "epoch": 1.6129905277401895,
+      "grad_norm": 0.48441273918020705,
+      "learning_rate": 1.0982172119015594e-06,
+      "loss": 0.2457,
+      "step": 2384
+    },
+    {
+      "epoch": 1.6136671177266577,
+      "grad_norm": 0.4989957447608218,
+      "learning_rate": 1.0945271759504083e-06,
+      "loss": 0.2503,
+      "step": 2385
+    },
+    {
+      "epoch": 1.6143437077131257,
+      "grad_norm": 0.46957061362484553,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.2333,
+      "step": 2386
+    },
+    {
+      "epoch": 1.6150202976995942,
+      "grad_norm": 0.49016714062868816,
+      "learning_rate": 1.0871634521694412e-06,
+      "loss": 0.2478,
+      "step": 2387
+    },
+    {
+      "epoch": 1.6156968876860622,
+      "grad_norm": 0.5284880339902153,
+      "learning_rate": 1.0834897746111233e-06,
+      "loss": 0.2554,
+      "step": 2388
+    },
+    {
+      "epoch": 1.6163734776725305,
+      "grad_norm": 0.4891721046213172,
+      "learning_rate": 1.0798215601092354e-06,
+      "loss": 0.2518,
+      "step": 2389
+    },
+    {
+      "epoch": 1.6170500676589987,
+      "grad_norm": 0.5044285234144619,
+      "learning_rate": 1.0761588137804896e-06,
+      "loss": 0.2376,
+      "step": 2390
+    },
+    {
+      "epoch": 1.6177266576454667,
+      "grad_norm": 0.5067031139653726,
+      "learning_rate": 1.0725015407339718e-06,
+      "loss": 0.2589,
+      "step": 2391
+    },
+    {
+      "epoch": 1.618403247631935,
+      "grad_norm": 0.5048120655034204,
+      "learning_rate": 1.0688497460711345e-06,
+      "loss": 0.2579,
+      "step": 2392
+    },
+    {
+      "epoch": 1.6190798376184032,
+      "grad_norm": 0.494371253409698,
+      "learning_rate": 1.0652034348857904e-06,
+      "loss": 0.2507,
+      "step": 2393
+    },
+    {
+      "epoch": 1.6197564276048715,
+      "grad_norm": 0.5175808884788287,
+      "learning_rate": 1.0615626122640988e-06,
+      "loss": 0.2552,
+      "step": 2394
+    },
+    {
+      "epoch": 1.6204330175913397,
+      "grad_norm": 0.4999004297893581,
+      "learning_rate": 1.0579272832845644e-06,
+      "loss": 0.2482,
+      "step": 2395
+    },
+    {
+      "epoch": 1.6211096075778078,
+      "grad_norm": 0.5006389305106579,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.2471,
+      "step": 2396
+    },
+    {
+      "epoch": 1.621786197564276,
+      "grad_norm": 0.4915720757093139,
+      "learning_rate": 1.050673126527675e-06,
+      "loss": 0.2492,
+      "step": 2397
+    },
+    {
+      "epoch": 1.6224627875507442,
+      "grad_norm": 0.4849484996805809,
+      "learning_rate": 1.0470543088689855e-06,
+      "loss": 0.2452,
+      "step": 2398
+    },
+    {
+      "epoch": 1.6231393775372125,
+      "grad_norm": 0.49755632509167586,
+      "learning_rate": 1.043441005089777e-06,
+      "loss": 0.2583,
+      "step": 2399
+    },
+    {
+      "epoch": 1.6238159675236807,
+      "grad_norm": 0.5055647045326658,
+      "learning_rate": 1.0398332202301708e-06,
+      "loss": 0.2481,
+      "step": 2400
+    },
+    {
+      "epoch": 1.6244925575101488,
+      "grad_norm": 0.5380287458610089,
+      "learning_rate": 1.0362309593225877e-06,
+      "loss": 0.2739,
+      "step": 2401
+    },
+    {
+      "epoch": 1.625169147496617,
+      "grad_norm": 0.5018299746581325,
+      "learning_rate": 1.0326342273917432e-06,
+      "loss": 0.2509,
+      "step": 2402
+    },
+    {
+      "epoch": 1.6258457374830853,
+      "grad_norm": 0.5135676725702435,
+      "learning_rate": 1.0290430294546449e-06,
+      "loss": 0.2628,
+      "step": 2403
+    },
+    {
+      "epoch": 1.6265223274695535,
+      "grad_norm": 0.5235451416743837,
+      "learning_rate": 1.0254573705205751e-06,
+      "loss": 0.252,
+      "step": 2404
+    },
+    {
+      "epoch": 1.6271989174560217,
+      "grad_norm": 0.5064124965277469,
+      "learning_rate": 1.0218772555910955e-06,
+      "loss": 0.2394,
+      "step": 2405
+    },
+    {
+      "epoch": 1.6278755074424898,
+      "grad_norm": 0.4902739458605531,
+      "learning_rate": 1.0183026896600284e-06,
+      "loss": 0.2453,
+      "step": 2406
+    },
+    {
+      "epoch": 1.628552097428958,
+      "grad_norm": 0.4871472595301355,
+      "learning_rate": 1.0147336777134636e-06,
+      "loss": 0.2489,
+      "step": 2407
+    },
+    {
+      "epoch": 1.6292286874154263,
+      "grad_norm": 0.5072738467967105,
+      "learning_rate": 1.0111702247297372e-06,
+      "loss": 0.251,
+      "step": 2408
+    },
+    {
+      "epoch": 1.6299052774018945,
+      "grad_norm": 0.5071314608524446,
+      "learning_rate": 1.0076123356794342e-06,
+      "loss": 0.2561,
+      "step": 2409
+    },
+    {
+      "epoch": 1.6305818673883627,
+      "grad_norm": 0.4977658604052108,
+      "learning_rate": 1.0040600155253766e-06,
+      "loss": 0.2519,
+      "step": 2410
+    },
+    {
+      "epoch": 1.6312584573748308,
+      "grad_norm": 0.5024488259516454,
+      "learning_rate": 1.000513269222621e-06,
+      "loss": 0.2462,
+      "step": 2411
+    },
+    {
+      "epoch": 1.631935047361299,
+      "grad_norm": 0.4807574151807117,
+      "learning_rate": 9.969721017184492e-07,
+      "loss": 0.2347,
+      "step": 2412
+    },
+    {
+      "epoch": 1.6326116373477673,
+      "grad_norm": 0.49264026338664585,
+      "learning_rate": 9.934365179523576e-07,
+      "loss": 0.2478,
+      "step": 2413
+    },
+    {
+      "epoch": 1.6332882273342353,
+      "grad_norm": 0.49967405680337057,
+      "learning_rate": 9.899065228560596e-07,
+      "loss": 0.2535,
+      "step": 2414
+    },
+    {
+      "epoch": 1.6339648173207038,
+      "grad_norm": 0.49021830304951536,
+      "learning_rate": 9.86382121353468e-07,
+      "loss": 0.2525,
+      "step": 2415
+    },
+    {
+      "epoch": 1.6346414073071718,
+      "grad_norm": 0.4885734192519705,
+      "learning_rate": 9.82863318360695e-07,
+      "loss": 0.2434,
+      "step": 2416
+    },
+    {
+      "epoch": 1.63531799729364,
+      "grad_norm": 0.48161478246330014,
+      "learning_rate": 9.793501187860432e-07,
+      "loss": 0.2491,
+      "step": 2417
+    },
+    {
+      "epoch": 1.6359945872801083,
+      "grad_norm": 0.5044935076642106,
+      "learning_rate": 9.758425275299998e-07,
+      "loss": 0.2503,
+      "step": 2418
+    },
+    {
+      "epoch": 1.6366711772665763,
+      "grad_norm": 0.5065803997041026,
+      "learning_rate": 9.723405494852317e-07,
+      "loss": 0.2468,
+      "step": 2419
+    },
+    {
+      "epoch": 1.6373477672530448,
+      "grad_norm": 0.5157551971496974,
+      "learning_rate": 9.688441895365708e-07,
+      "loss": 0.2549,
+      "step": 2420
+    },
+    {
+      "epoch": 1.6380243572395128,
+      "grad_norm": 0.5008427188109924,
+      "learning_rate": 9.653534525610137e-07,
+      "loss": 0.2454,
+      "step": 2421
+    },
+    {
+      "epoch": 1.638700947225981,
+      "grad_norm": 0.5153115652667144,
+      "learning_rate": 9.618683434277176e-07,
+      "loss": 0.2625,
+      "step": 2422
+    },
+    {
+      "epoch": 1.6393775372124493,
+      "grad_norm": 0.49830946533407394,
+      "learning_rate": 9.58388866997985e-07,
+      "loss": 0.2437,
+      "step": 2423
+    },
+    {
+      "epoch": 1.6400541271989173,
+      "grad_norm": 0.49521987058224964,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.2499,
+      "step": 2424
+    },
+    {
+      "epoch": 1.6407307171853858,
+      "grad_norm": 0.514096069280857,
+      "learning_rate": 9.514468316551362e-07,
+      "loss": 0.2543,
+      "step": 2425
+    },
+    {
+      "epoch": 1.6414073071718538,
+      "grad_norm": 0.5089851593538368,
+      "learning_rate": 9.479842824253182e-07,
+      "loss": 0.2546,
+      "step": 2426
+    },
+    {
+      "epoch": 1.642083897158322,
+      "grad_norm": 0.475558915194692,
+      "learning_rate": 9.445273852656445e-07,
+      "loss": 0.2368,
+      "step": 2427
+    },
+    {
+      "epoch": 1.6427604871447903,
+      "grad_norm": 0.5042750527316645,
+      "learning_rate": 9.410761449980654e-07,
+      "loss": 0.2493,
+      "step": 2428
+    },
+    {
+      "epoch": 1.6434370771312583,
+      "grad_norm": 0.4987155239123901,
+      "learning_rate": 9.376305664366447e-07,
+      "loss": 0.2458,
+      "step": 2429
+    },
+    {
+      "epoch": 1.6441136671177268,
+      "grad_norm": 0.5028395843709255,
+      "learning_rate": 9.341906543875451e-07,
+      "loss": 0.2476,
+      "step": 2430
+    },
+    {
+      "epoch": 1.6447902571041948,
+      "grad_norm": 0.5028082302304264,
+      "learning_rate": 9.307564136490255e-07,
+      "loss": 0.2453,
+      "step": 2431
+    },
+    {
+      "epoch": 1.645466847090663,
+      "grad_norm": 0.5065850051947194,
+      "learning_rate": 9.273278490114357e-07,
+      "loss": 0.2478,
+      "step": 2432
+    },
+    {
+      "epoch": 1.6461434370771313,
+      "grad_norm": 0.49687980585256003,
+      "learning_rate": 9.23904965257209e-07,
+      "loss": 0.242,
+      "step": 2433
+    },
+    {
+      "epoch": 1.6468200270635993,
+      "grad_norm": 0.5238676870203345,
+      "learning_rate": 9.204877671608515e-07,
+      "loss": 0.2474,
+      "step": 2434
+    },
+    {
+      "epoch": 1.6474966170500678,
+      "grad_norm": 0.5040088849568352,
+      "learning_rate": 9.170762594889376e-07,
+      "loss": 0.2489,
+      "step": 2435
+    },
+    {
+      "epoch": 1.6481732070365358,
+      "grad_norm": 0.510748759756837,
+      "learning_rate": 9.136704470001101e-07,
+      "loss": 0.2537,
+      "step": 2436
+    },
+    {
+      "epoch": 1.648849797023004,
+      "grad_norm": 0.46982726093443317,
+      "learning_rate": 9.102703344450625e-07,
+      "loss": 0.2341,
+      "step": 2437
+    },
+    {
+      "epoch": 1.6495263870094723,
+      "grad_norm": 0.5078937140831491,
+      "learning_rate": 9.068759265665384e-07,
+      "loss": 0.2609,
+      "step": 2438
+    },
+    {
+      "epoch": 1.6502029769959403,
+      "grad_norm": 0.4869522232539573,
+      "learning_rate": 9.034872280993245e-07,
+      "loss": 0.2491,
+      "step": 2439
+    },
+    {
+      "epoch": 1.6508795669824088,
+      "grad_norm": 0.5013437144296375,
+      "learning_rate": 9.001042437702468e-07,
+      "loss": 0.2467,
+      "step": 2440
+    },
+    {
+      "epoch": 1.6515561569688768,
+      "grad_norm": 0.49414363323241023,
+      "learning_rate": 8.967269782981558e-07,
+      "loss": 0.2571,
+      "step": 2441
+    },
+    {
+      "epoch": 1.652232746955345,
+      "grad_norm": 0.5031503318040169,
+      "learning_rate": 8.933554363939256e-07,
+      "loss": 0.2368,
+      "step": 2442
+    },
+    {
+      "epoch": 1.6529093369418133,
+      "grad_norm": 0.5219986753712459,
+      "learning_rate": 8.899896227604509e-07,
+      "loss": 0.2598,
+      "step": 2443
+    },
+    {
+      "epoch": 1.6535859269282813,
+      "grad_norm": 0.4936135466227145,
+      "learning_rate": 8.866295420926319e-07,
+      "loss": 0.2541,
+      "step": 2444
+    },
+    {
+      "epoch": 1.6542625169147498,
+      "grad_norm": 0.5105210877293747,
+      "learning_rate": 8.832751990773714e-07,
+      "loss": 0.2415,
+      "step": 2445
+    },
+    {
+      "epoch": 1.6549391069012178,
+      "grad_norm": 0.5034050141147028,
+      "learning_rate": 8.799265983935734e-07,
+      "loss": 0.2552,
+      "step": 2446
+    },
+    {
+      "epoch": 1.655615696887686,
+      "grad_norm": 0.5127269611431764,
+      "learning_rate": 8.765837447121295e-07,
+      "loss": 0.2506,
+      "step": 2447
+    },
+    {
+      "epoch": 1.6562922868741543,
+      "grad_norm": 0.5010002602552861,
+      "learning_rate": 8.732466426959135e-07,
+      "loss": 0.2519,
+      "step": 2448
+    },
+    {
+      "epoch": 1.6569688768606223,
+      "grad_norm": 0.5142722363966798,
+      "learning_rate": 8.699152969997787e-07,
+      "loss": 0.2573,
+      "step": 2449
+    },
+    {
+      "epoch": 1.6576454668470908,
+      "grad_norm": 0.4933599804566414,
+      "learning_rate": 8.665897122705463e-07,
+      "loss": 0.2389,
+      "step": 2450
+    },
+    {
+      "epoch": 1.6583220568335588,
+      "grad_norm": 0.5176087983232446,
+      "learning_rate": 8.632698931470063e-07,
+      "loss": 0.2508,
+      "step": 2451
+    },
+    {
+      "epoch": 1.658998646820027,
+      "grad_norm": 0.5146222041320239,
+      "learning_rate": 8.599558442598998e-07,
+      "loss": 0.2654,
+      "step": 2452
+    },
+    {
+      "epoch": 1.6596752368064953,
+      "grad_norm": 0.5246285564566604,
+      "learning_rate": 8.566475702319249e-07,
+      "loss": 0.2526,
+      "step": 2453
+    },
+    {
+      "epoch": 1.6603518267929633,
+      "grad_norm": 0.4936639233001636,
+      "learning_rate": 8.53345075677724e-07,
+      "loss": 0.2488,
+      "step": 2454
+    },
+    {
+      "epoch": 1.6610284167794318,
+      "grad_norm": 0.4918123438484583,
+      "learning_rate": 8.500483652038738e-07,
+      "loss": 0.2411,
+      "step": 2455
+    },
+    {
+      "epoch": 1.6617050067658998,
+      "grad_norm": 0.4842622717953389,
+      "learning_rate": 8.46757443408886e-07,
+      "loss": 0.2541,
+      "step": 2456
+    },
+    {
+      "epoch": 1.662381596752368,
+      "grad_norm": 0.5016582490727748,
+      "learning_rate": 8.434723148831959e-07,
+      "loss": 0.2519,
+      "step": 2457
+    },
+    {
+      "epoch": 1.6630581867388363,
+      "grad_norm": 0.4829418742144453,
+      "learning_rate": 8.401929842091616e-07,
+      "loss": 0.2447,
+      "step": 2458
+    },
+    {
+      "epoch": 1.6637347767253043,
+      "grad_norm": 0.4988726027461327,
+      "learning_rate": 8.369194559610483e-07,
+      "loss": 0.2624,
+      "step": 2459
+    },
+    {
+      "epoch": 1.6644113667117728,
+      "grad_norm": 0.5043748237030066,
+      "learning_rate": 8.336517347050327e-07,
+      "loss": 0.2517,
+      "step": 2460
+    },
+    {
+      "epoch": 1.6650879566982408,
+      "grad_norm": 0.5037548604978515,
+      "learning_rate": 8.303898249991899e-07,
+      "loss": 0.2597,
+      "step": 2461
+    },
+    {
+      "epoch": 1.665764546684709,
+      "grad_norm": 0.49718388332803587,
+      "learning_rate": 8.271337313934869e-07,
+      "loss": 0.2448,
+      "step": 2462
+    },
+    {
+      "epoch": 1.6664411366711773,
+      "grad_norm": 0.4753648904013494,
+      "learning_rate": 8.238834584297794e-07,
+      "loss": 0.2394,
+      "step": 2463
+    },
+    {
+      "epoch": 1.6671177266576453,
+      "grad_norm": 0.5092388099672063,
+      "learning_rate": 8.206390106418028e-07,
+      "loss": 0.2594,
+      "step": 2464
+    },
+    {
+      "epoch": 1.6677943166441138,
+      "grad_norm": 0.5004530533015025,
+      "learning_rate": 8.174003925551699e-07,
+      "loss": 0.2427,
+      "step": 2465
+    },
+    {
+      "epoch": 1.6684709066305818,
+      "grad_norm": 0.48651309291681316,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.2446,
+      "step": 2466
+    },
+    {
+      "epoch": 1.66914749661705,
+      "grad_norm": 0.4965680394358086,
+      "learning_rate": 8.109406635477085e-07,
+      "loss": 0.2508,
+      "step": 2467
+    },
+    {
+      "epoch": 1.6698240866035183,
+      "grad_norm": 0.5054343888075985,
+      "learning_rate": 8.077195616374184e-07,
+      "loss": 0.2487,
+      "step": 2468
+    },
+    {
+      "epoch": 1.6705006765899864,
+      "grad_norm": 0.5145208631440018,
+      "learning_rate": 8.045043074495357e-07,
+      "loss": 0.2626,
+      "step": 2469
+    },
+    {
+      "epoch": 1.6711772665764548,
+      "grad_norm": 0.5099412554416078,
+      "learning_rate": 8.012949054689484e-07,
+      "loss": 0.2615,
+      "step": 2470
+    },
+    {
+      "epoch": 1.6718538565629228,
+      "grad_norm": 0.5095603321437306,
+      "learning_rate": 7.980913601723811e-07,
+      "loss": 0.258,
+      "step": 2471
+    },
+    {
+      "epoch": 1.672530446549391,
+      "grad_norm": 0.49120875963761756,
+      "learning_rate": 7.948936760283937e-07,
+      "loss": 0.2406,
+      "step": 2472
+    },
+    {
+      "epoch": 1.6732070365358593,
+      "grad_norm": 0.49480578754036597,
+      "learning_rate": 7.917018574973646e-07,
+      "loss": 0.2548,
+      "step": 2473
+    },
+    {
+      "epoch": 1.6738836265223274,
+      "grad_norm": 0.48128535313442167,
+      "learning_rate": 7.885159090314959e-07,
+      "loss": 0.2406,
+      "step": 2474
+    },
+    {
+      "epoch": 1.6745602165087958,
+      "grad_norm": 0.4890615220845373,
+      "learning_rate": 7.853358350747969e-07,
+      "loss": 0.2585,
+      "step": 2475
+    },
+    {
+      "epoch": 1.6752368064952639,
+      "grad_norm": 0.5087705681779945,
+      "learning_rate": 7.821616400630866e-07,
+      "loss": 0.2476,
+      "step": 2476
+    },
+    {
+      "epoch": 1.675913396481732,
+      "grad_norm": 0.4926549741668967,
+      "learning_rate": 7.789933284239814e-07,
+      "loss": 0.246,
+      "step": 2477
+    },
+    {
+      "epoch": 1.6765899864682003,
+      "grad_norm": 0.5009164163093539,
+      "learning_rate": 7.758309045768908e-07,
+      "loss": 0.2539,
+      "step": 2478
+    },
+    {
+      "epoch": 1.6772665764546684,
+      "grad_norm": 0.48548410503179923,
+      "learning_rate": 7.726743729330111e-07,
+      "loss": 0.2355,
+      "step": 2479
+    },
+    {
+      "epoch": 1.6779431664411368,
+      "grad_norm": 0.4857357512971529,
+      "learning_rate": 7.695237378953224e-07,
+      "loss": 0.244,
+      "step": 2480
+    },
+    {
+      "epoch": 1.6786197564276049,
+      "grad_norm": 0.5005519620676971,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.2593,
+      "step": 2481
+    },
+    {
+      "epoch": 1.679296346414073,
+      "grad_norm": 0.49533876313379044,
+      "learning_rate": 7.632401752093016e-07,
+      "loss": 0.2419,
+      "step": 2482
+    },
+    {
+      "epoch": 1.6799729364005414,
+      "grad_norm": 0.5047302931709279,
+      "learning_rate": 7.601072563257772e-07,
+      "loss": 0.2542,
+      "step": 2483
+    },
+    {
+      "epoch": 1.6806495263870094,
+      "grad_norm": 0.5078864996400942,
+      "learning_rate": 7.569802515780455e-07,
+      "loss": 0.2476,
+      "step": 2484
+    },
+    {
+      "epoch": 1.6813261163734776,
+      "grad_norm": 0.4794025734773629,
+      "learning_rate": 7.538591653278993e-07,
+      "loss": 0.2395,
+      "step": 2485
+    },
+    {
+      "epoch": 1.6820027063599459,
+      "grad_norm": 0.49820452086048356,
+      "learning_rate": 7.507440019288742e-07,
+      "loss": 0.2511,
+      "step": 2486
+    },
+    {
+      "epoch": 1.682679296346414,
+      "grad_norm": 0.4980936507340043,
+      "learning_rate": 7.476347657262456e-07,
+      "loss": 0.2533,
+      "step": 2487
+    },
+    {
+      "epoch": 1.6833558863328824,
+      "grad_norm": 0.48330054569147807,
+      "learning_rate": 7.44531461057022e-07,
+      "loss": 0.2443,
+      "step": 2488
+    },
+    {
+      "epoch": 1.6840324763193504,
+      "grad_norm": 0.49758757232788675,
+      "learning_rate": 7.414340922499358e-07,
+      "loss": 0.2445,
+      "step": 2489
+    },
+    {
+      "epoch": 1.6847090663058186,
+      "grad_norm": 0.5251112666099677,
+      "learning_rate": 7.383426636254392e-07,
+      "loss": 0.2513,
+      "step": 2490
+    },
+    {
+      "epoch": 1.6853856562922869,
+      "grad_norm": 0.5035350955303267,
+      "learning_rate": 7.352571794957025e-07,
+      "loss": 0.2431,
+      "step": 2491
+    },
+    {
+      "epoch": 1.6860622462787551,
+      "grad_norm": 0.5251426833942452,
+      "learning_rate": 7.321776441646001e-07,
+      "loss": 0.2418,
+      "step": 2492
+    },
+    {
+      "epoch": 1.6867388362652234,
+      "grad_norm": 0.499865593790622,
+      "learning_rate": 7.29104061927709e-07,
+      "loss": 0.242,
+      "step": 2493
+    },
+    {
+      "epoch": 1.6874154262516914,
+      "grad_norm": 0.4971024793816913,
+      "learning_rate": 7.260364370723044e-07,
+      "loss": 0.2499,
+      "step": 2494
+    },
+    {
+      "epoch": 1.6880920162381596,
+      "grad_norm": 0.48936768945734566,
+      "learning_rate": 7.229747738773524e-07,
+      "loss": 0.2391,
+      "step": 2495
+    },
+    {
+      "epoch": 1.6887686062246279,
+      "grad_norm": 0.4771888965268285,
+      "learning_rate": 7.199190766135001e-07,
+      "loss": 0.2383,
+      "step": 2496
+    },
+    {
+      "epoch": 1.6894451962110961,
+      "grad_norm": 0.48764255871245255,
+      "learning_rate": 7.168693495430729e-07,
+      "loss": 0.2465,
+      "step": 2497
+    },
+    {
+      "epoch": 1.6901217861975644,
+      "grad_norm": 0.49308333592196496,
+      "learning_rate": 7.138255969200724e-07,
+      "loss": 0.2406,
+      "step": 2498
+    },
+    {
+      "epoch": 1.6907983761840324,
+      "grad_norm": 0.5047880129902625,
+      "learning_rate": 7.107878229901632e-07,
+      "loss": 0.2349,
+      "step": 2499
+    },
+    {
+      "epoch": 1.6914749661705006,
+      "grad_norm": 0.48094328214688,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.2498,
+      "step": 2500
+    },
+    {
+      "epoch": 1.6921515561569689,
+      "grad_norm": 0.5112123464006819,
+      "learning_rate": 7.047302281505735e-07,
+      "loss": 0.245,
+      "step": 2501
+    },
+    {
+      "epoch": 1.6928281461434371,
+      "grad_norm": 0.4954954321500193,
+      "learning_rate": 7.017104156905058e-07,
+      "loss": 0.249,
+      "step": 2502
+    },
+    {
+      "epoch": 1.6935047361299054,
+      "grad_norm": 0.4968169100181153,
+      "learning_rate": 6.986965988227356e-07,
+      "loss": 0.2399,
+      "step": 2503
+    },
+    {
+      "epoch": 1.6941813261163734,
+      "grad_norm": 0.4863728785333933,
+      "learning_rate": 6.95688781751172e-07,
+      "loss": 0.2496,
+      "step": 2504
+    },
+    {
+      "epoch": 1.6948579161028416,
+      "grad_norm": 0.5183695555225226,
+      "learning_rate": 6.926869686713566e-07,
+      "loss": 0.2565,
+      "step": 2505
+    },
+    {
+      "epoch": 1.69553450608931,
+      "grad_norm": 0.4787144787799259,
+      "learning_rate": 6.896911637704534e-07,
+      "loss": 0.2509,
+      "step": 2506
+    },
+    {
+      "epoch": 1.696211096075778,
+      "grad_norm": 0.5317227491716305,
+      "learning_rate": 6.867013712272447e-07,
+      "loss": 0.2523,
+      "step": 2507
+    },
+    {
+      "epoch": 1.6968876860622464,
+      "grad_norm": 0.4821105978931383,
+      "learning_rate": 6.837175952121305e-07,
+      "loss": 0.2455,
+      "step": 2508
+    },
+    {
+      "epoch": 1.6975642760487144,
+      "grad_norm": 0.5077885489476185,
+      "learning_rate": 6.807398398871162e-07,
+      "loss": 0.2525,
+      "step": 2509
+    },
+    {
+      "epoch": 1.6982408660351827,
+      "grad_norm": 0.4967645038229651,
+      "learning_rate": 6.777681094058087e-07,
+      "loss": 0.2555,
+      "step": 2510
+    },
+    {
+      "epoch": 1.698917456021651,
+      "grad_norm": 0.4813344889981423,
+      "learning_rate": 6.748024079134102e-07,
+      "loss": 0.2405,
+      "step": 2511
+    },
+    {
+      "epoch": 1.699594046008119,
+      "grad_norm": 0.5169035526166853,
+      "learning_rate": 6.718427395467165e-07,
+      "loss": 0.2529,
+      "step": 2512
+    },
+    {
+      "epoch": 1.7002706359945874,
+      "grad_norm": 0.5283261689786055,
+      "learning_rate": 6.688891084341042e-07,
+      "loss": 0.2593,
+      "step": 2513
+    },
+    {
+      "epoch": 1.7009472259810554,
+      "grad_norm": 0.5154525130563101,
+      "learning_rate": 6.659415186955298e-07,
+      "loss": 0.26,
+      "step": 2514
+    },
+    {
+      "epoch": 1.7016238159675237,
+      "grad_norm": 0.48494833098684836,
+      "learning_rate": 6.629999744425236e-07,
+      "loss": 0.2452,
+      "step": 2515
+    },
+    {
+      "epoch": 1.702300405953992,
+      "grad_norm": 0.5023810002888528,
+      "learning_rate": 6.600644797781847e-07,
+      "loss": 0.243,
+      "step": 2516
+    },
+    {
+      "epoch": 1.70297699594046,
+      "grad_norm": 0.48874102955935883,
+      "learning_rate": 6.571350387971703e-07,
+      "loss": 0.2491,
+      "step": 2517
+    },
+    {
+      "epoch": 1.7036535859269284,
+      "grad_norm": 0.4989820775693793,
+      "learning_rate": 6.542116555856953e-07,
+      "loss": 0.2575,
+      "step": 2518
+    },
+    {
+      "epoch": 1.7043301759133964,
+      "grad_norm": 0.5119196744804391,
+      "learning_rate": 6.512943342215234e-07,
+      "loss": 0.2492,
+      "step": 2519
+    },
+    {
+      "epoch": 1.7050067658998647,
+      "grad_norm": 0.5203032153784719,
+      "learning_rate": 6.483830787739659e-07,
+      "loss": 0.2622,
+      "step": 2520
+    },
+    {
+      "epoch": 1.705683355886333,
+      "grad_norm": 0.48182979040033663,
+      "learning_rate": 6.454778933038681e-07,
+      "loss": 0.2461,
+      "step": 2521
+    },
+    {
+      "epoch": 1.706359945872801,
+      "grad_norm": 0.49338521585363354,
+      "learning_rate": 6.425787818636131e-07,
+      "loss": 0.246,
+      "step": 2522
+    },
+    {
+      "epoch": 1.7070365358592694,
+      "grad_norm": 0.49441379994768353,
+      "learning_rate": 6.396857484971103e-07,
+      "loss": 0.2449,
+      "step": 2523
+    },
+    {
+      "epoch": 1.7077131258457374,
+      "grad_norm": 0.4784945346310669,
+      "learning_rate": 6.367987972397887e-07,
+      "loss": 0.2346,
+      "step": 2524
+    },
+    {
+      "epoch": 1.7083897158322057,
+      "grad_norm": 0.4953389056295345,
+      "learning_rate": 6.339179321185956e-07,
+      "loss": 0.2362,
+      "step": 2525
+    },
+    {
+      "epoch": 1.709066305818674,
+      "grad_norm": 0.4885969723122,
+      "learning_rate": 6.310431571519865e-07,
+      "loss": 0.2456,
+      "step": 2526
+    },
+    {
+      "epoch": 1.709742895805142,
+      "grad_norm": 0.5228087055107105,
+      "learning_rate": 6.281744763499259e-07,
+      "loss": 0.2489,
+      "step": 2527
+    },
+    {
+      "epoch": 1.7104194857916104,
+      "grad_norm": 0.49209216154375046,
+      "learning_rate": 6.25311893713873e-07,
+      "loss": 0.2497,
+      "step": 2528
+    },
+    {
+      "epoch": 1.7110960757780784,
+      "grad_norm": 0.48193703018573447,
+      "learning_rate": 6.224554132367861e-07,
+      "loss": 0.2459,
+      "step": 2529
+    },
+    {
+      "epoch": 1.7117726657645467,
+      "grad_norm": 0.49551649813254206,
+      "learning_rate": 6.196050389031061e-07,
+      "loss": 0.2518,
+      "step": 2530
+    },
+    {
+      "epoch": 1.712449255751015,
+      "grad_norm": 0.4982770025635861,
+      "learning_rate": 6.16760774688761e-07,
+      "loss": 0.253,
+      "step": 2531
+    },
+    {
+      "epoch": 1.713125845737483,
+      "grad_norm": 0.503666410800622,
+      "learning_rate": 6.139226245611535e-07,
+      "loss": 0.2508,
+      "step": 2532
+    },
+    {
+      "epoch": 1.7138024357239514,
+      "grad_norm": 0.48191066993461507,
+      "learning_rate": 6.110905924791577e-07,
+      "loss": 0.2441,
+      "step": 2533
+    },
+    {
+      "epoch": 1.7144790257104194,
+      "grad_norm": 0.5006000365639539,
+      "learning_rate": 6.082646823931165e-07,
+      "loss": 0.2533,
+      "step": 2534
+    },
+    {
+      "epoch": 1.7151556156968877,
+      "grad_norm": 0.5063681194364289,
+      "learning_rate": 6.054448982448291e-07,
+      "loss": 0.2461,
+      "step": 2535
+    },
+    {
+      "epoch": 1.715832205683356,
+      "grad_norm": 0.49403575586654497,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.2513,
+      "step": 2536
+    },
+    {
+      "epoch": 1.716508795669824,
+      "grad_norm": 0.4950327880662957,
+      "learning_rate": 5.998237234859977e-07,
+      "loss": 0.2532,
+      "step": 2537
+    },
+    {
+      "epoch": 1.7171853856562924,
+      "grad_norm": 0.48915728121438024,
+      "learning_rate": 5.9702234071631e-07,
+      "loss": 0.2508,
+      "step": 2538
+    },
+    {
+      "epoch": 1.7178619756427604,
+      "grad_norm": 0.4978367726823817,
+      "learning_rate": 5.942270995660799e-07,
+      "loss": 0.2482,
+      "step": 2539
+    },
+    {
+      "epoch": 1.7185385656292287,
+      "grad_norm": 0.49177895189725873,
+      "learning_rate": 5.914380039343281e-07,
+      "loss": 0.2508,
+      "step": 2540
+    },
+    {
+      "epoch": 1.719215155615697,
+      "grad_norm": 0.5059340348312885,
+      "learning_rate": 5.886550577115069e-07,
+      "loss": 0.254,
+      "step": 2541
+    },
+    {
+      "epoch": 1.719891745602165,
+      "grad_norm": 0.4994919802592363,
+      "learning_rate": 5.858782647794864e-07,
+      "loss": 0.2532,
+      "step": 2542
+    },
+    {
+      "epoch": 1.7205683355886334,
+      "grad_norm": 0.5044237094741355,
+      "learning_rate": 5.831076290115572e-07,
+      "loss": 0.2511,
+      "step": 2543
+    },
+    {
+      "epoch": 1.7212449255751014,
+      "grad_norm": 0.4800426372087279,
+      "learning_rate": 5.803431542724192e-07,
+      "loss": 0.2464,
+      "step": 2544
+    },
+    {
+      "epoch": 1.7219215155615697,
+      "grad_norm": 0.5016643542940342,
+      "learning_rate": 5.775848444181797e-07,
+      "loss": 0.2464,
+      "step": 2545
+    },
+    {
+      "epoch": 1.722598105548038,
+      "grad_norm": 0.48963099377957736,
+      "learning_rate": 5.748327032963464e-07,
+      "loss": 0.2433,
+      "step": 2546
+    },
+    {
+      "epoch": 1.723274695534506,
+      "grad_norm": 0.4829195221007469,
+      "learning_rate": 5.720867347458209e-07,
+      "loss": 0.2543,
+      "step": 2547
+    },
+    {
+      "epoch": 1.7239512855209744,
+      "grad_norm": 0.5043868897602848,
+      "learning_rate": 5.693469425968962e-07,
+      "loss": 0.2383,
+      "step": 2548
+    },
+    {
+      "epoch": 1.7246278755074425,
+      "grad_norm": 0.5113203517527847,
+      "learning_rate": 5.666133306712501e-07,
+      "loss": 0.2506,
+      "step": 2549
+    },
+    {
+      "epoch": 1.7253044654939107,
+      "grad_norm": 0.48539128392899517,
+      "learning_rate": 5.63885902781941e-07,
+      "loss": 0.2429,
+      "step": 2550
+    },
+    {
+      "epoch": 1.725981055480379,
+      "grad_norm": 0.4891669662608059,
+      "learning_rate": 5.611646627333977e-07,
+      "loss": 0.2348,
+      "step": 2551
+    },
+    {
+      "epoch": 1.726657645466847,
+      "grad_norm": 0.5147168643752245,
+      "learning_rate": 5.584496143214213e-07,
+      "loss": 0.2501,
+      "step": 2552
+    },
+    {
+      "epoch": 1.7273342354533154,
+      "grad_norm": 0.5153657636599415,
+      "learning_rate": 5.557407613331745e-07,
+      "loss": 0.2605,
+      "step": 2553
+    },
+    {
+      "epoch": 1.7280108254397835,
+      "grad_norm": 0.4982208249247303,
+      "learning_rate": 5.530381075471775e-07,
+      "loss": 0.2511,
+      "step": 2554
+    },
+    {
+      "epoch": 1.7286874154262517,
+      "grad_norm": 0.4988767894746393,
+      "learning_rate": 5.503416567333042e-07,
+      "loss": 0.2447,
+      "step": 2555
+    },
+    {
+      "epoch": 1.72936400541272,
+      "grad_norm": 0.5065694053574913,
+      "learning_rate": 5.476514126527771e-07,
+      "loss": 0.2587,
+      "step": 2556
+    },
+    {
+      "epoch": 1.730040595399188,
+      "grad_norm": 0.49172319300875605,
+      "learning_rate": 5.449673790581611e-07,
+      "loss": 0.2423,
+      "step": 2557
+    },
+    {
+      "epoch": 1.7307171853856564,
+      "grad_norm": 0.5025269841446781,
+      "learning_rate": 5.422895596933559e-07,
+      "loss": 0.2535,
+      "step": 2558
+    },
+    {
+      "epoch": 1.7313937753721245,
+      "grad_norm": 0.4999436445555782,
+      "learning_rate": 5.39617958293594e-07,
+      "loss": 0.2568,
+      "step": 2559
+    },
+    {
+      "epoch": 1.7320703653585927,
+      "grad_norm": 0.47835394166119605,
+      "learning_rate": 5.369525785854368e-07,
+      "loss": 0.2416,
+      "step": 2560
+    },
+    {
+      "epoch": 1.732746955345061,
+      "grad_norm": 0.5160405367152339,
+      "learning_rate": 5.342934242867648e-07,
+      "loss": 0.2482,
+      "step": 2561
+    },
+    {
+      "epoch": 1.733423545331529,
+      "grad_norm": 0.5030884526173469,
+      "learning_rate": 5.316404991067747e-07,
+      "loss": 0.2369,
+      "step": 2562
+    },
+    {
+      "epoch": 1.7341001353179974,
+      "grad_norm": 0.5123145622692458,
+      "learning_rate": 5.289938067459755e-07,
+      "loss": 0.2463,
+      "step": 2563
+    },
+    {
+      "epoch": 1.7347767253044655,
+      "grad_norm": 0.4996737944943131,
+      "learning_rate": 5.263533508961827e-07,
+      "loss": 0.2454,
+      "step": 2564
+    },
+    {
+      "epoch": 1.7354533152909337,
+      "grad_norm": 0.5090274452524801,
+      "learning_rate": 5.237191352405118e-07,
+      "loss": 0.2558,
+      "step": 2565
+    },
+    {
+      "epoch": 1.736129905277402,
+      "grad_norm": 0.5119498337674819,
+      "learning_rate": 5.210911634533722e-07,
+      "loss": 0.2595,
+      "step": 2566
+    },
+    {
+      "epoch": 1.73680649526387,
+      "grad_norm": 0.5092249381927475,
+      "learning_rate": 5.184694392004675e-07,
+      "loss": 0.2457,
+      "step": 2567
+    },
+    {
+      "epoch": 1.7374830852503385,
+      "grad_norm": 0.49512913975701583,
+      "learning_rate": 5.15853966138784e-07,
+      "loss": 0.2412,
+      "step": 2568
+    },
+    {
+      "epoch": 1.7381596752368065,
+      "grad_norm": 0.47280729892869494,
+      "learning_rate": 5.132447479165887e-07,
+      "loss": 0.2355,
+      "step": 2569
+    },
+    {
+      "epoch": 1.7388362652232747,
+      "grad_norm": 0.47792775199886883,
+      "learning_rate": 5.106417881734244e-07,
+      "loss": 0.2327,
+      "step": 2570
+    },
+    {
+      "epoch": 1.739512855209743,
+      "grad_norm": 0.49342004445888255,
+      "learning_rate": 5.080450905401057e-07,
+      "loss": 0.2534,
+      "step": 2571
+    },
+    {
+      "epoch": 1.740189445196211,
+      "grad_norm": 0.5112526554866993,
+      "learning_rate": 5.054546586387093e-07,
+      "loss": 0.2554,
+      "step": 2572
+    },
+    {
+      "epoch": 1.7408660351826795,
+      "grad_norm": 0.5181706229679123,
+      "learning_rate": 5.028704960825731e-07,
+      "loss": 0.2551,
+      "step": 2573
+    },
+    {
+      "epoch": 1.7415426251691475,
+      "grad_norm": 0.5056700827350163,
+      "learning_rate": 5.002926064762908e-07,
+      "loss": 0.257,
+      "step": 2574
+    },
+    {
+      "epoch": 1.7422192151556157,
+      "grad_norm": 0.5230942625986709,
+      "learning_rate": 4.977209934157057e-07,
+      "loss": 0.2579,
+      "step": 2575
+    },
+    {
+      "epoch": 1.742895805142084,
+      "grad_norm": 0.5043069504815765,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.2487,
+      "step": 2576
+    },
+    {
+      "epoch": 1.743572395128552,
+      "grad_norm": 0.48542988740114595,
+      "learning_rate": 4.925966112712166e-07,
+      "loss": 0.2498,
+      "step": 2577
+    },
+    {
+      "epoch": 1.7442489851150202,
+      "grad_norm": 0.4936539828007256,
+      "learning_rate": 4.900438493352056e-07,
+      "loss": 0.2468,
+      "step": 2578
+    },
+    {
+      "epoch": 1.7449255751014885,
+      "grad_norm": 0.47533287028041477,
+      "learning_rate": 4.874973782406634e-07,
+      "loss": 0.2409,
+      "step": 2579
+    },
+    {
+      "epoch": 1.7456021650879567,
+      "grad_norm": 0.477759714550434,
+      "learning_rate": 4.849572015396081e-07,
+      "loss": 0.246,
+      "step": 2580
+    },
+    {
+      "epoch": 1.746278755074425,
+      "grad_norm": 0.5065162403694692,
+      "learning_rate": 4.824233227752789e-07,
+      "loss": 0.2414,
+      "step": 2581
+    },
+    {
+      "epoch": 1.746955345060893,
+      "grad_norm": 0.4935984953119808,
+      "learning_rate": 4.798957454821285e-07,
+      "loss": 0.2434,
+      "step": 2582
+    },
+    {
+      "epoch": 1.7476319350473613,
+      "grad_norm": 0.4882334164016001,
+      "learning_rate": 4.773744731858199e-07,
+      "loss": 0.2619,
+      "step": 2583
+    },
+    {
+      "epoch": 1.7483085250338295,
+      "grad_norm": 0.49149443777152624,
+      "learning_rate": 4.748595094032221e-07,
+      "loss": 0.2492,
+      "step": 2584
+    },
+    {
+      "epoch": 1.7489851150202977,
+      "grad_norm": 0.5180554215612304,
+      "learning_rate": 4.723508576424063e-07,
+      "loss": 0.2611,
+      "step": 2585
+    },
+    {
+      "epoch": 1.749661705006766,
+      "grad_norm": 0.49718579940206586,
+      "learning_rate": 4.698485214026349e-07,
+      "loss": 0.2495,
+      "step": 2586
+    },
+    {
+      "epoch": 1.750338294993234,
+      "grad_norm": 0.4960124842497286,
+      "learning_rate": 4.6735250417436485e-07,
+      "loss": 0.2436,
+      "step": 2587
+    },
+    {
+      "epoch": 1.7510148849797023,
+      "grad_norm": 0.5089388097395485,
+      "learning_rate": 4.6486280943923547e-07,
+      "loss": 0.2645,
+      "step": 2588
+    },
+    {
+      "epoch": 1.7516914749661705,
+      "grad_norm": 0.5032759953229424,
+      "learning_rate": 4.623794406700705e-07,
+      "loss": 0.2382,
+      "step": 2589
+    },
+    {
+      "epoch": 1.7523680649526387,
+      "grad_norm": 0.5090795357423945,
+      "learning_rate": 4.5990240133086617e-07,
+      "loss": 0.2503,
+      "step": 2590
+    },
+    {
+      "epoch": 1.753044654939107,
+      "grad_norm": 0.4928645839018117,
+      "learning_rate": 4.574316948767932e-07,
+      "loss": 0.2473,
+      "step": 2591
+    },
+    {
+      "epoch": 1.753721244925575,
+      "grad_norm": 0.4940373242313454,
+      "learning_rate": 4.549673247541875e-07,
+      "loss": 0.2499,
+      "step": 2592
+    },
+    {
+      "epoch": 1.7543978349120433,
+      "grad_norm": 0.4915143058701345,
+      "learning_rate": 4.5250929440054547e-07,
+      "loss": 0.2528,
+      "step": 2593
+    },
+    {
+      "epoch": 1.7550744248985115,
+      "grad_norm": 0.4784631177282373,
+      "learning_rate": 4.5005760724452173e-07,
+      "loss": 0.2352,
+      "step": 2594
+    },
+    {
+      "epoch": 1.7557510148849798,
+      "grad_norm": 0.5074878045527873,
+      "learning_rate": 4.4761226670592074e-07,
+      "loss": 0.2486,
+      "step": 2595
+    },
+    {
+      "epoch": 1.756427604871448,
+      "grad_norm": 0.48433392533627534,
+      "learning_rate": 4.4517327619569784e-07,
+      "loss": 0.2416,
+      "step": 2596
+    },
+    {
+      "epoch": 1.757104194857916,
+      "grad_norm": 0.5087253270337099,
+      "learning_rate": 4.427406391159467e-07,
+      "loss": 0.2537,
+      "step": 2597
+    },
+    {
+      "epoch": 1.7577807848443843,
+      "grad_norm": 0.4986915591027291,
+      "learning_rate": 4.403143588599029e-07,
+      "loss": 0.2482,
+      "step": 2598
+    },
+    {
+      "epoch": 1.7584573748308525,
+      "grad_norm": 0.49592688791275147,
+      "learning_rate": 4.3789443881193107e-07,
+      "loss": 0.2535,
+      "step": 2599
+    },
+    {
+      "epoch": 1.7591339648173205,
+      "grad_norm": 0.4931463360118562,
+      "learning_rate": 4.3548088234752814e-07,
+      "loss": 0.2445,
+      "step": 2600
+    },
+    {
+      "epoch": 1.759810554803789,
+      "grad_norm": 0.5099905050767143,
+      "learning_rate": 4.330736928333107e-07,
+      "loss": 0.245,
+      "step": 2601
+    },
+    {
+      "epoch": 1.760487144790257,
+      "grad_norm": 0.4888837496641677,
+      "learning_rate": 4.3067287362701606e-07,
+      "loss": 0.2411,
+      "step": 2602
+    },
+    {
+      "epoch": 1.7611637347767253,
+      "grad_norm": 0.5061789207845503,
+      "learning_rate": 4.282784280774965e-07,
+      "loss": 0.2429,
+      "step": 2603
+    },
+    {
+      "epoch": 1.7618403247631935,
+      "grad_norm": 0.5111960410936642,
+      "learning_rate": 4.258903595247116e-07,
+      "loss": 0.2543,
+      "step": 2604
+    },
+    {
+      "epoch": 1.7625169147496615,
+      "grad_norm": 0.5092738652728804,
+      "learning_rate": 4.235086712997288e-07,
+      "loss": 0.2575,
+      "step": 2605
+    },
+    {
+      "epoch": 1.76319350473613,
+      "grad_norm": 0.4927693068819805,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.2364,
+      "step": 2606
+    },
+    {
+      "epoch": 1.763870094722598,
+      "grad_norm": 0.48548366546094956,
+      "learning_rate": 4.1876444911292613e-07,
+      "loss": 0.2374,
+      "step": 2607
+    },
+    {
+      "epoch": 1.7645466847090663,
+      "grad_norm": 0.5039545367269136,
+      "learning_rate": 4.164019217687215e-07,
+      "loss": 0.2515,
+      "step": 2608
+    },
+    {
+      "epoch": 1.7652232746955345,
+      "grad_norm": 0.4789685895701932,
+      "learning_rate": 4.140457879875365e-07,
+      "loss": 0.2486,
+      "step": 2609
+    },
+    {
+      "epoch": 1.7658998646820026,
+      "grad_norm": 0.49975464497301136,
+      "learning_rate": 4.1169605105589315e-07,
+      "loss": 0.2629,
+      "step": 2610
+    },
+    {
+      "epoch": 1.766576454668471,
+      "grad_norm": 0.5003792902766644,
+      "learning_rate": 4.0935271425138757e-07,
+      "loss": 0.2369,
+      "step": 2611
+    },
+    {
+      "epoch": 1.767253044654939,
+      "grad_norm": 0.5027995800389407,
+      "learning_rate": 4.070157808426928e-07,
+      "loss": 0.2422,
+      "step": 2612
+    },
+    {
+      "epoch": 1.7679296346414073,
+      "grad_norm": 0.47639821084560513,
+      "learning_rate": 4.046852540895446e-07,
+      "loss": 0.248,
+      "step": 2613
+    },
+    {
+      "epoch": 1.7686062246278755,
+      "grad_norm": 0.49400779843307363,
+      "learning_rate": 4.0236113724274716e-07,
+      "loss": 0.24,
+      "step": 2614
+    },
+    {
+      "epoch": 1.7692828146143436,
+      "grad_norm": 0.49685766911825385,
+      "learning_rate": 4.0004343354416053e-07,
+      "loss": 0.2427,
+      "step": 2615
+    },
+    {
+      "epoch": 1.769959404600812,
+      "grad_norm": 0.4976554409845932,
+      "learning_rate": 3.9773214622669974e-07,
+      "loss": 0.2491,
+      "step": 2616
+    },
+    {
+      "epoch": 1.77063599458728,
+      "grad_norm": 0.5075964857656241,
+      "learning_rate": 3.954272785143298e-07,
+      "loss": 0.2616,
+      "step": 2617
+    },
+    {
+      "epoch": 1.7713125845737483,
+      "grad_norm": 0.4953333249201334,
+      "learning_rate": 3.9312883362206177e-07,
+      "loss": 0.2422,
+      "step": 2618
+    },
+    {
+      "epoch": 1.7719891745602165,
+      "grad_norm": 0.4781949268881124,
+      "learning_rate": 3.9083681475594827e-07,
+      "loss": 0.245,
+      "step": 2619
+    },
+    {
+      "epoch": 1.7726657645466846,
+      "grad_norm": 0.4865749262698487,
+      "learning_rate": 3.885512251130763e-07,
+      "loss": 0.2426,
+      "step": 2620
+    },
+    {
+      "epoch": 1.773342354533153,
+      "grad_norm": 0.4903377355868679,
+      "learning_rate": 3.862720678815668e-07,
+      "loss": 0.2555,
+      "step": 2621
+    },
+    {
+      "epoch": 1.774018944519621,
+      "grad_norm": 0.4924516685784948,
+      "learning_rate": 3.839993462405678e-07,
+      "loss": 0.2468,
+      "step": 2622
+    },
+    {
+      "epoch": 1.7746955345060893,
+      "grad_norm": 0.5060253448447951,
+      "learning_rate": 3.8173306336024784e-07,
+      "loss": 0.2555,
+      "step": 2623
+    },
+    {
+      "epoch": 1.7753721244925575,
+      "grad_norm": 0.4928368633091723,
+      "learning_rate": 3.794732224017994e-07,
+      "loss": 0.2382,
+      "step": 2624
+    },
+    {
+      "epoch": 1.7760487144790256,
+      "grad_norm": 0.507028764459079,
+      "learning_rate": 3.772198265174243e-07,
+      "loss": 0.2581,
+      "step": 2625
+    },
+    {
+      "epoch": 1.776725304465494,
+      "grad_norm": 0.4960872388519864,
+      "learning_rate": 3.7497287885033763e-07,
+      "loss": 0.2439,
+      "step": 2626
+    },
+    {
+      "epoch": 1.777401894451962,
+      "grad_norm": 0.493953466949582,
+      "learning_rate": 3.7273238253475783e-07,
+      "loss": 0.2397,
+      "step": 2627
+    },
+    {
+      "epoch": 1.7780784844384303,
+      "grad_norm": 0.4988248538423828,
+      "learning_rate": 3.7049834069590507e-07,
+      "loss": 0.2483,
+      "step": 2628
+    },
+    {
+      "epoch": 1.7787550744248986,
+      "grad_norm": 0.490056081957018,
+      "learning_rate": 3.682707564499982e-07,
+      "loss": 0.2446,
+      "step": 2629
+    },
+    {
+      "epoch": 1.7794316644113666,
+      "grad_norm": 0.5009485844462291,
+      "learning_rate": 3.6604963290424453e-07,
+      "loss": 0.2583,
+      "step": 2630
+    },
+    {
+      "epoch": 1.780108254397835,
+      "grad_norm": 0.492487677761043,
+      "learning_rate": 3.638349731568436e-07,
+      "loss": 0.2552,
+      "step": 2631
+    },
+    {
+      "epoch": 1.780784844384303,
+      "grad_norm": 0.5012971521706225,
+      "learning_rate": 3.6162678029697696e-07,
+      "loss": 0.2406,
+      "step": 2632
+    },
+    {
+      "epoch": 1.7814614343707713,
+      "grad_norm": 0.48722885090227797,
+      "learning_rate": 3.5942505740480583e-07,
+      "loss": 0.2408,
+      "step": 2633
+    },
+    {
+      "epoch": 1.7821380243572396,
+      "grad_norm": 0.5118606813357625,
+      "learning_rate": 3.572298075514652e-07,
+      "loss": 0.2516,
+      "step": 2634
+    },
+    {
+      "epoch": 1.7828146143437076,
+      "grad_norm": 0.5012800386556492,
+      "learning_rate": 3.5504103379906174e-07,
+      "loss": 0.2464,
+      "step": 2635
+    },
+    {
+      "epoch": 1.783491204330176,
+      "grad_norm": 0.4976126741550675,
+      "learning_rate": 3.528587392006716e-07,
+      "loss": 0.2517,
+      "step": 2636
+    },
+    {
+      "epoch": 1.784167794316644,
+      "grad_norm": 0.4808798943082307,
+      "learning_rate": 3.506829268003281e-07,
+      "loss": 0.2349,
+      "step": 2637
+    },
+    {
+      "epoch": 1.7848443843031123,
+      "grad_norm": 0.5132669160127712,
+      "learning_rate": 3.48513599633028e-07,
+      "loss": 0.2531,
+      "step": 2638
+    },
+    {
+      "epoch": 1.7855209742895806,
+      "grad_norm": 0.4892428123954744,
+      "learning_rate": 3.463507607247174e-07,
+      "loss": 0.2509,
+      "step": 2639
+    },
+    {
+      "epoch": 1.7861975642760486,
+      "grad_norm": 0.5161541310249723,
+      "learning_rate": 3.4419441309229587e-07,
+      "loss": 0.2598,
+      "step": 2640
+    },
+    {
+      "epoch": 1.786874154262517,
+      "grad_norm": 0.5109755045348078,
+      "learning_rate": 3.420445597436056e-07,
+      "loss": 0.2511,
+      "step": 2641
+    },
+    {
+      "epoch": 1.787550744248985,
+      "grad_norm": 0.509143877480822,
+      "learning_rate": 3.3990120367743074e-07,
+      "loss": 0.2552,
+      "step": 2642
+    },
+    {
+      "epoch": 1.7882273342354533,
+      "grad_norm": 0.47765295871841007,
+      "learning_rate": 3.3776434788349355e-07,
+      "loss": 0.2468,
+      "step": 2643
+    },
+    {
+      "epoch": 1.7889039242219216,
+      "grad_norm": 0.48169883950771575,
+      "learning_rate": 3.356339953424481e-07,
+      "loss": 0.2361,
+      "step": 2644
+    },
+    {
+      "epoch": 1.7895805142083896,
+      "grad_norm": 0.5014801910507384,
+      "learning_rate": 3.3351014902587795e-07,
+      "loss": 0.2415,
+      "step": 2645
+    },
+    {
+      "epoch": 1.790257104194858,
+      "grad_norm": 0.5045891790102259,
+      "learning_rate": 3.313928118962906e-07,
+      "loss": 0.2484,
+      "step": 2646
+    },
+    {
+      "epoch": 1.790933694181326,
+      "grad_norm": 0.5096888147194478,
+      "learning_rate": 3.2928198690711564e-07,
+      "loss": 0.2552,
+      "step": 2647
+    },
+    {
+      "epoch": 1.7916102841677943,
+      "grad_norm": 0.5114230754430588,
+      "learning_rate": 3.271776770026963e-07,
+      "loss": 0.2592,
+      "step": 2648
+    },
+    {
+      "epoch": 1.7922868741542626,
+      "grad_norm": 0.48712827158262195,
+      "learning_rate": 3.2507988511828905e-07,
+      "loss": 0.2384,
+      "step": 2649
+    },
+    {
+      "epoch": 1.7929634641407306,
+      "grad_norm": 0.5056066660628573,
+      "learning_rate": 3.229886141800609e-07,
+      "loss": 0.2498,
+      "step": 2650
+    },
+    {
+      "epoch": 1.793640054127199,
+      "grad_norm": 0.49781402075014725,
+      "learning_rate": 3.2090386710507906e-07,
+      "loss": 0.2505,
+      "step": 2651
+    },
+    {
+      "epoch": 1.794316644113667,
+      "grad_norm": 0.4759857040213447,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.2412,
+      "step": 2652
+    },
+    {
+      "epoch": 1.7949932341001353,
+      "grad_norm": 0.4994029417051804,
+      "learning_rate": 3.167539561676303e-07,
+      "loss": 0.2602,
+      "step": 2653
+    },
+    {
+      "epoch": 1.7956698240866036,
+      "grad_norm": 0.5065909241562929,
+      "learning_rate": 3.146887980937852e-07,
+      "loss": 0.2385,
+      "step": 2654
+    },
+    {
+      "epoch": 1.7963464140730716,
+      "grad_norm": 0.49696658706554037,
+      "learning_rate": 3.126301754604233e-07,
+      "loss": 0.2373,
+      "step": 2655
+    },
+    {
+      "epoch": 1.79702300405954,
+      "grad_norm": 0.502959425168816,
+      "learning_rate": 3.105780911390738e-07,
+      "loss": 0.2589,
+      "step": 2656
+    },
+    {
+      "epoch": 1.797699594046008,
+      "grad_norm": 0.4879717453909071,
+      "learning_rate": 3.0853254799214414e-07,
+      "loss": 0.2424,
+      "step": 2657
+    },
+    {
+      "epoch": 1.7983761840324763,
+      "grad_norm": 0.4894633467801478,
+      "learning_rate": 3.0649354887291927e-07,
+      "loss": 0.2544,
+      "step": 2658
+    },
+    {
+      "epoch": 1.7990527740189446,
+      "grad_norm": 0.4954359958616379,
+      "learning_rate": 3.044610966255568e-07,
+      "loss": 0.2447,
+      "step": 2659
+    },
+    {
+      "epoch": 1.7997293640054126,
+      "grad_norm": 0.4971426600408802,
+      "learning_rate": 3.0243519408507894e-07,
+      "loss": 0.2488,
+      "step": 2660
+    },
+    {
+      "epoch": 1.800405953991881,
+      "grad_norm": 0.47506196666423933,
+      "learning_rate": 3.0041584407737577e-07,
+      "loss": 0.2457,
+      "step": 2661
+    },
+    {
+      "epoch": 1.801082543978349,
+      "grad_norm": 0.49487892578786374,
+      "learning_rate": 2.984030494191942e-07,
+      "loss": 0.2507,
+      "step": 2662
+    },
+    {
+      "epoch": 1.8017591339648173,
+      "grad_norm": 0.5021523352655755,
+      "learning_rate": 2.9639681291813883e-07,
+      "loss": 0.2551,
+      "step": 2663
+    },
+    {
+      "epoch": 1.8024357239512856,
+      "grad_norm": 0.490382800199004,
+      "learning_rate": 2.9439713737266504e-07,
+      "loss": 0.2379,
+      "step": 2664
+    },
+    {
+      "epoch": 1.8031123139377536,
+      "grad_norm": 0.4798350442944653,
+      "learning_rate": 2.9240402557207895e-07,
+      "loss": 0.2424,
+      "step": 2665
+    },
+    {
+      "epoch": 1.803788903924222,
+      "grad_norm": 0.4896281841449204,
+      "learning_rate": 2.904174802965293e-07,
+      "loss": 0.2502,
+      "step": 2666
+    },
+    {
+      "epoch": 1.80446549391069,
+      "grad_norm": 0.48550964025679416,
+      "learning_rate": 2.884375043170062e-07,
+      "loss": 0.2407,
+      "step": 2667
+    },
+    {
+      "epoch": 1.8051420838971584,
+      "grad_norm": 0.5040783204701812,
+      "learning_rate": 2.864641003953339e-07,
+      "loss": 0.25,
+      "step": 2668
+    },
+    {
+      "epoch": 1.8058186738836266,
+      "grad_norm": 0.49469896217059045,
+      "learning_rate": 2.844972712841737e-07,
+      "loss": 0.248,
+      "step": 2669
+    },
+    {
+      "epoch": 1.8064952638700946,
+      "grad_norm": 0.5069188098439675,
+      "learning_rate": 2.8253701972701275e-07,
+      "loss": 0.2492,
+      "step": 2670
+    },
+    {
+      "epoch": 1.8071718538565629,
+      "grad_norm": 0.5030653215601004,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.2498,
+      "step": 2671
+    },
+    {
+      "epoch": 1.8078484438430311,
+      "grad_norm": 0.47775744688484895,
+      "learning_rate": 2.7863626020275867e-07,
+      "loss": 0.2354,
+      "step": 2672
+    },
+    {
+      "epoch": 1.8085250338294994,
+      "grad_norm": 0.4846183827406555,
+      "learning_rate": 2.766957576767537e-07,
+      "loss": 0.2431,
+      "step": 2673
+    },
+    {
+      "epoch": 1.8092016238159676,
+      "grad_norm": 0.4889130966147435,
+      "learning_rate": 2.7476184358691206e-07,
+      "loss": 0.2555,
+      "step": 2674
+    },
+    {
+      "epoch": 1.8098782138024356,
+      "grad_norm": 0.506211370217711,
+      "learning_rate": 2.728345206308086e-07,
+      "loss": 0.2475,
+      "step": 2675
+    },
+    {
+      "epoch": 1.8105548037889039,
+      "grad_norm": 0.5270520266161699,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.2586,
+      "step": 2676
+    },
+    {
+      "epoch": 1.8112313937753721,
+      "grad_norm": 0.5039959961387973,
+      "learning_rate": 2.689996588641497e-07,
+      "loss": 0.2405,
+      "step": 2677
+    },
+    {
+      "epoch": 1.8119079837618404,
+      "grad_norm": 0.504803789636751,
+      "learning_rate": 2.670921254027592e-07,
+      "loss": 0.2607,
+      "step": 2678
+    },
+    {
+      "epoch": 1.8125845737483086,
+      "grad_norm": 0.4980640162100982,
+      "learning_rate": 2.651911937734336e-07,
+      "loss": 0.2427,
+      "step": 2679
+    },
+    {
+      "epoch": 1.8132611637347766,
+      "grad_norm": 0.46541760935138293,
+      "learning_rate": 2.6329686662774247e-07,
+      "loss": 0.2476,
+      "step": 2680
+    },
+    {
+      "epoch": 1.8139377537212449,
+      "grad_norm": 0.4936611722532536,
+      "learning_rate": 2.6140914660804205e-07,
+      "loss": 0.2419,
+      "step": 2681
+    },
+    {
+      "epoch": 1.8146143437077131,
+      "grad_norm": 0.48554306473308545,
+      "learning_rate": 2.595280363474717e-07,
+      "loss": 0.2479,
+      "step": 2682
+    },
+    {
+      "epoch": 1.8152909336941814,
+      "grad_norm": 0.4704882874821065,
+      "learning_rate": 2.57653538469953e-07,
+      "loss": 0.2366,
+      "step": 2683
+    },
+    {
+      "epoch": 1.8159675236806496,
+      "grad_norm": 0.5102687489259601,
+      "learning_rate": 2.5578565559018276e-07,
+      "loss": 0.2421,
+      "step": 2684
+    },
+    {
+      "epoch": 1.8166441136671176,
+      "grad_norm": 0.49716599676048695,
+      "learning_rate": 2.539243903136296e-07,
+      "loss": 0.2483,
+      "step": 2685
+    },
+    {
+      "epoch": 1.817320703653586,
+      "grad_norm": 0.5102531611975376,
+      "learning_rate": 2.520697452365345e-07,
+      "loss": 0.2578,
+      "step": 2686
+    },
+    {
+      "epoch": 1.8179972936400541,
+      "grad_norm": 0.5240007418895813,
+      "learning_rate": 2.502217229459014e-07,
+      "loss": 0.2486,
+      "step": 2687
+    },
+    {
+      "epoch": 1.8186738836265224,
+      "grad_norm": 0.4975360323588128,
+      "learning_rate": 2.483803260194978e-07,
+      "loss": 0.2508,
+      "step": 2688
+    },
+    {
+      "epoch": 1.8193504736129906,
+      "grad_norm": 0.4843469472499859,
+      "learning_rate": 2.465455570258479e-07,
+      "loss": 0.2461,
+      "step": 2689
+    },
+    {
+      "epoch": 1.8200270635994586,
+      "grad_norm": 0.4959957990822598,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.2563,
+      "step": 2690
+    },
+    {
+      "epoch": 1.820703653585927,
+      "grad_norm": 0.49859796706174087,
+      "learning_rate": 2.4289591306468244e-07,
+      "loss": 0.2432,
+      "step": 2691
+    },
+    {
+      "epoch": 1.8213802435723951,
+      "grad_norm": 0.4934904685531084,
+      "learning_rate": 2.4108104318797674e-07,
+      "loss": 0.2565,
+      "step": 2692
+    },
+    {
+      "epoch": 1.8220568335588632,
+      "grad_norm": 0.5010844274764045,
+      "learning_rate": 2.3927281142563874e-07,
+      "loss": 0.2462,
+      "step": 2693
+    },
+    {
+      "epoch": 1.8227334235453316,
+      "grad_norm": 0.4908152399661423,
+      "learning_rate": 2.3747122029993296e-07,
+      "loss": 0.2539,
+      "step": 2694
+    },
+    {
+      "epoch": 1.8234100135317997,
+      "grad_norm": 0.48812105999319,
+      "learning_rate": 2.3567627232385982e-07,
+      "loss": 0.2481,
+      "step": 2695
+    },
+    {
+      "epoch": 1.824086603518268,
+      "grad_norm": 0.490405678023693,
+      "learning_rate": 2.3388797000115427e-07,
+      "loss": 0.2431,
+      "step": 2696
+    },
+    {
+      "epoch": 1.8247631935047361,
+      "grad_norm": 0.5090069651228606,
+      "learning_rate": 2.321063158262793e-07,
+      "loss": 0.2505,
+      "step": 2697
+    },
+    {
+      "epoch": 1.8254397834912042,
+      "grad_norm": 0.4813882798289277,
+      "learning_rate": 2.3033131228442863e-07,
+      "loss": 0.2371,
+      "step": 2698
+    },
+    {
+      "epoch": 1.8261163734776726,
+      "grad_norm": 0.5083827777943416,
+      "learning_rate": 2.2856296185151505e-07,
+      "loss": 0.2461,
+      "step": 2699
+    },
+    {
+      "epoch": 1.8267929634641407,
+      "grad_norm": 0.5225168094963644,
+      "learning_rate": 2.2680126699417383e-07,
+      "loss": 0.2606,
+      "step": 2700
+    },
+    {
+      "epoch": 1.827469553450609,
+      "grad_norm": 0.5200703156569669,
+      "learning_rate": 2.2504623016975536e-07,
+      "loss": 0.2477,
+      "step": 2701
+    },
+    {
+      "epoch": 1.8281461434370772,
+      "grad_norm": 0.47886991647800065,
+      "learning_rate": 2.2329785382632253e-07,
+      "loss": 0.2485,
+      "step": 2702
+    },
+    {
+      "epoch": 1.8288227334235452,
+      "grad_norm": 0.47770704675343106,
+      "learning_rate": 2.2155614040264895e-07,
+      "loss": 0.235,
+      "step": 2703
+    },
+    {
+      "epoch": 1.8294993234100136,
+      "grad_norm": 0.4926125465978069,
+      "learning_rate": 2.198210923282118e-07,
+      "loss": 0.2535,
+      "step": 2704
+    },
+    {
+      "epoch": 1.8301759133964817,
+      "grad_norm": 0.5023266264988802,
+      "learning_rate": 2.180927120231946e-07,
+      "loss": 0.2498,
+      "step": 2705
+    },
+    {
+      "epoch": 1.83085250338295,
+      "grad_norm": 0.4922304713286355,
+      "learning_rate": 2.163710018984766e-07,
+      "loss": 0.2435,
+      "step": 2706
+    },
+    {
+      "epoch": 1.8315290933694182,
+      "grad_norm": 0.48105449165668895,
+      "learning_rate": 2.1465596435563508e-07,
+      "loss": 0.2451,
+      "step": 2707
+    },
+    {
+      "epoch": 1.8322056833558862,
+      "grad_norm": 0.502616462326218,
+      "learning_rate": 2.1294760178693918e-07,
+      "loss": 0.2386,
+      "step": 2708
+    },
+    {
+      "epoch": 1.8328822733423547,
+      "grad_norm": 0.49712701494655104,
+      "learning_rate": 2.1124591657534776e-07,
+      "loss": 0.2462,
+      "step": 2709
+    },
+    {
+      "epoch": 1.8335588633288227,
+      "grad_norm": 0.49625575726703736,
+      "learning_rate": 2.0955091109450488e-07,
+      "loss": 0.2474,
+      "step": 2710
+    },
+    {
+      "epoch": 1.834235453315291,
+      "grad_norm": 0.4908237640736268,
+      "learning_rate": 2.0786258770873647e-07,
+      "loss": 0.2362,
+      "step": 2711
+    },
+    {
+      "epoch": 1.8349120433017592,
+      "grad_norm": 0.5246091195752569,
+      "learning_rate": 2.061809487730504e-07,
+      "loss": 0.2655,
+      "step": 2712
+    },
+    {
+      "epoch": 1.8355886332882272,
+      "grad_norm": 0.5004289404804811,
+      "learning_rate": 2.0450599663312754e-07,
+      "loss": 0.2449,
+      "step": 2713
+    },
+    {
+      "epoch": 1.8362652232746957,
+      "grad_norm": 0.49358684477135223,
+      "learning_rate": 2.0283773362532455e-07,
+      "loss": 0.2447,
+      "step": 2714
+    },
+    {
+      "epoch": 1.8369418132611637,
+      "grad_norm": 0.4926168755075107,
+      "learning_rate": 2.0117616207666334e-07,
+      "loss": 0.2475,
+      "step": 2715
+    },
+    {
+      "epoch": 1.837618403247632,
+      "grad_norm": 0.49465614201614405,
+      "learning_rate": 1.9952128430483718e-07,
+      "loss": 0.2572,
+      "step": 2716
+    },
+    {
+      "epoch": 1.8382949932341002,
+      "grad_norm": 0.4799958441553541,
+      "learning_rate": 1.9787310261819793e-07,
+      "loss": 0.2469,
+      "step": 2717
+    },
+    {
+      "epoch": 1.8389715832205682,
+      "grad_norm": 0.486155447694846,
+      "learning_rate": 1.962316193157593e-07,
+      "loss": 0.2491,
+      "step": 2718
+    },
+    {
+      "epoch": 1.8396481732070367,
+      "grad_norm": 0.4891326331762417,
+      "learning_rate": 1.9459683668719197e-07,
+      "loss": 0.2526,
+      "step": 2719
+    },
+    {
+      "epoch": 1.8403247631935047,
+      "grad_norm": 0.5080157941452895,
+      "learning_rate": 1.9296875701281858e-07,
+      "loss": 0.245,
+      "step": 2720
+    },
+    {
+      "epoch": 1.841001353179973,
+      "grad_norm": 0.501965801974377,
+      "learning_rate": 1.9134738256361306e-07,
+      "loss": 0.2483,
+      "step": 2721
+    },
+    {
+      "epoch": 1.8416779431664412,
+      "grad_norm": 0.479945244116184,
+      "learning_rate": 1.8973271560119576e-07,
+      "loss": 0.2477,
+      "step": 2722
+    },
+    {
+      "epoch": 1.8423545331529092,
+      "grad_norm": 0.509587546245771,
+      "learning_rate": 1.8812475837783172e-07,
+      "loss": 0.2463,
+      "step": 2723
+    },
+    {
+      "epoch": 1.8430311231393777,
+      "grad_norm": 0.4878714911576229,
+      "learning_rate": 1.8652351313642568e-07,
+      "loss": 0.2487,
+      "step": 2724
+    },
+    {
+      "epoch": 1.8437077131258457,
+      "grad_norm": 0.49257026949445865,
+      "learning_rate": 1.849289821105199e-07,
+      "loss": 0.2422,
+      "step": 2725
+    },
+    {
+      "epoch": 1.844384303112314,
+      "grad_norm": 0.499526450903088,
+      "learning_rate": 1.8334116752429243e-07,
+      "loss": 0.2488,
+      "step": 2726
+    },
+    {
+      "epoch": 1.8450608930987822,
+      "grad_norm": 0.4899674432648164,
+      "learning_rate": 1.8176007159255106e-07,
+      "loss": 0.2482,
+      "step": 2727
+    },
+    {
+      "epoch": 1.8457374830852502,
+      "grad_norm": 0.5108340878184533,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.2452,
+      "step": 2728
+    },
+    {
+      "epoch": 1.8464140730717187,
+      "grad_norm": 0.4868916201148893,
+      "learning_rate": 1.7861804450490183e-07,
+      "loss": 0.2301,
+      "step": 2729
+    },
+    {
+      "epoch": 1.8470906630581867,
+      "grad_norm": 0.5068937736083843,
+      "learning_rate": 1.770571177317404e-07,
+      "loss": 0.242,
+      "step": 2730
+    },
+    {
+      "epoch": 1.847767253044655,
+      "grad_norm": 0.5001646973053091,
+      "learning_rate": 1.7550291837855226e-07,
+      "loss": 0.2455,
+      "step": 2731
+    },
+    {
+      "epoch": 1.8484438430311232,
+      "grad_norm": 0.4934030299893805,
+      "learning_rate": 1.7395544861325718e-07,
+      "loss": 0.2394,
+      "step": 2732
+    },
+    {
+      "epoch": 1.8491204330175912,
+      "grad_norm": 0.4884580703607249,
+      "learning_rate": 1.7241471059438685e-07,
+      "loss": 0.2465,
+      "step": 2733
+    },
+    {
+      "epoch": 1.8497970230040597,
+      "grad_norm": 0.5145172841785638,
+      "learning_rate": 1.7088070647108433e-07,
+      "loss": 0.2425,
+      "step": 2734
+    },
+    {
+      "epoch": 1.8504736129905277,
+      "grad_norm": 0.5166939709746506,
+      "learning_rate": 1.6935343838310027e-07,
+      "loss": 0.2491,
+      "step": 2735
+    },
+    {
+      "epoch": 1.851150202976996,
+      "grad_norm": 0.4692866705616545,
+      "learning_rate": 1.6783290846078714e-07,
+      "loss": 0.2432,
+      "step": 2736
+    },
+    {
+      "epoch": 1.8518267929634642,
+      "grad_norm": 0.5097420135846584,
+      "learning_rate": 1.6631911882509956e-07,
+      "loss": 0.2512,
+      "step": 2737
+    },
+    {
+      "epoch": 1.8525033829499322,
+      "grad_norm": 0.49595910561027234,
+      "learning_rate": 1.648120715875906e-07,
+      "loss": 0.2427,
+      "step": 2738
+    },
+    {
+      "epoch": 1.8531799729364007,
+      "grad_norm": 0.5033094921745446,
+      "learning_rate": 1.6331176885040878e-07,
+      "loss": 0.2469,
+      "step": 2739
+    },
+    {
+      "epoch": 1.8538565629228687,
+      "grad_norm": 0.5128335823879966,
+      "learning_rate": 1.618182127062934e-07,
+      "loss": 0.2448,
+      "step": 2740
+    },
+    {
+      "epoch": 1.854533152909337,
+      "grad_norm": 0.4806864824416147,
+      "learning_rate": 1.6033140523857405e-07,
+      "loss": 0.2472,
+      "step": 2741
+    },
+    {
+      "epoch": 1.8552097428958052,
+      "grad_norm": 0.5097849844751589,
+      "learning_rate": 1.588513485211679e-07,
+      "loss": 0.2553,
+      "step": 2742
+    },
+    {
+      "epoch": 1.8558863328822732,
+      "grad_norm": 0.5029410545052978,
+      "learning_rate": 1.5737804461857355e-07,
+      "loss": 0.2575,
+      "step": 2743
+    },
+    {
+      "epoch": 1.8565629228687417,
+      "grad_norm": 0.49229617019123123,
+      "learning_rate": 1.5591149558587037e-07,
+      "loss": 0.2457,
+      "step": 2744
+    },
+    {
+      "epoch": 1.8572395128552097,
+      "grad_norm": 0.4867425420007886,
+      "learning_rate": 1.544517034687182e-07,
+      "loss": 0.2409,
+      "step": 2745
+    },
+    {
+      "epoch": 1.857916102841678,
+      "grad_norm": 0.4820171393556296,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.2505,
+      "step": 2746
+    },
+    {
+      "epoch": 1.8585926928281462,
+      "grad_norm": 0.4727158127632365,
+      "learning_rate": 1.5155239811656562e-07,
+      "loss": 0.2311,
+      "step": 2747
+    },
+    {
+      "epoch": 1.8592692828146142,
+      "grad_norm": 0.4992916522667062,
+      "learning_rate": 1.5011288892574526e-07,
+      "loss": 0.25,
+      "step": 2748
+    },
+    {
+      "epoch": 1.8599458728010827,
+      "grad_norm": 0.4972877492360314,
+      "learning_rate": 1.4868014473882753e-07,
+      "loss": 0.243,
+      "step": 2749
+    },
+    {
+      "epoch": 1.8606224627875507,
+      "grad_norm": 0.48648399537327525,
+      "learning_rate": 1.4725416755431655e-07,
+      "loss": 0.2418,
+      "step": 2750
+    },
+    {
+      "epoch": 1.861299052774019,
+      "grad_norm": 0.4790302881421719,
+      "learning_rate": 1.4583495936127678e-07,
+      "loss": 0.2494,
+      "step": 2751
+    },
+    {
+      "epoch": 1.8619756427604872,
+      "grad_norm": 0.5126616283998403,
+      "learning_rate": 1.444225221393325e-07,
+      "loss": 0.2575,
+      "step": 2752
+    },
+    {
+      "epoch": 1.8626522327469552,
+      "grad_norm": 0.49430872102821793,
+      "learning_rate": 1.4301685785866215e-07,
+      "loss": 0.2429,
+      "step": 2753
+    },
+    {
+      "epoch": 1.8633288227334237,
+      "grad_norm": 0.4951039280804687,
+      "learning_rate": 1.4161796847999566e-07,
+      "loss": 0.236,
+      "step": 2754
+    },
+    {
+      "epoch": 1.8640054127198917,
+      "grad_norm": 0.4917364070345142,
+      "learning_rate": 1.4022585595461434e-07,
+      "loss": 0.2554,
+      "step": 2755
+    },
+    {
+      "epoch": 1.86468200270636,
+      "grad_norm": 0.48691521129311033,
+      "learning_rate": 1.388405222243472e-07,
+      "loss": 0.2427,
+      "step": 2756
+    },
+    {
+      "epoch": 1.8653585926928282,
+      "grad_norm": 0.4821844756021286,
+      "learning_rate": 1.3746196922156617e-07,
+      "loss": 0.2463,
+      "step": 2757
+    },
+    {
+      "epoch": 1.8660351826792962,
+      "grad_norm": 0.4992767316369565,
+      "learning_rate": 1.360901988691843e-07,
+      "loss": 0.2469,
+      "step": 2758
+    },
+    {
+      "epoch": 1.8667117726657647,
+      "grad_norm": 0.5115146063211774,
+      "learning_rate": 1.3472521308065645e-07,
+      "loss": 0.2389,
+      "step": 2759
+    },
+    {
+      "epoch": 1.8673883626522327,
+      "grad_norm": 0.4909862155309222,
+      "learning_rate": 1.333670137599713e-07,
+      "loss": 0.2533,
+      "step": 2760
+    },
+    {
+      "epoch": 1.868064952638701,
+      "grad_norm": 0.5146847521135632,
+      "learning_rate": 1.3201560280165117e-07,
+      "loss": 0.2438,
+      "step": 2761
+    },
+    {
+      "epoch": 1.8687415426251692,
+      "grad_norm": 0.4892129950279416,
+      "learning_rate": 1.3067098209075202e-07,
+      "loss": 0.2442,
+      "step": 2762
+    },
+    {
+      "epoch": 1.8694181326116373,
+      "grad_norm": 0.49593568579735425,
+      "learning_rate": 1.2933315350285525e-07,
+      "loss": 0.2427,
+      "step": 2763
+    },
+    {
+      "epoch": 1.8700947225981055,
+      "grad_norm": 0.46442452500076087,
+      "learning_rate": 1.280021189040709e-07,
+      "loss": 0.2376,
+      "step": 2764
+    },
+    {
+      "epoch": 1.8707713125845737,
+      "grad_norm": 0.5111218418902066,
+      "learning_rate": 1.2667788015102879e-07,
+      "loss": 0.2495,
+      "step": 2765
+    },
+    {
+      "epoch": 1.871447902571042,
+      "grad_norm": 0.4845883873102546,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.2538,
+      "step": 2766
+    },
+    {
+      "epoch": 1.8721244925575102,
+      "grad_norm": 0.5057642119055453,
+      "learning_rate": 1.2404979756130142e-07,
+      "loss": 0.2367,
+      "step": 2767
+    },
+    {
+      "epoch": 1.8728010825439783,
+      "grad_norm": 0.4745318914627002,
+      "learning_rate": 1.2274595739047267e-07,
+      "loss": 0.235,
+      "step": 2768
+    },
+    {
+      "epoch": 1.8734776725304465,
+      "grad_norm": 0.5042755201150829,
+      "learning_rate": 1.2144892039709367e-07,
+      "loss": 0.2623,
+      "step": 2769
+    },
+    {
+      "epoch": 1.8741542625169147,
+      "grad_norm": 0.5164282296272626,
+      "learning_rate": 1.2015868839037492e-07,
+      "loss": 0.2525,
+      "step": 2770
+    },
+    {
+      "epoch": 1.874830852503383,
+      "grad_norm": 0.49666901314919026,
+      "learning_rate": 1.1887526317003351e-07,
+      "loss": 0.2592,
+      "step": 2771
+    },
+    {
+      "epoch": 1.8755074424898512,
+      "grad_norm": 0.4818404086410922,
+      "learning_rate": 1.1759864652629072e-07,
+      "loss": 0.2485,
+      "step": 2772
+    },
+    {
+      "epoch": 1.8761840324763193,
+      "grad_norm": 0.5063021547567165,
+      "learning_rate": 1.1632884023987212e-07,
+      "loss": 0.2449,
+      "step": 2773
+    },
+    {
+      "epoch": 1.8768606224627875,
+      "grad_norm": 0.48757771748996764,
+      "learning_rate": 1.1506584608200366e-07,
+      "loss": 0.2517,
+      "step": 2774
+    },
+    {
+      "epoch": 1.8775372124492558,
+      "grad_norm": 0.4890840139384109,
+      "learning_rate": 1.1380966581440889e-07,
+      "loss": 0.2486,
+      "step": 2775
+    },
+    {
+      "epoch": 1.878213802435724,
+      "grad_norm": 0.5073895649811323,
+      "learning_rate": 1.1256030118930727e-07,
+      "loss": 0.2518,
+      "step": 2776
+    },
+    {
+      "epoch": 1.8788903924221922,
+      "grad_norm": 0.49119421420993836,
+      "learning_rate": 1.113177539494098e-07,
+      "loss": 0.2494,
+      "step": 2777
+    },
+    {
+      "epoch": 1.8795669824086603,
+      "grad_norm": 0.4880004301308363,
+      "learning_rate": 1.1008202582792005e-07,
+      "loss": 0.2545,
+      "step": 2778
+    },
+    {
+      "epoch": 1.8802435723951285,
+      "grad_norm": 0.5059925568810683,
+      "learning_rate": 1.0885311854852809e-07,
+      "loss": 0.2541,
+      "step": 2779
+    },
+    {
+      "epoch": 1.8809201623815968,
+      "grad_norm": 0.5065082162744158,
+      "learning_rate": 1.0763103382541052e-07,
+      "loss": 0.2547,
+      "step": 2780
+    },
+    {
+      "epoch": 1.881596752368065,
+      "grad_norm": 0.4803156673233187,
+      "learning_rate": 1.0641577336322761e-07,
+      "loss": 0.2453,
+      "step": 2781
+    },
+    {
+      "epoch": 1.8822733423545333,
+      "grad_norm": 0.49303144602334975,
+      "learning_rate": 1.0520733885712008e-07,
+      "loss": 0.2462,
+      "step": 2782
+    },
+    {
+      "epoch": 1.8829499323410013,
+      "grad_norm": 0.5173461069270066,
+      "learning_rate": 1.0400573199270736e-07,
+      "loss": 0.2535,
+      "step": 2783
+    },
+    {
+      "epoch": 1.8836265223274695,
+      "grad_norm": 0.49653050777845253,
+      "learning_rate": 1.0281095444608425e-07,
+      "loss": 0.2466,
+      "step": 2784
+    },
+    {
+      "epoch": 1.8843031123139378,
+      "grad_norm": 0.5078625828234871,
+      "learning_rate": 1.0162300788382263e-07,
+      "loss": 0.251,
+      "step": 2785
+    },
+    {
+      "epoch": 1.8849797023004058,
+      "grad_norm": 0.49720362322328415,
+      "learning_rate": 1.0044189396296144e-07,
+      "loss": 0.2516,
+      "step": 2786
+    },
+    {
+      "epoch": 1.8856562922868743,
+      "grad_norm": 0.5129915009917727,
+      "learning_rate": 9.926761433101284e-08,
+      "loss": 0.2449,
+      "step": 2787
+    },
+    {
+      "epoch": 1.8863328822733423,
+      "grad_norm": 0.5002767594924182,
+      "learning_rate": 9.810017062595322e-08,
+      "loss": 0.2453,
+      "step": 2788
+    },
+    {
+      "epoch": 1.8870094722598105,
+      "grad_norm": 0.5051838990632037,
+      "learning_rate": 9.693956447622555e-08,
+      "loss": 0.247,
+      "step": 2789
+    },
+    {
+      "epoch": 1.8876860622462788,
+      "grad_norm": 0.47713148782906956,
+      "learning_rate": 9.57857975007348e-08,
+      "loss": 0.2482,
+      "step": 2790
+    },
+    {
+      "epoch": 1.8883626522327468,
+      "grad_norm": 0.509568993992654,
+      "learning_rate": 9.46388713088453e-08,
+      "loss": 0.2565,
+      "step": 2791
+    },
+    {
+      "epoch": 1.8890392422192153,
+      "grad_norm": 0.511684481443198,
+      "learning_rate": 9.349878750038067e-08,
+      "loss": 0.2548,
+      "step": 2792
+    },
+    {
+      "epoch": 1.8897158322056833,
+      "grad_norm": 0.4848597112557326,
+      "learning_rate": 9.236554766561823e-08,
+      "loss": 0.2422,
+      "step": 2793
+    },
+    {
+      "epoch": 1.8903924221921515,
+      "grad_norm": 0.4903012955670563,
+      "learning_rate": 9.123915338529132e-08,
+      "loss": 0.2476,
+      "step": 2794
+    },
+    {
+      "epoch": 1.8910690121786198,
+      "grad_norm": 0.5012516945893771,
+      "learning_rate": 9.011960623058202e-08,
+      "loss": 0.2441,
+      "step": 2795
+    },
+    {
+      "epoch": 1.8917456021650878,
+      "grad_norm": 0.49341517981305094,
+      "learning_rate": 8.900690776312282e-08,
+      "loss": 0.2409,
+      "step": 2796
+    },
+    {
+      "epoch": 1.8924221921515563,
+      "grad_norm": 0.49034073135876355,
+      "learning_rate": 8.790105953499439e-08,
+      "loss": 0.2469,
+      "step": 2797
+    },
+    {
+      "epoch": 1.8930987821380243,
+      "grad_norm": 0.48293443233242467,
+      "learning_rate": 8.680206308871953e-08,
+      "loss": 0.2369,
+      "step": 2798
+    },
+    {
+      "epoch": 1.8937753721244925,
+      "grad_norm": 0.47676557556146815,
+      "learning_rate": 8.570991995726586e-08,
+      "loss": 0.2456,
+      "step": 2799
+    },
+    {
+      "epoch": 1.8944519621109608,
+      "grad_norm": 0.5111764442366136,
+      "learning_rate": 8.462463166403978e-08,
+      "loss": 0.2421,
+      "step": 2800
+    },
+    {
+      "epoch": 1.8951285520974288,
+      "grad_norm": 0.4925035570449802,
+      "learning_rate": 8.354619972288703e-08,
+      "loss": 0.2467,
+      "step": 2801
+    },
+    {
+      "epoch": 1.8958051420838973,
+      "grad_norm": 0.48889890214709114,
+      "learning_rate": 8.247462563808816e-08,
+      "loss": 0.2449,
+      "step": 2802
+    },
+    {
+      "epoch": 1.8964817320703653,
+      "grad_norm": 0.48658675066090856,
+      "learning_rate": 8.14099109043609e-08,
+      "loss": 0.2501,
+      "step": 2803
+    },
+    {
+      "epoch": 1.8971583220568335,
+      "grad_norm": 0.5031206741591998,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.2516,
+      "step": 2804
+    },
+    {
+      "epoch": 1.8978349120433018,
+      "grad_norm": 0.49155903655182,
+      "learning_rate": 7.930106542113958e-08,
+      "loss": 0.2429,
+      "step": 2805
+    },
+    {
+      "epoch": 1.8985115020297698,
+      "grad_norm": 0.5006075652088146,
+      "learning_rate": 7.825693761322861e-08,
+      "loss": 0.2519,
+      "step": 2806
+    },
+    {
+      "epoch": 1.8991880920162383,
+      "grad_norm": 0.48349482537427474,
+      "learning_rate": 7.721967503955153e-08,
+      "loss": 0.2427,
+      "step": 2807
+    },
+    {
+      "epoch": 1.8998646820027063,
+      "grad_norm": 0.5095896305504259,
+      "learning_rate": 7.618927914696372e-08,
+      "loss": 0.247,
+      "step": 2808
+    },
+    {
+      "epoch": 1.9005412719891746,
+      "grad_norm": 0.49450481685579,
+      "learning_rate": 7.516575137274162e-08,
+      "loss": 0.2484,
+      "step": 2809
+    },
+    {
+      "epoch": 1.9012178619756428,
+      "grad_norm": 0.5036752684485468,
+      "learning_rate": 7.414909314458263e-08,
+      "loss": 0.2523,
+      "step": 2810
+    },
+    {
+      "epoch": 1.9018944519621108,
+      "grad_norm": 0.5079983947404492,
+      "learning_rate": 7.31393058806018e-08,
+      "loss": 0.25,
+      "step": 2811
+    },
+    {
+      "epoch": 1.9025710419485793,
+      "grad_norm": 0.492273284748527,
+      "learning_rate": 7.213639098933022e-08,
+      "loss": 0.2478,
+      "step": 2812
+    },
+    {
+      "epoch": 1.9032476319350473,
+      "grad_norm": 0.4990563383068816,
+      "learning_rate": 7.114034986971163e-08,
+      "loss": 0.2443,
+      "step": 2813
+    },
+    {
+      "epoch": 1.9039242219215156,
+      "grad_norm": 0.496267873463442,
+      "learning_rate": 7.015118391110299e-08,
+      "loss": 0.2441,
+      "step": 2814
+    },
+    {
+      "epoch": 1.9046008119079838,
+      "grad_norm": 0.4820428509069645,
+      "learning_rate": 6.916889449327058e-08,
+      "loss": 0.2399,
+      "step": 2815
+    },
+    {
+      "epoch": 1.9052774018944518,
+      "grad_norm": 0.49495627429482697,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.2516,
+      "step": 2816
+    },
+    {
+      "epoch": 1.9059539918809203,
+      "grad_norm": 0.4882254266036918,
+      "learning_rate": 6.722495075103752e-08,
+      "loss": 0.2384,
+      "step": 2817
+    },
+    {
+      "epoch": 1.9066305818673883,
+      "grad_norm": 0.503382111370038,
+      "learning_rate": 6.626329913820339e-08,
+      "loss": 0.2508,
+      "step": 2818
+    },
+    {
+      "epoch": 1.9073071718538566,
+      "grad_norm": 0.5163766622963248,
+      "learning_rate": 6.530852948927302e-08,
+      "loss": 0.2559,
+      "step": 2819
+    },
+    {
+      "epoch": 1.9079837618403248,
+      "grad_norm": 0.49738237844813493,
+      "learning_rate": 6.436064313603385e-08,
+      "loss": 0.2435,
+      "step": 2820
+    },
+    {
+      "epoch": 1.9086603518267928,
+      "grad_norm": 0.5011780215091824,
+      "learning_rate": 6.34196414006727e-08,
+      "loss": 0.2455,
+      "step": 2821
+    },
+    {
+      "epoch": 1.9093369418132613,
+      "grad_norm": 0.4945377563406494,
+      "learning_rate": 6.248552559577292e-08,
+      "loss": 0.2563,
+      "step": 2822
+    },
+    {
+      "epoch": 1.9100135317997293,
+      "grad_norm": 0.502159606963705,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.2442,
+      "step": 2823
+    },
+    {
+      "epoch": 1.9106901217861976,
+      "grad_norm": 0.4919212996062236,
+      "learning_rate": 6.063795697966057e-08,
+      "loss": 0.2421,
+      "step": 2824
+    },
+    {
+      "epoch": 1.9113667117726658,
+      "grad_norm": 0.5058685996938302,
+      "learning_rate": 5.972450674558373e-08,
+      "loss": 0.2453,
+      "step": 2825
+    },
+    {
+      "epoch": 1.9120433017591338,
+      "grad_norm": 0.5129684906694052,
+      "learning_rate": 5.881794759623194e-08,
+      "loss": 0.2561,
+      "step": 2826
+    },
+    {
+      "epoch": 1.9127198917456023,
+      "grad_norm": 0.4967341740538279,
+      "learning_rate": 5.7918280796145365e-08,
+      "loss": 0.2352,
+      "step": 2827
+    },
+    {
+      "epoch": 1.9133964817320703,
+      "grad_norm": 0.5062909141186087,
+      "learning_rate": 5.7025507600250165e-08,
+      "loss": 0.2491,
+      "step": 2828
+    },
+    {
+      "epoch": 1.9140730717185386,
+      "grad_norm": 0.49083888463389325,
+      "learning_rate": 5.6139629253856856e-08,
+      "loss": 0.2522,
+      "step": 2829
+    },
+    {
+      "epoch": 1.9147496617050068,
+      "grad_norm": 0.49707293993495216,
+      "learning_rate": 5.526064699265754e-08,
+      "loss": 0.2377,
+      "step": 2830
+    },
+    {
+      "epoch": 1.9154262516914748,
+      "grad_norm": 0.47856744730754347,
+      "learning_rate": 5.438856204272647e-08,
+      "loss": 0.2369,
+      "step": 2831
+    },
+    {
+      "epoch": 1.9161028416779433,
+      "grad_norm": 0.4974080633888079,
+      "learning_rate": 5.352337562051613e-08,
+      "loss": 0.2471,
+      "step": 2832
+    },
+    {
+      "epoch": 1.9167794316644113,
+      "grad_norm": 0.4954128076097202,
+      "learning_rate": 5.266508893285671e-08,
+      "loss": 0.2529,
+      "step": 2833
+    },
+    {
+      "epoch": 1.9174560216508796,
+      "grad_norm": 0.4991651198451824,
+      "learning_rate": 5.181370317695389e-08,
+      "loss": 0.2506,
+      "step": 2834
+    },
+    {
+      "epoch": 1.9181326116373478,
+      "grad_norm": 0.4761073587541816,
+      "learning_rate": 5.0969219540386584e-08,
+      "loss": 0.2423,
+      "step": 2835
+    },
+    {
+      "epoch": 1.9188092016238159,
+      "grad_norm": 0.5007203456956018,
+      "learning_rate": 5.013163920110864e-08,
+      "loss": 0.2587,
+      "step": 2836
+    },
+    {
+      "epoch": 1.9194857916102843,
+      "grad_norm": 0.4843125252484714,
+      "learning_rate": 4.930096332744105e-08,
+      "loss": 0.249,
+      "step": 2837
+    },
+    {
+      "epoch": 1.9201623815967523,
+      "grad_norm": 0.509480854301639,
+      "learning_rate": 4.847719307807752e-08,
+      "loss": 0.2555,
+      "step": 2838
+    },
+    {
+      "epoch": 1.9208389715832206,
+      "grad_norm": 0.5073143669644843,
+      "learning_rate": 4.766032960207667e-08,
+      "loss": 0.2553,
+      "step": 2839
+    },
+    {
+      "epoch": 1.9215155615696888,
+      "grad_norm": 0.49595677990579695,
+      "learning_rate": 4.685037403886483e-08,
+      "loss": 0.2447,
+      "step": 2840
+    },
+    {
+      "epoch": 1.9221921515561569,
+      "grad_norm": 0.48969276513853416,
+      "learning_rate": 4.604732751823049e-08,
+      "loss": 0.2443,
+      "step": 2841
+    },
+    {
+      "epoch": 1.9228687415426253,
+      "grad_norm": 0.48743930447076317,
+      "learning_rate": 4.52511911603265e-08,
+      "loss": 0.2336,
+      "step": 2842
+    },
+    {
+      "epoch": 1.9235453315290933,
+      "grad_norm": 0.49202867357399266,
+      "learning_rate": 4.446196607566677e-08,
+      "loss": 0.2524,
+      "step": 2843
+    },
+    {
+      "epoch": 1.9242219215155616,
+      "grad_norm": 0.4887204438242513,
+      "learning_rate": 4.367965336512403e-08,
+      "loss": 0.2443,
+      "step": 2844
+    },
+    {
+      "epoch": 1.9248985115020298,
+      "grad_norm": 0.5061838821820128,
+      "learning_rate": 4.290425411992982e-08,
+      "loss": 0.2458,
+      "step": 2845
+    },
+    {
+      "epoch": 1.9255751014884979,
+      "grad_norm": 0.49870320988895334,
+      "learning_rate": 4.21357694216723e-08,
+      "loss": 0.2395,
+      "step": 2846
+    },
+    {
+      "epoch": 1.9262516914749663,
+      "grad_norm": 0.508027798062586,
+      "learning_rate": 4.1374200342294e-08,
+      "loss": 0.2466,
+      "step": 2847
+    },
+    {
+      "epoch": 1.9269282814614344,
+      "grad_norm": 0.48871367141470645,
+      "learning_rate": 4.061954794409184e-08,
+      "loss": 0.2482,
+      "step": 2848
+    },
+    {
+      "epoch": 1.9276048714479026,
+      "grad_norm": 0.5157369841127691,
+      "learning_rate": 3.987181327971323e-08,
+      "loss": 0.2506,
+      "step": 2849
+    },
+    {
+      "epoch": 1.9282814614343708,
+      "grad_norm": 0.5164337144948753,
+      "learning_rate": 3.9130997392157756e-08,
+      "loss": 0.2617,
+      "step": 2850
+    },
+    {
+      "epoch": 1.9289580514208389,
+      "grad_norm": 0.4836822150150405,
+      "learning_rate": 3.839710131477492e-08,
+      "loss": 0.2433,
+      "step": 2851
+    },
+    {
+      "epoch": 1.9296346414073073,
+      "grad_norm": 0.5036517178007416,
+      "learning_rate": 3.7670126071259194e-08,
+      "loss": 0.2445,
+      "step": 2852
+    },
+    {
+      "epoch": 1.9303112313937754,
+      "grad_norm": 0.5033791911347771,
+      "learning_rate": 3.695007267565276e-08,
+      "loss": 0.2403,
+      "step": 2853
+    },
+    {
+      "epoch": 1.9309878213802436,
+      "grad_norm": 0.5168082117970442,
+      "learning_rate": 3.62369421323433e-08,
+      "loss": 0.2502,
+      "step": 2854
+    },
+    {
+      "epoch": 1.9316644113667119,
+      "grad_norm": 0.5214793256390812,
+      "learning_rate": 3.553073543606178e-08,
+      "loss": 0.2514,
+      "step": 2855
+    },
+    {
+      "epoch": 1.9323410013531799,
+      "grad_norm": 0.4732051098035832,
+      "learning_rate": 3.483145357187967e-08,
+      "loss": 0.2343,
+      "step": 2856
+    },
+    {
+      "epoch": 1.9330175913396481,
+      "grad_norm": 0.4979770215360456,
+      "learning_rate": 3.413909751521116e-08,
+      "loss": 0.2497,
+      "step": 2857
+    },
+    {
+      "epoch": 1.9336941813261164,
+      "grad_norm": 0.5168945791069994,
+      "learning_rate": 3.345366823180929e-08,
+      "loss": 0.2454,
+      "step": 2858
+    },
+    {
+      "epoch": 1.9343707713125846,
+      "grad_norm": 0.4754016651905845,
+      "learning_rate": 3.2775166677763704e-08,
+      "loss": 0.2421,
+      "step": 2859
+    },
+    {
+      "epoch": 1.9350473612990529,
+      "grad_norm": 0.47601582925769614,
+      "learning_rate": 3.2103593799501786e-08,
+      "loss": 0.2449,
+      "step": 2860
+    },
+    {
+      "epoch": 1.9357239512855209,
+      "grad_norm": 0.5104601336317238,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.2438,
+      "step": 2861
+    },
+    {
+      "epoch": 1.9364005412719891,
+      "grad_norm": 0.4969692495894916,
+      "learning_rate": 3.078123780771602e-08,
+      "loss": 0.2431,
+      "step": 2862
+    },
+    {
+      "epoch": 1.9370771312584574,
+      "grad_norm": 0.5135280573740428,
+      "learning_rate": 3.013045653871671e-08,
+      "loss": 0.253,
+      "step": 2863
+    },
+    {
+      "epoch": 1.9377537212449256,
+      "grad_norm": 0.46692191185022514,
+      "learning_rate": 2.94866076345518e-08,
+      "loss": 0.2445,
+      "step": 2864
+    },
+    {
+      "epoch": 1.9384303112313939,
+      "grad_norm": 0.486817777783252,
+      "learning_rate": 2.884969199331178e-08,
+      "loss": 0.2487,
+      "step": 2865
+    },
+    {
+      "epoch": 1.939106901217862,
+      "grad_norm": 0.48190346832023384,
+      "learning_rate": 2.8219710503416543e-08,
+      "loss": 0.2368,
+      "step": 2866
+    },
+    {
+      "epoch": 1.9397834912043301,
+      "grad_norm": 0.48414846987389876,
+      "learning_rate": 2.7596664043613718e-08,
+      "loss": 0.2404,
+      "step": 2867
+    },
+    {
+      "epoch": 1.9404600811907984,
+      "grad_norm": 0.4950227358568949,
+      "learning_rate": 2.6980553482977566e-08,
+      "loss": 0.2489,
+      "step": 2868
+    },
+    {
+      "epoch": 1.9411366711772666,
+      "grad_norm": 0.4664156108286562,
+      "learning_rate": 2.6371379680907307e-08,
+      "loss": 0.2434,
+      "step": 2869
+    },
+    {
+      "epoch": 1.9418132611637349,
+      "grad_norm": 0.5036598671071537,
+      "learning_rate": 2.5769143487127113e-08,
+      "loss": 0.2435,
+      "step": 2870
+    },
+    {
+      "epoch": 1.942489851150203,
+      "grad_norm": 0.48504360896960824,
+      "learning_rate": 2.5173845741682802e-08,
+      "loss": 0.2436,
+      "step": 2871
+    },
+    {
+      "epoch": 1.9431664411366711,
+      "grad_norm": 0.4970881024930273,
+      "learning_rate": 2.4585487274942922e-08,
+      "loss": 0.2584,
+      "step": 2872
+    },
+    {
+      "epoch": 1.9438430311231394,
+      "grad_norm": 0.501357122602246,
+      "learning_rate": 2.400406890759599e-08,
+      "loss": 0.2492,
+      "step": 2873
+    },
+    {
+      "epoch": 1.9445196211096076,
+      "grad_norm": 0.5183480636560841,
+      "learning_rate": 2.3429591450649934e-08,
+      "loss": 0.2417,
+      "step": 2874
+    },
+    {
+      "epoch": 1.9451962110960759,
+      "grad_norm": 0.48850492884075575,
+      "learning_rate": 2.286205570543043e-08,
+      "loss": 0.2355,
+      "step": 2875
+    },
+    {
+      "epoch": 1.945872801082544,
+      "grad_norm": 0.4990838743205616,
+      "learning_rate": 2.230146246358256e-08,
+      "loss": 0.2509,
+      "step": 2876
+    },
+    {
+      "epoch": 1.9465493910690121,
+      "grad_norm": 0.46937441607849295,
+      "learning_rate": 2.1747812507064148e-08,
+      "loss": 0.2372,
+      "step": 2877
+    },
+    {
+      "epoch": 1.9472259810554804,
+      "grad_norm": 0.4790787316684193,
+      "learning_rate": 2.120110660815078e-08,
+      "loss": 0.2391,
+      "step": 2878
+    },
+    {
+      "epoch": 1.9479025710419484,
+      "grad_norm": 0.4766320271458893,
+      "learning_rate": 2.0661345529430777e-08,
+      "loss": 0.2367,
+      "step": 2879
+    },
+    {
+      "epoch": 1.9485791610284169,
+      "grad_norm": 0.478446042788754,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.2457,
+      "step": 2880
+    },
+    {
+      "epoch": 1.949255751014885,
+      "grad_norm": 0.4998427098931567,
+      "learning_rate": 1.9602660834486253e-08,
+      "loss": 0.2523,
+      "step": 2881
+    },
+    {
+      "epoch": 1.9499323410013532,
+      "grad_norm": 0.48818914486332693,
+      "learning_rate": 1.9083738694998798e-08,
+      "loss": 0.2492,
+      "step": 2882
+    },
+    {
+      "epoch": 1.9506089309878214,
+      "grad_norm": 0.49791830814515875,
+      "learning_rate": 1.8571764329175512e-08,
+      "loss": 0.2537,
+      "step": 2883
+    },
+    {
+      "epoch": 1.9512855209742894,
+      "grad_norm": 0.47061086884028813,
+      "learning_rate": 1.8066738451159028e-08,
+      "loss": 0.2433,
+      "step": 2884
+    },
+    {
+      "epoch": 1.951962110960758,
+      "grad_norm": 0.48189529029961153,
+      "learning_rate": 1.756866176539862e-08,
+      "loss": 0.2501,
+      "step": 2885
+    },
+    {
+      "epoch": 1.952638700947226,
+      "grad_norm": 0.4999127706008267,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.2441,
+      "step": 2886
+    },
+    {
+      "epoch": 1.9533152909336942,
+      "grad_norm": 0.5024224434291705,
+      "learning_rate": 1.6593358739978028e-08,
+      "loss": 0.2447,
+      "step": 2887
+    },
+    {
+      "epoch": 1.9539918809201624,
+      "grad_norm": 0.493321218718305,
+      "learning_rate": 1.6116133760747944e-08,
+      "loss": 0.2467,
+      "step": 2888
+    },
+    {
+      "epoch": 1.9546684709066304,
+      "grad_norm": 0.4821621565652405,
+      "learning_rate": 1.564586069463081e-08,
+      "loss": 0.2462,
+      "step": 2889
+    },
+    {
+      "epoch": 1.955345060893099,
+      "grad_norm": 0.5138300138956365,
+      "learning_rate": 1.5182540197600237e-08,
+      "loss": 0.251,
+      "step": 2890
+    },
+    {
+      "epoch": 1.956021650879567,
+      "grad_norm": 0.5089452148264361,
+      "learning_rate": 1.4726172915933146e-08,
+      "loss": 0.2514,
+      "step": 2891
+    },
+    {
+      "epoch": 1.9566982408660352,
+      "grad_norm": 0.5026653292885015,
+      "learning_rate": 1.4276759486205328e-08,
+      "loss": 0.248,
+      "step": 2892
+    },
+    {
+      "epoch": 1.9573748308525034,
+      "grad_norm": 0.48920227692563584,
+      "learning_rate": 1.383430053529422e-08,
+      "loss": 0.2492,
+      "step": 2893
+    },
+    {
+      "epoch": 1.9580514208389714,
+      "grad_norm": 0.5143497974910732,
+      "learning_rate": 1.3398796680377245e-08,
+      "loss": 0.2502,
+      "step": 2894
+    },
+    {
+      "epoch": 1.95872801082544,
+      "grad_norm": 0.4912516003908781,
+      "learning_rate": 1.2970248528928476e-08,
+      "loss": 0.2485,
+      "step": 2895
+    },
+    {
+      "epoch": 1.959404600811908,
+      "grad_norm": 0.49264183351817703,
+      "learning_rate": 1.2548656678721404e-08,
+      "loss": 0.246,
+      "step": 2896
+    },
+    {
+      "epoch": 1.9600811907983762,
+      "grad_norm": 0.4912784877492902,
+      "learning_rate": 1.2134021717824518e-08,
+      "loss": 0.2435,
+      "step": 2897
+    },
+    {
+      "epoch": 1.9607577807848444,
+      "grad_norm": 0.4646279046430629,
+      "learning_rate": 1.1726344224603504e-08,
+      "loss": 0.233,
+      "step": 2898
+    },
+    {
+      "epoch": 1.9614343707713124,
+      "grad_norm": 0.4854082387874046,
+      "learning_rate": 1.132562476771959e-08,
+      "loss": 0.2461,
+      "step": 2899
+    },
+    {
+      "epoch": 1.962110960757781,
+      "grad_norm": 0.48101783543801596,
+      "learning_rate": 1.0931863906127327e-08,
+      "loss": 0.247,
+      "step": 2900
+    },
+    {
+      "epoch": 1.962787550744249,
+      "grad_norm": 0.5012427932037977,
+      "learning_rate": 1.0545062189075139e-08,
+      "loss": 0.2523,
+      "step": 2901
+    },
+    {
+      "epoch": 1.9634641407307172,
+      "grad_norm": 0.488898538344022,
+      "learning_rate": 1.0165220156105326e-08,
+      "loss": 0.2375,
+      "step": 2902
+    },
+    {
+      "epoch": 1.9641407307171854,
+      "grad_norm": 0.4961000607141083,
+      "learning_rate": 9.792338337050179e-09,
+      "loss": 0.246,
+      "step": 2903
+    },
+    {
+      "epoch": 1.9648173207036534,
+      "grad_norm": 0.49162744399344,
+      "learning_rate": 9.426417252035858e-09,
+      "loss": 0.2495,
+      "step": 2904
+    },
+    {
+      "epoch": 1.965493910690122,
+      "grad_norm": 0.4785554303916671,
+      "learning_rate": 9.067457411476854e-09,
+      "loss": 0.2424,
+      "step": 2905
+    },
+    {
+      "epoch": 1.96617050067659,
+      "grad_norm": 0.48404564570409264,
+      "learning_rate": 8.715459316078756e-09,
+      "loss": 0.2354,
+      "step": 2906
+    },
+    {
+      "epoch": 1.9668470906630582,
+      "grad_norm": 0.4810882991813076,
+      "learning_rate": 8.37042345683714e-09,
+      "loss": 0.2448,
+      "step": 2907
+    },
+    {
+      "epoch": 1.9675236806495264,
+      "grad_norm": 0.5003630824846942,
+      "learning_rate": 8.032350315033688e-09,
+      "loss": 0.2477,
+      "step": 2908
+    },
+    {
+      "epoch": 1.9682002706359945,
+      "grad_norm": 0.482111140020172,
+      "learning_rate": 7.701240362240626e-09,
+      "loss": 0.2371,
+      "step": 2909
+    },
+    {
+      "epoch": 1.968876860622463,
+      "grad_norm": 0.49451148944335455,
+      "learning_rate": 7.377094060315726e-09,
+      "loss": 0.2434,
+      "step": 2910
+    },
+    {
+      "epoch": 1.969553450608931,
+      "grad_norm": 0.48340391467839594,
+      "learning_rate": 7.0599118614034235e-09,
+      "loss": 0.2489,
+      "step": 2911
+    },
+    {
+      "epoch": 1.9702300405953992,
+      "grad_norm": 0.48634974319107793,
+      "learning_rate": 6.7496942079342546e-09,
+      "loss": 0.241,
+      "step": 2912
+    },
+    {
+      "epoch": 1.9709066305818674,
+      "grad_norm": 0.5102164699270759,
+      "learning_rate": 6.446441532625414e-09,
+      "loss": 0.2502,
+      "step": 2913
+    },
+    {
+      "epoch": 1.9715832205683355,
+      "grad_norm": 0.4827031447913692,
+      "learning_rate": 6.150154258476315e-09,
+      "loss": 0.2447,
+      "step": 2914
+    },
+    {
+      "epoch": 1.972259810554804,
+      "grad_norm": 0.4983647582469813,
+      "learning_rate": 5.8608327987719205e-09,
+      "loss": 0.2343,
+      "step": 2915
+    },
+    {
+      "epoch": 1.972936400541272,
+      "grad_norm": 0.5142969761038668,
+      "learning_rate": 5.578477557081074e-09,
+      "loss": 0.2499,
+      "step": 2916
+    },
+    {
+      "epoch": 1.9736129905277402,
+      "grad_norm": 0.4963064047823833,
+      "learning_rate": 5.30308892725484e-09,
+      "loss": 0.255,
+      "step": 2917
+    },
+    {
+      "epoch": 1.9742895805142084,
+      "grad_norm": 0.48945922512619633,
+      "learning_rate": 5.034667293427053e-09,
+      "loss": 0.2431,
+      "step": 2918
+    },
+    {
+      "epoch": 1.9749661705006765,
+      "grad_norm": 0.4783710901373366,
+      "learning_rate": 4.773213030012658e-09,
+      "loss": 0.2452,
+      "step": 2919
+    },
+    {
+      "epoch": 1.975642760487145,
+      "grad_norm": 0.5224318265069263,
+      "learning_rate": 4.51872650170937e-09,
+      "loss": 0.2578,
+      "step": 2920
+    },
+    {
+      "epoch": 1.976319350473613,
+      "grad_norm": 0.5053313180316535,
+      "learning_rate": 4.2712080634949024e-09,
+      "loss": 0.2597,
+      "step": 2921
+    },
+    {
+      "epoch": 1.9769959404600812,
+      "grad_norm": 0.4992598693638966,
+      "learning_rate": 4.030658060626969e-09,
+      "loss": 0.2484,
+      "step": 2922
+    },
+    {
+      "epoch": 1.9776725304465494,
+      "grad_norm": 0.4770059144967861,
+      "learning_rate": 3.797076828644386e-09,
+      "loss": 0.2464,
+      "step": 2923
+    },
+    {
+      "epoch": 1.9783491204330175,
+      "grad_norm": 0.505667664220048,
+      "learning_rate": 3.570464693364306e-09,
+      "loss": 0.2522,
+      "step": 2924
+    },
+    {
+      "epoch": 1.979025710419486,
+      "grad_norm": 0.4740767230373284,
+      "learning_rate": 3.3508219708822122e-09,
+      "loss": 0.2421,
+      "step": 2925
+    },
+    {
+      "epoch": 1.979702300405954,
+      "grad_norm": 0.5007362803558659,
+      "learning_rate": 3.1381489675746946e-09,
+      "loss": 0.2523,
+      "step": 2926
+    },
+    {
+      "epoch": 1.9803788903924222,
+      "grad_norm": 0.49686880911383086,
+      "learning_rate": 2.9324459800933457e-09,
+      "loss": 0.2409,
+      "step": 2927
+    },
+    {
+      "epoch": 1.9810554803788905,
+      "grad_norm": 0.4954856233237567,
+      "learning_rate": 2.7337132953697555e-09,
+      "loss": 0.2493,
+      "step": 2928
+    },
+    {
+      "epoch": 1.9817320703653585,
+      "grad_norm": 0.4771912770255598,
+      "learning_rate": 2.5419511906116244e-09,
+      "loss": 0.2364,
+      "step": 2929
+    },
+    {
+      "epoch": 1.982408660351827,
+      "grad_norm": 0.4857159469126075,
+      "learning_rate": 2.3571599333038765e-09,
+      "loss": 0.2494,
+      "step": 2930
+    },
+    {
+      "epoch": 1.983085250338295,
+      "grad_norm": 0.4761218183880434,
+      "learning_rate": 2.179339781208101e-09,
+      "loss": 0.2406,
+      "step": 2931
+    },
+    {
+      "epoch": 1.9837618403247632,
+      "grad_norm": 0.46888439425941003,
+      "learning_rate": 2.008490982360889e-09,
+      "loss": 0.2344,
+      "step": 2932
+    },
+    {
+      "epoch": 1.9844384303112315,
+      "grad_norm": 0.5048785472051432,
+      "learning_rate": 1.8446137750766092e-09,
+      "loss": 0.2514,
+      "step": 2933
+    },
+    {
+      "epoch": 1.9851150202976995,
+      "grad_norm": 0.48351046118425767,
+      "learning_rate": 1.687708387944076e-09,
+      "loss": 0.2406,
+      "step": 2934
+    },
+    {
+      "epoch": 1.985791610284168,
+      "grad_norm": 0.49604941304295097,
+      "learning_rate": 1.53777503982655e-09,
+      "loss": 0.2431,
+      "step": 2935
+    },
+    {
+      "epoch": 1.986468200270636,
+      "grad_norm": 0.5019072911460812,
+      "learning_rate": 1.3948139398628492e-09,
+      "loss": 0.243,
+      "step": 2936
+    },
+    {
+      "epoch": 1.9871447902571042,
+      "grad_norm": 0.4780058891930668,
+      "learning_rate": 1.2588252874673469e-09,
+      "loss": 0.2398,
+      "step": 2937
+    },
+    {
+      "epoch": 1.9878213802435725,
+      "grad_norm": 0.48689737010803646,
+      "learning_rate": 1.1298092723266429e-09,
+      "loss": 0.2483,
+      "step": 2938
+    },
+    {
+      "epoch": 1.9884979702300405,
+      "grad_norm": 0.5089865411792508,
+      "learning_rate": 1.0077660744028938e-09,
+      "loss": 0.244,
+      "step": 2939
+    },
+    {
+      "epoch": 1.989174560216509,
+      "grad_norm": 0.4979817881726077,
+      "learning_rate": 8.926958639315919e-10,
+      "loss": 0.2466,
+      "step": 2940
+    },
+    {
+      "epoch": 1.989851150202977,
+      "grad_norm": 0.5017821845354936,
+      "learning_rate": 7.845988014215655e-10,
+      "loss": 0.2438,
+      "step": 2941
+    },
+    {
+      "epoch": 1.9905277401894452,
+      "grad_norm": 0.49967232366580777,
+      "learning_rate": 6.834750376549793e-10,
+      "loss": 0.2461,
+      "step": 2942
+    },
+    {
+      "epoch": 1.9912043301759135,
+      "grad_norm": 0.4931404182115779,
+      "learning_rate": 5.893247136878888e-10,
+      "loss": 0.246,
+      "step": 2943
+    },
+    {
+      "epoch": 1.9918809201623815,
+      "grad_norm": 0.5070265981733995,
+      "learning_rate": 5.021479608474655e-10,
+      "loss": 0.2464,
+      "step": 2944
+    },
+    {
+      "epoch": 1.9925575101488497,
+      "grad_norm": 0.47916574361514197,
+      "learning_rate": 4.219449007358822e-10,
+      "loss": 0.2472,
+      "step": 2945
+    },
+    {
+      "epoch": 1.993234100135318,
+      "grad_norm": 0.47363338475168887,
+      "learning_rate": 3.487156452258722e-10,
+      "loss": 0.2442,
+      "step": 2946
+    },
+    {
+      "epoch": 1.9939106901217862,
+      "grad_norm": 0.4956475642405451,
+      "learning_rate": 2.824602964635048e-10,
+      "loss": 0.2422,
+      "step": 2947
+    },
+    {
+      "epoch": 1.9945872801082545,
+      "grad_norm": 0.49073324903809346,
+      "learning_rate": 2.231789468670753e-10,
+      "loss": 0.2398,
+      "step": 2948
+    },
+    {
+      "epoch": 1.9952638700947225,
+      "grad_norm": 0.5125716090777569,
+      "learning_rate": 1.7087167912710477e-10,
+      "loss": 0.2498,
+      "step": 2949
+    },
+    {
+      "epoch": 1.9959404600811907,
+      "grad_norm": 0.5039266387486624,
+      "learning_rate": 1.2553856620522997e-10,
+      "loss": 0.2484,
+      "step": 2950
+    },
+    {
+      "epoch": 1.996617050067659,
+      "grad_norm": 0.515447982216017,
+      "learning_rate": 8.717967133586857e-11,
+      "loss": 0.2566,
+      "step": 2951
+    },
+    {
+      "epoch": 1.9972936400541272,
+      "grad_norm": 0.48340547116025606,
+      "learning_rate": 5.579504802566416e-11,
+      "loss": 0.2415,
+      "step": 2952
+    },
+    {
+      "epoch": 1.9979702300405955,
+      "grad_norm": 0.5051218052502676,
+      "learning_rate": 3.1384740051265684e-11,
+      "loss": 0.249,
+      "step": 2953
+    },
+    {
+      "epoch": 1.9986468200270635,
+      "grad_norm": 0.5030179785805474,
+      "learning_rate": 1.3948781463213324e-11,
+      "loss": 0.2493,
+      "step": 2954
+    },
+    {
+      "epoch": 1.9993234100135318,
+      "grad_norm": 0.5078110148843704,
+      "learning_rate": 3.487196582052654e-12,
+      "loss": 0.249,
+      "step": 2955
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4808176875836281,
+      "learning_rate": 0.0,
+      "loss": 0.2388,
+      "step": 2956
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.2765226662158966,
+      "eval_runtime": 93.6575,
+      "eval_samples_per_second": 106.292,
+      "eval_steps_per_second": 0.833,
+      "step": 2956
+    },
+    {
+      "epoch": 2.0,
+      "step": 2956,
+      "total_flos": 1086769682448384.0,
+      "train_loss": 0.29042090329370573,
+      "train_runtime": 13931.3179,
+      "train_samples_per_second": 27.15,
+      "train_steps_per_second": 0.212
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2956,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1086769682448384.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}