diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18669 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9971830985915493,
+  "eval_steps": 500,
+  "global_step": 2661,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011267605633802818,
+      "grad_norm": 54.608844611192445,
+      "learning_rate": 1.8726591760299626e-07,
+      "loss": 11.127,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022535211267605635,
+      "grad_norm": 55.36601190256884,
+      "learning_rate": 3.7453183520599253e-07,
+      "loss": 11.0439,
+      "step": 2
+    },
+    {
+      "epoch": 0.0033802816901408453,
+      "grad_norm": 57.0063901888475,
+      "learning_rate": 5.617977528089887e-07,
+      "loss": 11.0579,
+      "step": 3
+    },
+    {
+      "epoch": 0.004507042253521127,
+      "grad_norm": 56.35810808194075,
+      "learning_rate": 7.490636704119851e-07,
+      "loss": 11.0157,
+      "step": 4
+    },
+    {
+      "epoch": 0.005633802816901409,
+      "grad_norm": 55.58312483866836,
+      "learning_rate": 9.363295880149814e-07,
+      "loss": 10.9188,
+      "step": 5
+    },
+    {
+      "epoch": 0.0067605633802816905,
+      "grad_norm": 53.77064024372009,
+      "learning_rate": 1.1235955056179775e-06,
+      "loss": 11.0218,
+      "step": 6
+    },
+    {
+      "epoch": 0.007887323943661971,
+      "grad_norm": 56.27070673773756,
+      "learning_rate": 1.310861423220974e-06,
+      "loss": 10.8824,
+      "step": 7
+    },
+    {
+      "epoch": 0.009014084507042254,
+      "grad_norm": 60.218078994693784,
+      "learning_rate": 1.4981273408239701e-06,
+      "loss": 10.7133,
+      "step": 8
+    },
+    {
+      "epoch": 0.010140845070422535,
+      "grad_norm": 61.64507136781477,
+      "learning_rate": 1.6853932584269663e-06,
+      "loss": 10.633,
+      "step": 9
+    },
+    {
+      "epoch": 0.011267605633802818,
+      "grad_norm": 61.97384688440408,
+      "learning_rate": 1.8726591760299627e-06,
+      "loss": 10.6376,
+      "step": 10
+    },
+    {
+      "epoch": 0.012394366197183098,
+      "grad_norm": 80.39629204947155,
+      "learning_rate": 2.0599250936329587e-06,
+      "loss": 9.5566,
+      "step": 11
+    },
+    {
+      "epoch": 0.013521126760563381,
+      "grad_norm": 86.33183007861814,
+      "learning_rate": 2.247191011235955e-06,
+      "loss": 9.394,
+      "step": 12
+    },
+    {
+      "epoch": 0.014647887323943662,
+      "grad_norm": 95.91392176867679,
+      "learning_rate": 2.4344569288389516e-06,
+      "loss": 8.8441,
+      "step": 13
+    },
+    {
+      "epoch": 0.015774647887323943,
+      "grad_norm": 103.31532121583676,
+      "learning_rate": 2.621722846441948e-06,
+      "loss": 8.6743,
+      "step": 14
+    },
+    {
+      "epoch": 0.016901408450704224,
+      "grad_norm": 65.1185503810022,
+      "learning_rate": 2.808988764044944e-06,
+      "loss": 3.687,
+      "step": 15
+    },
+    {
+      "epoch": 0.018028169014084508,
+      "grad_norm": 59.49401022581223,
+      "learning_rate": 2.9962546816479402e-06,
+      "loss": 3.345,
+      "step": 16
+    },
+    {
+      "epoch": 0.01915492957746479,
+      "grad_norm": 51.01948856952058,
+      "learning_rate": 3.1835205992509364e-06,
+      "loss": 3.1423,
+      "step": 17
+    },
+    {
+      "epoch": 0.02028169014084507,
+      "grad_norm": 37.11138317252327,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 2.5809,
+      "step": 18
+    },
+    {
+      "epoch": 0.02140845070422535,
+      "grad_norm": 31.987438014939944,
+      "learning_rate": 3.558052434456929e-06,
+      "loss": 2.3297,
+      "step": 19
+    },
+    {
+      "epoch": 0.022535211267605635,
+      "grad_norm": 6.423839988685857,
+      "learning_rate": 3.7453183520599255e-06,
+      "loss": 1.364,
+      "step": 20
+    },
+    {
+      "epoch": 0.023661971830985916,
+      "grad_norm": 5.045295439788782,
+      "learning_rate": 3.932584269662922e-06,
+      "loss": 1.2811,
+      "step": 21
+    },
+    {
+      "epoch": 0.024788732394366197,
+      "grad_norm": 4.170853030544661,
+      "learning_rate": 4.1198501872659175e-06,
+      "loss": 1.2669,
+      "step": 22
+    },
+    {
+      "epoch": 0.025915492957746478,
+      "grad_norm": 3.3349694092652165,
+      "learning_rate": 4.307116104868914e-06,
+      "loss": 1.1987,
+      "step": 23
+    },
+    {
+      "epoch": 0.027042253521126762,
+      "grad_norm": 2.533758218153574,
+      "learning_rate": 4.49438202247191e-06,
+      "loss": 1.0448,
+      "step": 24
+    },
+    {
+      "epoch": 0.028169014084507043,
+      "grad_norm": 2.155231765906746,
+      "learning_rate": 4.6816479400749066e-06,
+      "loss": 1.0711,
+      "step": 25
+    },
+    {
+      "epoch": 0.029295774647887324,
+      "grad_norm": 1.8042975450695744,
+      "learning_rate": 4.868913857677903e-06,
+      "loss": 1.0416,
+      "step": 26
+    },
+    {
+      "epoch": 0.030422535211267605,
+      "grad_norm": 1.3444349342706392,
+      "learning_rate": 5.056179775280899e-06,
+      "loss": 0.9522,
+      "step": 27
+    },
+    {
+      "epoch": 0.031549295774647886,
+      "grad_norm": 11.864058093901697,
+      "learning_rate": 5.243445692883896e-06,
+      "loss": 0.9194,
+      "step": 28
+    },
+    {
+      "epoch": 0.03267605633802817,
+      "grad_norm": 6.2119522899213475,
+      "learning_rate": 5.430711610486891e-06,
+      "loss": 0.8908,
+      "step": 29
+    },
+    {
+      "epoch": 0.03380281690140845,
+      "grad_norm": 1.4146654224158117,
+      "learning_rate": 5.617977528089888e-06,
+      "loss": 0.8566,
+      "step": 30
+    },
+    {
+      "epoch": 0.03492957746478873,
+      "grad_norm": 1.0329328673745717,
+      "learning_rate": 5.805243445692885e-06,
+      "loss": 0.8422,
+      "step": 31
+    },
+    {
+      "epoch": 0.036056338028169016,
+      "grad_norm": 0.9928258011925138,
+      "learning_rate": 5.9925093632958805e-06,
+      "loss": 0.8016,
+      "step": 32
+    },
+    {
+      "epoch": 0.03718309859154929,
+      "grad_norm": 0.8349238341651046,
+      "learning_rate": 6.179775280898876e-06,
+      "loss": 0.7795,
+      "step": 33
+    },
+    {
+      "epoch": 0.03830985915492958,
+      "grad_norm": 0.7208913642406135,
+      "learning_rate": 6.367041198501873e-06,
+      "loss": 0.7584,
+      "step": 34
+    },
+    {
+      "epoch": 0.03943661971830986,
+      "grad_norm": 0.7512239641572525,
+      "learning_rate": 6.554307116104869e-06,
+      "loss": 0.7576,
+      "step": 35
+    },
+    {
+      "epoch": 0.04056338028169014,
+      "grad_norm": 0.7667204457282304,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 0.7354,
+      "step": 36
+    },
+    {
+      "epoch": 0.041690140845070424,
+      "grad_norm": 0.5845579314338065,
+      "learning_rate": 6.928838951310862e-06,
+      "loss": 0.7496,
+      "step": 37
+    },
+    {
+      "epoch": 0.0428169014084507,
+      "grad_norm": 0.6139154497318552,
+      "learning_rate": 7.116104868913858e-06,
+      "loss": 0.7276,
+      "step": 38
+    },
+    {
+      "epoch": 0.043943661971830986,
+      "grad_norm": 0.6594049952301563,
+      "learning_rate": 7.303370786516854e-06,
+      "loss": 0.7092,
+      "step": 39
+    },
+    {
+      "epoch": 0.04507042253521127,
+      "grad_norm": 0.6420935015482123,
+      "learning_rate": 7.490636704119851e-06,
+      "loss": 0.7326,
+      "step": 40
+    },
+    {
+      "epoch": 0.04619718309859155,
+      "grad_norm": 0.5281343129019502,
+      "learning_rate": 7.677902621722846e-06,
+      "loss": 0.6908,
+      "step": 41
+    },
+    {
+      "epoch": 0.04732394366197183,
+      "grad_norm": 0.5290984786213891,
+      "learning_rate": 7.865168539325843e-06,
+      "loss": 0.6736,
+      "step": 42
+    },
+    {
+      "epoch": 0.048450704225352116,
+      "grad_norm": 0.4441023144930531,
+      "learning_rate": 8.05243445692884e-06,
+      "loss": 0.6482,
+      "step": 43
+    },
+    {
+      "epoch": 0.049577464788732394,
+      "grad_norm": 0.494465466454887,
+      "learning_rate": 8.239700374531835e-06,
+      "loss": 0.6487,
+      "step": 44
+    },
+    {
+      "epoch": 0.05070422535211268,
+      "grad_norm": 0.46992378200427287,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 0.6515,
+      "step": 45
+    },
+    {
+      "epoch": 0.051830985915492955,
+      "grad_norm": 0.48528089193349894,
+      "learning_rate": 8.614232209737828e-06,
+      "loss": 0.6652,
+      "step": 46
+    },
+    {
+      "epoch": 0.05295774647887324,
+      "grad_norm": 0.4309300506022204,
+      "learning_rate": 8.801498127340826e-06,
+      "loss": 0.6395,
+      "step": 47
+    },
+    {
+      "epoch": 0.054084507042253524,
+      "grad_norm": 0.46983425215048863,
+      "learning_rate": 8.98876404494382e-06,
+      "loss": 0.6498,
+      "step": 48
+    },
+    {
+      "epoch": 0.0552112676056338,
+      "grad_norm": 0.4232542955369631,
+      "learning_rate": 9.176029962546817e-06,
+      "loss": 0.6074,
+      "step": 49
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.4950602179089511,
+      "learning_rate": 9.363295880149813e-06,
+      "loss": 0.6537,
+      "step": 50
+    },
+    {
+      "epoch": 0.05746478873239436,
+      "grad_norm": 0.40781412615064017,
+      "learning_rate": 9.550561797752809e-06,
+      "loss": 0.6075,
+      "step": 51
+    },
+    {
+      "epoch": 0.05859154929577465,
+      "grad_norm": 0.36716946070233086,
+      "learning_rate": 9.737827715355806e-06,
+      "loss": 0.5996,
+      "step": 52
+    },
+    {
+      "epoch": 0.05971830985915493,
+      "grad_norm": 0.3285197883333934,
+      "learning_rate": 9.925093632958802e-06,
+      "loss": 0.5855,
+      "step": 53
+    },
+    {
+      "epoch": 0.06084507042253521,
+      "grad_norm": 0.37133706441288794,
+      "learning_rate": 1.0112359550561798e-05,
+      "loss": 0.6064,
+      "step": 54
+    },
+    {
+      "epoch": 0.061971830985915494,
+      "grad_norm": 0.39194875781607186,
+      "learning_rate": 1.0299625468164795e-05,
+      "loss": 0.6192,
+      "step": 55
+    },
+    {
+      "epoch": 0.06309859154929577,
+      "grad_norm": 0.35740092268419,
+      "learning_rate": 1.0486891385767791e-05,
+      "loss": 0.5782,
+      "step": 56
+    },
+    {
+      "epoch": 0.06422535211267606,
+      "grad_norm": 0.3684352593616929,
+      "learning_rate": 1.0674157303370787e-05,
+      "loss": 0.5872,
+      "step": 57
+    },
+    {
+      "epoch": 0.06535211267605634,
+      "grad_norm": 0.3025401149269732,
+      "learning_rate": 1.0861423220973783e-05,
+      "loss": 0.5798,
+      "step": 58
+    },
+    {
+      "epoch": 0.06647887323943662,
+      "grad_norm": 0.3590555360599909,
+      "learning_rate": 1.104868913857678e-05,
+      "loss": 0.5729,
+      "step": 59
+    },
+    {
+      "epoch": 0.0676056338028169,
+      "grad_norm": 0.34830243471060585,
+      "learning_rate": 1.1235955056179776e-05,
+      "loss": 0.5957,
+      "step": 60
+    },
+    {
+      "epoch": 0.06873239436619719,
+      "grad_norm": 0.3460855247633223,
+      "learning_rate": 1.1423220973782772e-05,
+      "loss": 0.6068,
+      "step": 61
+    },
+    {
+      "epoch": 0.06985915492957746,
+      "grad_norm": 0.35193542528227495,
+      "learning_rate": 1.161048689138577e-05,
+      "loss": 0.5972,
+      "step": 62
+    },
+    {
+      "epoch": 0.07098591549295774,
+      "grad_norm": 0.31692586734081885,
+      "learning_rate": 1.1797752808988765e-05,
+      "loss": 0.556,
+      "step": 63
+    },
+    {
+      "epoch": 0.07211267605633803,
+      "grad_norm": 0.37008606019965695,
+      "learning_rate": 1.1985018726591761e-05,
+      "loss": 0.5631,
+      "step": 64
+    },
+    {
+      "epoch": 0.07323943661971831,
+      "grad_norm": 0.3526075220813616,
+      "learning_rate": 1.2172284644194758e-05,
+      "loss": 0.5881,
+      "step": 65
+    },
+    {
+      "epoch": 0.07436619718309859,
+      "grad_norm": 0.2850697798787257,
+      "learning_rate": 1.2359550561797752e-05,
+      "loss": 0.5767,
+      "step": 66
+    },
+    {
+      "epoch": 0.07549295774647888,
+      "grad_norm": 0.3287887407319393,
+      "learning_rate": 1.254681647940075e-05,
+      "loss": 0.5817,
+      "step": 67
+    },
+    {
+      "epoch": 0.07661971830985916,
+      "grad_norm": 0.28737949408045615,
+      "learning_rate": 1.2734082397003746e-05,
+      "loss": 0.5433,
+      "step": 68
+    },
+    {
+      "epoch": 0.07774647887323943,
+      "grad_norm": 0.3319938188307383,
+      "learning_rate": 1.2921348314606743e-05,
+      "loss": 0.6024,
+      "step": 69
+    },
+    {
+      "epoch": 0.07887323943661972,
+      "grad_norm": 0.30064037993492726,
+      "learning_rate": 1.3108614232209737e-05,
+      "loss": 0.5634,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2990949423538674,
+      "learning_rate": 1.3295880149812733e-05,
+      "loss": 0.5624,
+      "step": 71
+    },
+    {
+      "epoch": 0.08112676056338028,
+      "grad_norm": 0.3711818897681024,
+      "learning_rate": 1.348314606741573e-05,
+      "loss": 0.5755,
+      "step": 72
+    },
+    {
+      "epoch": 0.08225352112676056,
+      "grad_norm": 0.30557570361690733,
+      "learning_rate": 1.3670411985018728e-05,
+      "loss": 0.5373,
+      "step": 73
+    },
+    {
+      "epoch": 0.08338028169014085,
+      "grad_norm": 0.28899560683771475,
+      "learning_rate": 1.3857677902621724e-05,
+      "loss": 0.5435,
+      "step": 74
+    },
+    {
+      "epoch": 0.08450704225352113,
+      "grad_norm": 0.33073697214599473,
+      "learning_rate": 1.4044943820224721e-05,
+      "loss": 0.5498,
+      "step": 75
+    },
+    {
+      "epoch": 0.0856338028169014,
+      "grad_norm": 0.34803723973746536,
+      "learning_rate": 1.4232209737827715e-05,
+      "loss": 0.5429,
+      "step": 76
+    },
+    {
+      "epoch": 0.0867605633802817,
+      "grad_norm": 0.3509830944935367,
+      "learning_rate": 1.4419475655430711e-05,
+      "loss": 0.5601,
+      "step": 77
+    },
+    {
+      "epoch": 0.08788732394366197,
+      "grad_norm": 0.2550870509542074,
+      "learning_rate": 1.4606741573033709e-05,
+      "loss": 0.5507,
+      "step": 78
+    },
+    {
+      "epoch": 0.08901408450704225,
+      "grad_norm": 0.35308533996212627,
+      "learning_rate": 1.4794007490636705e-05,
+      "loss": 0.5496,
+      "step": 79
+    },
+    {
+      "epoch": 0.09014084507042254,
+      "grad_norm": 0.3081240884981584,
+      "learning_rate": 1.4981273408239702e-05,
+      "loss": 0.5218,
+      "step": 80
+    },
+    {
+      "epoch": 0.09126760563380282,
+      "grad_norm": 0.33145750299141974,
+      "learning_rate": 1.5168539325842698e-05,
+      "loss": 0.5071,
+      "step": 81
+    },
+    {
+      "epoch": 0.0923943661971831,
+      "grad_norm": 0.26989050721529634,
+      "learning_rate": 1.5355805243445692e-05,
+      "loss": 0.5137,
+      "step": 82
+    },
+    {
+      "epoch": 0.09352112676056339,
+      "grad_norm": 0.3076846617884833,
+      "learning_rate": 1.554307116104869e-05,
+      "loss": 0.511,
+      "step": 83
+    },
+    {
+      "epoch": 0.09464788732394366,
+      "grad_norm": 0.34576789102985467,
+      "learning_rate": 1.5730337078651687e-05,
+      "loss": 0.5382,
+      "step": 84
+    },
+    {
+      "epoch": 0.09577464788732394,
+      "grad_norm": 0.3102686540327493,
+      "learning_rate": 1.591760299625468e-05,
+      "loss": 0.5391,
+      "step": 85
+    },
+    {
+      "epoch": 0.09690140845070423,
+      "grad_norm": 0.2907807397746215,
+      "learning_rate": 1.610486891385768e-05,
+      "loss": 0.5421,
+      "step": 86
+    },
+    {
+      "epoch": 0.09802816901408451,
+      "grad_norm": 0.3538353366658328,
+      "learning_rate": 1.6292134831460676e-05,
+      "loss": 0.54,
+      "step": 87
+    },
+    {
+      "epoch": 0.09915492957746479,
+      "grad_norm": 0.32089888058996763,
+      "learning_rate": 1.647940074906367e-05,
+      "loss": 0.5281,
+      "step": 88
+    },
+    {
+      "epoch": 0.10028169014084506,
+      "grad_norm": 0.30311268558188664,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.5136,
+      "step": 89
+    },
+    {
+      "epoch": 0.10140845070422536,
+      "grad_norm": 0.32804689946336774,
+      "learning_rate": 1.6853932584269665e-05,
+      "loss": 0.5344,
+      "step": 90
+    },
+    {
+      "epoch": 0.10253521126760563,
+      "grad_norm": 0.3634597759205688,
+      "learning_rate": 1.704119850187266e-05,
+      "loss": 0.5401,
+      "step": 91
+    },
+    {
+      "epoch": 0.10366197183098591,
+      "grad_norm": 0.3671107837131908,
+      "learning_rate": 1.7228464419475657e-05,
+      "loss": 0.5274,
+      "step": 92
+    },
+    {
+      "epoch": 0.1047887323943662,
+      "grad_norm": 0.3126280631131301,
+      "learning_rate": 1.7415730337078654e-05,
+      "loss": 0.5585,
+      "step": 93
+    },
+    {
+      "epoch": 0.10591549295774648,
+      "grad_norm": 0.3840901646645938,
+      "learning_rate": 1.760299625468165e-05,
+      "loss": 0.5543,
+      "step": 94
+    },
+    {
+      "epoch": 0.10704225352112676,
+      "grad_norm": 0.34406766779885367,
+      "learning_rate": 1.7790262172284646e-05,
+      "loss": 0.5155,
+      "step": 95
+    },
+    {
+      "epoch": 0.10816901408450705,
+      "grad_norm": 0.36234031937492145,
+      "learning_rate": 1.797752808988764e-05,
+      "loss": 0.5551,
+      "step": 96
+    },
+    {
+      "epoch": 0.10929577464788733,
+      "grad_norm": 0.36566069027244064,
+      "learning_rate": 1.8164794007490637e-05,
+      "loss": 0.5217,
+      "step": 97
+    },
+    {
+      "epoch": 0.1104225352112676,
+      "grad_norm": 0.3763936721342812,
+      "learning_rate": 1.8352059925093635e-05,
+      "loss": 0.5074,
+      "step": 98
+    },
+    {
+      "epoch": 0.1115492957746479,
+      "grad_norm": 0.3095163828509057,
+      "learning_rate": 1.8539325842696632e-05,
+      "loss": 0.5049,
+      "step": 99
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.32865175173186806,
+      "learning_rate": 1.8726591760299626e-05,
+      "loss": 0.5154,
+      "step": 100
+    },
+    {
+      "epoch": 0.11380281690140845,
+      "grad_norm": 0.3018690447008865,
+      "learning_rate": 1.891385767790262e-05,
+      "loss": 0.5111,
+      "step": 101
+    },
+    {
+      "epoch": 0.11492957746478873,
+      "grad_norm": 0.32640387711607555,
+      "learning_rate": 1.9101123595505618e-05,
+      "loss": 0.5083,
+      "step": 102
+    },
+    {
+      "epoch": 0.11605633802816902,
+      "grad_norm": 0.2905811810611324,
+      "learning_rate": 1.9288389513108615e-05,
+      "loss": 0.5241,
+      "step": 103
+    },
+    {
+      "epoch": 0.1171830985915493,
+      "grad_norm": 0.2871053244214264,
+      "learning_rate": 1.9475655430711613e-05,
+      "loss": 0.5136,
+      "step": 104
+    },
+    {
+      "epoch": 0.11830985915492957,
+      "grad_norm": 0.3024184852425039,
+      "learning_rate": 1.9662921348314607e-05,
+      "loss": 0.5238,
+      "step": 105
+    },
+    {
+      "epoch": 0.11943661971830986,
+      "grad_norm": 0.29008239870487684,
+      "learning_rate": 1.9850187265917604e-05,
+      "loss": 0.4903,
+      "step": 106
+    },
+    {
+      "epoch": 0.12056338028169014,
+      "grad_norm": 0.28736122238768774,
+      "learning_rate": 2.00374531835206e-05,
+      "loss": 0.5224,
+      "step": 107
+    },
+    {
+      "epoch": 0.12169014084507042,
+      "grad_norm": 0.33884824146385534,
+      "learning_rate": 2.0224719101123596e-05,
+      "loss": 0.5159,
+      "step": 108
+    },
+    {
+      "epoch": 0.12281690140845071,
+      "grad_norm": 0.33764167173784454,
+      "learning_rate": 2.0411985018726593e-05,
+      "loss": 0.4998,
+      "step": 109
+    },
+    {
+      "epoch": 0.12394366197183099,
+      "grad_norm": 0.28667282492308055,
+      "learning_rate": 2.059925093632959e-05,
+      "loss": 0.4886,
+      "step": 110
+    },
+    {
+      "epoch": 0.12507042253521128,
+      "grad_norm": 0.3586302431156048,
+      "learning_rate": 2.0786516853932585e-05,
+      "loss": 0.535,
+      "step": 111
+    },
+    {
+      "epoch": 0.12619718309859154,
+      "grad_norm": 0.29139506020299266,
+      "learning_rate": 2.0973782771535582e-05,
+      "loss": 0.5203,
+      "step": 112
+    },
+    {
+      "epoch": 0.12732394366197183,
+      "grad_norm": 0.32466629756389626,
+      "learning_rate": 2.1161048689138577e-05,
+      "loss": 0.4974,
+      "step": 113
+    },
+    {
+      "epoch": 0.12845070422535212,
+      "grad_norm": 0.319478891916245,
+      "learning_rate": 2.1348314606741574e-05,
+      "loss": 0.5037,
+      "step": 114
+    },
+    {
+      "epoch": 0.1295774647887324,
+      "grad_norm": 0.30895243254092425,
+      "learning_rate": 2.153558052434457e-05,
+      "loss": 0.506,
+      "step": 115
+    },
+    {
+      "epoch": 0.13070422535211268,
+      "grad_norm": 0.3081561843950191,
+      "learning_rate": 2.1722846441947566e-05,
+      "loss": 0.4933,
+      "step": 116
+    },
+    {
+      "epoch": 0.13183098591549297,
+      "grad_norm": 0.35620398362072353,
+      "learning_rate": 2.1910112359550563e-05,
+      "loss": 0.4953,
+      "step": 117
+    },
+    {
+      "epoch": 0.13295774647887323,
+      "grad_norm": 0.29870939906727606,
+      "learning_rate": 2.209737827715356e-05,
+      "loss": 0.4991,
+      "step": 118
+    },
+    {
+      "epoch": 0.13408450704225353,
+      "grad_norm": 0.34582247341621225,
+      "learning_rate": 2.2284644194756555e-05,
+      "loss": 0.5069,
+      "step": 119
+    },
+    {
+      "epoch": 0.1352112676056338,
+      "grad_norm": 0.3137351594240408,
+      "learning_rate": 2.2471910112359552e-05,
+      "loss": 0.4734,
+      "step": 120
+    },
+    {
+      "epoch": 0.13633802816901408,
+      "grad_norm": 0.3390962431731741,
+      "learning_rate": 2.2659176029962546e-05,
+      "loss": 0.4921,
+      "step": 121
+    },
+    {
+      "epoch": 0.13746478873239437,
+      "grad_norm": 0.35843870700131947,
+      "learning_rate": 2.2846441947565544e-05,
+      "loss": 0.4967,
+      "step": 122
+    },
+    {
+      "epoch": 0.13859154929577464,
+      "grad_norm": 0.3454143874126859,
+      "learning_rate": 2.303370786516854e-05,
+      "loss": 0.4949,
+      "step": 123
+    },
+    {
+      "epoch": 0.13971830985915493,
+      "grad_norm": 0.3793108056331812,
+      "learning_rate": 2.322097378277154e-05,
+      "loss": 0.4919,
+      "step": 124
+    },
+    {
+      "epoch": 0.14084507042253522,
+      "grad_norm": 0.33289196984913244,
+      "learning_rate": 2.3408239700374533e-05,
+      "loss": 0.4913,
+      "step": 125
+    },
+    {
+      "epoch": 0.14197183098591548,
+      "grad_norm": 0.3425874685500103,
+      "learning_rate": 2.359550561797753e-05,
+      "loss": 0.5134,
+      "step": 126
+    },
+    {
+      "epoch": 0.14309859154929577,
+      "grad_norm": 0.35144318220532755,
+      "learning_rate": 2.3782771535580524e-05,
+      "loss": 0.4847,
+      "step": 127
+    },
+    {
+      "epoch": 0.14422535211267606,
+      "grad_norm": 0.36947792672202345,
+      "learning_rate": 2.3970037453183522e-05,
+      "loss": 0.5171,
+      "step": 128
+    },
+    {
+      "epoch": 0.14535211267605633,
+      "grad_norm": 0.3525506964360257,
+      "learning_rate": 2.415730337078652e-05,
+      "loss": 0.5108,
+      "step": 129
+    },
+    {
+      "epoch": 0.14647887323943662,
+      "grad_norm": 0.36146203432070717,
+      "learning_rate": 2.4344569288389517e-05,
+      "loss": 0.5098,
+      "step": 130
+    },
+    {
+      "epoch": 0.1476056338028169,
+      "grad_norm": 0.35620473132880365,
+      "learning_rate": 2.453183520599251e-05,
+      "loss": 0.4795,
+      "step": 131
+    },
+    {
+      "epoch": 0.14873239436619717,
+      "grad_norm": 0.3630204881069924,
+      "learning_rate": 2.4719101123595505e-05,
+      "loss": 0.5187,
+      "step": 132
+    },
+    {
+      "epoch": 0.14985915492957746,
+      "grad_norm": 0.3482086458727476,
+      "learning_rate": 2.4906367041198502e-05,
+      "loss": 0.4977,
+      "step": 133
+    },
+    {
+      "epoch": 0.15098591549295776,
+      "grad_norm": 0.368109870697486,
+      "learning_rate": 2.50936329588015e-05,
+      "loss": 0.5202,
+      "step": 134
+    },
+    {
+      "epoch": 0.15211267605633802,
+      "grad_norm": 0.32473683043927754,
+      "learning_rate": 2.5280898876404497e-05,
+      "loss": 0.491,
+      "step": 135
+    },
+    {
+      "epoch": 0.1532394366197183,
+      "grad_norm": 0.39170978189739086,
+      "learning_rate": 2.546816479400749e-05,
+      "loss": 0.4819,
+      "step": 136
+    },
+    {
+      "epoch": 0.1543661971830986,
+      "grad_norm": 0.33051320119460775,
+      "learning_rate": 2.565543071161049e-05,
+      "loss": 0.4803,
+      "step": 137
+    },
+    {
+      "epoch": 0.15549295774647887,
+      "grad_norm": 0.39201639325298016,
+      "learning_rate": 2.5842696629213486e-05,
+      "loss": 0.4796,
+      "step": 138
+    },
+    {
+      "epoch": 0.15661971830985916,
+      "grad_norm": 0.3436149030584173,
+      "learning_rate": 2.6029962546816484e-05,
+      "loss": 0.4883,
+      "step": 139
+    },
+    {
+      "epoch": 0.15774647887323945,
+      "grad_norm": 0.366083590088758,
+      "learning_rate": 2.6217228464419475e-05,
+      "loss": 0.4894,
+      "step": 140
+    },
+    {
+      "epoch": 0.1588732394366197,
+      "grad_norm": 0.30735207167025824,
+      "learning_rate": 2.6404494382022472e-05,
+      "loss": 0.501,
+      "step": 141
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4232062943090275,
+      "learning_rate": 2.6591760299625466e-05,
+      "loss": 0.5067,
+      "step": 142
+    },
+    {
+      "epoch": 0.1611267605633803,
+      "grad_norm": 0.3322052225992824,
+      "learning_rate": 2.6779026217228464e-05,
+      "loss": 0.4802,
+      "step": 143
+    },
+    {
+      "epoch": 0.16225352112676056,
+      "grad_norm": 0.383878101272465,
+      "learning_rate": 2.696629213483146e-05,
+      "loss": 0.498,
+      "step": 144
+    },
+    {
+      "epoch": 0.16338028169014085,
+      "grad_norm": 0.34217239742624306,
+      "learning_rate": 2.715355805243446e-05,
+      "loss": 0.4757,
+      "step": 145
+    },
+    {
+      "epoch": 0.1645070422535211,
+      "grad_norm": 0.38048598444904064,
+      "learning_rate": 2.7340823970037456e-05,
+      "loss": 0.5133,
+      "step": 146
+    },
+    {
+      "epoch": 0.1656338028169014,
+      "grad_norm": 0.3704985899909714,
+      "learning_rate": 2.752808988764045e-05,
+      "loss": 0.5114,
+      "step": 147
+    },
+    {
+      "epoch": 0.1667605633802817,
+      "grad_norm": 0.347020727924182,
+      "learning_rate": 2.7715355805243448e-05,
+      "loss": 0.4881,
+      "step": 148
+    },
+    {
+      "epoch": 0.16788732394366196,
+      "grad_norm": 0.3636357042062005,
+      "learning_rate": 2.7902621722846445e-05,
+      "loss": 0.4777,
+      "step": 149
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.3634097825691515,
+      "learning_rate": 2.8089887640449443e-05,
+      "loss": 0.4889,
+      "step": 150
+    },
+    {
+      "epoch": 0.17014084507042254,
+      "grad_norm": 0.38132967790819516,
+      "learning_rate": 2.8277153558052437e-05,
+      "loss": 0.4746,
+      "step": 151
+    },
+    {
+      "epoch": 0.1712676056338028,
+      "grad_norm": 0.3685509255363722,
+      "learning_rate": 2.846441947565543e-05,
+      "loss": 0.4751,
+      "step": 152
+    },
+    {
+      "epoch": 0.1723943661971831,
+      "grad_norm": 0.3677251643575248,
+      "learning_rate": 2.8651685393258425e-05,
+      "loss": 0.4847,
+      "step": 153
+    },
+    {
+      "epoch": 0.1735211267605634,
+      "grad_norm": 0.3699177372205316,
+      "learning_rate": 2.8838951310861422e-05,
+      "loss": 0.4738,
+      "step": 154
+    },
+    {
+      "epoch": 0.17464788732394365,
+      "grad_norm": 0.3483871961624536,
+      "learning_rate": 2.902621722846442e-05,
+      "loss": 0.4805,
+      "step": 155
+    },
+    {
+      "epoch": 0.17577464788732394,
+      "grad_norm": 0.3920789722753322,
+      "learning_rate": 2.9213483146067417e-05,
+      "loss": 0.4712,
+      "step": 156
+    },
+    {
+      "epoch": 0.17690140845070423,
+      "grad_norm": 0.3551926481565757,
+      "learning_rate": 2.940074906367041e-05,
+      "loss": 0.5052,
+      "step": 157
+    },
+    {
+      "epoch": 0.1780281690140845,
+      "grad_norm": 0.36364018726777314,
+      "learning_rate": 2.958801498127341e-05,
+      "loss": 0.4584,
+      "step": 158
+    },
+    {
+      "epoch": 0.1791549295774648,
+      "grad_norm": 0.3972954045737472,
+      "learning_rate": 2.9775280898876406e-05,
+      "loss": 0.514,
+      "step": 159
+    },
+    {
+      "epoch": 0.18028169014084508,
+      "grad_norm": 0.4587946650119913,
+      "learning_rate": 2.9962546816479404e-05,
+      "loss": 0.4727,
+      "step": 160
+    },
+    {
+      "epoch": 0.18140845070422534,
+      "grad_norm": 0.351499766618042,
+      "learning_rate": 3.01498127340824e-05,
+      "loss": 0.4992,
+      "step": 161
+    },
+    {
+      "epoch": 0.18253521126760563,
+      "grad_norm": 0.5422721694050262,
+      "learning_rate": 3.0337078651685396e-05,
+      "loss": 0.4764,
+      "step": 162
+    },
+    {
+      "epoch": 0.18366197183098593,
+      "grad_norm": 0.5273905666988745,
+      "learning_rate": 3.052434456928839e-05,
+      "loss": 0.4953,
+      "step": 163
+    },
+    {
+      "epoch": 0.1847887323943662,
+      "grad_norm": 0.399247691576563,
+      "learning_rate": 3.0711610486891384e-05,
+      "loss": 0.4579,
+      "step": 164
+    },
+    {
+      "epoch": 0.18591549295774648,
+      "grad_norm": 0.43479889451138115,
+      "learning_rate": 3.089887640449438e-05,
+      "loss": 0.4643,
+      "step": 165
+    },
+    {
+      "epoch": 0.18704225352112677,
+      "grad_norm": 0.5429791467443568,
+      "learning_rate": 3.108614232209738e-05,
+      "loss": 0.5066,
+      "step": 166
+    },
+    {
+      "epoch": 0.18816901408450704,
+      "grad_norm": 0.4540405451965911,
+      "learning_rate": 3.1273408239700376e-05,
+      "loss": 0.4893,
+      "step": 167
+    },
+    {
+      "epoch": 0.18929577464788733,
+      "grad_norm": 0.5823441022105608,
+      "learning_rate": 3.1460674157303374e-05,
+      "loss": 0.4755,
+      "step": 168
+    },
+    {
+      "epoch": 0.19042253521126762,
+      "grad_norm": 0.7004005424258571,
+      "learning_rate": 3.164794007490637e-05,
+      "loss": 0.4891,
+      "step": 169
+    },
+    {
+      "epoch": 0.19154929577464788,
+      "grad_norm": 0.40931277269381217,
+      "learning_rate": 3.183520599250936e-05,
+      "loss": 0.4499,
+      "step": 170
+    },
+    {
+      "epoch": 0.19267605633802817,
+      "grad_norm": 0.6973058224617746,
+      "learning_rate": 3.202247191011236e-05,
+      "loss": 0.482,
+      "step": 171
+    },
+    {
+      "epoch": 0.19380281690140846,
+      "grad_norm": 0.7957866592242848,
+      "learning_rate": 3.220973782771536e-05,
+      "loss": 0.4844,
+      "step": 172
+    },
+    {
+      "epoch": 0.19492957746478873,
+      "grad_norm": 0.5275061025504003,
+      "learning_rate": 3.2397003745318354e-05,
+      "loss": 0.4652,
+      "step": 173
+    },
+    {
+      "epoch": 0.19605633802816902,
+      "grad_norm": 1.1842902748039579,
+      "learning_rate": 3.258426966292135e-05,
+      "loss": 0.4868,
+      "step": 174
+    },
+    {
+      "epoch": 0.19718309859154928,
+      "grad_norm": 0.6115542815114193,
+      "learning_rate": 3.277153558052435e-05,
+      "loss": 0.4724,
+      "step": 175
+    },
+    {
+      "epoch": 0.19830985915492957,
+      "grad_norm": 0.6492777695364972,
+      "learning_rate": 3.295880149812734e-05,
+      "loss": 0.441,
+      "step": 176
+    },
+    {
+      "epoch": 0.19943661971830987,
+      "grad_norm": 0.5988372578158707,
+      "learning_rate": 3.314606741573034e-05,
+      "loss": 0.4792,
+      "step": 177
+    },
+    {
+      "epoch": 0.20056338028169013,
+      "grad_norm": 0.5653319864850155,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4868,
+      "step": 178
+    },
+    {
+      "epoch": 0.20169014084507042,
+      "grad_norm": 0.5435905257267866,
+      "learning_rate": 3.352059925093633e-05,
+      "loss": 0.4747,
+      "step": 179
+    },
+    {
+      "epoch": 0.2028169014084507,
+      "grad_norm": 0.5572162635083778,
+      "learning_rate": 3.370786516853933e-05,
+      "loss": 0.4752,
+      "step": 180
+    },
+    {
+      "epoch": 0.20394366197183098,
+      "grad_norm": 0.5904914688697619,
+      "learning_rate": 3.389513108614232e-05,
+      "loss": 0.5,
+      "step": 181
+    },
+    {
+      "epoch": 0.20507042253521127,
+      "grad_norm": 0.5718739497949387,
+      "learning_rate": 3.408239700374532e-05,
+      "loss": 0.4834,
+      "step": 182
+    },
+    {
+      "epoch": 0.20619718309859156,
+      "grad_norm": 0.5136731587567704,
+      "learning_rate": 3.4269662921348316e-05,
+      "loss": 0.5129,
+      "step": 183
+    },
+    {
+      "epoch": 0.20732394366197182,
+      "grad_norm": 0.5553306091609278,
+      "learning_rate": 3.445692883895131e-05,
+      "loss": 0.4769,
+      "step": 184
+    },
+    {
+      "epoch": 0.2084507042253521,
+      "grad_norm": 0.5041333309153727,
+      "learning_rate": 3.464419475655431e-05,
+      "loss": 0.4673,
+      "step": 185
+    },
+    {
+      "epoch": 0.2095774647887324,
+      "grad_norm": 0.526418919436349,
+      "learning_rate": 3.483146067415731e-05,
+      "loss": 0.482,
+      "step": 186
+    },
+    {
+      "epoch": 0.21070422535211267,
+      "grad_norm": 0.5365226140722351,
+      "learning_rate": 3.5018726591760305e-05,
+      "loss": 0.4771,
+      "step": 187
+    },
+    {
+      "epoch": 0.21183098591549296,
+      "grad_norm": 0.456310763127666,
+      "learning_rate": 3.52059925093633e-05,
+      "loss": 0.4917,
+      "step": 188
+    },
+    {
+      "epoch": 0.21295774647887325,
+      "grad_norm": 0.5038033687577937,
+      "learning_rate": 3.5393258426966294e-05,
+      "loss": 0.4648,
+      "step": 189
+    },
+    {
+      "epoch": 0.2140845070422535,
+      "grad_norm": 0.3999119007137244,
+      "learning_rate": 3.558052434456929e-05,
+      "loss": 0.4736,
+      "step": 190
+    },
+    {
+      "epoch": 0.2152112676056338,
+      "grad_norm": 0.5384182633924429,
+      "learning_rate": 3.576779026217228e-05,
+      "loss": 0.4614,
+      "step": 191
+    },
+    {
+      "epoch": 0.2163380281690141,
+      "grad_norm": 0.3938798987316564,
+      "learning_rate": 3.595505617977528e-05,
+      "loss": 0.4824,
+      "step": 192
+    },
+    {
+      "epoch": 0.21746478873239436,
+      "grad_norm": 0.47825665590170224,
+      "learning_rate": 3.614232209737828e-05,
+      "loss": 0.4944,
+      "step": 193
+    },
+    {
+      "epoch": 0.21859154929577465,
+      "grad_norm": 0.4739210393787863,
+      "learning_rate": 3.6329588014981274e-05,
+      "loss": 0.4854,
+      "step": 194
+    },
+    {
+      "epoch": 0.21971830985915494,
+      "grad_norm": 0.4000560342009286,
+      "learning_rate": 3.651685393258427e-05,
+      "loss": 0.4537,
+      "step": 195
+    },
+    {
+      "epoch": 0.2208450704225352,
+      "grad_norm": 0.4745488866447151,
+      "learning_rate": 3.670411985018727e-05,
+      "loss": 0.4925,
+      "step": 196
+    },
+    {
+      "epoch": 0.2219718309859155,
+      "grad_norm": 0.40427608903936857,
+      "learning_rate": 3.689138576779027e-05,
+      "loss": 0.4759,
+      "step": 197
+    },
+    {
+      "epoch": 0.2230985915492958,
+      "grad_norm": 0.36830227225633555,
+      "learning_rate": 3.7078651685393264e-05,
+      "loss": 0.4736,
+      "step": 198
+    },
+    {
+      "epoch": 0.22422535211267605,
+      "grad_norm": 0.35731267597771954,
+      "learning_rate": 3.726591760299626e-05,
+      "loss": 0.4655,
+      "step": 199
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.36040514424773884,
+      "learning_rate": 3.745318352059925e-05,
+      "loss": 0.5123,
+      "step": 200
+    },
+    {
+      "epoch": 0.2264788732394366,
+      "grad_norm": 0.341416480034844,
+      "learning_rate": 3.764044943820225e-05,
+      "loss": 0.4817,
+      "step": 201
+    },
+    {
+      "epoch": 0.2276056338028169,
+      "grad_norm": 0.38281780990844233,
+      "learning_rate": 3.782771535580524e-05,
+      "loss": 0.4464,
+      "step": 202
+    },
+    {
+      "epoch": 0.2287323943661972,
+      "grad_norm": 0.39091583875242464,
+      "learning_rate": 3.801498127340824e-05,
+      "loss": 0.459,
+      "step": 203
+    },
+    {
+      "epoch": 0.22985915492957745,
+      "grad_norm": 0.4484240195809876,
+      "learning_rate": 3.8202247191011236e-05,
+      "loss": 0.4722,
+      "step": 204
+    },
+    {
+      "epoch": 0.23098591549295774,
+      "grad_norm": 0.3557165364605743,
+      "learning_rate": 3.838951310861423e-05,
+      "loss": 0.4481,
+      "step": 205
+    },
+    {
+      "epoch": 0.23211267605633804,
+      "grad_norm": 0.4351009901681479,
+      "learning_rate": 3.857677902621723e-05,
+      "loss": 0.471,
+      "step": 206
+    },
+    {
+      "epoch": 0.2332394366197183,
+      "grad_norm": 0.38692072509933034,
+      "learning_rate": 3.876404494382023e-05,
+      "loss": 0.4286,
+      "step": 207
+    },
+    {
+      "epoch": 0.2343661971830986,
+      "grad_norm": 0.4675415655077608,
+      "learning_rate": 3.8951310861423226e-05,
+      "loss": 0.4576,
+      "step": 208
+    },
+    {
+      "epoch": 0.23549295774647888,
+      "grad_norm": 0.39761524002543014,
+      "learning_rate": 3.913857677902622e-05,
+      "loss": 0.4794,
+      "step": 209
+    },
+    {
+      "epoch": 0.23661971830985915,
+      "grad_norm": 0.4540277389661571,
+      "learning_rate": 3.9325842696629214e-05,
+      "loss": 0.4759,
+      "step": 210
+    },
+    {
+      "epoch": 0.23774647887323944,
+      "grad_norm": 0.4440657059092484,
+      "learning_rate": 3.951310861423221e-05,
+      "loss": 0.4534,
+      "step": 211
+    },
+    {
+      "epoch": 0.23887323943661973,
+      "grad_norm": 0.3939795789051039,
+      "learning_rate": 3.970037453183521e-05,
+      "loss": 0.4612,
+      "step": 212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4762851872749466,
+      "learning_rate": 3.98876404494382e-05,
+      "loss": 0.4566,
+      "step": 213
+    },
+    {
+      "epoch": 0.24112676056338028,
+      "grad_norm": 0.41770476809342594,
+      "learning_rate": 4.00749063670412e-05,
+      "loss": 0.4532,
+      "step": 214
+    },
+    {
+      "epoch": 0.24225352112676057,
+      "grad_norm": 0.48777063502172047,
+      "learning_rate": 4.0262172284644194e-05,
+      "loss": 0.4432,
+      "step": 215
+    },
+    {
+      "epoch": 0.24338028169014084,
+      "grad_norm": 0.3690439757602066,
+      "learning_rate": 4.044943820224719e-05,
+      "loss": 0.4408,
+      "step": 216
+    },
+    {
+      "epoch": 0.24450704225352113,
+      "grad_norm": 0.49741320292973695,
+      "learning_rate": 4.063670411985019e-05,
+      "loss": 0.4659,
+      "step": 217
+    },
+    {
+      "epoch": 0.24563380281690142,
+      "grad_norm": 0.522242568777597,
+      "learning_rate": 4.082397003745319e-05,
+      "loss": 0.4474,
+      "step": 218
+    },
+    {
+      "epoch": 0.24676056338028168,
+      "grad_norm": 0.37597023810209546,
+      "learning_rate": 4.1011235955056184e-05,
+      "loss": 0.4863,
+      "step": 219
+    },
+    {
+      "epoch": 0.24788732394366197,
+      "grad_norm": 0.5868290025238638,
+      "learning_rate": 4.119850187265918e-05,
+      "loss": 0.4829,
+      "step": 220
+    },
+    {
+      "epoch": 0.24901408450704227,
+      "grad_norm": 0.5148405897733493,
+      "learning_rate": 4.138576779026217e-05,
+      "loss": 0.4518,
+      "step": 221
+    },
+    {
+      "epoch": 0.25014084507042256,
+      "grad_norm": 0.44695171811561535,
+      "learning_rate": 4.157303370786517e-05,
+      "loss": 0.4548,
+      "step": 222
+    },
+    {
+      "epoch": 0.2512676056338028,
+      "grad_norm": 0.5467900665481616,
+      "learning_rate": 4.176029962546817e-05,
+      "loss": 0.4716,
+      "step": 223
+    },
+    {
+      "epoch": 0.2523943661971831,
+      "grad_norm": 0.41846730812607025,
+      "learning_rate": 4.1947565543071165e-05,
+      "loss": 0.4661,
+      "step": 224
+    },
+    {
+      "epoch": 0.2535211267605634,
+      "grad_norm": 0.45662423635892097,
+      "learning_rate": 4.2134831460674156e-05,
+      "loss": 0.4534,
+      "step": 225
+    },
+    {
+      "epoch": 0.25464788732394367,
+      "grad_norm": 0.6325159390990795,
+      "learning_rate": 4.232209737827715e-05,
+      "loss": 0.4512,
+      "step": 226
+    },
+    {
+      "epoch": 0.25577464788732396,
+      "grad_norm": 0.3891612621892786,
+      "learning_rate": 4.250936329588015e-05,
+      "loss": 0.456,
+      "step": 227
+    },
+    {
+      "epoch": 0.25690140845070425,
+      "grad_norm": 0.49703727890231636,
+      "learning_rate": 4.269662921348315e-05,
+      "loss": 0.4563,
+      "step": 228
+    },
+    {
+      "epoch": 0.2580281690140845,
+      "grad_norm": 0.4866546409272407,
+      "learning_rate": 4.2883895131086146e-05,
+      "loss": 0.4717,
+      "step": 229
+    },
+    {
+      "epoch": 0.2591549295774648,
+      "grad_norm": 0.4531740529278661,
+      "learning_rate": 4.307116104868914e-05,
+      "loss": 0.4446,
+      "step": 230
+    },
+    {
+      "epoch": 0.26028169014084507,
+      "grad_norm": 0.5386248688297244,
+      "learning_rate": 4.3258426966292134e-05,
+      "loss": 0.4674,
+      "step": 231
+    },
+    {
+      "epoch": 0.26140845070422536,
+      "grad_norm": 0.42614317035006366,
+      "learning_rate": 4.344569288389513e-05,
+      "loss": 0.4583,
+      "step": 232
+    },
+    {
+      "epoch": 0.26253521126760565,
+      "grad_norm": 0.5720498969311514,
+      "learning_rate": 4.363295880149813e-05,
+      "loss": 0.4501,
+      "step": 233
+    },
+    {
+      "epoch": 0.26366197183098594,
+      "grad_norm": 0.5617737038927428,
+      "learning_rate": 4.3820224719101126e-05,
+      "loss": 0.4759,
+      "step": 234
+    },
+    {
+      "epoch": 0.2647887323943662,
+      "grad_norm": 0.4432550193181308,
+      "learning_rate": 4.4007490636704124e-05,
+      "loss": 0.4658,
+      "step": 235
+    },
+    {
+      "epoch": 0.26591549295774647,
+      "grad_norm": 0.5733406217420002,
+      "learning_rate": 4.419475655430712e-05,
+      "loss": 0.4739,
+      "step": 236
+    },
+    {
+      "epoch": 0.26704225352112676,
+      "grad_norm": 0.4228087393204907,
+      "learning_rate": 4.438202247191011e-05,
+      "loss": 0.4549,
+      "step": 237
+    },
+    {
+      "epoch": 0.26816901408450705,
+      "grad_norm": 0.4865503254588127,
+      "learning_rate": 4.456928838951311e-05,
+      "loss": 0.4413,
+      "step": 238
+    },
+    {
+      "epoch": 0.26929577464788734,
+      "grad_norm": 0.5412922689255075,
+      "learning_rate": 4.475655430711611e-05,
+      "loss": 0.4627,
+      "step": 239
+    },
+    {
+      "epoch": 0.2704225352112676,
+      "grad_norm": 0.42876550267809127,
+      "learning_rate": 4.4943820224719104e-05,
+      "loss": 0.4547,
+      "step": 240
+    },
+    {
+      "epoch": 0.27154929577464787,
+      "grad_norm": 0.5373861362677359,
+      "learning_rate": 4.51310861423221e-05,
+      "loss": 0.4612,
+      "step": 241
+    },
+    {
+      "epoch": 0.27267605633802816,
+      "grad_norm": 0.5804462215718947,
+      "learning_rate": 4.531835205992509e-05,
+      "loss": 0.4645,
+      "step": 242
+    },
+    {
+      "epoch": 0.27380281690140845,
+      "grad_norm": 0.4122842878113385,
+      "learning_rate": 4.550561797752809e-05,
+      "loss": 0.4451,
+      "step": 243
+    },
+    {
+      "epoch": 0.27492957746478874,
+      "grad_norm": 0.5336968424824022,
+      "learning_rate": 4.569288389513109e-05,
+      "loss": 0.4481,
+      "step": 244
+    },
+    {
+      "epoch": 0.27605633802816903,
+      "grad_norm": 0.42951583225710527,
+      "learning_rate": 4.5880149812734085e-05,
+      "loss": 0.4567,
+      "step": 245
+    },
+    {
+      "epoch": 0.27718309859154927,
+      "grad_norm": 0.48609553240950315,
+      "learning_rate": 4.606741573033708e-05,
+      "loss": 0.447,
+      "step": 246
+    },
+    {
+      "epoch": 0.27830985915492956,
+      "grad_norm": 0.5840157329385284,
+      "learning_rate": 4.625468164794008e-05,
+      "loss": 0.4683,
+      "step": 247
+    },
+    {
+      "epoch": 0.27943661971830985,
+      "grad_norm": 0.464308042264615,
+      "learning_rate": 4.644194756554308e-05,
+      "loss": 0.4435,
+      "step": 248
+    },
+    {
+      "epoch": 0.28056338028169014,
+      "grad_norm": 0.48568184619644433,
+      "learning_rate": 4.662921348314607e-05,
+      "loss": 0.4202,
+      "step": 249
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.5421087431500514,
+      "learning_rate": 4.6816479400749066e-05,
+      "loss": 0.4802,
+      "step": 250
+    },
+    {
+      "epoch": 0.2828169014084507,
+      "grad_norm": 0.5726166139181046,
+      "learning_rate": 4.700374531835206e-05,
+      "loss": 0.4082,
+      "step": 251
+    },
+    {
+      "epoch": 0.28394366197183096,
+      "grad_norm": 0.38708381470148295,
+      "learning_rate": 4.719101123595506e-05,
+      "loss": 0.4201,
+      "step": 252
+    },
+    {
+      "epoch": 0.28507042253521125,
+      "grad_norm": 0.6599296604136585,
+      "learning_rate": 4.737827715355805e-05,
+      "loss": 0.4574,
+      "step": 253
+    },
+    {
+      "epoch": 0.28619718309859155,
+      "grad_norm": 0.3689919503814012,
+      "learning_rate": 4.756554307116105e-05,
+      "loss": 0.443,
+      "step": 254
+    },
+    {
+      "epoch": 0.28732394366197184,
+      "grad_norm": 0.6554671527107907,
+      "learning_rate": 4.7752808988764046e-05,
+      "loss": 0.4521,
+      "step": 255
+    },
+    {
+      "epoch": 0.28845070422535213,
+      "grad_norm": 0.4732513325463823,
+      "learning_rate": 4.7940074906367044e-05,
+      "loss": 0.4612,
+      "step": 256
+    },
+    {
+      "epoch": 0.2895774647887324,
+      "grad_norm": 0.6422946109709861,
+      "learning_rate": 4.812734082397004e-05,
+      "loss": 0.4592,
+      "step": 257
+    },
+    {
+      "epoch": 0.29070422535211266,
+      "grad_norm": 0.5006208962628628,
+      "learning_rate": 4.831460674157304e-05,
+      "loss": 0.4591,
+      "step": 258
+    },
+    {
+      "epoch": 0.29183098591549295,
+      "grad_norm": 0.6352179330776139,
+      "learning_rate": 4.8501872659176036e-05,
+      "loss": 0.4421,
+      "step": 259
+    },
+    {
+      "epoch": 0.29295774647887324,
+      "grad_norm": 0.47815163623650375,
+      "learning_rate": 4.8689138576779034e-05,
+      "loss": 0.4416,
+      "step": 260
+    },
+    {
+      "epoch": 0.29408450704225353,
+      "grad_norm": 0.5761319167007566,
+      "learning_rate": 4.8876404494382024e-05,
+      "loss": 0.4423,
+      "step": 261
+    },
+    {
+      "epoch": 0.2952112676056338,
+      "grad_norm": 0.664916170697062,
+      "learning_rate": 4.906367041198502e-05,
+      "loss": 0.4752,
+      "step": 262
+    },
+    {
+      "epoch": 0.2963380281690141,
+      "grad_norm": 0.4686347528012791,
+      "learning_rate": 4.925093632958801e-05,
+      "loss": 0.4549,
+      "step": 263
+    },
+    {
+      "epoch": 0.29746478873239435,
+      "grad_norm": 0.6535509019931558,
+      "learning_rate": 4.943820224719101e-05,
+      "loss": 0.4185,
+      "step": 264
+    },
+    {
+      "epoch": 0.29859154929577464,
+      "grad_norm": 0.4418063367523136,
+      "learning_rate": 4.962546816479401e-05,
+      "loss": 0.445,
+      "step": 265
+    },
+    {
+      "epoch": 0.29971830985915493,
+      "grad_norm": 0.4780381630674905,
+      "learning_rate": 4.9812734082397005e-05,
+      "loss": 0.4396,
+      "step": 266
+    },
+    {
+      "epoch": 0.3008450704225352,
+      "grad_norm": 0.5799759746220542,
+      "learning_rate": 5e-05,
+      "loss": 0.4581,
+      "step": 267
+    },
+    {
+      "epoch": 0.3019718309859155,
+      "grad_norm": 0.5140227732840926,
+      "learning_rate": 4.997911445279866e-05,
+      "loss": 0.4258,
+      "step": 268
+    },
+    {
+      "epoch": 0.30309859154929575,
+      "grad_norm": 0.5093369543361972,
+      "learning_rate": 4.995822890559733e-05,
+      "loss": 0.447,
+      "step": 269
+    },
+    {
+      "epoch": 0.30422535211267604,
+      "grad_norm": 0.536665447881301,
+      "learning_rate": 4.9937343358395996e-05,
+      "loss": 0.4375,
+      "step": 270
+    },
+    {
+      "epoch": 0.30535211267605633,
+      "grad_norm": 0.6045725882311949,
+      "learning_rate": 4.9916457811194656e-05,
+      "loss": 0.4596,
+      "step": 271
+    },
+    {
+      "epoch": 0.3064788732394366,
+      "grad_norm": 0.4513638559820733,
+      "learning_rate": 4.9895572263993316e-05,
+      "loss": 0.4213,
+      "step": 272
+    },
+    {
+      "epoch": 0.3076056338028169,
+      "grad_norm": 0.5579346841610208,
+      "learning_rate": 4.987468671679198e-05,
+      "loss": 0.4496,
+      "step": 273
+    },
+    {
+      "epoch": 0.3087323943661972,
+      "grad_norm": 0.721853421554717,
+      "learning_rate": 4.985380116959065e-05,
+      "loss": 0.4447,
+      "step": 274
+    },
+    {
+      "epoch": 0.30985915492957744,
+      "grad_norm": 0.6403095657433422,
+      "learning_rate": 4.983291562238931e-05,
+      "loss": 0.4502,
+      "step": 275
+    },
+    {
+      "epoch": 0.31098591549295773,
+      "grad_norm": 0.6518049489007518,
+      "learning_rate": 4.981203007518797e-05,
+      "loss": 0.4632,
+      "step": 276
+    },
+    {
+      "epoch": 0.312112676056338,
+      "grad_norm": 0.8147228654510137,
+      "learning_rate": 4.9791144527986636e-05,
+      "loss": 0.4759,
+      "step": 277
+    },
+    {
+      "epoch": 0.3132394366197183,
+      "grad_norm": 0.9543881154945372,
+      "learning_rate": 4.97702589807853e-05,
+      "loss": 0.466,
+      "step": 278
+    },
+    {
+      "epoch": 0.3143661971830986,
+      "grad_norm": 0.711807527160549,
+      "learning_rate": 4.974937343358396e-05,
+      "loss": 0.4458,
+      "step": 279
+    },
+    {
+      "epoch": 0.3154929577464789,
+      "grad_norm": 0.9703166053609333,
+      "learning_rate": 4.972848788638262e-05,
+      "loss": 0.4548,
+      "step": 280
+    },
+    {
+      "epoch": 0.31661971830985913,
+      "grad_norm": 0.8568539146313636,
+      "learning_rate": 4.970760233918128e-05,
+      "loss": 0.4397,
+      "step": 281
+    },
+    {
+      "epoch": 0.3177464788732394,
+      "grad_norm": 0.8062901175799372,
+      "learning_rate": 4.9686716791979956e-05,
+      "loss": 0.4529,
+      "step": 282
+    },
+    {
+      "epoch": 0.3188732394366197,
+      "grad_norm": 0.8565696839712605,
+      "learning_rate": 4.9665831244778616e-05,
+      "loss": 0.4577,
+      "step": 283
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9299611575587775,
+      "learning_rate": 4.9644945697577276e-05,
+      "loss": 0.4351,
+      "step": 284
+    },
+    {
+      "epoch": 0.3211267605633803,
+      "grad_norm": 0.7290633308595302,
+      "learning_rate": 4.9624060150375936e-05,
+      "loss": 0.4337,
+      "step": 285
+    },
+    {
+      "epoch": 0.3222535211267606,
+      "grad_norm": 0.7827462663607737,
+      "learning_rate": 4.960317460317461e-05,
+      "loss": 0.4419,
+      "step": 286
+    },
+    {
+      "epoch": 0.3233802816901408,
+      "grad_norm": 0.8289797573873895,
+      "learning_rate": 4.958228905597327e-05,
+      "loss": 0.4313,
+      "step": 287
+    },
+    {
+      "epoch": 0.3245070422535211,
+      "grad_norm": 0.44578585581376295,
+      "learning_rate": 4.956140350877193e-05,
+      "loss": 0.4597,
+      "step": 288
+    },
+    {
+      "epoch": 0.3256338028169014,
+      "grad_norm": 0.8766044076142936,
+      "learning_rate": 4.954051796157059e-05,
+      "loss": 0.4812,
+      "step": 289
+    },
+    {
+      "epoch": 0.3267605633802817,
+      "grad_norm": 0.4914683054076262,
+      "learning_rate": 4.9519632414369263e-05,
+      "loss": 0.4461,
+      "step": 290
+    },
+    {
+      "epoch": 0.327887323943662,
+      "grad_norm": 0.7848770318839384,
+      "learning_rate": 4.9498746867167923e-05,
+      "loss": 0.4615,
+      "step": 291
+    },
+    {
+      "epoch": 0.3290140845070422,
+      "grad_norm": 0.5878732648655055,
+      "learning_rate": 4.947786131996658e-05,
+      "loss": 0.4487,
+      "step": 292
+    },
+    {
+      "epoch": 0.3301408450704225,
+      "grad_norm": 0.6268082137586906,
+      "learning_rate": 4.945697577276524e-05,
+      "loss": 0.4317,
+      "step": 293
+    },
+    {
+      "epoch": 0.3312676056338028,
+      "grad_norm": 0.5757058247458672,
+      "learning_rate": 4.943609022556392e-05,
+      "loss": 0.4319,
+      "step": 294
+    },
+    {
+      "epoch": 0.3323943661971831,
+      "grad_norm": 0.5478319671642797,
+      "learning_rate": 4.941520467836258e-05,
+      "loss": 0.4414,
+      "step": 295
+    },
+    {
+      "epoch": 0.3335211267605634,
+      "grad_norm": 0.6073166154579848,
+      "learning_rate": 4.939431913116124e-05,
+      "loss": 0.437,
+      "step": 296
+    },
+    {
+      "epoch": 0.3346478873239437,
+      "grad_norm": 0.5288488847365398,
+      "learning_rate": 4.93734335839599e-05,
+      "loss": 0.4634,
+      "step": 297
+    },
+    {
+      "epoch": 0.3357746478873239,
+      "grad_norm": 0.665813028870457,
+      "learning_rate": 4.9352548036758564e-05,
+      "loss": 0.4169,
+      "step": 298
+    },
+    {
+      "epoch": 0.3369014084507042,
+      "grad_norm": 0.5021503452806892,
+      "learning_rate": 4.933166248955723e-05,
+      "loss": 0.4362,
+      "step": 299
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.62033447140205,
+      "learning_rate": 4.931077694235589e-05,
+      "loss": 0.4563,
+      "step": 300
+    },
+    {
+      "epoch": 0.3391549295774648,
+      "grad_norm": 0.5274838444890766,
+      "learning_rate": 4.928989139515455e-05,
+      "loss": 0.4441,
+      "step": 301
+    },
+    {
+      "epoch": 0.3402816901408451,
+      "grad_norm": 0.612414648195828,
+      "learning_rate": 4.926900584795322e-05,
+      "loss": 0.4527,
+      "step": 302
+    },
+    {
+      "epoch": 0.3414084507042254,
+      "grad_norm": 0.575726524433347,
+      "learning_rate": 4.9248120300751884e-05,
+      "loss": 0.4618,
+      "step": 303
+    },
+    {
+      "epoch": 0.3425352112676056,
+      "grad_norm": 0.4857128803230017,
+      "learning_rate": 4.9227234753550544e-05,
+      "loss": 0.455,
+      "step": 304
+    },
+    {
+      "epoch": 0.3436619718309859,
+      "grad_norm": 0.4651119564318662,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 0.4415,
+      "step": 305
+    },
+    {
+      "epoch": 0.3447887323943662,
+      "grad_norm": 0.43833590031904573,
+      "learning_rate": 4.918546365914787e-05,
+      "loss": 0.4461,
+      "step": 306
+    },
+    {
+      "epoch": 0.3459154929577465,
+      "grad_norm": 0.39155081982450585,
+      "learning_rate": 4.916457811194654e-05,
+      "loss": 0.4385,
+      "step": 307
+    },
+    {
+      "epoch": 0.3470422535211268,
+      "grad_norm": 0.4515617162267813,
+      "learning_rate": 4.91436925647452e-05,
+      "loss": 0.4397,
+      "step": 308
+    },
+    {
+      "epoch": 0.34816901408450707,
+      "grad_norm": 0.39034501069443966,
+      "learning_rate": 4.912280701754386e-05,
+      "loss": 0.4275,
+      "step": 309
+    },
+    {
+      "epoch": 0.3492957746478873,
+      "grad_norm": 0.434941416892916,
+      "learning_rate": 4.9101921470342524e-05,
+      "loss": 0.4516,
+      "step": 310
+    },
+    {
+      "epoch": 0.3504225352112676,
+      "grad_norm": 0.405450124656521,
+      "learning_rate": 4.908103592314119e-05,
+      "loss": 0.4444,
+      "step": 311
+    },
+    {
+      "epoch": 0.3515492957746479,
+      "grad_norm": 0.3890876078168863,
+      "learning_rate": 4.906015037593985e-05,
+      "loss": 0.4337,
+      "step": 312
+    },
+    {
+      "epoch": 0.3526760563380282,
+      "grad_norm": 0.4368474819090546,
+      "learning_rate": 4.903926482873851e-05,
+      "loss": 0.4465,
+      "step": 313
+    },
+    {
+      "epoch": 0.35380281690140847,
+      "grad_norm": 0.49321100467381773,
+      "learning_rate": 4.901837928153718e-05,
+      "loss": 0.4418,
+      "step": 314
+    },
+    {
+      "epoch": 0.35492957746478876,
+      "grad_norm": 0.4007160642334723,
+      "learning_rate": 4.8997493734335844e-05,
+      "loss": 0.4354,
+      "step": 315
+    },
+    {
+      "epoch": 0.356056338028169,
+      "grad_norm": 0.34955955912720427,
+      "learning_rate": 4.8976608187134504e-05,
+      "loss": 0.4651,
+      "step": 316
+    },
+    {
+      "epoch": 0.3571830985915493,
+      "grad_norm": 0.45158037335045736,
+      "learning_rate": 4.8955722639933164e-05,
+      "loss": 0.4476,
+      "step": 317
+    },
+    {
+      "epoch": 0.3583098591549296,
+      "grad_norm": 0.3750679094506128,
+      "learning_rate": 4.893483709273183e-05,
+      "loss": 0.4296,
+      "step": 318
+    },
+    {
+      "epoch": 0.35943661971830987,
+      "grad_norm": 0.3949654475565995,
+      "learning_rate": 4.89139515455305e-05,
+      "loss": 0.4274,
+      "step": 319
+    },
+    {
+      "epoch": 0.36056338028169016,
+      "grad_norm": 0.4279246749378893,
+      "learning_rate": 4.889306599832916e-05,
+      "loss": 0.4531,
+      "step": 320
+    },
+    {
+      "epoch": 0.3616901408450704,
+      "grad_norm": 0.4464764928096698,
+      "learning_rate": 4.887218045112782e-05,
+      "loss": 0.4626,
+      "step": 321
+    },
+    {
+      "epoch": 0.3628169014084507,
+      "grad_norm": 0.4094055448948909,
+      "learning_rate": 4.8851294903926485e-05,
+      "loss": 0.4574,
+      "step": 322
+    },
+    {
+      "epoch": 0.363943661971831,
+      "grad_norm": 0.433464486212505,
+      "learning_rate": 4.883040935672515e-05,
+      "loss": 0.4419,
+      "step": 323
+    },
+    {
+      "epoch": 0.36507042253521127,
+      "grad_norm": 0.3943121841578179,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 0.4379,
+      "step": 324
+    },
+    {
+      "epoch": 0.36619718309859156,
+      "grad_norm": 0.43899232439651487,
+      "learning_rate": 4.878863826232247e-05,
+      "loss": 0.4465,
+      "step": 325
+    },
+    {
+      "epoch": 0.36732394366197185,
+      "grad_norm": 0.4193670484154254,
+      "learning_rate": 4.876775271512114e-05,
+      "loss": 0.4518,
+      "step": 326
+    },
+    {
+      "epoch": 0.3684507042253521,
+      "grad_norm": 0.4047850180694318,
+      "learning_rate": 4.8746867167919805e-05,
+      "loss": 0.4237,
+      "step": 327
+    },
+    {
+      "epoch": 0.3695774647887324,
+      "grad_norm": 0.32930872482740814,
+      "learning_rate": 4.8725981620718465e-05,
+      "loss": 0.438,
+      "step": 328
+    },
+    {
+      "epoch": 0.37070422535211267,
+      "grad_norm": 0.4319956598214786,
+      "learning_rate": 4.8705096073517125e-05,
+      "loss": 0.4342,
+      "step": 329
+    },
+    {
+      "epoch": 0.37183098591549296,
+      "grad_norm": 0.39983873887542376,
+      "learning_rate": 4.868421052631579e-05,
+      "loss": 0.4323,
+      "step": 330
+    },
+    {
+      "epoch": 0.37295774647887325,
+      "grad_norm": 0.42371687498447236,
+      "learning_rate": 4.866332497911446e-05,
+      "loss": 0.4601,
+      "step": 331
+    },
+    {
+      "epoch": 0.37408450704225354,
+      "grad_norm": 0.48052857625941175,
+      "learning_rate": 4.864243943191312e-05,
+      "loss": 0.4791,
+      "step": 332
+    },
+    {
+      "epoch": 0.3752112676056338,
+      "grad_norm": 0.40445062983039526,
+      "learning_rate": 4.862155388471178e-05,
+      "loss": 0.457,
+      "step": 333
+    },
+    {
+      "epoch": 0.37633802816901407,
+      "grad_norm": 0.4419492432404764,
+      "learning_rate": 4.8600668337510445e-05,
+      "loss": 0.4396,
+      "step": 334
+    },
+    {
+      "epoch": 0.37746478873239436,
+      "grad_norm": 0.3831902844997573,
+      "learning_rate": 4.857978279030911e-05,
+      "loss": 0.4775,
+      "step": 335
+    },
+    {
+      "epoch": 0.37859154929577465,
+      "grad_norm": 0.46441513775498133,
+      "learning_rate": 4.855889724310777e-05,
+      "loss": 0.4292,
+      "step": 336
+    },
+    {
+      "epoch": 0.37971830985915495,
+      "grad_norm": 0.4558569399558645,
+      "learning_rate": 4.853801169590643e-05,
+      "loss": 0.4321,
+      "step": 337
+    },
+    {
+      "epoch": 0.38084507042253524,
+      "grad_norm": 0.46584322528549255,
+      "learning_rate": 4.85171261487051e-05,
+      "loss": 0.4381,
+      "step": 338
+    },
+    {
+      "epoch": 0.3819718309859155,
+      "grad_norm": 0.4477280961298829,
+      "learning_rate": 4.849624060150376e-05,
+      "loss": 0.4149,
+      "step": 339
+    },
+    {
+      "epoch": 0.38309859154929576,
+      "grad_norm": 0.4000418205043348,
+      "learning_rate": 4.8475355054302425e-05,
+      "loss": 0.4278,
+      "step": 340
+    },
+    {
+      "epoch": 0.38422535211267606,
+      "grad_norm": 0.43292724989992054,
+      "learning_rate": 4.8454469507101085e-05,
+      "loss": 0.4582,
+      "step": 341
+    },
+    {
+      "epoch": 0.38535211267605635,
+      "grad_norm": 0.6115776129330851,
+      "learning_rate": 4.843358395989975e-05,
+      "loss": 0.4502,
+      "step": 342
+    },
+    {
+      "epoch": 0.38647887323943664,
+      "grad_norm": 0.4753406969860461,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.4546,
+      "step": 343
+    },
+    {
+      "epoch": 0.38760563380281693,
+      "grad_norm": 0.6453081760079101,
+      "learning_rate": 4.839181286549708e-05,
+      "loss": 0.4647,
+      "step": 344
+    },
+    {
+      "epoch": 0.38873239436619716,
+      "grad_norm": 0.6874748404475204,
+      "learning_rate": 4.837092731829574e-05,
+      "loss": 0.4372,
+      "step": 345
+    },
+    {
+      "epoch": 0.38985915492957746,
+      "grad_norm": 0.40443377712949796,
+      "learning_rate": 4.8350041771094406e-05,
+      "loss": 0.4357,
+      "step": 346
+    },
+    {
+      "epoch": 0.39098591549295775,
+      "grad_norm": 0.5991286906041724,
+      "learning_rate": 4.8329156223893066e-05,
+      "loss": 0.4635,
+      "step": 347
+    },
+    {
+      "epoch": 0.39211267605633804,
+      "grad_norm": 0.4722265848003038,
+      "learning_rate": 4.830827067669173e-05,
+      "loss": 0.4489,
+      "step": 348
+    },
+    {
+      "epoch": 0.39323943661971833,
+      "grad_norm": 0.5058106439502263,
+      "learning_rate": 4.828738512949039e-05,
+      "loss": 0.4393,
+      "step": 349
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.5121245224918637,
+      "learning_rate": 4.826649958228906e-05,
+      "loss": 0.4495,
+      "step": 350
+    },
+    {
+      "epoch": 0.39549295774647886,
+      "grad_norm": 0.45671179599721246,
+      "learning_rate": 4.824561403508772e-05,
+      "loss": 0.4507,
+      "step": 351
+    },
+    {
+      "epoch": 0.39661971830985915,
+      "grad_norm": 0.45937663080556684,
+      "learning_rate": 4.8224728487886386e-05,
+      "loss": 0.4209,
+      "step": 352
+    },
+    {
+      "epoch": 0.39774647887323944,
+      "grad_norm": 0.6104067835573581,
+      "learning_rate": 4.8203842940685046e-05,
+      "loss": 0.4418,
+      "step": 353
+    },
+    {
+      "epoch": 0.39887323943661973,
+      "grad_norm": 0.420277131165378,
+      "learning_rate": 4.818295739348371e-05,
+      "loss": 0.4243,
+      "step": 354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5086625172059411,
+      "learning_rate": 4.816207184628237e-05,
+      "loss": 0.443,
+      "step": 355
+    },
+    {
+      "epoch": 0.40112676056338026,
+      "grad_norm": 0.4470016709990632,
+      "learning_rate": 4.814118629908104e-05,
+      "loss": 0.4179,
+      "step": 356
+    },
+    {
+      "epoch": 0.40225352112676055,
+      "grad_norm": 0.44297926877619226,
+      "learning_rate": 4.81203007518797e-05,
+      "loss": 0.4402,
+      "step": 357
+    },
+    {
+      "epoch": 0.40338028169014084,
+      "grad_norm": 0.4010493977276293,
+      "learning_rate": 4.8099415204678366e-05,
+      "loss": 0.4588,
+      "step": 358
+    },
+    {
+      "epoch": 0.40450704225352113,
+      "grad_norm": 0.3506045656678495,
+      "learning_rate": 4.8078529657477026e-05,
+      "loss": 0.4151,
+      "step": 359
+    },
+    {
+      "epoch": 0.4056338028169014,
+      "grad_norm": 0.41248134787582563,
+      "learning_rate": 4.805764411027569e-05,
+      "loss": 0.4308,
+      "step": 360
+    },
+    {
+      "epoch": 0.4067605633802817,
+      "grad_norm": 0.42669347587090345,
+      "learning_rate": 4.803675856307435e-05,
+      "loss": 0.4395,
+      "step": 361
+    },
+    {
+      "epoch": 0.40788732394366195,
+      "grad_norm": 0.37272375532810725,
+      "learning_rate": 4.801587301587302e-05,
+      "loss": 0.4364,
+      "step": 362
+    },
+    {
+      "epoch": 0.40901408450704224,
+      "grad_norm": 0.4570821983785425,
+      "learning_rate": 4.799498746867168e-05,
+      "loss": 0.4418,
+      "step": 363
+    },
+    {
+      "epoch": 0.41014084507042253,
+      "grad_norm": 0.398872273320948,
+      "learning_rate": 4.7974101921470346e-05,
+      "loss": 0.4436,
+      "step": 364
+    },
+    {
+      "epoch": 0.4112676056338028,
+      "grad_norm": 0.511403087399427,
+      "learning_rate": 4.7953216374269006e-05,
+      "loss": 0.4255,
+      "step": 365
+    },
+    {
+      "epoch": 0.4123943661971831,
+      "grad_norm": 0.3713182569695666,
+      "learning_rate": 4.793233082706767e-05,
+      "loss": 0.4411,
+      "step": 366
+    },
+    {
+      "epoch": 0.4135211267605634,
+      "grad_norm": 0.4584153156070977,
+      "learning_rate": 4.791144527986633e-05,
+      "loss": 0.4372,
+      "step": 367
+    },
+    {
+      "epoch": 0.41464788732394364,
+      "grad_norm": 0.42578321791979223,
+      "learning_rate": 4.7890559732665e-05,
+      "loss": 0.4291,
+      "step": 368
+    },
+    {
+      "epoch": 0.41577464788732393,
+      "grad_norm": 0.4177132674416258,
+      "learning_rate": 4.786967418546366e-05,
+      "loss": 0.4204,
+      "step": 369
+    },
+    {
+      "epoch": 0.4169014084507042,
+      "grad_norm": 0.4666753953413797,
+      "learning_rate": 4.784878863826233e-05,
+      "loss": 0.443,
+      "step": 370
+    },
+    {
+      "epoch": 0.4180281690140845,
+      "grad_norm": 0.48466090823607716,
+      "learning_rate": 4.782790309106099e-05,
+      "loss": 0.451,
+      "step": 371
+    },
+    {
+      "epoch": 0.4191549295774648,
+      "grad_norm": 0.46837220316117795,
+      "learning_rate": 4.780701754385965e-05,
+      "loss": 0.4063,
+      "step": 372
+    },
+    {
+      "epoch": 0.42028169014084504,
+      "grad_norm": 0.5287131069297187,
+      "learning_rate": 4.778613199665831e-05,
+      "loss": 0.4419,
+      "step": 373
+    },
+    {
+      "epoch": 0.42140845070422533,
+      "grad_norm": 0.4684724501520272,
+      "learning_rate": 4.776524644945698e-05,
+      "loss": 0.4197,
+      "step": 374
+    },
+    {
+      "epoch": 0.4225352112676056,
+      "grad_norm": 0.45380336381720876,
+      "learning_rate": 4.774436090225564e-05,
+      "loss": 0.4391,
+      "step": 375
+    },
+    {
+      "epoch": 0.4236619718309859,
+      "grad_norm": 0.44925990919244896,
+      "learning_rate": 4.77234753550543e-05,
+      "loss": 0.4406,
+      "step": 376
+    },
+    {
+      "epoch": 0.4247887323943662,
+      "grad_norm": 0.3873662129780452,
+      "learning_rate": 4.770258980785297e-05,
+      "loss": 0.4088,
+      "step": 377
+    },
+    {
+      "epoch": 0.4259154929577465,
+      "grad_norm": 0.37778608304923517,
+      "learning_rate": 4.7681704260651634e-05,
+      "loss": 0.4254,
+      "step": 378
+    },
+    {
+      "epoch": 0.42704225352112674,
+      "grad_norm": 0.44423925673574166,
+      "learning_rate": 4.7660818713450294e-05,
+      "loss": 0.4595,
+      "step": 379
+    },
+    {
+      "epoch": 0.428169014084507,
+      "grad_norm": 0.3663991698191898,
+      "learning_rate": 4.7639933166248954e-05,
+      "loss": 0.4286,
+      "step": 380
+    },
+    {
+      "epoch": 0.4292957746478873,
+      "grad_norm": 0.4099685876093145,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.4363,
+      "step": 381
+    },
+    {
+      "epoch": 0.4304225352112676,
+      "grad_norm": 0.4309445114432002,
+      "learning_rate": 4.759816207184629e-05,
+      "loss": 0.4276,
+      "step": 382
+    },
+    {
+      "epoch": 0.4315492957746479,
+      "grad_norm": 0.4082416446454451,
+      "learning_rate": 4.757727652464495e-05,
+      "loss": 0.4468,
+      "step": 383
+    },
+    {
+      "epoch": 0.4326760563380282,
+      "grad_norm": 0.3823707441932576,
+      "learning_rate": 4.755639097744361e-05,
+      "loss": 0.4353,
+      "step": 384
+    },
+    {
+      "epoch": 0.43380281690140843,
+      "grad_norm": 0.4984043757512769,
+      "learning_rate": 4.7535505430242274e-05,
+      "loss": 0.4578,
+      "step": 385
+    },
+    {
+      "epoch": 0.4349295774647887,
+      "grad_norm": 0.43584279497841244,
+      "learning_rate": 4.751461988304094e-05,
+      "loss": 0.4258,
+      "step": 386
+    },
+    {
+      "epoch": 0.436056338028169,
+      "grad_norm": 0.43822399751636953,
+      "learning_rate": 4.74937343358396e-05,
+      "loss": 0.4353,
+      "step": 387
+    },
+    {
+      "epoch": 0.4371830985915493,
+      "grad_norm": 0.42344449072896734,
+      "learning_rate": 4.747284878863826e-05,
+      "loss": 0.4778,
+      "step": 388
+    },
+    {
+      "epoch": 0.4383098591549296,
+      "grad_norm": 0.4056081290584808,
+      "learning_rate": 4.745196324143693e-05,
+      "loss": 0.4237,
+      "step": 389
+    },
+    {
+      "epoch": 0.4394366197183099,
+      "grad_norm": 0.3732430125687137,
+      "learning_rate": 4.7431077694235594e-05,
+      "loss": 0.4554,
+      "step": 390
+    },
+    {
+      "epoch": 0.4405633802816901,
+      "grad_norm": 0.3647079020199324,
+      "learning_rate": 4.7410192147034254e-05,
+      "loss": 0.439,
+      "step": 391
+    },
+    {
+      "epoch": 0.4416901408450704,
+      "grad_norm": 0.4000271217336532,
+      "learning_rate": 4.7389306599832914e-05,
+      "loss": 0.4393,
+      "step": 392
+    },
+    {
+      "epoch": 0.4428169014084507,
+      "grad_norm": 0.374753617359853,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 0.4256,
+      "step": 393
+    },
+    {
+      "epoch": 0.443943661971831,
+      "grad_norm": 0.3305173070867681,
+      "learning_rate": 4.734753550543025e-05,
+      "loss": 0.4329,
+      "step": 394
+    },
+    {
+      "epoch": 0.4450704225352113,
+      "grad_norm": 0.36944667056910313,
+      "learning_rate": 4.732664995822891e-05,
+      "loss": 0.4695,
+      "step": 395
+    },
+    {
+      "epoch": 0.4461971830985916,
+      "grad_norm": 0.3928435655072016,
+      "learning_rate": 4.730576441102757e-05,
+      "loss": 0.4319,
+      "step": 396
+    },
+    {
+      "epoch": 0.4473239436619718,
+      "grad_norm": 0.336375328408579,
+      "learning_rate": 4.7284878863826234e-05,
+      "loss": 0.425,
+      "step": 397
+    },
+    {
+      "epoch": 0.4484507042253521,
+      "grad_norm": 0.39132740901645385,
+      "learning_rate": 4.72639933166249e-05,
+      "loss": 0.4431,
+      "step": 398
+    },
+    {
+      "epoch": 0.4495774647887324,
+      "grad_norm": 0.369926735816109,
+      "learning_rate": 4.724310776942356e-05,
+      "loss": 0.429,
+      "step": 399
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.39992553614477316,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.4484,
+      "step": 400
+    },
+    {
+      "epoch": 0.451830985915493,
+      "grad_norm": 0.4458014520019311,
+      "learning_rate": 4.720133667502089e-05,
+      "loss": 0.4483,
+      "step": 401
+    },
+    {
+      "epoch": 0.4529577464788732,
+      "grad_norm": 0.4549007653653714,
+      "learning_rate": 4.7180451127819555e-05,
+      "loss": 0.4745,
+      "step": 402
+    },
+    {
+      "epoch": 0.4540845070422535,
+      "grad_norm": 0.43605986381583667,
+      "learning_rate": 4.7159565580618215e-05,
+      "loss": 0.4234,
+      "step": 403
+    },
+    {
+      "epoch": 0.4552112676056338,
+      "grad_norm": 0.43415229649505843,
+      "learning_rate": 4.7138680033416875e-05,
+      "loss": 0.417,
+      "step": 404
+    },
+    {
+      "epoch": 0.4563380281690141,
+      "grad_norm": 0.45516649193571096,
+      "learning_rate": 4.711779448621554e-05,
+      "loss": 0.4116,
+      "step": 405
+    },
+    {
+      "epoch": 0.4574647887323944,
+      "grad_norm": 0.47614856097142705,
+      "learning_rate": 4.709690893901421e-05,
+      "loss": 0.4447,
+      "step": 406
+    },
+    {
+      "epoch": 0.45859154929577467,
+      "grad_norm": 0.3842022687208916,
+      "learning_rate": 4.707602339181287e-05,
+      "loss": 0.4265,
+      "step": 407
+    },
+    {
+      "epoch": 0.4597183098591549,
+      "grad_norm": 0.5003337639832289,
+      "learning_rate": 4.705513784461153e-05,
+      "loss": 0.4576,
+      "step": 408
+    },
+    {
+      "epoch": 0.4608450704225352,
+      "grad_norm": 0.371920341514909,
+      "learning_rate": 4.7034252297410195e-05,
+      "loss": 0.4139,
+      "step": 409
+    },
+    {
+      "epoch": 0.4619718309859155,
+      "grad_norm": 0.4327864270513005,
+      "learning_rate": 4.701336675020886e-05,
+      "loss": 0.4482,
+      "step": 410
+    },
+    {
+      "epoch": 0.4630985915492958,
+      "grad_norm": 0.4134719706292766,
+      "learning_rate": 4.699248120300752e-05,
+      "loss": 0.4187,
+      "step": 411
+    },
+    {
+      "epoch": 0.46422535211267607,
+      "grad_norm": 0.40524197759616115,
+      "learning_rate": 4.697159565580618e-05,
+      "loss": 0.4435,
+      "step": 412
+    },
+    {
+      "epoch": 0.46535211267605636,
+      "grad_norm": 0.42067638203215985,
+      "learning_rate": 4.695071010860485e-05,
+      "loss": 0.4363,
+      "step": 413
+    },
+    {
+      "epoch": 0.4664788732394366,
+      "grad_norm": 0.48630981183960437,
+      "learning_rate": 4.6929824561403515e-05,
+      "loss": 0.4577,
+      "step": 414
+    },
+    {
+      "epoch": 0.4676056338028169,
+      "grad_norm": 0.36522530679978066,
+      "learning_rate": 4.6908939014202175e-05,
+      "loss": 0.4353,
+      "step": 415
+    },
+    {
+      "epoch": 0.4687323943661972,
+      "grad_norm": 0.5227222722102871,
+      "learning_rate": 4.6888053467000835e-05,
+      "loss": 0.4226,
+      "step": 416
+    },
+    {
+      "epoch": 0.46985915492957747,
+      "grad_norm": 0.5069277128907972,
+      "learning_rate": 4.6867167919799495e-05,
+      "loss": 0.4316,
+      "step": 417
+    },
+    {
+      "epoch": 0.47098591549295776,
+      "grad_norm": 0.484118433212617,
+      "learning_rate": 4.684628237259817e-05,
+      "loss": 0.4224,
+      "step": 418
+    },
+    {
+      "epoch": 0.47211267605633805,
+      "grad_norm": 0.6464454503406327,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.4267,
+      "step": 419
+    },
+    {
+      "epoch": 0.4732394366197183,
+      "grad_norm": 0.3914137430386243,
+      "learning_rate": 4.680451127819549e-05,
+      "loss": 0.4344,
+      "step": 420
+    },
+    {
+      "epoch": 0.4743661971830986,
+      "grad_norm": 0.6288927819005267,
+      "learning_rate": 4.678362573099415e-05,
+      "loss": 0.4471,
+      "step": 421
+    },
+    {
+      "epoch": 0.4754929577464789,
+      "grad_norm": 0.42140273410218787,
+      "learning_rate": 4.676274018379282e-05,
+      "loss": 0.4376,
+      "step": 422
+    },
+    {
+      "epoch": 0.47661971830985916,
+      "grad_norm": 0.5218529452733929,
+      "learning_rate": 4.674185463659148e-05,
+      "loss": 0.4329,
+      "step": 423
+    },
+    {
+      "epoch": 0.47774647887323946,
+      "grad_norm": 0.4462368667411693,
+      "learning_rate": 4.672096908939014e-05,
+      "loss": 0.4427,
+      "step": 424
+    },
+    {
+      "epoch": 0.4788732394366197,
+      "grad_norm": 0.41668970796138727,
+      "learning_rate": 4.67000835421888e-05,
+      "loss": 0.4405,
+      "step": 425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3831817162113262,
+      "learning_rate": 4.6679197994987476e-05,
+      "loss": 0.4284,
+      "step": 426
+    },
+    {
+      "epoch": 0.4811267605633803,
+      "grad_norm": 0.4229704513884789,
+      "learning_rate": 4.6658312447786136e-05,
+      "loss": 0.4454,
+      "step": 427
+    },
+    {
+      "epoch": 0.48225352112676056,
+      "grad_norm": 0.48032056597692396,
+      "learning_rate": 4.6637426900584796e-05,
+      "loss": 0.4377,
+      "step": 428
+    },
+    {
+      "epoch": 0.48338028169014086,
+      "grad_norm": 0.3352986960279401,
+      "learning_rate": 4.6616541353383456e-05,
+      "loss": 0.4259,
+      "step": 429
+    },
+    {
+      "epoch": 0.48450704225352115,
+      "grad_norm": 0.4131630884201645,
+      "learning_rate": 4.659565580618213e-05,
+      "loss": 0.4317,
+      "step": 430
+    },
+    {
+      "epoch": 0.4856338028169014,
+      "grad_norm": 0.45751748634481637,
+      "learning_rate": 4.657477025898079e-05,
+      "loss": 0.4405,
+      "step": 431
+    },
+    {
+      "epoch": 0.4867605633802817,
+      "grad_norm": 0.4139862920321763,
+      "learning_rate": 4.655388471177945e-05,
+      "loss": 0.4353,
+      "step": 432
+    },
+    {
+      "epoch": 0.48788732394366197,
+      "grad_norm": 0.43396018489841076,
+      "learning_rate": 4.653299916457811e-05,
+      "loss": 0.4182,
+      "step": 433
+    },
+    {
+      "epoch": 0.48901408450704226,
+      "grad_norm": 0.4406093388366263,
+      "learning_rate": 4.6512113617376776e-05,
+      "loss": 0.4457,
+      "step": 434
+    },
+    {
+      "epoch": 0.49014084507042255,
+      "grad_norm": 0.43843357456964716,
+      "learning_rate": 4.649122807017544e-05,
+      "loss": 0.4339,
+      "step": 435
+    },
+    {
+      "epoch": 0.49126760563380284,
+      "grad_norm": 0.3239492046043941,
+      "learning_rate": 4.64703425229741e-05,
+      "loss": 0.4134,
+      "step": 436
+    },
+    {
+      "epoch": 0.4923943661971831,
+      "grad_norm": 0.40897075153266427,
+      "learning_rate": 4.644945697577276e-05,
+      "loss": 0.4339,
+      "step": 437
+    },
+    {
+      "epoch": 0.49352112676056337,
+      "grad_norm": 0.3331343883758221,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 0.409,
+      "step": 438
+    },
+    {
+      "epoch": 0.49464788732394366,
+      "grad_norm": 0.3994719175224885,
+      "learning_rate": 4.6407685881370096e-05,
+      "loss": 0.4189,
+      "step": 439
+    },
+    {
+      "epoch": 0.49577464788732395,
+      "grad_norm": 0.387575870558901,
+      "learning_rate": 4.6386800334168756e-05,
+      "loss": 0.4342,
+      "step": 440
+    },
+    {
+      "epoch": 0.49690140845070424,
+      "grad_norm": 0.38418921533926903,
+      "learning_rate": 4.6365914786967416e-05,
+      "loss": 0.4294,
+      "step": 441
+    },
+    {
+      "epoch": 0.49802816901408453,
+      "grad_norm": 0.515380262957802,
+      "learning_rate": 4.634502923976608e-05,
+      "loss": 0.4215,
+      "step": 442
+    },
+    {
+      "epoch": 0.49915492957746477,
+      "grad_norm": 0.3892122015570572,
+      "learning_rate": 4.632414369256475e-05,
+      "loss": 0.436,
+      "step": 443
+    },
+    {
+      "epoch": 0.5002816901408451,
+      "grad_norm": 0.43122533906327076,
+      "learning_rate": 4.630325814536341e-05,
+      "loss": 0.4174,
+      "step": 444
+    },
+    {
+      "epoch": 0.5014084507042254,
+      "grad_norm": 0.3825183575537722,
+      "learning_rate": 4.628237259816207e-05,
+      "loss": 0.4204,
+      "step": 445
+    },
+    {
+      "epoch": 0.5025352112676056,
+      "grad_norm": 0.3951031876799554,
+      "learning_rate": 4.6261487050960736e-05,
+      "loss": 0.4524,
+      "step": 446
+    },
+    {
+      "epoch": 0.5036619718309859,
+      "grad_norm": 0.3993720699665561,
+      "learning_rate": 4.62406015037594e-05,
+      "loss": 0.4535,
+      "step": 447
+    },
+    {
+      "epoch": 0.5047887323943662,
+      "grad_norm": 0.3702213447503512,
+      "learning_rate": 4.621971595655806e-05,
+      "loss": 0.4467,
+      "step": 448
+    },
+    {
+      "epoch": 0.5059154929577465,
+      "grad_norm": 0.42987431158793665,
+      "learning_rate": 4.619883040935672e-05,
+      "loss": 0.4491,
+      "step": 449
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 0.39840069823036856,
+      "learning_rate": 4.617794486215539e-05,
+      "loss": 0.4452,
+      "step": 450
+    },
+    {
+      "epoch": 0.508169014084507,
+      "grad_norm": 0.3774377587435373,
+      "learning_rate": 4.6157059314954057e-05,
+      "loss": 0.4151,
+      "step": 451
+    },
+    {
+      "epoch": 0.5092957746478873,
+      "grad_norm": 0.32586659282974617,
+      "learning_rate": 4.6136173767752717e-05,
+      "loss": 0.4047,
+      "step": 452
+    },
+    {
+      "epoch": 0.5104225352112676,
+      "grad_norm": 0.370312046890077,
+      "learning_rate": 4.6115288220551377e-05,
+      "loss": 0.4163,
+      "step": 453
+    },
+    {
+      "epoch": 0.5115492957746479,
+      "grad_norm": 0.3819499089937181,
+      "learning_rate": 4.609440267335004e-05,
+      "loss": 0.4144,
+      "step": 454
+    },
+    {
+      "epoch": 0.5126760563380282,
+      "grad_norm": 0.43036152053924126,
+      "learning_rate": 4.607351712614871e-05,
+      "loss": 0.4317,
+      "step": 455
+    },
+    {
+      "epoch": 0.5138028169014085,
+      "grad_norm": 0.5255656991540114,
+      "learning_rate": 4.605263157894737e-05,
+      "loss": 0.4429,
+      "step": 456
+    },
+    {
+      "epoch": 0.5149295774647887,
+      "grad_norm": 0.39606254003461006,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.4415,
+      "step": 457
+    },
+    {
+      "epoch": 0.516056338028169,
+      "grad_norm": 0.4396753982202345,
+      "learning_rate": 4.60108604845447e-05,
+      "loss": 0.4216,
+      "step": 458
+    },
+    {
+      "epoch": 0.5171830985915493,
+      "grad_norm": 0.42929853762727704,
+      "learning_rate": 4.5989974937343364e-05,
+      "loss": 0.4164,
+      "step": 459
+    },
+    {
+      "epoch": 0.5183098591549296,
+      "grad_norm": 0.43170237614681684,
+      "learning_rate": 4.5969089390142024e-05,
+      "loss": 0.4448,
+      "step": 460
+    },
+    {
+      "epoch": 0.5194366197183099,
+      "grad_norm": 0.48778911427761545,
+      "learning_rate": 4.5948203842940684e-05,
+      "loss": 0.432,
+      "step": 461
+    },
+    {
+      "epoch": 0.5205633802816901,
+      "grad_norm": 0.4312484290328229,
+      "learning_rate": 4.592731829573935e-05,
+      "loss": 0.4354,
+      "step": 462
+    },
+    {
+      "epoch": 0.5216901408450704,
+      "grad_norm": 0.41655513515023973,
+      "learning_rate": 4.590643274853802e-05,
+      "loss": 0.4378,
+      "step": 463
+    },
+    {
+      "epoch": 0.5228169014084507,
+      "grad_norm": 0.535031469710637,
+      "learning_rate": 4.588554720133668e-05,
+      "loss": 0.456,
+      "step": 464
+    },
+    {
+      "epoch": 0.523943661971831,
+      "grad_norm": 0.4089281972043279,
+      "learning_rate": 4.586466165413534e-05,
+      "loss": 0.4412,
+      "step": 465
+    },
+    {
+      "epoch": 0.5250704225352113,
+      "grad_norm": 0.49326718973583,
+      "learning_rate": 4.5843776106934004e-05,
+      "loss": 0.4505,
+      "step": 466
+    },
+    {
+      "epoch": 0.5261971830985915,
+      "grad_norm": 0.4220477561910598,
+      "learning_rate": 4.582289055973267e-05,
+      "loss": 0.4277,
+      "step": 467
+    },
+    {
+      "epoch": 0.5273239436619719,
+      "grad_norm": 0.46060462678182534,
+      "learning_rate": 4.580200501253133e-05,
+      "loss": 0.4593,
+      "step": 468
+    },
+    {
+      "epoch": 0.5284507042253521,
+      "grad_norm": 0.4370238774092618,
+      "learning_rate": 4.578111946532999e-05,
+      "loss": 0.4173,
+      "step": 469
+    },
+    {
+      "epoch": 0.5295774647887324,
+      "grad_norm": 0.4313253955555467,
+      "learning_rate": 4.576023391812866e-05,
+      "loss": 0.4289,
+      "step": 470
+    },
+    {
+      "epoch": 0.5307042253521127,
+      "grad_norm": 0.3539426843281356,
+      "learning_rate": 4.5739348370927324e-05,
+      "loss": 0.4154,
+      "step": 471
+    },
+    {
+      "epoch": 0.5318309859154929,
+      "grad_norm": 0.4487503536152153,
+      "learning_rate": 4.5718462823725984e-05,
+      "loss": 0.4304,
+      "step": 472
+    },
+    {
+      "epoch": 0.5329577464788733,
+      "grad_norm": 0.3698974487994367,
+      "learning_rate": 4.5697577276524644e-05,
+      "loss": 0.435,
+      "step": 473
+    },
+    {
+      "epoch": 0.5340845070422535,
+      "grad_norm": 0.41550377526255394,
+      "learning_rate": 4.567669172932331e-05,
+      "loss": 0.4502,
+      "step": 474
+    },
+    {
+      "epoch": 0.5352112676056338,
+      "grad_norm": 0.3800550727022986,
+      "learning_rate": 4.565580618212197e-05,
+      "loss": 0.427,
+      "step": 475
+    },
+    {
+      "epoch": 0.5363380281690141,
+      "grad_norm": 0.3988520807240111,
+      "learning_rate": 4.563492063492064e-05,
+      "loss": 0.4127,
+      "step": 476
+    },
+    {
+      "epoch": 0.5374647887323943,
+      "grad_norm": 0.4376872891197101,
+      "learning_rate": 4.56140350877193e-05,
+      "loss": 0.4291,
+      "step": 477
+    },
+    {
+      "epoch": 0.5385915492957747,
+      "grad_norm": 0.4572428687430911,
+      "learning_rate": 4.5593149540517964e-05,
+      "loss": 0.4216,
+      "step": 478
+    },
+    {
+      "epoch": 0.5397183098591549,
+      "grad_norm": 0.49237250245493225,
+      "learning_rate": 4.5572263993316624e-05,
+      "loss": 0.4511,
+      "step": 479
+    },
+    {
+      "epoch": 0.5408450704225352,
+      "grad_norm": 0.4082354204406026,
+      "learning_rate": 4.555137844611529e-05,
+      "loss": 0.4344,
+      "step": 480
+    },
+    {
+      "epoch": 0.5419718309859155,
+      "grad_norm": 0.5330130407437706,
+      "learning_rate": 4.553049289891395e-05,
+      "loss": 0.4268,
+      "step": 481
+    },
+    {
+      "epoch": 0.5430985915492957,
+      "grad_norm": 0.3877512017370143,
+      "learning_rate": 4.550960735171262e-05,
+      "loss": 0.4339,
+      "step": 482
+    },
+    {
+      "epoch": 0.5442253521126761,
+      "grad_norm": 0.496803144386499,
+      "learning_rate": 4.548872180451128e-05,
+      "loss": 0.4265,
+      "step": 483
+    },
+    {
+      "epoch": 0.5453521126760563,
+      "grad_norm": 0.3924339898963181,
+      "learning_rate": 4.5467836257309945e-05,
+      "loss": 0.4316,
+      "step": 484
+    },
+    {
+      "epoch": 0.5464788732394367,
+      "grad_norm": 0.5523374929860176,
+      "learning_rate": 4.5446950710108605e-05,
+      "loss": 0.4374,
+      "step": 485
+    },
+    {
+      "epoch": 0.5476056338028169,
+      "grad_norm": 0.39714016214787484,
+      "learning_rate": 4.542606516290727e-05,
+      "loss": 0.4141,
+      "step": 486
+    },
+    {
+      "epoch": 0.5487323943661971,
+      "grad_norm": 0.4672623284669598,
+      "learning_rate": 4.540517961570593e-05,
+      "loss": 0.4338,
+      "step": 487
+    },
+    {
+      "epoch": 0.5498591549295775,
+      "grad_norm": 0.45532761050056775,
+      "learning_rate": 4.53842940685046e-05,
+      "loss": 0.4282,
+      "step": 488
+    },
+    {
+      "epoch": 0.5509859154929577,
+      "grad_norm": 0.3512267218354597,
+      "learning_rate": 4.536340852130326e-05,
+      "loss": 0.4262,
+      "step": 489
+    },
+    {
+      "epoch": 0.5521126760563381,
+      "grad_norm": 0.5344471882676834,
+      "learning_rate": 4.5342522974101925e-05,
+      "loss": 0.4214,
+      "step": 490
+    },
+    {
+      "epoch": 0.5532394366197183,
+      "grad_norm": 0.40804809863465386,
+      "learning_rate": 4.5321637426900585e-05,
+      "loss": 0.4176,
+      "step": 491
+    },
+    {
+      "epoch": 0.5543661971830985,
+      "grad_norm": 0.5026100013021888,
+      "learning_rate": 4.530075187969925e-05,
+      "loss": 0.4435,
+      "step": 492
+    },
+    {
+      "epoch": 0.5554929577464789,
+      "grad_norm": 0.4298804130144407,
+      "learning_rate": 4.527986633249791e-05,
+      "loss": 0.42,
+      "step": 493
+    },
+    {
+      "epoch": 0.5566197183098591,
+      "grad_norm": 0.4450683590913135,
+      "learning_rate": 4.525898078529658e-05,
+      "loss": 0.4361,
+      "step": 494
+    },
+    {
+      "epoch": 0.5577464788732395,
+      "grad_norm": 0.5168808901708248,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.4044,
+      "step": 495
+    },
+    {
+      "epoch": 0.5588732394366197,
+      "grad_norm": 0.3796230305562614,
+      "learning_rate": 4.5217209690893905e-05,
+      "loss": 0.3923,
+      "step": 496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.494439451537962,
+      "learning_rate": 4.5196324143692565e-05,
+      "loss": 0.4134,
+      "step": 497
+    },
+    {
+      "epoch": 0.5611267605633803,
+      "grad_norm": 0.38887094714361475,
+      "learning_rate": 4.517543859649123e-05,
+      "loss": 0.4277,
+      "step": 498
+    },
+    {
+      "epoch": 0.5622535211267605,
+      "grad_norm": 0.4771133935931392,
+      "learning_rate": 4.515455304928989e-05,
+      "loss": 0.4332,
+      "step": 499
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 0.5833210814037667,
+      "learning_rate": 4.513366750208856e-05,
+      "loss": 0.4479,
+      "step": 500
+    },
+    {
+      "epoch": 0.5645070422535211,
+      "grad_norm": 0.4650742185978965,
+      "learning_rate": 4.511278195488722e-05,
+      "loss": 0.4276,
+      "step": 501
+    },
+    {
+      "epoch": 0.5656338028169015,
+      "grad_norm": 0.7649017257631372,
+      "learning_rate": 4.5091896407685885e-05,
+      "loss": 0.4271,
+      "step": 502
+    },
+    {
+      "epoch": 0.5667605633802817,
+      "grad_norm": 0.40039301536077476,
+      "learning_rate": 4.5071010860484545e-05,
+      "loss": 0.4291,
+      "step": 503
+    },
+    {
+      "epoch": 0.5678873239436619,
+      "grad_norm": 0.6664943828112884,
+      "learning_rate": 4.505012531328321e-05,
+      "loss": 0.4256,
+      "step": 504
+    },
+    {
+      "epoch": 0.5690140845070423,
+      "grad_norm": 0.4303326479225263,
+      "learning_rate": 4.502923976608187e-05,
+      "loss": 0.4366,
+      "step": 505
+    },
+    {
+      "epoch": 0.5701408450704225,
+      "grad_norm": 0.48633413358954836,
+      "learning_rate": 4.500835421888054e-05,
+      "loss": 0.4264,
+      "step": 506
+    },
+    {
+      "epoch": 0.5712676056338029,
+      "grad_norm": 0.48788986503015436,
+      "learning_rate": 4.49874686716792e-05,
+      "loss": 0.4094,
+      "step": 507
+    },
+    {
+      "epoch": 0.5723943661971831,
+      "grad_norm": 0.44298895531473165,
+      "learning_rate": 4.4966583124477866e-05,
+      "loss": 0.4142,
+      "step": 508
+    },
+    {
+      "epoch": 0.5735211267605633,
+      "grad_norm": 0.42806203997959974,
+      "learning_rate": 4.4945697577276526e-05,
+      "loss": 0.4358,
+      "step": 509
+    },
+    {
+      "epoch": 0.5746478873239437,
+      "grad_norm": 0.43778101558906973,
+      "learning_rate": 4.492481203007519e-05,
+      "loss": 0.4124,
+      "step": 510
+    },
+    {
+      "epoch": 0.5757746478873239,
+      "grad_norm": 0.3604903511146798,
+      "learning_rate": 4.490392648287385e-05,
+      "loss": 0.4039,
+      "step": 511
+    },
+    {
+      "epoch": 0.5769014084507043,
+      "grad_norm": 2.205061903497407,
+      "learning_rate": 4.488304093567251e-05,
+      "loss": 0.4032,
+      "step": 512
+    },
+    {
+      "epoch": 0.5780281690140845,
+      "grad_norm": 0.5118994703596127,
+      "learning_rate": 4.486215538847118e-05,
+      "loss": 0.4059,
+      "step": 513
+    },
+    {
+      "epoch": 0.5791549295774648,
+      "grad_norm": 0.40456548089618005,
+      "learning_rate": 4.4841269841269846e-05,
+      "loss": 0.4197,
+      "step": 514
+    },
+    {
+      "epoch": 0.5802816901408451,
+      "grad_norm": 0.4337732467996675,
+      "learning_rate": 4.4820384294068506e-05,
+      "loss": 0.4215,
+      "step": 515
+    },
+    {
+      "epoch": 0.5814084507042253,
+      "grad_norm": 0.37286043099281574,
+      "learning_rate": 4.4799498746867166e-05,
+      "loss": 0.4133,
+      "step": 516
+    },
+    {
+      "epoch": 0.5825352112676057,
+      "grad_norm": 0.4430381139696347,
+      "learning_rate": 4.477861319966583e-05,
+      "loss": 0.421,
+      "step": 517
+    },
+    {
+      "epoch": 0.5836619718309859,
+      "grad_norm": 0.38762202455986466,
+      "learning_rate": 4.47577276524645e-05,
+      "loss": 0.4464,
+      "step": 518
+    },
+    {
+      "epoch": 0.5847887323943662,
+      "grad_norm": 0.4254747897345159,
+      "learning_rate": 4.473684210526316e-05,
+      "loss": 0.4464,
+      "step": 519
+    },
+    {
+      "epoch": 0.5859154929577465,
+      "grad_norm": 0.36818503891060694,
+      "learning_rate": 4.471595655806182e-05,
+      "loss": 0.4359,
+      "step": 520
+    },
+    {
+      "epoch": 0.5870422535211267,
+      "grad_norm": 0.4396858568603212,
+      "learning_rate": 4.4695071010860486e-05,
+      "loss": 0.4524,
+      "step": 521
+    },
+    {
+      "epoch": 0.5881690140845071,
+      "grad_norm": 0.47575935108191253,
+      "learning_rate": 4.467418546365915e-05,
+      "loss": 0.4408,
+      "step": 522
+    },
+    {
+      "epoch": 0.5892957746478873,
+      "grad_norm": 0.4798203797254968,
+      "learning_rate": 4.465329991645781e-05,
+      "loss": 0.3955,
+      "step": 523
+    },
+    {
+      "epoch": 0.5904225352112676,
+      "grad_norm": 0.491404025492639,
+      "learning_rate": 4.463241436925647e-05,
+      "loss": 0.4347,
+      "step": 524
+    },
+    {
+      "epoch": 0.5915492957746479,
+      "grad_norm": 0.49123207445955547,
+      "learning_rate": 4.461152882205514e-05,
+      "loss": 0.4356,
+      "step": 525
+    },
+    {
+      "epoch": 0.5926760563380282,
+      "grad_norm": 0.6105485825942882,
+      "learning_rate": 4.4590643274853806e-05,
+      "loss": 0.409,
+      "step": 526
+    },
+    {
+      "epoch": 0.5938028169014085,
+      "grad_norm": 0.5207042930483079,
+      "learning_rate": 4.4569757727652466e-05,
+      "loss": 0.441,
+      "step": 527
+    },
+    {
+      "epoch": 0.5949295774647887,
+      "grad_norm": 0.5149266515759818,
+      "learning_rate": 4.4548872180451126e-05,
+      "loss": 0.4273,
+      "step": 528
+    },
+    {
+      "epoch": 0.596056338028169,
+      "grad_norm": 0.5477491737233843,
+      "learning_rate": 4.452798663324979e-05,
+      "loss": 0.4263,
+      "step": 529
+    },
+    {
+      "epoch": 0.5971830985915493,
+      "grad_norm": 0.468103294840018,
+      "learning_rate": 4.450710108604846e-05,
+      "loss": 0.4218,
+      "step": 530
+    },
+    {
+      "epoch": 0.5983098591549296,
+      "grad_norm": 0.45956460347799083,
+      "learning_rate": 4.448621553884712e-05,
+      "loss": 0.4334,
+      "step": 531
+    },
+    {
+      "epoch": 0.5994366197183099,
+      "grad_norm": 0.4111204934729319,
+      "learning_rate": 4.446532999164578e-05,
+      "loss": 0.4324,
+      "step": 532
+    },
+    {
+      "epoch": 0.6005633802816901,
+      "grad_norm": 0.36963074782643357,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.431,
+      "step": 533
+    },
+    {
+      "epoch": 0.6016901408450704,
+      "grad_norm": 0.4104529558446069,
+      "learning_rate": 4.442355889724311e-05,
+      "loss": 0.4372,
+      "step": 534
+    },
+    {
+      "epoch": 0.6028169014084507,
+      "grad_norm": 0.37824181533897855,
+      "learning_rate": 4.440267335004177e-05,
+      "loss": 0.4626,
+      "step": 535
+    },
+    {
+      "epoch": 0.603943661971831,
+      "grad_norm": 0.3483670591865026,
+      "learning_rate": 4.438178780284043e-05,
+      "loss": 0.4034,
+      "step": 536
+    },
+    {
+      "epoch": 0.6050704225352113,
+      "grad_norm": 0.392492115921629,
+      "learning_rate": 4.43609022556391e-05,
+      "loss": 0.4165,
+      "step": 537
+    },
+    {
+      "epoch": 0.6061971830985915,
+      "grad_norm": 0.38568538328258095,
+      "learning_rate": 4.434001670843777e-05,
+      "loss": 0.427,
+      "step": 538
+    },
+    {
+      "epoch": 0.6073239436619718,
+      "grad_norm": 0.40601482709549314,
+      "learning_rate": 4.431913116123643e-05,
+      "loss": 0.4251,
+      "step": 539
+    },
+    {
+      "epoch": 0.6084507042253521,
+      "grad_norm": 0.4320403980156557,
+      "learning_rate": 4.429824561403509e-05,
+      "loss": 0.4463,
+      "step": 540
+    },
+    {
+      "epoch": 0.6095774647887324,
+      "grad_norm": 0.3739618657407569,
+      "learning_rate": 4.4277360066833754e-05,
+      "loss": 0.4003,
+      "step": 541
+    },
+    {
+      "epoch": 0.6107042253521127,
+      "grad_norm": 0.36832449593704425,
+      "learning_rate": 4.425647451963242e-05,
+      "loss": 0.4245,
+      "step": 542
+    },
+    {
+      "epoch": 0.611830985915493,
+      "grad_norm": 0.3614080383744093,
+      "learning_rate": 4.423558897243108e-05,
+      "loss": 0.4178,
+      "step": 543
+    },
+    {
+      "epoch": 0.6129577464788732,
+      "grad_norm": 0.37450999344316493,
+      "learning_rate": 4.421470342522974e-05,
+      "loss": 0.4145,
+      "step": 544
+    },
+    {
+      "epoch": 0.6140845070422535,
+      "grad_norm": 0.36930155146025684,
+      "learning_rate": 4.419381787802841e-05,
+      "loss": 0.4226,
+      "step": 545
+    },
+    {
+      "epoch": 0.6152112676056338,
+      "grad_norm": 0.37304790770081386,
+      "learning_rate": 4.4172932330827074e-05,
+      "loss": 0.4184,
+      "step": 546
+    },
+    {
+      "epoch": 0.6163380281690141,
+      "grad_norm": 0.37345149852575577,
+      "learning_rate": 4.4152046783625734e-05,
+      "loss": 0.4419,
+      "step": 547
+    },
+    {
+      "epoch": 0.6174647887323944,
+      "grad_norm": 0.3942758512347128,
+      "learning_rate": 4.4131161236424394e-05,
+      "loss": 0.4573,
+      "step": 548
+    },
+    {
+      "epoch": 0.6185915492957746,
+      "grad_norm": 0.3407591248518673,
+      "learning_rate": 4.411027568922306e-05,
+      "loss": 0.412,
+      "step": 549
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.4414476222318839,
+      "learning_rate": 4.408939014202173e-05,
+      "loss": 0.4467,
+      "step": 550
+    },
+    {
+      "epoch": 0.6208450704225352,
+      "grad_norm": 0.3927007732686169,
+      "learning_rate": 4.406850459482039e-05,
+      "loss": 0.43,
+      "step": 551
+    },
+    {
+      "epoch": 0.6219718309859155,
+      "grad_norm": 0.31673151571390507,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 0.3833,
+      "step": 552
+    },
+    {
+      "epoch": 0.6230985915492958,
+      "grad_norm": 0.34613960365197394,
+      "learning_rate": 4.402673350041771e-05,
+      "loss": 0.3923,
+      "step": 553
+    },
+    {
+      "epoch": 0.624225352112676,
+      "grad_norm": 0.3562786927865309,
+      "learning_rate": 4.400584795321638e-05,
+      "loss": 0.4188,
+      "step": 554
+    },
+    {
+      "epoch": 0.6253521126760564,
+      "grad_norm": 0.3559363238612658,
+      "learning_rate": 4.398496240601504e-05,
+      "loss": 0.4437,
+      "step": 555
+    },
+    {
+      "epoch": 0.6264788732394366,
+      "grad_norm": 0.3843902875225514,
+      "learning_rate": 4.39640768588137e-05,
+      "loss": 0.42,
+      "step": 556
+    },
+    {
+      "epoch": 0.6276056338028169,
+      "grad_norm": 0.36186469271967187,
+      "learning_rate": 4.394319131161236e-05,
+      "loss": 0.3991,
+      "step": 557
+    },
+    {
+      "epoch": 0.6287323943661972,
+      "grad_norm": 0.33224653105469515,
+      "learning_rate": 4.3922305764411034e-05,
+      "loss": 0.4303,
+      "step": 558
+    },
+    {
+      "epoch": 0.6298591549295774,
+      "grad_norm": 0.4442020175280569,
+      "learning_rate": 4.3901420217209694e-05,
+      "loss": 0.4325,
+      "step": 559
+    },
+    {
+      "epoch": 0.6309859154929578,
+      "grad_norm": 0.32939188578737433,
+      "learning_rate": 4.3880534670008354e-05,
+      "loss": 0.4145,
+      "step": 560
+    },
+    {
+      "epoch": 0.632112676056338,
+      "grad_norm": 0.3941484544311852,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.4449,
+      "step": 561
+    },
+    {
+      "epoch": 0.6332394366197183,
+      "grad_norm": 0.37320706101610485,
+      "learning_rate": 4.383876357560569e-05,
+      "loss": 0.429,
+      "step": 562
+    },
+    {
+      "epoch": 0.6343661971830986,
+      "grad_norm": 0.37437865774917695,
+      "learning_rate": 4.381787802840435e-05,
+      "loss": 0.4185,
+      "step": 563
+    },
+    {
+      "epoch": 0.6354929577464788,
+      "grad_norm": 0.4286549189097408,
+      "learning_rate": 4.379699248120301e-05,
+      "loss": 0.4292,
+      "step": 564
+    },
+    {
+      "epoch": 0.6366197183098592,
+      "grad_norm": 0.35232103572447265,
+      "learning_rate": 4.377610693400167e-05,
+      "loss": 0.4129,
+      "step": 565
+    },
+    {
+      "epoch": 0.6377464788732394,
+      "grad_norm": 0.37846014546497503,
+      "learning_rate": 4.375522138680034e-05,
+      "loss": 0.4141,
+      "step": 566
+    },
+    {
+      "epoch": 0.6388732394366197,
+      "grad_norm": 0.4434692424607111,
+      "learning_rate": 4.3734335839599e-05,
+      "loss": 0.4075,
+      "step": 567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37421587532396255,
+      "learning_rate": 4.371345029239766e-05,
+      "loss": 0.4087,
+      "step": 568
+    },
+    {
+      "epoch": 0.6411267605633802,
+      "grad_norm": 0.39080496222014216,
+      "learning_rate": 4.369256474519632e-05,
+      "loss": 0.4394,
+      "step": 569
+    },
+    {
+      "epoch": 0.6422535211267606,
+      "grad_norm": 0.3066071449286599,
+      "learning_rate": 4.367167919799499e-05,
+      "loss": 0.4259,
+      "step": 570
+    },
+    {
+      "epoch": 0.6433802816901408,
+      "grad_norm": 0.43110701849869826,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.4267,
+      "step": 571
+    },
+    {
+      "epoch": 0.6445070422535212,
+      "grad_norm": 0.3124331941008899,
+      "learning_rate": 4.3629908103592315e-05,
+      "loss": 0.4092,
+      "step": 572
+    },
+    {
+      "epoch": 0.6456338028169014,
+      "grad_norm": 0.4571726835635694,
+      "learning_rate": 4.3609022556390975e-05,
+      "loss": 0.4033,
+      "step": 573
+    },
+    {
+      "epoch": 0.6467605633802817,
+      "grad_norm": 0.3705030567782246,
+      "learning_rate": 4.358813700918964e-05,
+      "loss": 0.4112,
+      "step": 574
+    },
+    {
+      "epoch": 0.647887323943662,
+      "grad_norm": 0.44415346849679094,
+      "learning_rate": 4.356725146198831e-05,
+      "loss": 0.4459,
+      "step": 575
+    },
+    {
+      "epoch": 0.6490140845070422,
+      "grad_norm": 0.4080778529925401,
+      "learning_rate": 4.354636591478697e-05,
+      "loss": 0.4,
+      "step": 576
+    },
+    {
+      "epoch": 0.6501408450704226,
+      "grad_norm": 0.4163294923487941,
+      "learning_rate": 4.352548036758563e-05,
+      "loss": 0.4256,
+      "step": 577
+    },
+    {
+      "epoch": 0.6512676056338028,
+      "grad_norm": 0.4200663678367453,
+      "learning_rate": 4.3504594820384295e-05,
+      "loss": 0.4517,
+      "step": 578
+    },
+    {
+      "epoch": 0.652394366197183,
+      "grad_norm": 0.42091769996589545,
+      "learning_rate": 4.348370927318296e-05,
+      "loss": 0.4177,
+      "step": 579
+    },
+    {
+      "epoch": 0.6535211267605634,
+      "grad_norm": 0.3307175111414965,
+      "learning_rate": 4.346282372598162e-05,
+      "loss": 0.4279,
+      "step": 580
+    },
+    {
+      "epoch": 0.6546478873239436,
+      "grad_norm": 0.4406406069874574,
+      "learning_rate": 4.344193817878028e-05,
+      "loss": 0.4478,
+      "step": 581
+    },
+    {
+      "epoch": 0.655774647887324,
+      "grad_norm": 0.37920787525299443,
+      "learning_rate": 4.342105263157895e-05,
+      "loss": 0.4318,
+      "step": 582
+    },
+    {
+      "epoch": 0.6569014084507042,
+      "grad_norm": 0.41822171166089017,
+      "learning_rate": 4.3400167084377615e-05,
+      "loss": 0.409,
+      "step": 583
+    },
+    {
+      "epoch": 0.6580281690140845,
+      "grad_norm": 0.3671624982481095,
+      "learning_rate": 4.3379281537176275e-05,
+      "loss": 0.3919,
+      "step": 584
+    },
+    {
+      "epoch": 0.6591549295774648,
+      "grad_norm": 0.4322278406739707,
+      "learning_rate": 4.3358395989974935e-05,
+      "loss": 0.4238,
+      "step": 585
+    },
+    {
+      "epoch": 0.660281690140845,
+      "grad_norm": 0.39790250618223566,
+      "learning_rate": 4.33375104427736e-05,
+      "loss": 0.4391,
+      "step": 586
+    },
+    {
+      "epoch": 0.6614084507042254,
+      "grad_norm": 0.4065653425921878,
+      "learning_rate": 4.331662489557227e-05,
+      "loss": 0.433,
+      "step": 587
+    },
+    {
+      "epoch": 0.6625352112676056,
+      "grad_norm": 0.38700056294166,
+      "learning_rate": 4.329573934837093e-05,
+      "loss": 0.419,
+      "step": 588
+    },
+    {
+      "epoch": 0.663661971830986,
+      "grad_norm": 0.387974246136351,
+      "learning_rate": 4.327485380116959e-05,
+      "loss": 0.4237,
+      "step": 589
+    },
+    {
+      "epoch": 0.6647887323943662,
+      "grad_norm": 0.3990357887007478,
+      "learning_rate": 4.3253968253968256e-05,
+      "loss": 0.4194,
+      "step": 590
+    },
+    {
+      "epoch": 0.6659154929577464,
+      "grad_norm": 0.3496417106380883,
+      "learning_rate": 4.323308270676692e-05,
+      "loss": 0.4079,
+      "step": 591
+    },
+    {
+      "epoch": 0.6670422535211268,
+      "grad_norm": 0.5495630885527686,
+      "learning_rate": 4.321219715956558e-05,
+      "loss": 0.4357,
+      "step": 592
+    },
+    {
+      "epoch": 0.668169014084507,
+      "grad_norm": 0.2900877904850997,
+      "learning_rate": 4.319131161236424e-05,
+      "loss": 0.4043,
+      "step": 593
+    },
+    {
+      "epoch": 0.6692957746478874,
+      "grad_norm": 0.5601474985977506,
+      "learning_rate": 4.317042606516291e-05,
+      "loss": 0.41,
+      "step": 594
+    },
+    {
+      "epoch": 0.6704225352112676,
+      "grad_norm": 0.3021224989529405,
+      "learning_rate": 4.3149540517961576e-05,
+      "loss": 0.408,
+      "step": 595
+    },
+    {
+      "epoch": 0.6715492957746478,
+      "grad_norm": 0.4829631218423333,
+      "learning_rate": 4.3128654970760236e-05,
+      "loss": 0.4386,
+      "step": 596
+    },
+    {
+      "epoch": 0.6726760563380282,
+      "grad_norm": 0.359916714114907,
+      "learning_rate": 4.3107769423558896e-05,
+      "loss": 0.3914,
+      "step": 597
+    },
+    {
+      "epoch": 0.6738028169014084,
+      "grad_norm": 0.4765698815625955,
+      "learning_rate": 4.308688387635756e-05,
+      "loss": 0.4005,
+      "step": 598
+    },
+    {
+      "epoch": 0.6749295774647888,
+      "grad_norm": 1.237364168501525,
+      "learning_rate": 4.306599832915623e-05,
+      "loss": 0.4228,
+      "step": 599
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.5930283676015237,
+      "learning_rate": 4.304511278195489e-05,
+      "loss": 0.4148,
+      "step": 600
+    },
+    {
+      "epoch": 0.6771830985915493,
+      "grad_norm": 0.5209464944628275,
+      "learning_rate": 4.302422723475355e-05,
+      "loss": 0.4123,
+      "step": 601
+    },
+    {
+      "epoch": 0.6783098591549296,
+      "grad_norm": 0.5466948649482853,
+      "learning_rate": 4.3003341687552216e-05,
+      "loss": 0.4427,
+      "step": 602
+    },
+    {
+      "epoch": 0.6794366197183098,
+      "grad_norm": 0.45120228009864977,
+      "learning_rate": 4.298245614035088e-05,
+      "loss": 0.4303,
+      "step": 603
+    },
+    {
+      "epoch": 0.6805633802816902,
+      "grad_norm": 0.4505983118265598,
+      "learning_rate": 4.296157059314954e-05,
+      "loss": 0.4043,
+      "step": 604
+    },
+    {
+      "epoch": 0.6816901408450704,
+      "grad_norm": 0.3585135118397449,
+      "learning_rate": 4.29406850459482e-05,
+      "loss": 0.4061,
+      "step": 605
+    },
+    {
+      "epoch": 0.6828169014084507,
+      "grad_norm": 0.46174446600580643,
+      "learning_rate": 4.291979949874687e-05,
+      "loss": 0.4247,
+      "step": 606
+    },
+    {
+      "epoch": 0.683943661971831,
+      "grad_norm": 0.5222240839271706,
+      "learning_rate": 4.2898913951545536e-05,
+      "loss": 0.4256,
+      "step": 607
+    },
+    {
+      "epoch": 0.6850704225352112,
+      "grad_norm": 0.3614268309232858,
+      "learning_rate": 4.2878028404344196e-05,
+      "loss": 0.4367,
+      "step": 608
+    },
+    {
+      "epoch": 0.6861971830985916,
+      "grad_norm": 0.5036220549816727,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.4162,
+      "step": 609
+    },
+    {
+      "epoch": 0.6873239436619718,
+      "grad_norm": 0.3363253080485147,
+      "learning_rate": 4.283625730994152e-05,
+      "loss": 0.4245,
+      "step": 610
+    },
+    {
+      "epoch": 0.6884507042253522,
+      "grad_norm": 0.46033573962506164,
+      "learning_rate": 4.281537176274018e-05,
+      "loss": 0.4215,
+      "step": 611
+    },
+    {
+      "epoch": 0.6895774647887324,
+      "grad_norm": 0.3247787059168572,
+      "learning_rate": 4.279448621553885e-05,
+      "loss": 0.419,
+      "step": 612
+    },
+    {
+      "epoch": 0.6907042253521126,
+      "grad_norm": 0.4447384588517809,
+      "learning_rate": 4.277360066833751e-05,
+      "loss": 0.4254,
+      "step": 613
+    },
+    {
+      "epoch": 0.691830985915493,
+      "grad_norm": 0.32317206065539833,
+      "learning_rate": 4.2752715121136177e-05,
+      "loss": 0.4352,
+      "step": 614
+    },
+    {
+      "epoch": 0.6929577464788732,
+      "grad_norm": 0.4461717956132611,
+      "learning_rate": 4.2731829573934837e-05,
+      "loss": 0.3882,
+      "step": 615
+    },
+    {
+      "epoch": 0.6940845070422536,
+      "grad_norm": 0.40128470765817387,
+      "learning_rate": 4.27109440267335e-05,
+      "loss": 0.4254,
+      "step": 616
+    },
+    {
+      "epoch": 0.6952112676056338,
+      "grad_norm": 0.40703672121281326,
+      "learning_rate": 4.269005847953216e-05,
+      "loss": 0.4026,
+      "step": 617
+    },
+    {
+      "epoch": 0.6963380281690141,
+      "grad_norm": 0.3542224954088699,
+      "learning_rate": 4.266917293233083e-05,
+      "loss": 0.41,
+      "step": 618
+    },
+    {
+      "epoch": 0.6974647887323944,
+      "grad_norm": 0.45578234504211723,
+      "learning_rate": 4.264828738512949e-05,
+      "loss": 0.41,
+      "step": 619
+    },
+    {
+      "epoch": 0.6985915492957746,
+      "grad_norm": 0.4080317444602256,
+      "learning_rate": 4.262740183792816e-05,
+      "loss": 0.4192,
+      "step": 620
+    },
+    {
+      "epoch": 0.699718309859155,
+      "grad_norm": 0.4274600706527524,
+      "learning_rate": 4.260651629072682e-05,
+      "loss": 0.4449,
+      "step": 621
+    },
+    {
+      "epoch": 0.7008450704225352,
+      "grad_norm": 0.4349579841624264,
+      "learning_rate": 4.2585630743525484e-05,
+      "loss": 0.4139,
+      "step": 622
+    },
+    {
+      "epoch": 0.7019718309859155,
+      "grad_norm": 0.370472320194136,
+      "learning_rate": 4.2564745196324144e-05,
+      "loss": 0.409,
+      "step": 623
+    },
+    {
+      "epoch": 0.7030985915492958,
+      "grad_norm": 0.3718077496564049,
+      "learning_rate": 4.254385964912281e-05,
+      "loss": 0.4175,
+      "step": 624
+    },
+    {
+      "epoch": 0.704225352112676,
+      "grad_norm": 0.4276658047382789,
+      "learning_rate": 4.252297410192147e-05,
+      "loss": 0.4424,
+      "step": 625
+    },
+    {
+      "epoch": 0.7053521126760564,
+      "grad_norm": 0.3194214007233317,
+      "learning_rate": 4.250208855472014e-05,
+      "loss": 0.4014,
+      "step": 626
+    },
+    {
+      "epoch": 0.7064788732394366,
+      "grad_norm": 0.48810927800348675,
+      "learning_rate": 4.24812030075188e-05,
+      "loss": 0.4408,
+      "step": 627
+    },
+    {
+      "epoch": 0.7076056338028169,
+      "grad_norm": 0.41652474245146415,
+      "learning_rate": 4.2460317460317464e-05,
+      "loss": 0.4301,
+      "step": 628
+    },
+    {
+      "epoch": 0.7087323943661972,
+      "grad_norm": 0.3835655312696598,
+      "learning_rate": 4.2439431913116124e-05,
+      "loss": 0.4206,
+      "step": 629
+    },
+    {
+      "epoch": 0.7098591549295775,
+      "grad_norm": 0.45662664167441197,
+      "learning_rate": 4.241854636591479e-05,
+      "loss": 0.4017,
+      "step": 630
+    },
+    {
+      "epoch": 0.7109859154929578,
+      "grad_norm": 0.42170162230275204,
+      "learning_rate": 4.239766081871345e-05,
+      "loss": 0.4262,
+      "step": 631
+    },
+    {
+      "epoch": 0.712112676056338,
+      "grad_norm": 0.5617019107854335,
+      "learning_rate": 4.237677527151212e-05,
+      "loss": 0.413,
+      "step": 632
+    },
+    {
+      "epoch": 0.7132394366197183,
+      "grad_norm": 0.44358045519069617,
+      "learning_rate": 4.235588972431078e-05,
+      "loss": 0.3937,
+      "step": 633
+    },
+    {
+      "epoch": 0.7143661971830986,
+      "grad_norm": 0.38692996860487894,
+      "learning_rate": 4.2335004177109444e-05,
+      "loss": 0.4267,
+      "step": 634
+    },
+    {
+      "epoch": 0.7154929577464789,
+      "grad_norm": 0.4348368744099167,
+      "learning_rate": 4.2314118629908104e-05,
+      "loss": 0.4096,
+      "step": 635
+    },
+    {
+      "epoch": 0.7166197183098592,
+      "grad_norm": 0.3630151334921331,
+      "learning_rate": 4.229323308270677e-05,
+      "loss": 0.4127,
+      "step": 636
+    },
+    {
+      "epoch": 0.7177464788732394,
+      "grad_norm": 0.34479478722909607,
+      "learning_rate": 4.227234753550543e-05,
+      "loss": 0.4125,
+      "step": 637
+    },
+    {
+      "epoch": 0.7188732394366197,
+      "grad_norm": 0.33161591976676635,
+      "learning_rate": 4.22514619883041e-05,
+      "loss": 0.4234,
+      "step": 638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35947364399414766,
+      "learning_rate": 4.223057644110276e-05,
+      "loss": 0.4,
+      "step": 639
+    },
+    {
+      "epoch": 0.7211267605633803,
+      "grad_norm": 0.3871104208671629,
+      "learning_rate": 4.2209690893901424e-05,
+      "loss": 0.4337,
+      "step": 640
+    },
+    {
+      "epoch": 0.7222535211267606,
+      "grad_norm": 0.42153964326186993,
+      "learning_rate": 4.2188805346700084e-05,
+      "loss": 0.4334,
+      "step": 641
+    },
+    {
+      "epoch": 0.7233802816901408,
+      "grad_norm": 0.34333286440381905,
+      "learning_rate": 4.216791979949875e-05,
+      "loss": 0.4161,
+      "step": 642
+    },
+    {
+      "epoch": 0.7245070422535211,
+      "grad_norm": 0.37528246725660674,
+      "learning_rate": 4.214703425229741e-05,
+      "loss": 0.4154,
+      "step": 643
+    },
+    {
+      "epoch": 0.7256338028169014,
+      "grad_norm": 0.390514048224168,
+      "learning_rate": 4.212614870509608e-05,
+      "loss": 0.4154,
+      "step": 644
+    },
+    {
+      "epoch": 0.7267605633802817,
+      "grad_norm": 0.3000205982909856,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 0.4087,
+      "step": 645
+    },
+    {
+      "epoch": 0.727887323943662,
+      "grad_norm": 0.3672658159191762,
+      "learning_rate": 4.2084377610693405e-05,
+      "loss": 0.4308,
+      "step": 646
+    },
+    {
+      "epoch": 0.7290140845070423,
+      "grad_norm": 0.41532606867614813,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.4084,
+      "step": 647
+    },
+    {
+      "epoch": 0.7301408450704225,
+      "grad_norm": 0.35732639249142395,
+      "learning_rate": 4.2042606516290725e-05,
+      "loss": 0.4204,
+      "step": 648
+    },
+    {
+      "epoch": 0.7312676056338028,
+      "grad_norm": 0.3849385876548144,
+      "learning_rate": 4.202172096908939e-05,
+      "loss": 0.4281,
+      "step": 649
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 0.38047640795559845,
+      "learning_rate": 4.200083542188806e-05,
+      "loss": 0.4059,
+      "step": 650
+    },
+    {
+      "epoch": 0.7335211267605634,
+      "grad_norm": 0.3837923395192255,
+      "learning_rate": 4.197994987468672e-05,
+      "loss": 0.4313,
+      "step": 651
+    },
+    {
+      "epoch": 0.7346478873239437,
+      "grad_norm": 0.3774699264614354,
+      "learning_rate": 4.195906432748538e-05,
+      "loss": 0.3941,
+      "step": 652
+    },
+    {
+      "epoch": 0.7357746478873239,
+      "grad_norm": 0.2985838697242696,
+      "learning_rate": 4.1938178780284045e-05,
+      "loss": 0.4153,
+      "step": 653
+    },
+    {
+      "epoch": 0.7369014084507042,
+      "grad_norm": 0.38164814201138764,
+      "learning_rate": 4.191729323308271e-05,
+      "loss": 0.4143,
+      "step": 654
+    },
+    {
+      "epoch": 0.7380281690140845,
+      "grad_norm": 0.3620891871835965,
+      "learning_rate": 4.189640768588137e-05,
+      "loss": 0.4231,
+      "step": 655
+    },
+    {
+      "epoch": 0.7391549295774648,
+      "grad_norm": 0.3834560816016984,
+      "learning_rate": 4.187552213868003e-05,
+      "loss": 0.392,
+      "step": 656
+    },
+    {
+      "epoch": 0.7402816901408451,
+      "grad_norm": 0.40285992655625297,
+      "learning_rate": 4.18546365914787e-05,
+      "loss": 0.4297,
+      "step": 657
+    },
+    {
+      "epoch": 0.7414084507042253,
+      "grad_norm": 0.4044847340584241,
+      "learning_rate": 4.1833751044277365e-05,
+      "loss": 0.4161,
+      "step": 658
+    },
+    {
+      "epoch": 0.7425352112676057,
+      "grad_norm": 0.4195935788052179,
+      "learning_rate": 4.1812865497076025e-05,
+      "loss": 0.4137,
+      "step": 659
+    },
+    {
+      "epoch": 0.7436619718309859,
+      "grad_norm": 0.3665940252411504,
+      "learning_rate": 4.1791979949874685e-05,
+      "loss": 0.4427,
+      "step": 660
+    },
+    {
+      "epoch": 0.7447887323943662,
+      "grad_norm": 0.49282768353460193,
+      "learning_rate": 4.177109440267335e-05,
+      "loss": 0.3939,
+      "step": 661
+    },
+    {
+      "epoch": 0.7459154929577465,
+      "grad_norm": 0.3789428349328341,
+      "learning_rate": 4.175020885547202e-05,
+      "loss": 0.3978,
+      "step": 662
+    },
+    {
+      "epoch": 0.7470422535211267,
+      "grad_norm": 0.4423781388327986,
+      "learning_rate": 4.172932330827068e-05,
+      "loss": 0.423,
+      "step": 663
+    },
+    {
+      "epoch": 0.7481690140845071,
+      "grad_norm": 0.38796006159727037,
+      "learning_rate": 4.170843776106934e-05,
+      "loss": 0.4138,
+      "step": 664
+    },
+    {
+      "epoch": 0.7492957746478873,
+      "grad_norm": 0.34306544342853107,
+      "learning_rate": 4.1687552213868005e-05,
+      "loss": 0.3977,
+      "step": 665
+    },
+    {
+      "epoch": 0.7504225352112676,
+      "grad_norm": 0.4288434223869902,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4003,
+      "step": 666
+    },
+    {
+      "epoch": 0.7515492957746479,
+      "grad_norm": 0.4006094181432751,
+      "learning_rate": 4.164578111946533e-05,
+      "loss": 0.4048,
+      "step": 667
+    },
+    {
+      "epoch": 0.7526760563380281,
+      "grad_norm": 0.40733982774726674,
+      "learning_rate": 4.162489557226399e-05,
+      "loss": 0.3939,
+      "step": 668
+    },
+    {
+      "epoch": 0.7538028169014085,
+      "grad_norm": 0.4063549027527979,
+      "learning_rate": 4.160401002506266e-05,
+      "loss": 0.4223,
+      "step": 669
+    },
+    {
+      "epoch": 0.7549295774647887,
+      "grad_norm": 0.45687019171833976,
+      "learning_rate": 4.1583124477861326e-05,
+      "loss": 0.4185,
+      "step": 670
+    },
+    {
+      "epoch": 0.756056338028169,
+      "grad_norm": 0.36681775649393633,
+      "learning_rate": 4.1562238930659986e-05,
+      "loss": 0.4083,
+      "step": 671
+    },
+    {
+      "epoch": 0.7571830985915493,
+      "grad_norm": 0.5537705015482887,
+      "learning_rate": 4.1541353383458646e-05,
+      "loss": 0.4225,
+      "step": 672
+    },
+    {
+      "epoch": 0.7583098591549295,
+      "grad_norm": 0.4108896243173496,
+      "learning_rate": 4.152046783625731e-05,
+      "loss": 0.3993,
+      "step": 673
+    },
+    {
+      "epoch": 0.7594366197183099,
+      "grad_norm": 0.4180315200055127,
+      "learning_rate": 4.149958228905598e-05,
+      "loss": 0.4069,
+      "step": 674
+    },
+    {
+      "epoch": 0.7605633802816901,
+      "grad_norm": 0.43974251385905005,
+      "learning_rate": 4.147869674185464e-05,
+      "loss": 0.4036,
+      "step": 675
+    },
+    {
+      "epoch": 0.7616901408450705,
+      "grad_norm": 0.46133112591194336,
+      "learning_rate": 4.14578111946533e-05,
+      "loss": 0.4151,
+      "step": 676
+    },
+    {
+      "epoch": 0.7628169014084507,
+      "grad_norm": 0.4285557758637548,
+      "learning_rate": 4.1436925647451966e-05,
+      "loss": 0.4091,
+      "step": 677
+    },
+    {
+      "epoch": 0.763943661971831,
+      "grad_norm": 0.45188384956398886,
+      "learning_rate": 4.141604010025063e-05,
+      "loss": 0.4013,
+      "step": 678
+    },
+    {
+      "epoch": 0.7650704225352113,
+      "grad_norm": 0.378944363878999,
+      "learning_rate": 4.139515455304929e-05,
+      "loss": 0.3954,
+      "step": 679
+    },
+    {
+      "epoch": 0.7661971830985915,
+      "grad_norm": 0.4652707802166915,
+      "learning_rate": 4.137426900584795e-05,
+      "loss": 0.4059,
+      "step": 680
+    },
+    {
+      "epoch": 0.7673239436619719,
+      "grad_norm": 0.3315608943628699,
+      "learning_rate": 4.135338345864662e-05,
+      "loss": 0.4301,
+      "step": 681
+    },
+    {
+      "epoch": 0.7684507042253521,
+      "grad_norm": 0.3764928938013357,
+      "learning_rate": 4.1332497911445286e-05,
+      "loss": 0.4014,
+      "step": 682
+    },
+    {
+      "epoch": 0.7695774647887323,
+      "grad_norm": 0.28199550514715443,
+      "learning_rate": 4.1311612364243946e-05,
+      "loss": 0.3836,
+      "step": 683
+    },
+    {
+      "epoch": 0.7707042253521127,
+      "grad_norm": 0.3754349031218491,
+      "learning_rate": 4.1290726817042606e-05,
+      "loss": 0.4017,
+      "step": 684
+    },
+    {
+      "epoch": 0.7718309859154929,
+      "grad_norm": 0.29913461312362266,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 0.384,
+      "step": 685
+    },
+    {
+      "epoch": 0.7729577464788733,
+      "grad_norm": 0.45858396872072255,
+      "learning_rate": 4.124895572263994e-05,
+      "loss": 0.4308,
+      "step": 686
+    },
+    {
+      "epoch": 0.7740845070422535,
+      "grad_norm": 0.3412505295746519,
+      "learning_rate": 4.12280701754386e-05,
+      "loss": 0.4168,
+      "step": 687
+    },
+    {
+      "epoch": 0.7752112676056339,
+      "grad_norm": 0.3858887385455696,
+      "learning_rate": 4.120718462823726e-05,
+      "loss": 0.406,
+      "step": 688
+    },
+    {
+      "epoch": 0.7763380281690141,
+      "grad_norm": 0.3345909319603766,
+      "learning_rate": 4.118629908103592e-05,
+      "loss": 0.3944,
+      "step": 689
+    },
+    {
+      "epoch": 0.7774647887323943,
+      "grad_norm": 0.40269397288356046,
+      "learning_rate": 4.116541353383459e-05,
+      "loss": 0.405,
+      "step": 690
+    },
+    {
+      "epoch": 0.7785915492957747,
+      "grad_norm": 0.3737994454866607,
+      "learning_rate": 4.114452798663325e-05,
+      "loss": 0.4206,
+      "step": 691
+    },
+    {
+      "epoch": 0.7797183098591549,
+      "grad_norm": 0.45903125849438714,
+      "learning_rate": 4.112364243943191e-05,
+      "loss": 0.4065,
+      "step": 692
+    },
+    {
+      "epoch": 0.7808450704225353,
+      "grad_norm": 0.4034087006226384,
+      "learning_rate": 4.110275689223057e-05,
+      "loss": 0.4383,
+      "step": 693
+    },
+    {
+      "epoch": 0.7819718309859155,
+      "grad_norm": 0.4349274997587744,
+      "learning_rate": 4.1081871345029247e-05,
+      "loss": 0.42,
+      "step": 694
+    },
+    {
+      "epoch": 0.7830985915492957,
+      "grad_norm": 0.3930297481138719,
+      "learning_rate": 4.1060985797827907e-05,
+      "loss": 0.4071,
+      "step": 695
+    },
+    {
+      "epoch": 0.7842253521126761,
+      "grad_norm": 0.5192874311515382,
+      "learning_rate": 4.1040100250626567e-05,
+      "loss": 0.4117,
+      "step": 696
+    },
+    {
+      "epoch": 0.7853521126760563,
+      "grad_norm": 0.38351615215668994,
+      "learning_rate": 4.1019214703425227e-05,
+      "loss": 0.3994,
+      "step": 697
+    },
+    {
+      "epoch": 0.7864788732394367,
+      "grad_norm": 0.4572728634907756,
+      "learning_rate": 4.09983291562239e-05,
+      "loss": 0.3895,
+      "step": 698
+    },
+    {
+      "epoch": 0.7876056338028169,
+      "grad_norm": 0.46697618002820673,
+      "learning_rate": 4.097744360902256e-05,
+      "loss": 0.4148,
+      "step": 699
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.45965026461144826,
+      "learning_rate": 4.095655806182122e-05,
+      "loss": 0.4156,
+      "step": 700
+    },
+    {
+      "epoch": 0.7898591549295775,
+      "grad_norm": 0.4093107614744624,
+      "learning_rate": 4.093567251461988e-05,
+      "loss": 0.4292,
+      "step": 701
+    },
+    {
+      "epoch": 0.7909859154929577,
+      "grad_norm": 0.4486831843348481,
+      "learning_rate": 4.0914786967418554e-05,
+      "loss": 0.4073,
+      "step": 702
+    },
+    {
+      "epoch": 0.7921126760563381,
+      "grad_norm": 0.48346726109054283,
+      "learning_rate": 4.0893901420217214e-05,
+      "loss": 0.4462,
+      "step": 703
+    },
+    {
+      "epoch": 0.7932394366197183,
+      "grad_norm": 0.4946289570640352,
+      "learning_rate": 4.0873015873015874e-05,
+      "loss": 0.4141,
+      "step": 704
+    },
+    {
+      "epoch": 0.7943661971830986,
+      "grad_norm": 0.43699923683252617,
+      "learning_rate": 4.0852130325814534e-05,
+      "loss": 0.4139,
+      "step": 705
+    },
+    {
+      "epoch": 0.7954929577464789,
+      "grad_norm": 0.447209015951833,
+      "learning_rate": 4.08312447786132e-05,
+      "loss": 0.4063,
+      "step": 706
+    },
+    {
+      "epoch": 0.7966197183098591,
+      "grad_norm": 0.41188678028975295,
+      "learning_rate": 4.081035923141187e-05,
+      "loss": 0.4195,
+      "step": 707
+    },
+    {
+      "epoch": 0.7977464788732395,
+      "grad_norm": 0.4360319548470077,
+      "learning_rate": 4.078947368421053e-05,
+      "loss": 0.4181,
+      "step": 708
+    },
+    {
+      "epoch": 0.7988732394366197,
+      "grad_norm": 0.44710924458871487,
+      "learning_rate": 4.076858813700919e-05,
+      "loss": 0.4165,
+      "step": 709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35901189271573075,
+      "learning_rate": 4.0747702589807854e-05,
+      "loss": 0.4042,
+      "step": 710
+    },
+    {
+      "epoch": 0.8011267605633803,
+      "grad_norm": 0.37421550134735554,
+      "learning_rate": 4.072681704260652e-05,
+      "loss": 0.4133,
+      "step": 711
+    },
+    {
+      "epoch": 0.8022535211267605,
+      "grad_norm": 0.38518246190045313,
+      "learning_rate": 4.070593149540518e-05,
+      "loss": 0.4252,
+      "step": 712
+    },
+    {
+      "epoch": 0.8033802816901409,
+      "grad_norm": 0.35316967468640464,
+      "learning_rate": 4.068504594820384e-05,
+      "loss": 0.3921,
+      "step": 713
+    },
+    {
+      "epoch": 0.8045070422535211,
+      "grad_norm": 0.45881821126358313,
+      "learning_rate": 4.066416040100251e-05,
+      "loss": 0.4247,
+      "step": 714
+    },
+    {
+      "epoch": 0.8056338028169014,
+      "grad_norm": 0.35069913984167855,
+      "learning_rate": 4.0643274853801174e-05,
+      "loss": 0.4269,
+      "step": 715
+    },
+    {
+      "epoch": 0.8067605633802817,
+      "grad_norm": 0.3835410130576258,
+      "learning_rate": 4.0622389306599834e-05,
+      "loss": 0.3898,
+      "step": 716
+    },
+    {
+      "epoch": 0.8078873239436619,
+      "grad_norm": 0.4704560512938768,
+      "learning_rate": 4.0601503759398494e-05,
+      "loss": 0.4128,
+      "step": 717
+    },
+    {
+      "epoch": 0.8090140845070423,
+      "grad_norm": 0.36196175261639196,
+      "learning_rate": 4.058061821219716e-05,
+      "loss": 0.3914,
+      "step": 718
+    },
+    {
+      "epoch": 0.8101408450704225,
+      "grad_norm": 0.5037656792609231,
+      "learning_rate": 4.055973266499583e-05,
+      "loss": 0.4097,
+      "step": 719
+    },
+    {
+      "epoch": 0.8112676056338028,
+      "grad_norm": 0.4524163850817384,
+      "learning_rate": 4.053884711779449e-05,
+      "loss": 0.4108,
+      "step": 720
+    },
+    {
+      "epoch": 0.8123943661971831,
+      "grad_norm": 0.4349260980067756,
+      "learning_rate": 4.051796157059315e-05,
+      "loss": 0.4144,
+      "step": 721
+    },
+    {
+      "epoch": 0.8135211267605634,
+      "grad_norm": 0.5243382584421026,
+      "learning_rate": 4.0497076023391814e-05,
+      "loss": 0.4142,
+      "step": 722
+    },
+    {
+      "epoch": 0.8146478873239437,
+      "grad_norm": 0.391790515054285,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.4012,
+      "step": 723
+    },
+    {
+      "epoch": 0.8157746478873239,
+      "grad_norm": 0.48320887589903105,
+      "learning_rate": 4.045530492898914e-05,
+      "loss": 0.4117,
+      "step": 724
+    },
+    {
+      "epoch": 0.8169014084507042,
+      "grad_norm": 0.41303488238708935,
+      "learning_rate": 4.04344193817878e-05,
+      "loss": 0.3809,
+      "step": 725
+    },
+    {
+      "epoch": 0.8180281690140845,
+      "grad_norm": 0.4348424868731927,
+      "learning_rate": 4.041353383458647e-05,
+      "loss": 0.4766,
+      "step": 726
+    },
+    {
+      "epoch": 0.8191549295774648,
+      "grad_norm": 0.37121791299850376,
+      "learning_rate": 4.0392648287385135e-05,
+      "loss": 0.394,
+      "step": 727
+    },
+    {
+      "epoch": 0.8202816901408451,
+      "grad_norm": 0.4798401011692672,
+      "learning_rate": 4.0371762740183795e-05,
+      "loss": 0.4227,
+      "step": 728
+    },
+    {
+      "epoch": 0.8214084507042253,
+      "grad_norm": 0.39082215709390694,
+      "learning_rate": 4.0350877192982455e-05,
+      "loss": 0.4209,
+      "step": 729
+    },
+    {
+      "epoch": 0.8225352112676056,
+      "grad_norm": 0.43375941551110303,
+      "learning_rate": 4.032999164578112e-05,
+      "loss": 0.4225,
+      "step": 730
+    },
+    {
+      "epoch": 0.8236619718309859,
+      "grad_norm": 0.3237228830696122,
+      "learning_rate": 4.030910609857979e-05,
+      "loss": 0.3872,
+      "step": 731
+    },
+    {
+      "epoch": 0.8247887323943662,
+      "grad_norm": 0.34724647024342176,
+      "learning_rate": 4.028822055137845e-05,
+      "loss": 0.4032,
+      "step": 732
+    },
+    {
+      "epoch": 0.8259154929577465,
+      "grad_norm": 0.31407638077523553,
+      "learning_rate": 4.026733500417711e-05,
+      "loss": 0.3883,
+      "step": 733
+    },
+    {
+      "epoch": 0.8270422535211268,
+      "grad_norm": 0.3613395057755356,
+      "learning_rate": 4.0246449456975775e-05,
+      "loss": 0.4154,
+      "step": 734
+    },
+    {
+      "epoch": 0.828169014084507,
+      "grad_norm": 0.32893242131018746,
+      "learning_rate": 4.022556390977444e-05,
+      "loss": 0.425,
+      "step": 735
+    },
+    {
+      "epoch": 0.8292957746478873,
+      "grad_norm": 0.3969348591140389,
+      "learning_rate": 4.02046783625731e-05,
+      "loss": 0.4305,
+      "step": 736
+    },
+    {
+      "epoch": 0.8304225352112676,
+      "grad_norm": 0.3481060320315369,
+      "learning_rate": 4.018379281537176e-05,
+      "loss": 0.3991,
+      "step": 737
+    },
+    {
+      "epoch": 0.8315492957746479,
+      "grad_norm": 0.31650648249000474,
+      "learning_rate": 4.016290726817043e-05,
+      "loss": 0.4033,
+      "step": 738
+    },
+    {
+      "epoch": 0.8326760563380282,
+      "grad_norm": 0.3965538972614481,
+      "learning_rate": 4.0142021720969095e-05,
+      "loss": 0.3981,
+      "step": 739
+    },
+    {
+      "epoch": 0.8338028169014085,
+      "grad_norm": 0.3406795949884074,
+      "learning_rate": 4.0121136173767755e-05,
+      "loss": 0.3969,
+      "step": 740
+    },
+    {
+      "epoch": 0.8349295774647887,
+      "grad_norm": 0.3635583402480603,
+      "learning_rate": 4.0100250626566415e-05,
+      "loss": 0.4061,
+      "step": 741
+    },
+    {
+      "epoch": 0.836056338028169,
+      "grad_norm": 0.40519489587384744,
+      "learning_rate": 4.007936507936508e-05,
+      "loss": 0.4143,
+      "step": 742
+    },
+    {
+      "epoch": 0.8371830985915493,
+      "grad_norm": 0.3593266084385044,
+      "learning_rate": 4.005847953216375e-05,
+      "loss": 0.3991,
+      "step": 743
+    },
+    {
+      "epoch": 0.8383098591549296,
+      "grad_norm": 0.34564675037719295,
+      "learning_rate": 4.003759398496241e-05,
+      "loss": 0.4058,
+      "step": 744
+    },
+    {
+      "epoch": 0.8394366197183099,
+      "grad_norm": 0.3889580008045617,
+      "learning_rate": 4.001670843776107e-05,
+      "loss": 0.4357,
+      "step": 745
+    },
+    {
+      "epoch": 0.8405633802816901,
+      "grad_norm": 0.3076419722480114,
+      "learning_rate": 3.9995822890559735e-05,
+      "loss": 0.4215,
+      "step": 746
+    },
+    {
+      "epoch": 0.8416901408450704,
+      "grad_norm": 0.3953958807659964,
+      "learning_rate": 3.9974937343358395e-05,
+      "loss": 0.4282,
+      "step": 747
+    },
+    {
+      "epoch": 0.8428169014084507,
+      "grad_norm": 0.3510246787976059,
+      "learning_rate": 3.995405179615706e-05,
+      "loss": 0.405,
+      "step": 748
+    },
+    {
+      "epoch": 0.843943661971831,
+      "grad_norm": 0.3904858752741403,
+      "learning_rate": 3.993316624895572e-05,
+      "loss": 0.4017,
+      "step": 749
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.3466183465990686,
+      "learning_rate": 3.991228070175439e-05,
+      "loss": 0.4063,
+      "step": 750
+    },
+    {
+      "epoch": 0.8461971830985916,
+      "grad_norm": 0.4145982550809233,
+      "learning_rate": 3.989139515455305e-05,
+      "loss": 0.3943,
+      "step": 751
+    },
+    {
+      "epoch": 0.8473239436619718,
+      "grad_norm": 0.3056074309746521,
+      "learning_rate": 3.9870509607351716e-05,
+      "loss": 0.4096,
+      "step": 752
+    },
+    {
+      "epoch": 0.8484507042253521,
+      "grad_norm": 0.40235174127762036,
+      "learning_rate": 3.9849624060150376e-05,
+      "loss": 0.4344,
+      "step": 753
+    },
+    {
+      "epoch": 0.8495774647887324,
+      "grad_norm": 0.3331530224010519,
+      "learning_rate": 3.982873851294904e-05,
+      "loss": 0.4019,
+      "step": 754
+    },
+    {
+      "epoch": 0.8507042253521127,
+      "grad_norm": 0.3395688761807231,
+      "learning_rate": 3.98078529657477e-05,
+      "loss": 0.4223,
+      "step": 755
+    },
+    {
+      "epoch": 0.851830985915493,
+      "grad_norm": 0.31344411916177906,
+      "learning_rate": 3.978696741854637e-05,
+      "loss": 0.41,
+      "step": 756
+    },
+    {
+      "epoch": 0.8529577464788732,
+      "grad_norm": 0.31052339494796105,
+      "learning_rate": 3.976608187134503e-05,
+      "loss": 0.3863,
+      "step": 757
+    },
+    {
+      "epoch": 0.8540845070422535,
+      "grad_norm": 0.3505115182352671,
+      "learning_rate": 3.9745196324143696e-05,
+      "loss": 0.4043,
+      "step": 758
+    },
+    {
+      "epoch": 0.8552112676056338,
+      "grad_norm": 0.2989368214428119,
+      "learning_rate": 3.9724310776942356e-05,
+      "loss": 0.4027,
+      "step": 759
+    },
+    {
+      "epoch": 0.856338028169014,
+      "grad_norm": 0.3583425560695883,
+      "learning_rate": 3.970342522974102e-05,
+      "loss": 0.4232,
+      "step": 760
+    },
+    {
+      "epoch": 0.8574647887323944,
+      "grad_norm": 0.4429332714250863,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.4031,
+      "step": 761
+    },
+    {
+      "epoch": 0.8585915492957746,
+      "grad_norm": 0.34496606200669994,
+      "learning_rate": 3.966165413533835e-05,
+      "loss": 0.388,
+      "step": 762
+    },
+    {
+      "epoch": 0.859718309859155,
+      "grad_norm": 0.34447312913304845,
+      "learning_rate": 3.964076858813701e-05,
+      "loss": 0.4015,
+      "step": 763
+    },
+    {
+      "epoch": 0.8608450704225352,
+      "grad_norm": 0.4109964850785612,
+      "learning_rate": 3.9619883040935676e-05,
+      "loss": 0.399,
+      "step": 764
+    },
+    {
+      "epoch": 0.8619718309859155,
+      "grad_norm": 0.34614031909152393,
+      "learning_rate": 3.9598997493734336e-05,
+      "loss": 0.4203,
+      "step": 765
+    },
+    {
+      "epoch": 0.8630985915492958,
+      "grad_norm": 0.3786215940986347,
+      "learning_rate": 3.9578111946533e-05,
+      "loss": 0.4048,
+      "step": 766
+    },
+    {
+      "epoch": 0.864225352112676,
+      "grad_norm": 0.3485285199072157,
+      "learning_rate": 3.955722639933166e-05,
+      "loss": 0.3923,
+      "step": 767
+    },
+    {
+      "epoch": 0.8653521126760564,
+      "grad_norm": 0.32117325622732684,
+      "learning_rate": 3.953634085213033e-05,
+      "loss": 0.396,
+      "step": 768
+    },
+    {
+      "epoch": 0.8664788732394366,
+      "grad_norm": 0.3288721214913486,
+      "learning_rate": 3.951545530492899e-05,
+      "loss": 0.4073,
+      "step": 769
+    },
+    {
+      "epoch": 0.8676056338028169,
+      "grad_norm": 0.3553541184041968,
+      "learning_rate": 3.9494569757727656e-05,
+      "loss": 0.3791,
+      "step": 770
+    },
+    {
+      "epoch": 0.8687323943661972,
+      "grad_norm": 0.36597469077692873,
+      "learning_rate": 3.9473684210526316e-05,
+      "loss": 0.3823,
+      "step": 771
+    },
+    {
+      "epoch": 0.8698591549295774,
+      "grad_norm": 0.36679352760459877,
+      "learning_rate": 3.945279866332498e-05,
+      "loss": 0.3898,
+      "step": 772
+    },
+    {
+      "epoch": 0.8709859154929578,
+      "grad_norm": 0.3862878116970243,
+      "learning_rate": 3.943191311612364e-05,
+      "loss": 0.4099,
+      "step": 773
+    },
+    {
+      "epoch": 0.872112676056338,
+      "grad_norm": 0.34475847881480404,
+      "learning_rate": 3.941102756892231e-05,
+      "loss": 0.4084,
+      "step": 774
+    },
+    {
+      "epoch": 0.8732394366197183,
+      "grad_norm": 0.45638654676684215,
+      "learning_rate": 3.939014202172097e-05,
+      "loss": 0.3986,
+      "step": 775
+    },
+    {
+      "epoch": 0.8743661971830986,
+      "grad_norm": 0.35807833510998827,
+      "learning_rate": 3.9369256474519637e-05,
+      "loss": 0.4081,
+      "step": 776
+    },
+    {
+      "epoch": 0.8754929577464788,
+      "grad_norm": 0.5409597238956879,
+      "learning_rate": 3.9348370927318297e-05,
+      "loss": 0.4147,
+      "step": 777
+    },
+    {
+      "epoch": 0.8766197183098592,
+      "grad_norm": 0.3789429877677125,
+      "learning_rate": 3.932748538011696e-05,
+      "loss": 0.4195,
+      "step": 778
+    },
+    {
+      "epoch": 0.8777464788732394,
+      "grad_norm": 0.3709352534507574,
+      "learning_rate": 3.930659983291562e-05,
+      "loss": 0.3851,
+      "step": 779
+    },
+    {
+      "epoch": 0.8788732394366198,
+      "grad_norm": 0.45932668818057254,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 0.4018,
+      "step": 780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3189973459204549,
+      "learning_rate": 3.926482873851295e-05,
+      "loss": 0.4321,
+      "step": 781
+    },
+    {
+      "epoch": 0.8811267605633802,
+      "grad_norm": 0.5214079758077962,
+      "learning_rate": 3.924394319131162e-05,
+      "loss": 0.4237,
+      "step": 782
+    },
+    {
+      "epoch": 0.8822535211267606,
+      "grad_norm": 0.42449649095885933,
+      "learning_rate": 3.922305764411028e-05,
+      "loss": 0.4028,
+      "step": 783
+    },
+    {
+      "epoch": 0.8833802816901408,
+      "grad_norm": 0.42873753000780335,
+      "learning_rate": 3.920217209690894e-05,
+      "loss": 0.3954,
+      "step": 784
+    },
+    {
+      "epoch": 0.8845070422535212,
+      "grad_norm": 0.49566264876473903,
+      "learning_rate": 3.9181286549707604e-05,
+      "loss": 0.3963,
+      "step": 785
+    },
+    {
+      "epoch": 0.8856338028169014,
+      "grad_norm": 0.2995262123704027,
+      "learning_rate": 3.916040100250627e-05,
+      "loss": 0.3963,
+      "step": 786
+    },
+    {
+      "epoch": 0.8867605633802816,
+      "grad_norm": 0.4903480183321258,
+      "learning_rate": 3.913951545530493e-05,
+      "loss": 0.4457,
+      "step": 787
+    },
+    {
+      "epoch": 0.887887323943662,
+      "grad_norm": 0.46991879985880375,
+      "learning_rate": 3.911862990810359e-05,
+      "loss": 0.427,
+      "step": 788
+    },
+    {
+      "epoch": 0.8890140845070422,
+      "grad_norm": 0.3134823915048358,
+      "learning_rate": 3.909774436090226e-05,
+      "loss": 0.3867,
+      "step": 789
+    },
+    {
+      "epoch": 0.8901408450704226,
+      "grad_norm": 0.5545602314022595,
+      "learning_rate": 3.9076858813700924e-05,
+      "loss": 0.4171,
+      "step": 790
+    },
+    {
+      "epoch": 0.8912676056338028,
+      "grad_norm": 0.41881667647515214,
+      "learning_rate": 3.9055973266499584e-05,
+      "loss": 0.4053,
+      "step": 791
+    },
+    {
+      "epoch": 0.8923943661971832,
+      "grad_norm": 0.44463249533252486,
+      "learning_rate": 3.9035087719298244e-05,
+      "loss": 0.405,
+      "step": 792
+    },
+    {
+      "epoch": 0.8935211267605634,
+      "grad_norm": 0.4890949744923107,
+      "learning_rate": 3.901420217209691e-05,
+      "loss": 0.4142,
+      "step": 793
+    },
+    {
+      "epoch": 0.8946478873239436,
+      "grad_norm": 0.38382002153711203,
+      "learning_rate": 3.899331662489558e-05,
+      "loss": 0.3991,
+      "step": 794
+    },
+    {
+      "epoch": 0.895774647887324,
+      "grad_norm": 0.41032789060356845,
+      "learning_rate": 3.897243107769424e-05,
+      "loss": 0.4121,
+      "step": 795
+    },
+    {
+      "epoch": 0.8969014084507042,
+      "grad_norm": 0.44242188192656345,
+      "learning_rate": 3.89515455304929e-05,
+      "loss": 0.412,
+      "step": 796
+    },
+    {
+      "epoch": 0.8980281690140846,
+      "grad_norm": 0.4999281545425411,
+      "learning_rate": 3.8930659983291564e-05,
+      "loss": 0.4168,
+      "step": 797
+    },
+    {
+      "epoch": 0.8991549295774648,
+      "grad_norm": 0.38598639385702344,
+      "learning_rate": 3.890977443609023e-05,
+      "loss": 0.4099,
+      "step": 798
+    },
+    {
+      "epoch": 0.900281690140845,
+      "grad_norm": 0.34803726788055417,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.4274,
+      "step": 799
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.3249736560218246,
+      "learning_rate": 3.886800334168755e-05,
+      "loss": 0.3987,
+      "step": 800
+    },
+    {
+      "epoch": 0.9025352112676056,
+      "grad_norm": 0.37525912151977125,
+      "learning_rate": 3.884711779448622e-05,
+      "loss": 0.4337,
+      "step": 801
+    },
+    {
+      "epoch": 0.903661971830986,
+      "grad_norm": 0.3450630535655991,
+      "learning_rate": 3.8826232247284884e-05,
+      "loss": 0.409,
+      "step": 802
+    },
+    {
+      "epoch": 0.9047887323943662,
+      "grad_norm": 0.41490433438630725,
+      "learning_rate": 3.8805346700083544e-05,
+      "loss": 0.4196,
+      "step": 803
+    },
+    {
+      "epoch": 0.9059154929577464,
+      "grad_norm": 0.3918643716520213,
+      "learning_rate": 3.8784461152882204e-05,
+      "loss": 0.4178,
+      "step": 804
+    },
+    {
+      "epoch": 0.9070422535211268,
+      "grad_norm": 0.3031248324523258,
+      "learning_rate": 3.876357560568087e-05,
+      "loss": 0.3882,
+      "step": 805
+    },
+    {
+      "epoch": 0.908169014084507,
+      "grad_norm": 0.3612974091721475,
+      "learning_rate": 3.874269005847954e-05,
+      "loss": 0.3831,
+      "step": 806
+    },
+    {
+      "epoch": 0.9092957746478874,
+      "grad_norm": 0.3356819962030845,
+      "learning_rate": 3.87218045112782e-05,
+      "loss": 0.4013,
+      "step": 807
+    },
+    {
+      "epoch": 0.9104225352112676,
+      "grad_norm": 0.34249347780367584,
+      "learning_rate": 3.870091896407686e-05,
+      "loss": 0.4033,
+      "step": 808
+    },
+    {
+      "epoch": 0.9115492957746479,
+      "grad_norm": 0.34376175177747587,
+      "learning_rate": 3.8680033416875525e-05,
+      "loss": 0.3801,
+      "step": 809
+    },
+    {
+      "epoch": 0.9126760563380282,
+      "grad_norm": 0.4014809133874393,
+      "learning_rate": 3.865914786967419e-05,
+      "loss": 0.4096,
+      "step": 810
+    },
+    {
+      "epoch": 0.9138028169014084,
+      "grad_norm": 0.39950046560062835,
+      "learning_rate": 3.863826232247285e-05,
+      "loss": 0.3909,
+      "step": 811
+    },
+    {
+      "epoch": 0.9149295774647888,
+      "grad_norm": 0.3209448071853923,
+      "learning_rate": 3.861737677527151e-05,
+      "loss": 0.3981,
+      "step": 812
+    },
+    {
+      "epoch": 0.916056338028169,
+      "grad_norm": 0.38532834965240903,
+      "learning_rate": 3.859649122807018e-05,
+      "loss": 0.3903,
+      "step": 813
+    },
+    {
+      "epoch": 0.9171830985915493,
+      "grad_norm": 0.34248793319108745,
+      "learning_rate": 3.8575605680868845e-05,
+      "loss": 0.4094,
+      "step": 814
+    },
+    {
+      "epoch": 0.9183098591549296,
+      "grad_norm": 0.3360813492097211,
+      "learning_rate": 3.8554720133667505e-05,
+      "loss": 0.4159,
+      "step": 815
+    },
+    {
+      "epoch": 0.9194366197183098,
+      "grad_norm": 0.34850592514190076,
+      "learning_rate": 3.8533834586466165e-05,
+      "loss": 0.4025,
+      "step": 816
+    },
+    {
+      "epoch": 0.9205633802816902,
+      "grad_norm": 0.3708447582880061,
+      "learning_rate": 3.851294903926483e-05,
+      "loss": 0.4182,
+      "step": 817
+    },
+    {
+      "epoch": 0.9216901408450704,
+      "grad_norm": 0.4013945982859129,
+      "learning_rate": 3.84920634920635e-05,
+      "loss": 0.4198,
+      "step": 818
+    },
+    {
+      "epoch": 0.9228169014084507,
+      "grad_norm": 0.35166439502151015,
+      "learning_rate": 3.847117794486216e-05,
+      "loss": 0.4127,
+      "step": 819
+    },
+    {
+      "epoch": 0.923943661971831,
+      "grad_norm": 0.4182281253626696,
+      "learning_rate": 3.845029239766082e-05,
+      "loss": 0.4213,
+      "step": 820
+    },
+    {
+      "epoch": 0.9250704225352113,
+      "grad_norm": 0.45579640979574715,
+      "learning_rate": 3.8429406850459485e-05,
+      "loss": 0.4104,
+      "step": 821
+    },
+    {
+      "epoch": 0.9261971830985916,
+      "grad_norm": 0.3339252060540916,
+      "learning_rate": 3.840852130325815e-05,
+      "loss": 0.4067,
+      "step": 822
+    },
+    {
+      "epoch": 0.9273239436619718,
+      "grad_norm": 0.4981318249725421,
+      "learning_rate": 3.838763575605681e-05,
+      "loss": 0.3988,
+      "step": 823
+    },
+    {
+      "epoch": 0.9284507042253521,
+      "grad_norm": 0.3792710144886669,
+      "learning_rate": 3.836675020885547e-05,
+      "loss": 0.3925,
+      "step": 824
+    },
+    {
+      "epoch": 0.9295774647887324,
+      "grad_norm": 0.46368311223715153,
+      "learning_rate": 3.834586466165413e-05,
+      "loss": 0.4222,
+      "step": 825
+    },
+    {
+      "epoch": 0.9307042253521127,
+      "grad_norm": 0.362841062246029,
+      "learning_rate": 3.8324979114452805e-05,
+      "loss": 0.3936,
+      "step": 826
+    },
+    {
+      "epoch": 0.931830985915493,
+      "grad_norm": 0.43378036177185536,
+      "learning_rate": 3.8304093567251465e-05,
+      "loss": 0.3861,
+      "step": 827
+    },
+    {
+      "epoch": 0.9329577464788732,
+      "grad_norm": 0.33713557262630695,
+      "learning_rate": 3.8283208020050125e-05,
+      "loss": 0.4063,
+      "step": 828
+    },
+    {
+      "epoch": 0.9340845070422535,
+      "grad_norm": 0.45656551755730873,
+      "learning_rate": 3.8262322472848785e-05,
+      "loss": 0.4202,
+      "step": 829
+    },
+    {
+      "epoch": 0.9352112676056338,
+      "grad_norm": 0.4351376754181724,
+      "learning_rate": 3.824143692564746e-05,
+      "loss": 0.3832,
+      "step": 830
+    },
+    {
+      "epoch": 0.9363380281690141,
+      "grad_norm": 0.3703220173375687,
+      "learning_rate": 3.822055137844612e-05,
+      "loss": 0.406,
+      "step": 831
+    },
+    {
+      "epoch": 0.9374647887323944,
+      "grad_norm": 0.44651146569953,
+      "learning_rate": 3.819966583124478e-05,
+      "loss": 0.4045,
+      "step": 832
+    },
+    {
+      "epoch": 0.9385915492957746,
+      "grad_norm": 0.4189212862088426,
+      "learning_rate": 3.817878028404344e-05,
+      "loss": 0.4244,
+      "step": 833
+    },
+    {
+      "epoch": 0.9397183098591549,
+      "grad_norm": 0.45820610740313783,
+      "learning_rate": 3.815789473684211e-05,
+      "loss": 0.4163,
+      "step": 834
+    },
+    {
+      "epoch": 0.9408450704225352,
+      "grad_norm": 0.380217918734662,
+      "learning_rate": 3.813700918964077e-05,
+      "loss": 0.4205,
+      "step": 835
+    },
+    {
+      "epoch": 0.9419718309859155,
+      "grad_norm": 0.407312064783555,
+      "learning_rate": 3.811612364243943e-05,
+      "loss": 0.3974,
+      "step": 836
+    },
+    {
+      "epoch": 0.9430985915492958,
+      "grad_norm": 0.35017334450149956,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.4092,
+      "step": 837
+    },
+    {
+      "epoch": 0.9442253521126761,
+      "grad_norm": 0.38343367952612833,
+      "learning_rate": 3.8074352548036766e-05,
+      "loss": 0.3923,
+      "step": 838
+    },
+    {
+      "epoch": 0.9453521126760563,
+      "grad_norm": 0.3889582171194483,
+      "learning_rate": 3.8053467000835426e-05,
+      "loss": 0.4173,
+      "step": 839
+    },
+    {
+      "epoch": 0.9464788732394366,
+      "grad_norm": 0.45980181954652527,
+      "learning_rate": 3.8032581453634086e-05,
+      "loss": 0.422,
+      "step": 840
+    },
+    {
+      "epoch": 0.9476056338028169,
+      "grad_norm": 0.3297731615524771,
+      "learning_rate": 3.8011695906432746e-05,
+      "loss": 0.404,
+      "step": 841
+    },
+    {
+      "epoch": 0.9487323943661972,
+      "grad_norm": 0.38624507916793943,
+      "learning_rate": 3.799081035923141e-05,
+      "loss": 0.4159,
+      "step": 842
+    },
+    {
+      "epoch": 0.9498591549295775,
+      "grad_norm": 0.32519279849185767,
+      "learning_rate": 3.796992481203008e-05,
+      "loss": 0.4072,
+      "step": 843
+    },
+    {
+      "epoch": 0.9509859154929577,
+      "grad_norm": 0.42925411014050585,
+      "learning_rate": 3.794903926482874e-05,
+      "loss": 0.4143,
+      "step": 844
+    },
+    {
+      "epoch": 0.952112676056338,
+      "grad_norm": 0.31153163118316113,
+      "learning_rate": 3.79281537176274e-05,
+      "loss": 0.4076,
+      "step": 845
+    },
+    {
+      "epoch": 0.9532394366197183,
+      "grad_norm": 0.43824860641584196,
+      "learning_rate": 3.7907268170426066e-05,
+      "loss": 0.3763,
+      "step": 846
+    },
+    {
+      "epoch": 0.9543661971830986,
+      "grad_norm": 0.31884786049993985,
+      "learning_rate": 3.788638262322473e-05,
+      "loss": 0.4044,
+      "step": 847
+    },
+    {
+      "epoch": 0.9554929577464789,
+      "grad_norm": 0.4107221881754472,
+      "learning_rate": 3.786549707602339e-05,
+      "loss": 0.4054,
+      "step": 848
+    },
+    {
+      "epoch": 0.9566197183098591,
+      "grad_norm": 0.34924994082934024,
+      "learning_rate": 3.784461152882205e-05,
+      "loss": 0.3931,
+      "step": 849
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 0.3864084211943787,
+      "learning_rate": 3.782372598162072e-05,
+      "loss": 0.4181,
+      "step": 850
+    },
+    {
+      "epoch": 0.9588732394366197,
+      "grad_norm": 0.31475807218831736,
+      "learning_rate": 3.7802840434419386e-05,
+      "loss": 0.3796,
+      "step": 851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3679848297154997,
+      "learning_rate": 3.7781954887218046e-05,
+      "loss": 0.3829,
+      "step": 852
+    },
+    {
+      "epoch": 0.9611267605633803,
+      "grad_norm": 0.2816290223191142,
+      "learning_rate": 3.7761069340016706e-05,
+      "loss": 0.3892,
+      "step": 853
+    },
+    {
+      "epoch": 0.9622535211267605,
+      "grad_norm": 0.4296998369649018,
+      "learning_rate": 3.774018379281537e-05,
+      "loss": 0.3983,
+      "step": 854
+    },
+    {
+      "epoch": 0.9633802816901409,
+      "grad_norm": 0.2973109212493355,
+      "learning_rate": 3.771929824561404e-05,
+      "loss": 0.3917,
+      "step": 855
+    },
+    {
+      "epoch": 0.9645070422535211,
+      "grad_norm": 0.41637877482516356,
+      "learning_rate": 3.76984126984127e-05,
+      "loss": 0.4151,
+      "step": 856
+    },
+    {
+      "epoch": 0.9656338028169014,
+      "grad_norm": 0.4014759551588234,
+      "learning_rate": 3.767752715121136e-05,
+      "loss": 0.4026,
+      "step": 857
+    },
+    {
+      "epoch": 0.9667605633802817,
+      "grad_norm": 0.4896006055867873,
+      "learning_rate": 3.7656641604010026e-05,
+      "loss": 0.4078,
+      "step": 858
+    },
+    {
+      "epoch": 0.967887323943662,
+      "grad_norm": 0.3818461047798166,
+      "learning_rate": 3.763575605680869e-05,
+      "loss": 0.3896,
+      "step": 859
+    },
+    {
+      "epoch": 0.9690140845070423,
+      "grad_norm": 0.4857861818312063,
+      "learning_rate": 3.761487050960735e-05,
+      "loss": 0.4065,
+      "step": 860
+    },
+    {
+      "epoch": 0.9701408450704225,
+      "grad_norm": 0.3852300916675479,
+      "learning_rate": 3.759398496240601e-05,
+      "loss": 0.4376,
+      "step": 861
+    },
+    {
+      "epoch": 0.9712676056338028,
+      "grad_norm": 0.40146277023613025,
+      "learning_rate": 3.757309941520468e-05,
+      "loss": 0.3928,
+      "step": 862
+    },
+    {
+      "epoch": 0.9723943661971831,
+      "grad_norm": 0.3914507746385292,
+      "learning_rate": 3.755221386800335e-05,
+      "loss": 0.4047,
+      "step": 863
+    },
+    {
+      "epoch": 0.9735211267605633,
+      "grad_norm": 0.382869425291017,
+      "learning_rate": 3.753132832080201e-05,
+      "loss": 0.393,
+      "step": 864
+    },
+    {
+      "epoch": 0.9746478873239437,
+      "grad_norm": 0.358535369989854,
+      "learning_rate": 3.751044277360067e-05,
+      "loss": 0.447,
+      "step": 865
+    },
+    {
+      "epoch": 0.9757746478873239,
+      "grad_norm": 0.38266539367069063,
+      "learning_rate": 3.7489557226399333e-05,
+      "loss": 0.4214,
+      "step": 866
+    },
+    {
+      "epoch": 0.9769014084507043,
+      "grad_norm": 0.3417675891619715,
+      "learning_rate": 3.7468671679198e-05,
+      "loss": 0.414,
+      "step": 867
+    },
+    {
+      "epoch": 0.9780281690140845,
+      "grad_norm": 0.40431141770165596,
+      "learning_rate": 3.744778613199666e-05,
+      "loss": 0.3971,
+      "step": 868
+    },
+    {
+      "epoch": 0.9791549295774647,
+      "grad_norm": 0.3604611534772216,
+      "learning_rate": 3.742690058479532e-05,
+      "loss": 0.3979,
+      "step": 869
+    },
+    {
+      "epoch": 0.9802816901408451,
+      "grad_norm": 0.45047472284441686,
+      "learning_rate": 3.740601503759399e-05,
+      "loss": 0.4047,
+      "step": 870
+    },
+    {
+      "epoch": 0.9814084507042253,
+      "grad_norm": 0.4281416434513922,
+      "learning_rate": 3.7385129490392654e-05,
+      "loss": 0.43,
+      "step": 871
+    },
+    {
+      "epoch": 0.9825352112676057,
+      "grad_norm": 0.4839616963151465,
+      "learning_rate": 3.7364243943191314e-05,
+      "loss": 0.4165,
+      "step": 872
+    },
+    {
+      "epoch": 0.9836619718309859,
+      "grad_norm": 0.3517165408652703,
+      "learning_rate": 3.7343358395989974e-05,
+      "loss": 0.4066,
+      "step": 873
+    },
+    {
+      "epoch": 0.9847887323943662,
+      "grad_norm": 0.4233470102370198,
+      "learning_rate": 3.732247284878864e-05,
+      "loss": 0.4214,
+      "step": 874
+    },
+    {
+      "epoch": 0.9859154929577465,
+      "grad_norm": 0.38852278302910803,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.4253,
+      "step": 875
+    },
+    {
+      "epoch": 0.9870422535211267,
+      "grad_norm": 0.373573815310203,
+      "learning_rate": 3.728070175438597e-05,
+      "loss": 0.4159,
+      "step": 876
+    },
+    {
+      "epoch": 0.9881690140845071,
+      "grad_norm": 0.3748762280861789,
+      "learning_rate": 3.725981620718463e-05,
+      "loss": 0.4303,
+      "step": 877
+    },
+    {
+      "epoch": 0.9892957746478873,
+      "grad_norm": 0.4530223205377112,
+      "learning_rate": 3.7238930659983294e-05,
+      "loss": 0.4173,
+      "step": 878
+    },
+    {
+      "epoch": 0.9904225352112676,
+      "grad_norm": 0.32117261713447776,
+      "learning_rate": 3.721804511278196e-05,
+      "loss": 0.4012,
+      "step": 879
+    },
+    {
+      "epoch": 0.9915492957746479,
+      "grad_norm": 0.47606239155488134,
+      "learning_rate": 3.719715956558062e-05,
+      "loss": 0.3991,
+      "step": 880
+    },
+    {
+      "epoch": 0.9926760563380281,
+      "grad_norm": 0.4081652325703798,
+      "learning_rate": 3.717627401837928e-05,
+      "loss": 0.4031,
+      "step": 881
+    },
+    {
+      "epoch": 0.9938028169014085,
+      "grad_norm": 0.3323321538622951,
+      "learning_rate": 3.715538847117795e-05,
+      "loss": 0.4102,
+      "step": 882
+    },
+    {
+      "epoch": 0.9949295774647887,
+      "grad_norm": 0.45502275344767745,
+      "learning_rate": 3.713450292397661e-05,
+      "loss": 0.3854,
+      "step": 883
+    },
+    {
+      "epoch": 0.9960563380281691,
+      "grad_norm": 0.3425952963375361,
+      "learning_rate": 3.7113617376775274e-05,
+      "loss": 0.4216,
+      "step": 884
+    },
+    {
+      "epoch": 0.9971830985915493,
+      "grad_norm": 0.33252625920951534,
+      "learning_rate": 3.7092731829573934e-05,
+      "loss": 0.4233,
+      "step": 885
+    },
+    {
+      "epoch": 0.9983098591549295,
+      "grad_norm": 0.3060932623255807,
+      "learning_rate": 3.70718462823726e-05,
+      "loss": 0.4147,
+      "step": 886
+    },
+    {
+      "epoch": 0.9994366197183099,
+      "grad_norm": 0.34051331038310045,
+      "learning_rate": 3.705096073517126e-05,
+      "loss": 0.402,
+      "step": 887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.34051331038310045,
+      "learning_rate": 3.703007518796993e-05,
+      "loss": 0.4178,
+      "step": 888
+    },
+    {
+      "epoch": 1.0011267605633802,
+      "grad_norm": 0.45164933939808577,
+      "learning_rate": 3.700918964076859e-05,
+      "loss": 0.3601,
+      "step": 889
+    },
+    {
+      "epoch": 1.0022535211267605,
+      "grad_norm": 0.3324150998169332,
+      "learning_rate": 3.6988304093567254e-05,
+      "loss": 0.3333,
+      "step": 890
+    },
+    {
+      "epoch": 1.003380281690141,
+      "grad_norm": 0.37670491841487863,
+      "learning_rate": 3.6967418546365914e-05,
+      "loss": 0.3264,
+      "step": 891
+    },
+    {
+      "epoch": 1.0045070422535212,
+      "grad_norm": 0.35763553244923446,
+      "learning_rate": 3.694653299916458e-05,
+      "loss": 0.341,
+      "step": 892
+    },
+    {
+      "epoch": 1.0056338028169014,
+      "grad_norm": 0.3519446554726191,
+      "learning_rate": 3.692564745196324e-05,
+      "loss": 0.3544,
+      "step": 893
+    },
+    {
+      "epoch": 1.0067605633802816,
+      "grad_norm": 0.3247266894304795,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 0.3539,
+      "step": 894
+    },
+    {
+      "epoch": 1.0078873239436619,
+      "grad_norm": 0.39108519251934243,
+      "learning_rate": 3.688387635756057e-05,
+      "loss": 0.3298,
+      "step": 895
+    },
+    {
+      "epoch": 1.0090140845070423,
+      "grad_norm": 0.3865567028361942,
+      "learning_rate": 3.6862990810359235e-05,
+      "loss": 0.3355,
+      "step": 896
+    },
+    {
+      "epoch": 1.0101408450704226,
+      "grad_norm": 0.3681661446703723,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 0.3686,
+      "step": 897
+    },
+    {
+      "epoch": 1.0112676056338028,
+      "grad_norm": 0.416773726387342,
+      "learning_rate": 3.682121971595656e-05,
+      "loss": 0.3456,
+      "step": 898
+    },
+    {
+      "epoch": 1.012394366197183,
+      "grad_norm": 0.38573812250069184,
+      "learning_rate": 3.680033416875522e-05,
+      "loss": 0.3264,
+      "step": 899
+    },
+    {
+      "epoch": 1.0135211267605633,
+      "grad_norm": 0.32759729545292754,
+      "learning_rate": 3.677944862155389e-05,
+      "loss": 0.3504,
+      "step": 900
+    },
+    {
+      "epoch": 1.0146478873239437,
+      "grad_norm": 0.4785684697978304,
+      "learning_rate": 3.675856307435255e-05,
+      "loss": 0.3545,
+      "step": 901
+    },
+    {
+      "epoch": 1.015774647887324,
+      "grad_norm": 0.4082888514252489,
+      "learning_rate": 3.6737677527151215e-05,
+      "loss": 0.3428,
+      "step": 902
+    },
+    {
+      "epoch": 1.0169014084507042,
+      "grad_norm": 0.44634616820631956,
+      "learning_rate": 3.6716791979949875e-05,
+      "loss": 0.337,
+      "step": 903
+    },
+    {
+      "epoch": 1.0180281690140844,
+      "grad_norm": 0.3283715545248377,
+      "learning_rate": 3.669590643274854e-05,
+      "loss": 0.3533,
+      "step": 904
+    },
+    {
+      "epoch": 1.019154929577465,
+      "grad_norm": 0.46773983776072087,
+      "learning_rate": 3.66750208855472e-05,
+      "loss": 0.3576,
+      "step": 905
+    },
+    {
+      "epoch": 1.0202816901408451,
+      "grad_norm": 0.32934659204004435,
+      "learning_rate": 3.665413533834587e-05,
+      "loss": 0.3521,
+      "step": 906
+    },
+    {
+      "epoch": 1.0214084507042254,
+      "grad_norm": 0.4097291624527684,
+      "learning_rate": 3.663324979114453e-05,
+      "loss": 0.339,
+      "step": 907
+    },
+    {
+      "epoch": 1.0225352112676056,
+      "grad_norm": 0.36518890704682266,
+      "learning_rate": 3.6612364243943195e-05,
+      "loss": 0.3214,
+      "step": 908
+    },
+    {
+      "epoch": 1.0236619718309858,
+      "grad_norm": 0.37007347409989166,
+      "learning_rate": 3.6591478696741855e-05,
+      "loss": 0.3511,
+      "step": 909
+    },
+    {
+      "epoch": 1.0247887323943663,
+      "grad_norm": 0.39681642474464907,
+      "learning_rate": 3.657059314954052e-05,
+      "loss": 0.3384,
+      "step": 910
+    },
+    {
+      "epoch": 1.0259154929577465,
+      "grad_norm": 0.40990269096333465,
+      "learning_rate": 3.654970760233918e-05,
+      "loss": 0.3646,
+      "step": 911
+    },
+    {
+      "epoch": 1.0270422535211268,
+      "grad_norm": 0.31420500028287357,
+      "learning_rate": 3.652882205513785e-05,
+      "loss": 0.3197,
+      "step": 912
+    },
+    {
+      "epoch": 1.028169014084507,
+      "grad_norm": 0.4022121307134004,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.3697,
+      "step": 913
+    },
+    {
+      "epoch": 1.0292957746478872,
+      "grad_norm": 0.3436605082848157,
+      "learning_rate": 3.6487050960735175e-05,
+      "loss": 0.3484,
+      "step": 914
+    },
+    {
+      "epoch": 1.0304225352112677,
+      "grad_norm": 0.3464021769519033,
+      "learning_rate": 3.6466165413533835e-05,
+      "loss": 0.3479,
+      "step": 915
+    },
+    {
+      "epoch": 1.031549295774648,
+      "grad_norm": 0.3138471218213921,
+      "learning_rate": 3.64452798663325e-05,
+      "loss": 0.3288,
+      "step": 916
+    },
+    {
+      "epoch": 1.0326760563380282,
+      "grad_norm": 0.34546317527627096,
+      "learning_rate": 3.642439431913116e-05,
+      "loss": 0.3397,
+      "step": 917
+    },
+    {
+      "epoch": 1.0338028169014084,
+      "grad_norm": 0.28791110851255375,
+      "learning_rate": 3.640350877192983e-05,
+      "loss": 0.337,
+      "step": 918
+    },
+    {
+      "epoch": 1.0349295774647886,
+      "grad_norm": 0.35902903221755456,
+      "learning_rate": 3.638262322472849e-05,
+      "loss": 0.3469,
+      "step": 919
+    },
+    {
+      "epoch": 1.036056338028169,
+      "grad_norm": 0.31866269461936547,
+      "learning_rate": 3.636173767752715e-05,
+      "loss": 0.3597,
+      "step": 920
+    },
+    {
+      "epoch": 1.0371830985915493,
+      "grad_norm": 0.29209923379950026,
+      "learning_rate": 3.6340852130325816e-05,
+      "loss": 0.3268,
+      "step": 921
+    },
+    {
+      "epoch": 1.0383098591549296,
+      "grad_norm": 0.311877341299532,
+      "learning_rate": 3.631996658312448e-05,
+      "loss": 0.339,
+      "step": 922
+    },
+    {
+      "epoch": 1.0394366197183098,
+      "grad_norm": 0.3592110411387359,
+      "learning_rate": 3.629908103592314e-05,
+      "loss": 0.3421,
+      "step": 923
+    },
+    {
+      "epoch": 1.04056338028169,
+      "grad_norm": 0.2874188388571024,
+      "learning_rate": 3.62781954887218e-05,
+      "loss": 0.3659,
+      "step": 924
+    },
+    {
+      "epoch": 1.0416901408450705,
+      "grad_norm": 0.3493121303234514,
+      "learning_rate": 3.625730994152047e-05,
+      "loss": 0.3742,
+      "step": 925
+    },
+    {
+      "epoch": 1.0428169014084507,
+      "grad_norm": 0.34023873349599343,
+      "learning_rate": 3.6236424394319136e-05,
+      "loss": 0.3526,
+      "step": 926
+    },
+    {
+      "epoch": 1.043943661971831,
+      "grad_norm": 0.30697286116775674,
+      "learning_rate": 3.6215538847117796e-05,
+      "loss": 0.3617,
+      "step": 927
+    },
+    {
+      "epoch": 1.0450704225352112,
+      "grad_norm": 0.3608790433993316,
+      "learning_rate": 3.6194653299916456e-05,
+      "loss": 0.333,
+      "step": 928
+    },
+    {
+      "epoch": 1.0461971830985914,
+      "grad_norm": 0.38038596969857236,
+      "learning_rate": 3.617376775271512e-05,
+      "loss": 0.337,
+      "step": 929
+    },
+    {
+      "epoch": 1.047323943661972,
+      "grad_norm": 0.31211340626839723,
+      "learning_rate": 3.615288220551379e-05,
+      "loss": 0.352,
+      "step": 930
+    },
+    {
+      "epoch": 1.0484507042253521,
+      "grad_norm": 0.3365065293288928,
+      "learning_rate": 3.613199665831245e-05,
+      "loss": 0.3533,
+      "step": 931
+    },
+    {
+      "epoch": 1.0495774647887324,
+      "grad_norm": 0.3197898561220735,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.3401,
+      "step": 932
+    },
+    {
+      "epoch": 1.0507042253521126,
+      "grad_norm": 0.30282863716780495,
+      "learning_rate": 3.6090225563909776e-05,
+      "loss": 0.3721,
+      "step": 933
+    },
+    {
+      "epoch": 1.051830985915493,
+      "grad_norm": 0.3817497816730595,
+      "learning_rate": 3.606934001670844e-05,
+      "loss": 0.3355,
+      "step": 934
+    },
+    {
+      "epoch": 1.0529577464788733,
+      "grad_norm": 0.2633891541035518,
+      "learning_rate": 3.60484544695071e-05,
+      "loss": 0.3253,
+      "step": 935
+    },
+    {
+      "epoch": 1.0540845070422535,
+      "grad_norm": 0.3546137378241751,
+      "learning_rate": 3.602756892230576e-05,
+      "loss": 0.3592,
+      "step": 936
+    },
+    {
+      "epoch": 1.0552112676056338,
+      "grad_norm": 0.3192527668437388,
+      "learning_rate": 3.600668337510443e-05,
+      "loss": 0.3628,
+      "step": 937
+    },
+    {
+      "epoch": 1.056338028169014,
+      "grad_norm": 0.2941649387397572,
+      "learning_rate": 3.5985797827903096e-05,
+      "loss": 0.3479,
+      "step": 938
+    },
+    {
+      "epoch": 1.0574647887323945,
+      "grad_norm": 0.31290100122573195,
+      "learning_rate": 3.5964912280701756e-05,
+      "loss": 0.3252,
+      "step": 939
+    },
+    {
+      "epoch": 1.0585915492957747,
+      "grad_norm": 0.3320108281653657,
+      "learning_rate": 3.5944026733500416e-05,
+      "loss": 0.3572,
+      "step": 940
+    },
+    {
+      "epoch": 1.059718309859155,
+      "grad_norm": 0.27602331372293537,
+      "learning_rate": 3.592314118629908e-05,
+      "loss": 0.3621,
+      "step": 941
+    },
+    {
+      "epoch": 1.0608450704225352,
+      "grad_norm": 0.36796858266701027,
+      "learning_rate": 3.590225563909775e-05,
+      "loss": 0.3487,
+      "step": 942
+    },
+    {
+      "epoch": 1.0619718309859154,
+      "grad_norm": 0.36233861800684203,
+      "learning_rate": 3.588137009189641e-05,
+      "loss": 0.3419,
+      "step": 943
+    },
+    {
+      "epoch": 1.0630985915492959,
+      "grad_norm": 0.32452120940977225,
+      "learning_rate": 3.586048454469507e-05,
+      "loss": 0.3267,
+      "step": 944
+    },
+    {
+      "epoch": 1.064225352112676,
+      "grad_norm": 0.39936202937932747,
+      "learning_rate": 3.583959899749374e-05,
+      "loss": 0.3688,
+      "step": 945
+    },
+    {
+      "epoch": 1.0653521126760563,
+      "grad_norm": 0.32428851970148864,
+      "learning_rate": 3.5818713450292403e-05,
+      "loss": 0.356,
+      "step": 946
+    },
+    {
+      "epoch": 1.0664788732394366,
+      "grad_norm": 0.38698310508892947,
+      "learning_rate": 3.5797827903091063e-05,
+      "loss": 0.3671,
+      "step": 947
+    },
+    {
+      "epoch": 1.0676056338028168,
+      "grad_norm": 0.3994091204255308,
+      "learning_rate": 3.5776942355889723e-05,
+      "loss": 0.3308,
+      "step": 948
+    },
+    {
+      "epoch": 1.0687323943661973,
+      "grad_norm": 0.3570173882044036,
+      "learning_rate": 3.575605680868839e-05,
+      "loss": 0.3278,
+      "step": 949
+    },
+    {
+      "epoch": 1.0698591549295775,
+      "grad_norm": 0.3984032734080762,
+      "learning_rate": 3.573517126148706e-05,
+      "loss": 0.3364,
+      "step": 950
+    },
+    {
+      "epoch": 1.0709859154929577,
+      "grad_norm": 0.41969619013730747,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.352,
+      "step": 951
+    },
+    {
+      "epoch": 1.072112676056338,
+      "grad_norm": 0.3103306898949888,
+      "learning_rate": 3.569340016708438e-05,
+      "loss": 0.3655,
+      "step": 952
+    },
+    {
+      "epoch": 1.0732394366197182,
+      "grad_norm": 0.4202159085269893,
+      "learning_rate": 3.5672514619883044e-05,
+      "loss": 0.3619,
+      "step": 953
+    },
+    {
+      "epoch": 1.0743661971830987,
+      "grad_norm": 0.3390271655464648,
+      "learning_rate": 3.565162907268171e-05,
+      "loss": 0.3374,
+      "step": 954
+    },
+    {
+      "epoch": 1.075492957746479,
+      "grad_norm": 0.3979837332740325,
+      "learning_rate": 3.563074352548037e-05,
+      "loss": 0.3422,
+      "step": 955
+    },
+    {
+      "epoch": 1.0766197183098591,
+      "grad_norm": 0.34921767225311867,
+      "learning_rate": 3.560985797827903e-05,
+      "loss": 0.3291,
+      "step": 956
+    },
+    {
+      "epoch": 1.0777464788732394,
+      "grad_norm": 0.35651371731998577,
+      "learning_rate": 3.55889724310777e-05,
+      "loss": 0.3386,
+      "step": 957
+    },
+    {
+      "epoch": 1.0788732394366196,
+      "grad_norm": 0.288890974136275,
+      "learning_rate": 3.5568086883876364e-05,
+      "loss": 0.3284,
+      "step": 958
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3038814223753594,
+      "learning_rate": 3.5547201336675024e-05,
+      "loss": 0.3581,
+      "step": 959
+    },
+    {
+      "epoch": 1.0811267605633803,
+      "grad_norm": 0.3764557737078917,
+      "learning_rate": 3.5526315789473684e-05,
+      "loss": 0.3523,
+      "step": 960
+    },
+    {
+      "epoch": 1.0822535211267605,
+      "grad_norm": 0.32601737650410584,
+      "learning_rate": 3.5505430242272344e-05,
+      "loss": 0.3638,
+      "step": 961
+    },
+    {
+      "epoch": 1.0833802816901408,
+      "grad_norm": 0.40463015180513584,
+      "learning_rate": 3.548454469507102e-05,
+      "loss": 0.3682,
+      "step": 962
+    },
+    {
+      "epoch": 1.084507042253521,
+      "grad_norm": 0.27853954158074656,
+      "learning_rate": 3.546365914786968e-05,
+      "loss": 0.3341,
+      "step": 963
+    },
+    {
+      "epoch": 1.0856338028169015,
+      "grad_norm": 0.4034672865399672,
+      "learning_rate": 3.544277360066834e-05,
+      "loss": 0.3625,
+      "step": 964
+    },
+    {
+      "epoch": 1.0867605633802817,
+      "grad_norm": 0.2958763421888676,
+      "learning_rate": 3.5421888053467e-05,
+      "loss": 0.3645,
+      "step": 965
+    },
+    {
+      "epoch": 1.087887323943662,
+      "grad_norm": 0.3589284532436751,
+      "learning_rate": 3.540100250626567e-05,
+      "loss": 0.3535,
+      "step": 966
+    },
+    {
+      "epoch": 1.0890140845070422,
+      "grad_norm": 0.38048224390427376,
+      "learning_rate": 3.538011695906433e-05,
+      "loss": 0.3306,
+      "step": 967
+    },
+    {
+      "epoch": 1.0901408450704226,
+      "grad_norm": 0.34312664242838625,
+      "learning_rate": 3.535923141186299e-05,
+      "loss": 0.3486,
+      "step": 968
+    },
+    {
+      "epoch": 1.0912676056338029,
+      "grad_norm": 0.371785849240145,
+      "learning_rate": 3.533834586466165e-05,
+      "loss": 0.3468,
+      "step": 969
+    },
+    {
+      "epoch": 1.092394366197183,
+      "grad_norm": 0.40756311653422916,
+      "learning_rate": 3.5317460317460324e-05,
+      "loss": 0.3481,
+      "step": 970
+    },
+    {
+      "epoch": 1.0935211267605633,
+      "grad_norm": 0.3644784761338962,
+      "learning_rate": 3.5296574770258984e-05,
+      "loss": 0.3556,
+      "step": 971
+    },
+    {
+      "epoch": 1.0946478873239436,
+      "grad_norm": 0.36577915822034535,
+      "learning_rate": 3.5275689223057644e-05,
+      "loss": 0.3482,
+      "step": 972
+    },
+    {
+      "epoch": 1.095774647887324,
+      "grad_norm": 0.35163451346027164,
+      "learning_rate": 3.5254803675856304e-05,
+      "loss": 0.3384,
+      "step": 973
+    },
+    {
+      "epoch": 1.0969014084507043,
+      "grad_norm": 0.33307728406450554,
+      "learning_rate": 3.523391812865498e-05,
+      "loss": 0.3325,
+      "step": 974
+    },
+    {
+      "epoch": 1.0980281690140845,
+      "grad_norm": 0.3623939639719022,
+      "learning_rate": 3.521303258145364e-05,
+      "loss": 0.3346,
+      "step": 975
+    },
+    {
+      "epoch": 1.0991549295774647,
+      "grad_norm": 0.3236939339415892,
+      "learning_rate": 3.51921470342523e-05,
+      "loss": 0.335,
+      "step": 976
+    },
+    {
+      "epoch": 1.100281690140845,
+      "grad_norm": 0.30628231696632063,
+      "learning_rate": 3.517126148705096e-05,
+      "loss": 0.3704,
+      "step": 977
+    },
+    {
+      "epoch": 1.1014084507042254,
+      "grad_norm": 0.3438025646505993,
+      "learning_rate": 3.5150375939849625e-05,
+      "loss": 0.3474,
+      "step": 978
+    },
+    {
+      "epoch": 1.1025352112676057,
+      "grad_norm": 0.31351680352585487,
+      "learning_rate": 3.512949039264829e-05,
+      "loss": 0.3275,
+      "step": 979
+    },
+    {
+      "epoch": 1.103661971830986,
+      "grad_norm": 0.3152564819386366,
+      "learning_rate": 3.510860484544695e-05,
+      "loss": 0.3448,
+      "step": 980
+    },
+    {
+      "epoch": 1.1047887323943661,
+      "grad_norm": 0.30276251407844496,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.3397,
+      "step": 981
+    },
+    {
+      "epoch": 1.1059154929577464,
+      "grad_norm": 0.33042618325454126,
+      "learning_rate": 3.506683375104428e-05,
+      "loss": 0.3323,
+      "step": 982
+    },
+    {
+      "epoch": 1.1070422535211268,
+      "grad_norm": 0.32059827191068474,
+      "learning_rate": 3.5045948203842945e-05,
+      "loss": 0.3745,
+      "step": 983
+    },
+    {
+      "epoch": 1.108169014084507,
+      "grad_norm": 0.3571444063980653,
+      "learning_rate": 3.5025062656641605e-05,
+      "loss": 0.3448,
+      "step": 984
+    },
+    {
+      "epoch": 1.1092957746478873,
+      "grad_norm": 0.30726538847486445,
+      "learning_rate": 3.5004177109440265e-05,
+      "loss": 0.3346,
+      "step": 985
+    },
+    {
+      "epoch": 1.1104225352112675,
+      "grad_norm": 0.3626977765380658,
+      "learning_rate": 3.498329156223893e-05,
+      "loss": 0.3387,
+      "step": 986
+    },
+    {
+      "epoch": 1.111549295774648,
+      "grad_norm": 0.3393776360689493,
+      "learning_rate": 3.49624060150376e-05,
+      "loss": 0.3276,
+      "step": 987
+    },
+    {
+      "epoch": 1.1126760563380282,
+      "grad_norm": 0.2927132796531137,
+      "learning_rate": 3.494152046783626e-05,
+      "loss": 0.3182,
+      "step": 988
+    },
+    {
+      "epoch": 1.1138028169014085,
+      "grad_norm": 0.362124017997493,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.3504,
+      "step": 989
+    },
+    {
+      "epoch": 1.1149295774647887,
+      "grad_norm": 0.31528677471051464,
+      "learning_rate": 3.4899749373433585e-05,
+      "loss": 0.3269,
+      "step": 990
+    },
+    {
+      "epoch": 1.116056338028169,
+      "grad_norm": 0.3118000729629574,
+      "learning_rate": 3.487886382623225e-05,
+      "loss": 0.3547,
+      "step": 991
+    },
+    {
+      "epoch": 1.1171830985915494,
+      "grad_norm": 0.3687927756216515,
+      "learning_rate": 3.485797827903091e-05,
+      "loss": 0.3452,
+      "step": 992
+    },
+    {
+      "epoch": 1.1183098591549296,
+      "grad_norm": 0.31161259789155016,
+      "learning_rate": 3.483709273182957e-05,
+      "loss": 0.3521,
+      "step": 993
+    },
+    {
+      "epoch": 1.1194366197183099,
+      "grad_norm": 0.3234835784243064,
+      "learning_rate": 3.481620718462824e-05,
+      "loss": 0.3464,
+      "step": 994
+    },
+    {
+      "epoch": 1.1205633802816901,
+      "grad_norm": 0.3088864882657699,
+      "learning_rate": 3.4795321637426905e-05,
+      "loss": 0.3542,
+      "step": 995
+    },
+    {
+      "epoch": 1.1216901408450703,
+      "grad_norm": 0.25277381440828456,
+      "learning_rate": 3.4774436090225565e-05,
+      "loss": 0.3619,
+      "step": 996
+    },
+    {
+      "epoch": 1.1228169014084508,
+      "grad_norm": 0.3160344031005969,
+      "learning_rate": 3.4753550543024225e-05,
+      "loss": 0.3474,
+      "step": 997
+    },
+    {
+      "epoch": 1.123943661971831,
+      "grad_norm": 0.301588939897409,
+      "learning_rate": 3.473266499582289e-05,
+      "loss": 0.3503,
+      "step": 998
+    },
+    {
+      "epoch": 1.1250704225352113,
+      "grad_norm": 0.28935381172209435,
+      "learning_rate": 3.471177944862156e-05,
+      "loss": 0.3575,
+      "step": 999
+    },
+    {
+      "epoch": 1.1261971830985915,
+      "grad_norm": 0.31640598782618606,
+      "learning_rate": 3.469089390142022e-05,
+      "loss": 0.3503,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1273239436619718,
+      "grad_norm": 0.3002177232042162,
+      "learning_rate": 3.467000835421888e-05,
+      "loss": 0.3347,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1284507042253522,
+      "grad_norm": 0.3179693477403671,
+      "learning_rate": 3.4649122807017546e-05,
+      "loss": 0.352,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1295774647887324,
+      "grad_norm": 0.5729433235749711,
+      "learning_rate": 3.462823725981621e-05,
+      "loss": 0.344,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1307042253521127,
+      "grad_norm": 0.3239148096356237,
+      "learning_rate": 3.460735171261487e-05,
+      "loss": 0.3406,
+      "step": 1004
+    },
+    {
+      "epoch": 1.131830985915493,
+      "grad_norm": 0.36993062213931016,
+      "learning_rate": 3.458646616541353e-05,
+      "loss": 0.3519,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1329577464788732,
+      "grad_norm": 0.36413925386071816,
+      "learning_rate": 3.45655806182122e-05,
+      "loss": 0.3613,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1340845070422536,
+      "grad_norm": 0.29549142844737325,
+      "learning_rate": 3.4544695071010866e-05,
+      "loss": 0.3378,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1352112676056338,
+      "grad_norm": 0.3460169384070374,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 0.3584,
+      "step": 1008
+    },
+    {
+      "epoch": 1.136338028169014,
+      "grad_norm": 0.34099155583527796,
+      "learning_rate": 3.4502923976608186e-05,
+      "loss": 0.3475,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1374647887323943,
+      "grad_norm": 0.2889062780758684,
+      "learning_rate": 3.448203842940685e-05,
+      "loss": 0.35,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1385915492957746,
+      "grad_norm": 0.3352673980415479,
+      "learning_rate": 3.446115288220552e-05,
+      "loss": 0.3297,
+      "step": 1011
+    },
+    {
+      "epoch": 1.139718309859155,
+      "grad_norm": 0.32690036991244015,
+      "learning_rate": 3.444026733500418e-05,
+      "loss": 0.3415,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1408450704225352,
+      "grad_norm": 0.3570171231337064,
+      "learning_rate": 3.441938178780284e-05,
+      "loss": 0.3472,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1419718309859155,
+      "grad_norm": 0.3001542919735517,
+      "learning_rate": 3.4398496240601506e-05,
+      "loss": 0.3363,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1430985915492957,
+      "grad_norm": 0.332950416261565,
+      "learning_rate": 3.437761069340017e-05,
+      "loss": 0.3411,
+      "step": 1015
+    },
+    {
+      "epoch": 1.144225352112676,
+      "grad_norm": 0.34943372991524113,
+      "learning_rate": 3.435672514619883e-05,
+      "loss": 0.3434,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1453521126760564,
+      "grad_norm": 0.4687340030778913,
+      "learning_rate": 3.433583959899749e-05,
+      "loss": 0.3199,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1464788732394366,
+      "grad_norm": 0.32490609856126923,
+      "learning_rate": 3.431495405179616e-05,
+      "loss": 0.3498,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1476056338028169,
+      "grad_norm": 0.311611088242007,
+      "learning_rate": 3.429406850459482e-05,
+      "loss": 0.3575,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1487323943661971,
+      "grad_norm": 0.2990062423503099,
+      "learning_rate": 3.4273182957393486e-05,
+      "loss": 0.3373,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1498591549295774,
+      "grad_norm": 0.38353184002896906,
+      "learning_rate": 3.4252297410192146e-05,
+      "loss": 0.3483,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1509859154929578,
+      "grad_norm": 0.34451561538155556,
+      "learning_rate": 3.423141186299081e-05,
+      "loss": 0.3719,
+      "step": 1022
+    },
+    {
+      "epoch": 1.152112676056338,
+      "grad_norm": 0.3358121138275392,
+      "learning_rate": 3.421052631578947e-05,
+      "loss": 0.3362,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1532394366197183,
+      "grad_norm": 0.34065054887892876,
+      "learning_rate": 3.418964076858814e-05,
+      "loss": 0.3565,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1543661971830985,
+      "grad_norm": 0.27997379236251685,
+      "learning_rate": 3.41687552213868e-05,
+      "loss": 0.3403,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1554929577464788,
+      "grad_norm": 0.30141207345682136,
+      "learning_rate": 3.414786967418547e-05,
+      "loss": 0.3471,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1566197183098592,
+      "grad_norm": 0.337034300756015,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.3615,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1577464788732394,
+      "grad_norm": 0.30674246250268733,
+      "learning_rate": 3.4106098579782793e-05,
+      "loss": 0.3462,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1588732394366197,
+      "grad_norm": 0.3173684428076217,
+      "learning_rate": 3.4085213032581453e-05,
+      "loss": 0.3516,
+      "step": 1029
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.37769826101860116,
+      "learning_rate": 3.406432748538012e-05,
+      "loss": 0.3524,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1611267605633804,
+      "grad_norm": 0.28073869279700486,
+      "learning_rate": 3.404344193817878e-05,
+      "loss": 0.3531,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1622535211267606,
+      "grad_norm": 0.30120610684344723,
+      "learning_rate": 3.402255639097745e-05,
+      "loss": 0.3446,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1633802816901408,
+      "grad_norm": 0.35522297671980546,
+      "learning_rate": 3.400167084377611e-05,
+      "loss": 0.3458,
+      "step": 1033
+    },
+    {
+      "epoch": 1.164507042253521,
+      "grad_norm": 0.2838997521983254,
+      "learning_rate": 3.3980785296574774e-05,
+      "loss": 0.371,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1656338028169013,
+      "grad_norm": 0.3611036407312474,
+      "learning_rate": 3.3959899749373434e-05,
+      "loss": 0.334,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1667605633802818,
+      "grad_norm": 0.2950730591608434,
+      "learning_rate": 3.39390142021721e-05,
+      "loss": 0.3434,
+      "step": 1036
+    },
+    {
+      "epoch": 1.167887323943662,
+      "grad_norm": 0.28270863416618186,
+      "learning_rate": 3.391812865497076e-05,
+      "loss": 0.35,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1690140845070423,
+      "grad_norm": 0.3661793145277532,
+      "learning_rate": 3.389724310776943e-05,
+      "loss": 0.3614,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1701408450704225,
+      "grad_norm": 0.31362783654640425,
+      "learning_rate": 3.387635756056809e-05,
+      "loss": 0.3499,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1712676056338027,
+      "grad_norm": 0.2971686710050231,
+      "learning_rate": 3.3855472013366754e-05,
+      "loss": 0.3569,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1723943661971832,
+      "grad_norm": 0.352731153270978,
+      "learning_rate": 3.3834586466165414e-05,
+      "loss": 0.3524,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1735211267605634,
+      "grad_norm": 0.2936538359382206,
+      "learning_rate": 3.381370091896408e-05,
+      "loss": 0.3376,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1746478873239437,
+      "grad_norm": 0.3330092908628216,
+      "learning_rate": 3.379281537176274e-05,
+      "loss": 0.3464,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1757746478873239,
+      "grad_norm": 0.30245046256286734,
+      "learning_rate": 3.377192982456141e-05,
+      "loss": 0.3455,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1769014084507043,
+      "grad_norm": 0.3760828172052888,
+      "learning_rate": 3.375104427736007e-05,
+      "loss": 0.3364,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1780281690140846,
+      "grad_norm": 0.30918561156039337,
+      "learning_rate": 3.3730158730158734e-05,
+      "loss": 0.339,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1791549295774648,
+      "grad_norm": 0.334820663107568,
+      "learning_rate": 3.3709273182957394e-05,
+      "loss": 0.3452,
+      "step": 1047
+    },
+    {
+      "epoch": 1.180281690140845,
+      "grad_norm": 0.3235665593969472,
+      "learning_rate": 3.368838763575606e-05,
+      "loss": 0.353,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1814084507042253,
+      "grad_norm": 0.3683834175550552,
+      "learning_rate": 3.366750208855472e-05,
+      "loss": 0.353,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1825352112676057,
+      "grad_norm": 0.30202094720462735,
+      "learning_rate": 3.364661654135339e-05,
+      "loss": 0.3319,
+      "step": 1050
+    },
+    {
+      "epoch": 1.183661971830986,
+      "grad_norm": 0.3240517238787127,
+      "learning_rate": 3.362573099415205e-05,
+      "loss": 0.3328,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1847887323943662,
+      "grad_norm": 0.35263531537635145,
+      "learning_rate": 3.3604845446950714e-05,
+      "loss": 0.3475,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1859154929577465,
+      "grad_norm": 0.30604588229024526,
+      "learning_rate": 3.3583959899749374e-05,
+      "loss": 0.334,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1870422535211267,
+      "grad_norm": 0.3464586434956338,
+      "learning_rate": 3.356307435254804e-05,
+      "loss": 0.3412,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1881690140845071,
+      "grad_norm": 0.32135691756390916,
+      "learning_rate": 3.35421888053467e-05,
+      "loss": 0.361,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1892957746478874,
+      "grad_norm": 0.28987551299344433,
+      "learning_rate": 3.352130325814536e-05,
+      "loss": 0.3351,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1904225352112676,
+      "grad_norm": 0.3919076009259972,
+      "learning_rate": 3.350041771094403e-05,
+      "loss": 0.3589,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1915492957746479,
+      "grad_norm": 0.275569997096992,
+      "learning_rate": 3.3479532163742695e-05,
+      "loss": 0.3489,
+      "step": 1058
+    },
+    {
+      "epoch": 1.192676056338028,
+      "grad_norm": 0.3382553487679515,
+      "learning_rate": 3.3458646616541355e-05,
+      "loss": 0.3549,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1938028169014085,
+      "grad_norm": 0.33618071363176094,
+      "learning_rate": 3.3437761069340015e-05,
+      "loss": 0.3319,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1949295774647888,
+      "grad_norm": 0.3035047482300514,
+      "learning_rate": 3.341687552213868e-05,
+      "loss": 0.3385,
+      "step": 1061
+    },
+    {
+      "epoch": 1.196056338028169,
+      "grad_norm": 0.27465118295033303,
+      "learning_rate": 3.339598997493735e-05,
+      "loss": 0.3356,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1971830985915493,
+      "grad_norm": 0.3427740400251671,
+      "learning_rate": 3.337510442773601e-05,
+      "loss": 0.3359,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1983098591549295,
+      "grad_norm": 0.3228672170869719,
+      "learning_rate": 3.335421888053467e-05,
+      "loss": 0.344,
+      "step": 1064
+    },
+    {
+      "epoch": 1.19943661971831,
+      "grad_norm": 0.3162251658775707,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3395,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2005633802816902,
+      "grad_norm": 0.36784522815336534,
+      "learning_rate": 3.3312447786132e-05,
+      "loss": 0.3539,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2016901408450704,
+      "grad_norm": 0.3498666413543401,
+      "learning_rate": 3.329156223893066e-05,
+      "loss": 0.3411,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2028169014084507,
+      "grad_norm": 0.40603387109599365,
+      "learning_rate": 3.327067669172932e-05,
+      "loss": 0.339,
+      "step": 1068
+    },
+    {
+      "epoch": 1.203943661971831,
+      "grad_norm": 0.28466237503238295,
+      "learning_rate": 3.324979114452799e-05,
+      "loss": 0.3374,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2050704225352113,
+      "grad_norm": 0.43643291002776546,
+      "learning_rate": 3.3228905597326655e-05,
+      "loss": 0.3421,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2061971830985916,
+      "grad_norm": 0.3209330822309557,
+      "learning_rate": 3.3208020050125315e-05,
+      "loss": 0.3263,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2073239436619718,
+      "grad_norm": 0.3170979690800217,
+      "learning_rate": 3.3187134502923975e-05,
+      "loss": 0.349,
+      "step": 1072
+    },
+    {
+      "epoch": 1.208450704225352,
+      "grad_norm": 0.29620537314071277,
+      "learning_rate": 3.316624895572264e-05,
+      "loss": 0.3415,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2095774647887323,
+      "grad_norm": 0.2814235048086694,
+      "learning_rate": 3.314536340852131e-05,
+      "loss": 0.3533,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2107042253521128,
+      "grad_norm": 0.30429985408344573,
+      "learning_rate": 3.312447786131997e-05,
+      "loss": 0.3403,
+      "step": 1075
+    },
+    {
+      "epoch": 1.211830985915493,
+      "grad_norm": 0.2776747142048821,
+      "learning_rate": 3.310359231411863e-05,
+      "loss": 0.3489,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2129577464788732,
+      "grad_norm": 0.3117922804791841,
+      "learning_rate": 3.3082706766917295e-05,
+      "loss": 0.3589,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2140845070422535,
+      "grad_norm": 0.28206225975246074,
+      "learning_rate": 3.306182121971596e-05,
+      "loss": 0.3219,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2152112676056337,
+      "grad_norm": 0.3339551879612691,
+      "learning_rate": 3.304093567251462e-05,
+      "loss": 0.3487,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2163380281690142,
+      "grad_norm": 0.2470821047506818,
+      "learning_rate": 3.302005012531328e-05,
+      "loss": 0.3293,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2174647887323944,
+      "grad_norm": 0.31331363910902593,
+      "learning_rate": 3.299916457811195e-05,
+      "loss": 0.3384,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2185915492957746,
+      "grad_norm": 0.3173999509165457,
+      "learning_rate": 3.2978279030910616e-05,
+      "loss": 0.3519,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2197183098591549,
+      "grad_norm": 0.3472442907380024,
+      "learning_rate": 3.2957393483709276e-05,
+      "loss": 0.3453,
+      "step": 1083
+    },
+    {
+      "epoch": 1.220845070422535,
+      "grad_norm": 0.31667306436662185,
+      "learning_rate": 3.2936507936507936e-05,
+      "loss": 0.3632,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2219718309859156,
+      "grad_norm": 0.3633369367271794,
+      "learning_rate": 3.29156223893066e-05,
+      "loss": 0.3346,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2230985915492958,
+      "grad_norm": 0.2735176852696013,
+      "learning_rate": 3.289473684210527e-05,
+      "loss": 0.3561,
+      "step": 1086
+    },
+    {
+      "epoch": 1.224225352112676,
+      "grad_norm": 0.3880089164415539,
+      "learning_rate": 3.287385129490393e-05,
+      "loss": 0.3209,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2253521126760563,
+      "grad_norm": 0.26906175661116216,
+      "learning_rate": 3.285296574770259e-05,
+      "loss": 0.3465,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2264788732394365,
+      "grad_norm": 0.3284498960738571,
+      "learning_rate": 3.2832080200501256e-05,
+      "loss": 0.3457,
+      "step": 1089
+    },
+    {
+      "epoch": 1.227605633802817,
+      "grad_norm": 0.3667097517230621,
+      "learning_rate": 3.281119465329992e-05,
+      "loss": 0.3605,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2287323943661972,
+      "grad_norm": 0.28752992668073607,
+      "learning_rate": 3.279030910609858e-05,
+      "loss": 0.3635,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2298591549295774,
+      "grad_norm": 0.3946314540545924,
+      "learning_rate": 3.276942355889724e-05,
+      "loss": 0.3493,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2309859154929577,
+      "grad_norm": 0.2789906483434097,
+      "learning_rate": 3.274853801169591e-05,
+      "loss": 0.3347,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2321126760563381,
+      "grad_norm": 0.3225093109594014,
+      "learning_rate": 3.2727652464494576e-05,
+      "loss": 0.341,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2332394366197184,
+      "grad_norm": 0.32862283959211835,
+      "learning_rate": 3.2706766917293236e-05,
+      "loss": 0.3448,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2343661971830986,
+      "grad_norm": 0.30766363411442,
+      "learning_rate": 3.2685881370091896e-05,
+      "loss": 0.345,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2354929577464788,
+      "grad_norm": 0.33495055865479145,
+      "learning_rate": 3.2664995822890556e-05,
+      "loss": 0.3482,
+      "step": 1097
+    },
+    {
+      "epoch": 1.236619718309859,
+      "grad_norm": 0.4100596803077402,
+      "learning_rate": 3.264411027568923e-05,
+      "loss": 0.3776,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2377464788732395,
+      "grad_norm": 0.29849674067446036,
+      "learning_rate": 3.262322472848789e-05,
+      "loss": 0.3478,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2388732394366198,
+      "grad_norm": 0.4752815995738534,
+      "learning_rate": 3.260233918128655e-05,
+      "loss": 0.3413,
+      "step": 1100
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2841451269993109,
+      "learning_rate": 3.258145363408521e-05,
+      "loss": 0.3311,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2411267605633802,
+      "grad_norm": 0.43287277582944,
+      "learning_rate": 3.256056808688388e-05,
+      "loss": 0.3404,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2422535211267607,
+      "grad_norm": 0.27842035168195645,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.3364,
+      "step": 1103
+    },
+    {
+      "epoch": 1.243380281690141,
+      "grad_norm": 0.3580502578234584,
+      "learning_rate": 3.25187969924812e-05,
+      "loss": 0.3304,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2445070422535212,
+      "grad_norm": 0.33254398767744164,
+      "learning_rate": 3.249791144527986e-05,
+      "loss": 0.3371,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2456338028169014,
+      "grad_norm": 0.32142884321075016,
+      "learning_rate": 3.247702589807854e-05,
+      "loss": 0.3604,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2467605633802816,
+      "grad_norm": 0.39048564971156885,
+      "learning_rate": 3.24561403508772e-05,
+      "loss": 0.3665,
+      "step": 1107
+    },
+    {
+      "epoch": 1.247887323943662,
+      "grad_norm": 0.47432868715894116,
+      "learning_rate": 3.243525480367586e-05,
+      "loss": 0.3379,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2490140845070423,
+      "grad_norm": 0.5373426510405537,
+      "learning_rate": 3.241436925647452e-05,
+      "loss": 0.352,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2501408450704226,
+      "grad_norm": 0.28347871588612267,
+      "learning_rate": 3.239348370927319e-05,
+      "loss": 0.3209,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2512676056338028,
+      "grad_norm": 0.39484792243555067,
+      "learning_rate": 3.237259816207185e-05,
+      "loss": 0.3388,
+      "step": 1111
+    },
+    {
+      "epoch": 1.252394366197183,
+      "grad_norm": 0.3107741346812873,
+      "learning_rate": 3.235171261487051e-05,
+      "loss": 0.3366,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2535211267605635,
+      "grad_norm": 0.3898314974601013,
+      "learning_rate": 3.233082706766917e-05,
+      "loss": 0.3604,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2546478873239437,
+      "grad_norm": 0.3408581558334333,
+      "learning_rate": 3.230994152046784e-05,
+      "loss": 0.3492,
+      "step": 1114
+    },
+    {
+      "epoch": 1.255774647887324,
+      "grad_norm": 0.3618348938509109,
+      "learning_rate": 3.2289055973266504e-05,
+      "loss": 0.3617,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2569014084507042,
+      "grad_norm": 0.2817036619413573,
+      "learning_rate": 3.2268170426065164e-05,
+      "loss": 0.3517,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2580281690140844,
+      "grad_norm": 0.3070164964104573,
+      "learning_rate": 3.2247284878863824e-05,
+      "loss": 0.3683,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2591549295774649,
+      "grad_norm": 0.31552633114716866,
+      "learning_rate": 3.222639933166249e-05,
+      "loss": 0.3657,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2602816901408451,
+      "grad_norm": 0.3175500946772457,
+      "learning_rate": 3.220551378446116e-05,
+      "loss": 0.3625,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2614084507042254,
+      "grad_norm": 0.28442849784940555,
+      "learning_rate": 3.218462823725982e-05,
+      "loss": 0.3461,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2625352112676056,
+      "grad_norm": 0.3041046845176297,
+      "learning_rate": 3.216374269005848e-05,
+      "loss": 0.3325,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2636619718309858,
+      "grad_norm": 0.3065101535698693,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 0.3328,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2647887323943663,
+      "grad_norm": 0.32118291745639427,
+      "learning_rate": 3.212197159565581e-05,
+      "loss": 0.3516,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2659154929577465,
+      "grad_norm": 0.2706365873502233,
+      "learning_rate": 3.210108604845447e-05,
+      "loss": 0.3402,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2670422535211268,
+      "grad_norm": 0.3395123585654144,
+      "learning_rate": 3.208020050125313e-05,
+      "loss": 0.3488,
+      "step": 1125
+    },
+    {
+      "epoch": 1.268169014084507,
+      "grad_norm": 0.3192115703388621,
+      "learning_rate": 3.20593149540518e-05,
+      "loss": 0.3321,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2692957746478872,
+      "grad_norm": 0.2748915319828329,
+      "learning_rate": 3.2038429406850464e-05,
+      "loss": 0.36,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2704225352112677,
+      "grad_norm": 0.33790604735267227,
+      "learning_rate": 3.2017543859649124e-05,
+      "loss": 0.359,
+      "step": 1128
+    },
+    {
+      "epoch": 1.271549295774648,
+      "grad_norm": 0.28636007427713034,
+      "learning_rate": 3.1996658312447784e-05,
+      "loss": 0.3697,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2726760563380282,
+      "grad_norm": 0.3147026101699195,
+      "learning_rate": 3.197577276524645e-05,
+      "loss": 0.3566,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2738028169014084,
+      "grad_norm": 0.2939860946022757,
+      "learning_rate": 3.195488721804512e-05,
+      "loss": 0.3308,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2749295774647886,
+      "grad_norm": 0.29802348283544033,
+      "learning_rate": 3.193400167084378e-05,
+      "loss": 0.3563,
+      "step": 1132
+    },
+    {
+      "epoch": 1.276056338028169,
+      "grad_norm": 0.29506623728971354,
+      "learning_rate": 3.191311612364244e-05,
+      "loss": 0.3337,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2771830985915493,
+      "grad_norm": 0.2822855549021315,
+      "learning_rate": 3.1892230576441104e-05,
+      "loss": 0.3489,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2783098591549296,
+      "grad_norm": 0.28426720383589205,
+      "learning_rate": 3.187134502923977e-05,
+      "loss": 0.3338,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2794366197183098,
+      "grad_norm": 0.24335968053060297,
+      "learning_rate": 3.185045948203843e-05,
+      "loss": 0.3586,
+      "step": 1136
+    },
+    {
+      "epoch": 1.28056338028169,
+      "grad_norm": 0.34492248613575655,
+      "learning_rate": 3.182957393483709e-05,
+      "loss": 0.3643,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2816901408450705,
+      "grad_norm": 0.28163190816338013,
+      "learning_rate": 3.180868838763576e-05,
+      "loss": 0.3551,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2828169014084507,
+      "grad_norm": 0.3038306599947534,
+      "learning_rate": 3.1787802840434425e-05,
+      "loss": 0.3261,
+      "step": 1139
+    },
+    {
+      "epoch": 1.283943661971831,
+      "grad_norm": 0.27177700302131724,
+      "learning_rate": 3.1766917293233085e-05,
+      "loss": 0.3351,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2850704225352112,
+      "grad_norm": 0.26269138228106,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.3549,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2861971830985914,
+      "grad_norm": 0.2913648915223674,
+      "learning_rate": 3.172514619883041e-05,
+      "loss": 0.3436,
+      "step": 1142
+    },
+    {
+      "epoch": 1.287323943661972,
+      "grad_norm": 0.27276203661405635,
+      "learning_rate": 3.170426065162908e-05,
+      "loss": 0.3368,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2884507042253521,
+      "grad_norm": 0.356736914318009,
+      "learning_rate": 3.168337510442774e-05,
+      "loss": 0.3465,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2895774647887324,
+      "grad_norm": 0.28475878519907033,
+      "learning_rate": 3.16624895572264e-05,
+      "loss": 0.3569,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2907042253521126,
+      "grad_norm": 0.3514003335437731,
+      "learning_rate": 3.1641604010025065e-05,
+      "loss": 0.33,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2918309859154928,
+      "grad_norm": 0.3458744624264682,
+      "learning_rate": 3.162071846282373e-05,
+      "loss": 0.3775,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2929577464788733,
+      "grad_norm": 0.31156956928774027,
+      "learning_rate": 3.159983291562239e-05,
+      "loss": 0.3355,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2940845070422535,
+      "grad_norm": 0.2790592740208634,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 0.3183,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2952112676056338,
+      "grad_norm": 0.31894750285146517,
+      "learning_rate": 3.155806182121972e-05,
+      "loss": 0.3424,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2963380281690142,
+      "grad_norm": 0.2653718691621682,
+      "learning_rate": 3.1537176274018385e-05,
+      "loss": 0.3448,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2974647887323942,
+      "grad_norm": 0.30262000464687844,
+      "learning_rate": 3.1516290726817045e-05,
+      "loss": 0.3321,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2985915492957747,
+      "grad_norm": 0.3287674540594305,
+      "learning_rate": 3.1495405179615705e-05,
+      "loss": 0.3368,
+      "step": 1153
+    },
+    {
+      "epoch": 1.299718309859155,
+      "grad_norm": 0.2944767444308635,
+      "learning_rate": 3.147451963241437e-05,
+      "loss": 0.3563,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3008450704225352,
+      "grad_norm": 0.30510447934543955,
+      "learning_rate": 3.145363408521303e-05,
+      "loss": 0.3464,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3019718309859156,
+      "grad_norm": 0.3375919433957859,
+      "learning_rate": 3.14327485380117e-05,
+      "loss": 0.3423,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3030985915492956,
+      "grad_norm": 0.35168934931086815,
+      "learning_rate": 3.141186299081036e-05,
+      "loss": 0.3545,
+      "step": 1157
+    },
+    {
+      "epoch": 1.304225352112676,
+      "grad_norm": 0.3677607944955644,
+      "learning_rate": 3.1390977443609025e-05,
+      "loss": 0.3654,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3053521126760563,
+      "grad_norm": 0.35260259876267547,
+      "learning_rate": 3.1370091896407685e-05,
+      "loss": 0.3486,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3064788732394366,
+      "grad_norm": 0.35691937485605885,
+      "learning_rate": 3.134920634920635e-05,
+      "loss": 0.3265,
+      "step": 1160
+    },
+    {
+      "epoch": 1.307605633802817,
+      "grad_norm": 0.3512055814657964,
+      "learning_rate": 3.132832080200501e-05,
+      "loss": 0.366,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3087323943661973,
+      "grad_norm": 0.42092551860734834,
+      "learning_rate": 3.130743525480368e-05,
+      "loss": 0.3159,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3098591549295775,
+      "grad_norm": 0.3699923682926425,
+      "learning_rate": 3.128654970760234e-05,
+      "loss": 0.3393,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3109859154929577,
+      "grad_norm": 0.3399828357531795,
+      "learning_rate": 3.1265664160401006e-05,
+      "loss": 0.3457,
+      "step": 1164
+    },
+    {
+      "epoch": 1.312112676056338,
+      "grad_norm": 0.4585972251711265,
+      "learning_rate": 3.1244778613199666e-05,
+      "loss": 0.3458,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3132394366197184,
+      "grad_norm": 0.34287449191225977,
+      "learning_rate": 3.122389306599833e-05,
+      "loss": 0.3667,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3143661971830987,
+      "grad_norm": 0.4194731523616112,
+      "learning_rate": 3.120300751879699e-05,
+      "loss": 0.3601,
+      "step": 1167
+    },
+    {
+      "epoch": 1.315492957746479,
+      "grad_norm": 0.3294653640015108,
+      "learning_rate": 3.118212197159566e-05,
+      "loss": 0.3568,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3166197183098591,
+      "grad_norm": 0.47350063514750734,
+      "learning_rate": 3.116123642439432e-05,
+      "loss": 0.3477,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3177464788732394,
+      "grad_norm": 0.280534203327362,
+      "learning_rate": 3.1140350877192986e-05,
+      "loss": 0.3583,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3188732394366198,
+      "grad_norm": 0.49437365334095373,
+      "learning_rate": 3.1119465329991646e-05,
+      "loss": 0.3516,
+      "step": 1171
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3016444543070627,
+      "learning_rate": 3.109857978279031e-05,
+      "loss": 0.3479,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3211267605633803,
+      "grad_norm": 0.36752735629356575,
+      "learning_rate": 3.107769423558897e-05,
+      "loss": 0.3776,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3222535211267605,
+      "grad_norm": 0.2998746868582936,
+      "learning_rate": 3.105680868838764e-05,
+      "loss": 0.3319,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3233802816901408,
+      "grad_norm": 0.4052736495459436,
+      "learning_rate": 3.10359231411863e-05,
+      "loss": 0.3318,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3245070422535212,
+      "grad_norm": 0.29620311313799436,
+      "learning_rate": 3.1015037593984966e-05,
+      "loss": 0.3327,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3256338028169015,
+      "grad_norm": 0.35940054984699693,
+      "learning_rate": 3.0994152046783626e-05,
+      "loss": 0.3601,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3267605633802817,
+      "grad_norm": 0.2998670016480465,
+      "learning_rate": 3.097326649958229e-05,
+      "loss": 0.3358,
+      "step": 1178
+    },
+    {
+      "epoch": 1.327887323943662,
+      "grad_norm": 0.2613240139906125,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.3528,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3290140845070422,
+      "grad_norm": 0.312707479713965,
+      "learning_rate": 3.093149540517962e-05,
+      "loss": 0.3432,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3301408450704226,
+      "grad_norm": 0.2952059364485859,
+      "learning_rate": 3.091060985797828e-05,
+      "loss": 0.3591,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3312676056338029,
+      "grad_norm": 0.30308204907758685,
+      "learning_rate": 3.0889724310776946e-05,
+      "loss": 0.3544,
+      "step": 1182
+    },
+    {
+      "epoch": 1.332394366197183,
+      "grad_norm": 0.28585642966780267,
+      "learning_rate": 3.0868838763575606e-05,
+      "loss": 0.3189,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3335211267605633,
+      "grad_norm": 0.2726099821036104,
+      "learning_rate": 3.084795321637427e-05,
+      "loss": 0.3381,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3346478873239436,
+      "grad_norm": 0.32729439406058286,
+      "learning_rate": 3.082706766917293e-05,
+      "loss": 0.337,
+      "step": 1185
+    },
+    {
+      "epoch": 1.335774647887324,
+      "grad_norm": 0.30702486116129707,
+      "learning_rate": 3.08061821219716e-05,
+      "loss": 0.3469,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3369014084507043,
+      "grad_norm": 0.2919163761992772,
+      "learning_rate": 3.078529657477026e-05,
+      "loss": 0.3614,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3380281690140845,
+      "grad_norm": 0.3362453020110343,
+      "learning_rate": 3.076441102756893e-05,
+      "loss": 0.3506,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3391549295774647,
+      "grad_norm": 0.3726011350376432,
+      "learning_rate": 3.074352548036759e-05,
+      "loss": 0.3604,
+      "step": 1189
+    },
+    {
+      "epoch": 1.340281690140845,
+      "grad_norm": 0.4541280728678603,
+      "learning_rate": 3.0722639933166253e-05,
+      "loss": 0.3764,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3414084507042254,
+      "grad_norm": 1.3936880238257683,
+      "learning_rate": 3.0701754385964913e-05,
+      "loss": 0.3615,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3425352112676057,
+      "grad_norm": 0.4258102093027606,
+      "learning_rate": 3.068086883876357e-05,
+      "loss": 0.3543,
+      "step": 1192
+    },
+    {
+      "epoch": 1.343661971830986,
+      "grad_norm": 0.3575350282116202,
+      "learning_rate": 3.065998329156224e-05,
+      "loss": 0.3367,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3447887323943661,
+      "grad_norm": 0.32348813463671583,
+      "learning_rate": 3.063909774436091e-05,
+      "loss": 0.3254,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3459154929577464,
+      "grad_norm": 0.2620930397025272,
+      "learning_rate": 3.061821219715957e-05,
+      "loss": 0.3286,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3470422535211268,
+      "grad_norm": 0.4043629956242075,
+      "learning_rate": 3.059732664995823e-05,
+      "loss": 0.35,
+      "step": 1196
+    },
+    {
+      "epoch": 1.348169014084507,
+      "grad_norm": 0.3120790464669233,
+      "learning_rate": 3.0576441102756894e-05,
+      "loss": 0.3529,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3492957746478873,
+      "grad_norm": 0.3456006649337656,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.3422,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3504225352112675,
+      "grad_norm": 0.3203816325309061,
+      "learning_rate": 3.053467000835422e-05,
+      "loss": 0.3565,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3515492957746478,
+      "grad_norm": 0.3196109116354312,
+      "learning_rate": 3.0513784461152884e-05,
+      "loss": 0.3525,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3526760563380282,
+      "grad_norm": 0.3430290246706473,
+      "learning_rate": 3.0492898913951544e-05,
+      "loss": 0.333,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3538028169014085,
+      "grad_norm": 0.3306918690981812,
+      "learning_rate": 3.0472013366750214e-05,
+      "loss": 0.3389,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3549295774647887,
+      "grad_norm": 0.3451869621368286,
+      "learning_rate": 3.0451127819548874e-05,
+      "loss": 0.3397,
+      "step": 1203
+    },
+    {
+      "epoch": 1.356056338028169,
+      "grad_norm": 0.28806868402300256,
+      "learning_rate": 3.0430242272347537e-05,
+      "loss": 0.3501,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3571830985915492,
+      "grad_norm": 0.30789548763765795,
+      "learning_rate": 3.0409356725146197e-05,
+      "loss": 0.3637,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3583098591549296,
+      "grad_norm": 0.2714479894177003,
+      "learning_rate": 3.0388471177944867e-05,
+      "loss": 0.3519,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3594366197183099,
+      "grad_norm": 0.3590544930015899,
+      "learning_rate": 3.0367585630743527e-05,
+      "loss": 0.3502,
+      "step": 1207
+    },
+    {
+      "epoch": 1.36056338028169,
+      "grad_norm": 0.3214469345004128,
+      "learning_rate": 3.034670008354219e-05,
+      "loss": 0.3236,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3616901408450703,
+      "grad_norm": 0.3042782868325741,
+      "learning_rate": 3.032581453634085e-05,
+      "loss": 0.3452,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3628169014084506,
+      "grad_norm": 0.33936594312067697,
+      "learning_rate": 3.030492898913952e-05,
+      "loss": 0.344,
+      "step": 1210
+    },
+    {
+      "epoch": 1.363943661971831,
+      "grad_norm": 0.3043390659124525,
+      "learning_rate": 3.028404344193818e-05,
+      "loss": 0.3575,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3650704225352113,
+      "grad_norm": 0.3357631465279863,
+      "learning_rate": 3.0263157894736844e-05,
+      "loss": 0.3334,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3661971830985915,
+      "grad_norm": 0.3177753566079832,
+      "learning_rate": 3.0242272347535504e-05,
+      "loss": 0.3585,
+      "step": 1213
+    },
+    {
+      "epoch": 1.367323943661972,
+      "grad_norm": 0.31936174408730955,
+      "learning_rate": 3.022138680033417e-05,
+      "loss": 0.327,
+      "step": 1214
+    },
+    {
+      "epoch": 1.368450704225352,
+      "grad_norm": 0.315057693985089,
+      "learning_rate": 3.0200501253132834e-05,
+      "loss": 0.3462,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3695774647887324,
+      "grad_norm": 0.2681371385737636,
+      "learning_rate": 3.0179615705931498e-05,
+      "loss": 0.3207,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3707042253521127,
+      "grad_norm": 0.29792906360473065,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 0.3357,
+      "step": 1217
+    },
+    {
+      "epoch": 1.371830985915493,
+      "grad_norm": 0.3207461742321883,
+      "learning_rate": 3.0137844611528825e-05,
+      "loss": 0.3352,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3729577464788734,
+      "grad_norm": 0.2783192420898195,
+      "learning_rate": 3.0116959064327488e-05,
+      "loss": 0.3451,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3740845070422536,
+      "grad_norm": 0.32591368421816236,
+      "learning_rate": 3.009607351712615e-05,
+      "loss": 0.353,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3752112676056338,
+      "grad_norm": 0.3066561562944395,
+      "learning_rate": 3.007518796992481e-05,
+      "loss": 0.3243,
+      "step": 1221
+    },
+    {
+      "epoch": 1.376338028169014,
+      "grad_norm": 2.2660560288471254,
+      "learning_rate": 3.0054302422723478e-05,
+      "loss": 0.4137,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3774647887323943,
+      "grad_norm": 0.39210870348356686,
+      "learning_rate": 3.003341687552214e-05,
+      "loss": 0.348,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3785915492957748,
+      "grad_norm": 0.2962128553173713,
+      "learning_rate": 3.0012531328320805e-05,
+      "loss": 0.3435,
+      "step": 1224
+    },
+    {
+      "epoch": 1.379718309859155,
+      "grad_norm": 0.33275477838598055,
+      "learning_rate": 2.9991645781119465e-05,
+      "loss": 0.3515,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3808450704225352,
+      "grad_norm": 0.3352961753145922,
+      "learning_rate": 2.997076023391813e-05,
+      "loss": 0.3636,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3819718309859155,
+      "grad_norm": 0.35096015184771046,
+      "learning_rate": 2.9949874686716795e-05,
+      "loss": 0.3648,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3830985915492957,
+      "grad_norm": 0.3168355004880717,
+      "learning_rate": 2.9928989139515455e-05,
+      "loss": 0.3643,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3842253521126762,
+      "grad_norm": 0.32634015086711105,
+      "learning_rate": 2.9908103592314118e-05,
+      "loss": 0.3492,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3853521126760564,
+      "grad_norm": 0.33810275384931515,
+      "learning_rate": 2.9887218045112785e-05,
+      "loss": 0.3492,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3864788732394366,
+      "grad_norm": 0.33464097541186544,
+      "learning_rate": 2.986633249791145e-05,
+      "loss": 0.3519,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3876056338028169,
+      "grad_norm": 0.31725298321868456,
+      "learning_rate": 2.984544695071011e-05,
+      "loss": 0.3472,
+      "step": 1232
+    },
+    {
+      "epoch": 1.388732394366197,
+      "grad_norm": 0.30368942357679957,
+      "learning_rate": 2.9824561403508772e-05,
+      "loss": 0.3572,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3898591549295776,
+      "grad_norm": 0.3137955769099765,
+      "learning_rate": 2.980367585630744e-05,
+      "loss": 0.366,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3909859154929578,
+      "grad_norm": 0.32494170437175546,
+      "learning_rate": 2.9782790309106102e-05,
+      "loss": 0.3451,
+      "step": 1235
+    },
+    {
+      "epoch": 1.392112676056338,
+      "grad_norm": 0.289152507208471,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 0.3453,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3932394366197183,
+      "grad_norm": 0.30367923467257946,
+      "learning_rate": 2.9741019214703425e-05,
+      "loss": 0.3474,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3943661971830985,
+      "grad_norm": 0.27456188122911507,
+      "learning_rate": 2.9720133667502092e-05,
+      "loss": 0.3472,
+      "step": 1238
+    },
+    {
+      "epoch": 1.395492957746479,
+      "grad_norm": 0.3338851968314126,
+      "learning_rate": 2.9699248120300755e-05,
+      "loss": 0.3536,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3966197183098592,
+      "grad_norm": 0.2909887050565231,
+      "learning_rate": 2.9678362573099415e-05,
+      "loss": 0.3427,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3977464788732394,
+      "grad_norm": 0.2808404219130732,
+      "learning_rate": 2.965747702589808e-05,
+      "loss": 0.3349,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3988732394366197,
+      "grad_norm": 0.2974931938145654,
+      "learning_rate": 2.9636591478696746e-05,
+      "loss": 0.3258,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.30446773483609063,
+      "learning_rate": 2.961570593149541e-05,
+      "loss": 0.3451,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4011267605633804,
+      "grad_norm": 0.2629360082929095,
+      "learning_rate": 2.959482038429407e-05,
+      "loss": 0.3621,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4022535211267606,
+      "grad_norm": 0.33634058178632664,
+      "learning_rate": 2.9573934837092732e-05,
+      "loss": 0.3265,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4033802816901408,
+      "grad_norm": 0.25671340810425175,
+      "learning_rate": 2.95530492898914e-05,
+      "loss": 0.3596,
+      "step": 1246
+    },
+    {
+      "epoch": 1.404507042253521,
+      "grad_norm": 0.30371346506641733,
+      "learning_rate": 2.9532163742690062e-05,
+      "loss": 0.3373,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4056338028169013,
+      "grad_norm": 0.33309340268956333,
+      "learning_rate": 2.9511278195488722e-05,
+      "loss": 0.3524,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4067605633802818,
+      "grad_norm": 0.35312534002615037,
+      "learning_rate": 2.9490392648287386e-05,
+      "loss": 0.3554,
+      "step": 1249
+    },
+    {
+      "epoch": 1.407887323943662,
+      "grad_norm": 0.268431248391172,
+      "learning_rate": 2.9469507101086053e-05,
+      "loss": 0.3428,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4090140845070422,
+      "grad_norm": 0.29984418244715005,
+      "learning_rate": 2.9448621553884716e-05,
+      "loss": 0.3434,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4101408450704225,
+      "grad_norm": 0.3255596377490089,
+      "learning_rate": 2.9427736006683376e-05,
+      "loss": 0.3322,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4112676056338027,
+      "grad_norm": 0.29044414091181564,
+      "learning_rate": 2.940685045948204e-05,
+      "loss": 0.3617,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4123943661971832,
+      "grad_norm": 0.3137082885510621,
+      "learning_rate": 2.9385964912280706e-05,
+      "loss": 0.3359,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4135211267605634,
+      "grad_norm": 0.3818804237490533,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.353,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4146478873239436,
+      "grad_norm": 0.30290564364536143,
+      "learning_rate": 2.934419381787803e-05,
+      "loss": 0.3236,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4157746478873239,
+      "grad_norm": 0.32553490755642506,
+      "learning_rate": 2.9323308270676693e-05,
+      "loss": 0.3352,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4169014084507041,
+      "grad_norm": 0.3678190946467779,
+      "learning_rate": 2.930242272347536e-05,
+      "loss": 0.3299,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4180281690140846,
+      "grad_norm": 0.33853643840663594,
+      "learning_rate": 2.928153717627402e-05,
+      "loss": 0.3609,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4191549295774648,
+      "grad_norm": 0.3450814972809518,
+      "learning_rate": 2.9260651629072683e-05,
+      "loss": 0.3628,
+      "step": 1260
+    },
+    {
+      "epoch": 1.420281690140845,
+      "grad_norm": 0.4342951041395845,
+      "learning_rate": 2.9239766081871346e-05,
+      "loss": 0.3575,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4214084507042253,
+      "grad_norm": 0.28784037040527977,
+      "learning_rate": 2.9218880534670013e-05,
+      "loss": 0.3454,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4225352112676055,
+      "grad_norm": 0.3702939745854903,
+      "learning_rate": 2.9197994987468673e-05,
+      "loss": 0.3366,
+      "step": 1263
+    },
+    {
+      "epoch": 1.423661971830986,
+      "grad_norm": 0.36297638741464655,
+      "learning_rate": 2.9177109440267336e-05,
+      "loss": 0.3436,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4247887323943662,
+      "grad_norm": 0.2932884280851587,
+      "learning_rate": 2.9156223893066e-05,
+      "loss": 0.3375,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4259154929577464,
+      "grad_norm": 0.40800534520034465,
+      "learning_rate": 2.9135338345864667e-05,
+      "loss": 0.3667,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4270422535211267,
+      "grad_norm": 0.378230060871396,
+      "learning_rate": 2.9114452798663327e-05,
+      "loss": 0.3279,
+      "step": 1267
+    },
+    {
+      "epoch": 1.428169014084507,
+      "grad_norm": 0.34509394026907525,
+      "learning_rate": 2.909356725146199e-05,
+      "loss": 0.3475,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4292957746478874,
+      "grad_norm": 0.4153322561933915,
+      "learning_rate": 2.907268170426065e-05,
+      "loss": 0.3438,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4304225352112676,
+      "grad_norm": 0.27853930771661345,
+      "learning_rate": 2.905179615705932e-05,
+      "loss": 0.3363,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4315492957746478,
+      "grad_norm": 0.32099780030598213,
+      "learning_rate": 2.903091060985798e-05,
+      "loss": 0.3376,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4326760563380283,
+      "grad_norm": 0.37729012390993694,
+      "learning_rate": 2.9010025062656643e-05,
+      "loss": 0.3626,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4338028169014083,
+      "grad_norm": 0.3412079603451582,
+      "learning_rate": 2.8989139515455303e-05,
+      "loss": 0.347,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4349295774647888,
+      "grad_norm": 0.3165021784985063,
+      "learning_rate": 2.8968253968253974e-05,
+      "loss": 0.3464,
+      "step": 1274
+    },
+    {
+      "epoch": 1.436056338028169,
+      "grad_norm": 0.3591571483688786,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 0.37,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4371830985915492,
+      "grad_norm": 0.3562250029512977,
+      "learning_rate": 2.8926482873851297e-05,
+      "loss": 0.333,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4383098591549297,
+      "grad_norm": 0.25149984510123385,
+      "learning_rate": 2.8905597326649957e-05,
+      "loss": 0.3468,
+      "step": 1277
+    },
+    {
+      "epoch": 1.43943661971831,
+      "grad_norm": 0.30250034342485427,
+      "learning_rate": 2.8884711779448627e-05,
+      "loss": 0.357,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4405633802816902,
+      "grad_norm": 0.34875139847130576,
+      "learning_rate": 2.8863826232247287e-05,
+      "loss": 0.3625,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4416901408450704,
+      "grad_norm": 0.30334645075958366,
+      "learning_rate": 2.884294068504595e-05,
+      "loss": 0.3253,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4428169014084506,
+      "grad_norm": 0.31229266086668034,
+      "learning_rate": 2.882205513784461e-05,
+      "loss": 0.3534,
+      "step": 1281
+    },
+    {
+      "epoch": 1.443943661971831,
+      "grad_norm": 0.33041544401134926,
+      "learning_rate": 2.8801169590643277e-05,
+      "loss": 0.3352,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4450704225352113,
+      "grad_norm": 0.34674132045981165,
+      "learning_rate": 2.878028404344194e-05,
+      "loss": 0.3311,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4461971830985916,
+      "grad_norm": 0.26156829898934003,
+      "learning_rate": 2.8759398496240604e-05,
+      "loss": 0.3432,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4473239436619718,
+      "grad_norm": 0.3136080304572592,
+      "learning_rate": 2.8738512949039264e-05,
+      "loss": 0.3569,
+      "step": 1285
+    },
+    {
+      "epoch": 1.448450704225352,
+      "grad_norm": 0.30710210539173416,
+      "learning_rate": 2.871762740183793e-05,
+      "loss": 0.3405,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4495774647887325,
+      "grad_norm": 0.2894300410033628,
+      "learning_rate": 2.8696741854636594e-05,
+      "loss": 0.3423,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4507042253521127,
+      "grad_norm": 0.29605019545996425,
+      "learning_rate": 2.8675856307435257e-05,
+      "loss": 0.3548,
+      "step": 1288
+    },
+    {
+      "epoch": 1.451830985915493,
+      "grad_norm": 0.33202235824154624,
+      "learning_rate": 2.8654970760233917e-05,
+      "loss": 0.3642,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4529577464788732,
+      "grad_norm": 0.28873570052106284,
+      "learning_rate": 2.8634085213032584e-05,
+      "loss": 0.3352,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4540845070422534,
+      "grad_norm": 0.32778904766572686,
+      "learning_rate": 2.8613199665831247e-05,
+      "loss": 0.345,
+      "step": 1291
+    },
+    {
+      "epoch": 1.455211267605634,
+      "grad_norm": 0.30770894411783706,
+      "learning_rate": 2.859231411862991e-05,
+      "loss": 0.3484,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4563380281690141,
+      "grad_norm": 0.2833498740974589,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.352,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4574647887323944,
+      "grad_norm": 0.31814781815089144,
+      "learning_rate": 2.8550543024227238e-05,
+      "loss": 0.3541,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4585915492957746,
+      "grad_norm": 0.2770971173774878,
+      "learning_rate": 2.85296574770259e-05,
+      "loss": 0.3657,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4597183098591549,
+      "grad_norm": 0.36066551582561196,
+      "learning_rate": 2.850877192982456e-05,
+      "loss": 0.3313,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4608450704225353,
+      "grad_norm": 0.30952046262144167,
+      "learning_rate": 2.8487886382623224e-05,
+      "loss": 0.3514,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4619718309859155,
+      "grad_norm": 0.2748856964533165,
+      "learning_rate": 2.846700083542189e-05,
+      "loss": 0.3573,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4630985915492958,
+      "grad_norm": 0.35452544836499567,
+      "learning_rate": 2.8446115288220554e-05,
+      "loss": 0.3482,
+      "step": 1299
+    },
+    {
+      "epoch": 1.464225352112676,
+      "grad_norm": 0.2768770488408414,
+      "learning_rate": 2.8425229741019214e-05,
+      "loss": 0.3645,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4653521126760563,
+      "grad_norm": 0.2827464532913941,
+      "learning_rate": 2.8404344193817878e-05,
+      "loss": 0.3351,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4664788732394367,
+      "grad_norm": 0.37765012973788603,
+      "learning_rate": 2.8383458646616545e-05,
+      "loss": 0.357,
+      "step": 1302
+    },
+    {
+      "epoch": 1.467605633802817,
+      "grad_norm": 0.299044625170323,
+      "learning_rate": 2.8362573099415208e-05,
+      "loss": 0.3456,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4687323943661972,
+      "grad_norm": 0.2753644903042144,
+      "learning_rate": 2.8341687552213868e-05,
+      "loss": 0.3496,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4698591549295774,
+      "grad_norm": 0.3444703076328087,
+      "learning_rate": 2.832080200501253e-05,
+      "loss": 0.3258,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4709859154929577,
+      "grad_norm": 0.2481634802989212,
+      "learning_rate": 2.8299916457811198e-05,
+      "loss": 0.3355,
+      "step": 1306
+    },
+    {
+      "epoch": 1.472112676056338,
+      "grad_norm": 0.319902114656331,
+      "learning_rate": 2.827903091060986e-05,
+      "loss": 0.323,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4732394366197183,
+      "grad_norm": 0.2994178943222728,
+      "learning_rate": 2.825814536340852e-05,
+      "loss": 0.3484,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4743661971830986,
+      "grad_norm": 0.30917647854179314,
+      "learning_rate": 2.8237259816207185e-05,
+      "loss": 0.345,
+      "step": 1309
+    },
+    {
+      "epoch": 1.4754929577464788,
+      "grad_norm": 0.2531647281154903,
+      "learning_rate": 2.821637426900585e-05,
+      "loss": 0.3287,
+      "step": 1310
+    },
+    {
+      "epoch": 1.476619718309859,
+      "grad_norm": 0.2866526733970423,
+      "learning_rate": 2.8195488721804515e-05,
+      "loss": 0.3373,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4777464788732395,
+      "grad_norm": 0.3075467523111137,
+      "learning_rate": 2.8174603174603175e-05,
+      "loss": 0.3357,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4788732394366197,
+      "grad_norm": 0.272832856249605,
+      "learning_rate": 2.815371762740184e-05,
+      "loss": 0.3433,
+      "step": 1313
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2729760741101419,
+      "learning_rate": 2.8132832080200505e-05,
+      "loss": 0.3382,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4811267605633802,
+      "grad_norm": 0.28571235772706455,
+      "learning_rate": 2.811194653299917e-05,
+      "loss": 0.3425,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4822535211267605,
+      "grad_norm": 0.2736751524988955,
+      "learning_rate": 2.809106098579783e-05,
+      "loss": 0.3207,
+      "step": 1316
+    },
+    {
+      "epoch": 1.483380281690141,
+      "grad_norm": 0.33812848916965643,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 0.3468,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4845070422535211,
+      "grad_norm": 0.2873317112935958,
+      "learning_rate": 2.804928989139516e-05,
+      "loss": 0.3486,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4856338028169014,
+      "grad_norm": 0.27183076296760217,
+      "learning_rate": 2.8028404344193822e-05,
+      "loss": 0.3472,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4867605633802816,
+      "grad_norm": 0.3445807209520114,
+      "learning_rate": 2.8007518796992482e-05,
+      "loss": 0.3438,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4878873239436619,
+      "grad_norm": 0.3127876717078221,
+      "learning_rate": 2.7986633249791145e-05,
+      "loss": 0.3343,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4890140845070423,
+      "grad_norm": 0.2911551101410927,
+      "learning_rate": 2.7965747702589812e-05,
+      "loss": 0.3386,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4901408450704225,
+      "grad_norm": 0.3141573720998868,
+      "learning_rate": 2.7944862155388472e-05,
+      "loss": 0.3448,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4912676056338028,
+      "grad_norm": 0.28291618873548724,
+      "learning_rate": 2.7923976608187135e-05,
+      "loss": 0.3426,
+      "step": 1324
+    },
+    {
+      "epoch": 1.492394366197183,
+      "grad_norm": 0.2692167845293607,
+      "learning_rate": 2.79030910609858e-05,
+      "loss": 0.3371,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4935211267605633,
+      "grad_norm": 0.27872207571615465,
+      "learning_rate": 2.7882205513784466e-05,
+      "loss": 0.3259,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4946478873239437,
+      "grad_norm": 0.31338604494476663,
+      "learning_rate": 2.7861319966583126e-05,
+      "loss": 0.3457,
+      "step": 1327
+    },
+    {
+      "epoch": 1.495774647887324,
+      "grad_norm": 0.2928463795031251,
+      "learning_rate": 2.784043441938179e-05,
+      "loss": 0.3391,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4969014084507042,
+      "grad_norm": 0.34764955906498535,
+      "learning_rate": 2.7819548872180452e-05,
+      "loss": 0.3351,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4980281690140846,
+      "grad_norm": 0.3023292227508466,
+      "learning_rate": 2.779866332497912e-05,
+      "loss": 0.3275,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4991549295774647,
+      "grad_norm": 0.3091796807737989,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3579,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5002816901408451,
+      "grad_norm": 0.3095335257232928,
+      "learning_rate": 2.7756892230576442e-05,
+      "loss": 0.3384,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5014084507042254,
+      "grad_norm": 0.299616166937427,
+      "learning_rate": 2.7736006683375106e-05,
+      "loss": 0.3334,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5025352112676056,
+      "grad_norm": 0.2891915110352774,
+      "learning_rate": 2.7715121136173773e-05,
+      "loss": 0.3356,
+      "step": 1334
+    },
+    {
+      "epoch": 1.503661971830986,
+      "grad_norm": 0.3323177952735513,
+      "learning_rate": 2.7694235588972433e-05,
+      "loss": 0.3428,
+      "step": 1335
+    },
+    {
+      "epoch": 1.504788732394366,
+      "grad_norm": 0.35973144909936317,
+      "learning_rate": 2.7673350041771096e-05,
+      "loss": 0.3653,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5059154929577465,
+      "grad_norm": 0.28987961239262905,
+      "learning_rate": 2.7652464494569756e-05,
+      "loss": 0.3412,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5070422535211268,
+      "grad_norm": 0.3386510034404898,
+      "learning_rate": 2.7631578947368426e-05,
+      "loss": 0.3136,
+      "step": 1338
+    },
+    {
+      "epoch": 1.508169014084507,
+      "grad_norm": 0.31260198405738565,
+      "learning_rate": 2.7610693400167086e-05,
+      "loss": 0.3317,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5092957746478874,
+      "grad_norm": 0.27033647337294403,
+      "learning_rate": 2.758980785296575e-05,
+      "loss": 0.3232,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5104225352112675,
+      "grad_norm": 0.2918364615372896,
+      "learning_rate": 2.756892230576441e-05,
+      "loss": 0.337,
+      "step": 1341
+    },
+    {
+      "epoch": 1.511549295774648,
+      "grad_norm": 0.32001229005415976,
+      "learning_rate": 2.754803675856308e-05,
+      "loss": 0.345,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5126760563380282,
+      "grad_norm": 0.27427834567161835,
+      "learning_rate": 2.752715121136174e-05,
+      "loss": 0.3633,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5138028169014084,
+      "grad_norm": 0.2489172670773644,
+      "learning_rate": 2.7506265664160403e-05,
+      "loss": 0.3437,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5149295774647888,
+      "grad_norm": 0.2909539905928386,
+      "learning_rate": 2.7485380116959063e-05,
+      "loss": 0.3281,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5160563380281689,
+      "grad_norm": 0.28314692739296915,
+      "learning_rate": 2.7464494569757733e-05,
+      "loss": 0.3218,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5171830985915493,
+      "grad_norm": 0.2874575696178154,
+      "learning_rate": 2.7443609022556393e-05,
+      "loss": 0.362,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5183098591549296,
+      "grad_norm": 0.2974835890176962,
+      "learning_rate": 2.7422723475355056e-05,
+      "loss": 0.3545,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5194366197183098,
+      "grad_norm": 0.2835406784466393,
+      "learning_rate": 2.7401837928153716e-05,
+      "loss": 0.3653,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5205633802816902,
+      "grad_norm": 0.2844679548913541,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 0.3365,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5216901408450703,
+      "grad_norm": 0.3180498946601223,
+      "learning_rate": 2.7360066833751047e-05,
+      "loss": 0.3312,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5228169014084507,
+      "grad_norm": 0.3023117732604468,
+      "learning_rate": 2.733918128654971e-05,
+      "loss": 0.3696,
+      "step": 1352
+    },
+    {
+      "epoch": 1.523943661971831,
+      "grad_norm": 0.33024560678427756,
+      "learning_rate": 2.731829573934837e-05,
+      "loss": 0.3535,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5250704225352112,
+      "grad_norm": 0.32889562258435656,
+      "learning_rate": 2.7297410192147037e-05,
+      "loss": 0.3464,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5261971830985916,
+      "grad_norm": 0.3164497735318053,
+      "learning_rate": 2.72765246449457e-05,
+      "loss": 0.3438,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5273239436619719,
+      "grad_norm": 0.31319837575238246,
+      "learning_rate": 2.7255639097744363e-05,
+      "loss": 0.3225,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5284507042253521,
+      "grad_norm": 0.32188854803462924,
+      "learning_rate": 2.7234753550543023e-05,
+      "loss": 0.3483,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5295774647887324,
+      "grad_norm": 0.3193960591385271,
+      "learning_rate": 2.721386800334169e-05,
+      "loss": 0.3493,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5307042253521126,
+      "grad_norm": 0.30668644315358606,
+      "learning_rate": 2.7192982456140354e-05,
+      "loss": 0.3455,
+      "step": 1359
+    },
+    {
+      "epoch": 1.531830985915493,
+      "grad_norm": 0.29429059378149947,
+      "learning_rate": 2.7172096908939017e-05,
+      "loss": 0.3502,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5329577464788733,
+      "grad_norm": 0.34593884556412324,
+      "learning_rate": 2.7151211361737677e-05,
+      "loss": 0.3419,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5340845070422535,
+      "grad_norm": 0.2697471446325152,
+      "learning_rate": 2.7130325814536344e-05,
+      "loss": 0.3289,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5352112676056338,
+      "grad_norm": 0.2897128031295572,
+      "learning_rate": 2.7109440267335007e-05,
+      "loss": 0.3413,
+      "step": 1363
+    },
+    {
+      "epoch": 1.536338028169014,
+      "grad_norm": 0.34365216670244914,
+      "learning_rate": 2.7088554720133667e-05,
+      "loss": 0.3671,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5374647887323944,
+      "grad_norm": 0.26927575252902075,
+      "learning_rate": 2.706766917293233e-05,
+      "loss": 0.3667,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5385915492957747,
+      "grad_norm": 0.3063326672240967,
+      "learning_rate": 2.7046783625730997e-05,
+      "loss": 0.3515,
+      "step": 1366
+    },
+    {
+      "epoch": 1.539718309859155,
+      "grad_norm": 0.3648933491827071,
+      "learning_rate": 2.702589807852966e-05,
+      "loss": 0.3756,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5408450704225352,
+      "grad_norm": 0.300755843638833,
+      "learning_rate": 2.700501253132832e-05,
+      "loss": 0.3547,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5419718309859154,
+      "grad_norm": 0.3250180654384015,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.352,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5430985915492959,
+      "grad_norm": 0.30252245989727555,
+      "learning_rate": 2.696324143692565e-05,
+      "loss": 0.3533,
+      "step": 1370
+    },
+    {
+      "epoch": 1.544225352112676,
+      "grad_norm": 0.2845024385448824,
+      "learning_rate": 2.6942355889724314e-05,
+      "loss": 0.3388,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5453521126760563,
+      "grad_norm": 0.3180568115975252,
+      "learning_rate": 2.6921470342522974e-05,
+      "loss": 0.3546,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5464788732394368,
+      "grad_norm": 0.32527654151497487,
+      "learning_rate": 2.6900584795321637e-05,
+      "loss": 0.348,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5476056338028168,
+      "grad_norm": 0.3522036911658601,
+      "learning_rate": 2.6879699248120304e-05,
+      "loss": 0.3582,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5487323943661973,
+      "grad_norm": 0.35935681942406394,
+      "learning_rate": 2.6858813700918968e-05,
+      "loss": 0.3471,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5498591549295775,
+      "grad_norm": 0.26552218639595676,
+      "learning_rate": 2.6837928153717628e-05,
+      "loss": 0.3448,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5509859154929577,
+      "grad_norm": 0.32808185473297674,
+      "learning_rate": 2.681704260651629e-05,
+      "loss": 0.3368,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5521126760563382,
+      "grad_norm": 0.27214208030441817,
+      "learning_rate": 2.6796157059314958e-05,
+      "loss": 0.3314,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5532394366197182,
+      "grad_norm": 0.26073918239703964,
+      "learning_rate": 2.677527151211362e-05,
+      "loss": 0.3408,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5543661971830987,
+      "grad_norm": 0.3421765223133862,
+      "learning_rate": 2.675438596491228e-05,
+      "loss": 0.3595,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5554929577464789,
+      "grad_norm": 0.2869738964669758,
+      "learning_rate": 2.6733500417710944e-05,
+      "loss": 0.3349,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5566197183098591,
+      "grad_norm": 0.3181298736789464,
+      "learning_rate": 2.671261487050961e-05,
+      "loss": 0.351,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5577464788732396,
+      "grad_norm": 0.2565318489948106,
+      "learning_rate": 2.6691729323308275e-05,
+      "loss": 0.3349,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5588732394366196,
+      "grad_norm": 0.30534670746119047,
+      "learning_rate": 2.6670843776106935e-05,
+      "loss": 0.3517,
+      "step": 1384
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2588139544758108,
+      "learning_rate": 2.6649958228905598e-05,
+      "loss": 0.347,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5611267605633803,
+      "grad_norm": 0.2861119417513974,
+      "learning_rate": 2.6629072681704265e-05,
+      "loss": 0.3302,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5622535211267605,
+      "grad_norm": 0.2661849430364748,
+      "learning_rate": 2.6608187134502928e-05,
+      "loss": 0.3384,
+      "step": 1387
+    },
+    {
+      "epoch": 1.563380281690141,
+      "grad_norm": 0.28779496971243823,
+      "learning_rate": 2.6587301587301588e-05,
+      "loss": 0.3489,
+      "step": 1388
+    },
+    {
+      "epoch": 1.564507042253521,
+      "grad_norm": 0.280641359638324,
+      "learning_rate": 2.656641604010025e-05,
+      "loss": 0.3316,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5656338028169015,
+      "grad_norm": 0.254891395855962,
+      "learning_rate": 2.6545530492898918e-05,
+      "loss": 0.341,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5667605633802817,
+      "grad_norm": 0.2807897886759735,
+      "learning_rate": 2.6524644945697578e-05,
+      "loss": 0.3415,
+      "step": 1391
+    },
+    {
+      "epoch": 1.567887323943662,
+      "grad_norm": 0.2768837322927675,
+      "learning_rate": 2.650375939849624e-05,
+      "loss": 0.3426,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5690140845070424,
+      "grad_norm": 0.3101082631830913,
+      "learning_rate": 2.6482873851294905e-05,
+      "loss": 0.3695,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5701408450704224,
+      "grad_norm": 0.6820410639958185,
+      "learning_rate": 2.6461988304093572e-05,
+      "loss": 0.3432,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5712676056338029,
+      "grad_norm": 0.362099966934023,
+      "learning_rate": 2.6441102756892232e-05,
+      "loss": 0.3352,
+      "step": 1395
+    },
+    {
+      "epoch": 1.572394366197183,
+      "grad_norm": 0.30087023193343065,
+      "learning_rate": 2.6420217209690895e-05,
+      "loss": 0.3482,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5735211267605633,
+      "grad_norm": 0.38399351692851613,
+      "learning_rate": 2.639933166248956e-05,
+      "loss": 0.3354,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5746478873239438,
+      "grad_norm": 0.31915287347120525,
+      "learning_rate": 2.6378446115288225e-05,
+      "loss": 0.3519,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5757746478873238,
+      "grad_norm": 0.35631760611896585,
+      "learning_rate": 2.6357560568086885e-05,
+      "loss": 0.3486,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5769014084507043,
+      "grad_norm": 0.24857670158956344,
+      "learning_rate": 2.633667502088555e-05,
+      "loss": 0.3465,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5780281690140845,
+      "grad_norm": 0.32180056116060374,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.3211,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5791549295774647,
+      "grad_norm": 0.3912795354364383,
+      "learning_rate": 2.629490392648288e-05,
+      "loss": 0.3663,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5802816901408452,
+      "grad_norm": 0.26651798682062733,
+      "learning_rate": 2.627401837928154e-05,
+      "loss": 0.3627,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5814084507042252,
+      "grad_norm": 0.2936149895825024,
+      "learning_rate": 2.6253132832080202e-05,
+      "loss": 0.3507,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5825352112676057,
+      "grad_norm": 0.30696718635511294,
+      "learning_rate": 2.6232247284878862e-05,
+      "loss": 0.38,
+      "step": 1405
+    },
+    {
+      "epoch": 1.583661971830986,
+      "grad_norm": 0.33154457165579876,
+      "learning_rate": 2.6211361737677532e-05,
+      "loss": 0.3513,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5847887323943661,
+      "grad_norm": 0.31443266029042044,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.3487,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5859154929577466,
+      "grad_norm": 0.33892978419544095,
+      "learning_rate": 2.6169590643274856e-05,
+      "loss": 0.3643,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5870422535211266,
+      "grad_norm": 0.3070809683604584,
+      "learning_rate": 2.6148705096073516e-05,
+      "loss": 0.3596,
+      "step": 1409
+    },
+    {
+      "epoch": 1.588169014084507,
+      "grad_norm": 0.2827064976381111,
+      "learning_rate": 2.6127819548872186e-05,
+      "loss": 0.3573,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5892957746478873,
+      "grad_norm": 0.3229878235602582,
+      "learning_rate": 2.6106934001670846e-05,
+      "loss": 0.3219,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5904225352112675,
+      "grad_norm": 0.31093887040644913,
+      "learning_rate": 2.608604845446951e-05,
+      "loss": 0.3593,
+      "step": 1412
+    },
+    {
+      "epoch": 1.591549295774648,
+      "grad_norm": 0.2774968218487699,
+      "learning_rate": 2.606516290726817e-05,
+      "loss": 0.3315,
+      "step": 1413
+    },
+    {
+      "epoch": 1.5926760563380282,
+      "grad_norm": 0.25498385907155885,
+      "learning_rate": 2.604427736006684e-05,
+      "loss": 0.3223,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5938028169014085,
+      "grad_norm": 0.2727488829583004,
+      "learning_rate": 2.60233918128655e-05,
+      "loss": 0.337,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5949295774647887,
+      "grad_norm": 0.33554974496041423,
+      "learning_rate": 2.6002506265664163e-05,
+      "loss": 0.3484,
+      "step": 1416
+    },
+    {
+      "epoch": 1.596056338028169,
+      "grad_norm": 0.27591446012623233,
+      "learning_rate": 2.5981620718462823e-05,
+      "loss": 0.3501,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5971830985915494,
+      "grad_norm": 0.31941752603872553,
+      "learning_rate": 2.596073517126149e-05,
+      "loss": 0.3454,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5983098591549296,
+      "grad_norm": 0.30125833320515955,
+      "learning_rate": 2.5939849624060153e-05,
+      "loss": 0.3327,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5994366197183099,
+      "grad_norm": 0.2582237111436985,
+      "learning_rate": 2.5918964076858816e-05,
+      "loss": 0.3457,
+      "step": 1420
+    },
+    {
+      "epoch": 1.60056338028169,
+      "grad_norm": 0.31373164120604946,
+      "learning_rate": 2.5898078529657476e-05,
+      "loss": 0.3329,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6016901408450703,
+      "grad_norm": 0.32942272873563083,
+      "learning_rate": 2.5877192982456143e-05,
+      "loss": 0.3618,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6028169014084508,
+      "grad_norm": 0.2646664149122406,
+      "learning_rate": 2.5856307435254806e-05,
+      "loss": 0.3336,
+      "step": 1423
+    },
+    {
+      "epoch": 1.603943661971831,
+      "grad_norm": 0.3058840576523787,
+      "learning_rate": 2.583542188805347e-05,
+      "loss": 0.3402,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6050704225352113,
+      "grad_norm": 0.32158351015262177,
+      "learning_rate": 2.581453634085213e-05,
+      "loss": 0.3367,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6061971830985915,
+      "grad_norm": 0.2755437238431191,
+      "learning_rate": 2.5793650793650796e-05,
+      "loss": 0.3207,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6073239436619717,
+      "grad_norm": 0.30625595434371505,
+      "learning_rate": 2.577276524644946e-05,
+      "loss": 0.3697,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6084507042253522,
+      "grad_norm": 0.30284963531984754,
+      "learning_rate": 2.5751879699248123e-05,
+      "loss": 0.3295,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6095774647887324,
+      "grad_norm": 0.26367411671532615,
+      "learning_rate": 2.5730994152046783e-05,
+      "loss": 0.3218,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6107042253521127,
+      "grad_norm": 0.2723447457176679,
+      "learning_rate": 2.571010860484545e-05,
+      "loss": 0.3256,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6118309859154931,
+      "grad_norm": 0.32346483316378777,
+      "learning_rate": 2.5689223057644113e-05,
+      "loss": 0.3464,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6129577464788731,
+      "grad_norm": 0.2925095477776419,
+      "learning_rate": 2.5668337510442773e-05,
+      "loss": 0.3605,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6140845070422536,
+      "grad_norm": 0.30114671963002587,
+      "learning_rate": 2.5647451963241437e-05,
+      "loss": 0.3477,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6152112676056338,
+      "grad_norm": 0.2846035505506225,
+      "learning_rate": 2.5626566416040103e-05,
+      "loss": 0.3428,
+      "step": 1434
+    },
+    {
+      "epoch": 1.616338028169014,
+      "grad_norm": 0.2588210598131198,
+      "learning_rate": 2.5605680868838767e-05,
+      "loss": 0.3451,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6174647887323945,
+      "grad_norm": 0.2625763271243046,
+      "learning_rate": 2.5584795321637427e-05,
+      "loss": 0.3451,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6185915492957745,
+      "grad_norm": 0.24859397280546572,
+      "learning_rate": 2.556390977443609e-05,
+      "loss": 0.3631,
+      "step": 1437
+    },
+    {
+      "epoch": 1.619718309859155,
+      "grad_norm": 0.28890725030429065,
+      "learning_rate": 2.5543024227234757e-05,
+      "loss": 0.3338,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6208450704225352,
+      "grad_norm": 0.2729341226525535,
+      "learning_rate": 2.552213868003342e-05,
+      "loss": 0.3402,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6219718309859155,
+      "grad_norm": 0.321511849439831,
+      "learning_rate": 2.550125313283208e-05,
+      "loss": 0.3275,
+      "step": 1440
+    },
+    {
+      "epoch": 1.623098591549296,
+      "grad_norm": 0.2596677711472582,
+      "learning_rate": 2.5480367585630744e-05,
+      "loss": 0.3477,
+      "step": 1441
+    },
+    {
+      "epoch": 1.624225352112676,
+      "grad_norm": 0.3215232803975101,
+      "learning_rate": 2.545948203842941e-05,
+      "loss": 0.3475,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6253521126760564,
+      "grad_norm": 0.3049605657312551,
+      "learning_rate": 2.5438596491228074e-05,
+      "loss": 0.3408,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6264788732394366,
+      "grad_norm": 0.33000021262211476,
+      "learning_rate": 2.5417710944026734e-05,
+      "loss": 0.3451,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6276056338028169,
+      "grad_norm": 0.28250547117805813,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.3533,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6287323943661973,
+      "grad_norm": 0.27656167843203644,
+      "learning_rate": 2.5375939849624064e-05,
+      "loss": 0.3286,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6298591549295773,
+      "grad_norm": 0.37440105959190834,
+      "learning_rate": 2.5355054302422727e-05,
+      "loss": 0.3666,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6309859154929578,
+      "grad_norm": 0.2624422620925436,
+      "learning_rate": 2.5334168755221387e-05,
+      "loss": 0.3353,
+      "step": 1448
+    },
+    {
+      "epoch": 1.632112676056338,
+      "grad_norm": 0.29009307096198833,
+      "learning_rate": 2.531328320802005e-05,
+      "loss": 0.3394,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6332394366197183,
+      "grad_norm": 0.29225161048598075,
+      "learning_rate": 2.5292397660818717e-05,
+      "loss": 0.3654,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6343661971830987,
+      "grad_norm": 0.30721504708137215,
+      "learning_rate": 2.527151211361738e-05,
+      "loss": 0.3321,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6354929577464787,
+      "grad_norm": 0.2788064690368735,
+      "learning_rate": 2.525062656641604e-05,
+      "loss": 0.3417,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6366197183098592,
+      "grad_norm": 0.26995422234528577,
+      "learning_rate": 2.5229741019214704e-05,
+      "loss": 0.3484,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6377464788732394,
+      "grad_norm": 0.28226163647947505,
+      "learning_rate": 2.520885547201337e-05,
+      "loss": 0.346,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6388732394366197,
+      "grad_norm": 0.2945446918455596,
+      "learning_rate": 2.5187969924812034e-05,
+      "loss": 0.3518,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.259788327412407,
+      "learning_rate": 2.5167084377610694e-05,
+      "loss": 0.3494,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6411267605633801,
+      "grad_norm": 0.270194461267937,
+      "learning_rate": 2.5146198830409358e-05,
+      "loss": 0.3504,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6422535211267606,
+      "grad_norm": 0.2950037384175335,
+      "learning_rate": 2.5125313283208024e-05,
+      "loss": 0.3393,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6433802816901408,
+      "grad_norm": 0.23883286607922005,
+      "learning_rate": 2.5104427736006684e-05,
+      "loss": 0.3405,
+      "step": 1459
+    },
+    {
+      "epoch": 1.644507042253521,
+      "grad_norm": 0.2630432210821515,
+      "learning_rate": 2.5083542188805348e-05,
+      "loss": 0.3357,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6456338028169015,
+      "grad_norm": 0.25060028578291704,
+      "learning_rate": 2.506265664160401e-05,
+      "loss": 0.3473,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6467605633802815,
+      "grad_norm": 0.25358297437939425,
+      "learning_rate": 2.5041771094402678e-05,
+      "loss": 0.3281,
+      "step": 1462
+    },
+    {
+      "epoch": 1.647887323943662,
+      "grad_norm": 0.27110843160789827,
+      "learning_rate": 2.5020885547201338e-05,
+      "loss": 0.364,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6490140845070422,
+      "grad_norm": 0.27859901156381256,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3275,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6501408450704225,
+      "grad_norm": 0.2590260996612864,
+      "learning_rate": 2.4979114452798665e-05,
+      "loss": 0.3339,
+      "step": 1465
+    },
+    {
+      "epoch": 1.651267605633803,
+      "grad_norm": 0.28615345729601793,
+      "learning_rate": 2.4958228905597328e-05,
+      "loss": 0.3186,
+      "step": 1466
+    },
+    {
+      "epoch": 1.652394366197183,
+      "grad_norm": 0.2561561406004291,
+      "learning_rate": 2.493734335839599e-05,
+      "loss": 0.3441,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6535211267605634,
+      "grad_norm": 0.28420263034336,
+      "learning_rate": 2.4916457811194655e-05,
+      "loss": 0.3487,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6546478873239436,
+      "grad_norm": 0.26810924062603336,
+      "learning_rate": 2.4895572263993318e-05,
+      "loss": 0.3218,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6557746478873239,
+      "grad_norm": 0.2833833485230433,
+      "learning_rate": 2.487468671679198e-05,
+      "loss": 0.3459,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6569014084507043,
+      "grad_norm": 0.31905201414375717,
+      "learning_rate": 2.485380116959064e-05,
+      "loss": 0.3619,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6580281690140843,
+      "grad_norm": 0.2858551553408948,
+      "learning_rate": 2.4832915622389308e-05,
+      "loss": 0.3495,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6591549295774648,
+      "grad_norm": 0.3322138453327257,
+      "learning_rate": 2.4812030075187968e-05,
+      "loss": 0.3401,
+      "step": 1473
+    },
+    {
+      "epoch": 1.660281690140845,
+      "grad_norm": 0.32119712572911885,
+      "learning_rate": 2.4791144527986635e-05,
+      "loss": 0.3395,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6614084507042253,
+      "grad_norm": 0.2970697742218709,
+      "learning_rate": 2.4770258980785295e-05,
+      "loss": 0.3528,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6625352112676057,
+      "grad_norm": 0.2899062355623534,
+      "learning_rate": 2.4749373433583962e-05,
+      "loss": 0.3456,
+      "step": 1476
+    },
+    {
+      "epoch": 1.663661971830986,
+      "grad_norm": 0.3241336123208444,
+      "learning_rate": 2.472848788638262e-05,
+      "loss": 0.3517,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6647887323943662,
+      "grad_norm": 0.24657467228328783,
+      "learning_rate": 2.470760233918129e-05,
+      "loss": 0.3398,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6659154929577464,
+      "grad_norm": 0.2580275287950773,
+      "learning_rate": 2.468671679197995e-05,
+      "loss": 0.3308,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6670422535211267,
+      "grad_norm": 0.2634110979008221,
+      "learning_rate": 2.4665831244778615e-05,
+      "loss": 0.3478,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6681690140845071,
+      "grad_norm": 0.2788858596184465,
+      "learning_rate": 2.4644945697577275e-05,
+      "loss": 0.3398,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6692957746478874,
+      "grad_norm": 0.31372425383389874,
+      "learning_rate": 2.4624060150375942e-05,
+      "loss": 0.337,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6704225352112676,
+      "grad_norm": 0.2509975030925098,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.3441,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6715492957746478,
+      "grad_norm": 0.2522862475159247,
+      "learning_rate": 2.458228905597327e-05,
+      "loss": 0.3338,
+      "step": 1484
+    },
+    {
+      "epoch": 1.672676056338028,
+      "grad_norm": 0.2813497727876396,
+      "learning_rate": 2.456140350877193e-05,
+      "loss": 0.3525,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6738028169014085,
+      "grad_norm": 0.31413834254162076,
+      "learning_rate": 2.4540517961570595e-05,
+      "loss": 0.3394,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6749295774647888,
+      "grad_norm": 0.26360158583575644,
+      "learning_rate": 2.4519632414369255e-05,
+      "loss": 0.3374,
+      "step": 1487
+    },
+    {
+      "epoch": 1.676056338028169,
+      "grad_norm": 0.28877767158248374,
+      "learning_rate": 2.4498746867167922e-05,
+      "loss": 0.3313,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6771830985915495,
+      "grad_norm": 0.24981169529637415,
+      "learning_rate": 2.4477861319966582e-05,
+      "loss": 0.3645,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6783098591549295,
+      "grad_norm": 0.32518492340841537,
+      "learning_rate": 2.445697577276525e-05,
+      "loss": 0.3818,
+      "step": 1490
+    },
+    {
+      "epoch": 1.67943661971831,
+      "grad_norm": 0.2849687146700346,
+      "learning_rate": 2.443609022556391e-05,
+      "loss": 0.3455,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6805633802816902,
+      "grad_norm": 0.2937565422277086,
+      "learning_rate": 2.4415204678362576e-05,
+      "loss": 0.3482,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6816901408450704,
+      "grad_norm": 0.27832882114852947,
+      "learning_rate": 2.4394319131161236e-05,
+      "loss": 0.3569,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6828169014084509,
+      "grad_norm": 0.2884288361313236,
+      "learning_rate": 2.4373433583959902e-05,
+      "loss": 0.3503,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6839436619718309,
+      "grad_norm": 0.30909359315178064,
+      "learning_rate": 2.4352548036758562e-05,
+      "loss": 0.3474,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6850704225352113,
+      "grad_norm": 0.28233050232701684,
+      "learning_rate": 2.433166248955723e-05,
+      "loss": 0.3278,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6861971830985916,
+      "grad_norm": 0.3222047849649389,
+      "learning_rate": 2.431077694235589e-05,
+      "loss": 0.3568,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6873239436619718,
+      "grad_norm": 0.25531140449026785,
+      "learning_rate": 2.4289891395154556e-05,
+      "loss": 0.3414,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6884507042253523,
+      "grad_norm": 0.3219325112504361,
+      "learning_rate": 2.4269005847953216e-05,
+      "loss": 0.3764,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6895774647887323,
+      "grad_norm": 0.3187727339219832,
+      "learning_rate": 2.424812030075188e-05,
+      "loss": 0.3404,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6907042253521127,
+      "grad_norm": 0.2586679108873934,
+      "learning_rate": 2.4227234753550543e-05,
+      "loss": 0.3321,
+      "step": 1501
+    },
+    {
+      "epoch": 1.691830985915493,
+      "grad_norm": 0.27144949193599927,
+      "learning_rate": 2.4206349206349206e-05,
+      "loss": 0.366,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6929577464788732,
+      "grad_norm": 0.30940358041853494,
+      "learning_rate": 2.418546365914787e-05,
+      "loss": 0.3305,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6940845070422537,
+      "grad_norm": 0.29242474439362176,
+      "learning_rate": 2.4164578111946533e-05,
+      "loss": 0.3509,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6952112676056337,
+      "grad_norm": 0.2930460128468937,
+      "learning_rate": 2.4143692564745196e-05,
+      "loss": 0.3562,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6963380281690141,
+      "grad_norm": 0.2935340603708641,
+      "learning_rate": 2.412280701754386e-05,
+      "loss": 0.3294,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6974647887323944,
+      "grad_norm": 0.36957759240564325,
+      "learning_rate": 2.4101921470342523e-05,
+      "loss": 0.3625,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6985915492957746,
+      "grad_norm": 0.26404789663599043,
+      "learning_rate": 2.4081035923141186e-05,
+      "loss": 0.3332,
+      "step": 1508
+    },
+    {
+      "epoch": 1.699718309859155,
+      "grad_norm": 0.322718769871893,
+      "learning_rate": 2.406015037593985e-05,
+      "loss": 0.358,
+      "step": 1509
+    },
+    {
+      "epoch": 1.700845070422535,
+      "grad_norm": 0.3190498928145156,
+      "learning_rate": 2.4039264828738513e-05,
+      "loss": 0.3455,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7019718309859155,
+      "grad_norm": 0.2799891294624286,
+      "learning_rate": 2.4018379281537176e-05,
+      "loss": 0.3622,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7030985915492958,
+      "grad_norm": 0.3247412168234658,
+      "learning_rate": 2.399749373433584e-05,
+      "loss": 0.3517,
+      "step": 1512
+    },
+    {
+      "epoch": 1.704225352112676,
+      "grad_norm": 0.3178652704989509,
+      "learning_rate": 2.3976608187134503e-05,
+      "loss": 0.3587,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7053521126760565,
+      "grad_norm": 0.30342422099773886,
+      "learning_rate": 2.3955722639933167e-05,
+      "loss": 0.3433,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7064788732394365,
+      "grad_norm": 0.3195855337345034,
+      "learning_rate": 2.393483709273183e-05,
+      "loss": 0.3254,
+      "step": 1515
+    },
+    {
+      "epoch": 1.707605633802817,
+      "grad_norm": 0.31217809950297026,
+      "learning_rate": 2.3913951545530493e-05,
+      "loss": 0.3579,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7087323943661972,
+      "grad_norm": 0.299422990510086,
+      "learning_rate": 2.3893065998329157e-05,
+      "loss": 0.3357,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7098591549295774,
+      "grad_norm": 0.3175609532668606,
+      "learning_rate": 2.387218045112782e-05,
+      "loss": 0.3371,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7109859154929579,
+      "grad_norm": 0.37076828925369637,
+      "learning_rate": 2.3851294903926483e-05,
+      "loss": 0.3743,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7121126760563379,
+      "grad_norm": 0.3391543904315033,
+      "learning_rate": 2.3830409356725147e-05,
+      "loss": 0.3319,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7132394366197183,
+      "grad_norm": 0.33903528196620386,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.3574,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7143661971830986,
+      "grad_norm": 0.2764464452991892,
+      "learning_rate": 2.3788638262322474e-05,
+      "loss": 0.3372,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7154929577464788,
+      "grad_norm": 0.3398776749901052,
+      "learning_rate": 2.3767752715121137e-05,
+      "loss": 0.3549,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7166197183098593,
+      "grad_norm": 0.28957579450024307,
+      "learning_rate": 2.37468671679198e-05,
+      "loss": 0.3444,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7177464788732393,
+      "grad_norm": 0.29884466819939715,
+      "learning_rate": 2.3725981620718464e-05,
+      "loss": 0.3347,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7188732394366197,
+      "grad_norm": 0.30274130464516896,
+      "learning_rate": 2.3705096073517127e-05,
+      "loss": 0.3585,
+      "step": 1526
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.305132519989941,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 0.3242,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7211267605633802,
+      "grad_norm": 0.34652155759374775,
+      "learning_rate": 2.3663324979114454e-05,
+      "loss": 0.3651,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7222535211267607,
+      "grad_norm": 0.27639427261114835,
+      "learning_rate": 2.3642439431913117e-05,
+      "loss": 0.3354,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7233802816901407,
+      "grad_norm": 0.3191386977012033,
+      "learning_rate": 2.362155388471178e-05,
+      "loss": 0.3437,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7245070422535211,
+      "grad_norm": 0.3363085997287805,
+      "learning_rate": 2.3600668337510444e-05,
+      "loss": 0.3337,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7256338028169014,
+      "grad_norm": 0.3038076959659683,
+      "learning_rate": 2.3579782790309107e-05,
+      "loss": 0.3626,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7267605633802816,
+      "grad_norm": 0.3295456357864944,
+      "learning_rate": 2.355889724310777e-05,
+      "loss": 0.329,
+      "step": 1533
+    },
+    {
+      "epoch": 1.727887323943662,
+      "grad_norm": 0.3608395454167289,
+      "learning_rate": 2.3538011695906434e-05,
+      "loss": 0.3434,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7290140845070423,
+      "grad_norm": 0.305318898574501,
+      "learning_rate": 2.3517126148705097e-05,
+      "loss": 0.3349,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7301408450704225,
+      "grad_norm": 0.33016464265419676,
+      "learning_rate": 2.349624060150376e-05,
+      "loss": 0.339,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7312676056338028,
+      "grad_norm": 0.2908830781714398,
+      "learning_rate": 2.3475355054302424e-05,
+      "loss": 0.3392,
+      "step": 1537
+    },
+    {
+      "epoch": 1.732394366197183,
+      "grad_norm": 0.3143787621188546,
+      "learning_rate": 2.3454469507101088e-05,
+      "loss": 0.3618,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7335211267605635,
+      "grad_norm": 0.3055235467368555,
+      "learning_rate": 2.3433583959899748e-05,
+      "loss": 0.3291,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7346478873239437,
+      "grad_norm": 0.33773067011488017,
+      "learning_rate": 2.3412698412698414e-05,
+      "loss": 0.3402,
+      "step": 1540
+    },
+    {
+      "epoch": 1.735774647887324,
+      "grad_norm": 0.3103493669644309,
+      "learning_rate": 2.3391812865497074e-05,
+      "loss": 0.3151,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7369014084507042,
+      "grad_norm": 0.2640626886368537,
+      "learning_rate": 2.337092731829574e-05,
+      "loss": 0.3479,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7380281690140844,
+      "grad_norm": 0.32926649547925485,
+      "learning_rate": 2.33500417710944e-05,
+      "loss": 0.3434,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7391549295774649,
+      "grad_norm": 0.28911071323480914,
+      "learning_rate": 2.3329156223893068e-05,
+      "loss": 0.3644,
+      "step": 1544
+    },
+    {
+      "epoch": 1.740281690140845,
+      "grad_norm": 0.3516859251363121,
+      "learning_rate": 2.3308270676691728e-05,
+      "loss": 0.3466,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7414084507042253,
+      "grad_norm": 0.2830494642914461,
+      "learning_rate": 2.3287385129490395e-05,
+      "loss": 0.3333,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7425352112676058,
+      "grad_norm": 0.35115973878932605,
+      "learning_rate": 2.3266499582289055e-05,
+      "loss": 0.3471,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7436619718309858,
+      "grad_norm": 0.29708723639195195,
+      "learning_rate": 2.324561403508772e-05,
+      "loss": 0.3327,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7447887323943663,
+      "grad_norm": 0.29149822506565004,
+      "learning_rate": 2.322472848788638e-05,
+      "loss": 0.3564,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7459154929577465,
+      "grad_norm": 0.3059464683757231,
+      "learning_rate": 2.3203842940685048e-05,
+      "loss": 0.374,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7470422535211267,
+      "grad_norm": 0.24945858165460774,
+      "learning_rate": 2.3182957393483708e-05,
+      "loss": 0.3373,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7481690140845072,
+      "grad_norm": 0.2837768348385519,
+      "learning_rate": 2.3162071846282375e-05,
+      "loss": 0.3421,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7492957746478872,
+      "grad_norm": 0.28899009777285367,
+      "learning_rate": 2.3141186299081035e-05,
+      "loss": 0.3422,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7504225352112677,
+      "grad_norm": 0.3185735213904924,
+      "learning_rate": 2.31203007518797e-05,
+      "loss": 0.3479,
+      "step": 1554
+    },
+    {
+      "epoch": 1.751549295774648,
+      "grad_norm": 0.2727834946078125,
+      "learning_rate": 2.309941520467836e-05,
+      "loss": 0.3515,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7526760563380281,
+      "grad_norm": 0.2920114525814956,
+      "learning_rate": 2.3078529657477028e-05,
+      "loss": 0.3259,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7538028169014086,
+      "grad_norm": 0.2774511777486092,
+      "learning_rate": 2.3057644110275688e-05,
+      "loss": 0.3444,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7549295774647886,
+      "grad_norm": 0.33462145841096125,
+      "learning_rate": 2.3036758563074355e-05,
+      "loss": 0.3394,
+      "step": 1558
+    },
+    {
+      "epoch": 1.756056338028169,
+      "grad_norm": 0.28517999320080967,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.347,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7571830985915493,
+      "grad_norm": 0.3207924838988884,
+      "learning_rate": 2.2994987468671682e-05,
+      "loss": 0.3458,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7583098591549295,
+      "grad_norm": 0.342362737976875,
+      "learning_rate": 2.2974101921470342e-05,
+      "loss": 0.3757,
+      "step": 1561
+    },
+    {
+      "epoch": 1.75943661971831,
+      "grad_norm": 0.3248576570360329,
+      "learning_rate": 2.295321637426901e-05,
+      "loss": 0.3516,
+      "step": 1562
+    },
+    {
+      "epoch": 1.76056338028169,
+      "grad_norm": 0.287455125644324,
+      "learning_rate": 2.293233082706767e-05,
+      "loss": 0.3393,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7616901408450705,
+      "grad_norm": 0.2868575359250464,
+      "learning_rate": 2.2911445279866335e-05,
+      "loss": 0.3365,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7628169014084507,
+      "grad_norm": 0.2549480950241948,
+      "learning_rate": 2.2890559732664995e-05,
+      "loss": 0.343,
+      "step": 1565
+    },
+    {
+      "epoch": 1.763943661971831,
+      "grad_norm": 0.282209289641925,
+      "learning_rate": 2.2869674185463662e-05,
+      "loss": 0.3366,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7650704225352114,
+      "grad_norm": 0.27361435419089486,
+      "learning_rate": 2.2848788638262322e-05,
+      "loss": 0.3505,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7661971830985914,
+      "grad_norm": 0.2619276537280242,
+      "learning_rate": 2.2827903091060985e-05,
+      "loss": 0.3459,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7673239436619719,
+      "grad_norm": 0.2746835426997974,
+      "learning_rate": 2.280701754385965e-05,
+      "loss": 0.3608,
+      "step": 1569
+    },
+    {
+      "epoch": 1.768450704225352,
+      "grad_norm": 0.2883899114472481,
+      "learning_rate": 2.2786131996658312e-05,
+      "loss": 0.3481,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7695774647887323,
+      "grad_norm": 0.27976723869096415,
+      "learning_rate": 2.2765246449456976e-05,
+      "loss": 0.3453,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7707042253521128,
+      "grad_norm": 0.29356343885407754,
+      "learning_rate": 2.274436090225564e-05,
+      "loss": 0.3225,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7718309859154928,
+      "grad_norm": 0.2548029064797793,
+      "learning_rate": 2.2723475355054302e-05,
+      "loss": 0.3452,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7729577464788733,
+      "grad_norm": 0.2887543722226567,
+      "learning_rate": 2.2702589807852966e-05,
+      "loss": 0.3425,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7740845070422535,
+      "grad_norm": 0.30230412782024907,
+      "learning_rate": 2.268170426065163e-05,
+      "loss": 0.3473,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7752112676056337,
+      "grad_norm": 0.31160638538614543,
+      "learning_rate": 2.2660818713450292e-05,
+      "loss": 0.3744,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7763380281690142,
+      "grad_norm": 0.29139656731845104,
+      "learning_rate": 2.2639933166248956e-05,
+      "loss": 0.3514,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7774647887323942,
+      "grad_norm": 0.33332671251383295,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 0.3432,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7785915492957747,
+      "grad_norm": 0.28747603115955583,
+      "learning_rate": 2.2598162071846283e-05,
+      "loss": 0.3304,
+      "step": 1579
+    },
+    {
+      "epoch": 1.779718309859155,
+      "grad_norm": 0.2779297515417481,
+      "learning_rate": 2.2577276524644946e-05,
+      "loss": 0.3297,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7808450704225351,
+      "grad_norm": 0.28845443740956794,
+      "learning_rate": 2.255639097744361e-05,
+      "loss": 0.3488,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7819718309859156,
+      "grad_norm": 0.28781351634594954,
+      "learning_rate": 2.2535505430242273e-05,
+      "loss": 0.3438,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7830985915492956,
+      "grad_norm": 0.26061432523931766,
+      "learning_rate": 2.2514619883040936e-05,
+      "loss": 0.345,
+      "step": 1583
+    },
+    {
+      "epoch": 1.784225352112676,
+      "grad_norm": 0.2668349047223478,
+      "learning_rate": 2.24937343358396e-05,
+      "loss": 0.34,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7853521126760563,
+      "grad_norm": 0.2639968324519894,
+      "learning_rate": 2.2472848788638263e-05,
+      "loss": 0.3329,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7864788732394365,
+      "grad_norm": 0.2723178234361553,
+      "learning_rate": 2.2451963241436926e-05,
+      "loss": 0.3263,
+      "step": 1586
+    },
+    {
+      "epoch": 1.787605633802817,
+      "grad_norm": 0.29903286947902974,
+      "learning_rate": 2.243107769423559e-05,
+      "loss": 0.3306,
+      "step": 1587
+    },
+    {
+      "epoch": 1.788732394366197,
+      "grad_norm": 0.24254004324999334,
+      "learning_rate": 2.2410192147034253e-05,
+      "loss": 0.3404,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7898591549295775,
+      "grad_norm": 0.7213667973586395,
+      "learning_rate": 2.2389306599832916e-05,
+      "loss": 0.3265,
+      "step": 1589
+    },
+    {
+      "epoch": 1.7909859154929577,
+      "grad_norm": 0.2833182002249716,
+      "learning_rate": 2.236842105263158e-05,
+      "loss": 0.3431,
+      "step": 1590
+    },
+    {
+      "epoch": 1.792112676056338,
+      "grad_norm": 0.2455273694404395,
+      "learning_rate": 2.2347535505430243e-05,
+      "loss": 0.331,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7932394366197184,
+      "grad_norm": 0.312552449696081,
+      "learning_rate": 2.2326649958228906e-05,
+      "loss": 0.3393,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7943661971830986,
+      "grad_norm": 0.31396411445311406,
+      "learning_rate": 2.230576441102757e-05,
+      "loss": 0.3467,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7954929577464789,
+      "grad_norm": 0.3056183583942239,
+      "learning_rate": 2.2284878863826233e-05,
+      "loss": 0.3624,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7966197183098591,
+      "grad_norm": 0.2761911063940829,
+      "learning_rate": 2.2263993316624897e-05,
+      "loss": 0.3358,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7977464788732394,
+      "grad_norm": 0.25345443456743505,
+      "learning_rate": 2.224310776942356e-05,
+      "loss": 0.3387,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7988732394366198,
+      "grad_norm": 0.33926576308711215,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3379,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2766434018395404,
+      "learning_rate": 2.2201336675020887e-05,
+      "loss": 0.3523,
+      "step": 1598
+    },
+    {
+      "epoch": 1.8011267605633803,
+      "grad_norm": 0.27350545871668236,
+      "learning_rate": 2.218045112781955e-05,
+      "loss": 0.3237,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8022535211267605,
+      "grad_norm": 0.29343193244243265,
+      "learning_rate": 2.2159565580618213e-05,
+      "loss": 0.3408,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8033802816901408,
+      "grad_norm": 0.30098082503394646,
+      "learning_rate": 2.2138680033416877e-05,
+      "loss": 0.3511,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8045070422535212,
+      "grad_norm": 0.31054186922907717,
+      "learning_rate": 2.211779448621554e-05,
+      "loss": 0.3353,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8056338028169014,
+      "grad_norm": 0.28301958875610483,
+      "learning_rate": 2.2096908939014204e-05,
+      "loss": 0.3485,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8067605633802817,
+      "grad_norm": 0.2942985022402299,
+      "learning_rate": 2.2076023391812867e-05,
+      "loss": 0.3377,
+      "step": 1604
+    },
+    {
+      "epoch": 1.807887323943662,
+      "grad_norm": 0.35468777953769726,
+      "learning_rate": 2.205513784461153e-05,
+      "loss": 0.3551,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8090140845070422,
+      "grad_norm": 0.3242292242406716,
+      "learning_rate": 2.2034252297410194e-05,
+      "loss": 0.3204,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8101408450704226,
+      "grad_norm": 0.3143821962891906,
+      "learning_rate": 2.2013366750208854e-05,
+      "loss": 0.3412,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8112676056338028,
+      "grad_norm": 0.27534323950223133,
+      "learning_rate": 2.199248120300752e-05,
+      "loss": 0.3577,
+      "step": 1608
+    },
+    {
+      "epoch": 1.812394366197183,
+      "grad_norm": 0.25069024865893086,
+      "learning_rate": 2.197159565580618e-05,
+      "loss": 0.3308,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8135211267605635,
+      "grad_norm": 0.33679641393934495,
+      "learning_rate": 2.1950710108604847e-05,
+      "loss": 0.3312,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8146478873239436,
+      "grad_norm": 0.26647094286661044,
+      "learning_rate": 2.1929824561403507e-05,
+      "loss": 0.3466,
+      "step": 1611
+    },
+    {
+      "epoch": 1.815774647887324,
+      "grad_norm": 0.30509100023725005,
+      "learning_rate": 2.1908939014202174e-05,
+      "loss": 0.3547,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8169014084507042,
+      "grad_norm": 0.3493657451369854,
+      "learning_rate": 2.1888053467000834e-05,
+      "loss": 0.3624,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8180281690140845,
+      "grad_norm": 0.3230058198263615,
+      "learning_rate": 2.18671679197995e-05,
+      "loss": 0.3588,
+      "step": 1614
+    },
+    {
+      "epoch": 1.819154929577465,
+      "grad_norm": 0.3677734400847891,
+      "learning_rate": 2.184628237259816e-05,
+      "loss": 0.3307,
+      "step": 1615
+    },
+    {
+      "epoch": 1.820281690140845,
+      "grad_norm": 0.26197131961421155,
+      "learning_rate": 2.1825396825396827e-05,
+      "loss": 0.3349,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8214084507042254,
+      "grad_norm": 0.30393753088662656,
+      "learning_rate": 2.1804511278195487e-05,
+      "loss": 0.3315,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8225352112676056,
+      "grad_norm": 0.33606822603694075,
+      "learning_rate": 2.1783625730994154e-05,
+      "loss": 0.328,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8236619718309859,
+      "grad_norm": 0.27706561665932133,
+      "learning_rate": 2.1762740183792814e-05,
+      "loss": 0.3339,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8247887323943663,
+      "grad_norm": 0.3554195135230648,
+      "learning_rate": 2.174185463659148e-05,
+      "loss": 0.358,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8259154929577464,
+      "grad_norm": 0.29712856717297026,
+      "learning_rate": 2.172096908939014e-05,
+      "loss": 0.3421,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8270422535211268,
+      "grad_norm": 0.2927915595603246,
+      "learning_rate": 2.1700083542188808e-05,
+      "loss": 0.3357,
+      "step": 1622
+    },
+    {
+      "epoch": 1.828169014084507,
+      "grad_norm": 0.35141700138951115,
+      "learning_rate": 2.1679197994987468e-05,
+      "loss": 0.3363,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8292957746478873,
+      "grad_norm": 0.279248687120629,
+      "learning_rate": 2.1658312447786134e-05,
+      "loss": 0.3311,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8304225352112677,
+      "grad_norm": 0.32879501257039334,
+      "learning_rate": 2.1637426900584794e-05,
+      "loss": 0.3285,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8315492957746478,
+      "grad_norm": 0.3127639986073736,
+      "learning_rate": 2.161654135338346e-05,
+      "loss": 0.3662,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8326760563380282,
+      "grad_norm": 0.2642635116143911,
+      "learning_rate": 2.159565580618212e-05,
+      "loss": 0.3502,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8338028169014085,
+      "grad_norm": 0.3184196432962934,
+      "learning_rate": 2.1574770258980788e-05,
+      "loss": 0.3413,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8349295774647887,
+      "grad_norm": 0.2762768573978393,
+      "learning_rate": 2.1553884711779448e-05,
+      "loss": 0.335,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8360563380281691,
+      "grad_norm": 0.2639598293830892,
+      "learning_rate": 2.1532999164578115e-05,
+      "loss": 0.3487,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8371830985915492,
+      "grad_norm": 0.2910237030621018,
+      "learning_rate": 2.1512113617376775e-05,
+      "loss": 0.3339,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8383098591549296,
+      "grad_norm": 0.3118135930021659,
+      "learning_rate": 2.149122807017544e-05,
+      "loss": 0.349,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8394366197183099,
+      "grad_norm": 0.2705892616745369,
+      "learning_rate": 2.14703425229741e-05,
+      "loss": 0.3525,
+      "step": 1633
+    },
+    {
+      "epoch": 1.84056338028169,
+      "grad_norm": 0.340698961529552,
+      "learning_rate": 2.1449456975772768e-05,
+      "loss": 0.3299,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8416901408450705,
+      "grad_norm": 0.3091514758989356,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.3379,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8428169014084506,
+      "grad_norm": 0.2643485419141822,
+      "learning_rate": 2.140768588137009e-05,
+      "loss": 0.3302,
+      "step": 1636
+    },
+    {
+      "epoch": 1.843943661971831,
+      "grad_norm": 0.31087625176013,
+      "learning_rate": 2.1386800334168755e-05,
+      "loss": 0.3325,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8450704225352113,
+      "grad_norm": 0.27948087905956615,
+      "learning_rate": 2.1365914786967418e-05,
+      "loss": 0.3604,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8461971830985915,
+      "grad_norm": 0.35227442456776953,
+      "learning_rate": 2.134502923976608e-05,
+      "loss": 0.3449,
+      "step": 1639
+    },
+    {
+      "epoch": 1.847323943661972,
+      "grad_norm": 0.3142701794790016,
+      "learning_rate": 2.1324143692564745e-05,
+      "loss": 0.3465,
+      "step": 1640
+    },
+    {
+      "epoch": 1.848450704225352,
+      "grad_norm": 0.27148992742513084,
+      "learning_rate": 2.130325814536341e-05,
+      "loss": 0.343,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8495774647887324,
+      "grad_norm": 0.3131614508725724,
+      "learning_rate": 2.1282372598162072e-05,
+      "loss": 0.3483,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8507042253521127,
+      "grad_norm": 0.2696953224596118,
+      "learning_rate": 2.1261487050960735e-05,
+      "loss": 0.3222,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8518309859154929,
+      "grad_norm": 0.3297224034001454,
+      "learning_rate": 2.12406015037594e-05,
+      "loss": 0.338,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8529577464788733,
+      "grad_norm": 0.2877479636683936,
+      "learning_rate": 2.1219715956558062e-05,
+      "loss": 0.3287,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8540845070422534,
+      "grad_norm": 0.287649253375458,
+      "learning_rate": 2.1198830409356725e-05,
+      "loss": 0.3228,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8552112676056338,
+      "grad_norm": 0.3053211833522019,
+      "learning_rate": 2.117794486215539e-05,
+      "loss": 0.3364,
+      "step": 1647
+    },
+    {
+      "epoch": 1.856338028169014,
+      "grad_norm": 0.26500299421684886,
+      "learning_rate": 2.1157059314954052e-05,
+      "loss": 0.3367,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8574647887323943,
+      "grad_norm": 0.301128723829993,
+      "learning_rate": 2.1136173767752715e-05,
+      "loss": 0.3668,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8585915492957747,
+      "grad_norm": 0.26457007250253695,
+      "learning_rate": 2.111528822055138e-05,
+      "loss": 0.3312,
+      "step": 1650
+    },
+    {
+      "epoch": 1.859718309859155,
+      "grad_norm": 0.28765531421267604,
+      "learning_rate": 2.1094402673350042e-05,
+      "loss": 0.3429,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8608450704225352,
+      "grad_norm": 0.304870943493138,
+      "learning_rate": 2.1073517126148706e-05,
+      "loss": 0.3426,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8619718309859155,
+      "grad_norm": 0.2795950818927515,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 0.3267,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8630985915492957,
+      "grad_norm": 0.2754880594324992,
+      "learning_rate": 2.1031746031746032e-05,
+      "loss": 0.3242,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8642253521126761,
+      "grad_norm": 0.3027378251702463,
+      "learning_rate": 2.1010860484544696e-05,
+      "loss": 0.3485,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8653521126760564,
+      "grad_norm": 0.3177812168120693,
+      "learning_rate": 2.098997493734336e-05,
+      "loss": 0.3513,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8664788732394366,
+      "grad_norm": 0.32938906771356163,
+      "learning_rate": 2.0969089390142022e-05,
+      "loss": 0.347,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8676056338028169,
+      "grad_norm": 0.2617718930009443,
+      "learning_rate": 2.0948203842940686e-05,
+      "loss": 0.3493,
+      "step": 1658
+    },
+    {
+      "epoch": 1.868732394366197,
+      "grad_norm": 0.33340860389167637,
+      "learning_rate": 2.092731829573935e-05,
+      "loss": 0.3351,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8698591549295775,
+      "grad_norm": 0.3161357707059111,
+      "learning_rate": 2.0906432748538013e-05,
+      "loss": 0.3255,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8709859154929578,
+      "grad_norm": 0.26257517402572905,
+      "learning_rate": 2.0885547201336676e-05,
+      "loss": 0.3625,
+      "step": 1661
+    },
+    {
+      "epoch": 1.872112676056338,
+      "grad_norm": 0.28303279003083226,
+      "learning_rate": 2.086466165413534e-05,
+      "loss": 0.3163,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8732394366197183,
+      "grad_norm": 0.29747456344610324,
+      "learning_rate": 2.0843776106934003e-05,
+      "loss": 0.3327,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8743661971830985,
+      "grad_norm": 0.25373424352702173,
+      "learning_rate": 2.0822890559732666e-05,
+      "loss": 0.3308,
+      "step": 1664
+    },
+    {
+      "epoch": 1.875492957746479,
+      "grad_norm": 0.2311080800384494,
+      "learning_rate": 2.080200501253133e-05,
+      "loss": 0.3387,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8766197183098592,
+      "grad_norm": 0.2837018423422604,
+      "learning_rate": 2.0781119465329993e-05,
+      "loss": 0.32,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8777464788732394,
+      "grad_norm": 0.26993270331175634,
+      "learning_rate": 2.0760233918128656e-05,
+      "loss": 0.3459,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8788732394366199,
+      "grad_norm": 0.23928714499375003,
+      "learning_rate": 2.073934837092732e-05,
+      "loss": 0.3301,
+      "step": 1668
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.3453714187721245,
+      "learning_rate": 2.0718462823725983e-05,
+      "loss": 0.3338,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8811267605633804,
+      "grad_norm": 0.3423733909657252,
+      "learning_rate": 2.0697577276524646e-05,
+      "loss": 0.3345,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8822535211267606,
+      "grad_norm": 0.2788575696634892,
+      "learning_rate": 2.067669172932331e-05,
+      "loss": 0.3378,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8833802816901408,
+      "grad_norm": 0.3538012744028978,
+      "learning_rate": 2.0655806182121973e-05,
+      "loss": 0.3564,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8845070422535213,
+      "grad_norm": 0.25591831678615323,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.3135,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8856338028169013,
+      "grad_norm": 0.26212868447836296,
+      "learning_rate": 2.06140350877193e-05,
+      "loss": 0.338,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8867605633802818,
+      "grad_norm": 0.29447227411910276,
+      "learning_rate": 2.059314954051796e-05,
+      "loss": 0.3283,
+      "step": 1675
+    },
+    {
+      "epoch": 1.887887323943662,
+      "grad_norm": 0.3015256669930087,
+      "learning_rate": 2.0572263993316627e-05,
+      "loss": 0.3528,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8890140845070422,
+      "grad_norm": 0.3176306050699436,
+      "learning_rate": 2.0551378446115287e-05,
+      "loss": 0.3181,
+      "step": 1677
+    },
+    {
+      "epoch": 1.8901408450704227,
+      "grad_norm": 0.2976333332404098,
+      "learning_rate": 2.0530492898913953e-05,
+      "loss": 0.3484,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8912676056338027,
+      "grad_norm": 0.3110422983664593,
+      "learning_rate": 2.0509607351712613e-05,
+      "loss": 0.3355,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8923943661971832,
+      "grad_norm": 0.29233125015587214,
+      "learning_rate": 2.048872180451128e-05,
+      "loss": 0.3411,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8935211267605634,
+      "grad_norm": 0.3111388968581442,
+      "learning_rate": 2.046783625730994e-05,
+      "loss": 0.3507,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8946478873239436,
+      "grad_norm": 0.3081020386333236,
+      "learning_rate": 2.0446950710108607e-05,
+      "loss": 0.3373,
+      "step": 1682
+    },
+    {
+      "epoch": 1.895774647887324,
+      "grad_norm": 0.2988072234644409,
+      "learning_rate": 2.0426065162907267e-05,
+      "loss": 0.3457,
+      "step": 1683
+    },
+    {
+      "epoch": 1.896901408450704,
+      "grad_norm": 0.2840073115940892,
+      "learning_rate": 2.0405179615705934e-05,
+      "loss": 0.3183,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8980281690140846,
+      "grad_norm": 0.2600353216994005,
+      "learning_rate": 2.0384294068504594e-05,
+      "loss": 0.3548,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8991549295774648,
+      "grad_norm": 0.26737930725428694,
+      "learning_rate": 2.036340852130326e-05,
+      "loss": 0.3314,
+      "step": 1686
+    },
+    {
+      "epoch": 1.900281690140845,
+      "grad_norm": 0.30275166487425725,
+      "learning_rate": 2.034252297410192e-05,
+      "loss": 0.3233,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9014084507042255,
+      "grad_norm": 0.30729359639027726,
+      "learning_rate": 2.0321637426900587e-05,
+      "loss": 0.3356,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9025352112676055,
+      "grad_norm": 0.31944513824067683,
+      "learning_rate": 2.0300751879699247e-05,
+      "loss": 0.3383,
+      "step": 1689
+    },
+    {
+      "epoch": 1.903661971830986,
+      "grad_norm": 0.26274255011466285,
+      "learning_rate": 2.0279866332497914e-05,
+      "loss": 0.34,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9047887323943662,
+      "grad_norm": 0.29026862877536747,
+      "learning_rate": 2.0258980785296574e-05,
+      "loss": 0.3479,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9059154929577464,
+      "grad_norm": 0.24240529796930577,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 0.3475,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9070422535211269,
+      "grad_norm": 0.2721316935600126,
+      "learning_rate": 2.02172096908939e-05,
+      "loss": 0.3283,
+      "step": 1693
+    },
+    {
+      "epoch": 1.908169014084507,
+      "grad_norm": 0.286075938733232,
+      "learning_rate": 2.0196324143692567e-05,
+      "loss": 0.3491,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9092957746478874,
+      "grad_norm": 0.25044492660792417,
+      "learning_rate": 2.0175438596491227e-05,
+      "loss": 0.3366,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9104225352112676,
+      "grad_norm": 0.2617118447514493,
+      "learning_rate": 2.0154553049289894e-05,
+      "loss": 0.3409,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9115492957746478,
+      "grad_norm": 0.2896232484551307,
+      "learning_rate": 2.0133667502088554e-05,
+      "loss": 0.362,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9126760563380283,
+      "grad_norm": 0.2820678153733654,
+      "learning_rate": 2.011278195488722e-05,
+      "loss": 0.3211,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9138028169014083,
+      "grad_norm": 0.25316052627447677,
+      "learning_rate": 2.009189640768588e-05,
+      "loss": 0.3418,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9149295774647888,
+      "grad_norm": 0.2969628596361171,
+      "learning_rate": 2.0071010860484548e-05,
+      "loss": 0.3403,
+      "step": 1700
+    },
+    {
+      "epoch": 1.916056338028169,
+      "grad_norm": 0.3103149867868271,
+      "learning_rate": 2.0050125313283208e-05,
+      "loss": 0.3286,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9171830985915492,
+      "grad_norm": 0.25684431514460954,
+      "learning_rate": 2.0029239766081874e-05,
+      "loss": 0.3312,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9183098591549297,
+      "grad_norm": 0.23745838856526574,
+      "learning_rate": 2.0008354218880534e-05,
+      "loss": 0.3097,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9194366197183097,
+      "grad_norm": 0.2873795363633347,
+      "learning_rate": 1.9987468671679198e-05,
+      "loss": 0.344,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9205633802816902,
+      "grad_norm": 0.29633915312398484,
+      "learning_rate": 1.996658312447786e-05,
+      "loss": 0.3307,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9216901408450704,
+      "grad_norm": 0.2810829629998503,
+      "learning_rate": 1.9945697577276524e-05,
+      "loss": 0.3419,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9228169014084506,
+      "grad_norm": 0.3281832059884901,
+      "learning_rate": 1.9924812030075188e-05,
+      "loss": 0.3453,
+      "step": 1707
+    },
+    {
+      "epoch": 1.923943661971831,
+      "grad_norm": 0.3350869651799835,
+      "learning_rate": 1.990392648287385e-05,
+      "loss": 0.3616,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9250704225352113,
+      "grad_norm": 0.27519285232473656,
+      "learning_rate": 1.9883040935672515e-05,
+      "loss": 0.3432,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9261971830985916,
+      "grad_norm": 0.36486612755178166,
+      "learning_rate": 1.9862155388471178e-05,
+      "loss": 0.3439,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9273239436619718,
+      "grad_norm": 0.2972926771765651,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.3291,
+      "step": 1711
+    },
+    {
+      "epoch": 1.928450704225352,
+      "grad_norm": 0.29236209232234134,
+      "learning_rate": 1.9820384294068505e-05,
+      "loss": 0.3264,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9295774647887325,
+      "grad_norm": 0.25086892831193985,
+      "learning_rate": 1.9799498746867168e-05,
+      "loss": 0.3605,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9307042253521127,
+      "grad_norm": 0.29506199630237756,
+      "learning_rate": 1.977861319966583e-05,
+      "loss": 0.3263,
+      "step": 1714
+    },
+    {
+      "epoch": 1.931830985915493,
+      "grad_norm": 0.27982886372693655,
+      "learning_rate": 1.9757727652464495e-05,
+      "loss": 0.3374,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9329577464788732,
+      "grad_norm": 0.29117276364325867,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 0.3723,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9340845070422534,
+      "grad_norm": 0.280093507811842,
+      "learning_rate": 1.971595655806182e-05,
+      "loss": 0.3454,
+      "step": 1717
+    },
+    {
+      "epoch": 1.935211267605634,
+      "grad_norm": 0.2793099946402611,
+      "learning_rate": 1.9695071010860485e-05,
+      "loss": 0.3298,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9363380281690141,
+      "grad_norm": 0.31700496486348656,
+      "learning_rate": 1.9674185463659148e-05,
+      "loss": 0.3861,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9374647887323944,
+      "grad_norm": 0.28444988922202363,
+      "learning_rate": 1.965329991645781e-05,
+      "loss": 0.3747,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9385915492957746,
+      "grad_norm": 0.3156855684456946,
+      "learning_rate": 1.9632414369256475e-05,
+      "loss": 0.349,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9397183098591548,
+      "grad_norm": 0.2684019030063972,
+      "learning_rate": 1.961152882205514e-05,
+      "loss": 0.3358,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9408450704225353,
+      "grad_norm": 0.24664955105702524,
+      "learning_rate": 1.9590643274853802e-05,
+      "loss": 0.3468,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9419718309859155,
+      "grad_norm": 0.30750144819353686,
+      "learning_rate": 1.9569757727652465e-05,
+      "loss": 0.3296,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9430985915492958,
+      "grad_norm": 0.28855748148807314,
+      "learning_rate": 1.954887218045113e-05,
+      "loss": 0.3298,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9442253521126762,
+      "grad_norm": 0.27663366620065105,
+      "learning_rate": 1.9527986633249792e-05,
+      "loss": 0.3484,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9453521126760562,
+      "grad_norm": 0.26053405371799565,
+      "learning_rate": 1.9507101086048455e-05,
+      "loss": 0.3435,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9464788732394367,
+      "grad_norm": 0.2789834306854869,
+      "learning_rate": 1.948621553884712e-05,
+      "loss": 0.3388,
+      "step": 1728
+    },
+    {
+      "epoch": 1.947605633802817,
+      "grad_norm": 0.33677228028455536,
+      "learning_rate": 1.9465329991645782e-05,
+      "loss": 0.3454,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9487323943661972,
+      "grad_norm": 0.3078151462694492,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.3434,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9498591549295776,
+      "grad_norm": 0.3089644977305877,
+      "learning_rate": 1.942355889724311e-05,
+      "loss": 0.3413,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9509859154929576,
+      "grad_norm": 0.2817600040007839,
+      "learning_rate": 1.9402673350041772e-05,
+      "loss": 0.3435,
+      "step": 1732
+    },
+    {
+      "epoch": 1.952112676056338,
+      "grad_norm": 0.360677722904184,
+      "learning_rate": 1.9381787802840436e-05,
+      "loss": 0.3497,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9532394366197183,
+      "grad_norm": 0.3332403942471745,
+      "learning_rate": 1.93609022556391e-05,
+      "loss": 0.3432,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9543661971830986,
+      "grad_norm": 0.3451996520792954,
+      "learning_rate": 1.9340016708437762e-05,
+      "loss": 0.3457,
+      "step": 1735
+    },
+    {
+      "epoch": 1.955492957746479,
+      "grad_norm": 0.37021919612796633,
+      "learning_rate": 1.9319131161236426e-05,
+      "loss": 0.364,
+      "step": 1736
+    },
+    {
+      "epoch": 1.956619718309859,
+      "grad_norm": 0.3141412057799376,
+      "learning_rate": 1.929824561403509e-05,
+      "loss": 0.3562,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9577464788732395,
+      "grad_norm": 0.2774486726808161,
+      "learning_rate": 1.9277360066833752e-05,
+      "loss": 0.3393,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9588732394366197,
+      "grad_norm": 0.3184098259456306,
+      "learning_rate": 1.9256474519632416e-05,
+      "loss": 0.3347,
+      "step": 1739
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.25779629730566117,
+      "learning_rate": 1.923558897243108e-05,
+      "loss": 0.34,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9611267605633804,
+      "grad_norm": 0.31507109794266863,
+      "learning_rate": 1.9214703425229743e-05,
+      "loss": 0.3306,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9622535211267604,
+      "grad_norm": 0.2427966113722695,
+      "learning_rate": 1.9193817878028406e-05,
+      "loss": 0.3141,
+      "step": 1742
+    },
+    {
+      "epoch": 1.963380281690141,
+      "grad_norm": 0.2542233643337515,
+      "learning_rate": 1.9172932330827066e-05,
+      "loss": 0.3344,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9645070422535211,
+      "grad_norm": 0.32081915488615875,
+      "learning_rate": 1.9152046783625733e-05,
+      "loss": 0.339,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9656338028169014,
+      "grad_norm": 0.27906537367605055,
+      "learning_rate": 1.9131161236424393e-05,
+      "loss": 0.3348,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9667605633802818,
+      "grad_norm": 0.274000810063737,
+      "learning_rate": 1.911027568922306e-05,
+      "loss": 0.3487,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9678873239436618,
+      "grad_norm": 0.2959088718509456,
+      "learning_rate": 1.908939014202172e-05,
+      "loss": 0.3488,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9690140845070423,
+      "grad_norm": 0.4473115218653689,
+      "learning_rate": 1.9068504594820386e-05,
+      "loss": 0.3512,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9701408450704225,
+      "grad_norm": 0.2669740194740602,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.3401,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9712676056338028,
+      "grad_norm": 0.2788668712741395,
+      "learning_rate": 1.9026733500417713e-05,
+      "loss": 0.3347,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9723943661971832,
+      "grad_norm": 0.2872455377014697,
+      "learning_rate": 1.9005847953216373e-05,
+      "loss": 0.3379,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9735211267605632,
+      "grad_norm": 0.27694769236197103,
+      "learning_rate": 1.898496240601504e-05,
+      "loss": 0.3526,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9746478873239437,
+      "grad_norm": 0.28203236948807303,
+      "learning_rate": 1.89640768588137e-05,
+      "loss": 0.3412,
+      "step": 1753
+    },
+    {
+      "epoch": 1.975774647887324,
+      "grad_norm": 0.24888801814995407,
+      "learning_rate": 1.8943191311612366e-05,
+      "loss": 0.3427,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9769014084507042,
+      "grad_norm": 0.287119114978281,
+      "learning_rate": 1.8922305764411026e-05,
+      "loss": 0.3446,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9780281690140846,
+      "grad_norm": 0.2766952677979089,
+      "learning_rate": 1.8901420217209693e-05,
+      "loss": 0.3505,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9791549295774646,
+      "grad_norm": 0.3045381617070715,
+      "learning_rate": 1.8880534670008353e-05,
+      "loss": 0.3469,
+      "step": 1757
+    },
+    {
+      "epoch": 1.980281690140845,
+      "grad_norm": 0.25307779996131646,
+      "learning_rate": 1.885964912280702e-05,
+      "loss": 0.3457,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9814084507042253,
+      "grad_norm": 0.2621907977207169,
+      "learning_rate": 1.883876357560568e-05,
+      "loss": 0.3414,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9825352112676056,
+      "grad_norm": 0.28413673846352067,
+      "learning_rate": 1.8817878028404347e-05,
+      "loss": 0.3219,
+      "step": 1760
+    },
+    {
+      "epoch": 1.983661971830986,
+      "grad_norm": 0.2477558078368733,
+      "learning_rate": 1.8796992481203007e-05,
+      "loss": 0.3418,
+      "step": 1761
+    },
+    {
+      "epoch": 1.984788732394366,
+      "grad_norm": 0.2623624152689003,
+      "learning_rate": 1.8776106934001673e-05,
+      "loss": 0.3472,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9859154929577465,
+      "grad_norm": 0.32582596048800616,
+      "learning_rate": 1.8755221386800333e-05,
+      "loss": 0.3547,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9870422535211267,
+      "grad_norm": 0.2560026865469511,
+      "learning_rate": 1.8734335839599e-05,
+      "loss": 0.3452,
+      "step": 1764
+    },
+    {
+      "epoch": 1.988169014084507,
+      "grad_norm": 0.3194965526847108,
+      "learning_rate": 1.871345029239766e-05,
+      "loss": 0.3401,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9892957746478874,
+      "grad_norm": 0.2552661250188611,
+      "learning_rate": 1.8692564745196327e-05,
+      "loss": 0.345,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9904225352112674,
+      "grad_norm": 0.30136152145840883,
+      "learning_rate": 1.8671679197994987e-05,
+      "loss": 0.3431,
+      "step": 1767
+    },
+    {
+      "epoch": 1.991549295774648,
+      "grad_norm": 0.386170303761133,
+      "learning_rate": 1.8650793650793654e-05,
+      "loss": 0.3329,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9926760563380281,
+      "grad_norm": 0.25735909944386404,
+      "learning_rate": 1.8629908103592314e-05,
+      "loss": 0.3656,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9938028169014084,
+      "grad_norm": 0.35370824580018567,
+      "learning_rate": 1.860902255639098e-05,
+      "loss": 0.3343,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9949295774647888,
+      "grad_norm": 0.36421894137805505,
+      "learning_rate": 1.858813700918964e-05,
+      "loss": 0.3448,
+      "step": 1771
+    },
+    {
+      "epoch": 1.996056338028169,
+      "grad_norm": 0.30126873361621703,
+      "learning_rate": 1.8567251461988304e-05,
+      "loss": 0.3739,
+      "step": 1772
+    },
+    {
+      "epoch": 1.9971830985915493,
+      "grad_norm": 0.3259500784443935,
+      "learning_rate": 1.8546365914786967e-05,
+      "loss": 0.3325,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9983098591549295,
+      "grad_norm": 0.32438298898201956,
+      "learning_rate": 1.852548036758563e-05,
+      "loss": 0.3416,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9994366197183098,
+      "grad_norm": 0.28319736419228,
+      "learning_rate": 1.8504594820384294e-05,
+      "loss": 0.316,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4400283988411228,
+      "learning_rate": 1.8483709273182957e-05,
+      "loss": 0.304,
+      "step": 1776
+    },
+    {
+      "epoch": 2.0011267605633805,
+      "grad_norm": 0.36193074426019617,
+      "learning_rate": 1.846282372598162e-05,
+      "loss": 0.2886,
+      "step": 1777
+    },
+    {
+      "epoch": 2.0022535211267605,
+      "grad_norm": 0.2898828417148018,
+      "learning_rate": 1.8441938178780284e-05,
+      "loss": 0.2844,
+      "step": 1778
+    },
+    {
+      "epoch": 2.003380281690141,
+      "grad_norm": 0.31537351348150766,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.2768,
+      "step": 1779
+    },
+    {
+      "epoch": 2.004507042253521,
+      "grad_norm": 0.30556540318006736,
+      "learning_rate": 1.840016708437761e-05,
+      "loss": 0.2828,
+      "step": 1780
+    },
+    {
+      "epoch": 2.0056338028169014,
+      "grad_norm": 0.2875419069796752,
+      "learning_rate": 1.8379281537176274e-05,
+      "loss": 0.2813,
+      "step": 1781
+    },
+    {
+      "epoch": 2.006760563380282,
+      "grad_norm": 0.2629141920619817,
+      "learning_rate": 1.8358395989974937e-05,
+      "loss": 0.2709,
+      "step": 1782
+    },
+    {
+      "epoch": 2.007887323943662,
+      "grad_norm": 0.3049128895618629,
+      "learning_rate": 1.83375104427736e-05,
+      "loss": 0.2932,
+      "step": 1783
+    },
+    {
+      "epoch": 2.0090140845070423,
+      "grad_norm": 0.3183992325698135,
+      "learning_rate": 1.8316624895572264e-05,
+      "loss": 0.2785,
+      "step": 1784
+    },
+    {
+      "epoch": 2.0101408450704223,
+      "grad_norm": 0.2710738584069236,
+      "learning_rate": 1.8295739348370928e-05,
+      "loss": 0.2846,
+      "step": 1785
+    },
+    {
+      "epoch": 2.011267605633803,
+      "grad_norm": 0.26450209061387736,
+      "learning_rate": 1.827485380116959e-05,
+      "loss": 0.2752,
+      "step": 1786
+    },
+    {
+      "epoch": 2.0123943661971833,
+      "grad_norm": 0.3420014204418473,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.2749,
+      "step": 1787
+    },
+    {
+      "epoch": 2.0135211267605633,
+      "grad_norm": 0.2804770684477757,
+      "learning_rate": 1.8233082706766918e-05,
+      "loss": 0.2659,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0146478873239437,
+      "grad_norm": 0.25143467930958646,
+      "learning_rate": 1.821219715956558e-05,
+      "loss": 0.2622,
+      "step": 1789
+    },
+    {
+      "epoch": 2.0157746478873237,
+      "grad_norm": 0.3047744277993807,
+      "learning_rate": 1.8191311612364244e-05,
+      "loss": 0.2755,
+      "step": 1790
+    },
+    {
+      "epoch": 2.016901408450704,
+      "grad_norm": 0.2885467754272612,
+      "learning_rate": 1.8170426065162908e-05,
+      "loss": 0.2674,
+      "step": 1791
+    },
+    {
+      "epoch": 2.0180281690140847,
+      "grad_norm": 0.25497456315156125,
+      "learning_rate": 1.814954051796157e-05,
+      "loss": 0.2674,
+      "step": 1792
+    },
+    {
+      "epoch": 2.0191549295774647,
+      "grad_norm": 0.26183676682755286,
+      "learning_rate": 1.8128654970760235e-05,
+      "loss": 0.2755,
+      "step": 1793
+    },
+    {
+      "epoch": 2.020281690140845,
+      "grad_norm": 0.2647018970469358,
+      "learning_rate": 1.8107769423558898e-05,
+      "loss": 0.2668,
+      "step": 1794
+    },
+    {
+      "epoch": 2.021408450704225,
+      "grad_norm": 0.2579179710408074,
+      "learning_rate": 1.808688387635756e-05,
+      "loss": 0.2892,
+      "step": 1795
+    },
+    {
+      "epoch": 2.0225352112676056,
+      "grad_norm": 0.26710273941351514,
+      "learning_rate": 1.8065998329156225e-05,
+      "loss": 0.2754,
+      "step": 1796
+    },
+    {
+      "epoch": 2.023661971830986,
+      "grad_norm": 0.25497985314228244,
+      "learning_rate": 1.8045112781954888e-05,
+      "loss": 0.2734,
+      "step": 1797
+    },
+    {
+      "epoch": 2.024788732394366,
+      "grad_norm": 0.23892301351613357,
+      "learning_rate": 1.802422723475355e-05,
+      "loss": 0.276,
+      "step": 1798
+    },
+    {
+      "epoch": 2.0259154929577465,
+      "grad_norm": 0.26278511985156594,
+      "learning_rate": 1.8003341687552215e-05,
+      "loss": 0.2723,
+      "step": 1799
+    },
+    {
+      "epoch": 2.0270422535211265,
+      "grad_norm": 0.24659026721963412,
+      "learning_rate": 1.7982456140350878e-05,
+      "loss": 0.2689,
+      "step": 1800
+    },
+    {
+      "epoch": 2.028169014084507,
+      "grad_norm": 0.24799078147607242,
+      "learning_rate": 1.796157059314954e-05,
+      "loss": 0.2776,
+      "step": 1801
+    },
+    {
+      "epoch": 2.0292957746478875,
+      "grad_norm": 0.22333418600097785,
+      "learning_rate": 1.7940685045948205e-05,
+      "loss": 0.2816,
+      "step": 1802
+    },
+    {
+      "epoch": 2.0304225352112675,
+      "grad_norm": 0.23522702564048342,
+      "learning_rate": 1.791979949874687e-05,
+      "loss": 0.2762,
+      "step": 1803
+    },
+    {
+      "epoch": 2.031549295774648,
+      "grad_norm": 0.25393937508910336,
+      "learning_rate": 1.7898913951545532e-05,
+      "loss": 0.2831,
+      "step": 1804
+    },
+    {
+      "epoch": 2.032676056338028,
+      "grad_norm": 0.23768502018015894,
+      "learning_rate": 1.7878028404344195e-05,
+      "loss": 0.2585,
+      "step": 1805
+    },
+    {
+      "epoch": 2.0338028169014084,
+      "grad_norm": 0.2503858979894685,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 0.2644,
+      "step": 1806
+    },
+    {
+      "epoch": 2.034929577464789,
+      "grad_norm": 0.26131973579511447,
+      "learning_rate": 1.7836257309941522e-05,
+      "loss": 0.2693,
+      "step": 1807
+    },
+    {
+      "epoch": 2.036056338028169,
+      "grad_norm": 0.22749664547284826,
+      "learning_rate": 1.7815371762740185e-05,
+      "loss": 0.2715,
+      "step": 1808
+    },
+    {
+      "epoch": 2.0371830985915493,
+      "grad_norm": 0.27026999257320183,
+      "learning_rate": 1.779448621553885e-05,
+      "loss": 0.2692,
+      "step": 1809
+    },
+    {
+      "epoch": 2.03830985915493,
+      "grad_norm": 0.22997409267610297,
+      "learning_rate": 1.7773600668337512e-05,
+      "loss": 0.2637,
+      "step": 1810
+    },
+    {
+      "epoch": 2.03943661971831,
+      "grad_norm": 0.23763229339076203,
+      "learning_rate": 1.7752715121136172e-05,
+      "loss": 0.2758,
+      "step": 1811
+    },
+    {
+      "epoch": 2.0405633802816903,
+      "grad_norm": 0.2569594050394131,
+      "learning_rate": 1.773182957393484e-05,
+      "loss": 0.2884,
+      "step": 1812
+    },
+    {
+      "epoch": 2.0416901408450703,
+      "grad_norm": 0.2225541191637147,
+      "learning_rate": 1.77109440267335e-05,
+      "loss": 0.2688,
+      "step": 1813
+    },
+    {
+      "epoch": 2.0428169014084507,
+      "grad_norm": 0.26774599371649005,
+      "learning_rate": 1.7690058479532165e-05,
+      "loss": 0.2585,
+      "step": 1814
+    },
+    {
+      "epoch": 2.043943661971831,
+      "grad_norm": 0.24749302500377496,
+      "learning_rate": 1.7669172932330825e-05,
+      "loss": 0.2678,
+      "step": 1815
+    },
+    {
+      "epoch": 2.045070422535211,
+      "grad_norm": 0.2101377864414631,
+      "learning_rate": 1.7648287385129492e-05,
+      "loss": 0.263,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0461971830985917,
+      "grad_norm": 0.25971366791251355,
+      "learning_rate": 1.7627401837928152e-05,
+      "loss": 0.2687,
+      "step": 1817
+    },
+    {
+      "epoch": 2.0473239436619717,
+      "grad_norm": 0.2639815940476354,
+      "learning_rate": 1.760651629072682e-05,
+      "loss": 0.2863,
+      "step": 1818
+    },
+    {
+      "epoch": 2.048450704225352,
+      "grad_norm": 0.24759316454790214,
+      "learning_rate": 1.758563074352548e-05,
+      "loss": 0.2618,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0495774647887326,
+      "grad_norm": 0.21908544281153025,
+      "learning_rate": 1.7564745196324146e-05,
+      "loss": 0.262,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0507042253521126,
+      "grad_norm": 0.24986814532092871,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 0.2752,
+      "step": 1821
+    },
+    {
+      "epoch": 2.051830985915493,
+      "grad_norm": 0.22903689190851212,
+      "learning_rate": 1.7522974101921472e-05,
+      "loss": 0.2748,
+      "step": 1822
+    },
+    {
+      "epoch": 2.052957746478873,
+      "grad_norm": 0.24241577149726312,
+      "learning_rate": 1.7502088554720132e-05,
+      "loss": 0.2747,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0540845070422535,
+      "grad_norm": 0.23394955802543638,
+      "learning_rate": 1.74812030075188e-05,
+      "loss": 0.2641,
+      "step": 1824
+    },
+    {
+      "epoch": 2.055211267605634,
+      "grad_norm": 0.2560912344902298,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.2754,
+      "step": 1825
+    },
+    {
+      "epoch": 2.056338028169014,
+      "grad_norm": 0.2598392341032889,
+      "learning_rate": 1.7439431913116126e-05,
+      "loss": 0.2872,
+      "step": 1826
+    },
+    {
+      "epoch": 2.0574647887323945,
+      "grad_norm": 0.22978850883466367,
+      "learning_rate": 1.7418546365914786e-05,
+      "loss": 0.2598,
+      "step": 1827
+    },
+    {
+      "epoch": 2.0585915492957745,
+      "grad_norm": 0.22254614744090132,
+      "learning_rate": 1.7397660818713453e-05,
+      "loss": 0.2517,
+      "step": 1828
+    },
+    {
+      "epoch": 2.059718309859155,
+      "grad_norm": 0.24504177887758496,
+      "learning_rate": 1.7376775271512113e-05,
+      "loss": 0.2756,
+      "step": 1829
+    },
+    {
+      "epoch": 2.0608450704225354,
+      "grad_norm": 0.23609279494765378,
+      "learning_rate": 1.735588972431078e-05,
+      "loss": 0.2529,
+      "step": 1830
+    },
+    {
+      "epoch": 2.0619718309859154,
+      "grad_norm": 0.2572397612590121,
+      "learning_rate": 1.733500417710944e-05,
+      "loss": 0.2736,
+      "step": 1831
+    },
+    {
+      "epoch": 2.063098591549296,
+      "grad_norm": 0.24132343329906028,
+      "learning_rate": 1.7314118629908106e-05,
+      "loss": 0.268,
+      "step": 1832
+    },
+    {
+      "epoch": 2.064225352112676,
+      "grad_norm": 0.2571060147938181,
+      "learning_rate": 1.7293233082706766e-05,
+      "loss": 0.2635,
+      "step": 1833
+    },
+    {
+      "epoch": 2.0653521126760563,
+      "grad_norm": 0.2207625937897721,
+      "learning_rate": 1.7272347535505433e-05,
+      "loss": 0.2652,
+      "step": 1834
+    },
+    {
+      "epoch": 2.066478873239437,
+      "grad_norm": 0.22787090805449173,
+      "learning_rate": 1.7251461988304093e-05,
+      "loss": 0.2598,
+      "step": 1835
+    },
+    {
+      "epoch": 2.067605633802817,
+      "grad_norm": 0.2357951714204803,
+      "learning_rate": 1.723057644110276e-05,
+      "loss": 0.2702,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0687323943661973,
+      "grad_norm": 0.26844254869193424,
+      "learning_rate": 1.720969089390142e-05,
+      "loss": 0.2688,
+      "step": 1837
+    },
+    {
+      "epoch": 2.0698591549295773,
+      "grad_norm": 0.24650719910111354,
+      "learning_rate": 1.7188805346700086e-05,
+      "loss": 0.2845,
+      "step": 1838
+    },
+    {
+      "epoch": 2.0709859154929577,
+      "grad_norm": 0.21170408427745419,
+      "learning_rate": 1.7167919799498746e-05,
+      "loss": 0.2722,
+      "step": 1839
+    },
+    {
+      "epoch": 2.072112676056338,
+      "grad_norm": 0.23270936286588711,
+      "learning_rate": 1.714703425229741e-05,
+      "loss": 0.2732,
+      "step": 1840
+    },
+    {
+      "epoch": 2.073239436619718,
+      "grad_norm": 0.2776846836673971,
+      "learning_rate": 1.7126148705096073e-05,
+      "loss": 0.275,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0743661971830987,
+      "grad_norm": 0.2425811214938422,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 0.2696,
+      "step": 1842
+    },
+    {
+      "epoch": 2.0754929577464787,
+      "grad_norm": 0.22164171885336945,
+      "learning_rate": 1.70843776106934e-05,
+      "loss": 0.2592,
+      "step": 1843
+    },
+    {
+      "epoch": 2.076619718309859,
+      "grad_norm": 0.23609097727556208,
+      "learning_rate": 1.7063492063492063e-05,
+      "loss": 0.2785,
+      "step": 1844
+    },
+    {
+      "epoch": 2.0777464788732396,
+      "grad_norm": 0.20312418904361224,
+      "learning_rate": 1.7042606516290727e-05,
+      "loss": 0.2538,
+      "step": 1845
+    },
+    {
+      "epoch": 2.0788732394366196,
+      "grad_norm": 0.21196290946837307,
+      "learning_rate": 1.702172096908939e-05,
+      "loss": 0.2567,
+      "step": 1846
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.229275235692593,
+      "learning_rate": 1.7000835421888053e-05,
+      "loss": 0.2825,
+      "step": 1847
+    },
+    {
+      "epoch": 2.08112676056338,
+      "grad_norm": 0.23106628116686928,
+      "learning_rate": 1.6979949874686717e-05,
+      "loss": 0.2672,
+      "step": 1848
+    },
+    {
+      "epoch": 2.0822535211267605,
+      "grad_norm": 0.22809768966738436,
+      "learning_rate": 1.695906432748538e-05,
+      "loss": 0.2758,
+      "step": 1849
+    },
+    {
+      "epoch": 2.083380281690141,
+      "grad_norm": 0.23270495974715888,
+      "learning_rate": 1.6938178780284044e-05,
+      "loss": 0.2824,
+      "step": 1850
+    },
+    {
+      "epoch": 2.084507042253521,
+      "grad_norm": 0.22996623133844957,
+      "learning_rate": 1.6917293233082707e-05,
+      "loss": 0.2601,
+      "step": 1851
+    },
+    {
+      "epoch": 2.0856338028169015,
+      "grad_norm": 0.22968083512308185,
+      "learning_rate": 1.689640768588137e-05,
+      "loss": 0.2729,
+      "step": 1852
+    },
+    {
+      "epoch": 2.0867605633802815,
+      "grad_norm": 0.24673638985141516,
+      "learning_rate": 1.6875522138680034e-05,
+      "loss": 0.2567,
+      "step": 1853
+    },
+    {
+      "epoch": 2.087887323943662,
+      "grad_norm": 0.23766764086953465,
+      "learning_rate": 1.6854636591478697e-05,
+      "loss": 0.2701,
+      "step": 1854
+    },
+    {
+      "epoch": 2.0890140845070424,
+      "grad_norm": 0.2399021649038314,
+      "learning_rate": 1.683375104427736e-05,
+      "loss": 0.2696,
+      "step": 1855
+    },
+    {
+      "epoch": 2.0901408450704224,
+      "grad_norm": 0.23486201387497946,
+      "learning_rate": 1.6812865497076024e-05,
+      "loss": 0.2721,
+      "step": 1856
+    },
+    {
+      "epoch": 2.091267605633803,
+      "grad_norm": 0.22721396393200446,
+      "learning_rate": 1.6791979949874687e-05,
+      "loss": 0.2733,
+      "step": 1857
+    },
+    {
+      "epoch": 2.092394366197183,
+      "grad_norm": 0.2355585301253839,
+      "learning_rate": 1.677109440267335e-05,
+      "loss": 0.278,
+      "step": 1858
+    },
+    {
+      "epoch": 2.0935211267605633,
+      "grad_norm": 0.2178378033963335,
+      "learning_rate": 1.6750208855472014e-05,
+      "loss": 0.2692,
+      "step": 1859
+    },
+    {
+      "epoch": 2.094647887323944,
+      "grad_norm": 0.24264463238025005,
+      "learning_rate": 1.6729323308270677e-05,
+      "loss": 0.2727,
+      "step": 1860
+    },
+    {
+      "epoch": 2.095774647887324,
+      "grad_norm": 0.21911277932782425,
+      "learning_rate": 1.670843776106934e-05,
+      "loss": 0.2505,
+      "step": 1861
+    },
+    {
+      "epoch": 2.0969014084507043,
+      "grad_norm": 0.23178142619843944,
+      "learning_rate": 1.6687552213868004e-05,
+      "loss": 0.2619,
+      "step": 1862
+    },
+    {
+      "epoch": 2.0980281690140847,
+      "grad_norm": 0.252731984105388,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2784,
+      "step": 1863
+    },
+    {
+      "epoch": 2.0991549295774647,
+      "grad_norm": 0.21672351302271223,
+      "learning_rate": 1.664578111946533e-05,
+      "loss": 0.2636,
+      "step": 1864
+    },
+    {
+      "epoch": 2.100281690140845,
+      "grad_norm": 0.25773586636908935,
+      "learning_rate": 1.6624895572263994e-05,
+      "loss": 0.2798,
+      "step": 1865
+    },
+    {
+      "epoch": 2.101408450704225,
+      "grad_norm": 0.25307351443490506,
+      "learning_rate": 1.6604010025062658e-05,
+      "loss": 0.2706,
+      "step": 1866
+    },
+    {
+      "epoch": 2.1025352112676057,
+      "grad_norm": 0.27472243989047346,
+      "learning_rate": 1.658312447786132e-05,
+      "loss": 0.2946,
+      "step": 1867
+    },
+    {
+      "epoch": 2.103661971830986,
+      "grad_norm": 0.24031345131054146,
+      "learning_rate": 1.6562238930659984e-05,
+      "loss": 0.2682,
+      "step": 1868
+    },
+    {
+      "epoch": 2.104788732394366,
+      "grad_norm": 0.25162922606579063,
+      "learning_rate": 1.6541353383458648e-05,
+      "loss": 0.283,
+      "step": 1869
+    },
+    {
+      "epoch": 2.1059154929577466,
+      "grad_norm": 0.24837686555353697,
+      "learning_rate": 1.652046783625731e-05,
+      "loss": 0.2799,
+      "step": 1870
+    },
+    {
+      "epoch": 2.1070422535211266,
+      "grad_norm": 0.23073933513436476,
+      "learning_rate": 1.6499582289055974e-05,
+      "loss": 0.2618,
+      "step": 1871
+    },
+    {
+      "epoch": 2.108169014084507,
+      "grad_norm": 0.2153699627040841,
+      "learning_rate": 1.6478696741854638e-05,
+      "loss": 0.2615,
+      "step": 1872
+    },
+    {
+      "epoch": 2.1092957746478875,
+      "grad_norm": 0.23948382707329402,
+      "learning_rate": 1.64578111946533e-05,
+      "loss": 0.2728,
+      "step": 1873
+    },
+    {
+      "epoch": 2.1104225352112675,
+      "grad_norm": 0.2520250671386909,
+      "learning_rate": 1.6436925647451965e-05,
+      "loss": 0.2685,
+      "step": 1874
+    },
+    {
+      "epoch": 2.111549295774648,
+      "grad_norm": 0.2553208905599901,
+      "learning_rate": 1.6416040100250628e-05,
+      "loss": 0.2694,
+      "step": 1875
+    },
+    {
+      "epoch": 2.112676056338028,
+      "grad_norm": 0.23661034067017267,
+      "learning_rate": 1.639515455304929e-05,
+      "loss": 0.2612,
+      "step": 1876
+    },
+    {
+      "epoch": 2.1138028169014085,
+      "grad_norm": 0.2751315951325099,
+      "learning_rate": 1.6374269005847955e-05,
+      "loss": 0.2683,
+      "step": 1877
+    },
+    {
+      "epoch": 2.114929577464789,
+      "grad_norm": 0.2518068149348879,
+      "learning_rate": 1.6353383458646618e-05,
+      "loss": 0.2836,
+      "step": 1878
+    },
+    {
+      "epoch": 2.116056338028169,
+      "grad_norm": 0.2611569444554147,
+      "learning_rate": 1.6332497911445278e-05,
+      "loss": 0.291,
+      "step": 1879
+    },
+    {
+      "epoch": 2.1171830985915494,
+      "grad_norm": 0.28586130206224214,
+      "learning_rate": 1.6311612364243945e-05,
+      "loss": 0.2582,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1183098591549294,
+      "grad_norm": 0.2683318383289673,
+      "learning_rate": 1.6290726817042605e-05,
+      "loss": 0.2725,
+      "step": 1881
+    },
+    {
+      "epoch": 2.11943661971831,
+      "grad_norm": 0.29793498860011597,
+      "learning_rate": 1.626984126984127e-05,
+      "loss": 0.2909,
+      "step": 1882
+    },
+    {
+      "epoch": 2.1205633802816903,
+      "grad_norm": 0.21648859977673668,
+      "learning_rate": 1.624895572263993e-05,
+      "loss": 0.2911,
+      "step": 1883
+    },
+    {
+      "epoch": 2.1216901408450703,
+      "grad_norm": 0.2519696022244229,
+      "learning_rate": 1.62280701754386e-05,
+      "loss": 0.2632,
+      "step": 1884
+    },
+    {
+      "epoch": 2.122816901408451,
+      "grad_norm": 0.2409211931064604,
+      "learning_rate": 1.620718462823726e-05,
+      "loss": 0.2569,
+      "step": 1885
+    },
+    {
+      "epoch": 2.123943661971831,
+      "grad_norm": 0.2405120416752815,
+      "learning_rate": 1.6186299081035925e-05,
+      "loss": 0.2773,
+      "step": 1886
+    },
+    {
+      "epoch": 2.1250704225352113,
+      "grad_norm": 0.23129455393921802,
+      "learning_rate": 1.6165413533834585e-05,
+      "loss": 0.2609,
+      "step": 1887
+    },
+    {
+      "epoch": 2.1261971830985917,
+      "grad_norm": 0.30413505581317013,
+      "learning_rate": 1.6144527986633252e-05,
+      "loss": 0.2961,
+      "step": 1888
+    },
+    {
+      "epoch": 2.1273239436619718,
+      "grad_norm": 0.22533717087570718,
+      "learning_rate": 1.6123642439431912e-05,
+      "loss": 0.273,
+      "step": 1889
+    },
+    {
+      "epoch": 2.128450704225352,
+      "grad_norm": 0.21954251927197982,
+      "learning_rate": 1.610275689223058e-05,
+      "loss": 0.2724,
+      "step": 1890
+    },
+    {
+      "epoch": 2.129577464788732,
+      "grad_norm": 0.22588652986281654,
+      "learning_rate": 1.608187134502924e-05,
+      "loss": 0.2752,
+      "step": 1891
+    },
+    {
+      "epoch": 2.1307042253521127,
+      "grad_norm": 0.2342667299365894,
+      "learning_rate": 1.6060985797827905e-05,
+      "loss": 0.2956,
+      "step": 1892
+    },
+    {
+      "epoch": 2.131830985915493,
+      "grad_norm": 0.22987006498758314,
+      "learning_rate": 1.6040100250626565e-05,
+      "loss": 0.2735,
+      "step": 1893
+    },
+    {
+      "epoch": 2.132957746478873,
+      "grad_norm": 0.2165995732033388,
+      "learning_rate": 1.6019214703425232e-05,
+      "loss": 0.2658,
+      "step": 1894
+    },
+    {
+      "epoch": 2.1340845070422536,
+      "grad_norm": 0.23369412691232638,
+      "learning_rate": 1.5998329156223892e-05,
+      "loss": 0.2768,
+      "step": 1895
+    },
+    {
+      "epoch": 2.1352112676056336,
+      "grad_norm": 0.22527571990992717,
+      "learning_rate": 1.597744360902256e-05,
+      "loss": 0.2803,
+      "step": 1896
+    },
+    {
+      "epoch": 2.136338028169014,
+      "grad_norm": 0.22307214273493317,
+      "learning_rate": 1.595655806182122e-05,
+      "loss": 0.2587,
+      "step": 1897
+    },
+    {
+      "epoch": 2.1374647887323945,
+      "grad_norm": 0.2308702963968873,
+      "learning_rate": 1.5935672514619886e-05,
+      "loss": 0.2667,
+      "step": 1898
+    },
+    {
+      "epoch": 2.1385915492957746,
+      "grad_norm": 0.21884535384212295,
+      "learning_rate": 1.5914786967418546e-05,
+      "loss": 0.2673,
+      "step": 1899
+    },
+    {
+      "epoch": 2.139718309859155,
+      "grad_norm": 0.23083539873570982,
+      "learning_rate": 1.5893901420217212e-05,
+      "loss": 0.2664,
+      "step": 1900
+    },
+    {
+      "epoch": 2.140845070422535,
+      "grad_norm": 0.24268574749356314,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.2776,
+      "step": 1901
+    },
+    {
+      "epoch": 2.1419718309859155,
+      "grad_norm": 0.22809985206158837,
+      "learning_rate": 1.585213032581454e-05,
+      "loss": 0.2832,
+      "step": 1902
+    },
+    {
+      "epoch": 2.143098591549296,
+      "grad_norm": 0.21387933791759006,
+      "learning_rate": 1.58312447786132e-05,
+      "loss": 0.251,
+      "step": 1903
+    },
+    {
+      "epoch": 2.144225352112676,
+      "grad_norm": 0.2353271860833369,
+      "learning_rate": 1.5810359231411866e-05,
+      "loss": 0.2675,
+      "step": 1904
+    },
+    {
+      "epoch": 2.1453521126760564,
+      "grad_norm": 0.23509764350054024,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 0.2615,
+      "step": 1905
+    },
+    {
+      "epoch": 2.1464788732394364,
+      "grad_norm": 0.22530232891108226,
+      "learning_rate": 1.5768588137009193e-05,
+      "loss": 0.2751,
+      "step": 1906
+    },
+    {
+      "epoch": 2.147605633802817,
+      "grad_norm": 0.23442604251542756,
+      "learning_rate": 1.5747702589807853e-05,
+      "loss": 0.2665,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1487323943661973,
+      "grad_norm": 1.2319080043249988,
+      "learning_rate": 1.5726817042606516e-05,
+      "loss": 0.2872,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1498591549295774,
+      "grad_norm": 0.2619177006852762,
+      "learning_rate": 1.570593149540518e-05,
+      "loss": 0.2799,
+      "step": 1909
+    },
+    {
+      "epoch": 2.150985915492958,
+      "grad_norm": 0.21937142768827655,
+      "learning_rate": 1.5685045948203843e-05,
+      "loss": 0.2815,
+      "step": 1910
+    },
+    {
+      "epoch": 2.152112676056338,
+      "grad_norm": 0.20711092031146155,
+      "learning_rate": 1.5664160401002506e-05,
+      "loss": 0.2655,
+      "step": 1911
+    },
+    {
+      "epoch": 2.1532394366197183,
+      "grad_norm": 0.24552607736755352,
+      "learning_rate": 1.564327485380117e-05,
+      "loss": 0.2786,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1543661971830987,
+      "grad_norm": 0.28078844381407797,
+      "learning_rate": 1.5622389306599833e-05,
+      "loss": 0.2875,
+      "step": 1913
+    },
+    {
+      "epoch": 2.1554929577464788,
+      "grad_norm": 0.21221403574579847,
+      "learning_rate": 1.5601503759398496e-05,
+      "loss": 0.2664,
+      "step": 1914
+    },
+    {
+      "epoch": 2.156619718309859,
+      "grad_norm": 0.2602669647184901,
+      "learning_rate": 1.558061821219716e-05,
+      "loss": 0.2756,
+      "step": 1915
+    },
+    {
+      "epoch": 2.1577464788732392,
+      "grad_norm": 0.25509195485127495,
+      "learning_rate": 1.5559732664995823e-05,
+      "loss": 0.266,
+      "step": 1916
+    },
+    {
+      "epoch": 2.1588732394366197,
+      "grad_norm": 0.2328826021778946,
+      "learning_rate": 1.5538847117794486e-05,
+      "loss": 0.2735,
+      "step": 1917
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.242712790494576,
+      "learning_rate": 1.551796157059315e-05,
+      "loss": 0.2759,
+      "step": 1918
+    },
+    {
+      "epoch": 2.16112676056338,
+      "grad_norm": 0.23179279279221807,
+      "learning_rate": 1.5497076023391813e-05,
+      "loss": 0.2617,
+      "step": 1919
+    },
+    {
+      "epoch": 2.1622535211267606,
+      "grad_norm": 0.26249600722408734,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 0.2723,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1633802816901406,
+      "grad_norm": 0.2256678161675639,
+      "learning_rate": 1.545530492898914e-05,
+      "loss": 0.2697,
+      "step": 1921
+    },
+    {
+      "epoch": 2.164507042253521,
+      "grad_norm": 0.22999297934710036,
+      "learning_rate": 1.5434419381787803e-05,
+      "loss": 0.2619,
+      "step": 1922
+    },
+    {
+      "epoch": 2.1656338028169015,
+      "grad_norm": 0.24393784025390963,
+      "learning_rate": 1.5413533834586467e-05,
+      "loss": 0.2768,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1667605633802816,
+      "grad_norm": 0.23896347948557023,
+      "learning_rate": 1.539264828738513e-05,
+      "loss": 0.2673,
+      "step": 1924
+    },
+    {
+      "epoch": 2.167887323943662,
+      "grad_norm": 0.21106922466282554,
+      "learning_rate": 1.5371762740183793e-05,
+      "loss": 0.269,
+      "step": 1925
+    },
+    {
+      "epoch": 2.169014084507042,
+      "grad_norm": 0.23530802533165923,
+      "learning_rate": 1.5350877192982457e-05,
+      "loss": 0.2835,
+      "step": 1926
+    },
+    {
+      "epoch": 2.1701408450704225,
+      "grad_norm": 0.23668876249523446,
+      "learning_rate": 1.532999164578112e-05,
+      "loss": 0.2899,
+      "step": 1927
+    },
+    {
+      "epoch": 2.171267605633803,
+      "grad_norm": 0.23153335234833616,
+      "learning_rate": 1.5309106098579783e-05,
+      "loss": 0.2885,
+      "step": 1928
+    },
+    {
+      "epoch": 2.172394366197183,
+      "grad_norm": 0.23935356077783637,
+      "learning_rate": 1.5288220551378447e-05,
+      "loss": 0.262,
+      "step": 1929
+    },
+    {
+      "epoch": 2.1735211267605634,
+      "grad_norm": 0.24217670995829305,
+      "learning_rate": 1.526733500417711e-05,
+      "loss": 0.2661,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1746478873239434,
+      "grad_norm": 0.23240003368990678,
+      "learning_rate": 1.5246449456975772e-05,
+      "loss": 0.2725,
+      "step": 1931
+    },
+    {
+      "epoch": 2.175774647887324,
+      "grad_norm": 0.23175726446485997,
+      "learning_rate": 1.5225563909774437e-05,
+      "loss": 0.2638,
+      "step": 1932
+    },
+    {
+      "epoch": 2.1769014084507043,
+      "grad_norm": 0.2111102205308594,
+      "learning_rate": 1.5204678362573099e-05,
+      "loss": 0.2673,
+      "step": 1933
+    },
+    {
+      "epoch": 2.1780281690140844,
+      "grad_norm": 0.23634239259785408,
+      "learning_rate": 1.5183792815371764e-05,
+      "loss": 0.2817,
+      "step": 1934
+    },
+    {
+      "epoch": 2.179154929577465,
+      "grad_norm": 0.225470067506582,
+      "learning_rate": 1.5162907268170425e-05,
+      "loss": 0.2791,
+      "step": 1935
+    },
+    {
+      "epoch": 2.1802816901408453,
+      "grad_norm": 0.23640365976070718,
+      "learning_rate": 1.514202172096909e-05,
+      "loss": 0.2739,
+      "step": 1936
+    },
+    {
+      "epoch": 2.1814084507042253,
+      "grad_norm": 0.22139305552383293,
+      "learning_rate": 1.5121136173767752e-05,
+      "loss": 0.2649,
+      "step": 1937
+    },
+    {
+      "epoch": 2.1825352112676057,
+      "grad_norm": 0.23104278414531945,
+      "learning_rate": 1.5100250626566417e-05,
+      "loss": 0.2585,
+      "step": 1938
+    },
+    {
+      "epoch": 2.1836619718309858,
+      "grad_norm": 0.20957672014576867,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.2797,
+      "step": 1939
+    },
+    {
+      "epoch": 2.184788732394366,
+      "grad_norm": 0.34330613589580256,
+      "learning_rate": 1.5058479532163744e-05,
+      "loss": 0.2806,
+      "step": 1940
+    },
+    {
+      "epoch": 2.1859154929577467,
+      "grad_norm": 0.20503771036948973,
+      "learning_rate": 1.5037593984962406e-05,
+      "loss": 0.269,
+      "step": 1941
+    },
+    {
+      "epoch": 2.1870422535211267,
+      "grad_norm": 0.24007031747179594,
+      "learning_rate": 1.501670843776107e-05,
+      "loss": 0.2728,
+      "step": 1942
+    },
+    {
+      "epoch": 2.188169014084507,
+      "grad_norm": 0.25402861782615715,
+      "learning_rate": 1.4995822890559732e-05,
+      "loss": 0.2851,
+      "step": 1943
+    },
+    {
+      "epoch": 2.189295774647887,
+      "grad_norm": 0.23051212385499653,
+      "learning_rate": 1.4974937343358397e-05,
+      "loss": 0.2699,
+      "step": 1944
+    },
+    {
+      "epoch": 2.1904225352112676,
+      "grad_norm": 0.23558216483964192,
+      "learning_rate": 1.4954051796157059e-05,
+      "loss": 0.2665,
+      "step": 1945
+    },
+    {
+      "epoch": 2.191549295774648,
+      "grad_norm": 0.2515237013022344,
+      "learning_rate": 1.4933166248955724e-05,
+      "loss": 0.278,
+      "step": 1946
+    },
+    {
+      "epoch": 2.192676056338028,
+      "grad_norm": 0.2375005933780065,
+      "learning_rate": 1.4912280701754386e-05,
+      "loss": 0.2669,
+      "step": 1947
+    },
+    {
+      "epoch": 2.1938028169014085,
+      "grad_norm": 0.23052546797566495,
+      "learning_rate": 1.4891395154553051e-05,
+      "loss": 0.2736,
+      "step": 1948
+    },
+    {
+      "epoch": 2.1949295774647886,
+      "grad_norm": 0.24459193123583572,
+      "learning_rate": 1.4870509607351713e-05,
+      "loss": 0.2789,
+      "step": 1949
+    },
+    {
+      "epoch": 2.196056338028169,
+      "grad_norm": 0.22373640891967844,
+      "learning_rate": 1.4849624060150378e-05,
+      "loss": 0.2708,
+      "step": 1950
+    },
+    {
+      "epoch": 2.1971830985915495,
+      "grad_norm": 0.23185887909498595,
+      "learning_rate": 1.482873851294904e-05,
+      "loss": 0.2872,
+      "step": 1951
+    },
+    {
+      "epoch": 2.1983098591549295,
+      "grad_norm": 0.23657985465235495,
+      "learning_rate": 1.4807852965747704e-05,
+      "loss": 0.2675,
+      "step": 1952
+    },
+    {
+      "epoch": 2.19943661971831,
+      "grad_norm": 0.22371982233885945,
+      "learning_rate": 1.4786967418546366e-05,
+      "loss": 0.2743,
+      "step": 1953
+    },
+    {
+      "epoch": 2.20056338028169,
+      "grad_norm": 0.21136209655985594,
+      "learning_rate": 1.4766081871345031e-05,
+      "loss": 0.2542,
+      "step": 1954
+    },
+    {
+      "epoch": 2.2016901408450704,
+      "grad_norm": 0.21509580106115853,
+      "learning_rate": 1.4745196324143693e-05,
+      "loss": 0.2659,
+      "step": 1955
+    },
+    {
+      "epoch": 2.202816901408451,
+      "grad_norm": 0.23707195283541888,
+      "learning_rate": 1.4724310776942358e-05,
+      "loss": 0.2646,
+      "step": 1956
+    },
+    {
+      "epoch": 2.203943661971831,
+      "grad_norm": 0.2061422131991942,
+      "learning_rate": 1.470342522974102e-05,
+      "loss": 0.2616,
+      "step": 1957
+    },
+    {
+      "epoch": 2.2050704225352113,
+      "grad_norm": 0.23775130186960383,
+      "learning_rate": 1.4682539682539683e-05,
+      "loss": 0.2899,
+      "step": 1958
+    },
+    {
+      "epoch": 2.2061971830985914,
+      "grad_norm": 0.24788983547937893,
+      "learning_rate": 1.4661654135338346e-05,
+      "loss": 0.276,
+      "step": 1959
+    },
+    {
+      "epoch": 2.207323943661972,
+      "grad_norm": 0.27365040649107353,
+      "learning_rate": 1.464076858813701e-05,
+      "loss": 0.2742,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2084507042253523,
+      "grad_norm": 0.21230372072638573,
+      "learning_rate": 1.4619883040935673e-05,
+      "loss": 0.2651,
+      "step": 1961
+    },
+    {
+      "epoch": 2.2095774647887323,
+      "grad_norm": 0.21542674747871746,
+      "learning_rate": 1.4598997493734337e-05,
+      "loss": 0.2746,
+      "step": 1962
+    },
+    {
+      "epoch": 2.2107042253521128,
+      "grad_norm": 0.23177548101881454,
+      "learning_rate": 1.4578111946533e-05,
+      "loss": 0.2641,
+      "step": 1963
+    },
+    {
+      "epoch": 2.2118309859154928,
+      "grad_norm": 0.22616947980308638,
+      "learning_rate": 1.4557226399331663e-05,
+      "loss": 0.2725,
+      "step": 1964
+    },
+    {
+      "epoch": 2.2129577464788732,
+      "grad_norm": 0.22554747009445886,
+      "learning_rate": 1.4536340852130325e-05,
+      "loss": 0.2939,
+      "step": 1965
+    },
+    {
+      "epoch": 2.2140845070422537,
+      "grad_norm": 0.22472896879143436,
+      "learning_rate": 1.451545530492899e-05,
+      "loss": 0.2681,
+      "step": 1966
+    },
+    {
+      "epoch": 2.2152112676056337,
+      "grad_norm": 0.20992257256681324,
+      "learning_rate": 1.4494569757727652e-05,
+      "loss": 0.2596,
+      "step": 1967
+    },
+    {
+      "epoch": 2.216338028169014,
+      "grad_norm": 0.22871327607085135,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 0.262,
+      "step": 1968
+    },
+    {
+      "epoch": 2.217464788732394,
+      "grad_norm": 0.23046973280343408,
+      "learning_rate": 1.4452798663324978e-05,
+      "loss": 0.2824,
+      "step": 1969
+    },
+    {
+      "epoch": 2.2185915492957746,
+      "grad_norm": 0.2096113182042097,
+      "learning_rate": 1.4431913116123644e-05,
+      "loss": 0.272,
+      "step": 1970
+    },
+    {
+      "epoch": 2.219718309859155,
+      "grad_norm": 0.20335565864164426,
+      "learning_rate": 1.4411027568922305e-05,
+      "loss": 0.2658,
+      "step": 1971
+    },
+    {
+      "epoch": 2.220845070422535,
+      "grad_norm": 0.21750740359226126,
+      "learning_rate": 1.439014202172097e-05,
+      "loss": 0.2744,
+      "step": 1972
+    },
+    {
+      "epoch": 2.2219718309859156,
+      "grad_norm": 0.22524066181597557,
+      "learning_rate": 1.4369256474519632e-05,
+      "loss": 0.2753,
+      "step": 1973
+    },
+    {
+      "epoch": 2.223098591549296,
+      "grad_norm": 0.23929892166160505,
+      "learning_rate": 1.4348370927318297e-05,
+      "loss": 0.28,
+      "step": 1974
+    },
+    {
+      "epoch": 2.224225352112676,
+      "grad_norm": 0.2248746927793624,
+      "learning_rate": 1.4327485380116959e-05,
+      "loss": 0.2734,
+      "step": 1975
+    },
+    {
+      "epoch": 2.2253521126760565,
+      "grad_norm": 0.20356256188239322,
+      "learning_rate": 1.4306599832915624e-05,
+      "loss": 0.2675,
+      "step": 1976
+    },
+    {
+      "epoch": 2.2264788732394365,
+      "grad_norm": 0.21154119418913328,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.2658,
+      "step": 1977
+    },
+    {
+      "epoch": 2.227605633802817,
+      "grad_norm": 0.21918481100532844,
+      "learning_rate": 1.426482873851295e-05,
+      "loss": 0.2819,
+      "step": 1978
+    },
+    {
+      "epoch": 2.2287323943661974,
+      "grad_norm": 0.21407891485923058,
+      "learning_rate": 1.4243943191311612e-05,
+      "loss": 0.2668,
+      "step": 1979
+    },
+    {
+      "epoch": 2.2298591549295774,
+      "grad_norm": 0.21161000632068389,
+      "learning_rate": 1.4223057644110277e-05,
+      "loss": 0.2736,
+      "step": 1980
+    },
+    {
+      "epoch": 2.230985915492958,
+      "grad_norm": 0.22280466855017053,
+      "learning_rate": 1.4202172096908939e-05,
+      "loss": 0.2734,
+      "step": 1981
+    },
+    {
+      "epoch": 2.232112676056338,
+      "grad_norm": 0.23417364326010645,
+      "learning_rate": 1.4181286549707604e-05,
+      "loss": 0.2717,
+      "step": 1982
+    },
+    {
+      "epoch": 2.2332394366197184,
+      "grad_norm": 0.21160121555283712,
+      "learning_rate": 1.4160401002506266e-05,
+      "loss": 0.2765,
+      "step": 1983
+    },
+    {
+      "epoch": 2.234366197183099,
+      "grad_norm": 0.23573792906579782,
+      "learning_rate": 1.413951545530493e-05,
+      "loss": 0.2843,
+      "step": 1984
+    },
+    {
+      "epoch": 2.235492957746479,
+      "grad_norm": 0.23981660549956635,
+      "learning_rate": 1.4118629908103592e-05,
+      "loss": 0.2916,
+      "step": 1985
+    },
+    {
+      "epoch": 2.2366197183098593,
+      "grad_norm": 0.2106101201254402,
+      "learning_rate": 1.4097744360902257e-05,
+      "loss": 0.2841,
+      "step": 1986
+    },
+    {
+      "epoch": 2.2377464788732393,
+      "grad_norm": 0.23639851155741806,
+      "learning_rate": 1.407685881370092e-05,
+      "loss": 0.277,
+      "step": 1987
+    },
+    {
+      "epoch": 2.2388732394366198,
+      "grad_norm": 0.22583083758980332,
+      "learning_rate": 1.4055973266499584e-05,
+      "loss": 0.2823,
+      "step": 1988
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2223738892554726,
+      "learning_rate": 1.4035087719298246e-05,
+      "loss": 0.271,
+      "step": 1989
+    },
+    {
+      "epoch": 2.2411267605633802,
+      "grad_norm": 0.22063232817325545,
+      "learning_rate": 1.4014202172096911e-05,
+      "loss": 0.2734,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2422535211267607,
+      "grad_norm": 0.2312948424666875,
+      "learning_rate": 1.3993316624895573e-05,
+      "loss": 0.2699,
+      "step": 1991
+    },
+    {
+      "epoch": 2.2433802816901407,
+      "grad_norm": 0.2338652743050092,
+      "learning_rate": 1.3972431077694236e-05,
+      "loss": 0.2778,
+      "step": 1992
+    },
+    {
+      "epoch": 2.244507042253521,
+      "grad_norm": 0.24131557161139464,
+      "learning_rate": 1.39515455304929e-05,
+      "loss": 0.2664,
+      "step": 1993
+    },
+    {
+      "epoch": 2.2456338028169016,
+      "grad_norm": 0.2489872832324652,
+      "learning_rate": 1.3930659983291563e-05,
+      "loss": 0.2807,
+      "step": 1994
+    },
+    {
+      "epoch": 2.2467605633802816,
+      "grad_norm": 0.22613350399229276,
+      "learning_rate": 1.3909774436090226e-05,
+      "loss": 0.2798,
+      "step": 1995
+    },
+    {
+      "epoch": 2.247887323943662,
+      "grad_norm": 0.22409833184125685,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.2815,
+      "step": 1996
+    },
+    {
+      "epoch": 2.249014084507042,
+      "grad_norm": 0.21011752202705167,
+      "learning_rate": 1.3868003341687553e-05,
+      "loss": 0.269,
+      "step": 1997
+    },
+    {
+      "epoch": 2.2501408450704226,
+      "grad_norm": 0.21734688044205402,
+      "learning_rate": 1.3847117794486216e-05,
+      "loss": 0.274,
+      "step": 1998
+    },
+    {
+      "epoch": 2.251267605633803,
+      "grad_norm": 0.2450339290069644,
+      "learning_rate": 1.3826232247284878e-05,
+      "loss": 0.2767,
+      "step": 1999
+    },
+    {
+      "epoch": 2.252394366197183,
+      "grad_norm": 0.21843275358835337,
+      "learning_rate": 1.3805346700083543e-05,
+      "loss": 0.2776,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2535211267605635,
+      "grad_norm": 0.22482809048283803,
+      "learning_rate": 1.3784461152882205e-05,
+      "loss": 0.2724,
+      "step": 2001
+    },
+    {
+      "epoch": 2.2546478873239435,
+      "grad_norm": 0.21721684317508147,
+      "learning_rate": 1.376357560568087e-05,
+      "loss": 0.279,
+      "step": 2002
+    },
+    {
+      "epoch": 2.255774647887324,
+      "grad_norm": 0.24301326583221047,
+      "learning_rate": 1.3742690058479531e-05,
+      "loss": 0.2588,
+      "step": 2003
+    },
+    {
+      "epoch": 2.2569014084507044,
+      "grad_norm": 0.19746584879494172,
+      "learning_rate": 1.3721804511278197e-05,
+      "loss": 0.253,
+      "step": 2004
+    },
+    {
+      "epoch": 2.2580281690140844,
+      "grad_norm": 0.2390184562770489,
+      "learning_rate": 1.3700918964076858e-05,
+      "loss": 0.2822,
+      "step": 2005
+    },
+    {
+      "epoch": 2.259154929577465,
+      "grad_norm": 0.23217244184241698,
+      "learning_rate": 1.3680033416875523e-05,
+      "loss": 0.29,
+      "step": 2006
+    },
+    {
+      "epoch": 2.260281690140845,
+      "grad_norm": 0.22506809172530984,
+      "learning_rate": 1.3659147869674185e-05,
+      "loss": 0.2676,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2614084507042254,
+      "grad_norm": 0.21517913805104338,
+      "learning_rate": 1.363826232247285e-05,
+      "loss": 0.2792,
+      "step": 2008
+    },
+    {
+      "epoch": 2.262535211267606,
+      "grad_norm": 0.24456781917421866,
+      "learning_rate": 1.3617376775271512e-05,
+      "loss": 0.2944,
+      "step": 2009
+    },
+    {
+      "epoch": 2.263661971830986,
+      "grad_norm": 0.23981555634152454,
+      "learning_rate": 1.3596491228070177e-05,
+      "loss": 0.2708,
+      "step": 2010
+    },
+    {
+      "epoch": 2.2647887323943663,
+      "grad_norm": 0.22061927832632905,
+      "learning_rate": 1.3575605680868838e-05,
+      "loss": 0.2522,
+      "step": 2011
+    },
+    {
+      "epoch": 2.2659154929577463,
+      "grad_norm": 0.2300492242646862,
+      "learning_rate": 1.3554720133667504e-05,
+      "loss": 0.29,
+      "step": 2012
+    },
+    {
+      "epoch": 2.2670422535211268,
+      "grad_norm": 0.22722745124981358,
+      "learning_rate": 1.3533834586466165e-05,
+      "loss": 0.2646,
+      "step": 2013
+    },
+    {
+      "epoch": 2.268169014084507,
+      "grad_norm": 0.22878725894793864,
+      "learning_rate": 1.351294903926483e-05,
+      "loss": 0.2623,
+      "step": 2014
+    },
+    {
+      "epoch": 2.2692957746478872,
+      "grad_norm": 0.2124223044313935,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.2906,
+      "step": 2015
+    },
+    {
+      "epoch": 2.2704225352112677,
+      "grad_norm": 0.2036252987427459,
+      "learning_rate": 1.3471177944862157e-05,
+      "loss": 0.2547,
+      "step": 2016
+    },
+    {
+      "epoch": 2.2715492957746477,
+      "grad_norm": 0.2399812429436266,
+      "learning_rate": 1.3450292397660819e-05,
+      "loss": 0.2819,
+      "step": 2017
+    },
+    {
+      "epoch": 2.272676056338028,
+      "grad_norm": 0.25316518682365097,
+      "learning_rate": 1.3429406850459484e-05,
+      "loss": 0.2706,
+      "step": 2018
+    },
+    {
+      "epoch": 2.2738028169014086,
+      "grad_norm": 0.2252764130337124,
+      "learning_rate": 1.3408521303258145e-05,
+      "loss": 0.273,
+      "step": 2019
+    },
+    {
+      "epoch": 2.2749295774647886,
+      "grad_norm": 0.20614835043806035,
+      "learning_rate": 1.338763575605681e-05,
+      "loss": 0.2687,
+      "step": 2020
+    },
+    {
+      "epoch": 2.276056338028169,
+      "grad_norm": 0.23543495966359457,
+      "learning_rate": 1.3366750208855472e-05,
+      "loss": 0.2784,
+      "step": 2021
+    },
+    {
+      "epoch": 2.277183098591549,
+      "grad_norm": 0.23084199615754333,
+      "learning_rate": 1.3345864661654137e-05,
+      "loss": 0.2521,
+      "step": 2022
+    },
+    {
+      "epoch": 2.2783098591549296,
+      "grad_norm": 0.23191754255578015,
+      "learning_rate": 1.3324979114452799e-05,
+      "loss": 0.2834,
+      "step": 2023
+    },
+    {
+      "epoch": 2.27943661971831,
+      "grad_norm": 0.21490002373296088,
+      "learning_rate": 1.3304093567251464e-05,
+      "loss": 0.264,
+      "step": 2024
+    },
+    {
+      "epoch": 2.28056338028169,
+      "grad_norm": 0.2285277364686723,
+      "learning_rate": 1.3283208020050126e-05,
+      "loss": 0.2961,
+      "step": 2025
+    },
+    {
+      "epoch": 2.2816901408450705,
+      "grad_norm": 0.22139617147513796,
+      "learning_rate": 1.3262322472848789e-05,
+      "loss": 0.2493,
+      "step": 2026
+    },
+    {
+      "epoch": 2.2828169014084505,
+      "grad_norm": 0.2266693203995399,
+      "learning_rate": 1.3241436925647452e-05,
+      "loss": 0.2513,
+      "step": 2027
+    },
+    {
+      "epoch": 2.283943661971831,
+      "grad_norm": 0.21507464866756967,
+      "learning_rate": 1.3220551378446116e-05,
+      "loss": 0.2727,
+      "step": 2028
+    },
+    {
+      "epoch": 2.2850704225352114,
+      "grad_norm": 0.21760306384693703,
+      "learning_rate": 1.319966583124478e-05,
+      "loss": 0.2837,
+      "step": 2029
+    },
+    {
+      "epoch": 2.2861971830985914,
+      "grad_norm": 0.22955926412886085,
+      "learning_rate": 1.3178780284043443e-05,
+      "loss": 0.2792,
+      "step": 2030
+    },
+    {
+      "epoch": 2.287323943661972,
+      "grad_norm": 0.22954629278050026,
+      "learning_rate": 1.3157894736842106e-05,
+      "loss": 0.2781,
+      "step": 2031
+    },
+    {
+      "epoch": 2.288450704225352,
+      "grad_norm": 0.21520773497636855,
+      "learning_rate": 1.313700918964077e-05,
+      "loss": 0.2681,
+      "step": 2032
+    },
+    {
+      "epoch": 2.2895774647887324,
+      "grad_norm": 0.21696543795200143,
+      "learning_rate": 1.3116123642439431e-05,
+      "loss": 0.2595,
+      "step": 2033
+    },
+    {
+      "epoch": 2.290704225352113,
+      "grad_norm": 0.22952681284858015,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 0.282,
+      "step": 2034
+    },
+    {
+      "epoch": 2.291830985915493,
+      "grad_norm": 0.33170692132136104,
+      "learning_rate": 1.3074352548036758e-05,
+      "loss": 0.2919,
+      "step": 2035
+    },
+    {
+      "epoch": 2.2929577464788733,
+      "grad_norm": 0.2533477575243943,
+      "learning_rate": 1.3053467000835423e-05,
+      "loss": 0.2962,
+      "step": 2036
+    },
+    {
+      "epoch": 2.2940845070422533,
+      "grad_norm": 0.2168408112742443,
+      "learning_rate": 1.3032581453634085e-05,
+      "loss": 0.275,
+      "step": 2037
+    },
+    {
+      "epoch": 2.2952112676056338,
+      "grad_norm": 0.21500736585640562,
+      "learning_rate": 1.301169590643275e-05,
+      "loss": 0.2686,
+      "step": 2038
+    },
+    {
+      "epoch": 2.2963380281690142,
+      "grad_norm": 0.2676815326816881,
+      "learning_rate": 1.2990810359231411e-05,
+      "loss": 0.2796,
+      "step": 2039
+    },
+    {
+      "epoch": 2.2974647887323942,
+      "grad_norm": 0.2263627947111304,
+      "learning_rate": 1.2969924812030076e-05,
+      "loss": 0.2656,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2985915492957747,
+      "grad_norm": 0.22688933416191495,
+      "learning_rate": 1.2949039264828738e-05,
+      "loss": 0.277,
+      "step": 2041
+    },
+    {
+      "epoch": 2.2997183098591547,
+      "grad_norm": 0.2257086440618408,
+      "learning_rate": 1.2928153717627403e-05,
+      "loss": 0.2674,
+      "step": 2042
+    },
+    {
+      "epoch": 2.300845070422535,
+      "grad_norm": 0.2416933950813198,
+      "learning_rate": 1.2907268170426065e-05,
+      "loss": 0.2687,
+      "step": 2043
+    },
+    {
+      "epoch": 2.3019718309859156,
+      "grad_norm": 0.2323204387295286,
+      "learning_rate": 1.288638262322473e-05,
+      "loss": 0.2774,
+      "step": 2044
+    },
+    {
+      "epoch": 2.3030985915492956,
+      "grad_norm": 0.21946992538397853,
+      "learning_rate": 1.2865497076023392e-05,
+      "loss": 0.264,
+      "step": 2045
+    },
+    {
+      "epoch": 2.304225352112676,
+      "grad_norm": 0.24515276245328027,
+      "learning_rate": 1.2844611528822057e-05,
+      "loss": 0.2667,
+      "step": 2046
+    },
+    {
+      "epoch": 2.305352112676056,
+      "grad_norm": 0.2464566315493969,
+      "learning_rate": 1.2823725981620718e-05,
+      "loss": 0.2721,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3064788732394366,
+      "grad_norm": 0.22334469262497983,
+      "learning_rate": 1.2802840434419383e-05,
+      "loss": 0.2812,
+      "step": 2048
+    },
+    {
+      "epoch": 2.307605633802817,
+      "grad_norm": 0.25873138298084397,
+      "learning_rate": 1.2781954887218045e-05,
+      "loss": 0.2917,
+      "step": 2049
+    },
+    {
+      "epoch": 2.308732394366197,
+      "grad_norm": 0.2778202804249837,
+      "learning_rate": 1.276106934001671e-05,
+      "loss": 0.273,
+      "step": 2050
+    },
+    {
+      "epoch": 2.3098591549295775,
+      "grad_norm": 0.21671557928806906,
+      "learning_rate": 1.2740183792815372e-05,
+      "loss": 0.2621,
+      "step": 2051
+    },
+    {
+      "epoch": 2.3109859154929575,
+      "grad_norm": 0.22789737210449681,
+      "learning_rate": 1.2719298245614037e-05,
+      "loss": 0.2576,
+      "step": 2052
+    },
+    {
+      "epoch": 2.312112676056338,
+      "grad_norm": 0.26173819893725,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.2638,
+      "step": 2053
+    },
+    {
+      "epoch": 2.3132394366197184,
+      "grad_norm": 0.24649801061804577,
+      "learning_rate": 1.2677527151211364e-05,
+      "loss": 0.2784,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3143661971830984,
+      "grad_norm": 0.23514114032370073,
+      "learning_rate": 1.2656641604010025e-05,
+      "loss": 0.2651,
+      "step": 2055
+    },
+    {
+      "epoch": 2.315492957746479,
+      "grad_norm": 0.2466224146716087,
+      "learning_rate": 1.263575605680869e-05,
+      "loss": 0.2917,
+      "step": 2056
+    },
+    {
+      "epoch": 2.316619718309859,
+      "grad_norm": 0.2872020468405315,
+      "learning_rate": 1.2614870509607352e-05,
+      "loss": 0.2536,
+      "step": 2057
+    },
+    {
+      "epoch": 2.3177464788732394,
+      "grad_norm": 0.21720874528836792,
+      "learning_rate": 1.2593984962406017e-05,
+      "loss": 0.28,
+      "step": 2058
+    },
+    {
+      "epoch": 2.31887323943662,
+      "grad_norm": 0.20741119190594928,
+      "learning_rate": 1.2573099415204679e-05,
+      "loss": 0.2718,
+      "step": 2059
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2560372734108066,
+      "learning_rate": 1.2552213868003342e-05,
+      "loss": 0.2789,
+      "step": 2060
+    },
+    {
+      "epoch": 2.3211267605633803,
+      "grad_norm": 0.24952747418775886,
+      "learning_rate": 1.2531328320802006e-05,
+      "loss": 0.2825,
+      "step": 2061
+    },
+    {
+      "epoch": 2.3222535211267608,
+      "grad_norm": 0.22738294194395234,
+      "learning_rate": 1.2510442773600669e-05,
+      "loss": 0.2937,
+      "step": 2062
+    },
+    {
+      "epoch": 2.3233802816901408,
+      "grad_norm": 0.22446187495202183,
+      "learning_rate": 1.2489557226399332e-05,
+      "loss": 0.2713,
+      "step": 2063
+    },
+    {
+      "epoch": 2.3245070422535212,
+      "grad_norm": 0.24144984465835664,
+      "learning_rate": 1.2468671679197996e-05,
+      "loss": 0.2756,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3256338028169012,
+      "grad_norm": 0.23689819143768417,
+      "learning_rate": 1.2447786131996659e-05,
+      "loss": 0.2546,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3267605633802817,
+      "grad_norm": 0.2611020677993963,
+      "learning_rate": 1.242690058479532e-05,
+      "loss": 0.2809,
+      "step": 2066
+    },
+    {
+      "epoch": 2.327887323943662,
+      "grad_norm": 0.2339885651831768,
+      "learning_rate": 1.2406015037593984e-05,
+      "loss": 0.2801,
+      "step": 2067
+    },
+    {
+      "epoch": 2.329014084507042,
+      "grad_norm": 0.2334254544471349,
+      "learning_rate": 1.2385129490392647e-05,
+      "loss": 0.2725,
+      "step": 2068
+    },
+    {
+      "epoch": 2.3301408450704226,
+      "grad_norm": 0.2246880889088873,
+      "learning_rate": 1.236424394319131e-05,
+      "loss": 0.2664,
+      "step": 2069
+    },
+    {
+      "epoch": 2.3312676056338026,
+      "grad_norm": 0.2507553837232819,
+      "learning_rate": 1.2343358395989974e-05,
+      "loss": 0.2788,
+      "step": 2070
+    },
+    {
+      "epoch": 2.332394366197183,
+      "grad_norm": 0.21634555471769967,
+      "learning_rate": 1.2322472848788638e-05,
+      "loss": 0.2635,
+      "step": 2071
+    },
+    {
+      "epoch": 2.3335211267605636,
+      "grad_norm": 0.2331492863630478,
+      "learning_rate": 1.2301587301587301e-05,
+      "loss": 0.2717,
+      "step": 2072
+    },
+    {
+      "epoch": 2.3346478873239436,
+      "grad_norm": 0.23764686043090247,
+      "learning_rate": 1.2280701754385964e-05,
+      "loss": 0.2719,
+      "step": 2073
+    },
+    {
+      "epoch": 2.335774647887324,
+      "grad_norm": 0.20696704110699093,
+      "learning_rate": 1.2259816207184628e-05,
+      "loss": 0.2623,
+      "step": 2074
+    },
+    {
+      "epoch": 2.336901408450704,
+      "grad_norm": 0.23715745041246034,
+      "learning_rate": 1.2238930659983291e-05,
+      "loss": 0.2804,
+      "step": 2075
+    },
+    {
+      "epoch": 2.3380281690140845,
+      "grad_norm": 0.23109068722265308,
+      "learning_rate": 1.2218045112781954e-05,
+      "loss": 0.2709,
+      "step": 2076
+    },
+    {
+      "epoch": 2.339154929577465,
+      "grad_norm": 0.1991020046023143,
+      "learning_rate": 1.2197159565580618e-05,
+      "loss": 0.2616,
+      "step": 2077
+    },
+    {
+      "epoch": 2.340281690140845,
+      "grad_norm": 0.20836991373278604,
+      "learning_rate": 1.2176274018379281e-05,
+      "loss": 0.2803,
+      "step": 2078
+    },
+    {
+      "epoch": 2.3414084507042254,
+      "grad_norm": 0.23228249419993582,
+      "learning_rate": 1.2155388471177945e-05,
+      "loss": 0.2775,
+      "step": 2079
+    },
+    {
+      "epoch": 2.3425352112676054,
+      "grad_norm": 0.23298791905416905,
+      "learning_rate": 1.2134502923976608e-05,
+      "loss": 0.2735,
+      "step": 2080
+    },
+    {
+      "epoch": 2.343661971830986,
+      "grad_norm": 0.20806123139516708,
+      "learning_rate": 1.2113617376775271e-05,
+      "loss": 0.2781,
+      "step": 2081
+    },
+    {
+      "epoch": 2.3447887323943664,
+      "grad_norm": 0.22064451961073028,
+      "learning_rate": 1.2092731829573935e-05,
+      "loss": 0.2734,
+      "step": 2082
+    },
+    {
+      "epoch": 2.3459154929577464,
+      "grad_norm": 0.21649854203207658,
+      "learning_rate": 1.2071846282372598e-05,
+      "loss": 0.2583,
+      "step": 2083
+    },
+    {
+      "epoch": 2.347042253521127,
+      "grad_norm": 0.21476663436618942,
+      "learning_rate": 1.2050960735171261e-05,
+      "loss": 0.2712,
+      "step": 2084
+    },
+    {
+      "epoch": 2.3481690140845073,
+      "grad_norm": 0.1982230142966004,
+      "learning_rate": 1.2030075187969925e-05,
+      "loss": 0.256,
+      "step": 2085
+    },
+    {
+      "epoch": 2.3492957746478873,
+      "grad_norm": 0.21709399899243265,
+      "learning_rate": 1.2009189640768588e-05,
+      "loss": 0.2653,
+      "step": 2086
+    },
+    {
+      "epoch": 2.3504225352112678,
+      "grad_norm": 0.23444500781279742,
+      "learning_rate": 1.1988304093567252e-05,
+      "loss": 0.283,
+      "step": 2087
+    },
+    {
+      "epoch": 2.3515492957746478,
+      "grad_norm": 0.22818369005780376,
+      "learning_rate": 1.1967418546365915e-05,
+      "loss": 0.2646,
+      "step": 2088
+    },
+    {
+      "epoch": 2.3526760563380282,
+      "grad_norm": 0.2028931812809309,
+      "learning_rate": 1.1946532999164578e-05,
+      "loss": 0.2748,
+      "step": 2089
+    },
+    {
+      "epoch": 2.3538028169014087,
+      "grad_norm": 0.2345211570550957,
+      "learning_rate": 1.1925647451963242e-05,
+      "loss": 0.2745,
+      "step": 2090
+    },
+    {
+      "epoch": 2.3549295774647887,
+      "grad_norm": 0.2753030384992383,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.2729,
+      "step": 2091
+    },
+    {
+      "epoch": 2.356056338028169,
+      "grad_norm": 0.21352984178296622,
+      "learning_rate": 1.1883876357560568e-05,
+      "loss": 0.2608,
+      "step": 2092
+    },
+    {
+      "epoch": 2.357183098591549,
+      "grad_norm": 0.2728726782141452,
+      "learning_rate": 1.1862990810359232e-05,
+      "loss": 0.2793,
+      "step": 2093
+    },
+    {
+      "epoch": 2.3583098591549296,
+      "grad_norm": 0.2410185342470983,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 0.2861,
+      "step": 2094
+    },
+    {
+      "epoch": 2.35943661971831,
+      "grad_norm": 0.24989805449908808,
+      "learning_rate": 1.1821219715956559e-05,
+      "loss": 0.2906,
+      "step": 2095
+    },
+    {
+      "epoch": 2.36056338028169,
+      "grad_norm": 0.23661941159683683,
+      "learning_rate": 1.1800334168755222e-05,
+      "loss": 0.2846,
+      "step": 2096
+    },
+    {
+      "epoch": 2.3616901408450706,
+      "grad_norm": 0.24911928401703032,
+      "learning_rate": 1.1779448621553885e-05,
+      "loss": 0.2842,
+      "step": 2097
+    },
+    {
+      "epoch": 2.3628169014084506,
+      "grad_norm": 0.2188034204109028,
+      "learning_rate": 1.1758563074352549e-05,
+      "loss": 0.2619,
+      "step": 2098
+    },
+    {
+      "epoch": 2.363943661971831,
+      "grad_norm": 0.23519778699583993,
+      "learning_rate": 1.1737677527151212e-05,
+      "loss": 0.2528,
+      "step": 2099
+    },
+    {
+      "epoch": 2.3650704225352115,
+      "grad_norm": 0.24858670761142815,
+      "learning_rate": 1.1716791979949874e-05,
+      "loss": 0.2764,
+      "step": 2100
+    },
+    {
+      "epoch": 2.3661971830985915,
+      "grad_norm": 0.2157432927871551,
+      "learning_rate": 1.1695906432748537e-05,
+      "loss": 0.2703,
+      "step": 2101
+    },
+    {
+      "epoch": 2.367323943661972,
+      "grad_norm": 0.20035197021874354,
+      "learning_rate": 1.16750208855472e-05,
+      "loss": 0.2635,
+      "step": 2102
+    },
+    {
+      "epoch": 2.368450704225352,
+      "grad_norm": 0.25971878714562346,
+      "learning_rate": 1.1654135338345864e-05,
+      "loss": 0.2766,
+      "step": 2103
+    },
+    {
+      "epoch": 2.3695774647887324,
+      "grad_norm": 0.22235337146940867,
+      "learning_rate": 1.1633249791144527e-05,
+      "loss": 0.2723,
+      "step": 2104
+    },
+    {
+      "epoch": 2.370704225352113,
+      "grad_norm": 0.19367345176317313,
+      "learning_rate": 1.161236424394319e-05,
+      "loss": 0.2672,
+      "step": 2105
+    },
+    {
+      "epoch": 2.371830985915493,
+      "grad_norm": 0.22489503194261767,
+      "learning_rate": 1.1591478696741854e-05,
+      "loss": 0.2705,
+      "step": 2106
+    },
+    {
+      "epoch": 2.3729577464788734,
+      "grad_norm": 0.22838289804635875,
+      "learning_rate": 1.1570593149540517e-05,
+      "loss": 0.2694,
+      "step": 2107
+    },
+    {
+      "epoch": 2.3740845070422534,
+      "grad_norm": 0.20986582906030665,
+      "learning_rate": 1.154970760233918e-05,
+      "loss": 0.2613,
+      "step": 2108
+    },
+    {
+      "epoch": 2.375211267605634,
+      "grad_norm": 0.2223157151237609,
+      "learning_rate": 1.1528822055137844e-05,
+      "loss": 0.2826,
+      "step": 2109
+    },
+    {
+      "epoch": 2.3763380281690143,
+      "grad_norm": 0.22977691867501657,
+      "learning_rate": 1.1507936507936508e-05,
+      "loss": 0.2789,
+      "step": 2110
+    },
+    {
+      "epoch": 2.3774647887323943,
+      "grad_norm": 0.21110855292816205,
+      "learning_rate": 1.1487050960735171e-05,
+      "loss": 0.2741,
+      "step": 2111
+    },
+    {
+      "epoch": 2.3785915492957748,
+      "grad_norm": 0.21244436892484309,
+      "learning_rate": 1.1466165413533834e-05,
+      "loss": 0.2709,
+      "step": 2112
+    },
+    {
+      "epoch": 2.379718309859155,
+      "grad_norm": 0.2175330654582257,
+      "learning_rate": 1.1445279866332498e-05,
+      "loss": 0.2622,
+      "step": 2113
+    },
+    {
+      "epoch": 2.3808450704225352,
+      "grad_norm": 0.23308321264773776,
+      "learning_rate": 1.1424394319131161e-05,
+      "loss": 0.2808,
+      "step": 2114
+    },
+    {
+      "epoch": 2.3819718309859157,
+      "grad_norm": 0.21031989758189668,
+      "learning_rate": 1.1403508771929824e-05,
+      "loss": 0.2802,
+      "step": 2115
+    },
+    {
+      "epoch": 2.3830985915492957,
+      "grad_norm": 0.20745254829798823,
+      "learning_rate": 1.1382623224728488e-05,
+      "loss": 0.2582,
+      "step": 2116
+    },
+    {
+      "epoch": 2.384225352112676,
+      "grad_norm": 0.2332964916151851,
+      "learning_rate": 1.1361737677527151e-05,
+      "loss": 0.2886,
+      "step": 2117
+    },
+    {
+      "epoch": 2.385352112676056,
+      "grad_norm": 0.2279416792446994,
+      "learning_rate": 1.1340852130325815e-05,
+      "loss": 0.2866,
+      "step": 2118
+    },
+    {
+      "epoch": 2.3864788732394366,
+      "grad_norm": 0.21711758088930125,
+      "learning_rate": 1.1319966583124478e-05,
+      "loss": 0.2805,
+      "step": 2119
+    },
+    {
+      "epoch": 2.387605633802817,
+      "grad_norm": 0.22213392090537115,
+      "learning_rate": 1.1299081035923141e-05,
+      "loss": 0.2561,
+      "step": 2120
+    },
+    {
+      "epoch": 2.388732394366197,
+      "grad_norm": 0.22722511191755496,
+      "learning_rate": 1.1278195488721805e-05,
+      "loss": 0.2651,
+      "step": 2121
+    },
+    {
+      "epoch": 2.3898591549295776,
+      "grad_norm": 0.24840655590846175,
+      "learning_rate": 1.1257309941520468e-05,
+      "loss": 0.2853,
+      "step": 2122
+    },
+    {
+      "epoch": 2.3909859154929576,
+      "grad_norm": 0.20463193123425508,
+      "learning_rate": 1.1236424394319131e-05,
+      "loss": 0.2638,
+      "step": 2123
+    },
+    {
+      "epoch": 2.392112676056338,
+      "grad_norm": 0.22768377246676877,
+      "learning_rate": 1.1215538847117795e-05,
+      "loss": 0.2814,
+      "step": 2124
+    },
+    {
+      "epoch": 2.3932394366197185,
+      "grad_norm": 0.24848832166520118,
+      "learning_rate": 1.1194653299916458e-05,
+      "loss": 0.2725,
+      "step": 2125
+    },
+    {
+      "epoch": 2.3943661971830985,
+      "grad_norm": 0.21778817501473433,
+      "learning_rate": 1.1173767752715122e-05,
+      "loss": 0.2763,
+      "step": 2126
+    },
+    {
+      "epoch": 2.395492957746479,
+      "grad_norm": 0.20652062321780992,
+      "learning_rate": 1.1152882205513785e-05,
+      "loss": 0.2626,
+      "step": 2127
+    },
+    {
+      "epoch": 2.396619718309859,
+      "grad_norm": 0.24174624714657003,
+      "learning_rate": 1.1131996658312448e-05,
+      "loss": 0.269,
+      "step": 2128
+    },
+    {
+      "epoch": 2.3977464788732394,
+      "grad_norm": 0.236893442627161,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2774,
+      "step": 2129
+    },
+    {
+      "epoch": 2.39887323943662,
+      "grad_norm": 0.21898107027200428,
+      "learning_rate": 1.1090225563909775e-05,
+      "loss": 0.2803,
+      "step": 2130
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.22808988943103967,
+      "learning_rate": 1.1069340016708438e-05,
+      "loss": 0.2626,
+      "step": 2131
+    },
+    {
+      "epoch": 2.4011267605633804,
+      "grad_norm": 0.24093717514555563,
+      "learning_rate": 1.1048454469507102e-05,
+      "loss": 0.2762,
+      "step": 2132
+    },
+    {
+      "epoch": 2.4022535211267604,
+      "grad_norm": 0.240001552288654,
+      "learning_rate": 1.1027568922305765e-05,
+      "loss": 0.2688,
+      "step": 2133
+    },
+    {
+      "epoch": 2.403380281690141,
+      "grad_norm": 0.20401632119073637,
+      "learning_rate": 1.1006683375104427e-05,
+      "loss": 0.2685,
+      "step": 2134
+    },
+    {
+      "epoch": 2.4045070422535213,
+      "grad_norm": 0.23215631653494903,
+      "learning_rate": 1.098579782790309e-05,
+      "loss": 0.2739,
+      "step": 2135
+    },
+    {
+      "epoch": 2.4056338028169013,
+      "grad_norm": 0.2438746235505259,
+      "learning_rate": 1.0964912280701754e-05,
+      "loss": 0.2798,
+      "step": 2136
+    },
+    {
+      "epoch": 2.4067605633802818,
+      "grad_norm": 0.2185948791096995,
+      "learning_rate": 1.0944026733500417e-05,
+      "loss": 0.2763,
+      "step": 2137
+    },
+    {
+      "epoch": 2.407887323943662,
+      "grad_norm": 0.22927415544945678,
+      "learning_rate": 1.092314118629908e-05,
+      "loss": 0.2741,
+      "step": 2138
+    },
+    {
+      "epoch": 2.4090140845070422,
+      "grad_norm": 0.2352880631874138,
+      "learning_rate": 1.0902255639097744e-05,
+      "loss": 0.2836,
+      "step": 2139
+    },
+    {
+      "epoch": 2.4101408450704227,
+      "grad_norm": 0.22718997230465515,
+      "learning_rate": 1.0881370091896407e-05,
+      "loss": 0.2773,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4112676056338027,
+      "grad_norm": 0.22040466124629957,
+      "learning_rate": 1.086048454469507e-05,
+      "loss": 0.2678,
+      "step": 2141
+    },
+    {
+      "epoch": 2.412394366197183,
+      "grad_norm": 0.23551257603799414,
+      "learning_rate": 1.0839598997493734e-05,
+      "loss": 0.269,
+      "step": 2142
+    },
+    {
+      "epoch": 2.413521126760563,
+      "grad_norm": 0.21954462480504305,
+      "learning_rate": 1.0818713450292397e-05,
+      "loss": 0.2731,
+      "step": 2143
+    },
+    {
+      "epoch": 2.4146478873239436,
+      "grad_norm": 0.22592886171330676,
+      "learning_rate": 1.079782790309106e-05,
+      "loss": 0.296,
+      "step": 2144
+    },
+    {
+      "epoch": 2.415774647887324,
+      "grad_norm": 0.2240484943300781,
+      "learning_rate": 1.0776942355889724e-05,
+      "loss": 0.2719,
+      "step": 2145
+    },
+    {
+      "epoch": 2.416901408450704,
+      "grad_norm": 0.22363256604144707,
+      "learning_rate": 1.0756056808688387e-05,
+      "loss": 0.2679,
+      "step": 2146
+    },
+    {
+      "epoch": 2.4180281690140846,
+      "grad_norm": 0.2154008784772404,
+      "learning_rate": 1.073517126148705e-05,
+      "loss": 0.2844,
+      "step": 2147
+    },
+    {
+      "epoch": 2.4191549295774646,
+      "grad_norm": 0.21320566200159038,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 0.2586,
+      "step": 2148
+    },
+    {
+      "epoch": 2.420281690140845,
+      "grad_norm": 0.20909928611938552,
+      "learning_rate": 1.0693400167084377e-05,
+      "loss": 0.2606,
+      "step": 2149
+    },
+    {
+      "epoch": 2.4214084507042255,
+      "grad_norm": 0.20981361725139566,
+      "learning_rate": 1.067251461988304e-05,
+      "loss": 0.2705,
+      "step": 2150
+    },
+    {
+      "epoch": 2.4225352112676055,
+      "grad_norm": 0.21510118347835527,
+      "learning_rate": 1.0651629072681704e-05,
+      "loss": 0.2696,
+      "step": 2151
+    },
+    {
+      "epoch": 2.423661971830986,
+      "grad_norm": 0.21114872083025085,
+      "learning_rate": 1.0630743525480368e-05,
+      "loss": 0.2695,
+      "step": 2152
+    },
+    {
+      "epoch": 2.424788732394366,
+      "grad_norm": 0.20265014743832913,
+      "learning_rate": 1.0609857978279031e-05,
+      "loss": 0.2654,
+      "step": 2153
+    },
+    {
+      "epoch": 2.4259154929577464,
+      "grad_norm": 0.21350093970564465,
+      "learning_rate": 1.0588972431077694e-05,
+      "loss": 0.2795,
+      "step": 2154
+    },
+    {
+      "epoch": 2.427042253521127,
+      "grad_norm": 0.2118394159905768,
+      "learning_rate": 1.0568086883876358e-05,
+      "loss": 0.2642,
+      "step": 2155
+    },
+    {
+      "epoch": 2.428169014084507,
+      "grad_norm": 0.23040409136612594,
+      "learning_rate": 1.0547201336675021e-05,
+      "loss": 0.2929,
+      "step": 2156
+    },
+    {
+      "epoch": 2.4292957746478874,
+      "grad_norm": 0.2276700680735358,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.258,
+      "step": 2157
+    },
+    {
+      "epoch": 2.4304225352112674,
+      "grad_norm": 0.23948540092079731,
+      "learning_rate": 1.0505430242272348e-05,
+      "loss": 0.2705,
+      "step": 2158
+    },
+    {
+      "epoch": 2.431549295774648,
+      "grad_norm": 0.22322257893047265,
+      "learning_rate": 1.0484544695071011e-05,
+      "loss": 0.273,
+      "step": 2159
+    },
+    {
+      "epoch": 2.4326760563380283,
+      "grad_norm": 0.2698826968184876,
+      "learning_rate": 1.0463659147869675e-05,
+      "loss": 0.2743,
+      "step": 2160
+    },
+    {
+      "epoch": 2.4338028169014083,
+      "grad_norm": 0.21669959538878936,
+      "learning_rate": 1.0442773600668338e-05,
+      "loss": 0.2655,
+      "step": 2161
+    },
+    {
+      "epoch": 2.4349295774647888,
+      "grad_norm": 0.2409956078704885,
+      "learning_rate": 1.0421888053467001e-05,
+      "loss": 0.2832,
+      "step": 2162
+    },
+    {
+      "epoch": 2.436056338028169,
+      "grad_norm": 0.20476452459540187,
+      "learning_rate": 1.0401002506265665e-05,
+      "loss": 0.2809,
+      "step": 2163
+    },
+    {
+      "epoch": 2.4371830985915492,
+      "grad_norm": 0.20986579303419386,
+      "learning_rate": 1.0380116959064328e-05,
+      "loss": 0.2789,
+      "step": 2164
+    },
+    {
+      "epoch": 2.4383098591549297,
+      "grad_norm": 0.2352840245085601,
+      "learning_rate": 1.0359231411862991e-05,
+      "loss": 0.2816,
+      "step": 2165
+    },
+    {
+      "epoch": 2.4394366197183097,
+      "grad_norm": 0.22355582613204722,
+      "learning_rate": 1.0338345864661655e-05,
+      "loss": 0.2738,
+      "step": 2166
+    },
+    {
+      "epoch": 2.44056338028169,
+      "grad_norm": 0.21389291437331132,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.2727,
+      "step": 2167
+    },
+    {
+      "epoch": 2.44169014084507,
+      "grad_norm": 0.23646272987408282,
+      "learning_rate": 1.029657477025898e-05,
+      "loss": 0.271,
+      "step": 2168
+    },
+    {
+      "epoch": 2.4428169014084506,
+      "grad_norm": 0.23027802072594214,
+      "learning_rate": 1.0275689223057643e-05,
+      "loss": 0.2896,
+      "step": 2169
+    },
+    {
+      "epoch": 2.443943661971831,
+      "grad_norm": 0.2274161999681408,
+      "learning_rate": 1.0254803675856307e-05,
+      "loss": 0.2811,
+      "step": 2170
+    },
+    {
+      "epoch": 2.445070422535211,
+      "grad_norm": 0.22538041115209936,
+      "learning_rate": 1.023391812865497e-05,
+      "loss": 0.277,
+      "step": 2171
+    },
+    {
+      "epoch": 2.4461971830985916,
+      "grad_norm": 0.24038478413990452,
+      "learning_rate": 1.0213032581453633e-05,
+      "loss": 0.2848,
+      "step": 2172
+    },
+    {
+      "epoch": 2.4473239436619716,
+      "grad_norm": 0.244583573845353,
+      "learning_rate": 1.0192147034252297e-05,
+      "loss": 0.2825,
+      "step": 2173
+    },
+    {
+      "epoch": 2.448450704225352,
+      "grad_norm": 0.2093797766595799,
+      "learning_rate": 1.017126148705096e-05,
+      "loss": 0.248,
+      "step": 2174
+    },
+    {
+      "epoch": 2.4495774647887325,
+      "grad_norm": 0.23293668786136626,
+      "learning_rate": 1.0150375939849624e-05,
+      "loss": 0.282,
+      "step": 2175
+    },
+    {
+      "epoch": 2.4507042253521125,
+      "grad_norm": 0.22928470807328025,
+      "learning_rate": 1.0129490392648287e-05,
+      "loss": 0.2703,
+      "step": 2176
+    },
+    {
+      "epoch": 2.451830985915493,
+      "grad_norm": 0.23828534953004338,
+      "learning_rate": 1.010860484544695e-05,
+      "loss": 0.261,
+      "step": 2177
+    },
+    {
+      "epoch": 2.452957746478873,
+      "grad_norm": 0.22773354216922517,
+      "learning_rate": 1.0087719298245614e-05,
+      "loss": 0.2791,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4540845070422534,
+      "grad_norm": 0.22213705135610468,
+      "learning_rate": 1.0066833751044277e-05,
+      "loss": 0.2748,
+      "step": 2179
+    },
+    {
+      "epoch": 2.455211267605634,
+      "grad_norm": 0.23231356179846382,
+      "learning_rate": 1.004594820384294e-05,
+      "loss": 0.2705,
+      "step": 2180
+    },
+    {
+      "epoch": 2.456338028169014,
+      "grad_norm": 0.22561566979144523,
+      "learning_rate": 1.0025062656641604e-05,
+      "loss": 0.2774,
+      "step": 2181
+    },
+    {
+      "epoch": 2.4574647887323944,
+      "grad_norm": 0.22022567511058663,
+      "learning_rate": 1.0004177109440267e-05,
+      "loss": 0.2406,
+      "step": 2182
+    },
+    {
+      "epoch": 2.458591549295775,
+      "grad_norm": 0.20961456153746924,
+      "learning_rate": 9.98329156223893e-06,
+      "loss": 0.2658,
+      "step": 2183
+    },
+    {
+      "epoch": 2.459718309859155,
+      "grad_norm": 0.2059883661597407,
+      "learning_rate": 9.962406015037594e-06,
+      "loss": 0.2775,
+      "step": 2184
+    },
+    {
+      "epoch": 2.4608450704225353,
+      "grad_norm": 0.256771919518465,
+      "learning_rate": 9.941520467836257e-06,
+      "loss": 0.2743,
+      "step": 2185
+    },
+    {
+      "epoch": 2.4619718309859153,
+      "grad_norm": 0.23952040840749259,
+      "learning_rate": 9.92063492063492e-06,
+      "loss": 0.2741,
+      "step": 2186
+    },
+    {
+      "epoch": 2.463098591549296,
+      "grad_norm": 0.21297152510598305,
+      "learning_rate": 9.899749373433584e-06,
+      "loss": 0.2775,
+      "step": 2187
+    },
+    {
+      "epoch": 2.4642253521126762,
+      "grad_norm": 0.2066578541585253,
+      "learning_rate": 9.878863826232247e-06,
+      "loss": 0.2639,
+      "step": 2188
+    },
+    {
+      "epoch": 2.4653521126760563,
+      "grad_norm": 0.2137411082320371,
+      "learning_rate": 9.85797827903091e-06,
+      "loss": 0.2779,
+      "step": 2189
+    },
+    {
+      "epoch": 2.4664788732394367,
+      "grad_norm": 0.23503564988189526,
+      "learning_rate": 9.837092731829574e-06,
+      "loss": 0.2734,
+      "step": 2190
+    },
+    {
+      "epoch": 2.4676056338028167,
+      "grad_norm": 0.22828241783204792,
+      "learning_rate": 9.816207184628238e-06,
+      "loss": 0.2559,
+      "step": 2191
+    },
+    {
+      "epoch": 2.468732394366197,
+      "grad_norm": 0.23122651593415716,
+      "learning_rate": 9.795321637426901e-06,
+      "loss": 0.2862,
+      "step": 2192
+    },
+    {
+      "epoch": 2.4698591549295776,
+      "grad_norm": 0.2323362117389597,
+      "learning_rate": 9.774436090225564e-06,
+      "loss": 0.291,
+      "step": 2193
+    },
+    {
+      "epoch": 2.4709859154929577,
+      "grad_norm": 0.20386352123726093,
+      "learning_rate": 9.753550543024228e-06,
+      "loss": 0.266,
+      "step": 2194
+    },
+    {
+      "epoch": 2.472112676056338,
+      "grad_norm": 0.23765114384248068,
+      "learning_rate": 9.732664995822891e-06,
+      "loss": 0.265,
+      "step": 2195
+    },
+    {
+      "epoch": 2.473239436619718,
+      "grad_norm": 0.22209086256665098,
+      "learning_rate": 9.711779448621554e-06,
+      "loss": 0.2649,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4743661971830986,
+      "grad_norm": 0.22644727329815742,
+      "learning_rate": 9.690893901420218e-06,
+      "loss": 0.2635,
+      "step": 2197
+    },
+    {
+      "epoch": 2.475492957746479,
+      "grad_norm": 0.22367553789132952,
+      "learning_rate": 9.670008354218881e-06,
+      "loss": 0.2699,
+      "step": 2198
+    },
+    {
+      "epoch": 2.476619718309859,
+      "grad_norm": 0.22888798104554653,
+      "learning_rate": 9.649122807017545e-06,
+      "loss": 0.2705,
+      "step": 2199
+    },
+    {
+      "epoch": 2.4777464788732395,
+      "grad_norm": 0.22848461387767924,
+      "learning_rate": 9.628237259816208e-06,
+      "loss": 0.2809,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4788732394366195,
+      "grad_norm": 0.248504426413411,
+      "learning_rate": 9.607351712614871e-06,
+      "loss": 0.29,
+      "step": 2201
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.23464770896979678,
+      "learning_rate": 9.586466165413533e-06,
+      "loss": 0.2803,
+      "step": 2202
+    },
+    {
+      "epoch": 2.4811267605633804,
+      "grad_norm": 0.216404731594425,
+      "learning_rate": 9.565580618212196e-06,
+      "loss": 0.288,
+      "step": 2203
+    },
+    {
+      "epoch": 2.4822535211267605,
+      "grad_norm": 0.21606120576921523,
+      "learning_rate": 9.54469507101086e-06,
+      "loss": 0.276,
+      "step": 2204
+    },
+    {
+      "epoch": 2.483380281690141,
+      "grad_norm": 0.23724905187991394,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.2821,
+      "step": 2205
+    },
+    {
+      "epoch": 2.4845070422535214,
+      "grad_norm": 0.2213014785466372,
+      "learning_rate": 9.502923976608186e-06,
+      "loss": 0.2703,
+      "step": 2206
+    },
+    {
+      "epoch": 2.4856338028169014,
+      "grad_norm": 0.20649440183174905,
+      "learning_rate": 9.48203842940685e-06,
+      "loss": 0.2566,
+      "step": 2207
+    },
+    {
+      "epoch": 2.486760563380282,
+      "grad_norm": 0.2188221728681939,
+      "learning_rate": 9.461152882205513e-06,
+      "loss": 0.2829,
+      "step": 2208
+    },
+    {
+      "epoch": 2.487887323943662,
+      "grad_norm": 0.20971132597538833,
+      "learning_rate": 9.440267335004177e-06,
+      "loss": 0.2636,
+      "step": 2209
+    },
+    {
+      "epoch": 2.4890140845070423,
+      "grad_norm": 0.22017712989793567,
+      "learning_rate": 9.41938178780284e-06,
+      "loss": 0.2819,
+      "step": 2210
+    },
+    {
+      "epoch": 2.4901408450704228,
+      "grad_norm": 0.21163799913141193,
+      "learning_rate": 9.398496240601503e-06,
+      "loss": 0.2693,
+      "step": 2211
+    },
+    {
+      "epoch": 2.491267605633803,
+      "grad_norm": 0.22189885631357084,
+      "learning_rate": 9.377610693400167e-06,
+      "loss": 0.2704,
+      "step": 2212
+    },
+    {
+      "epoch": 2.4923943661971832,
+      "grad_norm": 0.218520608239522,
+      "learning_rate": 9.35672514619883e-06,
+      "loss": 0.2767,
+      "step": 2213
+    },
+    {
+      "epoch": 2.4935211267605633,
+      "grad_norm": 0.215359628487211,
+      "learning_rate": 9.335839598997493e-06,
+      "loss": 0.2783,
+      "step": 2214
+    },
+    {
+      "epoch": 2.4946478873239437,
+      "grad_norm": 0.22803232795887976,
+      "learning_rate": 9.314954051796157e-06,
+      "loss": 0.2525,
+      "step": 2215
+    },
+    {
+      "epoch": 2.495774647887324,
+      "grad_norm": 0.23560760814110923,
+      "learning_rate": 9.29406850459482e-06,
+      "loss": 0.2831,
+      "step": 2216
+    },
+    {
+      "epoch": 2.496901408450704,
+      "grad_norm": 0.22343667946076662,
+      "learning_rate": 9.273182957393484e-06,
+      "loss": 0.278,
+      "step": 2217
+    },
+    {
+      "epoch": 2.4980281690140846,
+      "grad_norm": 0.23027385335442674,
+      "learning_rate": 9.252297410192147e-06,
+      "loss": 0.2761,
+      "step": 2218
+    },
+    {
+      "epoch": 2.4991549295774647,
+      "grad_norm": 0.20879502549623066,
+      "learning_rate": 9.23141186299081e-06,
+      "loss": 0.284,
+      "step": 2219
+    },
+    {
+      "epoch": 2.500281690140845,
+      "grad_norm": 0.20835602045508547,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.2676,
+      "step": 2220
+    },
+    {
+      "epoch": 2.5014084507042256,
+      "grad_norm": 0.21675622542416495,
+      "learning_rate": 9.189640768588137e-06,
+      "loss": 0.2851,
+      "step": 2221
+    },
+    {
+      "epoch": 2.5025352112676056,
+      "grad_norm": 0.21434267814700203,
+      "learning_rate": 9.1687552213868e-06,
+      "loss": 0.2714,
+      "step": 2222
+    },
+    {
+      "epoch": 2.503661971830986,
+      "grad_norm": 0.21218470836676095,
+      "learning_rate": 9.147869674185464e-06,
+      "loss": 0.2752,
+      "step": 2223
+    },
+    {
+      "epoch": 2.504788732394366,
+      "grad_norm": 0.223752505258997,
+      "learning_rate": 9.126984126984127e-06,
+      "loss": 0.2776,
+      "step": 2224
+    },
+    {
+      "epoch": 2.5059154929577465,
+      "grad_norm": 0.22561799691711334,
+      "learning_rate": 9.10609857978279e-06,
+      "loss": 0.2733,
+      "step": 2225
+    },
+    {
+      "epoch": 2.507042253521127,
+      "grad_norm": 0.22188044096598836,
+      "learning_rate": 9.085213032581454e-06,
+      "loss": 0.2934,
+      "step": 2226
+    },
+    {
+      "epoch": 2.508169014084507,
+      "grad_norm": 0.2400515222028944,
+      "learning_rate": 9.064327485380117e-06,
+      "loss": 0.2751,
+      "step": 2227
+    },
+    {
+      "epoch": 2.5092957746478874,
+      "grad_norm": 1.1185047955601348,
+      "learning_rate": 9.04344193817878e-06,
+      "loss": 0.314,
+      "step": 2228
+    },
+    {
+      "epoch": 2.5104225352112675,
+      "grad_norm": 0.212220685527622,
+      "learning_rate": 9.022556390977444e-06,
+      "loss": 0.2547,
+      "step": 2229
+    },
+    {
+      "epoch": 2.511549295774648,
+      "grad_norm": 0.19524098475255758,
+      "learning_rate": 9.001670843776107e-06,
+      "loss": 0.2591,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5126760563380284,
+      "grad_norm": 0.21921017676076446,
+      "learning_rate": 8.98078529657477e-06,
+      "loss": 0.2566,
+      "step": 2231
+    },
+    {
+      "epoch": 2.5138028169014084,
+      "grad_norm": 0.24290751995043525,
+      "learning_rate": 8.959899749373434e-06,
+      "loss": 0.2835,
+      "step": 2232
+    },
+    {
+      "epoch": 2.514929577464789,
+      "grad_norm": 0.2103964108730701,
+      "learning_rate": 8.939014202172098e-06,
+      "loss": 0.2669,
+      "step": 2233
+    },
+    {
+      "epoch": 2.516056338028169,
+      "grad_norm": 0.24321395392230583,
+      "learning_rate": 8.918128654970761e-06,
+      "loss": 0.2777,
+      "step": 2234
+    },
+    {
+      "epoch": 2.5171830985915493,
+      "grad_norm": 0.2193688002097274,
+      "learning_rate": 8.897243107769424e-06,
+      "loss": 0.2853,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5183098591549298,
+      "grad_norm": 0.20127959943533763,
+      "learning_rate": 8.876357560568086e-06,
+      "loss": 0.2641,
+      "step": 2236
+    },
+    {
+      "epoch": 2.51943661971831,
+      "grad_norm": 0.23285952320693487,
+      "learning_rate": 8.85547201336675e-06,
+      "loss": 0.2828,
+      "step": 2237
+    },
+    {
+      "epoch": 2.5205633802816902,
+      "grad_norm": 0.24374568774168048,
+      "learning_rate": 8.834586466165413e-06,
+      "loss": 0.288,
+      "step": 2238
+    },
+    {
+      "epoch": 2.5216901408450703,
+      "grad_norm": 0.2076800484914963,
+      "learning_rate": 8.813700918964076e-06,
+      "loss": 0.2701,
+      "step": 2239
+    },
+    {
+      "epoch": 2.5228169014084507,
+      "grad_norm": 0.22275413320481321,
+      "learning_rate": 8.79281537176274e-06,
+      "loss": 0.2812,
+      "step": 2240
+    },
+    {
+      "epoch": 2.523943661971831,
+      "grad_norm": 0.2321323473115153,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 0.2847,
+      "step": 2241
+    },
+    {
+      "epoch": 2.525070422535211,
+      "grad_norm": 0.2261219354028944,
+      "learning_rate": 8.751044277360066e-06,
+      "loss": 0.272,
+      "step": 2242
+    },
+    {
+      "epoch": 2.5261971830985916,
+      "grad_norm": 0.2160864640984234,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.2738,
+      "step": 2243
+    },
+    {
+      "epoch": 2.5273239436619717,
+      "grad_norm": 0.20292547199970007,
+      "learning_rate": 8.709273182957393e-06,
+      "loss": 0.2775,
+      "step": 2244
+    },
+    {
+      "epoch": 2.528450704225352,
+      "grad_norm": 0.22177204904918363,
+      "learning_rate": 8.688387635756056e-06,
+      "loss": 0.2775,
+      "step": 2245
+    },
+    {
+      "epoch": 2.5295774647887326,
+      "grad_norm": 0.21710872413385202,
+      "learning_rate": 8.66750208855472e-06,
+      "loss": 0.2674,
+      "step": 2246
+    },
+    {
+      "epoch": 2.5307042253521126,
+      "grad_norm": 0.22683414573268926,
+      "learning_rate": 8.646616541353383e-06,
+      "loss": 0.282,
+      "step": 2247
+    },
+    {
+      "epoch": 2.531830985915493,
+      "grad_norm": 0.21072187393255432,
+      "learning_rate": 8.625730994152046e-06,
+      "loss": 0.2798,
+      "step": 2248
+    },
+    {
+      "epoch": 2.532957746478873,
+      "grad_norm": 0.2120038777625571,
+      "learning_rate": 8.60484544695071e-06,
+      "loss": 0.2687,
+      "step": 2249
+    },
+    {
+      "epoch": 2.5340845070422535,
+      "grad_norm": 0.21026382448537614,
+      "learning_rate": 8.583959899749373e-06,
+      "loss": 0.2743,
+      "step": 2250
+    },
+    {
+      "epoch": 2.535211267605634,
+      "grad_norm": 0.2106368497175309,
+      "learning_rate": 8.563074352548037e-06,
+      "loss": 0.2773,
+      "step": 2251
+    },
+    {
+      "epoch": 2.536338028169014,
+      "grad_norm": 0.20603896149924772,
+      "learning_rate": 8.5421888053467e-06,
+      "loss": 0.259,
+      "step": 2252
+    },
+    {
+      "epoch": 2.5374647887323944,
+      "grad_norm": 0.2118135718024562,
+      "learning_rate": 8.521303258145363e-06,
+      "loss": 0.2649,
+      "step": 2253
+    },
+    {
+      "epoch": 2.5385915492957745,
+      "grad_norm": 0.20585213804063746,
+      "learning_rate": 8.500417710944027e-06,
+      "loss": 0.2732,
+      "step": 2254
+    },
+    {
+      "epoch": 2.539718309859155,
+      "grad_norm": 0.20091087601640018,
+      "learning_rate": 8.47953216374269e-06,
+      "loss": 0.2717,
+      "step": 2255
+    },
+    {
+      "epoch": 2.5408450704225354,
+      "grad_norm": 0.21649323641482876,
+      "learning_rate": 8.458646616541353e-06,
+      "loss": 0.2666,
+      "step": 2256
+    },
+    {
+      "epoch": 2.5419718309859154,
+      "grad_norm": 0.20791601769314685,
+      "learning_rate": 8.437761069340017e-06,
+      "loss": 0.257,
+      "step": 2257
+    },
+    {
+      "epoch": 2.543098591549296,
+      "grad_norm": 0.20369955712165008,
+      "learning_rate": 8.41687552213868e-06,
+      "loss": 0.2602,
+      "step": 2258
+    },
+    {
+      "epoch": 2.544225352112676,
+      "grad_norm": 0.20343660503076466,
+      "learning_rate": 8.395989974937344e-06,
+      "loss": 0.2667,
+      "step": 2259
+    },
+    {
+      "epoch": 2.5453521126760563,
+      "grad_norm": 0.22171507016332392,
+      "learning_rate": 8.375104427736007e-06,
+      "loss": 0.2855,
+      "step": 2260
+    },
+    {
+      "epoch": 2.546478873239437,
+      "grad_norm": 0.21406067163893322,
+      "learning_rate": 8.35421888053467e-06,
+      "loss": 0.2695,
+      "step": 2261
+    },
+    {
+      "epoch": 2.547605633802817,
+      "grad_norm": 0.20333696608016158,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2582,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5487323943661973,
+      "grad_norm": 0.21157404363540527,
+      "learning_rate": 8.312447786131997e-06,
+      "loss": 0.2817,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5498591549295773,
+      "grad_norm": 0.19044739916367748,
+      "learning_rate": 8.29156223893066e-06,
+      "loss": 0.2703,
+      "step": 2264
+    },
+    {
+      "epoch": 2.5509859154929577,
+      "grad_norm": 0.22007179198673688,
+      "learning_rate": 8.270676691729324e-06,
+      "loss": 0.278,
+      "step": 2265
+    },
+    {
+      "epoch": 2.552112676056338,
+      "grad_norm": 0.22748155158872424,
+      "learning_rate": 8.249791144527987e-06,
+      "loss": 0.2771,
+      "step": 2266
+    },
+    {
+      "epoch": 2.553239436619718,
+      "grad_norm": 0.22399759601611058,
+      "learning_rate": 8.22890559732665e-06,
+      "loss": 0.29,
+      "step": 2267
+    },
+    {
+      "epoch": 2.5543661971830987,
+      "grad_norm": 0.21124861709091,
+      "learning_rate": 8.208020050125314e-06,
+      "loss": 0.2808,
+      "step": 2268
+    },
+    {
+      "epoch": 2.5554929577464787,
+      "grad_norm": 0.20934774955328989,
+      "learning_rate": 8.187134502923977e-06,
+      "loss": 0.2775,
+      "step": 2269
+    },
+    {
+      "epoch": 2.556619718309859,
+      "grad_norm": 0.20286481973644913,
+      "learning_rate": 8.166248955722639e-06,
+      "loss": 0.2573,
+      "step": 2270
+    },
+    {
+      "epoch": 2.5577464788732396,
+      "grad_norm": 0.21768694947303274,
+      "learning_rate": 8.145363408521302e-06,
+      "loss": 0.2828,
+      "step": 2271
+    },
+    {
+      "epoch": 2.5588732394366196,
+      "grad_norm": 0.21611062790309854,
+      "learning_rate": 8.124477861319966e-06,
+      "loss": 0.2682,
+      "step": 2272
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.20753855733262955,
+      "learning_rate": 8.10359231411863e-06,
+      "loss": 0.2581,
+      "step": 2273
+    },
+    {
+      "epoch": 2.56112676056338,
+      "grad_norm": 0.2263658555970021,
+      "learning_rate": 8.082706766917293e-06,
+      "loss": 0.2787,
+      "step": 2274
+    },
+    {
+      "epoch": 2.5622535211267605,
+      "grad_norm": 0.21119974183836335,
+      "learning_rate": 8.061821219715956e-06,
+      "loss": 0.2727,
+      "step": 2275
+    },
+    {
+      "epoch": 2.563380281690141,
+      "grad_norm": 0.24289625766805403,
+      "learning_rate": 8.04093567251462e-06,
+      "loss": 0.292,
+      "step": 2276
+    },
+    {
+      "epoch": 2.564507042253521,
+      "grad_norm": 0.20730419219969304,
+      "learning_rate": 8.020050125313283e-06,
+      "loss": 0.2759,
+      "step": 2277
+    },
+    {
+      "epoch": 2.5656338028169015,
+      "grad_norm": 0.20268603012888006,
+      "learning_rate": 7.999164578111946e-06,
+      "loss": 0.2846,
+      "step": 2278
+    },
+    {
+      "epoch": 2.5667605633802815,
+      "grad_norm": 0.24266035890675589,
+      "learning_rate": 7.97827903091061e-06,
+      "loss": 0.2744,
+      "step": 2279
+    },
+    {
+      "epoch": 2.567887323943662,
+      "grad_norm": 0.2255216230584472,
+      "learning_rate": 7.957393483709273e-06,
+      "loss": 0.2722,
+      "step": 2280
+    },
+    {
+      "epoch": 2.5690140845070424,
+      "grad_norm": 0.26969813040912644,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.2839,
+      "step": 2281
+    },
+    {
+      "epoch": 2.5701408450704224,
+      "grad_norm": 0.212315699062885,
+      "learning_rate": 7.9156223893066e-06,
+      "loss": 0.2742,
+      "step": 2282
+    },
+    {
+      "epoch": 2.571267605633803,
+      "grad_norm": 0.22140384088683024,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 0.2754,
+      "step": 2283
+    },
+    {
+      "epoch": 2.572394366197183,
+      "grad_norm": 0.22774101820964093,
+      "learning_rate": 7.873851294903926e-06,
+      "loss": 0.2759,
+      "step": 2284
+    },
+    {
+      "epoch": 2.5735211267605633,
+      "grad_norm": 0.21452111492793702,
+      "learning_rate": 7.85296574770259e-06,
+      "loss": 0.2711,
+      "step": 2285
+    },
+    {
+      "epoch": 2.574647887323944,
+      "grad_norm": 0.24422191612219893,
+      "learning_rate": 7.832080200501253e-06,
+      "loss": 0.2578,
+      "step": 2286
+    },
+    {
+      "epoch": 2.575774647887324,
+      "grad_norm": 0.21323333386316548,
+      "learning_rate": 7.811194653299916e-06,
+      "loss": 0.2771,
+      "step": 2287
+    },
+    {
+      "epoch": 2.5769014084507043,
+      "grad_norm": 0.23961948141565678,
+      "learning_rate": 7.79030910609858e-06,
+      "loss": 0.2861,
+      "step": 2288
+    },
+    {
+      "epoch": 2.5780281690140843,
+      "grad_norm": 0.23425870356594633,
+      "learning_rate": 7.769423558897243e-06,
+      "loss": 0.2421,
+      "step": 2289
+    },
+    {
+      "epoch": 2.5791549295774647,
+      "grad_norm": 0.1996419422515128,
+      "learning_rate": 7.748538011695907e-06,
+      "loss": 0.273,
+      "step": 2290
+    },
+    {
+      "epoch": 2.580281690140845,
+      "grad_norm": 0.20233534762369687,
+      "learning_rate": 7.72765246449457e-06,
+      "loss": 0.2703,
+      "step": 2291
+    },
+    {
+      "epoch": 2.581408450704225,
+      "grad_norm": 0.21712525623306356,
+      "learning_rate": 7.706766917293233e-06,
+      "loss": 0.2759,
+      "step": 2292
+    },
+    {
+      "epoch": 2.5825352112676057,
+      "grad_norm": 0.22883204279526698,
+      "learning_rate": 7.685881370091897e-06,
+      "loss": 0.2678,
+      "step": 2293
+    },
+    {
+      "epoch": 2.5836619718309857,
+      "grad_norm": 0.2204878352077323,
+      "learning_rate": 7.66499582289056e-06,
+      "loss": 0.275,
+      "step": 2294
+    },
+    {
+      "epoch": 2.584788732394366,
+      "grad_norm": 0.22141537977546438,
+      "learning_rate": 7.644110275689223e-06,
+      "loss": 0.2698,
+      "step": 2295
+    },
+    {
+      "epoch": 2.5859154929577466,
+      "grad_norm": 0.19423299809216446,
+      "learning_rate": 7.623224728487886e-06,
+      "loss": 0.2689,
+      "step": 2296
+    },
+    {
+      "epoch": 2.5870422535211266,
+      "grad_norm": 0.21596012544263102,
+      "learning_rate": 7.602339181286549e-06,
+      "loss": 0.2718,
+      "step": 2297
+    },
+    {
+      "epoch": 2.588169014084507,
+      "grad_norm": 0.21361017588707046,
+      "learning_rate": 7.581453634085213e-06,
+      "loss": 0.257,
+      "step": 2298
+    },
+    {
+      "epoch": 2.589295774647887,
+      "grad_norm": 0.2100094729735693,
+      "learning_rate": 7.560568086883876e-06,
+      "loss": 0.2728,
+      "step": 2299
+    },
+    {
+      "epoch": 2.5904225352112675,
+      "grad_norm": 0.23462890669836461,
+      "learning_rate": 7.5396825396825394e-06,
+      "loss": 0.2754,
+      "step": 2300
+    },
+    {
+      "epoch": 2.591549295774648,
+      "grad_norm": 0.22123571932368516,
+      "learning_rate": 7.518796992481203e-06,
+      "loss": 0.2865,
+      "step": 2301
+    },
+    {
+      "epoch": 2.5926760563380284,
+      "grad_norm": 0.20183983085558269,
+      "learning_rate": 7.497911445279866e-06,
+      "loss": 0.2539,
+      "step": 2302
+    },
+    {
+      "epoch": 2.5938028169014085,
+      "grad_norm": 0.21065368643205074,
+      "learning_rate": 7.4770258980785296e-06,
+      "loss": 0.2837,
+      "step": 2303
+    },
+    {
+      "epoch": 2.5949295774647885,
+      "grad_norm": 0.21056747302003648,
+      "learning_rate": 7.456140350877193e-06,
+      "loss": 0.2625,
+      "step": 2304
+    },
+    {
+      "epoch": 2.596056338028169,
+      "grad_norm": 0.2030094546961027,
+      "learning_rate": 7.435254803675856e-06,
+      "loss": 0.2737,
+      "step": 2305
+    },
+    {
+      "epoch": 2.5971830985915494,
+      "grad_norm": 0.21033048117190295,
+      "learning_rate": 7.41436925647452e-06,
+      "loss": 0.2701,
+      "step": 2306
+    },
+    {
+      "epoch": 2.59830985915493,
+      "grad_norm": 0.208205304479834,
+      "learning_rate": 7.393483709273183e-06,
+      "loss": 0.2885,
+      "step": 2307
+    },
+    {
+      "epoch": 2.59943661971831,
+      "grad_norm": 0.19802754212323007,
+      "learning_rate": 7.3725981620718464e-06,
+      "loss": 0.2615,
+      "step": 2308
+    },
+    {
+      "epoch": 2.60056338028169,
+      "grad_norm": 0.19863847632704087,
+      "learning_rate": 7.35171261487051e-06,
+      "loss": 0.2595,
+      "step": 2309
+    },
+    {
+      "epoch": 2.6016901408450703,
+      "grad_norm": 0.2060235191172473,
+      "learning_rate": 7.330827067669173e-06,
+      "loss": 0.2708,
+      "step": 2310
+    },
+    {
+      "epoch": 2.602816901408451,
+      "grad_norm": 0.19033175410373906,
+      "learning_rate": 7.3099415204678366e-06,
+      "loss": 0.2511,
+      "step": 2311
+    },
+    {
+      "epoch": 2.6039436619718312,
+      "grad_norm": 0.2273535376707046,
+      "learning_rate": 7.2890559732665e-06,
+      "loss": 0.305,
+      "step": 2312
+    },
+    {
+      "epoch": 2.6050704225352113,
+      "grad_norm": 0.21774376489548256,
+      "learning_rate": 7.2681704260651625e-06,
+      "loss": 0.2717,
+      "step": 2313
+    },
+    {
+      "epoch": 2.6061971830985913,
+      "grad_norm": 0.2123574175700146,
+      "learning_rate": 7.247284878863826e-06,
+      "loss": 0.2663,
+      "step": 2314
+    },
+    {
+      "epoch": 2.6073239436619717,
+      "grad_norm": 0.20251054768809612,
+      "learning_rate": 7.226399331662489e-06,
+      "loss": 0.2722,
+      "step": 2315
+    },
+    {
+      "epoch": 2.608450704225352,
+      "grad_norm": 0.20004193052090175,
+      "learning_rate": 7.205513784461153e-06,
+      "loss": 0.2744,
+      "step": 2316
+    },
+    {
+      "epoch": 2.6095774647887326,
+      "grad_norm": 0.2008765929340839,
+      "learning_rate": 7.184628237259816e-06,
+      "loss": 0.2707,
+      "step": 2317
+    },
+    {
+      "epoch": 2.6107042253521127,
+      "grad_norm": 0.20943784373994273,
+      "learning_rate": 7.163742690058479e-06,
+      "loss": 0.2806,
+      "step": 2318
+    },
+    {
+      "epoch": 2.611830985915493,
+      "grad_norm": 0.19336187833106966,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.2567,
+      "step": 2319
+    },
+    {
+      "epoch": 2.612957746478873,
+      "grad_norm": 0.20249874567077011,
+      "learning_rate": 7.121971595655806e-06,
+      "loss": 0.2821,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6140845070422536,
+      "grad_norm": 0.2038952398146593,
+      "learning_rate": 7.1010860484544695e-06,
+      "loss": 0.2647,
+      "step": 2321
+    },
+    {
+      "epoch": 2.615211267605634,
+      "grad_norm": 0.199090713946215,
+      "learning_rate": 7.080200501253133e-06,
+      "loss": 0.2597,
+      "step": 2322
+    },
+    {
+      "epoch": 2.616338028169014,
+      "grad_norm": 0.20062706496466548,
+      "learning_rate": 7.059314954051796e-06,
+      "loss": 0.2536,
+      "step": 2323
+    },
+    {
+      "epoch": 2.6174647887323945,
+      "grad_norm": 0.21265638038127999,
+      "learning_rate": 7.03842940685046e-06,
+      "loss": 0.2735,
+      "step": 2324
+    },
+    {
+      "epoch": 2.6185915492957745,
+      "grad_norm": 0.2269643737061253,
+      "learning_rate": 7.017543859649123e-06,
+      "loss": 0.2748,
+      "step": 2325
+    },
+    {
+      "epoch": 2.619718309859155,
+      "grad_norm": 0.21587988309043532,
+      "learning_rate": 6.996658312447786e-06,
+      "loss": 0.2694,
+      "step": 2326
+    },
+    {
+      "epoch": 2.6208450704225354,
+      "grad_norm": 0.21524910286307244,
+      "learning_rate": 6.97577276524645e-06,
+      "loss": 0.289,
+      "step": 2327
+    },
+    {
+      "epoch": 2.6219718309859155,
+      "grad_norm": 0.2131855827627971,
+      "learning_rate": 6.954887218045113e-06,
+      "loss": 0.2662,
+      "step": 2328
+    },
+    {
+      "epoch": 2.623098591549296,
+      "grad_norm": 0.22605902643475048,
+      "learning_rate": 6.9340016708437765e-06,
+      "loss": 0.2739,
+      "step": 2329
+    },
+    {
+      "epoch": 2.624225352112676,
+      "grad_norm": 0.21724567124408525,
+      "learning_rate": 6.913116123642439e-06,
+      "loss": 0.2742,
+      "step": 2330
+    },
+    {
+      "epoch": 2.6253521126760564,
+      "grad_norm": 0.2462430805017593,
+      "learning_rate": 6.892230576441102e-06,
+      "loss": 0.2803,
+      "step": 2331
+    },
+    {
+      "epoch": 2.626478873239437,
+      "grad_norm": 0.2518622667237682,
+      "learning_rate": 6.871345029239766e-06,
+      "loss": 0.2846,
+      "step": 2332
+    },
+    {
+      "epoch": 2.627605633802817,
+      "grad_norm": 0.21629583479800257,
+      "learning_rate": 6.850459482038429e-06,
+      "loss": 0.2822,
+      "step": 2333
+    },
+    {
+      "epoch": 2.6287323943661973,
+      "grad_norm": 0.1996507832762662,
+      "learning_rate": 6.8295739348370925e-06,
+      "loss": 0.2508,
+      "step": 2334
+    },
+    {
+      "epoch": 2.6298591549295773,
+      "grad_norm": 0.20795118877473084,
+      "learning_rate": 6.808688387635756e-06,
+      "loss": 0.2697,
+      "step": 2335
+    },
+    {
+      "epoch": 2.630985915492958,
+      "grad_norm": 0.22770627889631342,
+      "learning_rate": 6.787802840434419e-06,
+      "loss": 0.2612,
+      "step": 2336
+    },
+    {
+      "epoch": 2.6321126760563383,
+      "grad_norm": 0.20644557543972056,
+      "learning_rate": 6.766917293233083e-06,
+      "loss": 0.2565,
+      "step": 2337
+    },
+    {
+      "epoch": 2.6332394366197183,
+      "grad_norm": 0.21461015321494206,
+      "learning_rate": 6.746031746031746e-06,
+      "loss": 0.27,
+      "step": 2338
+    },
+    {
+      "epoch": 2.6343661971830987,
+      "grad_norm": 0.19732503779094568,
+      "learning_rate": 6.725146198830409e-06,
+      "loss": 0.2634,
+      "step": 2339
+    },
+    {
+      "epoch": 2.6354929577464787,
+      "grad_norm": 0.26017499658400894,
+      "learning_rate": 6.704260651629073e-06,
+      "loss": 0.2706,
+      "step": 2340
+    },
+    {
+      "epoch": 2.636619718309859,
+      "grad_norm": 0.21222354482286887,
+      "learning_rate": 6.683375104427736e-06,
+      "loss": 0.2671,
+      "step": 2341
+    },
+    {
+      "epoch": 2.6377464788732397,
+      "grad_norm": 0.20910643969965562,
+      "learning_rate": 6.6624895572263995e-06,
+      "loss": 0.2754,
+      "step": 2342
+    },
+    {
+      "epoch": 2.6388732394366197,
+      "grad_norm": 0.2191016125969615,
+      "learning_rate": 6.641604010025063e-06,
+      "loss": 0.2666,
+      "step": 2343
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2335065855062414,
+      "learning_rate": 6.620718462823726e-06,
+      "loss": 0.2685,
+      "step": 2344
+    },
+    {
+      "epoch": 2.64112676056338,
+      "grad_norm": 0.2510005721444893,
+      "learning_rate": 6.59983291562239e-06,
+      "loss": 0.2665,
+      "step": 2345
+    },
+    {
+      "epoch": 2.6422535211267606,
+      "grad_norm": 0.20280531897037118,
+      "learning_rate": 6.578947368421053e-06,
+      "loss": 0.2803,
+      "step": 2346
+    },
+    {
+      "epoch": 2.643380281690141,
+      "grad_norm": 0.20905561403592426,
+      "learning_rate": 6.5580618212197155e-06,
+      "loss": 0.2668,
+      "step": 2347
+    },
+    {
+      "epoch": 2.644507042253521,
+      "grad_norm": 0.2263015920784223,
+      "learning_rate": 6.537176274018379e-06,
+      "loss": 0.2829,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6456338028169015,
+      "grad_norm": 0.20955235757852655,
+      "learning_rate": 6.516290726817042e-06,
+      "loss": 0.2805,
+      "step": 2349
+    },
+    {
+      "epoch": 2.6467605633802815,
+      "grad_norm": 0.224891839804344,
+      "learning_rate": 6.495405179615706e-06,
+      "loss": 0.2785,
+      "step": 2350
+    },
+    {
+      "epoch": 2.647887323943662,
+      "grad_norm": 0.2514145879480467,
+      "learning_rate": 6.474519632414369e-06,
+      "loss": 0.2702,
+      "step": 2351
+    },
+    {
+      "epoch": 2.6490140845070425,
+      "grad_norm": 0.20702753379991554,
+      "learning_rate": 6.453634085213032e-06,
+      "loss": 0.2782,
+      "step": 2352
+    },
+    {
+      "epoch": 2.6501408450704225,
+      "grad_norm": 0.2076677000213851,
+      "learning_rate": 6.432748538011696e-06,
+      "loss": 0.2789,
+      "step": 2353
+    },
+    {
+      "epoch": 2.651267605633803,
+      "grad_norm": 0.20277332239505097,
+      "learning_rate": 6.411862990810359e-06,
+      "loss": 0.2735,
+      "step": 2354
+    },
+    {
+      "epoch": 2.652394366197183,
+      "grad_norm": 0.20395286758663664,
+      "learning_rate": 6.3909774436090225e-06,
+      "loss": 0.2563,
+      "step": 2355
+    },
+    {
+      "epoch": 2.6535211267605634,
+      "grad_norm": 0.20906280386651896,
+      "learning_rate": 6.370091896407686e-06,
+      "loss": 0.2599,
+      "step": 2356
+    },
+    {
+      "epoch": 2.654647887323944,
+      "grad_norm": 0.19817976429154502,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.266,
+      "step": 2357
+    },
+    {
+      "epoch": 2.655774647887324,
+      "grad_norm": 0.20810058909422713,
+      "learning_rate": 6.328320802005013e-06,
+      "loss": 0.2783,
+      "step": 2358
+    },
+    {
+      "epoch": 2.6569014084507043,
+      "grad_norm": 0.20241203375584763,
+      "learning_rate": 6.307435254803676e-06,
+      "loss": 0.261,
+      "step": 2359
+    },
+    {
+      "epoch": 2.6580281690140843,
+      "grad_norm": 0.2118363720707042,
+      "learning_rate": 6.286549707602339e-06,
+      "loss": 0.2678,
+      "step": 2360
+    },
+    {
+      "epoch": 2.659154929577465,
+      "grad_norm": 0.19848641252564173,
+      "learning_rate": 6.265664160401003e-06,
+      "loss": 0.271,
+      "step": 2361
+    },
+    {
+      "epoch": 2.6602816901408453,
+      "grad_norm": 0.20026019201154394,
+      "learning_rate": 6.244778613199666e-06,
+      "loss": 0.2791,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6614084507042253,
+      "grad_norm": 0.20953873693762937,
+      "learning_rate": 6.2238930659983295e-06,
+      "loss": 0.2867,
+      "step": 2363
+    },
+    {
+      "epoch": 2.6625352112676057,
+      "grad_norm": 0.20698398472666224,
+      "learning_rate": 6.203007518796992e-06,
+      "loss": 0.278,
+      "step": 2364
+    },
+    {
+      "epoch": 2.6636619718309857,
+      "grad_norm": 0.19878641417622853,
+      "learning_rate": 6.182121971595655e-06,
+      "loss": 0.2652,
+      "step": 2365
+    },
+    {
+      "epoch": 2.664788732394366,
+      "grad_norm": 0.201533174808496,
+      "learning_rate": 6.161236424394319e-06,
+      "loss": 0.2692,
+      "step": 2366
+    },
+    {
+      "epoch": 2.6659154929577467,
+      "grad_norm": 0.1972095561695125,
+      "learning_rate": 6.140350877192982e-06,
+      "loss": 0.2772,
+      "step": 2367
+    },
+    {
+      "epoch": 2.6670422535211267,
+      "grad_norm": 0.20294816458612344,
+      "learning_rate": 6.1194653299916455e-06,
+      "loss": 0.2634,
+      "step": 2368
+    },
+    {
+      "epoch": 2.668169014084507,
+      "grad_norm": 0.2039384476289958,
+      "learning_rate": 6.098579782790309e-06,
+      "loss": 0.2702,
+      "step": 2369
+    },
+    {
+      "epoch": 2.669295774647887,
+      "grad_norm": 0.20279326725097785,
+      "learning_rate": 6.077694235588972e-06,
+      "loss": 0.2618,
+      "step": 2370
+    },
+    {
+      "epoch": 2.6704225352112676,
+      "grad_norm": 0.18879396983382535,
+      "learning_rate": 6.056808688387636e-06,
+      "loss": 0.2528,
+      "step": 2371
+    },
+    {
+      "epoch": 2.671549295774648,
+      "grad_norm": 0.1959472061397774,
+      "learning_rate": 6.035923141186299e-06,
+      "loss": 0.2681,
+      "step": 2372
+    },
+    {
+      "epoch": 2.672676056338028,
+      "grad_norm": 0.2021374855693646,
+      "learning_rate": 6.015037593984962e-06,
+      "loss": 0.2621,
+      "step": 2373
+    },
+    {
+      "epoch": 2.6738028169014085,
+      "grad_norm": 0.20173353152168114,
+      "learning_rate": 5.994152046783626e-06,
+      "loss": 0.2814,
+      "step": 2374
+    },
+    {
+      "epoch": 2.6749295774647885,
+      "grad_norm": 0.189284667490437,
+      "learning_rate": 5.973266499582289e-06,
+      "loss": 0.2591,
+      "step": 2375
+    },
+    {
+      "epoch": 2.676056338028169,
+      "grad_norm": 0.19970317017407754,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 0.2679,
+      "step": 2376
+    },
+    {
+      "epoch": 2.6771830985915495,
+      "grad_norm": 0.20682008244281408,
+      "learning_rate": 5.931495405179616e-06,
+      "loss": 0.2551,
+      "step": 2377
+    },
+    {
+      "epoch": 2.6783098591549295,
+      "grad_norm": 0.19441488497018497,
+      "learning_rate": 5.910609857978279e-06,
+      "loss": 0.2694,
+      "step": 2378
+    },
+    {
+      "epoch": 2.67943661971831,
+      "grad_norm": 0.2339475027598514,
+      "learning_rate": 5.889724310776943e-06,
+      "loss": 0.2738,
+      "step": 2379
+    },
+    {
+      "epoch": 2.68056338028169,
+      "grad_norm": 0.18963715125286926,
+      "learning_rate": 5.868838763575606e-06,
+      "loss": 0.269,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6816901408450704,
+      "grad_norm": 0.19746564345134807,
+      "learning_rate": 5.8479532163742686e-06,
+      "loss": 0.2539,
+      "step": 2381
+    },
+    {
+      "epoch": 2.682816901408451,
+      "grad_norm": 0.21340096415199547,
+      "learning_rate": 5.827067669172932e-06,
+      "loss": 0.2496,
+      "step": 2382
+    },
+    {
+      "epoch": 2.683943661971831,
+      "grad_norm": 0.2076517092942985,
+      "learning_rate": 5.806182121971595e-06,
+      "loss": 0.2711,
+      "step": 2383
+    },
+    {
+      "epoch": 2.6850704225352113,
+      "grad_norm": 0.21085166652762866,
+      "learning_rate": 5.785296574770259e-06,
+      "loss": 0.2678,
+      "step": 2384
+    },
+    {
+      "epoch": 2.6861971830985913,
+      "grad_norm": 0.20858055414942492,
+      "learning_rate": 5.764411027568922e-06,
+      "loss": 0.2768,
+      "step": 2385
+    },
+    {
+      "epoch": 2.687323943661972,
+      "grad_norm": 0.2020164011881153,
+      "learning_rate": 5.7435254803675854e-06,
+      "loss": 0.2806,
+      "step": 2386
+    },
+    {
+      "epoch": 2.6884507042253523,
+      "grad_norm": 0.199726718314877,
+      "learning_rate": 5.722639933166249e-06,
+      "loss": 0.2715,
+      "step": 2387
+    },
+    {
+      "epoch": 2.6895774647887323,
+      "grad_norm": 0.2080107369414713,
+      "learning_rate": 5.701754385964912e-06,
+      "loss": 0.276,
+      "step": 2388
+    },
+    {
+      "epoch": 2.6907042253521127,
+      "grad_norm": 0.1979896245434823,
+      "learning_rate": 5.6808688387635756e-06,
+      "loss": 0.2817,
+      "step": 2389
+    },
+    {
+      "epoch": 2.6918309859154927,
+      "grad_norm": 0.20562546122155004,
+      "learning_rate": 5.659983291562239e-06,
+      "loss": 0.2704,
+      "step": 2390
+    },
+    {
+      "epoch": 2.692957746478873,
+      "grad_norm": 0.20470550805401863,
+      "learning_rate": 5.639097744360902e-06,
+      "loss": 0.2771,
+      "step": 2391
+    },
+    {
+      "epoch": 2.6940845070422537,
+      "grad_norm": 0.19781324737141767,
+      "learning_rate": 5.618212197159566e-06,
+      "loss": 0.2759,
+      "step": 2392
+    },
+    {
+      "epoch": 2.6952112676056337,
+      "grad_norm": 0.20254476618474254,
+      "learning_rate": 5.597326649958229e-06,
+      "loss": 0.2704,
+      "step": 2393
+    },
+    {
+      "epoch": 2.696338028169014,
+      "grad_norm": 0.20992790173543557,
+      "learning_rate": 5.5764411027568924e-06,
+      "loss": 0.2839,
+      "step": 2394
+    },
+    {
+      "epoch": 2.697464788732394,
+      "grad_norm": 0.1954454347239477,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.2721,
+      "step": 2395
+    },
+    {
+      "epoch": 2.6985915492957746,
+      "grad_norm": 0.2071654200170682,
+      "learning_rate": 5.534670008354219e-06,
+      "loss": 0.2752,
+      "step": 2396
+    },
+    {
+      "epoch": 2.699718309859155,
+      "grad_norm": 0.2242536610374017,
+      "learning_rate": 5.5137844611528826e-06,
+      "loss": 0.2762,
+      "step": 2397
+    },
+    {
+      "epoch": 2.700845070422535,
+      "grad_norm": 0.1907324431506242,
+      "learning_rate": 5.492898913951545e-06,
+      "loss": 0.2771,
+      "step": 2398
+    },
+    {
+      "epoch": 2.7019718309859155,
+      "grad_norm": 0.19484330126498606,
+      "learning_rate": 5.4720133667502085e-06,
+      "loss": 0.2706,
+      "step": 2399
+    },
+    {
+      "epoch": 2.7030985915492955,
+      "grad_norm": 0.20695536111287274,
+      "learning_rate": 5.451127819548872e-06,
+      "loss": 0.281,
+      "step": 2400
+    },
+    {
+      "epoch": 2.704225352112676,
+      "grad_norm": 0.20319374269659216,
+      "learning_rate": 5.430242272347535e-06,
+      "loss": 0.2685,
+      "step": 2401
+    },
+    {
+      "epoch": 2.7053521126760565,
+      "grad_norm": 0.21148004686086,
+      "learning_rate": 5.409356725146199e-06,
+      "loss": 0.2799,
+      "step": 2402
+    },
+    {
+      "epoch": 2.7064788732394365,
+      "grad_norm": 0.21311727421898696,
+      "learning_rate": 5.388471177944862e-06,
+      "loss": 0.2714,
+      "step": 2403
+    },
+    {
+      "epoch": 2.707605633802817,
+      "grad_norm": 0.22574161475977417,
+      "learning_rate": 5.367585630743525e-06,
+      "loss": 0.2767,
+      "step": 2404
+    },
+    {
+      "epoch": 2.708732394366197,
+      "grad_norm": 0.20029944977930578,
+      "learning_rate": 5.346700083542189e-06,
+      "loss": 0.2728,
+      "step": 2405
+    },
+    {
+      "epoch": 2.7098591549295774,
+      "grad_norm": 1.1118092180835053,
+      "learning_rate": 5.325814536340852e-06,
+      "loss": 0.3041,
+      "step": 2406
+    },
+    {
+      "epoch": 2.710985915492958,
+      "grad_norm": 0.19540775371133393,
+      "learning_rate": 5.3049289891395155e-06,
+      "loss": 0.2557,
+      "step": 2407
+    },
+    {
+      "epoch": 2.712112676056338,
+      "grad_norm": 0.23050010758305498,
+      "learning_rate": 5.284043441938179e-06,
+      "loss": 0.2738,
+      "step": 2408
+    },
+    {
+      "epoch": 2.7132394366197183,
+      "grad_norm": 0.21861879662954065,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.2627,
+      "step": 2409
+    },
+    {
+      "epoch": 2.7143661971830984,
+      "grad_norm": 0.2412625976478493,
+      "learning_rate": 5.242272347535506e-06,
+      "loss": 0.2972,
+      "step": 2410
+    },
+    {
+      "epoch": 2.715492957746479,
+      "grad_norm": 0.21315491181399193,
+      "learning_rate": 5.221386800334169e-06,
+      "loss": 0.2598,
+      "step": 2411
+    },
+    {
+      "epoch": 2.7166197183098593,
+      "grad_norm": 0.21801887717412052,
+      "learning_rate": 5.200501253132832e-06,
+      "loss": 0.277,
+      "step": 2412
+    },
+    {
+      "epoch": 2.7177464788732393,
+      "grad_norm": 0.2288078040701529,
+      "learning_rate": 5.179615705931496e-06,
+      "loss": 0.2709,
+      "step": 2413
+    },
+    {
+      "epoch": 2.7188732394366197,
+      "grad_norm": 0.24229940991558882,
+      "learning_rate": 5.158730158730159e-06,
+      "loss": 0.2763,
+      "step": 2414
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.20780929808380763,
+      "learning_rate": 5.137844611528822e-06,
+      "loss": 0.2587,
+      "step": 2415
+    },
+    {
+      "epoch": 2.72112676056338,
+      "grad_norm": 0.1965927790337949,
+      "learning_rate": 5.116959064327485e-06,
+      "loss": 0.2559,
+      "step": 2416
+    },
+    {
+      "epoch": 2.7222535211267607,
+      "grad_norm": 0.21812181825805288,
+      "learning_rate": 5.096073517126148e-06,
+      "loss": 0.2717,
+      "step": 2417
+    },
+    {
+      "epoch": 2.7233802816901407,
+      "grad_norm": 0.20356344576897678,
+      "learning_rate": 5.075187969924812e-06,
+      "loss": 0.2644,
+      "step": 2418
+    },
+    {
+      "epoch": 2.724507042253521,
+      "grad_norm": 0.21794435243405022,
+      "learning_rate": 5.054302422723475e-06,
+      "loss": 0.2972,
+      "step": 2419
+    },
+    {
+      "epoch": 2.725633802816901,
+      "grad_norm": 0.20482210124602934,
+      "learning_rate": 5.0334168755221385e-06,
+      "loss": 0.26,
+      "step": 2420
+    },
+    {
+      "epoch": 2.7267605633802816,
+      "grad_norm": 0.22141978206467378,
+      "learning_rate": 5.012531328320802e-06,
+      "loss": 0.2792,
+      "step": 2421
+    },
+    {
+      "epoch": 2.727887323943662,
+      "grad_norm": 0.20919522927581857,
+      "learning_rate": 4.991645781119465e-06,
+      "loss": 0.2684,
+      "step": 2422
+    },
+    {
+      "epoch": 2.7290140845070425,
+      "grad_norm": 0.2216358246660207,
+      "learning_rate": 4.970760233918129e-06,
+      "loss": 0.2699,
+      "step": 2423
+    },
+    {
+      "epoch": 2.7301408450704225,
+      "grad_norm": 0.1969750748283934,
+      "learning_rate": 4.949874686716792e-06,
+      "loss": 0.2809,
+      "step": 2424
+    },
+    {
+      "epoch": 2.7312676056338026,
+      "grad_norm": 0.2079510314461725,
+      "learning_rate": 4.928989139515455e-06,
+      "loss": 0.2696,
+      "step": 2425
+    },
+    {
+      "epoch": 2.732394366197183,
+      "grad_norm": 0.194241562661248,
+      "learning_rate": 4.908103592314119e-06,
+      "loss": 0.2774,
+      "step": 2426
+    },
+    {
+      "epoch": 2.7335211267605635,
+      "grad_norm": 0.2026621133239319,
+      "learning_rate": 4.887218045112782e-06,
+      "loss": 0.2725,
+      "step": 2427
+    },
+    {
+      "epoch": 2.734647887323944,
+      "grad_norm": 0.20101449945353245,
+      "learning_rate": 4.8663324979114455e-06,
+      "loss": 0.2653,
+      "step": 2428
+    },
+    {
+      "epoch": 2.735774647887324,
+      "grad_norm": 0.2006548082586194,
+      "learning_rate": 4.845446950710109e-06,
+      "loss": 0.2726,
+      "step": 2429
+    },
+    {
+      "epoch": 2.736901408450704,
+      "grad_norm": 0.20515002494448348,
+      "learning_rate": 4.824561403508772e-06,
+      "loss": 0.2752,
+      "step": 2430
+    },
+    {
+      "epoch": 2.7380281690140844,
+      "grad_norm": 0.19443305928747673,
+      "learning_rate": 4.803675856307436e-06,
+      "loss": 0.26,
+      "step": 2431
+    },
+    {
+      "epoch": 2.739154929577465,
+      "grad_norm": 0.24936926145200294,
+      "learning_rate": 4.782790309106098e-06,
+      "loss": 0.2781,
+      "step": 2432
+    },
+    {
+      "epoch": 2.7402816901408453,
+      "grad_norm": 0.19672033980425266,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.259,
+      "step": 2433
+    },
+    {
+      "epoch": 2.7414084507042253,
+      "grad_norm": 0.1925496189912426,
+      "learning_rate": 4.741019214703425e-06,
+      "loss": 0.2696,
+      "step": 2434
+    },
+    {
+      "epoch": 2.742535211267606,
+      "grad_norm": 0.2025168089756278,
+      "learning_rate": 4.720133667502088e-06,
+      "loss": 0.2702,
+      "step": 2435
+    },
+    {
+      "epoch": 2.743661971830986,
+      "grad_norm": 0.19985176043299094,
+      "learning_rate": 4.699248120300752e-06,
+      "loss": 0.2745,
+      "step": 2436
+    },
+    {
+      "epoch": 2.7447887323943663,
+      "grad_norm": 0.19981795334353009,
+      "learning_rate": 4.678362573099415e-06,
+      "loss": 0.2629,
+      "step": 2437
+    },
+    {
+      "epoch": 2.7459154929577467,
+      "grad_norm": 0.19927027839419753,
+      "learning_rate": 4.657477025898078e-06,
+      "loss": 0.2831,
+      "step": 2438
+    },
+    {
+      "epoch": 2.7470422535211267,
+      "grad_norm": 0.19225763956546674,
+      "learning_rate": 4.636591478696742e-06,
+      "loss": 0.262,
+      "step": 2439
+    },
+    {
+      "epoch": 2.748169014084507,
+      "grad_norm": 0.2040307365886815,
+      "learning_rate": 4.615705931495405e-06,
+      "loss": 0.2765,
+      "step": 2440
+    },
+    {
+      "epoch": 2.749295774647887,
+      "grad_norm": 0.19287934723693517,
+      "learning_rate": 4.5948203842940685e-06,
+      "loss": 0.2668,
+      "step": 2441
+    },
+    {
+      "epoch": 2.7504225352112677,
+      "grad_norm": 0.19461542697134238,
+      "learning_rate": 4.573934837092732e-06,
+      "loss": 0.2711,
+      "step": 2442
+    },
+    {
+      "epoch": 2.751549295774648,
+      "grad_norm": 0.1941443062206399,
+      "learning_rate": 4.553049289891395e-06,
+      "loss": 0.2757,
+      "step": 2443
+    },
+    {
+      "epoch": 2.752676056338028,
+      "grad_norm": 0.20276456368730925,
+      "learning_rate": 4.532163742690059e-06,
+      "loss": 0.2645,
+      "step": 2444
+    },
+    {
+      "epoch": 2.7538028169014086,
+      "grad_norm": 0.20240283977029078,
+      "learning_rate": 4.511278195488722e-06,
+      "loss": 0.2737,
+      "step": 2445
+    },
+    {
+      "epoch": 2.7549295774647886,
+      "grad_norm": 0.19047012180382308,
+      "learning_rate": 4.490392648287385e-06,
+      "loss": 0.2797,
+      "step": 2446
+    },
+    {
+      "epoch": 2.756056338028169,
+      "grad_norm": 0.19677082956627245,
+      "learning_rate": 4.469507101086049e-06,
+      "loss": 0.2697,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7571830985915495,
+      "grad_norm": 0.18796618933611656,
+      "learning_rate": 4.448621553884712e-06,
+      "loss": 0.2751,
+      "step": 2448
+    },
+    {
+      "epoch": 2.7583098591549295,
+      "grad_norm": 0.18958326345231097,
+      "learning_rate": 4.427736006683375e-06,
+      "loss": 0.2764,
+      "step": 2449
+    },
+    {
+      "epoch": 2.75943661971831,
+      "grad_norm": 0.20261814792556176,
+      "learning_rate": 4.406850459482038e-06,
+      "loss": 0.271,
+      "step": 2450
+    },
+    {
+      "epoch": 2.76056338028169,
+      "grad_norm": 0.1807739197575019,
+      "learning_rate": 4.3859649122807014e-06,
+      "loss": 0.2473,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7616901408450705,
+      "grad_norm": 0.20748841468743903,
+      "learning_rate": 4.365079365079365e-06,
+      "loss": 0.2913,
+      "step": 2452
+    },
+    {
+      "epoch": 2.762816901408451,
+      "grad_norm": 0.20357770127659672,
+      "learning_rate": 4.344193817878028e-06,
+      "loss": 0.2691,
+      "step": 2453
+    },
+    {
+      "epoch": 2.763943661971831,
+      "grad_norm": 0.1954035900950154,
+      "learning_rate": 4.3233082706766916e-06,
+      "loss": 0.2687,
+      "step": 2454
+    },
+    {
+      "epoch": 2.7650704225352114,
+      "grad_norm": 0.20788295930137504,
+      "learning_rate": 4.302422723475355e-06,
+      "loss": 0.2782,
+      "step": 2455
+    },
+    {
+      "epoch": 2.7661971830985914,
+      "grad_norm": 0.20075359582721036,
+      "learning_rate": 4.281537176274018e-06,
+      "loss": 0.2785,
+      "step": 2456
+    },
+    {
+      "epoch": 2.767323943661972,
+      "grad_norm": 0.1948372109824818,
+      "learning_rate": 4.260651629072682e-06,
+      "loss": 0.2773,
+      "step": 2457
+    },
+    {
+      "epoch": 2.7684507042253523,
+      "grad_norm": 0.18984572990853224,
+      "learning_rate": 4.239766081871345e-06,
+      "loss": 0.2715,
+      "step": 2458
+    },
+    {
+      "epoch": 2.7695774647887323,
+      "grad_norm": 0.2867708967764196,
+      "learning_rate": 4.2188805346700084e-06,
+      "loss": 0.2768,
+      "step": 2459
+    },
+    {
+      "epoch": 2.770704225352113,
+      "grad_norm": 0.18287769839140158,
+      "learning_rate": 4.197994987468672e-06,
+      "loss": 0.2643,
+      "step": 2460
+    },
+    {
+      "epoch": 2.771830985915493,
+      "grad_norm": 0.20179759422784257,
+      "learning_rate": 4.177109440267335e-06,
+      "loss": 0.2711,
+      "step": 2461
+    },
+    {
+      "epoch": 2.7729577464788733,
+      "grad_norm": 0.18468214491025223,
+      "learning_rate": 4.1562238930659986e-06,
+      "loss": 0.2591,
+      "step": 2462
+    },
+    {
+      "epoch": 2.7740845070422537,
+      "grad_norm": 0.19385557248568827,
+      "learning_rate": 4.135338345864662e-06,
+      "loss": 0.27,
+      "step": 2463
+    },
+    {
+      "epoch": 2.7752112676056337,
+      "grad_norm": 0.19088025686792837,
+      "learning_rate": 4.114452798663325e-06,
+      "loss": 0.276,
+      "step": 2464
+    },
+    {
+      "epoch": 2.776338028169014,
+      "grad_norm": 0.19939344539211365,
+      "learning_rate": 4.093567251461989e-06,
+      "loss": 0.2698,
+      "step": 2465
+    },
+    {
+      "epoch": 2.777464788732394,
+      "grad_norm": 0.20468133899712737,
+      "learning_rate": 4.072681704260651e-06,
+      "loss": 0.2623,
+      "step": 2466
+    },
+    {
+      "epoch": 2.7785915492957747,
+      "grad_norm": 0.19499115099357528,
+      "learning_rate": 4.051796157059315e-06,
+      "loss": 0.2758,
+      "step": 2467
+    },
+    {
+      "epoch": 2.779718309859155,
+      "grad_norm": 0.19242817971755558,
+      "learning_rate": 4.030910609857978e-06,
+      "loss": 0.2738,
+      "step": 2468
+    },
+    {
+      "epoch": 2.780845070422535,
+      "grad_norm": 0.2064820572185044,
+      "learning_rate": 4.010025062656641e-06,
+      "loss": 0.2658,
+      "step": 2469
+    },
+    {
+      "epoch": 2.7819718309859156,
+      "grad_norm": 0.18478105216990628,
+      "learning_rate": 3.989139515455305e-06,
+      "loss": 0.2525,
+      "step": 2470
+    },
+    {
+      "epoch": 2.7830985915492956,
+      "grad_norm": 0.1923498212779712,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.2715,
+      "step": 2471
+    },
+    {
+      "epoch": 2.784225352112676,
+      "grad_norm": 0.18828664531802047,
+      "learning_rate": 3.9473684210526315e-06,
+      "loss": 0.2728,
+      "step": 2472
+    },
+    {
+      "epoch": 2.7853521126760565,
+      "grad_norm": 0.20210623933644906,
+      "learning_rate": 3.926482873851295e-06,
+      "loss": 0.2861,
+      "step": 2473
+    },
+    {
+      "epoch": 2.7864788732394365,
+      "grad_norm": 0.19619598558040613,
+      "learning_rate": 3.905597326649958e-06,
+      "loss": 0.2789,
+      "step": 2474
+    },
+    {
+      "epoch": 2.787605633802817,
+      "grad_norm": 0.18677663597929384,
+      "learning_rate": 3.884711779448622e-06,
+      "loss": 0.2557,
+      "step": 2475
+    },
+    {
+      "epoch": 2.788732394366197,
+      "grad_norm": 0.19370640944340184,
+      "learning_rate": 3.863826232247285e-06,
+      "loss": 0.261,
+      "step": 2476
+    },
+    {
+      "epoch": 2.7898591549295775,
+      "grad_norm": 0.18917738667503978,
+      "learning_rate": 3.842940685045948e-06,
+      "loss": 0.2846,
+      "step": 2477
+    },
+    {
+      "epoch": 2.790985915492958,
+      "grad_norm": 0.19624087872984702,
+      "learning_rate": 3.822055137844612e-06,
+      "loss": 0.2705,
+      "step": 2478
+    },
+    {
+      "epoch": 2.792112676056338,
+      "grad_norm": 0.2113346802276601,
+      "learning_rate": 3.8011695906432747e-06,
+      "loss": 0.2821,
+      "step": 2479
+    },
+    {
+      "epoch": 2.7932394366197184,
+      "grad_norm": 0.22120899061938676,
+      "learning_rate": 3.780284043441938e-06,
+      "loss": 0.2784,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7943661971830984,
+      "grad_norm": 0.18299334408238654,
+      "learning_rate": 3.7593984962406014e-06,
+      "loss": 0.2691,
+      "step": 2481
+    },
+    {
+      "epoch": 2.795492957746479,
+      "grad_norm": 0.1990833006682935,
+      "learning_rate": 3.7385129490392648e-06,
+      "loss": 0.2652,
+      "step": 2482
+    },
+    {
+      "epoch": 2.7966197183098593,
+      "grad_norm": 0.1987106354910443,
+      "learning_rate": 3.717627401837928e-06,
+      "loss": 0.2632,
+      "step": 2483
+    },
+    {
+      "epoch": 2.7977464788732394,
+      "grad_norm": 0.18900103030662868,
+      "learning_rate": 3.6967418546365915e-06,
+      "loss": 0.2697,
+      "step": 2484
+    },
+    {
+      "epoch": 2.79887323943662,
+      "grad_norm": 0.2043997405036503,
+      "learning_rate": 3.675856307435255e-06,
+      "loss": 0.2877,
+      "step": 2485
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.1967067147976069,
+      "learning_rate": 3.6549707602339183e-06,
+      "loss": 0.2565,
+      "step": 2486
+    },
+    {
+      "epoch": 2.8011267605633803,
+      "grad_norm": 0.19951594066115694,
+      "learning_rate": 3.6340852130325812e-06,
+      "loss": 0.2621,
+      "step": 2487
+    },
+    {
+      "epoch": 2.8022535211267607,
+      "grad_norm": 0.19440731106003983,
+      "learning_rate": 3.6131996658312446e-06,
+      "loss": 0.2583,
+      "step": 2488
+    },
+    {
+      "epoch": 2.8033802816901408,
+      "grad_norm": 0.20889642401441064,
+      "learning_rate": 3.592314118629908e-06,
+      "loss": 0.2769,
+      "step": 2489
+    },
+    {
+      "epoch": 2.804507042253521,
+      "grad_norm": 0.21369620152884275,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 0.2608,
+      "step": 2490
+    },
+    {
+      "epoch": 2.8056338028169012,
+      "grad_norm": 0.20041112286087975,
+      "learning_rate": 3.5505430242272347e-06,
+      "loss": 0.2658,
+      "step": 2491
+    },
+    {
+      "epoch": 2.8067605633802817,
+      "grad_norm": 0.18226035415201983,
+      "learning_rate": 3.529657477025898e-06,
+      "loss": 0.2674,
+      "step": 2492
+    },
+    {
+      "epoch": 2.807887323943662,
+      "grad_norm": 0.2148547507215812,
+      "learning_rate": 3.5087719298245615e-06,
+      "loss": 0.2715,
+      "step": 2493
+    },
+    {
+      "epoch": 2.809014084507042,
+      "grad_norm": 0.19484195408298655,
+      "learning_rate": 3.487886382623225e-06,
+      "loss": 0.2699,
+      "step": 2494
+    },
+    {
+      "epoch": 2.8101408450704226,
+      "grad_norm": 0.20732677521305964,
+      "learning_rate": 3.4670008354218882e-06,
+      "loss": 0.2722,
+      "step": 2495
+    },
+    {
+      "epoch": 2.8112676056338026,
+      "grad_norm": 0.1955496382063183,
+      "learning_rate": 3.446115288220551e-06,
+      "loss": 0.2822,
+      "step": 2496
+    },
+    {
+      "epoch": 2.812394366197183,
+      "grad_norm": 0.19650717309897933,
+      "learning_rate": 3.4252297410192146e-06,
+      "loss": 0.2735,
+      "step": 2497
+    },
+    {
+      "epoch": 2.8135211267605635,
+      "grad_norm": 0.19374665236211555,
+      "learning_rate": 3.404344193817878e-06,
+      "loss": 0.2651,
+      "step": 2498
+    },
+    {
+      "epoch": 2.8146478873239436,
+      "grad_norm": 0.19408012707468997,
+      "learning_rate": 3.3834586466165413e-06,
+      "loss": 0.2736,
+      "step": 2499
+    },
+    {
+      "epoch": 2.815774647887324,
+      "grad_norm": 0.22293055676113496,
+      "learning_rate": 3.3625730994152047e-06,
+      "loss": 0.2942,
+      "step": 2500
+    },
+    {
+      "epoch": 2.816901408450704,
+      "grad_norm": 0.18823841692528656,
+      "learning_rate": 3.341687552213868e-06,
+      "loss": 0.2711,
+      "step": 2501
+    },
+    {
+      "epoch": 2.8180281690140845,
+      "grad_norm": 0.20046700932004663,
+      "learning_rate": 3.3208020050125314e-06,
+      "loss": 0.2839,
+      "step": 2502
+    },
+    {
+      "epoch": 2.819154929577465,
+      "grad_norm": 0.1993489442831212,
+      "learning_rate": 3.299916457811195e-06,
+      "loss": 0.2591,
+      "step": 2503
+    },
+    {
+      "epoch": 2.820281690140845,
+      "grad_norm": 0.21648223238685965,
+      "learning_rate": 3.2790309106098578e-06,
+      "loss": 0.285,
+      "step": 2504
+    },
+    {
+      "epoch": 2.8214084507042254,
+      "grad_norm": 0.1973020058583428,
+      "learning_rate": 3.258145363408521e-06,
+      "loss": 0.2843,
+      "step": 2505
+    },
+    {
+      "epoch": 2.8225352112676054,
+      "grad_norm": 0.19872411461682024,
+      "learning_rate": 3.2372598162071845e-06,
+      "loss": 0.2641,
+      "step": 2506
+    },
+    {
+      "epoch": 2.823661971830986,
+      "grad_norm": 0.20169863379712635,
+      "learning_rate": 3.216374269005848e-06,
+      "loss": 0.2656,
+      "step": 2507
+    },
+    {
+      "epoch": 2.8247887323943663,
+      "grad_norm": 0.1888517270151852,
+      "learning_rate": 3.1954887218045113e-06,
+      "loss": 0.2653,
+      "step": 2508
+    },
+    {
+      "epoch": 2.8259154929577464,
+      "grad_norm": 0.20216080686759164,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.2744,
+      "step": 2509
+    },
+    {
+      "epoch": 2.827042253521127,
+      "grad_norm": 0.20181970891877923,
+      "learning_rate": 3.153717627401838e-06,
+      "loss": 0.2829,
+      "step": 2510
+    },
+    {
+      "epoch": 2.828169014084507,
+      "grad_norm": 0.20300929741881887,
+      "learning_rate": 3.1328320802005014e-06,
+      "loss": 0.2563,
+      "step": 2511
+    },
+    {
+      "epoch": 2.8292957746478873,
+      "grad_norm": 0.19968083349162075,
+      "learning_rate": 3.1119465329991648e-06,
+      "loss": 0.2627,
+      "step": 2512
+    },
+    {
+      "epoch": 2.8304225352112677,
+      "grad_norm": 0.20002016294034097,
+      "learning_rate": 3.0910609857978277e-06,
+      "loss": 0.2857,
+      "step": 2513
+    },
+    {
+      "epoch": 2.8315492957746478,
+      "grad_norm": 0.19989915565084954,
+      "learning_rate": 3.070175438596491e-06,
+      "loss": 0.2635,
+      "step": 2514
+    },
+    {
+      "epoch": 2.832676056338028,
+      "grad_norm": 0.1931276854494213,
+      "learning_rate": 3.0492898913951545e-06,
+      "loss": 0.2692,
+      "step": 2515
+    },
+    {
+      "epoch": 2.8338028169014082,
+      "grad_norm": 0.20152537458785733,
+      "learning_rate": 3.028404344193818e-06,
+      "loss": 0.2718,
+      "step": 2516
+    },
+    {
+      "epoch": 2.8349295774647887,
+      "grad_norm": 0.2167040764450326,
+      "learning_rate": 3.007518796992481e-06,
+      "loss": 0.2638,
+      "step": 2517
+    },
+    {
+      "epoch": 2.836056338028169,
+      "grad_norm": 0.19834634026979267,
+      "learning_rate": 2.9866332497911446e-06,
+      "loss": 0.2605,
+      "step": 2518
+    },
+    {
+      "epoch": 2.837183098591549,
+      "grad_norm": 0.20216873108165062,
+      "learning_rate": 2.965747702589808e-06,
+      "loss": 0.2831,
+      "step": 2519
+    },
+    {
+      "epoch": 2.8383098591549296,
+      "grad_norm": 0.19829831038853793,
+      "learning_rate": 2.9448621553884713e-06,
+      "loss": 0.2704,
+      "step": 2520
+    },
+    {
+      "epoch": 2.8394366197183096,
+      "grad_norm": 0.18583659228225755,
+      "learning_rate": 2.9239766081871343e-06,
+      "loss": 0.2686,
+      "step": 2521
+    },
+    {
+      "epoch": 2.84056338028169,
+      "grad_norm": 0.18548834091769986,
+      "learning_rate": 2.9030910609857977e-06,
+      "loss": 0.2611,
+      "step": 2522
+    },
+    {
+      "epoch": 2.8416901408450705,
+      "grad_norm": 0.18533496072735434,
+      "learning_rate": 2.882205513784461e-06,
+      "loss": 0.2518,
+      "step": 2523
+    },
+    {
+      "epoch": 2.8428169014084506,
+      "grad_norm": 0.23150060299736044,
+      "learning_rate": 2.8613199665831244e-06,
+      "loss": 0.2827,
+      "step": 2524
+    },
+    {
+      "epoch": 2.843943661971831,
+      "grad_norm": 0.1886977964132381,
+      "learning_rate": 2.8404344193817878e-06,
+      "loss": 0.265,
+      "step": 2525
+    },
+    {
+      "epoch": 2.845070422535211,
+      "grad_norm": 0.2013505228422611,
+      "learning_rate": 2.819548872180451e-06,
+      "loss": 0.2674,
+      "step": 2526
+    },
+    {
+      "epoch": 2.8461971830985915,
+      "grad_norm": 0.2007294096521873,
+      "learning_rate": 2.7986633249791145e-06,
+      "loss": 0.291,
+      "step": 2527
+    },
+    {
+      "epoch": 2.847323943661972,
+      "grad_norm": 0.183745634563008,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.2589,
+      "step": 2528
+    },
+    {
+      "epoch": 2.848450704225352,
+      "grad_norm": 0.18212939429070288,
+      "learning_rate": 2.7568922305764413e-06,
+      "loss": 0.2697,
+      "step": 2529
+    },
+    {
+      "epoch": 2.8495774647887324,
+      "grad_norm": 0.18979119231193187,
+      "learning_rate": 2.7360066833751042e-06,
+      "loss": 0.2701,
+      "step": 2530
+    },
+    {
+      "epoch": 2.8507042253521124,
+      "grad_norm": 0.1972009385195689,
+      "learning_rate": 2.7151211361737676e-06,
+      "loss": 0.2739,
+      "step": 2531
+    },
+    {
+      "epoch": 2.851830985915493,
+      "grad_norm": 0.19723175664487233,
+      "learning_rate": 2.694235588972431e-06,
+      "loss": 0.2812,
+      "step": 2532
+    },
+    {
+      "epoch": 2.8529577464788733,
+      "grad_norm": 0.1937792010166364,
+      "learning_rate": 2.6733500417710944e-06,
+      "loss": 0.2745,
+      "step": 2533
+    },
+    {
+      "epoch": 2.8540845070422534,
+      "grad_norm": 0.3750659868633327,
+      "learning_rate": 2.6524644945697577e-06,
+      "loss": 0.2743,
+      "step": 2534
+    },
+    {
+      "epoch": 2.855211267605634,
+      "grad_norm": 0.1945898375011954,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.2746,
+      "step": 2535
+    },
+    {
+      "epoch": 2.856338028169014,
+      "grad_norm": 0.2150933913305423,
+      "learning_rate": 2.6106934001670845e-06,
+      "loss": 0.2916,
+      "step": 2536
+    },
+    {
+      "epoch": 2.8574647887323943,
+      "grad_norm": 0.21264387967988305,
+      "learning_rate": 2.589807852965748e-06,
+      "loss": 0.2811,
+      "step": 2537
+    },
+    {
+      "epoch": 2.8585915492957747,
+      "grad_norm": 0.23079403788796102,
+      "learning_rate": 2.568922305764411e-06,
+      "loss": 0.285,
+      "step": 2538
+    },
+    {
+      "epoch": 2.859718309859155,
+      "grad_norm": 0.20558029784877604,
+      "learning_rate": 2.548036758563074e-06,
+      "loss": 0.2759,
+      "step": 2539
+    },
+    {
+      "epoch": 2.860845070422535,
+      "grad_norm": 0.20785868781053038,
+      "learning_rate": 2.5271512113617376e-06,
+      "loss": 0.2679,
+      "step": 2540
+    },
+    {
+      "epoch": 2.8619718309859152,
+      "grad_norm": 0.19430868517354405,
+      "learning_rate": 2.506265664160401e-06,
+      "loss": 0.2885,
+      "step": 2541
+    },
+    {
+      "epoch": 2.8630985915492957,
+      "grad_norm": 0.1828472425239832,
+      "learning_rate": 2.4853801169590643e-06,
+      "loss": 0.2638,
+      "step": 2542
+    },
+    {
+      "epoch": 2.864225352112676,
+      "grad_norm": 0.5729244338095114,
+      "learning_rate": 2.4644945697577277e-06,
+      "loss": 0.2849,
+      "step": 2543
+    },
+    {
+      "epoch": 2.8653521126760566,
+      "grad_norm": 0.22029494980868425,
+      "learning_rate": 2.443609022556391e-06,
+      "loss": 0.272,
+      "step": 2544
+    },
+    {
+      "epoch": 2.8664788732394366,
+      "grad_norm": 0.22565936277758128,
+      "learning_rate": 2.4227234753550544e-06,
+      "loss": 0.2675,
+      "step": 2545
+    },
+    {
+      "epoch": 2.8676056338028166,
+      "grad_norm": 0.1924515218434605,
+      "learning_rate": 2.401837928153718e-06,
+      "loss": 0.2669,
+      "step": 2546
+    },
+    {
+      "epoch": 2.868732394366197,
+      "grad_norm": 0.1970704537926288,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.2772,
+      "step": 2547
+    },
+    {
+      "epoch": 2.8698591549295775,
+      "grad_norm": 0.1968336302115501,
+      "learning_rate": 2.360066833751044e-06,
+      "loss": 0.2733,
+      "step": 2548
+    },
+    {
+      "epoch": 2.870985915492958,
+      "grad_norm": 0.19951417016284853,
+      "learning_rate": 2.3391812865497075e-06,
+      "loss": 0.2748,
+      "step": 2549
+    },
+    {
+      "epoch": 2.872112676056338,
+      "grad_norm": 0.19731362576994593,
+      "learning_rate": 2.318295739348371e-06,
+      "loss": 0.2701,
+      "step": 2550
+    },
+    {
+      "epoch": 2.873239436619718,
+      "grad_norm": 0.1992218992308437,
+      "learning_rate": 2.2974101921470343e-06,
+      "loss": 0.2683,
+      "step": 2551
+    },
+    {
+      "epoch": 2.8743661971830985,
+      "grad_norm": 0.19687956744752266,
+      "learning_rate": 2.2765246449456976e-06,
+      "loss": 0.269,
+      "step": 2552
+    },
+    {
+      "epoch": 2.875492957746479,
+      "grad_norm": 0.20090642987027768,
+      "learning_rate": 2.255639097744361e-06,
+      "loss": 0.281,
+      "step": 2553
+    },
+    {
+      "epoch": 2.8766197183098594,
+      "grad_norm": 0.18297310576256712,
+      "learning_rate": 2.2347535505430244e-06,
+      "loss": 0.2654,
+      "step": 2554
+    },
+    {
+      "epoch": 2.8777464788732394,
+      "grad_norm": 0.19179779807002117,
+      "learning_rate": 2.2138680033416873e-06,
+      "loss": 0.2741,
+      "step": 2555
+    },
+    {
+      "epoch": 2.87887323943662,
+      "grad_norm": 0.1875378969243087,
+      "learning_rate": 2.1929824561403507e-06,
+      "loss": 0.2732,
+      "step": 2556
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.20155911135753982,
+      "learning_rate": 2.172096908939014e-06,
+      "loss": 0.2643,
+      "step": 2557
+    },
+    {
+      "epoch": 2.8811267605633804,
+      "grad_norm": 0.17436401072988408,
+      "learning_rate": 2.1512113617376775e-06,
+      "loss": 0.2576,
+      "step": 2558
+    },
+    {
+      "epoch": 2.882253521126761,
+      "grad_norm": 0.1888243340130565,
+      "learning_rate": 2.130325814536341e-06,
+      "loss": 0.2687,
+      "step": 2559
+    },
+    {
+      "epoch": 2.883380281690141,
+      "grad_norm": 0.19001310090354823,
+      "learning_rate": 2.1094402673350042e-06,
+      "loss": 0.2845,
+      "step": 2560
+    },
+    {
+      "epoch": 2.8845070422535213,
+      "grad_norm": 0.20043987055068513,
+      "learning_rate": 2.0885547201336676e-06,
+      "loss": 0.2591,
+      "step": 2561
+    },
+    {
+      "epoch": 2.8856338028169013,
+      "grad_norm": 0.18832534506583973,
+      "learning_rate": 2.067669172932331e-06,
+      "loss": 0.2703,
+      "step": 2562
+    },
+    {
+      "epoch": 2.8867605633802818,
+      "grad_norm": 0.18059472089368747,
+      "learning_rate": 2.0467836257309943e-06,
+      "loss": 0.2619,
+      "step": 2563
+    },
+    {
+      "epoch": 2.887887323943662,
+      "grad_norm": 0.18024598808699047,
+      "learning_rate": 2.0258980785296573e-06,
+      "loss": 0.2738,
+      "step": 2564
+    },
+    {
+      "epoch": 2.8890140845070422,
+      "grad_norm": 0.19845150497229927,
+      "learning_rate": 2.0050125313283207e-06,
+      "loss": 0.2684,
+      "step": 2565
+    },
+    {
+      "epoch": 2.8901408450704227,
+      "grad_norm": 0.19833786406257187,
+      "learning_rate": 1.984126984126984e-06,
+      "loss": 0.2783,
+      "step": 2566
+    },
+    {
+      "epoch": 2.8912676056338027,
+      "grad_norm": 0.19626086542361154,
+      "learning_rate": 1.9632414369256474e-06,
+      "loss": 0.2596,
+      "step": 2567
+    },
+    {
+      "epoch": 2.892394366197183,
+      "grad_norm": 0.20338902305801612,
+      "learning_rate": 1.942355889724311e-06,
+      "loss": 0.2764,
+      "step": 2568
+    },
+    {
+      "epoch": 2.8935211267605636,
+      "grad_norm": 0.19781084939411023,
+      "learning_rate": 1.921470342522974e-06,
+      "loss": 0.2778,
+      "step": 2569
+    },
+    {
+      "epoch": 2.8946478873239436,
+      "grad_norm": 0.19561908254215962,
+      "learning_rate": 1.9005847953216373e-06,
+      "loss": 0.2699,
+      "step": 2570
+    },
+    {
+      "epoch": 2.895774647887324,
+      "grad_norm": 0.1935427762584011,
+      "learning_rate": 1.8796992481203007e-06,
+      "loss": 0.2838,
+      "step": 2571
+    },
+    {
+      "epoch": 2.896901408450704,
+      "grad_norm": 0.19901870718500267,
+      "learning_rate": 1.858813700918964e-06,
+      "loss": 0.2607,
+      "step": 2572
+    },
+    {
+      "epoch": 2.8980281690140846,
+      "grad_norm": 0.2041526853343621,
+      "learning_rate": 1.8379281537176275e-06,
+      "loss": 0.2632,
+      "step": 2573
+    },
+    {
+      "epoch": 2.899154929577465,
+      "grad_norm": 0.20369090183683206,
+      "learning_rate": 1.8170426065162906e-06,
+      "loss": 0.2775,
+      "step": 2574
+    },
+    {
+      "epoch": 2.900281690140845,
+      "grad_norm": 0.19548877752031224,
+      "learning_rate": 1.796157059314954e-06,
+      "loss": 0.2744,
+      "step": 2575
+    },
+    {
+      "epoch": 2.9014084507042255,
+      "grad_norm": 0.1966975673022479,
+      "learning_rate": 1.7752715121136174e-06,
+      "loss": 0.2833,
+      "step": 2576
+    },
+    {
+      "epoch": 2.9025352112676055,
+      "grad_norm": 0.17786942131721542,
+      "learning_rate": 1.7543859649122807e-06,
+      "loss": 0.2723,
+      "step": 2577
+    },
+    {
+      "epoch": 2.903661971830986,
+      "grad_norm": 0.19925521250604009,
+      "learning_rate": 1.7335004177109441e-06,
+      "loss": 0.2636,
+      "step": 2578
+    },
+    {
+      "epoch": 2.9047887323943664,
+      "grad_norm": 0.23409309646200915,
+      "learning_rate": 1.7126148705096073e-06,
+      "loss": 0.2921,
+      "step": 2579
+    },
+    {
+      "epoch": 2.9059154929577464,
+      "grad_norm": 0.18937871949826304,
+      "learning_rate": 1.6917293233082707e-06,
+      "loss": 0.2813,
+      "step": 2580
+    },
+    {
+      "epoch": 2.907042253521127,
+      "grad_norm": 0.17928190877065076,
+      "learning_rate": 1.670843776106934e-06,
+      "loss": 0.2581,
+      "step": 2581
+    },
+    {
+      "epoch": 2.908169014084507,
+      "grad_norm": 0.19881594048337264,
+      "learning_rate": 1.6499582289055974e-06,
+      "loss": 0.2791,
+      "step": 2582
+    },
+    {
+      "epoch": 2.9092957746478874,
+      "grad_norm": 0.18798769970014184,
+      "learning_rate": 1.6290726817042606e-06,
+      "loss": 0.2692,
+      "step": 2583
+    },
+    {
+      "epoch": 2.910422535211268,
+      "grad_norm": 0.18484773070092542,
+      "learning_rate": 1.608187134502924e-06,
+      "loss": 0.2718,
+      "step": 2584
+    },
+    {
+      "epoch": 2.911549295774648,
+      "grad_norm": 0.1968185538288094,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.2581,
+      "step": 2585
+    },
+    {
+      "epoch": 2.9126760563380283,
+      "grad_norm": 0.1799512428536084,
+      "learning_rate": 1.5664160401002507e-06,
+      "loss": 0.2494,
+      "step": 2586
+    },
+    {
+      "epoch": 2.9138028169014083,
+      "grad_norm": 0.19311808887546641,
+      "learning_rate": 1.5455304928989139e-06,
+      "loss": 0.2612,
+      "step": 2587
+    },
+    {
+      "epoch": 2.9149295774647888,
+      "grad_norm": 0.1977538053337172,
+      "learning_rate": 1.5246449456975772e-06,
+      "loss": 0.2791,
+      "step": 2588
+    },
+    {
+      "epoch": 2.916056338028169,
+      "grad_norm": 0.19783783260320706,
+      "learning_rate": 1.5037593984962406e-06,
+      "loss": 0.286,
+      "step": 2589
+    },
+    {
+      "epoch": 2.9171830985915492,
+      "grad_norm": 0.18923595980392527,
+      "learning_rate": 1.482873851294904e-06,
+      "loss": 0.2756,
+      "step": 2590
+    },
+    {
+      "epoch": 2.9183098591549297,
+      "grad_norm": 0.1903715800057653,
+      "learning_rate": 1.4619883040935671e-06,
+      "loss": 0.2527,
+      "step": 2591
+    },
+    {
+      "epoch": 2.9194366197183097,
+      "grad_norm": 0.17790209769789897,
+      "learning_rate": 1.4411027568922305e-06,
+      "loss": 0.2518,
+      "step": 2592
+    },
+    {
+      "epoch": 2.92056338028169,
+      "grad_norm": 0.1935001923293554,
+      "learning_rate": 1.4202172096908939e-06,
+      "loss": 0.2765,
+      "step": 2593
+    },
+    {
+      "epoch": 2.9216901408450706,
+      "grad_norm": 0.19754748528918087,
+      "learning_rate": 1.3993316624895573e-06,
+      "loss": 0.2857,
+      "step": 2594
+    },
+    {
+      "epoch": 2.9228169014084506,
+      "grad_norm": 0.20086815277045125,
+      "learning_rate": 1.3784461152882206e-06,
+      "loss": 0.2751,
+      "step": 2595
+    },
+    {
+      "epoch": 2.923943661971831,
+      "grad_norm": 0.18549765450299266,
+      "learning_rate": 1.3575605680868838e-06,
+      "loss": 0.2519,
+      "step": 2596
+    },
+    {
+      "epoch": 2.925070422535211,
+      "grad_norm": 0.20347648261235304,
+      "learning_rate": 1.3366750208855472e-06,
+      "loss": 0.2703,
+      "step": 2597
+    },
+    {
+      "epoch": 2.9261971830985916,
+      "grad_norm": 0.18863661663852818,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.2815,
+      "step": 2598
+    },
+    {
+      "epoch": 2.927323943661972,
+      "grad_norm": 0.1909766983577419,
+      "learning_rate": 1.294903926482874e-06,
+      "loss": 0.2898,
+      "step": 2599
+    },
+    {
+      "epoch": 2.928450704225352,
+      "grad_norm": 0.1999440510850279,
+      "learning_rate": 1.274018379281537e-06,
+      "loss": 0.2755,
+      "step": 2600
+    },
+    {
+      "epoch": 2.9295774647887325,
+      "grad_norm": 0.20261542499258492,
+      "learning_rate": 1.2531328320802005e-06,
+      "loss": 0.2816,
+      "step": 2601
+    },
+    {
+      "epoch": 2.9307042253521125,
+      "grad_norm": 0.18547039231567655,
+      "learning_rate": 1.2322472848788638e-06,
+      "loss": 0.2706,
+      "step": 2602
+    },
+    {
+      "epoch": 2.931830985915493,
+      "grad_norm": 0.1769366856054714,
+      "learning_rate": 1.2113617376775272e-06,
+      "loss": 0.2477,
+      "step": 2603
+    },
+    {
+      "epoch": 2.9329577464788734,
+      "grad_norm": 0.18674158365631324,
+      "learning_rate": 1.1904761904761904e-06,
+      "loss": 0.2719,
+      "step": 2604
+    },
+    {
+      "epoch": 2.9340845070422534,
+      "grad_norm": 0.18119752904576938,
+      "learning_rate": 1.1695906432748538e-06,
+      "loss": 0.2601,
+      "step": 2605
+    },
+    {
+      "epoch": 2.935211267605634,
+      "grad_norm": 0.17563623868661624,
+      "learning_rate": 1.1487050960735171e-06,
+      "loss": 0.2541,
+      "step": 2606
+    },
+    {
+      "epoch": 2.936338028169014,
+      "grad_norm": 0.18921047383106981,
+      "learning_rate": 1.1278195488721805e-06,
+      "loss": 0.2641,
+      "step": 2607
+    },
+    {
+      "epoch": 2.9374647887323944,
+      "grad_norm": 0.18681988729634014,
+      "learning_rate": 1.1069340016708437e-06,
+      "loss": 0.2758,
+      "step": 2608
+    },
+    {
+      "epoch": 2.938591549295775,
+      "grad_norm": 0.19596833579258421,
+      "learning_rate": 1.086048454469507e-06,
+      "loss": 0.2726,
+      "step": 2609
+    },
+    {
+      "epoch": 2.939718309859155,
+      "grad_norm": 0.1825674306794061,
+      "learning_rate": 1.0651629072681704e-06,
+      "loss": 0.2698,
+      "step": 2610
+    },
+    {
+      "epoch": 2.9408450704225353,
+      "grad_norm": 0.18430912912728567,
+      "learning_rate": 1.0442773600668338e-06,
+      "loss": 0.2685,
+      "step": 2611
+    },
+    {
+      "epoch": 2.9419718309859153,
+      "grad_norm": 0.19244209622286726,
+      "learning_rate": 1.0233918128654972e-06,
+      "loss": 0.2806,
+      "step": 2612
+    },
+    {
+      "epoch": 2.9430985915492958,
+      "grad_norm": 0.20302337096941844,
+      "learning_rate": 1.0025062656641603e-06,
+      "loss": 0.2716,
+      "step": 2613
+    },
+    {
+      "epoch": 2.944225352112676,
+      "grad_norm": 0.18903859461596323,
+      "learning_rate": 9.816207184628237e-07,
+      "loss": 0.2642,
+      "step": 2614
+    },
+    {
+      "epoch": 2.9453521126760562,
+      "grad_norm": 0.18638171625330813,
+      "learning_rate": 9.60735171261487e-07,
+      "loss": 0.2591,
+      "step": 2615
+    },
+    {
+      "epoch": 2.9464788732394367,
+      "grad_norm": 0.1833750682955007,
+      "learning_rate": 9.398496240601504e-07,
+      "loss": 0.2716,
+      "step": 2616
+    },
+    {
+      "epoch": 2.9476056338028167,
+      "grad_norm": 0.18228414426130218,
+      "learning_rate": 9.189640768588137e-07,
+      "loss": 0.2515,
+      "step": 2617
+    },
+    {
+      "epoch": 2.948732394366197,
+      "grad_norm": 0.1798516100935427,
+      "learning_rate": 8.98078529657477e-07,
+      "loss": 0.259,
+      "step": 2618
+    },
+    {
+      "epoch": 2.9498591549295776,
+      "grad_norm": 0.18413985904175006,
+      "learning_rate": 8.771929824561404e-07,
+      "loss": 0.2607,
+      "step": 2619
+    },
+    {
+      "epoch": 2.9509859154929576,
+      "grad_norm": 0.1840718951735798,
+      "learning_rate": 8.563074352548036e-07,
+      "loss": 0.2589,
+      "step": 2620
+    },
+    {
+      "epoch": 2.952112676056338,
+      "grad_norm": 0.18727995547384207,
+      "learning_rate": 8.35421888053467e-07,
+      "loss": 0.2723,
+      "step": 2621
+    },
+    {
+      "epoch": 2.953239436619718,
+      "grad_norm": 0.18116030630941662,
+      "learning_rate": 8.145363408521303e-07,
+      "loss": 0.2612,
+      "step": 2622
+    },
+    {
+      "epoch": 2.9543661971830986,
+      "grad_norm": 0.18810161068825337,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.2861,
+      "step": 2623
+    },
+    {
+      "epoch": 2.955492957746479,
+      "grad_norm": 0.18331900835667217,
+      "learning_rate": 7.727652464494569e-07,
+      "loss": 0.2792,
+      "step": 2624
+    },
+    {
+      "epoch": 2.956619718309859,
+      "grad_norm": 0.182289911621126,
+      "learning_rate": 7.518796992481203e-07,
+      "loss": 0.2651,
+      "step": 2625
+    },
+    {
+      "epoch": 2.9577464788732395,
+      "grad_norm": 0.18706591671355421,
+      "learning_rate": 7.309941520467836e-07,
+      "loss": 0.2797,
+      "step": 2626
+    },
+    {
+      "epoch": 2.9588732394366195,
+      "grad_norm": 0.18099206297431417,
+      "learning_rate": 7.101086048454469e-07,
+      "loss": 0.2731,
+      "step": 2627
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.1953971293038329,
+      "learning_rate": 6.892230576441103e-07,
+      "loss": 0.2729,
+      "step": 2628
+    },
+    {
+      "epoch": 2.9611267605633804,
+      "grad_norm": 0.20373835878570956,
+      "learning_rate": 6.683375104427736e-07,
+      "loss": 0.2632,
+      "step": 2629
+    },
+    {
+      "epoch": 2.9622535211267604,
+      "grad_norm": 0.17540883485928005,
+      "learning_rate": 6.47451963241437e-07,
+      "loss": 0.2615,
+      "step": 2630
+    },
+    {
+      "epoch": 2.963380281690141,
+      "grad_norm": 0.18358270628844636,
+      "learning_rate": 6.265664160401002e-07,
+      "loss": 0.2597,
+      "step": 2631
+    },
+    {
+      "epoch": 2.964507042253521,
+      "grad_norm": 0.1839418900023207,
+      "learning_rate": 6.056808688387636e-07,
+      "loss": 0.2629,
+      "step": 2632
+    },
+    {
+      "epoch": 2.9656338028169014,
+      "grad_norm": 0.1901007031151732,
+      "learning_rate": 5.847953216374269e-07,
+      "loss": 0.283,
+      "step": 2633
+    },
+    {
+      "epoch": 2.966760563380282,
+      "grad_norm": 0.191252903805876,
+      "learning_rate": 5.639097744360903e-07,
+      "loss": 0.2696,
+      "step": 2634
+    },
+    {
+      "epoch": 2.967887323943662,
+      "grad_norm": 0.1873473949837531,
+      "learning_rate": 5.430242272347535e-07,
+      "loss": 0.2744,
+      "step": 2635
+    },
+    {
+      "epoch": 2.9690140845070423,
+      "grad_norm": 0.20423475149382705,
+      "learning_rate": 5.221386800334169e-07,
+      "loss": 0.2816,
+      "step": 2636
+    },
+    {
+      "epoch": 2.9701408450704223,
+      "grad_norm": 0.18777274154193785,
+      "learning_rate": 5.012531328320802e-07,
+      "loss": 0.2845,
+      "step": 2637
+    },
+    {
+      "epoch": 2.9712676056338028,
+      "grad_norm": 0.18260602994902567,
+      "learning_rate": 4.803675856307435e-07,
+      "loss": 0.2581,
+      "step": 2638
+    },
+    {
+      "epoch": 2.9723943661971832,
+      "grad_norm": 0.1802755775949997,
+      "learning_rate": 4.5948203842940686e-07,
+      "loss": 0.2682,
+      "step": 2639
+    },
+    {
+      "epoch": 2.9735211267605632,
+      "grad_norm": 0.18791079252994222,
+      "learning_rate": 4.385964912280702e-07,
+      "loss": 0.2744,
+      "step": 2640
+    },
+    {
+      "epoch": 2.9746478873239437,
+      "grad_norm": 0.18215279311987378,
+      "learning_rate": 4.177109440267335e-07,
+      "loss": 0.2727,
+      "step": 2641
+    },
+    {
+      "epoch": 2.9757746478873237,
+      "grad_norm": 0.1927350306855454,
+      "learning_rate": 3.9682539682539683e-07,
+      "loss": 0.2625,
+      "step": 2642
+    },
+    {
+      "epoch": 2.976901408450704,
+      "grad_norm": 0.1924027562763381,
+      "learning_rate": 3.7593984962406015e-07,
+      "loss": 0.2782,
+      "step": 2643
+    },
+    {
+      "epoch": 2.9780281690140846,
+      "grad_norm": 0.179203470085156,
+      "learning_rate": 3.5505430242272347e-07,
+      "loss": 0.2719,
+      "step": 2644
+    },
+    {
+      "epoch": 2.9791549295774646,
+      "grad_norm": 0.18815877380613574,
+      "learning_rate": 3.341687552213868e-07,
+      "loss": 0.2835,
+      "step": 2645
+    },
+    {
+      "epoch": 2.980281690140845,
+      "grad_norm": 0.1863822803452641,
+      "learning_rate": 3.132832080200501e-07,
+      "loss": 0.2762,
+      "step": 2646
+    },
+    {
+      "epoch": 2.981408450704225,
+      "grad_norm": 0.18297056791628377,
+      "learning_rate": 2.9239766081871344e-07,
+      "loss": 0.2758,
+      "step": 2647
+    },
+    {
+      "epoch": 2.9825352112676056,
+      "grad_norm": 0.18737249129907566,
+      "learning_rate": 2.7151211361737676e-07,
+      "loss": 0.2562,
+      "step": 2648
+    },
+    {
+      "epoch": 2.983661971830986,
+      "grad_norm": 0.19612228107081542,
+      "learning_rate": 2.506265664160401e-07,
+      "loss": 0.2766,
+      "step": 2649
+    },
+    {
+      "epoch": 2.984788732394366,
+      "grad_norm": 0.1854161232189919,
+      "learning_rate": 2.2974101921470343e-07,
+      "loss": 0.2752,
+      "step": 2650
+    },
+    {
+      "epoch": 2.9859154929577465,
+      "grad_norm": 0.20183176340985043,
+      "learning_rate": 2.0885547201336675e-07,
+      "loss": 0.2534,
+      "step": 2651
+    },
+    {
+      "epoch": 2.9870422535211265,
+      "grad_norm": 0.18318455589567129,
+      "learning_rate": 1.8796992481203008e-07,
+      "loss": 0.2667,
+      "step": 2652
+    },
+    {
+      "epoch": 2.988169014084507,
+      "grad_norm": 0.18114138777343547,
+      "learning_rate": 1.670843776106934e-07,
+      "loss": 0.2649,
+      "step": 2653
+    },
+    {
+      "epoch": 2.9892957746478874,
+      "grad_norm": 0.18142311856831886,
+      "learning_rate": 1.4619883040935672e-07,
+      "loss": 0.2541,
+      "step": 2654
+    },
+    {
+      "epoch": 2.9904225352112674,
+      "grad_norm": 0.19061057401571532,
+      "learning_rate": 1.2531328320802004e-07,
+      "loss": 0.2723,
+      "step": 2655
+    },
+    {
+      "epoch": 2.991549295774648,
+      "grad_norm": 0.17998880127291522,
+      "learning_rate": 1.0442773600668338e-07,
+      "loss": 0.2723,
+      "step": 2656
+    },
+    {
+      "epoch": 2.992676056338028,
+      "grad_norm": 0.19166047768195757,
+      "learning_rate": 8.35421888053467e-08,
+      "loss": 0.2698,
+      "step": 2657
+    },
+    {
+      "epoch": 2.9938028169014084,
+      "grad_norm": 0.20913016553214908,
+      "learning_rate": 6.265664160401002e-08,
+      "loss": 0.2941,
+      "step": 2658
+    },
+    {
+      "epoch": 2.994929577464789,
+      "grad_norm": 0.187476388632694,
+      "learning_rate": 4.177109440267335e-08,
+      "loss": 0.2747,
+      "step": 2659
+    },
+    {
+      "epoch": 2.9960563380281693,
+      "grad_norm": 0.18706857618185885,
+      "learning_rate": 2.0885547201336675e-08,
+      "loss": 0.2755,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9971830985915493,
+      "grad_norm": 0.1863776553285958,
+      "learning_rate": 0.0,
+      "loss": 0.2707,
+      "step": 2661
+    },
+    {
+      "epoch": 2.9971830985915493,
+      "step": 2661,
+      "total_flos": 2.2756805496474173e+18,
+      "train_loss": 0.4145932649813483,
+      "train_runtime": 154577.0289,
+      "train_samples_per_second": 0.275,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2661,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2756805496474173e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}