diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.25852485716501644,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010340994286600657,
+      "grad_norm": 5.403522491455078,
+      "learning_rate": 2e-05,
+      "loss": 2.3818,
+      "step": 1
+    },
+    {
+      "epoch": 0.00020681988573201314,
+      "grad_norm": 8.3225679397583,
+      "learning_rate": 4e-05,
+      "loss": 2.5235,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003102298285980197,
+      "grad_norm": 3.45357608795166,
+      "learning_rate": 6e-05,
+      "loss": 2.1258,
+      "step": 3
+    },
+    {
+      "epoch": 0.0004136397714640263,
+      "grad_norm": 10.966958045959473,
+      "learning_rate": 8e-05,
+      "loss": 3.0596,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005170497143300328,
+      "grad_norm": 1.818668246269226,
+      "learning_rate": 0.0001,
+      "loss": 2.5631,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006204596571960394,
+      "grad_norm": 1.0356301069259644,
+      "learning_rate": 0.00012,
+      "loss": 1.683,
+      "step": 6
+    },
+    {
+      "epoch": 0.000723869600062046,
+      "grad_norm": 1.3918240070343018,
+      "learning_rate": 0.00014,
+      "loss": 2.5425,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008272795429280526,
+      "grad_norm": 3.50264835357666,
+      "learning_rate": 0.00016,
+      "loss": 2.705,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009306894857940591,
+      "grad_norm": 2.2157135009765625,
+      "learning_rate": 0.00018,
+      "loss": 2.5137,
+      "step": 9
+    },
+    {
+      "epoch": 0.0010340994286600657,
+      "grad_norm": 1.1943459510803223,
+      "learning_rate": 0.0002,
+      "loss": 2.2908,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011375093715260721,
+      "grad_norm": 0.9670085906982422,
+      "learning_rate": 0.00019998965338851528,
+      "loss": 1.8234,
+      "step": 11
+    },
+    {
+      "epoch": 0.0012409193143920788,
+      "grad_norm": 0.866587221622467,
+      "learning_rate": 0.00019997930677703054,
+      "loss": 1.9754,
+      "step": 12
+    },
+    {
+      "epoch": 0.0013443292572580853,
+      "grad_norm": 0.941832423210144,
+      "learning_rate": 0.00019996896016554578,
+      "loss": 1.8104,
+      "step": 13
+    },
+    {
+      "epoch": 0.001447739200124092,
+      "grad_norm": 1.8255549669265747,
+      "learning_rate": 0.00019995861355406104,
+      "loss": 1.9344,
+      "step": 14
+    },
+    {
+      "epoch": 0.0015511491429900984,
+      "grad_norm": 2.044438600540161,
+      "learning_rate": 0.0001999482669425763,
+      "loss": 1.8897,
+      "step": 15
+    },
+    {
+      "epoch": 0.001654559085856105,
+      "grad_norm": 1.520342469215393,
+      "learning_rate": 0.00019993792033109158,
+      "loss": 1.8936,
+      "step": 16
+    },
+    {
+      "epoch": 0.0017579690287221116,
+      "grad_norm": 1.0684723854064941,
+      "learning_rate": 0.00019992757371960684,
+      "loss": 2.1489,
+      "step": 17
+    },
+    {
+      "epoch": 0.0018613789715881182,
+      "grad_norm": 0.8357504606246948,
+      "learning_rate": 0.0001999172271081221,
+      "loss": 1.9924,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019647889144541247,
+      "grad_norm": 0.7619360089302063,
+      "learning_rate": 0.00019990688049663735,
+      "loss": 2.19,
+      "step": 19
+    },
+    {
+      "epoch": 0.0020681988573201314,
+      "grad_norm": 0.9019061923027039,
+      "learning_rate": 0.0001998965338851526,
+      "loss": 2.0112,
+      "step": 20
+    },
+    {
+      "epoch": 0.002171608800186138,
+      "grad_norm": 0.8690137267112732,
+      "learning_rate": 0.00019988618727366788,
+      "loss": 2.0031,
+      "step": 21
+    },
+    {
+      "epoch": 0.0022750187430521443,
+      "grad_norm": 1.88344407081604,
+      "learning_rate": 0.00019987584066218314,
+      "loss": 2.6006,
+      "step": 22
+    },
+    {
+      "epoch": 0.002378428685918151,
+      "grad_norm": 0.9320490956306458,
+      "learning_rate": 0.0001998654940506984,
+      "loss": 1.8907,
+      "step": 23
+    },
+    {
+      "epoch": 0.0024818386287841577,
+      "grad_norm": 1.128214955329895,
+      "learning_rate": 0.00019985514743921367,
+      "loss": 1.6415,
+      "step": 24
+    },
+    {
+      "epoch": 0.0025852485716501643,
+      "grad_norm": 1.0375274419784546,
+      "learning_rate": 0.0001998448008277289,
+      "loss": 1.9145,
+      "step": 25
+    },
+    {
+      "epoch": 0.0026886585145161706,
+      "grad_norm": 4.119718551635742,
+      "learning_rate": 0.00019983445421624418,
+      "loss": 1.6368,
+      "step": 26
+    },
+    {
+      "epoch": 0.0027920684573821773,
+      "grad_norm": 0.958248496055603,
+      "learning_rate": 0.00019982410760475944,
+      "loss": 1.632,
+      "step": 27
+    },
+    {
+      "epoch": 0.002895478400248184,
+      "grad_norm": 2.7749078273773193,
+      "learning_rate": 0.0001998137609932747,
+      "loss": 2.4571,
+      "step": 28
+    },
+    {
+      "epoch": 0.0029988883431141906,
+      "grad_norm": 1.2027294635772705,
+      "learning_rate": 0.00019980341438178997,
+      "loss": 1.8688,
+      "step": 29
+    },
+    {
+      "epoch": 0.003102298285980197,
+      "grad_norm": 1.0721759796142578,
+      "learning_rate": 0.00019979306777030524,
+      "loss": 1.9768,
+      "step": 30
+    },
+    {
+      "epoch": 0.0032057082288462035,
+      "grad_norm": 1.3779335021972656,
+      "learning_rate": 0.00019978272115882048,
+      "loss": 1.7229,
+      "step": 31
+    },
+    {
+      "epoch": 0.00330911817171221,
+      "grad_norm": 1.5010607242584229,
+      "learning_rate": 0.00019977237454733574,
+      "loss": 1.9824,
+      "step": 32
+    },
+    {
+      "epoch": 0.003412528114578217,
+      "grad_norm": 0.9950757026672363,
+      "learning_rate": 0.000199762027935851,
+      "loss": 1.1941,
+      "step": 33
+    },
+    {
+      "epoch": 0.003515938057444223,
+      "grad_norm": 0.9842276573181152,
+      "learning_rate": 0.00019975168132436627,
+      "loss": 1.6324,
+      "step": 34
+    },
+    {
+      "epoch": 0.00361934800031023,
+      "grad_norm": 1.6210508346557617,
+      "learning_rate": 0.00019974133471288154,
+      "loss": 2.2909,
+      "step": 35
+    },
+    {
+      "epoch": 0.0037227579431762365,
+      "grad_norm": 1.318981647491455,
+      "learning_rate": 0.0001997309881013968,
+      "loss": 2.4224,
+      "step": 36
+    },
+    {
+      "epoch": 0.003826167886042243,
+      "grad_norm": 0.8251323103904724,
+      "learning_rate": 0.00019972064148991204,
+      "loss": 2.1714,
+      "step": 37
+    },
+    {
+      "epoch": 0.003929577828908249,
+      "grad_norm": 0.9951919317245483,
+      "learning_rate": 0.0001997102948784273,
+      "loss": 1.819,
+      "step": 38
+    },
+    {
+      "epoch": 0.0040329877717742565,
+      "grad_norm": 0.8593564629554749,
+      "learning_rate": 0.00019969994826694258,
+      "loss": 1.8302,
+      "step": 39
+    },
+    {
+      "epoch": 0.004136397714640263,
+      "grad_norm": 1.3132660388946533,
+      "learning_rate": 0.00019968960165545784,
+      "loss": 2.0428,
+      "step": 40
+    },
+    {
+      "epoch": 0.004239807657506269,
+      "grad_norm": 0.806970477104187,
+      "learning_rate": 0.0001996792550439731,
+      "loss": 2.0075,
+      "step": 41
+    },
+    {
+      "epoch": 0.004343217600372276,
+      "grad_norm": 0.8618819713592529,
+      "learning_rate": 0.00019966890843248837,
+      "loss": 1.3391,
+      "step": 42
+    },
+    {
+      "epoch": 0.004446627543238282,
+      "grad_norm": 0.966143786907196,
+      "learning_rate": 0.0001996585618210036,
+      "loss": 1.6868,
+      "step": 43
+    },
+    {
+      "epoch": 0.004550037486104289,
+      "grad_norm": 1.2480261325836182,
+      "learning_rate": 0.00019964821520951888,
+      "loss": 2.3074,
+      "step": 44
+    },
+    {
+      "epoch": 0.004653447428970296,
+      "grad_norm": 0.9969954490661621,
+      "learning_rate": 0.00019963786859803414,
+      "loss": 2.0236,
+      "step": 45
+    },
+    {
+      "epoch": 0.004756857371836302,
+      "grad_norm": 1.7677679061889648,
+      "learning_rate": 0.0001996275219865494,
+      "loss": 2.1244,
+      "step": 46
+    },
+    {
+      "epoch": 0.004860267314702308,
+      "grad_norm": 0.9171494841575623,
+      "learning_rate": 0.00019961717537506467,
+      "loss": 1.6648,
+      "step": 47
+    },
+    {
+      "epoch": 0.004963677257568315,
+      "grad_norm": 1.048912763595581,
+      "learning_rate": 0.00019960682876357994,
+      "loss": 2.1855,
+      "step": 48
+    },
+    {
+      "epoch": 0.0050670872004343216,
+      "grad_norm": 1.4205477237701416,
+      "learning_rate": 0.00019959648215209518,
+      "loss": 2.0607,
+      "step": 49
+    },
+    {
+      "epoch": 0.005170497143300329,
+      "grad_norm": 1.180935025215149,
+      "learning_rate": 0.00019958613554061044,
+      "loss": 2.405,
+      "step": 50
+    },
+    {
+      "epoch": 0.005273907086166335,
+      "grad_norm": 2.3712823390960693,
+      "learning_rate": 0.0001995757889291257,
+      "loss": 1.7894,
+      "step": 51
+    },
+    {
+      "epoch": 0.005377317029032341,
+      "grad_norm": 0.8639737367630005,
+      "learning_rate": 0.00019956544231764097,
+      "loss": 1.488,
+      "step": 52
+    },
+    {
+      "epoch": 0.005480726971898348,
+      "grad_norm": 1.143019199371338,
+      "learning_rate": 0.00019955509570615624,
+      "loss": 2.0396,
+      "step": 53
+    },
+    {
+      "epoch": 0.0055841369147643545,
+      "grad_norm": 0.9509019255638123,
+      "learning_rate": 0.0001995447490946715,
+      "loss": 2.1223,
+      "step": 54
+    },
+    {
+      "epoch": 0.005687546857630361,
+      "grad_norm": 1.8030970096588135,
+      "learning_rate": 0.00019953440248318674,
+      "loss": 2.2948,
+      "step": 55
+    },
+    {
+      "epoch": 0.005790956800496368,
+      "grad_norm": 1.2299774885177612,
+      "learning_rate": 0.000199524055871702,
+      "loss": 2.2414,
+      "step": 56
+    },
+    {
+      "epoch": 0.005894366743362374,
+      "grad_norm": 0.9413872957229614,
+      "learning_rate": 0.00019951370926021727,
+      "loss": 2.2899,
+      "step": 57
+    },
+    {
+      "epoch": 0.005997776686228381,
+      "grad_norm": 1.6003267765045166,
+      "learning_rate": 0.00019950336264873254,
+      "loss": 1.8357,
+      "step": 58
+    },
+    {
+      "epoch": 0.0061011866290943875,
+      "grad_norm": 1.1218072175979614,
+      "learning_rate": 0.0001994930160372478,
+      "loss": 1.8433,
+      "step": 59
+    },
+    {
+      "epoch": 0.006204596571960394,
+      "grad_norm": 0.9314822554588318,
+      "learning_rate": 0.00019948266942576307,
+      "loss": 2.0523,
+      "step": 60
+    },
+    {
+      "epoch": 0.006308006514826401,
+      "grad_norm": 1.321779489517212,
+      "learning_rate": 0.0001994723228142783,
+      "loss": 1.6102,
+      "step": 61
+    },
+    {
+      "epoch": 0.006411416457692407,
+      "grad_norm": 0.9747542142868042,
+      "learning_rate": 0.00019946197620279358,
+      "loss": 2.0,
+      "step": 62
+    },
+    {
+      "epoch": 0.006514826400558413,
+      "grad_norm": 0.7083417177200317,
+      "learning_rate": 0.00019945162959130884,
+      "loss": 2.023,
+      "step": 63
+    },
+    {
+      "epoch": 0.00661823634342442,
+      "grad_norm": 0.7545279860496521,
+      "learning_rate": 0.0001994412829798241,
+      "loss": 1.8848,
+      "step": 64
+    },
+    {
+      "epoch": 0.006721646286290427,
+      "grad_norm": 1.7224853038787842,
+      "learning_rate": 0.00019943093636833937,
+      "loss": 2.3156,
+      "step": 65
+    },
+    {
+      "epoch": 0.006825056229156434,
+      "grad_norm": 1.3995798826217651,
+      "learning_rate": 0.00019942058975685464,
+      "loss": 2.1194,
+      "step": 66
+    },
+    {
+      "epoch": 0.00692846617202244,
+      "grad_norm": 0.8144084811210632,
+      "learning_rate": 0.0001994102431453699,
+      "loss": 1.8109,
+      "step": 67
+    },
+    {
+      "epoch": 0.007031876114888446,
+      "grad_norm": 1.196885585784912,
+      "learning_rate": 0.00019939989653388517,
+      "loss": 2.022,
+      "step": 68
+    },
+    {
+      "epoch": 0.007135286057754453,
+      "grad_norm": 0.7056652903556824,
+      "learning_rate": 0.0001993895499224004,
+      "loss": 2.086,
+      "step": 69
+    },
+    {
+      "epoch": 0.00723869600062046,
+      "grad_norm": 1.2433514595031738,
+      "learning_rate": 0.00019937920331091567,
+      "loss": 2.0888,
+      "step": 70
+    },
+    {
+      "epoch": 0.007342105943486466,
+      "grad_norm": 0.6289485692977905,
+      "learning_rate": 0.00019936885669943094,
+      "loss": 2.3105,
+      "step": 71
+    },
+    {
+      "epoch": 0.007445515886352473,
+      "grad_norm": 0.8752896189689636,
+      "learning_rate": 0.0001993585100879462,
+      "loss": 1.7598,
+      "step": 72
+    },
+    {
+      "epoch": 0.007548925829218479,
+      "grad_norm": 0.6081548929214478,
+      "learning_rate": 0.00019934816347646147,
+      "loss": 1.8297,
+      "step": 73
+    },
+    {
+      "epoch": 0.007652335772084486,
+      "grad_norm": 0.6561211347579956,
+      "learning_rate": 0.00019933781686497674,
+      "loss": 2.0692,
+      "step": 74
+    },
+    {
+      "epoch": 0.0077557457149504926,
+      "grad_norm": 0.7748151421546936,
+      "learning_rate": 0.000199327470253492,
+      "loss": 1.6976,
+      "step": 75
+    },
+    {
+      "epoch": 0.007859155657816499,
+      "grad_norm": 0.9392269849777222,
+      "learning_rate": 0.00019931712364200727,
+      "loss": 1.657,
+      "step": 76
+    },
+    {
+      "epoch": 0.007962565600682505,
+      "grad_norm": 1.0892373323440552,
+      "learning_rate": 0.0001993067770305225,
+      "loss": 2.0677,
+      "step": 77
+    },
+    {
+      "epoch": 0.008065975543548513,
+      "grad_norm": 1.633618950843811,
+      "learning_rate": 0.00019929643041903777,
+      "loss": 2.1268,
+      "step": 78
+    },
+    {
+      "epoch": 0.00816938548641452,
+      "grad_norm": 1.1079933643341064,
+      "learning_rate": 0.00019928608380755304,
+      "loss": 2.4634,
+      "step": 79
+    },
+    {
+      "epoch": 0.008272795429280526,
+      "grad_norm": 0.7377402782440186,
+      "learning_rate": 0.0001992757371960683,
+      "loss": 2.0904,
+      "step": 80
+    },
+    {
+      "epoch": 0.008376205372146532,
+      "grad_norm": 1.0623070001602173,
+      "learning_rate": 0.00019926539058458357,
+      "loss": 2.0502,
+      "step": 81
+    },
+    {
+      "epoch": 0.008479615315012538,
+      "grad_norm": 1.7848681211471558,
+      "learning_rate": 0.00019925504397309883,
+      "loss": 2.0149,
+      "step": 82
+    },
+    {
+      "epoch": 0.008583025257878544,
+      "grad_norm": 1.0376352071762085,
+      "learning_rate": 0.0001992446973616141,
+      "loss": 1.8225,
+      "step": 83
+    },
+    {
+      "epoch": 0.008686435200744552,
+      "grad_norm": 1.0023243427276611,
+      "learning_rate": 0.00019923435075012936,
+      "loss": 2.3898,
+      "step": 84
+    },
+    {
+      "epoch": 0.008789845143610558,
+      "grad_norm": 0.7046536803245544,
+      "learning_rate": 0.0001992240041386446,
+      "loss": 2.2428,
+      "step": 85
+    },
+    {
+      "epoch": 0.008893255086476565,
+      "grad_norm": 0.605064332485199,
+      "learning_rate": 0.00019921365752715987,
+      "loss": 2.0025,
+      "step": 86
+    },
+    {
+      "epoch": 0.008996665029342571,
+      "grad_norm": 0.8229315876960754,
+      "learning_rate": 0.00019920331091567513,
+      "loss": 1.5395,
+      "step": 87
+    },
+    {
+      "epoch": 0.009100074972208577,
+      "grad_norm": 0.5166494250297546,
+      "learning_rate": 0.0001991929643041904,
+      "loss": 2.0568,
+      "step": 88
+    },
+    {
+      "epoch": 0.009203484915074585,
+      "grad_norm": 1.8681402206420898,
+      "learning_rate": 0.00019918261769270566,
+      "loss": 2.483,
+      "step": 89
+    },
+    {
+      "epoch": 0.009306894857940591,
+      "grad_norm": 0.937924861907959,
+      "learning_rate": 0.00019917227108122093,
+      "loss": 2.0965,
+      "step": 90
+    },
+    {
+      "epoch": 0.009410304800806598,
+      "grad_norm": 1.0499122142791748,
+      "learning_rate": 0.00019916192446973617,
+      "loss": 1.8826,
+      "step": 91
+    },
+    {
+      "epoch": 0.009513714743672604,
+      "grad_norm": 0.8156949877738953,
+      "learning_rate": 0.00019915157785825143,
+      "loss": 2.261,
+      "step": 92
+    },
+    {
+      "epoch": 0.00961712468653861,
+      "grad_norm": 1.9210288524627686,
+      "learning_rate": 0.0001991412312467667,
+      "loss": 2.4876,
+      "step": 93
+    },
+    {
+      "epoch": 0.009720534629404616,
+      "grad_norm": 1.3630799055099487,
+      "learning_rate": 0.00019913088463528197,
+      "loss": 1.7244,
+      "step": 94
+    },
+    {
+      "epoch": 0.009823944572270624,
+      "grad_norm": 1.2075848579406738,
+      "learning_rate": 0.00019912053802379723,
+      "loss": 1.4987,
+      "step": 95
+    },
+    {
+      "epoch": 0.00992735451513663,
+      "grad_norm": 1.182338833808899,
+      "learning_rate": 0.0001991101914123125,
+      "loss": 2.0565,
+      "step": 96
+    },
+    {
+      "epoch": 0.010030764458002637,
+      "grad_norm": 1.4979480504989624,
+      "learning_rate": 0.00019909984480082773,
+      "loss": 2.0842,
+      "step": 97
+    },
+    {
+      "epoch": 0.010134174400868643,
+      "grad_norm": 0.6272477507591248,
+      "learning_rate": 0.000199089498189343,
+      "loss": 2.2594,
+      "step": 98
+    },
+    {
+      "epoch": 0.01023758434373465,
+      "grad_norm": 1.4160089492797852,
+      "learning_rate": 0.00019907915157785827,
+      "loss": 2.2122,
+      "step": 99
+    },
+    {
+      "epoch": 0.010340994286600657,
+      "grad_norm": 0.9657629132270813,
+      "learning_rate": 0.00019906880496637353,
+      "loss": 1.6975,
+      "step": 100
+    },
+    {
+      "epoch": 0.010444404229466664,
+      "grad_norm": 0.8677767515182495,
+      "learning_rate": 0.0001990584583548888,
+      "loss": 2.3514,
+      "step": 101
+    },
+    {
+      "epoch": 0.01054781417233267,
+      "grad_norm": 1.6429847478866577,
+      "learning_rate": 0.00019904811174340406,
+      "loss": 1.798,
+      "step": 102
+    },
+    {
+      "epoch": 0.010651224115198676,
+      "grad_norm": 0.6432476043701172,
+      "learning_rate": 0.0001990377651319193,
+      "loss": 1.7637,
+      "step": 103
+    },
+    {
+      "epoch": 0.010754634058064682,
+      "grad_norm": 0.9115468263626099,
+      "learning_rate": 0.00019902741852043457,
+      "loss": 1.8993,
+      "step": 104
+    },
+    {
+      "epoch": 0.01085804400093069,
+      "grad_norm": 1.1197313070297241,
+      "learning_rate": 0.00019901707190894983,
+      "loss": 1.5482,
+      "step": 105
+    },
+    {
+      "epoch": 0.010961453943796697,
+      "grad_norm": 0.8248776197433472,
+      "learning_rate": 0.0001990067252974651,
+      "loss": 2.2015,
+      "step": 106
+    },
+    {
+      "epoch": 0.011064863886662703,
+      "grad_norm": 0.7735565304756165,
+      "learning_rate": 0.00019899637868598036,
+      "loss": 1.4849,
+      "step": 107
+    },
+    {
+      "epoch": 0.011168273829528709,
+      "grad_norm": 1.956572413444519,
+      "learning_rate": 0.00019898603207449563,
+      "loss": 2.1429,
+      "step": 108
+    },
+    {
+      "epoch": 0.011271683772394715,
+      "grad_norm": 0.8068015575408936,
+      "learning_rate": 0.00019897568546301087,
+      "loss": 1.4472,
+      "step": 109
+    },
+    {
+      "epoch": 0.011375093715260721,
+      "grad_norm": 1.3157227039337158,
+      "learning_rate": 0.00019896533885152613,
+      "loss": 1.7154,
+      "step": 110
+    },
+    {
+      "epoch": 0.01147850365812673,
+      "grad_norm": 0.8677852153778076,
+      "learning_rate": 0.0001989549922400414,
+      "loss": 1.5613,
+      "step": 111
+    },
+    {
+      "epoch": 0.011581913600992736,
+      "grad_norm": 1.2042537927627563,
+      "learning_rate": 0.00019894464562855666,
+      "loss": 1.9679,
+      "step": 112
+    },
+    {
+      "epoch": 0.011685323543858742,
+      "grad_norm": 0.7814430594444275,
+      "learning_rate": 0.00019893429901707193,
+      "loss": 1.6496,
+      "step": 113
+    },
+    {
+      "epoch": 0.011788733486724748,
+      "grad_norm": 0.8422583341598511,
+      "learning_rate": 0.0001989239524055872,
+      "loss": 1.672,
+      "step": 114
+    },
+    {
+      "epoch": 0.011892143429590754,
+      "grad_norm": 0.829519510269165,
+      "learning_rate": 0.00019891360579410243,
+      "loss": 1.9504,
+      "step": 115
+    },
+    {
+      "epoch": 0.011995553372456762,
+      "grad_norm": 0.7655951976776123,
+      "learning_rate": 0.0001989032591826177,
+      "loss": 1.779,
+      "step": 116
+    },
+    {
+      "epoch": 0.012098963315322769,
+      "grad_norm": 1.0722001791000366,
+      "learning_rate": 0.00019889291257113296,
+      "loss": 1.948,
+      "step": 117
+    },
+    {
+      "epoch": 0.012202373258188775,
+      "grad_norm": 0.9230548143386841,
+      "learning_rate": 0.00019888256595964823,
+      "loss": 1.0467,
+      "step": 118
+    },
+    {
+      "epoch": 0.012305783201054781,
+      "grad_norm": 1.7951265573501587,
+      "learning_rate": 0.0001988722193481635,
+      "loss": 1.9929,
+      "step": 119
+    },
+    {
+      "epoch": 0.012409193143920787,
+      "grad_norm": 1.0556249618530273,
+      "learning_rate": 0.00019886187273667876,
+      "loss": 1.8817,
+      "step": 120
+    },
+    {
+      "epoch": 0.012512603086786795,
+      "grad_norm": 0.8999210000038147,
+      "learning_rate": 0.000198851526125194,
+      "loss": 2.1797,
+      "step": 121
+    },
+    {
+      "epoch": 0.012616013029652802,
+      "grad_norm": 0.6396268606185913,
+      "learning_rate": 0.00019884117951370927,
+      "loss": 1.8188,
+      "step": 122
+    },
+    {
+      "epoch": 0.012719422972518808,
+      "grad_norm": 1.6357016563415527,
+      "learning_rate": 0.00019883083290222453,
+      "loss": 2.4962,
+      "step": 123
+    },
+    {
+      "epoch": 0.012822832915384814,
+      "grad_norm": 1.6592292785644531,
+      "learning_rate": 0.0001988204862907398,
+      "loss": 1.768,
+      "step": 124
+    },
+    {
+      "epoch": 0.01292624285825082,
+      "grad_norm": 1.0383470058441162,
+      "learning_rate": 0.00019881013967925506,
+      "loss": 1.5453,
+      "step": 125
+    },
+    {
+      "epoch": 0.013029652801116827,
+      "grad_norm": 0.6433761119842529,
+      "learning_rate": 0.00019879979306777033,
+      "loss": 1.7572,
+      "step": 126
+    },
+    {
+      "epoch": 0.013133062743982835,
+      "grad_norm": 0.8064449429512024,
+      "learning_rate": 0.00019878944645628557,
+      "loss": 1.6774,
+      "step": 127
+    },
+    {
+      "epoch": 0.01323647268684884,
+      "grad_norm": 1.010185956954956,
+      "learning_rate": 0.00019877909984480083,
+      "loss": 1.9185,
+      "step": 128
+    },
+    {
+      "epoch": 0.013339882629714847,
+      "grad_norm": 0.8461198806762695,
+      "learning_rate": 0.0001987687532333161,
+      "loss": 1.7703,
+      "step": 129
+    },
+    {
+      "epoch": 0.013443292572580853,
+      "grad_norm": 1.4564467668533325,
+      "learning_rate": 0.00019875840662183136,
+      "loss": 2.3164,
+      "step": 130
+    },
+    {
+      "epoch": 0.01354670251544686,
+      "grad_norm": 0.7336029410362244,
+      "learning_rate": 0.00019874806001034663,
+      "loss": 1.8283,
+      "step": 131
+    },
+    {
+      "epoch": 0.013650112458312868,
+      "grad_norm": 2.1804862022399902,
+      "learning_rate": 0.0001987377133988619,
+      "loss": 1.9357,
+      "step": 132
+    },
+    {
+      "epoch": 0.013753522401178874,
+      "grad_norm": 1.1863375902175903,
+      "learning_rate": 0.00019872736678737713,
+      "loss": 1.8582,
+      "step": 133
+    },
+    {
+      "epoch": 0.01385693234404488,
+      "grad_norm": 1.062116265296936,
+      "learning_rate": 0.0001987170201758924,
+      "loss": 2.0296,
+      "step": 134
+    },
+    {
+      "epoch": 0.013960342286910886,
+      "grad_norm": 1.7113271951675415,
+      "learning_rate": 0.00019870667356440766,
+      "loss": 1.8241,
+      "step": 135
+    },
+    {
+      "epoch": 0.014063752229776893,
+      "grad_norm": 1.9041301012039185,
+      "learning_rate": 0.00019869632695292293,
+      "loss": 2.0608,
+      "step": 136
+    },
+    {
+      "epoch": 0.0141671621726429,
+      "grad_norm": 1.0523148775100708,
+      "learning_rate": 0.0001986859803414382,
+      "loss": 1.9497,
+      "step": 137
+    },
+    {
+      "epoch": 0.014270572115508907,
+      "grad_norm": 0.8391201496124268,
+      "learning_rate": 0.00019867563372995346,
+      "loss": 1.5612,
+      "step": 138
+    },
+    {
+      "epoch": 0.014373982058374913,
+      "grad_norm": 1.2530022859573364,
+      "learning_rate": 0.0001986652871184687,
+      "loss": 1.9835,
+      "step": 139
+    },
+    {
+      "epoch": 0.01447739200124092,
+      "grad_norm": 0.7818213701248169,
+      "learning_rate": 0.00019865494050698396,
+      "loss": 1.7641,
+      "step": 140
+    },
+    {
+      "epoch": 0.014580801944106925,
+      "grad_norm": 1.1031138896942139,
+      "learning_rate": 0.00019864459389549923,
+      "loss": 2.028,
+      "step": 141
+    },
+    {
+      "epoch": 0.014684211886972932,
+      "grad_norm": 0.6917052268981934,
+      "learning_rate": 0.0001986342472840145,
+      "loss": 1.7597,
+      "step": 142
+    },
+    {
+      "epoch": 0.01478762182983894,
+      "grad_norm": 1.2647300958633423,
+      "learning_rate": 0.00019862390067252976,
+      "loss": 2.0946,
+      "step": 143
+    },
+    {
+      "epoch": 0.014891031772704946,
+      "grad_norm": 0.819813072681427,
+      "learning_rate": 0.00019861355406104503,
+      "loss": 1.786,
+      "step": 144
+    },
+    {
+      "epoch": 0.014994441715570952,
+      "grad_norm": 2.5735700130462646,
+      "learning_rate": 0.00019860320744956027,
+      "loss": 1.8916,
+      "step": 145
+    },
+    {
+      "epoch": 0.015097851658436958,
+      "grad_norm": 1.1350979804992676,
+      "learning_rate": 0.00019859286083807553,
+      "loss": 2.223,
+      "step": 146
+    },
+    {
+      "epoch": 0.015201261601302965,
+      "grad_norm": 0.9886088967323303,
+      "learning_rate": 0.0001985825142265908,
+      "loss": 1.3458,
+      "step": 147
+    },
+    {
+      "epoch": 0.015304671544168973,
+      "grad_norm": 0.7377306222915649,
+      "learning_rate": 0.00019857216761510606,
+      "loss": 1.5796,
+      "step": 148
+    },
+    {
+      "epoch": 0.015408081487034979,
+      "grad_norm": 1.4361933469772339,
+      "learning_rate": 0.00019856182100362133,
+      "loss": 2.2587,
+      "step": 149
+    },
+    {
+      "epoch": 0.015511491429900985,
+      "grad_norm": 1.8182250261306763,
+      "learning_rate": 0.0001985514743921366,
+      "loss": 2.4222,
+      "step": 150
+    },
+    {
+      "epoch": 0.015614901372766991,
+      "grad_norm": 1.7325825691223145,
+      "learning_rate": 0.00019854112778065183,
+      "loss": 1.5374,
+      "step": 151
+    },
+    {
+      "epoch": 0.015718311315632998,
+      "grad_norm": 0.7336366176605225,
+      "learning_rate": 0.0001985307811691671,
+      "loss": 1.9757,
+      "step": 152
+    },
+    {
+      "epoch": 0.015821721258499004,
+      "grad_norm": 1.2688993215560913,
+      "learning_rate": 0.00019852043455768236,
+      "loss": 2.3533,
+      "step": 153
+    },
+    {
+      "epoch": 0.01592513120136501,
+      "grad_norm": 1.637938141822815,
+      "learning_rate": 0.00019851008794619763,
+      "loss": 2.4375,
+      "step": 154
+    },
+    {
+      "epoch": 0.016028541144231016,
+      "grad_norm": 0.6742689609527588,
+      "learning_rate": 0.0001984997413347129,
+      "loss": 2.0748,
+      "step": 155
+    },
+    {
+      "epoch": 0.016131951087097026,
+      "grad_norm": 1.2148876190185547,
+      "learning_rate": 0.00019848939472322816,
+      "loss": 1.8236,
+      "step": 156
+    },
+    {
+      "epoch": 0.016235361029963032,
+      "grad_norm": 2.3418567180633545,
+      "learning_rate": 0.0001984790481117434,
+      "loss": 2.413,
+      "step": 157
+    },
+    {
+      "epoch": 0.01633877097282904,
+      "grad_norm": 0.7256979942321777,
+      "learning_rate": 0.00019846870150025866,
+      "loss": 1.6911,
+      "step": 158
+    },
+    {
+      "epoch": 0.016442180915695045,
+      "grad_norm": 1.3450243473052979,
+      "learning_rate": 0.00019845835488877393,
+      "loss": 1.7432,
+      "step": 159
+    },
+    {
+      "epoch": 0.01654559085856105,
+      "grad_norm": 1.6073719263076782,
+      "learning_rate": 0.0001984480082772892,
+      "loss": 2.2007,
+      "step": 160
+    },
+    {
+      "epoch": 0.016649000801427057,
+      "grad_norm": 1.153110146522522,
+      "learning_rate": 0.00019843766166580446,
+      "loss": 1.715,
+      "step": 161
+    },
+    {
+      "epoch": 0.016752410744293064,
+      "grad_norm": 0.807807445526123,
+      "learning_rate": 0.00019842731505431973,
+      "loss": 1.5903,
+      "step": 162
+    },
+    {
+      "epoch": 0.01685582068715907,
+      "grad_norm": 1.3365882635116577,
+      "learning_rate": 0.00019841696844283496,
+      "loss": 1.978,
+      "step": 163
+    },
+    {
+      "epoch": 0.016959230630025076,
+      "grad_norm": 0.7147388458251953,
+      "learning_rate": 0.00019840662183135023,
+      "loss": 1.6717,
+      "step": 164
+    },
+    {
+      "epoch": 0.017062640572891082,
+      "grad_norm": 0.6310922503471375,
+      "learning_rate": 0.0001983962752198655,
+      "loss": 1.9129,
+      "step": 165
+    },
+    {
+      "epoch": 0.01716605051575709,
+      "grad_norm": 0.8110767602920532,
+      "learning_rate": 0.00019838592860838076,
+      "loss": 1.6107,
+      "step": 166
+    },
+    {
+      "epoch": 0.017269460458623098,
+      "grad_norm": 0.9394704103469849,
+      "learning_rate": 0.00019837558199689603,
+      "loss": 2.2022,
+      "step": 167
+    },
+    {
+      "epoch": 0.017372870401489104,
+      "grad_norm": 0.6974437236785889,
+      "learning_rate": 0.0001983652353854113,
+      "loss": 2.0294,
+      "step": 168
+    },
+    {
+      "epoch": 0.01747628034435511,
+      "grad_norm": 0.9328587055206299,
+      "learning_rate": 0.00019835488877392653,
+      "loss": 1.9777,
+      "step": 169
+    },
+    {
+      "epoch": 0.017579690287221117,
+      "grad_norm": 1.6323524713516235,
+      "learning_rate": 0.0001983445421624418,
+      "loss": 2.2248,
+      "step": 170
+    },
+    {
+      "epoch": 0.017683100230087123,
+      "grad_norm": 1.306799292564392,
+      "learning_rate": 0.00019833419555095706,
+      "loss": 1.5457,
+      "step": 171
+    },
+    {
+      "epoch": 0.01778651017295313,
+      "grad_norm": 1.1161526441574097,
+      "learning_rate": 0.00019832384893947233,
+      "loss": 1.9586,
+      "step": 172
+    },
+    {
+      "epoch": 0.017889920115819136,
+      "grad_norm": 0.6718397736549377,
+      "learning_rate": 0.0001983135023279876,
+      "loss": 1.901,
+      "step": 173
+    },
+    {
+      "epoch": 0.017993330058685142,
+      "grad_norm": 0.7336934208869934,
+      "learning_rate": 0.00019830315571650286,
+      "loss": 1.7648,
+      "step": 174
+    },
+    {
+      "epoch": 0.018096740001551148,
+      "grad_norm": 0.5712825655937195,
+      "learning_rate": 0.0001982928091050181,
+      "loss": 1.8274,
+      "step": 175
+    },
+    {
+      "epoch": 0.018200149944417154,
+      "grad_norm": 0.7899829149246216,
+      "learning_rate": 0.00019828246249353336,
+      "loss": 1.8218,
+      "step": 176
+    },
+    {
+      "epoch": 0.01830355988728316,
+      "grad_norm": 2.2078018188476562,
+      "learning_rate": 0.00019827211588204863,
+      "loss": 2.0636,
+      "step": 177
+    },
+    {
+      "epoch": 0.01840696983014917,
+      "grad_norm": 1.9544669389724731,
+      "learning_rate": 0.0001982617692705639,
+      "loss": 1.9228,
+      "step": 178
+    },
+    {
+      "epoch": 0.018510379773015177,
+      "grad_norm": 1.212548017501831,
+      "learning_rate": 0.00019825142265907916,
+      "loss": 1.407,
+      "step": 179
+    },
+    {
+      "epoch": 0.018613789715881183,
+      "grad_norm": 1.0474966764450073,
+      "learning_rate": 0.00019824107604759442,
+      "loss": 1.8511,
+      "step": 180
+    },
+    {
+      "epoch": 0.01871719965874719,
+      "grad_norm": 0.9905005097389221,
+      "learning_rate": 0.00019823072943610966,
+      "loss": 1.8197,
+      "step": 181
+    },
+    {
+      "epoch": 0.018820609601613195,
+      "grad_norm": 1.660206913948059,
+      "learning_rate": 0.00019822038282462493,
+      "loss": 2.2478,
+      "step": 182
+    },
+    {
+      "epoch": 0.0189240195444792,
+      "grad_norm": 0.8422425985336304,
+      "learning_rate": 0.0001982100362131402,
+      "loss": 1.6032,
+      "step": 183
+    },
+    {
+      "epoch": 0.019027429487345208,
+      "grad_norm": 0.9838465452194214,
+      "learning_rate": 0.00019819968960165546,
+      "loss": 2.0711,
+      "step": 184
+    },
+    {
+      "epoch": 0.019130839430211214,
+      "grad_norm": 1.9396435022354126,
+      "learning_rate": 0.00019818934299017073,
+      "loss": 2.2734,
+      "step": 185
+    },
+    {
+      "epoch": 0.01923424937307722,
+      "grad_norm": 1.4743622541427612,
+      "learning_rate": 0.000198178996378686,
+      "loss": 1.7227,
+      "step": 186
+    },
+    {
+      "epoch": 0.019337659315943227,
+      "grad_norm": 1.6646116971969604,
+      "learning_rate": 0.00019816864976720123,
+      "loss": 2.3045,
+      "step": 187
+    },
+    {
+      "epoch": 0.019441069258809233,
+      "grad_norm": 0.8518794178962708,
+      "learning_rate": 0.0001981583031557165,
+      "loss": 1.9421,
+      "step": 188
+    },
+    {
+      "epoch": 0.019544479201675242,
+      "grad_norm": 1.869166374206543,
+      "learning_rate": 0.00019814795654423176,
+      "loss": 1.6767,
+      "step": 189
+    },
+    {
+      "epoch": 0.01964788914454125,
+      "grad_norm": 0.9898916482925415,
+      "learning_rate": 0.00019813760993274703,
+      "loss": 1.7501,
+      "step": 190
+    },
+    {
+      "epoch": 0.019751299087407255,
+      "grad_norm": 0.7842496037483215,
+      "learning_rate": 0.0001981272633212623,
+      "loss": 2.2322,
+      "step": 191
+    },
+    {
+      "epoch": 0.01985470903027326,
+      "grad_norm": 1.4266057014465332,
+      "learning_rate": 0.00019811691670977756,
+      "loss": 1.9491,
+      "step": 192
+    },
+    {
+      "epoch": 0.019958118973139267,
+      "grad_norm": 0.8259785771369934,
+      "learning_rate": 0.00019810657009829282,
+      "loss": 1.6481,
+      "step": 193
+    },
+    {
+      "epoch": 0.020061528916005274,
+      "grad_norm": 0.8917863965034485,
+      "learning_rate": 0.00019809622348680806,
+      "loss": 1.7006,
+      "step": 194
+    },
+    {
+      "epoch": 0.02016493885887128,
+      "grad_norm": 1.5615893602371216,
+      "learning_rate": 0.00019808587687532333,
+      "loss": 1.8371,
+      "step": 195
+    },
+    {
+      "epoch": 0.020268348801737286,
+      "grad_norm": 0.999208927154541,
+      "learning_rate": 0.0001980755302638386,
+      "loss": 1.9719,
+      "step": 196
+    },
+    {
+      "epoch": 0.020371758744603292,
+      "grad_norm": 1.0224575996398926,
+      "learning_rate": 0.00019806518365235386,
+      "loss": 1.9708,
+      "step": 197
+    },
+    {
+      "epoch": 0.0204751686874693,
+      "grad_norm": 0.6324958205223083,
+      "learning_rate": 0.00019805483704086912,
+      "loss": 2.2463,
+      "step": 198
+    },
+    {
+      "epoch": 0.02057857863033531,
+      "grad_norm": 1.042873501777649,
+      "learning_rate": 0.0001980444904293844,
+      "loss": 1.1634,
+      "step": 199
+    },
+    {
+      "epoch": 0.020681988573201315,
+      "grad_norm": 1.5669771432876587,
+      "learning_rate": 0.00019803414381789966,
+      "loss": 1.8985,
+      "step": 200
+    },
+    {
+      "epoch": 0.02078539851606732,
+      "grad_norm": 1.6116214990615845,
+      "learning_rate": 0.00019802379720641492,
+      "loss": 1.9126,
+      "step": 201
+    },
+    {
+      "epoch": 0.020888808458933327,
+      "grad_norm": 1.1320672035217285,
+      "learning_rate": 0.00019801345059493016,
+      "loss": 1.96,
+      "step": 202
+    },
+    {
+      "epoch": 0.020992218401799333,
+      "grad_norm": 1.2705937623977661,
+      "learning_rate": 0.00019800310398344542,
+      "loss": 1.8843,
+      "step": 203
+    },
+    {
+      "epoch": 0.02109562834466534,
+      "grad_norm": 1.1148872375488281,
+      "learning_rate": 0.0001979927573719607,
+      "loss": 1.5164,
+      "step": 204
+    },
+    {
+      "epoch": 0.021199038287531346,
+      "grad_norm": 2.2498512268066406,
+      "learning_rate": 0.00019798241076047596,
+      "loss": 2.4194,
+      "step": 205
+    },
+    {
+      "epoch": 0.021302448230397352,
+      "grad_norm": 2.049044609069824,
+      "learning_rate": 0.00019797206414899122,
+      "loss": 1.8792,
+      "step": 206
+    },
+    {
+      "epoch": 0.02140585817326336,
+      "grad_norm": 0.7835369110107422,
+      "learning_rate": 0.0001979617175375065,
+      "loss": 1.7925,
+      "step": 207
+    },
+    {
+      "epoch": 0.021509268116129365,
+      "grad_norm": 0.814444363117218,
+      "learning_rate": 0.00019795137092602175,
+      "loss": 2.0929,
+      "step": 208
+    },
+    {
+      "epoch": 0.02161267805899537,
+      "grad_norm": 0.6900445222854614,
+      "learning_rate": 0.00019794102431453702,
+      "loss": 1.6241,
+      "step": 209
+    },
+    {
+      "epoch": 0.02171608800186138,
+      "grad_norm": 0.9963400959968567,
+      "learning_rate": 0.00019793067770305226,
+      "loss": 1.5765,
+      "step": 210
+    },
+    {
+      "epoch": 0.021819497944727387,
+      "grad_norm": 0.8830106258392334,
+      "learning_rate": 0.00019792033109156752,
+      "loss": 2.2472,
+      "step": 211
+    },
+    {
+      "epoch": 0.021922907887593393,
+      "grad_norm": 1.1107029914855957,
+      "learning_rate": 0.0001979099844800828,
+      "loss": 1.9912,
+      "step": 212
+    },
+    {
+      "epoch": 0.0220263178304594,
+      "grad_norm": 1.394497036933899,
+      "learning_rate": 0.00019789963786859805,
+      "loss": 2.049,
+      "step": 213
+    },
+    {
+      "epoch": 0.022129727773325406,
+      "grad_norm": 1.3322755098342896,
+      "learning_rate": 0.00019788929125711332,
+      "loss": 1.4923,
+      "step": 214
+    },
+    {
+      "epoch": 0.022233137716191412,
+      "grad_norm": 1.0636125802993774,
+      "learning_rate": 0.00019787894464562858,
+      "loss": 2.1262,
+      "step": 215
+    },
+    {
+      "epoch": 0.022336547659057418,
+      "grad_norm": 0.9886292219161987,
+      "learning_rate": 0.00019786859803414385,
+      "loss": 1.6698,
+      "step": 216
+    },
+    {
+      "epoch": 0.022439957601923424,
+      "grad_norm": 0.9913386106491089,
+      "learning_rate": 0.0001978582514226591,
+      "loss": 1.9735,
+      "step": 217
+    },
+    {
+      "epoch": 0.02254336754478943,
+      "grad_norm": 0.4542105197906494,
+      "learning_rate": 0.00019784790481117435,
+      "loss": 1.9013,
+      "step": 218
+    },
+    {
+      "epoch": 0.022646777487655437,
+      "grad_norm": 0.6689864993095398,
+      "learning_rate": 0.00019783755819968962,
+      "loss": 1.9576,
+      "step": 219
+    },
+    {
+      "epoch": 0.022750187430521443,
+      "grad_norm": 0.8507350087165833,
+      "learning_rate": 0.00019782721158820489,
+      "loss": 1.8583,
+      "step": 220
+    },
+    {
+      "epoch": 0.022853597373387453,
+      "grad_norm": 1.187589168548584,
+      "learning_rate": 0.00019781686497672015,
+      "loss": 1.6836,
+      "step": 221
+    },
+    {
+      "epoch": 0.02295700731625346,
+      "grad_norm": 1.2565590143203735,
+      "learning_rate": 0.00019780651836523542,
+      "loss": 2.155,
+      "step": 222
+    },
+    {
+      "epoch": 0.023060417259119465,
+      "grad_norm": 0.8139156699180603,
+      "learning_rate": 0.00019779617175375065,
+      "loss": 1.9202,
+      "step": 223
+    },
+    {
+      "epoch": 0.02316382720198547,
+      "grad_norm": 0.6143559217453003,
+      "learning_rate": 0.00019778582514226592,
+      "loss": 2.0392,
+      "step": 224
+    },
+    {
+      "epoch": 0.023267237144851478,
+      "grad_norm": 0.992732584476471,
+      "learning_rate": 0.00019777547853078119,
+      "loss": 1.8389,
+      "step": 225
+    },
+    {
+      "epoch": 0.023370647087717484,
+      "grad_norm": 0.8133312463760376,
+      "learning_rate": 0.00019776513191929645,
+      "loss": 2.2321,
+      "step": 226
+    },
+    {
+      "epoch": 0.02347405703058349,
+      "grad_norm": 0.5798576474189758,
+      "learning_rate": 0.00019775478530781172,
+      "loss": 1.905,
+      "step": 227
+    },
+    {
+      "epoch": 0.023577466973449496,
+      "grad_norm": 0.9979951977729797,
+      "learning_rate": 0.00019774443869632698,
+      "loss": 1.7011,
+      "step": 228
+    },
+    {
+      "epoch": 0.023680876916315503,
+      "grad_norm": 1.0430858135223389,
+      "learning_rate": 0.00019773409208484222,
+      "loss": 1.8414,
+      "step": 229
+    },
+    {
+      "epoch": 0.02378428685918151,
+      "grad_norm": 0.9151175022125244,
+      "learning_rate": 0.0001977237454733575,
+      "loss": 1.711,
+      "step": 230
+    },
+    {
+      "epoch": 0.023887696802047515,
+      "grad_norm": 1.0341556072235107,
+      "learning_rate": 0.00019771339886187275,
+      "loss": 1.5063,
+      "step": 231
+    },
+    {
+      "epoch": 0.023991106744913525,
+      "grad_norm": 1.1679956912994385,
+      "learning_rate": 0.00019770305225038802,
+      "loss": 2.314,
+      "step": 232
+    },
+    {
+      "epoch": 0.02409451668777953,
+      "grad_norm": 0.9437854886054993,
+      "learning_rate": 0.00019769270563890328,
+      "loss": 1.9386,
+      "step": 233
+    },
+    {
+      "epoch": 0.024197926630645537,
+      "grad_norm": 0.7488498687744141,
+      "learning_rate": 0.00019768235902741855,
+      "loss": 2.3507,
+      "step": 234
+    },
+    {
+      "epoch": 0.024301336573511544,
+      "grad_norm": 1.0003606081008911,
+      "learning_rate": 0.0001976720124159338,
+      "loss": 1.7557,
+      "step": 235
+    },
+    {
+      "epoch": 0.02440474651637755,
+      "grad_norm": 1.38227117061615,
+      "learning_rate": 0.00019766166580444905,
+      "loss": 1.7559,
+      "step": 236
+    },
+    {
+      "epoch": 0.024508156459243556,
+      "grad_norm": 0.732622504234314,
+      "learning_rate": 0.00019765131919296432,
+      "loss": 2.0015,
+      "step": 237
+    },
+    {
+      "epoch": 0.024611566402109562,
+      "grad_norm": 0.7456156015396118,
+      "learning_rate": 0.00019764097258147958,
+      "loss": 2.6102,
+      "step": 238
+    },
+    {
+      "epoch": 0.02471497634497557,
+      "grad_norm": 1.0653510093688965,
+      "learning_rate": 0.00019763062596999485,
+      "loss": 1.9302,
+      "step": 239
+    },
+    {
+      "epoch": 0.024818386287841575,
+      "grad_norm": 1.5663673877716064,
+      "learning_rate": 0.00019762027935851012,
+      "loss": 1.9772,
+      "step": 240
+    },
+    {
+      "epoch": 0.02492179623070758,
+      "grad_norm": 7.048519134521484,
+      "learning_rate": 0.00019760993274702535,
+      "loss": 2.0148,
+      "step": 241
+    },
+    {
+      "epoch": 0.02502520617357359,
+      "grad_norm": 1.1092578172683716,
+      "learning_rate": 0.00019759958613554062,
+      "loss": 1.4961,
+      "step": 242
+    },
+    {
+      "epoch": 0.025128616116439597,
+      "grad_norm": 1.3526990413665771,
+      "learning_rate": 0.00019758923952405588,
+      "loss": 2.0189,
+      "step": 243
+    },
+    {
+      "epoch": 0.025232026059305603,
+      "grad_norm": 0.9486410021781921,
+      "learning_rate": 0.00019757889291257115,
+      "loss": 2.1304,
+      "step": 244
+    },
+    {
+      "epoch": 0.02533543600217161,
+      "grad_norm": 1.422751545906067,
+      "learning_rate": 0.00019756854630108642,
+      "loss": 2.103,
+      "step": 245
+    },
+    {
+      "epoch": 0.025438845945037616,
+      "grad_norm": 1.5062353610992432,
+      "learning_rate": 0.00019755819968960168,
+      "loss": 1.7464,
+      "step": 246
+    },
+    {
+      "epoch": 0.025542255887903622,
+      "grad_norm": 1.2952933311462402,
+      "learning_rate": 0.00019754785307811692,
+      "loss": 1.6458,
+      "step": 247
+    },
+    {
+      "epoch": 0.025645665830769628,
+      "grad_norm": 0.8431225419044495,
+      "learning_rate": 0.00019753750646663219,
+      "loss": 1.6088,
+      "step": 248
+    },
+    {
+      "epoch": 0.025749075773635634,
+      "grad_norm": 0.7888919711112976,
+      "learning_rate": 0.00019752715985514745,
+      "loss": 1.9032,
+      "step": 249
+    },
+    {
+      "epoch": 0.02585248571650164,
+      "grad_norm": 0.90091872215271,
+      "learning_rate": 0.00019751681324366272,
+      "loss": 1.3846,
+      "step": 250
+    },
+    {
+      "epoch": 0.025955895659367647,
+      "grad_norm": 1.8097118139266968,
+      "learning_rate": 0.00019750646663217798,
+      "loss": 2.2661,
+      "step": 251
+    },
+    {
+      "epoch": 0.026059305602233653,
+      "grad_norm": 0.6663314700126648,
+      "learning_rate": 0.00019749612002069325,
+      "loss": 1.7202,
+      "step": 252
+    },
+    {
+      "epoch": 0.026162715545099663,
+      "grad_norm": 5.989753246307373,
+      "learning_rate": 0.00019748577340920849,
+      "loss": 1.879,
+      "step": 253
+    },
+    {
+      "epoch": 0.02626612548796567,
+      "grad_norm": 1.0955287218093872,
+      "learning_rate": 0.00019747542679772375,
+      "loss": 1.9042,
+      "step": 254
+    },
+    {
+      "epoch": 0.026369535430831675,
+      "grad_norm": 1.4397646188735962,
+      "learning_rate": 0.00019746508018623902,
+      "loss": 2.2163,
+      "step": 255
+    },
+    {
+      "epoch": 0.02647294537369768,
+      "grad_norm": 1.1022082567214966,
+      "learning_rate": 0.00019745473357475428,
+      "loss": 1.7654,
+      "step": 256
+    },
+    {
+      "epoch": 0.026576355316563688,
+      "grad_norm": 0.7697210907936096,
+      "learning_rate": 0.00019744438696326955,
+      "loss": 1.4869,
+      "step": 257
+    },
+    {
+      "epoch": 0.026679765259429694,
+      "grad_norm": 1.1588715314865112,
+      "learning_rate": 0.00019743404035178481,
+      "loss": 2.1267,
+      "step": 258
+    },
+    {
+      "epoch": 0.0267831752022957,
+      "grad_norm": 0.8320289254188538,
+      "learning_rate": 0.00019742369374030005,
+      "loss": 1.6688,
+      "step": 259
+    },
+    {
+      "epoch": 0.026886585145161707,
+      "grad_norm": 1.0717252492904663,
+      "learning_rate": 0.00019741334712881532,
+      "loss": 2.0394,
+      "step": 260
+    },
+    {
+      "epoch": 0.026989995088027713,
+      "grad_norm": 0.7327308654785156,
+      "learning_rate": 0.00019740300051733058,
+      "loss": 1.8524,
+      "step": 261
+    },
+    {
+      "epoch": 0.02709340503089372,
+      "grad_norm": 0.9403550624847412,
+      "learning_rate": 0.00019739265390584585,
+      "loss": 2.0664,
+      "step": 262
+    },
+    {
+      "epoch": 0.027196814973759725,
+      "grad_norm": 1.1287577152252197,
+      "learning_rate": 0.00019738230729436112,
+      "loss": 1.9978,
+      "step": 263
+    },
+    {
+      "epoch": 0.027300224916625735,
+      "grad_norm": 3.3969926834106445,
+      "learning_rate": 0.00019737196068287638,
+      "loss": 2.55,
+      "step": 264
+    },
+    {
+      "epoch": 0.02740363485949174,
+      "grad_norm": 1.8311293125152588,
+      "learning_rate": 0.00019736161407139162,
+      "loss": 2.335,
+      "step": 265
+    },
+    {
+      "epoch": 0.027507044802357748,
+      "grad_norm": 0.8647951483726501,
+      "learning_rate": 0.00019735126745990688,
+      "loss": 1.8155,
+      "step": 266
+    },
+    {
+      "epoch": 0.027610454745223754,
+      "grad_norm": 2.12625789642334,
+      "learning_rate": 0.00019734092084842215,
+      "loss": 1.7278,
+      "step": 267
+    },
+    {
+      "epoch": 0.02771386468808976,
+      "grad_norm": 1.0068492889404297,
+      "learning_rate": 0.00019733057423693742,
+      "loss": 1.8327,
+      "step": 268
+    },
+    {
+      "epoch": 0.027817274630955766,
+      "grad_norm": 1.791537880897522,
+      "learning_rate": 0.00019732022762545268,
+      "loss": 2.2064,
+      "step": 269
+    },
+    {
+      "epoch": 0.027920684573821773,
+      "grad_norm": 0.6931219100952148,
+      "learning_rate": 0.00019730988101396795,
+      "loss": 1.6676,
+      "step": 270
+    },
+    {
+      "epoch": 0.02802409451668778,
+      "grad_norm": 1.1089061498641968,
+      "learning_rate": 0.00019729953440248319,
+      "loss": 1.6938,
+      "step": 271
+    },
+    {
+      "epoch": 0.028127504459553785,
+      "grad_norm": 0.8395723104476929,
+      "learning_rate": 0.00019728918779099845,
+      "loss": 2.0785,
+      "step": 272
+    },
+    {
+      "epoch": 0.02823091440241979,
+      "grad_norm": 0.8383250832557678,
+      "learning_rate": 0.00019727884117951372,
+      "loss": 1.9188,
+      "step": 273
+    },
+    {
+      "epoch": 0.0283343243452858,
+      "grad_norm": 1.6984490156173706,
+      "learning_rate": 0.00019726849456802898,
+      "loss": 1.8758,
+      "step": 274
+    },
+    {
+      "epoch": 0.028437734288151807,
+      "grad_norm": 1.8110450506210327,
+      "learning_rate": 0.00019725814795654425,
+      "loss": 2.1544,
+      "step": 275
+    },
+    {
+      "epoch": 0.028541144231017813,
+      "grad_norm": 1.1028351783752441,
+      "learning_rate": 0.0001972478013450595,
+      "loss": 1.8011,
+      "step": 276
+    },
+    {
+      "epoch": 0.02864455417388382,
+      "grad_norm": 1.2571910619735718,
+      "learning_rate": 0.00019723745473357475,
+      "loss": 1.754,
+      "step": 277
+    },
+    {
+      "epoch": 0.028747964116749826,
+      "grad_norm": 0.7787155508995056,
+      "learning_rate": 0.00019722710812209002,
+      "loss": 1.847,
+      "step": 278
+    },
+    {
+      "epoch": 0.028851374059615832,
+      "grad_norm": 1.1525181531906128,
+      "learning_rate": 0.00019721676151060528,
+      "loss": 2.0515,
+      "step": 279
+    },
+    {
+      "epoch": 0.02895478400248184,
+      "grad_norm": 1.8915187120437622,
+      "learning_rate": 0.00019720641489912055,
+      "loss": 2.2906,
+      "step": 280
+    },
+    {
+      "epoch": 0.029058193945347845,
+      "grad_norm": 0.9054998755455017,
+      "learning_rate": 0.00019719606828763581,
+      "loss": 1.8102,
+      "step": 281
+    },
+    {
+      "epoch": 0.02916160388821385,
+      "grad_norm": 1.0367076396942139,
+      "learning_rate": 0.00019718572167615108,
+      "loss": 2.0181,
+      "step": 282
+    },
+    {
+      "epoch": 0.029265013831079857,
+      "grad_norm": 1.2361838817596436,
+      "learning_rate": 0.00019717537506466632,
+      "loss": 1.879,
+      "step": 283
+    },
+    {
+      "epoch": 0.029368423773945863,
+      "grad_norm": 0.7413283586502075,
+      "learning_rate": 0.00019716502845318158,
+      "loss": 1.7919,
+      "step": 284
+    },
+    {
+      "epoch": 0.029471833716811873,
+      "grad_norm": 0.993869960308075,
+      "learning_rate": 0.00019715468184169685,
+      "loss": 1.3424,
+      "step": 285
+    },
+    {
+      "epoch": 0.02957524365967788,
+      "grad_norm": 1.8364124298095703,
+      "learning_rate": 0.00019714433523021211,
+      "loss": 1.9326,
+      "step": 286
+    },
+    {
+      "epoch": 0.029678653602543886,
+      "grad_norm": 1.173508882522583,
+      "learning_rate": 0.00019713398861872738,
+      "loss": 2.6364,
+      "step": 287
+    },
+    {
+      "epoch": 0.029782063545409892,
+      "grad_norm": 0.8887481689453125,
+      "learning_rate": 0.00019712364200724265,
+      "loss": 1.5257,
+      "step": 288
+    },
+    {
+      "epoch": 0.029885473488275898,
+      "grad_norm": 1.2317752838134766,
+      "learning_rate": 0.00019711329539575788,
+      "loss": 2.1392,
+      "step": 289
+    },
+    {
+      "epoch": 0.029988883431141904,
+      "grad_norm": 1.7063419818878174,
+      "learning_rate": 0.00019710294878427315,
+      "loss": 2.3708,
+      "step": 290
+    },
+    {
+      "epoch": 0.03009229337400791,
+      "grad_norm": 0.9618234634399414,
+      "learning_rate": 0.00019709260217278842,
+      "loss": 1.2293,
+      "step": 291
+    },
+    {
+      "epoch": 0.030195703316873917,
+      "grad_norm": 1.4276773929595947,
+      "learning_rate": 0.00019708225556130368,
+      "loss": 1.6762,
+      "step": 292
+    },
+    {
+      "epoch": 0.030299113259739923,
+      "grad_norm": 1.2723143100738525,
+      "learning_rate": 0.00019707190894981895,
+      "loss": 2.0648,
+      "step": 293
+    },
+    {
+      "epoch": 0.03040252320260593,
+      "grad_norm": 3.1224710941314697,
+      "learning_rate": 0.0001970615623383342,
+      "loss": 2.1218,
+      "step": 294
+    },
+    {
+      "epoch": 0.030505933145471936,
+      "grad_norm": 1.720165729522705,
+      "learning_rate": 0.00019705121572684945,
+      "loss": 1.4688,
+      "step": 295
+    },
+    {
+      "epoch": 0.030609343088337945,
+      "grad_norm": 1.38371741771698,
+      "learning_rate": 0.00019704086911536472,
+      "loss": 1.7212,
+      "step": 296
+    },
+    {
+      "epoch": 0.03071275303120395,
+      "grad_norm": 0.995150625705719,
+      "learning_rate": 0.00019703052250387998,
+      "loss": 1.8055,
+      "step": 297
+    },
+    {
+      "epoch": 0.030816162974069958,
+      "grad_norm": 1.070515513420105,
+      "learning_rate": 0.00019702017589239525,
+      "loss": 1.5934,
+      "step": 298
+    },
+    {
+      "epoch": 0.030919572916935964,
+      "grad_norm": 1.0754715204238892,
+      "learning_rate": 0.0001970098292809105,
+      "loss": 1.4899,
+      "step": 299
+    },
+    {
+      "epoch": 0.03102298285980197,
+      "grad_norm": 1.0355829000473022,
+      "learning_rate": 0.00019699948266942578,
+      "loss": 2.0269,
+      "step": 300
+    },
+    {
+      "epoch": 0.031126392802667976,
+      "grad_norm": 1.7430510520935059,
+      "learning_rate": 0.00019698913605794102,
+      "loss": 1.6499,
+      "step": 301
+    },
+    {
+      "epoch": 0.031229802745533983,
+      "grad_norm": 1.3326469659805298,
+      "learning_rate": 0.00019697878944645628,
+      "loss": 2.3119,
+      "step": 302
+    },
+    {
+      "epoch": 0.03133321268839999,
+      "grad_norm": 1.1865754127502441,
+      "learning_rate": 0.00019696844283497155,
+      "loss": 1.9132,
+      "step": 303
+    },
+    {
+      "epoch": 0.031436622631265995,
+      "grad_norm": 0.8858489990234375,
+      "learning_rate": 0.00019695809622348681,
+      "loss": 1.7969,
+      "step": 304
+    },
+    {
+      "epoch": 0.031540032574132,
+      "grad_norm": 1.064029335975647,
+      "learning_rate": 0.00019694774961200208,
+      "loss": 2.0591,
+      "step": 305
+    },
+    {
+      "epoch": 0.03164344251699801,
+      "grad_norm": 0.7958260774612427,
+      "learning_rate": 0.00019693740300051734,
+      "loss": 1.7113,
+      "step": 306
+    },
+    {
+      "epoch": 0.031746852459864014,
+      "grad_norm": 0.9576128125190735,
+      "learning_rate": 0.00019692705638903258,
+      "loss": 1.4719,
+      "step": 307
+    },
+    {
+      "epoch": 0.03185026240273002,
+      "grad_norm": 0.8087986707687378,
+      "learning_rate": 0.00019691670977754785,
+      "loss": 2.1864,
+      "step": 308
+    },
+    {
+      "epoch": 0.031953672345596026,
+      "grad_norm": 1.0684764385223389,
+      "learning_rate": 0.00019690636316606311,
+      "loss": 2.1109,
+      "step": 309
+    },
+    {
+      "epoch": 0.03205708228846203,
+      "grad_norm": 1.1162570714950562,
+      "learning_rate": 0.00019689601655457838,
+      "loss": 1.6932,
+      "step": 310
+    },
+    {
+      "epoch": 0.03216049223132804,
+      "grad_norm": 0.9443684220314026,
+      "learning_rate": 0.00019688566994309365,
+      "loss": 1.9788,
+      "step": 311
+    },
+    {
+      "epoch": 0.03226390217419405,
+      "grad_norm": 0.8843461871147156,
+      "learning_rate": 0.0001968753233316089,
+      "loss": 1.5086,
+      "step": 312
+    },
+    {
+      "epoch": 0.03236731211706006,
+      "grad_norm": 1.1329318284988403,
+      "learning_rate": 0.00019686497672012415,
+      "loss": 1.5243,
+      "step": 313
+    },
+    {
+      "epoch": 0.032470722059926065,
+      "grad_norm": 0.6813123226165771,
+      "learning_rate": 0.00019685463010863942,
+      "loss": 2.0232,
+      "step": 314
+    },
+    {
+      "epoch": 0.03257413200279207,
+      "grad_norm": 1.022520661354065,
+      "learning_rate": 0.00019684428349715468,
+      "loss": 1.3452,
+      "step": 315
+    },
+    {
+      "epoch": 0.03267754194565808,
+      "grad_norm": 1.8104302883148193,
+      "learning_rate": 0.00019683393688566995,
+      "loss": 1.9257,
+      "step": 316
+    },
+    {
+      "epoch": 0.03278095188852408,
+      "grad_norm": 1.073581337928772,
+      "learning_rate": 0.0001968235902741852,
+      "loss": 1.5704,
+      "step": 317
+    },
+    {
+      "epoch": 0.03288436183139009,
+      "grad_norm": 2.10227632522583,
+      "learning_rate": 0.00019681324366270048,
+      "loss": 1.4086,
+      "step": 318
+    },
+    {
+      "epoch": 0.032987771774256096,
+      "grad_norm": 0.888225793838501,
+      "learning_rate": 0.00019680289705121572,
+      "loss": 1.8722,
+      "step": 319
+    },
+    {
+      "epoch": 0.0330911817171221,
+      "grad_norm": 0.8909750580787659,
+      "learning_rate": 0.00019679255043973098,
+      "loss": 1.8109,
+      "step": 320
+    },
+    {
+      "epoch": 0.03319459165998811,
+      "grad_norm": 1.5270005464553833,
+      "learning_rate": 0.00019678220382824625,
+      "loss": 2.4184,
+      "step": 321
+    },
+    {
+      "epoch": 0.033298001602854115,
+      "grad_norm": 1.501200795173645,
+      "learning_rate": 0.0001967718572167615,
+      "loss": 2.5458,
+      "step": 322
+    },
+    {
+      "epoch": 0.03340141154572012,
+      "grad_norm": 2.178788661956787,
+      "learning_rate": 0.00019676151060527678,
+      "loss": 1.6343,
+      "step": 323
+    },
+    {
+      "epoch": 0.03350482148858613,
+      "grad_norm": 2.2387502193450928,
+      "learning_rate": 0.00019675116399379204,
+      "loss": 2.2075,
+      "step": 324
+    },
+    {
+      "epoch": 0.03360823143145213,
+      "grad_norm": 0.9987953901290894,
+      "learning_rate": 0.0001967408173823073,
+      "loss": 1.8006,
+      "step": 325
+    },
+    {
+      "epoch": 0.03371164137431814,
+      "grad_norm": 1.5084731578826904,
+      "learning_rate": 0.00019673047077082255,
+      "loss": 1.9028,
+      "step": 326
+    },
+    {
+      "epoch": 0.033815051317184146,
+      "grad_norm": 0.7339261770248413,
+      "learning_rate": 0.0001967201241593378,
+      "loss": 1.8989,
+      "step": 327
+    },
+    {
+      "epoch": 0.03391846126005015,
+      "grad_norm": 0.7608749866485596,
+      "learning_rate": 0.00019670977754785308,
+      "loss": 1.4195,
+      "step": 328
+    },
+    {
+      "epoch": 0.03402187120291616,
+      "grad_norm": 1.1834888458251953,
+      "learning_rate": 0.00019669943093636834,
+      "loss": 1.5853,
+      "step": 329
+    },
+    {
+      "epoch": 0.034125281145782164,
+      "grad_norm": 0.7872140407562256,
+      "learning_rate": 0.0001966890843248836,
+      "loss": 1.6897,
+      "step": 330
+    },
+    {
+      "epoch": 0.03422869108864817,
+      "grad_norm": 1.3614797592163086,
+      "learning_rate": 0.00019667873771339888,
+      "loss": 1.6085,
+      "step": 331
+    },
+    {
+      "epoch": 0.03433210103151418,
+      "grad_norm": 1.9641176462173462,
+      "learning_rate": 0.00019666839110191414,
+      "loss": 2.2778,
+      "step": 332
+    },
+    {
+      "epoch": 0.03443551097438018,
+      "grad_norm": 1.391445517539978,
+      "learning_rate": 0.0001966580444904294,
+      "loss": 2.1415,
+      "step": 333
+    },
+    {
+      "epoch": 0.034538920917246196,
+      "grad_norm": 1.4932453632354736,
+      "learning_rate": 0.00019664769787894465,
+      "loss": 2.2698,
+      "step": 334
+    },
+    {
+      "epoch": 0.0346423308601122,
+      "grad_norm": 1.4438022375106812,
+      "learning_rate": 0.0001966373512674599,
+      "loss": 2.1187,
+      "step": 335
+    },
+    {
+      "epoch": 0.03474574080297821,
+      "grad_norm": 0.9921761751174927,
+      "learning_rate": 0.00019662700465597518,
+      "loss": 1.8119,
+      "step": 336
+    },
+    {
+      "epoch": 0.034849150745844215,
+      "grad_norm": 1.183098316192627,
+      "learning_rate": 0.00019661665804449044,
+      "loss": 1.7168,
+      "step": 337
+    },
+    {
+      "epoch": 0.03495256068871022,
+      "grad_norm": 0.990986168384552,
+      "learning_rate": 0.0001966063114330057,
+      "loss": 1.7015,
+      "step": 338
+    },
+    {
+      "epoch": 0.03505597063157623,
+      "grad_norm": 0.5842035412788391,
+      "learning_rate": 0.00019659596482152097,
+      "loss": 2.0659,
+      "step": 339
+    },
+    {
+      "epoch": 0.035159380574442234,
+      "grad_norm": 1.477404236793518,
+      "learning_rate": 0.00019658561821003624,
+      "loss": 1.9748,
+      "step": 340
+    },
+    {
+      "epoch": 0.03526279051730824,
+      "grad_norm": 0.8818874359130859,
+      "learning_rate": 0.0001965752715985515,
+      "loss": 1.9903,
+      "step": 341
+    },
+    {
+      "epoch": 0.035366200460174246,
+      "grad_norm": 1.2804322242736816,
+      "learning_rate": 0.00019656492498706674,
+      "loss": 1.9245,
+      "step": 342
+    },
+    {
+      "epoch": 0.03546961040304025,
+      "grad_norm": 1.5887545347213745,
+      "learning_rate": 0.000196554578375582,
+      "loss": 2.2518,
+      "step": 343
+    },
+    {
+      "epoch": 0.03557302034590626,
+      "grad_norm": 0.8990143537521362,
+      "learning_rate": 0.00019654423176409727,
+      "loss": 1.4269,
+      "step": 344
+    },
+    {
+      "epoch": 0.035676430288772265,
+      "grad_norm": 1.9320247173309326,
+      "learning_rate": 0.00019653388515261254,
+      "loss": 2.0075,
+      "step": 345
+    },
+    {
+      "epoch": 0.03577984023163827,
+      "grad_norm": 1.5792721509933472,
+      "learning_rate": 0.0001965235385411278,
+      "loss": 2.1097,
+      "step": 346
+    },
+    {
+      "epoch": 0.03588325017450428,
+      "grad_norm": 3.98584246635437,
+      "learning_rate": 0.00019651319192964307,
+      "loss": 1.951,
+      "step": 347
+    },
+    {
+      "epoch": 0.035986660117370284,
+      "grad_norm": 1.516032099723816,
+      "learning_rate": 0.0001965028453181583,
+      "loss": 2.3663,
+      "step": 348
+    },
+    {
+      "epoch": 0.03609007006023629,
+      "grad_norm": 2.0040533542633057,
+      "learning_rate": 0.00019649249870667357,
+      "loss": 1.8742,
+      "step": 349
+    },
+    {
+      "epoch": 0.036193480003102296,
+      "grad_norm": 1.3640230894088745,
+      "learning_rate": 0.00019648215209518884,
+      "loss": 2.1976,
+      "step": 350
+    },
+    {
+      "epoch": 0.0362968899459683,
+      "grad_norm": 1.2644320726394653,
+      "learning_rate": 0.0001964718054837041,
+      "loss": 1.9636,
+      "step": 351
+    },
+    {
+      "epoch": 0.03640029988883431,
+      "grad_norm": 1.351493239402771,
+      "learning_rate": 0.00019646145887221937,
+      "loss": 2.0332,
+      "step": 352
+    },
+    {
+      "epoch": 0.036503709831700315,
+      "grad_norm": 1.6324365139007568,
+      "learning_rate": 0.00019645111226073464,
+      "loss": 2.2028,
+      "step": 353
+    },
+    {
+      "epoch": 0.03660711977456632,
+      "grad_norm": 0.8964999318122864,
+      "learning_rate": 0.00019644076564924988,
+      "loss": 1.6669,
+      "step": 354
+    },
+    {
+      "epoch": 0.036710529717432334,
+      "grad_norm": 0.6752513647079468,
+      "learning_rate": 0.00019643041903776514,
+      "loss": 1.9048,
+      "step": 355
+    },
+    {
+      "epoch": 0.03681393966029834,
+      "grad_norm": 0.8167389631271362,
+      "learning_rate": 0.0001964200724262804,
+      "loss": 1.9142,
+      "step": 356
+    },
+    {
+      "epoch": 0.03691734960316435,
+      "grad_norm": 0.5024054646492004,
+      "learning_rate": 0.00019640972581479567,
+      "loss": 2.1693,
+      "step": 357
+    },
+    {
+      "epoch": 0.03702075954603035,
+      "grad_norm": 0.8729010224342346,
+      "learning_rate": 0.00019639937920331094,
+      "loss": 1.7099,
+      "step": 358
+    },
+    {
+      "epoch": 0.03712416948889636,
+      "grad_norm": 0.8799291849136353,
+      "learning_rate": 0.0001963890325918262,
+      "loss": 1.5546,
+      "step": 359
+    },
+    {
+      "epoch": 0.037227579431762366,
+      "grad_norm": 1.0670166015625,
+      "learning_rate": 0.00019637868598034144,
+      "loss": 1.6354,
+      "step": 360
+    },
+    {
+      "epoch": 0.03733098937462837,
+      "grad_norm": 0.9233925342559814,
+      "learning_rate": 0.0001963683393688567,
+      "loss": 2.1515,
+      "step": 361
+    },
+    {
+      "epoch": 0.03743439931749438,
+      "grad_norm": 1.1841235160827637,
+      "learning_rate": 0.00019635799275737197,
+      "loss": 1.7396,
+      "step": 362
+    },
+    {
+      "epoch": 0.037537809260360384,
+      "grad_norm": 0.8963435292243958,
+      "learning_rate": 0.00019634764614588724,
+      "loss": 2.1037,
+      "step": 363
+    },
+    {
+      "epoch": 0.03764121920322639,
+      "grad_norm": 1.3393969535827637,
+      "learning_rate": 0.0001963372995344025,
+      "loss": 2.0611,
+      "step": 364
+    },
+    {
+      "epoch": 0.0377446291460924,
+      "grad_norm": 1.2839521169662476,
+      "learning_rate": 0.00019632695292291777,
+      "loss": 1.2457,
+      "step": 365
+    },
+    {
+      "epoch": 0.0378480390889584,
+      "grad_norm": 0.8691089153289795,
+      "learning_rate": 0.000196316606311433,
+      "loss": 1.4168,
+      "step": 366
+    },
+    {
+      "epoch": 0.03795144903182441,
+      "grad_norm": 0.8430613279342651,
+      "learning_rate": 0.00019630625969994827,
+      "loss": 2.0188,
+      "step": 367
+    },
+    {
+      "epoch": 0.038054858974690416,
+      "grad_norm": 1.3562883138656616,
+      "learning_rate": 0.00019629591308846354,
+      "loss": 1.773,
+      "step": 368
+    },
+    {
+      "epoch": 0.03815826891755642,
+      "grad_norm": 1.4203057289123535,
+      "learning_rate": 0.0001962855664769788,
+      "loss": 2.0644,
+      "step": 369
+    },
+    {
+      "epoch": 0.03826167886042243,
+      "grad_norm": 1.5780055522918701,
+      "learning_rate": 0.00019627521986549407,
+      "loss": 1.8363,
+      "step": 370
+    },
+    {
+      "epoch": 0.038365088803288434,
+      "grad_norm": 1.713783621788025,
+      "learning_rate": 0.00019626487325400934,
+      "loss": 1.6578,
+      "step": 371
+    },
+    {
+      "epoch": 0.03846849874615444,
+      "grad_norm": 1.316636323928833,
+      "learning_rate": 0.00019625452664252457,
+      "loss": 1.571,
+      "step": 372
+    },
+    {
+      "epoch": 0.03857190868902045,
+      "grad_norm": 0.8482599854469299,
+      "learning_rate": 0.00019624418003103984,
+      "loss": 1.993,
+      "step": 373
+    },
+    {
+      "epoch": 0.03867531863188645,
+      "grad_norm": 0.9783570766448975,
+      "learning_rate": 0.0001962338334195551,
+      "loss": 1.7741,
+      "step": 374
+    },
+    {
+      "epoch": 0.03877872857475246,
+      "grad_norm": 1.0668089389801025,
+      "learning_rate": 0.00019622348680807037,
+      "loss": 2.0415,
+      "step": 375
+    },
+    {
+      "epoch": 0.038882138517618466,
+      "grad_norm": 0.7089089155197144,
+      "learning_rate": 0.00019621314019658564,
+      "loss": 1.9086,
+      "step": 376
+    },
+    {
+      "epoch": 0.03898554846048448,
+      "grad_norm": 1.2400895357131958,
+      "learning_rate": 0.0001962027935851009,
+      "loss": 1.9245,
+      "step": 377
+    },
+    {
+      "epoch": 0.039088958403350485,
+      "grad_norm": 0.638532280921936,
+      "learning_rate": 0.00019619244697361614,
+      "loss": 1.6457,
+      "step": 378
+    },
+    {
+      "epoch": 0.03919236834621649,
+      "grad_norm": 0.9810022711753845,
+      "learning_rate": 0.0001961821003621314,
+      "loss": 1.8889,
+      "step": 379
+    },
+    {
+      "epoch": 0.0392957782890825,
+      "grad_norm": 1.2179443836212158,
+      "learning_rate": 0.00019617175375064667,
+      "loss": 1.9645,
+      "step": 380
+    },
+    {
+      "epoch": 0.039399188231948504,
+      "grad_norm": 1.3626303672790527,
+      "learning_rate": 0.00019616140713916194,
+      "loss": 2.1616,
+      "step": 381
+    },
+    {
+      "epoch": 0.03950259817481451,
+      "grad_norm": 1.0680832862854004,
+      "learning_rate": 0.0001961510605276772,
+      "loss": 1.8425,
+      "step": 382
+    },
+    {
+      "epoch": 0.039606008117680516,
+      "grad_norm": 1.3585445880889893,
+      "learning_rate": 0.00019614071391619247,
+      "loss": 1.659,
+      "step": 383
+    },
+    {
+      "epoch": 0.03970941806054652,
+      "grad_norm": 0.6698444485664368,
+      "learning_rate": 0.0001961303673047077,
+      "loss": 1.8558,
+      "step": 384
+    },
+    {
+      "epoch": 0.03981282800341253,
+      "grad_norm": 1.7466821670532227,
+      "learning_rate": 0.00019612002069322297,
+      "loss": 1.7544,
+      "step": 385
+    },
+    {
+      "epoch": 0.039916237946278535,
+      "grad_norm": 1.762367844581604,
+      "learning_rate": 0.00019610967408173824,
+      "loss": 1.9057,
+      "step": 386
+    },
+    {
+      "epoch": 0.04001964788914454,
+      "grad_norm": 0.8794074654579163,
+      "learning_rate": 0.0001960993274702535,
+      "loss": 1.7722,
+      "step": 387
+    },
+    {
+      "epoch": 0.04012305783201055,
+      "grad_norm": 0.8490718007087708,
+      "learning_rate": 0.00019608898085876877,
+      "loss": 1.8479,
+      "step": 388
+    },
+    {
+      "epoch": 0.040226467774876554,
+      "grad_norm": 1.8691964149475098,
+      "learning_rate": 0.00019607863424728403,
+      "loss": 2.4129,
+      "step": 389
+    },
+    {
+      "epoch": 0.04032987771774256,
+      "grad_norm": 0.9891068339347839,
+      "learning_rate": 0.00019606828763579927,
+      "loss": 2.2297,
+      "step": 390
+    },
+    {
+      "epoch": 0.040433287660608566,
+      "grad_norm": 0.9857831001281738,
+      "learning_rate": 0.00019605794102431454,
+      "loss": 1.8453,
+      "step": 391
+    },
+    {
+      "epoch": 0.04053669760347457,
+      "grad_norm": 0.8795828223228455,
+      "learning_rate": 0.0001960475944128298,
+      "loss": 1.5699,
+      "step": 392
+    },
+    {
+      "epoch": 0.04064010754634058,
+      "grad_norm": 0.8345451951026917,
+      "learning_rate": 0.00019603724780134507,
+      "loss": 1.3905,
+      "step": 393
+    },
+    {
+      "epoch": 0.040743517489206585,
+      "grad_norm": 1.587632179260254,
+      "learning_rate": 0.00019602690118986034,
+      "loss": 1.6993,
+      "step": 394
+    },
+    {
+      "epoch": 0.04084692743207259,
+      "grad_norm": 0.9896488189697266,
+      "learning_rate": 0.0001960165545783756,
+      "loss": 1.9135,
+      "step": 395
+    },
+    {
+      "epoch": 0.0409503373749386,
+      "grad_norm": 2.5469861030578613,
+      "learning_rate": 0.00019600620796689084,
+      "loss": 2.1052,
+      "step": 396
+    },
+    {
+      "epoch": 0.041053747317804604,
+      "grad_norm": 3.350372314453125,
+      "learning_rate": 0.0001959958613554061,
+      "loss": 2.3579,
+      "step": 397
+    },
+    {
+      "epoch": 0.04115715726067062,
+      "grad_norm": 1.0747287273406982,
+      "learning_rate": 0.00019598551474392137,
+      "loss": 2.0841,
+      "step": 398
+    },
+    {
+      "epoch": 0.04126056720353662,
+      "grad_norm": 0.9288704991340637,
+      "learning_rate": 0.00019597516813243664,
+      "loss": 2.2219,
+      "step": 399
+    },
+    {
+      "epoch": 0.04136397714640263,
+      "grad_norm": 1.3250536918640137,
+      "learning_rate": 0.0001959648215209519,
+      "loss": 1.4469,
+      "step": 400
+    },
+    {
+      "epoch": 0.041467387089268636,
+      "grad_norm": 0.8420027494430542,
+      "learning_rate": 0.00019595447490946717,
+      "loss": 1.7364,
+      "step": 401
+    },
+    {
+      "epoch": 0.04157079703213464,
+      "grad_norm": 1.9357109069824219,
+      "learning_rate": 0.0001959441282979824,
+      "loss": 1.9886,
+      "step": 402
+    },
+    {
+      "epoch": 0.04167420697500065,
+      "grad_norm": 2.0010147094726562,
+      "learning_rate": 0.00019593378168649767,
+      "loss": 2.2205,
+      "step": 403
+    },
+    {
+      "epoch": 0.041777616917866654,
+      "grad_norm": 0.8120896220207214,
+      "learning_rate": 0.00019592343507501294,
+      "loss": 2.0961,
+      "step": 404
+    },
+    {
+      "epoch": 0.04188102686073266,
+      "grad_norm": 0.7679799795150757,
+      "learning_rate": 0.0001959130884635282,
+      "loss": 1.9252,
+      "step": 405
+    },
+    {
+      "epoch": 0.04198443680359867,
+      "grad_norm": 0.8940830826759338,
+      "learning_rate": 0.00019590274185204347,
+      "loss": 2.0255,
+      "step": 406
+    },
+    {
+      "epoch": 0.04208784674646467,
+      "grad_norm": 0.9653421640396118,
+      "learning_rate": 0.00019589239524055873,
+      "loss": 1.3163,
+      "step": 407
+    },
+    {
+      "epoch": 0.04219125668933068,
+      "grad_norm": 1.0259383916854858,
+      "learning_rate": 0.00019588204862907397,
+      "loss": 2.0588,
+      "step": 408
+    },
+    {
+      "epoch": 0.042294666632196685,
+      "grad_norm": 0.9519892930984497,
+      "learning_rate": 0.00019587170201758924,
+      "loss": 1.9092,
+      "step": 409
+    },
+    {
+      "epoch": 0.04239807657506269,
+      "grad_norm": 1.7076958417892456,
+      "learning_rate": 0.0001958613554061045,
+      "loss": 1.9263,
+      "step": 410
+    },
+    {
+      "epoch": 0.0425014865179287,
+      "grad_norm": 1.0932697057724,
+      "learning_rate": 0.00019585100879461977,
+      "loss": 1.796,
+      "step": 411
+    },
+    {
+      "epoch": 0.042604896460794704,
+      "grad_norm": 1.7663601636886597,
+      "learning_rate": 0.00019584066218313503,
+      "loss": 2.0681,
+      "step": 412
+    },
+    {
+      "epoch": 0.04270830640366071,
+      "grad_norm": 1.7543317079544067,
+      "learning_rate": 0.0001958303155716503,
+      "loss": 1.7257,
+      "step": 413
+    },
+    {
+      "epoch": 0.04281171634652672,
+      "grad_norm": 2.176375389099121,
+      "learning_rate": 0.00019581996896016554,
+      "loss": 2.515,
+      "step": 414
+    },
+    {
+      "epoch": 0.04291512628939272,
+      "grad_norm": 1.0832066535949707,
+      "learning_rate": 0.0001958096223486808,
+      "loss": 1.9952,
+      "step": 415
+    },
+    {
+      "epoch": 0.04301853623225873,
+      "grad_norm": 1.238050937652588,
+      "learning_rate": 0.00019579927573719607,
+      "loss": 1.8734,
+      "step": 416
+    },
+    {
+      "epoch": 0.043121946175124735,
+      "grad_norm": 1.0513511896133423,
+      "learning_rate": 0.00019578892912571134,
+      "loss": 2.0816,
+      "step": 417
+    },
+    {
+      "epoch": 0.04322535611799074,
+      "grad_norm": 0.7336549758911133,
+      "learning_rate": 0.0001957785825142266,
+      "loss": 1.6495,
+      "step": 418
+    },
+    {
+      "epoch": 0.04332876606085675,
+      "grad_norm": 1.76078200340271,
+      "learning_rate": 0.00019576823590274187,
+      "loss": 2.0322,
+      "step": 419
+    },
+    {
+      "epoch": 0.04343217600372276,
+      "grad_norm": 1.6764607429504395,
+      "learning_rate": 0.0001957578892912571,
+      "loss": 1.7698,
+      "step": 420
+    },
+    {
+      "epoch": 0.04353558594658877,
+      "grad_norm": 1.1513508558273315,
+      "learning_rate": 0.00019574754267977237,
+      "loss": 1.8382,
+      "step": 421
+    },
+    {
+      "epoch": 0.043638995889454774,
+      "grad_norm": 0.7220249176025391,
+      "learning_rate": 0.00019573719606828764,
+      "loss": 1.7605,
+      "step": 422
+    },
+    {
+      "epoch": 0.04374240583232078,
+      "grad_norm": 1.9611780643463135,
+      "learning_rate": 0.0001957268494568029,
+      "loss": 1.3147,
+      "step": 423
+    },
+    {
+      "epoch": 0.043845815775186786,
+      "grad_norm": 1.3807041645050049,
+      "learning_rate": 0.00019571650284531817,
+      "loss": 1.8702,
+      "step": 424
+    },
+    {
+      "epoch": 0.04394922571805279,
+      "grad_norm": 1.157392144203186,
+      "learning_rate": 0.00019570615623383343,
+      "loss": 1.9506,
+      "step": 425
+    },
+    {
+      "epoch": 0.0440526356609188,
+      "grad_norm": 2.0201995372772217,
+      "learning_rate": 0.00019569580962234867,
+      "loss": 2.2078,
+      "step": 426
+    },
+    {
+      "epoch": 0.044156045603784805,
+      "grad_norm": 1.2479861974716187,
+      "learning_rate": 0.00019568546301086394,
+      "loss": 2.5931,
+      "step": 427
+    },
+    {
+      "epoch": 0.04425945554665081,
+      "grad_norm": 1.2408305406570435,
+      "learning_rate": 0.0001956751163993792,
+      "loss": 1.7142,
+      "step": 428
+    },
+    {
+      "epoch": 0.04436286548951682,
+      "grad_norm": 1.2568728923797607,
+      "learning_rate": 0.00019566476978789447,
+      "loss": 1.6175,
+      "step": 429
+    },
+    {
+      "epoch": 0.044466275432382824,
+      "grad_norm": 0.9081295728683472,
+      "learning_rate": 0.00019565442317640973,
+      "loss": 1.6577,
+      "step": 430
+    },
+    {
+      "epoch": 0.04456968537524883,
+      "grad_norm": 2.198873281478882,
+      "learning_rate": 0.000195644076564925,
+      "loss": 1.851,
+      "step": 431
+    },
+    {
+      "epoch": 0.044673095318114836,
+      "grad_norm": 1.4093550443649292,
+      "learning_rate": 0.00019563372995344024,
+      "loss": 2.0104,
+      "step": 432
+    },
+    {
+      "epoch": 0.04477650526098084,
+      "grad_norm": 1.2647321224212646,
+      "learning_rate": 0.0001956233833419555,
+      "loss": 1.9424,
+      "step": 433
+    },
+    {
+      "epoch": 0.04487991520384685,
+      "grad_norm": 2.1517534255981445,
+      "learning_rate": 0.00019561303673047077,
+      "loss": 2.1392,
+      "step": 434
+    },
+    {
+      "epoch": 0.044983325146712855,
+      "grad_norm": 0.9314191937446594,
+      "learning_rate": 0.00019560269011898603,
+      "loss": 2.0857,
+      "step": 435
+    },
+    {
+      "epoch": 0.04508673508957886,
+      "grad_norm": 0.8166632056236267,
+      "learning_rate": 0.0001955923435075013,
+      "loss": 1.7981,
+      "step": 436
+    },
+    {
+      "epoch": 0.04519014503244487,
+      "grad_norm": 0.8604589104652405,
+      "learning_rate": 0.00019558199689601657,
+      "loss": 1.6313,
+      "step": 437
+    },
+    {
+      "epoch": 0.045293554975310873,
+      "grad_norm": 1.1065396070480347,
+      "learning_rate": 0.0001955716502845318,
+      "loss": 1.8895,
+      "step": 438
+    },
+    {
+      "epoch": 0.04539696491817688,
+      "grad_norm": 0.9663698077201843,
+      "learning_rate": 0.00019556130367304707,
+      "loss": 1.6672,
+      "step": 439
+    },
+    {
+      "epoch": 0.045500374861042886,
+      "grad_norm": 0.92950838804245,
+      "learning_rate": 0.00019555095706156234,
+      "loss": 1.8932,
+      "step": 440
+    },
+    {
+      "epoch": 0.0456037848039089,
+      "grad_norm": 1.708186388015747,
+      "learning_rate": 0.0001955406104500776,
+      "loss": 2.0815,
+      "step": 441
+    },
+    {
+      "epoch": 0.045707194746774905,
+      "grad_norm": 1.041289210319519,
+      "learning_rate": 0.00019553026383859287,
+      "loss": 1.8329,
+      "step": 442
+    },
+    {
+      "epoch": 0.04581060468964091,
+      "grad_norm": 2.087454080581665,
+      "learning_rate": 0.00019551991722710813,
+      "loss": 2.5975,
+      "step": 443
+    },
+    {
+      "epoch": 0.04591401463250692,
+      "grad_norm": 0.8211826682090759,
+      "learning_rate": 0.00019550957061562337,
+      "loss": 1.5402,
+      "step": 444
+    },
+    {
+      "epoch": 0.046017424575372924,
+      "grad_norm": 0.9692025184631348,
+      "learning_rate": 0.00019549922400413864,
+      "loss": 1.5424,
+      "step": 445
+    },
+    {
+      "epoch": 0.04612083451823893,
+      "grad_norm": 1.282361626625061,
+      "learning_rate": 0.0001954888773926539,
+      "loss": 1.1477,
+      "step": 446
+    },
+    {
+      "epoch": 0.04622424446110494,
+      "grad_norm": 1.3725165128707886,
+      "learning_rate": 0.00019547853078116917,
+      "loss": 1.7384,
+      "step": 447
+    },
+    {
+      "epoch": 0.04632765440397094,
+      "grad_norm": 1.2803922891616821,
+      "learning_rate": 0.00019546818416968443,
+      "loss": 1.5037,
+      "step": 448
+    },
+    {
+      "epoch": 0.04643106434683695,
+      "grad_norm": 1.339898943901062,
+      "learning_rate": 0.0001954578375581997,
+      "loss": 1.6536,
+      "step": 449
+    },
+    {
+      "epoch": 0.046534474289702955,
+      "grad_norm": 1.1587685346603394,
+      "learning_rate": 0.00019544749094671496,
+      "loss": 1.8648,
+      "step": 450
+    },
+    {
+      "epoch": 0.04663788423256896,
+      "grad_norm": 1.1297105550765991,
+      "learning_rate": 0.0001954371443352302,
+      "loss": 1.6633,
+      "step": 451
+    },
+    {
+      "epoch": 0.04674129417543497,
+      "grad_norm": 1.044912576675415,
+      "learning_rate": 0.00019542679772374547,
+      "loss": 1.6992,
+      "step": 452
+    },
+    {
+      "epoch": 0.046844704118300974,
+      "grad_norm": 1.1585153341293335,
+      "learning_rate": 0.00019541645111226073,
+      "loss": 1.5027,
+      "step": 453
+    },
+    {
+      "epoch": 0.04694811406116698,
+      "grad_norm": 2.7175047397613525,
+      "learning_rate": 0.000195406104500776,
+      "loss": 1.9774,
+      "step": 454
+    },
+    {
+      "epoch": 0.04705152400403299,
+      "grad_norm": 1.172574758529663,
+      "learning_rate": 0.00019539575788929126,
+      "loss": 1.8882,
+      "step": 455
+    },
+    {
+      "epoch": 0.04715493394689899,
+      "grad_norm": 1.0482293367385864,
+      "learning_rate": 0.00019538541127780653,
+      "loss": 1.6194,
+      "step": 456
+    },
+    {
+      "epoch": 0.047258343889765,
+      "grad_norm": 1.8754322528839111,
+      "learning_rate": 0.0001953750646663218,
+      "loss": 2.1531,
+      "step": 457
+    },
+    {
+      "epoch": 0.047361753832631005,
+      "grad_norm": 1.1492758989334106,
+      "learning_rate": 0.00019536471805483706,
+      "loss": 1.7342,
+      "step": 458
+    },
+    {
+      "epoch": 0.04746516377549701,
+      "grad_norm": 1.121586561203003,
+      "learning_rate": 0.0001953543714433523,
+      "loss": 1.3099,
+      "step": 459
+    },
+    {
+      "epoch": 0.04756857371836302,
+      "grad_norm": 0.9962121844291687,
+      "learning_rate": 0.00019534402483186757,
+      "loss": 1.738,
+      "step": 460
+    },
+    {
+      "epoch": 0.047671983661229024,
+      "grad_norm": 1.8688263893127441,
+      "learning_rate": 0.00019533367822038283,
+      "loss": 2.1001,
+      "step": 461
+    },
+    {
+      "epoch": 0.04777539360409503,
+      "grad_norm": 0.9260441660881042,
+      "learning_rate": 0.0001953233316088981,
+      "loss": 1.5776,
+      "step": 462
+    },
+    {
+      "epoch": 0.04787880354696104,
+      "grad_norm": 0.962735116481781,
+      "learning_rate": 0.00019531298499741336,
+      "loss": 1.3518,
+      "step": 463
+    },
+    {
+      "epoch": 0.04798221348982705,
+      "grad_norm": 3.1183736324310303,
+      "learning_rate": 0.00019530263838592863,
+      "loss": 2.299,
+      "step": 464
+    },
+    {
+      "epoch": 0.048085623432693056,
+      "grad_norm": 0.9161269664764404,
+      "learning_rate": 0.0001952922917744439,
+      "loss": 2.2219,
+      "step": 465
+    },
+    {
+      "epoch": 0.04818903337555906,
+      "grad_norm": 1.8039288520812988,
+      "learning_rate": 0.00019528194516295916,
+      "loss": 2.5586,
+      "step": 466
+    },
+    {
+      "epoch": 0.04829244331842507,
+      "grad_norm": 1.0817803144454956,
+      "learning_rate": 0.0001952715985514744,
+      "loss": 1.8248,
+      "step": 467
+    },
+    {
+      "epoch": 0.048395853261291075,
+      "grad_norm": 1.4744598865509033,
+      "learning_rate": 0.00019526125193998966,
+      "loss": 1.8484,
+      "step": 468
+    },
+    {
+      "epoch": 0.04849926320415708,
+      "grad_norm": 0.6107484698295593,
+      "learning_rate": 0.00019525090532850493,
+      "loss": 1.8709,
+      "step": 469
+    },
+    {
+      "epoch": 0.04860267314702309,
+      "grad_norm": 1.2721505165100098,
+      "learning_rate": 0.0001952405587170202,
+      "loss": 1.9078,
+      "step": 470
+    },
+    {
+      "epoch": 0.04870608308988909,
+      "grad_norm": 1.4992153644561768,
+      "learning_rate": 0.00019523021210553546,
+      "loss": 2.1692,
+      "step": 471
+    },
+    {
+      "epoch": 0.0488094930327551,
+      "grad_norm": 0.6918119788169861,
+      "learning_rate": 0.00019521986549405073,
+      "loss": 2.0205,
+      "step": 472
+    },
+    {
+      "epoch": 0.048912902975621106,
+      "grad_norm": 1.3048973083496094,
+      "learning_rate": 0.000195209518882566,
+      "loss": 1.5603,
+      "step": 473
+    },
+    {
+      "epoch": 0.04901631291848711,
+      "grad_norm": 2.52955961227417,
+      "learning_rate": 0.00019519917227108123,
+      "loss": 2.3796,
+      "step": 474
+    },
+    {
+      "epoch": 0.04911972286135312,
+      "grad_norm": 1.8823479413986206,
+      "learning_rate": 0.0001951888256595965,
+      "loss": 2.4266,
+      "step": 475
+    },
+    {
+      "epoch": 0.049223132804219125,
+      "grad_norm": 1.1469167470932007,
+      "learning_rate": 0.00019517847904811176,
+      "loss": 2.0518,
+      "step": 476
+    },
+    {
+      "epoch": 0.04932654274708513,
+      "grad_norm": 1.1344563961029053,
+      "learning_rate": 0.00019516813243662703,
+      "loss": 1.707,
+      "step": 477
+    },
+    {
+      "epoch": 0.04942995268995114,
+      "grad_norm": 2.04890513420105,
+      "learning_rate": 0.0001951577858251423,
+      "loss": 2.219,
+      "step": 478
+    },
+    {
+      "epoch": 0.04953336263281714,
+      "grad_norm": 1.6981679201126099,
+      "learning_rate": 0.00019514743921365756,
+      "loss": 2.0902,
+      "step": 479
+    },
+    {
+      "epoch": 0.04963677257568315,
+      "grad_norm": 0.7248592972755432,
+      "learning_rate": 0.0001951370926021728,
+      "loss": 1.8263,
+      "step": 480
+    },
+    {
+      "epoch": 0.049740182518549156,
+      "grad_norm": 1.1791532039642334,
+      "learning_rate": 0.00019512674599068806,
+      "loss": 1.7427,
+      "step": 481
+    },
+    {
+      "epoch": 0.04984359246141516,
+      "grad_norm": 1.0736463069915771,
+      "learning_rate": 0.00019511639937920333,
+      "loss": 1.7577,
+      "step": 482
+    },
+    {
+      "epoch": 0.04994700240428117,
+      "grad_norm": 0.7646867632865906,
+      "learning_rate": 0.0001951060527677186,
+      "loss": 1.9615,
+      "step": 483
+    },
+    {
+      "epoch": 0.05005041234714718,
+      "grad_norm": 2.399254083633423,
+      "learning_rate": 0.00019509570615623386,
+      "loss": 2.0014,
+      "step": 484
+    },
+    {
+      "epoch": 0.05015382229001319,
+      "grad_norm": 1.1391059160232544,
+      "learning_rate": 0.00019508535954474912,
+      "loss": 1.5579,
+      "step": 485
+    },
+    {
+      "epoch": 0.050257232232879194,
+      "grad_norm": 1.318179965019226,
+      "learning_rate": 0.00019507501293326436,
+      "loss": 1.7957,
+      "step": 486
+    },
+    {
+      "epoch": 0.0503606421757452,
+      "grad_norm": 3.170398473739624,
+      "learning_rate": 0.00019506466632177963,
+      "loss": 1.9489,
+      "step": 487
+    },
+    {
+      "epoch": 0.050464052118611206,
+      "grad_norm": 2.3561599254608154,
+      "learning_rate": 0.0001950543197102949,
+      "loss": 2.0789,
+      "step": 488
+    },
+    {
+      "epoch": 0.05056746206147721,
+      "grad_norm": 0.9455686807632446,
+      "learning_rate": 0.00019504397309881016,
+      "loss": 1.7533,
+      "step": 489
+    },
+    {
+      "epoch": 0.05067087200434322,
+      "grad_norm": 0.999107837677002,
+      "learning_rate": 0.00019503362648732542,
+      "loss": 1.8654,
+      "step": 490
+    },
+    {
+      "epoch": 0.050774281947209225,
+      "grad_norm": 1.1541388034820557,
+      "learning_rate": 0.0001950232798758407,
+      "loss": 1.9729,
+      "step": 491
+    },
+    {
+      "epoch": 0.05087769189007523,
+      "grad_norm": 1.5254406929016113,
+      "learning_rate": 0.00019501293326435593,
+      "loss": 2.4427,
+      "step": 492
+    },
+    {
+      "epoch": 0.05098110183294124,
+      "grad_norm": 4.286779403686523,
+      "learning_rate": 0.0001950025866528712,
+      "loss": 2.4384,
+      "step": 493
+    },
+    {
+      "epoch": 0.051084511775807244,
+      "grad_norm": 1.545657753944397,
+      "learning_rate": 0.00019499224004138646,
+      "loss": 1.6825,
+      "step": 494
+    },
+    {
+      "epoch": 0.05118792171867325,
+      "grad_norm": 1.4681485891342163,
+      "learning_rate": 0.00019498189342990172,
+      "loss": 1.9014,
+      "step": 495
+    },
+    {
+      "epoch": 0.051291331661539256,
+      "grad_norm": 1.6862431764602661,
+      "learning_rate": 0.000194971546818417,
+      "loss": 2.1929,
+      "step": 496
+    },
+    {
+      "epoch": 0.05139474160440526,
+      "grad_norm": 1.032334804534912,
+      "learning_rate": 0.00019496120020693226,
+      "loss": 1.8015,
+      "step": 497
+    },
+    {
+      "epoch": 0.05149815154727127,
+      "grad_norm": 1.258537769317627,
+      "learning_rate": 0.0001949508535954475,
+      "loss": 1.6544,
+      "step": 498
+    },
+    {
+      "epoch": 0.051601561490137275,
+      "grad_norm": 1.245613694190979,
+      "learning_rate": 0.00019494050698396276,
+      "loss": 1.8314,
+      "step": 499
+    },
+    {
+      "epoch": 0.05170497143300328,
+      "grad_norm": 1.1958906650543213,
+      "learning_rate": 0.00019493016037247803,
+      "loss": 2.0709,
+      "step": 500
+    },
+    {
+      "epoch": 0.05180838137586929,
+      "grad_norm": 1.1290735006332397,
+      "learning_rate": 0.0001949198137609933,
+      "loss": 1.6583,
+      "step": 501
+    },
+    {
+      "epoch": 0.051911791318735294,
+      "grad_norm": 0.966052234172821,
+      "learning_rate": 0.00019490946714950856,
+      "loss": 1.5282,
+      "step": 502
+    },
+    {
+      "epoch": 0.0520152012616013,
+      "grad_norm": 1.0488630533218384,
+      "learning_rate": 0.00019489912053802382,
+      "loss": 1.6947,
+      "step": 503
+    },
+    {
+      "epoch": 0.052118611204467306,
+      "grad_norm": 2.5863540172576904,
+      "learning_rate": 0.00019488877392653906,
+      "loss": 1.9402,
+      "step": 504
+    },
+    {
+      "epoch": 0.05222202114733331,
+      "grad_norm": 1.042441964149475,
+      "learning_rate": 0.00019487842731505433,
+      "loss": 1.7202,
+      "step": 505
+    },
+    {
+      "epoch": 0.052325431090199326,
+      "grad_norm": 1.457277774810791,
+      "learning_rate": 0.0001948680807035696,
+      "loss": 2.5209,
+      "step": 506
+    },
+    {
+      "epoch": 0.05242884103306533,
+      "grad_norm": 1.354345679283142,
+      "learning_rate": 0.00019485773409208486,
+      "loss": 2.0281,
+      "step": 507
+    },
+    {
+      "epoch": 0.05253225097593134,
+      "grad_norm": 1.181408166885376,
+      "learning_rate": 0.00019484738748060012,
+      "loss": 2.1902,
+      "step": 508
+    },
+    {
+      "epoch": 0.052635660918797345,
+      "grad_norm": 1.5439062118530273,
+      "learning_rate": 0.0001948370408691154,
+      "loss": 2.0531,
+      "step": 509
+    },
+    {
+      "epoch": 0.05273907086166335,
+      "grad_norm": 0.7987815141677856,
+      "learning_rate": 0.00019482669425763063,
+      "loss": 1.9365,
+      "step": 510
+    },
+    {
+      "epoch": 0.05284248080452936,
+      "grad_norm": 1.217344880104065,
+      "learning_rate": 0.0001948163476461459,
+      "loss": 1.5093,
+      "step": 511
+    },
+    {
+      "epoch": 0.05294589074739536,
+      "grad_norm": 1.0402098894119263,
+      "learning_rate": 0.00019480600103466116,
+      "loss": 1.6761,
+      "step": 512
+    },
+    {
+      "epoch": 0.05304930069026137,
+      "grad_norm": 0.9624907374382019,
+      "learning_rate": 0.00019479565442317642,
+      "loss": 2.379,
+      "step": 513
+    },
+    {
+      "epoch": 0.053152710633127376,
+      "grad_norm": 1.2226768732070923,
+      "learning_rate": 0.0001947853078116917,
+      "loss": 1.7264,
+      "step": 514
+    },
+    {
+      "epoch": 0.05325612057599338,
+      "grad_norm": 4.221198558807373,
+      "learning_rate": 0.00019477496120020695,
+      "loss": 1.6815,
+      "step": 515
+    },
+    {
+      "epoch": 0.05335953051885939,
+      "grad_norm": 1.2019705772399902,
+      "learning_rate": 0.0001947646145887222,
+      "loss": 1.5503,
+      "step": 516
+    },
+    {
+      "epoch": 0.053462940461725394,
+      "grad_norm": 1.3328227996826172,
+      "learning_rate": 0.00019475426797723746,
+      "loss": 1.5755,
+      "step": 517
+    },
+    {
+      "epoch": 0.0535663504045914,
+      "grad_norm": 1.030809760093689,
+      "learning_rate": 0.00019474392136575272,
+      "loss": 1.2846,
+      "step": 518
+    },
+    {
+      "epoch": 0.05366976034745741,
+      "grad_norm": 0.912787675857544,
+      "learning_rate": 0.000194733574754268,
+      "loss": 1.4728,
+      "step": 519
+    },
+    {
+      "epoch": 0.05377317029032341,
+      "grad_norm": 0.8509954214096069,
+      "learning_rate": 0.00019472322814278326,
+      "loss": 2.0389,
+      "step": 520
+    },
+    {
+      "epoch": 0.05387658023318942,
+      "grad_norm": 2.391146421432495,
+      "learning_rate": 0.00019471288153129852,
+      "loss": 1.894,
+      "step": 521
+    },
+    {
+      "epoch": 0.053979990176055426,
+      "grad_norm": 0.9386624693870544,
+      "learning_rate": 0.00019470253491981376,
+      "loss": 1.6739,
+      "step": 522
+    },
+    {
+      "epoch": 0.05408340011892143,
+      "grad_norm": 1.2840310335159302,
+      "learning_rate": 0.00019469218830832903,
+      "loss": 1.6585,
+      "step": 523
+    },
+    {
+      "epoch": 0.05418681006178744,
+      "grad_norm": 1.0667188167572021,
+      "learning_rate": 0.0001946818416968443,
+      "loss": 1.8078,
+      "step": 524
+    },
+    {
+      "epoch": 0.054290220004653444,
+      "grad_norm": 1.4864556789398193,
+      "learning_rate": 0.00019467149508535956,
+      "loss": 1.4433,
+      "step": 525
+    },
+    {
+      "epoch": 0.05439362994751945,
+      "grad_norm": 2.4991726875305176,
+      "learning_rate": 0.00019466114847387482,
+      "loss": 1.8232,
+      "step": 526
+    },
+    {
+      "epoch": 0.054497039890385464,
+      "grad_norm": 1.7041082382202148,
+      "learning_rate": 0.0001946508018623901,
+      "loss": 2.0924,
+      "step": 527
+    },
+    {
+      "epoch": 0.05460044983325147,
+      "grad_norm": 1.8495500087738037,
+      "learning_rate": 0.00019464045525090533,
+      "loss": 1.7738,
+      "step": 528
+    },
+    {
+      "epoch": 0.054703859776117476,
+      "grad_norm": 1.5319288969039917,
+      "learning_rate": 0.0001946301086394206,
+      "loss": 1.6553,
+      "step": 529
+    },
+    {
+      "epoch": 0.05480726971898348,
+      "grad_norm": 1.266295313835144,
+      "learning_rate": 0.00019461976202793586,
+      "loss": 2.0879,
+      "step": 530
+    },
+    {
+      "epoch": 0.05491067966184949,
+      "grad_norm": 1.0957818031311035,
+      "learning_rate": 0.00019460941541645112,
+      "loss": 1.8462,
+      "step": 531
+    },
+    {
+      "epoch": 0.055014089604715495,
+      "grad_norm": 1.0180859565734863,
+      "learning_rate": 0.0001945990688049664,
+      "loss": 1.6908,
+      "step": 532
+    },
+    {
+      "epoch": 0.0551174995475815,
+      "grad_norm": 3.8691227436065674,
+      "learning_rate": 0.00019458872219348165,
+      "loss": 1.7842,
+      "step": 533
+    },
+    {
+      "epoch": 0.05522090949044751,
+      "grad_norm": 1.1823201179504395,
+      "learning_rate": 0.0001945783755819969,
+      "loss": 1.9832,
+      "step": 534
+    },
+    {
+      "epoch": 0.055324319433313514,
+      "grad_norm": 1.3128019571304321,
+      "learning_rate": 0.00019456802897051216,
+      "loss": 2.1566,
+      "step": 535
+    },
+    {
+      "epoch": 0.05542772937617952,
+      "grad_norm": 1.2830859422683716,
+      "learning_rate": 0.00019455768235902742,
+      "loss": 1.817,
+      "step": 536
+    },
+    {
+      "epoch": 0.055531139319045526,
+      "grad_norm": 1.5610332489013672,
+      "learning_rate": 0.0001945473357475427,
+      "loss": 2.0753,
+      "step": 537
+    },
+    {
+      "epoch": 0.05563454926191153,
+      "grad_norm": 1.1201059818267822,
+      "learning_rate": 0.00019453698913605795,
+      "loss": 1.6957,
+      "step": 538
+    },
+    {
+      "epoch": 0.05573795920477754,
+      "grad_norm": 1.3312937021255493,
+      "learning_rate": 0.00019452664252457322,
+      "loss": 1.8485,
+      "step": 539
+    },
+    {
+      "epoch": 0.055841369147643545,
+      "grad_norm": 1.670798659324646,
+      "learning_rate": 0.00019451629591308846,
+      "loss": 1.5113,
+      "step": 540
+    },
+    {
+      "epoch": 0.05594477909050955,
+      "grad_norm": 1.1942532062530518,
+      "learning_rate": 0.00019450594930160372,
+      "loss": 1.7411,
+      "step": 541
+    },
+    {
+      "epoch": 0.05604818903337556,
+      "grad_norm": 0.8941847681999207,
+      "learning_rate": 0.000194495602690119,
+      "loss": 2.2823,
+      "step": 542
+    },
+    {
+      "epoch": 0.056151598976241564,
+      "grad_norm": 1.014508605003357,
+      "learning_rate": 0.00019448525607863426,
+      "loss": 1.1381,
+      "step": 543
+    },
+    {
+      "epoch": 0.05625500891910757,
+      "grad_norm": 1.2505890130996704,
+      "learning_rate": 0.00019447490946714952,
+      "loss": 1.8727,
+      "step": 544
+    },
+    {
+      "epoch": 0.056358418861973576,
+      "grad_norm": 1.0327749252319336,
+      "learning_rate": 0.0001944645628556648,
+      "loss": 1.4399,
+      "step": 545
+    },
+    {
+      "epoch": 0.05646182880483958,
+      "grad_norm": 1.232862949371338,
+      "learning_rate": 0.00019445421624418002,
+      "loss": 1.7911,
+      "step": 546
+    },
+    {
+      "epoch": 0.05656523874770559,
+      "grad_norm": 1.9683369398117065,
+      "learning_rate": 0.0001944438696326953,
+      "loss": 2.122,
+      "step": 547
+    },
+    {
+      "epoch": 0.0566686486905716,
+      "grad_norm": 1.4104269742965698,
+      "learning_rate": 0.00019443352302121056,
+      "loss": 1.4098,
+      "step": 548
+    },
+    {
+      "epoch": 0.05677205863343761,
+      "grad_norm": 1.673534631729126,
+      "learning_rate": 0.00019442317640972582,
+      "loss": 1.9461,
+      "step": 549
+    },
+    {
+      "epoch": 0.056875468576303614,
+      "grad_norm": 0.748426616191864,
+      "learning_rate": 0.0001944128297982411,
+      "loss": 1.6687,
+      "step": 550
+    },
+    {
+      "epoch": 0.05697887851916962,
+      "grad_norm": 1.689274549484253,
+      "learning_rate": 0.00019440248318675635,
+      "loss": 1.8469,
+      "step": 551
+    },
+    {
+      "epoch": 0.05708228846203563,
+      "grad_norm": 1.210861086845398,
+      "learning_rate": 0.0001943921365752716,
+      "loss": 1.7772,
+      "step": 552
+    },
+    {
+      "epoch": 0.05718569840490163,
+      "grad_norm": 1.6195064783096313,
+      "learning_rate": 0.00019438178996378686,
+      "loss": 1.6484,
+      "step": 553
+    },
+    {
+      "epoch": 0.05728910834776764,
+      "grad_norm": 2.027785062789917,
+      "learning_rate": 0.00019437144335230212,
+      "loss": 2.0626,
+      "step": 554
+    },
+    {
+      "epoch": 0.057392518290633646,
+      "grad_norm": 2.145552158355713,
+      "learning_rate": 0.0001943610967408174,
+      "loss": 1.6507,
+      "step": 555
+    },
+    {
+      "epoch": 0.05749592823349965,
+      "grad_norm": 0.8149853348731995,
+      "learning_rate": 0.00019435075012933265,
+      "loss": 2.1143,
+      "step": 556
+    },
+    {
+      "epoch": 0.05759933817636566,
+      "grad_norm": 1.6035088300704956,
+      "learning_rate": 0.00019434040351784792,
+      "loss": 2.1081,
+      "step": 557
+    },
+    {
+      "epoch": 0.057702748119231664,
+      "grad_norm": 0.7236506938934326,
+      "learning_rate": 0.00019433005690636316,
+      "loss": 1.7452,
+      "step": 558
+    },
+    {
+      "epoch": 0.05780615806209767,
+      "grad_norm": 1.3249518871307373,
+      "learning_rate": 0.00019431971029487842,
+      "loss": 1.7061,
+      "step": 559
+    },
+    {
+      "epoch": 0.05790956800496368,
+      "grad_norm": 0.9372283220291138,
+      "learning_rate": 0.0001943093636833937,
+      "loss": 1.8231,
+      "step": 560
+    },
+    {
+      "epoch": 0.05801297794782968,
+      "grad_norm": 1.004478096961975,
+      "learning_rate": 0.00019429901707190895,
+      "loss": 1.4882,
+      "step": 561
+    },
+    {
+      "epoch": 0.05811638789069569,
+      "grad_norm": 1.0969711542129517,
+      "learning_rate": 0.00019428867046042422,
+      "loss": 1.8174,
+      "step": 562
+    },
+    {
+      "epoch": 0.058219797833561696,
+      "grad_norm": 1.0373283624649048,
+      "learning_rate": 0.00019427832384893949,
+      "loss": 1.5651,
+      "step": 563
+    },
+    {
+      "epoch": 0.0583232077764277,
+      "grad_norm": 0.7008000612258911,
+      "learning_rate": 0.00019426797723745472,
+      "loss": 2.0102,
+      "step": 564
+    },
+    {
+      "epoch": 0.05842661771929371,
+      "grad_norm": 1.4896759986877441,
+      "learning_rate": 0.00019425763062597,
+      "loss": 1.4347,
+      "step": 565
+    },
+    {
+      "epoch": 0.058530027662159714,
+      "grad_norm": 1.116275668144226,
+      "learning_rate": 0.00019424728401448526,
+      "loss": 2.1501,
+      "step": 566
+    },
+    {
+      "epoch": 0.05863343760502572,
+      "grad_norm": 1.1160902976989746,
+      "learning_rate": 0.00019423693740300052,
+      "loss": 1.7055,
+      "step": 567
+    },
+    {
+      "epoch": 0.05873684754789173,
+      "grad_norm": 1.161346673965454,
+      "learning_rate": 0.00019422659079151579,
+      "loss": 1.7059,
+      "step": 568
+    },
+    {
+      "epoch": 0.05884025749075773,
+      "grad_norm": 1.6276146173477173,
+      "learning_rate": 0.00019421624418003105,
+      "loss": 2.172,
+      "step": 569
+    },
+    {
+      "epoch": 0.058943667433623746,
+      "grad_norm": 1.095842957496643,
+      "learning_rate": 0.0001942058975685463,
+      "loss": 1.8135,
+      "step": 570
+    },
+    {
+      "epoch": 0.05904707737648975,
+      "grad_norm": 1.5054162740707397,
+      "learning_rate": 0.00019419555095706156,
+      "loss": 1.5956,
+      "step": 571
+    },
+    {
+      "epoch": 0.05915048731935576,
+      "grad_norm": 1.1078232526779175,
+      "learning_rate": 0.00019418520434557682,
+      "loss": 2.1434,
+      "step": 572
+    },
+    {
+      "epoch": 0.059253897262221765,
+      "grad_norm": 1.3417181968688965,
+      "learning_rate": 0.0001941748577340921,
+      "loss": 2.0869,
+      "step": 573
+    },
+    {
+      "epoch": 0.05935730720508777,
+      "grad_norm": 1.2181316614151,
+      "learning_rate": 0.00019416451112260735,
+      "loss": 1.8027,
+      "step": 574
+    },
+    {
+      "epoch": 0.05946071714795378,
+      "grad_norm": 1.519131064414978,
+      "learning_rate": 0.00019415416451112262,
+      "loss": 1.9721,
+      "step": 575
+    },
+    {
+      "epoch": 0.059564127090819784,
+      "grad_norm": 1.165762186050415,
+      "learning_rate": 0.00019414381789963786,
+      "loss": 1.0573,
+      "step": 576
+    },
+    {
+      "epoch": 0.05966753703368579,
+      "grad_norm": 0.9382355213165283,
+      "learning_rate": 0.00019413347128815312,
+      "loss": 1.8539,
+      "step": 577
+    },
+    {
+      "epoch": 0.059770946976551796,
+      "grad_norm": 2.0592143535614014,
+      "learning_rate": 0.0001941231246766684,
+      "loss": 1.8897,
+      "step": 578
+    },
+    {
+      "epoch": 0.0598743569194178,
+      "grad_norm": 1.1367623805999756,
+      "learning_rate": 0.00019411277806518365,
+      "loss": 2.2578,
+      "step": 579
+    },
+    {
+      "epoch": 0.05997776686228381,
+      "grad_norm": 1.5440205335617065,
+      "learning_rate": 0.00019410243145369892,
+      "loss": 1.8463,
+      "step": 580
+    },
+    {
+      "epoch": 0.060081176805149815,
+      "grad_norm": 1.3517295122146606,
+      "learning_rate": 0.00019409208484221418,
+      "loss": 1.7255,
+      "step": 581
+    },
+    {
+      "epoch": 0.06018458674801582,
+      "grad_norm": 1.735840916633606,
+      "learning_rate": 0.00019408173823072945,
+      "loss": 1.8489,
+      "step": 582
+    },
+    {
+      "epoch": 0.06028799669088183,
+      "grad_norm": 1.2492977380752563,
+      "learning_rate": 0.00019407139161924472,
+      "loss": 1.679,
+      "step": 583
+    },
+    {
+      "epoch": 0.060391406633747834,
+      "grad_norm": 1.0322548151016235,
+      "learning_rate": 0.00019406104500775995,
+      "loss": 1.7664,
+      "step": 584
+    },
+    {
+      "epoch": 0.06049481657661384,
+      "grad_norm": 1.5941745042800903,
+      "learning_rate": 0.00019405069839627522,
+      "loss": 2.0438,
+      "step": 585
+    },
+    {
+      "epoch": 0.060598226519479846,
+      "grad_norm": 0.8804157376289368,
+      "learning_rate": 0.00019404035178479049,
+      "loss": 1.7586,
+      "step": 586
+    },
+    {
+      "epoch": 0.06070163646234585,
+      "grad_norm": 1.5512956380844116,
+      "learning_rate": 0.00019403000517330575,
+      "loss": 2.1695,
+      "step": 587
+    },
+    {
+      "epoch": 0.06080504640521186,
+      "grad_norm": 1.4347621202468872,
+      "learning_rate": 0.00019401965856182102,
+      "loss": 1.8846,
+      "step": 588
+    },
+    {
+      "epoch": 0.060908456348077865,
+      "grad_norm": 1.970655083656311,
+      "learning_rate": 0.00019400931195033628,
+      "loss": 2.0841,
+      "step": 589
+    },
+    {
+      "epoch": 0.06101186629094387,
+      "grad_norm": 0.9423110485076904,
+      "learning_rate": 0.00019399896533885155,
+      "loss": 1.6569,
+      "step": 590
+    },
+    {
+      "epoch": 0.061115276233809884,
+      "grad_norm": 0.7332543730735779,
+      "learning_rate": 0.0001939886187273668,
+      "loss": 1.9821,
+      "step": 591
+    },
+    {
+      "epoch": 0.06121868617667589,
+      "grad_norm": 1.3857728242874146,
+      "learning_rate": 0.00019397827211588205,
+      "loss": 2.016,
+      "step": 592
+    },
+    {
+      "epoch": 0.0613220961195419,
+      "grad_norm": 1.6231861114501953,
+      "learning_rate": 0.00019396792550439732,
+      "loss": 1.6667,
+      "step": 593
+    },
+    {
+      "epoch": 0.0614255060624079,
+      "grad_norm": 2.417095422744751,
+      "learning_rate": 0.00019395757889291258,
+      "loss": 2.3043,
+      "step": 594
+    },
+    {
+      "epoch": 0.06152891600527391,
+      "grad_norm": 0.9848974943161011,
+      "learning_rate": 0.00019394723228142785,
+      "loss": 1.7858,
+      "step": 595
+    },
+    {
+      "epoch": 0.061632325948139915,
+      "grad_norm": 1.6868712902069092,
+      "learning_rate": 0.00019393688566994311,
+      "loss": 1.6796,
+      "step": 596
+    },
+    {
+      "epoch": 0.06173573589100592,
+      "grad_norm": 1.9725900888442993,
+      "learning_rate": 0.00019392653905845838,
+      "loss": 1.6825,
+      "step": 597
+    },
+    {
+      "epoch": 0.06183914583387193,
+      "grad_norm": 1.056667447090149,
+      "learning_rate": 0.00019391619244697364,
+      "loss": 1.9901,
+      "step": 598
+    },
+    {
+      "epoch": 0.061942555776737934,
+      "grad_norm": 1.6670137643814087,
+      "learning_rate": 0.0001939058458354889,
+      "loss": 1.517,
+      "step": 599
+    },
+    {
+      "epoch": 0.06204596571960394,
+      "grad_norm": 1.2356425523757935,
+      "learning_rate": 0.00019389549922400415,
+      "loss": 1.9377,
+      "step": 600
+    },
+    {
+      "epoch": 0.06214937566246995,
+      "grad_norm": 1.196425199508667,
+      "learning_rate": 0.00019388515261251941,
+      "loss": 1.6639,
+      "step": 601
+    },
+    {
+      "epoch": 0.06225278560533595,
+      "grad_norm": 0.9325123429298401,
+      "learning_rate": 0.00019387480600103468,
+      "loss": 1.6499,
+      "step": 602
+    },
+    {
+      "epoch": 0.06235619554820196,
+      "grad_norm": 1.1611047983169556,
+      "learning_rate": 0.00019386445938954995,
+      "loss": 2.0172,
+      "step": 603
+    },
+    {
+      "epoch": 0.062459605491067965,
+      "grad_norm": 2.512784957885742,
+      "learning_rate": 0.0001938541127780652,
+      "loss": 2.5452,
+      "step": 604
+    },
+    {
+      "epoch": 0.06256301543393397,
+      "grad_norm": 1.321507215499878,
+      "learning_rate": 0.00019384376616658048,
+      "loss": 1.3509,
+      "step": 605
+    },
+    {
+      "epoch": 0.06266642537679998,
+      "grad_norm": 0.9709866046905518,
+      "learning_rate": 0.00019383341955509572,
+      "loss": 1.9421,
+      "step": 606
+    },
+    {
+      "epoch": 0.06276983531966598,
+      "grad_norm": 1.2269256114959717,
+      "learning_rate": 0.00019382307294361098,
+      "loss": 1.7247,
+      "step": 607
+    },
+    {
+      "epoch": 0.06287324526253199,
+      "grad_norm": 1.642680287361145,
+      "learning_rate": 0.00019381272633212625,
+      "loss": 1.9324,
+      "step": 608
+    },
+    {
+      "epoch": 0.062976655205398,
+      "grad_norm": 1.0991445779800415,
+      "learning_rate": 0.0001938023797206415,
+      "loss": 1.5484,
+      "step": 609
+    },
+    {
+      "epoch": 0.063080065148264,
+      "grad_norm": 1.4630918502807617,
+      "learning_rate": 0.00019379203310915678,
+      "loss": 1.7564,
+      "step": 610
+    },
+    {
+      "epoch": 0.06318347509113001,
+      "grad_norm": 1.593808889389038,
+      "learning_rate": 0.00019378168649767204,
+      "loss": 1.6729,
+      "step": 611
+    },
+    {
+      "epoch": 0.06328688503399602,
+      "grad_norm": 2.05412220954895,
+      "learning_rate": 0.00019377133988618728,
+      "loss": 1.9342,
+      "step": 612
+    },
+    {
+      "epoch": 0.06339029497686202,
+      "grad_norm": 2.5476601123809814,
+      "learning_rate": 0.00019376099327470255,
+      "loss": 2.4104,
+      "step": 613
+    },
+    {
+      "epoch": 0.06349370491972803,
+      "grad_norm": 1.3254003524780273,
+      "learning_rate": 0.0001937506466632178,
+      "loss": 1.7209,
+      "step": 614
+    },
+    {
+      "epoch": 0.06359711486259403,
+      "grad_norm": 1.5243037939071655,
+      "learning_rate": 0.00019374030005173308,
+      "loss": 1.8476,
+      "step": 615
+    },
+    {
+      "epoch": 0.06370052480546004,
+      "grad_norm": 1.1326062679290771,
+      "learning_rate": 0.00019372995344024834,
+      "loss": 1.3442,
+      "step": 616
+    },
+    {
+      "epoch": 0.06380393474832605,
+      "grad_norm": 1.375542402267456,
+      "learning_rate": 0.0001937196068287636,
+      "loss": 1.6413,
+      "step": 617
+    },
+    {
+      "epoch": 0.06390734469119205,
+      "grad_norm": 1.1514317989349365,
+      "learning_rate": 0.00019370926021727885,
+      "loss": 1.2693,
+      "step": 618
+    },
+    {
+      "epoch": 0.06401075463405806,
+      "grad_norm": 1.629643201828003,
+      "learning_rate": 0.00019369891360579411,
+      "loss": 1.8193,
+      "step": 619
+    },
+    {
+      "epoch": 0.06411416457692407,
+      "grad_norm": 9.200542449951172,
+      "learning_rate": 0.00019368856699430938,
+      "loss": 1.8048,
+      "step": 620
+    },
+    {
+      "epoch": 0.06421757451979007,
+      "grad_norm": 1.2734607458114624,
+      "learning_rate": 0.00019367822038282464,
+      "loss": 2.2,
+      "step": 621
+    },
+    {
+      "epoch": 0.06432098446265608,
+      "grad_norm": 1.6576577425003052,
+      "learning_rate": 0.0001936678737713399,
+      "loss": 2.0564,
+      "step": 622
+    },
+    {
+      "epoch": 0.06442439440552208,
+      "grad_norm": 1.359603762626648,
+      "learning_rate": 0.00019365752715985518,
+      "loss": 1.9281,
+      "step": 623
+    },
+    {
+      "epoch": 0.0645278043483881,
+      "grad_norm": 0.9970663785934448,
+      "learning_rate": 0.00019364718054837041,
+      "loss": 1.6042,
+      "step": 624
+    },
+    {
+      "epoch": 0.06463121429125411,
+      "grad_norm": 1.8795320987701416,
+      "learning_rate": 0.00019363683393688568,
+      "loss": 2.0144,
+      "step": 625
+    },
+    {
+      "epoch": 0.06473462423412012,
+      "grad_norm": 1.8463833332061768,
+      "learning_rate": 0.00019362648732540095,
+      "loss": 1.9506,
+      "step": 626
+    },
+    {
+      "epoch": 0.06483803417698612,
+      "grad_norm": 1.0363043546676636,
+      "learning_rate": 0.0001936161407139162,
+      "loss": 2.2462,
+      "step": 627
+    },
+    {
+      "epoch": 0.06494144411985213,
+      "grad_norm": 1.2393428087234497,
+      "learning_rate": 0.00019360579410243148,
+      "loss": 1.2618,
+      "step": 628
+    },
+    {
+      "epoch": 0.06504485406271814,
+      "grad_norm": 1.0203514099121094,
+      "learning_rate": 0.00019359544749094674,
+      "loss": 1.9836,
+      "step": 629
+    },
+    {
+      "epoch": 0.06514826400558414,
+      "grad_norm": 1.408217191696167,
+      "learning_rate": 0.00019358510087946198,
+      "loss": 1.9312,
+      "step": 630
+    },
+    {
+      "epoch": 0.06525167394845015,
+      "grad_norm": 1.4327986240386963,
+      "learning_rate": 0.00019357475426797725,
+      "loss": 1.9366,
+      "step": 631
+    },
+    {
+      "epoch": 0.06535508389131615,
+      "grad_norm": 0.8558879494667053,
+      "learning_rate": 0.0001935644076564925,
+      "loss": 1.7253,
+      "step": 632
+    },
+    {
+      "epoch": 0.06545849383418216,
+      "grad_norm": 1.3604400157928467,
+      "learning_rate": 0.00019355406104500778,
+      "loss": 2.0347,
+      "step": 633
+    },
+    {
+      "epoch": 0.06556190377704817,
+      "grad_norm": 1.0031954050064087,
+      "learning_rate": 0.00019354371443352304,
+      "loss": 1.8156,
+      "step": 634
+    },
+    {
+      "epoch": 0.06566531371991417,
+      "grad_norm": 1.2598495483398438,
+      "learning_rate": 0.00019353336782203828,
+      "loss": 1.5517,
+      "step": 635
+    },
+    {
+      "epoch": 0.06576872366278018,
+      "grad_norm": 0.7164933681488037,
+      "learning_rate": 0.00019352302121055355,
+      "loss": 1.7041,
+      "step": 636
+    },
+    {
+      "epoch": 0.06587213360564619,
+      "grad_norm": 1.795189380645752,
+      "learning_rate": 0.0001935126745990688,
+      "loss": 1.7281,
+      "step": 637
+    },
+    {
+      "epoch": 0.06597554354851219,
+      "grad_norm": 1.1797122955322266,
+      "learning_rate": 0.00019350232798758408,
+      "loss": 1.6707,
+      "step": 638
+    },
+    {
+      "epoch": 0.0660789534913782,
+      "grad_norm": 1.0083849430084229,
+      "learning_rate": 0.00019349198137609934,
+      "loss": 1.6134,
+      "step": 639
+    },
+    {
+      "epoch": 0.0661823634342442,
+      "grad_norm": 0.823398768901825,
+      "learning_rate": 0.0001934816347646146,
+      "loss": 2.1252,
+      "step": 640
+    },
+    {
+      "epoch": 0.06628577337711021,
+      "grad_norm": 1.0543327331542969,
+      "learning_rate": 0.00019347128815312985,
+      "loss": 1.4388,
+      "step": 641
+    },
+    {
+      "epoch": 0.06638918331997622,
+      "grad_norm": 1.2253812551498413,
+      "learning_rate": 0.0001934609415416451,
+      "loss": 1.6264,
+      "step": 642
+    },
+    {
+      "epoch": 0.06649259326284222,
+      "grad_norm": 0.5971275568008423,
+      "learning_rate": 0.00019345059493016038,
+      "loss": 2.0363,
+      "step": 643
+    },
+    {
+      "epoch": 0.06659600320570823,
+      "grad_norm": 1.0794010162353516,
+      "learning_rate": 0.00019344024831867564,
+      "loss": 1.8093,
+      "step": 644
+    },
+    {
+      "epoch": 0.06669941314857424,
+      "grad_norm": 0.8961673378944397,
+      "learning_rate": 0.0001934299017071909,
+      "loss": 1.8438,
+      "step": 645
+    },
+    {
+      "epoch": 0.06680282309144024,
+      "grad_norm": 0.8820412158966064,
+      "learning_rate": 0.00019341955509570618,
+      "loss": 2.2632,
+      "step": 646
+    },
+    {
+      "epoch": 0.06690623303430625,
+      "grad_norm": 1.2584704160690308,
+      "learning_rate": 0.00019340920848422141,
+      "loss": 1.7135,
+      "step": 647
+    },
+    {
+      "epoch": 0.06700964297717225,
+      "grad_norm": 1.5364105701446533,
+      "learning_rate": 0.00019339886187273668,
+      "loss": 2.0992,
+      "step": 648
+    },
+    {
+      "epoch": 0.06711305292003826,
+      "grad_norm": 1.1961735486984253,
+      "learning_rate": 0.00019338851526125195,
+      "loss": 1.7492,
+      "step": 649
+    },
+    {
+      "epoch": 0.06721646286290427,
+      "grad_norm": 1.824004888534546,
+      "learning_rate": 0.0001933781686497672,
+      "loss": 2.0148,
+      "step": 650
+    },
+    {
+      "epoch": 0.06731987280577027,
+      "grad_norm": 1.5567946434020996,
+      "learning_rate": 0.00019336782203828248,
+      "loss": 1.4838,
+      "step": 651
+    },
+    {
+      "epoch": 0.06742328274863628,
+      "grad_norm": 1.0475163459777832,
+      "learning_rate": 0.00019335747542679774,
+      "loss": 1.7052,
+      "step": 652
+    },
+    {
+      "epoch": 0.06752669269150229,
+      "grad_norm": 1.7999539375305176,
+      "learning_rate": 0.00019334712881531298,
+      "loss": 1.8598,
+      "step": 653
+    },
+    {
+      "epoch": 0.06763010263436829,
+      "grad_norm": 1.4463770389556885,
+      "learning_rate": 0.00019333678220382825,
+      "loss": 1.6957,
+      "step": 654
+    },
+    {
+      "epoch": 0.0677335125772343,
+      "grad_norm": 1.0363894701004028,
+      "learning_rate": 0.0001933264355923435,
+      "loss": 1.9333,
+      "step": 655
+    },
+    {
+      "epoch": 0.0678369225201003,
+      "grad_norm": 0.8181893825531006,
+      "learning_rate": 0.00019331608898085878,
+      "loss": 1.7909,
+      "step": 656
+    },
+    {
+      "epoch": 0.06794033246296631,
+      "grad_norm": 1.2460615634918213,
+      "learning_rate": 0.00019330574236937404,
+      "loss": 2.0109,
+      "step": 657
+    },
+    {
+      "epoch": 0.06804374240583232,
+      "grad_norm": 1.011409878730774,
+      "learning_rate": 0.0001932953957578893,
+      "loss": 2.0692,
+      "step": 658
+    },
+    {
+      "epoch": 0.06814715234869832,
+      "grad_norm": 1.0723906755447388,
+      "learning_rate": 0.00019328504914640455,
+      "loss": 2.0749,
+      "step": 659
+    },
+    {
+      "epoch": 0.06825056229156433,
+      "grad_norm": 1.5507792234420776,
+      "learning_rate": 0.0001932747025349198,
+      "loss": 2.0177,
+      "step": 660
+    },
+    {
+      "epoch": 0.06835397223443034,
+      "grad_norm": 1.6804261207580566,
+      "learning_rate": 0.00019326435592343508,
+      "loss": 1.7871,
+      "step": 661
+    },
+    {
+      "epoch": 0.06845738217729634,
+      "grad_norm": 1.0971132516860962,
+      "learning_rate": 0.00019325400931195034,
+      "loss": 1.7657,
+      "step": 662
+    },
+    {
+      "epoch": 0.06856079212016235,
+      "grad_norm": 0.9800976514816284,
+      "learning_rate": 0.0001932436627004656,
+      "loss": 1.6668,
+      "step": 663
+    },
+    {
+      "epoch": 0.06866420206302835,
+      "grad_norm": 1.3540741205215454,
+      "learning_rate": 0.00019323331608898087,
+      "loss": 1.5822,
+      "step": 664
+    },
+    {
+      "epoch": 0.06876761200589436,
+      "grad_norm": 1.2192633152008057,
+      "learning_rate": 0.0001932229694774961,
+      "loss": 1.8653,
+      "step": 665
+    },
+    {
+      "epoch": 0.06887102194876037,
+      "grad_norm": 1.444132924079895,
+      "learning_rate": 0.00019321262286601138,
+      "loss": 2.2671,
+      "step": 666
+    },
+    {
+      "epoch": 0.06897443189162639,
+      "grad_norm": 1.5947513580322266,
+      "learning_rate": 0.00019320227625452664,
+      "loss": 1.4868,
+      "step": 667
+    },
+    {
+      "epoch": 0.06907784183449239,
+      "grad_norm": 1.0117969512939453,
+      "learning_rate": 0.0001931919296430419,
+      "loss": 1.2683,
+      "step": 668
+    },
+    {
+      "epoch": 0.0691812517773584,
+      "grad_norm": 1.3698185682296753,
+      "learning_rate": 0.00019318158303155718,
+      "loss": 1.8173,
+      "step": 669
+    },
+    {
+      "epoch": 0.0692846617202244,
+      "grad_norm": 2.321216344833374,
+      "learning_rate": 0.00019317123642007244,
+      "loss": 1.4577,
+      "step": 670
+    },
+    {
+      "epoch": 0.06938807166309041,
+      "grad_norm": 1.6404632329940796,
+      "learning_rate": 0.00019316088980858768,
+      "loss": 1.4561,
+      "step": 671
+    },
+    {
+      "epoch": 0.06949148160595642,
+      "grad_norm": 1.2985002994537354,
+      "learning_rate": 0.00019315054319710294,
+      "loss": 2.0832,
+      "step": 672
+    },
+    {
+      "epoch": 0.06959489154882242,
+      "grad_norm": 1.020167589187622,
+      "learning_rate": 0.0001931401965856182,
+      "loss": 2.0277,
+      "step": 673
+    },
+    {
+      "epoch": 0.06969830149168843,
+      "grad_norm": 1.8142248392105103,
+      "learning_rate": 0.00019312984997413348,
+      "loss": 1.7957,
+      "step": 674
+    },
+    {
+      "epoch": 0.06980171143455444,
+      "grad_norm": 2.099975824356079,
+      "learning_rate": 0.00019311950336264874,
+      "loss": 1.9989,
+      "step": 675
+    },
+    {
+      "epoch": 0.06990512137742044,
+      "grad_norm": 2.1778149604797363,
+      "learning_rate": 0.000193109156751164,
+      "loss": 1.7527,
+      "step": 676
+    },
+    {
+      "epoch": 0.07000853132028645,
+      "grad_norm": 2.6961352825164795,
+      "learning_rate": 0.00019309881013967925,
+      "loss": 2.3416,
+      "step": 677
+    },
+    {
+      "epoch": 0.07011194126315246,
+      "grad_norm": 1.3259974718093872,
+      "learning_rate": 0.0001930884635281945,
+      "loss": 1.9032,
+      "step": 678
+    },
+    {
+      "epoch": 0.07021535120601846,
+      "grad_norm": 1.0762745141983032,
+      "learning_rate": 0.00019307811691670978,
+      "loss": 1.2353,
+      "step": 679
+    },
+    {
+      "epoch": 0.07031876114888447,
+      "grad_norm": 0.88735431432724,
+      "learning_rate": 0.00019306777030522504,
+      "loss": 1.756,
+      "step": 680
+    },
+    {
+      "epoch": 0.07042217109175047,
+      "grad_norm": 1.5669922828674316,
+      "learning_rate": 0.0001930574236937403,
+      "loss": 1.9952,
+      "step": 681
+    },
+    {
+      "epoch": 0.07052558103461648,
+      "grad_norm": 1.8280024528503418,
+      "learning_rate": 0.00019304707708225557,
+      "loss": 2.0567,
+      "step": 682
+    },
+    {
+      "epoch": 0.07062899097748249,
+      "grad_norm": 1.9424211978912354,
+      "learning_rate": 0.0001930367304707708,
+      "loss": 1.873,
+      "step": 683
+    },
+    {
+      "epoch": 0.07073240092034849,
+      "grad_norm": 1.0285444259643555,
+      "learning_rate": 0.00019302638385928608,
+      "loss": 1.4914,
+      "step": 684
+    },
+    {
+      "epoch": 0.0708358108632145,
+      "grad_norm": 1.020849347114563,
+      "learning_rate": 0.00019301603724780134,
+      "loss": 1.7896,
+      "step": 685
+    },
+    {
+      "epoch": 0.0709392208060805,
+      "grad_norm": 1.274097204208374,
+      "learning_rate": 0.0001930056906363166,
+      "loss": 1.6604,
+      "step": 686
+    },
+    {
+      "epoch": 0.07104263074894651,
+      "grad_norm": 2.1440742015838623,
+      "learning_rate": 0.00019299534402483187,
+      "loss": 1.6177,
+      "step": 687
+    },
+    {
+      "epoch": 0.07114604069181252,
+      "grad_norm": 1.0610333681106567,
+      "learning_rate": 0.00019298499741334714,
+      "loss": 1.6835,
+      "step": 688
+    },
+    {
+      "epoch": 0.07124945063467852,
+      "grad_norm": 1.0177770853042603,
+      "learning_rate": 0.00019297465080186238,
+      "loss": 1.5577,
+      "step": 689
+    },
+    {
+      "epoch": 0.07135286057754453,
+      "grad_norm": 1.0054686069488525,
+      "learning_rate": 0.00019296430419037764,
+      "loss": 1.4735,
+      "step": 690
+    },
+    {
+      "epoch": 0.07145627052041054,
+      "grad_norm": 0.7136942148208618,
+      "learning_rate": 0.0001929539575788929,
+      "loss": 2.0558,
+      "step": 691
+    },
+    {
+      "epoch": 0.07155968046327654,
+      "grad_norm": 2.260354995727539,
+      "learning_rate": 0.00019294361096740818,
+      "loss": 1.7685,
+      "step": 692
+    },
+    {
+      "epoch": 0.07166309040614255,
+      "grad_norm": 1.2539150714874268,
+      "learning_rate": 0.00019293326435592344,
+      "loss": 1.8324,
+      "step": 693
+    },
+    {
+      "epoch": 0.07176650034900856,
+      "grad_norm": 1.8409538269042969,
+      "learning_rate": 0.0001929229177444387,
+      "loss": 1.9525,
+      "step": 694
+    },
+    {
+      "epoch": 0.07186991029187456,
+      "grad_norm": 1.0681891441345215,
+      "learning_rate": 0.00019291257113295394,
+      "loss": 1.5554,
+      "step": 695
+    },
+    {
+      "epoch": 0.07197332023474057,
+      "grad_norm": 2.5881447792053223,
+      "learning_rate": 0.0001929022245214692,
+      "loss": 2.1174,
+      "step": 696
+    },
+    {
+      "epoch": 0.07207673017760657,
+      "grad_norm": 0.9745953679084778,
+      "learning_rate": 0.00019289187790998448,
+      "loss": 1.3149,
+      "step": 697
+    },
+    {
+      "epoch": 0.07218014012047258,
+      "grad_norm": 1.1194660663604736,
+      "learning_rate": 0.00019288153129849974,
+      "loss": 1.3463,
+      "step": 698
+    },
+    {
+      "epoch": 0.07228355006333859,
+      "grad_norm": 1.7407701015472412,
+      "learning_rate": 0.000192871184687015,
+      "loss": 2.2298,
+      "step": 699
+    },
+    {
+      "epoch": 0.07238696000620459,
+      "grad_norm": 1.0828900337219238,
+      "learning_rate": 0.00019286083807553027,
+      "loss": 1.7538,
+      "step": 700
+    },
+    {
+      "epoch": 0.0724903699490706,
+      "grad_norm": 2.003937005996704,
+      "learning_rate": 0.0001928504914640455,
+      "loss": 1.5031,
+      "step": 701
+    },
+    {
+      "epoch": 0.0725937798919366,
+      "grad_norm": 1.4628642797470093,
+      "learning_rate": 0.00019284014485256078,
+      "loss": 1.7986,
+      "step": 702
+    },
+    {
+      "epoch": 0.07269718983480261,
+      "grad_norm": 1.0137124061584473,
+      "learning_rate": 0.00019282979824107604,
+      "loss": 1.5951,
+      "step": 703
+    },
+    {
+      "epoch": 0.07280059977766862,
+      "grad_norm": 1.588923692703247,
+      "learning_rate": 0.0001928194516295913,
+      "loss": 1.8639,
+      "step": 704
+    },
+    {
+      "epoch": 0.07290400972053462,
+      "grad_norm": 0.8961623311042786,
+      "learning_rate": 0.00019280910501810657,
+      "loss": 1.5859,
+      "step": 705
+    },
+    {
+      "epoch": 0.07300741966340063,
+      "grad_norm": 0.8769466280937195,
+      "learning_rate": 0.00019279875840662184,
+      "loss": 1.3753,
+      "step": 706
+    },
+    {
+      "epoch": 0.07311082960626664,
+      "grad_norm": 0.658501148223877,
+      "learning_rate": 0.0001927884117951371,
+      "loss": 2.1286,
+      "step": 707
+    },
+    {
+      "epoch": 0.07321423954913264,
+      "grad_norm": 1.503748893737793,
+      "learning_rate": 0.00019277806518365237,
+      "loss": 2.365,
+      "step": 708
+    },
+    {
+      "epoch": 0.07331764949199865,
+      "grad_norm": 1.359349250793457,
+      "learning_rate": 0.0001927677185721676,
+      "loss": 1.9966,
+      "step": 709
+    },
+    {
+      "epoch": 0.07342105943486467,
+      "grad_norm": 1.5836228132247925,
+      "learning_rate": 0.00019275737196068287,
+      "loss": 1.2079,
+      "step": 710
+    },
+    {
+      "epoch": 0.07352446937773068,
+      "grad_norm": 1.0415390729904175,
+      "learning_rate": 0.00019274702534919814,
+      "loss": 1.8506,
+      "step": 711
+    },
+    {
+      "epoch": 0.07362787932059668,
+      "grad_norm": 1.3095402717590332,
+      "learning_rate": 0.0001927366787377134,
+      "loss": 1.2082,
+      "step": 712
+    },
+    {
+      "epoch": 0.07373128926346269,
+      "grad_norm": 1.356926441192627,
+      "learning_rate": 0.00019272633212622867,
+      "loss": 1.341,
+      "step": 713
+    },
+    {
+      "epoch": 0.0738346992063287,
+      "grad_norm": 1.0612878799438477,
+      "learning_rate": 0.00019271598551474394,
+      "loss": 1.7926,
+      "step": 714
+    },
+    {
+      "epoch": 0.0739381091491947,
+      "grad_norm": 1.9951964616775513,
+      "learning_rate": 0.0001927056389032592,
+      "loss": 1.6938,
+      "step": 715
+    },
+    {
+      "epoch": 0.0740415190920607,
+      "grad_norm": 1.6049528121948242,
+      "learning_rate": 0.00019269529229177447,
+      "loss": 1.602,
+      "step": 716
+    },
+    {
+      "epoch": 0.07414492903492671,
+      "grad_norm": 1.646397590637207,
+      "learning_rate": 0.0001926849456802897,
+      "loss": 1.9062,
+      "step": 717
+    },
+    {
+      "epoch": 0.07424833897779272,
+      "grad_norm": 1.7734266519546509,
+      "learning_rate": 0.00019267459906880497,
+      "loss": 1.9716,
+      "step": 718
+    },
+    {
+      "epoch": 0.07435174892065873,
+      "grad_norm": 3.2258822917938232,
+      "learning_rate": 0.00019266425245732024,
+      "loss": 2.0379,
+      "step": 719
+    },
+    {
+      "epoch": 0.07445515886352473,
+      "grad_norm": 1.35995614528656,
+      "learning_rate": 0.0001926539058458355,
+      "loss": 1.5753,
+      "step": 720
+    },
+    {
+      "epoch": 0.07455856880639074,
+      "grad_norm": 1.1354846954345703,
+      "learning_rate": 0.00019264355923435077,
+      "loss": 1.5182,
+      "step": 721
+    },
+    {
+      "epoch": 0.07466197874925674,
+      "grad_norm": 0.9102991223335266,
+      "learning_rate": 0.00019263321262286603,
+      "loss": 1.3167,
+      "step": 722
+    },
+    {
+      "epoch": 0.07476538869212275,
+      "grad_norm": 1.0951752662658691,
+      "learning_rate": 0.0001926228660113813,
+      "loss": 2.0027,
+      "step": 723
+    },
+    {
+      "epoch": 0.07486879863498876,
+      "grad_norm": 1.243021845817566,
+      "learning_rate": 0.00019261251939989656,
+      "loss": 1.8353,
+      "step": 724
+    },
+    {
+      "epoch": 0.07497220857785476,
+      "grad_norm": 1.185826063156128,
+      "learning_rate": 0.0001926021727884118,
+      "loss": 1.4974,
+      "step": 725
+    },
+    {
+      "epoch": 0.07507561852072077,
+      "grad_norm": 1.537313461303711,
+      "learning_rate": 0.00019259182617692707,
+      "loss": 1.7462,
+      "step": 726
+    },
+    {
+      "epoch": 0.07517902846358678,
+      "grad_norm": 1.8373098373413086,
+      "learning_rate": 0.00019258147956544233,
+      "loss": 2.0801,
+      "step": 727
+    },
+    {
+      "epoch": 0.07528243840645278,
+      "grad_norm": 1.0397356748580933,
+      "learning_rate": 0.0001925711329539576,
+      "loss": 1.3511,
+      "step": 728
+    },
+    {
+      "epoch": 0.07538584834931879,
+      "grad_norm": 1.2884058952331543,
+      "learning_rate": 0.00019256078634247287,
+      "loss": 2.1643,
+      "step": 729
+    },
+    {
+      "epoch": 0.0754892582921848,
+      "grad_norm": 1.4546620845794678,
+      "learning_rate": 0.00019255043973098813,
+      "loss": 2.0311,
+      "step": 730
+    },
+    {
+      "epoch": 0.0755926682350508,
+      "grad_norm": 1.3401422500610352,
+      "learning_rate": 0.00019254009311950337,
+      "loss": 2.1767,
+      "step": 731
+    },
+    {
+      "epoch": 0.0756960781779168,
+      "grad_norm": 2.024405002593994,
+      "learning_rate": 0.00019252974650801864,
+      "loss": 1.8924,
+      "step": 732
+    },
+    {
+      "epoch": 0.07579948812078281,
+      "grad_norm": 1.3837287425994873,
+      "learning_rate": 0.0001925193998965339,
+      "loss": 1.6319,
+      "step": 733
+    },
+    {
+      "epoch": 0.07590289806364882,
+      "grad_norm": 1.0432416200637817,
+      "learning_rate": 0.00019250905328504917,
+      "loss": 1.3704,
+      "step": 734
+    },
+    {
+      "epoch": 0.07600630800651482,
+      "grad_norm": 2.283970594406128,
+      "learning_rate": 0.00019249870667356443,
+      "loss": 2.1359,
+      "step": 735
+    },
+    {
+      "epoch": 0.07610971794938083,
+      "grad_norm": 1.2128592729568481,
+      "learning_rate": 0.0001924883600620797,
+      "loss": 1.5521,
+      "step": 736
+    },
+    {
+      "epoch": 0.07621312789224684,
+      "grad_norm": 1.692131757736206,
+      "learning_rate": 0.00019247801345059494,
+      "loss": 2.5598,
+      "step": 737
+    },
+    {
+      "epoch": 0.07631653783511284,
+      "grad_norm": 1.76670241355896,
+      "learning_rate": 0.0001924676668391102,
+      "loss": 1.6109,
+      "step": 738
+    },
+    {
+      "epoch": 0.07641994777797885,
+      "grad_norm": 2.0876705646514893,
+      "learning_rate": 0.00019245732022762547,
+      "loss": 1.9295,
+      "step": 739
+    },
+    {
+      "epoch": 0.07652335772084486,
+      "grad_norm": 1.60186767578125,
+      "learning_rate": 0.00019244697361614073,
+      "loss": 1.5051,
+      "step": 740
+    },
+    {
+      "epoch": 0.07662676766371086,
+      "grad_norm": 0.9031509160995483,
+      "learning_rate": 0.000192436627004656,
+      "loss": 1.8877,
+      "step": 741
+    },
+    {
+      "epoch": 0.07673017760657687,
+      "grad_norm": 2.985736131668091,
+      "learning_rate": 0.00019242628039317126,
+      "loss": 2.0977,
+      "step": 742
+    },
+    {
+      "epoch": 0.07683358754944287,
+      "grad_norm": 1.0996371507644653,
+      "learning_rate": 0.0001924159337816865,
+      "loss": 1.7847,
+      "step": 743
+    },
+    {
+      "epoch": 0.07693699749230888,
+      "grad_norm": 2.194692611694336,
+      "learning_rate": 0.00019240558717020177,
+      "loss": 2.3167,
+      "step": 744
+    },
+    {
+      "epoch": 0.07704040743517489,
+      "grad_norm": 0.8932716846466064,
+      "learning_rate": 0.00019239524055871703,
+      "loss": 1.9315,
+      "step": 745
+    },
+    {
+      "epoch": 0.0771438173780409,
+      "grad_norm": 1.4136930704116821,
+      "learning_rate": 0.0001923848939472323,
+      "loss": 1.7161,
+      "step": 746
+    },
+    {
+      "epoch": 0.0772472273209069,
+      "grad_norm": 1.594767689704895,
+      "learning_rate": 0.00019237454733574756,
+      "loss": 2.0131,
+      "step": 747
+    },
+    {
+      "epoch": 0.0773506372637729,
+      "grad_norm": 1.9824858903884888,
+      "learning_rate": 0.00019236420072426283,
+      "loss": 2.0987,
+      "step": 748
+    },
+    {
+      "epoch": 0.07745404720663891,
+      "grad_norm": 1.2606532573699951,
+      "learning_rate": 0.00019235385411277807,
+      "loss": 1.3413,
+      "step": 749
+    },
+    {
+      "epoch": 0.07755745714950492,
+      "grad_norm": 2.217545509338379,
+      "learning_rate": 0.00019234350750129333,
+      "loss": 1.7555,
+      "step": 750
+    },
+    {
+      "epoch": 0.07766086709237092,
+      "grad_norm": 1.3210011720657349,
+      "learning_rate": 0.0001923331608898086,
+      "loss": 2.0739,
+      "step": 751
+    },
+    {
+      "epoch": 0.07776427703523693,
+      "grad_norm": 1.381223440170288,
+      "learning_rate": 0.00019232281427832387,
+      "loss": 1.3535,
+      "step": 752
+    },
+    {
+      "epoch": 0.07786768697810295,
+      "grad_norm": 0.9292250275611877,
+      "learning_rate": 0.00019231246766683913,
+      "loss": 1.1796,
+      "step": 753
+    },
+    {
+      "epoch": 0.07797109692096896,
+      "grad_norm": 2.1304092407226562,
+      "learning_rate": 0.0001923021210553544,
+      "loss": 2.2769,
+      "step": 754
+    },
+    {
+      "epoch": 0.07807450686383496,
+      "grad_norm": 1.3840636014938354,
+      "learning_rate": 0.00019229177444386964,
+      "loss": 1.3881,
+      "step": 755
+    },
+    {
+      "epoch": 0.07817791680670097,
+      "grad_norm": 1.8769537210464478,
+      "learning_rate": 0.0001922814278323849,
+      "loss": 1.5471,
+      "step": 756
+    },
+    {
+      "epoch": 0.07828132674956698,
+      "grad_norm": 1.668870210647583,
+      "learning_rate": 0.00019227108122090017,
+      "loss": 2.2528,
+      "step": 757
+    },
+    {
+      "epoch": 0.07838473669243298,
+      "grad_norm": 1.6583913564682007,
+      "learning_rate": 0.00019226073460941543,
+      "loss": 1.835,
+      "step": 758
+    },
+    {
+      "epoch": 0.07848814663529899,
+      "grad_norm": 2.5908656120300293,
+      "learning_rate": 0.0001922503879979307,
+      "loss": 2.1578,
+      "step": 759
+    },
+    {
+      "epoch": 0.078591556578165,
+      "grad_norm": 0.833453357219696,
+      "learning_rate": 0.00019224004138644596,
+      "loss": 1.8684,
+      "step": 760
+    },
+    {
+      "epoch": 0.078694966521031,
+      "grad_norm": 0.9646596908569336,
+      "learning_rate": 0.0001922296947749612,
+      "loss": 1.2843,
+      "step": 761
+    },
+    {
+      "epoch": 0.07879837646389701,
+      "grad_norm": 1.4253886938095093,
+      "learning_rate": 0.00019221934816347647,
+      "loss": 1.8455,
+      "step": 762
+    },
+    {
+      "epoch": 0.07890178640676301,
+      "grad_norm": 1.565880537033081,
+      "learning_rate": 0.00019220900155199173,
+      "loss": 2.1664,
+      "step": 763
+    },
+    {
+      "epoch": 0.07900519634962902,
+      "grad_norm": 2.0248756408691406,
+      "learning_rate": 0.000192198654940507,
+      "loss": 2.2098,
+      "step": 764
+    },
+    {
+      "epoch": 0.07910860629249503,
+      "grad_norm": 1.5047415494918823,
+      "learning_rate": 0.00019218830832902226,
+      "loss": 1.9411,
+      "step": 765
+    },
+    {
+      "epoch": 0.07921201623536103,
+      "grad_norm": 3.2473227977752686,
+      "learning_rate": 0.00019217796171753753,
+      "loss": 2.1491,
+      "step": 766
+    },
+    {
+      "epoch": 0.07931542617822704,
+      "grad_norm": 1.174277424812317,
+      "learning_rate": 0.00019216761510605277,
+      "loss": 1.403,
+      "step": 767
+    },
+    {
+      "epoch": 0.07941883612109304,
+      "grad_norm": 1.26174795627594,
+      "learning_rate": 0.00019215726849456803,
+      "loss": 1.6089,
+      "step": 768
+    },
+    {
+      "epoch": 0.07952224606395905,
+      "grad_norm": 1.7186594009399414,
+      "learning_rate": 0.0001921469218830833,
+      "loss": 1.7814,
+      "step": 769
+    },
+    {
+      "epoch": 0.07962565600682506,
+      "grad_norm": 1.6304243803024292,
+      "learning_rate": 0.00019213657527159856,
+      "loss": 1.458,
+      "step": 770
+    },
+    {
+      "epoch": 0.07972906594969106,
+      "grad_norm": 1.331419825553894,
+      "learning_rate": 0.00019212622866011383,
+      "loss": 2.0749,
+      "step": 771
+    },
+    {
+      "epoch": 0.07983247589255707,
+      "grad_norm": 1.0822752714157104,
+      "learning_rate": 0.0001921158820486291,
+      "loss": 1.8906,
+      "step": 772
+    },
+    {
+      "epoch": 0.07993588583542308,
+      "grad_norm": 1.3944919109344482,
+      "learning_rate": 0.00019210553543714433,
+      "loss": 1.9101,
+      "step": 773
+    },
+    {
+      "epoch": 0.08003929577828908,
+      "grad_norm": 1.3132551908493042,
+      "learning_rate": 0.0001920951888256596,
+      "loss": 1.4764,
+      "step": 774
+    },
+    {
+      "epoch": 0.08014270572115509,
+      "grad_norm": 1.2457085847854614,
+      "learning_rate": 0.00019208484221417487,
+      "loss": 2.1376,
+      "step": 775
+    },
+    {
+      "epoch": 0.0802461156640211,
+      "grad_norm": 1.6716758012771606,
+      "learning_rate": 0.00019207449560269013,
+      "loss": 1.9652,
+      "step": 776
+    },
+    {
+      "epoch": 0.0803495256068871,
+      "grad_norm": 2.1182303428649902,
+      "learning_rate": 0.0001920641489912054,
+      "loss": 2.3112,
+      "step": 777
+    },
+    {
+      "epoch": 0.08045293554975311,
+      "grad_norm": 1.169825553894043,
+      "learning_rate": 0.00019205380237972066,
+      "loss": 1.7178,
+      "step": 778
+    },
+    {
+      "epoch": 0.08055634549261911,
+      "grad_norm": 1.0207164287567139,
+      "learning_rate": 0.0001920434557682359,
+      "loss": 1.3879,
+      "step": 779
+    },
+    {
+      "epoch": 0.08065975543548512,
+      "grad_norm": 0.6265839338302612,
+      "learning_rate": 0.00019203310915675117,
+      "loss": 1.9532,
+      "step": 780
+    },
+    {
+      "epoch": 0.08076316537835113,
+      "grad_norm": 1.588606834411621,
+      "learning_rate": 0.00019202276254526643,
+      "loss": 1.8934,
+      "step": 781
+    },
+    {
+      "epoch": 0.08086657532121713,
+      "grad_norm": 1.160733699798584,
+      "learning_rate": 0.0001920124159337817,
+      "loss": 1.7211,
+      "step": 782
+    },
+    {
+      "epoch": 0.08096998526408314,
+      "grad_norm": 0.9673109650611877,
+      "learning_rate": 0.00019200206932229696,
+      "loss": 1.7659,
+      "step": 783
+    },
+    {
+      "epoch": 0.08107339520694914,
+      "grad_norm": 1.6592562198638916,
+      "learning_rate": 0.00019199172271081223,
+      "loss": 2.2091,
+      "step": 784
+    },
+    {
+      "epoch": 0.08117680514981515,
+      "grad_norm": 1.3717565536499023,
+      "learning_rate": 0.00019198137609932747,
+      "loss": 1.7133,
+      "step": 785
+    },
+    {
+      "epoch": 0.08128021509268116,
+      "grad_norm": 1.01557457447052,
+      "learning_rate": 0.00019197102948784273,
+      "loss": 1.6405,
+      "step": 786
+    },
+    {
+      "epoch": 0.08138362503554716,
+      "grad_norm": 1.504614233970642,
+      "learning_rate": 0.000191960682876358,
+      "loss": 2.049,
+      "step": 787
+    },
+    {
+      "epoch": 0.08148703497841317,
+      "grad_norm": 1.606438159942627,
+      "learning_rate": 0.00019195033626487326,
+      "loss": 1.5679,
+      "step": 788
+    },
+    {
+      "epoch": 0.08159044492127918,
+      "grad_norm": 1.1667381525039673,
+      "learning_rate": 0.00019193998965338853,
+      "loss": 1.507,
+      "step": 789
+    },
+    {
+      "epoch": 0.08169385486414518,
+      "grad_norm": 1.2914198637008667,
+      "learning_rate": 0.0001919296430419038,
+      "loss": 1.9261,
+      "step": 790
+    },
+    {
+      "epoch": 0.08179726480701119,
+      "grad_norm": 1.4520224332809448,
+      "learning_rate": 0.00019191929643041903,
+      "loss": 1.7537,
+      "step": 791
+    },
+    {
+      "epoch": 0.0819006747498772,
+      "grad_norm": 0.614142894744873,
+      "learning_rate": 0.0001919089498189343,
+      "loss": 2.1076,
+      "step": 792
+    },
+    {
+      "epoch": 0.0820040846927432,
+      "grad_norm": 1.1694809198379517,
+      "learning_rate": 0.00019189860320744956,
+      "loss": 1.7364,
+      "step": 793
+    },
+    {
+      "epoch": 0.08210749463560921,
+      "grad_norm": 1.0019803047180176,
+      "learning_rate": 0.00019188825659596483,
+      "loss": 1.8743,
+      "step": 794
+    },
+    {
+      "epoch": 0.08221090457847521,
+      "grad_norm": 1.2464485168457031,
+      "learning_rate": 0.0001918779099844801,
+      "loss": 1.9216,
+      "step": 795
+    },
+    {
+      "epoch": 0.08231431452134123,
+      "grad_norm": 1.84528648853302,
+      "learning_rate": 0.00019186756337299536,
+      "loss": 2.1066,
+      "step": 796
+    },
+    {
+      "epoch": 0.08241772446420724,
+      "grad_norm": 1.9971226453781128,
+      "learning_rate": 0.0001918572167615106,
+      "loss": 1.7721,
+      "step": 797
+    },
+    {
+      "epoch": 0.08252113440707325,
+      "grad_norm": 1.5557864904403687,
+      "learning_rate": 0.00019184687015002586,
+      "loss": 1.8283,
+      "step": 798
+    },
+    {
+      "epoch": 0.08262454434993925,
+      "grad_norm": 1.3879731893539429,
+      "learning_rate": 0.00019183652353854113,
+      "loss": 1.9447,
+      "step": 799
+    },
+    {
+      "epoch": 0.08272795429280526,
+      "grad_norm": 1.3920087814331055,
+      "learning_rate": 0.0001918261769270564,
+      "loss": 1.9993,
+      "step": 800
+    },
+    {
+      "epoch": 0.08283136423567126,
+      "grad_norm": 1.2388873100280762,
+      "learning_rate": 0.00019181583031557166,
+      "loss": 1.4231,
+      "step": 801
+    },
+    {
+      "epoch": 0.08293477417853727,
+      "grad_norm": 0.550944983959198,
+      "learning_rate": 0.00019180548370408693,
+      "loss": 1.907,
+      "step": 802
+    },
+    {
+      "epoch": 0.08303818412140328,
+      "grad_norm": 1.2141319513320923,
+      "learning_rate": 0.00019179513709260217,
+      "loss": 1.5638,
+      "step": 803
+    },
+    {
+      "epoch": 0.08314159406426928,
+      "grad_norm": 1.085733413696289,
+      "learning_rate": 0.00019178479048111743,
+      "loss": 1.3835,
+      "step": 804
+    },
+    {
+      "epoch": 0.08324500400713529,
+      "grad_norm": 1.7833837270736694,
+      "learning_rate": 0.0001917744438696327,
+      "loss": 1.62,
+      "step": 805
+    },
+    {
+      "epoch": 0.0833484139500013,
+      "grad_norm": 1.7468349933624268,
+      "learning_rate": 0.00019176409725814796,
+      "loss": 2.0823,
+      "step": 806
+    },
+    {
+      "epoch": 0.0834518238928673,
+      "grad_norm": 2.2203943729400635,
+      "learning_rate": 0.00019175375064666323,
+      "loss": 2.3832,
+      "step": 807
+    },
+    {
+      "epoch": 0.08355523383573331,
+      "grad_norm": 1.2291004657745361,
+      "learning_rate": 0.0001917434040351785,
+      "loss": 1.3023,
+      "step": 808
+    },
+    {
+      "epoch": 0.08365864377859931,
+      "grad_norm": 0.8754696249961853,
+      "learning_rate": 0.00019173305742369373,
+      "loss": 1.7436,
+      "step": 809
+    },
+    {
+      "epoch": 0.08376205372146532,
+      "grad_norm": 1.8157202005386353,
+      "learning_rate": 0.000191722710812209,
+      "loss": 1.6954,
+      "step": 810
+    },
+    {
+      "epoch": 0.08386546366433133,
+      "grad_norm": 1.9625682830810547,
+      "learning_rate": 0.00019171236420072426,
+      "loss": 2.0076,
+      "step": 811
+    },
+    {
+      "epoch": 0.08396887360719733,
+      "grad_norm": 1.1999704837799072,
+      "learning_rate": 0.00019170201758923953,
+      "loss": 1.3628,
+      "step": 812
+    },
+    {
+      "epoch": 0.08407228355006334,
+      "grad_norm": 0.9686172008514404,
+      "learning_rate": 0.0001916916709777548,
+      "loss": 1.5073,
+      "step": 813
+    },
+    {
+      "epoch": 0.08417569349292935,
+      "grad_norm": 1.4953769445419312,
+      "learning_rate": 0.00019168132436627006,
+      "loss": 2.0297,
+      "step": 814
+    },
+    {
+      "epoch": 0.08427910343579535,
+      "grad_norm": 1.7770730257034302,
+      "learning_rate": 0.0001916709777547853,
+      "loss": 1.9318,
+      "step": 815
+    },
+    {
+      "epoch": 0.08438251337866136,
+      "grad_norm": 1.8911982774734497,
+      "learning_rate": 0.00019166063114330056,
+      "loss": 1.8299,
+      "step": 816
+    },
+    {
+      "epoch": 0.08448592332152736,
+      "grad_norm": 1.0817290544509888,
+      "learning_rate": 0.00019165028453181583,
+      "loss": 1.9069,
+      "step": 817
+    },
+    {
+      "epoch": 0.08458933326439337,
+      "grad_norm": 1.6773991584777832,
+      "learning_rate": 0.0001916399379203311,
+      "loss": 1.5428,
+      "step": 818
+    },
+    {
+      "epoch": 0.08469274320725938,
+      "grad_norm": 2.850167989730835,
+      "learning_rate": 0.00019162959130884636,
+      "loss": 1.9894,
+      "step": 819
+    },
+    {
+      "epoch": 0.08479615315012538,
+      "grad_norm": 1.3555854558944702,
+      "learning_rate": 0.00019161924469736163,
+      "loss": 2.0496,
+      "step": 820
+    },
+    {
+      "epoch": 0.08489956309299139,
+      "grad_norm": 1.4591540098190308,
+      "learning_rate": 0.00019160889808587686,
+      "loss": 1.6269,
+      "step": 821
+    },
+    {
+      "epoch": 0.0850029730358574,
+      "grad_norm": 1.1350617408752441,
+      "learning_rate": 0.00019159855147439213,
+      "loss": 1.5491,
+      "step": 822
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 1.1935083866119385,
+      "learning_rate": 0.0001915882048629074,
+      "loss": 1.8258,
+      "step": 823
+    },
+    {
+      "epoch": 0.08520979292158941,
+      "grad_norm": 1.8028544187545776,
+      "learning_rate": 0.00019157785825142266,
+      "loss": 1.374,
+      "step": 824
+    },
+    {
+      "epoch": 0.08531320286445541,
+      "grad_norm": 1.5393927097320557,
+      "learning_rate": 0.00019156751163993793,
+      "loss": 1.7656,
+      "step": 825
+    },
+    {
+      "epoch": 0.08541661280732142,
+      "grad_norm": 1.5210120677947998,
+      "learning_rate": 0.0001915571650284532,
+      "loss": 1.6707,
+      "step": 826
+    },
+    {
+      "epoch": 0.08552002275018743,
+      "grad_norm": 1.7695602178573608,
+      "learning_rate": 0.00019154681841696843,
+      "loss": 1.8571,
+      "step": 827
+    },
+    {
+      "epoch": 0.08562343269305343,
+      "grad_norm": 2.1058201789855957,
+      "learning_rate": 0.0001915364718054837,
+      "loss": 1.6645,
+      "step": 828
+    },
+    {
+      "epoch": 0.08572684263591944,
+      "grad_norm": 2.380413055419922,
+      "learning_rate": 0.00019152612519399896,
+      "loss": 2.0669,
+      "step": 829
+    },
+    {
+      "epoch": 0.08583025257878545,
+      "grad_norm": 1.2920933961868286,
+      "learning_rate": 0.00019151577858251423,
+      "loss": 1.6571,
+      "step": 830
+    },
+    {
+      "epoch": 0.08593366252165145,
+      "grad_norm": 1.4352431297302246,
+      "learning_rate": 0.0001915054319710295,
+      "loss": 1.7653,
+      "step": 831
+    },
+    {
+      "epoch": 0.08603707246451746,
+      "grad_norm": 1.5805858373641968,
+      "learning_rate": 0.00019149508535954476,
+      "loss": 1.684,
+      "step": 832
+    },
+    {
+      "epoch": 0.08614048240738346,
+      "grad_norm": 2.331017255783081,
+      "learning_rate": 0.00019148473874806002,
+      "loss": 2.1492,
+      "step": 833
+    },
+    {
+      "epoch": 0.08624389235024947,
+      "grad_norm": 1.6652169227600098,
+      "learning_rate": 0.00019147439213657526,
+      "loss": 1.5581,
+      "step": 834
+    },
+    {
+      "epoch": 0.08634730229311548,
+      "grad_norm": 0.9818973541259766,
+      "learning_rate": 0.00019146404552509053,
+      "loss": 1.4011,
+      "step": 835
+    },
+    {
+      "epoch": 0.08645071223598148,
+      "grad_norm": 1.396761178970337,
+      "learning_rate": 0.0001914536989136058,
+      "loss": 1.8978,
+      "step": 836
+    },
+    {
+      "epoch": 0.08655412217884749,
+      "grad_norm": 1.8270074129104614,
+      "learning_rate": 0.00019144335230212106,
+      "loss": 2.036,
+      "step": 837
+    },
+    {
+      "epoch": 0.0866575321217135,
+      "grad_norm": 1.182669997215271,
+      "learning_rate": 0.00019143300569063633,
+      "loss": 1.4941,
+      "step": 838
+    },
+    {
+      "epoch": 0.08676094206457952,
+      "grad_norm": 1.0010217428207397,
+      "learning_rate": 0.0001914226590791516,
+      "loss": 1.8623,
+      "step": 839
+    },
+    {
+      "epoch": 0.08686435200744552,
+      "grad_norm": 1.0710371732711792,
+      "learning_rate": 0.00019141231246766686,
+      "loss": 2.0832,
+      "step": 840
+    },
+    {
+      "epoch": 0.08696776195031153,
+      "grad_norm": 1.0800421237945557,
+      "learning_rate": 0.00019140196585618212,
+      "loss": 1.6501,
+      "step": 841
+    },
+    {
+      "epoch": 0.08707117189317753,
+      "grad_norm": 1.3268680572509766,
+      "learning_rate": 0.00019139161924469736,
+      "loss": 2.1682,
+      "step": 842
+    },
+    {
+      "epoch": 0.08717458183604354,
+      "grad_norm": 1.0593661069869995,
+      "learning_rate": 0.00019138127263321263,
+      "loss": 1.5648,
+      "step": 843
+    },
+    {
+      "epoch": 0.08727799177890955,
+      "grad_norm": 7.249431610107422,
+      "learning_rate": 0.0001913709260217279,
+      "loss": 2.1704,
+      "step": 844
+    },
+    {
+      "epoch": 0.08738140172177555,
+      "grad_norm": 1.202315092086792,
+      "learning_rate": 0.00019136057941024316,
+      "loss": 1.5081,
+      "step": 845
+    },
+    {
+      "epoch": 0.08748481166464156,
+      "grad_norm": 3.4483230113983154,
+      "learning_rate": 0.00019135023279875842,
+      "loss": 1.5247,
+      "step": 846
+    },
+    {
+      "epoch": 0.08758822160750757,
+      "grad_norm": 2.4174540042877197,
+      "learning_rate": 0.0001913398861872737,
+      "loss": 1.9511,
+      "step": 847
+    },
+    {
+      "epoch": 0.08769163155037357,
+      "grad_norm": 1.351784586906433,
+      "learning_rate": 0.00019132953957578895,
+      "loss": 1.4813,
+      "step": 848
+    },
+    {
+      "epoch": 0.08779504149323958,
+      "grad_norm": 1.2987369298934937,
+      "learning_rate": 0.00019131919296430422,
+      "loss": 1.7586,
+      "step": 849
+    },
+    {
+      "epoch": 0.08789845143610558,
+      "grad_norm": 3.5988855361938477,
+      "learning_rate": 0.00019130884635281946,
+      "loss": 2.0485,
+      "step": 850
+    },
+    {
+      "epoch": 0.08800186137897159,
+      "grad_norm": 0.9274793863296509,
+      "learning_rate": 0.00019129849974133472,
+      "loss": 1.5124,
+      "step": 851
+    },
+    {
+      "epoch": 0.0881052713218376,
+      "grad_norm": 2.8831684589385986,
+      "learning_rate": 0.00019128815312985,
+      "loss": 1.9516,
+      "step": 852
+    },
+    {
+      "epoch": 0.0882086812647036,
+      "grad_norm": 1.241784691810608,
+      "learning_rate": 0.00019127780651836525,
+      "loss": 1.998,
+      "step": 853
+    },
+    {
+      "epoch": 0.08831209120756961,
+      "grad_norm": 1.2777564525604248,
+      "learning_rate": 0.00019126745990688052,
+      "loss": 1.6914,
+      "step": 854
+    },
+    {
+      "epoch": 0.08841550115043562,
+      "grad_norm": 1.3454850912094116,
+      "learning_rate": 0.00019125711329539579,
+      "loss": 2.0849,
+      "step": 855
+    },
+    {
+      "epoch": 0.08851891109330162,
+      "grad_norm": 1.694756031036377,
+      "learning_rate": 0.00019124676668391105,
+      "loss": 1.4139,
+      "step": 856
+    },
+    {
+      "epoch": 0.08862232103616763,
+      "grad_norm": 1.0123131275177002,
+      "learning_rate": 0.0001912364200724263,
+      "loss": 1.4719,
+      "step": 857
+    },
+    {
+      "epoch": 0.08872573097903363,
+      "grad_norm": 2.1002085208892822,
+      "learning_rate": 0.00019122607346094156,
+      "loss": 1.4789,
+      "step": 858
+    },
+    {
+      "epoch": 0.08882914092189964,
+      "grad_norm": 1.3821834325790405,
+      "learning_rate": 0.00019121572684945682,
+      "loss": 1.8582,
+      "step": 859
+    },
+    {
+      "epoch": 0.08893255086476565,
+      "grad_norm": 0.9540427923202515,
+      "learning_rate": 0.00019120538023797209,
+      "loss": 1.2874,
+      "step": 860
+    },
+    {
+      "epoch": 0.08903596080763165,
+      "grad_norm": 1.8231464624404907,
+      "learning_rate": 0.00019119503362648735,
+      "loss": 1.5624,
+      "step": 861
+    },
+    {
+      "epoch": 0.08913937075049766,
+      "grad_norm": 1.7630882263183594,
+      "learning_rate": 0.00019118468701500262,
+      "loss": 2.1599,
+      "step": 862
+    },
+    {
+      "epoch": 0.08924278069336367,
+      "grad_norm": 2.143895387649536,
+      "learning_rate": 0.00019117434040351786,
+      "loss": 2.201,
+      "step": 863
+    },
+    {
+      "epoch": 0.08934619063622967,
+      "grad_norm": 1.8420048952102661,
+      "learning_rate": 0.00019116399379203312,
+      "loss": 1.4954,
+      "step": 864
+    },
+    {
+      "epoch": 0.08944960057909568,
+      "grad_norm": 0.9784311056137085,
+      "learning_rate": 0.0001911536471805484,
+      "loss": 1.9278,
+      "step": 865
+    },
+    {
+      "epoch": 0.08955301052196168,
+      "grad_norm": 1.2139637470245361,
+      "learning_rate": 0.00019114330056906365,
+      "loss": 1.5343,
+      "step": 866
+    },
+    {
+      "epoch": 0.08965642046482769,
+      "grad_norm": 1.4579898118972778,
+      "learning_rate": 0.00019113295395757892,
+      "loss": 1.7,
+      "step": 867
+    },
+    {
+      "epoch": 0.0897598304076937,
+      "grad_norm": 2.5946152210235596,
+      "learning_rate": 0.00019112260734609418,
+      "loss": 1.2391,
+      "step": 868
+    },
+    {
+      "epoch": 0.0898632403505597,
+      "grad_norm": 1.2225054502487183,
+      "learning_rate": 0.00019111226073460942,
+      "loss": 1.558,
+      "step": 869
+    },
+    {
+      "epoch": 0.08996665029342571,
+      "grad_norm": 1.6341155767440796,
+      "learning_rate": 0.0001911019141231247,
+      "loss": 1.4779,
+      "step": 870
+    },
+    {
+      "epoch": 0.09007006023629172,
+      "grad_norm": 0.8429521918296814,
+      "learning_rate": 0.00019109156751163995,
+      "loss": 2.0981,
+      "step": 871
+    },
+    {
+      "epoch": 0.09017347017915772,
+      "grad_norm": 1.4018882513046265,
+      "learning_rate": 0.00019108122090015522,
+      "loss": 1.6935,
+      "step": 872
+    },
+    {
+      "epoch": 0.09027688012202373,
+      "grad_norm": 1.764603853225708,
+      "learning_rate": 0.00019107087428867048,
+      "loss": 1.6275,
+      "step": 873
+    },
+    {
+      "epoch": 0.09038029006488973,
+      "grad_norm": 2.3118224143981934,
+      "learning_rate": 0.00019106052767718575,
+      "loss": 1.8666,
+      "step": 874
+    },
+    {
+      "epoch": 0.09048370000775574,
+      "grad_norm": 1.2632291316986084,
+      "learning_rate": 0.000191050181065701,
+      "loss": 1.6197,
+      "step": 875
+    },
+    {
+      "epoch": 0.09058710995062175,
+      "grad_norm": 1.9582924842834473,
+      "learning_rate": 0.00019103983445421625,
+      "loss": 1.5178,
+      "step": 876
+    },
+    {
+      "epoch": 0.09069051989348775,
+      "grad_norm": 1.8235286474227905,
+      "learning_rate": 0.00019102948784273152,
+      "loss": 1.7107,
+      "step": 877
+    },
+    {
+      "epoch": 0.09079392983635376,
+      "grad_norm": 0.9742588400840759,
+      "learning_rate": 0.00019101914123124679,
+      "loss": 1.3714,
+      "step": 878
+    },
+    {
+      "epoch": 0.09089733977921977,
+      "grad_norm": 1.9299086332321167,
+      "learning_rate": 0.00019100879461976205,
+      "loss": 2.1482,
+      "step": 879
+    },
+    {
+      "epoch": 0.09100074972208577,
+      "grad_norm": 1.1086747646331787,
+      "learning_rate": 0.00019099844800827732,
+      "loss": 1.5429,
+      "step": 880
+    },
+    {
+      "epoch": 0.09110415966495178,
+      "grad_norm": 0.9190935492515564,
+      "learning_rate": 0.00019098810139679255,
+      "loss": 1.5422,
+      "step": 881
+    },
+    {
+      "epoch": 0.0912075696078178,
+      "grad_norm": 1.3607590198516846,
+      "learning_rate": 0.00019097775478530782,
+      "loss": 1.5797,
+      "step": 882
+    },
+    {
+      "epoch": 0.0913109795506838,
+      "grad_norm": 1.1328152418136597,
+      "learning_rate": 0.00019096740817382309,
+      "loss": 1.6915,
+      "step": 883
+    },
+    {
+      "epoch": 0.09141438949354981,
+      "grad_norm": 2.0909512042999268,
+      "learning_rate": 0.00019095706156233835,
+      "loss": 1.6088,
+      "step": 884
+    },
+    {
+      "epoch": 0.09151779943641582,
+      "grad_norm": 5.501047134399414,
+      "learning_rate": 0.00019094671495085362,
+      "loss": 1.5214,
+      "step": 885
+    },
+    {
+      "epoch": 0.09162120937928182,
+      "grad_norm": 1.4709688425064087,
+      "learning_rate": 0.00019093636833936888,
+      "loss": 1.8307,
+      "step": 886
+    },
+    {
+      "epoch": 0.09172461932214783,
+      "grad_norm": 1.181794285774231,
+      "learning_rate": 0.00019092602172788412,
+      "loss": 2.4404,
+      "step": 887
+    },
+    {
+      "epoch": 0.09182802926501384,
+      "grad_norm": 1.3347965478897095,
+      "learning_rate": 0.0001909156751163994,
+      "loss": 1.6549,
+      "step": 888
+    },
+    {
+      "epoch": 0.09193143920787984,
+      "grad_norm": 1.237500786781311,
+      "learning_rate": 0.00019090532850491465,
+      "loss": 1.4831,
+      "step": 889
+    },
+    {
+      "epoch": 0.09203484915074585,
+      "grad_norm": 3.037428140640259,
+      "learning_rate": 0.00019089498189342992,
+      "loss": 2.5507,
+      "step": 890
+    },
+    {
+      "epoch": 0.09213825909361185,
+      "grad_norm": 1.1561611890792847,
+      "learning_rate": 0.00019088463528194518,
+      "loss": 1.5395,
+      "step": 891
+    },
+    {
+      "epoch": 0.09224166903647786,
+      "grad_norm": 0.9404453039169312,
+      "learning_rate": 0.00019087428867046045,
+      "loss": 1.3764,
+      "step": 892
+    },
+    {
+      "epoch": 0.09234507897934387,
+      "grad_norm": 0.9672577381134033,
+      "learning_rate": 0.0001908639420589757,
+      "loss": 1.4578,
+      "step": 893
+    },
+    {
+      "epoch": 0.09244848892220987,
+      "grad_norm": 1.6051990985870361,
+      "learning_rate": 0.00019085359544749095,
+      "loss": 1.9744,
+      "step": 894
+    },
+    {
+      "epoch": 0.09255189886507588,
+      "grad_norm": 1.4234753847122192,
+      "learning_rate": 0.00019084324883600622,
+      "loss": 1.904,
+      "step": 895
+    },
+    {
+      "epoch": 0.09265530880794189,
+      "grad_norm": 1.7959076166152954,
+      "learning_rate": 0.00019083290222452148,
+      "loss": 1.6987,
+      "step": 896
+    },
+    {
+      "epoch": 0.09275871875080789,
+      "grad_norm": 2.001089572906494,
+      "learning_rate": 0.00019082255561303675,
+      "loss": 1.8458,
+      "step": 897
+    },
+    {
+      "epoch": 0.0928621286936739,
+      "grad_norm": 1.192054271697998,
+      "learning_rate": 0.00019081220900155202,
+      "loss": 1.8139,
+      "step": 898
+    },
+    {
+      "epoch": 0.0929655386365399,
+      "grad_norm": 1.8499089479446411,
+      "learning_rate": 0.00019080186239006725,
+      "loss": 1.6258,
+      "step": 899
+    },
+    {
+      "epoch": 0.09306894857940591,
+      "grad_norm": 1.2967066764831543,
+      "learning_rate": 0.00019079151577858252,
+      "loss": 1.8121,
+      "step": 900
+    },
+    {
+      "epoch": 0.09317235852227192,
+      "grad_norm": 0.9628750681877136,
+      "learning_rate": 0.00019078116916709779,
+      "loss": 1.9695,
+      "step": 901
+    },
+    {
+      "epoch": 0.09327576846513792,
+      "grad_norm": 1.2634750604629517,
+      "learning_rate": 0.00019077082255561305,
+      "loss": 2.051,
+      "step": 902
+    },
+    {
+      "epoch": 0.09337917840800393,
+      "grad_norm": 0.887198805809021,
+      "learning_rate": 0.00019076047594412832,
+      "loss": 1.6507,
+      "step": 903
+    },
+    {
+      "epoch": 0.09348258835086994,
+      "grad_norm": 1.443992257118225,
+      "learning_rate": 0.00019075012933264358,
+      "loss": 1.5173,
+      "step": 904
+    },
+    {
+      "epoch": 0.09358599829373594,
+      "grad_norm": 2.711820125579834,
+      "learning_rate": 0.00019073978272115882,
+      "loss": 2.4954,
+      "step": 905
+    },
+    {
+      "epoch": 0.09368940823660195,
+      "grad_norm": 1.4171301126480103,
+      "learning_rate": 0.00019072943610967409,
+      "loss": 1.5614,
+      "step": 906
+    },
+    {
+      "epoch": 0.09379281817946795,
+      "grad_norm": 1.58612859249115,
+      "learning_rate": 0.00019071908949818935,
+      "loss": 1.9968,
+      "step": 907
+    },
+    {
+      "epoch": 0.09389622812233396,
+      "grad_norm": 2.6168100833892822,
+      "learning_rate": 0.00019070874288670462,
+      "loss": 1.9727,
+      "step": 908
+    },
+    {
+      "epoch": 0.09399963806519997,
+      "grad_norm": 1.3814314603805542,
+      "learning_rate": 0.00019069839627521988,
+      "loss": 1.4318,
+      "step": 909
+    },
+    {
+      "epoch": 0.09410304800806597,
+      "grad_norm": 1.9172375202178955,
+      "learning_rate": 0.00019068804966373515,
+      "loss": 1.8765,
+      "step": 910
+    },
+    {
+      "epoch": 0.09420645795093198,
+      "grad_norm": 0.6962313055992126,
+      "learning_rate": 0.0001906777030522504,
+      "loss": 1.9879,
+      "step": 911
+    },
+    {
+      "epoch": 0.09430986789379799,
+      "grad_norm": 2.2384886741638184,
+      "learning_rate": 0.00019066735644076565,
+      "loss": 2.0699,
+      "step": 912
+    },
+    {
+      "epoch": 0.09441327783666399,
+      "grad_norm": 1.6673318147659302,
+      "learning_rate": 0.00019065700982928092,
+      "loss": 1.8016,
+      "step": 913
+    },
+    {
+      "epoch": 0.09451668777953,
+      "grad_norm": 1.3593833446502686,
+      "learning_rate": 0.00019064666321779618,
+      "loss": 1.5257,
+      "step": 914
+    },
+    {
+      "epoch": 0.094620097722396,
+      "grad_norm": 1.5814733505249023,
+      "learning_rate": 0.00019063631660631145,
+      "loss": 1.796,
+      "step": 915
+    },
+    {
+      "epoch": 0.09472350766526201,
+      "grad_norm": 2.525911569595337,
+      "learning_rate": 0.00019062596999482671,
+      "loss": 2.0192,
+      "step": 916
+    },
+    {
+      "epoch": 0.09482691760812802,
+      "grad_norm": 1.1196669340133667,
+      "learning_rate": 0.00019061562338334195,
+      "loss": 1.7643,
+      "step": 917
+    },
+    {
+      "epoch": 0.09493032755099402,
+      "grad_norm": 1.3181729316711426,
+      "learning_rate": 0.00019060527677185722,
+      "loss": 1.4779,
+      "step": 918
+    },
+    {
+      "epoch": 0.09503373749386003,
+      "grad_norm": 1.9003279209136963,
+      "learning_rate": 0.00019059493016037248,
+      "loss": 1.8914,
+      "step": 919
+    },
+    {
+      "epoch": 0.09513714743672604,
+      "grad_norm": 1.2940086126327515,
+      "learning_rate": 0.00019058458354888775,
+      "loss": 1.5011,
+      "step": 920
+    },
+    {
+      "epoch": 0.09524055737959204,
+      "grad_norm": 1.673016905784607,
+      "learning_rate": 0.00019057423693740302,
+      "loss": 1.9298,
+      "step": 921
+    },
+    {
+      "epoch": 0.09534396732245805,
+      "grad_norm": 2.4916605949401855,
+      "learning_rate": 0.00019056389032591828,
+      "loss": 1.4196,
+      "step": 922
+    },
+    {
+      "epoch": 0.09544737726532405,
+      "grad_norm": 0.787038266658783,
+      "learning_rate": 0.00019055354371443352,
+      "loss": 1.9657,
+      "step": 923
+    },
+    {
+      "epoch": 0.09555078720819006,
+      "grad_norm": 2.5524959564208984,
+      "learning_rate": 0.00019054319710294878,
+      "loss": 1.543,
+      "step": 924
+    },
+    {
+      "epoch": 0.09565419715105608,
+      "grad_norm": 1.0950359106063843,
+      "learning_rate": 0.00019053285049146405,
+      "loss": 1.0407,
+      "step": 925
+    },
+    {
+      "epoch": 0.09575760709392209,
+      "grad_norm": 1.2391451597213745,
+      "learning_rate": 0.00019052250387997932,
+      "loss": 1.2096,
+      "step": 926
+    },
+    {
+      "epoch": 0.0958610170367881,
+      "grad_norm": 0.8148314952850342,
+      "learning_rate": 0.00019051215726849458,
+      "loss": 1.675,
+      "step": 927
+    },
+    {
+      "epoch": 0.0959644269796541,
+      "grad_norm": 2.42737078666687,
+      "learning_rate": 0.00019050181065700985,
+      "loss": 1.4995,
+      "step": 928
+    },
+    {
+      "epoch": 0.0960678369225201,
+      "grad_norm": 1.2046351432800293,
+      "learning_rate": 0.00019049146404552509,
+      "loss": 1.5453,
+      "step": 929
+    },
+    {
+      "epoch": 0.09617124686538611,
+      "grad_norm": 1.1635297536849976,
+      "learning_rate": 0.00019048111743404035,
+      "loss": 1.6753,
+      "step": 930
+    },
+    {
+      "epoch": 0.09627465680825212,
+      "grad_norm": 0.8624559640884399,
+      "learning_rate": 0.00019047077082255562,
+      "loss": 1.9709,
+      "step": 931
+    },
+    {
+      "epoch": 0.09637806675111812,
+      "grad_norm": 2.0648558139801025,
+      "learning_rate": 0.00019046042421107088,
+      "loss": 1.93,
+      "step": 932
+    },
+    {
+      "epoch": 0.09648147669398413,
+      "grad_norm": 1.3905847072601318,
+      "learning_rate": 0.00019045007759958615,
+      "loss": 1.309,
+      "step": 933
+    },
+    {
+      "epoch": 0.09658488663685014,
+      "grad_norm": 0.8313616514205933,
+      "learning_rate": 0.0001904397309881014,
+      "loss": 1.8303,
+      "step": 934
+    },
+    {
+      "epoch": 0.09668829657971614,
+      "grad_norm": 1.6961098909378052,
+      "learning_rate": 0.00019042938437661665,
+      "loss": 1.5906,
+      "step": 935
+    },
+    {
+      "epoch": 0.09679170652258215,
+      "grad_norm": 2.953028917312622,
+      "learning_rate": 0.00019041903776513192,
+      "loss": 1.5128,
+      "step": 936
+    },
+    {
+      "epoch": 0.09689511646544816,
+      "grad_norm": 1.1944273710250854,
+      "learning_rate": 0.00019040869115364718,
+      "loss": 1.2601,
+      "step": 937
+    },
+    {
+      "epoch": 0.09699852640831416,
+      "grad_norm": 1.3880339860916138,
+      "learning_rate": 0.00019039834454216245,
+      "loss": 1.3335,
+      "step": 938
+    },
+    {
+      "epoch": 0.09710193635118017,
+      "grad_norm": 3.0873427391052246,
+      "learning_rate": 0.00019038799793067771,
+      "loss": 2.2271,
+      "step": 939
+    },
+    {
+      "epoch": 0.09720534629404617,
+      "grad_norm": 1.1165976524353027,
+      "learning_rate": 0.00019037765131919298,
+      "loss": 0.925,
+      "step": 940
+    },
+    {
+      "epoch": 0.09730875623691218,
+      "grad_norm": 1.3360462188720703,
+      "learning_rate": 0.00019036730470770822,
+      "loss": 1.9376,
+      "step": 941
+    },
+    {
+      "epoch": 0.09741216617977819,
+      "grad_norm": 1.3079710006713867,
+      "learning_rate": 0.00019035695809622348,
+      "loss": 2.2023,
+      "step": 942
+    },
+    {
+      "epoch": 0.09751557612264419,
+      "grad_norm": 3.9090816974639893,
+      "learning_rate": 0.00019034661148473875,
+      "loss": 1.9705,
+      "step": 943
+    },
+    {
+      "epoch": 0.0976189860655102,
+      "grad_norm": 1.266180396080017,
+      "learning_rate": 0.00019033626487325401,
+      "loss": 2.0189,
+      "step": 944
+    },
+    {
+      "epoch": 0.0977223960083762,
+      "grad_norm": 1.4128590822219849,
+      "learning_rate": 0.00019032591826176928,
+      "loss": 1.9973,
+      "step": 945
+    },
+    {
+      "epoch": 0.09782580595124221,
+      "grad_norm": 1.5954068899154663,
+      "learning_rate": 0.00019031557165028452,
+      "loss": 1.6339,
+      "step": 946
+    },
+    {
+      "epoch": 0.09792921589410822,
+      "grad_norm": 2.1788735389709473,
+      "learning_rate": 0.00019030522503879978,
+      "loss": 1.2789,
+      "step": 947
+    },
+    {
+      "epoch": 0.09803262583697422,
+      "grad_norm": 2.4323537349700928,
+      "learning_rate": 0.00019029487842731505,
+      "loss": 1.3597,
+      "step": 948
+    },
+    {
+      "epoch": 0.09813603577984023,
+      "grad_norm": 1.620073676109314,
+      "learning_rate": 0.00019028453181583032,
+      "loss": 1.7057,
+      "step": 949
+    },
+    {
+      "epoch": 0.09823944572270624,
+      "grad_norm": 1.2310841083526611,
+      "learning_rate": 0.00019027418520434558,
+      "loss": 1.7842,
+      "step": 950
+    },
+    {
+      "epoch": 0.09834285566557224,
+      "grad_norm": 2.3200294971466064,
+      "learning_rate": 0.00019026383859286085,
+      "loss": 1.3738,
+      "step": 951
+    },
+    {
+      "epoch": 0.09844626560843825,
+      "grad_norm": 1.0328644514083862,
+      "learning_rate": 0.00019025349198137609,
+      "loss": 1.1602,
+      "step": 952
+    },
+    {
+      "epoch": 0.09854967555130426,
+      "grad_norm": 3.512075662612915,
+      "learning_rate": 0.00019024314536989135,
+      "loss": 1.6666,
+      "step": 953
+    },
+    {
+      "epoch": 0.09865308549417026,
+      "grad_norm": 2.2993133068084717,
+      "learning_rate": 0.00019023279875840662,
+      "loss": 2.9702,
+      "step": 954
+    },
+    {
+      "epoch": 0.09875649543703627,
+      "grad_norm": 1.472238302230835,
+      "learning_rate": 0.00019022245214692188,
+      "loss": 1.7172,
+      "step": 955
+    },
+    {
+      "epoch": 0.09885990537990227,
+      "grad_norm": 1.6245614290237427,
+      "learning_rate": 0.00019021210553543715,
+      "loss": 2.1302,
+      "step": 956
+    },
+    {
+      "epoch": 0.09896331532276828,
+      "grad_norm": 2.2646124362945557,
+      "learning_rate": 0.0001902017589239524,
+      "loss": 1.2524,
+      "step": 957
+    },
+    {
+      "epoch": 0.09906672526563429,
+      "grad_norm": 1.8233280181884766,
+      "learning_rate": 0.00019019141231246765,
+      "loss": 1.6103,
+      "step": 958
+    },
+    {
+      "epoch": 0.09917013520850029,
+      "grad_norm": 1.894910216331482,
+      "learning_rate": 0.00019018106570098292,
+      "loss": 1.8498,
+      "step": 959
+    },
+    {
+      "epoch": 0.0992735451513663,
+      "grad_norm": 2.561936378479004,
+      "learning_rate": 0.00019017071908949818,
+      "loss": 2.3839,
+      "step": 960
+    },
+    {
+      "epoch": 0.0993769550942323,
+      "grad_norm": 1.281097650527954,
+      "learning_rate": 0.00019016037247801345,
+      "loss": 1.4926,
+      "step": 961
+    },
+    {
+      "epoch": 0.09948036503709831,
+      "grad_norm": 1.704836130142212,
+      "learning_rate": 0.00019015002586652871,
+      "loss": 1.9475,
+      "step": 962
+    },
+    {
+      "epoch": 0.09958377497996432,
+      "grad_norm": 1.2036210298538208,
+      "learning_rate": 0.00019013967925504398,
+      "loss": 1.5725,
+      "step": 963
+    },
+    {
+      "epoch": 0.09968718492283032,
+      "grad_norm": 0.9029375314712524,
+      "learning_rate": 0.00019012933264355925,
+      "loss": 1.3477,
+      "step": 964
+    },
+    {
+      "epoch": 0.09979059486569633,
+      "grad_norm": 1.436606526374817,
+      "learning_rate": 0.0001901189860320745,
+      "loss": 1.7976,
+      "step": 965
+    },
+    {
+      "epoch": 0.09989400480856234,
+      "grad_norm": 1.4428085088729858,
+      "learning_rate": 0.00019010863942058975,
+      "loss": 1.8221,
+      "step": 966
+    },
+    {
+      "epoch": 0.09999741475142834,
+      "grad_norm": 1.1217411756515503,
+      "learning_rate": 0.00019009829280910501,
+      "loss": 1.3546,
+      "step": 967
+    },
+    {
+      "epoch": 0.10010082469429436,
+      "grad_norm": 1.180086612701416,
+      "learning_rate": 0.00019008794619762028,
+      "loss": 2.22,
+      "step": 968
+    },
+    {
+      "epoch": 0.10020423463716037,
+      "grad_norm": 2.1352503299713135,
+      "learning_rate": 0.00019007759958613555,
+      "loss": 2.0902,
+      "step": 969
+    },
+    {
+      "epoch": 0.10030764458002638,
+      "grad_norm": 0.9205876588821411,
+      "learning_rate": 0.0001900672529746508,
+      "loss": 1.7533,
+      "step": 970
+    },
+    {
+      "epoch": 0.10041105452289238,
+      "grad_norm": 1.8197256326675415,
+      "learning_rate": 0.00019005690636316608,
+      "loss": 1.2994,
+      "step": 971
+    },
+    {
+      "epoch": 0.10051446446575839,
+      "grad_norm": 2.683286428451538,
+      "learning_rate": 0.00019004655975168134,
+      "loss": 2.3546,
+      "step": 972
+    },
+    {
+      "epoch": 0.1006178744086244,
+      "grad_norm": 1.520284652709961,
+      "learning_rate": 0.0001900362131401966,
+      "loss": 1.4579,
+      "step": 973
+    },
+    {
+      "epoch": 0.1007212843514904,
+      "grad_norm": 1.7554701566696167,
+      "learning_rate": 0.00019002586652871185,
+      "loss": 1.664,
+      "step": 974
+    },
+    {
+      "epoch": 0.1008246942943564,
+      "grad_norm": 0.9988136887550354,
+      "learning_rate": 0.0001900155199172271,
+      "loss": 1.6349,
+      "step": 975
+    },
+    {
+      "epoch": 0.10092810423722241,
+      "grad_norm": 1.7587372064590454,
+      "learning_rate": 0.00019000517330574238,
+      "loss": 1.3191,
+      "step": 976
+    },
+    {
+      "epoch": 0.10103151418008842,
+      "grad_norm": 0.7632730603218079,
+      "learning_rate": 0.00018999482669425764,
+      "loss": 1.8607,
+      "step": 977
+    },
+    {
+      "epoch": 0.10113492412295443,
+      "grad_norm": 1.414077639579773,
+      "learning_rate": 0.0001899844800827729,
+      "loss": 1.6789,
+      "step": 978
+    },
+    {
+      "epoch": 0.10123833406582043,
+      "grad_norm": 0.9060714840888977,
+      "learning_rate": 0.00018997413347128817,
+      "loss": 1.9993,
+      "step": 979
+    },
+    {
+      "epoch": 0.10134174400868644,
+      "grad_norm": 1.177000641822815,
+      "learning_rate": 0.00018996378685980344,
+      "loss": 1.6802,
+      "step": 980
+    },
+    {
+      "epoch": 0.10144515395155244,
+      "grad_norm": 0.9905523061752319,
+      "learning_rate": 0.0001899534402483187,
+      "loss": 1.7459,
+      "step": 981
+    },
+    {
+      "epoch": 0.10154856389441845,
+      "grad_norm": 1.1813868284225464,
+      "learning_rate": 0.00018994309363683394,
+      "loss": 1.4899,
+      "step": 982
+    },
+    {
+      "epoch": 0.10165197383728446,
+      "grad_norm": 1.2635287046432495,
+      "learning_rate": 0.0001899327470253492,
+      "loss": 1.53,
+      "step": 983
+    },
+    {
+      "epoch": 0.10175538378015046,
+      "grad_norm": 0.9415909051895142,
+      "learning_rate": 0.00018992240041386448,
+      "loss": 1.7562,
+      "step": 984
+    },
+    {
+      "epoch": 0.10185879372301647,
+      "grad_norm": 1.4737257957458496,
+      "learning_rate": 0.00018991205380237974,
+      "loss": 1.5612,
+      "step": 985
+    },
+    {
+      "epoch": 0.10196220366588248,
+      "grad_norm": 1.445164680480957,
+      "learning_rate": 0.000189901707190895,
+      "loss": 1.6512,
+      "step": 986
+    },
+    {
+      "epoch": 0.10206561360874848,
+      "grad_norm": 1.29860520362854,
+      "learning_rate": 0.00018989136057941027,
+      "loss": 1.903,
+      "step": 987
+    },
+    {
+      "epoch": 0.10216902355161449,
+      "grad_norm": 1.0900709629058838,
+      "learning_rate": 0.0001898810139679255,
+      "loss": 1.4451,
+      "step": 988
+    },
+    {
+      "epoch": 0.1022724334944805,
+      "grad_norm": 1.3902963399887085,
+      "learning_rate": 0.00018987066735644078,
+      "loss": 1.8913,
+      "step": 989
+    },
+    {
+      "epoch": 0.1023758434373465,
+      "grad_norm": 2.240596055984497,
+      "learning_rate": 0.00018986032074495604,
+      "loss": 1.9735,
+      "step": 990
+    },
+    {
+      "epoch": 0.1024792533802125,
+      "grad_norm": 3.0811617374420166,
+      "learning_rate": 0.0001898499741334713,
+      "loss": 1.7042,
+      "step": 991
+    },
+    {
+      "epoch": 0.10258266332307851,
+      "grad_norm": 1.270492434501648,
+      "learning_rate": 0.00018983962752198657,
+      "loss": 1.541,
+      "step": 992
+    },
+    {
+      "epoch": 0.10268607326594452,
+      "grad_norm": 1.7597604990005493,
+      "learning_rate": 0.00018982928091050184,
+      "loss": 2.216,
+      "step": 993
+    },
+    {
+      "epoch": 0.10278948320881053,
+      "grad_norm": 1.8272831439971924,
+      "learning_rate": 0.00018981893429901708,
+      "loss": 1.5311,
+      "step": 994
+    },
+    {
+      "epoch": 0.10289289315167653,
+      "grad_norm": 1.4568592309951782,
+      "learning_rate": 0.00018980858768753234,
+      "loss": 1.3844,
+      "step": 995
+    },
+    {
+      "epoch": 0.10299630309454254,
+      "grad_norm": 1.969649076461792,
+      "learning_rate": 0.0001897982410760476,
+      "loss": 1.8859,
+      "step": 996
+    },
+    {
+      "epoch": 0.10309971303740854,
+      "grad_norm": 1.2497482299804688,
+      "learning_rate": 0.00018978789446456287,
+      "loss": 1.455,
+      "step": 997
+    },
+    {
+      "epoch": 0.10320312298027455,
+      "grad_norm": 2.641781806945801,
+      "learning_rate": 0.00018977754785307814,
+      "loss": 2.2136,
+      "step": 998
+    },
+    {
+      "epoch": 0.10330653292314056,
+      "grad_norm": 0.9556031227111816,
+      "learning_rate": 0.0001897672012415934,
+      "loss": 1.3249,
+      "step": 999
+    },
+    {
+      "epoch": 0.10340994286600656,
+      "grad_norm": 1.0540037155151367,
+      "learning_rate": 0.00018975685463010864,
+      "loss": 1.1152,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10351335280887257,
+      "grad_norm": 1.8548526763916016,
+      "learning_rate": 0.0001897465080186239,
+      "loss": 1.6326,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10361676275173858,
+      "grad_norm": 1.2354981899261475,
+      "learning_rate": 0.00018973616140713917,
+      "loss": 1.9958,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10372017269460458,
+      "grad_norm": 1.3923211097717285,
+      "learning_rate": 0.00018972581479565444,
+      "loss": 1.601,
+      "step": 1003
+    },
+    {
+      "epoch": 0.10382358263747059,
+      "grad_norm": 0.9945183396339417,
+      "learning_rate": 0.0001897154681841697,
+      "loss": 1.5537,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1039269925803366,
+      "grad_norm": 2.5964150428771973,
+      "learning_rate": 0.00018970512157268497,
+      "loss": 2.0066,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1040304025232026,
+      "grad_norm": 1.4605211019515991,
+      "learning_rate": 0.0001896947749612002,
+      "loss": 1.6974,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1041338124660686,
+      "grad_norm": 0.8577373623847961,
+      "learning_rate": 0.00018968442834971547,
+      "loss": 2.3723,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10423722240893461,
+      "grad_norm": 1.6100918054580688,
+      "learning_rate": 0.00018967408173823074,
+      "loss": 1.8351,
+      "step": 1008
+    },
+    {
+      "epoch": 0.10434063235180062,
+      "grad_norm": 1.1363835334777832,
+      "learning_rate": 0.000189663735126746,
+      "loss": 2.0254,
+      "step": 1009
+    },
+    {
+      "epoch": 0.10444404229466663,
+      "grad_norm": 1.3265516757965088,
+      "learning_rate": 0.00018965338851526127,
+      "loss": 1.8254,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10454745223753265,
+      "grad_norm": 1.4067211151123047,
+      "learning_rate": 0.00018964304190377654,
+      "loss": 1.2216,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10465086218039865,
+      "grad_norm": 1.7004269361495972,
+      "learning_rate": 0.00018963269529229178,
+      "loss": 1.3757,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10475427212326466,
+      "grad_norm": 1.0695922374725342,
+      "learning_rate": 0.00018962234868080704,
+      "loss": 1.2822,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10485768206613066,
+      "grad_norm": 1.6438508033752441,
+      "learning_rate": 0.0001896120020693223,
+      "loss": 1.7428,
+      "step": 1014
+    },
+    {
+      "epoch": 0.10496109200899667,
+      "grad_norm": 1.1417630910873413,
+      "learning_rate": 0.00018960165545783757,
+      "loss": 2.2032,
+      "step": 1015
+    },
+    {
+      "epoch": 0.10506450195186268,
+      "grad_norm": 4.685209274291992,
+      "learning_rate": 0.00018959130884635284,
+      "loss": 1.7533,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10516791189472868,
+      "grad_norm": 1.016170859336853,
+      "learning_rate": 0.0001895809622348681,
+      "loss": 2.2786,
+      "step": 1017
+    },
+    {
+      "epoch": 0.10527132183759469,
+      "grad_norm": 2.0337867736816406,
+      "learning_rate": 0.00018957061562338334,
+      "loss": 1.9138,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1053747317804607,
+      "grad_norm": 1.1036525964736938,
+      "learning_rate": 0.0001895602690118986,
+      "loss": 1.5866,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1054781417233267,
+      "grad_norm": 1.8529484272003174,
+      "learning_rate": 0.00018954992240041387,
+      "loss": 2.0264,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10558155166619271,
+      "grad_norm": 1.2001843452453613,
+      "learning_rate": 0.00018953957578892914,
+      "loss": 1.5934,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10568496160905871,
+      "grad_norm": 1.453425407409668,
+      "learning_rate": 0.0001895292291774444,
+      "loss": 1.9904,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10578837155192472,
+      "grad_norm": 1.6289454698562622,
+      "learning_rate": 0.00018951888256595967,
+      "loss": 1.8033,
+      "step": 1023
+    },
+    {
+      "epoch": 0.10589178149479073,
+      "grad_norm": 1.375352144241333,
+      "learning_rate": 0.0001895085359544749,
+      "loss": 1.0417,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10599519143765673,
+      "grad_norm": 1.941213846206665,
+      "learning_rate": 0.00018949818934299017,
+      "loss": 2.1872,
+      "step": 1025
+    },
+    {
+      "epoch": 0.10609860138052274,
+      "grad_norm": 1.7311123609542847,
+      "learning_rate": 0.00018948784273150544,
+      "loss": 1.9615,
+      "step": 1026
+    },
+    {
+      "epoch": 0.10620201132338875,
+      "grad_norm": 2.0823657512664795,
+      "learning_rate": 0.0001894774961200207,
+      "loss": 1.4303,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10630542126625475,
+      "grad_norm": 2.074422597885132,
+      "learning_rate": 0.00018946714950853597,
+      "loss": 2.1897,
+      "step": 1028
+    },
+    {
+      "epoch": 0.10640883120912076,
+      "grad_norm": 2.0139143466949463,
+      "learning_rate": 0.00018945680289705124,
+      "loss": 1.3624,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10651224115198676,
+      "grad_norm": 1.2934083938598633,
+      "learning_rate": 0.00018944645628556647,
+      "loss": 1.5068,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10661565109485277,
+      "grad_norm": 1.888208270072937,
+      "learning_rate": 0.00018943610967408174,
+      "loss": 2.3472,
+      "step": 1031
+    },
+    {
+      "epoch": 0.10671906103771878,
+      "grad_norm": 1.3515228033065796,
+      "learning_rate": 0.000189425763062597,
+      "loss": 1.7646,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10682247098058478,
+      "grad_norm": 1.068301796913147,
+      "learning_rate": 0.00018941541645111227,
+      "loss": 1.8304,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10692588092345079,
+      "grad_norm": 0.8828536868095398,
+      "learning_rate": 0.00018940506983962754,
+      "loss": 1.6585,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1070292908663168,
+      "grad_norm": 1.5016223192214966,
+      "learning_rate": 0.0001893947232281428,
+      "loss": 1.8102,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1071327008091828,
+      "grad_norm": 1.7466976642608643,
+      "learning_rate": 0.00018938437661665804,
+      "loss": 1.6282,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10723611075204881,
+      "grad_norm": 1.166925072669983,
+      "learning_rate": 0.0001893740300051733,
+      "loss": 1.5768,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10733952069491481,
+      "grad_norm": 1.8839147090911865,
+      "learning_rate": 0.00018936368339368857,
+      "loss": 2.4981,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10744293063778082,
+      "grad_norm": 1.049673080444336,
+      "learning_rate": 0.00018935333678220384,
+      "loss": 1.8435,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10754634058064683,
+      "grad_norm": 2.3705356121063232,
+      "learning_rate": 0.0001893429901707191,
+      "loss": 1.9116,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10764975052351283,
+      "grad_norm": 1.645258903503418,
+      "learning_rate": 0.00018933264355923437,
+      "loss": 1.6528,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10775316046637884,
+      "grad_norm": 1.6526823043823242,
+      "learning_rate": 0.0001893222969477496,
+      "loss": 2.0118,
+      "step": 1042
+    },
+    {
+      "epoch": 0.10785657040924485,
+      "grad_norm": 1.2598350048065186,
+      "learning_rate": 0.00018931195033626487,
+      "loss": 2.2473,
+      "step": 1043
+    },
+    {
+      "epoch": 0.10795998035211085,
+      "grad_norm": 1.955881953239441,
+      "learning_rate": 0.00018930160372478014,
+      "loss": 1.8537,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10806339029497686,
+      "grad_norm": 2.739224433898926,
+      "learning_rate": 0.0001892912571132954,
+      "loss": 2.3352,
+      "step": 1045
+    },
+    {
+      "epoch": 0.10816680023784286,
+      "grad_norm": 2.8407015800476074,
+      "learning_rate": 0.00018928091050181067,
+      "loss": 2.006,
+      "step": 1046
+    },
+    {
+      "epoch": 0.10827021018070887,
+      "grad_norm": 1.3017852306365967,
+      "learning_rate": 0.00018927056389032594,
+      "loss": 1.8562,
+      "step": 1047
+    },
+    {
+      "epoch": 0.10837362012357488,
+      "grad_norm": 1.8522440195083618,
+      "learning_rate": 0.00018926021727884117,
+      "loss": 1.3757,
+      "step": 1048
+    },
+    {
+      "epoch": 0.10847703006644088,
+      "grad_norm": 1.512387990951538,
+      "learning_rate": 0.00018924987066735644,
+      "loss": 1.7554,
+      "step": 1049
+    },
+    {
+      "epoch": 0.10858044000930689,
+      "grad_norm": 1.8739479780197144,
+      "learning_rate": 0.0001892395240558717,
+      "loss": 2.3464,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1086838499521729,
+      "grad_norm": 1.4831976890563965,
+      "learning_rate": 0.00018922917744438697,
+      "loss": 1.9555,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1087872598950389,
+      "grad_norm": 1.5891178846359253,
+      "learning_rate": 0.00018921883083290224,
+      "loss": 1.3746,
+      "step": 1052
+    },
+    {
+      "epoch": 0.10889066983790492,
+      "grad_norm": 2.351069688796997,
+      "learning_rate": 0.0001892084842214175,
+      "loss": 1.3839,
+      "step": 1053
+    },
+    {
+      "epoch": 0.10899407978077093,
+      "grad_norm": 2.2983150482177734,
+      "learning_rate": 0.00018919813760993274,
+      "loss": 2.013,
+      "step": 1054
+    },
+    {
+      "epoch": 0.10909748972363693,
+      "grad_norm": 0.9400889277458191,
+      "learning_rate": 0.000189187790998448,
+      "loss": 1.8208,
+      "step": 1055
+    },
+    {
+      "epoch": 0.10920089966650294,
+      "grad_norm": 1.7346934080123901,
+      "learning_rate": 0.00018917744438696327,
+      "loss": 1.6008,
+      "step": 1056
+    },
+    {
+      "epoch": 0.10930430960936895,
+      "grad_norm": 2.0144057273864746,
+      "learning_rate": 0.00018916709777547854,
+      "loss": 1.5488,
+      "step": 1057
+    },
+    {
+      "epoch": 0.10940771955223495,
+      "grad_norm": 1.8916040658950806,
+      "learning_rate": 0.0001891567511639938,
+      "loss": 1.9123,
+      "step": 1058
+    },
+    {
+      "epoch": 0.10951112949510096,
+      "grad_norm": 0.820121169090271,
+      "learning_rate": 0.00018914640455250907,
+      "loss": 1.5304,
+      "step": 1059
+    },
+    {
+      "epoch": 0.10961453943796697,
+      "grad_norm": 1.2862852811813354,
+      "learning_rate": 0.0001891360579410243,
+      "loss": 1.7054,
+      "step": 1060
+    },
+    {
+      "epoch": 0.10971794938083297,
+      "grad_norm": 1.6166162490844727,
+      "learning_rate": 0.00018912571132953957,
+      "loss": 1.4308,
+      "step": 1061
+    },
+    {
+      "epoch": 0.10982135932369898,
+      "grad_norm": 1.2072175741195679,
+      "learning_rate": 0.00018911536471805484,
+      "loss": 1.937,
+      "step": 1062
+    },
+    {
+      "epoch": 0.10992476926656498,
+      "grad_norm": 1.044589638710022,
+      "learning_rate": 0.0001891050181065701,
+      "loss": 1.6156,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11002817920943099,
+      "grad_norm": 1.2547587156295776,
+      "learning_rate": 0.00018909467149508537,
+      "loss": 1.9699,
+      "step": 1064
+    },
+    {
+      "epoch": 0.110131589152297,
+      "grad_norm": 0.9746593832969666,
+      "learning_rate": 0.00018908432488360063,
+      "loss": 1.4966,
+      "step": 1065
+    },
+    {
+      "epoch": 0.110234999095163,
+      "grad_norm": 1.631894826889038,
+      "learning_rate": 0.00018907397827211587,
+      "loss": 1.586,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11033840903802901,
+      "grad_norm": 1.65902578830719,
+      "learning_rate": 0.00018906363166063114,
+      "loss": 1.7788,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11044181898089502,
+      "grad_norm": 0.8059775829315186,
+      "learning_rate": 0.0001890532850491464,
+      "loss": 1.7275,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11054522892376102,
+      "grad_norm": 1.8470535278320312,
+      "learning_rate": 0.00018904293843766167,
+      "loss": 1.1985,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11064863886662703,
+      "grad_norm": 1.7774137258529663,
+      "learning_rate": 0.00018903259182617693,
+      "loss": 2.0377,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11075204880949303,
+      "grad_norm": 1.8821941614151,
+      "learning_rate": 0.0001890222452146922,
+      "loss": 1.8849,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11085545875235904,
+      "grad_norm": 1.5304746627807617,
+      "learning_rate": 0.00018901189860320744,
+      "loss": 2.0822,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11095886869522505,
+      "grad_norm": 1.7353805303573608,
+      "learning_rate": 0.0001890015519917227,
+      "loss": 1.7504,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11106227863809105,
+      "grad_norm": 1.6357505321502686,
+      "learning_rate": 0.00018899120538023797,
+      "loss": 1.99,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11116568858095706,
+      "grad_norm": 1.29119074344635,
+      "learning_rate": 0.00018898085876875324,
+      "loss": 1.5753,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11126909852382307,
+      "grad_norm": 0.8210546374320984,
+      "learning_rate": 0.0001889705121572685,
+      "loss": 1.4492,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11137250846668907,
+      "grad_norm": 1.6336994171142578,
+      "learning_rate": 0.00018896016554578377,
+      "loss": 2.0778,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11147591840955508,
+      "grad_norm": 1.7355693578720093,
+      "learning_rate": 0.000188949818934299,
+      "loss": 1.7414,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11157932835242108,
+      "grad_norm": 0.9598165154457092,
+      "learning_rate": 0.00018893947232281427,
+      "loss": 1.8316,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11168273829528709,
+      "grad_norm": 1.4467240571975708,
+      "learning_rate": 0.00018892912571132954,
+      "loss": 1.3253,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1117861482381531,
+      "grad_norm": 1.258633017539978,
+      "learning_rate": 0.0001889187790998448,
+      "loss": 2.1484,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1118895581810191,
+      "grad_norm": 3.974945545196533,
+      "learning_rate": 0.00018890843248836007,
+      "loss": 2.4456,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11199296812388511,
+      "grad_norm": 3.577840805053711,
+      "learning_rate": 0.00018889808587687533,
+      "loss": 2.3568,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11209637806675112,
+      "grad_norm": 2.9973061084747314,
+      "learning_rate": 0.00018888773926539057,
+      "loss": 1.9609,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11219978800961712,
+      "grad_norm": 2.482928514480591,
+      "learning_rate": 0.00018887739265390584,
+      "loss": 2.2426,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11230319795248313,
+      "grad_norm": 1.4588831663131714,
+      "learning_rate": 0.0001888670460424211,
+      "loss": 1.3213,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11240660789534913,
+      "grad_norm": 1.1756937503814697,
+      "learning_rate": 0.00018885669943093637,
+      "loss": 1.6379,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11251001783821514,
+      "grad_norm": 1.3494646549224854,
+      "learning_rate": 0.00018884635281945163,
+      "loss": 1.4225,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11261342778108115,
+      "grad_norm": 1.6202143430709839,
+      "learning_rate": 0.0001888360062079669,
+      "loss": 1.8345,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11271683772394715,
+      "grad_norm": 1.5296990871429443,
+      "learning_rate": 0.00018882565959648216,
+      "loss": 2.0233,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11282024766681316,
+      "grad_norm": 1.0927106142044067,
+      "learning_rate": 0.0001888153129849974,
+      "loss": 1.2746,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11292365760967916,
+      "grad_norm": 1.003934383392334,
+      "learning_rate": 0.00018880496637351267,
+      "loss": 1.5895,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11302706755254517,
+      "grad_norm": 1.3455843925476074,
+      "learning_rate": 0.00018879461976202793,
+      "loss": 2.2522,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11313047749541118,
+      "grad_norm": 1.3665963411331177,
+      "learning_rate": 0.0001887842731505432,
+      "loss": 1.6567,
+      "step": 1094
+    },
+    {
+      "epoch": 0.11323388743827718,
+      "grad_norm": 3.7127087116241455,
+      "learning_rate": 0.00018877392653905847,
+      "loss": 1.7663,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1133372973811432,
+      "grad_norm": 1.8103179931640625,
+      "learning_rate": 0.00018876357992757373,
+      "loss": 1.8739,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11344070732400921,
+      "grad_norm": 2.501490592956543,
+      "learning_rate": 0.000188753233316089,
+      "loss": 2.0688,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11354411726687522,
+      "grad_norm": 2.178652763366699,
+      "learning_rate": 0.00018874288670460426,
+      "loss": 1.7537,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11364752720974122,
+      "grad_norm": 1.7944008111953735,
+      "learning_rate": 0.0001887325400931195,
+      "loss": 1.1038,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11375093715260723,
+      "grad_norm": 1.4743938446044922,
+      "learning_rate": 0.00018872219348163477,
+      "loss": 1.8314,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11385434709547324,
+      "grad_norm": 2.165440082550049,
+      "learning_rate": 0.00018871184687015003,
+      "loss": 1.7616,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11395775703833924,
+      "grad_norm": 2.110471487045288,
+      "learning_rate": 0.0001887015002586653,
+      "loss": 2.0106,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11406116698120525,
+      "grad_norm": 1.5662785768508911,
+      "learning_rate": 0.00018869115364718056,
+      "loss": 1.5418,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11416457692407125,
+      "grad_norm": 2.1364023685455322,
+      "learning_rate": 0.00018868080703569583,
+      "loss": 2.4005,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11426798686693726,
+      "grad_norm": 1.4032834768295288,
+      "learning_rate": 0.0001886704604242111,
+      "loss": 1.6216,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11437139680980327,
+      "grad_norm": 1.2782148122787476,
+      "learning_rate": 0.00018866011381272636,
+      "loss": 1.3291,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11447480675266927,
+      "grad_norm": 0.9738740921020508,
+      "learning_rate": 0.0001886497672012416,
+      "loss": 1.3762,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11457821669553528,
+      "grad_norm": 2.0070621967315674,
+      "learning_rate": 0.00018863942058975686,
+      "loss": 1.9043,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11468162663840129,
+      "grad_norm": 2.14795184135437,
+      "learning_rate": 0.00018862907397827213,
+      "loss": 1.5564,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11478503658126729,
+      "grad_norm": 2.192723512649536,
+      "learning_rate": 0.0001886187273667874,
+      "loss": 2.1081,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1148884465241333,
+      "grad_norm": 1.234890103340149,
+      "learning_rate": 0.00018860838075530266,
+      "loss": 1.4527,
+      "step": 1111
+    },
+    {
+      "epoch": 0.1149918564669993,
+      "grad_norm": 1.0031839609146118,
+      "learning_rate": 0.00018859803414381793,
+      "loss": 1.5052,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11509526640986531,
+      "grad_norm": 1.5561741590499878,
+      "learning_rate": 0.0001885876875323332,
+      "loss": 1.9241,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11519867635273132,
+      "grad_norm": 2.3440773487091064,
+      "learning_rate": 0.00018857734092084843,
+      "loss": 1.9541,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11530208629559732,
+      "grad_norm": 1.510392427444458,
+      "learning_rate": 0.0001885669943093637,
+      "loss": 2.0317,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11540549623846333,
+      "grad_norm": 1.3775566816329956,
+      "learning_rate": 0.00018855664769787896,
+      "loss": 1.7841,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11550890618132933,
+      "grad_norm": 0.9024312496185303,
+      "learning_rate": 0.00018854630108639423,
+      "loss": 1.974,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11561231612419534,
+      "grad_norm": 1.6050846576690674,
+      "learning_rate": 0.0001885359544749095,
+      "loss": 1.9041,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11571572606706135,
+      "grad_norm": 1.6221837997436523,
+      "learning_rate": 0.00018852560786342476,
+      "loss": 1.8008,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11581913600992735,
+      "grad_norm": 1.4674919843673706,
+      "learning_rate": 0.00018851526125194,
+      "loss": 1.0687,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11592254595279336,
+      "grad_norm": 2.9665956497192383,
+      "learning_rate": 0.00018850491464045526,
+      "loss": 1.9465,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11602595589565937,
+      "grad_norm": 2.3202719688415527,
+      "learning_rate": 0.00018849456802897053,
+      "loss": 2.2519,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11612936583852537,
+      "grad_norm": 1.2720879316329956,
+      "learning_rate": 0.0001884842214174858,
+      "loss": 2.0585,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11623277578139138,
+      "grad_norm": 1.8549882173538208,
+      "learning_rate": 0.00018847387480600106,
+      "loss": 1.8807,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11633618572425738,
+      "grad_norm": 1.7395340204238892,
+      "learning_rate": 0.00018846352819451632,
+      "loss": 1.7122,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11643959566712339,
+      "grad_norm": 1.8677852153778076,
+      "learning_rate": 0.00018845318158303156,
+      "loss": 1.8792,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1165430056099894,
+      "grad_norm": 1.4431513547897339,
+      "learning_rate": 0.00018844283497154683,
+      "loss": 2.1237,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1166464155528554,
+      "grad_norm": 1.1699776649475098,
+      "learning_rate": 0.0001884324883600621,
+      "loss": 1.2709,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11674982549572141,
+      "grad_norm": 1.6420997381210327,
+      "learning_rate": 0.00018842214174857736,
+      "loss": 1.7907,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11685323543858742,
+      "grad_norm": 1.390235424041748,
+      "learning_rate": 0.00018841179513709263,
+      "loss": 1.7772,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11695664538145342,
+      "grad_norm": 1.4697620868682861,
+      "learning_rate": 0.0001884014485256079,
+      "loss": 1.5639,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11706005532431943,
+      "grad_norm": 1.740312933921814,
+      "learning_rate": 0.00018839110191412313,
+      "loss": 1.7756,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11716346526718543,
+      "grad_norm": 2.991058111190796,
+      "learning_rate": 0.0001883807553026384,
+      "loss": 1.7446,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11726687521005144,
+      "grad_norm": 2.671280860900879,
+      "learning_rate": 0.00018837040869115366,
+      "loss": 1.9445,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11737028515291745,
+      "grad_norm": 2.767537832260132,
+      "learning_rate": 0.00018836006207966893,
+      "loss": 2.0817,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11747369509578345,
+      "grad_norm": 2.4302000999450684,
+      "learning_rate": 0.0001883497154681842,
+      "loss": 2.3611,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11757710503864946,
+      "grad_norm": 5.114750862121582,
+      "learning_rate": 0.00018833936885669946,
+      "loss": 2.3867,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11768051498151547,
+      "grad_norm": 1.0005320310592651,
+      "learning_rate": 0.0001883290222452147,
+      "loss": 2.0295,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11778392492438149,
+      "grad_norm": 1.4860533475875854,
+      "learning_rate": 0.00018831867563372996,
+      "loss": 2.0879,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11788733486724749,
+      "grad_norm": 0.8643640279769897,
+      "learning_rate": 0.00018830832902224523,
+      "loss": 2.0423,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1179907448101135,
+      "grad_norm": 2.4749903678894043,
+      "learning_rate": 0.0001882979824107605,
+      "loss": 1.5508,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1180941547529795,
+      "grad_norm": 1.414445400238037,
+      "learning_rate": 0.00018828763579927576,
+      "loss": 2.1075,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11819756469584551,
+      "grad_norm": 2.0021872520446777,
+      "learning_rate": 0.00018827728918779102,
+      "loss": 1.0144,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11830097463871152,
+      "grad_norm": 1.3872566223144531,
+      "learning_rate": 0.00018826694257630626,
+      "loss": 1.3635,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11840438458157752,
+      "grad_norm": 1.5357730388641357,
+      "learning_rate": 0.00018825659596482153,
+      "loss": 1.7582,
+      "step": 1145
+    },
+    {
+      "epoch": 0.11850779452444353,
+      "grad_norm": 1.2004647254943848,
+      "learning_rate": 0.0001882462493533368,
+      "loss": 1.5143,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11861120446730954,
+      "grad_norm": 1.6172764301300049,
+      "learning_rate": 0.00018823590274185206,
+      "loss": 1.2178,
+      "step": 1147
+    },
+    {
+      "epoch": 0.11871461441017554,
+      "grad_norm": 1.283896803855896,
+      "learning_rate": 0.00018822555613036732,
+      "loss": 2.0604,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11881802435304155,
+      "grad_norm": 1.100882887840271,
+      "learning_rate": 0.0001882152095188826,
+      "loss": 1.5466,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11892143429590755,
+      "grad_norm": 3.3976354598999023,
+      "learning_rate": 0.00018820486290739783,
+      "loss": 1.6791,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11902484423877356,
+      "grad_norm": 1.0436819791793823,
+      "learning_rate": 0.0001881945162959131,
+      "loss": 1.5231,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11912825418163957,
+      "grad_norm": 1.857042670249939,
+      "learning_rate": 0.00018818416968442836,
+      "loss": 2.3335,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11923166412450557,
+      "grad_norm": 1.8413466215133667,
+      "learning_rate": 0.00018817382307294362,
+      "loss": 1.6003,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11933507406737158,
+      "grad_norm": 1.1375494003295898,
+      "learning_rate": 0.0001881634764614589,
+      "loss": 2.1043,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11943848401023759,
+      "grad_norm": 1.7668977975845337,
+      "learning_rate": 0.00018815312984997416,
+      "loss": 2.2028,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11954189395310359,
+      "grad_norm": 1.776978611946106,
+      "learning_rate": 0.0001881427832384894,
+      "loss": 1.7239,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1196453038959696,
+      "grad_norm": 2.0679118633270264,
+      "learning_rate": 0.00018813243662700466,
+      "loss": 1.4295,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1197487138388356,
+      "grad_norm": 2.2436676025390625,
+      "learning_rate": 0.00018812209001551993,
+      "loss": 2.1156,
+      "step": 1158
+    },
+    {
+      "epoch": 0.11985212378170161,
+      "grad_norm": 1.7714333534240723,
+      "learning_rate": 0.0001881117434040352,
+      "loss": 1.8211,
+      "step": 1159
+    },
+    {
+      "epoch": 0.11995553372456762,
+      "grad_norm": 1.1076422929763794,
+      "learning_rate": 0.00018810139679255046,
+      "loss": 1.384,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12005894366743362,
+      "grad_norm": 3.0645408630371094,
+      "learning_rate": 0.00018809105018106572,
+      "loss": 1.8257,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12016235361029963,
+      "grad_norm": 1.4861654043197632,
+      "learning_rate": 0.00018808070356958096,
+      "loss": 2.0568,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12026576355316564,
+      "grad_norm": 1.040371298789978,
+      "learning_rate": 0.00018807035695809623,
+      "loss": 1.6845,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12036917349603164,
+      "grad_norm": 1.0902824401855469,
+      "learning_rate": 0.0001880600103466115,
+      "loss": 1.8769,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12047258343889765,
+      "grad_norm": 1.154840111732483,
+      "learning_rate": 0.00018804966373512676,
+      "loss": 1.6885,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12057599338176365,
+      "grad_norm": 1.073218584060669,
+      "learning_rate": 0.00018803931712364202,
+      "loss": 1.7502,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12067940332462966,
+      "grad_norm": 0.9243725538253784,
+      "learning_rate": 0.0001880289705121573,
+      "loss": 1.4066,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12078281326749567,
+      "grad_norm": 1.23936927318573,
+      "learning_rate": 0.00018801862390067253,
+      "loss": 2.0428,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12088622321036167,
+      "grad_norm": 1.1509627103805542,
+      "learning_rate": 0.0001880082772891878,
+      "loss": 1.6855,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12098963315322768,
+      "grad_norm": 1.6797055006027222,
+      "learning_rate": 0.00018799793067770306,
+      "loss": 2.1733,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12109304309609369,
+      "grad_norm": 1.2919670343399048,
+      "learning_rate": 0.00018798758406621832,
+      "loss": 1.8092,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12119645303895969,
+      "grad_norm": 0.9914805293083191,
+      "learning_rate": 0.0001879772374547336,
+      "loss": 1.621,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1212998629818257,
+      "grad_norm": 1.9691003561019897,
+      "learning_rate": 0.00018796689084324886,
+      "loss": 2.0663,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1214032729246917,
+      "grad_norm": 1.3074166774749756,
+      "learning_rate": 0.0001879565442317641,
+      "loss": 2.0947,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12150668286755771,
+      "grad_norm": 1.9575916528701782,
+      "learning_rate": 0.00018794619762027936,
+      "loss": 1.2152,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12161009281042372,
+      "grad_norm": 1.3250031471252441,
+      "learning_rate": 0.00018793585100879462,
+      "loss": 1.8787,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12171350275328972,
+      "grad_norm": 1.2709016799926758,
+      "learning_rate": 0.0001879255043973099,
+      "loss": 1.8267,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12181691269615573,
+      "grad_norm": 2.2685484886169434,
+      "learning_rate": 0.00018791515778582516,
+      "loss": 2.0463,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12192032263902174,
+      "grad_norm": 2.0677502155303955,
+      "learning_rate": 0.00018790481117434042,
+      "loss": 1.9328,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12202373258188774,
+      "grad_norm": 2.4347801208496094,
+      "learning_rate": 0.00018789446456285566,
+      "loss": 2.56,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12212714252475375,
+      "grad_norm": 3.401825428009033,
+      "learning_rate": 0.00018788411795137093,
+      "loss": 1.718,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12223055246761977,
+      "grad_norm": 2.031280040740967,
+      "learning_rate": 0.0001878737713398862,
+      "loss": 2.2049,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12233396241048577,
+      "grad_norm": 1.5570838451385498,
+      "learning_rate": 0.00018786342472840146,
+      "loss": 1.5248,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12243737235335178,
+      "grad_norm": 1.4443638324737549,
+      "learning_rate": 0.00018785307811691672,
+      "loss": 1.7989,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12254078229621779,
+      "grad_norm": 2.5985703468322754,
+      "learning_rate": 0.000187842731505432,
+      "loss": 1.7115,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1226441922390838,
+      "grad_norm": 1.325608730316162,
+      "learning_rate": 0.00018783238489394723,
+      "loss": 1.4347,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1227476021819498,
+      "grad_norm": 1.1276994943618774,
+      "learning_rate": 0.0001878220382824625,
+      "loss": 1.1806,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1228510121248158,
+      "grad_norm": 2.02768874168396,
+      "learning_rate": 0.00018781169167097776,
+      "loss": 1.9614,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12295442206768181,
+      "grad_norm": 1.9223341941833496,
+      "learning_rate": 0.00018780134505949302,
+      "loss": 2.0055,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12305783201054782,
+      "grad_norm": 1.4801243543624878,
+      "learning_rate": 0.0001877909984480083,
+      "loss": 1.5568,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12316124195341382,
+      "grad_norm": 1.6072427034378052,
+      "learning_rate": 0.00018778065183652355,
+      "loss": 1.6918,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12326465189627983,
+      "grad_norm": 1.8966267108917236,
+      "learning_rate": 0.0001877703052250388,
+      "loss": 1.7092,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12336806183914584,
+      "grad_norm": 1.2562758922576904,
+      "learning_rate": 0.00018775995861355406,
+      "loss": 1.3349,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12347147178201184,
+      "grad_norm": 1.6847071647644043,
+      "learning_rate": 0.00018774961200206932,
+      "loss": 1.2752,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12357488172487785,
+      "grad_norm": 2.11639666557312,
+      "learning_rate": 0.0001877392653905846,
+      "loss": 2.1486,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12367829166774386,
+      "grad_norm": 1.4768255949020386,
+      "learning_rate": 0.00018772891877909985,
+      "loss": 1.9611,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12378170161060986,
+      "grad_norm": 0.8098143935203552,
+      "learning_rate": 0.00018771857216761512,
+      "loss": 1.5275,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12388511155347587,
+      "grad_norm": 1.0442907810211182,
+      "learning_rate": 0.00018770822555613036,
+      "loss": 1.3282,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12398852149634187,
+      "grad_norm": 2.2555179595947266,
+      "learning_rate": 0.00018769787894464562,
+      "loss": 2.0941,
+      "step": 1199
+    },
+    {
+      "epoch": 0.12409193143920788,
+      "grad_norm": 1.3613994121551514,
+      "learning_rate": 0.0001876875323331609,
+      "loss": 1.7308,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12419534138207389,
+      "grad_norm": 2.940253496170044,
+      "learning_rate": 0.00018767718572167616,
+      "loss": 1.9223,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1242987513249399,
+      "grad_norm": 2.203312873840332,
+      "learning_rate": 0.00018766683911019142,
+      "loss": 2.1979,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1244021612678059,
+      "grad_norm": 1.6097736358642578,
+      "learning_rate": 0.0001876564924987067,
+      "loss": 1.169,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1245055712106719,
+      "grad_norm": 2.578017234802246,
+      "learning_rate": 0.00018764614588722193,
+      "loss": 1.5018,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12460898115353791,
+      "grad_norm": 1.4888848066329956,
+      "learning_rate": 0.0001876357992757372,
+      "loss": 1.8575,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12471239109640392,
+      "grad_norm": 1.513687252998352,
+      "learning_rate": 0.00018762545266425246,
+      "loss": 1.5326,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12481580103926992,
+      "grad_norm": 1.3281147480010986,
+      "learning_rate": 0.00018761510605276772,
+      "loss": 2.1136,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12491921098213593,
+      "grad_norm": 1.496387004852295,
+      "learning_rate": 0.000187604759441283,
+      "loss": 1.8321,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12502262092500194,
+      "grad_norm": 1.2002882957458496,
+      "learning_rate": 0.00018759441282979825,
+      "loss": 2.0025,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12512603086786794,
+      "grad_norm": 1.6012659072875977,
+      "learning_rate": 0.0001875840662183135,
+      "loss": 1.3442,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12522944081073395,
+      "grad_norm": 1.4702413082122803,
+      "learning_rate": 0.00018757371960682876,
+      "loss": 2.041,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12533285075359996,
+      "grad_norm": 1.0756279230117798,
+      "learning_rate": 0.00018756337299534402,
+      "loss": 1.7976,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12543626069646596,
+      "grad_norm": 4.028149604797363,
+      "learning_rate": 0.0001875530263838593,
+      "loss": 1.5223,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12553967063933197,
+      "grad_norm": 2.3542799949645996,
+      "learning_rate": 0.00018754267977237455,
+      "loss": 1.5814,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12564308058219797,
+      "grad_norm": 1.941663384437561,
+      "learning_rate": 0.00018753233316088982,
+      "loss": 2.1115,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12574649052506398,
+      "grad_norm": 2.03517484664917,
+      "learning_rate": 0.00018752198654940506,
+      "loss": 1.8151,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12584990046793,
+      "grad_norm": 1.52018141746521,
+      "learning_rate": 0.00018751163993792032,
+      "loss": 1.7026,
+      "step": 1217
+    },
+    {
+      "epoch": 0.125953310410796,
+      "grad_norm": 2.1755878925323486,
+      "learning_rate": 0.0001875012933264356,
+      "loss": 2.7846,
+      "step": 1218
+    },
+    {
+      "epoch": 0.126056720353662,
+      "grad_norm": 0.7671854496002197,
+      "learning_rate": 0.00018749094671495085,
+      "loss": 1.8867,
+      "step": 1219
+    },
+    {
+      "epoch": 0.126160130296528,
+      "grad_norm": 1.6932541131973267,
+      "learning_rate": 0.00018748060010346612,
+      "loss": 2.2631,
+      "step": 1220
+    },
+    {
+      "epoch": 0.126263540239394,
+      "grad_norm": 1.4602147340774536,
+      "learning_rate": 0.00018747025349198139,
+      "loss": 1.7261,
+      "step": 1221
+    },
+    {
+      "epoch": 0.12636695018226002,
+      "grad_norm": 1.7497947216033936,
+      "learning_rate": 0.00018745990688049665,
+      "loss": 2.4626,
+      "step": 1222
+    },
+    {
+      "epoch": 0.12647036012512602,
+      "grad_norm": 1.4249109029769897,
+      "learning_rate": 0.00018744956026901192,
+      "loss": 1.7258,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12657377006799203,
+      "grad_norm": 1.2902969121932983,
+      "learning_rate": 0.00018743921365752716,
+      "loss": 1.2497,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12667718001085804,
+      "grad_norm": 1.8637944459915161,
+      "learning_rate": 0.00018742886704604242,
+      "loss": 1.686,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12678058995372404,
+      "grad_norm": 1.095707893371582,
+      "learning_rate": 0.00018741852043455769,
+      "loss": 2.192,
+      "step": 1226
+    },
+    {
+      "epoch": 0.12688399989659005,
+      "grad_norm": 1.8274673223495483,
+      "learning_rate": 0.00018740817382307295,
+      "loss": 1.5955,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12698740983945606,
+      "grad_norm": 1.9945489168167114,
+      "learning_rate": 0.00018739782721158822,
+      "loss": 2.1984,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12709081978232206,
+      "grad_norm": 1.4088261127471924,
+      "learning_rate": 0.00018738748060010348,
+      "loss": 2.0431,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12719422972518807,
+      "grad_norm": 1.128542423248291,
+      "learning_rate": 0.00018737713398861875,
+      "loss": 2.0949,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12729763966805407,
+      "grad_norm": 1.2845412492752075,
+      "learning_rate": 0.00018736678737713401,
+      "loss": 1.2357,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12740104961092008,
+      "grad_norm": 1.4940705299377441,
+      "learning_rate": 0.00018735644076564925,
+      "loss": 1.8014,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1275044595537861,
+      "grad_norm": 1.725649118423462,
+      "learning_rate": 0.00018734609415416452,
+      "loss": 1.8969,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1276078694966521,
+      "grad_norm": 1.2723456621170044,
+      "learning_rate": 0.00018733574754267978,
+      "loss": 1.4834,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1277112794395181,
+      "grad_norm": 0.8685710430145264,
+      "learning_rate": 0.00018732540093119505,
+      "loss": 1.5972,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1278146893823841,
+      "grad_norm": 1.1668968200683594,
+      "learning_rate": 0.00018731505431971032,
+      "loss": 1.6169,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1279180993252501,
+      "grad_norm": 1.3985564708709717,
+      "learning_rate": 0.00018730470770822558,
+      "loss": 1.5132,
+      "step": 1237
+    },
+    {
+      "epoch": 0.12802150926811612,
+      "grad_norm": 1.8650096654891968,
+      "learning_rate": 0.00018729436109674085,
+      "loss": 1.5906,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12812491921098212,
+      "grad_norm": 1.2535521984100342,
+      "learning_rate": 0.0001872840144852561,
+      "loss": 1.6443,
+      "step": 1239
+    },
+    {
+      "epoch": 0.12822832915384813,
+      "grad_norm": 1.6224900484085083,
+      "learning_rate": 0.00018727366787377135,
+      "loss": 1.8258,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12833173909671414,
+      "grad_norm": 1.5536644458770752,
+      "learning_rate": 0.00018726332126228662,
+      "loss": 1.8278,
+      "step": 1241
+    },
+    {
+      "epoch": 0.12843514903958014,
+      "grad_norm": 1.2493494749069214,
+      "learning_rate": 0.00018725297465080188,
+      "loss": 1.9223,
+      "step": 1242
+    },
+    {
+      "epoch": 0.12853855898244615,
+      "grad_norm": 1.945225715637207,
+      "learning_rate": 0.00018724262803931715,
+      "loss": 2.0077,
+      "step": 1243
+    },
+    {
+      "epoch": 0.12864196892531216,
+      "grad_norm": 0.7477271556854248,
+      "learning_rate": 0.0001872322814278324,
+      "loss": 1.8014,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12874537886817816,
+      "grad_norm": 1.8180761337280273,
+      "learning_rate": 0.00018722193481634768,
+      "loss": 1.4772,
+      "step": 1245
+    },
+    {
+      "epoch": 0.12884878881104417,
+      "grad_norm": 1.7646000385284424,
+      "learning_rate": 0.00018721158820486292,
+      "loss": 1.598,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1289521987539102,
+      "grad_norm": 1.0384913682937622,
+      "learning_rate": 0.00018720124159337818,
+      "loss": 1.5885,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1290556086967762,
+      "grad_norm": 1.9679780006408691,
+      "learning_rate": 0.00018719089498189345,
+      "loss": 1.2981,
+      "step": 1248
+    },
+    {
+      "epoch": 0.12915901863964221,
+      "grad_norm": 1.365814447402954,
+      "learning_rate": 0.0001871805483704087,
+      "loss": 1.7245,
+      "step": 1249
+    },
+    {
+      "epoch": 0.12926242858250822,
+      "grad_norm": 1.7487189769744873,
+      "learning_rate": 0.00018717020175892398,
+      "loss": 1.6709,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12936583852537423,
+      "grad_norm": 0.9565590620040894,
+      "learning_rate": 0.00018715985514743924,
+      "loss": 1.6439,
+      "step": 1251
+    },
+    {
+      "epoch": 0.12946924846824023,
+      "grad_norm": 2.015857696533203,
+      "learning_rate": 0.00018714950853595448,
+      "loss": 1.8146,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12957265841110624,
+      "grad_norm": 0.9637671113014221,
+      "learning_rate": 0.00018713916192446975,
+      "loss": 1.7245,
+      "step": 1253
+    },
+    {
+      "epoch": 0.12967606835397225,
+      "grad_norm": 1.3832422494888306,
+      "learning_rate": 0.00018712881531298501,
+      "loss": 1.3993,
+      "step": 1254
+    },
+    {
+      "epoch": 0.12977947829683825,
+      "grad_norm": 2.1820902824401855,
+      "learning_rate": 0.00018711846870150028,
+      "loss": 1.7568,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12988288823970426,
+      "grad_norm": 1.573229432106018,
+      "learning_rate": 0.00018710812209001555,
+      "loss": 1.415,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12998629818257026,
+      "grad_norm": 1.5266555547714233,
+      "learning_rate": 0.0001870977754785308,
+      "loss": 1.5005,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13008970812543627,
+      "grad_norm": 0.9026231169700623,
+      "learning_rate": 0.00018708742886704605,
+      "loss": 2.0375,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13019311806830228,
+      "grad_norm": 1.4228705167770386,
+      "learning_rate": 0.00018707708225556131,
+      "loss": 1.8817,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13029652801116828,
+      "grad_norm": 1.961610198020935,
+      "learning_rate": 0.00018706673564407658,
+      "loss": 1.7003,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1303999379540343,
+      "grad_norm": 1.7642508745193481,
+      "learning_rate": 0.00018705638903259185,
+      "loss": 2.0717,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1305033478969003,
+      "grad_norm": 2.054243564605713,
+      "learning_rate": 0.0001870460424211071,
+      "loss": 2.1404,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1306067578397663,
+      "grad_norm": 2.2590620517730713,
+      "learning_rate": 0.00018703569580962235,
+      "loss": 2.1784,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1307101677826323,
+      "grad_norm": 1.1175493001937866,
+      "learning_rate": 0.00018702534919813762,
+      "loss": 1.8119,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13081357772549831,
+      "grad_norm": 2.2815165519714355,
+      "learning_rate": 0.00018701500258665288,
+      "loss": 2.1407,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13091698766836432,
+      "grad_norm": 1.36077082157135,
+      "learning_rate": 0.00018700465597516815,
+      "loss": 2.338,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13102039761123033,
+      "grad_norm": 1.8549292087554932,
+      "learning_rate": 0.0001869943093636834,
+      "loss": 2.447,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13112380755409633,
+      "grad_norm": 2.007368564605713,
+      "learning_rate": 0.00018698396275219868,
+      "loss": 1.9328,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13122721749696234,
+      "grad_norm": 1.925146222114563,
+      "learning_rate": 0.00018697361614071392,
+      "loss": 1.6689,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13133062743982835,
+      "grad_norm": 1.3448445796966553,
+      "learning_rate": 0.00018696326952922918,
+      "loss": 1.6307,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13143403738269435,
+      "grad_norm": 1.1631834506988525,
+      "learning_rate": 0.00018695292291774445,
+      "loss": 1.3101,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13153744732556036,
+      "grad_norm": 1.2478338479995728,
+      "learning_rate": 0.0001869425763062597,
+      "loss": 2.0214,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13164085726842636,
+      "grad_norm": 1.4405758380889893,
+      "learning_rate": 0.00018693222969477498,
+      "loss": 1.9482,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13174426721129237,
+      "grad_norm": 1.3760428428649902,
+      "learning_rate": 0.00018692188308329024,
+      "loss": 1.3526,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13184767715415838,
+      "grad_norm": 2.5238633155822754,
+      "learning_rate": 0.00018691153647180548,
+      "loss": 1.8157,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13195108709702438,
+      "grad_norm": 0.7784034013748169,
+      "learning_rate": 0.00018690118986032075,
+      "loss": 1.7506,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1320544970398904,
+      "grad_norm": 2.70515513420105,
+      "learning_rate": 0.00018689084324883601,
+      "loss": 1.3656,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1321579069827564,
+      "grad_norm": 1.7337223291397095,
+      "learning_rate": 0.00018688049663735128,
+      "loss": 1.8144,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1322613169256224,
+      "grad_norm": 1.2779606580734253,
+      "learning_rate": 0.00018687015002586654,
+      "loss": 1.5246,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1323647268684884,
+      "grad_norm": 1.8382599353790283,
+      "learning_rate": 0.0001868598034143818,
+      "loss": 1.9191,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13246813681135441,
+      "grad_norm": 2.009275197982788,
+      "learning_rate": 0.00018684945680289705,
+      "loss": 1.9812,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13257154675422042,
+      "grad_norm": 1.8259612321853638,
+      "learning_rate": 0.00018683911019141231,
+      "loss": 1.6244,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13267495669708643,
+      "grad_norm": 1.3301429748535156,
+      "learning_rate": 0.00018682876357992758,
+      "loss": 1.5117,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13277836663995243,
+      "grad_norm": 1.6860365867614746,
+      "learning_rate": 0.00018681841696844285,
+      "loss": 1.6663,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13288177658281844,
+      "grad_norm": 2.0984020233154297,
+      "learning_rate": 0.0001868080703569581,
+      "loss": 1.8559,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13298518652568445,
+      "grad_norm": 1.4987295866012573,
+      "learning_rate": 0.00018679772374547338,
+      "loss": 1.8326,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13308859646855045,
+      "grad_norm": 1.5269889831542969,
+      "learning_rate": 0.00018678737713398862,
+      "loss": 1.4678,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13319200641141646,
+      "grad_norm": 1.3404724597930908,
+      "learning_rate": 0.00018677703052250388,
+      "loss": 1.4516,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13329541635428246,
+      "grad_norm": 2.1874942779541016,
+      "learning_rate": 0.00018676668391101915,
+      "loss": 1.8405,
+      "step": 1289
+    },
+    {
+      "epoch": 0.13339882629714847,
+      "grad_norm": 1.327823519706726,
+      "learning_rate": 0.0001867563372995344,
+      "loss": 1.7545,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13350223624001448,
+      "grad_norm": 1.8989559412002563,
+      "learning_rate": 0.00018674599068804968,
+      "loss": 2.4508,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13360564618288048,
+      "grad_norm": 1.538395643234253,
+      "learning_rate": 0.00018673564407656494,
+      "loss": 1.9593,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1337090561257465,
+      "grad_norm": 2.1572530269622803,
+      "learning_rate": 0.00018672529746508018,
+      "loss": 1.7497,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1338124660686125,
+      "grad_norm": 1.3159223794937134,
+      "learning_rate": 0.00018671495085359545,
+      "loss": 2.1502,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1339158760114785,
+      "grad_norm": 2.51725697517395,
+      "learning_rate": 0.0001867046042421107,
+      "loss": 1.6545,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1340192859543445,
+      "grad_norm": 1.1315113306045532,
+      "learning_rate": 0.00018669425763062598,
+      "loss": 1.2736,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13412269589721051,
+      "grad_norm": 1.8445188999176025,
+      "learning_rate": 0.00018668391101914124,
+      "loss": 2.1056,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13422610584007652,
+      "grad_norm": 1.004447340965271,
+      "learning_rate": 0.0001866735644076565,
+      "loss": 1.9898,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13432951578294253,
+      "grad_norm": 1.5629618167877197,
+      "learning_rate": 0.00018666321779617175,
+      "loss": 1.9748,
+      "step": 1299
+    },
+    {
+      "epoch": 0.13443292572580853,
+      "grad_norm": 1.6147910356521606,
+      "learning_rate": 0.000186652871184687,
+      "loss": 2.041,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13453633566867454,
+      "grad_norm": 1.707336187362671,
+      "learning_rate": 0.00018664252457320228,
+      "loss": 1.4538,
+      "step": 1301
+    },
+    {
+      "epoch": 0.13463974561154055,
+      "grad_norm": 2.5680580139160156,
+      "learning_rate": 0.00018663217796171754,
+      "loss": 2.3289,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13474315555440655,
+      "grad_norm": 2.430309772491455,
+      "learning_rate": 0.0001866218313502328,
+      "loss": 1.6057,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13484656549727256,
+      "grad_norm": 1.7052457332611084,
+      "learning_rate": 0.00018661148473874808,
+      "loss": 2.0138,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13494997544013856,
+      "grad_norm": 1.389034390449524,
+      "learning_rate": 0.00018660113812726331,
+      "loss": 1.3975,
+      "step": 1305
+    },
+    {
+      "epoch": 0.13505338538300457,
+      "grad_norm": 1.011025309562683,
+      "learning_rate": 0.00018659079151577858,
+      "loss": 1.7537,
+      "step": 1306
+    },
+    {
+      "epoch": 0.13515679532587058,
+      "grad_norm": 1.0741398334503174,
+      "learning_rate": 0.00018658044490429385,
+      "loss": 1.3626,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13526020526873658,
+      "grad_norm": 1.2746872901916504,
+      "learning_rate": 0.0001865700982928091,
+      "loss": 1.0874,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1353636152116026,
+      "grad_norm": 3.1592459678649902,
+      "learning_rate": 0.00018655975168132438,
+      "loss": 1.8081,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1354670251544686,
+      "grad_norm": 1.210808515548706,
+      "learning_rate": 0.00018654940506983964,
+      "loss": 2.2929,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1355704350973346,
+      "grad_norm": 1.5508512258529663,
+      "learning_rate": 0.00018653905845835488,
+      "loss": 1.7335,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1356738450402006,
+      "grad_norm": 1.1455914974212646,
+      "learning_rate": 0.00018652871184687015,
+      "loss": 1.6225,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13577725498306661,
+      "grad_norm": 0.8236886262893677,
+      "learning_rate": 0.0001865183652353854,
+      "loss": 1.6923,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13588066492593262,
+      "grad_norm": 0.9368153214454651,
+      "learning_rate": 0.00018650801862390068,
+      "loss": 1.3802,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13598407486879863,
+      "grad_norm": 2.053213357925415,
+      "learning_rate": 0.00018649767201241594,
+      "loss": 1.6705,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13608748481166463,
+      "grad_norm": 2.63618803024292,
+      "learning_rate": 0.0001864873254009312,
+      "loss": 1.4974,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13619089475453064,
+      "grad_norm": 0.9730865359306335,
+      "learning_rate": 0.00018647697878944645,
+      "loss": 1.6048,
+      "step": 1317
+    },
+    {
+      "epoch": 0.13629430469739665,
+      "grad_norm": 2.2292282581329346,
+      "learning_rate": 0.0001864666321779617,
+      "loss": 1.6644,
+      "step": 1318
+    },
+    {
+      "epoch": 0.13639771464026265,
+      "grad_norm": 1.701492428779602,
+      "learning_rate": 0.00018645628556647698,
+      "loss": 1.4162,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13650112458312866,
+      "grad_norm": 2.6111085414886475,
+      "learning_rate": 0.00018644593895499224,
+      "loss": 1.8539,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13660453452599466,
+      "grad_norm": 0.8146743178367615,
+      "learning_rate": 0.0001864355923435075,
+      "loss": 1.8789,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13670794446886067,
+      "grad_norm": 1.8751240968704224,
+      "learning_rate": 0.00018642524573202277,
+      "loss": 2.0231,
+      "step": 1322
+    },
+    {
+      "epoch": 0.13681135441172668,
+      "grad_norm": 2.0674078464508057,
+      "learning_rate": 0.000186414899120538,
+      "loss": 1.5609,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13691476435459268,
+      "grad_norm": 1.0525996685028076,
+      "learning_rate": 0.00018640455250905328,
+      "loss": 1.4097,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1370181742974587,
+      "grad_norm": 4.645424842834473,
+      "learning_rate": 0.00018639420589756854,
+      "loss": 1.2942,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1371215842403247,
+      "grad_norm": 2.035285472869873,
+      "learning_rate": 0.0001863838592860838,
+      "loss": 1.3918,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1372249941831907,
+      "grad_norm": 1.863338828086853,
+      "learning_rate": 0.00018637351267459908,
+      "loss": 1.7489,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1373284041260567,
+      "grad_norm": 1.5353502035140991,
+      "learning_rate": 0.00018636316606311434,
+      "loss": 1.4828,
+      "step": 1328
+    },
+    {
+      "epoch": 0.13743181406892271,
+      "grad_norm": 1.3041024208068848,
+      "learning_rate": 0.00018635281945162958,
+      "loss": 1.3859,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13753522401178872,
+      "grad_norm": 1.15968918800354,
+      "learning_rate": 0.00018634247284014485,
+      "loss": 1.9243,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13763863395465473,
+      "grad_norm": 1.0997211933135986,
+      "learning_rate": 0.0001863321262286601,
+      "loss": 1.5236,
+      "step": 1331
+    },
+    {
+      "epoch": 0.13774204389752073,
+      "grad_norm": 1.8653374910354614,
+      "learning_rate": 0.00018632177961717538,
+      "loss": 1.4485,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13784545384038677,
+      "grad_norm": 2.3222272396087646,
+      "learning_rate": 0.00018631143300569064,
+      "loss": 1.9068,
+      "step": 1333
+    },
+    {
+      "epoch": 0.13794886378325277,
+      "grad_norm": 2.070629596710205,
+      "learning_rate": 0.0001863010863942059,
+      "loss": 1.1053,
+      "step": 1334
+    },
+    {
+      "epoch": 0.13805227372611878,
+      "grad_norm": 0.7526354789733887,
+      "learning_rate": 0.00018629073978272115,
+      "loss": 2.0752,
+      "step": 1335
+    },
+    {
+      "epoch": 0.13815568366898479,
+      "grad_norm": 1.8768320083618164,
+      "learning_rate": 0.0001862803931712364,
+      "loss": 1.8098,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1382590936118508,
+      "grad_norm": 2.4717142581939697,
+      "learning_rate": 0.00018627004655975168,
+      "loss": 1.967,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1383625035547168,
+      "grad_norm": 0.8968499302864075,
+      "learning_rate": 0.00018625969994826694,
+      "loss": 1.4093,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1384659134975828,
+      "grad_norm": 1.5168501138687134,
+      "learning_rate": 0.0001862493533367822,
+      "loss": 0.7085,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1385693234404488,
+      "grad_norm": 1.7330024242401123,
+      "learning_rate": 0.00018623900672529747,
+      "loss": 2.2376,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13867273338331482,
+      "grad_norm": 2.234142780303955,
+      "learning_rate": 0.0001862286601138127,
+      "loss": 1.507,
+      "step": 1341
+    },
+    {
+      "epoch": 0.13877614332618082,
+      "grad_norm": 1.6087912321090698,
+      "learning_rate": 0.00018621831350232798,
+      "loss": 1.6281,
+      "step": 1342
+    },
+    {
+      "epoch": 0.13887955326904683,
+      "grad_norm": 1.7449696063995361,
+      "learning_rate": 0.00018620796689084324,
+      "loss": 1.8542,
+      "step": 1343
+    },
+    {
+      "epoch": 0.13898296321191284,
+      "grad_norm": 1.2030277252197266,
+      "learning_rate": 0.0001861976202793585,
+      "loss": 1.485,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13908637315477884,
+      "grad_norm": 1.4781368970870972,
+      "learning_rate": 0.00018618727366787377,
+      "loss": 1.742,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13918978309764485,
+      "grad_norm": 1.9535484313964844,
+      "learning_rate": 0.00018617692705638904,
+      "loss": 1.876,
+      "step": 1346
+    },
+    {
+      "epoch": 0.13929319304051085,
+      "grad_norm": 1.9578993320465088,
+      "learning_rate": 0.0001861665804449043,
+      "loss": 1.7654,
+      "step": 1347
+    },
+    {
+      "epoch": 0.13939660298337686,
+      "grad_norm": 1.1052072048187256,
+      "learning_rate": 0.00018615623383341957,
+      "loss": 1.5072,
+      "step": 1348
+    },
+    {
+      "epoch": 0.13950001292624287,
+      "grad_norm": 1.2735376358032227,
+      "learning_rate": 0.0001861458872219348,
+      "loss": 1.9359,
+      "step": 1349
+    },
+    {
+      "epoch": 0.13960342286910887,
+      "grad_norm": 1.1834639310836792,
+      "learning_rate": 0.00018613554061045008,
+      "loss": 1.4385,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13970683281197488,
+      "grad_norm": 0.7228855490684509,
+      "learning_rate": 0.00018612519399896534,
+      "loss": 1.8304,
+      "step": 1351
+    },
+    {
+      "epoch": 0.13981024275484089,
+      "grad_norm": 1.2559309005737305,
+      "learning_rate": 0.0001861148473874806,
+      "loss": 1.6177,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1399136526977069,
+      "grad_norm": 2.237501621246338,
+      "learning_rate": 0.00018610450077599587,
+      "loss": 1.9076,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1400170626405729,
+      "grad_norm": 3.862377643585205,
+      "learning_rate": 0.00018609415416451114,
+      "loss": 1.7243,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1401204725834389,
+      "grad_norm": 1.3240326642990112,
+      "learning_rate": 0.0001860838075530264,
+      "loss": 1.5081,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1402238825263049,
+      "grad_norm": 1.5742530822753906,
+      "learning_rate": 0.00018607346094154167,
+      "loss": 1.422,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14032729246917092,
+      "grad_norm": 1.4533076286315918,
+      "learning_rate": 0.0001860631143300569,
+      "loss": 1.9615,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14043070241203692,
+      "grad_norm": 3.0016930103302,
+      "learning_rate": 0.00018605276771857217,
+      "loss": 1.9092,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14053411235490293,
+      "grad_norm": 1.1997783184051514,
+      "learning_rate": 0.00018604242110708744,
+      "loss": 1.2373,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14063752229776894,
+      "grad_norm": 0.9452020525932312,
+      "learning_rate": 0.0001860320744956027,
+      "loss": 1.9416,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14074093224063494,
+      "grad_norm": 1.4949263334274292,
+      "learning_rate": 0.00018602172788411797,
+      "loss": 1.9632,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14084434218350095,
+      "grad_norm": 1.2022387981414795,
+      "learning_rate": 0.00018601138127263323,
+      "loss": 1.3986,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14094775212636695,
+      "grad_norm": 1.3742767572402954,
+      "learning_rate": 0.0001860010346611485,
+      "loss": 1.4181,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14105116206923296,
+      "grad_norm": 1.5255154371261597,
+      "learning_rate": 0.00018599068804966377,
+      "loss": 2.0134,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14115457201209897,
+      "grad_norm": 1.4387857913970947,
+      "learning_rate": 0.000185980341438179,
+      "loss": 1.4114,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14125798195496497,
+      "grad_norm": 2.9835379123687744,
+      "learning_rate": 0.00018596999482669427,
+      "loss": 2.0213,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14136139189783098,
+      "grad_norm": 1.301589846611023,
+      "learning_rate": 0.00018595964821520954,
+      "loss": 1.7207,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14146480184069699,
+      "grad_norm": 1.2321486473083496,
+      "learning_rate": 0.0001859493016037248,
+      "loss": 1.6375,
+      "step": 1368
+    },
+    {
+      "epoch": 0.141568211783563,
+      "grad_norm": 2.0843982696533203,
+      "learning_rate": 0.00018593895499224007,
+      "loss": 2.2411,
+      "step": 1369
+    },
+    {
+      "epoch": 0.141671621726429,
+      "grad_norm": 1.382814645767212,
+      "learning_rate": 0.00018592860838075533,
+      "loss": 1.7965,
+      "step": 1370
+    },
+    {
+      "epoch": 0.141775031669295,
+      "grad_norm": 1.7248510122299194,
+      "learning_rate": 0.00018591826176927057,
+      "loss": 1.5689,
+      "step": 1371
+    },
+    {
+      "epoch": 0.141878441612161,
+      "grad_norm": 1.425204873085022,
+      "learning_rate": 0.00018590791515778584,
+      "loss": 1.7564,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14198185155502702,
+      "grad_norm": 1.2911546230316162,
+      "learning_rate": 0.0001858975685463011,
+      "loss": 1.2647,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14208526149789302,
+      "grad_norm": 1.1772571802139282,
+      "learning_rate": 0.00018588722193481637,
+      "loss": 1.8914,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14218867144075903,
+      "grad_norm": 1.4799343347549438,
+      "learning_rate": 0.00018587687532333163,
+      "loss": 1.0355,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14229208138362504,
+      "grad_norm": 2.084376811981201,
+      "learning_rate": 0.0001858665287118469,
+      "loss": 1.063,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14239549132649104,
+      "grad_norm": 1.7417645454406738,
+      "learning_rate": 0.00018585618210036214,
+      "loss": 1.7731,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14249890126935705,
+      "grad_norm": 1.8002780675888062,
+      "learning_rate": 0.0001858458354888774,
+      "loss": 1.9528,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14260231121222305,
+      "grad_norm": 3.58180570602417,
+      "learning_rate": 0.00018583548887739267,
+      "loss": 1.6168,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14270572115508906,
+      "grad_norm": 1.7094149589538574,
+      "learning_rate": 0.00018582514226590793,
+      "loss": 1.0873,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14280913109795507,
+      "grad_norm": 2.5685203075408936,
+      "learning_rate": 0.0001858147956544232,
+      "loss": 1.7104,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14291254104082107,
+      "grad_norm": 3.272225856781006,
+      "learning_rate": 0.00018580444904293847,
+      "loss": 1.8178,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14301595098368708,
+      "grad_norm": 3.2546474933624268,
+      "learning_rate": 0.0001857941024314537,
+      "loss": 2.2033,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14311936092655309,
+      "grad_norm": 1.2203083038330078,
+      "learning_rate": 0.00018578375581996897,
+      "loss": 1.7435,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1432227708694191,
+      "grad_norm": 1.4671968221664429,
+      "learning_rate": 0.00018577340920848423,
+      "loss": 1.6157,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1433261808122851,
+      "grad_norm": 0.9329012036323547,
+      "learning_rate": 0.0001857630625969995,
+      "loss": 1.9851,
+      "step": 1386
+    },
+    {
+      "epoch": 0.1434295907551511,
+      "grad_norm": 1.7417888641357422,
+      "learning_rate": 0.00018575271598551477,
+      "loss": 2.3805,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1435330006980171,
+      "grad_norm": 1.4650931358337402,
+      "learning_rate": 0.00018574236937403003,
+      "loss": 1.7815,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14363641064088312,
+      "grad_norm": 2.144329071044922,
+      "learning_rate": 0.00018573202276254527,
+      "loss": 1.751,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14373982058374912,
+      "grad_norm": 1.9252647161483765,
+      "learning_rate": 0.00018572167615106054,
+      "loss": 2.0645,
+      "step": 1390
+    },
+    {
+      "epoch": 0.14384323052661513,
+      "grad_norm": 0.8894973397254944,
+      "learning_rate": 0.0001857113295395758,
+      "loss": 1.8464,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14394664046948114,
+      "grad_norm": 3.1202118396759033,
+      "learning_rate": 0.00018570098292809107,
+      "loss": 2.0249,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14405005041234714,
+      "grad_norm": 1.7900038957595825,
+      "learning_rate": 0.00018569063631660633,
+      "loss": 1.4751,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14415346035521315,
+      "grad_norm": 1.6321499347686768,
+      "learning_rate": 0.0001856802897051216,
+      "loss": 1.7489,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14425687029807915,
+      "grad_norm": 2.1138758659362793,
+      "learning_rate": 0.00018566994309363684,
+      "loss": 1.6695,
+      "step": 1395
+    },
+    {
+      "epoch": 0.14436028024094516,
+      "grad_norm": 1.5029677152633667,
+      "learning_rate": 0.0001856595964821521,
+      "loss": 1.4675,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14446369018381117,
+      "grad_norm": 4.420119762420654,
+      "learning_rate": 0.00018564924987066737,
+      "loss": 2.0966,
+      "step": 1397
+    },
+    {
+      "epoch": 0.14456710012667717,
+      "grad_norm": 0.8471847772598267,
+      "learning_rate": 0.00018563890325918263,
+      "loss": 1.5981,
+      "step": 1398
+    },
+    {
+      "epoch": 0.14467051006954318,
+      "grad_norm": 1.1944760084152222,
+      "learning_rate": 0.0001856285566476979,
+      "loss": 1.6597,
+      "step": 1399
+    },
+    {
+      "epoch": 0.14477392001240919,
+      "grad_norm": 2.2949111461639404,
+      "learning_rate": 0.00018561821003621316,
+      "loss": 1.5731,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1448773299552752,
+      "grad_norm": 3.328265428543091,
+      "learning_rate": 0.0001856078634247284,
+      "loss": 1.9217,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1449807398981412,
+      "grad_norm": 0.9115211963653564,
+      "learning_rate": 0.00018559751681324367,
+      "loss": 1.5536,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1450841498410072,
+      "grad_norm": 1.1472818851470947,
+      "learning_rate": 0.00018558717020175893,
+      "loss": 1.9001,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1451875597838732,
+      "grad_norm": 1.5385702848434448,
+      "learning_rate": 0.0001855768235902742,
+      "loss": 1.187,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14529096972673922,
+      "grad_norm": 1.0141092538833618,
+      "learning_rate": 0.00018556647697878946,
+      "loss": 1.1727,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14539437966960522,
+      "grad_norm": 1.044488549232483,
+      "learning_rate": 0.00018555613036730473,
+      "loss": 1.6773,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14549778961247123,
+      "grad_norm": 1.4678056240081787,
+      "learning_rate": 0.00018554578375581997,
+      "loss": 1.9119,
+      "step": 1407
+    },
+    {
+      "epoch": 0.14560119955533724,
+      "grad_norm": 4.0386738777160645,
+      "learning_rate": 0.00018553543714433523,
+      "loss": 2.031,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14570460949820324,
+      "grad_norm": 1.2234376668930054,
+      "learning_rate": 0.0001855250905328505,
+      "loss": 1.4436,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14580801944106925,
+      "grad_norm": 1.4218946695327759,
+      "learning_rate": 0.00018551474392136577,
+      "loss": 1.9956,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14591142938393525,
+      "grad_norm": 1.7964671850204468,
+      "learning_rate": 0.00018550439730988103,
+      "loss": 1.362,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14601483932680126,
+      "grad_norm": 0.8776262998580933,
+      "learning_rate": 0.0001854940506983963,
+      "loss": 1.718,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14611824926966727,
+      "grad_norm": 0.9666465520858765,
+      "learning_rate": 0.00018548370408691154,
+      "loss": 1.6304,
+      "step": 1413
+    },
+    {
+      "epoch": 0.14622165921253327,
+      "grad_norm": 1.7246460914611816,
+      "learning_rate": 0.0001854733574754268,
+      "loss": 1.8842,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14632506915539928,
+      "grad_norm": 1.5267548561096191,
+      "learning_rate": 0.00018546301086394207,
+      "loss": 1.8565,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14642847909826529,
+      "grad_norm": 2.6059603691101074,
+      "learning_rate": 0.00018545266425245733,
+      "loss": 1.7332,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1465318890411313,
+      "grad_norm": 1.2317285537719727,
+      "learning_rate": 0.0001854423176409726,
+      "loss": 1.2607,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1466352989839973,
+      "grad_norm": 1.1376893520355225,
+      "learning_rate": 0.00018543197102948786,
+      "loss": 1.7747,
+      "step": 1418
+    },
+    {
+      "epoch": 0.14673870892686333,
+      "grad_norm": 1.2814754247665405,
+      "learning_rate": 0.0001854216244180031,
+      "loss": 1.6338,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14684211886972934,
+      "grad_norm": 1.9180951118469238,
+      "learning_rate": 0.00018541127780651837,
+      "loss": 1.7912,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14694552881259534,
+      "grad_norm": 1.478571891784668,
+      "learning_rate": 0.00018540093119503363,
+      "loss": 1.771,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14704893875546135,
+      "grad_norm": 1.6875447034835815,
+      "learning_rate": 0.0001853905845835489,
+      "loss": 0.9275,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14715234869832736,
+      "grad_norm": 1.1770519018173218,
+      "learning_rate": 0.00018538023797206416,
+      "loss": 1.1857,
+      "step": 1423
+    },
+    {
+      "epoch": 0.14725575864119336,
+      "grad_norm": 1.1245090961456299,
+      "learning_rate": 0.00018536989136057943,
+      "loss": 1.6672,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14735916858405937,
+      "grad_norm": 2.2437310218811035,
+      "learning_rate": 0.00018535954474909467,
+      "loss": 2.0515,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14746257852692538,
+      "grad_norm": 2.4972667694091797,
+      "learning_rate": 0.00018534919813760993,
+      "loss": 1.848,
+      "step": 1426
+    },
+    {
+      "epoch": 0.14756598846979138,
+      "grad_norm": 1.7464969158172607,
+      "learning_rate": 0.0001853388515261252,
+      "loss": 2.007,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1476693984126574,
+      "grad_norm": 1.0928596258163452,
+      "learning_rate": 0.00018532850491464046,
+      "loss": 1.1851,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1477728083555234,
+      "grad_norm": 1.876115322113037,
+      "learning_rate": 0.00018531815830315573,
+      "loss": 1.2047,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1478762182983894,
+      "grad_norm": 2.2480833530426025,
+      "learning_rate": 0.000185307811691671,
+      "loss": 1.8043,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1479796282412554,
+      "grad_norm": 1.9176514148712158,
+      "learning_rate": 0.00018529746508018623,
+      "loss": 2.0757,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1480830381841214,
+      "grad_norm": 1.2820544242858887,
+      "learning_rate": 0.0001852871184687015,
+      "loss": 1.7068,
+      "step": 1432
+    },
+    {
+      "epoch": 0.14818644812698742,
+      "grad_norm": 1.16887366771698,
+      "learning_rate": 0.00018527677185721677,
+      "loss": 0.8879,
+      "step": 1433
+    },
+    {
+      "epoch": 0.14828985806985343,
+      "grad_norm": 1.29765784740448,
+      "learning_rate": 0.00018526642524573203,
+      "loss": 1.4978,
+      "step": 1434
+    },
+    {
+      "epoch": 0.14839326801271943,
+      "grad_norm": 1.439942479133606,
+      "learning_rate": 0.0001852560786342473,
+      "loss": 1.332,
+      "step": 1435
+    },
+    {
+      "epoch": 0.14849667795558544,
+      "grad_norm": 1.7839429378509521,
+      "learning_rate": 0.00018524573202276256,
+      "loss": 1.9976,
+      "step": 1436
+    },
+    {
+      "epoch": 0.14860008789845144,
+      "grad_norm": 1.2119537591934204,
+      "learning_rate": 0.0001852353854112778,
+      "loss": 2.2104,
+      "step": 1437
+    },
+    {
+      "epoch": 0.14870349784131745,
+      "grad_norm": 1.0513869524002075,
+      "learning_rate": 0.00018522503879979307,
+      "loss": 1.8778,
+      "step": 1438
+    },
+    {
+      "epoch": 0.14880690778418346,
+      "grad_norm": 2.588869571685791,
+      "learning_rate": 0.00018521469218830833,
+      "loss": 1.7518,
+      "step": 1439
+    },
+    {
+      "epoch": 0.14891031772704946,
+      "grad_norm": 2.166912794113159,
+      "learning_rate": 0.0001852043455768236,
+      "loss": 2.1437,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14901372766991547,
+      "grad_norm": 1.8580693006515503,
+      "learning_rate": 0.00018519399896533886,
+      "loss": 1.5795,
+      "step": 1441
+    },
+    {
+      "epoch": 0.14911713761278148,
+      "grad_norm": 1.4996910095214844,
+      "learning_rate": 0.00018518365235385413,
+      "loss": 1.4934,
+      "step": 1442
+    },
+    {
+      "epoch": 0.14922054755564748,
+      "grad_norm": 2.2671191692352295,
+      "learning_rate": 0.00018517330574236937,
+      "loss": 2.2631,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1493239574985135,
+      "grad_norm": 2.4487438201904297,
+      "learning_rate": 0.00018516295913088463,
+      "loss": 2.6226,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1494273674413795,
+      "grad_norm": 0.7984973788261414,
+      "learning_rate": 0.0001851526125193999,
+      "loss": 2.0354,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1495307773842455,
+      "grad_norm": 2.1399385929107666,
+      "learning_rate": 0.00018514226590791516,
+      "loss": 1.4461,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1496341873271115,
+      "grad_norm": 1.6950798034667969,
+      "learning_rate": 0.00018513191929643043,
+      "loss": 1.6589,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1497375972699775,
+      "grad_norm": 1.0480849742889404,
+      "learning_rate": 0.0001851215726849457,
+      "loss": 1.2511,
+      "step": 1448
+    },
+    {
+      "epoch": 0.14984100721284352,
+      "grad_norm": 1.5871855020523071,
+      "learning_rate": 0.00018511122607346093,
+      "loss": 1.4684,
+      "step": 1449
+    },
+    {
+      "epoch": 0.14994441715570953,
+      "grad_norm": 1.3230794668197632,
+      "learning_rate": 0.0001851008794619762,
+      "loss": 1.5079,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15004782709857553,
+      "grad_norm": 2.08754825592041,
+      "learning_rate": 0.00018509053285049146,
+      "loss": 2.3247,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15015123704144154,
+      "grad_norm": 3.3206236362457275,
+      "learning_rate": 0.00018508018623900673,
+      "loss": 1.5526,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15025464698430754,
+      "grad_norm": 1.5044193267822266,
+      "learning_rate": 0.000185069839627522,
+      "loss": 1.8548,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15035805692717355,
+      "grad_norm": 1.271353006362915,
+      "learning_rate": 0.00018505949301603726,
+      "loss": 1.7336,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15046146687003956,
+      "grad_norm": 1.4992772340774536,
+      "learning_rate": 0.0001850491464045525,
+      "loss": 2.1556,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15056487681290556,
+      "grad_norm": 1.797611951828003,
+      "learning_rate": 0.00018503879979306777,
+      "loss": 2.1368,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15066828675577157,
+      "grad_norm": 2.4767348766326904,
+      "learning_rate": 0.00018502845318158303,
+      "loss": 1.3575,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15077169669863758,
+      "grad_norm": 1.9797945022583008,
+      "learning_rate": 0.0001850181065700983,
+      "loss": 1.968,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15087510664150358,
+      "grad_norm": 1.4756824970245361,
+      "learning_rate": 0.00018500775995861356,
+      "loss": 0.9718,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1509785165843696,
+      "grad_norm": 1.1938824653625488,
+      "learning_rate": 0.00018499741334712883,
+      "loss": 1.5162,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1510819265272356,
+      "grad_norm": 2.0596578121185303,
+      "learning_rate": 0.00018498706673564407,
+      "loss": 1.5841,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1511853364701016,
+      "grad_norm": 0.9470666646957397,
+      "learning_rate": 0.00018497672012415933,
+      "loss": 1.5052,
+      "step": 1462
+    },
+    {
+      "epoch": 0.1512887464129676,
+      "grad_norm": 1.30280339717865,
+      "learning_rate": 0.0001849663735126746,
+      "loss": 1.0616,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1513921563558336,
+      "grad_norm": 1.7117403745651245,
+      "learning_rate": 0.00018495602690118986,
+      "loss": 1.6116,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15149556629869962,
+      "grad_norm": 1.2089602947235107,
+      "learning_rate": 0.00018494568028970513,
+      "loss": 1.7289,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15159897624156563,
+      "grad_norm": 2.146265983581543,
+      "learning_rate": 0.0001849353336782204,
+      "loss": 2.1897,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15170238618443163,
+      "grad_norm": 1.0412538051605225,
+      "learning_rate": 0.00018492498706673563,
+      "loss": 1.3649,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15180579612729764,
+      "grad_norm": 1.6170674562454224,
+      "learning_rate": 0.0001849146404552509,
+      "loss": 1.5992,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15190920607016364,
+      "grad_norm": 2.0406131744384766,
+      "learning_rate": 0.00018490429384376616,
+      "loss": 1.7176,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15201261601302965,
+      "grad_norm": 1.5884300470352173,
+      "learning_rate": 0.00018489394723228143,
+      "loss": 1.7171,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15211602595589566,
+      "grad_norm": 1.7351653575897217,
+      "learning_rate": 0.0001848836006207967,
+      "loss": 2.1659,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15221943589876166,
+      "grad_norm": 1.464855670928955,
+      "learning_rate": 0.00018487325400931196,
+      "loss": 1.8101,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15232284584162767,
+      "grad_norm": 1.5113881826400757,
+      "learning_rate": 0.0001848629073978272,
+      "loss": 1.5864,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15242625578449367,
+      "grad_norm": 1.0496164560317993,
+      "learning_rate": 0.00018485256078634246,
+      "loss": 1.3613,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15252966572735968,
+      "grad_norm": 1.4251512289047241,
+      "learning_rate": 0.00018484221417485773,
+      "loss": 0.8508,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1526330756702257,
+      "grad_norm": 1.001116394996643,
+      "learning_rate": 0.000184831867563373,
+      "loss": 1.3093,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1527364856130917,
+      "grad_norm": 2.355113983154297,
+      "learning_rate": 0.00018482152095188826,
+      "loss": 1.7889,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1528398955559577,
+      "grad_norm": 0.9703680276870728,
+      "learning_rate": 0.00018481117434040353,
+      "loss": 2.0384,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1529433054988237,
+      "grad_norm": 1.182063102722168,
+      "learning_rate": 0.0001848008277289188,
+      "loss": 1.3248,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1530467154416897,
+      "grad_norm": 2.8416783809661865,
+      "learning_rate": 0.00018479048111743406,
+      "loss": 2.1734,
+      "step": 1480
+    },
+    {
+      "epoch": 0.15315012538455572,
+      "grad_norm": 1.6548497676849365,
+      "learning_rate": 0.0001847801345059493,
+      "loss": 1.9716,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15325353532742172,
+      "grad_norm": 1.6484485864639282,
+      "learning_rate": 0.00018476978789446456,
+      "loss": 0.5344,
+      "step": 1482
+    },
+    {
+      "epoch": 0.15335694527028773,
+      "grad_norm": 1.1464232206344604,
+      "learning_rate": 0.00018475944128297983,
+      "loss": 1.4424,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15346035521315374,
+      "grad_norm": 1.261054515838623,
+      "learning_rate": 0.0001847490946714951,
+      "loss": 1.8385,
+      "step": 1484
+    },
+    {
+      "epoch": 0.15356376515601974,
+      "grad_norm": 2.0900843143463135,
+      "learning_rate": 0.00018473874806001036,
+      "loss": 2.3219,
+      "step": 1485
+    },
+    {
+      "epoch": 0.15366717509888575,
+      "grad_norm": 2.0020716190338135,
+      "learning_rate": 0.00018472840144852562,
+      "loss": 1.9556,
+      "step": 1486
+    },
+    {
+      "epoch": 0.15377058504175176,
+      "grad_norm": 0.9590654969215393,
+      "learning_rate": 0.0001847180548370409,
+      "loss": 2.0801,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15387399498461776,
+      "grad_norm": 1.3336544036865234,
+      "learning_rate": 0.00018470770822555615,
+      "loss": 1.3413,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15397740492748377,
+      "grad_norm": 1.676986813545227,
+      "learning_rate": 0.00018469736161407142,
+      "loss": 1.1669,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15408081487034977,
+      "grad_norm": 1.1323169469833374,
+      "learning_rate": 0.00018468701500258666,
+      "loss": 1.5916,
+      "step": 1490
+    },
+    {
+      "epoch": 0.15418422481321578,
+      "grad_norm": 1.1335134506225586,
+      "learning_rate": 0.00018467666839110192,
+      "loss": 1.3671,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1542876347560818,
+      "grad_norm": 1.59001886844635,
+      "learning_rate": 0.0001846663217796172,
+      "loss": 1.6888,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1543910446989478,
+      "grad_norm": 0.6160492300987244,
+      "learning_rate": 0.00018465597516813246,
+      "loss": 1.6795,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1544944546418138,
+      "grad_norm": 1.1683189868927002,
+      "learning_rate": 0.00018464562855664772,
+      "loss": 1.5374,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1545978645846798,
+      "grad_norm": 1.4470138549804688,
+      "learning_rate": 0.000184635281945163,
+      "loss": 1.7904,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1547012745275458,
+      "grad_norm": 1.729058027267456,
+      "learning_rate": 0.00018462493533367825,
+      "loss": 1.8934,
+      "step": 1496
+    },
+    {
+      "epoch": 0.15480468447041182,
+      "grad_norm": 1.35962975025177,
+      "learning_rate": 0.0001846145887221935,
+      "loss": 1.4041,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15490809441327782,
+      "grad_norm": 3.6339924335479736,
+      "learning_rate": 0.00018460424211070876,
+      "loss": 1.7157,
+      "step": 1498
+    },
+    {
+      "epoch": 0.15501150435614383,
+      "grad_norm": 1.2886037826538086,
+      "learning_rate": 0.00018459389549922402,
+      "loss": 1.6121,
+      "step": 1499
+    },
+    {
+      "epoch": 0.15511491429900984,
+      "grad_norm": 1.1662096977233887,
+      "learning_rate": 0.0001845835488877393,
+      "loss": 1.1437,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15521832424187584,
+      "grad_norm": 1.9506018161773682,
+      "learning_rate": 0.00018457320227625455,
+      "loss": 2.239,
+      "step": 1501
+    },
+    {
+      "epoch": 0.15532173418474185,
+      "grad_norm": 3.311880111694336,
+      "learning_rate": 0.00018456285566476982,
+      "loss": 1.9514,
+      "step": 1502
+    },
+    {
+      "epoch": 0.15542514412760786,
+      "grad_norm": 0.8730196952819824,
+      "learning_rate": 0.00018455250905328506,
+      "loss": 1.7644,
+      "step": 1503
+    },
+    {
+      "epoch": 0.15552855407047386,
+      "grad_norm": 1.311930775642395,
+      "learning_rate": 0.00018454216244180032,
+      "loss": 1.6487,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1556319640133399,
+      "grad_norm": 1.7224714756011963,
+      "learning_rate": 0.0001845318158303156,
+      "loss": 1.1888,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1557353739562059,
+      "grad_norm": 1.223656415939331,
+      "learning_rate": 0.00018452146921883085,
+      "loss": 1.8018,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1558387838990719,
+      "grad_norm": 1.7321665287017822,
+      "learning_rate": 0.00018451112260734612,
+      "loss": 1.5763,
+      "step": 1507
+    },
+    {
+      "epoch": 0.15594219384193791,
+      "grad_norm": 1.9610759019851685,
+      "learning_rate": 0.00018450077599586138,
+      "loss": 2.2778,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15604560378480392,
+      "grad_norm": 2.585524797439575,
+      "learning_rate": 0.00018449042938437662,
+      "loss": 1.9698,
+      "step": 1509
+    },
+    {
+      "epoch": 0.15614901372766993,
+      "grad_norm": 0.9229252338409424,
+      "learning_rate": 0.0001844800827728919,
+      "loss": 1.4744,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15625242367053593,
+      "grad_norm": 1.3276410102844238,
+      "learning_rate": 0.00018446973616140715,
+      "loss": 1.2333,
+      "step": 1511
+    },
+    {
+      "epoch": 0.15635583361340194,
+      "grad_norm": 1.4182887077331543,
+      "learning_rate": 0.00018445938954992242,
+      "loss": 1.314,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15645924355626795,
+      "grad_norm": 1.714599609375,
+      "learning_rate": 0.00018444904293843769,
+      "loss": 1.7817,
+      "step": 1513
+    },
+    {
+      "epoch": 0.15656265349913395,
+      "grad_norm": 3.0684444904327393,
+      "learning_rate": 0.00018443869632695295,
+      "loss": 1.7853,
+      "step": 1514
+    },
+    {
+      "epoch": 0.15666606344199996,
+      "grad_norm": 0.8052710890769958,
+      "learning_rate": 0.0001844283497154682,
+      "loss": 1.972,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15676947338486596,
+      "grad_norm": 1.2668997049331665,
+      "learning_rate": 0.00018441800310398346,
+      "loss": 2.2012,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15687288332773197,
+      "grad_norm": 2.7253763675689697,
+      "learning_rate": 0.00018440765649249872,
+      "loss": 2.2218,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15697629327059798,
+      "grad_norm": 1.9879379272460938,
+      "learning_rate": 0.000184397309881014,
+      "loss": 1.9497,
+      "step": 1518
+    },
+    {
+      "epoch": 0.15707970321346398,
+      "grad_norm": 1.4316390752792358,
+      "learning_rate": 0.00018438696326952925,
+      "loss": 1.9449,
+      "step": 1519
+    },
+    {
+      "epoch": 0.15718311315633,
+      "grad_norm": 0.9651163816452026,
+      "learning_rate": 0.00018437661665804452,
+      "loss": 1.857,
+      "step": 1520
+    },
+    {
+      "epoch": 0.157286523099196,
+      "grad_norm": 2.6256558895111084,
+      "learning_rate": 0.00018436627004655976,
+      "loss": 1.7782,
+      "step": 1521
+    },
+    {
+      "epoch": 0.157389933042062,
+      "grad_norm": 1.571499228477478,
+      "learning_rate": 0.00018435592343507502,
+      "loss": 1.5711,
+      "step": 1522
+    },
+    {
+      "epoch": 0.157493342984928,
+      "grad_norm": 1.6658003330230713,
+      "learning_rate": 0.0001843455768235903,
+      "loss": 2.0324,
+      "step": 1523
+    },
+    {
+      "epoch": 0.15759675292779401,
+      "grad_norm": 1.4340386390686035,
+      "learning_rate": 0.00018433523021210555,
+      "loss": 1.9085,
+      "step": 1524
+    },
+    {
+      "epoch": 0.15770016287066002,
+      "grad_norm": 1.508362054824829,
+      "learning_rate": 0.00018432488360062082,
+      "loss": 1.6622,
+      "step": 1525
+    },
+    {
+      "epoch": 0.15780357281352603,
+      "grad_norm": 2.414968729019165,
+      "learning_rate": 0.00018431453698913608,
+      "loss": 1.7528,
+      "step": 1526
+    },
+    {
+      "epoch": 0.15790698275639203,
+      "grad_norm": 1.766019344329834,
+      "learning_rate": 0.00018430419037765132,
+      "loss": 1.4326,
+      "step": 1527
+    },
+    {
+      "epoch": 0.15801039269925804,
+      "grad_norm": 1.2096456289291382,
+      "learning_rate": 0.0001842938437661666,
+      "loss": 1.7347,
+      "step": 1528
+    },
+    {
+      "epoch": 0.15811380264212405,
+      "grad_norm": 0.8431787490844727,
+      "learning_rate": 0.00018428349715468185,
+      "loss": 1.3484,
+      "step": 1529
+    },
+    {
+      "epoch": 0.15821721258499005,
+      "grad_norm": 1.123042106628418,
+      "learning_rate": 0.00018427315054319712,
+      "loss": 1.6327,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15832062252785606,
+      "grad_norm": 1.5339807271957397,
+      "learning_rate": 0.00018426280393171238,
+      "loss": 2.3622,
+      "step": 1531
+    },
+    {
+      "epoch": 0.15842403247072206,
+      "grad_norm": 1.0583611726760864,
+      "learning_rate": 0.00018425245732022765,
+      "loss": 1.2064,
+      "step": 1532
+    },
+    {
+      "epoch": 0.15852744241358807,
+      "grad_norm": 1.5183857679367065,
+      "learning_rate": 0.0001842421107087429,
+      "loss": 1.8776,
+      "step": 1533
+    },
+    {
+      "epoch": 0.15863085235645408,
+      "grad_norm": 1.4887058734893799,
+      "learning_rate": 0.00018423176409725815,
+      "loss": 1.762,
+      "step": 1534
+    },
+    {
+      "epoch": 0.15873426229932008,
+      "grad_norm": 1.5032808780670166,
+      "learning_rate": 0.00018422141748577342,
+      "loss": 1.5944,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1588376722421861,
+      "grad_norm": 1.552122950553894,
+      "learning_rate": 0.00018421107087428869,
+      "loss": 2.2603,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1589410821850521,
+      "grad_norm": 1.0997596979141235,
+      "learning_rate": 0.00018420072426280395,
+      "loss": 1.7179,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1590444921279181,
+      "grad_norm": 2.1846163272857666,
+      "learning_rate": 0.00018419037765131922,
+      "loss": 1.8186,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1591479020707841,
+      "grad_norm": 1.4159355163574219,
+      "learning_rate": 0.00018418003103983446,
+      "loss": 1.9963,
+      "step": 1539
+    },
+    {
+      "epoch": 0.15925131201365011,
+      "grad_norm": 1.7774709463119507,
+      "learning_rate": 0.00018416968442834972,
+      "loss": 2.39,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15935472195651612,
+      "grad_norm": 1.3332961797714233,
+      "learning_rate": 0.00018415933781686499,
+      "loss": 1.4559,
+      "step": 1541
+    },
+    {
+      "epoch": 0.15945813189938213,
+      "grad_norm": 1.3671313524246216,
+      "learning_rate": 0.00018414899120538025,
+      "loss": 1.9792,
+      "step": 1542
+    },
+    {
+      "epoch": 0.15956154184224813,
+      "grad_norm": 0.9749310612678528,
+      "learning_rate": 0.00018413864459389552,
+      "loss": 1.8855,
+      "step": 1543
+    },
+    {
+      "epoch": 0.15966495178511414,
+      "grad_norm": 1.3629378080368042,
+      "learning_rate": 0.00018412829798241078,
+      "loss": 2.0187,
+      "step": 1544
+    },
+    {
+      "epoch": 0.15976836172798015,
+      "grad_norm": 2.236741304397583,
+      "learning_rate": 0.00018411795137092602,
+      "loss": 1.6343,
+      "step": 1545
+    },
+    {
+      "epoch": 0.15987177167084615,
+      "grad_norm": 1.9610300064086914,
+      "learning_rate": 0.0001841076047594413,
+      "loss": 1.6565,
+      "step": 1546
+    },
+    {
+      "epoch": 0.15997518161371216,
+      "grad_norm": 1.3427401781082153,
+      "learning_rate": 0.00018409725814795655,
+      "loss": 1.7781,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16007859155657816,
+      "grad_norm": 1.5893080234527588,
+      "learning_rate": 0.00018408691153647182,
+      "loss": 1.6186,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16018200149944417,
+      "grad_norm": 1.7143442630767822,
+      "learning_rate": 0.00018407656492498708,
+      "loss": 1.726,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16028541144231018,
+      "grad_norm": 1.6941561698913574,
+      "learning_rate": 0.00018406621831350235,
+      "loss": 1.4204,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16038882138517618,
+      "grad_norm": 2.1100339889526367,
+      "learning_rate": 0.0001840558717020176,
+      "loss": 2.1381,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1604922313280422,
+      "grad_norm": 1.1537351608276367,
+      "learning_rate": 0.00018404552509053285,
+      "loss": 1.1759,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1605956412709082,
+      "grad_norm": 1.5180721282958984,
+      "learning_rate": 0.00018403517847904812,
+      "loss": 1.6175,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1606990512137742,
+      "grad_norm": 2.0618677139282227,
+      "learning_rate": 0.00018402483186756338,
+      "loss": 1.5854,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1608024611566402,
+      "grad_norm": 1.102594017982483,
+      "learning_rate": 0.00018401448525607865,
+      "loss": 1.3946,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16090587109950621,
+      "grad_norm": 1.2164007425308228,
+      "learning_rate": 0.00018400413864459392,
+      "loss": 1.449,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16100928104237222,
+      "grad_norm": 1.3186836242675781,
+      "learning_rate": 0.00018399379203310915,
+      "loss": 1.5946,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16111269098523823,
+      "grad_norm": 2.05452036857605,
+      "learning_rate": 0.00018398344542162442,
+      "loss": 1.7512,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16121610092810423,
+      "grad_norm": 1.482884168624878,
+      "learning_rate": 0.00018397309881013969,
+      "loss": 1.5238,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16131951087097024,
+      "grad_norm": 1.144921898841858,
+      "learning_rate": 0.00018396275219865495,
+      "loss": 1.598,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16142292081383625,
+      "grad_norm": 0.9143129587173462,
+      "learning_rate": 0.00018395240558717022,
+      "loss": 1.3763,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16152633075670225,
+      "grad_norm": 1.4671123027801514,
+      "learning_rate": 0.00018394205897568548,
+      "loss": 1.8797,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16162974069956826,
+      "grad_norm": 0.8517533540725708,
+      "learning_rate": 0.00018393171236420072,
+      "loss": 1.8004,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16173315064243426,
+      "grad_norm": 1.7459747791290283,
+      "learning_rate": 0.00018392136575271599,
+      "loss": 1.5426,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16183656058530027,
+      "grad_norm": 1.2846120595932007,
+      "learning_rate": 0.00018391101914123125,
+      "loss": 1.2563,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16193997052816628,
+      "grad_norm": 1.0815187692642212,
+      "learning_rate": 0.00018390067252974652,
+      "loss": 1.9735,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16204338047103228,
+      "grad_norm": 1.4935349225997925,
+      "learning_rate": 0.00018389032591826178,
+      "loss": 1.9829,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1621467904138983,
+      "grad_norm": 0.7376902103424072,
+      "learning_rate": 0.00018387997930677705,
+      "loss": 2.0292,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1622502003567643,
+      "grad_norm": 0.8843541741371155,
+      "learning_rate": 0.0001838696326952923,
+      "loss": 1.3294,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1623536102996303,
+      "grad_norm": 0.8407658934593201,
+      "learning_rate": 0.00018385928608380755,
+      "loss": 1.499,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1624570202424963,
+      "grad_norm": 1.1754120588302612,
+      "learning_rate": 0.00018384893947232282,
+      "loss": 1.832,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16256043018536231,
+      "grad_norm": 2.4817240238189697,
+      "learning_rate": 0.00018383859286083808,
+      "loss": 2.1077,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16266384012822832,
+      "grad_norm": 1.4242192506790161,
+      "learning_rate": 0.00018382824624935335,
+      "loss": 1.9378,
+      "step": 1573
+    },
+    {
+      "epoch": 0.16276725007109433,
+      "grad_norm": 1.4254143238067627,
+      "learning_rate": 0.0001838178996378686,
+      "loss": 1.5059,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16287066001396033,
+      "grad_norm": 1.4265427589416504,
+      "learning_rate": 0.00018380755302638385,
+      "loss": 1.9646,
+      "step": 1575
+    },
+    {
+      "epoch": 0.16297406995682634,
+      "grad_norm": 1.5007380247116089,
+      "learning_rate": 0.00018379720641489912,
+      "loss": 1.1932,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16307747989969235,
+      "grad_norm": 1.8024120330810547,
+      "learning_rate": 0.00018378685980341438,
+      "loss": 1.8475,
+      "step": 1577
+    },
+    {
+      "epoch": 0.16318088984255835,
+      "grad_norm": 1.2183353900909424,
+      "learning_rate": 0.00018377651319192965,
+      "loss": 1.9732,
+      "step": 1578
+    },
+    {
+      "epoch": 0.16328429978542436,
+      "grad_norm": 1.5071296691894531,
+      "learning_rate": 0.00018376616658044492,
+      "loss": 2.4146,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16338770972829036,
+      "grad_norm": 1.172339916229248,
+      "learning_rate": 0.00018375581996896015,
+      "loss": 1.4546,
+      "step": 1580
+    },
+    {
+      "epoch": 0.16349111967115637,
+      "grad_norm": 2.6015713214874268,
+      "learning_rate": 0.00018374547335747542,
+      "loss": 2.5725,
+      "step": 1581
+    },
+    {
+      "epoch": 0.16359452961402238,
+      "grad_norm": 1.1628609895706177,
+      "learning_rate": 0.00018373512674599068,
+      "loss": 1.3454,
+      "step": 1582
+    },
+    {
+      "epoch": 0.16369793955688838,
+      "grad_norm": 1.3187514543533325,
+      "learning_rate": 0.00018372478013450595,
+      "loss": 1.5801,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1638013494997544,
+      "grad_norm": 1.2284473180770874,
+      "learning_rate": 0.00018371443352302122,
+      "loss": 1.7379,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1639047594426204,
+      "grad_norm": 1.388885259628296,
+      "learning_rate": 0.00018370408691153648,
+      "loss": 1.3364,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1640081693854864,
+      "grad_norm": 1.2158303260803223,
+      "learning_rate": 0.00018369374030005172,
+      "loss": 1.6606,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1641115793283524,
+      "grad_norm": 1.6560564041137695,
+      "learning_rate": 0.00018368339368856699,
+      "loss": 1.8817,
+      "step": 1587
+    },
+    {
+      "epoch": 0.16421498927121841,
+      "grad_norm": 1.2663594484329224,
+      "learning_rate": 0.00018367304707708225,
+      "loss": 1.6547,
+      "step": 1588
+    },
+    {
+      "epoch": 0.16431839921408442,
+      "grad_norm": 1.790147304534912,
+      "learning_rate": 0.00018366270046559752,
+      "loss": 1.5503,
+      "step": 1589
+    },
+    {
+      "epoch": 0.16442180915695043,
+      "grad_norm": 1.1513880491256714,
+      "learning_rate": 0.00018365235385411278,
+      "loss": 1.6339,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16452521909981646,
+      "grad_norm": 1.9630792140960693,
+      "learning_rate": 0.00018364200724262805,
+      "loss": 2.0134,
+      "step": 1591
+    },
+    {
+      "epoch": 0.16462862904268247,
+      "grad_norm": 1.1871670484542847,
+      "learning_rate": 0.00018363166063114329,
+      "loss": 1.5574,
+      "step": 1592
+    },
+    {
+      "epoch": 0.16473203898554847,
+      "grad_norm": 1.6941906213760376,
+      "learning_rate": 0.00018362131401965855,
+      "loss": 1.566,
+      "step": 1593
+    },
+    {
+      "epoch": 0.16483544892841448,
+      "grad_norm": 1.7763041257858276,
+      "learning_rate": 0.00018361096740817382,
+      "loss": 1.933,
+      "step": 1594
+    },
+    {
+      "epoch": 0.16493885887128049,
+      "grad_norm": 1.50661039352417,
+      "learning_rate": 0.00018360062079668908,
+      "loss": 1.3962,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1650422688141465,
+      "grad_norm": 1.118114709854126,
+      "learning_rate": 0.00018359027418520435,
+      "loss": 1.6802,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1651456787570125,
+      "grad_norm": 1.108808159828186,
+      "learning_rate": 0.00018357992757371961,
+      "loss": 1.2169,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1652490886998785,
+      "grad_norm": 1.5891281366348267,
+      "learning_rate": 0.00018356958096223485,
+      "loss": 1.9502,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1653524986427445,
+      "grad_norm": 0.7753489017486572,
+      "learning_rate": 0.00018355923435075012,
+      "loss": 1.577,
+      "step": 1599
+    },
+    {
+      "epoch": 0.16545590858561052,
+      "grad_norm": 1.6518728733062744,
+      "learning_rate": 0.00018354888773926538,
+      "loss": 1.9462,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16555931852847652,
+      "grad_norm": 2.548527240753174,
+      "learning_rate": 0.00018353854112778065,
+      "loss": 1.8724,
+      "step": 1601
+    },
+    {
+      "epoch": 0.16566272847134253,
+      "grad_norm": 1.5010498762130737,
+      "learning_rate": 0.00018352819451629592,
+      "loss": 1.8285,
+      "step": 1602
+    },
+    {
+      "epoch": 0.16576613841420854,
+      "grad_norm": 1.1613878011703491,
+      "learning_rate": 0.00018351784790481118,
+      "loss": 2.121,
+      "step": 1603
+    },
+    {
+      "epoch": 0.16586954835707454,
+      "grad_norm": 1.3867918252944946,
+      "learning_rate": 0.00018350750129332645,
+      "loss": 1.1178,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16597295829994055,
+      "grad_norm": 2.6823570728302,
+      "learning_rate": 0.0001834971546818417,
+      "loss": 2.086,
+      "step": 1605
+    },
+    {
+      "epoch": 0.16607636824280655,
+      "grad_norm": 1.298323392868042,
+      "learning_rate": 0.00018348680807035695,
+      "loss": 1.8797,
+      "step": 1606
+    },
+    {
+      "epoch": 0.16617977818567256,
+      "grad_norm": 1.6981557607650757,
+      "learning_rate": 0.00018347646145887222,
+      "loss": 1.7928,
+      "step": 1607
+    },
+    {
+      "epoch": 0.16628318812853857,
+      "grad_norm": 1.8808917999267578,
+      "learning_rate": 0.00018346611484738748,
+      "loss": 1.7693,
+      "step": 1608
+    },
+    {
+      "epoch": 0.16638659807140457,
+      "grad_norm": 1.246436595916748,
+      "learning_rate": 0.00018345576823590275,
+      "loss": 1.6371,
+      "step": 1609
+    },
+    {
+      "epoch": 0.16649000801427058,
+      "grad_norm": 1.5189836025238037,
+      "learning_rate": 0.000183445421624418,
+      "loss": 1.8056,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16659341795713659,
+      "grad_norm": 1.1304303407669067,
+      "learning_rate": 0.00018343507501293328,
+      "loss": 1.7153,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1666968279000026,
+      "grad_norm": 1.4506385326385498,
+      "learning_rate": 0.00018342472840144854,
+      "loss": 2.0781,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1668002378428686,
+      "grad_norm": 2.3718414306640625,
+      "learning_rate": 0.0001834143817899638,
+      "loss": 2.0175,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1669036477857346,
+      "grad_norm": 1.657726526260376,
+      "learning_rate": 0.00018340403517847905,
+      "loss": 1.8869,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1670070577286006,
+      "grad_norm": 1.561111569404602,
+      "learning_rate": 0.0001833936885669943,
+      "loss": 1.9702,
+      "step": 1615
+    },
+    {
+      "epoch": 0.16711046767146662,
+      "grad_norm": 1.083419919013977,
+      "learning_rate": 0.00018338334195550958,
+      "loss": 1.4907,
+      "step": 1616
+    },
+    {
+      "epoch": 0.16721387761433262,
+      "grad_norm": 1.3286840915679932,
+      "learning_rate": 0.00018337299534402484,
+      "loss": 1.436,
+      "step": 1617
+    },
+    {
+      "epoch": 0.16731728755719863,
+      "grad_norm": 1.2173808813095093,
+      "learning_rate": 0.0001833626487325401,
+      "loss": 1.8981,
+      "step": 1618
+    },
+    {
+      "epoch": 0.16742069750006464,
+      "grad_norm": 1.6411387920379639,
+      "learning_rate": 0.00018335230212105538,
+      "loss": 1.9403,
+      "step": 1619
+    },
+    {
+      "epoch": 0.16752410744293064,
+      "grad_norm": 1.6110321283340454,
+      "learning_rate": 0.00018334195550957064,
+      "loss": 2.2896,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16762751738579665,
+      "grad_norm": 2.330904245376587,
+      "learning_rate": 0.0001833316088980859,
+      "loss": 1.8244,
+      "step": 1621
+    },
+    {
+      "epoch": 0.16773092732866265,
+      "grad_norm": 0.8903806209564209,
+      "learning_rate": 0.00018332126228660115,
+      "loss": 1.7901,
+      "step": 1622
+    },
+    {
+      "epoch": 0.16783433727152866,
+      "grad_norm": 1.515162467956543,
+      "learning_rate": 0.0001833109156751164,
+      "loss": 1.739,
+      "step": 1623
+    },
+    {
+      "epoch": 0.16793774721439467,
+      "grad_norm": 1.395464539527893,
+      "learning_rate": 0.00018330056906363168,
+      "loss": 1.5509,
+      "step": 1624
+    },
+    {
+      "epoch": 0.16804115715726067,
+      "grad_norm": 1.7974237203598022,
+      "learning_rate": 0.00018329022245214694,
+      "loss": 1.8329,
+      "step": 1625
+    },
+    {
+      "epoch": 0.16814456710012668,
+      "grad_norm": 1.6956497430801392,
+      "learning_rate": 0.0001832798758406622,
+      "loss": 1.5083,
+      "step": 1626
+    },
+    {
+      "epoch": 0.16824797704299269,
+      "grad_norm": 2.477332592010498,
+      "learning_rate": 0.00018326952922917747,
+      "loss": 2.1227,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1683513869858587,
+      "grad_norm": 1.1041282415390015,
+      "learning_rate": 0.0001832591826176927,
+      "loss": 2.3492,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1684547969287247,
+      "grad_norm": 2.6464221477508545,
+      "learning_rate": 0.00018324883600620798,
+      "loss": 2.2591,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1685582068715907,
+      "grad_norm": 1.3996168375015259,
+      "learning_rate": 0.00018323848939472324,
+      "loss": 0.9682,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1686616168144567,
+      "grad_norm": 1.6426746845245361,
+      "learning_rate": 0.0001832281427832385,
+      "loss": 1.9295,
+      "step": 1631
+    },
+    {
+      "epoch": 0.16876502675732272,
+      "grad_norm": 1.202686071395874,
+      "learning_rate": 0.00018321779617175377,
+      "loss": 1.3054,
+      "step": 1632
+    },
+    {
+      "epoch": 0.16886843670018872,
+      "grad_norm": 1.330463171005249,
+      "learning_rate": 0.00018320744956026904,
+      "loss": 1.2563,
+      "step": 1633
+    },
+    {
+      "epoch": 0.16897184664305473,
+      "grad_norm": 1.0541915893554688,
+      "learning_rate": 0.00018319710294878428,
+      "loss": 1.7607,
+      "step": 1634
+    },
+    {
+      "epoch": 0.16907525658592074,
+      "grad_norm": 0.961632251739502,
+      "learning_rate": 0.00018318675633729954,
+      "loss": 1.555,
+      "step": 1635
+    },
+    {
+      "epoch": 0.16917866652878674,
+      "grad_norm": 1.850738286972046,
+      "learning_rate": 0.0001831764097258148,
+      "loss": 1.6219,
+      "step": 1636
+    },
+    {
+      "epoch": 0.16928207647165275,
+      "grad_norm": 1.7717586755752563,
+      "learning_rate": 0.00018316606311433007,
+      "loss": 2.3512,
+      "step": 1637
+    },
+    {
+      "epoch": 0.16938548641451875,
+      "grad_norm": 0.7864888310432434,
+      "learning_rate": 0.00018315571650284534,
+      "loss": 1.3779,
+      "step": 1638
+    },
+    {
+      "epoch": 0.16948889635738476,
+      "grad_norm": 1.4093188047409058,
+      "learning_rate": 0.0001831453698913606,
+      "loss": 1.5036,
+      "step": 1639
+    },
+    {
+      "epoch": 0.16959230630025077,
+      "grad_norm": 1.5282201766967773,
+      "learning_rate": 0.00018313502327987584,
+      "loss": 1.4186,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16969571624311677,
+      "grad_norm": 1.574284553527832,
+      "learning_rate": 0.0001831246766683911,
+      "loss": 1.483,
+      "step": 1641
+    },
+    {
+      "epoch": 0.16979912618598278,
+      "grad_norm": 2.263618230819702,
+      "learning_rate": 0.00018311433005690638,
+      "loss": 2.1512,
+      "step": 1642
+    },
+    {
+      "epoch": 0.16990253612884879,
+      "grad_norm": 1.4073622226715088,
+      "learning_rate": 0.00018310398344542164,
+      "loss": 2.0069,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1700059460717148,
+      "grad_norm": 2.1453514099121094,
+      "learning_rate": 0.0001830936368339369,
+      "loss": 2.1053,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1701093560145808,
+      "grad_norm": 2.447608709335327,
+      "learning_rate": 0.00018308329022245217,
+      "loss": 1.8623,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 1.4036715030670166,
+      "learning_rate": 0.0001830729436109674,
+      "loss": 1.3103,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1703161759003128,
+      "grad_norm": 1.4812512397766113,
+      "learning_rate": 0.00018306259699948268,
+      "loss": 1.4172,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17041958584317882,
+      "grad_norm": 1.730870008468628,
+      "learning_rate": 0.00018305225038799794,
+      "loss": 1.5538,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17052299578604482,
+      "grad_norm": 1.4224426746368408,
+      "learning_rate": 0.0001830419037765132,
+      "loss": 2.0328,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17062640572891083,
+      "grad_norm": 1.2195708751678467,
+      "learning_rate": 0.00018303155716502847,
+      "loss": 1.5262,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17072981567177684,
+      "grad_norm": 1.5029513835906982,
+      "learning_rate": 0.00018302121055354374,
+      "loss": 1.6855,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17083322561464284,
+      "grad_norm": 1.6222515106201172,
+      "learning_rate": 0.00018301086394205898,
+      "loss": 1.6238,
+      "step": 1652
+    },
+    {
+      "epoch": 0.17093663555750885,
+      "grad_norm": 1.0669522285461426,
+      "learning_rate": 0.00018300051733057424,
+      "loss": 1.3903,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17104004550037485,
+      "grad_norm": 2.3513834476470947,
+      "learning_rate": 0.0001829901707190895,
+      "loss": 1.9285,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17114345544324086,
+      "grad_norm": 1.5260512828826904,
+      "learning_rate": 0.00018297982410760477,
+      "loss": 1.1363,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17124686538610687,
+      "grad_norm": 2.7969038486480713,
+      "learning_rate": 0.00018296947749612004,
+      "loss": 2.721,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17135027532897287,
+      "grad_norm": 0.9958081245422363,
+      "learning_rate": 0.0001829591308846353,
+      "loss": 1.535,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17145368527183888,
+      "grad_norm": 2.141313314437866,
+      "learning_rate": 0.00018294878427315054,
+      "loss": 2.1867,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17155709521470489,
+      "grad_norm": 1.7324568033218384,
+      "learning_rate": 0.0001829384376616658,
+      "loss": 2.0558,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1716605051575709,
+      "grad_norm": 2.449826240539551,
+      "learning_rate": 0.00018292809105018107,
+      "loss": 1.8396,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1717639151004369,
+      "grad_norm": 1.0371530055999756,
+      "learning_rate": 0.00018291774443869634,
+      "loss": 1.5168,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1718673250433029,
+      "grad_norm": 2.267475128173828,
+      "learning_rate": 0.0001829073978272116,
+      "loss": 1.443,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1719707349861689,
+      "grad_norm": 1.4712553024291992,
+      "learning_rate": 0.00018289705121572687,
+      "loss": 1.5452,
+      "step": 1663
+    },
+    {
+      "epoch": 0.17207414492903492,
+      "grad_norm": 2.187161445617676,
+      "learning_rate": 0.0001828867046042421,
+      "loss": 2.2943,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17217755487190092,
+      "grad_norm": 1.8305400609970093,
+      "learning_rate": 0.00018287635799275738,
+      "loss": 1.5445,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17228096481476693,
+      "grad_norm": 2.1179354190826416,
+      "learning_rate": 0.00018286601138127264,
+      "loss": 1.921,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17238437475763294,
+      "grad_norm": 1.11764395236969,
+      "learning_rate": 0.0001828556647697879,
+      "loss": 1.6441,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17248778470049894,
+      "grad_norm": 1.8130611181259155,
+      "learning_rate": 0.00018284531815830317,
+      "loss": 1.5732,
+      "step": 1668
+    },
+    {
+      "epoch": 0.17259119464336495,
+      "grad_norm": 1.2046823501586914,
+      "learning_rate": 0.00018283497154681844,
+      "loss": 1.3871,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17269460458623095,
+      "grad_norm": 1.1269457340240479,
+      "learning_rate": 0.00018282462493533368,
+      "loss": 1.7667,
+      "step": 1670
+    },
+    {
+      "epoch": 0.17279801452909696,
+      "grad_norm": 0.9750357270240784,
+      "learning_rate": 0.00018281427832384894,
+      "loss": 1.4863,
+      "step": 1671
+    },
+    {
+      "epoch": 0.17290142447196297,
+      "grad_norm": 2.1386210918426514,
+      "learning_rate": 0.0001828039317123642,
+      "loss": 1.6579,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17300483441482897,
+      "grad_norm": 1.1148815155029297,
+      "learning_rate": 0.00018279358510087947,
+      "loss": 0.9662,
+      "step": 1673
+    },
+    {
+      "epoch": 0.17310824435769498,
+      "grad_norm": 1.1114896535873413,
+      "learning_rate": 0.00018278323848939474,
+      "loss": 1.2727,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17321165430056099,
+      "grad_norm": 1.411158561706543,
+      "learning_rate": 0.00018277289187791,
+      "loss": 1.8358,
+      "step": 1675
+    },
+    {
+      "epoch": 0.173315064243427,
+      "grad_norm": 1.026279091835022,
+      "learning_rate": 0.00018276254526642524,
+      "loss": 1.4396,
+      "step": 1676
+    },
+    {
+      "epoch": 0.17341847418629303,
+      "grad_norm": 1.8113138675689697,
+      "learning_rate": 0.0001827521986549405,
+      "loss": 1.0483,
+      "step": 1677
+    },
+    {
+      "epoch": 0.17352188412915903,
+      "grad_norm": 1.5766301155090332,
+      "learning_rate": 0.00018274185204345577,
+      "loss": 1.8646,
+      "step": 1678
+    },
+    {
+      "epoch": 0.17362529407202504,
+      "grad_norm": 1.361986517906189,
+      "learning_rate": 0.00018273150543197104,
+      "loss": 1.3236,
+      "step": 1679
+    },
+    {
+      "epoch": 0.17372870401489104,
+      "grad_norm": 2.4059231281280518,
+      "learning_rate": 0.0001827211588204863,
+      "loss": 2.188,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17383211395775705,
+      "grad_norm": 5.3925886154174805,
+      "learning_rate": 0.00018271081220900157,
+      "loss": 1.6441,
+      "step": 1681
+    },
+    {
+      "epoch": 0.17393552390062306,
+      "grad_norm": 1.831649661064148,
+      "learning_rate": 0.0001827004655975168,
+      "loss": 1.9166,
+      "step": 1682
+    },
+    {
+      "epoch": 0.17403893384348906,
+      "grad_norm": 1.7696832418441772,
+      "learning_rate": 0.00018269011898603207,
+      "loss": 1.6002,
+      "step": 1683
+    },
+    {
+      "epoch": 0.17414234378635507,
+      "grad_norm": 1.0492238998413086,
+      "learning_rate": 0.00018267977237454734,
+      "loss": 1.6034,
+      "step": 1684
+    },
+    {
+      "epoch": 0.17424575372922108,
+      "grad_norm": 2.082101821899414,
+      "learning_rate": 0.0001826694257630626,
+      "loss": 1.5458,
+      "step": 1685
+    },
+    {
+      "epoch": 0.17434916367208708,
+      "grad_norm": 1.129961609840393,
+      "learning_rate": 0.00018265907915157787,
+      "loss": 0.9186,
+      "step": 1686
+    },
+    {
+      "epoch": 0.1744525736149531,
+      "grad_norm": 1.0595530271530151,
+      "learning_rate": 0.00018264873254009314,
+      "loss": 1.5479,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1745559835578191,
+      "grad_norm": 0.6668352484703064,
+      "learning_rate": 0.00018263838592860837,
+      "loss": 1.4539,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1746593935006851,
+      "grad_norm": 1.6207376718521118,
+      "learning_rate": 0.00018262803931712364,
+      "loss": 1.7107,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1747628034435511,
+      "grad_norm": 1.0871083736419678,
+      "learning_rate": 0.0001826176927056389,
+      "loss": 1.9602,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1748662133864171,
+      "grad_norm": 1.131425380706787,
+      "learning_rate": 0.00018260734609415417,
+      "loss": 1.2928,
+      "step": 1691
+    },
+    {
+      "epoch": 0.17496962332928312,
+      "grad_norm": 1.9855254888534546,
+      "learning_rate": 0.00018259699948266944,
+      "loss": 1.5173,
+      "step": 1692
+    },
+    {
+      "epoch": 0.17507303327214913,
+      "grad_norm": 1.17534601688385,
+      "learning_rate": 0.0001825866528711847,
+      "loss": 1.185,
+      "step": 1693
+    },
+    {
+      "epoch": 0.17517644321501513,
+      "grad_norm": 1.2976510524749756,
+      "learning_rate": 0.00018257630625969994,
+      "loss": 1.7015,
+      "step": 1694
+    },
+    {
+      "epoch": 0.17527985315788114,
+      "grad_norm": 1.7299031019210815,
+      "learning_rate": 0.0001825659596482152,
+      "loss": 1.9024,
+      "step": 1695
+    },
+    {
+      "epoch": 0.17538326310074714,
+      "grad_norm": 1.3706202507019043,
+      "learning_rate": 0.00018255561303673047,
+      "loss": 0.9326,
+      "step": 1696
+    },
+    {
+      "epoch": 0.17548667304361315,
+      "grad_norm": 1.2045725584030151,
+      "learning_rate": 0.00018254526642524574,
+      "loss": 1.2952,
+      "step": 1697
+    },
+    {
+      "epoch": 0.17559008298647916,
+      "grad_norm": 1.2185454368591309,
+      "learning_rate": 0.000182534919813761,
+      "loss": 1.9565,
+      "step": 1698
+    },
+    {
+      "epoch": 0.17569349292934516,
+      "grad_norm": 2.578016996383667,
+      "learning_rate": 0.00018252457320227627,
+      "loss": 1.8957,
+      "step": 1699
+    },
+    {
+      "epoch": 0.17579690287221117,
+      "grad_norm": 1.152272343635559,
+      "learning_rate": 0.0001825142265907915,
+      "loss": 1.6591,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17590031281507718,
+      "grad_norm": 2.588287591934204,
+      "learning_rate": 0.00018250387997930677,
+      "loss": 1.9043,
+      "step": 1701
+    },
+    {
+      "epoch": 0.17600372275794318,
+      "grad_norm": 1.1727057695388794,
+      "learning_rate": 0.00018249353336782204,
+      "loss": 1.464,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1761071327008092,
+      "grad_norm": 1.9140554666519165,
+      "learning_rate": 0.0001824831867563373,
+      "loss": 2.1905,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1762105426436752,
+      "grad_norm": 1.4966853857040405,
+      "learning_rate": 0.00018247284014485257,
+      "loss": 1.7903,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1763139525865412,
+      "grad_norm": 1.9793884754180908,
+      "learning_rate": 0.00018246249353336784,
+      "loss": 0.965,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1764173625294072,
+      "grad_norm": 1.5424928665161133,
+      "learning_rate": 0.00018245214692188307,
+      "loss": 2.1276,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1765207724722732,
+      "grad_norm": 2.2728068828582764,
+      "learning_rate": 0.00018244180031039834,
+      "loss": 2.4757,
+      "step": 1707
+    },
+    {
+      "epoch": 0.17662418241513922,
+      "grad_norm": 1.7877531051635742,
+      "learning_rate": 0.0001824314536989136,
+      "loss": 1.708,
+      "step": 1708
+    },
+    {
+      "epoch": 0.17672759235800523,
+      "grad_norm": 1.9923186302185059,
+      "learning_rate": 0.00018242110708742887,
+      "loss": 1.5047,
+      "step": 1709
+    },
+    {
+      "epoch": 0.17683100230087123,
+      "grad_norm": 2.096897840499878,
+      "learning_rate": 0.00018241076047594414,
+      "loss": 1.7008,
+      "step": 1710
+    },
+    {
+      "epoch": 0.17693441224373724,
+      "grad_norm": 0.9570733904838562,
+      "learning_rate": 0.0001824004138644594,
+      "loss": 1.6933,
+      "step": 1711
+    },
+    {
+      "epoch": 0.17703782218660324,
+      "grad_norm": 1.9453891515731812,
+      "learning_rate": 0.00018239006725297464,
+      "loss": 1.769,
+      "step": 1712
+    },
+    {
+      "epoch": 0.17714123212946925,
+      "grad_norm": 0.9500687718391418,
+      "learning_rate": 0.0001823797206414899,
+      "loss": 1.6701,
+      "step": 1713
+    },
+    {
+      "epoch": 0.17724464207233526,
+      "grad_norm": 1.359328031539917,
+      "learning_rate": 0.00018236937403000517,
+      "loss": 1.742,
+      "step": 1714
+    },
+    {
+      "epoch": 0.17734805201520126,
+      "grad_norm": 1.8505967855453491,
+      "learning_rate": 0.00018235902741852044,
+      "loss": 1.4796,
+      "step": 1715
+    },
+    {
+      "epoch": 0.17745146195806727,
+      "grad_norm": 2.4523441791534424,
+      "learning_rate": 0.0001823486808070357,
+      "loss": 2.1651,
+      "step": 1716
+    },
+    {
+      "epoch": 0.17755487190093328,
+      "grad_norm": 1.6013096570968628,
+      "learning_rate": 0.00018233833419555097,
+      "loss": 1.6705,
+      "step": 1717
+    },
+    {
+      "epoch": 0.17765828184379928,
+      "grad_norm": 1.9895267486572266,
+      "learning_rate": 0.0001823279875840662,
+      "loss": 1.3338,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1777616917866653,
+      "grad_norm": 1.7524160146713257,
+      "learning_rate": 0.00018231764097258147,
+      "loss": 2.2723,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1778651017295313,
+      "grad_norm": 2.6437292098999023,
+      "learning_rate": 0.00018230729436109674,
+      "loss": 1.9029,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1779685116723973,
+      "grad_norm": 1.6161471605300903,
+      "learning_rate": 0.000182296947749612,
+      "loss": 1.7033,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1780719216152633,
+      "grad_norm": 1.4976915121078491,
+      "learning_rate": 0.00018228660113812727,
+      "loss": 1.7068,
+      "step": 1722
+    },
+    {
+      "epoch": 0.1781753315581293,
+      "grad_norm": 1.0355263948440552,
+      "learning_rate": 0.00018227625452664253,
+      "loss": 1.9508,
+      "step": 1723
+    },
+    {
+      "epoch": 0.17827874150099532,
+      "grad_norm": 0.9397589564323425,
+      "learning_rate": 0.00018226590791515777,
+      "loss": 1.8944,
+      "step": 1724
+    },
+    {
+      "epoch": 0.17838215144386133,
+      "grad_norm": 1.7493468523025513,
+      "learning_rate": 0.00018225556130367304,
+      "loss": 1.7938,
+      "step": 1725
+    },
+    {
+      "epoch": 0.17848556138672733,
+      "grad_norm": 2.3740999698638916,
+      "learning_rate": 0.0001822452146921883,
+      "loss": 1.9897,
+      "step": 1726
+    },
+    {
+      "epoch": 0.17858897132959334,
+      "grad_norm": 1.8972642421722412,
+      "learning_rate": 0.00018223486808070357,
+      "loss": 2.481,
+      "step": 1727
+    },
+    {
+      "epoch": 0.17869238127245934,
+      "grad_norm": 1.6285275220870972,
+      "learning_rate": 0.00018222452146921883,
+      "loss": 1.9144,
+      "step": 1728
+    },
+    {
+      "epoch": 0.17879579121532535,
+      "grad_norm": 3.3694674968719482,
+      "learning_rate": 0.0001822141748577341,
+      "loss": 2.0067,
+      "step": 1729
+    },
+    {
+      "epoch": 0.17889920115819136,
+      "grad_norm": 1.9680372476577759,
+      "learning_rate": 0.00018220382824624937,
+      "loss": 2.1982,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17900261110105736,
+      "grad_norm": 1.4508297443389893,
+      "learning_rate": 0.0001821934816347646,
+      "loss": 1.5265,
+      "step": 1731
+    },
+    {
+      "epoch": 0.17910602104392337,
+      "grad_norm": 1.2594949007034302,
+      "learning_rate": 0.00018218313502327987,
+      "loss": 1.5713,
+      "step": 1732
+    },
+    {
+      "epoch": 0.17920943098678938,
+      "grad_norm": 1.9360672235488892,
+      "learning_rate": 0.00018217278841179514,
+      "loss": 1.9253,
+      "step": 1733
+    },
+    {
+      "epoch": 0.17931284092965538,
+      "grad_norm": 3.0524022579193115,
+      "learning_rate": 0.0001821624418003104,
+      "loss": 2.1621,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1794162508725214,
+      "grad_norm": 1.465190052986145,
+      "learning_rate": 0.00018215209518882567,
+      "loss": 1.3045,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1795196608153874,
+      "grad_norm": 1.1517199277877808,
+      "learning_rate": 0.00018214174857734093,
+      "loss": 1.381,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1796230707582534,
+      "grad_norm": 3.004659414291382,
+      "learning_rate": 0.0001821314019658562,
+      "loss": 2.0144,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1797264807011194,
+      "grad_norm": 1.5032134056091309,
+      "learning_rate": 0.00018212105535437146,
+      "loss": 1.3108,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1798298906439854,
+      "grad_norm": 1.3385944366455078,
+      "learning_rate": 0.0001821107087428867,
+      "loss": 1.4403,
+      "step": 1739
+    },
+    {
+      "epoch": 0.17993330058685142,
+      "grad_norm": 1.471063494682312,
+      "learning_rate": 0.00018210036213140197,
+      "loss": 1.6935,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18003671052971743,
+      "grad_norm": 2.064852714538574,
+      "learning_rate": 0.00018209001551991723,
+      "loss": 1.8696,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18014012047258343,
+      "grad_norm": 2.8570311069488525,
+      "learning_rate": 0.0001820796689084325,
+      "loss": 1.6901,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18024353041544944,
+      "grad_norm": 1.3590514659881592,
+      "learning_rate": 0.00018206932229694776,
+      "loss": 1.61,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18034694035831544,
+      "grad_norm": 1.277525544166565,
+      "learning_rate": 0.00018205897568546303,
+      "loss": 1.8969,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18045035030118145,
+      "grad_norm": 1.2931078672409058,
+      "learning_rate": 0.0001820486290739783,
+      "loss": 1.4851,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18055376024404746,
+      "grad_norm": 1.664851188659668,
+      "learning_rate": 0.00018203828246249356,
+      "loss": 1.8105,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18065717018691346,
+      "grad_norm": 1.2162920236587524,
+      "learning_rate": 0.0001820279358510088,
+      "loss": 1.3794,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18076058012977947,
+      "grad_norm": 0.8630446791648865,
+      "learning_rate": 0.00018201758923952407,
+      "loss": 1.7145,
+      "step": 1748
+    },
+    {
+      "epoch": 0.18086399007264548,
+      "grad_norm": 2.054414749145508,
+      "learning_rate": 0.00018200724262803933,
+      "loss": 1.4697,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18096740001551148,
+      "grad_norm": 1.0530321598052979,
+      "learning_rate": 0.0001819968960165546,
+      "loss": 1.249,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1810708099583775,
+      "grad_norm": 1.1370590925216675,
+      "learning_rate": 0.00018198654940506986,
+      "loss": 1.5279,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1811742199012435,
+      "grad_norm": 1.3252718448638916,
+      "learning_rate": 0.00018197620279358513,
+      "loss": 1.8875,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1812776298441095,
+      "grad_norm": 1.1777899265289307,
+      "learning_rate": 0.0001819658561821004,
+      "loss": 1.4416,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1813810397869755,
+      "grad_norm": 1.79104745388031,
+      "learning_rate": 0.00018195550957061563,
+      "loss": 1.3529,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1814844497298415,
+      "grad_norm": 1.3128929138183594,
+      "learning_rate": 0.0001819451629591309,
+      "loss": 1.2693,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18158785967270752,
+      "grad_norm": 1.3561937808990479,
+      "learning_rate": 0.00018193481634764616,
+      "loss": 1.1964,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18169126961557353,
+      "grad_norm": 3.990429162979126,
+      "learning_rate": 0.00018192446973616143,
+      "loss": 1.7025,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18179467955843953,
+      "grad_norm": 2.079883575439453,
+      "learning_rate": 0.0001819141231246767,
+      "loss": 2.1624,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18189808950130554,
+      "grad_norm": 3.3075170516967773,
+      "learning_rate": 0.00018190377651319196,
+      "loss": 2.1052,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18200149944417154,
+      "grad_norm": 1.6594444513320923,
+      "learning_rate": 0.0001818934299017072,
+      "loss": 1.8497,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18210490938703755,
+      "grad_norm": 1.8769367933273315,
+      "learning_rate": 0.00018188308329022246,
+      "loss": 1.6261,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18220831932990356,
+      "grad_norm": 1.0080485343933105,
+      "learning_rate": 0.00018187273667873773,
+      "loss": 1.7013,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1823117292727696,
+      "grad_norm": 1.650251030921936,
+      "learning_rate": 0.000181862390067253,
+      "loss": 1.2848,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1824151392156356,
+      "grad_norm": 1.2735741138458252,
+      "learning_rate": 0.00018185204345576826,
+      "loss": 1.7479,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1825185491585016,
+      "grad_norm": 1.7841858863830566,
+      "learning_rate": 0.00018184169684428353,
+      "loss": 2.2005,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1826219591013676,
+      "grad_norm": 1.6187489032745361,
+      "learning_rate": 0.00018183135023279876,
+      "loss": 2.1096,
+      "step": 1766
+    },
+    {
+      "epoch": 0.18272536904423362,
+      "grad_norm": 1.2000303268432617,
+      "learning_rate": 0.00018182100362131403,
+      "loss": 1.3292,
+      "step": 1767
+    },
+    {
+      "epoch": 0.18282877898709962,
+      "grad_norm": 1.7545541524887085,
+      "learning_rate": 0.0001818106570098293,
+      "loss": 1.621,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18293218892996563,
+      "grad_norm": 1.455883502960205,
+      "learning_rate": 0.00018180031039834456,
+      "loss": 1.5392,
+      "step": 1769
+    },
+    {
+      "epoch": 0.18303559887283163,
+      "grad_norm": 1.1145455837249756,
+      "learning_rate": 0.00018178996378685983,
+      "loss": 1.4463,
+      "step": 1770
+    },
+    {
+      "epoch": 0.18313900881569764,
+      "grad_norm": 3.161123514175415,
+      "learning_rate": 0.0001817796171753751,
+      "loss": 1.4145,
+      "step": 1771
+    },
+    {
+      "epoch": 0.18324241875856365,
+      "grad_norm": 1.0690654516220093,
+      "learning_rate": 0.00018176927056389033,
+      "loss": 1.2464,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18334582870142965,
+      "grad_norm": 2.340733528137207,
+      "learning_rate": 0.0001817589239524056,
+      "loss": 1.3952,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18344923864429566,
+      "grad_norm": 1.527890920639038,
+      "learning_rate": 0.00018174857734092086,
+      "loss": 1.7334,
+      "step": 1774
+    },
+    {
+      "epoch": 0.18355264858716167,
+      "grad_norm": 2.002925157546997,
+      "learning_rate": 0.00018173823072943613,
+      "loss": 2.08,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18365605853002767,
+      "grad_norm": 1.2468239068984985,
+      "learning_rate": 0.0001817278841179514,
+      "loss": 1.8638,
+      "step": 1776
+    },
+    {
+      "epoch": 0.18375946847289368,
+      "grad_norm": 1.9202920198440552,
+      "learning_rate": 0.00018171753750646666,
+      "loss": 2.5098,
+      "step": 1777
+    },
+    {
+      "epoch": 0.18386287841575968,
+      "grad_norm": 1.4701899290084839,
+      "learning_rate": 0.0001817071908949819,
+      "loss": 1.783,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1839662883586257,
+      "grad_norm": 1.1845182180404663,
+      "learning_rate": 0.00018169684428349716,
+      "loss": 2.0948,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1840696983014917,
+      "grad_norm": 2.3378961086273193,
+      "learning_rate": 0.00018168649767201243,
+      "loss": 1.5978,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1841731082443577,
+      "grad_norm": 1.632500171661377,
+      "learning_rate": 0.0001816761510605277,
+      "loss": 1.7837,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1842765181872237,
+      "grad_norm": 1.8515764474868774,
+      "learning_rate": 0.00018166580444904296,
+      "loss": 2.0129,
+      "step": 1782
+    },
+    {
+      "epoch": 0.18437992813008972,
+      "grad_norm": 1.1611318588256836,
+      "learning_rate": 0.00018165545783755822,
+      "loss": 2.0427,
+      "step": 1783
+    },
+    {
+      "epoch": 0.18448333807295572,
+      "grad_norm": 0.8150402307510376,
+      "learning_rate": 0.00018164511122607346,
+      "loss": 1.9761,
+      "step": 1784
+    },
+    {
+      "epoch": 0.18458674801582173,
+      "grad_norm": 2.1012911796569824,
+      "learning_rate": 0.00018163476461458873,
+      "loss": 1.3257,
+      "step": 1785
+    },
+    {
+      "epoch": 0.18469015795868773,
+      "grad_norm": 1.231393575668335,
+      "learning_rate": 0.000181624418003104,
+      "loss": 1.8023,
+      "step": 1786
+    },
+    {
+      "epoch": 0.18479356790155374,
+      "grad_norm": 1.4408806562423706,
+      "learning_rate": 0.00018161407139161926,
+      "loss": 1.6601,
+      "step": 1787
+    },
+    {
+      "epoch": 0.18489697784441975,
+      "grad_norm": 1.223767876625061,
+      "learning_rate": 0.00018160372478013453,
+      "loss": 1.972,
+      "step": 1788
+    },
+    {
+      "epoch": 0.18500038778728575,
+      "grad_norm": 1.9878640174865723,
+      "learning_rate": 0.0001815933781686498,
+      "loss": 2.1535,
+      "step": 1789
+    },
+    {
+      "epoch": 0.18510379773015176,
+      "grad_norm": 1.076229214668274,
+      "learning_rate": 0.00018158303155716503,
+      "loss": 1.533,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18520720767301777,
+      "grad_norm": 2.2140488624572754,
+      "learning_rate": 0.0001815726849456803,
+      "loss": 1.9387,
+      "step": 1791
+    },
+    {
+      "epoch": 0.18531061761588377,
+      "grad_norm": 0.9972267150878906,
+      "learning_rate": 0.00018156233833419556,
+      "loss": 1.4952,
+      "step": 1792
+    },
+    {
+      "epoch": 0.18541402755874978,
+      "grad_norm": 2.2801876068115234,
+      "learning_rate": 0.00018155199172271083,
+      "loss": 1.9597,
+      "step": 1793
+    },
+    {
+      "epoch": 0.18551743750161578,
+      "grad_norm": 1.1939889192581177,
+      "learning_rate": 0.0001815416451112261,
+      "loss": 1.3651,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1856208474444818,
+      "grad_norm": 1.5104225873947144,
+      "learning_rate": 0.00018153129849974136,
+      "loss": 1.9471,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1857242573873478,
+      "grad_norm": 1.1902467012405396,
+      "learning_rate": 0.0001815209518882566,
+      "loss": 1.4968,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1858276673302138,
+      "grad_norm": 1.8906666040420532,
+      "learning_rate": 0.00018151060527677186,
+      "loss": 1.0471,
+      "step": 1797
+    },
+    {
+      "epoch": 0.1859310772730798,
+      "grad_norm": 1.378791093826294,
+      "learning_rate": 0.00018150025866528713,
+      "loss": 1.6083,
+      "step": 1798
+    },
+    {
+      "epoch": 0.18603448721594582,
+      "grad_norm": 1.4541735649108887,
+      "learning_rate": 0.0001814899120538024,
+      "loss": 2.0237,
+      "step": 1799
+    },
+    {
+      "epoch": 0.18613789715881182,
+      "grad_norm": 1.8095440864562988,
+      "learning_rate": 0.00018147956544231766,
+      "loss": 1.9967,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18624130710167783,
+      "grad_norm": 1.2104191780090332,
+      "learning_rate": 0.00018146921883083292,
+      "loss": 1.5496,
+      "step": 1801
+    },
+    {
+      "epoch": 0.18634471704454383,
+      "grad_norm": 2.2273075580596924,
+      "learning_rate": 0.00018145887221934816,
+      "loss": 2.3056,
+      "step": 1802
+    },
+    {
+      "epoch": 0.18644812698740984,
+      "grad_norm": 2.3855462074279785,
+      "learning_rate": 0.00018144852560786343,
+      "loss": 2.286,
+      "step": 1803
+    },
+    {
+      "epoch": 0.18655153693027585,
+      "grad_norm": 1.2031198740005493,
+      "learning_rate": 0.0001814381789963787,
+      "loss": 2.1857,
+      "step": 1804
+    },
+    {
+      "epoch": 0.18665494687314185,
+      "grad_norm": 0.9996086955070496,
+      "learning_rate": 0.00018142783238489396,
+      "loss": 2.0501,
+      "step": 1805
+    },
+    {
+      "epoch": 0.18675835681600786,
+      "grad_norm": 0.9292736053466797,
+      "learning_rate": 0.00018141748577340922,
+      "loss": 1.7131,
+      "step": 1806
+    },
+    {
+      "epoch": 0.18686176675887387,
+      "grad_norm": 1.5405597686767578,
+      "learning_rate": 0.0001814071391619245,
+      "loss": 2.1909,
+      "step": 1807
+    },
+    {
+      "epoch": 0.18696517670173987,
+      "grad_norm": 2.568983793258667,
+      "learning_rate": 0.00018139679255043973,
+      "loss": 1.8599,
+      "step": 1808
+    },
+    {
+      "epoch": 0.18706858664460588,
+      "grad_norm": 1.7339938879013062,
+      "learning_rate": 0.000181386445938955,
+      "loss": 1.1257,
+      "step": 1809
+    },
+    {
+      "epoch": 0.18717199658747188,
+      "grad_norm": 1.8022384643554688,
+      "learning_rate": 0.00018137609932747026,
+      "loss": 1.4831,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1872754065303379,
+      "grad_norm": 0.9867877960205078,
+      "learning_rate": 0.00018136575271598553,
+      "loss": 1.0952,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1873788164732039,
+      "grad_norm": 1.9051878452301025,
+      "learning_rate": 0.0001813554061045008,
+      "loss": 1.9558,
+      "step": 1812
+    },
+    {
+      "epoch": 0.1874822264160699,
+      "grad_norm": 1.5351530313491821,
+      "learning_rate": 0.00018134505949301606,
+      "loss": 1.4955,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1875856363589359,
+      "grad_norm": 1.1694519519805908,
+      "learning_rate": 0.0001813347128815313,
+      "loss": 1.533,
+      "step": 1814
+    },
+    {
+      "epoch": 0.18768904630180192,
+      "grad_norm": 0.9428067803382874,
+      "learning_rate": 0.00018132436627004656,
+      "loss": 1.5396,
+      "step": 1815
+    },
+    {
+      "epoch": 0.18779245624466792,
+      "grad_norm": 1.915429949760437,
+      "learning_rate": 0.00018131401965856183,
+      "loss": 1.621,
+      "step": 1816
+    },
+    {
+      "epoch": 0.18789586618753393,
+      "grad_norm": 2.1339144706726074,
+      "learning_rate": 0.0001813036730470771,
+      "loss": 2.2716,
+      "step": 1817
+    },
+    {
+      "epoch": 0.18799927613039993,
+      "grad_norm": 1.354615569114685,
+      "learning_rate": 0.00018129332643559236,
+      "loss": 1.3368,
+      "step": 1818
+    },
+    {
+      "epoch": 0.18810268607326594,
+      "grad_norm": 1.710594892501831,
+      "learning_rate": 0.00018128297982410762,
+      "loss": 1.6227,
+      "step": 1819
+    },
+    {
+      "epoch": 0.18820609601613195,
+      "grad_norm": 0.7719640135765076,
+      "learning_rate": 0.00018127263321262286,
+      "loss": 1.7036,
+      "step": 1820
+    },
+    {
+      "epoch": 0.18830950595899795,
+      "grad_norm": 1.9698446989059448,
+      "learning_rate": 0.00018126228660113813,
+      "loss": 1.8552,
+      "step": 1821
+    },
+    {
+      "epoch": 0.18841291590186396,
+      "grad_norm": 1.904042363166809,
+      "learning_rate": 0.0001812519399896534,
+      "loss": 1.9086,
+      "step": 1822
+    },
+    {
+      "epoch": 0.18851632584472997,
+      "grad_norm": 1.8710474967956543,
+      "learning_rate": 0.00018124159337816866,
+      "loss": 1.9492,
+      "step": 1823
+    },
+    {
+      "epoch": 0.18861973578759597,
+      "grad_norm": 1.190084457397461,
+      "learning_rate": 0.00018123124676668392,
+      "loss": 1.8335,
+      "step": 1824
+    },
+    {
+      "epoch": 0.18872314573046198,
+      "grad_norm": 4.177672386169434,
+      "learning_rate": 0.0001812209001551992,
+      "loss": 2.4607,
+      "step": 1825
+    },
+    {
+      "epoch": 0.18882655567332798,
+      "grad_norm": 3.2717580795288086,
+      "learning_rate": 0.00018121055354371443,
+      "loss": 1.9554,
+      "step": 1826
+    },
+    {
+      "epoch": 0.188929965616194,
+      "grad_norm": 1.0546072721481323,
+      "learning_rate": 0.0001812002069322297,
+      "loss": 1.3306,
+      "step": 1827
+    },
+    {
+      "epoch": 0.18903337555906,
+      "grad_norm": 1.9088751077651978,
+      "learning_rate": 0.00018118986032074496,
+      "loss": 2.0673,
+      "step": 1828
+    },
+    {
+      "epoch": 0.189136785501926,
+      "grad_norm": 1.713178277015686,
+      "learning_rate": 0.00018117951370926022,
+      "loss": 1.864,
+      "step": 1829
+    },
+    {
+      "epoch": 0.189240195444792,
+      "grad_norm": 1.981321930885315,
+      "learning_rate": 0.0001811691670977755,
+      "loss": 1.7863,
+      "step": 1830
+    },
+    {
+      "epoch": 0.18934360538765801,
+      "grad_norm": 1.5771139860153198,
+      "learning_rate": 0.00018115882048629076,
+      "loss": 2.094,
+      "step": 1831
+    },
+    {
+      "epoch": 0.18944701533052402,
+      "grad_norm": 1.2328580617904663,
+      "learning_rate": 0.000181148473874806,
+      "loss": 1.9157,
+      "step": 1832
+    },
+    {
+      "epoch": 0.18955042527339003,
+      "grad_norm": 1.098201870918274,
+      "learning_rate": 0.00018113812726332126,
+      "loss": 1.3232,
+      "step": 1833
+    },
+    {
+      "epoch": 0.18965383521625603,
+      "grad_norm": 1.778517246246338,
+      "learning_rate": 0.00018112778065183652,
+      "loss": 1.9743,
+      "step": 1834
+    },
+    {
+      "epoch": 0.18975724515912204,
+      "grad_norm": 1.304283618927002,
+      "learning_rate": 0.0001811174340403518,
+      "loss": 1.9143,
+      "step": 1835
+    },
+    {
+      "epoch": 0.18986065510198805,
+      "grad_norm": 1.3429310321807861,
+      "learning_rate": 0.00018110708742886706,
+      "loss": 1.7206,
+      "step": 1836
+    },
+    {
+      "epoch": 0.18996406504485405,
+      "grad_norm": 1.1537504196166992,
+      "learning_rate": 0.00018109674081738232,
+      "loss": 1.7866,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19006747498772006,
+      "grad_norm": 1.7021723985671997,
+      "learning_rate": 0.00018108639420589756,
+      "loss": 1.9366,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19017088493058606,
+      "grad_norm": 1.0984216928482056,
+      "learning_rate": 0.00018107604759441283,
+      "loss": 1.344,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19027429487345207,
+      "grad_norm": 1.413723111152649,
+      "learning_rate": 0.0001810657009829281,
+      "loss": 1.7328,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19037770481631808,
+      "grad_norm": 1.7750047445297241,
+      "learning_rate": 0.00018105535437144336,
+      "loss": 1.584,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19048111475918408,
+      "grad_norm": 1.1813859939575195,
+      "learning_rate": 0.00018104500775995862,
+      "loss": 2.1179,
+      "step": 1842
+    },
+    {
+      "epoch": 0.1905845247020501,
+      "grad_norm": 1.3422218561172485,
+      "learning_rate": 0.0001810346611484739,
+      "loss": 1.5151,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1906879346449161,
+      "grad_norm": 1.4470585584640503,
+      "learning_rate": 0.00018102431453698913,
+      "loss": 2.2249,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1907913445877821,
+      "grad_norm": 1.856718897819519,
+      "learning_rate": 0.0001810139679255044,
+      "loss": 1.36,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1908947545306481,
+      "grad_norm": 1.5910100936889648,
+      "learning_rate": 0.00018100362131401966,
+      "loss": 1.3646,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19099816447351411,
+      "grad_norm": 1.5017138719558716,
+      "learning_rate": 0.00018099327470253492,
+      "loss": 1.3302,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19110157441638012,
+      "grad_norm": 1.379167079925537,
+      "learning_rate": 0.0001809829280910502,
+      "loss": 1.4517,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19120498435924616,
+      "grad_norm": 2.3777546882629395,
+      "learning_rate": 0.00018097258147956545,
+      "loss": 1.7635,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19130839430211216,
+      "grad_norm": 1.7774989604949951,
+      "learning_rate": 0.0001809622348680807,
+      "loss": 1.8686,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19141180424497817,
+      "grad_norm": 1.2162162065505981,
+      "learning_rate": 0.00018095188825659596,
+      "loss": 1.3562,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19151521418784417,
+      "grad_norm": 1.2358673810958862,
+      "learning_rate": 0.00018094154164511122,
+      "loss": 1.38,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19161862413071018,
+      "grad_norm": 1.128541350364685,
+      "learning_rate": 0.0001809311950336265,
+      "loss": 1.9068,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1917220340735762,
+      "grad_norm": 1.5324431657791138,
+      "learning_rate": 0.00018092084842214175,
+      "loss": 1.5886,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1918254440164422,
+      "grad_norm": 1.675620436668396,
+      "learning_rate": 0.00018091050181065702,
+      "loss": 1.8084,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1919288539593082,
+      "grad_norm": 2.1495304107666016,
+      "learning_rate": 0.00018090015519917226,
+      "loss": 1.8716,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1920322639021742,
+      "grad_norm": 2.3184926509857178,
+      "learning_rate": 0.00018088980858768752,
+      "loss": 1.6612,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1921356738450402,
+      "grad_norm": 1.2949163913726807,
+      "learning_rate": 0.0001808794619762028,
+      "loss": 1.3865,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19223908378790622,
+      "grad_norm": 1.916689157485962,
+      "learning_rate": 0.00018086911536471806,
+      "loss": 1.3509,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19234249373077222,
+      "grad_norm": 1.8200879096984863,
+      "learning_rate": 0.00018085876875323332,
+      "loss": 1.8744,
+      "step": 1860
+    },
+    {
+      "epoch": 0.19244590367363823,
+      "grad_norm": 1.343815565109253,
+      "learning_rate": 0.0001808484221417486,
+      "loss": 1.5256,
+      "step": 1861
+    },
+    {
+      "epoch": 0.19254931361650424,
+      "grad_norm": 2.418281078338623,
+      "learning_rate": 0.00018083807553026385,
+      "loss": 1.5507,
+      "step": 1862
+    },
+    {
+      "epoch": 0.19265272355937024,
+      "grad_norm": 2.052889585494995,
+      "learning_rate": 0.00018082772891877912,
+      "loss": 1.2509,
+      "step": 1863
+    },
+    {
+      "epoch": 0.19275613350223625,
+      "grad_norm": 2.2111313343048096,
+      "learning_rate": 0.00018081738230729436,
+      "loss": 1.5071,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19285954344510225,
+      "grad_norm": 1.393479347229004,
+      "learning_rate": 0.00018080703569580962,
+      "loss": 1.3836,
+      "step": 1865
+    },
+    {
+      "epoch": 0.19296295338796826,
+      "grad_norm": 1.9624559879302979,
+      "learning_rate": 0.0001807966890843249,
+      "loss": 1.4573,
+      "step": 1866
+    },
+    {
+      "epoch": 0.19306636333083427,
+      "grad_norm": 1.0306349992752075,
+      "learning_rate": 0.00018078634247284015,
+      "loss": 1.514,
+      "step": 1867
+    },
+    {
+      "epoch": 0.19316977327370027,
+      "grad_norm": 1.0501223802566528,
+      "learning_rate": 0.00018077599586135542,
+      "loss": 1.702,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19327318321656628,
+      "grad_norm": 1.1913334131240845,
+      "learning_rate": 0.00018076564924987068,
+      "loss": 1.5761,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1933765931594323,
+      "grad_norm": 2.000929832458496,
+      "learning_rate": 0.00018075530263838595,
+      "loss": 2.0399,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1934800031022983,
+      "grad_norm": 1.3432594537734985,
+      "learning_rate": 0.00018074495602690122,
+      "loss": 0.6957,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1935834130451643,
+      "grad_norm": 1.617430329322815,
+      "learning_rate": 0.00018073460941541645,
+      "loss": 1.5554,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1936868229880303,
+      "grad_norm": 1.2280590534210205,
+      "learning_rate": 0.00018072426280393172,
+      "loss": 1.4669,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1937902329308963,
+      "grad_norm": 2.365945816040039,
+      "learning_rate": 0.00018071391619244699,
+      "loss": 1.5908,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19389364287376232,
+      "grad_norm": 2.2847042083740234,
+      "learning_rate": 0.00018070356958096225,
+      "loss": 1.329,
+      "step": 1875
+    },
+    {
+      "epoch": 0.19399705281662832,
+      "grad_norm": 1.594077229499817,
+      "learning_rate": 0.00018069322296947752,
+      "loss": 2.0627,
+      "step": 1876
+    },
+    {
+      "epoch": 0.19410046275949433,
+      "grad_norm": 1.4474533796310425,
+      "learning_rate": 0.00018068287635799278,
+      "loss": 1.9983,
+      "step": 1877
+    },
+    {
+      "epoch": 0.19420387270236034,
+      "grad_norm": 1.8211748600006104,
+      "learning_rate": 0.00018067252974650805,
+      "loss": 1.3836,
+      "step": 1878
+    },
+    {
+      "epoch": 0.19430728264522634,
+      "grad_norm": 1.5337395668029785,
+      "learning_rate": 0.0001806621831350233,
+      "loss": 1.8984,
+      "step": 1879
+    },
+    {
+      "epoch": 0.19441069258809235,
+      "grad_norm": 1.8015854358673096,
+      "learning_rate": 0.00018065183652353855,
+      "loss": 1.5139,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19451410253095835,
+      "grad_norm": 1.4770269393920898,
+      "learning_rate": 0.00018064148991205382,
+      "loss": 1.801,
+      "step": 1881
+    },
+    {
+      "epoch": 0.19461751247382436,
+      "grad_norm": 2.826849937438965,
+      "learning_rate": 0.00018063114330056908,
+      "loss": 1.9032,
+      "step": 1882
+    },
+    {
+      "epoch": 0.19472092241669037,
+      "grad_norm": 1.6297709941864014,
+      "learning_rate": 0.00018062079668908435,
+      "loss": 1.89,
+      "step": 1883
+    },
+    {
+      "epoch": 0.19482433235955637,
+      "grad_norm": 1.8109357357025146,
+      "learning_rate": 0.00018061045007759961,
+      "loss": 1.4969,
+      "step": 1884
+    },
+    {
+      "epoch": 0.19492774230242238,
+      "grad_norm": 1.7631341218948364,
+      "learning_rate": 0.00018060010346611485,
+      "loss": 2.0972,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19503115224528839,
+      "grad_norm": 1.25094735622406,
+      "learning_rate": 0.00018058975685463012,
+      "loss": 1.1867,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1951345621881544,
+      "grad_norm": 2.0920863151550293,
+      "learning_rate": 0.00018057941024314538,
+      "loss": 2.0022,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1952379721310204,
+      "grad_norm": 1.4094507694244385,
+      "learning_rate": 0.00018056906363166065,
+      "loss": 1.168,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1953413820738864,
+      "grad_norm": 1.4328402280807495,
+      "learning_rate": 0.00018055871702017591,
+      "loss": 1.7634,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1954447920167524,
+      "grad_norm": 1.9557263851165771,
+      "learning_rate": 0.00018054837040869118,
+      "loss": 1.9204,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19554820195961842,
+      "grad_norm": 1.4277377128601074,
+      "learning_rate": 0.00018053802379720642,
+      "loss": 1.5392,
+      "step": 1891
+    },
+    {
+      "epoch": 0.19565161190248442,
+      "grad_norm": 2.212749481201172,
+      "learning_rate": 0.00018052767718572168,
+      "loss": 1.9089,
+      "step": 1892
+    },
+    {
+      "epoch": 0.19575502184535043,
+      "grad_norm": 1.0590177774429321,
+      "learning_rate": 0.00018051733057423695,
+      "loss": 1.4042,
+      "step": 1893
+    },
+    {
+      "epoch": 0.19585843178821644,
+      "grad_norm": 1.6970980167388916,
+      "learning_rate": 0.00018050698396275222,
+      "loss": 1.638,
+      "step": 1894
+    },
+    {
+      "epoch": 0.19596184173108244,
+      "grad_norm": 1.8155356645584106,
+      "learning_rate": 0.00018049663735126748,
+      "loss": 1.5066,
+      "step": 1895
+    },
+    {
+      "epoch": 0.19606525167394845,
+      "grad_norm": 2.029043674468994,
+      "learning_rate": 0.00018048629073978275,
+      "loss": 2.2207,
+      "step": 1896
+    },
+    {
+      "epoch": 0.19616866161681445,
+      "grad_norm": 1.168373465538025,
+      "learning_rate": 0.00018047594412829798,
+      "loss": 1.3578,
+      "step": 1897
+    },
+    {
+      "epoch": 0.19627207155968046,
+      "grad_norm": 2.446021318435669,
+      "learning_rate": 0.00018046559751681325,
+      "loss": 1.6108,
+      "step": 1898
+    },
+    {
+      "epoch": 0.19637548150254647,
+      "grad_norm": 1.6077264547348022,
+      "learning_rate": 0.00018045525090532852,
+      "loss": 1.5486,
+      "step": 1899
+    },
+    {
+      "epoch": 0.19647889144541247,
+      "grad_norm": 1.242552638053894,
+      "learning_rate": 0.00018044490429384378,
+      "loss": 1.7297,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19658230138827848,
+      "grad_norm": 1.8686028718948364,
+      "learning_rate": 0.00018043455768235905,
+      "loss": 1.758,
+      "step": 1901
+    },
+    {
+      "epoch": 0.19668571133114449,
+      "grad_norm": 1.9251303672790527,
+      "learning_rate": 0.0001804242110708743,
+      "loss": 1.4774,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1967891212740105,
+      "grad_norm": 4.7904133796691895,
+      "learning_rate": 0.00018041386445938955,
+      "loss": 1.8559,
+      "step": 1903
+    },
+    {
+      "epoch": 0.1968925312168765,
+      "grad_norm": 0.9137360453605652,
+      "learning_rate": 0.00018040351784790482,
+      "loss": 1.5535,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1969959411597425,
+      "grad_norm": 1.5711393356323242,
+      "learning_rate": 0.00018039317123642008,
+      "loss": 1.03,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1970993511026085,
+      "grad_norm": 1.640053153038025,
+      "learning_rate": 0.00018038282462493535,
+      "loss": 1.836,
+      "step": 1906
+    },
+    {
+      "epoch": 0.19720276104547452,
+      "grad_norm": 0.7787024974822998,
+      "learning_rate": 0.0001803724780134506,
+      "loss": 1.9089,
+      "step": 1907
+    },
+    {
+      "epoch": 0.19730617098834052,
+      "grad_norm": 2.0896191596984863,
+      "learning_rate": 0.00018036213140196588,
+      "loss": 1.9978,
+      "step": 1908
+    },
+    {
+      "epoch": 0.19740958093120653,
+      "grad_norm": 2.179332971572876,
+      "learning_rate": 0.00018035178479048112,
+      "loss": 1.9776,
+      "step": 1909
+    },
+    {
+      "epoch": 0.19751299087407254,
+      "grad_norm": 1.585008978843689,
+      "learning_rate": 0.00018034143817899638,
+      "loss": 1.1899,
+      "step": 1910
+    },
+    {
+      "epoch": 0.19761640081693854,
+      "grad_norm": 1.303787350654602,
+      "learning_rate": 0.00018033109156751165,
+      "loss": 1.7906,
+      "step": 1911
+    },
+    {
+      "epoch": 0.19771981075980455,
+      "grad_norm": 1.9878695011138916,
+      "learning_rate": 0.00018032074495602691,
+      "loss": 1.8106,
+      "step": 1912
+    },
+    {
+      "epoch": 0.19782322070267055,
+      "grad_norm": 1.501775860786438,
+      "learning_rate": 0.00018031039834454218,
+      "loss": 1.4732,
+      "step": 1913
+    },
+    {
+      "epoch": 0.19792663064553656,
+      "grad_norm": 0.9017894864082336,
+      "learning_rate": 0.00018030005173305745,
+      "loss": 1.4282,
+      "step": 1914
+    },
+    {
+      "epoch": 0.19803004058840257,
+      "grad_norm": 2.1266887187957764,
+      "learning_rate": 0.00018028970512157268,
+      "loss": 2.2244,
+      "step": 1915
+    },
+    {
+      "epoch": 0.19813345053126857,
+      "grad_norm": 1.604373812675476,
+      "learning_rate": 0.00018027935851008795,
+      "loss": 1.6581,
+      "step": 1916
+    },
+    {
+      "epoch": 0.19823686047413458,
+      "grad_norm": 1.8775768280029297,
+      "learning_rate": 0.00018026901189860321,
+      "loss": 1.5546,
+      "step": 1917
+    },
+    {
+      "epoch": 0.19834027041700059,
+      "grad_norm": 1.487886905670166,
+      "learning_rate": 0.00018025866528711848,
+      "loss": 1.5156,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1984436803598666,
+      "grad_norm": 3.0149381160736084,
+      "learning_rate": 0.00018024831867563375,
+      "loss": 1.7735,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1985470903027326,
+      "grad_norm": 1.967848300933838,
+      "learning_rate": 0.000180237972064149,
+      "loss": 1.4067,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1986505002455986,
+      "grad_norm": 1.7828800678253174,
+      "learning_rate": 0.00018022762545266425,
+      "loss": 1.3988,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1987539101884646,
+      "grad_norm": 1.206902265548706,
+      "learning_rate": 0.00018021727884117952,
+      "loss": 1.9073,
+      "step": 1922
+    },
+    {
+      "epoch": 0.19885732013133062,
+      "grad_norm": 1.424898624420166,
+      "learning_rate": 0.00018020693222969478,
+      "loss": 1.6547,
+      "step": 1923
+    },
+    {
+      "epoch": 0.19896073007419662,
+      "grad_norm": 1.5171235799789429,
+      "learning_rate": 0.00018019658561821005,
+      "loss": 1.2257,
+      "step": 1924
+    },
+    {
+      "epoch": 0.19906414001706263,
+      "grad_norm": 2.3045217990875244,
+      "learning_rate": 0.0001801862390067253,
+      "loss": 1.5398,
+      "step": 1925
+    },
+    {
+      "epoch": 0.19916754995992864,
+      "grad_norm": 1.2358486652374268,
+      "learning_rate": 0.00018017589239524058,
+      "loss": 1.2342,
+      "step": 1926
+    },
+    {
+      "epoch": 0.19927095990279464,
+      "grad_norm": 3.0135512351989746,
+      "learning_rate": 0.00018016554578375582,
+      "loss": 2.1923,
+      "step": 1927
+    },
+    {
+      "epoch": 0.19937436984566065,
+      "grad_norm": 1.4758455753326416,
+      "learning_rate": 0.00018015519917227108,
+      "loss": 1.5922,
+      "step": 1928
+    },
+    {
+      "epoch": 0.19947777978852665,
+      "grad_norm": 1.570439100265503,
+      "learning_rate": 0.00018014485256078635,
+      "loss": 1.8605,
+      "step": 1929
+    },
+    {
+      "epoch": 0.19958118973139266,
+      "grad_norm": 1.461016058921814,
+      "learning_rate": 0.0001801345059493016,
+      "loss": 1.5143,
+      "step": 1930
+    },
+    {
+      "epoch": 0.19968459967425867,
+      "grad_norm": 0.7076051831245422,
+      "learning_rate": 0.00018012415933781688,
+      "loss": 1.5049,
+      "step": 1931
+    },
+    {
+      "epoch": 0.19978800961712467,
+      "grad_norm": 2.8588709831237793,
+      "learning_rate": 0.00018011381272633214,
+      "loss": 1.7324,
+      "step": 1932
+    },
+    {
+      "epoch": 0.19989141955999068,
+      "grad_norm": 0.7718111276626587,
+      "learning_rate": 0.00018010346611484738,
+      "loss": 1.6924,
+      "step": 1933
+    },
+    {
+      "epoch": 0.19999482950285669,
+      "grad_norm": 1.3416244983673096,
+      "learning_rate": 0.00018009311950336265,
+      "loss": 1.7367,
+      "step": 1934
+    },
+    {
+      "epoch": 0.20009823944572272,
+      "grad_norm": 1.6038373708724976,
+      "learning_rate": 0.00018008277289187791,
+      "loss": 1.4978,
+      "step": 1935
+    },
+    {
+      "epoch": 0.20020164938858873,
+      "grad_norm": 0.9637541770935059,
+      "learning_rate": 0.00018007242628039318,
+      "loss": 1.5899,
+      "step": 1936
+    },
+    {
+      "epoch": 0.20030505933145473,
+      "grad_norm": 2.398042678833008,
+      "learning_rate": 0.00018006207966890845,
+      "loss": 1.684,
+      "step": 1937
+    },
+    {
+      "epoch": 0.20040846927432074,
+      "grad_norm": 0.8711498379707336,
+      "learning_rate": 0.0001800517330574237,
+      "loss": 1.6272,
+      "step": 1938
+    },
+    {
+      "epoch": 0.20051187921718674,
+      "grad_norm": 1.1462258100509644,
+      "learning_rate": 0.00018004138644593895,
+      "loss": 1.5288,
+      "step": 1939
+    },
+    {
+      "epoch": 0.20061528916005275,
+      "grad_norm": 0.9858958125114441,
+      "learning_rate": 0.00018003103983445421,
+      "loss": 1.3255,
+      "step": 1940
+    },
+    {
+      "epoch": 0.20071869910291876,
+      "grad_norm": 2.7075014114379883,
+      "learning_rate": 0.00018002069322296948,
+      "loss": 1.7802,
+      "step": 1941
+    },
+    {
+      "epoch": 0.20082210904578476,
+      "grad_norm": 1.764085054397583,
+      "learning_rate": 0.00018001034661148475,
+      "loss": 1.7139,
+      "step": 1942
+    },
+    {
+      "epoch": 0.20092551898865077,
+      "grad_norm": 2.7296926975250244,
+      "learning_rate": 0.00018,
+      "loss": 1.9334,
+      "step": 1943
+    },
+    {
+      "epoch": 0.20102892893151678,
+      "grad_norm": 2.1738665103912354,
+      "learning_rate": 0.00017998965338851528,
+      "loss": 1.5643,
+      "step": 1944
+    },
+    {
+      "epoch": 0.20113233887438278,
+      "grad_norm": 1.6310564279556274,
+      "learning_rate": 0.00017997930677703052,
+      "loss": 1.8784,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2012357488172488,
+      "grad_norm": 2.148175001144409,
+      "learning_rate": 0.00017996896016554578,
+      "loss": 2.0443,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2013391587601148,
+      "grad_norm": 1.1791040897369385,
+      "learning_rate": 0.00017995861355406105,
+      "loss": 0.9135,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2014425687029808,
+      "grad_norm": 0.9954102635383606,
+      "learning_rate": 0.0001799482669425763,
+      "loss": 1.2455,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2015459786458468,
+      "grad_norm": 2.1388514041900635,
+      "learning_rate": 0.00017993792033109158,
+      "loss": 1.6639,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2016493885887128,
+      "grad_norm": 1.2391835451126099,
+      "learning_rate": 0.00017992757371960684,
+      "loss": 1.5007,
+      "step": 1950
+    },
+    {
+      "epoch": 0.20175279853157882,
+      "grad_norm": 0.8275509476661682,
+      "learning_rate": 0.00017991722710812208,
+      "loss": 1.9589,
+      "step": 1951
+    },
+    {
+      "epoch": 0.20185620847444483,
+      "grad_norm": 1.2318191528320312,
+      "learning_rate": 0.00017990688049663735,
+      "loss": 1.4067,
+      "step": 1952
+    },
+    {
+      "epoch": 0.20195961841731083,
+      "grad_norm": 1.2213613986968994,
+      "learning_rate": 0.0001798965338851526,
+      "loss": 2.0376,
+      "step": 1953
+    },
+    {
+      "epoch": 0.20206302836017684,
+      "grad_norm": 1.7849206924438477,
+      "learning_rate": 0.00017988618727366788,
+      "loss": 2.1899,
+      "step": 1954
+    },
+    {
+      "epoch": 0.20216643830304284,
+      "grad_norm": 1.1603035926818848,
+      "learning_rate": 0.00017987584066218314,
+      "loss": 1.5716,
+      "step": 1955
+    },
+    {
+      "epoch": 0.20226984824590885,
+      "grad_norm": 1.1252952814102173,
+      "learning_rate": 0.0001798654940506984,
+      "loss": 1.9331,
+      "step": 1956
+    },
+    {
+      "epoch": 0.20237325818877486,
+      "grad_norm": 1.050102710723877,
+      "learning_rate": 0.00017985514743921365,
+      "loss": 2.0236,
+      "step": 1957
+    },
+    {
+      "epoch": 0.20247666813164086,
+      "grad_norm": 1.4389870166778564,
+      "learning_rate": 0.00017984480082772891,
+      "loss": 1.4933,
+      "step": 1958
+    },
+    {
+      "epoch": 0.20258007807450687,
+      "grad_norm": 1.7062374353408813,
+      "learning_rate": 0.00017983445421624418,
+      "loss": 1.843,
+      "step": 1959
+    },
+    {
+      "epoch": 0.20268348801737288,
+      "grad_norm": 1.5522654056549072,
+      "learning_rate": 0.00017982410760475944,
+      "loss": 1.8029,
+      "step": 1960
+    },
+    {
+      "epoch": 0.20278689796023888,
+      "grad_norm": 0.9937103390693665,
+      "learning_rate": 0.0001798137609932747,
+      "loss": 1.3794,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2028903079031049,
+      "grad_norm": 2.6092488765716553,
+      "learning_rate": 0.00017980341438178998,
+      "loss": 1.7977,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2029937178459709,
+      "grad_norm": 1.2122944593429565,
+      "learning_rate": 0.00017979306777030521,
+      "loss": 1.532,
+      "step": 1963
+    },
+    {
+      "epoch": 0.2030971277888369,
+      "grad_norm": 2.4709153175354004,
+      "learning_rate": 0.00017978272115882048,
+      "loss": 1.8433,
+      "step": 1964
+    },
+    {
+      "epoch": 0.2032005377317029,
+      "grad_norm": 1.01822829246521,
+      "learning_rate": 0.00017977237454733575,
+      "loss": 1.6264,
+      "step": 1965
+    },
+    {
+      "epoch": 0.2033039476745689,
+      "grad_norm": 1.2235360145568848,
+      "learning_rate": 0.000179762027935851,
+      "loss": 1.6075,
+      "step": 1966
+    },
+    {
+      "epoch": 0.20340735761743492,
+      "grad_norm": 1.8367936611175537,
+      "learning_rate": 0.00017975168132436628,
+      "loss": 1.5176,
+      "step": 1967
+    },
+    {
+      "epoch": 0.20351076756030093,
+      "grad_norm": 0.938755989074707,
+      "learning_rate": 0.00017974133471288154,
+      "loss": 1.5801,
+      "step": 1968
+    },
+    {
+      "epoch": 0.20361417750316693,
+      "grad_norm": 1.2774001359939575,
+      "learning_rate": 0.00017973098810139678,
+      "loss": 2.0368,
+      "step": 1969
+    },
+    {
+      "epoch": 0.20371758744603294,
+      "grad_norm": 1.1897248029708862,
+      "learning_rate": 0.00017972064148991205,
+      "loss": 1.599,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20382099738889894,
+      "grad_norm": 2.377195119857788,
+      "learning_rate": 0.0001797102948784273,
+      "loss": 1.7758,
+      "step": 1971
+    },
+    {
+      "epoch": 0.20392440733176495,
+      "grad_norm": 0.9643241763114929,
+      "learning_rate": 0.00017969994826694258,
+      "loss": 1.4347,
+      "step": 1972
+    },
+    {
+      "epoch": 0.20402781727463096,
+      "grad_norm": 1.207589030265808,
+      "learning_rate": 0.00017968960165545784,
+      "loss": 1.8052,
+      "step": 1973
+    },
+    {
+      "epoch": 0.20413122721749696,
+      "grad_norm": 1.32759690284729,
+      "learning_rate": 0.0001796792550439731,
+      "loss": 1.6897,
+      "step": 1974
+    },
+    {
+      "epoch": 0.20423463716036297,
+      "grad_norm": 2.175830364227295,
+      "learning_rate": 0.00017966890843248835,
+      "loss": 1.6727,
+      "step": 1975
+    },
+    {
+      "epoch": 0.20433804710322898,
+      "grad_norm": 1.3569698333740234,
+      "learning_rate": 0.0001796585618210036,
+      "loss": 1.6497,
+      "step": 1976
+    },
+    {
+      "epoch": 0.20444145704609498,
+      "grad_norm": 1.3463072776794434,
+      "learning_rate": 0.00017964821520951888,
+      "loss": 1.4355,
+      "step": 1977
+    },
+    {
+      "epoch": 0.204544866988961,
+      "grad_norm": 1.4692943096160889,
+      "learning_rate": 0.00017963786859803414,
+      "loss": 1.7329,
+      "step": 1978
+    },
+    {
+      "epoch": 0.204648276931827,
+      "grad_norm": 2.3857581615448,
+      "learning_rate": 0.0001796275219865494,
+      "loss": 1.7336,
+      "step": 1979
+    },
+    {
+      "epoch": 0.204751686874693,
+      "grad_norm": 1.526208758354187,
+      "learning_rate": 0.00017961717537506467,
+      "loss": 1.3837,
+      "step": 1980
+    },
+    {
+      "epoch": 0.204855096817559,
+      "grad_norm": 3.072439193725586,
+      "learning_rate": 0.0001796068287635799,
+      "loss": 2.5553,
+      "step": 1981
+    },
+    {
+      "epoch": 0.204958506760425,
+      "grad_norm": 1.1713007688522339,
+      "learning_rate": 0.00017959648215209518,
+      "loss": 1.4069,
+      "step": 1982
+    },
+    {
+      "epoch": 0.20506191670329102,
+      "grad_norm": 1.4028162956237793,
+      "learning_rate": 0.00017958613554061044,
+      "loss": 1.4357,
+      "step": 1983
+    },
+    {
+      "epoch": 0.20516532664615703,
+      "grad_norm": 1.3035467863082886,
+      "learning_rate": 0.0001795757889291257,
+      "loss": 1.8673,
+      "step": 1984
+    },
+    {
+      "epoch": 0.20526873658902303,
+      "grad_norm": 1.277909755706787,
+      "learning_rate": 0.00017956544231764098,
+      "loss": 1.7968,
+      "step": 1985
+    },
+    {
+      "epoch": 0.20537214653188904,
+      "grad_norm": 2.3015072345733643,
+      "learning_rate": 0.00017955509570615624,
+      "loss": 1.6155,
+      "step": 1986
+    },
+    {
+      "epoch": 0.20547555647475504,
+      "grad_norm": 1.4320391416549683,
+      "learning_rate": 0.0001795447490946715,
+      "loss": 1.2945,
+      "step": 1987
+    },
+    {
+      "epoch": 0.20557896641762105,
+      "grad_norm": 1.892540454864502,
+      "learning_rate": 0.00017953440248318677,
+      "loss": 1.854,
+      "step": 1988
+    },
+    {
+      "epoch": 0.20568237636048706,
+      "grad_norm": 1.6273618936538696,
+      "learning_rate": 0.000179524055871702,
+      "loss": 1.8503,
+      "step": 1989
+    },
+    {
+      "epoch": 0.20578578630335306,
+      "grad_norm": 1.4557535648345947,
+      "learning_rate": 0.00017951370926021728,
+      "loss": 1.5845,
+      "step": 1990
+    },
+    {
+      "epoch": 0.20588919624621907,
+      "grad_norm": 1.505989909172058,
+      "learning_rate": 0.00017950336264873254,
+      "loss": 1.2067,
+      "step": 1991
+    },
+    {
+      "epoch": 0.20599260618908508,
+      "grad_norm": 0.9464778900146484,
+      "learning_rate": 0.0001794930160372478,
+      "loss": 1.1262,
+      "step": 1992
+    },
+    {
+      "epoch": 0.20609601613195108,
+      "grad_norm": 1.0865939855575562,
+      "learning_rate": 0.00017948266942576307,
+      "loss": 1.8715,
+      "step": 1993
+    },
+    {
+      "epoch": 0.2061994260748171,
+      "grad_norm": 2.5264039039611816,
+      "learning_rate": 0.00017947232281427834,
+      "loss": 1.3308,
+      "step": 1994
+    },
+    {
+      "epoch": 0.2063028360176831,
+      "grad_norm": 2.007601261138916,
+      "learning_rate": 0.0001794619762027936,
+      "loss": 1.3582,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2064062459605491,
+      "grad_norm": 1.6429976224899292,
+      "learning_rate": 0.00017945162959130887,
+      "loss": 1.9568,
+      "step": 1996
+    },
+    {
+      "epoch": 0.2065096559034151,
+      "grad_norm": 1.2347596883773804,
+      "learning_rate": 0.0001794412829798241,
+      "loss": 0.9477,
+      "step": 1997
+    },
+    {
+      "epoch": 0.2066130658462811,
+      "grad_norm": 1.348817229270935,
+      "learning_rate": 0.00017943093636833937,
+      "loss": 1.5172,
+      "step": 1998
+    },
+    {
+      "epoch": 0.20671647578914712,
+      "grad_norm": 1.0015040636062622,
+      "learning_rate": 0.00017942058975685464,
+      "loss": 1.5398,
+      "step": 1999
+    },
+    {
+      "epoch": 0.20681988573201313,
+      "grad_norm": 1.2908525466918945,
+      "learning_rate": 0.0001794102431453699,
+      "loss": 1.6978,
+      "step": 2000
+    },
+    {
+      "epoch": 0.20692329567487913,
+      "grad_norm": 1.1875195503234863,
+      "learning_rate": 0.00017939989653388517,
+      "loss": 1.533,
+      "step": 2001
+    },
+    {
+      "epoch": 0.20702670561774514,
+      "grad_norm": 2.341791868209839,
+      "learning_rate": 0.00017938954992240044,
+      "loss": 1.9741,
+      "step": 2002
+    },
+    {
+      "epoch": 0.20713011556061114,
+      "grad_norm": 0.8359478116035461,
+      "learning_rate": 0.0001793792033109157,
+      "loss": 1.3734,
+      "step": 2003
+    },
+    {
+      "epoch": 0.20723352550347715,
+      "grad_norm": 0.9825917482376099,
+      "learning_rate": 0.00017936885669943097,
+      "loss": 1.5892,
+      "step": 2004
+    },
+    {
+      "epoch": 0.20733693544634316,
+      "grad_norm": 2.1322312355041504,
+      "learning_rate": 0.0001793585100879462,
+      "loss": 2.1437,
+      "step": 2005
+    },
+    {
+      "epoch": 0.20744034538920916,
+      "grad_norm": 2.004436731338501,
+      "learning_rate": 0.00017934816347646147,
+      "loss": 2.1666,
+      "step": 2006
+    },
+    {
+      "epoch": 0.20754375533207517,
+      "grad_norm": 2.5417864322662354,
+      "learning_rate": 0.00017933781686497674,
+      "loss": 1.3547,
+      "step": 2007
+    },
+    {
+      "epoch": 0.20764716527494118,
+      "grad_norm": 1.3734756708145142,
+      "learning_rate": 0.000179327470253492,
+      "loss": 2.0243,
+      "step": 2008
+    },
+    {
+      "epoch": 0.20775057521780718,
+      "grad_norm": 2.404106378555298,
+      "learning_rate": 0.00017931712364200727,
+      "loss": 1.8643,
+      "step": 2009
+    },
+    {
+      "epoch": 0.2078539851606732,
+      "grad_norm": 1.595017910003662,
+      "learning_rate": 0.00017930677703052253,
+      "loss": 1.7677,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2079573951035392,
+      "grad_norm": 2.3860955238342285,
+      "learning_rate": 0.00017929643041903777,
+      "loss": 2.1223,
+      "step": 2011
+    },
+    {
+      "epoch": 0.2080608050464052,
+      "grad_norm": 1.7865846157073975,
+      "learning_rate": 0.00017928608380755304,
+      "loss": 1.5736,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2081642149892712,
+      "grad_norm": 1.4908995628356934,
+      "learning_rate": 0.0001792757371960683,
+      "loss": 0.9999,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2082676249321372,
+      "grad_norm": 1.2065404653549194,
+      "learning_rate": 0.00017926539058458357,
+      "loss": 1.3708,
+      "step": 2014
+    },
+    {
+      "epoch": 0.20837103487500322,
+      "grad_norm": 1.7153881788253784,
+      "learning_rate": 0.00017925504397309883,
+      "loss": 1.2165,
+      "step": 2015
+    },
+    {
+      "epoch": 0.20847444481786923,
+      "grad_norm": 2.4799675941467285,
+      "learning_rate": 0.0001792446973616141,
+      "loss": 2.138,
+      "step": 2016
+    },
+    {
+      "epoch": 0.20857785476073523,
+      "grad_norm": 2.0555872917175293,
+      "learning_rate": 0.00017923435075012934,
+      "loss": 1.4837,
+      "step": 2017
+    },
+    {
+      "epoch": 0.20868126470360124,
+      "grad_norm": 1.3154330253601074,
+      "learning_rate": 0.0001792240041386446,
+      "loss": 1.3167,
+      "step": 2018
+    },
+    {
+      "epoch": 0.20878467464646724,
+      "grad_norm": 1.519303798675537,
+      "learning_rate": 0.00017921365752715987,
+      "loss": 1.4666,
+      "step": 2019
+    },
+    {
+      "epoch": 0.20888808458933325,
+      "grad_norm": 1.393334984779358,
+      "learning_rate": 0.00017920331091567514,
+      "loss": 1.8231,
+      "step": 2020
+    },
+    {
+      "epoch": 0.20899149453219928,
+      "grad_norm": 2.9100759029388428,
+      "learning_rate": 0.0001791929643041904,
+      "loss": 2.0966,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2090949044750653,
+      "grad_norm": 1.2485828399658203,
+      "learning_rate": 0.00017918261769270567,
+      "loss": 2.3454,
+      "step": 2022
+    },
+    {
+      "epoch": 0.2091983144179313,
+      "grad_norm": 1.3368488550186157,
+      "learning_rate": 0.0001791722710812209,
+      "loss": 1.5258,
+      "step": 2023
+    },
+    {
+      "epoch": 0.2093017243607973,
+      "grad_norm": 1.3732163906097412,
+      "learning_rate": 0.00017916192446973617,
+      "loss": 1.4365,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2094051343036633,
+      "grad_norm": 2.5866189002990723,
+      "learning_rate": 0.00017915157785825144,
+      "loss": 2.0169,
+      "step": 2025
+    },
+    {
+      "epoch": 0.20950854424652932,
+      "grad_norm": 1.5932780504226685,
+      "learning_rate": 0.0001791412312467667,
+      "loss": 2.3346,
+      "step": 2026
+    },
+    {
+      "epoch": 0.20961195418939532,
+      "grad_norm": 1.9270954132080078,
+      "learning_rate": 0.00017913088463528197,
+      "loss": 1.405,
+      "step": 2027
+    },
+    {
+      "epoch": 0.20971536413226133,
+      "grad_norm": 1.3831976652145386,
+      "learning_rate": 0.00017912053802379723,
+      "loss": 1.543,
+      "step": 2028
+    },
+    {
+      "epoch": 0.20981877407512733,
+      "grad_norm": 2.5507442951202393,
+      "learning_rate": 0.00017911019141231247,
+      "loss": 1.992,
+      "step": 2029
+    },
+    {
+      "epoch": 0.20992218401799334,
+      "grad_norm": 2.1970303058624268,
+      "learning_rate": 0.00017909984480082774,
+      "loss": 1.2845,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21002559396085935,
+      "grad_norm": 2.447885751724243,
+      "learning_rate": 0.000179089498189343,
+      "loss": 2.5404,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21012900390372535,
+      "grad_norm": 2.6955151557922363,
+      "learning_rate": 0.00017907915157785827,
+      "loss": 2.1718,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21023241384659136,
+      "grad_norm": 1.2139887809753418,
+      "learning_rate": 0.00017906880496637353,
+      "loss": 1.5857,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21033582378945737,
+      "grad_norm": 3.4483728408813477,
+      "learning_rate": 0.0001790584583548888,
+      "loss": 1.2517,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21043923373232337,
+      "grad_norm": 1.2948105335235596,
+      "learning_rate": 0.00017904811174340404,
+      "loss": 2.0194,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21054264367518938,
+      "grad_norm": 1.39495050907135,
+      "learning_rate": 0.0001790377651319193,
+      "loss": 1.2252,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21064605361805538,
+      "grad_norm": 1.9834434986114502,
+      "learning_rate": 0.00017902741852043457,
+      "loss": 1.8306,
+      "step": 2037
+    },
+    {
+      "epoch": 0.2107494635609214,
+      "grad_norm": 1.5110279321670532,
+      "learning_rate": 0.00017901707190894983,
+      "loss": 1.5356,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2108528735037874,
+      "grad_norm": 1.726402997970581,
+      "learning_rate": 0.0001790067252974651,
+      "loss": 1.5534,
+      "step": 2039
+    },
+    {
+      "epoch": 0.2109562834466534,
+      "grad_norm": 1.5266774892807007,
+      "learning_rate": 0.00017899637868598037,
+      "loss": 1.9585,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2110596933895194,
+      "grad_norm": 2.4746997356414795,
+      "learning_rate": 0.0001789860320744956,
+      "loss": 1.8594,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21116310333238542,
+      "grad_norm": 1.94597589969635,
+      "learning_rate": 0.00017897568546301087,
+      "loss": 1.9834,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21126651327525142,
+      "grad_norm": 1.2754744291305542,
+      "learning_rate": 0.00017896533885152613,
+      "loss": 1.2801,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21136992321811743,
+      "grad_norm": 1.6256976127624512,
+      "learning_rate": 0.0001789549922400414,
+      "loss": 1.6023,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21147333316098343,
+      "grad_norm": 1.609365463256836,
+      "learning_rate": 0.00017894464562855667,
+      "loss": 2.1025,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21157674310384944,
+      "grad_norm": 1.7030891180038452,
+      "learning_rate": 0.00017893429901707193,
+      "loss": 1.7133,
+      "step": 2046
+    },
+    {
+      "epoch": 0.21168015304671545,
+      "grad_norm": 1.521166443824768,
+      "learning_rate": 0.00017892395240558717,
+      "loss": 1.7595,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21178356298958145,
+      "grad_norm": 2.1347200870513916,
+      "learning_rate": 0.00017891360579410244,
+      "loss": 1.8925,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21188697293244746,
+      "grad_norm": 1.0615814924240112,
+      "learning_rate": 0.0001789032591826177,
+      "loss": 1.2334,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21199038287531347,
+      "grad_norm": 1.3307974338531494,
+      "learning_rate": 0.00017889291257113297,
+      "loss": 1.3501,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21209379281817947,
+      "grad_norm": 2.896918773651123,
+      "learning_rate": 0.00017888256595964823,
+      "loss": 1.9585,
+      "step": 2051
+    },
+    {
+      "epoch": 0.21219720276104548,
+      "grad_norm": 1.7504158020019531,
+      "learning_rate": 0.0001788722193481635,
+      "loss": 1.6612,
+      "step": 2052
+    },
+    {
+      "epoch": 0.21230061270391148,
+      "grad_norm": 1.9615195989608765,
+      "learning_rate": 0.00017886187273667874,
+      "loss": 1.6103,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2124040226467775,
+      "grad_norm": 1.6140780448913574,
+      "learning_rate": 0.000178851526125194,
+      "loss": 1.6658,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2125074325896435,
+      "grad_norm": 2.134024143218994,
+      "learning_rate": 0.00017884117951370927,
+      "loss": 2.0824,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2126108425325095,
+      "grad_norm": 1.1607773303985596,
+      "learning_rate": 0.00017883083290222453,
+      "loss": 1.5691,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2127142524753755,
+      "grad_norm": 1.261562466621399,
+      "learning_rate": 0.0001788204862907398,
+      "loss": 1.8007,
+      "step": 2057
+    },
+    {
+      "epoch": 0.21281766241824152,
+      "grad_norm": 2.1017324924468994,
+      "learning_rate": 0.00017881013967925506,
+      "loss": 1.6821,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21292107236110752,
+      "grad_norm": 1.7821030616760254,
+      "learning_rate": 0.0001787997930677703,
+      "loss": 1.5448,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21302448230397353,
+      "grad_norm": 0.8674912452697754,
+      "learning_rate": 0.00017878944645628557,
+      "loss": 2.0453,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21312789224683953,
+      "grad_norm": 1.2638546228408813,
+      "learning_rate": 0.00017877909984480083,
+      "loss": 1.4075,
+      "step": 2061
+    },
+    {
+      "epoch": 0.21323130218970554,
+      "grad_norm": 1.0837128162384033,
+      "learning_rate": 0.0001787687532333161,
+      "loss": 2.0548,
+      "step": 2062
+    },
+    {
+      "epoch": 0.21333471213257155,
+      "grad_norm": 3.9489150047302246,
+      "learning_rate": 0.00017875840662183136,
+      "loss": 2.059,
+      "step": 2063
+    },
+    {
+      "epoch": 0.21343812207543755,
+      "grad_norm": 3.7381279468536377,
+      "learning_rate": 0.00017874806001034663,
+      "loss": 2.097,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21354153201830356,
+      "grad_norm": 1.6482309103012085,
+      "learning_rate": 0.00017873771339886187,
+      "loss": 1.8298,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21364494196116957,
+      "grad_norm": 1.2123875617980957,
+      "learning_rate": 0.00017872736678737713,
+      "loss": 1.4125,
+      "step": 2066
+    },
+    {
+      "epoch": 0.21374835190403557,
+      "grad_norm": 2.2748680114746094,
+      "learning_rate": 0.0001787170201758924,
+      "loss": 1.8227,
+      "step": 2067
+    },
+    {
+      "epoch": 0.21385176184690158,
+      "grad_norm": 1.461757779121399,
+      "learning_rate": 0.00017870667356440767,
+      "loss": 1.7162,
+      "step": 2068
+    },
+    {
+      "epoch": 0.21395517178976758,
+      "grad_norm": 1.454840898513794,
+      "learning_rate": 0.00017869632695292293,
+      "loss": 1.9135,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2140585817326336,
+      "grad_norm": 3.0303759574890137,
+      "learning_rate": 0.0001786859803414382,
+      "loss": 1.9702,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2141619916754996,
+      "grad_norm": 1.0980781316757202,
+      "learning_rate": 0.00017867563372995344,
+      "loss": 1.3306,
+      "step": 2071
+    },
+    {
+      "epoch": 0.2142654016183656,
+      "grad_norm": 1.069142460823059,
+      "learning_rate": 0.0001786652871184687,
+      "loss": 0.7896,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2143688115612316,
+      "grad_norm": 1.9988906383514404,
+      "learning_rate": 0.00017865494050698397,
+      "loss": 2.5048,
+      "step": 2073
+    },
+    {
+      "epoch": 0.21447222150409762,
+      "grad_norm": 1.0396260023117065,
+      "learning_rate": 0.00017864459389549923,
+      "loss": 1.173,
+      "step": 2074
+    },
+    {
+      "epoch": 0.21457563144696362,
+      "grad_norm": 1.9060814380645752,
+      "learning_rate": 0.0001786342472840145,
+      "loss": 2.4794,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21467904138982963,
+      "grad_norm": 2.0201363563537598,
+      "learning_rate": 0.00017862390067252976,
+      "loss": 1.7278,
+      "step": 2076
+    },
+    {
+      "epoch": 0.21478245133269563,
+      "grad_norm": 1.4735513925552368,
+      "learning_rate": 0.000178613554061045,
+      "loss": 1.7763,
+      "step": 2077
+    },
+    {
+      "epoch": 0.21488586127556164,
+      "grad_norm": 1.1059455871582031,
+      "learning_rate": 0.00017860320744956027,
+      "loss": 1.5732,
+      "step": 2078
+    },
+    {
+      "epoch": 0.21498927121842765,
+      "grad_norm": 1.4942630529403687,
+      "learning_rate": 0.00017859286083807553,
+      "loss": 1.3362,
+      "step": 2079
+    },
+    {
+      "epoch": 0.21509268116129365,
+      "grad_norm": 1.017965316772461,
+      "learning_rate": 0.0001785825142265908,
+      "loss": 1.3988,
+      "step": 2080
+    },
+    {
+      "epoch": 0.21519609110415966,
+      "grad_norm": 1.3145697116851807,
+      "learning_rate": 0.00017857216761510606,
+      "loss": 1.2764,
+      "step": 2081
+    },
+    {
+      "epoch": 0.21529950104702567,
+      "grad_norm": 1.3202226161956787,
+      "learning_rate": 0.00017856182100362133,
+      "loss": 1.9009,
+      "step": 2082
+    },
+    {
+      "epoch": 0.21540291098989167,
+      "grad_norm": 2.01487398147583,
+      "learning_rate": 0.00017855147439213657,
+      "loss": 1.3367,
+      "step": 2083
+    },
+    {
+      "epoch": 0.21550632093275768,
+      "grad_norm": 0.839141845703125,
+      "learning_rate": 0.00017854112778065183,
+      "loss": 2.0631,
+      "step": 2084
+    },
+    {
+      "epoch": 0.21560973087562368,
+      "grad_norm": 1.0034396648406982,
+      "learning_rate": 0.0001785307811691671,
+      "loss": 1.7007,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2157131408184897,
+      "grad_norm": 1.548252820968628,
+      "learning_rate": 0.00017852043455768236,
+      "loss": 1.1115,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2158165507613557,
+      "grad_norm": 1.6088694334030151,
+      "learning_rate": 0.00017851008794619763,
+      "loss": 1.7482,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2159199607042217,
+      "grad_norm": 1.0032658576965332,
+      "learning_rate": 0.0001784997413347129,
+      "loss": 1.952,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2160233706470877,
+      "grad_norm": 1.0490684509277344,
+      "learning_rate": 0.00017848939472322813,
+      "loss": 2.0505,
+      "step": 2089
+    },
+    {
+      "epoch": 0.21612678058995372,
+      "grad_norm": 1.1849490404129028,
+      "learning_rate": 0.0001784790481117434,
+      "loss": 1.6461,
+      "step": 2090
+    },
+    {
+      "epoch": 0.21623019053281972,
+      "grad_norm": 1.3508158922195435,
+      "learning_rate": 0.00017846870150025867,
+      "loss": 0.9686,
+      "step": 2091
+    },
+    {
+      "epoch": 0.21633360047568573,
+      "grad_norm": 2.8345625400543213,
+      "learning_rate": 0.00017845835488877393,
+      "loss": 1.6479,
+      "step": 2092
+    },
+    {
+      "epoch": 0.21643701041855173,
+      "grad_norm": 1.3022091388702393,
+      "learning_rate": 0.0001784480082772892,
+      "loss": 1.609,
+      "step": 2093
+    },
+    {
+      "epoch": 0.21654042036141774,
+      "grad_norm": 1.6842771768569946,
+      "learning_rate": 0.00017843766166580446,
+      "loss": 1.1587,
+      "step": 2094
+    },
+    {
+      "epoch": 0.21664383030428375,
+      "grad_norm": 1.806605577468872,
+      "learning_rate": 0.0001784273150543197,
+      "loss": 1.8456,
+      "step": 2095
+    },
+    {
+      "epoch": 0.21674724024714975,
+      "grad_norm": 1.597494125366211,
+      "learning_rate": 0.00017841696844283497,
+      "loss": 1.3116,
+      "step": 2096
+    },
+    {
+      "epoch": 0.21685065019001576,
+      "grad_norm": 1.737967848777771,
+      "learning_rate": 0.00017840662183135023,
+      "loss": 1.2401,
+      "step": 2097
+    },
+    {
+      "epoch": 0.21695406013288177,
+      "grad_norm": 0.8340157270431519,
+      "learning_rate": 0.0001783962752198655,
+      "loss": 1.4509,
+      "step": 2098
+    },
+    {
+      "epoch": 0.21705747007574777,
+      "grad_norm": 1.295029640197754,
+      "learning_rate": 0.00017838592860838076,
+      "loss": 1.3269,
+      "step": 2099
+    },
+    {
+      "epoch": 0.21716088001861378,
+      "grad_norm": 1.8555551767349243,
+      "learning_rate": 0.00017837558199689603,
+      "loss": 1.8658,
+      "step": 2100
+    },
+    {
+      "epoch": 0.21726428996147978,
+      "grad_norm": 1.1895952224731445,
+      "learning_rate": 0.00017836523538541127,
+      "loss": 1.4318,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2173676999043458,
+      "grad_norm": 1.3722859621047974,
+      "learning_rate": 0.00017835488877392653,
+      "loss": 1.5428,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2174711098472118,
+      "grad_norm": 0.909134030342102,
+      "learning_rate": 0.0001783445421624418,
+      "loss": 1.7826,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2175745197900778,
+      "grad_norm": 3.7859716415405273,
+      "learning_rate": 0.00017833419555095706,
+      "loss": 1.5052,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2176779297329438,
+      "grad_norm": 2.5243780612945557,
+      "learning_rate": 0.00017832384893947233,
+      "loss": 2.3734,
+      "step": 2105
+    },
+    {
+      "epoch": 0.21778133967580984,
+      "grad_norm": 1.4078128337860107,
+      "learning_rate": 0.0001783135023279876,
+      "loss": 2.0086,
+      "step": 2106
+    },
+    {
+      "epoch": 0.21788474961867585,
+      "grad_norm": 1.2615593671798706,
+      "learning_rate": 0.00017830315571650283,
+      "loss": 1.6874,
+      "step": 2107
+    },
+    {
+      "epoch": 0.21798815956154186,
+      "grad_norm": 2.134047031402588,
+      "learning_rate": 0.0001782928091050181,
+      "loss": 1.9679,
+      "step": 2108
+    },
+    {
+      "epoch": 0.21809156950440786,
+      "grad_norm": 1.8929492235183716,
+      "learning_rate": 0.00017828246249353336,
+      "loss": 1.5872,
+      "step": 2109
+    },
+    {
+      "epoch": 0.21819497944727387,
+      "grad_norm": 1.081035852432251,
+      "learning_rate": 0.00017827211588204863,
+      "loss": 1.2067,
+      "step": 2110
+    },
+    {
+      "epoch": 0.21829838939013987,
+      "grad_norm": 1.5089231729507446,
+      "learning_rate": 0.0001782617692705639,
+      "loss": 1.6677,
+      "step": 2111
+    },
+    {
+      "epoch": 0.21840179933300588,
+      "grad_norm": 1.6253288984298706,
+      "learning_rate": 0.00017825142265907916,
+      "loss": 1.5697,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2185052092758719,
+      "grad_norm": 1.9758878946304321,
+      "learning_rate": 0.0001782410760475944,
+      "loss": 1.7322,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2186086192187379,
+      "grad_norm": 3.3685495853424072,
+      "learning_rate": 0.00017823072943610967,
+      "loss": 2.1612,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2187120291616039,
+      "grad_norm": 0.7514423131942749,
+      "learning_rate": 0.00017822038282462493,
+      "loss": 1.8737,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2188154391044699,
+      "grad_norm": 1.4034844636917114,
+      "learning_rate": 0.0001782100362131402,
+      "loss": 1.5761,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2189188490473359,
+      "grad_norm": 2.4100217819213867,
+      "learning_rate": 0.00017819968960165546,
+      "loss": 1.6346,
+      "step": 2117
+    },
+    {
+      "epoch": 0.21902225899020192,
+      "grad_norm": 0.9915169477462769,
+      "learning_rate": 0.00017818934299017073,
+      "loss": 1.3795,
+      "step": 2118
+    },
+    {
+      "epoch": 0.21912566893306792,
+      "grad_norm": 1.9723891019821167,
+      "learning_rate": 0.000178178996378686,
+      "loss": 1.4034,
+      "step": 2119
+    },
+    {
+      "epoch": 0.21922907887593393,
+      "grad_norm": 1.6976430416107178,
+      "learning_rate": 0.00017816864976720126,
+      "loss": 1.9929,
+      "step": 2120
+    },
+    {
+      "epoch": 0.21933248881879994,
+      "grad_norm": 2.5415894985198975,
+      "learning_rate": 0.0001781583031557165,
+      "loss": 1.4351,
+      "step": 2121
+    },
+    {
+      "epoch": 0.21943589876166594,
+      "grad_norm": 2.1857051849365234,
+      "learning_rate": 0.00017814795654423176,
+      "loss": 1.3509,
+      "step": 2122
+    },
+    {
+      "epoch": 0.21953930870453195,
+      "grad_norm": 1.309126853942871,
+      "learning_rate": 0.00017813760993274703,
+      "loss": 2.0322,
+      "step": 2123
+    },
+    {
+      "epoch": 0.21964271864739796,
+      "grad_norm": 1.8605514764785767,
+      "learning_rate": 0.0001781272633212623,
+      "loss": 1.831,
+      "step": 2124
+    },
+    {
+      "epoch": 0.21974612859026396,
+      "grad_norm": 1.8593494892120361,
+      "learning_rate": 0.00017811691670977756,
+      "loss": 1.3086,
+      "step": 2125
+    },
+    {
+      "epoch": 0.21984953853312997,
+      "grad_norm": 1.4725764989852905,
+      "learning_rate": 0.00017810657009829282,
+      "loss": 2.0356,
+      "step": 2126
+    },
+    {
+      "epoch": 0.21995294847599597,
+      "grad_norm": 1.6571112871170044,
+      "learning_rate": 0.0001780962234868081,
+      "loss": 1.4547,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22005635841886198,
+      "grad_norm": 1.592392086982727,
+      "learning_rate": 0.00017808587687532336,
+      "loss": 1.8363,
+      "step": 2128
+    },
+    {
+      "epoch": 0.220159768361728,
+      "grad_norm": 1.4015467166900635,
+      "learning_rate": 0.0001780755302638386,
+      "loss": 1.6763,
+      "step": 2129
+    },
+    {
+      "epoch": 0.220263178304594,
+      "grad_norm": 1.4670350551605225,
+      "learning_rate": 0.00017806518365235386,
+      "loss": 1.8672,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22036658824746,
+      "grad_norm": 1.1072840690612793,
+      "learning_rate": 0.00017805483704086913,
+      "loss": 1.4605,
+      "step": 2131
+    },
+    {
+      "epoch": 0.220469998190326,
+      "grad_norm": 2.009456157684326,
+      "learning_rate": 0.0001780444904293844,
+      "loss": 2.1299,
+      "step": 2132
+    },
+    {
+      "epoch": 0.220573408133192,
+      "grad_norm": 2.815999746322632,
+      "learning_rate": 0.00017803414381789966,
+      "loss": 1.8397,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22067681807605802,
+      "grad_norm": 2.340900182723999,
+      "learning_rate": 0.00017802379720641492,
+      "loss": 1.9299,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22078022801892402,
+      "grad_norm": 1.3540948629379272,
+      "learning_rate": 0.0001780134505949302,
+      "loss": 1.8427,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22088363796179003,
+      "grad_norm": 1.0881321430206299,
+      "learning_rate": 0.00017800310398344545,
+      "loss": 1.4275,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22098704790465604,
+      "grad_norm": 1.0537045001983643,
+      "learning_rate": 0.0001779927573719607,
+      "loss": 1.662,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22109045784752204,
+      "grad_norm": 2.121901273727417,
+      "learning_rate": 0.00017798241076047596,
+      "loss": 1.5767,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22119386779038805,
+      "grad_norm": 1.318829894065857,
+      "learning_rate": 0.00017797206414899122,
+      "loss": 1.9675,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22129727773325406,
+      "grad_norm": 2.5058493614196777,
+      "learning_rate": 0.0001779617175375065,
+      "loss": 1.6812,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22140068767612006,
+      "grad_norm": 1.9833593368530273,
+      "learning_rate": 0.00017795137092602175,
+      "loss": 1.672,
+      "step": 2141
+    },
+    {
+      "epoch": 0.22150409761898607,
+      "grad_norm": 1.5669039487838745,
+      "learning_rate": 0.00017794102431453702,
+      "loss": 1.4147,
+      "step": 2142
+    },
+    {
+      "epoch": 0.22160750756185207,
+      "grad_norm": 1.6391706466674805,
+      "learning_rate": 0.00017793067770305226,
+      "loss": 0.935,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22171091750471808,
+      "grad_norm": 2.4117825031280518,
+      "learning_rate": 0.00017792033109156752,
+      "loss": 1.5555,
+      "step": 2144
+    },
+    {
+      "epoch": 0.2218143274475841,
+      "grad_norm": 1.7879774570465088,
+      "learning_rate": 0.0001779099844800828,
+      "loss": 1.7854,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2219177373904501,
+      "grad_norm": 1.3162338733673096,
+      "learning_rate": 0.00017789963786859806,
+      "loss": 1.9177,
+      "step": 2146
+    },
+    {
+      "epoch": 0.2220211473333161,
+      "grad_norm": 2.763244152069092,
+      "learning_rate": 0.00017788929125711332,
+      "loss": 2.4389,
+      "step": 2147
+    },
+    {
+      "epoch": 0.2221245572761821,
+      "grad_norm": 1.8798487186431885,
+      "learning_rate": 0.00017787894464562859,
+      "loss": 1.8481,
+      "step": 2148
+    },
+    {
+      "epoch": 0.2222279672190481,
+      "grad_norm": 1.089530348777771,
+      "learning_rate": 0.00017786859803414382,
+      "loss": 1.4647,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22233137716191412,
+      "grad_norm": 1.5578677654266357,
+      "learning_rate": 0.0001778582514226591,
+      "loss": 2.0841,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22243478710478012,
+      "grad_norm": 1.403699278831482,
+      "learning_rate": 0.00017784790481117436,
+      "loss": 2.2018,
+      "step": 2151
+    },
+    {
+      "epoch": 0.22253819704764613,
+      "grad_norm": 1.9366167783737183,
+      "learning_rate": 0.00017783755819968962,
+      "loss": 1.8533,
+      "step": 2152
+    },
+    {
+      "epoch": 0.22264160699051214,
+      "grad_norm": 1.0486398935317993,
+      "learning_rate": 0.0001778272115882049,
+      "loss": 2.0046,
+      "step": 2153
+    },
+    {
+      "epoch": 0.22274501693337814,
+      "grad_norm": 2.3693816661834717,
+      "learning_rate": 0.00017781686497672015,
+      "loss": 1.5874,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22284842687624415,
+      "grad_norm": 1.4884207248687744,
+      "learning_rate": 0.0001778065183652354,
+      "loss": 1.4696,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22295183681911016,
+      "grad_norm": 1.5521442890167236,
+      "learning_rate": 0.00017779617175375066,
+      "loss": 1.6104,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22305524676197616,
+      "grad_norm": 2.393778085708618,
+      "learning_rate": 0.00017778582514226592,
+      "loss": 2.2676,
+      "step": 2157
+    },
+    {
+      "epoch": 0.22315865670484217,
+      "grad_norm": 1.3362990617752075,
+      "learning_rate": 0.0001777754785307812,
+      "loss": 1.5341,
+      "step": 2158
+    },
+    {
+      "epoch": 0.22326206664770817,
+      "grad_norm": 1.3599947690963745,
+      "learning_rate": 0.00017776513191929645,
+      "loss": 1.6264,
+      "step": 2159
+    },
+    {
+      "epoch": 0.22336547659057418,
+      "grad_norm": 1.904009222984314,
+      "learning_rate": 0.00017775478530781172,
+      "loss": 1.5065,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2234688865334402,
+      "grad_norm": 1.6462085247039795,
+      "learning_rate": 0.00017774443869632696,
+      "loss": 1.5861,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2235722964763062,
+      "grad_norm": 1.7757351398468018,
+      "learning_rate": 0.00017773409208484222,
+      "loss": 2.2307,
+      "step": 2162
+    },
+    {
+      "epoch": 0.2236757064191722,
+      "grad_norm": 1.2958214282989502,
+      "learning_rate": 0.0001777237454733575,
+      "loss": 1.764,
+      "step": 2163
+    },
+    {
+      "epoch": 0.2237791163620382,
+      "grad_norm": 1.368507981300354,
+      "learning_rate": 0.00017771339886187275,
+      "loss": 2.3469,
+      "step": 2164
+    },
+    {
+      "epoch": 0.2238825263049042,
+      "grad_norm": 1.3048007488250732,
+      "learning_rate": 0.00017770305225038802,
+      "loss": 1.8811,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22398593624777022,
+      "grad_norm": 1.4118998050689697,
+      "learning_rate": 0.00017769270563890329,
+      "loss": 1.2182,
+      "step": 2166
+    },
+    {
+      "epoch": 0.22408934619063622,
+      "grad_norm": 2.8922696113586426,
+      "learning_rate": 0.00017768235902741852,
+      "loss": 1.9509,
+      "step": 2167
+    },
+    {
+      "epoch": 0.22419275613350223,
+      "grad_norm": 1.4795236587524414,
+      "learning_rate": 0.0001776720124159338,
+      "loss": 1.8401,
+      "step": 2168
+    },
+    {
+      "epoch": 0.22429616607636824,
+      "grad_norm": 2.1654906272888184,
+      "learning_rate": 0.00017766166580444905,
+      "loss": 1.8052,
+      "step": 2169
+    },
+    {
+      "epoch": 0.22439957601923424,
+      "grad_norm": 1.3528178930282593,
+      "learning_rate": 0.00017765131919296432,
+      "loss": 1.5025,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22450298596210025,
+      "grad_norm": 1.5375784635543823,
+      "learning_rate": 0.00017764097258147959,
+      "loss": 2.1606,
+      "step": 2171
+    },
+    {
+      "epoch": 0.22460639590496626,
+      "grad_norm": 1.4318784475326538,
+      "learning_rate": 0.00017763062596999485,
+      "loss": 1.6414,
+      "step": 2172
+    },
+    {
+      "epoch": 0.22470980584783226,
+      "grad_norm": 2.0012366771698,
+      "learning_rate": 0.0001776202793585101,
+      "loss": 1.8316,
+      "step": 2173
+    },
+    {
+      "epoch": 0.22481321579069827,
+      "grad_norm": 1.7899354696273804,
+      "learning_rate": 0.00017760993274702536,
+      "loss": 1.3864,
+      "step": 2174
+    },
+    {
+      "epoch": 0.22491662573356427,
+      "grad_norm": 1.3995294570922852,
+      "learning_rate": 0.00017759958613554062,
+      "loss": 1.7188,
+      "step": 2175
+    },
+    {
+      "epoch": 0.22502003567643028,
+      "grad_norm": 2.419296979904175,
+      "learning_rate": 0.0001775892395240559,
+      "loss": 1.9595,
+      "step": 2176
+    },
+    {
+      "epoch": 0.2251234456192963,
+      "grad_norm": 2.1976547241210938,
+      "learning_rate": 0.00017757889291257115,
+      "loss": 1.7631,
+      "step": 2177
+    },
+    {
+      "epoch": 0.2252268555621623,
+      "grad_norm": 1.25783371925354,
+      "learning_rate": 0.00017756854630108642,
+      "loss": 1.0977,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2253302655050283,
+      "grad_norm": 2.3799781799316406,
+      "learning_rate": 0.00017755819968960166,
+      "loss": 1.6176,
+      "step": 2179
+    },
+    {
+      "epoch": 0.2254336754478943,
+      "grad_norm": 1.4703243970870972,
+      "learning_rate": 0.00017754785307811692,
+      "loss": 1.1459,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2255370853907603,
+      "grad_norm": 1.6233093738555908,
+      "learning_rate": 0.0001775375064666322,
+      "loss": 1.6363,
+      "step": 2181
+    },
+    {
+      "epoch": 0.22564049533362632,
+      "grad_norm": 1.394158959388733,
+      "learning_rate": 0.00017752715985514745,
+      "loss": 1.4521,
+      "step": 2182
+    },
+    {
+      "epoch": 0.22574390527649232,
+      "grad_norm": 2.117434501647949,
+      "learning_rate": 0.00017751681324366272,
+      "loss": 1.7419,
+      "step": 2183
+    },
+    {
+      "epoch": 0.22584731521935833,
+      "grad_norm": 1.5713629722595215,
+      "learning_rate": 0.00017750646663217798,
+      "loss": 1.6596,
+      "step": 2184
+    },
+    {
+      "epoch": 0.22595072516222434,
+      "grad_norm": 1.4215084314346313,
+      "learning_rate": 0.00017749612002069322,
+      "loss": 1.6273,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22605413510509034,
+      "grad_norm": 2.092386484146118,
+      "learning_rate": 0.0001774857734092085,
+      "loss": 1.6127,
+      "step": 2186
+    },
+    {
+      "epoch": 0.22615754504795635,
+      "grad_norm": 1.4384335279464722,
+      "learning_rate": 0.00017747542679772375,
+      "loss": 2.0603,
+      "step": 2187
+    },
+    {
+      "epoch": 0.22626095499082235,
+      "grad_norm": 1.191169261932373,
+      "learning_rate": 0.00017746508018623902,
+      "loss": 0.8099,
+      "step": 2188
+    },
+    {
+      "epoch": 0.22636436493368836,
+      "grad_norm": 1.1086655855178833,
+      "learning_rate": 0.00017745473357475428,
+      "loss": 1.9708,
+      "step": 2189
+    },
+    {
+      "epoch": 0.22646777487655437,
+      "grad_norm": 1.1252591609954834,
+      "learning_rate": 0.00017744438696326955,
+      "loss": 1.4266,
+      "step": 2190
+    },
+    {
+      "epoch": 0.22657118481942037,
+      "grad_norm": 2.127028226852417,
+      "learning_rate": 0.0001774340403517848,
+      "loss": 1.4415,
+      "step": 2191
+    },
+    {
+      "epoch": 0.2266745947622864,
+      "grad_norm": 1.0664880275726318,
+      "learning_rate": 0.00017742369374030005,
+      "loss": 1.7211,
+      "step": 2192
+    },
+    {
+      "epoch": 0.22677800470515241,
+      "grad_norm": 1.391688585281372,
+      "learning_rate": 0.00017741334712881532,
+      "loss": 2.242,
+      "step": 2193
+    },
+    {
+      "epoch": 0.22688141464801842,
+      "grad_norm": 1.4060182571411133,
+      "learning_rate": 0.00017740300051733059,
+      "loss": 1.7529,
+      "step": 2194
+    },
+    {
+      "epoch": 0.22698482459088443,
+      "grad_norm": 1.3024177551269531,
+      "learning_rate": 0.00017739265390584585,
+      "loss": 1.2435,
+      "step": 2195
+    },
+    {
+      "epoch": 0.22708823453375043,
+      "grad_norm": 0.9886972308158875,
+      "learning_rate": 0.0001773823072943611,
+      "loss": 1.177,
+      "step": 2196
+    },
+    {
+      "epoch": 0.22719164447661644,
+      "grad_norm": 1.9852739572525024,
+      "learning_rate": 0.00017737196068287636,
+      "loss": 1.4258,
+      "step": 2197
+    },
+    {
+      "epoch": 0.22729505441948245,
+      "grad_norm": 1.706716537475586,
+      "learning_rate": 0.00017736161407139162,
+      "loss": 1.7284,
+      "step": 2198
+    },
+    {
+      "epoch": 0.22739846436234845,
+      "grad_norm": 2.0738205909729004,
+      "learning_rate": 0.00017735126745990689,
+      "loss": 2.6703,
+      "step": 2199
+    },
+    {
+      "epoch": 0.22750187430521446,
+      "grad_norm": 2.022379159927368,
+      "learning_rate": 0.00017734092084842215,
+      "loss": 1.1353,
+      "step": 2200
+    },
+    {
+      "epoch": 0.22760528424808046,
+      "grad_norm": 2.1549413204193115,
+      "learning_rate": 0.00017733057423693742,
+      "loss": 1.8087,
+      "step": 2201
+    },
+    {
+      "epoch": 0.22770869419094647,
+      "grad_norm": 1.0443583726882935,
+      "learning_rate": 0.00017732022762545266,
+      "loss": 1.1172,
+      "step": 2202
+    },
+    {
+      "epoch": 0.22781210413381248,
+      "grad_norm": 2.4402153491973877,
+      "learning_rate": 0.00017730988101396792,
+      "loss": 1.6493,
+      "step": 2203
+    },
+    {
+      "epoch": 0.22791551407667848,
+      "grad_norm": 1.2988420724868774,
+      "learning_rate": 0.0001772995344024832,
+      "loss": 1.4614,
+      "step": 2204
+    },
+    {
+      "epoch": 0.2280189240195445,
+      "grad_norm": 1.1931614875793457,
+      "learning_rate": 0.00017728918779099845,
+      "loss": 1.9354,
+      "step": 2205
+    },
+    {
+      "epoch": 0.2281223339624105,
+      "grad_norm": 1.1833059787750244,
+      "learning_rate": 0.00017727884117951372,
+      "loss": 1.4399,
+      "step": 2206
+    },
+    {
+      "epoch": 0.2282257439052765,
+      "grad_norm": 3.9384918212890625,
+      "learning_rate": 0.00017726849456802898,
+      "loss": 2.5865,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2283291538481425,
+      "grad_norm": 1.3188074827194214,
+      "learning_rate": 0.00017725814795654422,
+      "loss": 1.9454,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2284325637910085,
+      "grad_norm": 1.1684150695800781,
+      "learning_rate": 0.0001772478013450595,
+      "loss": 1.5983,
+      "step": 2209
+    },
+    {
+      "epoch": 0.22853597373387452,
+      "grad_norm": 1.6735270023345947,
+      "learning_rate": 0.00017723745473357475,
+      "loss": 1.4366,
+      "step": 2210
+    },
+    {
+      "epoch": 0.22863938367674053,
+      "grad_norm": 1.5910207033157349,
+      "learning_rate": 0.00017722710812209002,
+      "loss": 2.0415,
+      "step": 2211
+    },
+    {
+      "epoch": 0.22874279361960653,
+      "grad_norm": 1.5946006774902344,
+      "learning_rate": 0.00017721676151060528,
+      "loss": 1.7567,
+      "step": 2212
+    },
+    {
+      "epoch": 0.22884620356247254,
+      "grad_norm": 1.840343713760376,
+      "learning_rate": 0.00017720641489912055,
+      "loss": 1.3738,
+      "step": 2213
+    },
+    {
+      "epoch": 0.22894961350533855,
+      "grad_norm": 1.5357155799865723,
+      "learning_rate": 0.0001771960682876358,
+      "loss": 1.8191,
+      "step": 2214
+    },
+    {
+      "epoch": 0.22905302344820455,
+      "grad_norm": 1.179421067237854,
+      "learning_rate": 0.00017718572167615105,
+      "loss": 1.5523,
+      "step": 2215
+    },
+    {
+      "epoch": 0.22915643339107056,
+      "grad_norm": 1.764402151107788,
+      "learning_rate": 0.00017717537506466632,
+      "loss": 1.1171,
+      "step": 2216
+    },
+    {
+      "epoch": 0.22925984333393656,
+      "grad_norm": 1.2253751754760742,
+      "learning_rate": 0.00017716502845318159,
+      "loss": 1.751,
+      "step": 2217
+    },
+    {
+      "epoch": 0.22936325327680257,
+      "grad_norm": 2.0649373531341553,
+      "learning_rate": 0.00017715468184169685,
+      "loss": 1.865,
+      "step": 2218
+    },
+    {
+      "epoch": 0.22946666321966858,
+      "grad_norm": 1.24588143825531,
+      "learning_rate": 0.00017714433523021212,
+      "loss": 1.9165,
+      "step": 2219
+    },
+    {
+      "epoch": 0.22957007316253458,
+      "grad_norm": 1.8557204008102417,
+      "learning_rate": 0.00017713398861872735,
+      "loss": 1.5765,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2296734831054006,
+      "grad_norm": 1.494800090789795,
+      "learning_rate": 0.00017712364200724262,
+      "loss": 1.8745,
+      "step": 2221
+    },
+    {
+      "epoch": 0.2297768930482666,
+      "grad_norm": 1.9378629922866821,
+      "learning_rate": 0.00017711329539575789,
+      "loss": 2.0202,
+      "step": 2222
+    },
+    {
+      "epoch": 0.2298803029911326,
+      "grad_norm": 2.8973958492279053,
+      "learning_rate": 0.00017710294878427315,
+      "loss": 1.1802,
+      "step": 2223
+    },
+    {
+      "epoch": 0.2299837129339986,
+      "grad_norm": 2.0168254375457764,
+      "learning_rate": 0.00017709260217278842,
+      "loss": 1.5332,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2300871228768646,
+      "grad_norm": 0.7732300758361816,
+      "learning_rate": 0.00017708225556130368,
+      "loss": 0.937,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23019053281973062,
+      "grad_norm": 2.062314987182617,
+      "learning_rate": 0.00017707190894981892,
+      "loss": 2.1582,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23029394276259663,
+      "grad_norm": 1.475707769393921,
+      "learning_rate": 0.0001770615623383342,
+      "loss": 1.8616,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23039735270546263,
+      "grad_norm": 1.7562530040740967,
+      "learning_rate": 0.00017705121572684945,
+      "loss": 1.6349,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23050076264832864,
+      "grad_norm": 3.099172592163086,
+      "learning_rate": 0.00017704086911536472,
+      "loss": 2.1556,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23060417259119464,
+      "grad_norm": 1.312379240989685,
+      "learning_rate": 0.00017703052250387998,
+      "loss": 1.5145,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23070758253406065,
+      "grad_norm": 1.0672568082809448,
+      "learning_rate": 0.00017702017589239525,
+      "loss": 1.5323,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23081099247692666,
+      "grad_norm": 2.077061891555786,
+      "learning_rate": 0.0001770098292809105,
+      "loss": 1.9191,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23091440241979266,
+      "grad_norm": 3.2172346115112305,
+      "learning_rate": 0.00017699948266942575,
+      "loss": 1.3069,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23101781236265867,
+      "grad_norm": 1.9069652557373047,
+      "learning_rate": 0.00017698913605794102,
+      "loss": 1.8306,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23112122230552468,
+      "grad_norm": 1.4115225076675415,
+      "learning_rate": 0.00017697878944645628,
+      "loss": 1.1319,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23122463224839068,
+      "grad_norm": 1.0676754713058472,
+      "learning_rate": 0.00017696844283497155,
+      "loss": 1.8715,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2313280421912567,
+      "grad_norm": 1.7754651308059692,
+      "learning_rate": 0.00017695809622348682,
+      "loss": 1.6673,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2314314521341227,
+      "grad_norm": 1.3493293523788452,
+      "learning_rate": 0.00017694774961200205,
+      "loss": 1.7102,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2315348620769887,
+      "grad_norm": 1.344414472579956,
+      "learning_rate": 0.00017693740300051732,
+      "loss": 1.5833,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2316382720198547,
+      "grad_norm": 2.3626956939697266,
+      "learning_rate": 0.00017692705638903259,
+      "loss": 1.6388,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2317416819627207,
+      "grad_norm": 1.318768858909607,
+      "learning_rate": 0.00017691670977754785,
+      "loss": 1.4585,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23184509190558672,
+      "grad_norm": 1.23952054977417,
+      "learning_rate": 0.00017690636316606312,
+      "loss": 1.211,
+      "step": 2242
+    },
+    {
+      "epoch": 0.23194850184845273,
+      "grad_norm": 3.935603380203247,
+      "learning_rate": 0.00017689601655457838,
+      "loss": 2.0167,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23205191179131873,
+      "grad_norm": 1.5189486742019653,
+      "learning_rate": 0.00017688566994309365,
+      "loss": 2.0937,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23215532173418474,
+      "grad_norm": 2.4079957008361816,
+      "learning_rate": 0.0001768753233316089,
+      "loss": 1.1639,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23225873167705074,
+      "grad_norm": 1.9767324924468994,
+      "learning_rate": 0.00017686497672012415,
+      "loss": 1.7567,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23236214161991675,
+      "grad_norm": 2.0366764068603516,
+      "learning_rate": 0.00017685463010863942,
+      "loss": 1.8011,
+      "step": 2247
+    },
+    {
+      "epoch": 0.23246555156278276,
+      "grad_norm": 1.6386555433273315,
+      "learning_rate": 0.00017684428349715468,
+      "loss": 1.8492,
+      "step": 2248
+    },
+    {
+      "epoch": 0.23256896150564876,
+      "grad_norm": 1.8099335432052612,
+      "learning_rate": 0.00017683393688566995,
+      "loss": 1.5167,
+      "step": 2249
+    },
+    {
+      "epoch": 0.23267237144851477,
+      "grad_norm": 1.8088555335998535,
+      "learning_rate": 0.00017682359027418521,
+      "loss": 1.8067,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23277578139138078,
+      "grad_norm": 1.6619300842285156,
+      "learning_rate": 0.00017681324366270048,
+      "loss": 2.3162,
+      "step": 2251
+    },
+    {
+      "epoch": 0.23287919133424678,
+      "grad_norm": 1.1905879974365234,
+      "learning_rate": 0.00017680289705121574,
+      "loss": 1.6536,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2329826012771128,
+      "grad_norm": 1.5096153020858765,
+      "learning_rate": 0.000176792550439731,
+      "loss": 2.1118,
+      "step": 2253
+    },
+    {
+      "epoch": 0.2330860112199788,
+      "grad_norm": 0.9325785040855408,
+      "learning_rate": 0.00017678220382824625,
+      "loss": 2.137,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2331894211628448,
+      "grad_norm": 2.137453079223633,
+      "learning_rate": 0.00017677185721676151,
+      "loss": 2.3314,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2332928311057108,
+      "grad_norm": 1.8949631452560425,
+      "learning_rate": 0.00017676151060527678,
+      "loss": 1.5151,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2333962410485768,
+      "grad_norm": 1.6897765398025513,
+      "learning_rate": 0.00017675116399379205,
+      "loss": 1.5015,
+      "step": 2257
+    },
+    {
+      "epoch": 0.23349965099144282,
+      "grad_norm": 1.646003246307373,
+      "learning_rate": 0.0001767408173823073,
+      "loss": 1.7934,
+      "step": 2258
+    },
+    {
+      "epoch": 0.23360306093430883,
+      "grad_norm": 2.5955355167388916,
+      "learning_rate": 0.00017673047077082258,
+      "loss": 2.2967,
+      "step": 2259
+    },
+    {
+      "epoch": 0.23370647087717483,
+      "grad_norm": 1.4860060214996338,
+      "learning_rate": 0.00017672012415933784,
+      "loss": 1.8974,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23380988082004084,
+      "grad_norm": 1.9003316164016724,
+      "learning_rate": 0.0001767097775478531,
+      "loss": 1.5562,
+      "step": 2261
+    },
+    {
+      "epoch": 0.23391329076290684,
+      "grad_norm": 2.3601090908050537,
+      "learning_rate": 0.00017669943093636835,
+      "loss": 2.0441,
+      "step": 2262
+    },
+    {
+      "epoch": 0.23401670070577285,
+      "grad_norm": 1.5946472883224487,
+      "learning_rate": 0.0001766890843248836,
+      "loss": 1.4005,
+      "step": 2263
+    },
+    {
+      "epoch": 0.23412011064863886,
+      "grad_norm": 2.192887783050537,
+      "learning_rate": 0.00017667873771339888,
+      "loss": 1.6484,
+      "step": 2264
+    },
+    {
+      "epoch": 0.23422352059150486,
+      "grad_norm": 2.673560619354248,
+      "learning_rate": 0.00017666839110191414,
+      "loss": 1.7329,
+      "step": 2265
+    },
+    {
+      "epoch": 0.23432693053437087,
+      "grad_norm": 1.700700044631958,
+      "learning_rate": 0.0001766580444904294,
+      "loss": 1.4666,
+      "step": 2266
+    },
+    {
+      "epoch": 0.23443034047723688,
+      "grad_norm": 2.216442346572876,
+      "learning_rate": 0.00017664769787894467,
+      "loss": 1.9816,
+      "step": 2267
+    },
+    {
+      "epoch": 0.23453375042010288,
+      "grad_norm": 2.568310022354126,
+      "learning_rate": 0.0001766373512674599,
+      "loss": 1.6597,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2346371603629689,
+      "grad_norm": 2.262301206588745,
+      "learning_rate": 0.00017662700465597518,
+      "loss": 2.1794,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2347405703058349,
+      "grad_norm": 1.755656361579895,
+      "learning_rate": 0.00017661665804449044,
+      "loss": 2.4144,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2348439802487009,
+      "grad_norm": 1.864424705505371,
+      "learning_rate": 0.0001766063114330057,
+      "loss": 1.8568,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2349473901915669,
+      "grad_norm": 0.9477588534355164,
+      "learning_rate": 0.00017659596482152097,
+      "loss": 1.8716,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2350508001344329,
+      "grad_norm": 1.9028419256210327,
+      "learning_rate": 0.00017658561821003624,
+      "loss": 1.9195,
+      "step": 2273
+    },
+    {
+      "epoch": 0.23515421007729892,
+      "grad_norm": 1.5173331499099731,
+      "learning_rate": 0.00017657527159855148,
+      "loss": 1.8441,
+      "step": 2274
+    },
+    {
+      "epoch": 0.23525762002016493,
+      "grad_norm": 1.5247029066085815,
+      "learning_rate": 0.00017656492498706674,
+      "loss": 1.1869,
+      "step": 2275
+    },
+    {
+      "epoch": 0.23536102996303093,
+      "grad_norm": 1.830028772354126,
+      "learning_rate": 0.000176554578375582,
+      "loss": 1.7555,
+      "step": 2276
+    },
+    {
+      "epoch": 0.23546443990589694,
+      "grad_norm": 2.476260185241699,
+      "learning_rate": 0.00017654423176409728,
+      "loss": 1.7501,
+      "step": 2277
+    },
+    {
+      "epoch": 0.23556784984876297,
+      "grad_norm": 1.660536289215088,
+      "learning_rate": 0.00017653388515261254,
+      "loss": 1.5538,
+      "step": 2278
+    },
+    {
+      "epoch": 0.23567125979162898,
+      "grad_norm": 2.4655914306640625,
+      "learning_rate": 0.0001765235385411278,
+      "loss": 2.124,
+      "step": 2279
+    },
+    {
+      "epoch": 0.23577466973449498,
+      "grad_norm": 2.0129129886627197,
+      "learning_rate": 0.00017651319192964305,
+      "loss": 1.624,
+      "step": 2280
+    },
+    {
+      "epoch": 0.235878079677361,
+      "grad_norm": 1.2003356218338013,
+      "learning_rate": 0.0001765028453181583,
+      "loss": 1.4315,
+      "step": 2281
+    },
+    {
+      "epoch": 0.235981489620227,
+      "grad_norm": 1.2782875299453735,
+      "learning_rate": 0.00017649249870667358,
+      "loss": 1.7513,
+      "step": 2282
+    },
+    {
+      "epoch": 0.236084899563093,
+      "grad_norm": 2.0260257720947266,
+      "learning_rate": 0.00017648215209518884,
+      "loss": 1.9921,
+      "step": 2283
+    },
+    {
+      "epoch": 0.236188309505959,
+      "grad_norm": 1.1111077070236206,
+      "learning_rate": 0.0001764718054837041,
+      "loss": 1.8772,
+      "step": 2284
+    },
+    {
+      "epoch": 0.23629171944882502,
+      "grad_norm": 1.1391185522079468,
+      "learning_rate": 0.00017646145887221937,
+      "loss": 1.3482,
+      "step": 2285
+    },
+    {
+      "epoch": 0.23639512939169102,
+      "grad_norm": 1.0886589288711548,
+      "learning_rate": 0.0001764511122607346,
+      "loss": 1.0132,
+      "step": 2286
+    },
+    {
+      "epoch": 0.23649853933455703,
+      "grad_norm": 1.4947338104248047,
+      "learning_rate": 0.00017644076564924988,
+      "loss": 1.5809,
+      "step": 2287
+    },
+    {
+      "epoch": 0.23660194927742303,
+      "grad_norm": 1.7177505493164062,
+      "learning_rate": 0.00017643041903776514,
+      "loss": 1.2868,
+      "step": 2288
+    },
+    {
+      "epoch": 0.23670535922028904,
+      "grad_norm": 1.0610607862472534,
+      "learning_rate": 0.0001764200724262804,
+      "loss": 2.027,
+      "step": 2289
+    },
+    {
+      "epoch": 0.23680876916315505,
+      "grad_norm": 1.152755856513977,
+      "learning_rate": 0.00017640972581479567,
+      "loss": 1.4798,
+      "step": 2290
+    },
+    {
+      "epoch": 0.23691217910602105,
+      "grad_norm": 1.1068141460418701,
+      "learning_rate": 0.00017639937920331094,
+      "loss": 2.0155,
+      "step": 2291
+    },
+    {
+      "epoch": 0.23701558904888706,
+      "grad_norm": 1.3579416275024414,
+      "learning_rate": 0.00017638903259182618,
+      "loss": 1.878,
+      "step": 2292
+    },
+    {
+      "epoch": 0.23711899899175307,
+      "grad_norm": 1.3110994100570679,
+      "learning_rate": 0.00017637868598034144,
+      "loss": 1.5488,
+      "step": 2293
+    },
+    {
+      "epoch": 0.23722240893461907,
+      "grad_norm": 3.31453275680542,
+      "learning_rate": 0.0001763683393688567,
+      "loss": 2.1308,
+      "step": 2294
+    },
+    {
+      "epoch": 0.23732581887748508,
+      "grad_norm": 0.9859296679496765,
+      "learning_rate": 0.00017635799275737197,
+      "loss": 1.7665,
+      "step": 2295
+    },
+    {
+      "epoch": 0.23742922882035108,
+      "grad_norm": 1.7569581270217896,
+      "learning_rate": 0.00017634764614588724,
+      "loss": 1.7785,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2375326387632171,
+      "grad_norm": 1.4133583307266235,
+      "learning_rate": 0.0001763372995344025,
+      "loss": 1.7305,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2376360487060831,
+      "grad_norm": 3.925902843475342,
+      "learning_rate": 0.00017632695292291774,
+      "loss": 2.0507,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2377394586489491,
+      "grad_norm": 1.5905970335006714,
+      "learning_rate": 0.000176316606311433,
+      "loss": 1.6149,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2378428685918151,
+      "grad_norm": 1.6597696542739868,
+      "learning_rate": 0.00017630625969994828,
+      "loss": 1.4731,
+      "step": 2300
+    },
+    {
+      "epoch": 0.23794627853468112,
+      "grad_norm": 1.602620005607605,
+      "learning_rate": 0.00017629591308846354,
+      "loss": 1.8871,
+      "step": 2301
+    },
+    {
+      "epoch": 0.23804968847754712,
+      "grad_norm": 1.0693349838256836,
+      "learning_rate": 0.0001762855664769788,
+      "loss": 1.3862,
+      "step": 2302
+    },
+    {
+      "epoch": 0.23815309842041313,
+      "grad_norm": 1.631103277206421,
+      "learning_rate": 0.00017627521986549407,
+      "loss": 1.8944,
+      "step": 2303
+    },
+    {
+      "epoch": 0.23825650836327913,
+      "grad_norm": 2.4888055324554443,
+      "learning_rate": 0.0001762648732540093,
+      "loss": 2.5028,
+      "step": 2304
+    },
+    {
+      "epoch": 0.23835991830614514,
+      "grad_norm": 1.9068800210952759,
+      "learning_rate": 0.00017625452664252458,
+      "loss": 1.3553,
+      "step": 2305
+    },
+    {
+      "epoch": 0.23846332824901115,
+      "grad_norm": 2.230555295944214,
+      "learning_rate": 0.00017624418003103984,
+      "loss": 1.9119,
+      "step": 2306
+    },
+    {
+      "epoch": 0.23856673819187715,
+      "grad_norm": 1.1259522438049316,
+      "learning_rate": 0.0001762338334195551,
+      "loss": 1.7393,
+      "step": 2307
+    },
+    {
+      "epoch": 0.23867014813474316,
+      "grad_norm": 2.369147777557373,
+      "learning_rate": 0.00017622348680807037,
+      "loss": 1.8111,
+      "step": 2308
+    },
+    {
+      "epoch": 0.23877355807760917,
+      "grad_norm": 1.7751003503799438,
+      "learning_rate": 0.00017621314019658564,
+      "loss": 1.5761,
+      "step": 2309
+    },
+    {
+      "epoch": 0.23887696802047517,
+      "grad_norm": 2.357508420944214,
+      "learning_rate": 0.00017620279358510088,
+      "loss": 1.8912,
+      "step": 2310
+    },
+    {
+      "epoch": 0.23898037796334118,
+      "grad_norm": 0.9444859027862549,
+      "learning_rate": 0.00017619244697361614,
+      "loss": 2.177,
+      "step": 2311
+    },
+    {
+      "epoch": 0.23908378790620718,
+      "grad_norm": 1.4139807224273682,
+      "learning_rate": 0.0001761821003621314,
+      "loss": 1.5666,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2391871978490732,
+      "grad_norm": 2.163088321685791,
+      "learning_rate": 0.00017617175375064667,
+      "loss": 1.5789,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2392906077919392,
+      "grad_norm": 1.5765900611877441,
+      "learning_rate": 0.00017616140713916194,
+      "loss": 1.7379,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2393940177348052,
+      "grad_norm": 2.842447280883789,
+      "learning_rate": 0.0001761510605276772,
+      "loss": 1.711,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2394974276776712,
+      "grad_norm": 2.930110216140747,
+      "learning_rate": 0.00017614071391619244,
+      "loss": 2.089,
+      "step": 2316
+    },
+    {
+      "epoch": 0.23960083762053722,
+      "grad_norm": 1.4796249866485596,
+      "learning_rate": 0.0001761303673047077,
+      "loss": 1.9833,
+      "step": 2317
+    },
+    {
+      "epoch": 0.23970424756340322,
+      "grad_norm": 1.5853078365325928,
+      "learning_rate": 0.00017612002069322297,
+      "loss": 1.8592,
+      "step": 2318
+    },
+    {
+      "epoch": 0.23980765750626923,
+      "grad_norm": 0.9812987446784973,
+      "learning_rate": 0.00017610967408173824,
+      "loss": 1.7803,
+      "step": 2319
+    },
+    {
+      "epoch": 0.23991106744913523,
+      "grad_norm": 1.229555606842041,
+      "learning_rate": 0.0001760993274702535,
+      "loss": 1.0883,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24001447739200124,
+      "grad_norm": 2.2885689735412598,
+      "learning_rate": 0.00017608898085876877,
+      "loss": 1.7889,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24011788733486725,
+      "grad_norm": 1.1627824306488037,
+      "learning_rate": 0.000176078634247284,
+      "loss": 1.106,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24022129727773325,
+      "grad_norm": 1.3515424728393555,
+      "learning_rate": 0.00017606828763579928,
+      "loss": 1.2491,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24032470722059926,
+      "grad_norm": 0.9530225396156311,
+      "learning_rate": 0.00017605794102431454,
+      "loss": 1.7492,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24042811716346527,
+      "grad_norm": 1.797919750213623,
+      "learning_rate": 0.0001760475944128298,
+      "loss": 1.4208,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24053152710633127,
+      "grad_norm": 1.2609739303588867,
+      "learning_rate": 0.00017603724780134507,
+      "loss": 1.2555,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24063493704919728,
+      "grad_norm": 1.9391148090362549,
+      "learning_rate": 0.00017602690118986034,
+      "loss": 1.0435,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24073834699206328,
+      "grad_norm": 1.8738831281661987,
+      "learning_rate": 0.00017601655457837558,
+      "loss": 1.8572,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2408417569349293,
+      "grad_norm": 2.168691396713257,
+      "learning_rate": 0.00017600620796689084,
+      "loss": 1.7667,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2409451668777953,
+      "grad_norm": 2.458089828491211,
+      "learning_rate": 0.0001759958613554061,
+      "loss": 1.871,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2410485768206613,
+      "grad_norm": 1.9956097602844238,
+      "learning_rate": 0.00017598551474392137,
+      "loss": 2.3792,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2411519867635273,
+      "grad_norm": 1.3749737739562988,
+      "learning_rate": 0.00017597516813243664,
+      "loss": 1.6698,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24125539670639332,
+      "grad_norm": 1.7299331426620483,
+      "learning_rate": 0.0001759648215209519,
+      "loss": 1.64,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24135880664925932,
+      "grad_norm": 2.118608236312866,
+      "learning_rate": 0.00017595447490946714,
+      "loss": 1.3257,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24146221659212533,
+      "grad_norm": 1.570468544960022,
+      "learning_rate": 0.0001759441282979824,
+      "loss": 1.5102,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24156562653499133,
+      "grad_norm": 2.2476089000701904,
+      "learning_rate": 0.00017593378168649767,
+      "loss": 1.6563,
+      "step": 2336
+    },
+    {
+      "epoch": 0.24166903647785734,
+      "grad_norm": 2.5278682708740234,
+      "learning_rate": 0.00017592343507501294,
+      "loss": 1.7088,
+      "step": 2337
+    },
+    {
+      "epoch": 0.24177244642072335,
+      "grad_norm": 1.7787667512893677,
+      "learning_rate": 0.0001759130884635282,
+      "loss": 1.7681,
+      "step": 2338
+    },
+    {
+      "epoch": 0.24187585636358935,
+      "grad_norm": 1.1270442008972168,
+      "learning_rate": 0.00017590274185204347,
+      "loss": 1.477,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24197926630645536,
+      "grad_norm": 1.9922478199005127,
+      "learning_rate": 0.0001758923952405587,
+      "loss": 2.2118,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24208267624932137,
+      "grad_norm": 1.2730507850646973,
+      "learning_rate": 0.00017588204862907397,
+      "loss": 1.5687,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24218608619218737,
+      "grad_norm": 2.0439341068267822,
+      "learning_rate": 0.00017587170201758924,
+      "loss": 1.6693,
+      "step": 2342
+    },
+    {
+      "epoch": 0.24228949613505338,
+      "grad_norm": 1.3062502145767212,
+      "learning_rate": 0.0001758613554061045,
+      "loss": 1.497,
+      "step": 2343
+    },
+    {
+      "epoch": 0.24239290607791938,
+      "grad_norm": 2.715073823928833,
+      "learning_rate": 0.00017585100879461977,
+      "loss": 1.6284,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2424963160207854,
+      "grad_norm": 1.6569184064865112,
+      "learning_rate": 0.00017584066218313504,
+      "loss": 1.6966,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2425997259636514,
+      "grad_norm": 1.6366918087005615,
+      "learning_rate": 0.00017583031557165027,
+      "loss": 1.8435,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2427031359065174,
+      "grad_norm": 1.7843844890594482,
+      "learning_rate": 0.00017581996896016554,
+      "loss": 1.941,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2428065458493834,
+      "grad_norm": 1.6880625486373901,
+      "learning_rate": 0.0001758096223486808,
+      "loss": 1.2436,
+      "step": 2348
+    },
+    {
+      "epoch": 0.24290995579224942,
+      "grad_norm": 1.39772629737854,
+      "learning_rate": 0.00017579927573719607,
+      "loss": 1.3784,
+      "step": 2349
+    },
+    {
+      "epoch": 0.24301336573511542,
+      "grad_norm": 1.4256283044815063,
+      "learning_rate": 0.00017578892912571134,
+      "loss": 1.4929,
+      "step": 2350
+    },
+    {
+      "epoch": 0.24311677567798143,
+      "grad_norm": 1.4295189380645752,
+      "learning_rate": 0.0001757785825142266,
+      "loss": 1.6985,
+      "step": 2351
+    },
+    {
+      "epoch": 0.24322018562084743,
+      "grad_norm": 2.4880120754241943,
+      "learning_rate": 0.00017576823590274184,
+      "loss": 1.9306,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24332359556371344,
+      "grad_norm": 1.9826879501342773,
+      "learning_rate": 0.0001757578892912571,
+      "loss": 1.9971,
+      "step": 2353
+    },
+    {
+      "epoch": 0.24342700550657945,
+      "grad_norm": 1.8703138828277588,
+      "learning_rate": 0.00017574754267977237,
+      "loss": 1.2286,
+      "step": 2354
+    },
+    {
+      "epoch": 0.24353041544944545,
+      "grad_norm": 2.05326509475708,
+      "learning_rate": 0.00017573719606828764,
+      "loss": 1.7075,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24363382539231146,
+      "grad_norm": 1.259153127670288,
+      "learning_rate": 0.0001757268494568029,
+      "loss": 1.7123,
+      "step": 2356
+    },
+    {
+      "epoch": 0.24373723533517747,
+      "grad_norm": 1.4366209506988525,
+      "learning_rate": 0.00017571650284531817,
+      "loss": 1.4503,
+      "step": 2357
+    },
+    {
+      "epoch": 0.24384064527804347,
+      "grad_norm": 1.6645400524139404,
+      "learning_rate": 0.0001757061562338334,
+      "loss": 1.1578,
+      "step": 2358
+    },
+    {
+      "epoch": 0.24394405522090948,
+      "grad_norm": 1.6528881788253784,
+      "learning_rate": 0.00017569580962234867,
+      "loss": 1.6437,
+      "step": 2359
+    },
+    {
+      "epoch": 0.24404746516377548,
+      "grad_norm": 1.262462854385376,
+      "learning_rate": 0.00017568546301086394,
+      "loss": 1.6392,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2441508751066415,
+      "grad_norm": 1.7609543800354004,
+      "learning_rate": 0.0001756751163993792,
+      "loss": 2.3627,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2442542850495075,
+      "grad_norm": 0.7221395969390869,
+      "learning_rate": 0.00017566476978789447,
+      "loss": 1.6682,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2443576949923735,
+      "grad_norm": 2.4060263633728027,
+      "learning_rate": 0.00017565442317640974,
+      "loss": 1.7926,
+      "step": 2363
+    },
+    {
+      "epoch": 0.24446110493523954,
+      "grad_norm": 1.0417516231536865,
+      "learning_rate": 0.00017564407656492497,
+      "loss": 1.544,
+      "step": 2364
+    },
+    {
+      "epoch": 0.24456451487810554,
+      "grad_norm": 0.9361560344696045,
+      "learning_rate": 0.00017563372995344024,
+      "loss": 1.5395,
+      "step": 2365
+    },
+    {
+      "epoch": 0.24466792482097155,
+      "grad_norm": 1.880168080329895,
+      "learning_rate": 0.0001756233833419555,
+      "loss": 2.4219,
+      "step": 2366
+    },
+    {
+      "epoch": 0.24477133476383756,
+      "grad_norm": 1.9342352151870728,
+      "learning_rate": 0.00017561303673047077,
+      "loss": 1.3834,
+      "step": 2367
+    },
+    {
+      "epoch": 0.24487474470670356,
+      "grad_norm": 2.5511040687561035,
+      "learning_rate": 0.00017560269011898604,
+      "loss": 1.7103,
+      "step": 2368
+    },
+    {
+      "epoch": 0.24497815464956957,
+      "grad_norm": 0.9870395064353943,
+      "learning_rate": 0.0001755923435075013,
+      "loss": 1.3327,
+      "step": 2369
+    },
+    {
+      "epoch": 0.24508156459243557,
+      "grad_norm": 1.2001618146896362,
+      "learning_rate": 0.00017558199689601657,
+      "loss": 1.7844,
+      "step": 2370
+    },
+    {
+      "epoch": 0.24518497453530158,
+      "grad_norm": 1.2673696279525757,
+      "learning_rate": 0.0001755716502845318,
+      "loss": 1.6301,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2452883844781676,
+      "grad_norm": 1.7478164434432983,
+      "learning_rate": 0.00017556130367304707,
+      "loss": 1.4385,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2453917944210336,
+      "grad_norm": 1.680871844291687,
+      "learning_rate": 0.00017555095706156234,
+      "loss": 2.3595,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2454952043638996,
+      "grad_norm": 1.445009469985962,
+      "learning_rate": 0.0001755406104500776,
+      "loss": 1.4895,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2455986143067656,
+      "grad_norm": 1.6499723196029663,
+      "learning_rate": 0.00017553026383859287,
+      "loss": 1.0861,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2457020242496316,
+      "grad_norm": 2.56699538230896,
+      "learning_rate": 0.00017551991722710813,
+      "loss": 1.8802,
+      "step": 2376
+    },
+    {
+      "epoch": 0.24580543419249762,
+      "grad_norm": 1.9932339191436768,
+      "learning_rate": 0.0001755095706156234,
+      "loss": 1.8787,
+      "step": 2377
+    },
+    {
+      "epoch": 0.24590884413536362,
+      "grad_norm": 1.4826245307922363,
+      "learning_rate": 0.00017549922400413866,
+      "loss": 1.3792,
+      "step": 2378
+    },
+    {
+      "epoch": 0.24601225407822963,
+      "grad_norm": 1.8881828784942627,
+      "learning_rate": 0.0001754888773926539,
+      "loss": 1.9625,
+      "step": 2379
+    },
+    {
+      "epoch": 0.24611566402109564,
+      "grad_norm": 3.217658519744873,
+      "learning_rate": 0.00017547853078116917,
+      "loss": 1.5442,
+      "step": 2380
+    },
+    {
+      "epoch": 0.24621907396396164,
+      "grad_norm": 1.5086668729782104,
+      "learning_rate": 0.00017546818416968443,
+      "loss": 1.3905,
+      "step": 2381
+    },
+    {
+      "epoch": 0.24632248390682765,
+      "grad_norm": 1.2646621465682983,
+      "learning_rate": 0.0001754578375581997,
+      "loss": 1.0604,
+      "step": 2382
+    },
+    {
+      "epoch": 0.24642589384969366,
+      "grad_norm": 1.091963529586792,
+      "learning_rate": 0.00017544749094671497,
+      "loss": 1.1739,
+      "step": 2383
+    },
+    {
+      "epoch": 0.24652930379255966,
+      "grad_norm": 1.5914099216461182,
+      "learning_rate": 0.00017543714433523023,
+      "loss": 2.022,
+      "step": 2384
+    },
+    {
+      "epoch": 0.24663271373542567,
+      "grad_norm": 1.4844694137573242,
+      "learning_rate": 0.0001754267977237455,
+      "loss": 2.1174,
+      "step": 2385
+    },
+    {
+      "epoch": 0.24673612367829167,
+      "grad_norm": 1.486586570739746,
+      "learning_rate": 0.00017541645111226076,
+      "loss": 1.4655,
+      "step": 2386
+    },
+    {
+      "epoch": 0.24683953362115768,
+      "grad_norm": 1.22221839427948,
+      "learning_rate": 0.000175406104500776,
+      "loss": 1.6983,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2469429435640237,
+      "grad_norm": 1.0208945274353027,
+      "learning_rate": 0.00017539575788929127,
+      "loss": 1.2739,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2470463535068897,
+      "grad_norm": 1.514829397201538,
+      "learning_rate": 0.00017538541127780653,
+      "loss": 0.8628,
+      "step": 2389
+    },
+    {
+      "epoch": 0.2471497634497557,
+      "grad_norm": 1.2276806831359863,
+      "learning_rate": 0.0001753750646663218,
+      "loss": 1.4324,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2472531733926217,
+      "grad_norm": 1.2919999361038208,
+      "learning_rate": 0.00017536471805483706,
+      "loss": 1.4985,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2473565833354877,
+      "grad_norm": 1.6556484699249268,
+      "learning_rate": 0.00017535437144335233,
+      "loss": 1.6762,
+      "step": 2392
+    },
+    {
+      "epoch": 0.24745999327835372,
+      "grad_norm": 1.1307460069656372,
+      "learning_rate": 0.0001753440248318676,
+      "loss": 1.6692,
+      "step": 2393
+    },
+    {
+      "epoch": 0.24756340322121972,
+      "grad_norm": 1.1884139776229858,
+      "learning_rate": 0.00017533367822038283,
+      "loss": 1.3586,
+      "step": 2394
+    },
+    {
+      "epoch": 0.24766681316408573,
+      "grad_norm": 1.539978265762329,
+      "learning_rate": 0.0001753233316088981,
+      "loss": 1.4159,
+      "step": 2395
+    },
+    {
+      "epoch": 0.24777022310695174,
+      "grad_norm": 2.753964424133301,
+      "learning_rate": 0.00017531298499741336,
+      "loss": 1.3736,
+      "step": 2396
+    },
+    {
+      "epoch": 0.24787363304981774,
+      "grad_norm": 1.3421005010604858,
+      "learning_rate": 0.00017530263838592863,
+      "loss": 1.8879,
+      "step": 2397
+    },
+    {
+      "epoch": 0.24797704299268375,
+      "grad_norm": 1.2711719274520874,
+      "learning_rate": 0.0001752922917744439,
+      "loss": 1.1862,
+      "step": 2398
+    },
+    {
+      "epoch": 0.24808045293554976,
+      "grad_norm": 1.870287537574768,
+      "learning_rate": 0.00017528194516295916,
+      "loss": 0.6339,
+      "step": 2399
+    },
+    {
+      "epoch": 0.24818386287841576,
+      "grad_norm": 1.4832231998443604,
+      "learning_rate": 0.0001752715985514744,
+      "loss": 1.7517,
+      "step": 2400
+    },
+    {
+      "epoch": 0.24828727282128177,
+      "grad_norm": 2.3762567043304443,
+      "learning_rate": 0.00017526125193998966,
+      "loss": 1.3174,
+      "step": 2401
+    },
+    {
+      "epoch": 0.24839068276414777,
+      "grad_norm": 1.2148253917694092,
+      "learning_rate": 0.00017525090532850493,
+      "loss": 1.4028,
+      "step": 2402
+    },
+    {
+      "epoch": 0.24849409270701378,
+      "grad_norm": 1.2104130983352661,
+      "learning_rate": 0.0001752405587170202,
+      "loss": 1.347,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2485975026498798,
+      "grad_norm": 0.9523527026176453,
+      "learning_rate": 0.00017523021210553546,
+      "loss": 1.9917,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2487009125927458,
+      "grad_norm": 1.6441329717636108,
+      "learning_rate": 0.00017521986549405073,
+      "loss": 1.6389,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2488043225356118,
+      "grad_norm": 2.3193979263305664,
+      "learning_rate": 0.00017520951888256597,
+      "loss": 1.8484,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2489077324784778,
+      "grad_norm": 1.6989390850067139,
+      "learning_rate": 0.00017519917227108123,
+      "loss": 0.9939,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2490111424213438,
+      "grad_norm": 1.290727138519287,
+      "learning_rate": 0.0001751888256595965,
+      "loss": 1.7555,
+      "step": 2408
+    },
+    {
+      "epoch": 0.24911455236420982,
+      "grad_norm": 1.2122284173965454,
+      "learning_rate": 0.00017517847904811176,
+      "loss": 1.4804,
+      "step": 2409
+    },
+    {
+      "epoch": 0.24921796230707582,
+      "grad_norm": 1.3853946924209595,
+      "learning_rate": 0.00017516813243662703,
+      "loss": 1.897,
+      "step": 2410
+    },
+    {
+      "epoch": 0.24932137224994183,
+      "grad_norm": 2.634765148162842,
+      "learning_rate": 0.0001751577858251423,
+      "loss": 1.3514,
+      "step": 2411
+    },
+    {
+      "epoch": 0.24942478219280784,
+      "grad_norm": 2.0169286727905273,
+      "learning_rate": 0.00017514743921365753,
+      "loss": 2.0064,
+      "step": 2412
+    },
+    {
+      "epoch": 0.24952819213567384,
+      "grad_norm": 2.7714152336120605,
+      "learning_rate": 0.0001751370926021728,
+      "loss": 1.7758,
+      "step": 2413
+    },
+    {
+      "epoch": 0.24963160207853985,
+      "grad_norm": 1.5576125383377075,
+      "learning_rate": 0.00017512674599068806,
+      "loss": 1.4461,
+      "step": 2414
+    },
+    {
+      "epoch": 0.24973501202140586,
+      "grad_norm": 1.512837290763855,
+      "learning_rate": 0.00017511639937920333,
+      "loss": 1.7374,
+      "step": 2415
+    },
+    {
+      "epoch": 0.24983842196427186,
+      "grad_norm": 0.8817567825317383,
+      "learning_rate": 0.0001751060527677186,
+      "loss": 1.4742,
+      "step": 2416
+    },
+    {
+      "epoch": 0.24994183190713787,
+      "grad_norm": 0.9802753329277039,
+      "learning_rate": 0.00017509570615623386,
+      "loss": 1.0384,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2500452418500039,
+      "grad_norm": 1.5185511112213135,
+      "learning_rate": 0.0001750853595447491,
+      "loss": 1.6683,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2501486517928699,
+      "grad_norm": 1.5050675868988037,
+      "learning_rate": 0.00017507501293326436,
+      "loss": 1.8603,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2502520617357359,
+      "grad_norm": 1.4089784622192383,
+      "learning_rate": 0.00017506466632177963,
+      "loss": 1.1532,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2503554716786019,
+      "grad_norm": 1.0789377689361572,
+      "learning_rate": 0.0001750543197102949,
+      "loss": 1.1915,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2504588816214679,
+      "grad_norm": 1.937451958656311,
+      "learning_rate": 0.00017504397309881016,
+      "loss": 1.8906,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2505622915643339,
+      "grad_norm": 1.167373776435852,
+      "learning_rate": 0.00017503362648732543,
+      "loss": 2.0633,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2506657015071999,
+      "grad_norm": 1.1927114725112915,
+      "learning_rate": 0.00017502327987584066,
+      "loss": 1.7121,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2507691114500659,
+      "grad_norm": 1.8490036725997925,
+      "learning_rate": 0.00017501293326435593,
+      "loss": 1.7096,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2508725213929319,
+      "grad_norm": 2.367034673690796,
+      "learning_rate": 0.0001750025866528712,
+      "loss": 2.011,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25097593133579793,
+      "grad_norm": 1.748449444770813,
+      "learning_rate": 0.00017499224004138646,
+      "loss": 1.3284,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25107934127866394,
+      "grad_norm": 1.6005412340164185,
+      "learning_rate": 0.00017498189342990173,
+      "loss": 1.866,
+      "step": 2428
+    },
+    {
+      "epoch": 0.25118275122152994,
+      "grad_norm": 1.6830029487609863,
+      "learning_rate": 0.000174971546818417,
+      "loss": 1.596,
+      "step": 2429
+    },
+    {
+      "epoch": 0.25128616116439595,
+      "grad_norm": 1.3933378458023071,
+      "learning_rate": 0.00017496120020693223,
+      "loss": 1.3813,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25138957110726196,
+      "grad_norm": 2.1896424293518066,
+      "learning_rate": 0.0001749508535954475,
+      "loss": 1.8016,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25149298105012796,
+      "grad_norm": 1.2955199480056763,
+      "learning_rate": 0.00017494050698396276,
+      "loss": 1.4218,
+      "step": 2432
+    },
+    {
+      "epoch": 0.25159639099299397,
+      "grad_norm": 1.6225368976593018,
+      "learning_rate": 0.00017493016037247803,
+      "loss": 1.5471,
+      "step": 2433
+    },
+    {
+      "epoch": 0.25169980093586,
+      "grad_norm": 1.3177118301391602,
+      "learning_rate": 0.0001749198137609933,
+      "loss": 1.3752,
+      "step": 2434
+    },
+    {
+      "epoch": 0.251803210878726,
+      "grad_norm": 2.3560526371002197,
+      "learning_rate": 0.00017490946714950856,
+      "loss": 1.2315,
+      "step": 2435
+    },
+    {
+      "epoch": 0.251906620821592,
+      "grad_norm": 2.910111665725708,
+      "learning_rate": 0.0001748991205380238,
+      "loss": 1.7354,
+      "step": 2436
+    },
+    {
+      "epoch": 0.252010030764458,
+      "grad_norm": 1.5888240337371826,
+      "learning_rate": 0.00017488877392653906,
+      "loss": 1.2371,
+      "step": 2437
+    },
+    {
+      "epoch": 0.252113440707324,
+      "grad_norm": 1.467577576637268,
+      "learning_rate": 0.00017487842731505433,
+      "loss": 1.2963,
+      "step": 2438
+    },
+    {
+      "epoch": 0.25221685065019,
+      "grad_norm": 0.8557147979736328,
+      "learning_rate": 0.0001748680807035696,
+      "loss": 1.392,
+      "step": 2439
+    },
+    {
+      "epoch": 0.252320260593056,
+      "grad_norm": 1.261374592781067,
+      "learning_rate": 0.00017485773409208486,
+      "loss": 1.3752,
+      "step": 2440
+    },
+    {
+      "epoch": 0.252423670535922,
+      "grad_norm": 1.7545433044433594,
+      "learning_rate": 0.00017484738748060012,
+      "loss": 1.375,
+      "step": 2441
+    },
+    {
+      "epoch": 0.252527080478788,
+      "grad_norm": 2.189995288848877,
+      "learning_rate": 0.00017483704086911536,
+      "loss": 1.9399,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25263049042165403,
+      "grad_norm": 1.4490615129470825,
+      "learning_rate": 0.00017482669425763063,
+      "loss": 1.8315,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25273390036452004,
+      "grad_norm": 1.2145003080368042,
+      "learning_rate": 0.0001748163476461459,
+      "loss": 1.0258,
+      "step": 2444
+    },
+    {
+      "epoch": 0.25283731030738604,
+      "grad_norm": 1.8237292766571045,
+      "learning_rate": 0.00017480600103466116,
+      "loss": 1.3875,
+      "step": 2445
+    },
+    {
+      "epoch": 0.25294072025025205,
+      "grad_norm": 1.7630832195281982,
+      "learning_rate": 0.00017479565442317643,
+      "loss": 1.7942,
+      "step": 2446
+    },
+    {
+      "epoch": 0.25304413019311806,
+      "grad_norm": 1.2417792081832886,
+      "learning_rate": 0.0001747853078116917,
+      "loss": 1.5995,
+      "step": 2447
+    },
+    {
+      "epoch": 0.25314754013598406,
+      "grad_norm": 0.8555414080619812,
+      "learning_rate": 0.00017477496120020693,
+      "loss": 1.4727,
+      "step": 2448
+    },
+    {
+      "epoch": 0.25325095007885007,
+      "grad_norm": 1.3805317878723145,
+      "learning_rate": 0.0001747646145887222,
+      "loss": 1.457,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2533543600217161,
+      "grad_norm": 2.568117141723633,
+      "learning_rate": 0.00017475426797723746,
+      "loss": 1.5291,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2534577699645821,
+      "grad_norm": 1.2383670806884766,
+      "learning_rate": 0.00017474392136575273,
+      "loss": 1.1544,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2535611799074481,
+      "grad_norm": 1.880109429359436,
+      "learning_rate": 0.000174733574754268,
+      "loss": 1.7525,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2536645898503141,
+      "grad_norm": 0.9020270109176636,
+      "learning_rate": 0.00017472322814278326,
+      "loss": 1.9656,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2537679997931801,
+      "grad_norm": 0.9956245422363281,
+      "learning_rate": 0.0001747128815312985,
+      "loss": 1.5702,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2538714097360461,
+      "grad_norm": 1.4423266649246216,
+      "learning_rate": 0.00017470253491981376,
+      "loss": 1.2459,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2539748196789121,
+      "grad_norm": 1.7919795513153076,
+      "learning_rate": 0.00017469218830832903,
+      "loss": 1.6817,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2540782296217781,
+      "grad_norm": 4.085237979888916,
+      "learning_rate": 0.0001746818416968443,
+      "loss": 1.8207,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2541816395646441,
+      "grad_norm": 1.3783023357391357,
+      "learning_rate": 0.00017467149508535956,
+      "loss": 1.5989,
+      "step": 2458
+    },
+    {
+      "epoch": 0.25428504950751013,
+      "grad_norm": 1.3352357149124146,
+      "learning_rate": 0.00017466114847387482,
+      "loss": 1.5502,
+      "step": 2459
+    },
+    {
+      "epoch": 0.25438845945037614,
+      "grad_norm": 1.0194011926651,
+      "learning_rate": 0.00017465080186239006,
+      "loss": 1.7048,
+      "step": 2460
+    },
+    {
+      "epoch": 0.25449186939324214,
+      "grad_norm": 1.5900938510894775,
+      "learning_rate": 0.00017464045525090533,
+      "loss": 1.3563,
+      "step": 2461
+    },
+    {
+      "epoch": 0.25459527933610815,
+      "grad_norm": 2.041964530944824,
+      "learning_rate": 0.0001746301086394206,
+      "loss": 1.4161,
+      "step": 2462
+    },
+    {
+      "epoch": 0.25469868927897416,
+      "grad_norm": 3.3470609188079834,
+      "learning_rate": 0.00017461976202793586,
+      "loss": 2.2123,
+      "step": 2463
+    },
+    {
+      "epoch": 0.25480209922184016,
+      "grad_norm": 1.328002691268921,
+      "learning_rate": 0.00017460941541645112,
+      "loss": 1.4335,
+      "step": 2464
+    },
+    {
+      "epoch": 0.25490550916470617,
+      "grad_norm": 2.004028558731079,
+      "learning_rate": 0.0001745990688049664,
+      "loss": 1.5092,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2550089191075722,
+      "grad_norm": 1.4580358266830444,
+      "learning_rate": 0.00017458872219348163,
+      "loss": 2.0,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2551123290504382,
+      "grad_norm": 1.3243385553359985,
+      "learning_rate": 0.0001745783755819969,
+      "loss": 1.8759,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2552157389933042,
+      "grad_norm": 2.5180490016937256,
+      "learning_rate": 0.00017456802897051216,
+      "loss": 1.5182,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 1.6826629638671875,
+      "learning_rate": 0.00017455768235902743,
+      "loss": 1.4656,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2554225588790362,
+      "grad_norm": 2.267388105392456,
+      "learning_rate": 0.0001745473357475427,
+      "loss": 2.3572,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2555259688219022,
+      "grad_norm": 2.7292869091033936,
+      "learning_rate": 0.00017453698913605796,
+      "loss": 2.0746,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2556293787647682,
+      "grad_norm": 2.772287607192993,
+      "learning_rate": 0.0001745266425245732,
+      "loss": 1.9232,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2557327887076342,
+      "grad_norm": 1.6850965023040771,
+      "learning_rate": 0.00017451629591308846,
+      "loss": 2.2237,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2558361986505002,
+      "grad_norm": 2.2228617668151855,
+      "learning_rate": 0.00017450594930160373,
+      "loss": 2.0034,
+      "step": 2474
+    },
+    {
+      "epoch": 0.25593960859336623,
+      "grad_norm": 1.7392048835754395,
+      "learning_rate": 0.000174495602690119,
+      "loss": 1.2657,
+      "step": 2475
+    },
+    {
+      "epoch": 0.25604301853623224,
+      "grad_norm": 1.7491132020950317,
+      "learning_rate": 0.00017448525607863426,
+      "loss": 0.9487,
+      "step": 2476
+    },
+    {
+      "epoch": 0.25614642847909824,
+      "grad_norm": 1.7165073156356812,
+      "learning_rate": 0.00017447490946714952,
+      "loss": 1.463,
+      "step": 2477
+    },
+    {
+      "epoch": 0.25624983842196425,
+      "grad_norm": 1.6869577169418335,
+      "learning_rate": 0.00017446456285566476,
+      "loss": 1.8076,
+      "step": 2478
+    },
+    {
+      "epoch": 0.25635324836483026,
+      "grad_norm": 1.95469331741333,
+      "learning_rate": 0.00017445421624418003,
+      "loss": 1.8281,
+      "step": 2479
+    },
+    {
+      "epoch": 0.25645665830769626,
+      "grad_norm": 0.9262880086898804,
+      "learning_rate": 0.0001744438696326953,
+      "loss": 1.4683,
+      "step": 2480
+    },
+    {
+      "epoch": 0.25656006825056227,
+      "grad_norm": 1.4505789279937744,
+      "learning_rate": 0.00017443352302121056,
+      "loss": 1.425,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2566634781934283,
+      "grad_norm": 1.750234842300415,
+      "learning_rate": 0.00017442317640972582,
+      "loss": 2.1092,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2567668881362943,
+      "grad_norm": 1.320644497871399,
+      "learning_rate": 0.0001744128297982411,
+      "loss": 1.3916,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2568702980791603,
+      "grad_norm": 0.9432196617126465,
+      "learning_rate": 0.00017440248318675633,
+      "loss": 1.3987,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2569737080220263,
+      "grad_norm": 2.0192582607269287,
+      "learning_rate": 0.0001743921365752716,
+      "loss": 1.9502,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2570771179648923,
+      "grad_norm": 1.798473834991455,
+      "learning_rate": 0.00017438178996378686,
+      "loss": 1.6392,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2571805279077583,
+      "grad_norm": 1.092153549194336,
+      "learning_rate": 0.00017437144335230212,
+      "loss": 1.702,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2572839378506243,
+      "grad_norm": 0.8232825398445129,
+      "learning_rate": 0.0001743610967408174,
+      "loss": 1.6369,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2573873477934903,
+      "grad_norm": 2.7390313148498535,
+      "learning_rate": 0.00017435075012933266,
+      "loss": 2.2424,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2574907577363563,
+      "grad_norm": 1.3584825992584229,
+      "learning_rate": 0.0001743404035178479,
+      "loss": 1.5879,
+      "step": 2490
+    },
+    {
+      "epoch": 0.25759416767922233,
+      "grad_norm": 1.8954455852508545,
+      "learning_rate": 0.00017433005690636316,
+      "loss": 2.0067,
+      "step": 2491
+    },
+    {
+      "epoch": 0.25769757762208834,
+      "grad_norm": 1.358513593673706,
+      "learning_rate": 0.00017431971029487842,
+      "loss": 1.8005,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2578009875649544,
+      "grad_norm": 1.0788158178329468,
+      "learning_rate": 0.0001743093636833937,
+      "loss": 1.5718,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2579043975078204,
+      "grad_norm": 1.8200621604919434,
+      "learning_rate": 0.00017429901707190896,
+      "loss": 0.9247,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2580078074506864,
+      "grad_norm": 1.5555850267410278,
+      "learning_rate": 0.00017428867046042422,
+      "loss": 1.5324,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2581112173935524,
+      "grad_norm": 1.2410125732421875,
+      "learning_rate": 0.00017427832384893946,
+      "loss": 0.8922,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2582146273364184,
+      "grad_norm": 2.1035003662109375,
+      "learning_rate": 0.00017426797723745473,
+      "loss": 1.7946,
+      "step": 2497
+    },
+    {
+      "epoch": 0.25831803727928443,
+      "grad_norm": 1.2379070520401,
+      "learning_rate": 0.00017425763062597,
+      "loss": 1.4737,
+      "step": 2498
+    },
+    {
+      "epoch": 0.25842144722215044,
+      "grad_norm": 3.038174867630005,
+      "learning_rate": 0.00017424728401448526,
+      "loss": 1.7392,
+      "step": 2499
+    },
+    {
+      "epoch": 0.25852485716501644,
+      "grad_norm": 2.1067824363708496,
+      "learning_rate": 0.00017423693740300052,
+      "loss": 1.2049,
+      "step": 2500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 19340,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 250,
+  "total_flos": 7.247812094145331e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}