{
  "best_metric": 0.031198250129818916,
  "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold9/checkpoint-1850",
  "epoch": 4.994228549442093,
  "eval_steps": 50,
  "global_step": 3245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015390534821085032,
      "grad_norm": 3.7126495838165283,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.9976,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 4.087608337402344,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.9289,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.9696438312530518,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.7183,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 3.3904402256011963,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.3793,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 2.1097986698150635,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8724,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.6930168867111206,
      "eval_runtime": 153.845,
      "eval_samples_per_second": 7.514,
      "eval_steps_per_second": 7.514,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.8988815546035767,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5891,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.223275065422058,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.407,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.5957785844802856,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.2129,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.9629207253456116,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.218,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.6734359860420227,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1565,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.10799121856689453,
      "eval_runtime": 153.9089,
      "eval_samples_per_second": 7.511,
      "eval_steps_per_second": 7.511,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8578782677650452,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1103,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8439019322395325,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1095,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.261668086051941,
      "learning_rate": 4e-05,
      "loss": 0.0785,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.9868234395980835,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0838,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.7067152261734009,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0877,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.07301879674196243,
      "eval_runtime": 153.7871,
      "eval_samples_per_second": 7.517,
      "eval_steps_per_second": 7.517,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7437458634376526,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0826,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 1.305187702178955,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0755,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.7209121584892273,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0598,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.6427532434463501,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0519,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5510497689247131,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0732,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.05595707520842552,
      "eval_runtime": 153.5914,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.6868012547492981,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0653,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.7613152861595154,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0615,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.34094569087028503,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0462,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.7298648357391357,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0586,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.503629207611084,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0544,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.05310456082224846,
      "eval_runtime": 153.5432,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.37196189165115356,
      "learning_rate": 8e-05,
      "loss": 0.0524,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.7513332366943359,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0522,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.5176823735237122,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0585,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.3480306565761566,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0623,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.4438271224498749,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0517,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04978759214282036,
      "eval_runtime": 153.6339,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.45750996470451355,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0583,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 1.0053805112838745,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.063,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.32346051931381226,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0562,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.45163822174072266,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.053,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.4021495282649994,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0468,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04682430624961853,
      "eval_runtime": 153.7949,
      "eval_samples_per_second": 7.517,
      "eval_steps_per_second": 7.517,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.2811560332775116,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0587,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.4460107386112213,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0445,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.38643085956573486,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0453,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.6017974615097046,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0418,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.2464153915643692,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.065,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.045511145144701004,
      "eval_runtime": 153.6925,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.17885592579841614,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0439,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.47445446252822876,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0534,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.5449044108390808,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0475,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.34884852170944214,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0605,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.2872726023197174,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0349,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.043942831456661224,
      "eval_runtime": 153.5486,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.3480148911476135,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0363,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.24586477875709534,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0303,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.45910361409187317,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0352,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.3279295563697815,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0393,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.29057052731513977,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0292,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.040697645395994186,
      "eval_runtime": 153.4966,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.5438590049743652,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0326,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.19197799265384674,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0497,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.39244726300239563,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0629,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.37133127450942993,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0417,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.22155240178108215,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0477,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03853714093565941,
      "eval_runtime": 153.7008,
      "eval_samples_per_second": 7.521,
      "eval_steps_per_second": 7.521,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.20890064537525177,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0326,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.34260958433151245,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0317,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.352001428604126,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0383,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.28014665842056274,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.054,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.30864518880844116,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0372,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.037615351378917694,
      "eval_runtime": 153.5805,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.250641793012619,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0359,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.19472698867321014,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0399,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.36603400111198425,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0437,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.8467999696731567,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0358,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.6848512291908264,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0309,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03592747822403908,
      "eval_runtime": 153.5608,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.45682844519615173,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0369,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.45806455612182617,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0379,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.29280415177345276,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.036,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.44070157408714294,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0475,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.22153230011463165,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0363,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.038189180195331573,
      "eval_runtime": 153.5273,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.25434547662734985,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0313,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.1719035804271698,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0266,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.25739869475364685,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0319,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.515324592590332,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0396,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.251718133687973,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0306,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03562772646546364,
      "eval_runtime": 153.6109,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.24031510949134827,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0373,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.23968929052352905,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0398,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.24071623384952545,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0316,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.3068773150444031,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0341,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.2881113588809967,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0318,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03550774231553078,
      "eval_runtime": 153.4758,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.18504729866981506,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0354,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.20161668956279755,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0264,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.3000791668891907,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0308,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.4419153332710266,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.03,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.5741679072380066,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0339,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03561116009950638,
      "eval_runtime": 153.5069,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.20193594694137573,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0335,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.31052979826927185,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.03,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.13428319990634918,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0279,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.21019649505615234,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0354,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.23262323439121246,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0209,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03455716371536255,
      "eval_runtime": 153.6557,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.22337669134140015,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0283,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.2844817340373993,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0352,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.24523243308067322,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0277,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.4137607514858246,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.037,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.07586589455604553,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0399,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.034909505397081375,
      "eval_runtime": 153.455,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2908670902252197,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0286,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.08356165885925293,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0243,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.300259530544281,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0363,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.18209409713745117,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0339,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.2502767741680145,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0326,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.033195313066244125,
      "eval_runtime": 153.6191,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.2148541510105133,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0346,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.2552015483379364,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0405,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.15273122489452362,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0312,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.12820535898208618,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.033,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.24438832700252533,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0331,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03342430666089058,
      "eval_runtime": 153.5577,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.19123180210590363,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0275,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.16545648872852325,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0277,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.16582709550857544,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0235,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.5007317066192627,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.031,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.17510643601417542,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0324,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.03433471545577049,
      "eval_runtime": 153.4796,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.6270409226417542,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0384,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.19247688353061676,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.027,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.25986501574516296,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0401,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.15622414648532867,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0477,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.20874176919460297,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0213,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.03305734321475029,
      "eval_runtime": 153.4439,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.1636982411146164,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0294,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.14659443497657776,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0238,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.19172105193138123,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0254,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.3553459942340851,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0373,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.09117433428764343,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0217,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.0347236767411232,
      "eval_runtime": 153.4161,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.2408464252948761,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0287,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.15327930450439453,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0269,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.11014704406261444,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0307,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.2290666401386261,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0351,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.3346557319164276,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0289,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.03386509791016579,
      "eval_runtime": 153.5881,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.3548568785190582,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0332,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.21676219999790192,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0281,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.16335730254650116,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0238,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.21953915059566498,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0325,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.16354408860206604,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0227,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.03279327601194382,
      "eval_runtime": 153.3451,
      "eval_samples_per_second": 7.539,
      "eval_steps_per_second": 7.539,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.2507956922054291,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0213,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.2832318842411041,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0243,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.14541825652122498,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.022,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.14678367972373962,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0202,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.27238062024116516,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0163,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.033340852707624435,
      "eval_runtime": 153.5167,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.11591003090143204,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0214,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.4332759380340576,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0182,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.258517861366272,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0139,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.21928441524505615,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.016,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.6055142283439636,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0288,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.034971561282873154,
      "eval_runtime": 153.4188,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.09305189549922943,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0158,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.3016238808631897,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0255,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.3424316942691803,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0151,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.1434495598077774,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.021,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.11851069331169128,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0139,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.036287665367126465,
      "eval_runtime": 153.4724,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.22188839316368103,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0207,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.17747120559215546,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0225,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.25305667519569397,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0236,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.1354154497385025,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0243,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.29097962379455566,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0205,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.033679693937301636,
      "eval_runtime": 153.468,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.3821086883544922,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.022,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.18589210510253906,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0227,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.220177561044693,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0188,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.29140767455101013,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0186,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.20807486772537231,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0164,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.035624731332063675,
      "eval_runtime": 153.5255,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.23355218768119812,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0134,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.3649499714374542,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0248,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.11541513353586197,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.021,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.19494101405143738,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0192,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.22867737710475922,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0147,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.033948931843042374,
      "eval_runtime": 153.4734,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.22062183916568756,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.016,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.1727512925863266,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0222,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.13478009402751923,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0177,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.34471556544303894,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.02,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.33221709728240967,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0223,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.03265100717544556,
      "eval_runtime": 153.5117,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.4092860519886017,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0171,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.33218109607696533,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0187,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.24247807264328003,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0252,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.23863863945007324,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0226,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.3615785241127014,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0179,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03302457183599472,
      "eval_runtime": 153.4755,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.25043681263923645,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.019,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.18690428137779236,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0307,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.41168877482414246,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0242,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.13710512220859528,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0175,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.16903425753116608,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0261,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03208105266094208,
      "eval_runtime": 153.6694,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.28284743428230286,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.015,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.34749895334243774,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0252,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.40931135416030884,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.024,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.08158578723669052,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0192,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.2976609170436859,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0235,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03284631296992302,
      "eval_runtime": 153.4294,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.23181110620498657,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0198,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.27695050835609436,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0254,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.7658836841583252,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0249,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.2545831501483917,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0258,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.2002500295639038,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0238,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.031198250129818916,
      "eval_runtime": 153.5667,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.28489193320274353,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0143,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.31040656566619873,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0253,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.6301350593566895,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0141,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.21027325093746185,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0179,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.29681673645973206,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0181,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03257755935192108,
      "eval_runtime": 153.4962,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.32226938009262085,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0234,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.3540765941143036,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0184,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.16609501838684082,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0173,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.17475761473178864,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0149,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.23754218220710754,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0279,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03226524963974953,
      "eval_runtime": 153.4723,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.10268258303403854,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0068,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.22028988599777222,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.011,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.2139747440814972,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0099,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.3226635456085205,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0091,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.1942538321018219,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0106,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03540215268731117,
      "eval_runtime": 153.5721,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 2.1021478176116943,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0075,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.05760245770215988,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0092,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.2452974021434784,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0099,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.045641012489795685,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0076,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.42202338576316833,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.009,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.03886837139725685,
      "eval_runtime": 153.6385,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.09824363142251968,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.007,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.42016705870628357,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0089,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.26800215244293213,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0094,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.3631320893764496,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.01,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.30696097016334534,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0071,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.039117343723773956,
      "eval_runtime": 153.4025,
      "eval_samples_per_second": 7.536,
      "eval_steps_per_second": 7.536,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.10165564715862274,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0093,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.3360536992549896,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.011,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.4711909890174866,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0081,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.20146417617797852,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0067,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.3740951120853424,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0098,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.038514357060194016,
      "eval_runtime": 153.4219,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.1507674902677536,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0099,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.4039727747440338,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0144,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.03309769183397293,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0098,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.7046623229980469,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0092,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.13739725947380066,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0161,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.03872603923082352,
      "eval_runtime": 153.4475,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.25823473930358887,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0105,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.2748499810695648,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0082,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.09509546309709549,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0127,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.27710622549057007,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0129,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.3034014403820038,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0094,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.03822940215468407,
      "eval_runtime": 153.4571,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.1106347069144249,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0096,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.4596875011920929,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0066,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.24233205616474152,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0134,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.4554457366466522,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0076,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.23505744338035583,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0099,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.038470253348350525,
      "eval_runtime": 153.4782,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.1141432598233223,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0133,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.49946677684783936,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.013,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.14206841588020325,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0084,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.24162252247333527,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0117,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.09814824908971786,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0053,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.03881305456161499,
      "eval_runtime": 153.4737,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.2978641390800476,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0133,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.3559443950653076,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0095,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.18104760348796844,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0091,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.24279925227165222,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0113,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.15692494809627533,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0122,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.0382477305829525,
      "eval_runtime": 153.525,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.2750367820262909,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0104,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.25488975644111633,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0069,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.25898972153663635,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0086,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.5144363045692444,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0085,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.14448030292987823,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0077,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.03859272971749306,
      "eval_runtime": 153.3812,
      "eval_samples_per_second": 7.537,
      "eval_steps_per_second": 7.537,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.1518450677394867,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0099,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.10527992248535156,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0049,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.1166950911283493,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0096,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.28515174984931946,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0111,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.23182082176208496,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0082,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.039548709988594055,
      "eval_runtime": 153.4546,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.04112737998366356,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0098,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.5165395736694336,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.004,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.08130805939435959,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0065,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.45289987325668335,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0068,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.515565812587738,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0122,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.040204811841249466,
      "eval_runtime": 153.4568,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.3798106014728546,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.007,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.6449206471443176,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0052,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 1.285163402557373,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0142,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.4362947344779968,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0081,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.0601702556014061,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0089,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.040835246443748474,
      "eval_runtime": 153.6576,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.16174668073654175,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0047,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.2250216007232666,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0018,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.08969392627477646,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.002,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.06381311267614365,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0048,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.03750935569405556,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.002,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.04213506728410721,
      "eval_runtime": 153.4238,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.17056500911712646,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0048,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.5444008708000183,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.003,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.06367257237434387,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0022,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.07164566218852997,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0044,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.04249047115445137,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0048,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.04472676292061806,
      "eval_runtime": 153.5005,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.07885459065437317,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0049,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.1322193592786789,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.004,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.05429103225469589,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0042,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.4150710999965668,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0054,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.10260337591171265,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0021,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.04568769037723541,
      "eval_runtime": 153.4731,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.05860987678170204,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0018,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.03835854306817055,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0019,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.28779852390289307,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0023,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.0977441594004631,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0035,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.028815653175115585,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.004,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04634019732475281,
      "eval_runtime": 153.3933,
      "eval_samples_per_second": 7.536,
      "eval_steps_per_second": 7.536,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.03354586660861969,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0025,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.027734043076634407,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0027,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.04675295576453209,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0025,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.48769840598106384,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0036,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.09698507934808731,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0035,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.047643765807151794,
      "eval_runtime": 153.4342,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.20398098230361938,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0017,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.07381857186555862,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0015,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.06577713042497635,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0077,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.026341058313846588,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0016,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.2532847225666046,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0046,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04775525629520416,
      "eval_runtime": 153.391,
      "eval_samples_per_second": 7.536,
      "eval_steps_per_second": 7.536,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.0765874832868576,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0075,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.04009735584259033,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0042,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.03760379180312157,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0024,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.46977555751800537,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0035,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.01115656178444624,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0019,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.0476050078868866,
      "eval_runtime": 153.4015,
      "eval_samples_per_second": 7.536,
      "eval_steps_per_second": 7.536,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.0846967101097107,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0007,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.3001258075237274,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0033,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.026126157492399216,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0024,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.3240077793598175,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0031,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.36474794149398804,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0031,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.047984521836042404,
      "eval_runtime": 153.4315,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.08020499348640442,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0018,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.05722915008664131,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0029,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.22595173120498657,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0036,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.6162356734275818,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0034,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.008868547156453133,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.003,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.04777419939637184,
      "eval_runtime": 153.4819,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.11587874591350555,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.002,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.5329419374465942,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0032,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.5119339823722839,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0035,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.05603371188044548,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.003,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.05833933502435684,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.002,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.048142001032829285,
      "eval_runtime": 153.4684,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.11830338090658188,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.005,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.2384611815214157,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0019,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.006224968936294317,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0035,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.006869935430586338,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0022,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.34784266352653503,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0064,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.048092927783727646,
      "eval_runtime": 153.4737,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.06563670933246613,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0048,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.2973856031894684,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0024,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.10954169929027557,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0025,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.11312809586524963,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0024,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.03980771079659462,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0022,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.04806261882185936,
      "eval_runtime": 153.4574,
      "eval_samples_per_second": 7.533,
      "eval_steps_per_second": 7.533,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.127711683511734,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0041,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.019069330766797066,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0026,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.4193945825099945,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0054,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.11006677895784378,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0036,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 7.696760527880847e+17,
      "train_loss": 0.051293578231491964,
      "train_runtime": 36396.8521,
      "train_samples_per_second": 1.428,
      "train_steps_per_second": 0.089
    }
  ],
  "logging_steps": 10,
  "max_steps": 3245,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.696760527880847e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}