{
  "best_metric": 0.028574174270033836,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold1/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": 1.91911780834198,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.7638,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 2.2050745487213135,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.7203,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 2.5624966621398926,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.5894,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 2.5878520011901855,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.2764,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 1.0483590364456177,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.7547,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.6286114454269409,
      "eval_runtime": 174.249,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 6.634,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.6310367584228516,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5908,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 0.8355681896209717,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4822,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.49773526191711426,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.2557,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.5860103964805603,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.2361,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.724571704864502,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.1561,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.1106639876961708,
      "eval_runtime": 174.5465,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.772670567035675,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1137,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.3965820372104645,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1185,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 0.6853440403938293,
      "learning_rate": 4e-05,
      "loss": 0.0844,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.6801953911781311,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0831,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.6663649678230286,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0844,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.06864361464977264,
      "eval_runtime": 174.5434,
      "eval_samples_per_second": 6.623,
      "eval_steps_per_second": 6.623,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.725919783115387,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0828,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 0.5786092877388,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0739,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.7779836654663086,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0609,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.5991057753562927,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0565,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.529734194278717,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.072,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.053990062326192856,
      "eval_runtime": 174.4709,
      "eval_samples_per_second": 6.626,
      "eval_steps_per_second": 6.626,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.3646191358566284,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.058,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.4933874309062958,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0547,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.4272030293941498,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0489,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.5146765112876892,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0552,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.3383745849132538,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0458,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.05263851210474968,
      "eval_runtime": 174.3561,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.3861671984195709,
      "learning_rate": 8e-05,
      "loss": 0.0502,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.6642241477966309,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0499,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.6194387674331665,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0532,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.2517557740211487,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0577,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.5964483618736267,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0511,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.04897654429078102,
      "eval_runtime": 174.3622,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4947182238101959,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0535,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 0.3285829722881317,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0561,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.5888502597808838,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.054,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.498468816280365,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0489,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.42100989818573,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0479,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.0436147004365921,
      "eval_runtime": 174.4364,
      "eval_samples_per_second": 6.627,
      "eval_steps_per_second": 6.627,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.2836369574069977,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0588,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.5293593406677246,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0507,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.2520653009414673,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0431,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.3996206223964691,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0359,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.30496442317962646,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0648,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.04576963558793068,
      "eval_runtime": 174.387,
      "eval_samples_per_second": 6.629,
      "eval_steps_per_second": 6.629,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.1969025582075119,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0415,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.3383856415748596,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0475,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.5390830039978027,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0452,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.35218295454978943,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0536,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.4012250304222107,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0355,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.041032228618860245,
      "eval_runtime": 174.4197,
      "eval_samples_per_second": 6.628,
      "eval_steps_per_second": 6.628,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.21302421391010284,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0414,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.40680834650993347,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0314,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.5074829459190369,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0346,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.201634019613266,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0415,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.22618559002876282,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0335,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.03908700868487358,
      "eval_runtime": 174.3199,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.42317214608192444,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0307,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.2541219890117645,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.053,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.3020005524158478,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0602,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.29475560784339905,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0396,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.2962871789932251,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0478,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.035114649683237076,
      "eval_runtime": 174.3456,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 6.631,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.26232391595840454,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0351,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.415580689907074,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0301,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3246370255947113,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0344,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.2644323706626892,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.052,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.2802543342113495,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0332,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.03356001526117325,
      "eval_runtime": 174.3516,
      "eval_samples_per_second": 6.63,
      "eval_steps_per_second": 6.63,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.2146507054567337,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0343,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.32996946573257446,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0403,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.3721659481525421,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0424,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.22549860179424286,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0321,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.3647274076938629,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0275,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.03331815451383591,
      "eval_runtime": 174.566,
      "eval_samples_per_second": 6.622,
      "eval_steps_per_second": 6.622,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.3034367263317108,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0285,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.3316395878791809,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0321,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.16369342803955078,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0325,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.3651125431060791,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0398,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.2514151334762573,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0361,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03538798168301582,
      "eval_runtime": 174.7603,
      "eval_samples_per_second": 6.615,
      "eval_steps_per_second": 6.615,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.26945796608924866,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0282,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.20300740003585815,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0296,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.23300029337406158,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0288,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.36307772994041443,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0406,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.19166801869869232,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0303,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.03269141912460327,
      "eval_runtime": 174.945,
      "eval_samples_per_second": 6.608,
      "eval_steps_per_second": 6.608,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.16024506092071533,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0413,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.24341633915901184,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0339,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.20774884521961212,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0331,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.2876737117767334,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0349,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.23228268325328827,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0288,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.03259655460715294,
      "eval_runtime": 175.1441,
      "eval_samples_per_second": 6.6,
      "eval_steps_per_second": 6.6,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.24187685549259186,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0329,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.29656919836997986,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.028,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.335854172706604,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0328,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.3435249328613281,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0279,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.3224927484989166,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0316,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.033617354929447174,
      "eval_runtime": 175.3612,
      "eval_samples_per_second": 6.592,
      "eval_steps_per_second": 6.592,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.13063621520996094,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0315,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.2139972597360611,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0289,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.18014025688171387,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0261,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.17632456123828888,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0336,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.2335646003484726,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0222,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.032529912889003754,
      "eval_runtime": 175.7486,
      "eval_samples_per_second": 6.578,
      "eval_steps_per_second": 6.578,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.1565573811531067,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.029,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.11942538619041443,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0354,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.1982511579990387,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0296,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.28440022468566895,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0321,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.05921921133995056,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0373,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03450460359454155,
      "eval_runtime": 175.949,
      "eval_samples_per_second": 6.57,
      "eval_steps_per_second": 6.57,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.2323029637336731,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0304,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.07920199632644653,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0204,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.20067308843135834,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0284,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.18132272362709045,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0305,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.33372020721435547,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0335,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.0324813537299633,
      "eval_runtime": 176.0437,
      "eval_samples_per_second": 6.567,
      "eval_steps_per_second": 6.567,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.13886448740959167,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0339,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.22245754301548004,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0323,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.13712306320667267,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0277,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.21789470314979553,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0283,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.19365765154361725,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0316,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03077256865799427,
      "eval_runtime": 176.1271,
      "eval_samples_per_second": 6.563,
      "eval_steps_per_second": 6.563,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.1837930977344513,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0282,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.1351374238729477,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0307,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.13289318978786469,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0233,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.3027450442314148,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0282,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.18188823759555817,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0317,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.031701479107141495,
      "eval_runtime": 176.2605,
      "eval_samples_per_second": 6.558,
      "eval_steps_per_second": 6.558,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.575790286064148,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0345,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.14114218950271606,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0255,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.20832759141921997,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0381,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.10318358987569809,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0461,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.19055433571338654,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0214,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.03095921128988266,
      "eval_runtime": 176.3888,
      "eval_samples_per_second": 6.554,
      "eval_steps_per_second": 6.554,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.12193877249956131,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0301,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.1311451941728592,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0237,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.14820480346679688,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0238,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.2328910231590271,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0286,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.13607022166252136,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0203,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.031919609755277634,
      "eval_runtime": 177.0513,
      "eval_samples_per_second": 6.529,
      "eval_steps_per_second": 6.529,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.5148403644561768,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0271,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.19450101256370544,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0242,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.08071769773960114,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0296,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.2114274501800537,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0364,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.22326155006885529,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.027,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.029949208721518517,
      "eval_runtime": 177.2572,
      "eval_samples_per_second": 6.522,
      "eval_steps_per_second": 6.522,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.3004878759384155,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0301,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.21572935581207275,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0295,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.11787977069616318,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0224,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.15495480597019196,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0304,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.1819297820329666,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0202,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.02913297526538372,
      "eval_runtime": 177.2995,
      "eval_samples_per_second": 6.52,
      "eval_steps_per_second": 6.52,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.2882908582687378,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0222,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.22883860766887665,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.0227,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.11047745496034622,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0238,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.10832877457141876,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0214,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.16084590554237366,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0184,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.02983141504228115,
      "eval_runtime": 177.4605,
      "eval_samples_per_second": 6.514,
      "eval_steps_per_second": 6.514,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.15730451047420502,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0214,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.2245597094297409,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0182,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.2549389600753784,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.015,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.1015252098441124,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0171,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.505249559879303,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0297,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.03001730516552925,
      "eval_runtime": 177.5138,
      "eval_samples_per_second": 6.512,
      "eval_steps_per_second": 6.512,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.0905173271894455,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0161,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.20880304276943207,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.02,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.17935968935489655,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0133,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.20665256679058075,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0237,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.10122594982385635,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.015,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.033047374337911606,
      "eval_runtime": 177.5437,
      "eval_samples_per_second": 6.511,
      "eval_steps_per_second": 6.511,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.17196600139141083,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0201,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.1958017349243164,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0201,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.17364785075187683,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0181,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.07149443030357361,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0227,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.3328922390937805,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0205,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.030856935307383537,
      "eval_runtime": 177.4879,
      "eval_samples_per_second": 6.513,
      "eval_steps_per_second": 6.513,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.2552804946899414,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0217,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.13563519716262817,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0217,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.12015320360660553,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0171,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.22560076415538788,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0189,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.14802168309688568,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0169,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.03161758929491043,
      "eval_runtime": 177.5857,
      "eval_samples_per_second": 6.51,
      "eval_steps_per_second": 6.51,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.1654222160577774,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0142,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.24380093812942505,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0196,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.12394367903470993,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0207,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.21143899857997894,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0179,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.3284677565097809,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0155,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.032331354916095734,
      "eval_runtime": 177.7288,
      "eval_samples_per_second": 6.504,
      "eval_steps_per_second": 6.504,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.1364264339208603,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.017,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.1283983588218689,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0233,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.10850122570991516,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.019,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.26383689045906067,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.022,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.2121564894914627,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0216,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.029794415459036827,
      "eval_runtime": 177.7907,
      "eval_samples_per_second": 6.502,
      "eval_steps_per_second": 6.502,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.28200432658195496,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0168,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.2761673033237457,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0209,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.2146574705839157,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.022,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.25720420479774475,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0196,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.20952045917510986,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.016,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.028928693383932114,
      "eval_runtime": 178.0224,
      "eval_samples_per_second": 6.494,
      "eval_steps_per_second": 6.494,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.2755894362926483,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0207,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.20959150791168213,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0281,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.29663607478141785,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0203,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.11737760901451111,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0155,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.13971927762031555,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0255,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.029521094635128975,
      "eval_runtime": 178.0494,
      "eval_samples_per_second": 6.493,
      "eval_steps_per_second": 6.493,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.08941899240016937,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0151,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.2656846046447754,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0228,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.25469648838043213,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0283,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.1437469720840454,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0187,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.1965341567993164,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0212,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.029723171144723892,
      "eval_runtime": 178.1547,
      "eval_samples_per_second": 6.489,
      "eval_steps_per_second": 6.489,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.15759004652500153,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0178,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.2479798048734665,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0239,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.39559921622276306,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0199,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.16699132323265076,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0221,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.19198571145534515,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0236,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.028574174270033836,
      "eval_runtime": 178.5138,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 0.2197525054216385,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0168,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.2501000463962555,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0253,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.2751866579055786,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0146,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.1775154322385788,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0179,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.24539782106876373,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0184,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.02906928025186062,
      "eval_runtime": 178.368,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.24070227146148682,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.021,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.19709298014640808,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0196,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.09389526396989822,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0177,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.14415249228477478,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0151,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.1684339940547943,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0255,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.02895752526819706,
      "eval_runtime": 178.3448,
      "eval_samples_per_second": 6.482,
      "eval_steps_per_second": 6.482,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.07023809105157852,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0087,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.13149429857730865,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.0114,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.21794962882995605,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0129,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.1960616111755371,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.011,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.15295001864433289,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0122,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.029189225286245346,
      "eval_runtime": 178.2572,
      "eval_samples_per_second": 6.485,
      "eval_steps_per_second": 6.485,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 0.9009066224098206,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0116,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.10128714889287949,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0142,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.047336652874946594,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0121,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.09581335633993149,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0103,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.1117752268910408,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0108,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.030642051249742508,
      "eval_runtime": 178.1861,
      "eval_samples_per_second": 6.488,
      "eval_steps_per_second": 6.488,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.14221416413784027,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0086,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.227350115776062,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0093,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.276828408241272,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0118,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.23060496151447296,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0119,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.2514233887195587,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0087,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.032412089407444,
      "eval_runtime": 178.2179,
      "eval_samples_per_second": 6.486,
      "eval_steps_per_second": 6.486,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.21293283998966217,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0126,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.15282544493675232,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.01,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.23623453080654144,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0123,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.1641853004693985,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.008,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.2445630431175232,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.009,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.03302958607673645,
      "eval_runtime": 178.3484,
      "eval_samples_per_second": 6.482,
      "eval_steps_per_second": 6.482,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.191659614443779,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0114,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.3134976029396057,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0158,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.11693289875984192,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0085,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.3104563057422638,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0082,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.13702411949634552,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0149,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.0316716767847538,
      "eval_runtime": 178.2316,
      "eval_samples_per_second": 6.486,
      "eval_steps_per_second": 6.486,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.14635716378688812,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0122,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.2247380018234253,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0093,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.24413704872131348,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0137,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.11154405027627945,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0126,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.20430442690849304,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0127,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.03190552815794945,
      "eval_runtime": 178.3234,
      "eval_samples_per_second": 6.483,
      "eval_steps_per_second": 6.483,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.10168766230344772,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.01,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.12406831234693527,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0079,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.13969552516937256,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0114,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.21292313933372498,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0088,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.2173164039850235,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0139,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.032915081828832626,
      "eval_runtime": 178.2658,
      "eval_samples_per_second": 6.485,
      "eval_steps_per_second": 6.485,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.14659535884857178,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0193,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.3260945975780487,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0135,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.11929278820753098,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0111,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.1494680792093277,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0136,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.25308865308761597,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.008,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.03248561918735504,
      "eval_runtime": 178.3701,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.23371288180351257,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0159,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.18893377482891083,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0091,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.19870726764202118,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0097,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.3336198329925537,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0129,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.28578025102615356,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0118,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.03351206332445145,
      "eval_runtime": 178.4618,
      "eval_samples_per_second": 6.478,
      "eval_steps_per_second": 6.478,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.3268282413482666,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0081,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.1091109961271286,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0083,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.1856333613395691,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0106,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.2309342324733734,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0078,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.12330292165279388,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.008,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.03310447931289673,
      "eval_runtime": 178.3143,
      "eval_samples_per_second": 6.483,
      "eval_steps_per_second": 6.483,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.05731755122542381,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0077,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.11073189228773117,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0072,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.10200586915016174,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0134,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.17372234165668488,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0142,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.21851395070552826,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0092,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.033498842269182205,
      "eval_runtime": 178.4058,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.05207385495305061,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0141,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.22672376036643982,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.006,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.014472682029008865,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0066,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.15640006959438324,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0059,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.260066419839859,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.01,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.0338662825524807,
      "eval_runtime": 178.3673,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.2558554410934448,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0087,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.31148645281791687,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0084,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.05014743283390999,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0119,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.48116248846054077,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0091,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.3713391423225403,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0128,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.03427472338080406,
      "eval_runtime": 178.4309,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.07122137397527695,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0045,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.2939665615558624,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0038,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.12382407486438751,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0036,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.17299573123455048,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0085,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.11822900176048279,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0028,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.035295337438583374,
      "eval_runtime": 178.4312,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.326088547706604,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0068,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.07961667329072952,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0033,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.14925013482570648,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0039,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.07986024022102356,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0063,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.06009113788604736,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0072,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.03678782284259796,
      "eval_runtime": 178.4786,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.045810941606760025,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0055,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.1355172097682953,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0072,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.2088068723678589,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0033,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.15902523696422577,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0055,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.10189412534236908,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.004,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.037185050547122955,
      "eval_runtime": 178.5413,
      "eval_samples_per_second": 6.475,
      "eval_steps_per_second": 6.475,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.10047200322151184,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0028,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.210731640458107,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0043,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.12152950465679169,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0046,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.12959113717079163,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0046,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.07909118384122849,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.004,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.038275256752967834,
      "eval_runtime": 178.4695,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.1813175082206726,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0045,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.04787240922451019,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0033,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.10137149691581726,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0032,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.2671409845352173,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.005,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.13451577723026276,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0042,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.0391458235681057,
      "eval_runtime": 178.5306,
      "eval_samples_per_second": 6.475,
      "eval_steps_per_second": 6.475,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.23198837041854858,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0044,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.05521807819604874,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0028,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.040397126227617264,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0073,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.021665485575795174,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0093,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.34407833218574524,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0047,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.03926192969083786,
      "eval_runtime": 178.5053,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.16064494848251343,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0079,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.08187097311019897,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.005,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.09601439535617828,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0066,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.3348727226257324,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0062,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.08337458223104477,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0029,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.039283424615859985,
      "eval_runtime": 178.5158,
      "eval_samples_per_second": 6.476,
      "eval_steps_per_second": 6.476,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.24738673865795135,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0031,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.07796851545572281,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0063,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.02726847492158413,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.004,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.21389120817184448,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0041,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.10532323271036148,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0043,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.03927818313241005,
      "eval_runtime": 178.5353,
      "eval_samples_per_second": 6.475,
      "eval_steps_per_second": 6.475,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.10616503655910492,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0054,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.22982020676136017,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0063,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.3210495114326477,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0066,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.3766602575778961,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0053,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.010918395593762398,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.005,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.03917413577437401,
      "eval_runtime": 178.6674,
      "eval_samples_per_second": 6.47,
      "eval_steps_per_second": 6.47,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.06418674439191818,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0038,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.24328982830047607,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0049,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.32226595282554626,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0076,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.21441389620304108,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0035,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.1904405653476715,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0065,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.03925997018814087,
      "eval_runtime": 178.6602,
      "eval_samples_per_second": 6.47,
      "eval_steps_per_second": 6.47,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.133952334523201,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0077,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.0544293038547039,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0051,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.05452466011047363,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0057,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.05347217991948128,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0036,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.22037625312805176,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.01,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.039222352206707,
      "eval_runtime": 178.5733,
      "eval_samples_per_second": 6.474,
      "eval_steps_per_second": 6.474,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.06674408167600632,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0055,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.3585244119167328,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0049,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.13992421329021454,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0037,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.1115778237581253,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0049,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.04023246467113495,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0031,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.03929106518626213,
      "eval_runtime": 178.6627,
      "eval_samples_per_second": 6.47,
      "eval_steps_per_second": 6.47,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.17467571794986725,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0056,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.053972214460372925,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.003,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.4406251907348633,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0059,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.31173989176750183,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0044,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 8.777347846375342e+17,
      "train_loss": 0.04924881559766496,
      "train_runtime": 36995.9986,
      "train_samples_per_second": 1.405,
      "train_steps_per_second": 0.088
    }
  ],
  "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": 8.777347846375342e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}