{
  "best_metric": 0.03199377283453941,
  "best_model_checkpoint": "saves/psy-course/Qwen2.5-7B-Instruct/train/fold7/checkpoint-1750",
  "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": 3.7242681980133057,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.9983,
      "step": 10
    },
    {
      "epoch": 0.030781069642170065,
      "grad_norm": 4.060908794403076,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.9251,
      "step": 20
    },
    {
      "epoch": 0.0461716044632551,
      "grad_norm": 3.9718127250671387,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.7145,
      "step": 30
    },
    {
      "epoch": 0.06156213928434013,
      "grad_norm": 3.4009063243865967,
      "learning_rate": 1.230769230769231e-05,
      "loss": 1.3802,
      "step": 40
    },
    {
      "epoch": 0.07695267410542517,
      "grad_norm": 2.0971736907958984,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.8743,
      "step": 50
    },
    {
      "epoch": 0.07695267410542517,
      "eval_loss": 0.6952459812164307,
      "eval_runtime": 152.6263,
      "eval_samples_per_second": 7.574,
      "eval_steps_per_second": 7.574,
      "step": 50
    },
    {
      "epoch": 0.0923432089265102,
      "grad_norm": 0.893717885017395,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.5914,
      "step": 60
    },
    {
      "epoch": 0.10773374374759523,
      "grad_norm": 1.2102733850479126,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.4083,
      "step": 70
    },
    {
      "epoch": 0.12312427856868026,
      "grad_norm": 0.6031086444854736,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.2135,
      "step": 80
    },
    {
      "epoch": 0.1385148133897653,
      "grad_norm": 0.9773842096328735,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.2181,
      "step": 90
    },
    {
      "epoch": 0.15390534821085033,
      "grad_norm": 0.6637694239616394,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.156,
      "step": 100
    },
    {
      "epoch": 0.15390534821085033,
      "eval_loss": 0.10799720883369446,
      "eval_runtime": 152.9511,
      "eval_samples_per_second": 7.558,
      "eval_steps_per_second": 7.558,
      "step": 100
    },
    {
      "epoch": 0.16929588303193535,
      "grad_norm": 0.8486623764038086,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.1094,
      "step": 110
    },
    {
      "epoch": 0.1846864178530204,
      "grad_norm": 0.8181426525115967,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.1091,
      "step": 120
    },
    {
      "epoch": 0.2000769526741054,
      "grad_norm": 1.299336552619934,
      "learning_rate": 4e-05,
      "loss": 0.079,
      "step": 130
    },
    {
      "epoch": 0.21546748749519046,
      "grad_norm": 0.9502790570259094,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0835,
      "step": 140
    },
    {
      "epoch": 0.2308580223162755,
      "grad_norm": 0.7055000066757202,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0872,
      "step": 150
    },
    {
      "epoch": 0.2308580223162755,
      "eval_loss": 0.07323972135782242,
      "eval_runtime": 153.062,
      "eval_samples_per_second": 7.552,
      "eval_steps_per_second": 7.552,
      "step": 150
    },
    {
      "epoch": 0.24624855713736052,
      "grad_norm": 0.7500439286231995,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0828,
      "step": 160
    },
    {
      "epoch": 0.26163909195844554,
      "grad_norm": 1.3891620635986328,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0752,
      "step": 170
    },
    {
      "epoch": 0.2770296267795306,
      "grad_norm": 0.7466645240783691,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0595,
      "step": 180
    },
    {
      "epoch": 0.2924201616006156,
      "grad_norm": 0.6291409134864807,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0522,
      "step": 190
    },
    {
      "epoch": 0.30781069642170067,
      "grad_norm": 0.5699347853660583,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0732,
      "step": 200
    },
    {
      "epoch": 0.30781069642170067,
      "eval_loss": 0.055596888065338135,
      "eval_runtime": 153.0422,
      "eval_samples_per_second": 7.553,
      "eval_steps_per_second": 7.553,
      "step": 200
    },
    {
      "epoch": 0.3232012312427857,
      "grad_norm": 0.6821591854095459,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0646,
      "step": 210
    },
    {
      "epoch": 0.3385917660638707,
      "grad_norm": 0.7783260941505432,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0616,
      "step": 220
    },
    {
      "epoch": 0.35398230088495575,
      "grad_norm": 0.319352924823761,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0452,
      "step": 230
    },
    {
      "epoch": 0.3693728357060408,
      "grad_norm": 0.6423914432525635,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0581,
      "step": 240
    },
    {
      "epoch": 0.38476337052712584,
      "grad_norm": 0.4901570677757263,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0551,
      "step": 250
    },
    {
      "epoch": 0.38476337052712584,
      "eval_loss": 0.0529809296131134,
      "eval_runtime": 153.1271,
      "eval_samples_per_second": 7.549,
      "eval_steps_per_second": 7.549,
      "step": 250
    },
    {
      "epoch": 0.4001539053482108,
      "grad_norm": 0.35566428303718567,
      "learning_rate": 8e-05,
      "loss": 0.0522,
      "step": 260
    },
    {
      "epoch": 0.41554444016929587,
      "grad_norm": 0.8170797824859619,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0512,
      "step": 270
    },
    {
      "epoch": 0.4309349749903809,
      "grad_norm": 0.506725549697876,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0575,
      "step": 280
    },
    {
      "epoch": 0.44632550981146596,
      "grad_norm": 0.38407424092292786,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0619,
      "step": 290
    },
    {
      "epoch": 0.461716044632551,
      "grad_norm": 0.4513370394706726,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0528,
      "step": 300
    },
    {
      "epoch": 0.461716044632551,
      "eval_loss": 0.0502796396613121,
      "eval_runtime": 153.1739,
      "eval_samples_per_second": 7.547,
      "eval_steps_per_second": 7.547,
      "step": 300
    },
    {
      "epoch": 0.477106579453636,
      "grad_norm": 0.4808270037174225,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0587,
      "step": 310
    },
    {
      "epoch": 0.49249711427472104,
      "grad_norm": 1.0640689134597778,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.063,
      "step": 320
    },
    {
      "epoch": 0.5078876490958061,
      "grad_norm": 0.3364812433719635,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0574,
      "step": 330
    },
    {
      "epoch": 0.5232781839168911,
      "grad_norm": 0.42063969373703003,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0523,
      "step": 340
    },
    {
      "epoch": 0.5386687187379762,
      "grad_norm": 0.401907354593277,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.046,
      "step": 350
    },
    {
      "epoch": 0.5386687187379762,
      "eval_loss": 0.04721313342452049,
      "eval_runtime": 153.2316,
      "eval_samples_per_second": 7.544,
      "eval_steps_per_second": 7.544,
      "step": 350
    },
    {
      "epoch": 0.5540592535590612,
      "grad_norm": 0.2865579426288605,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.059,
      "step": 360
    },
    {
      "epoch": 0.5694497883801463,
      "grad_norm": 0.472827672958374,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0441,
      "step": 370
    },
    {
      "epoch": 0.5848403232012312,
      "grad_norm": 0.43391329050064087,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0456,
      "step": 380
    },
    {
      "epoch": 0.6002308580223162,
      "grad_norm": 0.573106586933136,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0422,
      "step": 390
    },
    {
      "epoch": 0.6156213928434013,
      "grad_norm": 0.2550828456878662,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0625,
      "step": 400
    },
    {
      "epoch": 0.6156213928434013,
      "eval_loss": 0.045130085200071335,
      "eval_runtime": 153.2505,
      "eval_samples_per_second": 7.543,
      "eval_steps_per_second": 7.543,
      "step": 400
    },
    {
      "epoch": 0.6310119276644863,
      "grad_norm": 0.17762994766235352,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0422,
      "step": 410
    },
    {
      "epoch": 0.6464024624855714,
      "grad_norm": 0.4644079804420471,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.054,
      "step": 420
    },
    {
      "epoch": 0.6617929973066564,
      "grad_norm": 0.56646728515625,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0473,
      "step": 430
    },
    {
      "epoch": 0.6771835321277414,
      "grad_norm": 0.3449053466320038,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0608,
      "step": 440
    },
    {
      "epoch": 0.6925740669488265,
      "grad_norm": 0.2865304946899414,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0349,
      "step": 450
    },
    {
      "epoch": 0.6925740669488265,
      "eval_loss": 0.04394521191716194,
      "eval_runtime": 153.2495,
      "eval_samples_per_second": 7.543,
      "eval_steps_per_second": 7.543,
      "step": 450
    },
    {
      "epoch": 0.7079646017699115,
      "grad_norm": 0.3648393154144287,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0365,
      "step": 460
    },
    {
      "epoch": 0.7233551365909965,
      "grad_norm": 0.2595067024230957,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0304,
      "step": 470
    },
    {
      "epoch": 0.7387456714120816,
      "grad_norm": 0.49874451756477356,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0354,
      "step": 480
    },
    {
      "epoch": 0.7541362062331666,
      "grad_norm": 0.28884512186050415,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0402,
      "step": 490
    },
    {
      "epoch": 0.7695267410542517,
      "grad_norm": 0.4333209693431854,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0303,
      "step": 500
    },
    {
      "epoch": 0.7695267410542517,
      "eval_loss": 0.04045053571462631,
      "eval_runtime": 153.2372,
      "eval_samples_per_second": 7.544,
      "eval_steps_per_second": 7.544,
      "step": 500
    },
    {
      "epoch": 0.7849172758753367,
      "grad_norm": 0.49973243474960327,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0319,
      "step": 510
    },
    {
      "epoch": 0.8003078106964217,
      "grad_norm": 0.14952120184898376,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0496,
      "step": 520
    },
    {
      "epoch": 0.8156983455175068,
      "grad_norm": 0.38538721203804016,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.062,
      "step": 530
    },
    {
      "epoch": 0.8310888803385917,
      "grad_norm": 0.35999152064323425,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0416,
      "step": 540
    },
    {
      "epoch": 0.8464794151596768,
      "grad_norm": 0.24905063211917877,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0468,
      "step": 550
    },
    {
      "epoch": 0.8464794151596768,
      "eval_loss": 0.03800861909985542,
      "eval_runtime": 153.2454,
      "eval_samples_per_second": 7.543,
      "eval_steps_per_second": 7.543,
      "step": 550
    },
    {
      "epoch": 0.8618699499807618,
      "grad_norm": 0.21499799191951752,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0319,
      "step": 560
    },
    {
      "epoch": 0.8772604848018468,
      "grad_norm": 0.34962165355682373,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0321,
      "step": 570
    },
    {
      "epoch": 0.8926510196229319,
      "grad_norm": 0.3447439670562744,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0382,
      "step": 580
    },
    {
      "epoch": 0.9080415544440169,
      "grad_norm": 0.29191455245018005,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0536,
      "step": 590
    },
    {
      "epoch": 0.923432089265102,
      "grad_norm": 0.3137579560279846,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.0383,
      "step": 600
    },
    {
      "epoch": 0.923432089265102,
      "eval_loss": 0.03731103986501694,
      "eval_runtime": 153.3431,
      "eval_samples_per_second": 7.539,
      "eval_steps_per_second": 7.539,
      "step": 600
    },
    {
      "epoch": 0.938822624086187,
      "grad_norm": 0.26628851890563965,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0354,
      "step": 610
    },
    {
      "epoch": 0.954213158907272,
      "grad_norm": 0.17702606320381165,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0397,
      "step": 620
    },
    {
      "epoch": 0.9696036937283571,
      "grad_norm": 0.35290348529815674,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.0429,
      "step": 630
    },
    {
      "epoch": 0.9849942285494421,
      "grad_norm": 0.9227323532104492,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0362,
      "step": 640
    },
    {
      "epoch": 1.0003847633705272,
      "grad_norm": 0.5537252426147461,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0307,
      "step": 650
    },
    {
      "epoch": 1.0003847633705272,
      "eval_loss": 0.036164309829473495,
      "eval_runtime": 153.3321,
      "eval_samples_per_second": 7.539,
      "eval_steps_per_second": 7.539,
      "step": 650
    },
    {
      "epoch": 1.0157752981916122,
      "grad_norm": 0.44172441959381104,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0372,
      "step": 660
    },
    {
      "epoch": 1.0311658330126972,
      "grad_norm": 0.40371954441070557,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.039,
      "step": 670
    },
    {
      "epoch": 1.0465563678337821,
      "grad_norm": 0.2652777433395386,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0338,
      "step": 680
    },
    {
      "epoch": 1.0619469026548674,
      "grad_norm": 0.4269815981388092,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.048,
      "step": 690
    },
    {
      "epoch": 1.0773374374759523,
      "grad_norm": 0.2099122256040573,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0357,
      "step": 700
    },
    {
      "epoch": 1.0773374374759523,
      "eval_loss": 0.03784062713384628,
      "eval_runtime": 153.4131,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 700
    },
    {
      "epoch": 1.0927279722970373,
      "grad_norm": 0.3871307373046875,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.0318,
      "step": 710
    },
    {
      "epoch": 1.1081185071181223,
      "grad_norm": 0.1988789290189743,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.0259,
      "step": 720
    },
    {
      "epoch": 1.1235090419392073,
      "grad_norm": 0.26087290048599243,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0315,
      "step": 730
    },
    {
      "epoch": 1.1388995767602923,
      "grad_norm": 0.5976797342300415,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.038,
      "step": 740
    },
    {
      "epoch": 1.1542901115813775,
      "grad_norm": 0.2549760937690735,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.03,
      "step": 750
    },
    {
      "epoch": 1.1542901115813775,
      "eval_loss": 0.035586223006248474,
      "eval_runtime": 153.4685,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 750
    },
    {
      "epoch": 1.1696806464024625,
      "grad_norm": 0.23747926950454712,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0373,
      "step": 760
    },
    {
      "epoch": 1.1850711812235475,
      "grad_norm": 0.2744942605495453,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0388,
      "step": 770
    },
    {
      "epoch": 1.2004617160446325,
      "grad_norm": 0.20212800800800323,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.032,
      "step": 780
    },
    {
      "epoch": 1.2158522508657175,
      "grad_norm": 0.2919940650463104,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.034,
      "step": 790
    },
    {
      "epoch": 1.2312427856868027,
      "grad_norm": 0.2869492471218109,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.0303,
      "step": 800
    },
    {
      "epoch": 1.2312427856868027,
      "eval_loss": 0.035416457802057266,
      "eval_runtime": 153.6345,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 800
    },
    {
      "epoch": 1.2466333205078877,
      "grad_norm": 0.16437041759490967,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0359,
      "step": 810
    },
    {
      "epoch": 1.2620238553289727,
      "grad_norm": 0.21569447219371796,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.026,
      "step": 820
    },
    {
      "epoch": 1.2774143901500576,
      "grad_norm": 0.3021999001502991,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0312,
      "step": 830
    },
    {
      "epoch": 1.2928049249711426,
      "grad_norm": 0.45706892013549805,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0284,
      "step": 840
    },
    {
      "epoch": 1.3081954597922278,
      "grad_norm": 0.5673989057540894,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0334,
      "step": 850
    },
    {
      "epoch": 1.3081954597922278,
      "eval_loss": 0.03567519783973694,
      "eval_runtime": 153.5184,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 850
    },
    {
      "epoch": 1.3235859946133128,
      "grad_norm": 0.21376289427280426,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0326,
      "step": 860
    },
    {
      "epoch": 1.3389765294343978,
      "grad_norm": 0.333779513835907,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.0294,
      "step": 870
    },
    {
      "epoch": 1.3543670642554828,
      "grad_norm": 0.12364228069782257,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0292,
      "step": 880
    },
    {
      "epoch": 1.3697575990765678,
      "grad_norm": 0.18600425124168396,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0353,
      "step": 890
    },
    {
      "epoch": 1.385148133897653,
      "grad_norm": 0.260386198759079,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.0203,
      "step": 900
    },
    {
      "epoch": 1.385148133897653,
      "eval_loss": 0.03431593254208565,
      "eval_runtime": 153.4965,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 900
    },
    {
      "epoch": 1.400538668718738,
      "grad_norm": 0.23011595010757446,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0293,
      "step": 910
    },
    {
      "epoch": 1.415929203539823,
      "grad_norm": 0.2681461274623871,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0352,
      "step": 920
    },
    {
      "epoch": 1.431319738360908,
      "grad_norm": 0.224797785282135,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0275,
      "step": 930
    },
    {
      "epoch": 1.446710273181993,
      "grad_norm": 0.43936702609062195,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0353,
      "step": 940
    },
    {
      "epoch": 1.4621008080030782,
      "grad_norm": 0.07843472063541412,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0394,
      "step": 950
    },
    {
      "epoch": 1.4621008080030782,
      "eval_loss": 0.03500496223568916,
      "eval_runtime": 153.4389,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 950
    },
    {
      "epoch": 1.4774913428241632,
      "grad_norm": 0.30399879813194275,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.03,
      "step": 960
    },
    {
      "epoch": 1.4928818776452482,
      "grad_norm": 0.0947657898068428,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0241,
      "step": 970
    },
    {
      "epoch": 1.5082724124663334,
      "grad_norm": 0.3287079632282257,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0352,
      "step": 980
    },
    {
      "epoch": 1.5236629472874181,
      "grad_norm": 0.20038163661956787,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0345,
      "step": 990
    },
    {
      "epoch": 1.5390534821085033,
      "grad_norm": 0.23154109716415405,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0321,
      "step": 1000
    },
    {
      "epoch": 1.5390534821085033,
      "eval_loss": 0.03294072300195694,
      "eval_runtime": 153.4172,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 1000
    },
    {
      "epoch": 1.5544440169295883,
      "grad_norm": 0.19439604878425598,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0346,
      "step": 1010
    },
    {
      "epoch": 1.5698345517506733,
      "grad_norm": 0.2487119734287262,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.0394,
      "step": 1020
    },
    {
      "epoch": 1.5852250865717585,
      "grad_norm": 0.14983823895454407,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0319,
      "step": 1030
    },
    {
      "epoch": 1.6006156213928433,
      "grad_norm": 0.12596146762371063,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0322,
      "step": 1040
    },
    {
      "epoch": 1.6160061562139285,
      "grad_norm": 0.25006648898124695,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0331,
      "step": 1050
    },
    {
      "epoch": 1.6160061562139285,
      "eval_loss": 0.03370964899659157,
      "eval_runtime": 153.4104,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 1050
    },
    {
      "epoch": 1.6313966910350135,
      "grad_norm": 0.19626592099666595,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0268,
      "step": 1060
    },
    {
      "epoch": 1.6467872258560985,
      "grad_norm": 0.16402246057987213,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0283,
      "step": 1070
    },
    {
      "epoch": 1.6621777606771835,
      "grad_norm": 0.16251496970653534,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0244,
      "step": 1080
    },
    {
      "epoch": 1.6775682954982685,
      "grad_norm": 0.5160397887229919,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.03,
      "step": 1090
    },
    {
      "epoch": 1.6929588303193537,
      "grad_norm": 0.1784953773021698,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0328,
      "step": 1100
    },
    {
      "epoch": 1.6929588303193537,
      "eval_loss": 0.034384939819574356,
      "eval_runtime": 153.4141,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 7.535,
      "step": 1100
    },
    {
      "epoch": 1.7083493651404387,
      "grad_norm": 0.4627922773361206,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0377,
      "step": 1110
    },
    {
      "epoch": 1.7237398999615237,
      "grad_norm": 0.1881691813468933,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0271,
      "step": 1120
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.2602936029434204,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0402,
      "step": 1130
    },
    {
      "epoch": 1.7545209696036936,
      "grad_norm": 0.13404856622219086,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0468,
      "step": 1140
    },
    {
      "epoch": 1.7699115044247788,
      "grad_norm": 0.2135487049818039,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0213,
      "step": 1150
    },
    {
      "epoch": 1.7699115044247788,
      "eval_loss": 0.0335485003888607,
      "eval_runtime": 153.3703,
      "eval_samples_per_second": 7.537,
      "eval_steps_per_second": 7.537,
      "step": 1150
    },
    {
      "epoch": 1.7853020392458638,
      "grad_norm": 0.16438467800617218,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0296,
      "step": 1160
    },
    {
      "epoch": 1.8006925740669488,
      "grad_norm": 0.15055839717388153,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.0238,
      "step": 1170
    },
    {
      "epoch": 1.8160831088880338,
      "grad_norm": 0.15962374210357666,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0254,
      "step": 1180
    },
    {
      "epoch": 1.8314736437091188,
      "grad_norm": 0.35032787919044495,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0368,
      "step": 1190
    },
    {
      "epoch": 1.846864178530204,
      "grad_norm": 0.09017729759216309,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0216,
      "step": 1200
    },
    {
      "epoch": 1.846864178530204,
      "eval_loss": 0.035347830504179,
      "eval_runtime": 153.4758,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1200
    },
    {
      "epoch": 1.862254713351289,
      "grad_norm": 0.3004724979400635,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.0299,
      "step": 1210
    },
    {
      "epoch": 1.877645248172374,
      "grad_norm": 0.13141077756881714,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0271,
      "step": 1220
    },
    {
      "epoch": 1.893035782993459,
      "grad_norm": 0.11445621401071548,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.031,
      "step": 1230
    },
    {
      "epoch": 1.908426317814544,
      "grad_norm": 0.20827576518058777,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0346,
      "step": 1240
    },
    {
      "epoch": 1.9238168526356292,
      "grad_norm": 0.35453861951828003,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0295,
      "step": 1250
    },
    {
      "epoch": 1.9238168526356292,
      "eval_loss": 0.03391389176249504,
      "eval_runtime": 153.4937,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1250
    },
    {
      "epoch": 1.9392073874567142,
      "grad_norm": 0.35258570313453674,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.0321,
      "step": 1260
    },
    {
      "epoch": 1.9545979222777992,
      "grad_norm": 0.19581347703933716,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0284,
      "step": 1270
    },
    {
      "epoch": 1.9699884570988841,
      "grad_norm": 0.1685507595539093,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.0247,
      "step": 1280
    },
    {
      "epoch": 1.9853789919199691,
      "grad_norm": 0.21214863657951355,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0313,
      "step": 1290
    },
    {
      "epoch": 2.0007695267410543,
      "grad_norm": 0.14349879324436188,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0224,
      "step": 1300
    },
    {
      "epoch": 2.0007695267410543,
      "eval_loss": 0.032222457230091095,
      "eval_runtime": 153.5029,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1300
    },
    {
      "epoch": 2.016160061562139,
      "grad_norm": 0.24724121391773224,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0211,
      "step": 1310
    },
    {
      "epoch": 2.0315505963832243,
      "grad_norm": 0.2738001048564911,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.022,
      "step": 1320
    },
    {
      "epoch": 2.0469411312043095,
      "grad_norm": 0.13100820779800415,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.022,
      "step": 1330
    },
    {
      "epoch": 2.0623316660253943,
      "grad_norm": 0.1414738893508911,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0201,
      "step": 1340
    },
    {
      "epoch": 2.0777222008464795,
      "grad_norm": 0.2568168640136719,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0157,
      "step": 1350
    },
    {
      "epoch": 2.0777222008464795,
      "eval_loss": 0.033858660608530045,
      "eval_runtime": 153.4369,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 1350
    },
    {
      "epoch": 2.0931127356675643,
      "grad_norm": 0.12020983546972275,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0212,
      "step": 1360
    },
    {
      "epoch": 2.1085032704886495,
      "grad_norm": 0.517400324344635,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.0173,
      "step": 1370
    },
    {
      "epoch": 2.1238938053097347,
      "grad_norm": 0.251701682806015,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0141,
      "step": 1380
    },
    {
      "epoch": 2.1392843401308195,
      "grad_norm": 0.2363581359386444,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0174,
      "step": 1390
    },
    {
      "epoch": 2.1546748749519047,
      "grad_norm": 0.6419008374214172,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0288,
      "step": 1400
    },
    {
      "epoch": 2.1546748749519047,
      "eval_loss": 0.0355730764567852,
      "eval_runtime": 153.5741,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 1400
    },
    {
      "epoch": 2.1700654097729895,
      "grad_norm": 0.0650162398815155,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0156,
      "step": 1410
    },
    {
      "epoch": 2.1854559445940747,
      "grad_norm": 0.3059469759464264,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0246,
      "step": 1420
    },
    {
      "epoch": 2.20084647941516,
      "grad_norm": 0.3678857088088989,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.0163,
      "step": 1430
    },
    {
      "epoch": 2.2162370142362446,
      "grad_norm": 0.17199157178401947,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0202,
      "step": 1440
    },
    {
      "epoch": 2.23162754905733,
      "grad_norm": 0.1235027089715004,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.0142,
      "step": 1450
    },
    {
      "epoch": 2.23162754905733,
      "eval_loss": 0.03511354699730873,
      "eval_runtime": 153.5029,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1450
    },
    {
      "epoch": 2.2470180838784146,
      "grad_norm": 0.24692963063716888,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0206,
      "step": 1460
    },
    {
      "epoch": 2.2624086186995,
      "grad_norm": 0.16651657223701477,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0222,
      "step": 1470
    },
    {
      "epoch": 2.2777991535205846,
      "grad_norm": 0.25574541091918945,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0238,
      "step": 1480
    },
    {
      "epoch": 2.29318968834167,
      "grad_norm": 0.13530392944812775,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0248,
      "step": 1490
    },
    {
      "epoch": 2.308580223162755,
      "grad_norm": 0.24548520147800446,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0186,
      "step": 1500
    },
    {
      "epoch": 2.308580223162755,
      "eval_loss": 0.03335192799568176,
      "eval_runtime": 153.6065,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 1500
    },
    {
      "epoch": 2.32397075798384,
      "grad_norm": 0.2126888781785965,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0221,
      "step": 1510
    },
    {
      "epoch": 2.339361292804925,
      "grad_norm": 0.12796470522880554,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.021,
      "step": 1520
    },
    {
      "epoch": 2.3547518276260098,
      "grad_norm": 0.23904035985469818,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0203,
      "step": 1530
    },
    {
      "epoch": 2.370142362447095,
      "grad_norm": 0.2982531785964966,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0191,
      "step": 1540
    },
    {
      "epoch": 2.38553289726818,
      "grad_norm": 0.18962526321411133,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0145,
      "step": 1550
    },
    {
      "epoch": 2.38553289726818,
      "eval_loss": 0.03628704696893692,
      "eval_runtime": 153.5799,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 1550
    },
    {
      "epoch": 2.400923432089265,
      "grad_norm": 0.48935335874557495,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0139,
      "step": 1560
    },
    {
      "epoch": 2.41631396691035,
      "grad_norm": 0.4216659665107727,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0242,
      "step": 1570
    },
    {
      "epoch": 2.431704501731435,
      "grad_norm": 0.14698106050491333,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0191,
      "step": 1580
    },
    {
      "epoch": 2.44709503655252,
      "grad_norm": 0.1931188553571701,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0179,
      "step": 1590
    },
    {
      "epoch": 2.4624855713736054,
      "grad_norm": 0.3148833215236664,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0148,
      "step": 1600
    },
    {
      "epoch": 2.4624855713736054,
      "eval_loss": 0.03412811458110809,
      "eval_runtime": 153.5482,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 1600
    },
    {
      "epoch": 2.47787610619469,
      "grad_norm": 0.28911468386650085,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.018,
      "step": 1610
    },
    {
      "epoch": 2.4932666410157753,
      "grad_norm": 0.16387534141540527,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0203,
      "step": 1620
    },
    {
      "epoch": 2.50865717583686,
      "grad_norm": 0.13904359936714172,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0189,
      "step": 1630
    },
    {
      "epoch": 2.5240477106579453,
      "grad_norm": 0.35930711030960083,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0189,
      "step": 1640
    },
    {
      "epoch": 2.5394382454790305,
      "grad_norm": 0.29709628224372864,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0208,
      "step": 1650
    },
    {
      "epoch": 2.5394382454790305,
      "eval_loss": 0.03251875191926956,
      "eval_runtime": 153.4794,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1650
    },
    {
      "epoch": 2.5548287803001153,
      "grad_norm": 0.5938408970832825,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0166,
      "step": 1660
    },
    {
      "epoch": 2.5702193151212005,
      "grad_norm": 0.2963032126426697,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.018,
      "step": 1670
    },
    {
      "epoch": 2.5856098499422853,
      "grad_norm": 0.2618955373764038,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0262,
      "step": 1680
    },
    {
      "epoch": 2.6010003847633705,
      "grad_norm": 0.20557081699371338,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0233,
      "step": 1690
    },
    {
      "epoch": 2.6163909195844557,
      "grad_norm": 0.188442662358284,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0178,
      "step": 1700
    },
    {
      "epoch": 2.6163909195844557,
      "eval_loss": 0.03273773565888405,
      "eval_runtime": 153.5007,
      "eval_samples_per_second": 7.531,
      "eval_steps_per_second": 7.531,
      "step": 1700
    },
    {
      "epoch": 2.6317814544055405,
      "grad_norm": 0.2138468325138092,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0189,
      "step": 1710
    },
    {
      "epoch": 2.6471719892266257,
      "grad_norm": 0.18362319469451904,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0307,
      "step": 1720
    },
    {
      "epoch": 2.6625625240477104,
      "grad_norm": 0.38724571466445923,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0234,
      "step": 1730
    },
    {
      "epoch": 2.6779530588687956,
      "grad_norm": 0.13776294887065887,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0163,
      "step": 1740
    },
    {
      "epoch": 2.693343593689881,
      "grad_norm": 0.18619877099990845,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.026,
      "step": 1750
    },
    {
      "epoch": 2.693343593689881,
      "eval_loss": 0.03199377283453941,
      "eval_runtime": 153.5102,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1750
    },
    {
      "epoch": 2.7087341285109656,
      "grad_norm": 0.3150208294391632,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0153,
      "step": 1760
    },
    {
      "epoch": 2.724124663332051,
      "grad_norm": 0.6441848874092102,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0202,
      "step": 1770
    },
    {
      "epoch": 2.7395151981531356,
      "grad_norm": 0.3958023190498352,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0274,
      "step": 1780
    },
    {
      "epoch": 2.754905732974221,
      "grad_norm": 0.1302323192358017,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0196,
      "step": 1790
    },
    {
      "epoch": 2.770296267795306,
      "grad_norm": 0.3273060917854309,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0239,
      "step": 1800
    },
    {
      "epoch": 2.770296267795306,
      "eval_loss": 0.03334551677107811,
      "eval_runtime": 153.4821,
      "eval_samples_per_second": 7.532,
      "eval_steps_per_second": 7.532,
      "step": 1800
    },
    {
      "epoch": 2.785686802616391,
      "grad_norm": 0.24538679420948029,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0202,
      "step": 1810
    },
    {
      "epoch": 2.801077337437476,
      "grad_norm": 0.4235229790210724,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0248,
      "step": 1820
    },
    {
      "epoch": 2.8164678722585608,
      "grad_norm": 0.6049575209617615,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0238,
      "step": 1830
    },
    {
      "epoch": 2.831858407079646,
      "grad_norm": 0.2025410681962967,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0262,
      "step": 1840
    },
    {
      "epoch": 2.847248941900731,
      "grad_norm": 0.21600890159606934,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0227,
      "step": 1850
    },
    {
      "epoch": 2.847248941900731,
      "eval_loss": 0.03210474178195,
      "eval_runtime": 153.5798,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 1850
    },
    {
      "epoch": 2.862639476721816,
      "grad_norm": 1.0758823156356812,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0145,
      "step": 1860
    },
    {
      "epoch": 2.878030011542901,
      "grad_norm": 0.30649304389953613,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.027,
      "step": 1870
    },
    {
      "epoch": 2.893420546363986,
      "grad_norm": 0.48852482438087463,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.0136,
      "step": 1880
    },
    {
      "epoch": 2.908811081185071,
      "grad_norm": 0.20170648396015167,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.018,
      "step": 1890
    },
    {
      "epoch": 2.9242016160061564,
      "grad_norm": 0.22177694737911224,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0166,
      "step": 1900
    },
    {
      "epoch": 2.9242016160061564,
      "eval_loss": 0.03265943005681038,
      "eval_runtime": 153.518,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1900
    },
    {
      "epoch": 2.939592150827241,
      "grad_norm": 0.28209346532821655,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0228,
      "step": 1910
    },
    {
      "epoch": 2.9549826856483263,
      "grad_norm": 0.4292439818382263,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0184,
      "step": 1920
    },
    {
      "epoch": 2.970373220469411,
      "grad_norm": 0.15051664412021637,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.0163,
      "step": 1930
    },
    {
      "epoch": 2.9857637552904963,
      "grad_norm": 0.18560007214546204,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0138,
      "step": 1940
    },
    {
      "epoch": 3.0011542901115815,
      "grad_norm": 0.26750048995018005,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0281,
      "step": 1950
    },
    {
      "epoch": 3.0011542901115815,
      "eval_loss": 0.03369417041540146,
      "eval_runtime": 153.5195,
      "eval_samples_per_second": 7.53,
      "eval_steps_per_second": 7.53,
      "step": 1950
    },
    {
      "epoch": 3.0165448249326663,
      "grad_norm": 0.15498404204845428,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0065,
      "step": 1960
    },
    {
      "epoch": 3.0319353597537515,
      "grad_norm": 0.20060932636260986,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.011,
      "step": 1970
    },
    {
      "epoch": 3.0473258945748363,
      "grad_norm": 0.16343066096305847,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0097,
      "step": 1980
    },
    {
      "epoch": 3.0627164293959215,
      "grad_norm": 0.2700605094432831,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0089,
      "step": 1990
    },
    {
      "epoch": 3.0781069642170067,
      "grad_norm": 0.15850695967674255,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0121,
      "step": 2000
    },
    {
      "epoch": 3.0781069642170067,
      "eval_loss": 0.03507872298359871,
      "eval_runtime": 153.4414,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 2000
    },
    {
      "epoch": 3.0934974990380915,
      "grad_norm": 1.2655847072601318,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.007,
      "step": 2010
    },
    {
      "epoch": 3.1088880338591767,
      "grad_norm": 0.07978175580501556,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0112,
      "step": 2020
    },
    {
      "epoch": 3.1242785686802614,
      "grad_norm": 0.06244011968374252,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.009,
      "step": 2030
    },
    {
      "epoch": 3.1396691035013466,
      "grad_norm": 0.1545596420764923,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.0072,
      "step": 2040
    },
    {
      "epoch": 3.155059638322432,
      "grad_norm": 0.3066238462924957,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0084,
      "step": 2050
    },
    {
      "epoch": 3.155059638322432,
      "eval_loss": 0.03844831883907318,
      "eval_runtime": 153.62,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 2050
    },
    {
      "epoch": 3.1704501731435166,
      "grad_norm": 0.18462346494197845,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0073,
      "step": 2060
    },
    {
      "epoch": 3.185840707964602,
      "grad_norm": 0.48185572028160095,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0078,
      "step": 2070
    },
    {
      "epoch": 3.2012312427856866,
      "grad_norm": 0.25839656591415405,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0096,
      "step": 2080
    },
    {
      "epoch": 3.216621777606772,
      "grad_norm": 0.31359630823135376,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0092,
      "step": 2090
    },
    {
      "epoch": 3.232012312427857,
      "grad_norm": 0.13891646265983582,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.009,
      "step": 2100
    },
    {
      "epoch": 3.232012312427857,
      "eval_loss": 0.03841811791062355,
      "eval_runtime": 153.5857,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2100
    },
    {
      "epoch": 3.247402847248942,
      "grad_norm": 0.10538731515407562,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0107,
      "step": 2110
    },
    {
      "epoch": 3.262793382070027,
      "grad_norm": 0.3056080639362335,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0099,
      "step": 2120
    },
    {
      "epoch": 3.2781839168911118,
      "grad_norm": 0.43214520812034607,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0087,
      "step": 2130
    },
    {
      "epoch": 3.293574451712197,
      "grad_norm": 0.188354030251503,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0055,
      "step": 2140
    },
    {
      "epoch": 3.308964986533282,
      "grad_norm": 0.33608776330947876,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0085,
      "step": 2150
    },
    {
      "epoch": 3.308964986533282,
      "eval_loss": 0.038668908178806305,
      "eval_runtime": 153.6417,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2150
    },
    {
      "epoch": 3.324355521354367,
      "grad_norm": 0.1601262390613556,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0103,
      "step": 2160
    },
    {
      "epoch": 3.339746056175452,
      "grad_norm": 0.44842690229415894,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.014,
      "step": 2170
    },
    {
      "epoch": 3.355136590996537,
      "grad_norm": 0.05282491073012352,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.009,
      "step": 2180
    },
    {
      "epoch": 3.370527125817622,
      "grad_norm": 0.4253879189491272,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0086,
      "step": 2190
    },
    {
      "epoch": 3.3859176606387074,
      "grad_norm": 0.15728415548801422,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0122,
      "step": 2200
    },
    {
      "epoch": 3.3859176606387074,
      "eval_loss": 0.038225095719099045,
      "eval_runtime": 153.5555,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 2200
    },
    {
      "epoch": 3.401308195459792,
      "grad_norm": 0.2565001845359802,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0097,
      "step": 2210
    },
    {
      "epoch": 3.4166987302808773,
      "grad_norm": 0.24468065798282623,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0073,
      "step": 2220
    },
    {
      "epoch": 3.432089265101962,
      "grad_norm": 0.1760941445827484,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0108,
      "step": 2230
    },
    {
      "epoch": 3.4474797999230473,
      "grad_norm": 0.2724066376686096,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0124,
      "step": 2240
    },
    {
      "epoch": 3.4628703347441325,
      "grad_norm": 0.24038079380989075,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0098,
      "step": 2250
    },
    {
      "epoch": 3.4628703347441325,
      "eval_loss": 0.03855520859360695,
      "eval_runtime": 153.5926,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2250
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.13553006947040558,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0087,
      "step": 2260
    },
    {
      "epoch": 3.4936514043863025,
      "grad_norm": 0.29440829157829285,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0051,
      "step": 2270
    },
    {
      "epoch": 3.5090419392073873,
      "grad_norm": 0.25422459840774536,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0115,
      "step": 2280
    },
    {
      "epoch": 3.5244324740284725,
      "grad_norm": 0.4125272333621979,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0072,
      "step": 2290
    },
    {
      "epoch": 3.5398230088495577,
      "grad_norm": 0.12208764255046844,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0098,
      "step": 2300
    },
    {
      "epoch": 3.5398230088495577,
      "eval_loss": 0.04005374014377594,
      "eval_runtime": 153.5778,
      "eval_samples_per_second": 7.527,
      "eval_steps_per_second": 7.527,
      "step": 2300
    },
    {
      "epoch": 3.5552135436706425,
      "grad_norm": 0.13058073818683624,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0119,
      "step": 2310
    },
    {
      "epoch": 3.5706040784917277,
      "grad_norm": 0.5856406688690186,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0125,
      "step": 2320
    },
    {
      "epoch": 3.5859946133128124,
      "grad_norm": 0.1644304394721985,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0089,
      "step": 2330
    },
    {
      "epoch": 3.6013851481338977,
      "grad_norm": 0.21768446266651154,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0114,
      "step": 2340
    },
    {
      "epoch": 3.616775682954983,
      "grad_norm": 0.10144747793674469,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0042,
      "step": 2350
    },
    {
      "epoch": 3.616775682954983,
      "eval_loss": 0.040071211755275726,
      "eval_runtime": 153.5656,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 2350
    },
    {
      "epoch": 3.6321662177760676,
      "grad_norm": 0.32862967252731323,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0116,
      "step": 2360
    },
    {
      "epoch": 3.647556752597153,
      "grad_norm": 0.23250256478786469,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0073,
      "step": 2370
    },
    {
      "epoch": 3.6629472874182376,
      "grad_norm": 0.21265877783298492,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0083,
      "step": 2380
    },
    {
      "epoch": 3.678337822239323,
      "grad_norm": 0.3839889168739319,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0127,
      "step": 2390
    },
    {
      "epoch": 3.693728357060408,
      "grad_norm": 0.11790256202220917,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.0136,
      "step": 2400
    },
    {
      "epoch": 3.693728357060408,
      "eval_loss": 0.03891169652342796,
      "eval_runtime": 153.6468,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2400
    },
    {
      "epoch": 3.709118891881493,
      "grad_norm": 0.4756694436073303,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.0094,
      "step": 2410
    },
    {
      "epoch": 3.724509426702578,
      "grad_norm": 0.2919870913028717,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0064,
      "step": 2420
    },
    {
      "epoch": 3.7398999615236628,
      "grad_norm": 0.13588456809520721,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0083,
      "step": 2430
    },
    {
      "epoch": 3.755290496344748,
      "grad_norm": 0.4328075647354126,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.0074,
      "step": 2440
    },
    {
      "epoch": 3.770681031165833,
      "grad_norm": 0.20062220096588135,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0069,
      "step": 2450
    },
    {
      "epoch": 3.770681031165833,
      "eval_loss": 0.039667315781116486,
      "eval_runtime": 153.6248,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 2450
    },
    {
      "epoch": 3.786071565986918,
      "grad_norm": 0.17748044431209564,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0098,
      "step": 2460
    },
    {
      "epoch": 3.801462100808003,
      "grad_norm": 0.20702221989631653,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0045,
      "step": 2470
    },
    {
      "epoch": 3.816852635629088,
      "grad_norm": 0.13460060954093933,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.0097,
      "step": 2480
    },
    {
      "epoch": 3.832243170450173,
      "grad_norm": 0.3496813476085663,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0128,
      "step": 2490
    },
    {
      "epoch": 3.8476337052712584,
      "grad_norm": 0.19876216351985931,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0103,
      "step": 2500
    },
    {
      "epoch": 3.8476337052712584,
      "eval_loss": 0.040010035037994385,
      "eval_runtime": 153.6847,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 2500
    },
    {
      "epoch": 3.863024240092343,
      "grad_norm": 0.05878753960132599,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.0095,
      "step": 2510
    },
    {
      "epoch": 3.8784147749134283,
      "grad_norm": 0.4279109835624695,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.006,
      "step": 2520
    },
    {
      "epoch": 3.893805309734513,
      "grad_norm": 0.048280470073223114,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0062,
      "step": 2530
    },
    {
      "epoch": 3.9091958445555983,
      "grad_norm": 0.146065354347229,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0064,
      "step": 2540
    },
    {
      "epoch": 3.9245863793766835,
      "grad_norm": 0.47686049342155457,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0127,
      "step": 2550
    },
    {
      "epoch": 3.9245863793766835,
      "eval_loss": 0.03990347683429718,
      "eval_runtime": 153.6454,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2550
    },
    {
      "epoch": 3.9399769141977683,
      "grad_norm": 0.36422210931777954,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0071,
      "step": 2560
    },
    {
      "epoch": 3.9553674490188535,
      "grad_norm": 0.412509024143219,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0067,
      "step": 2570
    },
    {
      "epoch": 3.9707579838399383,
      "grad_norm": 0.28886809945106506,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0137,
      "step": 2580
    },
    {
      "epoch": 3.9861485186610235,
      "grad_norm": 0.4501112103462219,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0071,
      "step": 2590
    },
    {
      "epoch": 4.001539053482109,
      "grad_norm": 0.05743866413831711,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0094,
      "step": 2600
    },
    {
      "epoch": 4.001539053482109,
      "eval_loss": 0.04044051468372345,
      "eval_runtime": 153.5381,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 2600
    },
    {
      "epoch": 4.016929588303194,
      "grad_norm": 0.18015027046203613,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.005,
      "step": 2610
    },
    {
      "epoch": 4.032320123124278,
      "grad_norm": 0.2031249850988388,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0018,
      "step": 2620
    },
    {
      "epoch": 4.047710657945363,
      "grad_norm": 0.040425654500722885,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0021,
      "step": 2630
    },
    {
      "epoch": 4.063101192766449,
      "grad_norm": 0.05413879081606865,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0062,
      "step": 2640
    },
    {
      "epoch": 4.078491727587534,
      "grad_norm": 0.053463660180568695,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0021,
      "step": 2650
    },
    {
      "epoch": 4.078491727587534,
      "eval_loss": 0.04159535467624664,
      "eval_runtime": 153.5542,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 2650
    },
    {
      "epoch": 4.093882262408619,
      "grad_norm": 0.2915383279323578,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0047,
      "step": 2660
    },
    {
      "epoch": 4.109272797229703,
      "grad_norm": 0.20806525647640228,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0038,
      "step": 2670
    },
    {
      "epoch": 4.124663332050789,
      "grad_norm": 0.0508367158472538,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.003,
      "step": 2680
    },
    {
      "epoch": 4.140053866871874,
      "grad_norm": 0.11718433350324631,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0045,
      "step": 2690
    },
    {
      "epoch": 4.155444401692959,
      "grad_norm": 0.04133271425962448,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0041,
      "step": 2700
    },
    {
      "epoch": 4.155444401692959,
      "eval_loss": 0.044051170349121094,
      "eval_runtime": 153.5585,
      "eval_samples_per_second": 7.528,
      "eval_steps_per_second": 7.528,
      "step": 2700
    },
    {
      "epoch": 4.170834936514044,
      "grad_norm": 0.05952709540724754,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.004,
      "step": 2710
    },
    {
      "epoch": 4.186225471335129,
      "grad_norm": 0.14487892389297485,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0037,
      "step": 2720
    },
    {
      "epoch": 4.201616006156214,
      "grad_norm": 0.07504426687955856,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0032,
      "step": 2730
    },
    {
      "epoch": 4.217006540977299,
      "grad_norm": 0.4350045919418335,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0041,
      "step": 2740
    },
    {
      "epoch": 4.232397075798384,
      "grad_norm": 0.10965240746736526,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.0018,
      "step": 2750
    },
    {
      "epoch": 4.232397075798384,
      "eval_loss": 0.045383669435977936,
      "eval_runtime": 153.6568,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2750
    },
    {
      "epoch": 4.247787610619469,
      "grad_norm": 0.15645179152488708,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0024,
      "step": 2760
    },
    {
      "epoch": 4.263178145440554,
      "grad_norm": 0.3239983320236206,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0018,
      "step": 2770
    },
    {
      "epoch": 4.278568680261639,
      "grad_norm": 0.40874114632606506,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0028,
      "step": 2780
    },
    {
      "epoch": 4.293959215082724,
      "grad_norm": 0.05759256333112717,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0035,
      "step": 2790
    },
    {
      "epoch": 4.309349749903809,
      "grad_norm": 0.022221652790904045,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.0031,
      "step": 2800
    },
    {
      "epoch": 4.309349749903809,
      "eval_loss": 0.04668058082461357,
      "eval_runtime": 153.643,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2800
    },
    {
      "epoch": 4.324740284724895,
      "grad_norm": 0.02655138447880745,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0024,
      "step": 2810
    },
    {
      "epoch": 4.340130819545979,
      "grad_norm": 0.03543700650334358,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.003,
      "step": 2820
    },
    {
      "epoch": 4.355521354367064,
      "grad_norm": 0.07309824228286743,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0025,
      "step": 2830
    },
    {
      "epoch": 4.370911889188149,
      "grad_norm": 0.5029100775718689,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0032,
      "step": 2840
    },
    {
      "epoch": 4.3863024240092345,
      "grad_norm": 0.04499899223446846,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0032,
      "step": 2850
    },
    {
      "epoch": 4.3863024240092345,
      "eval_loss": 0.04750298708677292,
      "eval_runtime": 153.6046,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 2850
    },
    {
      "epoch": 4.40169295883032,
      "grad_norm": 0.2527826130390167,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0018,
      "step": 2860
    },
    {
      "epoch": 4.417083493651404,
      "grad_norm": 0.0479939766228199,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0014,
      "step": 2870
    },
    {
      "epoch": 4.432474028472489,
      "grad_norm": 0.060761477798223495,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0077,
      "step": 2880
    },
    {
      "epoch": 4.4478645632935745,
      "grad_norm": 0.023320671170949936,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0018,
      "step": 2890
    },
    {
      "epoch": 4.46325509811466,
      "grad_norm": 0.30546990036964417,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0041,
      "step": 2900
    },
    {
      "epoch": 4.46325509811466,
      "eval_loss": 0.04849005490541458,
      "eval_runtime": 153.6455,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 2900
    },
    {
      "epoch": 4.478645632935745,
      "grad_norm": 0.09365452080965042,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.0062,
      "step": 2910
    },
    {
      "epoch": 4.494036167756829,
      "grad_norm": 0.054107315838336945,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0038,
      "step": 2920
    },
    {
      "epoch": 4.5094267025779144,
      "grad_norm": 0.06470002979040146,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.003,
      "step": 2930
    },
    {
      "epoch": 4.524817237399,
      "grad_norm": 0.498833566904068,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0033,
      "step": 2940
    },
    {
      "epoch": 4.540207772220085,
      "grad_norm": 0.0228235125541687,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0022,
      "step": 2950
    },
    {
      "epoch": 4.540207772220085,
      "eval_loss": 0.04833614081144333,
      "eval_runtime": 153.6168,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 2950
    },
    {
      "epoch": 4.555598307041169,
      "grad_norm": 0.15385878086090088,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.001,
      "step": 2960
    },
    {
      "epoch": 4.570988841862254,
      "grad_norm": 0.05598286911845207,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0031,
      "step": 2970
    },
    {
      "epoch": 4.58637937668334,
      "grad_norm": 0.027815211564302444,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0019,
      "step": 2980
    },
    {
      "epoch": 4.601769911504425,
      "grad_norm": 0.41858261823654175,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0035,
      "step": 2990
    },
    {
      "epoch": 4.61716044632551,
      "grad_norm": 0.21508339047431946,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0025,
      "step": 3000
    },
    {
      "epoch": 4.61716044632551,
      "eval_loss": 0.04863893240690231,
      "eval_runtime": 153.7,
      "eval_samples_per_second": 7.521,
      "eval_steps_per_second": 7.521,
      "step": 3000
    },
    {
      "epoch": 4.632550981146595,
      "grad_norm": 0.10608770698308945,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0022,
      "step": 3010
    },
    {
      "epoch": 4.64794151596768,
      "grad_norm": 0.11644887924194336,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0038,
      "step": 3020
    },
    {
      "epoch": 4.663332050788765,
      "grad_norm": 0.28463560342788696,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0061,
      "step": 3030
    },
    {
      "epoch": 4.67872258560985,
      "grad_norm": 0.23935408890247345,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0024,
      "step": 3040
    },
    {
      "epoch": 4.694113120430935,
      "grad_norm": 0.004781701602041721,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0018,
      "step": 3050
    },
    {
      "epoch": 4.694113120430935,
      "eval_loss": 0.04872981086373329,
      "eval_runtime": 153.7354,
      "eval_samples_per_second": 7.519,
      "eval_steps_per_second": 7.519,
      "step": 3050
    },
    {
      "epoch": 4.7095036552520195,
      "grad_norm": 0.15920807421207428,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.0037,
      "step": 3060
    },
    {
      "epoch": 4.724894190073105,
      "grad_norm": 0.6049684286117554,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0027,
      "step": 3070
    },
    {
      "epoch": 4.74028472489419,
      "grad_norm": 0.8507621884346008,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0031,
      "step": 3080
    },
    {
      "epoch": 4.755675259715275,
      "grad_norm": 0.04301118105649948,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0034,
      "step": 3090
    },
    {
      "epoch": 4.77106579453636,
      "grad_norm": 0.02284870482981205,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0026,
      "step": 3100
    },
    {
      "epoch": 4.77106579453636,
      "eval_loss": 0.04887767136096954,
      "eval_runtime": 153.7167,
      "eval_samples_per_second": 7.52,
      "eval_steps_per_second": 7.52,
      "step": 3100
    },
    {
      "epoch": 4.786456329357446,
      "grad_norm": 0.06778880953788757,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0047,
      "step": 3110
    },
    {
      "epoch": 4.80184686417853,
      "grad_norm": 0.1492983102798462,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0022,
      "step": 3120
    },
    {
      "epoch": 4.817237398999615,
      "grad_norm": 0.009499853476881981,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0033,
      "step": 3130
    },
    {
      "epoch": 4.8326279338207,
      "grad_norm": 0.0072215707041323185,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0025,
      "step": 3140
    },
    {
      "epoch": 4.8480184686417855,
      "grad_norm": 0.274718314409256,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0063,
      "step": 3150
    },
    {
      "epoch": 4.8480184686417855,
      "eval_loss": 0.04920244961977005,
      "eval_runtime": 153.6753,
      "eval_samples_per_second": 7.522,
      "eval_steps_per_second": 7.522,
      "step": 3150
    },
    {
      "epoch": 4.86340900346287,
      "grad_norm": 0.0887475311756134,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.0037,
      "step": 3160
    },
    {
      "epoch": 4.878799538283955,
      "grad_norm": 0.7433437705039978,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0036,
      "step": 3170
    },
    {
      "epoch": 4.89419007310504,
      "grad_norm": 0.09217419475317001,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0031,
      "step": 3180
    },
    {
      "epoch": 4.9095806079261255,
      "grad_norm": 0.07938030362129211,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0026,
      "step": 3190
    },
    {
      "epoch": 4.924971142747211,
      "grad_norm": 0.07195305079221725,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0029,
      "step": 3200
    },
    {
      "epoch": 4.924971142747211,
      "eval_loss": 0.04898969456553459,
      "eval_runtime": 153.5913,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 3200
    },
    {
      "epoch": 4.940361677568296,
      "grad_norm": 0.03176508843898773,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0035,
      "step": 3210
    },
    {
      "epoch": 4.95575221238938,
      "grad_norm": 0.018604934215545654,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0028,
      "step": 3220
    },
    {
      "epoch": 4.9711427472104655,
      "grad_norm": 0.4045256972312927,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0054,
      "step": 3230
    },
    {
      "epoch": 4.986533282031551,
      "grad_norm": 0.12830907106399536,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0023,
      "step": 3240
    },
    {
      "epoch": 4.994228549442093,
      "step": 3245,
      "total_flos": 7.696760527880847e+17,
      "train_loss": 0.0511217040796183,
      "train_runtime": 36319.0813,
      "train_samples_per_second": 1.431,
      "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
}