{
  "best_metric": 0.0350230410695076,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold3/checkpoint-1300",
  "epoch": 4.9976479443033215,
  "eval_steps": 50,
  "global_step": 3320,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015053156458744943,
      "grad_norm": 6.178123474121094,
      "learning_rate": 3.0120481927710846e-06,
      "loss": 1.3483,
      "step": 10
    },
    {
      "epoch": 0.030106312917489886,
      "grad_norm": 7.668521404266357,
      "learning_rate": 6.024096385542169e-06,
      "loss": 1.2949,
      "step": 20
    },
    {
      "epoch": 0.04515946937623483,
      "grad_norm": 4.06071138381958,
      "learning_rate": 9.036144578313253e-06,
      "loss": 1.1399,
      "step": 30
    },
    {
      "epoch": 0.06021262583497977,
      "grad_norm": 2.9999659061431885,
      "learning_rate": 1.2048192771084338e-05,
      "loss": 0.9064,
      "step": 40
    },
    {
      "epoch": 0.07526578229372471,
      "grad_norm": 1.4494060277938843,
      "learning_rate": 1.5060240963855424e-05,
      "loss": 0.4566,
      "step": 50
    },
    {
      "epoch": 0.07526578229372471,
      "eval_loss": 0.31506142020225525,
      "eval_runtime": 156.762,
      "eval_samples_per_second": 7.534,
      "eval_steps_per_second": 7.534,
      "step": 50
    },
    {
      "epoch": 0.09031893875246966,
      "grad_norm": 0.9608872532844543,
      "learning_rate": 1.8072289156626505e-05,
      "loss": 0.2427,
      "step": 60
    },
    {
      "epoch": 0.1053720952112146,
      "grad_norm": 2.3095591068267822,
      "learning_rate": 2.1084337349397593e-05,
      "loss": 0.2342,
      "step": 70
    },
    {
      "epoch": 0.12042525166995954,
      "grad_norm": 0.8325700163841248,
      "learning_rate": 2.4096385542168677e-05,
      "loss": 0.1697,
      "step": 80
    },
    {
      "epoch": 0.1354784081287045,
      "grad_norm": 1.9299920797348022,
      "learning_rate": 2.7108433734939758e-05,
      "loss": 0.1242,
      "step": 90
    },
    {
      "epoch": 0.15053156458744943,
      "grad_norm": 1.137899398803711,
      "learning_rate": 3.012048192771085e-05,
      "loss": 0.1039,
      "step": 100
    },
    {
      "epoch": 0.15053156458744943,
      "eval_loss": 0.09335148334503174,
      "eval_runtime": 156.9195,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 100
    },
    {
      "epoch": 0.16558472104619437,
      "grad_norm": 2.3025782108306885,
      "learning_rate": 3.313253012048193e-05,
      "loss": 0.0837,
      "step": 110
    },
    {
      "epoch": 0.18063787750493931,
      "grad_norm": 1.0972262620925903,
      "learning_rate": 3.614457831325301e-05,
      "loss": 0.1008,
      "step": 120
    },
    {
      "epoch": 0.19569103396368426,
      "grad_norm": 0.8031754493713379,
      "learning_rate": 3.91566265060241e-05,
      "loss": 0.0816,
      "step": 130
    },
    {
      "epoch": 0.2107441904224292,
      "grad_norm": 2.3960483074188232,
      "learning_rate": 4.2168674698795186e-05,
      "loss": 0.0828,
      "step": 140
    },
    {
      "epoch": 0.22579734688117414,
      "grad_norm": 1.7464011907577515,
      "learning_rate": 4.5180722891566266e-05,
      "loss": 0.0879,
      "step": 150
    },
    {
      "epoch": 0.22579734688117414,
      "eval_loss": 0.07289445400238037,
      "eval_runtime": 156.9574,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 150
    },
    {
      "epoch": 0.2408505033399191,
      "grad_norm": 0.8709592223167419,
      "learning_rate": 4.8192771084337354e-05,
      "loss": 0.0821,
      "step": 160
    },
    {
      "epoch": 0.25590365979866403,
      "grad_norm": 0.9007524847984314,
      "learning_rate": 5.120481927710844e-05,
      "loss": 0.0799,
      "step": 170
    },
    {
      "epoch": 0.270956816257409,
      "grad_norm": 1.1023762226104736,
      "learning_rate": 5.4216867469879516e-05,
      "loss": 0.0729,
      "step": 180
    },
    {
      "epoch": 0.2860099727161539,
      "grad_norm": 1.223743200302124,
      "learning_rate": 5.72289156626506e-05,
      "loss": 0.0699,
      "step": 190
    },
    {
      "epoch": 0.30106312917489886,
      "grad_norm": 1.030595064163208,
      "learning_rate": 6.02409638554217e-05,
      "loss": 0.0684,
      "step": 200
    },
    {
      "epoch": 0.30106312917489886,
      "eval_loss": 0.06437338143587112,
      "eval_runtime": 156.9724,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 7.524,
      "step": 200
    },
    {
      "epoch": 0.3161162856336438,
      "grad_norm": 0.8338972926139832,
      "learning_rate": 6.325301204819278e-05,
      "loss": 0.0706,
      "step": 210
    },
    {
      "epoch": 0.33116944209238874,
      "grad_norm": 1.3395456075668335,
      "learning_rate": 6.626506024096386e-05,
      "loss": 0.066,
      "step": 220
    },
    {
      "epoch": 0.3462225985511337,
      "grad_norm": 0.8639088869094849,
      "learning_rate": 6.927710843373494e-05,
      "loss": 0.0741,
      "step": 230
    },
    {
      "epoch": 0.36127575500987863,
      "grad_norm": 0.9407922029495239,
      "learning_rate": 7.228915662650602e-05,
      "loss": 0.0612,
      "step": 240
    },
    {
      "epoch": 0.3763289114686236,
      "grad_norm": 1.3013204336166382,
      "learning_rate": 7.530120481927712e-05,
      "loss": 0.0696,
      "step": 250
    },
    {
      "epoch": 0.3763289114686236,
      "eval_loss": 0.060398537665605545,
      "eval_runtime": 156.4906,
      "eval_samples_per_second": 7.547,
      "eval_steps_per_second": 7.547,
      "step": 250
    },
    {
      "epoch": 0.3913820679273685,
      "grad_norm": 0.6307721138000488,
      "learning_rate": 7.83132530120482e-05,
      "loss": 0.0436,
      "step": 260
    },
    {
      "epoch": 0.40643522438611346,
      "grad_norm": 0.4275800585746765,
      "learning_rate": 8.132530120481928e-05,
      "loss": 0.0543,
      "step": 270
    },
    {
      "epoch": 0.4214883808448584,
      "grad_norm": 0.3509841561317444,
      "learning_rate": 8.433734939759037e-05,
      "loss": 0.0613,
      "step": 280
    },
    {
      "epoch": 0.43654153730360334,
      "grad_norm": 0.2633926272392273,
      "learning_rate": 8.734939759036145e-05,
      "loss": 0.0506,
      "step": 290
    },
    {
      "epoch": 0.4515946937623483,
      "grad_norm": 1.0434097051620483,
      "learning_rate": 9.036144578313253e-05,
      "loss": 0.064,
      "step": 300
    },
    {
      "epoch": 0.4515946937623483,
      "eval_loss": 0.053826671093702316,
      "eval_runtime": 156.4877,
      "eval_samples_per_second": 7.547,
      "eval_steps_per_second": 7.547,
      "step": 300
    },
    {
      "epoch": 0.46664785022109323,
      "grad_norm": 0.2717237174510956,
      "learning_rate": 9.337349397590361e-05,
      "loss": 0.0469,
      "step": 310
    },
    {
      "epoch": 0.4817010066798382,
      "grad_norm": 0.8223183155059814,
      "learning_rate": 9.638554216867471e-05,
      "loss": 0.0784,
      "step": 320
    },
    {
      "epoch": 0.4967541631385831,
      "grad_norm": 0.9409707188606262,
      "learning_rate": 9.939759036144579e-05,
      "loss": 0.056,
      "step": 330
    },
    {
      "epoch": 0.5118073195973281,
      "grad_norm": 0.4314805865287781,
      "learning_rate": 9.999823129264712e-05,
      "loss": 0.0396,
      "step": 340
    },
    {
      "epoch": 0.526860476056073,
      "grad_norm": 0.5084919333457947,
      "learning_rate": 9.999104613348688e-05,
      "loss": 0.048,
      "step": 350
    },
    {
      "epoch": 0.526860476056073,
      "eval_loss": 0.055311419069767,
      "eval_runtime": 156.4635,
      "eval_samples_per_second": 7.548,
      "eval_steps_per_second": 7.548,
      "step": 350
    },
    {
      "epoch": 0.541913632514818,
      "grad_norm": 0.37716764211654663,
      "learning_rate": 9.997833477197385e-05,
      "loss": 0.0501,
      "step": 360
    },
    {
      "epoch": 0.5569667889735629,
      "grad_norm": 0.7053455114364624,
      "learning_rate": 9.996009861327077e-05,
      "loss": 0.0474,
      "step": 370
    },
    {
      "epoch": 0.5720199454323078,
      "grad_norm": 0.5760583877563477,
      "learning_rate": 9.993633967327269e-05,
      "loss": 0.0538,
      "step": 380
    },
    {
      "epoch": 0.5870731018910528,
      "grad_norm": 0.6930764317512512,
      "learning_rate": 9.990706057838416e-05,
      "loss": 0.0417,
      "step": 390
    },
    {
      "epoch": 0.6021262583497977,
      "grad_norm": 0.5816176533699036,
      "learning_rate": 9.987226456522884e-05,
      "loss": 0.0568,
      "step": 400
    },
    {
      "epoch": 0.6021262583497977,
      "eval_loss": 0.05042395368218422,
      "eval_runtime": 156.5219,
      "eval_samples_per_second": 7.545,
      "eval_steps_per_second": 7.545,
      "step": 400
    },
    {
      "epoch": 0.6171794148085427,
      "grad_norm": 0.5168653726577759,
      "learning_rate": 9.983195548029173e-05,
      "loss": 0.0403,
      "step": 410
    },
    {
      "epoch": 0.6322325712672876,
      "grad_norm": 0.9114458560943604,
      "learning_rate": 9.9786137779494e-05,
      "loss": 0.0613,
      "step": 420
    },
    {
      "epoch": 0.6472857277260325,
      "grad_norm": 0.3248330354690552,
      "learning_rate": 9.973481652770038e-05,
      "loss": 0.0427,
      "step": 430
    },
    {
      "epoch": 0.6623388841847775,
      "grad_norm": 0.35154685378074646,
      "learning_rate": 9.967799739815925e-05,
      "loss": 0.0413,
      "step": 440
    },
    {
      "epoch": 0.6773920406435224,
      "grad_norm": 0.813867449760437,
      "learning_rate": 9.961568667187556e-05,
      "loss": 0.0548,
      "step": 450
    },
    {
      "epoch": 0.6773920406435224,
      "eval_loss": 0.046242497861385345,
      "eval_runtime": 156.7192,
      "eval_samples_per_second": 7.536,
      "eval_steps_per_second": 7.536,
      "step": 450
    },
    {
      "epoch": 0.6924451971022674,
      "grad_norm": 0.8682682514190674,
      "learning_rate": 9.954789123691642e-05,
      "loss": 0.0444,
      "step": 460
    },
    {
      "epoch": 0.7074983535610123,
      "grad_norm": 0.4860692322254181,
      "learning_rate": 9.947461858764978e-05,
      "loss": 0.0486,
      "step": 470
    },
    {
      "epoch": 0.7225515100197573,
      "grad_norm": 0.49391481280326843,
      "learning_rate": 9.939587682391586e-05,
      "loss": 0.0423,
      "step": 480
    },
    {
      "epoch": 0.7376046664785022,
      "grad_norm": 0.6448398232460022,
      "learning_rate": 9.931167465013182e-05,
      "loss": 0.0604,
      "step": 490
    },
    {
      "epoch": 0.7526578229372471,
      "grad_norm": 0.7871343493461609,
      "learning_rate": 9.922202137432955e-05,
      "loss": 0.0387,
      "step": 500
    },
    {
      "epoch": 0.7526578229372471,
      "eval_loss": 0.04570082202553749,
      "eval_runtime": 157.0432,
      "eval_samples_per_second": 7.52,
      "eval_steps_per_second": 7.52,
      "step": 500
    },
    {
      "epoch": 0.7677109793959921,
      "grad_norm": 0.16567155718803406,
      "learning_rate": 9.912692690712665e-05,
      "loss": 0.0443,
      "step": 510
    },
    {
      "epoch": 0.782764135854737,
      "grad_norm": 0.49851304292678833,
      "learning_rate": 9.902640176063103e-05,
      "loss": 0.0482,
      "step": 520
    },
    {
      "epoch": 0.797817292313482,
      "grad_norm": 0.27470889687538147,
      "learning_rate": 9.892045704727864e-05,
      "loss": 0.0529,
      "step": 530
    },
    {
      "epoch": 0.8128704487722269,
      "grad_norm": 0.4213567078113556,
      "learning_rate": 9.880910447860527e-05,
      "loss": 0.0636,
      "step": 540
    },
    {
      "epoch": 0.8279236052309719,
      "grad_norm": 0.42453011870384216,
      "learning_rate": 9.869235636395177e-05,
      "loss": 0.0454,
      "step": 550
    },
    {
      "epoch": 0.8279236052309719,
      "eval_loss": 0.043903350830078125,
      "eval_runtime": 157.299,
      "eval_samples_per_second": 7.508,
      "eval_steps_per_second": 7.508,
      "step": 550
    },
    {
      "epoch": 0.8429767616897168,
      "grad_norm": 0.7841992378234863,
      "learning_rate": 9.857022560910338e-05,
      "loss": 0.0639,
      "step": 560
    },
    {
      "epoch": 0.8580299181484617,
      "grad_norm": 0.6617533564567566,
      "learning_rate": 9.844272571486311e-05,
      "loss": 0.0414,
      "step": 570
    },
    {
      "epoch": 0.8730830746072067,
      "grad_norm": 0.23420529067516327,
      "learning_rate": 9.830987077555924e-05,
      "loss": 0.0569,
      "step": 580
    },
    {
      "epoch": 0.8881362310659516,
      "grad_norm": 0.5472472906112671,
      "learning_rate": 9.817167547748729e-05,
      "loss": 0.0629,
      "step": 590
    },
    {
      "epoch": 0.9031893875246966,
      "grad_norm": 0.2491326779127121,
      "learning_rate": 9.802815509728662e-05,
      "loss": 0.0343,
      "step": 600
    },
    {
      "epoch": 0.9031893875246966,
      "eval_loss": 0.042122405022382736,
      "eval_runtime": 157.5417,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "step": 600
    },
    {
      "epoch": 0.9182425439834415,
      "grad_norm": 0.2855018377304077,
      "learning_rate": 9.787932550025158e-05,
      "loss": 0.0373,
      "step": 610
    },
    {
      "epoch": 0.9332957004421865,
      "grad_norm": 0.36227959394454956,
      "learning_rate": 9.772520313857775e-05,
      "loss": 0.0413,
      "step": 620
    },
    {
      "epoch": 0.9483488569009314,
      "grad_norm": 0.425972044467926,
      "learning_rate": 9.756580504954334e-05,
      "loss": 0.0462,
      "step": 630
    },
    {
      "epoch": 0.9634020133596763,
      "grad_norm": 0.35800591111183167,
      "learning_rate": 9.740114885362562e-05,
      "loss": 0.047,
      "step": 640
    },
    {
      "epoch": 0.9784551698184213,
      "grad_norm": 0.28469201922416687,
      "learning_rate": 9.723125275255325e-05,
      "loss": 0.0363,
      "step": 650
    },
    {
      "epoch": 0.9784551698184213,
      "eval_loss": 0.0401780903339386,
      "eval_runtime": 157.68,
      "eval_samples_per_second": 7.49,
      "eval_steps_per_second": 7.49,
      "step": 650
    },
    {
      "epoch": 0.9935083262771662,
      "grad_norm": 0.2893456220626831,
      "learning_rate": 9.705613552729415e-05,
      "loss": 0.0306,
      "step": 660
    },
    {
      "epoch": 1.0085614827359113,
      "grad_norm": 0.29287856817245483,
      "learning_rate": 9.68758165359794e-05,
      "loss": 0.0412,
      "step": 670
    },
    {
      "epoch": 1.0236146391946561,
      "grad_norm": 0.30157166719436646,
      "learning_rate": 9.669031571176322e-05,
      "loss": 0.0404,
      "step": 680
    },
    {
      "epoch": 1.0386677956534012,
      "grad_norm": 0.29122719168663025,
      "learning_rate": 9.64996535606196e-05,
      "loss": 0.0377,
      "step": 690
    },
    {
      "epoch": 1.053720952112146,
      "grad_norm": 0.1511625498533249,
      "learning_rate": 9.630385115907545e-05,
      "loss": 0.0272,
      "step": 700
    },
    {
      "epoch": 1.053720952112146,
      "eval_loss": 0.04149024188518524,
      "eval_runtime": 157.9474,
      "eval_samples_per_second": 7.477,
      "eval_steps_per_second": 7.477,
      "step": 700
    },
    {
      "epoch": 1.068774108570891,
      "grad_norm": 0.19845999777317047,
      "learning_rate": 9.610293015188067e-05,
      "loss": 0.0345,
      "step": 710
    },
    {
      "epoch": 1.083827265029636,
      "grad_norm": 0.3642478287220001,
      "learning_rate": 9.589691274961556e-05,
      "loss": 0.0271,
      "step": 720
    },
    {
      "epoch": 1.098880421488381,
      "grad_norm": 0.3590589761734009,
      "learning_rate": 9.568582172623544e-05,
      "loss": 0.0437,
      "step": 730
    },
    {
      "epoch": 1.1139335779471258,
      "grad_norm": 0.5430022478103638,
      "learning_rate": 9.546968041655326e-05,
      "loss": 0.0355,
      "step": 740
    },
    {
      "epoch": 1.1289867344058706,
      "grad_norm": 0.18729464709758759,
      "learning_rate": 9.524851271366001e-05,
      "loss": 0.0324,
      "step": 750
    },
    {
      "epoch": 1.1289867344058706,
      "eval_loss": 0.03844917565584183,
      "eval_runtime": 158.1863,
      "eval_samples_per_second": 7.466,
      "eval_steps_per_second": 7.466,
      "step": 750
    },
    {
      "epoch": 1.1440398908646157,
      "grad_norm": 0.26876088976860046,
      "learning_rate": 9.502234306628355e-05,
      "loss": 0.0288,
      "step": 760
    },
    {
      "epoch": 1.1590930473233607,
      "grad_norm": 0.3723165690898895,
      "learning_rate": 9.47911964760858e-05,
      "loss": 0.0327,
      "step": 770
    },
    {
      "epoch": 1.1741462037821055,
      "grad_norm": 0.4213339686393738,
      "learning_rate": 9.455509849489915e-05,
      "loss": 0.0284,
      "step": 780
    },
    {
      "epoch": 1.1891993602408506,
      "grad_norm": 0.3251422047615051,
      "learning_rate": 9.431407522190175e-05,
      "loss": 0.0361,
      "step": 790
    },
    {
      "epoch": 1.2042525166995954,
      "grad_norm": 0.23341499269008636,
      "learning_rate": 9.406815330073244e-05,
      "loss": 0.0394,
      "step": 800
    },
    {
      "epoch": 1.2042525166995954,
      "eval_loss": 0.042881306260824203,
      "eval_runtime": 158.3508,
      "eval_samples_per_second": 7.458,
      "eval_steps_per_second": 7.458,
      "step": 800
    },
    {
      "epoch": 1.2193056731583405,
      "grad_norm": 0.24621306359767914,
      "learning_rate": 9.381735991654546e-05,
      "loss": 0.0394,
      "step": 810
    },
    {
      "epoch": 1.2343588296170853,
      "grad_norm": 0.4738430380821228,
      "learning_rate": 9.356172279300528e-05,
      "loss": 0.038,
      "step": 820
    },
    {
      "epoch": 1.2494119860758304,
      "grad_norm": 0.16725154221057892,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.0351,
      "step": 830
    },
    {
      "epoch": 1.2644651425345752,
      "grad_norm": 0.1837790161371231,
      "learning_rate": 9.303603089662716e-05,
      "loss": 0.0379,
      "step": 840
    },
    {
      "epoch": 1.2795182989933203,
      "grad_norm": 0.2514703869819641,
      "learning_rate": 9.276603423579164e-05,
      "loss": 0.0297,
      "step": 850
    },
    {
      "epoch": 1.2795182989933203,
      "eval_loss": 0.04107122868299484,
      "eval_runtime": 158.521,
      "eval_samples_per_second": 7.45,
      "eval_steps_per_second": 7.45,
      "step": 850
    },
    {
      "epoch": 1.294571455452065,
      "grad_norm": 0.3693402111530304,
      "learning_rate": 9.249131005318387e-05,
      "loss": 0.0429,
      "step": 860
    },
    {
      "epoch": 1.30962461191081,
      "grad_norm": 0.16316485404968262,
      "learning_rate": 9.221188871787075e-05,
      "loss": 0.0349,
      "step": 870
    },
    {
      "epoch": 1.324677768369555,
      "grad_norm": 0.2155531495809555,
      "learning_rate": 9.192780111816047e-05,
      "loss": 0.037,
      "step": 880
    },
    {
      "epoch": 1.3397309248283,
      "grad_norm": 0.2901955544948578,
      "learning_rate": 9.163907865818806e-05,
      "loss": 0.0327,
      "step": 890
    },
    {
      "epoch": 1.3547840812870449,
      "grad_norm": 0.5211817622184753,
      "learning_rate": 9.134575325444376e-05,
      "loss": 0.0423,
      "step": 900
    },
    {
      "epoch": 1.3547840812870449,
      "eval_loss": 0.03963349014520645,
      "eval_runtime": 158.651,
      "eval_samples_per_second": 7.444,
      "eval_steps_per_second": 7.444,
      "step": 900
    },
    {
      "epoch": 1.3698372377457897,
      "grad_norm": 0.27670544385910034,
      "learning_rate": 9.104785733224496e-05,
      "loss": 0.0292,
      "step": 910
    },
    {
      "epoch": 1.3848903942045347,
      "grad_norm": 0.45413723587989807,
      "learning_rate": 9.07454238221517e-05,
      "loss": 0.0358,
      "step": 920
    },
    {
      "epoch": 1.3999435506632798,
      "grad_norm": 0.2523304522037506,
      "learning_rate": 9.043848615632642e-05,
      "loss": 0.0294,
      "step": 930
    },
    {
      "epoch": 1.4149967071220246,
      "grad_norm": 0.2198895364999771,
      "learning_rate": 9.012707826483823e-05,
      "loss": 0.0375,
      "step": 940
    },
    {
      "epoch": 1.4300498635807695,
      "grad_norm": 0.3737075626850128,
      "learning_rate": 8.98112345719122e-05,
      "loss": 0.0324,
      "step": 950
    },
    {
      "epoch": 1.4300498635807695,
      "eval_loss": 0.03718731552362442,
      "eval_runtime": 158.7997,
      "eval_samples_per_second": 7.437,
      "eval_steps_per_second": 7.437,
      "step": 950
    },
    {
      "epoch": 1.4451030200395145,
      "grad_norm": 0.6216887831687927,
      "learning_rate": 8.949098999212391e-05,
      "loss": 0.0385,
      "step": 960
    },
    {
      "epoch": 1.4601561764982596,
      "grad_norm": 0.2569785416126251,
      "learning_rate": 8.916637992653991e-05,
      "loss": 0.0321,
      "step": 970
    },
    {
      "epoch": 1.4752093329570044,
      "grad_norm": 0.5239596366882324,
      "learning_rate": 8.883744025880428e-05,
      "loss": 0.0299,
      "step": 980
    },
    {
      "epoch": 1.4902624894157492,
      "grad_norm": 0.4125153422355652,
      "learning_rate": 8.850420735117202e-05,
      "loss": 0.0304,
      "step": 990
    },
    {
      "epoch": 1.5053156458744943,
      "grad_norm": 0.23545460402965546,
      "learning_rate": 8.816671804048933e-05,
      "loss": 0.0311,
      "step": 1000
    },
    {
      "epoch": 1.5053156458744943,
      "eval_loss": 0.03950938582420349,
      "eval_runtime": 158.844,
      "eval_samples_per_second": 7.435,
      "eval_steps_per_second": 7.435,
      "step": 1000
    },
    {
      "epoch": 1.5203688023332393,
      "grad_norm": 0.2283906191587448,
      "learning_rate": 8.782500963412156e-05,
      "loss": 0.0358,
      "step": 1010
    },
    {
      "epoch": 1.5354219587919842,
      "grad_norm": 0.3116263449192047,
      "learning_rate": 8.747911990582912e-05,
      "loss": 0.029,
      "step": 1020
    },
    {
      "epoch": 1.550475115250729,
      "grad_norm": 0.36042335629463196,
      "learning_rate": 8.712908709159183e-05,
      "loss": 0.0288,
      "step": 1030
    },
    {
      "epoch": 1.565528271709474,
      "grad_norm": 0.5503274202346802,
      "learning_rate": 8.677494988538211e-05,
      "loss": 0.0422,
      "step": 1040
    },
    {
      "epoch": 1.5805814281682191,
      "grad_norm": 0.30002403259277344,
      "learning_rate": 8.641674743488769e-05,
      "loss": 0.0309,
      "step": 1050
    },
    {
      "epoch": 1.5805814281682191,
      "eval_loss": 0.04065566882491112,
      "eval_runtime": 159.2321,
      "eval_samples_per_second": 7.417,
      "eval_steps_per_second": 7.417,
      "step": 1050
    },
    {
      "epoch": 1.595634584626964,
      "grad_norm": 0.34152716398239136,
      "learning_rate": 8.605451933718397e-05,
      "loss": 0.0377,
      "step": 1060
    },
    {
      "epoch": 1.6106877410857088,
      "grad_norm": 0.3016291558742523,
      "learning_rate": 8.568830563435694e-05,
      "loss": 0.0387,
      "step": 1070
    },
    {
      "epoch": 1.6257408975444538,
      "grad_norm": 0.5669217109680176,
      "learning_rate": 8.531814680907664e-05,
      "loss": 0.0254,
      "step": 1080
    },
    {
      "epoch": 1.6407940540031989,
      "grad_norm": 0.15337613224983215,
      "learning_rate": 8.494408378012209e-05,
      "loss": 0.0284,
      "step": 1090
    },
    {
      "epoch": 1.6558472104619437,
      "grad_norm": 0.14289391040802002,
      "learning_rate": 8.456615789785804e-05,
      "loss": 0.0233,
      "step": 1100
    },
    {
      "epoch": 1.6558472104619437,
      "eval_loss": 0.03772934526205063,
      "eval_runtime": 159.1219,
      "eval_samples_per_second": 7.422,
      "eval_steps_per_second": 7.422,
      "step": 1100
    },
    {
      "epoch": 1.6709003669206886,
      "grad_norm": 0.24611565470695496,
      "learning_rate": 8.418441093966385e-05,
      "loss": 0.0445,
      "step": 1110
    },
    {
      "epoch": 1.6859535233794336,
      "grad_norm": 0.21417029201984406,
      "learning_rate": 8.379888510531535e-05,
      "loss": 0.029,
      "step": 1120
    },
    {
      "epoch": 1.7010066798381787,
      "grad_norm": 0.3725775182247162,
      "learning_rate": 8.340962301231981e-05,
      "loss": 0.0224,
      "step": 1130
    },
    {
      "epoch": 1.7160598362969235,
      "grad_norm": 0.16964785754680634,
      "learning_rate": 8.301666769120488e-05,
      "loss": 0.0311,
      "step": 1140
    },
    {
      "epoch": 1.7311129927556683,
      "grad_norm": 0.19215409457683563,
      "learning_rate": 8.262006258076187e-05,
      "loss": 0.0455,
      "step": 1150
    },
    {
      "epoch": 1.7311129927556683,
      "eval_loss": 0.035410501062870026,
      "eval_runtime": 159.3978,
      "eval_samples_per_second": 7.409,
      "eval_steps_per_second": 7.409,
      "step": 1150
    },
    {
      "epoch": 1.7461661492144134,
      "grad_norm": 0.1331547200679779,
      "learning_rate": 8.221985152324385e-05,
      "loss": 0.0307,
      "step": 1160
    },
    {
      "epoch": 1.7612193056731584,
      "grad_norm": 0.22927172482013702,
      "learning_rate": 8.18160787595191e-05,
      "loss": 0.0413,
      "step": 1170
    },
    {
      "epoch": 1.7762724621319033,
      "grad_norm": 0.3783109784126282,
      "learning_rate": 8.14087889241806e-05,
      "loss": 0.0364,
      "step": 1180
    },
    {
      "epoch": 1.791325618590648,
      "grad_norm": 0.211699977517128,
      "learning_rate": 8.099802704061195e-05,
      "loss": 0.0384,
      "step": 1190
    },
    {
      "epoch": 1.8063787750493931,
      "grad_norm": 0.44848278164863586,
      "learning_rate": 8.058383851601027e-05,
      "loss": 0.0329,
      "step": 1200
    },
    {
      "epoch": 1.8063787750493931,
      "eval_loss": 0.03639229014515877,
      "eval_runtime": 159.4908,
      "eval_samples_per_second": 7.405,
      "eval_steps_per_second": 7.405,
      "step": 1200
    },
    {
      "epoch": 1.8214319315081382,
      "grad_norm": 0.5605763792991638,
      "learning_rate": 8.01662691363668e-05,
      "loss": 0.0289,
      "step": 1210
    },
    {
      "epoch": 1.836485087966883,
      "grad_norm": 0.3262452483177185,
      "learning_rate": 7.974536506140547e-05,
      "loss": 0.0367,
      "step": 1220
    },
    {
      "epoch": 1.8515382444256279,
      "grad_norm": 0.2924893796443939,
      "learning_rate": 7.932117281948021e-05,
      "loss": 0.0328,
      "step": 1230
    },
    {
      "epoch": 1.866591400884373,
      "grad_norm": 0.277554452419281,
      "learning_rate": 7.889373930243164e-05,
      "loss": 0.0343,
      "step": 1240
    },
    {
      "epoch": 1.881644557343118,
      "grad_norm": 0.23780488967895508,
      "learning_rate": 7.846311176040331e-05,
      "loss": 0.0352,
      "step": 1250
    },
    {
      "epoch": 1.881644557343118,
      "eval_loss": 0.035072676837444305,
      "eval_runtime": 159.7659,
      "eval_samples_per_second": 7.392,
      "eval_steps_per_second": 7.392,
      "step": 1250
    },
    {
      "epoch": 1.8966977138018628,
      "grad_norm": 0.17802149057388306,
      "learning_rate": 7.802933779661859e-05,
      "loss": 0.0374,
      "step": 1260
    },
    {
      "epoch": 1.9117508702606076,
      "grad_norm": 0.19435422122478485,
      "learning_rate": 7.759246536211844e-05,
      "loss": 0.0307,
      "step": 1270
    },
    {
      "epoch": 1.9268040267193527,
      "grad_norm": 0.1490616649389267,
      "learning_rate": 7.715254275046062e-05,
      "loss": 0.0326,
      "step": 1280
    },
    {
      "epoch": 1.9418571831780977,
      "grad_norm": 0.22022828459739685,
      "learning_rate": 7.670961859238124e-05,
      "loss": 0.0297,
      "step": 1290
    },
    {
      "epoch": 1.9569103396368426,
      "grad_norm": 0.43419596552848816,
      "learning_rate": 7.626374185041886e-05,
      "loss": 0.029,
      "step": 1300
    },
    {
      "epoch": 1.9569103396368426,
      "eval_loss": 0.0350230410695076,
      "eval_runtime": 159.6805,
      "eval_samples_per_second": 7.396,
      "eval_steps_per_second": 7.396,
      "step": 1300
    },
    {
      "epoch": 1.9719634960955874,
      "grad_norm": 0.24266742169857025,
      "learning_rate": 7.581496181350203e-05,
      "loss": 0.0263,
      "step": 1310
    },
    {
      "epoch": 1.9870166525543325,
      "grad_norm": 0.17458631098270416,
      "learning_rate": 7.536332809150067e-05,
      "loss": 0.0281,
      "step": 1320
    },
    {
      "epoch": 2.0020698090130775,
      "grad_norm": 0.27361252903938293,
      "learning_rate": 7.490889060974201e-05,
      "loss": 0.0346,
      "step": 1330
    },
    {
      "epoch": 2.0171229654718226,
      "grad_norm": 0.27446067333221436,
      "learning_rate": 7.445169960349167e-05,
      "loss": 0.017,
      "step": 1340
    },
    {
      "epoch": 2.032176121930567,
      "grad_norm": 0.1635950803756714,
      "learning_rate": 7.399180561240044e-05,
      "loss": 0.0187,
      "step": 1350
    },
    {
      "epoch": 2.032176121930567,
      "eval_loss": 0.03626255691051483,
      "eval_runtime": 159.7835,
      "eval_samples_per_second": 7.391,
      "eval_steps_per_second": 7.391,
      "step": 1350
    },
    {
      "epoch": 2.0472292783893122,
      "grad_norm": 0.10786867886781693,
      "learning_rate": 7.352925947491746e-05,
      "loss": 0.0119,
      "step": 1360
    },
    {
      "epoch": 2.0622824348480573,
      "grad_norm": 0.13401247560977936,
      "learning_rate": 7.306411232267029e-05,
      "loss": 0.0175,
      "step": 1370
    },
    {
      "epoch": 2.0773355913068023,
      "grad_norm": 0.18004673719406128,
      "learning_rate": 7.259641557481269e-05,
      "loss": 0.0172,
      "step": 1380
    },
    {
      "epoch": 2.092388747765547,
      "grad_norm": 0.31352919340133667,
      "learning_rate": 7.212622093234049e-05,
      "loss": 0.0146,
      "step": 1390
    },
    {
      "epoch": 2.107441904224292,
      "grad_norm": 0.5032555460929871,
      "learning_rate": 7.165358037237643e-05,
      "loss": 0.025,
      "step": 1400
    },
    {
      "epoch": 2.107441904224292,
      "eval_loss": 0.038006313145160675,
      "eval_runtime": 159.9044,
      "eval_samples_per_second": 7.386,
      "eval_steps_per_second": 7.386,
      "step": 1400
    },
    {
      "epoch": 2.122495060683037,
      "grad_norm": 0.28312698006629944,
      "learning_rate": 7.117854614242434e-05,
      "loss": 0.0249,
      "step": 1410
    },
    {
      "epoch": 2.137548217141782,
      "grad_norm": 0.470378577709198,
      "learning_rate": 7.070117075459352e-05,
      "loss": 0.022,
      "step": 1420
    },
    {
      "epoch": 2.1526013736005267,
      "grad_norm": 0.1857473999261856,
      "learning_rate": 7.022150697979384e-05,
      "loss": 0.0244,
      "step": 1430
    },
    {
      "epoch": 2.167654530059272,
      "grad_norm": 0.3747231066226959,
      "learning_rate": 6.973960784190237e-05,
      "loss": 0.0283,
      "step": 1440
    },
    {
      "epoch": 2.182707686518017,
      "grad_norm": 0.2834364175796509,
      "learning_rate": 6.925552661190166e-05,
      "loss": 0.0209,
      "step": 1450
    },
    {
      "epoch": 2.182707686518017,
      "eval_loss": 0.03765711560845375,
      "eval_runtime": 160.0216,
      "eval_samples_per_second": 7.38,
      "eval_steps_per_second": 7.38,
      "step": 1450
    },
    {
      "epoch": 2.197760842976762,
      "grad_norm": 0.2789125144481659,
      "learning_rate": 6.876931680199121e-05,
      "loss": 0.0255,
      "step": 1460
    },
    {
      "epoch": 2.2128139994355065,
      "grad_norm": 0.16488228738307953,
      "learning_rate": 6.828103215967186e-05,
      "loss": 0.0202,
      "step": 1470
    },
    {
      "epoch": 2.2278671558942515,
      "grad_norm": 0.2753039300441742,
      "learning_rate": 6.779072666180446e-05,
      "loss": 0.0266,
      "step": 1480
    },
    {
      "epoch": 2.2429203123529966,
      "grad_norm": 0.08052641153335571,
      "learning_rate": 6.729845450864294e-05,
      "loss": 0.0166,
      "step": 1490
    },
    {
      "epoch": 2.257973468811741,
      "grad_norm": 0.18475492298603058,
      "learning_rate": 6.680427011784292e-05,
      "loss": 0.0224,
      "step": 1500
    },
    {
      "epoch": 2.257973468811741,
      "eval_loss": 0.0406508594751358,
      "eval_runtime": 160.1992,
      "eval_samples_per_second": 7.372,
      "eval_steps_per_second": 7.372,
      "step": 1500
    },
    {
      "epoch": 2.2730266252704863,
      "grad_norm": 0.4654618799686432,
      "learning_rate": 6.630822811844604e-05,
      "loss": 0.0187,
      "step": 1510
    },
    {
      "epoch": 2.2880797817292313,
      "grad_norm": 0.1831727921962738,
      "learning_rate": 6.58103833448412e-05,
      "loss": 0.0234,
      "step": 1520
    },
    {
      "epoch": 2.3031329381879764,
      "grad_norm": 0.3095850944519043,
      "learning_rate": 6.531079083070288e-05,
      "loss": 0.0241,
      "step": 1530
    },
    {
      "epoch": 2.3181860946467214,
      "grad_norm": 0.11199437081813812,
      "learning_rate": 6.480950580290752e-05,
      "loss": 0.0174,
      "step": 1540
    },
    {
      "epoch": 2.333239251105466,
      "grad_norm": 0.30752846598625183,
      "learning_rate": 6.430658367542843e-05,
      "loss": 0.0261,
      "step": 1550
    },
    {
      "epoch": 2.333239251105466,
      "eval_loss": 0.039291996508836746,
      "eval_runtime": 160.0709,
      "eval_samples_per_second": 7.378,
      "eval_steps_per_second": 7.378,
      "step": 1550
    },
    {
      "epoch": 2.348292407564211,
      "grad_norm": 0.1692783236503601,
      "learning_rate": 6.380208004321036e-05,
      "loss": 0.0239,
      "step": 1560
    },
    {
      "epoch": 2.363345564022956,
      "grad_norm": 0.24642743170261383,
      "learning_rate": 6.32960506760236e-05,
      "loss": 0.0182,
      "step": 1570
    },
    {
      "epoch": 2.378398720481701,
      "grad_norm": 0.2417881041765213,
      "learning_rate": 6.278855151229901e-05,
      "loss": 0.0169,
      "step": 1580
    },
    {
      "epoch": 2.393451876940446,
      "grad_norm": 0.1943085491657257,
      "learning_rate": 6.227963865294444e-05,
      "loss": 0.0126,
      "step": 1590
    },
    {
      "epoch": 2.408505033399191,
      "grad_norm": 0.29183781147003174,
      "learning_rate": 6.176936835514312e-05,
      "loss": 0.0156,
      "step": 1600
    },
    {
      "epoch": 2.408505033399191,
      "eval_loss": 0.038704484701156616,
      "eval_runtime": 160.2281,
      "eval_samples_per_second": 7.371,
      "eval_steps_per_second": 7.371,
      "step": 1600
    },
    {
      "epoch": 2.423558189857936,
      "grad_norm": 0.3246520459651947,
      "learning_rate": 6.125779702613471e-05,
      "loss": 0.0188,
      "step": 1610
    },
    {
      "epoch": 2.438611346316681,
      "grad_norm": 0.21692079305648804,
      "learning_rate": 6.074498121697983e-05,
      "loss": 0.0217,
      "step": 1620
    },
    {
      "epoch": 2.4536645027754256,
      "grad_norm": 0.29945269227027893,
      "learning_rate": 6.023097761630879e-05,
      "loss": 0.0261,
      "step": 1630
    },
    {
      "epoch": 2.4687176592341706,
      "grad_norm": 0.12020477652549744,
      "learning_rate": 5.971584304405489e-05,
      "loss": 0.016,
      "step": 1640
    },
    {
      "epoch": 2.4837708156929157,
      "grad_norm": 0.18562984466552734,
      "learning_rate": 5.919963444517338e-05,
      "loss": 0.0173,
      "step": 1650
    },
    {
      "epoch": 2.4837708156929157,
      "eval_loss": 0.038775425404310226,
      "eval_runtime": 160.3667,
      "eval_samples_per_second": 7.364,
      "eval_steps_per_second": 7.364,
      "step": 1650
    },
    {
      "epoch": 2.4988239721516607,
      "grad_norm": 0.45023879408836365,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0204,
      "step": 1660
    },
    {
      "epoch": 2.5138771286104054,
      "grad_norm": 0.40980616211891174,
      "learning_rate": 5.816422353467562e-05,
      "loss": 0.0163,
      "step": 1670
    },
    {
      "epoch": 2.5289302850691504,
      "grad_norm": 0.12173057347536087,
      "learning_rate": 5.7645135681360496e-05,
      "loss": 0.0207,
      "step": 1680
    },
    {
      "epoch": 2.5439834415278955,
      "grad_norm": 0.36546701192855835,
      "learning_rate": 5.7125202705367234e-05,
      "loss": 0.0216,
      "step": 1690
    },
    {
      "epoch": 2.5590365979866405,
      "grad_norm": 0.13067616522312164,
      "learning_rate": 5.660448208208513e-05,
      "loss": 0.0228,
      "step": 1700
    },
    {
      "epoch": 2.5590365979866405,
      "eval_loss": 0.03666573390364647,
      "eval_runtime": 160.2566,
      "eval_samples_per_second": 7.369,
      "eval_steps_per_second": 7.369,
      "step": 1700
    },
    {
      "epoch": 2.574089754445385,
      "grad_norm": 0.12506772577762604,
      "learning_rate": 5.608303137397294e-05,
      "loss": 0.0169,
      "step": 1710
    },
    {
      "epoch": 2.58914291090413,
      "grad_norm": 0.27075260877609253,
      "learning_rate": 5.5560908224195886e-05,
      "loss": 0.0218,
      "step": 1720
    },
    {
      "epoch": 2.6041960673628752,
      "grad_norm": 0.2712065279483795,
      "learning_rate": 5.503817035025342e-05,
      "loss": 0.0224,
      "step": 1730
    },
    {
      "epoch": 2.61924922382162,
      "grad_norm": 0.5131589770317078,
      "learning_rate": 5.4514875537598985e-05,
      "loss": 0.019,
      "step": 1740
    },
    {
      "epoch": 2.634302380280365,
      "grad_norm": 0.26789724826812744,
      "learning_rate": 5.399108163325217e-05,
      "loss": 0.0328,
      "step": 1750
    },
    {
      "epoch": 2.634302380280365,
      "eval_loss": 0.037083230912685394,
      "eval_runtime": 160.4016,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 1750
    },
    {
      "epoch": 2.64935553673911,
      "grad_norm": 0.20713073015213013,
      "learning_rate": 5.346684653940408e-05,
      "loss": 0.0188,
      "step": 1760
    },
    {
      "epoch": 2.664408693197855,
      "grad_norm": 0.16466517746448517,
      "learning_rate": 5.294222820701661e-05,
      "loss": 0.0203,
      "step": 1770
    },
    {
      "epoch": 2.6794618496566,
      "grad_norm": 0.3724817633628845,
      "learning_rate": 5.24172846294163e-05,
      "loss": 0.0162,
      "step": 1780
    },
    {
      "epoch": 2.6945150061153447,
      "grad_norm": 0.23705260455608368,
      "learning_rate": 5.1892073835883524e-05,
      "loss": 0.018,
      "step": 1790
    },
    {
      "epoch": 2.7095681625740897,
      "grad_norm": 0.8817890882492065,
      "learning_rate": 5.136665388523778e-05,
      "loss": 0.0264,
      "step": 1800
    },
    {
      "epoch": 2.7095681625740897,
      "eval_loss": 0.03928447514772415,
      "eval_runtime": 160.3293,
      "eval_samples_per_second": 7.366,
      "eval_steps_per_second": 7.366,
      "step": 1800
    },
    {
      "epoch": 2.7246213190328348,
      "grad_norm": 0.11891370266675949,
      "learning_rate": 5.0841082859419585e-05,
      "loss": 0.0268,
      "step": 1810
    },
    {
      "epoch": 2.7396744754915794,
      "grad_norm": 0.352401465177536,
      "learning_rate": 5.031541885706987e-05,
      "loss": 0.0255,
      "step": 1820
    },
    {
      "epoch": 2.7547276319503244,
      "grad_norm": 0.2503037750720978,
      "learning_rate": 4.9789719987107545e-05,
      "loss": 0.0214,
      "step": 1830
    },
    {
      "epoch": 2.7697807884090695,
      "grad_norm": 0.34862828254699707,
      "learning_rate": 4.926404436230596e-05,
      "loss": 0.0207,
      "step": 1840
    },
    {
      "epoch": 2.7848339448678145,
      "grad_norm": 0.2952202260494232,
      "learning_rate": 4.8738450092868785e-05,
      "loss": 0.0235,
      "step": 1850
    },
    {
      "epoch": 2.7848339448678145,
      "eval_loss": 0.036392927169799805,
      "eval_runtime": 160.407,
      "eval_samples_per_second": 7.363,
      "eval_steps_per_second": 7.363,
      "step": 1850
    },
    {
      "epoch": 2.7998871013265596,
      "grad_norm": 0.13503505289554596,
      "learning_rate": 4.8212995280006426e-05,
      "loss": 0.0189,
      "step": 1860
    },
    {
      "epoch": 2.814940257785304,
      "grad_norm": 0.2581332325935364,
      "learning_rate": 4.76877380095132e-05,
      "loss": 0.0201,
      "step": 1870
    },
    {
      "epoch": 2.8299934142440493,
      "grad_norm": 0.30006110668182373,
      "learning_rate": 4.7162736345346303e-05,
      "loss": 0.0154,
      "step": 1880
    },
    {
      "epoch": 2.8450465707027943,
      "grad_norm": 0.5561713576316833,
      "learning_rate": 4.663804832320726e-05,
      "loss": 0.0191,
      "step": 1890
    },
    {
      "epoch": 2.860099727161539,
      "grad_norm": 0.32717275619506836,
      "learning_rate": 4.6113731944126406e-05,
      "loss": 0.0204,
      "step": 1900
    },
    {
      "epoch": 2.860099727161539,
      "eval_loss": 0.038577429950237274,
      "eval_runtime": 160.6685,
      "eval_samples_per_second": 7.351,
      "eval_steps_per_second": 7.351,
      "step": 1900
    },
    {
      "epoch": 2.875152883620284,
      "grad_norm": 0.30127692222595215,
      "learning_rate": 4.558984516805118e-05,
      "loss": 0.0193,
      "step": 1910
    },
    {
      "epoch": 2.890206040079029,
      "grad_norm": 0.31373462080955505,
      "learning_rate": 4.5066445907439104e-05,
      "loss": 0.0143,
      "step": 1920
    },
    {
      "epoch": 2.905259196537774,
      "grad_norm": 0.47283080220222473,
      "learning_rate": 4.454359202085582e-05,
      "loss": 0.0242,
      "step": 1930
    },
    {
      "epoch": 2.920312352996519,
      "grad_norm": 0.47461745142936707,
      "learning_rate": 4.402134130657925e-05,
      "loss": 0.0187,
      "step": 1940
    },
    {
      "epoch": 2.9353655094552638,
      "grad_norm": 0.3949275016784668,
      "learning_rate": 4.349975149621039e-05,
      "loss": 0.0207,
      "step": 1950
    },
    {
      "epoch": 2.9353655094552638,
      "eval_loss": 0.03719981759786606,
      "eval_runtime": 160.6509,
      "eval_samples_per_second": 7.351,
      "eval_steps_per_second": 7.351,
      "step": 1950
    },
    {
      "epoch": 2.950418665914009,
      "grad_norm": 0.23788753151893616,
      "learning_rate": 4.297888024829126e-05,
      "loss": 0.0159,
      "step": 1960
    },
    {
      "epoch": 2.965471822372754,
      "grad_norm": 0.2292858064174652,
      "learning_rate": 4.2458785141931314e-05,
      "loss": 0.0198,
      "step": 1970
    },
    {
      "epoch": 2.9805249788314985,
      "grad_norm": 0.12435556948184967,
      "learning_rate": 4.1939523670442316e-05,
      "loss": 0.0161,
      "step": 1980
    },
    {
      "epoch": 2.9955781352902435,
      "grad_norm": 0.25317656993865967,
      "learning_rate": 4.14211532349828e-05,
      "loss": 0.0166,
      "step": 1990
    },
    {
      "epoch": 3.0106312917489886,
      "grad_norm": 0.03211146593093872,
      "learning_rate": 4.090373113821281e-05,
      "loss": 0.01,
      "step": 2000
    },
    {
      "epoch": 3.0106312917489886,
      "eval_loss": 0.03901160508394241,
      "eval_runtime": 160.8018,
      "eval_samples_per_second": 7.344,
      "eval_steps_per_second": 7.344,
      "step": 2000
    },
    {
      "epoch": 3.0256844482077336,
      "grad_norm": 0.23511217534542084,
      "learning_rate": 4.0387314577959315e-05,
      "loss": 0.0093,
      "step": 2010
    },
    {
      "epoch": 3.0407376046664787,
      "grad_norm": 0.06058523431420326,
      "learning_rate": 3.987196064089346e-05,
      "loss": 0.0063,
      "step": 2020
    },
    {
      "epoch": 3.0557907611252233,
      "grad_norm": 0.2121521681547165,
      "learning_rate": 3.935772629621995e-05,
      "loss": 0.0047,
      "step": 2030
    },
    {
      "epoch": 3.0708439175839684,
      "grad_norm": 0.11164995282888412,
      "learning_rate": 3.8844668389379396e-05,
      "loss": 0.0084,
      "step": 2040
    },
    {
      "epoch": 3.0858970740427134,
      "grad_norm": 0.3772049844264984,
      "learning_rate": 3.833284363576447e-05,
      "loss": 0.0103,
      "step": 2050
    },
    {
      "epoch": 3.0858970740427134,
      "eval_loss": 0.043147701770067215,
      "eval_runtime": 160.8171,
      "eval_samples_per_second": 7.344,
      "eval_steps_per_second": 7.344,
      "step": 2050
    },
    {
      "epoch": 3.1009502305014585,
      "grad_norm": 0.551115870475769,
      "learning_rate": 3.7822308614450406e-05,
      "loss": 0.0063,
      "step": 2060
    },
    {
      "epoch": 3.116003386960203,
      "grad_norm": 0.13338187336921692,
      "learning_rate": 3.7313119761940375e-05,
      "loss": 0.0136,
      "step": 2070
    },
    {
      "epoch": 3.131056543418948,
      "grad_norm": 0.2778257429599762,
      "learning_rate": 3.680533336592694e-05,
      "loss": 0.0074,
      "step": 2080
    },
    {
      "epoch": 3.146109699877693,
      "grad_norm": 0.10951809585094452,
      "learning_rate": 3.62990055590697e-05,
      "loss": 0.0101,
      "step": 2090
    },
    {
      "epoch": 3.1611628563364382,
      "grad_norm": 0.08265303820371628,
      "learning_rate": 3.579419231279023e-05,
      "loss": 0.0096,
      "step": 2100
    },
    {
      "epoch": 3.1611628563364382,
      "eval_loss": 0.041790470480918884,
      "eval_runtime": 160.7872,
      "eval_samples_per_second": 7.345,
      "eval_steps_per_second": 7.345,
      "step": 2100
    },
    {
      "epoch": 3.176216012795183,
      "grad_norm": 0.032768603414297104,
      "learning_rate": 3.529094943108475e-05,
      "loss": 0.0072,
      "step": 2110
    },
    {
      "epoch": 3.191269169253928,
      "grad_norm": 0.22910161316394806,
      "learning_rate": 3.478933254435534e-05,
      "loss": 0.0124,
      "step": 2120
    },
    {
      "epoch": 3.206322325712673,
      "grad_norm": 0.22300395369529724,
      "learning_rate": 3.4289397103260346e-05,
      "loss": 0.0116,
      "step": 2130
    },
    {
      "epoch": 3.221375482171418,
      "grad_norm": 0.23817072808742523,
      "learning_rate": 3.3791198372584664e-05,
      "loss": 0.0106,
      "step": 2140
    },
    {
      "epoch": 3.2364286386301626,
      "grad_norm": 0.12127862125635147,
      "learning_rate": 3.329479142513051e-05,
      "loss": 0.012,
      "step": 2150
    },
    {
      "epoch": 3.2364286386301626,
      "eval_loss": 0.043117210268974304,
      "eval_runtime": 160.8539,
      "eval_samples_per_second": 7.342,
      "eval_steps_per_second": 7.342,
      "step": 2150
    },
    {
      "epoch": 3.2514817950889077,
      "grad_norm": 0.28934839367866516,
      "learning_rate": 3.280023113562957e-05,
      "loss": 0.0115,
      "step": 2160
    },
    {
      "epoch": 3.2665349515476527,
      "grad_norm": 0.2645263671875,
      "learning_rate": 3.230757217467677e-05,
      "loss": 0.0163,
      "step": 2170
    },
    {
      "epoch": 3.2815881080063978,
      "grad_norm": 0.29075613617897034,
      "learning_rate": 3.1816869002686936e-05,
      "loss": 0.0105,
      "step": 2180
    },
    {
      "epoch": 3.2966412644651424,
      "grad_norm": 0.05602017417550087,
      "learning_rate": 3.1328175863874464e-05,
      "loss": 0.0067,
      "step": 2190
    },
    {
      "epoch": 3.3116944209238874,
      "grad_norm": 0.293961763381958,
      "learning_rate": 3.084154678025692e-05,
      "loss": 0.0098,
      "step": 2200
    },
    {
      "epoch": 3.3116944209238874,
      "eval_loss": 0.04433353245258331,
      "eval_runtime": 160.9207,
      "eval_samples_per_second": 7.339,
      "eval_steps_per_second": 7.339,
      "step": 2200
    },
    {
      "epoch": 3.3267475773826325,
      "grad_norm": 0.5569509863853455,
      "learning_rate": 3.035703554568331e-05,
      "loss": 0.0101,
      "step": 2210
    },
    {
      "epoch": 3.3418007338413775,
      "grad_norm": 0.23910270631313324,
      "learning_rate": 2.9874695719887464e-05,
      "loss": 0.0078,
      "step": 2220
    },
    {
      "epoch": 3.356853890300122,
      "grad_norm": 0.3301239013671875,
      "learning_rate": 2.9394580622567312e-05,
      "loss": 0.0049,
      "step": 2230
    },
    {
      "epoch": 3.371907046758867,
      "grad_norm": 0.25911498069763184,
      "learning_rate": 2.8916743327490803e-05,
      "loss": 0.0093,
      "step": 2240
    },
    {
      "epoch": 3.3869602032176123,
      "grad_norm": 0.22444899380207062,
      "learning_rate": 2.8441236656628828e-05,
      "loss": 0.0104,
      "step": 2250
    },
    {
      "epoch": 3.3869602032176123,
      "eval_loss": 0.04660218954086304,
      "eval_runtime": 161.0077,
      "eval_samples_per_second": 7.335,
      "eval_steps_per_second": 7.335,
      "step": 2250
    },
    {
      "epoch": 3.4020133596763573,
      "grad_norm": 0.18989907205104828,
      "learning_rate": 2.79681131743161e-05,
      "loss": 0.0038,
      "step": 2260
    },
    {
      "epoch": 3.417066516135102,
      "grad_norm": 0.3380717933177948,
      "learning_rate": 2.7497425181440607e-05,
      "loss": 0.0083,
      "step": 2270
    },
    {
      "epoch": 3.432119672593847,
      "grad_norm": 0.3728026747703552,
      "learning_rate": 2.702922470966187e-05,
      "loss": 0.0076,
      "step": 2280
    },
    {
      "epoch": 3.447172829052592,
      "grad_norm": 0.47184136509895325,
      "learning_rate": 2.6563563515659306e-05,
      "loss": 0.0094,
      "step": 2290
    },
    {
      "epoch": 3.462225985511337,
      "grad_norm": 0.5333015322685242,
      "learning_rate": 2.6100493075410848e-05,
      "loss": 0.0061,
      "step": 2300
    },
    {
      "epoch": 3.462225985511337,
      "eval_loss": 0.04809270426630974,
      "eval_runtime": 160.9513,
      "eval_samples_per_second": 7.338,
      "eval_steps_per_second": 7.338,
      "step": 2300
    },
    {
      "epoch": 3.4772791419700817,
      "grad_norm": 0.4247993230819702,
      "learning_rate": 2.5640064578502497e-05,
      "loss": 0.0135,
      "step": 2310
    },
    {
      "epoch": 3.4923322984288268,
      "grad_norm": 0.4623792767524719,
      "learning_rate": 2.5182328922469723e-05,
      "loss": 0.011,
      "step": 2320
    },
    {
      "epoch": 3.507385454887572,
      "grad_norm": 0.37935352325439453,
      "learning_rate": 2.4727336707170973e-05,
      "loss": 0.0118,
      "step": 2330
    },
    {
      "epoch": 3.522438611346317,
      "grad_norm": 0.13567650318145752,
      "learning_rate": 2.427513822919424e-05,
      "loss": 0.0097,
      "step": 2340
    },
    {
      "epoch": 3.5374917678050615,
      "grad_norm": 0.3444633483886719,
      "learning_rate": 2.3825783476297087e-05,
      "loss": 0.0051,
      "step": 2350
    },
    {
      "epoch": 3.5374917678050615,
      "eval_loss": 0.04625854268670082,
      "eval_runtime": 160.9561,
      "eval_samples_per_second": 7.337,
      "eval_steps_per_second": 7.337,
      "step": 2350
    },
    {
      "epoch": 3.5525449242638065,
      "grad_norm": 0.2048613429069519,
      "learning_rate": 2.337932212188073e-05,
      "loss": 0.0113,
      "step": 2360
    },
    {
      "epoch": 3.5675980807225516,
      "grad_norm": 0.30786871910095215,
      "learning_rate": 2.2935803519499e-05,
      "loss": 0.0086,
      "step": 2370
    },
    {
      "epoch": 3.582651237181296,
      "grad_norm": 0.3959846794605255,
      "learning_rate": 2.2495276697402662e-05,
      "loss": 0.0079,
      "step": 2380
    },
    {
      "epoch": 3.5977043936400412,
      "grad_norm": 0.08823379129171371,
      "learning_rate": 2.2057790353119535e-05,
      "loss": 0.0077,
      "step": 2390
    },
    {
      "epoch": 3.6127575500987863,
      "grad_norm": 0.09143965691328049,
      "learning_rate": 2.1623392848071354e-05,
      "loss": 0.0085,
      "step": 2400
    },
    {
      "epoch": 3.6127575500987863,
      "eval_loss": 0.045845236629247665,
      "eval_runtime": 161.0434,
      "eval_samples_per_second": 7.333,
      "eval_steps_per_second": 7.333,
      "step": 2400
    },
    {
      "epoch": 3.6278107065575314,
      "grad_norm": 0.031368013471364975,
      "learning_rate": 2.1192132202227677e-05,
      "loss": 0.0082,
      "step": 2410
    },
    {
      "epoch": 3.6428638630162764,
      "grad_norm": 0.35049495100975037,
      "learning_rate": 2.0764056088797645e-05,
      "loss": 0.0067,
      "step": 2420
    },
    {
      "epoch": 3.657917019475021,
      "grad_norm": 0.40344491600990295,
      "learning_rate": 2.0339211828959904e-05,
      "loss": 0.0057,
      "step": 2430
    },
    {
      "epoch": 3.672970175933766,
      "grad_norm": 0.22775594890117645,
      "learning_rate": 1.9917646386631577e-05,
      "loss": 0.0064,
      "step": 2440
    },
    {
      "epoch": 3.688023332392511,
      "grad_norm": 0.025965895503759384,
      "learning_rate": 1.949940636327671e-05,
      "loss": 0.0093,
      "step": 2450
    },
    {
      "epoch": 3.688023332392511,
      "eval_loss": 0.04670306667685509,
      "eval_runtime": 161.0288,
      "eval_samples_per_second": 7.334,
      "eval_steps_per_second": 7.334,
      "step": 2450
    },
    {
      "epoch": 3.7030764888512557,
      "grad_norm": 0.36296582221984863,
      "learning_rate": 1.9084537992754792e-05,
      "loss": 0.0064,
      "step": 2460
    },
    {
      "epoch": 3.718129645310001,
      "grad_norm": 0.05511193722486496,
      "learning_rate": 1.8673087136209803e-05,
      "loss": 0.0056,
      "step": 2470
    },
    {
      "epoch": 3.733182801768746,
      "grad_norm": 0.03270444646477699,
      "learning_rate": 1.8265099277000614e-05,
      "loss": 0.006,
      "step": 2480
    },
    {
      "epoch": 3.748235958227491,
      "grad_norm": 0.2531210780143738,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.0081,
      "step": 2490
    },
    {
      "epoch": 3.763289114686236,
      "grad_norm": 1.2437455654144287,
      "learning_rate": 1.7459692564974316e-05,
      "loss": 0.0103,
      "step": 2500
    },
    {
      "epoch": 3.763289114686236,
      "eval_loss": 0.04770328849554062,
      "eval_runtime": 161.0027,
      "eval_samples_per_second": 7.335,
      "eval_steps_per_second": 7.335,
      "step": 2500
    },
    {
      "epoch": 3.7783422711449806,
      "grad_norm": 0.21093598008155823,
      "learning_rate": 1.7062362744910322e-05,
      "loss": 0.0053,
      "step": 2510
    },
    {
      "epoch": 3.7933954276037256,
      "grad_norm": 0.33043161034584045,
      "learning_rate": 1.6668673977846254e-05,
      "loss": 0.0071,
      "step": 2520
    },
    {
      "epoch": 3.8084485840624707,
      "grad_norm": 0.1768001765012741,
      "learning_rate": 1.6278669783651395e-05,
      "loss": 0.0053,
      "step": 2530
    },
    {
      "epoch": 3.8235017405212153,
      "grad_norm": 0.09979119896888733,
      "learning_rate": 1.589239327488812e-05,
      "loss": 0.0065,
      "step": 2540
    },
    {
      "epoch": 3.8385548969799603,
      "grad_norm": 0.06995889544487,
      "learning_rate": 1.5509887152046137e-05,
      "loss": 0.0066,
      "step": 2550
    },
    {
      "epoch": 3.8385548969799603,
      "eval_loss": 0.048698194324970245,
      "eval_runtime": 161.1062,
      "eval_samples_per_second": 7.331,
      "eval_steps_per_second": 7.331,
      "step": 2550
    },
    {
      "epoch": 3.8536080534387054,
      "grad_norm": 0.30948641896247864,
      "learning_rate": 1.5131193698822232e-05,
      "loss": 0.0085,
      "step": 2560
    },
    {
      "epoch": 3.8686612098974504,
      "grad_norm": 0.21722477674484253,
      "learning_rate": 1.4756354777446001e-05,
      "loss": 0.007,
      "step": 2570
    },
    {
      "epoch": 3.8837143663561955,
      "grad_norm": 0.4803875684738159,
      "learning_rate": 1.4385411824052342e-05,
      "loss": 0.0069,
      "step": 2580
    },
    {
      "epoch": 3.89876752281494,
      "grad_norm": 0.31911638379096985,
      "learning_rate": 1.4018405844100812e-05,
      "loss": 0.0084,
      "step": 2590
    },
    {
      "epoch": 3.913820679273685,
      "grad_norm": 0.5358151197433472,
      "learning_rate": 1.3655377407842812e-05,
      "loss": 0.0101,
      "step": 2600
    },
    {
      "epoch": 3.913820679273685,
      "eval_loss": 0.04765944555401802,
      "eval_runtime": 161.2114,
      "eval_samples_per_second": 7.326,
      "eval_steps_per_second": 7.326,
      "step": 2600
    },
    {
      "epoch": 3.92887383573243,
      "grad_norm": 0.04455338418483734,
      "learning_rate": 1.3296366645836822e-05,
      "loss": 0.0067,
      "step": 2610
    },
    {
      "epoch": 3.943926992191175,
      "grad_norm": 0.34614667296409607,
      "learning_rate": 1.2941413244512113e-05,
      "loss": 0.0088,
      "step": 2620
    },
    {
      "epoch": 3.95898014864992,
      "grad_norm": 0.2596482038497925,
      "learning_rate": 1.2590556441781725e-05,
      "loss": 0.0078,
      "step": 2630
    },
    {
      "epoch": 3.974033305108665,
      "grad_norm": 0.4142569601535797,
      "learning_rate": 1.2243835022705003e-05,
      "loss": 0.0102,
      "step": 2640
    },
    {
      "epoch": 3.98908646156741,
      "grad_norm": 0.302739679813385,
      "learning_rate": 1.1901287315199977e-05,
      "loss": 0.0104,
      "step": 2650
    },
    {
      "epoch": 3.98908646156741,
      "eval_loss": 0.04554565250873566,
      "eval_runtime": 161.125,
      "eval_samples_per_second": 7.33,
      "eval_steps_per_second": 7.33,
      "step": 2650
    },
    {
      "epoch": 4.004139618026155,
      "grad_norm": 0.37217289209365845,
      "learning_rate": 1.1562951185806676e-05,
      "loss": 0.0066,
      "step": 2660
    },
    {
      "epoch": 4.0191927744849,
      "grad_norm": 0.054262690246105194,
      "learning_rate": 1.1228864035501069e-05,
      "loss": 0.0038,
      "step": 2670
    },
    {
      "epoch": 4.034245930943645,
      "grad_norm": 0.3357163965702057,
      "learning_rate": 1.0899062795560573e-05,
      "loss": 0.0033,
      "step": 2680
    },
    {
      "epoch": 4.04929908740239,
      "grad_norm": 0.25357311964035034,
      "learning_rate": 1.0573583923481711e-05,
      "loss": 0.0047,
      "step": 2690
    },
    {
      "epoch": 4.064352243861134,
      "grad_norm": 0.34490883350372314,
      "learning_rate": 1.0252463398949792e-05,
      "loss": 0.0073,
      "step": 2700
    },
    {
      "epoch": 4.064352243861134,
      "eval_loss": 0.046002693474292755,
      "eval_runtime": 161.1762,
      "eval_samples_per_second": 7.327,
      "eval_steps_per_second": 7.327,
      "step": 2700
    },
    {
      "epoch": 4.07940540031988,
      "grad_norm": 0.12939685583114624,
      "learning_rate": 9.935736719861622e-06,
      "loss": 0.0033,
      "step": 2710
    },
    {
      "epoch": 4.0944585567786245,
      "grad_norm": 0.0437207855284214,
      "learning_rate": 9.62343889840151e-06,
      "loss": 0.0043,
      "step": 2720
    },
    {
      "epoch": 4.109511713237369,
      "grad_norm": 0.10753034800291061,
      "learning_rate": 9.315604457170768e-06,
      "loss": 0.0028,
      "step": 2730
    },
    {
      "epoch": 4.124564869696115,
      "grad_norm": 0.04752466827630997,
      "learning_rate": 9.012267425371513e-06,
      "loss": 0.0024,
      "step": 2740
    },
    {
      "epoch": 4.139618026154859,
      "grad_norm": 0.11159101128578186,
      "learning_rate": 8.71346133504498e-06,
      "loss": 0.003,
      "step": 2750
    },
    {
      "epoch": 4.139618026154859,
      "eval_loss": 0.04775209724903107,
      "eval_runtime": 161.2317,
      "eval_samples_per_second": 7.325,
      "eval_steps_per_second": 7.325,
      "step": 2750
    },
    {
      "epoch": 4.154671182613605,
      "grad_norm": 0.06603502482175827,
      "learning_rate": 8.419219217364654e-06,
      "loss": 0.0023,
      "step": 2760
    },
    {
      "epoch": 4.169724339072349,
      "grad_norm": 0.08250313252210617,
      "learning_rate": 8.129573598984997e-06,
      "loss": 0.0031,
      "step": 2770
    },
    {
      "epoch": 4.184777495531094,
      "grad_norm": 0.33163315057754517,
      "learning_rate": 7.844556498445788e-06,
      "loss": 0.0023,
      "step": 2780
    },
    {
      "epoch": 4.199830651989839,
      "grad_norm": 0.32914066314697266,
      "learning_rate": 7.564199422632579e-06,
      "loss": 0.0053,
      "step": 2790
    },
    {
      "epoch": 4.214883808448584,
      "grad_norm": 0.027555787935853004,
      "learning_rate": 7.288533363293959e-06,
      "loss": 0.0025,
      "step": 2800
    },
    {
      "epoch": 4.214883808448584,
      "eval_loss": 0.04977192357182503,
      "eval_runtime": 161.302,
      "eval_samples_per_second": 7.322,
      "eval_steps_per_second": 7.322,
      "step": 2800
    },
    {
      "epoch": 4.229936964907329,
      "grad_norm": 0.4597860872745514,
      "learning_rate": 7.017588793615498e-06,
      "loss": 0.0045,
      "step": 2810
    },
    {
      "epoch": 4.244990121366074,
      "grad_norm": 0.0674595907330513,
      "learning_rate": 6.751395664851135e-06,
      "loss": 0.0037,
      "step": 2820
    },
    {
      "epoch": 4.260043277824819,
      "grad_norm": 0.09473542869091034,
      "learning_rate": 6.489983403012312e-06,
      "loss": 0.0021,
      "step": 2830
    },
    {
      "epoch": 4.275096434283564,
      "grad_norm": 0.4772437810897827,
      "learning_rate": 6.233380905615049e-06,
      "loss": 0.003,
      "step": 2840
    },
    {
      "epoch": 4.290149590742309,
      "grad_norm": 0.14856751263141632,
      "learning_rate": 5.981616538485496e-06,
      "loss": 0.0051,
      "step": 2850
    },
    {
      "epoch": 4.290149590742309,
      "eval_loss": 0.0505627803504467,
      "eval_runtime": 161.4919,
      "eval_samples_per_second": 7.313,
      "eval_steps_per_second": 7.313,
      "step": 2850
    },
    {
      "epoch": 4.3052027472010534,
      "grad_norm": 0.07821979373693466,
      "learning_rate": 5.73471813262435e-06,
      "loss": 0.0021,
      "step": 2860
    },
    {
      "epoch": 4.320255903659799,
      "grad_norm": 0.07666800916194916,
      "learning_rate": 5.4927129811301715e-06,
      "loss": 0.0019,
      "step": 2870
    },
    {
      "epoch": 4.335309060118544,
      "grad_norm": 0.06206662952899933,
      "learning_rate": 5.255627836182453e-06,
      "loss": 0.0021,
      "step": 2880
    },
    {
      "epoch": 4.350362216577288,
      "grad_norm": 0.17196908593177795,
      "learning_rate": 5.0234889060842176e-06,
      "loss": 0.0042,
      "step": 2890
    },
    {
      "epoch": 4.365415373036034,
      "grad_norm": 0.05123843625187874,
      "learning_rate": 4.796321852364877e-06,
      "loss": 0.0078,
      "step": 2900
    },
    {
      "epoch": 4.365415373036034,
      "eval_loss": 0.051590368151664734,
      "eval_runtime": 161.4863,
      "eval_samples_per_second": 7.313,
      "eval_steps_per_second": 7.313,
      "step": 2900
    },
    {
      "epoch": 4.380468529494778,
      "grad_norm": 0.09933144599199295,
      "learning_rate": 4.5741517869435706e-06,
      "loss": 0.0016,
      "step": 2910
    },
    {
      "epoch": 4.395521685953524,
      "grad_norm": 0.1274539679288864,
      "learning_rate": 4.357003269353105e-06,
      "loss": 0.003,
      "step": 2920
    },
    {
      "epoch": 4.410574842412268,
      "grad_norm": 0.0829009860754013,
      "learning_rate": 4.144900304025101e-06,
      "loss": 0.0039,
      "step": 2930
    },
    {
      "epoch": 4.425627998871013,
      "grad_norm": 0.0254693403840065,
      "learning_rate": 3.937866337636459e-06,
      "loss": 0.0042,
      "step": 2940
    },
    {
      "epoch": 4.4406811553297585,
      "grad_norm": 0.22674445807933807,
      "learning_rate": 3.7359242565174423e-06,
      "loss": 0.0019,
      "step": 2950
    },
    {
      "epoch": 4.4406811553297585,
      "eval_loss": 0.05175633728504181,
      "eval_runtime": 161.6011,
      "eval_samples_per_second": 7.308,
      "eval_steps_per_second": 7.308,
      "step": 2950
    },
    {
      "epoch": 4.455734311788503,
      "grad_norm": 0.04479767754673958,
      "learning_rate": 3.539096384121743e-06,
      "loss": 0.0052,
      "step": 2960
    },
    {
      "epoch": 4.470787468247248,
      "grad_norm": 0.050800859928131104,
      "learning_rate": 3.34740447855878e-06,
      "loss": 0.0021,
      "step": 2970
    },
    {
      "epoch": 4.485840624705993,
      "grad_norm": 0.3891144096851349,
      "learning_rate": 3.160869730188465e-06,
      "loss": 0.0038,
      "step": 2980
    },
    {
      "epoch": 4.500893781164738,
      "grad_norm": 0.05956409499049187,
      "learning_rate": 2.9795127592787186e-06,
      "loss": 0.0038,
      "step": 2990
    },
    {
      "epoch": 4.515946937623482,
      "grad_norm": 0.3809262812137604,
      "learning_rate": 2.803353613726056e-06,
      "loss": 0.0029,
      "step": 3000
    },
    {
      "epoch": 4.515946937623482,
      "eval_loss": 0.05206470191478729,
      "eval_runtime": 161.5497,
      "eval_samples_per_second": 7.31,
      "eval_steps_per_second": 7.31,
      "step": 3000
    },
    {
      "epoch": 4.531000094082228,
      "grad_norm": 0.08371811360120773,
      "learning_rate": 2.6324117668393877e-06,
      "loss": 0.0039,
      "step": 3010
    },
    {
      "epoch": 4.5460532505409725,
      "grad_norm": 0.2737991511821747,
      "learning_rate": 2.466706115187406e-06,
      "loss": 0.002,
      "step": 3020
    },
    {
      "epoch": 4.561106406999718,
      "grad_norm": 0.08303394168615341,
      "learning_rate": 2.3062549765096364e-06,
      "loss": 0.002,
      "step": 3030
    },
    {
      "epoch": 4.576159563458463,
      "grad_norm": 0.29287809133529663,
      "learning_rate": 2.1510760876915505e-06,
      "loss": 0.0065,
      "step": 3040
    },
    {
      "epoch": 4.591212719917207,
      "grad_norm": 0.22874212265014648,
      "learning_rate": 2.0011866028038617e-06,
      "loss": 0.0048,
      "step": 3050
    },
    {
      "epoch": 4.591212719917207,
      "eval_loss": 0.052139151841402054,
      "eval_runtime": 161.5292,
      "eval_samples_per_second": 7.311,
      "eval_steps_per_second": 7.311,
      "step": 3050
    },
    {
      "epoch": 4.606265876375953,
      "grad_norm": 0.13931158185005188,
      "learning_rate": 1.8566030912062549e-06,
      "loss": 0.0022,
      "step": 3060
    },
    {
      "epoch": 4.621319032834697,
      "grad_norm": 0.13953697681427002,
      "learning_rate": 1.717341535715733e-06,
      "loss": 0.0026,
      "step": 3070
    },
    {
      "epoch": 4.636372189293443,
      "grad_norm": 0.10214455425739288,
      "learning_rate": 1.5834173308397982e-06,
      "loss": 0.0029,
      "step": 3080
    },
    {
      "epoch": 4.6514253457521875,
      "grad_norm": 0.14687424898147583,
      "learning_rate": 1.4548452810747403e-06,
      "loss": 0.0039,
      "step": 3090
    },
    {
      "epoch": 4.666478502210932,
      "grad_norm": 0.17296817898750305,
      "learning_rate": 1.33163959926903e-06,
      "loss": 0.0033,
      "step": 3100
    },
    {
      "epoch": 4.666478502210932,
      "eval_loss": 0.05249761790037155,
      "eval_runtime": 161.6303,
      "eval_samples_per_second": 7.307,
      "eval_steps_per_second": 7.307,
      "step": 3100
    },
    {
      "epoch": 4.681531658669678,
      "grad_norm": 0.11973539739847183,
      "learning_rate": 1.2138139050522023e-06,
      "loss": 0.0041,
      "step": 3110
    },
    {
      "epoch": 4.696584815128422,
      "grad_norm": 0.16553695499897003,
      "learning_rate": 1.101381223329301e-06,
      "loss": 0.0028,
      "step": 3120
    },
    {
      "epoch": 4.711637971587168,
      "grad_norm": 0.036532752215862274,
      "learning_rate": 9.943539828410342e-07,
      "loss": 0.0027,
      "step": 3130
    },
    {
      "epoch": 4.726691128045912,
      "grad_norm": 0.08354666829109192,
      "learning_rate": 8.927440147898702e-07,
      "loss": 0.0015,
      "step": 3140
    },
    {
      "epoch": 4.741744284504657,
      "grad_norm": 0.14718425273895264,
      "learning_rate": 7.96562551532154e-07,
      "loss": 0.002,
      "step": 3150
    },
    {
      "epoch": 4.741744284504657,
      "eval_loss": 0.05251290276646614,
      "eval_runtime": 161.6656,
      "eval_samples_per_second": 7.305,
      "eval_steps_per_second": 7.305,
      "step": 3150
    },
    {
      "epoch": 4.756797440963402,
      "grad_norm": 0.019017163664102554,
      "learning_rate": 7.05820225336451e-07,
      "loss": 0.0022,
      "step": 3160
    },
    {
      "epoch": 4.771850597422147,
      "grad_norm": 0.06493131816387177,
      "learning_rate": 6.20527067208232e-07,
      "loss": 0.002,
      "step": 3170
    },
    {
      "epoch": 4.786903753880892,
      "grad_norm": 0.13360609114170074,
      "learning_rate": 5.406925057809653e-07,
      "loss": 0.0026,
      "step": 3180
    },
    {
      "epoch": 4.801956910339637,
      "grad_norm": 0.26471206545829773,
      "learning_rate": 4.6632536627386756e-07,
      "loss": 0.002,
      "step": 3190
    },
    {
      "epoch": 4.817010066798382,
      "grad_norm": 0.1764497309923172,
      "learning_rate": 3.974338695163393e-07,
      "loss": 0.0017,
      "step": 3200
    },
    {
      "epoch": 4.817010066798382,
      "eval_loss": 0.05277745798230171,
      "eval_runtime": 161.6536,
      "eval_samples_per_second": 7.306,
      "eval_steps_per_second": 7.306,
      "step": 3200
    },
    {
      "epoch": 4.832063223257126,
      "grad_norm": 0.055088795721530914,
      "learning_rate": 3.3402563103916984e-07,
      "loss": 0.001,
      "step": 3210
    },
    {
      "epoch": 4.847116379715872,
      "grad_norm": 0.059630122035741806,
      "learning_rate": 2.7610766023271615e-07,
      "loss": 0.0013,
      "step": 3220
    },
    {
      "epoch": 4.8621695361746164,
      "grad_norm": 0.031690843403339386,
      "learning_rate": 2.2368635957205618e-07,
      "loss": 0.0029,
      "step": 3230
    },
    {
      "epoch": 4.877222692633362,
      "grad_norm": 0.09654844552278519,
      "learning_rate": 1.7676752390920482e-07,
      "loss": 0.0038,
      "step": 3240
    },
    {
      "epoch": 4.8922758490921066,
      "grad_norm": 0.6978961825370789,
      "learning_rate": 1.3535633983257078e-07,
      "loss": 0.003,
      "step": 3250
    },
    {
      "epoch": 4.8922758490921066,
      "eval_loss": 0.05285593494772911,
      "eval_runtime": 161.6775,
      "eval_samples_per_second": 7.305,
      "eval_steps_per_second": 7.305,
      "step": 3250
    },
    {
      "epoch": 4.907329005550851,
      "grad_norm": 0.05225907266139984,
      "learning_rate": 9.945738509358205e-08,
      "loss": 0.0028,
      "step": 3260
    },
    {
      "epoch": 4.922382162009597,
      "grad_norm": 0.04852062091231346,
      "learning_rate": 6.907462810065158e-08,
      "loss": 0.0015,
      "step": 3270
    },
    {
      "epoch": 4.937435318468341,
      "grad_norm": 0.24554955959320068,
      "learning_rate": 4.4211427480500554e-08,
      "loss": 0.0056,
      "step": 3280
    },
    {
      "epoch": 4.952488474927087,
      "grad_norm": 0.01633327081799507,
      "learning_rate": 2.4870531706872034e-08,
      "loss": 0.0038,
      "step": 3290
    },
    {
      "epoch": 4.967541631385831,
      "grad_norm": 0.20463958382606506,
      "learning_rate": 1.105407879670728e-08,
      "loss": 0.0021,
      "step": 3300
    },
    {
      "epoch": 4.967541631385831,
      "eval_loss": 0.05255585163831711,
      "eval_runtime": 161.6589,
      "eval_samples_per_second": 7.306,
      "eval_steps_per_second": 7.306,
      "step": 3300
    },
    {
      "epoch": 4.982594787844576,
      "grad_norm": 0.07524765282869339,
      "learning_rate": 2.763596073807051e-09,
      "loss": 0.0022,
      "step": 3310
    },
    {
      "epoch": 4.9976479443033215,
      "grad_norm": 0.0663004145026207,
      "learning_rate": 0.0,
      "loss": 0.002,
      "step": 3320
    },
    {
      "epoch": 4.9976479443033215,
      "step": 3320,
      "total_flos": 8.339771893214085e+17,
      "train_loss": 0.04091864959056298,
      "train_runtime": 36302.3812,
      "train_samples_per_second": 1.464,
      "train_steps_per_second": 0.091
    }
  ],
  "logging_steps": 10,
  "max_steps": 3320,
  "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": 8.339771893214085e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}