{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 963,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03115264797507788,
      "grad_norm": 4.738249253201136,
      "learning_rate": 1.0309278350515464e-06,
      "loss": 2.1741,
      "step": 10
    },
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 3.2401014424140167,
      "learning_rate": 2.061855670103093e-06,
      "loss": 2.1229,
      "step": 20
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 2.8415277325627266,
      "learning_rate": 3.0927835051546395e-06,
      "loss": 1.9732,
      "step": 30
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 2.736037961913757,
      "learning_rate": 4.123711340206186e-06,
      "loss": 1.9484,
      "step": 40
    },
    {
      "epoch": 0.1557632398753894,
      "grad_norm": 2.4003619066518302,
      "learning_rate": 5.154639175257732e-06,
      "loss": 1.8748,
      "step": 50
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 2.6391276979901694,
      "learning_rate": 6.185567010309279e-06,
      "loss": 1.8787,
      "step": 60
    },
    {
      "epoch": 0.21806853582554517,
      "grad_norm": 2.4942487849758694,
      "learning_rate": 7.216494845360825e-06,
      "loss": 1.8368,
      "step": 70
    },
    {
      "epoch": 0.24922118380062305,
      "grad_norm": 2.822084088437886,
      "learning_rate": 8.247422680412371e-06,
      "loss": 1.8894,
      "step": 80
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 2.6875520332918503,
      "learning_rate": 9.278350515463918e-06,
      "loss": 1.8245,
      "step": 90
    },
    {
      "epoch": 0.3115264797507788,
      "grad_norm": 2.6263381435665116,
      "learning_rate": 9.999703897419048e-06,
      "loss": 1.7866,
      "step": 100
    },
    {
      "epoch": 0.3426791277258567,
      "grad_norm": 2.7744012568631478,
      "learning_rate": 9.99444082710777e-06,
      "loss": 1.8159,
      "step": 110
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 2.6658024247528034,
      "learning_rate": 9.982605671302293e-06,
      "loss": 1.8021,
      "step": 120
    },
    {
      "epoch": 0.40498442367601245,
      "grad_norm": 2.625120262587981,
      "learning_rate": 9.9642140036491e-06,
      "loss": 1.7743,
      "step": 130
    },
    {
      "epoch": 0.43613707165109034,
      "grad_norm": 2.5142261111844966,
      "learning_rate": 9.93929002537839e-06,
      "loss": 1.7878,
      "step": 140
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 2.854232906840411,
      "learning_rate": 9.90786653345818e-06,
      "loss": 1.7857,
      "step": 150
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 2.8406719564272014,
      "learning_rate": 9.869984877437413e-06,
      "loss": 1.8149,
      "step": 160
    },
    {
      "epoch": 0.5295950155763239,
      "grad_norm": 2.700981263790333,
      "learning_rate": 9.82569490503491e-06,
      "loss": 1.7803,
      "step": 170
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 2.530124037261927,
      "learning_rate": 9.775054896545755e-06,
      "loss": 1.7613,
      "step": 180
    },
    {
      "epoch": 0.5919003115264797,
      "grad_norm": 2.2222205932023793,
      "learning_rate": 9.718131488151399e-06,
      "loss": 1.7738,
      "step": 190
    },
    {
      "epoch": 0.6230529595015576,
      "grad_norm": 2.5500930798593413,
      "learning_rate": 9.654999584234444e-06,
      "loss": 1.7506,
      "step": 200
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 2.3248583237139,
      "learning_rate": 9.585742258813447e-06,
      "loss": 1.7656,
      "step": 210
    },
    {
      "epoch": 0.6853582554517134,
      "grad_norm": 2.5583210039836257,
      "learning_rate": 9.51045064622747e-06,
      "loss": 1.7601,
      "step": 220
    },
    {
      "epoch": 0.7165109034267912,
      "grad_norm": 2.5253607291187095,
      "learning_rate": 9.429223821214213e-06,
      "loss": 1.7587,
      "step": 230
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 3.599519999833458,
      "learning_rate": 9.34216866853954e-06,
      "loss": 1.7371,
      "step": 240
    },
    {
      "epoch": 0.778816199376947,
      "grad_norm": 3.112140263611012,
      "learning_rate": 9.249399742349928e-06,
      "loss": 1.7399,
      "step": 250
    },
    {
      "epoch": 0.8099688473520249,
      "grad_norm": 2.460048723197972,
      "learning_rate": 9.151039115432946e-06,
      "loss": 1.7792,
      "step": 260
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 2.7818108539377695,
      "learning_rate": 9.047216218584105e-06,
      "loss": 1.7368,
      "step": 270
    },
    {
      "epoch": 0.8722741433021807,
      "grad_norm": 2.6593659104855902,
      "learning_rate": 8.93806767029143e-06,
      "loss": 1.7264,
      "step": 280
    },
    {
      "epoch": 0.9034267912772586,
      "grad_norm": 2.5183656701737074,
      "learning_rate": 8.823737096961916e-06,
      "loss": 1.6964,
      "step": 290
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 2.6798445502428265,
      "learning_rate": 8.704374943926386e-06,
      "loss": 1.765,
      "step": 300
    },
    {
      "epoch": 0.9657320872274143,
      "grad_norm": 2.5879778784700394,
      "learning_rate": 8.580138277471476e-06,
      "loss": 1.7533,
      "step": 310
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 3.013949460294022,
      "learning_rate": 8.45119057815922e-06,
      "loss": 1.7242,
      "step": 320
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 2.4197649218374995,
      "learning_rate": 8.317701525706226e-06,
      "loss": 1.5794,
      "step": 330
    },
    {
      "epoch": 1.0591900311526479,
      "grad_norm": 2.9627624008639906,
      "learning_rate": 8.179846775705504e-06,
      "loss": 1.4717,
      "step": 340
    },
    {
      "epoch": 1.0903426791277258,
      "grad_norm": 2.455280291962598,
      "learning_rate": 8.03780772848477e-06,
      "loss": 1.5367,
      "step": 350
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 2.658888202364778,
      "learning_rate": 7.891771290405351e-06,
      "loss": 1.5353,
      "step": 360
    },
    {
      "epoch": 1.1526479750778815,
      "grad_norm": 2.8313334702049024,
      "learning_rate": 7.741929627915814e-06,
      "loss": 1.5069,
      "step": 370
    },
    {
      "epoch": 1.1838006230529594,
      "grad_norm": 2.5138362094593383,
      "learning_rate": 7.588479914683954e-06,
      "loss": 1.5002,
      "step": 380
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 2.653704835678763,
      "learning_rate": 7.431624072139884e-06,
      "loss": 1.5774,
      "step": 390
    },
    {
      "epoch": 1.2461059190031152,
      "grad_norm": 2.7594654733051924,
      "learning_rate": 7.271568503771632e-06,
      "loss": 1.5323,
      "step": 400
    },
    {
      "epoch": 1.277258566978193,
      "grad_norm": 2.6291361296605604,
      "learning_rate": 7.108523823522891e-06,
      "loss": 1.5193,
      "step": 410
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 2.4716101194608635,
      "learning_rate": 6.942704578650312e-06,
      "loss": 1.5163,
      "step": 420
    },
    {
      "epoch": 1.3395638629283488,
      "grad_norm": 2.6790323258843354,
      "learning_rate": 6.774328967405035e-06,
      "loss": 1.5731,
      "step": 430
    },
    {
      "epoch": 1.3707165109034267,
      "grad_norm": 2.4220245385790635,
      "learning_rate": 6.603618551909935e-06,
      "loss": 1.4769,
      "step": 440
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 2.9692192247657476,
      "learning_rate": 6.430797966610436e-06,
      "loss": 1.538,
      "step": 450
    },
    {
      "epoch": 1.4330218068535825,
      "grad_norm": 2.681036765738111,
      "learning_rate": 6.256094622682493e-06,
      "loss": 1.5295,
      "step": 460
    },
    {
      "epoch": 1.4641744548286604,
      "grad_norm": 2.5464332369331233,
      "learning_rate": 6.079738408786753e-06,
      "loss": 1.5375,
      "step": 470
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 2.918141623936573,
      "learning_rate": 5.9019613885626235e-06,
      "loss": 1.483,
      "step": 480
    },
    {
      "epoch": 1.5264797507788161,
      "grad_norm": 2.6056089657043024,
      "learning_rate": 5.722997495260348e-06,
      "loss": 1.5378,
      "step": 490
    },
    {
      "epoch": 1.557632398753894,
      "grad_norm": 2.682345830391196,
      "learning_rate": 5.543082223912875e-06,
      "loss": 1.5105,
      "step": 500
    },
    {
      "epoch": 1.588785046728972,
      "grad_norm": 2.418071341611168,
      "learning_rate": 5.362452321452636e-06,
      "loss": 1.5186,
      "step": 510
    },
    {
      "epoch": 1.6199376947040498,
      "grad_norm": 2.7761981641005105,
      "learning_rate": 5.181345475180941e-06,
      "loss": 1.4844,
      "step": 520
    },
    {
      "epoch": 1.6510903426791277,
      "grad_norm": 2.630104175186077,
      "learning_rate": 5e-06,
      "loss": 1.5189,
      "step": 530
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 2.7160409762576845,
      "learning_rate": 4.8186545248190604e-06,
      "loss": 1.5063,
      "step": 540
    },
    {
      "epoch": 1.7133956386292835,
      "grad_norm": 2.7941182969523237,
      "learning_rate": 4.637547678547366e-06,
      "loss": 1.5104,
      "step": 550
    },
    {
      "epoch": 1.7445482866043613,
      "grad_norm": 2.7635210894959186,
      "learning_rate": 4.4569177760871255e-06,
      "loss": 1.4964,
      "step": 560
    },
    {
      "epoch": 1.7757009345794392,
      "grad_norm": 2.7040452036881253,
      "learning_rate": 4.277002504739653e-06,
      "loss": 1.5406,
      "step": 570
    },
    {
      "epoch": 1.8068535825545171,
      "grad_norm": 3.0310876611935678,
      "learning_rate": 4.098038611437377e-06,
      "loss": 1.5634,
      "step": 580
    },
    {
      "epoch": 1.838006230529595,
      "grad_norm": 2.8756337522810007,
      "learning_rate": 3.920261591213249e-06,
      "loss": 1.4982,
      "step": 590
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 2.931182866108726,
      "learning_rate": 3.7439053773175092e-06,
      "loss": 1.525,
      "step": 600
    },
    {
      "epoch": 1.9003115264797508,
      "grad_norm": 2.618724233974035,
      "learning_rate": 3.569202033389565e-06,
      "loss": 1.5296,
      "step": 610
    },
    {
      "epoch": 1.9314641744548287,
      "grad_norm": 2.4527018980750133,
      "learning_rate": 3.3963814480900665e-06,
      "loss": 1.4938,
      "step": 620
    },
    {
      "epoch": 1.9626168224299065,
      "grad_norm": 2.5593961662264753,
      "learning_rate": 3.225671032594966e-06,
      "loss": 1.5196,
      "step": 630
    },
    {
      "epoch": 1.9937694704049844,
      "grad_norm": 2.6422408984177164,
      "learning_rate": 3.0572954213496897e-06,
      "loss": 1.5159,
      "step": 640
    },
    {
      "epoch": 2.0249221183800623,
      "grad_norm": 2.2409404747264996,
      "learning_rate": 2.8914761764771093e-06,
      "loss": 1.368,
      "step": 650
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 2.832245112153312,
      "learning_rate": 2.728431496228369e-06,
      "loss": 1.3305,
      "step": 660
    },
    {
      "epoch": 2.087227414330218,
      "grad_norm": 2.6890920333725554,
      "learning_rate": 2.5683759278601174e-06,
      "loss": 1.3275,
      "step": 670
    },
    {
      "epoch": 2.1183800623052957,
      "grad_norm": 2.7348466649941816,
      "learning_rate": 2.4115200853160475e-06,
      "loss": 1.3551,
      "step": 680
    },
    {
      "epoch": 2.149532710280374,
      "grad_norm": 2.7443347323397407,
      "learning_rate": 2.258070372084188e-06,
      "loss": 1.3282,
      "step": 690
    },
    {
      "epoch": 2.1806853582554515,
      "grad_norm": 2.58000682159267,
      "learning_rate": 2.108228709594649e-06,
      "loss": 1.3291,
      "step": 700
    },
    {
      "epoch": 2.2118380062305296,
      "grad_norm": 2.857457848907294,
      "learning_rate": 1.962192271515232e-06,
      "loss": 1.3544,
      "step": 710
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 2.9475901657988777,
      "learning_rate": 1.820153224294498e-06,
      "loss": 1.3479,
      "step": 720
    },
    {
      "epoch": 2.2741433021806854,
      "grad_norm": 2.2731328518925906,
      "learning_rate": 1.6822984742937764e-06,
      "loss": 1.3707,
      "step": 730
    },
    {
      "epoch": 2.305295950155763,
      "grad_norm": 2.7512861025748805,
      "learning_rate": 1.548809421840779e-06,
      "loss": 1.3778,
      "step": 740
    },
    {
      "epoch": 2.336448598130841,
      "grad_norm": 2.6460831192309184,
      "learning_rate": 1.4198617225285244e-06,
      "loss": 1.3371,
      "step": 750
    },
    {
      "epoch": 2.367601246105919,
      "grad_norm": 2.6430950689993016,
      "learning_rate": 1.2956250560736143e-06,
      "loss": 1.3529,
      "step": 760
    },
    {
      "epoch": 2.398753894080997,
      "grad_norm": 2.754383193984707,
      "learning_rate": 1.1762629030380867e-06,
      "loss": 1.3617,
      "step": 770
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 2.3375327059468627,
      "learning_rate": 1.061932329708572e-06,
      "loss": 1.3712,
      "step": 780
    },
    {
      "epoch": 2.4610591900311527,
      "grad_norm": 2.9641705209236675,
      "learning_rate": 9.527837814158963e-07,
      "loss": 1.3363,
      "step": 790
    },
    {
      "epoch": 2.4922118380062304,
      "grad_norm": 2.8810838394628457,
      "learning_rate": 8.489608845670527e-07,
      "loss": 1.3548,
      "step": 800
    },
    {
      "epoch": 2.5233644859813085,
      "grad_norm": 2.895341139217939,
      "learning_rate": 7.506002576500732e-07,
      "loss": 1.3483,
      "step": 810
    },
    {
      "epoch": 2.554517133956386,
      "grad_norm": 3.25688215356644,
      "learning_rate": 6.578313314604612e-07,
      "loss": 1.3509,
      "step": 820
    },
    {
      "epoch": 2.5856697819314642,
      "grad_norm": 2.265660300440991,
      "learning_rate": 5.707761787857879e-07,
      "loss": 1.3071,
      "step": 830
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 2.866761226251221,
      "learning_rate": 4.895493537725326e-07,
      "loss": 1.3367,
      "step": 840
    },
    {
      "epoch": 2.64797507788162,
      "grad_norm": 2.623512078489181,
      "learning_rate": 4.1425774118655505e-07,
      "loss": 1.3577,
      "step": 850
    },
    {
      "epoch": 2.6791277258566977,
      "grad_norm": 2.5389678031150305,
      "learning_rate": 3.4500041576555733e-07,
      "loss": 1.3349,
      "step": 860
    },
    {
      "epoch": 2.710280373831776,
      "grad_norm": 2.7619056373906736,
      "learning_rate": 2.818685118486025e-07,
      "loss": 1.36,
      "step": 870
    },
    {
      "epoch": 2.7414330218068534,
      "grad_norm": 2.6263120768496533,
      "learning_rate": 2.2494510345424657e-07,
      "loss": 1.3853,
      "step": 880
    },
    {
      "epoch": 2.7725856697819315,
      "grad_norm": 2.60206393215365,
      "learning_rate": 1.7430509496508985e-07,
      "loss": 1.3795,
      "step": 890
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 2.755154917816144,
      "learning_rate": 1.3001512256258841e-07,
      "loss": 1.3213,
      "step": 900
    },
    {
      "epoch": 2.8348909657320873,
      "grad_norm": 2.57799358687677,
      "learning_rate": 9.213346654182054e-08,
      "loss": 1.3557,
      "step": 910
    },
    {
      "epoch": 2.866043613707165,
      "grad_norm": 2.6277993338233108,
      "learning_rate": 6.070997462161055e-08,
      "loss": 1.3675,
      "step": 920
    },
    {
      "epoch": 2.897196261682243,
      "grad_norm": 2.3161497816289978,
      "learning_rate": 3.578599635090163e-08,
      "loss": 1.3401,
      "step": 930
    },
    {
      "epoch": 2.9283489096573208,
      "grad_norm": 2.383092955101212,
      "learning_rate": 1.7394328697707407e-08,
      "loss": 1.3287,
      "step": 940
    },
    {
      "epoch": 2.959501557632399,
      "grad_norm": 2.5504821012557124,
      "learning_rate": 5.5591728922316235e-09,
      "loss": 1.3864,
      "step": 950
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 2.459519377766809,
      "learning_rate": 2.9610258095169596e-10,
      "loss": 1.3746,
      "step": 960
    },
    {
      "epoch": 3.0,
      "step": 963,
      "total_flos": 5654321823744.0,
      "train_loss": 1.5619763468903791,
      "train_runtime": 395.7248,
      "train_samples_per_second": 155.411,
      "train_steps_per_second": 2.434
    }
  ],
  "logging_steps": 10,
  "max_steps": 963,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 20000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5654321823744.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}