{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.9921414538310414,
  "eval_steps": 50000,
  "global_step": 1016,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03929273084479371,
      "grad_norm": 4.240847110748291,
      "learning_rate": 1.6129032258064516e-06,
      "loss": 0.8413,
      "step": 10
    },
    {
      "epoch": 0.07858546168958742,
      "grad_norm": 2.775519371032715,
      "learning_rate": 3.225806451612903e-06,
      "loss": 0.8393,
      "step": 20
    },
    {
      "epoch": 0.11787819253438114,
      "grad_norm": 1.6855406761169434,
      "learning_rate": 4.838709677419355e-06,
      "loss": 0.6735,
      "step": 30
    },
    {
      "epoch": 0.15717092337917485,
      "grad_norm": 1.2230864763259888,
      "learning_rate": 4.998970106077018e-06,
      "loss": 0.5542,
      "step": 40
    },
    {
      "epoch": 0.19646365422396855,
      "grad_norm": 1.1485282182693481,
      "learning_rate": 4.99541106832608e-06,
      "loss": 0.5265,
      "step": 50
    },
    {
      "epoch": 0.2357563850687623,
      "grad_norm": 1.0626362562179565,
      "learning_rate": 4.989313791265896e-06,
      "loss": 0.6019,
      "step": 60
    },
    {
      "epoch": 0.275049115913556,
      "grad_norm": 1.3240628242492676,
      "learning_rate": 4.9806844768198724e-06,
      "loss": 0.5705,
      "step": 70
    },
    {
      "epoch": 0.3143418467583497,
      "grad_norm": 1.8527514934539795,
      "learning_rate": 4.969531902405652e-06,
      "loss": 0.5798,
      "step": 80
    },
    {
      "epoch": 0.35363457760314343,
      "grad_norm": 1.7867568731307983,
      "learning_rate": 4.955867412007052e-06,
      "loss": 0.5768,
      "step": 90
    },
    {
      "epoch": 0.3929273084479371,
      "grad_norm": 1.3669260740280151,
      "learning_rate": 4.939704904635388e-06,
      "loss": 0.5416,
      "step": 100
    },
    {
      "epoch": 0.43222003929273084,
      "grad_norm": 1.1892198324203491,
      "learning_rate": 4.921060820191909e-06,
      "loss": 0.5135,
      "step": 110
    },
    {
      "epoch": 0.4715127701375246,
      "grad_norm": 1.509191632270813,
      "learning_rate": 4.8999541227457514e-06,
      "loss": 0.5646,
      "step": 120
    },
    {
      "epoch": 0.5108055009823183,
      "grad_norm": 1.106804609298706,
      "learning_rate": 4.8764062812443875e-06,
      "loss": 0.512,
      "step": 130
    },
    {
      "epoch": 0.550098231827112,
      "grad_norm": 1.2564148902893066,
      "learning_rate": 4.8504412476762105e-06,
      "loss": 0.5396,
      "step": 140
    },
    {
      "epoch": 0.5893909626719057,
      "grad_norm": 1.2446550130844116,
      "learning_rate": 4.822085432707465e-06,
      "loss": 0.5246,
      "step": 150
    },
    {
      "epoch": 0.6286836935166994,
      "grad_norm": 0.9773355722427368,
      "learning_rate": 4.791367678818299e-06,
      "loss": 0.498,
      "step": 160
    },
    {
      "epoch": 0.6679764243614931,
      "grad_norm": 1.3318967819213867,
      "learning_rate": 4.758319230965267e-06,
      "loss": 0.5106,
      "step": 170
    },
    {
      "epoch": 0.7072691552062869,
      "grad_norm": 1.806362271308899,
      "learning_rate": 4.72297370480012e-06,
      "loss": 0.4969,
      "step": 180
    },
    {
      "epoch": 0.7465618860510805,
      "grad_norm": 1.4682146310806274,
      "learning_rate": 4.685367052477218e-06,
      "loss": 0.4937,
      "step": 190
    },
    {
      "epoch": 0.7858546168958742,
      "grad_norm": 1.461862325668335,
      "learning_rate": 4.645537526084331e-06,
      "loss": 0.4722,
      "step": 200
    },
    {
      "epoch": 0.825147347740668,
      "grad_norm": 1.3028712272644043,
      "learning_rate": 4.603525638734049e-06,
      "loss": 0.5044,
      "step": 210
    },
    {
      "epoch": 0.8644400785854617,
      "grad_norm": 1.2041152715682983,
      "learning_rate": 4.559374123355337e-06,
      "loss": 0.5001,
      "step": 220
    },
    {
      "epoch": 0.9037328094302554,
      "grad_norm": 1.3553584814071655,
      "learning_rate": 4.5131278892272e-06,
      "loss": 0.4411,
      "step": 230
    },
    {
      "epoch": 0.9430255402750491,
      "grad_norm": 1.3149176836013794,
      "learning_rate": 4.46483397629863e-06,
      "loss": 0.4714,
      "step": 240
    },
    {
      "epoch": 0.9823182711198428,
      "grad_norm": 1.2405178546905518,
      "learning_rate": 4.414541507341323e-06,
      "loss": 0.4632,
      "step": 250
    },
    {
      "epoch": 1.0216110019646365,
      "grad_norm": 1.4180119037628174,
      "learning_rate": 4.362301637983815e-06,
      "loss": 0.4177,
      "step": 260
    },
    {
      "epoch": 1.0609037328094302,
      "grad_norm": 1.6266894340515137,
      "learning_rate": 4.308167504677893e-06,
      "loss": 0.404,
      "step": 270
    },
    {
      "epoch": 1.1001964636542239,
      "grad_norm": 1.8860481977462769,
      "learning_rate": 4.2521941706501625e-06,
      "loss": 0.385,
      "step": 280
    },
    {
      "epoch": 1.1394891944990178,
      "grad_norm": 1.4382065534591675,
      "learning_rate": 4.194438569893784e-06,
      "loss": 0.3881,
      "step": 290
    },
    {
      "epoch": 1.1787819253438114,
      "grad_norm": 1.4116265773773193,
      "learning_rate": 4.134959449257335e-06,
      "loss": 0.3739,
      "step": 300
    },
    {
      "epoch": 1.218074656188605,
      "grad_norm": 1.192594289779663,
      "learning_rate": 4.0738173086896995e-06,
      "loss": 0.3687,
      "step": 310
    },
    {
      "epoch": 1.2573673870333988,
      "grad_norm": 1.2677867412567139,
      "learning_rate": 4.011074339701772e-06,
      "loss": 0.3822,
      "step": 320
    },
    {
      "epoch": 1.2966601178781925,
      "grad_norm": 1.2392891645431519,
      "learning_rate": 3.946794362107564e-06,
      "loss": 0.3787,
      "step": 330
    },
    {
      "epoch": 1.3359528487229864,
      "grad_norm": 1.810904622077942,
      "learning_rate": 3.8810427591090635e-06,
      "loss": 0.4118,
      "step": 340
    },
    {
      "epoch": 1.37524557956778,
      "grad_norm": 1.9088846445083618,
      "learning_rate": 3.813886410790879e-06,
      "loss": 0.3854,
      "step": 350
    },
    {
      "epoch": 1.4145383104125737,
      "grad_norm": 1.1440798044204712,
      "learning_rate": 3.7453936260922983e-06,
      "loss": 0.3636,
      "step": 360
    },
    {
      "epoch": 1.4538310412573674,
      "grad_norm": 1.5531620979309082,
      "learning_rate": 3.675634073325981e-06,
      "loss": 0.4106,
      "step": 370
    },
    {
      "epoch": 1.493123772102161,
      "grad_norm": 1.4519999027252197,
      "learning_rate": 3.6046787093139415e-06,
      "loss": 0.3809,
      "step": 380
    },
    {
      "epoch": 1.5324165029469548,
      "grad_norm": 1.8616143465042114,
      "learning_rate": 3.5325997072129066e-06,
      "loss": 0.3773,
      "step": 390
    },
    {
      "epoch": 1.5717092337917484,
      "grad_norm": 1.5181927680969238,
      "learning_rate": 3.4594703831024723e-06,
      "loss": 0.394,
      "step": 400
    },
    {
      "epoch": 1.611001964636542,
      "grad_norm": 1.901669979095459,
      "learning_rate": 3.385365121410706e-06,
      "loss": 0.366,
      "step": 410
    },
    {
      "epoch": 1.650294695481336,
      "grad_norm": 1.4574675559997559,
      "learning_rate": 3.310359299253082e-06,
      "loss": 0.3554,
      "step": 420
    },
    {
      "epoch": 1.6895874263261297,
      "grad_norm": 1.6998835802078247,
      "learning_rate": 3.234529209761676e-06,
      "loss": 0.3888,
      "step": 430
    },
    {
      "epoch": 1.7288801571709234,
      "grad_norm": 1.4896142482757568,
      "learning_rate": 3.157951984482635e-06,
      "loss": 0.3853,
      "step": 440
    },
    {
      "epoch": 1.768172888015717,
      "grad_norm": 1.1636215448379517,
      "learning_rate": 3.080705514920836e-06,
      "loss": 0.4516,
      "step": 450
    },
    {
      "epoch": 1.807465618860511,
      "grad_norm": 1.5409859418869019,
      "learning_rate": 3.0028683733115417e-06,
      "loss": 0.3768,
      "step": 460
    },
    {
      "epoch": 1.8467583497053046,
      "grad_norm": 1.0910699367523193,
      "learning_rate": 2.9245197326996515e-06,
      "loss": 0.3674,
      "step": 470
    },
    {
      "epoch": 1.8860510805500983,
      "grad_norm": 1.6574006080627441,
      "learning_rate": 2.845739286407821e-06,
      "loss": 0.3872,
      "step": 480
    },
    {
      "epoch": 1.925343811394892,
      "grad_norm": 1.2492740154266357,
      "learning_rate": 2.7666071669753807e-06,
      "loss": 0.3565,
      "step": 490
    },
    {
      "epoch": 1.9646365422396856,
      "grad_norm": 1.5630059242248535,
      "learning_rate": 2.687203864650497e-06,
      "loss": 0.3736,
      "step": 500
    },
    {
      "epoch": 2.0039292730844793,
      "grad_norm": 1.4160641431808472,
      "learning_rate": 2.6076101455184867e-06,
      "loss": 0.3559,
      "step": 510
    },
    {
      "epoch": 2.043222003929273,
      "grad_norm": 1.2741782665252686,
      "learning_rate": 2.527906969349559e-06,
      "loss": 0.3241,
      "step": 520
    },
    {
      "epoch": 2.0825147347740667,
      "grad_norm": 1.7951622009277344,
      "learning_rate": 2.44817540724955e-06,
      "loss": 0.2905,
      "step": 530
    },
    {
      "epoch": 2.1218074656188604,
      "grad_norm": 2.1442809104919434,
      "learning_rate": 2.3684965591974084e-06,
      "loss": 0.3063,
      "step": 540
    },
    {
      "epoch": 2.161100196463654,
      "grad_norm": 1.2380913496017456,
      "learning_rate": 2.288951471553316e-06,
      "loss": 0.3883,
      "step": 550
    },
    {
      "epoch": 2.2003929273084477,
      "grad_norm": 1.4480243921279907,
      "learning_rate": 2.2096210546213397e-06,
      "loss": 0.2879,
      "step": 560
    },
    {
      "epoch": 2.239685658153242,
      "grad_norm": 1.1371639966964722,
      "learning_rate": 2.130586000350486e-06,
      "loss": 0.288,
      "step": 570
    },
    {
      "epoch": 2.2789783889980355,
      "grad_norm": 2.1175849437713623,
      "learning_rate": 2.0519267002578517e-06,
      "loss": 0.2835,
      "step": 580
    },
    {
      "epoch": 2.318271119842829,
      "grad_norm": 2.0246710777282715,
      "learning_rate": 1.9737231636573595e-06,
      "loss": 0.3274,
      "step": 590
    },
    {
      "epoch": 2.357563850687623,
      "grad_norm": 1.6658718585968018,
      "learning_rate": 1.8960549362772618e-06,
      "loss": 0.284,
      "step": 600
    },
    {
      "epoch": 2.3968565815324165,
      "grad_norm": 2.166947364807129,
      "learning_rate": 1.8190010193491869e-06,
      "loss": 0.3257,
      "step": 610
    },
    {
      "epoch": 2.43614931237721,
      "grad_norm": 1.4202911853790283,
      "learning_rate": 1.7426397892510244e-06,
      "loss": 0.3166,
      "step": 620
    },
    {
      "epoch": 2.475442043222004,
      "grad_norm": 1.6804779767990112,
      "learning_rate": 1.6670489177853854e-06,
      "loss": 0.3084,
      "step": 630
    },
    {
      "epoch": 2.5147347740667976,
      "grad_norm": 1.5360515117645264,
      "learning_rate": 1.5923052931747408e-06,
      "loss": 0.2706,
      "step": 640
    },
    {
      "epoch": 2.5540275049115913,
      "grad_norm": 1.8666387796401978,
      "learning_rate": 1.5184849418535763e-06,
      "loss": 0.3063,
      "step": 650
    },
    {
      "epoch": 2.593320235756385,
      "grad_norm": 1.715324878692627,
      "learning_rate": 1.4456629511371346e-06,
      "loss": 0.2612,
      "step": 660
    },
    {
      "epoch": 2.6326129666011786,
      "grad_norm": 1.3589706420898438,
      "learning_rate": 1.3739133928453884e-06,
      "loss": 0.2843,
      "step": 670
    },
    {
      "epoch": 2.6719056974459727,
      "grad_norm": 1.6187039613723755,
      "learning_rate": 1.3033092479599453e-06,
      "loss": 0.2794,
      "step": 680
    },
    {
      "epoch": 2.711198428290766,
      "grad_norm": 1.5836079120635986,
      "learning_rate": 1.2339223323905026e-06,
      "loss": 0.2591,
      "step": 690
    },
    {
      "epoch": 2.75049115913556,
      "grad_norm": 1.5338759422302246,
      "learning_rate": 1.1658232239263815e-06,
      "loss": 0.2775,
      "step": 700
    },
    {
      "epoch": 2.7897838899803538,
      "grad_norm": 1.7079037427902222,
      "learning_rate": 1.099081190447418e-06,
      "loss": 0.2829,
      "step": 710
    },
    {
      "epoch": 2.8290766208251474,
      "grad_norm": 1.3403918743133545,
      "learning_rate": 1.0337641194672608e-06,
      "loss": 0.278,
      "step": 720
    },
    {
      "epoch": 2.868369351669941,
      "grad_norm": 1.9498125314712524,
      "learning_rate": 9.699384490807113e-07,
      "loss": 0.303,
      "step": 730
    },
    {
      "epoch": 2.907662082514735,
      "grad_norm": 1.7675586938858032,
      "learning_rate": 9.076691003853666e-07,
      "loss": 0.2701,
      "step": 740
    },
    {
      "epoch": 2.9469548133595285,
      "grad_norm": 1.6067301034927368,
      "learning_rate": 8.470194114462965e-07,
      "loss": 0.2734,
      "step": 750
    },
    {
      "epoch": 2.986247544204322,
      "grad_norm": 1.7567888498306274,
      "learning_rate": 7.880510728709176e-07,
      "loss": 0.3013,
      "step": 760
    },
    {
      "epoch": 3.025540275049116,
      "grad_norm": 1.446516990661621,
      "learning_rate": 7.30824065059603e-07,
      "loss": 0.2958,
      "step": 770
    },
    {
      "epoch": 3.0648330058939095,
      "grad_norm": 2.272646427154541,
      "learning_rate": 6.753965971958487e-07,
      "loss": 0.2528,
      "step": 780
    },
    {
      "epoch": 3.104125736738703,
      "grad_norm": 1.5716464519500732,
      "learning_rate": 6.218250480380611e-07,
      "loss": 0.257,
      "step": 790
    },
    {
      "epoch": 3.143418467583497,
      "grad_norm": 1.6481252908706665,
      "learning_rate": 5.701639085731786e-07,
      "loss": 0.2777,
      "step": 800
    },
    {
      "epoch": 3.182711198428291,
      "grad_norm": 1.6308358907699585,
      "learning_rate": 5.204657265904664e-07,
      "loss": 0.2403,
      "step": 810
    },
    {
      "epoch": 3.2220039292730847,
      "grad_norm": 1.8249069452285767,
      "learning_rate": 4.7278105323186244e-07,
      "loss": 0.264,
      "step": 820
    },
    {
      "epoch": 3.2612966601178783,
      "grad_norm": 1.533990740776062,
      "learning_rate": 4.2715839157323384e-07,
      "loss": 0.236,
      "step": 830
    },
    {
      "epoch": 3.300589390962672,
      "grad_norm": 2.055413007736206,
      "learning_rate": 3.8364414728885414e-07,
      "loss": 0.2239,
      "step": 840
    },
    {
      "epoch": 3.3398821218074657,
      "grad_norm": 1.7431176900863647,
      "learning_rate": 3.4228258144927587e-07,
      "loss": 0.2219,
      "step": 850
    },
    {
      "epoch": 3.3791748526522594,
      "grad_norm": 1.1028379201889038,
      "learning_rate": 3.031157655006187e-07,
      "loss": 0.2589,
      "step": 860
    },
    {
      "epoch": 3.418467583497053,
      "grad_norm": 1.7137134075164795,
      "learning_rate": 2.6618353847105703e-07,
      "loss": 0.2515,
      "step": 870
    },
    {
      "epoch": 3.4577603143418467,
      "grad_norm": 1.3465744256973267,
      "learning_rate": 2.315234664480448e-07,
      "loss": 0.2438,
      "step": 880
    },
    {
      "epoch": 3.4970530451866404,
      "grad_norm": 1.597948431968689,
      "learning_rate": 1.9917080436748615e-07,
      "loss": 0.2253,
      "step": 890
    },
    {
      "epoch": 3.536345776031434,
      "grad_norm": 1.588294506072998,
      "learning_rate": 1.6915846015372855e-07,
      "loss": 0.2254,
      "step": 900
    },
    {
      "epoch": 3.5756385068762278,
      "grad_norm": 1.6670293807983398,
      "learning_rate": 1.4151696124684504e-07,
      "loss": 0.2322,
      "step": 910
    },
    {
      "epoch": 3.614931237721022,
      "grad_norm": 1.8248482942581177,
      "learning_rate": 1.1627442355125834e-07,
      "loss": 0.2412,
      "step": 920
    },
    {
      "epoch": 3.654223968565815,
      "grad_norm": 2.2712173461914062,
      "learning_rate": 9.345652283728828e-08,
      "loss": 0.2338,
      "step": 930
    },
    {
      "epoch": 3.6935166994106092,
      "grad_norm": 1.6087329387664795,
      "learning_rate": 7.308646862471319e-08,
      "loss": 0.2435,
      "step": 940
    },
    {
      "epoch": 3.732809430255403,
      "grad_norm": 2.1418673992156982,
      "learning_rate": 5.5184980574911617e-08,
      "loss": 0.253,
      "step": 950
    },
    {
      "epoch": 3.7721021611001966,
      "grad_norm": 1.753722071647644,
      "learning_rate": 3.9770267415590866e-08,
      "loss": 0.2306,
      "step": 960
    },
    {
      "epoch": 3.8113948919449903,
      "grad_norm": 2.183093309402466,
      "learning_rate": 2.6858008419548e-08,
      "loss": 0.2403,
      "step": 970
    },
    {
      "epoch": 3.850687622789784,
      "grad_norm": 2.0719070434570312,
      "learning_rate": 1.6461337456297187e-08,
      "loss": 0.2371,
      "step": 980
    },
    {
      "epoch": 3.8899803536345776,
      "grad_norm": 1.8452452421188354,
      "learning_rate": 8.590829632785513e-09,
      "loss": 0.2388,
      "step": 990
    },
    {
      "epoch": 3.9292730844793713,
      "grad_norm": 2.089444875717163,
      "learning_rate": 3.2544905367876134e-09,
      "loss": 0.2568,
      "step": 1000
    },
    {
      "epoch": 3.968565815324165,
      "grad_norm": 1.6948010921478271,
      "learning_rate": 4.577480939194079e-10,
      "loss": 0.3124,
      "step": 1010
    },
    {
      "epoch": 3.9921414538310414,
      "step": 1016,
      "total_flos": 1.8674965730112307e+17,
      "train_loss": 0.3695383257283939,
      "train_runtime": 3236.8261,
      "train_samples_per_second": 5.023,
      "train_steps_per_second": 0.314
    }
  ],
  "logging_steps": 10,
  "max_steps": 1016,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.8674965730112307e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}