{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1022,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.019569471624266144,
      "grad_norm": 2.358055099231819,
      "learning_rate": 6.451612903225806e-06,
      "loss": 0.6115,
      "step": 10
    },
    {
      "epoch": 0.03913894324853229,
      "grad_norm": 2.4319114197234546,
      "learning_rate": 1.2903225806451613e-05,
      "loss": 0.5445,
      "step": 20
    },
    {
      "epoch": 0.05870841487279843,
      "grad_norm": 2.0017713808509323,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.5314,
      "step": 30
    },
    {
      "epoch": 0.07827788649706457,
      "grad_norm": 2.135041689320543,
      "learning_rate": 1.999593015378676e-05,
      "loss": 0.5605,
      "step": 40
    },
    {
      "epoch": 0.09784735812133072,
      "grad_norm": 1.886107920458415,
      "learning_rate": 1.9981865802584795e-05,
      "loss": 0.5519,
      "step": 50
    },
    {
      "epoch": 0.11741682974559686,
      "grad_norm": 1.7545253281285733,
      "learning_rate": 1.9957770831058518e-05,
      "loss": 0.5715,
      "step": 60
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 1.9425827544859249,
      "learning_rate": 1.9923669451866753e-05,
      "loss": 0.5751,
      "step": 70
    },
    {
      "epoch": 0.15655577299412915,
      "grad_norm": 1.9306565295676892,
      "learning_rate": 1.987959593295039e-05,
      "loss": 0.5686,
      "step": 80
    },
    {
      "epoch": 0.1761252446183953,
      "grad_norm": 1.7564236049441155,
      "learning_rate": 1.9825594563097043e-05,
      "loss": 0.5782,
      "step": 90
    },
    {
      "epoch": 0.19569471624266144,
      "grad_norm": 1.6232157228047726,
      "learning_rate": 1.9761719607435973e-05,
      "loss": 0.57,
      "step": 100
    },
    {
      "epoch": 0.21526418786692758,
      "grad_norm": 1.8228974550233925,
      "learning_rate": 1.9688035252907888e-05,
      "loss": 0.568,
      "step": 110
    },
    {
      "epoch": 0.23483365949119372,
      "grad_norm": 1.7488528306352211,
      "learning_rate": 1.9604615543764506e-05,
      "loss": 0.5719,
      "step": 120
    },
    {
      "epoch": 0.25440313111545987,
      "grad_norm": 1.6679225084385316,
      "learning_rate": 1.9511544307162656e-05,
      "loss": 0.5661,
      "step": 130
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 1.5735195680219156,
      "learning_rate": 1.9408915068927653e-05,
      "loss": 0.5751,
      "step": 140
    },
    {
      "epoch": 0.29354207436399216,
      "grad_norm": 1.881773729389861,
      "learning_rate": 1.9296830959570697e-05,
      "loss": 0.567,
      "step": 150
    },
    {
      "epoch": 0.3131115459882583,
      "grad_norm": 1.669906241360482,
      "learning_rate": 1.917540461065462e-05,
      "loss": 0.5652,
      "step": 160
    },
    {
      "epoch": 0.33268101761252444,
      "grad_norm": 1.4933017744550008,
      "learning_rate": 1.9044758041612207e-05,
      "loss": 0.5682,
      "step": 170
    },
    {
      "epoch": 0.3522504892367906,
      "grad_norm": 1.7120959396931783,
      "learning_rate": 1.8905022537130774e-05,
      "loss": 0.571,
      "step": 180
    },
    {
      "epoch": 0.37181996086105673,
      "grad_norm": 1.8277560782891744,
      "learning_rate": 1.875633851522625e-05,
      "loss": 0.5774,
      "step": 190
    },
    {
      "epoch": 0.3913894324853229,
      "grad_norm": 1.4948184800706057,
      "learning_rate": 1.859885538613932e-05,
      "loss": 0.5694,
      "step": 200
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 1.5917802493859525,
      "learning_rate": 1.843273140219541e-05,
      "loss": 0.5697,
      "step": 210
    },
    {
      "epoch": 0.43052837573385516,
      "grad_norm": 1.6561321000655207,
      "learning_rate": 1.8258133498779407e-05,
      "loss": 0.5664,
      "step": 220
    },
    {
      "epoch": 0.4500978473581213,
      "grad_norm": 1.6233826437999508,
      "learning_rate": 1.807523712658493e-05,
      "loss": 0.5543,
      "step": 230
    },
    {
      "epoch": 0.46966731898238745,
      "grad_norm": 1.5920461634982808,
      "learning_rate": 1.7884226075306652e-05,
      "loss": 0.5717,
      "step": 240
    },
    {
      "epoch": 0.4892367906066536,
      "grad_norm": 1.5142935183933208,
      "learning_rate": 1.768529228895294e-05,
      "loss": 0.5511,
      "step": 250
    },
    {
      "epoch": 0.5088062622309197,
      "grad_norm": 1.5778513000637655,
      "learning_rate": 1.7478635672964324e-05,
      "loss": 0.5704,
      "step": 260
    },
    {
      "epoch": 0.5283757338551859,
      "grad_norm": 1.4360677799071841,
      "learning_rate": 1.726446389333166e-05,
      "loss": 0.55,
      "step": 270
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 1.7777152878321898,
      "learning_rate": 1.7042992167915836e-05,
      "loss": 0.5601,
      "step": 280
    },
    {
      "epoch": 0.5675146771037182,
      "grad_norm": 1.6013006228828397,
      "learning_rate": 1.6814443050178713e-05,
      "loss": 0.5597,
      "step": 290
    },
    {
      "epoch": 0.5870841487279843,
      "grad_norm": 1.522062257999709,
      "learning_rate": 1.6579046205542656e-05,
      "loss": 0.5425,
      "step": 300
    },
    {
      "epoch": 0.6066536203522505,
      "grad_norm": 1.9444890864863704,
      "learning_rate": 1.6337038180603332e-05,
      "loss": 0.5588,
      "step": 310
    },
    {
      "epoch": 0.6262230919765166,
      "grad_norm": 1.6256702214704648,
      "learning_rate": 1.6088662165427767e-05,
      "loss": 0.5656,
      "step": 320
    },
    {
      "epoch": 0.6457925636007827,
      "grad_norm": 1.6987309599785507,
      "learning_rate": 1.583416774917647e-05,
      "loss": 0.5464,
      "step": 330
    },
    {
      "epoch": 0.6653620352250489,
      "grad_norm": 1.536714185544068,
      "learning_rate": 1.5573810669295176e-05,
      "loss": 0.5548,
      "step": 340
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 1.4517361145984635,
      "learning_rate": 1.5307852554528318e-05,
      "loss": 0.5412,
      "step": 350
    },
    {
      "epoch": 0.7045009784735812,
      "grad_norm": 1.379468851703595,
      "learning_rate": 1.5036560662012405e-05,
      "loss": 0.5548,
      "step": 360
    },
    {
      "epoch": 0.7240704500978473,
      "grad_norm": 1.560556194482889,
      "learning_rate": 1.4760207608713515e-05,
      "loss": 0.5468,
      "step": 370
    },
    {
      "epoch": 0.7436399217221135,
      "grad_norm": 1.5799605015754972,
      "learning_rate": 1.4479071097478778e-05,
      "loss": 0.5356,
      "step": 380
    },
    {
      "epoch": 0.7632093933463796,
      "grad_norm": 1.490930077385001,
      "learning_rate": 1.4193433637977165e-05,
      "loss": 0.548,
      "step": 390
    },
    {
      "epoch": 0.7827788649706457,
      "grad_norm": 1.493190089036989,
      "learning_rate": 1.3903582262809918e-05,
      "loss": 0.5452,
      "step": 400
    },
    {
      "epoch": 0.8023483365949119,
      "grad_norm": 1.5176481135073754,
      "learning_rate": 1.3609808239076025e-05,
      "loss": 0.5514,
      "step": 410
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 1.8050366348817912,
      "learning_rate": 1.3312406775682471e-05,
      "loss": 0.5472,
      "step": 420
    },
    {
      "epoch": 0.8414872798434442,
      "grad_norm": 1.5083842404171728,
      "learning_rate": 1.3011676726693432e-05,
      "loss": 0.5526,
      "step": 430
    },
    {
      "epoch": 0.8610567514677103,
      "grad_norm": 1.4905244037584433,
      "learning_rate": 1.2707920291016526e-05,
      "loss": 0.535,
      "step": 440
    },
    {
      "epoch": 0.8806262230919765,
      "grad_norm": 1.5693631315219312,
      "learning_rate": 1.2401442708727869e-05,
      "loss": 0.5338,
      "step": 450
    },
    {
      "epoch": 0.9001956947162426,
      "grad_norm": 1.4345201206970122,
      "learning_rate": 1.2092551954341104e-05,
      "loss": 0.5371,
      "step": 460
    },
    {
      "epoch": 0.9197651663405088,
      "grad_norm": 1.524988255690081,
      "learning_rate": 1.1781558427328662e-05,
      "loss": 0.5252,
      "step": 470
    },
    {
      "epoch": 0.9393346379647749,
      "grad_norm": 1.4177625942380498,
      "learning_rate": 1.146877464020618e-05,
      "loss": 0.5225,
      "step": 480
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 1.4410916116833656,
      "learning_rate": 1.1154514904493599e-05,
      "loss": 0.5265,
      "step": 490
    },
    {
      "epoch": 0.9784735812133072,
      "grad_norm": 1.455518822210514,
      "learning_rate": 1.083909501486844e-05,
      "loss": 0.5169,
      "step": 500
    },
    {
      "epoch": 0.9980430528375733,
      "grad_norm": 1.4301843665722596,
      "learning_rate": 1.0522831931828677e-05,
      "loss": 0.5229,
      "step": 510
    },
    {
      "epoch": 1.0176125244618395,
      "grad_norm": 1.8242107121619293,
      "learning_rate": 1.0206043463184127e-05,
      "loss": 0.2745,
      "step": 520
    },
    {
      "epoch": 1.0371819960861057,
      "grad_norm": 1.4326923249244197,
      "learning_rate": 9.889047944696354e-06,
      "loss": 0.2286,
      "step": 530
    },
    {
      "epoch": 1.0567514677103718,
      "grad_norm": 1.404291696764401,
      "learning_rate": 9.57216392018806e-06,
      "loss": 0.2274,
      "step": 540
    },
    {
      "epoch": 1.076320939334638,
      "grad_norm": 1.4383258106098322,
      "learning_rate": 9.255709821443399e-06,
      "loss": 0.2158,
      "step": 550
    },
    {
      "epoch": 1.095890410958904,
      "grad_norm": 1.5838369354979953,
      "learning_rate": 8.940003648220863e-06,
      "loss": 0.221,
      "step": 560
    },
    {
      "epoch": 1.1154598825831703,
      "grad_norm": 1.3260950619639504,
      "learning_rate": 8.625362648700332e-06,
      "loss": 0.2146,
      "step": 570
    },
    {
      "epoch": 1.1350293542074363,
      "grad_norm": 1.4928296689616758,
      "learning_rate": 8.31210300068534e-06,
      "loss": 0.2162,
      "step": 580
    },
    {
      "epoch": 1.1545988258317026,
      "grad_norm": 1.2720796331424087,
      "learning_rate": 8.000539493880972e-06,
      "loss": 0.2137,
      "step": 590
    },
    {
      "epoch": 1.1741682974559686,
      "grad_norm": 1.467401250476884,
      "learning_rate": 7.69098521356662e-06,
      "loss": 0.2075,
      "step": 600
    },
    {
      "epoch": 1.1937377690802349,
      "grad_norm": 1.6529789418415561,
      "learning_rate": 7.383751225981503e-06,
      "loss": 0.2079,
      "step": 610
    },
    {
      "epoch": 1.213307240704501,
      "grad_norm": 1.3912168807674363,
      "learning_rate": 7.079146265739079e-06,
      "loss": 0.1975,
      "step": 620
    },
    {
      "epoch": 1.2328767123287672,
      "grad_norm": 1.2032682160188106,
      "learning_rate": 6.777476425584486e-06,
      "loss": 0.1982,
      "step": 630
    },
    {
      "epoch": 1.2524461839530332,
      "grad_norm": 1.3762644242614392,
      "learning_rate": 6.479044848806739e-06,
      "loss": 0.1936,
      "step": 640
    },
    {
      "epoch": 1.2720156555772995,
      "grad_norm": 1.4113607994624038,
      "learning_rate": 6.184151424614795e-06,
      "loss": 0.1972,
      "step": 650
    },
    {
      "epoch": 1.2915851272015655,
      "grad_norm": 1.33710607093093,
      "learning_rate": 5.893092486783594e-06,
      "loss": 0.1959,
      "step": 660
    },
    {
      "epoch": 1.3111545988258317,
      "grad_norm": 1.2851278655561311,
      "learning_rate": 5.606160515872886e-06,
      "loss": 0.1967,
      "step": 670
    },
    {
      "epoch": 1.3307240704500978,
      "grad_norm": 1.354094894659071,
      "learning_rate": 5.323643845318135e-06,
      "loss": 0.1809,
      "step": 680
    },
    {
      "epoch": 1.350293542074364,
      "grad_norm": 1.3589718977357292,
      "learning_rate": 5.04582637168874e-06,
      "loss": 0.1797,
      "step": 690
    },
    {
      "epoch": 1.36986301369863,
      "grad_norm": 1.3320249853797883,
      "learning_rate": 4.772987269404855e-06,
      "loss": 0.1876,
      "step": 700
    },
    {
      "epoch": 1.3894324853228963,
      "grad_norm": 1.255711002395159,
      "learning_rate": 4.505400710199376e-06,
      "loss": 0.1738,
      "step": 710
    },
    {
      "epoch": 1.4090019569471623,
      "grad_norm": 1.3203552961467395,
      "learning_rate": 4.243335587607074e-06,
      "loss": 0.1731,
      "step": 720
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 1.4388318733364938,
      "learning_rate": 3.987055246757701e-06,
      "loss": 0.1748,
      "step": 730
    },
    {
      "epoch": 1.4481409001956946,
      "grad_norm": 1.1814268312854903,
      "learning_rate": 3.7368172197446007e-06,
      "loss": 0.1695,
      "step": 740
    },
    {
      "epoch": 1.467710371819961,
      "grad_norm": 1.4097124495060538,
      "learning_rate": 3.4928729668347616e-06,
      "loss": 0.1685,
      "step": 750
    },
    {
      "epoch": 1.487279843444227,
      "grad_norm": 1.378405688153761,
      "learning_rate": 3.2554676237803117e-06,
      "loss": 0.1684,
      "step": 760
    },
    {
      "epoch": 1.5068493150684932,
      "grad_norm": 1.229111519346911,
      "learning_rate": 3.0248397554854813e-06,
      "loss": 0.1683,
      "step": 770
    },
    {
      "epoch": 1.5264187866927594,
      "grad_norm": 1.3031304027710664,
      "learning_rate": 2.801221116276436e-06,
      "loss": 0.1655,
      "step": 780
    },
    {
      "epoch": 1.5459882583170255,
      "grad_norm": 1.2785430784043106,
      "learning_rate": 2.5848364170150307e-06,
      "loss": 0.1648,
      "step": 790
    },
    {
      "epoch": 1.5655577299412915,
      "grad_norm": 1.358399709979297,
      "learning_rate": 2.375903099290362e-06,
      "loss": 0.1597,
      "step": 800
    },
    {
      "epoch": 1.5851272015655578,
      "grad_norm": 1.2024572922385197,
      "learning_rate": 2.174631116915137e-06,
      "loss": 0.1561,
      "step": 810
    },
    {
      "epoch": 1.604696673189824,
      "grad_norm": 1.3460799576843692,
      "learning_rate": 1.981222724946383e-06,
      "loss": 0.1522,
      "step": 820
    },
    {
      "epoch": 1.62426614481409,
      "grad_norm": 1.3491783502711767,
      "learning_rate": 1.7958722764425119e-06,
      "loss": 0.1536,
      "step": 830
    },
    {
      "epoch": 1.643835616438356,
      "grad_norm": 1.3720639700420578,
      "learning_rate": 1.6187660271609773e-06,
      "loss": 0.147,
      "step": 840
    },
    {
      "epoch": 1.6634050880626223,
      "grad_norm": 1.3633119421184674,
      "learning_rate": 1.4500819483927898e-06,
      "loss": 0.1481,
      "step": 850
    },
    {
      "epoch": 1.6829745596868886,
      "grad_norm": 1.2542715045711834,
      "learning_rate": 1.2899895481219672e-06,
      "loss": 0.1509,
      "step": 860
    },
    {
      "epoch": 1.7025440313111546,
      "grad_norm": 1.2321547326015214,
      "learning_rate": 1.1386497006896058e-06,
      "loss": 0.1431,
      "step": 870
    },
    {
      "epoch": 1.7221135029354206,
      "grad_norm": 1.1527861182486887,
      "learning_rate": 9.962144851337863e-07,
      "loss": 0.1369,
      "step": 880
    },
    {
      "epoch": 1.741682974559687,
      "grad_norm": 1.1491235254021608,
      "learning_rate": 8.628270323677424e-07,
      "loss": 0.1469,
      "step": 890
    },
    {
      "epoch": 1.7612524461839532,
      "grad_norm": 1.4751342647673045,
      "learning_rate": 7.386213813498344e-07,
      "loss": 0.1533,
      "step": 900
    },
    {
      "epoch": 1.7808219178082192,
      "grad_norm": 1.2401144134897077,
      "learning_rate": 6.237223443899221e-07,
      "loss": 0.1413,
      "step": 910
    },
    {
      "epoch": 1.8003913894324852,
      "grad_norm": 1.1565076180160574,
      "learning_rate": 5.18245381727418e-07,
      "loss": 0.1448,
      "step": 920
    },
    {
      "epoch": 1.8199608610567515,
      "grad_norm": 1.162280466003293,
      "learning_rate": 4.222964855071154e-07,
      "loss": 0.1398,
      "step": 930
    },
    {
      "epoch": 1.8395303326810177,
      "grad_norm": 1.3768345558574515,
      "learning_rate": 3.359720732693361e-07,
      "loss": 0.1463,
      "step": 940
    },
    {
      "epoch": 1.8590998043052838,
      "grad_norm": 1.2842826772954274,
      "learning_rate": 2.5935889106146305e-07,
      "loss": 0.1452,
      "step": 950
    },
    {
      "epoch": 1.8786692759295498,
      "grad_norm": 1.27912475359879,
      "learning_rate": 1.9253392626819468e-07,
      "loss": 0.1409,
      "step": 960
    },
    {
      "epoch": 1.898238747553816,
      "grad_norm": 1.2881607138075466,
      "learning_rate": 1.3556433024813353e-07,
      "loss": 0.1456,
      "step": 970
    },
    {
      "epoch": 1.9178082191780823,
      "grad_norm": 1.1998515494235076,
      "learning_rate": 8.850735085443763e-08,
      "loss": 0.1456,
      "step": 980
    },
    {
      "epoch": 1.9373776908023483,
      "grad_norm": 1.2190320104381343,
      "learning_rate": 5.141027490735195e-08,
      "loss": 0.1488,
      "step": 990
    },
    {
      "epoch": 1.9569471624266144,
      "grad_norm": 1.1520725254335977,
      "learning_rate": 2.431038067642111e-08,
      "loss": 0.1424,
      "step": 1000
    },
    {
      "epoch": 1.9765166340508806,
      "grad_norm": 1.1950165625346185,
      "learning_rate": 7.234900420147739e-09,
      "loss": 0.1372,
      "step": 1010
    },
    {
      "epoch": 1.9960861056751469,
      "grad_norm": 1.124102075180773,
      "learning_rate": 2.0099302071807658e-10,
      "loss": 0.1414,
      "step": 1020
    },
    {
      "epoch": 2.0,
      "step": 1022,
      "total_flos": 183951682437120.0,
      "train_loss": 0.36399914105577713,
      "train_runtime": 9021.6815,
      "train_samples_per_second": 14.477,
      "train_steps_per_second": 0.113
    }
  ],
  "logging_steps": 10,
  "max_steps": 1022,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "total_flos": 183951682437120.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}