{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1088,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009191176470588236,
      "grad_norm": 11.266937255859375,
      "learning_rate": 8.256880733944956e-07,
      "loss": 0.578,
      "step": 10
    },
    {
      "epoch": 0.01838235294117647,
      "grad_norm": 4.71966552734375,
      "learning_rate": 1.743119266055046e-06,
      "loss": 0.4698,
      "step": 20
    },
    {
      "epoch": 0.027573529411764705,
      "grad_norm": 2.152991771697998,
      "learning_rate": 2.6605504587155968e-06,
      "loss": 0.254,
      "step": 30
    },
    {
      "epoch": 0.03676470588235294,
      "grad_norm": 1.5794391632080078,
      "learning_rate": 3.5779816513761473e-06,
      "loss": 0.1899,
      "step": 40
    },
    {
      "epoch": 0.04595588235294118,
      "grad_norm": 2.2544736862182617,
      "learning_rate": 4.4954128440366975e-06,
      "loss": 0.1712,
      "step": 50
    },
    {
      "epoch": 0.05514705882352941,
      "grad_norm": 1.6429979801177979,
      "learning_rate": 5.412844036697248e-06,
      "loss": 0.1554,
      "step": 60
    },
    {
      "epoch": 0.06433823529411764,
      "grad_norm": 1.428564190864563,
      "learning_rate": 6.330275229357799e-06,
      "loss": 0.1509,
      "step": 70
    },
    {
      "epoch": 0.07352941176470588,
      "grad_norm": 1.1606788635253906,
      "learning_rate": 7.247706422018349e-06,
      "loss": 0.144,
      "step": 80
    },
    {
      "epoch": 0.08272058823529412,
      "grad_norm": 1.4222913980484009,
      "learning_rate": 8.1651376146789e-06,
      "loss": 0.1377,
      "step": 90
    },
    {
      "epoch": 0.09191176470588236,
      "grad_norm": 1.059951663017273,
      "learning_rate": 9.08256880733945e-06,
      "loss": 0.128,
      "step": 100
    },
    {
      "epoch": 0.10110294117647059,
      "grad_norm": 1.0378493070602417,
      "learning_rate": 1e-05,
      "loss": 0.1237,
      "step": 110
    },
    {
      "epoch": 0.11029411764705882,
      "grad_norm": 1.9791302680969238,
      "learning_rate": 9.99742583072674e-06,
      "loss": 0.1317,
      "step": 120
    },
    {
      "epoch": 0.11948529411764706,
      "grad_norm": 1.28194260597229,
      "learning_rate": 9.98970597344593e-06,
      "loss": 0.1258,
      "step": 130
    },
    {
      "epoch": 0.12867647058823528,
      "grad_norm": 1.228013515472412,
      "learning_rate": 9.976848377045343e-06,
      "loss": 0.1259,
      "step": 140
    },
    {
      "epoch": 0.13786764705882354,
      "grad_norm": 0.9104840755462646,
      "learning_rate": 9.958866280576803e-06,
      "loss": 0.119,
      "step": 150
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 0.957346498966217,
      "learning_rate": 9.935778199624394e-06,
      "loss": 0.1193,
      "step": 160
    },
    {
      "epoch": 0.15625,
      "grad_norm": 1.0997158288955688,
      "learning_rate": 9.90760790723954e-06,
      "loss": 0.1159,
      "step": 170
    },
    {
      "epoch": 0.16544117647058823,
      "grad_norm": 1.1610311269760132,
      "learning_rate": 9.874384409462673e-06,
      "loss": 0.1198,
      "step": 180
    },
    {
      "epoch": 0.17463235294117646,
      "grad_norm": 0.9462277889251709,
      "learning_rate": 9.836141915456646e-06,
      "loss": 0.1109,
      "step": 190
    },
    {
      "epoch": 0.18382352941176472,
      "grad_norm": 0.793471097946167,
      "learning_rate": 9.792919802282656e-06,
      "loss": 0.1158,
      "step": 200
    },
    {
      "epoch": 0.19301470588235295,
      "grad_norm": 1.2811907529830933,
      "learning_rate": 9.744762574354967e-06,
      "loss": 0.1148,
      "step": 210
    },
    {
      "epoch": 0.20220588235294118,
      "grad_norm": 0.8056549429893494,
      "learning_rate": 9.691719817616148e-06,
      "loss": 0.1054,
      "step": 220
    },
    {
      "epoch": 0.2113970588235294,
      "grad_norm": 0.9176375269889832,
      "learning_rate": 9.633846148480024e-06,
      "loss": 0.1097,
      "step": 230
    },
    {
      "epoch": 0.22058823529411764,
      "grad_norm": 0.8464528918266296,
      "learning_rate": 9.571201157594925e-06,
      "loss": 0.1055,
      "step": 240
    },
    {
      "epoch": 0.22977941176470587,
      "grad_norm": 0.8142296671867371,
      "learning_rate": 9.503849348485112e-06,
      "loss": 0.1046,
      "step": 250
    },
    {
      "epoch": 0.23897058823529413,
      "grad_norm": 0.8536105751991272,
      "learning_rate": 9.431860071133592e-06,
      "loss": 0.0973,
      "step": 260
    },
    {
      "epoch": 0.24816176470588236,
      "grad_norm": 1.1437208652496338,
      "learning_rate": 9.355307450574666e-06,
      "loss": 0.1032,
      "step": 270
    },
    {
      "epoch": 0.25735294117647056,
      "grad_norm": 1.4142391681671143,
      "learning_rate": 9.27427031056979e-06,
      "loss": 0.1057,
      "step": 280
    },
    {
      "epoch": 0.2665441176470588,
      "grad_norm": 0.7978008389472961,
      "learning_rate": 9.188832092445281e-06,
      "loss": 0.105,
      "step": 290
    },
    {
      "epoch": 0.2757352941176471,
      "grad_norm": 1.3588639497756958,
      "learning_rate": 9.09908076917548e-06,
      "loss": 0.1048,
      "step": 300
    },
    {
      "epoch": 0.2849264705882353,
      "grad_norm": 0.6924364566802979,
      "learning_rate": 9.00510875479983e-06,
      "loss": 0.1001,
      "step": 310
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.7341586947441101,
      "learning_rate": 8.907012809267107e-06,
      "loss": 0.1045,
      "step": 320
    },
    {
      "epoch": 0.30330882352941174,
      "grad_norm": 0.703032910823822,
      "learning_rate": 8.804893938804839e-06,
      "loss": 0.1029,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.8211079239845276,
      "learning_rate": 8.698857291916456e-06,
      "loss": 0.1,
      "step": 340
    },
    {
      "epoch": 0.32169117647058826,
      "grad_norm": 0.7473457455635071,
      "learning_rate": 8.58901205111326e-06,
      "loss": 0.0923,
      "step": 350
    },
    {
      "epoch": 0.33088235294117646,
      "grad_norm": 1.1483430862426758,
      "learning_rate": 8.475471320492728e-06,
      "loss": 0.1128,
      "step": 360
    },
    {
      "epoch": 0.3400735294117647,
      "grad_norm": 0.8971941471099854,
      "learning_rate": 8.35835200927887e-06,
      "loss": 0.1044,
      "step": 370
    },
    {
      "epoch": 0.3492647058823529,
      "grad_norm": 0.7846776247024536,
      "learning_rate": 8.237774711444575e-06,
      "loss": 0.0983,
      "step": 380
    },
    {
      "epoch": 0.3584558823529412,
      "grad_norm": 0.7909674644470215,
      "learning_rate": 8.113863581539905e-06,
      "loss": 0.0908,
      "step": 390
    },
    {
      "epoch": 0.36764705882352944,
      "grad_norm": 0.8026263117790222,
      "learning_rate": 7.986746206854143e-06,
      "loss": 0.0939,
      "step": 400
    },
    {
      "epoch": 0.37683823529411764,
      "grad_norm": 0.6986457705497742,
      "learning_rate": 7.856553476043294e-06,
      "loss": 0.0967,
      "step": 410
    },
    {
      "epoch": 0.3860294117647059,
      "grad_norm": 0.7079197764396667,
      "learning_rate": 7.723419444358261e-06,
      "loss": 0.0953,
      "step": 420
    },
    {
      "epoch": 0.3952205882352941,
      "grad_norm": 1.0171252489089966,
      "learning_rate": 7.5874811956124805e-06,
      "loss": 0.0928,
      "step": 430
    },
    {
      "epoch": 0.40441176470588236,
      "grad_norm": 0.8423835635185242,
      "learning_rate": 7.4488787010311425e-06,
      "loss": 0.0938,
      "step": 440
    },
    {
      "epoch": 0.41360294117647056,
      "grad_norm": 1.0884760618209839,
      "learning_rate": 7.3077546751273494e-06,
      "loss": 0.0925,
      "step": 450
    },
    {
      "epoch": 0.4227941176470588,
      "grad_norm": 0.7641892433166504,
      "learning_rate": 7.164254428753581e-06,
      "loss": 0.0953,
      "step": 460
    },
    {
      "epoch": 0.4319852941176471,
      "grad_norm": 0.8474388122558594,
      "learning_rate": 7.018525719479805e-06,
      "loss": 0.0913,
      "step": 470
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 0.8151172399520874,
      "learning_rate": 6.870718599452279e-06,
      "loss": 0.0974,
      "step": 480
    },
    {
      "epoch": 0.45036764705882354,
      "grad_norm": 0.5582044124603271,
      "learning_rate": 6.7209852608897005e-06,
      "loss": 0.0901,
      "step": 490
    },
    {
      "epoch": 0.45955882352941174,
      "grad_norm": 0.9465530514717102,
      "learning_rate": 6.569479879375795e-06,
      "loss": 0.0936,
      "step": 500
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.7654430270195007,
      "learning_rate": 6.416358455109695e-06,
      "loss": 0.0933,
      "step": 510
    },
    {
      "epoch": 0.47794117647058826,
      "grad_norm": 0.7761212587356567,
      "learning_rate": 6.261778652277565e-06,
      "loss": 0.093,
      "step": 520
    },
    {
      "epoch": 0.48713235294117646,
      "grad_norm": 0.5924690365791321,
      "learning_rate": 6.105899636710895e-06,
      "loss": 0.0879,
      "step": 530
    },
    {
      "epoch": 0.4963235294117647,
      "grad_norm": 0.591857373714447,
      "learning_rate": 5.948881911998572e-06,
      "loss": 0.0904,
      "step": 540
    },
    {
      "epoch": 0.5055147058823529,
      "grad_norm": 0.7316629886627197,
      "learning_rate": 5.790887154221521e-06,
      "loss": 0.0883,
      "step": 550
    },
    {
      "epoch": 0.5147058823529411,
      "grad_norm": 0.8223432302474976,
      "learning_rate": 5.632078045480065e-06,
      "loss": 0.0883,
      "step": 560
    },
    {
      "epoch": 0.5238970588235294,
      "grad_norm": 0.7148771286010742,
      "learning_rate": 5.472618106385415e-06,
      "loss": 0.0814,
      "step": 570
    },
    {
      "epoch": 0.5330882352941176,
      "grad_norm": 0.6548059582710266,
      "learning_rate": 5.31267152768779e-06,
      "loss": 0.0884,
      "step": 580
    },
    {
      "epoch": 0.5422794117647058,
      "grad_norm": 0.7036318778991699,
      "learning_rate": 5.152403001214483e-06,
      "loss": 0.0882,
      "step": 590
    },
    {
      "epoch": 0.5514705882352942,
      "grad_norm": 0.8155568242073059,
      "learning_rate": 4.991977550292028e-06,
      "loss": 0.0873,
      "step": 600
    },
    {
      "epoch": 0.5606617647058824,
      "grad_norm": 0.654234766960144,
      "learning_rate": 4.831560359826985e-06,
      "loss": 0.086,
      "step": 610
    },
    {
      "epoch": 0.5698529411764706,
      "grad_norm": 0.8868378400802612,
      "learning_rate": 4.671316606220394e-06,
      "loss": 0.0905,
      "step": 620
    },
    {
      "epoch": 0.5790441176470589,
      "grad_norm": 0.5300816297531128,
      "learning_rate": 4.511411287290964e-06,
      "loss": 0.0881,
      "step": 630
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.9155387282371521,
      "learning_rate": 4.35200905238214e-06,
      "loss": 0.088,
      "step": 640
    },
    {
      "epoch": 0.5974264705882353,
      "grad_norm": 0.7309725880622864,
      "learning_rate": 4.193274032828e-06,
      "loss": 0.0873,
      "step": 650
    },
    {
      "epoch": 0.6066176470588235,
      "grad_norm": 0.5966044664382935,
      "learning_rate": 4.035369672952516e-06,
      "loss": 0.0878,
      "step": 660
    },
    {
      "epoch": 0.6158088235294118,
      "grad_norm": 0.924511194229126,
      "learning_rate": 3.8784585617762084e-06,
      "loss": 0.0849,
      "step": 670
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.1838688850402832,
      "learning_rate": 3.7227022656034873e-06,
      "loss": 0.0867,
      "step": 680
    },
    {
      "epoch": 0.6341911764705882,
      "grad_norm": 0.7663735747337341,
      "learning_rate": 3.568261161663042e-06,
      "loss": 0.0896,
      "step": 690
    },
    {
      "epoch": 0.6433823529411765,
      "grad_norm": 0.7709589600563049,
      "learning_rate": 3.4152942729725896e-06,
      "loss": 0.0893,
      "step": 700
    },
    {
      "epoch": 0.6525735294117647,
      "grad_norm": 0.7545693516731262,
      "learning_rate": 3.263959104598009e-06,
      "loss": 0.0861,
      "step": 710
    },
    {
      "epoch": 0.6617647058823529,
      "grad_norm": 0.6413795948028564,
      "learning_rate": 3.114411481475455e-06,
      "loss": 0.0776,
      "step": 720
    },
    {
      "epoch": 0.6709558823529411,
      "grad_norm": 0.7467756271362305,
      "learning_rate": 2.966805387963463e-06,
      "loss": 0.084,
      "step": 730
    },
    {
      "epoch": 0.6801470588235294,
      "grad_norm": 0.6672917008399963,
      "learning_rate": 2.821292809290217e-06,
      "loss": 0.0834,
      "step": 740
    },
    {
      "epoch": 0.6893382352941176,
      "grad_norm": 0.7798132300376892,
      "learning_rate": 2.678023575059274e-06,
      "loss": 0.083,
      "step": 750
    },
    {
      "epoch": 0.6985294117647058,
      "grad_norm": 0.729383647441864,
      "learning_rate": 2.5371452049748603e-06,
      "loss": 0.0842,
      "step": 760
    },
    {
      "epoch": 0.7077205882352942,
      "grad_norm": 0.8981873989105225,
      "learning_rate": 2.3988027569455895e-06,
      "loss": 0.0874,
      "step": 770
    },
    {
      "epoch": 0.7169117647058824,
      "grad_norm": 0.7742981314659119,
      "learning_rate": 2.2631386777230248e-06,
      "loss": 0.0815,
      "step": 780
    },
    {
      "epoch": 0.7261029411764706,
      "grad_norm": 0.6819539070129395,
      "learning_rate": 2.130292656228856e-06,
      "loss": 0.0814,
      "step": 790
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 0.7190724611282349,
      "learning_rate": 2.0004014797217207e-06,
      "loss": 0.0826,
      "step": 800
    },
    {
      "epoch": 0.7444852941176471,
      "grad_norm": 0.674633800983429,
      "learning_rate": 1.873598892951795e-06,
      "loss": 0.0851,
      "step": 810
    },
    {
      "epoch": 0.7536764705882353,
      "grad_norm": 0.7690839767456055,
      "learning_rate": 1.7500154604481312e-06,
      "loss": 0.0804,
      "step": 820
    },
    {
      "epoch": 0.7628676470588235,
      "grad_norm": 0.6933445930480957,
      "learning_rate": 1.629778432080586e-06,
      "loss": 0.084,
      "step": 830
    },
    {
      "epoch": 0.7720588235294118,
      "grad_norm": 0.944789707660675,
      "learning_rate": 1.513011612034726e-06,
      "loss": 0.0776,
      "step": 840
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.645876944065094,
      "learning_rate": 1.3998352313346768e-06,
      "loss": 0.0797,
      "step": 850
    },
    {
      "epoch": 0.7904411764705882,
      "grad_norm": 0.72840416431427,
      "learning_rate": 1.2903658240450989e-06,
      "loss": 0.0792,
      "step": 860
    },
    {
      "epoch": 0.7996323529411765,
      "grad_norm": 0.8772525191307068,
      "learning_rate": 1.184716107279837e-06,
      "loss": 0.0804,
      "step": 870
    },
    {
      "epoch": 0.8088235294117647,
      "grad_norm": 0.8419274091720581,
      "learning_rate": 1.0829948651407374e-06,
      "loss": 0.0817,
      "step": 880
    },
    {
      "epoch": 0.8180147058823529,
      "grad_norm": 0.6790868043899536,
      "learning_rate": 9.85306836706184e-07,
      "loss": 0.0798,
      "step": 890
    },
    {
      "epoch": 0.8272058823529411,
      "grad_norm": 0.7450520992279053,
      "learning_rate": 8.917526081846411e-07,
      "loss": 0.0772,
      "step": 900
    },
    {
      "epoch": 0.8363970588235294,
      "grad_norm": 0.9222913980484009,
      "learning_rate": 8.024285093442874e-07,
      "loss": 0.0796,
      "step": 910
    },
    {
      "epoch": 0.8455882352941176,
      "grad_norm": 0.5793068408966064,
      "learning_rate": 7.17426514325359e-07,
      "loss": 0.0754,
      "step": 920
    },
    {
      "epoch": 0.8547794117647058,
      "grad_norm": 0.8552721738815308,
      "learning_rate": 6.36834146937354e-07,
      "loss": 0.0816,
      "step": 930
    },
    {
      "epoch": 0.8639705882352942,
      "grad_norm": 0.6834543943405151,
      "learning_rate": 5.607343905385898e-07,
      "loss": 0.077,
      "step": 940
    },
    {
      "epoch": 0.8731617647058824,
      "grad_norm": 0.8566415309906006,
      "learning_rate": 4.892056025909148e-07,
      "loss": 0.0788,
      "step": 950
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 0.6394233107566833,
      "learning_rate": 4.2232143397756607e-07,
      "loss": 0.0761,
      "step": 960
    },
    {
      "epoch": 0.8915441176470589,
      "grad_norm": 0.6848531365394592,
      "learning_rate": 3.6015075316722605e-07,
      "loss": 0.0762,
      "step": 970
    },
    {
      "epoch": 0.9007352941176471,
      "grad_norm": 0.7035109996795654,
      "learning_rate": 3.02757575302392e-07,
      "loss": 0.0827,
      "step": 980
    },
    {
      "epoch": 0.9099264705882353,
      "grad_norm": 1.1025725603103638,
      "learning_rate": 2.5020099628504603e-07,
      "loss": 0.0786,
      "step": 990
    },
    {
      "epoch": 0.9191176470588235,
      "grad_norm": 0.9067655801773071,
      "learning_rate": 2.0253513192751374e-07,
      "loss": 0.0713,
      "step": 1000
    },
    {
      "epoch": 0.9283088235294118,
      "grad_norm": 2.0745654106140137,
      "learning_rate": 1.5980906223115933e-07,
      "loss": 0.0823,
      "step": 1010
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.77626633644104,
      "learning_rate": 1.220667808502951e-07,
      "loss": 0.0797,
      "step": 1020
    },
    {
      "epoch": 0.9466911764705882,
      "grad_norm": 0.6542201042175293,
      "learning_rate": 8.934714979333403e-08,
      "loss": 0.0813,
      "step": 1030
    },
    {
      "epoch": 0.9558823529411765,
      "grad_norm": 0.7697402238845825,
      "learning_rate": 6.168385940783727e-08,
      "loss": 0.0745,
      "step": 1040
    },
    {
      "epoch": 0.9650735294117647,
      "grad_norm": 1.0100551843643188,
      "learning_rate": 3.910539369064603e-08,
      "loss": 0.0794,
      "step": 1050
    },
    {
      "epoch": 0.9742647058823529,
      "grad_norm": 0.5885272026062012,
      "learning_rate": 2.1635000958836748e-08,
      "loss": 0.0788,
      "step": 1060
    },
    {
      "epoch": 0.9834558823529411,
      "grad_norm": 0.6646335124969482,
      "learning_rate": 9.290669911672934e-09,
      "loss": 0.0792,
      "step": 1070
    },
    {
      "epoch": 0.9926470588235294,
      "grad_norm": 0.5683983564376831,
      "learning_rate": 2.085111108227067e-09,
      "loss": 0.0762,
      "step": 1080
    },
    {
      "epoch": 1.0,
      "step": 1088,
      "total_flos": 1.9483786471612088e+18,
      "train_loss": 0.10519510776023655,
      "train_runtime": 10705.8335,
      "train_samples_per_second": 3.25,
      "train_steps_per_second": 0.102
    }
  ],
  "logging_steps": 10,
  "max_steps": 1088,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1.9483786471612088e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}