{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.991394148020654,
  "eval_steps": 500,
  "global_step": 870,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03442340791738382,
      "grad_norm": 1101.9962158203125,
      "learning_rate": 2.2988505747126437e-06,
      "loss": 2.9358,
      "step": 10
    },
    {
      "epoch": 0.06884681583476764,
      "grad_norm": 112.010009765625,
      "learning_rate": 8.045977011494253e-06,
      "loss": 2.8836,
      "step": 20
    },
    {
      "epoch": 0.10327022375215146,
      "grad_norm": 70.39278411865234,
      "learning_rate": 1.3793103448275863e-05,
      "loss": 2.5828,
      "step": 30
    },
    {
      "epoch": 0.13769363166953527,
      "grad_norm": 10.689151763916016,
      "learning_rate": 1.9540229885057475e-05,
      "loss": 2.0904,
      "step": 40
    },
    {
      "epoch": 0.1721170395869191,
      "grad_norm": 21.27436637878418,
      "learning_rate": 2.5287356321839083e-05,
      "loss": 1.0867,
      "step": 50
    },
    {
      "epoch": 0.20654044750430292,
      "grad_norm": 191.7716064453125,
      "learning_rate": 3.103448275862069e-05,
      "loss": 0.4477,
      "step": 60
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 11.0482816696167,
      "learning_rate": 3.67816091954023e-05,
      "loss": 0.5028,
      "step": 70
    },
    {
      "epoch": 0.27538726333907054,
      "grad_norm": 2.7239110469818115,
      "learning_rate": 4.1379310344827587e-05,
      "loss": 0.3772,
      "step": 80
    },
    {
      "epoch": 0.3098106712564544,
      "grad_norm": 1.2920331954956055,
      "learning_rate": 4.7126436781609195e-05,
      "loss": 0.3437,
      "step": 90
    },
    {
      "epoch": 0.3442340791738382,
      "grad_norm": 0.8523675799369812,
      "learning_rate": 4.999496949392727e-05,
      "loss": 0.3507,
      "step": 100
    },
    {
      "epoch": 0.37865748709122204,
      "grad_norm": 0.7762532234191895,
      "learning_rate": 4.9954737591406555e-05,
      "loss": 0.2585,
      "step": 110
    },
    {
      "epoch": 0.41308089500860584,
      "grad_norm": 0.6530282497406006,
      "learning_rate": 4.9874338543634794e-05,
      "loss": 0.2455,
      "step": 120
    },
    {
      "epoch": 0.4475043029259897,
      "grad_norm": 1.0387413501739502,
      "learning_rate": 4.975390176091809e-05,
      "loss": 0.24,
      "step": 130
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.45220646262168884,
      "learning_rate": 4.959362109830007e-05,
      "loss": 0.2177,
      "step": 140
    },
    {
      "epoch": 0.5163511187607573,
      "grad_norm": 0.3468196392059326,
      "learning_rate": 4.9393754543532886e-05,
      "loss": 0.2083,
      "step": 150
    },
    {
      "epoch": 0.5507745266781411,
      "grad_norm": 1.368167519569397,
      "learning_rate": 4.915462380182008e-05,
      "loss": 0.2119,
      "step": 160
    },
    {
      "epoch": 0.5851979345955249,
      "grad_norm": 0.3761076331138611,
      "learning_rate": 4.887661377799989e-05,
      "loss": 0.1688,
      "step": 170
    },
    {
      "epoch": 0.6196213425129088,
      "grad_norm": 0.5200079083442688,
      "learning_rate": 4.856017195700247e-05,
      "loss": 0.2084,
      "step": 180
    },
    {
      "epoch": 0.6540447504302926,
      "grad_norm": 0.7359570264816284,
      "learning_rate": 4.820580768357815e-05,
      "loss": 0.2355,
      "step": 190
    },
    {
      "epoch": 0.6884681583476764,
      "grad_norm": 0.6369001865386963,
      "learning_rate": 4.781409134245608e-05,
      "loss": 0.1958,
      "step": 200
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 0.6961612105369568,
      "learning_rate": 4.7385653440253026e-05,
      "loss": 0.222,
      "step": 210
    },
    {
      "epoch": 0.7573149741824441,
      "grad_norm": 0.2700839638710022,
      "learning_rate": 4.692118359060992e-05,
      "loss": 0.1754,
      "step": 220
    },
    {
      "epoch": 0.7917383820998278,
      "grad_norm": 0.5857542157173157,
      "learning_rate": 4.642142940418973e-05,
      "loss": 0.1893,
      "step": 230
    },
    {
      "epoch": 0.8261617900172117,
      "grad_norm": 1.13412344455719,
      "learning_rate": 4.588719528532342e-05,
      "loss": 0.1897,
      "step": 240
    },
    {
      "epoch": 0.8605851979345955,
      "grad_norm": 0.42694586515426636,
      "learning_rate": 4.5319341137240626e-05,
      "loss": 0.1583,
      "step": 250
    },
    {
      "epoch": 0.8950086058519794,
      "grad_norm": 0.8095172047615051,
      "learning_rate": 4.471878097796958e-05,
      "loss": 0.1879,
      "step": 260
    },
    {
      "epoch": 0.9294320137693631,
      "grad_norm": 0.5348218679428101,
      "learning_rate": 4.408648146913357e-05,
      "loss": 0.1788,
      "step": 270
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.42760810256004333,
      "learning_rate": 4.3423460360012495e-05,
      "loss": 0.1745,
      "step": 280
    },
    {
      "epoch": 0.9982788296041308,
      "grad_norm": 0.5095299482345581,
      "learning_rate": 4.2730784849373615e-05,
      "loss": 0.2198,
      "step": 290
    },
    {
      "epoch": 1.0309810671256454,
      "grad_norm": 0.24384614825248718,
      "learning_rate": 4.20095698677084e-05,
      "loss": 0.1812,
      "step": 300
    },
    {
      "epoch": 1.0654044750430294,
      "grad_norm": 0.40465590357780457,
      "learning_rate": 4.1260976282640466e-05,
      "loss": 0.1707,
      "step": 310
    },
    {
      "epoch": 1.0998278829604131,
      "grad_norm": 0.2638281285762787,
      "learning_rate": 4.048620903039308e-05,
      "loss": 0.1462,
      "step": 320
    },
    {
      "epoch": 1.1342512908777969,
      "grad_norm": 0.2884138524532318,
      "learning_rate": 3.968651517632376e-05,
      "loss": 0.1684,
      "step": 330
    },
    {
      "epoch": 1.1686746987951806,
      "grad_norm": 0.2671195864677429,
      "learning_rate": 3.886318190764797e-05,
      "loss": 0.1522,
      "step": 340
    },
    {
      "epoch": 1.2030981067125646,
      "grad_norm": 0.40242624282836914,
      "learning_rate": 3.801753446158256e-05,
      "loss": 0.1743,
      "step": 350
    },
    {
      "epoch": 1.2375215146299483,
      "grad_norm": 0.2599651515483856,
      "learning_rate": 3.715093399224398e-05,
      "loss": 0.1635,
      "step": 360
    },
    {
      "epoch": 1.2719449225473323,
      "grad_norm": 0.4389915466308594,
      "learning_rate": 3.626477537973462e-05,
      "loss": 0.2023,
      "step": 370
    },
    {
      "epoch": 1.306368330464716,
      "grad_norm": 0.6437344551086426,
      "learning_rate": 3.53604849849438e-05,
      "loss": 0.1639,
      "step": 380
    },
    {
      "epoch": 1.3407917383820998,
      "grad_norm": 0.2979753315448761,
      "learning_rate": 3.443951835367736e-05,
      "loss": 0.1639,
      "step": 390
    },
    {
      "epoch": 1.3752151462994837,
      "grad_norm": 1.0099725723266602,
      "learning_rate": 3.3503357873811105e-05,
      "loss": 0.1516,
      "step": 400
    },
    {
      "epoch": 1.4096385542168675,
      "grad_norm": 0.2964365780353546,
      "learning_rate": 3.2553510389239313e-05,
      "loss": 0.1872,
      "step": 410
    },
    {
      "epoch": 1.4440619621342514,
      "grad_norm": 0.32599982619285583,
      "learning_rate": 3.159150477445886e-05,
      "loss": 0.1705,
      "step": 420
    },
    {
      "epoch": 1.4784853700516352,
      "grad_norm": 0.46588459610939026,
      "learning_rate": 3.0618889473692756e-05,
      "loss": 0.1713,
      "step": 430
    },
    {
      "epoch": 1.512908777969019,
      "grad_norm": 0.4096398949623108,
      "learning_rate": 2.9637230008514494e-05,
      "loss": 0.1459,
      "step": 440
    },
    {
      "epoch": 1.5473321858864026,
      "grad_norm": 2.8078343868255615,
      "learning_rate": 2.864810645798442e-05,
      "loss": 0.166,
      "step": 450
    },
    {
      "epoch": 1.5817555938037866,
      "grad_norm": 0.25247690081596375,
      "learning_rate": 2.7653110915354542e-05,
      "loss": 0.1379,
      "step": 460
    },
    {
      "epoch": 1.6161790017211703,
      "grad_norm": 0.31816768646240234,
      "learning_rate": 2.6653844925435295e-05,
      "loss": 0.1596,
      "step": 470
    },
    {
      "epoch": 1.6506024096385543,
      "grad_norm": 0.29774752259254456,
      "learning_rate": 2.5651916906748995e-05,
      "loss": 0.1412,
      "step": 480
    },
    {
      "epoch": 1.685025817555938,
      "grad_norm": 0.4214092493057251,
      "learning_rate": 2.4648939562619423e-05,
      "loss": 0.1643,
      "step": 490
    },
    {
      "epoch": 1.7194492254733218,
      "grad_norm": 0.35638073086738586,
      "learning_rate": 2.3646527285364565e-05,
      "loss": 0.174,
      "step": 500
    },
    {
      "epoch": 1.7538726333907055,
      "grad_norm": 0.3502406179904938,
      "learning_rate": 2.2646293557770787e-05,
      "loss": 0.1862,
      "step": 510
    },
    {
      "epoch": 1.7882960413080895,
      "grad_norm": 0.2601950168609619,
      "learning_rate": 2.1649848356030945e-05,
      "loss": 0.1501,
      "step": 520
    },
    {
      "epoch": 1.8227194492254735,
      "grad_norm": 1.0533198118209839,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.1461,
      "step": 530
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.36168837547302246,
      "learning_rate": 1.967473036322644e-05,
      "loss": 0.1522,
      "step": 540
    },
    {
      "epoch": 1.891566265060241,
      "grad_norm": 0.4589601457118988,
      "learning_rate": 1.869923672205322e-05,
      "loss": 0.1682,
      "step": 550
    },
    {
      "epoch": 1.9259896729776247,
      "grad_norm": 0.2802368700504303,
      "learning_rate": 1.7733884789357226e-05,
      "loss": 0.1657,
      "step": 560
    },
    {
      "epoch": 1.9604130808950087,
      "grad_norm": 0.4951450526714325,
      "learning_rate": 1.6780228395594777e-05,
      "loss": 0.1534,
      "step": 570
    },
    {
      "epoch": 1.9948364888123924,
      "grad_norm": 0.3530052602291107,
      "learning_rate": 1.58398025460829e-05,
      "loss": 0.1904,
      "step": 580
    },
    {
      "epoch": 2.027538726333907,
      "grad_norm": 0.32037100195884705,
      "learning_rate": 1.491412095025479e-05,
      "loss": 0.1301,
      "step": 590
    },
    {
      "epoch": 2.061962134251291,
      "grad_norm": 0.4082101285457611,
      "learning_rate": 1.4004673585193145e-05,
      "loss": 0.141,
      "step": 600
    },
    {
      "epoch": 2.0963855421686746,
      "grad_norm": 0.5846110582351685,
      "learning_rate": 1.3112924297363066e-05,
      "loss": 0.1167,
      "step": 610
    },
    {
      "epoch": 2.1308089500860588,
      "grad_norm": 0.3282460570335388,
      "learning_rate": 1.2240308446404797e-05,
      "loss": 0.136,
      "step": 620
    },
    {
      "epoch": 2.1652323580034425,
      "grad_norm": 0.3618537187576294,
      "learning_rate": 1.1388230594778888e-05,
      "loss": 0.1321,
      "step": 630
    },
    {
      "epoch": 2.1996557659208262,
      "grad_norm": 0.4230036437511444,
      "learning_rate": 1.0558062246982484e-05,
      "loss": 0.1474,
      "step": 640
    },
    {
      "epoch": 2.23407917383821,
      "grad_norm": 0.6187330484390259,
      "learning_rate": 9.751139641975682e-06,
      "loss": 0.1216,
      "step": 650
    },
    {
      "epoch": 2.2685025817555937,
      "grad_norm": 0.6022536754608154,
      "learning_rate": 8.96876160237133e-06,
      "loss": 0.1328,
      "step": 660
    },
    {
      "epoch": 2.3029259896729775,
      "grad_norm": 0.47852516174316406,
      "learning_rate": 8.212187443850203e-06,
      "loss": 0.1244,
      "step": 670
    },
    {
      "epoch": 2.337349397590361,
      "grad_norm": 0.8470497727394104,
      "learning_rate": 7.482634948166442e-06,
      "loss": 0.1135,
      "step": 680
    },
    {
      "epoch": 2.3717728055077454,
      "grad_norm": 0.5010228157043457,
      "learning_rate": 6.7812784030061e-06,
      "loss": 0.115,
      "step": 690
    },
    {
      "epoch": 2.406196213425129,
      "grad_norm": 0.3755123019218445,
      "learning_rate": 6.109246711853711e-06,
      "loss": 0.1305,
      "step": 700
    },
    {
      "epoch": 2.440619621342513,
      "grad_norm": 0.464460164308548,
      "learning_rate": 5.467621576909218e-06,
      "loss": 0.11,
      "step": 710
    },
    {
      "epoch": 2.4750430292598966,
      "grad_norm": 0.8106716871261597,
      "learning_rate": 4.8574357579801495e-06,
      "loss": 0.1152,
      "step": 720
    },
    {
      "epoch": 2.509466437177281,
      "grad_norm": 0.6007969975471497,
      "learning_rate": 4.2796714101514725e-06,
      "loss": 0.1486,
      "step": 730
    },
    {
      "epoch": 2.5438898450946645,
      "grad_norm": 0.4238184690475464,
      "learning_rate": 3.735258502908737e-06,
      "loss": 0.1256,
      "step": 740
    },
    {
      "epoch": 2.5783132530120483,
      "grad_norm": 0.6052628755569458,
      "learning_rate": 3.2250733232592337e-06,
      "loss": 0.1188,
      "step": 750
    },
    {
      "epoch": 2.612736660929432,
      "grad_norm": 0.41783905029296875,
      "learning_rate": 2.749937065260444e-06,
      "loss": 0.1155,
      "step": 760
    },
    {
      "epoch": 2.6471600688468158,
      "grad_norm": 0.47119778394699097,
      "learning_rate": 2.310614508226078e-06,
      "loss": 0.122,
      "step": 770
    },
    {
      "epoch": 2.6815834767641995,
      "grad_norm": 0.37864789366722107,
      "learning_rate": 1.907812785737312e-06,
      "loss": 0.1164,
      "step": 780
    },
    {
      "epoch": 2.7160068846815832,
      "grad_norm": 0.5162110924720764,
      "learning_rate": 1.5421802474406027e-06,
      "loss": 0.13,
      "step": 790
    },
    {
      "epoch": 2.7504302925989674,
      "grad_norm": 0.7263408303260803,
      "learning_rate": 1.214305415464076e-06,
      "loss": 0.1139,
      "step": 800
    },
    {
      "epoch": 2.784853700516351,
      "grad_norm": 0.46103435754776,
      "learning_rate": 9.247160371323449e-07,
      "loss": 0.1424,
      "step": 810
    },
    {
      "epoch": 2.819277108433735,
      "grad_norm": 0.4251267910003662,
      "learning_rate": 6.738782355044049e-07,
      "loss": 0.1089,
      "step": 820
    },
    {
      "epoch": 2.8537005163511187,
      "grad_norm": 0.3590181767940521,
      "learning_rate": 4.6219575910197156e-07,
      "loss": 0.1212,
      "step": 830
    },
    {
      "epoch": 2.888123924268503,
      "grad_norm": 0.30818095803260803,
      "learning_rate": 2.9000933203588153e-07,
      "loss": 0.1245,
      "step": 840
    },
    {
      "epoch": 2.9225473321858866,
      "grad_norm": 0.43673816323280334,
      "learning_rate": 1.575961055765379e-07,
      "loss": 0.1263,
      "step": 850
    },
    {
      "epoch": 2.9569707401032703,
      "grad_norm": 0.5171876549720764,
      "learning_rate": 6.516921205125537e-08,
      "loss": 0.1232,
      "step": 860
    },
    {
      "epoch": 2.991394148020654,
      "grad_norm": 0.31435626745224,
      "learning_rate": 1.2877421786458566e-08,
      "loss": 0.114,
      "step": 870
    }
  ],
  "logging_steps": 10,
  "max_steps": 870,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 2.3941567684085023e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}