{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9988545246277205,
  "eval_steps": 500,
  "global_step": 654,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0015273004963726614,
      "grad_norm": 5.274756775371289,
      "learning_rate": 1.5151515151515152e-07,
      "loss": 1.3724,
      "step": 1
    },
    {
      "epoch": 0.015273004963726614,
      "grad_norm": 2.4356823026963044,
      "learning_rate": 1.5151515151515152e-06,
      "loss": 1.3215,
      "step": 10
    },
    {
      "epoch": 0.030546009927453228,
      "grad_norm": 0.8415959934340891,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 0.7041,
      "step": 20
    },
    {
      "epoch": 0.045819014891179836,
      "grad_norm": 0.3951472895674848,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.3621,
      "step": 30
    },
    {
      "epoch": 0.061092019854906456,
      "grad_norm": 0.3302170549897179,
      "learning_rate": 6.060606060606061e-06,
      "loss": 0.2555,
      "step": 40
    },
    {
      "epoch": 0.07636502481863307,
      "grad_norm": 0.3855993238810659,
      "learning_rate": 7.5757575757575764e-06,
      "loss": 0.21,
      "step": 50
    },
    {
      "epoch": 0.09163802978235967,
      "grad_norm": 0.3643941392356843,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.1707,
      "step": 60
    },
    {
      "epoch": 0.10691103474608629,
      "grad_norm": 0.24356795945348528,
      "learning_rate": 9.99885820390154e-06,
      "loss": 0.1628,
      "step": 70
    },
    {
      "epoch": 0.12218403970981291,
      "grad_norm": 0.18282398542383,
      "learning_rate": 9.986018985905901e-06,
      "loss": 0.1537,
      "step": 80
    },
    {
      "epoch": 0.13745704467353953,
      "grad_norm": 0.17824704057472698,
      "learning_rate": 9.95895006911623e-06,
      "loss": 0.1472,
      "step": 90
    },
    {
      "epoch": 0.15273004963726614,
      "grad_norm": 0.20039388213383333,
      "learning_rate": 9.917728706052765e-06,
      "loss": 0.1512,
      "step": 100
    },
    {
      "epoch": 0.16800305460099274,
      "grad_norm": 0.21394666608149668,
      "learning_rate": 9.862472539183757e-06,
      "loss": 0.1446,
      "step": 110
    },
    {
      "epoch": 0.18327605956471935,
      "grad_norm": 0.1399694568156125,
      "learning_rate": 9.793339265183303e-06,
      "loss": 0.1379,
      "step": 120
    },
    {
      "epoch": 0.19854906452844598,
      "grad_norm": 0.17756723955927178,
      "learning_rate": 9.710526184877667e-06,
      "loss": 0.1446,
      "step": 130
    },
    {
      "epoch": 0.21382206949217258,
      "grad_norm": 0.1256183224074801,
      "learning_rate": 9.61426964016452e-06,
      "loss": 0.1598,
      "step": 140
    },
    {
      "epoch": 0.2290950744558992,
      "grad_norm": 0.1925308832459223,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.1703,
      "step": 150
    },
    {
      "epoch": 0.24436807941962582,
      "grad_norm": 0.18425814277453983,
      "learning_rate": 9.382562573963238e-06,
      "loss": 0.1347,
      "step": 160
    },
    {
      "epoch": 0.2596410843833524,
      "grad_norm": 0.16364274896349304,
      "learning_rate": 9.24777332588177e-06,
      "loss": 0.1291,
      "step": 170
    },
    {
      "epoch": 0.27491408934707906,
      "grad_norm": 0.1575283767974318,
      "learning_rate": 9.10086127298478e-06,
      "loss": 0.1455,
      "step": 180
    },
    {
      "epoch": 0.29018709431080564,
      "grad_norm": 0.16299547711079815,
      "learning_rate": 8.94224569050324e-06,
      "loss": 0.1448,
      "step": 190
    },
    {
      "epoch": 0.30546009927453227,
      "grad_norm": 0.1678304349758231,
      "learning_rate": 8.772379254604074e-06,
      "loss": 0.1329,
      "step": 200
    },
    {
      "epoch": 0.3207331042382589,
      "grad_norm": 0.18133792526251943,
      "learning_rate": 8.591746750488639e-06,
      "loss": 0.1523,
      "step": 210
    },
    {
      "epoch": 0.3360061092019855,
      "grad_norm": 0.138988177087972,
      "learning_rate": 8.400863688854598e-06,
      "loss": 0.1487,
      "step": 220
    },
    {
      "epoch": 0.3512791141657121,
      "grad_norm": 0.10975239156978413,
      "learning_rate": 8.200274834669675e-06,
      "loss": 0.1454,
      "step": 230
    },
    {
      "epoch": 0.3665521191294387,
      "grad_norm": 0.13022143377921225,
      "learning_rate": 7.99055265245608e-06,
      "loss": 0.1414,
      "step": 240
    },
    {
      "epoch": 0.3818251240931653,
      "grad_norm": 0.09325598021130543,
      "learning_rate": 7.772295672522615e-06,
      "loss": 0.1406,
      "step": 250
    },
    {
      "epoch": 0.39709812905689196,
      "grad_norm": 0.10192417440664955,
      "learning_rate": 7.546126782807117e-06,
      "loss": 0.1243,
      "step": 260
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 0.10214456159318729,
      "learning_rate": 7.312691451204178e-06,
      "loss": 0.1381,
      "step": 270
    },
    {
      "epoch": 0.42764413898434517,
      "grad_norm": 0.11650402115334278,
      "learning_rate": 7.072655883451478e-06,
      "loss": 0.1155,
      "step": 280
    },
    {
      "epoch": 0.4429171439480718,
      "grad_norm": 0.09899147898887324,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.1372,
      "step": 290
    },
    {
      "epoch": 0.4581901489117984,
      "grad_norm": 0.11909991495605356,
      "learning_rate": 6.575541090118105e-06,
      "loss": 0.1449,
      "step": 300
    },
    {
      "epoch": 0.473463153875525,
      "grad_norm": 0.11324258672549159,
      "learning_rate": 6.319880590337549e-06,
      "loss": 0.1355,
      "step": 310
    },
    {
      "epoch": 0.48873615883925164,
      "grad_norm": 0.15684721556128514,
      "learning_rate": 6.060453257077686e-06,
      "loss": 0.1448,
      "step": 320
    },
    {
      "epoch": 0.5040091638029782,
      "grad_norm": 0.13428699506671887,
      "learning_rate": 5.797999475166897e-06,
      "loss": 0.161,
      "step": 330
    },
    {
      "epoch": 0.5192821687667049,
      "grad_norm": 0.16260953903247685,
      "learning_rate": 5.533268266675601e-06,
      "loss": 0.1301,
      "step": 340
    },
    {
      "epoch": 0.5345551737304315,
      "grad_norm": 0.138117775450493,
      "learning_rate": 5.267015153267246e-06,
      "loss": 0.1153,
      "step": 350
    },
    {
      "epoch": 0.5498281786941581,
      "grad_norm": 0.09342193422842268,
      "learning_rate": 5e-06,
      "loss": 0.1388,
      "step": 360
    },
    {
      "epoch": 0.5651011836578846,
      "grad_norm": 0.12521354455054481,
      "learning_rate": 4.732984846732755e-06,
      "loss": 0.1469,
      "step": 370
    },
    {
      "epoch": 0.5803741886216113,
      "grad_norm": 0.12074643796303608,
      "learning_rate": 4.466731733324399e-06,
      "loss": 0.1336,
      "step": 380
    },
    {
      "epoch": 0.5956471935853379,
      "grad_norm": 0.10811612216811932,
      "learning_rate": 4.2020005248331056e-06,
      "loss": 0.1444,
      "step": 390
    },
    {
      "epoch": 0.6109201985490645,
      "grad_norm": 0.07932149519427452,
      "learning_rate": 3.939546742922318e-06,
      "loss": 0.1217,
      "step": 400
    },
    {
      "epoch": 0.6261932035127912,
      "grad_norm": 0.13495076484320434,
      "learning_rate": 3.6801194096624515e-06,
      "loss": 0.1486,
      "step": 410
    },
    {
      "epoch": 0.6414662084765178,
      "grad_norm": 0.09577670394462665,
      "learning_rate": 3.424458909881897e-06,
      "loss": 0.1351,
      "step": 420
    },
    {
      "epoch": 0.6567392134402443,
      "grad_norm": 0.13052717933517727,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.1115,
      "step": 430
    },
    {
      "epoch": 0.672012218403971,
      "grad_norm": 0.10868560247173807,
      "learning_rate": 2.9273441165485227e-06,
      "loss": 0.1429,
      "step": 440
    },
    {
      "epoch": 0.6872852233676976,
      "grad_norm": 0.08492718279444092,
      "learning_rate": 2.687308548795825e-06,
      "loss": 0.1006,
      "step": 450
    },
    {
      "epoch": 0.7025582283314242,
      "grad_norm": 0.11879878396233616,
      "learning_rate": 2.4538732171928847e-06,
      "loss": 0.1441,
      "step": 460
    },
    {
      "epoch": 0.7178312332951509,
      "grad_norm": 0.08627366089386783,
      "learning_rate": 2.2277043274773856e-06,
      "loss": 0.121,
      "step": 470
    },
    {
      "epoch": 0.7331042382588774,
      "grad_norm": 0.11860493468588708,
      "learning_rate": 2.00944734754392e-06,
      "loss": 0.104,
      "step": 480
    },
    {
      "epoch": 0.748377243222604,
      "grad_norm": 0.13566719962532361,
      "learning_rate": 1.7997251653303249e-06,
      "loss": 0.1223,
      "step": 490
    },
    {
      "epoch": 0.7636502481863306,
      "grad_norm": 0.11713757986892694,
      "learning_rate": 1.5991363111454023e-06,
      "loss": 0.1297,
      "step": 500
    },
    {
      "epoch": 0.7789232531500573,
      "grad_norm": 0.0932474027124396,
      "learning_rate": 1.4082532495113627e-06,
      "loss": 0.129,
      "step": 510
    },
    {
      "epoch": 0.7941962581137839,
      "grad_norm": 0.06960734581371894,
      "learning_rate": 1.2276207453959283e-06,
      "loss": 0.102,
      "step": 520
    },
    {
      "epoch": 0.8094692630775105,
      "grad_norm": 0.1054188314922148,
      "learning_rate": 1.0577543094967613e-06,
      "loss": 0.1182,
      "step": 530
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.15433498716395788,
      "learning_rate": 8.991387270152202e-07,
      "loss": 0.1306,
      "step": 540
    },
    {
      "epoch": 0.8400152730049637,
      "grad_norm": 0.10854241303868371,
      "learning_rate": 7.522266741182305e-07,
      "loss": 0.11,
      "step": 550
    },
    {
      "epoch": 0.8552882779686903,
      "grad_norm": 0.13631634648737045,
      "learning_rate": 6.174374260367611e-07,
      "loss": 0.1162,
      "step": 560
    },
    {
      "epoch": 0.870561282932417,
      "grad_norm": 0.11447961702982919,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.1396,
      "step": 570
    },
    {
      "epoch": 0.8858342878961436,
      "grad_norm": 0.1527666931393435,
      "learning_rate": 3.8573035983548167e-07,
      "loss": 0.1353,
      "step": 580
    },
    {
      "epoch": 0.9011072928598702,
      "grad_norm": 0.13839722389025982,
      "learning_rate": 2.894738151223331e-07,
      "loss": 0.1193,
      "step": 590
    },
    {
      "epoch": 0.9163802978235968,
      "grad_norm": 0.13196785663691574,
      "learning_rate": 2.0666073481669714e-07,
      "loss": 0.1147,
      "step": 600
    },
    {
      "epoch": 0.9316533027873234,
      "grad_norm": 0.09878735062205103,
      "learning_rate": 1.375274608162447e-07,
      "loss": 0.1261,
      "step": 610
    },
    {
      "epoch": 0.94692630775105,
      "grad_norm": 0.07952980044315368,
      "learning_rate": 8.227129394723643e-08,
      "loss": 0.0992,
      "step": 620
    },
    {
      "epoch": 0.9621993127147767,
      "grad_norm": 0.09015417303190752,
      "learning_rate": 4.104993088376974e-08,
      "loss": 0.1212,
      "step": 630
    },
    {
      "epoch": 0.9774723176785033,
      "grad_norm": 0.12899056246523352,
      "learning_rate": 1.3981014094099354e-08,
      "loss": 0.131,
      "step": 640
    },
    {
      "epoch": 0.9927453226422298,
      "grad_norm": 0.08443201140015646,
      "learning_rate": 1.1417960984605459e-09,
      "loss": 0.1302,
      "step": 650
    }
  ],
  "logging_steps": 10,
  "max_steps": 654,
  "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": 32812144386048.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}