{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 369,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0408997955010225,
      "grad_norm": 9.2357177734375,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7318,
      "step": 5
    },
    {
      "epoch": 0.081799591002045,
      "grad_norm": 9.753511428833008,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.6559,
      "step": 10
    },
    {
      "epoch": 0.12269938650306748,
      "grad_norm": 8.666117668151855,
      "learning_rate": 9.999225623611208e-06,
      "loss": 0.702,
      "step": 15
    },
    {
      "epoch": 0.16359918200409,
      "grad_norm": 7.023216247558594,
      "learning_rate": 9.990516643685222e-06,
      "loss": 0.6079,
      "step": 20
    },
    {
      "epoch": 0.20449897750511248,
      "grad_norm": 8.54660415649414,
      "learning_rate": 9.972147627352593e-06,
      "loss": 0.7299,
      "step": 25
    },
    {
      "epoch": 0.24539877300613497,
      "grad_norm": 6.48859167098999,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.7183,
      "step": 30
    },
    {
      "epoch": 0.28629856850715746,
      "grad_norm": 7.0337629318237305,
      "learning_rate": 9.906590341416725e-06,
      "loss": 0.721,
      "step": 35
    },
    {
      "epoch": 0.32719836400818,
      "grad_norm": 6.987436294555664,
      "learning_rate": 9.859528969650739e-06,
      "loss": 0.6602,
      "step": 40
    },
    {
      "epoch": 0.36809815950920244,
      "grad_norm": 7.1492791175842285,
      "learning_rate": 9.803061111519188e-06,
      "loss": 0.6731,
      "step": 45
    },
    {
      "epoch": 0.40899795501022496,
      "grad_norm": 6.402842044830322,
      "learning_rate": 9.737296070648187e-06,
      "loss": 0.6957,
      "step": 50
    },
    {
      "epoch": 0.4498977505112474,
      "grad_norm": 6.457306385040283,
      "learning_rate": 9.66236114702178e-06,
      "loss": 0.6972,
      "step": 55
    },
    {
      "epoch": 0.49079754601226994,
      "grad_norm": 6.941644191741943,
      "learning_rate": 9.57840139057007e-06,
      "loss": 0.6772,
      "step": 60
    },
    {
      "epoch": 0.5316973415132924,
      "grad_norm": 7.459963321685791,
      "learning_rate": 9.485579320399184e-06,
      "loss": 0.639,
      "step": 65
    },
    {
      "epoch": 0.5725971370143149,
      "grad_norm": 7.600812911987305,
      "learning_rate": 9.384074610206495e-06,
      "loss": 0.6781,
      "step": 70
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 6.6164093017578125,
      "learning_rate": 9.274083740490097e-06,
      "loss": 0.7323,
      "step": 75
    },
    {
      "epoch": 0.65439672801636,
      "grad_norm": 6.039959907531738,
      "learning_rate": 9.155819618225707e-06,
      "loss": 0.69,
      "step": 80
    },
    {
      "epoch": 0.6952965235173824,
      "grad_norm": 6.699795246124268,
      "learning_rate": 9.029511164747175e-06,
      "loss": 0.6654,
      "step": 85
    },
    {
      "epoch": 0.7361963190184049,
      "grad_norm": 7.462390422821045,
      "learning_rate": 8.895402872628352e-06,
      "loss": 0.6825,
      "step": 90
    },
    {
      "epoch": 0.7770961145194274,
      "grad_norm": 7.326976299285889,
      "learning_rate": 8.753754332424047e-06,
      "loss": 0.6797,
      "step": 95
    },
    {
      "epoch": 0.8179959100204499,
      "grad_norm": 7.3194050788879395,
      "learning_rate": 8.604839730186125e-06,
      "loss": 0.7153,
      "step": 100
    },
    {
      "epoch": 0.8588957055214724,
      "grad_norm": 7.875491619110107,
      "learning_rate": 8.448947316727444e-06,
      "loss": 0.6896,
      "step": 105
    },
    {
      "epoch": 0.8997955010224948,
      "grad_norm": 7.002596378326416,
      "learning_rate": 8.286378849660895e-06,
      "loss": 0.6507,
      "step": 110
    },
    {
      "epoch": 0.9406952965235174,
      "grad_norm": 6.086474895477295,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.6434,
      "step": 115
    },
    {
      "epoch": 0.9815950920245399,
      "grad_norm": 6.344010353088379,
      "learning_rate": 7.942484789507282e-06,
      "loss": 0.6802,
      "step": 120
    },
    {
      "epoch": 1.016359918200409,
      "grad_norm": 4.62432861328125,
      "learning_rate": 7.76182486480253e-06,
      "loss": 0.4361,
      "step": 125
    },
    {
      "epoch": 1.0572597137014315,
      "grad_norm": 4.461228847503662,
      "learning_rate": 7.57581893473448e-06,
      "loss": 0.2795,
      "step": 130
    },
    {
      "epoch": 1.098159509202454,
      "grad_norm": 4.962184429168701,
      "learning_rate": 7.38482704700653e-06,
      "loss": 0.2937,
      "step": 135
    },
    {
      "epoch": 1.1390593047034765,
      "grad_norm": 6.302546501159668,
      "learning_rate": 7.1892189005337494e-06,
      "loss": 0.2469,
      "step": 140
    },
    {
      "epoch": 1.179959100204499,
      "grad_norm": 7.166332244873047,
      "learning_rate": 6.989373129824605e-06,
      "loss": 0.2656,
      "step": 145
    },
    {
      "epoch": 1.2208588957055215,
      "grad_norm": 5.206060409545898,
      "learning_rate": 6.785676572066225e-06,
      "loss": 0.2509,
      "step": 150
    },
    {
      "epoch": 1.261758691206544,
      "grad_norm": 5.387083053588867,
      "learning_rate": 6.5785235183319266e-06,
      "loss": 0.2365,
      "step": 155
    },
    {
      "epoch": 1.3026584867075663,
      "grad_norm": 5.5746917724609375,
      "learning_rate": 6.368314950360416e-06,
      "loss": 0.2332,
      "step": 160
    },
    {
      "epoch": 1.343558282208589,
      "grad_norm": 5.9678521156311035,
      "learning_rate": 6.155457764384001e-06,
      "loss": 0.2556,
      "step": 165
    },
    {
      "epoch": 1.3844580777096114,
      "grad_norm": 5.8653340339660645,
      "learning_rate": 5.940363983508257e-06,
      "loss": 0.2712,
      "step": 170
    },
    {
      "epoch": 1.425357873210634,
      "grad_norm": 5.267022132873535,
      "learning_rate": 5.723449960167703e-06,
      "loss": 0.2639,
      "step": 175
    },
    {
      "epoch": 1.4662576687116564,
      "grad_norm": 5.745113372802734,
      "learning_rate": 5.50513557020129e-06,
      "loss": 0.253,
      "step": 180
    },
    {
      "epoch": 1.5071574642126788,
      "grad_norm": 5.734601020812988,
      "learning_rate": 5.285843400107722e-06,
      "loss": 0.2505,
      "step": 185
    },
    {
      "epoch": 1.5480572597137015,
      "grad_norm": 5.375375270843506,
      "learning_rate": 5.065997929053795e-06,
      "loss": 0.2425,
      "step": 190
    },
    {
      "epoch": 1.588957055214724,
      "grad_norm": 6.181747913360596,
      "learning_rate": 4.846024707219149e-06,
      "loss": 0.2461,
      "step": 195
    },
    {
      "epoch": 1.6298568507157465,
      "grad_norm": 6.347278118133545,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.3003,
      "step": 200
    },
    {
      "epoch": 1.670756646216769,
      "grad_norm": 5.828954696655273,
      "learning_rate": 4.407397624141487e-06,
      "loss": 0.2714,
      "step": 205
    },
    {
      "epoch": 1.7116564417177913,
      "grad_norm": 6.085731029510498,
      "learning_rate": 4.1895928039685635e-06,
      "loss": 0.2678,
      "step": 210
    },
    {
      "epoch": 1.752556237218814,
      "grad_norm": 4.415971279144287,
      "learning_rate": 3.973356671684455e-06,
      "loss": 0.2438,
      "step": 215
    },
    {
      "epoch": 1.7934560327198366,
      "grad_norm": 6.427822113037109,
      "learning_rate": 3.7591077909488817e-06,
      "loss": 0.2649,
      "step": 220
    },
    {
      "epoch": 1.834355828220859,
      "grad_norm": 5.343576908111572,
      "learning_rate": 3.54726087874122e-06,
      "loss": 0.2457,
      "step": 225
    },
    {
      "epoch": 1.8752556237218814,
      "grad_norm": 5.313407897949219,
      "learning_rate": 3.3382260026017027e-06,
      "loss": 0.2533,
      "step": 230
    },
    {
      "epoch": 1.9161554192229038,
      "grad_norm": 5.814549922943115,
      "learning_rate": 3.132407786872442e-06,
      "loss": 0.2747,
      "step": 235
    },
    {
      "epoch": 1.9570552147239264,
      "grad_norm": 5.629500389099121,
      "learning_rate": 2.9302046294747498e-06,
      "loss": 0.2409,
      "step": 240
    },
    {
      "epoch": 1.997955010224949,
      "grad_norm": 6.071770668029785,
      "learning_rate": 2.7320079307387788e-06,
      "loss": 0.2861,
      "step": 245
    },
    {
      "epoch": 2.032719836400818,
      "grad_norm": 6.585735321044922,
      "learning_rate": 2.538201335778289e-06,
      "loss": 0.0862,
      "step": 250
    },
    {
      "epoch": 2.0736196319018405,
      "grad_norm": 2.7118561267852783,
      "learning_rate": 2.3491599918769854e-06,
      "loss": 0.0903,
      "step": 255
    },
    {
      "epoch": 2.114519427402863,
      "grad_norm": 3.2460713386535645,
      "learning_rate": 2.165249822323943e-06,
      "loss": 0.094,
      "step": 260
    },
    {
      "epoch": 2.1554192229038853,
      "grad_norm": 3.287091016769409,
      "learning_rate": 1.9868268181037186e-06,
      "loss": 0.0666,
      "step": 265
    },
    {
      "epoch": 2.196319018404908,
      "grad_norm": 3.842747926712036,
      "learning_rate": 1.814236348812211e-06,
      "loss": 0.0893,
      "step": 270
    },
    {
      "epoch": 2.2372188139059306,
      "grad_norm": 3.1263859272003174,
      "learning_rate": 1.6478124941321123e-06,
      "loss": 0.0834,
      "step": 275
    },
    {
      "epoch": 2.278118609406953,
      "grad_norm": 2.86702823638916,
      "learning_rate": 1.4878773971620076e-06,
      "loss": 0.0741,
      "step": 280
    },
    {
      "epoch": 2.3190184049079754,
      "grad_norm": 2.579695701599121,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.0738,
      "step": 285
    },
    {
      "epoch": 2.359918200408998,
      "grad_norm": 2.9244284629821777,
      "learning_rate": 1.1886986487449476e-06,
      "loss": 0.0794,
      "step": 290
    },
    {
      "epoch": 2.40081799591002,
      "grad_norm": 3.580735683441162,
      "learning_rate": 1.0500341112070605e-06,
      "loss": 0.0872,
      "step": 295
    },
    {
      "epoch": 2.441717791411043,
      "grad_norm": 3.844236373901367,
      "learning_rate": 9.190154382188921e-07,
      "loss": 0.067,
      "step": 300
    },
    {
      "epoch": 2.4826175869120655,
      "grad_norm": 3.944274425506592,
      "learning_rate": 7.958962398255293e-07,
      "loss": 0.0774,
      "step": 305
    },
    {
      "epoch": 2.523517382413088,
      "grad_norm": 6.193697929382324,
      "learning_rate": 6.809148352279182e-07,
      "loss": 0.0801,
      "step": 310
    },
    {
      "epoch": 2.5644171779141103,
      "grad_norm": 3.7585809230804443,
      "learning_rate": 5.742937914734853e-07,
      "loss": 0.0819,
      "step": 315
    },
    {
      "epoch": 2.6053169734151327,
      "grad_norm": 5.038247108459473,
      "learning_rate": 4.762394926378477e-07,
      "loss": 0.082,
      "step": 320
    },
    {
      "epoch": 2.6462167689161555,
      "grad_norm": 4.3431806564331055,
      "learning_rate": 3.869417403315856e-07,
      "loss": 0.0827,
      "step": 325
    },
    {
      "epoch": 2.687116564417178,
      "grad_norm": 4.178752899169922,
      "learning_rate": 3.065733863053072e-07,
      "loss": 0.0794,
      "step": 330
    },
    {
      "epoch": 2.7280163599182004,
      "grad_norm": 2.7639453411102295,
      "learning_rate": 2.3528999786421758e-07,
      "loss": 0.0699,
      "step": 335
    },
    {
      "epoch": 2.7689161554192228,
      "grad_norm": 3.0409483909606934,
      "learning_rate": 1.7322955673980678e-07,
      "loss": 0.0837,
      "step": 340
    },
    {
      "epoch": 2.809815950920245,
      "grad_norm": 3.909299850463867,
      "learning_rate": 1.2051219200156394e-07,
      "loss": 0.0984,
      "step": 345
    },
    {
      "epoch": 2.850715746421268,
      "grad_norm": 4.271017551422119,
      "learning_rate": 7.723994752570463e-08,
      "loss": 0.0728,
      "step": 350
    },
    {
      "epoch": 2.8916155419222904,
      "grad_norm": 3.3646178245544434,
      "learning_rate": 4.3496584471016125e-08,
      "loss": 0.0797,
      "step": 355
    },
    {
      "epoch": 2.932515337423313,
      "grad_norm": 3.125234365463257,
      "learning_rate": 1.9347419144180035e-08,
      "loss": 0.0782,
      "step": 360
    },
    {
      "epoch": 2.9734151329243352,
      "grad_norm": 5.080536365509033,
      "learning_rate": 4.839196568388049e-09,
      "loss": 0.0695,
      "step": 365
    }
  ],
  "logging_steps": 5,
  "max_steps": 369,
  "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": 4.9828469569644134e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}