{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 44368,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 1.9776866209880996e-05,
      "loss": 1.6544,
      "step": 500
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.9551478543094125e-05,
      "loss": 1.4272,
      "step": 1000
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.932609087630725e-05,
      "loss": 1.3783,
      "step": 1500
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9100703209520376e-05,
      "loss": 1.324,
      "step": 2000
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.8875315542733504e-05,
      "loss": 1.2637,
      "step": 2500
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.864992787594663e-05,
      "loss": 1.2255,
      "step": 3000
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.8424540209159756e-05,
      "loss": 1.211,
      "step": 3500
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.8199603317706455e-05,
      "loss": 1.1977,
      "step": 4000
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.7974215650919584e-05,
      "loss": 1.1879,
      "step": 4500
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.774882798413271e-05,
      "loss": 1.1487,
      "step": 5000
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.7523440317345835e-05,
      "loss": 1.1427,
      "step": 5500
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.7298052650558964e-05,
      "loss": 1.1066,
      "step": 6000
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.707266498377209e-05,
      "loss": 1.1576,
      "step": 6500
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.6847277316985215e-05,
      "loss": 1.0895,
      "step": 7000
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.662188965019834e-05,
      "loss": 1.0893,
      "step": 7500
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.639650198341147e-05,
      "loss": 1.0693,
      "step": 8000
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.6171114316624595e-05,
      "loss": 1.0979,
      "step": 8500
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.5945726649837724e-05,
      "loss": 1.0202,
      "step": 9000
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.572033898305085e-05,
      "loss": 1.0574,
      "step": 9500
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.549540209159755e-05,
      "loss": 0.9914,
      "step": 10000
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.5270014424810675e-05,
      "loss": 1.0377,
      "step": 10500
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5045077533357375e-05,
      "loss": 0.9989,
      "step": 11000
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.4820140641904077e-05,
      "loss": 0.8036,
      "step": 11500
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.4594752975117204e-05,
      "loss": 0.7445,
      "step": 12000
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.436936530833033e-05,
      "loss": 0.7566,
      "step": 12500
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.4143977641543456e-05,
      "loss": 0.7529,
      "step": 13000
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.3918589974756582e-05,
      "loss": 0.7446,
      "step": 13500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.369320230796971e-05,
      "loss": 0.7539,
      "step": 14000
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.3468265416516409e-05,
      "loss": 0.7389,
      "step": 14500
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.3242877749729536e-05,
      "loss": 0.7502,
      "step": 15000
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.3017940858276236e-05,
      "loss": 0.7319,
      "step": 15500
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.2792553191489363e-05,
      "loss": 0.7732,
      "step": 16000
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.2567165524702489e-05,
      "loss": 0.7458,
      "step": 16500
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.2341777857915616e-05,
      "loss": 0.7277,
      "step": 17000
    },
    {
      "epoch": 1.58,
      "learning_rate": 1.2116390191128742e-05,
      "loss": 0.7559,
      "step": 17500
    },
    {
      "epoch": 1.62,
      "learning_rate": 1.1891002524341869e-05,
      "loss": 0.7572,
      "step": 18000
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.1665614857554996e-05,
      "loss": 0.7387,
      "step": 18500
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1440227190768123e-05,
      "loss": 0.7502,
      "step": 19000
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.1214839523981249e-05,
      "loss": 0.7007,
      "step": 19500
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.0989451857194376e-05,
      "loss": 0.7435,
      "step": 20000
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.0764514965741076e-05,
      "loss": 0.7114,
      "step": 20500
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.0539127298954201e-05,
      "loss": 0.7574,
      "step": 21000
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.0313739632167327e-05,
      "loss": 0.7184,
      "step": 21500
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.0088351965380456e-05,
      "loss": 0.6812,
      "step": 22000
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.862964298593581e-06,
      "loss": 0.5916,
      "step": 22500
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.637576631806709e-06,
      "loss": 0.5152,
      "step": 23000
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.412188965019836e-06,
      "loss": 0.4892,
      "step": 23500
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.187252073566536e-06,
      "loss": 0.5003,
      "step": 24000
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.961864406779663e-06,
      "loss": 0.4962,
      "step": 24500
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.736476739992788e-06,
      "loss": 0.4955,
      "step": 25000
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.511089073205916e-06,
      "loss": 0.5183,
      "step": 25500
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.285701406419041e-06,
      "loss": 0.501,
      "step": 26000
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.060313739632168e-06,
      "loss": 0.4932,
      "step": 26500
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.835376848178868e-06,
      "loss": 0.5092,
      "step": 27000
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.609989181391995e-06,
      "loss": 0.5327,
      "step": 27500
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.384601514605121e-06,
      "loss": 0.4913,
      "step": 28000
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.159213847818248e-06,
      "loss": 0.4862,
      "step": 28500
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.9338261810313745e-06,
      "loss": 0.5005,
      "step": 29000
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.708889289578074e-06,
      "loss": 0.498,
      "step": 29500
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.483952398124775e-06,
      "loss": 0.5037,
      "step": 30000
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.2585647313379015e-06,
      "loss": 0.4922,
      "step": 30500
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.033177064551028e-06,
      "loss": 0.4944,
      "step": 31000
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.807789397764155e-06,
      "loss": 0.4795,
      "step": 31500
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.5824017309772815e-06,
      "loss": 0.498,
      "step": 32000
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.357014064190408e-06,
      "loss": 0.4985,
      "step": 32500
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.131626397403534e-06,
      "loss": 0.4915,
      "step": 33000
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.9062387306166605e-06,
      "loss": 0.4185,
      "step": 33500
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.680851063829788e-06,
      "loss": 0.3421,
      "step": 34000
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.455463397042914e-06,
      "loss": 0.353,
      "step": 34500
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.2300757302560405e-06,
      "loss": 0.33,
      "step": 35000
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.004688063469167e-06,
      "loss": 0.3277,
      "step": 35500
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.779300396682294e-06,
      "loss": 0.3456,
      "step": 36000
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.553912729895421e-06,
      "loss": 0.3295,
      "step": 36500
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.3289758384421207e-06,
      "loss": 0.3293,
      "step": 37000
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.1035881716552475e-06,
      "loss": 0.3478,
      "step": 37500
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8782005048683735e-06,
      "loss": 0.3639,
      "step": 38000
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6528128380815007e-06,
      "loss": 0.3333,
      "step": 38500
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.427425171294627e-06,
      "loss": 0.3324,
      "step": 39000
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.2024882798413274e-06,
      "loss": 0.3435,
      "step": 39500
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.9771006130544537e-06,
      "loss": 0.3512,
      "step": 40000
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.7517129462675803e-06,
      "loss": 0.3388,
      "step": 40500
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.526325279480707e-06,
      "loss": 0.3421,
      "step": 41000
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3009376126938337e-06,
      "loss": 0.3648,
      "step": 41500
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0760007212405338e-06,
      "loss": 0.3271,
      "step": 42000
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.506130544536603e-07,
      "loss": 0.321,
      "step": 42500
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.256761630003607e-07,
      "loss": 0.3432,
      "step": 43000
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.0028849621348723e-07,
      "loss": 0.322,
      "step": 43500
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.7490082942661378e-07,
      "loss": 0.3316,
      "step": 44000
    }
  ],
  "logging_steps": 500,
  "max_steps": 44368,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 6.955379978528563e+16,
  "trial_name": null,
  "trial_params": null
}