{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.96370235934664,
  "global_step": 22000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.45,
      "learning_rate": 1.9546279491833034e-05,
      "loss": 0.834,
      "step": 500
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.9092558983666063e-05,
      "loss": 0.8279,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8352304100990295,
      "eval_runtime": 300.0386,
      "eval_samples_per_second": 338.6,
      "eval_steps_per_second": 5.293,
      "step": 1102
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.8638838475499095e-05,
      "loss": 0.8249,
      "step": 1500
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.8185117967332124e-05,
      "loss": 0.8218,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8333624005317688,
      "eval_runtime": 299.0242,
      "eval_samples_per_second": 339.748,
      "eval_steps_per_second": 5.311,
      "step": 2204
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.7731397459165156e-05,
      "loss": 0.8171,
      "step": 2500
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.7277676950998185e-05,
      "loss": 0.8138,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.8279737234115601,
      "eval_runtime": 299.8977,
      "eval_samples_per_second": 338.759,
      "eval_steps_per_second": 5.295,
      "step": 3306
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.6823956442831217e-05,
      "loss": 0.8094,
      "step": 3500
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.637023593466425e-05,
      "loss": 0.8084,
      "step": 4000
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.8279556035995483,
      "eval_runtime": 305.7805,
      "eval_samples_per_second": 332.242,
      "eval_steps_per_second": 5.193,
      "step": 4408
    },
    {
      "epoch": 4.08,
      "learning_rate": 1.5916515426497278e-05,
      "loss": 0.8028,
      "step": 4500
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.546279491833031e-05,
      "loss": 0.7997,
      "step": 5000
    },
    {
      "epoch": 4.99,
      "learning_rate": 1.500907441016334e-05,
      "loss": 0.7975,
      "step": 5500
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.8251700401306152,
      "eval_runtime": 299.8544,
      "eval_samples_per_second": 338.808,
      "eval_steps_per_second": 5.296,
      "step": 5510
    },
    {
      "epoch": 5.44,
      "learning_rate": 1.4555353901996372e-05,
      "loss": 0.798,
      "step": 6000
    },
    {
      "epoch": 5.9,
      "learning_rate": 1.4101633393829401e-05,
      "loss": 0.7965,
      "step": 6500
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.8234975337982178,
      "eval_runtime": 299.5265,
      "eval_samples_per_second": 339.179,
      "eval_steps_per_second": 5.302,
      "step": 6612
    },
    {
      "epoch": 6.35,
      "learning_rate": 1.3647912885662433e-05,
      "loss": 0.7913,
      "step": 7000
    },
    {
      "epoch": 6.81,
      "learning_rate": 1.3194192377495462e-05,
      "loss": 0.7858,
      "step": 7500
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.8199203014373779,
      "eval_runtime": 298.6313,
      "eval_samples_per_second": 340.195,
      "eval_steps_per_second": 5.318,
      "step": 7714
    },
    {
      "epoch": 7.26,
      "learning_rate": 1.2740471869328494e-05,
      "loss": 0.7845,
      "step": 8000
    },
    {
      "epoch": 7.71,
      "learning_rate": 1.2286751361161527e-05,
      "loss": 0.7808,
      "step": 8500
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.8201999664306641,
      "eval_runtime": 305.3068,
      "eval_samples_per_second": 332.757,
      "eval_steps_per_second": 5.201,
      "step": 8816
    },
    {
      "epoch": 8.17,
      "learning_rate": 1.1833030852994555e-05,
      "loss": 0.7824,
      "step": 9000
    },
    {
      "epoch": 8.62,
      "learning_rate": 1.1379310344827587e-05,
      "loss": 0.7805,
      "step": 9500
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.8179581165313721,
      "eval_runtime": 299.8221,
      "eval_samples_per_second": 338.844,
      "eval_steps_per_second": 5.296,
      "step": 9918
    },
    {
      "epoch": 9.07,
      "learning_rate": 1.0925589836660618e-05,
      "loss": 0.7762,
      "step": 10000
    },
    {
      "epoch": 9.53,
      "learning_rate": 1.0471869328493648e-05,
      "loss": 0.7744,
      "step": 10500
    },
    {
      "epoch": 9.98,
      "learning_rate": 1.0018148820326679e-05,
      "loss": 0.7698,
      "step": 11000
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.813572347164154,
      "eval_runtime": 299.3306,
      "eval_samples_per_second": 339.401,
      "eval_steps_per_second": 5.305,
      "step": 11020
    },
    {
      "epoch": 10.44,
      "learning_rate": 9.564428312159711e-06,
      "loss": 0.768,
      "step": 11500
    },
    {
      "epoch": 10.89,
      "learning_rate": 9.110707803992742e-06,
      "loss": 0.7704,
      "step": 12000
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.8144657015800476,
      "eval_runtime": 299.3629,
      "eval_samples_per_second": 339.364,
      "eval_steps_per_second": 5.305,
      "step": 12122
    },
    {
      "epoch": 11.34,
      "learning_rate": 8.656987295825772e-06,
      "loss": 0.7649,
      "step": 12500
    },
    {
      "epoch": 11.8,
      "learning_rate": 8.203266787658803e-06,
      "loss": 0.7666,
      "step": 13000
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.8124042749404907,
      "eval_runtime": 306.0664,
      "eval_samples_per_second": 331.931,
      "eval_steps_per_second": 5.188,
      "step": 13224
    },
    {
      "epoch": 12.25,
      "learning_rate": 7.749546279491835e-06,
      "loss": 0.7638,
      "step": 13500
    },
    {
      "epoch": 12.7,
      "learning_rate": 7.295825771324865e-06,
      "loss": 0.7616,
      "step": 14000
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.8092049360275269,
      "eval_runtime": 300.1177,
      "eval_samples_per_second": 338.511,
      "eval_steps_per_second": 5.291,
      "step": 14326
    },
    {
      "epoch": 13.16,
      "learning_rate": 6.842105263157896e-06,
      "loss": 0.7597,
      "step": 14500
    },
    {
      "epoch": 13.61,
      "learning_rate": 6.388384754990926e-06,
      "loss": 0.7558,
      "step": 15000
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.8094133138656616,
      "eval_runtime": 299.6821,
      "eval_samples_per_second": 339.003,
      "eval_steps_per_second": 5.299,
      "step": 15428
    },
    {
      "epoch": 14.07,
      "learning_rate": 5.934664246823957e-06,
      "loss": 0.7575,
      "step": 15500
    },
    {
      "epoch": 14.52,
      "learning_rate": 5.480943738656987e-06,
      "loss": 0.7562,
      "step": 16000
    },
    {
      "epoch": 14.97,
      "learning_rate": 5.027223230490018e-06,
      "loss": 0.7564,
      "step": 16500
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.8088417053222656,
      "eval_runtime": 300.0855,
      "eval_samples_per_second": 338.547,
      "eval_steps_per_second": 5.292,
      "step": 16530
    },
    {
      "epoch": 15.43,
      "learning_rate": 4.57350272232305e-06,
      "loss": 0.7526,
      "step": 17000
    },
    {
      "epoch": 15.88,
      "learning_rate": 4.11978221415608e-06,
      "loss": 0.753,
      "step": 17500
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.8049781918525696,
      "eval_runtime": 304.2347,
      "eval_samples_per_second": 333.93,
      "eval_steps_per_second": 5.22,
      "step": 17632
    },
    {
      "epoch": 16.33,
      "learning_rate": 3.666061705989111e-06,
      "loss": 0.7524,
      "step": 18000
    },
    {
      "epoch": 16.79,
      "learning_rate": 3.212341197822142e-06,
      "loss": 0.7502,
      "step": 18500
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.8041799068450928,
      "eval_runtime": 301.0417,
      "eval_samples_per_second": 337.472,
      "eval_steps_per_second": 5.275,
      "step": 18734
    },
    {
      "epoch": 17.24,
      "learning_rate": 2.7586206896551725e-06,
      "loss": 0.7487,
      "step": 19000
    },
    {
      "epoch": 17.7,
      "learning_rate": 2.3049001814882034e-06,
      "loss": 0.7495,
      "step": 19500
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.8050508499145508,
      "eval_runtime": 298.8367,
      "eval_samples_per_second": 339.962,
      "eval_steps_per_second": 5.314,
      "step": 19836
    },
    {
      "epoch": 18.15,
      "learning_rate": 1.8511796733212343e-06,
      "loss": 0.7478,
      "step": 20000
    },
    {
      "epoch": 18.6,
      "learning_rate": 1.3974591651542652e-06,
      "loss": 0.7479,
      "step": 20500
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.8038392066955566,
      "eval_runtime": 299.5234,
      "eval_samples_per_second": 339.182,
      "eval_steps_per_second": 5.302,
      "step": 20938
    },
    {
      "epoch": 19.06,
      "learning_rate": 9.43738656987296e-07,
      "loss": 0.7469,
      "step": 21000
    },
    {
      "epoch": 19.51,
      "learning_rate": 4.900181488203267e-07,
      "loss": 0.7473,
      "step": 21500
    },
    {
      "epoch": 19.96,
      "learning_rate": 3.629764065335753e-08,
      "loss": 0.7475,
      "step": 22000
    }
  ],
  "max_steps": 22040,
  "num_train_epochs": 20,
  "total_flos": 2.622716924469212e+18,
  "trial_name": null,
  "trial_params": null
}