{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 1000,
  "global_step": 11038,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00036258158085569254,
      "grad_norm": 7.228061676025391,
      "learning_rate": 0.0,
      "loss": 4.1707,
      "step": 1
    },
    {
      "epoch": 0.36258158085569253,
      "grad_norm": 5.671091556549072,
      "learning_rate": 6.036253776435046e-06,
      "loss": 3.5534,
      "step": 1000
    },
    {
      "epoch": 0.36258158085569253,
      "eval_cosine_accuracy": 0.9460511207580566,
      "eval_loss": 0.562572181224823,
      "eval_runtime": 36.4701,
      "eval_samples_per_second": 260.734,
      "eval_steps_per_second": 1.042,
      "step": 1000
    },
    {
      "epoch": 0.7251631617113851,
      "grad_norm": 7.58541202545166,
      "learning_rate": 1.20785498489426e-05,
      "loss": 2.3098,
      "step": 2000
    },
    {
      "epoch": 0.7251631617113851,
      "eval_cosine_accuracy": 0.9515196084976196,
      "eval_loss": 0.48958125710487366,
      "eval_runtime": 34.58,
      "eval_samples_per_second": 274.986,
      "eval_steps_per_second": 1.099,
      "step": 2000
    },
    {
      "epoch": 1.0876811594202898,
      "grad_norm": 7.139122486114502,
      "learning_rate": 1.8120845921450153e-05,
      "loss": 1.7306,
      "step": 3000
    },
    {
      "epoch": 1.0876811594202898,
      "eval_cosine_accuracy": 0.9593017101287842,
      "eval_loss": 0.4472808539867401,
      "eval_runtime": 34.9782,
      "eval_samples_per_second": 271.855,
      "eval_steps_per_second": 1.086,
      "step": 3000
    },
    {
      "epoch": 1.45,
      "grad_norm": 6.1509857177734375,
      "learning_rate": 1.8959057259404747e-05,
      "loss": 1.8694,
      "step": 4000
    },
    {
      "epoch": 1.45,
      "eval_cosine_accuracy": 0.9605636596679688,
      "eval_loss": 0.4307531416416168,
      "eval_runtime": 34.7892,
      "eval_samples_per_second": 273.332,
      "eval_steps_per_second": 1.092,
      "step": 4000
    },
    {
      "epoch": 1.8123188405797102,
      "grad_norm": 5.870842933654785,
      "learning_rate": 1.7449765825653423e-05,
      "loss": 1.6628,
      "step": 5000
    },
    {
      "epoch": 1.8123188405797102,
      "eval_cosine_accuracy": 0.9643495678901672,
      "eval_loss": 0.4217974841594696,
      "eval_runtime": 34.6903,
      "eval_samples_per_second": 274.111,
      "eval_steps_per_second": 1.095,
      "step": 5000
    },
    {
      "epoch": 2.17463768115942,
      "grad_norm": 5.058931827545166,
      "learning_rate": 1.5938963589666113e-05,
      "loss": 1.5159,
      "step": 6000
    },
    {
      "epoch": 2.17463768115942,
      "eval_cosine_accuracy": 0.9647701978683472,
      "eval_loss": 0.4152602255344391,
      "eval_runtime": 36.1266,
      "eval_samples_per_second": 263.213,
      "eval_steps_per_second": 1.052,
      "step": 6000
    },
    {
      "epoch": 2.5369565217391306,
      "grad_norm": 4.676874160766602,
      "learning_rate": 1.4429672155914791e-05,
      "loss": 1.435,
      "step": 7000
    },
    {
      "epoch": 2.5369565217391306,
      "eval_cosine_accuracy": 0.9668734669685364,
      "eval_loss": 0.4095751941204071,
      "eval_runtime": 34.6705,
      "eval_samples_per_second": 274.268,
      "eval_steps_per_second": 1.096,
      "step": 7000
    },
    {
      "epoch": 2.8992753623188405,
      "grad_norm": 4.9811553955078125,
      "learning_rate": 1.2918869919927484e-05,
      "loss": 1.3973,
      "step": 8000
    },
    {
      "epoch": 2.8992753623188405,
      "eval_cosine_accuracy": 0.9683457612991333,
      "eval_loss": 0.3964327871799469,
      "eval_runtime": 35.1522,
      "eval_samples_per_second": 270.509,
      "eval_steps_per_second": 1.081,
      "step": 8000
    },
    {
      "epoch": 3.261594202898551,
      "grad_norm": 6.33431339263916,
      "learning_rate": 1.1409578486176161e-05,
      "loss": 1.3101,
      "step": 9000
    },
    {
      "epoch": 3.261594202898551,
      "eval_cosine_accuracy": 0.9673992991447449,
      "eval_loss": 0.39831292629241943,
      "eval_runtime": 35.4714,
      "eval_samples_per_second": 268.075,
      "eval_steps_per_second": 1.071,
      "step": 9000
    },
    {
      "epoch": 3.623913043478261,
      "grad_norm": 7.3262763023376465,
      "learning_rate": 9.89877625018885e-06,
      "loss": 1.3044,
      "step": 10000
    },
    {
      "epoch": 3.623913043478261,
      "eval_cosine_accuracy": 0.9680302739143372,
      "eval_loss": 0.3955402970314026,
      "eval_runtime": 35.2264,
      "eval_samples_per_second": 269.94,
      "eval_steps_per_second": 1.079,
      "step": 10000
    },
    {
      "epoch": 3.986231884057971,
      "grad_norm": 5.835394382476807,
      "learning_rate": 8.390995618673516e-06,
      "loss": 1.2367,
      "step": 11000
    },
    {
      "epoch": 3.986231884057971,
      "eval_cosine_accuracy": 0.9683457612991333,
      "eval_loss": 0.3905192017555237,
      "eval_runtime": 35.7011,
      "eval_samples_per_second": 266.351,
      "eval_steps_per_second": 1.064,
      "step": 11000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 16548,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}