{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998719590268886,
  "eval_steps": 1000,
  "global_step": 9366,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00032030749519538755,
      "grad_norm": 6.107327938079834,
      "learning_rate": 0.0,
      "loss": 2.2696,
      "step": 1
    },
    {
      "epoch": 0.3203074951953876,
      "grad_norm": 3.857839345932007,
      "learning_rate": 2.977933325424131e-05,
      "loss": 1.6542,
      "step": 1000
    },
    {
      "epoch": 0.3203074951953876,
      "eval_cosine_accuracy": 0.9641829133033752,
      "eval_loss": 0.5915806889533997,
      "eval_runtime": 28.6636,
      "eval_samples_per_second": 327.279,
      "eval_steps_per_second": 1.291,
      "step": 1000
    },
    {
      "epoch": 0.6406149903907752,
      "grad_norm": 3.1807026863098145,
      "learning_rate": 2.6220192193617276e-05,
      "loss": 1.0828,
      "step": 2000
    },
    {
      "epoch": 0.6406149903907752,
      "eval_cosine_accuracy": 0.9688732624053955,
      "eval_loss": 0.5482337474822998,
      "eval_runtime": 27.6361,
      "eval_samples_per_second": 339.447,
      "eval_steps_per_second": 1.339,
      "step": 2000
    },
    {
      "epoch": 0.9609224855861627,
      "grad_norm": 4.243162155151367,
      "learning_rate": 2.2661051132993238e-05,
      "loss": 0.8294,
      "step": 3000
    },
    {
      "epoch": 0.9609224855861627,
      "eval_cosine_accuracy": 0.971858024597168,
      "eval_loss": 0.5288547277450562,
      "eval_runtime": 27.5987,
      "eval_samples_per_second": 339.908,
      "eval_steps_per_second": 1.341,
      "step": 3000
    },
    {
      "epoch": 1.2810499359795133,
      "grad_norm": 5.0122575759887695,
      "learning_rate": 1.910902835449045e-05,
      "loss": 0.762,
      "step": 4000
    },
    {
      "epoch": 1.2810499359795133,
      "eval_cosine_accuracy": 0.9751625657081604,
      "eval_loss": 0.5255293250083923,
      "eval_runtime": 27.7311,
      "eval_samples_per_second": 338.284,
      "eval_steps_per_second": 1.334,
      "step": 4000
    },
    {
      "epoch": 1.6011523687580027,
      "grad_norm": 3.2703332901000977,
      "learning_rate": 1.5549887293866415e-05,
      "loss": 0.7273,
      "step": 5000
    },
    {
      "epoch": 1.6011523687580027,
      "eval_cosine_accuracy": 0.9755889773368835,
      "eval_loss": 0.5073443055152893,
      "eval_runtime": 27.5769,
      "eval_samples_per_second": 340.176,
      "eval_steps_per_second": 1.342,
      "step": 5000
    },
    {
      "epoch": 1.9212548015364916,
      "grad_norm": 3.8906168937683105,
      "learning_rate": 1.1990746233242378e-05,
      "loss": 0.6962,
      "step": 6000
    },
    {
      "epoch": 1.9212548015364916,
      "eval_cosine_accuracy": 0.9772945046424866,
      "eval_loss": 0.49614542722702026,
      "eval_runtime": 27.8537,
      "eval_samples_per_second": 336.795,
      "eval_steps_per_second": 1.328,
      "step": 6000
    },
    {
      "epoch": 2.2413572343149806,
      "grad_norm": 3.059576988220215,
      "learning_rate": 8.431605172618341e-06,
      "loss": 0.6648,
      "step": 7000
    },
    {
      "epoch": 2.2413572343149806,
      "eval_cosine_accuracy": 0.9775077104568481,
      "eval_loss": 0.5007394552230835,
      "eval_runtime": 28.1686,
      "eval_samples_per_second": 333.03,
      "eval_steps_per_second": 1.314,
      "step": 7000
    },
    {
      "epoch": 2.56145966709347,
      "grad_norm": 3.0882303714752197,
      "learning_rate": 4.872464111994306e-06,
      "loss": 0.6362,
      "step": 8000
    },
    {
      "epoch": 2.56145966709347,
      "eval_cosine_accuracy": 0.9779341220855713,
      "eval_loss": 0.5001775622367859,
      "eval_runtime": 27.7547,
      "eval_samples_per_second": 337.996,
      "eval_steps_per_second": 1.333,
      "step": 8000
    },
    {
      "epoch": 2.881562099871959,
      "grad_norm": 2.918884038925171,
      "learning_rate": 1.3133230513702693e-06,
      "loss": 0.6228,
      "step": 9000
    },
    {
      "epoch": 2.881562099871959,
      "eval_cosine_accuracy": 0.9784671068191528,
      "eval_loss": 0.4983352720737457,
      "eval_runtime": 28.0171,
      "eval_samples_per_second": 334.831,
      "eval_steps_per_second": 1.321,
      "step": 9000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 9366,
  "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": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}