{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 1000,
  "global_step": 6180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003237293622531564,
      "grad_norm": 5.294984340667725,
      "learning_rate": 0.0,
      "loss": 2.5131,
      "step": 1
    },
    {
      "epoch": 0.32372936225315635,
      "grad_norm": 3.888833522796631,
      "learning_rate": 2.424757281553398e-05,
      "loss": 1.8415,
      "step": 1000
    },
    {
      "epoch": 0.32372936225315635,
      "eval_cosine_accuracy": 0.9511937499046326,
      "eval_loss": 1.082350730895996,
      "eval_runtime": 25.045,
      "eval_samples_per_second": 377.959,
      "eval_steps_per_second": 1.477,
      "step": 1000
    },
    {
      "epoch": 0.6474587245063127,
      "grad_norm": 3.7192389965057373,
      "learning_rate": 2.7941546762589928e-05,
      "loss": 1.3696,
      "step": 2000
    },
    {
      "epoch": 0.6474587245063127,
      "eval_cosine_accuracy": 0.9616522192955017,
      "eval_loss": 0.9928585290908813,
      "eval_runtime": 24.1997,
      "eval_samples_per_second": 391.162,
      "eval_steps_per_second": 1.529,
      "step": 2000
    },
    {
      "epoch": 0.9711880867594691,
      "grad_norm": 3.8493621349334717,
      "learning_rate": 2.5243705035971223e-05,
      "loss": 1.4502,
      "step": 3000
    },
    {
      "epoch": 0.9711880867594691,
      "eval_cosine_accuracy": 0.9655609726905823,
      "eval_loss": 0.94865882396698,
      "eval_runtime": 24.3555,
      "eval_samples_per_second": 388.66,
      "eval_steps_per_second": 1.519,
      "step": 3000
    },
    {
      "epoch": 1.2947266256874799,
      "grad_norm": 3.0650620460510254,
      "learning_rate": 2.254586330935252e-05,
      "loss": 1.3141,
      "step": 4000
    },
    {
      "epoch": 1.2947266256874799,
      "eval_cosine_accuracy": 0.9704204797744751,
      "eval_loss": 0.8925007581710815,
      "eval_runtime": 24.3977,
      "eval_samples_per_second": 387.987,
      "eval_steps_per_second": 1.517,
      "step": 4000
    },
    {
      "epoch": 1.6182465221611129,
      "grad_norm": 2.806612968444824,
      "learning_rate": 1.985071942446043e-05,
      "loss": 1.1692,
      "step": 5000
    },
    {
      "epoch": 1.6182465221611129,
      "eval_cosine_accuracy": 0.9709486365318298,
      "eval_loss": 0.8780702352523804,
      "eval_runtime": 24.3201,
      "eval_samples_per_second": 389.226,
      "eval_steps_per_second": 1.521,
      "step": 5000
    },
    {
      "epoch": 1.941766418634746,
      "grad_norm": 3.0037014484405518,
      "learning_rate": 1.7152877697841724e-05,
      "loss": 1.1209,
      "step": 6000
    },
    {
      "epoch": 1.941766418634746,
      "eval_cosine_accuracy": 0.9717937707901001,
      "eval_loss": 0.8578657507896423,
      "eval_runtime": 24.5063,
      "eval_samples_per_second": 386.268,
      "eval_steps_per_second": 1.51,
      "step": 6000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 12356,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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
}