trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.999745123027154,
  "global_step": 9561,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "learning_rate": 0.0001,
      "loss": 4.7889,
      "step": 500
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001,
      "loss": 4.007,
      "step": 1000
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001,
      "loss": 3.8226,
      "step": 1500
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0001,
      "loss": 3.7134,
      "step": 2000
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0001,
      "loss": 3.643,
      "step": 2500
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0001,
      "loss": 3.5814,
      "step": 3000
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0001,
      "loss": 3.5365,
      "step": 3500
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0001,
      "loss": 3.4833,
      "step": 4000
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0001,
      "loss": 3.4472,
      "step": 4500
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0001,
      "loss": 3.4199,
      "step": 5000
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0001,
      "loss": 3.398,
      "step": 5500
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0001,
      "loss": 3.3639,
      "step": 6000
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0001,
      "loss": 3.3425,
      "step": 6500
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0001,
      "loss": 3.3083,
      "step": 7000
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0001,
      "loss": 3.2997,
      "step": 7500
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.0001,
      "loss": 3.2649,
      "step": 8000
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0001,
      "loss": 3.258,
      "step": 8500
    },
    {
      "epoch": 2.82,
      "learning_rate": 0.0001,
      "loss": 3.2514,
      "step": 9000
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.0001,
      "loss": 3.2301,
      "step": 9500
    },
    {
      "epoch": 3.0,
      "step": 9561,
      "total_flos": 8.150411322232013e+16,
      "train_loss": 3.5328583012638624,
      "train_runtime": 7182.8846,
      "train_samples_per_second": 85.21,
      "train_steps_per_second": 1.331
    }
  ],
  "max_steps": 9561,
  "num_train_epochs": 3,
  "total_flos": 8.150411322232013e+16,
  "trial_name": null,
  "trial_params": null
}