trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 5013,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3,
      "learning_rate": 4.501296628765211e-05,
      "loss": 3.2601,
      "step": 500
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.002593257530421e-05,
      "loss": 3.0867,
      "step": 1000
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.5038898862956315e-05,
      "loss": 2.9927,
      "step": 1500
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.005186515060842e-05,
      "loss": 2.8095,
      "step": 2000
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.506483143826052e-05,
      "loss": 2.6831,
      "step": 2500
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.0077797725912628e-05,
      "loss": 2.6323,
      "step": 3000
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5090764013564732e-05,
      "loss": 2.5364,
      "step": 3500
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0103730301216837e-05,
      "loss": 2.3822,
      "step": 4000
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.116696588868941e-06,
      "loss": 2.35,
      "step": 4500
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.296628765210453e-07,
      "loss": 2.332,
      "step": 5000
    },
    {
      "epoch": 3.0,
      "step": 5013,
      "total_flos": 5501794620801024.0,
      "train_loss": 2.7053030034203553,
      "train_runtime": 2600.5654,
      "train_samples_per_second": 7.708,
      "train_steps_per_second": 1.928
    }
  ],
  "max_steps": 5013,
  "num_train_epochs": 3,
  "total_flos": 5501794620801024.0,
  "trial_name": null,
  "trial_params": null
}