trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.992,
  "eval_steps": 500,
  "global_step": 62,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "grad_norm": 1.7386187314987183,
      "learning_rate": 1.9853538358476933e-05,
      "loss": 0.2613,
      "step": 10
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.1634316444396973,
      "learning_rate": 1.736741137876405e-05,
      "loss": 0.216,
      "step": 20
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.8627727031707764,
      "learning_rate": 1.2542183341934873e-05,
      "loss": 0.1431,
      "step": 30
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.0235531330108643,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.088,
      "step": 40
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.9142417907714844,
      "learning_rate": 2.2585838936091753e-06,
      "loss": 0.107,
      "step": 50
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.0446289777755737,
      "learning_rate": 6.51826465144978e-08,
      "loss": 0.118,
      "step": 60
    },
    {
      "epoch": 0.992,
      "step": 62,
      "total_flos": 1.6166347506450432e+16,
      "train_loss": 0.15166758986250048,
      "train_runtime": 892.7422,
      "train_samples_per_second": 0.56,
      "train_steps_per_second": 0.069
    }
  ],
  "logging_steps": 10,
  "max_steps": 62,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6166347506450432e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}