trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.953846153846154,
  "eval_steps": 500,
  "global_step": 72,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 1.7405303198094355,
      "learning_rate": 5e-06,
      "loss": 1.0638,
      "step": 10
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 1.0374735844199803,
      "learning_rate": 5e-06,
      "loss": 0.9425,
      "step": 20
    },
    {
      "epoch": 0.9846153846153847,
      "eval_loss": 0.915921151638031,
      "eval_runtime": 76.9595,
      "eval_samples_per_second": 8.537,
      "eval_steps_per_second": 0.273,
      "step": 24
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 2.1037944027880453,
      "learning_rate": 5e-06,
      "loss": 0.8961,
      "step": 30
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 1.0102534268182652,
      "learning_rate": 5e-06,
      "loss": 0.8596,
      "step": 40
    },
    {
      "epoch": 1.9692307692307693,
      "eval_loss": 0.8778233528137207,
      "eval_runtime": 76.3388,
      "eval_samples_per_second": 8.606,
      "eval_steps_per_second": 0.275,
      "step": 48
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 1.3638440690247002,
      "learning_rate": 5e-06,
      "loss": 0.8365,
      "step": 50
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 1.0944344666695314,
      "learning_rate": 5e-06,
      "loss": 0.7984,
      "step": 60
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 0.7516976026024018,
      "learning_rate": 5e-06,
      "loss": 0.7847,
      "step": 70
    },
    {
      "epoch": 2.953846153846154,
      "eval_loss": 0.8662426471710205,
      "eval_runtime": 76.2223,
      "eval_samples_per_second": 8.62,
      "eval_steps_per_second": 0.276,
      "step": 72
    },
    {
      "epoch": 2.953846153846154,
      "step": 72,
      "total_flos": 120497991843840.0,
      "train_loss": 0.8805405447880427,
      "train_runtime": 11751.6604,
      "train_samples_per_second": 3.185,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 10,
  "max_steps": 72,
  "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": 120497991843840.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}