last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01708671507902606,
  "eval_steps": 1,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017086715079026058,
      "grad_norm": 2.0610387325286865,
      "learning_rate": 2e-05,
      "loss": 2.7329,
      "step": 1
    },
    {
      "epoch": 0.0017086715079026058,
      "eval_loss": 2.804917097091675,
      "eval_runtime": 63.1764,
      "eval_samples_per_second": 3.91,
      "eval_steps_per_second": 1.963,
      "step": 1
    },
    {
      "epoch": 0.0034173430158052115,
      "grad_norm": 1.7274668216705322,
      "learning_rate": 4e-05,
      "loss": 2.4894,
      "step": 2
    },
    {
      "epoch": 0.0034173430158052115,
      "eval_loss": 2.803224563598633,
      "eval_runtime": 63.3099,
      "eval_samples_per_second": 3.901,
      "eval_steps_per_second": 1.959,
      "step": 2
    },
    {
      "epoch": 0.0051260145237078175,
      "grad_norm": 1.8122687339782715,
      "learning_rate": 6e-05,
      "loss": 2.4282,
      "step": 3
    },
    {
      "epoch": 0.0051260145237078175,
      "eval_loss": 2.8019886016845703,
      "eval_runtime": 63.3714,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 3
    },
    {
      "epoch": 0.006834686031610423,
      "grad_norm": 1.8031339645385742,
      "learning_rate": 8e-05,
      "loss": 2.5053,
      "step": 4
    },
    {
      "epoch": 0.006834686031610423,
      "eval_loss": 2.797816038131714,
      "eval_runtime": 63.365,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 4
    },
    {
      "epoch": 0.00854335753951303,
      "grad_norm": 2.084747314453125,
      "learning_rate": 0.0001,
      "loss": 2.8,
      "step": 5
    },
    {
      "epoch": 0.00854335753951303,
      "eval_loss": 2.7864952087402344,
      "eval_runtime": 63.3641,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 5
    },
    {
      "epoch": 0.010252029047415635,
      "grad_norm": 1.8946125507354736,
      "learning_rate": 0.00012,
      "loss": 2.8036,
      "step": 6
    },
    {
      "epoch": 0.010252029047415635,
      "eval_loss": 2.7658321857452393,
      "eval_runtime": 63.3725,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 6
    },
    {
      "epoch": 0.01196070055531824,
      "grad_norm": 1.428999423980713,
      "learning_rate": 0.00014,
      "loss": 1.9979,
      "step": 7
    },
    {
      "epoch": 0.01196070055531824,
      "eval_loss": 2.7274694442749023,
      "eval_runtime": 63.3666,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 7
    },
    {
      "epoch": 0.013669372063220846,
      "grad_norm": 2.2038583755493164,
      "learning_rate": 0.00016,
      "loss": 2.555,
      "step": 8
    },
    {
      "epoch": 0.013669372063220846,
      "eval_loss": 2.670942783355713,
      "eval_runtime": 63.3739,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 8
    },
    {
      "epoch": 0.015378043571123452,
      "grad_norm": 2.094712257385254,
      "learning_rate": 0.00018,
      "loss": 2.608,
      "step": 9
    },
    {
      "epoch": 0.015378043571123452,
      "eval_loss": 2.607150077819824,
      "eval_runtime": 63.3683,
      "eval_samples_per_second": 3.898,
      "eval_steps_per_second": 1.957,
      "step": 9
    },
    {
      "epoch": 0.01708671507902606,
      "grad_norm": 1.9949012994766235,
      "learning_rate": 0.0002,
      "loss": 2.4009,
      "step": 10
    },
    {
      "epoch": 0.01708671507902606,
      "eval_loss": 2.5534861087799072,
      "eval_runtime": 63.3498,
      "eval_samples_per_second": 3.899,
      "eval_steps_per_second": 1.957,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7398256992583680.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}