last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0490075961774075,
  "eval_steps": 10,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00098015192354815,
      "eval_loss": 0.7627389430999756,
      "eval_runtime": 26.8533,
      "eval_samples_per_second": 16.013,
      "eval_steps_per_second": 8.006,
      "step": 1
    },
    {
      "epoch": 0.00490075961774075,
      "grad_norm": 14.23064136505127,
      "learning_rate": 5e-05,
      "loss": 3.1124,
      "step": 5
    },
    {
      "epoch": 0.0098015192354815,
      "grad_norm": 9.638243675231934,
      "learning_rate": 0.0001,
      "loss": 2.274,
      "step": 10
    },
    {
      "epoch": 0.0098015192354815,
      "eval_loss": 0.39032769203186035,
      "eval_runtime": 27.1203,
      "eval_samples_per_second": 15.855,
      "eval_steps_per_second": 7.928,
      "step": 10
    },
    {
      "epoch": 0.014702278853222249,
      "grad_norm": 15.856855392456055,
      "learning_rate": 9.619397662556435e-05,
      "loss": 1.2127,
      "step": 15
    },
    {
      "epoch": 0.019603038470963,
      "grad_norm": 8.120400428771973,
      "learning_rate": 8.535533905932738e-05,
      "loss": 1.085,
      "step": 20
    },
    {
      "epoch": 0.019603038470963,
      "eval_loss": 0.26506921648979187,
      "eval_runtime": 27.2179,
      "eval_samples_per_second": 15.798,
      "eval_steps_per_second": 7.899,
      "step": 20
    },
    {
      "epoch": 0.02450379808870375,
      "grad_norm": 9.337193489074707,
      "learning_rate": 6.91341716182545e-05,
      "loss": 1.1599,
      "step": 25
    },
    {
      "epoch": 0.029404557706444498,
      "grad_norm": 6.683337688446045,
      "learning_rate": 5e-05,
      "loss": 1.0121,
      "step": 30
    },
    {
      "epoch": 0.029404557706444498,
      "eval_loss": 0.24212950468063354,
      "eval_runtime": 27.4102,
      "eval_samples_per_second": 15.688,
      "eval_steps_per_second": 7.844,
      "step": 30
    },
    {
      "epoch": 0.03430531732418525,
      "grad_norm": 7.745213985443115,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.9605,
      "step": 35
    },
    {
      "epoch": 0.039206076941926,
      "grad_norm": 10.360566139221191,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.7756,
      "step": 40
    },
    {
      "epoch": 0.039206076941926,
      "eval_loss": 0.23207147419452667,
      "eval_runtime": 27.4395,
      "eval_samples_per_second": 15.671,
      "eval_steps_per_second": 7.835,
      "step": 40
    },
    {
      "epoch": 0.04410683655966675,
      "grad_norm": 6.0309343338012695,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.7961,
      "step": 45
    },
    {
      "epoch": 0.0490075961774075,
      "grad_norm": 5.921755313873291,
      "learning_rate": 0.0,
      "loss": 0.835,
      "step": 50
    },
    {
      "epoch": 0.0490075961774075,
      "eval_loss": 0.2300577163696289,
      "eval_runtime": 27.2186,
      "eval_samples_per_second": 15.798,
      "eval_steps_per_second": 7.899,
      "step": 50
    }
  ],
  "logging_steps": 5,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 13,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8767214400307200.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}