last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.023391812865497075,
  "eval_steps": 8,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0011695906432748538,
      "eval_loss": 1.8938556909561157,
      "eval_runtime": 48.0467,
      "eval_samples_per_second": 7.514,
      "eval_steps_per_second": 3.767,
      "step": 1
    },
    {
      "epoch": 0.0035087719298245615,
      "grad_norm": 20.444992065429688,
      "learning_rate": 6e-05,
      "loss": 7.8612,
      "step": 3
    },
    {
      "epoch": 0.007017543859649123,
      "grad_norm": 19.376476287841797,
      "learning_rate": 0.00012,
      "loss": 7.2206,
      "step": 6
    },
    {
      "epoch": 0.00935672514619883,
      "eval_loss": 1.771126627922058,
      "eval_runtime": 47.7516,
      "eval_samples_per_second": 7.56,
      "eval_steps_per_second": 3.79,
      "step": 8
    },
    {
      "epoch": 0.010526315789473684,
      "grad_norm": 17.302108764648438,
      "learning_rate": 0.00018,
      "loss": 8.1698,
      "step": 9
    },
    {
      "epoch": 0.014035087719298246,
      "grad_norm": 12.537629127502441,
      "learning_rate": 0.00019510565162951537,
      "loss": 7.0124,
      "step": 12
    },
    {
      "epoch": 0.017543859649122806,
      "grad_norm": 13.109160423278809,
      "learning_rate": 0.00017071067811865476,
      "loss": 7.5845,
      "step": 15
    },
    {
      "epoch": 0.01871345029239766,
      "eval_loss": 1.7047653198242188,
      "eval_runtime": 47.9777,
      "eval_samples_per_second": 7.524,
      "eval_steps_per_second": 3.773,
      "step": 16
    },
    {
      "epoch": 0.021052631578947368,
      "grad_norm": 17.37651824951172,
      "learning_rate": 0.00013090169943749476,
      "loss": 7.0767,
      "step": 18
    }
  ],
  "logging_steps": 3,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7031303039877120.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}