last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0015222437873425429,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00015222437873425428,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 7.8275,
      "step": 1
    },
    {
      "epoch": 0.00015222437873425428,
      "eval_loss": 7.669393062591553,
      "eval_runtime": 93.5786,
      "eval_samples_per_second": 118.232,
      "eval_steps_per_second": 14.779,
      "step": 1
    },
    {
      "epoch": 0.00030444875746850857,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 7.9338,
      "step": 2
    },
    {
      "epoch": 0.0004566731362027629,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 7.5511,
      "step": 3
    },
    {
      "epoch": 0.0004566731362027629,
      "eval_loss": 7.669393062591553,
      "eval_runtime": 93.1619,
      "eval_samples_per_second": 118.761,
      "eval_steps_per_second": 14.845,
      "step": 3
    },
    {
      "epoch": 0.0006088975149370171,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 8.226,
      "step": 4
    },
    {
      "epoch": 0.0007611218936712714,
      "grad_norm": 26.795257568359375,
      "learning_rate": 2e-05,
      "loss": 7.7784,
      "step": 5
    },
    {
      "epoch": 0.0009133462724055258,
      "grad_norm": 26.446081161499023,
      "learning_rate": 4e-05,
      "loss": 7.7508,
      "step": 6
    },
    {
      "epoch": 0.0009133462724055258,
      "eval_loss": 7.27972412109375,
      "eval_runtime": 90.9362,
      "eval_samples_per_second": 121.668,
      "eval_steps_per_second": 15.208,
      "step": 6
    },
    {
      "epoch": 0.00106557065113978,
      "grad_norm": 25.43058204650879,
      "learning_rate": 6e-05,
      "loss": 7.1342,
      "step": 7
    },
    {
      "epoch": 0.0012177950298740343,
      "grad_norm": 20.734403610229492,
      "learning_rate": 8e-05,
      "loss": 6.4785,
      "step": 8
    },
    {
      "epoch": 0.0013700194086082886,
      "grad_norm": 19.443437576293945,
      "learning_rate": 0.0001,
      "loss": 5.6144,
      "step": 9
    },
    {
      "epoch": 0.0013700194086082886,
      "eval_loss": 4.204019069671631,
      "eval_runtime": 91.0755,
      "eval_samples_per_second": 121.482,
      "eval_steps_per_second": 15.185,
      "step": 9
    },
    {
      "epoch": 0.0015222437873425429,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 4.1238,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2883817203302400.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}