last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.006726080376660501,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006726080376660501,
      "grad_norm": 6.524783134460449,
      "learning_rate": 2e-05,
      "loss": 3.4708,
      "step": 1
    },
    {
      "epoch": 0.0006726080376660501,
      "eval_loss": 3.2175650596618652,
      "eval_runtime": 65.5188,
      "eval_samples_per_second": 4.793,
      "eval_steps_per_second": 4.793,
      "step": 1
    },
    {
      "epoch": 0.0013452160753321001,
      "grad_norm": 4.642040252685547,
      "learning_rate": 4e-05,
      "loss": 3.6476,
      "step": 2
    },
    {
      "epoch": 0.0020178241129981504,
      "grad_norm": 5.972542762756348,
      "learning_rate": 6e-05,
      "loss": 2.9136,
      "step": 3
    },
    {
      "epoch": 0.0020178241129981504,
      "eval_loss": 3.037116527557373,
      "eval_runtime": 65.6068,
      "eval_samples_per_second": 4.786,
      "eval_steps_per_second": 4.786,
      "step": 3
    },
    {
      "epoch": 0.0026904321506642003,
      "grad_norm": 5.8406782150268555,
      "learning_rate": 8e-05,
      "loss": 2.9974,
      "step": 4
    },
    {
      "epoch": 0.0033630401883302506,
      "grad_norm": 5.463853359222412,
      "learning_rate": 0.0001,
      "loss": 2.4602,
      "step": 5
    },
    {
      "epoch": 0.004035648225996301,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 2.7837,
      "step": 6
    },
    {
      "epoch": 0.004035648225996301,
      "eval_loss": 2.5530920028686523,
      "eval_runtime": 65.5723,
      "eval_samples_per_second": 4.789,
      "eval_steps_per_second": 4.789,
      "step": 6
    },
    {
      "epoch": 0.004708256263662351,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 2.541,
      "step": 7
    },
    {
      "epoch": 0.0053808643013284006,
      "grad_norm": 4.99736213684082,
      "learning_rate": 0.00012,
      "loss": 2.5739,
      "step": 8
    },
    {
      "epoch": 0.006053472338994451,
      "grad_norm": 4.587267875671387,
      "learning_rate": 0.00014,
      "loss": 2.1589,
      "step": 9
    },
    {
      "epoch": 0.006053472338994451,
      "eval_loss": 2.139094829559326,
      "eval_runtime": 124.0009,
      "eval_samples_per_second": 2.532,
      "eval_steps_per_second": 2.532,
      "step": 9
    },
    {
      "epoch": 0.006726080376660501,
      "grad_norm": 5.99055814743042,
      "learning_rate": 0.00016,
      "loss": 2.0694,
      "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": 7418872835604480.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}