last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.006426735218508998,
  "eval_steps": 10,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00012853470437017994,
      "eval_loss": NaN,
      "eval_runtime": 369.3591,
      "eval_samples_per_second": 8.869,
      "eval_steps_per_second": 4.435,
      "step": 1
    },
    {
      "epoch": 0.0006426735218508997,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 14.901,
      "step": 5
    },
    {
      "epoch": 0.0012853470437017994,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 10
    },
    {
      "epoch": 0.0012853470437017994,
      "eval_loss": NaN,
      "eval_runtime": 358.8544,
      "eval_samples_per_second": 9.129,
      "eval_steps_per_second": 4.565,
      "step": 10
    },
    {
      "epoch": 0.0019280205655526992,
      "grad_norm": NaN,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.002570694087403599,
      "grad_norm": NaN,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.0,
      "step": 20
    },
    {
      "epoch": 0.002570694087403599,
      "eval_loss": NaN,
      "eval_runtime": 408.3596,
      "eval_samples_per_second": 8.022,
      "eval_steps_per_second": 4.011,
      "step": 20
    },
    {
      "epoch": 0.003213367609254499,
      "grad_norm": NaN,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0,
      "step": 25
    },
    {
      "epoch": 0.0038560411311053984,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.0038560411311053984,
      "eval_loss": NaN,
      "eval_runtime": 439.1612,
      "eval_samples_per_second": 7.46,
      "eval_steps_per_second": 3.73,
      "step": 30
    },
    {
      "epoch": 0.004498714652956298,
      "grad_norm": NaN,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0,
      "step": 35
    },
    {
      "epoch": 0.005141388174807198,
      "grad_norm": NaN,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.0,
      "step": 40
    },
    {
      "epoch": 0.005141388174807198,
      "eval_loss": NaN,
      "eval_runtime": 425.7477,
      "eval_samples_per_second": 7.695,
      "eval_steps_per_second": 3.847,
      "step": 40
    },
    {
      "epoch": 0.005784061696658098,
      "grad_norm": NaN,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 0.006426735218508998,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.006426735218508998,
      "eval_loss": NaN,
      "eval_runtime": 431.409,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "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": 834732785664000.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}