last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.004011794676348464,
  "eval_steps": 10,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8.023589352696929e-05,
      "eval_loss": NaN,
      "eval_runtime": 358.6886,
      "eval_samples_per_second": 14.631,
      "eval_steps_per_second": 7.316,
      "step": 1
    },
    {
      "epoch": 0.00040117946763484647,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.0,
      "step": 5
    },
    {
      "epoch": 0.0008023589352696929,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 10
    },
    {
      "epoch": 0.0008023589352696929,
      "eval_loss": NaN,
      "eval_runtime": 304.5922,
      "eval_samples_per_second": 17.23,
      "eval_steps_per_second": 8.615,
      "step": 10
    },
    {
      "epoch": 0.0012035384029045392,
      "grad_norm": NaN,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.0016047178705393859,
      "grad_norm": NaN,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.0,
      "step": 20
    },
    {
      "epoch": 0.0016047178705393859,
      "eval_loss": NaN,
      "eval_runtime": 212.5377,
      "eval_samples_per_second": 24.692,
      "eval_steps_per_second": 12.346,
      "step": 20
    },
    {
      "epoch": 0.002005897338174232,
      "grad_norm": NaN,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0,
      "step": 25
    },
    {
      "epoch": 0.0024070768058090785,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.0024070768058090785,
      "eval_loss": NaN,
      "eval_runtime": 140.4237,
      "eval_samples_per_second": 37.373,
      "eval_steps_per_second": 18.686,
      "step": 30
    },
    {
      "epoch": 0.0028082562734439253,
      "grad_norm": NaN,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0,
      "step": 35
    },
    {
      "epoch": 0.0032094357410787718,
      "grad_norm": NaN,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.0,
      "step": 40
    },
    {
      "epoch": 0.0032094357410787718,
      "eval_loss": NaN,
      "eval_runtime": 129.5231,
      "eval_samples_per_second": 40.518,
      "eval_steps_per_second": 20.259,
      "step": 40
    },
    {
      "epoch": 0.003610615208713618,
      "grad_norm": NaN,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0,
      "step": 45
    },
    {
      "epoch": 0.004011794676348464,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.004011794676348464,
      "eval_loss": NaN,
      "eval_runtime": 130.1808,
      "eval_samples_per_second": 40.313,
      "eval_steps_per_second": 20.157,
      "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": 1621491292569600.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}