last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0008441668073611346,
  "eval_steps": 5,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.1104170184028365e-05,
      "eval_loss": NaN,
      "eval_runtime": 1034.4499,
      "eval_samples_per_second": 9.644,
      "eval_steps_per_second": 4.822,
      "step": 1
    },
    {
      "epoch": 6.331251055208509e-05,
      "grad_norm": NaN,
      "learning_rate": 3e-05,
      "loss": 0.0,
      "step": 3
    },
    {
      "epoch": 0.00010552085092014182,
      "eval_loss": NaN,
      "eval_runtime": 1035.3784,
      "eval_samples_per_second": 9.635,
      "eval_steps_per_second": 4.818,
      "step": 5
    },
    {
      "epoch": 0.00012662502110417017,
      "grad_norm": NaN,
      "learning_rate": 6e-05,
      "loss": 0.0,
      "step": 6
    },
    {
      "epoch": 0.00018993753165625528,
      "grad_norm": NaN,
      "learning_rate": 9e-05,
      "loss": 0.0,
      "step": 9
    },
    {
      "epoch": 0.00021104170184028364,
      "eval_loss": NaN,
      "eval_runtime": 1034.9765,
      "eval_samples_per_second": 9.639,
      "eval_steps_per_second": 4.819,
      "step": 10
    },
    {
      "epoch": 0.00025325004220834034,
      "grad_norm": NaN,
      "learning_rate": 0.00012,
      "loss": 0.0,
      "step": 12
    },
    {
      "epoch": 0.0003165625527604255,
      "grad_norm": NaN,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.0,
      "step": 15
    },
    {
      "epoch": 0.0003165625527604255,
      "eval_loss": NaN,
      "eval_runtime": 1035.1125,
      "eval_samples_per_second": 9.638,
      "eval_steps_per_second": 4.819,
      "step": 15
    },
    {
      "epoch": 0.00037987506331251057,
      "grad_norm": NaN,
      "learning_rate": 0.00018,
      "loss": 0.4772,
      "step": 18
    },
    {
      "epoch": 0.0004220834036805673,
      "eval_loss": NaN,
      "eval_runtime": 1035.0081,
      "eval_samples_per_second": 9.639,
      "eval_steps_per_second": 4.819,
      "step": 20
    },
    {
      "epoch": 0.00044318757386459565,
      "grad_norm": NaN,
      "learning_rate": 0.00019876883405951377,
      "loss": 0.0,
      "step": 21
    },
    {
      "epoch": 0.0005065000844166807,
      "grad_norm": NaN,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.0,
      "step": 24
    },
    {
      "epoch": 0.0005276042546007091,
      "eval_loss": NaN,
      "eval_runtime": 1035.7309,
      "eval_samples_per_second": 9.632,
      "eval_steps_per_second": 4.816,
      "step": 25
    },
    {
      "epoch": 0.0005698125949687658,
      "grad_norm": NaN,
      "learning_rate": 0.00014539904997395468,
      "loss": 0.0,
      "step": 27
    },
    {
      "epoch": 0.000633125105520851,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.0,
      "step": 30
    },
    {
      "epoch": 0.000633125105520851,
      "eval_loss": NaN,
      "eval_runtime": 1035.858,
      "eval_samples_per_second": 9.631,
      "eval_steps_per_second": 4.815,
      "step": 30
    },
    {
      "epoch": 0.000696437616072936,
      "grad_norm": NaN,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 0.0,
      "step": 33
    },
    {
      "epoch": 0.0007386459564409927,
      "eval_loss": NaN,
      "eval_runtime": 1035.6853,
      "eval_samples_per_second": 9.632,
      "eval_steps_per_second": 4.816,
      "step": 35
    },
    {
      "epoch": 0.0007597501266250211,
      "grad_norm": NaN,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 0.0,
      "step": 36
    },
    {
      "epoch": 0.0008230626371771062,
      "grad_norm": NaN,
      "learning_rate": 1.231165940486234e-06,
      "loss": 0.0,
      "step": 39
    },
    {
      "epoch": 0.0008441668073611346,
      "eval_loss": NaN,
      "eval_runtime": 1035.714,
      "eval_samples_per_second": 9.632,
      "eval_steps_per_second": 4.816,
      "step": 40
    }
  ],
  "logging_steps": 3,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9366326954950656.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}