last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.025506950644050504,
  "eval_steps": 10,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005101390128810101,
      "eval_loss": NaN,
      "eval_runtime": 84.0417,
      "eval_samples_per_second": 9.828,
      "eval_steps_per_second": 4.914,
      "step": 1
    },
    {
      "epoch": 0.0025506950644050503,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.0,
      "step": 5
    },
    {
      "epoch": 0.0051013901288101005,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 0.5415,
      "step": 10
    },
    {
      "epoch": 0.0051013901288101005,
      "eval_loss": NaN,
      "eval_runtime": 83.2748,
      "eval_samples_per_second": 9.919,
      "eval_steps_per_second": 4.959,
      "step": 10
    },
    {
      "epoch": 0.007652085193215151,
      "grad_norm": NaN,
      "learning_rate": 9.619397662556435e-05,
      "loss": 1.0469,
      "step": 15
    },
    {
      "epoch": 0.010202780257620201,
      "grad_norm": NaN,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.0,
      "step": 20
    },
    {
      "epoch": 0.010202780257620201,
      "eval_loss": NaN,
      "eval_runtime": 83.2813,
      "eval_samples_per_second": 9.918,
      "eval_steps_per_second": 4.959,
      "step": 20
    },
    {
      "epoch": 0.012753475322025252,
      "grad_norm": NaN,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0,
      "step": 25
    },
    {
      "epoch": 0.015304170386430302,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 13.312,
      "step": 30
    },
    {
      "epoch": 0.015304170386430302,
      "eval_loss": NaN,
      "eval_runtime": 83.3752,
      "eval_samples_per_second": 9.907,
      "eval_steps_per_second": 4.954,
      "step": 30
    },
    {
      "epoch": 0.017854865450835353,
      "grad_norm": NaN,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0,
      "step": 35
    },
    {
      "epoch": 0.020405560515240402,
      "grad_norm": NaN,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 7.5202,
      "step": 40
    },
    {
      "epoch": 0.020405560515240402,
      "eval_loss": NaN,
      "eval_runtime": 83.3864,
      "eval_samples_per_second": 9.906,
      "eval_steps_per_second": 4.953,
      "step": 40
    },
    {
      "epoch": 0.022956255579645455,
      "grad_norm": NaN,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 47.307,
      "step": 45
    },
    {
      "epoch": 0.025506950644050504,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 0.0,
      "step": 50
    },
    {
      "epoch": 0.025506950644050504,
      "eval_loss": NaN,
      "eval_runtime": 83.2927,
      "eval_samples_per_second": 9.917,
      "eval_steps_per_second": 4.958,
      "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": 1.546589490905088e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}