last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.028303555634176542,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0028303555634176542,
      "grad_norm": 0.5802844166755676,
      "learning_rate": 1e-05,
      "loss": 0.6877,
      "step": 1
    },
    {
      "epoch": 0.0028303555634176542,
      "eval_loss": NaN,
      "eval_runtime": 130.2723,
      "eval_samples_per_second": 4.567,
      "eval_steps_per_second": 2.288,
      "step": 1
    },
    {
      "epoch": 0.0056607111268353084,
      "grad_norm": 0.7363430857658386,
      "learning_rate": 2e-05,
      "loss": 0.826,
      "step": 2
    },
    {
      "epoch": 0.008491066690252963,
      "grad_norm": 0.730840802192688,
      "learning_rate": 3e-05,
      "loss": 0.8343,
      "step": 3
    },
    {
      "epoch": 0.008491066690252963,
      "eval_loss": NaN,
      "eval_runtime": 112.3005,
      "eval_samples_per_second": 5.298,
      "eval_steps_per_second": 2.654,
      "step": 3
    },
    {
      "epoch": 0.011321422253670617,
      "grad_norm": 0.6844142079353333,
      "learning_rate": 4e-05,
      "loss": 0.7993,
      "step": 4
    },
    {
      "epoch": 0.014151777817088271,
      "grad_norm": 0.6910650134086609,
      "learning_rate": 5e-05,
      "loss": 0.8568,
      "step": 5
    },
    {
      "epoch": 0.016982133380505925,
      "grad_norm": 0.5070165991783142,
      "learning_rate": 6e-05,
      "loss": 0.6597,
      "step": 6
    },
    {
      "epoch": 0.016982133380505925,
      "eval_loss": NaN,
      "eval_runtime": 112.4386,
      "eval_samples_per_second": 5.292,
      "eval_steps_per_second": 2.65,
      "step": 6
    },
    {
      "epoch": 0.01981248894392358,
      "grad_norm": 0.5357987284660339,
      "learning_rate": 7e-05,
      "loss": 0.7278,
      "step": 7
    },
    {
      "epoch": 0.022642844507341234,
      "grad_norm": 0.4978576898574829,
      "learning_rate": 8e-05,
      "loss": 0.6961,
      "step": 8
    },
    {
      "epoch": 0.02547320007075889,
      "grad_norm": 0.4354259967803955,
      "learning_rate": 9e-05,
      "loss": 0.7407,
      "step": 9
    },
    {
      "epoch": 0.02547320007075889,
      "eval_loss": NaN,
      "eval_runtime": 112.6245,
      "eval_samples_per_second": 5.283,
      "eval_steps_per_second": 2.646,
      "step": 9
    },
    {
      "epoch": 0.028303555634176542,
      "grad_norm": 0.39359647035598755,
      "learning_rate": 0.0001,
      "loss": 0.6108,
      "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": 6.588900336205824e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}