last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.2432432432432434,
  "eval_steps": 20,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02702702702702703,
      "eval_loss": 3.1186463832855225,
      "eval_runtime": 21.1695,
      "eval_samples_per_second": 70.904,
      "eval_steps_per_second": 3.59,
      "step": 1
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 6.0,
      "learning_rate": 8e-05,
      "loss": 2.1439,
      "step": 10
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.84375,
      "learning_rate": 0.00016,
      "loss": 1.9251,
      "step": 20
    },
    {
      "epoch": 0.5405405405405406,
      "eval_loss": 1.7902483940124512,
      "eval_runtime": 21.3533,
      "eval_samples_per_second": 70.294,
      "eval_steps_per_second": 3.559,
      "step": 20
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 1.515625,
      "learning_rate": 0.00019994532573409262,
      "loss": 1.9936,
      "step": 30
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 4.625,
      "learning_rate": 0.00019950829025450114,
      "loss": 1.8595,
      "step": 40
    },
    {
      "epoch": 1.0810810810810811,
      "eval_loss": 2.112858772277832,
      "eval_runtime": 21.4411,
      "eval_samples_per_second": 70.006,
      "eval_steps_per_second": 3.545,
      "step": 40
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 1.625,
      "learning_rate": 0.00019863613034027224,
      "loss": 1.3332,
      "step": 50
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0001973326597248006,
      "loss": 1.343,
      "step": 60
    },
    {
      "epoch": 1.6216216216216215,
      "eval_loss": 2.093151092529297,
      "eval_runtime": 22.3251,
      "eval_samples_per_second": 67.234,
      "eval_steps_per_second": 3.404,
      "step": 60
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 2.09375,
      "learning_rate": 0.00019560357815343577,
      "loss": 1.3674,
      "step": 70
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0001934564464599461,
      "loss": 0.9356,
      "step": 80
    },
    {
      "epoch": 2.1621621621621623,
      "eval_loss": 2.402878999710083,
      "eval_runtime": 21.51,
      "eval_samples_per_second": 69.782,
      "eval_steps_per_second": 3.533,
      "step": 80
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 33.0,
      "learning_rate": 0.00019090065350491626,
      "loss": 1.0747,
      "step": 90
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 1.40625,
      "learning_rate": 0.0001879473751206489,
      "loss": 0.6893,
      "step": 100
    },
    {
      "epoch": 2.7027027027027026,
      "eval_loss": 2.523198127746582,
      "eval_runtime": 21.2137,
      "eval_samples_per_second": 70.756,
      "eval_steps_per_second": 3.583,
      "step": 100
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 1.546875,
      "learning_rate": 0.00018460952524209355,
      "loss": 0.626,
      "step": 110
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 2.0625,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.3162,
      "step": 120
    },
    {
      "epoch": 3.2432432432432434,
      "eval_loss": 2.800004005432129,
      "eval_runtime": 21.4262,
      "eval_samples_per_second": 70.054,
      "eval_steps_per_second": 3.547,
      "step": 120
    }
  ],
  "logging_steps": 10,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 14,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.183140630029926e+16,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}