trainer_state.json

{
  "best_metric": 3.7005422115325928,
  "best_model_checkpoint": "./youtube/context/checkpoint-500",
  "epoch": 2.5915127955944284,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6478781988986071,
      "grad_norm": 3.625,
      "learning_rate": 4.4453199780850116e-05,
      "loss": 0.7716,
      "step": 500
    },
    {
      "epoch": 0.6478781988986071,
      "eval_accuracy": 0.4547751731606615,
      "eval_loss": 3.7005422115325928,
      "eval_runtime": 3.955,
      "eval_samples_per_second": 88.242,
      "eval_steps_per_second": 2.276,
      "step": 500
    },
    {
      "epoch": 1.2957563977972142,
      "grad_norm": 3.25,
      "learning_rate": 3.027415853709337e-05,
      "loss": 0.3675,
      "step": 1000
    },
    {
      "epoch": 1.2957563977972142,
      "eval_accuracy": 0.4565124201881642,
      "eval_loss": 3.8486993312835693,
      "eval_runtime": 2.4115,
      "eval_samples_per_second": 144.721,
      "eval_steps_per_second": 3.732,
      "step": 1000
    },
    {
      "epoch": 1.943634596695821,
      "grad_norm": 2.484375,
      "learning_rate": 1.3754740994986162e-05,
      "loss": 0.2102,
      "step": 1500
    },
    {
      "epoch": 1.943634596695821,
      "eval_accuracy": 0.45676059833495025,
      "eval_loss": 3.970470905303955,
      "eval_runtime": 2.3067,
      "eval_samples_per_second": 151.298,
      "eval_steps_per_second": 3.902,
      "step": 1500
    },
    {
      "epoch": 2.5915127955944284,
      "grad_norm": 5.53125,
      "learning_rate": 2.225339861951592e-06,
      "loss": 0.1164,
      "step": 2000
    },
    {
      "epoch": 2.5915127955944284,
      "eval_accuracy": 0.4569636531823207,
      "eval_loss": 4.1545634269714355,
      "eval_runtime": 2.3792,
      "eval_samples_per_second": 146.69,
      "eval_steps_per_second": 3.783,
      "step": 2000
    },
    {
      "epoch": 2.5915127955944284,
      "step": 2000,
      "total_flos": 4.747627355386675e+16,
      "train_loss": 0.366422456741333,
      "train_runtime": 2672.9336,
      "train_samples_per_second": 3.465,
      "train_steps_per_second": 0.865
    }
  ],
  "logging_steps": 500,
  "max_steps": 2313,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.747627355386675e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}