trainer_state.json

{
  "best_metric": 0.8730666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0991/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 30.986352920532227,
      "learning_rate": 0.0002851453301853628,
      "loss": 1.72,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7970666666666667,
      "eval_loss": 0.6638239026069641,
      "eval_runtime": 8.3471,
      "eval_samples_per_second": 449.259,
      "eval_steps_per_second": 7.068,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 34.668697357177734,
      "learning_rate": 0.00024352347027881003,
      "loss": 0.5333,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8317333333333333,
      "eval_loss": 0.5198476910591125,
      "eval_runtime": 8.709,
      "eval_samples_per_second": 430.589,
      "eval_steps_per_second": 6.775,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 36.814971923828125,
      "learning_rate": 0.00018337814009344714,
      "loss": 0.3078,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8504,
      "eval_loss": 0.4959512948989868,
      "eval_runtime": 9.2495,
      "eval_samples_per_second": 405.428,
      "eval_steps_per_second": 6.379,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 29.15039825439453,
      "learning_rate": 0.00011662185990655284,
      "loss": 0.1856,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.4874972105026245,
      "eval_runtime": 8.0464,
      "eval_samples_per_second": 466.044,
      "eval_steps_per_second": 7.332,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 37.43810272216797,
      "learning_rate": 5.6476529721189974e-05,
      "loss": 0.114,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.868,
      "eval_loss": 0.4644485116004944,
      "eval_runtime": 8.0146,
      "eval_samples_per_second": 467.896,
      "eval_steps_per_second": 7.362,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 36.63550567626953,
      "learning_rate": 1.4854669814637143e-05,
      "loss": 0.0782,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.47856101393699646,
      "eval_runtime": 7.9887,
      "eval_samples_per_second": 469.411,
      "eval_steps_per_second": 7.385,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 41.3846549987793,
      "learning_rate": 0.0,
      "loss": 0.0624,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.48232388496398926,
      "eval_runtime": 8.0534,
      "eval_samples_per_second": 465.641,
      "eval_steps_per_second": 7.326,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}