trainer_state.json

{
  "best_metric": 0.8562666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0256/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": 43.738670349121094,
      "learning_rate": 7.875e-05,
      "loss": 3.1369,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6528,
      "eval_loss": 1.4811153411865234,
      "eval_runtime": 9.1817,
      "eval_samples_per_second": 408.419,
      "eval_steps_per_second": 6.426,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 232.1677703857422,
      "learning_rate": 6.75e-05,
      "loss": 0.9828,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7866666666666666,
      "eval_loss": 0.7422832250595093,
      "eval_runtime": 8.344,
      "eval_samples_per_second": 449.426,
      "eval_steps_per_second": 7.071,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 110.77406311035156,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 0.6319,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8114666666666667,
      "eval_loss": 0.6047019958496094,
      "eval_runtime": 8.222,
      "eval_samples_per_second": 456.095,
      "eval_steps_per_second": 7.176,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 225.6870574951172,
      "learning_rate": 4.5e-05,
      "loss": 0.4869,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8346666666666667,
      "eval_loss": 0.5326733589172363,
      "eval_runtime": 8.494,
      "eval_samples_per_second": 441.49,
      "eval_steps_per_second": 6.946,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 115.34894561767578,
      "learning_rate": 3.375e-05,
      "loss": 0.3996,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8461333333333333,
      "eval_loss": 0.4901898503303528,
      "eval_runtime": 8.3808,
      "eval_samples_per_second": 447.451,
      "eval_steps_per_second": 7.04,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 78.4578857421875,
      "learning_rate": 2.25e-05,
      "loss": 0.3521,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8450666666666666,
      "eval_loss": 0.5104067325592041,
      "eval_runtime": 8.4158,
      "eval_samples_per_second": 445.588,
      "eval_steps_per_second": 7.011,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 78.55250549316406,
      "learning_rate": 1.125e-05,
      "loss": 0.321,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8562666666666666,
      "eval_loss": 0.4825674593448639,
      "eval_runtime": 8.6183,
      "eval_samples_per_second": 435.12,
      "eval_steps_per_second": 6.846,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}