trainer_state.json

{
  "best_metric": 0.8794666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0959/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": 197.4530029296875,
      "learning_rate": 5e-05,
      "loss": 3.6348,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5301333333333333,
      "eval_loss": 2.973487615585327,
      "eval_runtime": 8.6676,
      "eval_samples_per_second": 432.647,
      "eval_steps_per_second": 6.807,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 187.27658081054688,
      "learning_rate": 5e-05,
      "loss": 1.6096,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7749333333333334,
      "eval_loss": 0.8822299242019653,
      "eval_runtime": 7.8261,
      "eval_samples_per_second": 479.165,
      "eval_steps_per_second": 7.539,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 159.5889434814453,
      "learning_rate": 5e-05,
      "loss": 0.6534,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8309333333333333,
      "eval_loss": 0.5880273580551147,
      "eval_runtime": 8.3315,
      "eval_samples_per_second": 450.1,
      "eval_steps_per_second": 7.082,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 127.30839538574219,
      "learning_rate": 5e-05,
      "loss": 0.4457,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8530666666666666,
      "eval_loss": 0.4982427656650543,
      "eval_runtime": 8.0039,
      "eval_samples_per_second": 468.521,
      "eval_steps_per_second": 7.371,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 349.9974365234375,
      "learning_rate": 5e-05,
      "loss": 0.3294,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.4512905776500702,
      "eval_runtime": 9.0216,
      "eval_samples_per_second": 415.671,
      "eval_steps_per_second": 6.54,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 75.62332153320312,
      "learning_rate": 5e-05,
      "loss": 0.2415,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.42536917328834534,
      "eval_runtime": 7.697,
      "eval_samples_per_second": 487.2,
      "eval_steps_per_second": 7.665,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 96.31340026855469,
      "learning_rate": 5e-05,
      "loss": 0.1861,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8794666666666666,
      "eval_loss": 0.4215395152568817,
      "eval_runtime": 7.8821,
      "eval_samples_per_second": 475.762,
      "eval_steps_per_second": 7.485,
      "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
}