trainer_state.json

{
  "best_metric": 0.8952,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0953/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": 120.22265625,
      "learning_rate": 8.75e-05,
      "loss": 2.7673,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7578666666666667,
      "eval_loss": 0.9962217211723328,
      "eval_runtime": 7.7773,
      "eval_samples_per_second": 482.17,
      "eval_steps_per_second": 7.586,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 94.29248046875,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.7203,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8418666666666667,
      "eval_loss": 0.5253738164901733,
      "eval_runtime": 7.7176,
      "eval_samples_per_second": 485.9,
      "eval_steps_per_second": 7.645,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 46.7548942565918,
      "learning_rate": 6.25e-05,
      "loss": 0.4437,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8696,
      "eval_loss": 0.4451506435871124,
      "eval_runtime": 7.6892,
      "eval_samples_per_second": 487.697,
      "eval_steps_per_second": 7.673,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 189.05233764648438,
      "learning_rate": 5e-05,
      "loss": 0.3374,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.4018511474132538,
      "eval_runtime": 7.7355,
      "eval_samples_per_second": 484.777,
      "eval_steps_per_second": 7.627,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 68.28330993652344,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.2556,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.39204341173171997,
      "eval_runtime": 8.2562,
      "eval_samples_per_second": 454.205,
      "eval_steps_per_second": 7.146,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 77.26693725585938,
      "learning_rate": 2.5e-05,
      "loss": 0.2024,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8848,
      "eval_loss": 0.38335612416267395,
      "eval_runtime": 8.2858,
      "eval_samples_per_second": 452.581,
      "eval_steps_per_second": 7.121,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 55.462345123291016,
      "learning_rate": 1.25e-05,
      "loss": 0.1707,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8952,
      "eval_loss": 0.3578214943408966,
      "eval_runtime": 7.7688,
      "eval_samples_per_second": 482.698,
      "eval_steps_per_second": 7.594,
      "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
}