trainer_state.json

{
  "best_metric": 0.9117333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0628/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": 39.679439544677734,
      "learning_rate": 0.00042857142857142855,
      "loss": 1.3685,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8237333333333333,
      "eval_loss": 0.6076481342315674,
      "eval_runtime": 5.274,
      "eval_samples_per_second": 711.035,
      "eval_steps_per_second": 11.187,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 27.42409896850586,
      "learning_rate": 0.00035714285714285714,
      "loss": 0.4285,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8754666666666666,
      "eval_loss": 0.4362938702106476,
      "eval_runtime": 5.2489,
      "eval_samples_per_second": 714.438,
      "eval_steps_per_second": 11.24,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.513763427734375,
      "learning_rate": 0.0002857142857142857,
      "loss": 0.238,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8906666666666667,
      "eval_loss": 0.3920813500881195,
      "eval_runtime": 5.2543,
      "eval_samples_per_second": 713.7,
      "eval_steps_per_second": 11.229,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 15.094958305358887,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.138,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9002666666666667,
      "eval_loss": 0.3913276791572571,
      "eval_runtime": 5.2698,
      "eval_samples_per_second": 711.603,
      "eval_steps_per_second": 11.196,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.129653930664062,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.0758,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8989333333333334,
      "eval_loss": 0.4001148045063019,
      "eval_runtime": 5.2667,
      "eval_samples_per_second": 712.019,
      "eval_steps_per_second": 11.202,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 17.215864181518555,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.0423,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9074666666666666,
      "eval_loss": 0.3900512456893921,
      "eval_runtime": 5.2348,
      "eval_samples_per_second": 716.362,
      "eval_steps_per_second": 11.271,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 26.606304168701172,
      "learning_rate": 0.0,
      "loss": 0.0299,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9117333333333333,
      "eval_loss": 0.37934979796409607,
      "eval_runtime": 5.2505,
      "eval_samples_per_second": 714.219,
      "eval_steps_per_second": 11.237,
      "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
}