trainer_state.json

{
  "best_metric": 0.8914666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0220/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 144.86192321777344,
      "learning_rate": 6.222222222222222e-05,
      "loss": 3.3321,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6632,
      "eval_loss": 1.9351942539215088,
      "eval_runtime": 7.3342,
      "eval_samples_per_second": 511.3,
      "eval_steps_per_second": 8.044,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 240.2034454345703,
      "learning_rate": 5.444444444444444e-05,
      "loss": 1.0365,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8224,
      "eval_loss": 0.6331597566604614,
      "eval_runtime": 7.2668,
      "eval_samples_per_second": 516.044,
      "eval_steps_per_second": 8.119,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 120.08686065673828,
      "learning_rate": 4.666666666666666e-05,
      "loss": 0.5366,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8616,
      "eval_loss": 0.45949846506118774,
      "eval_runtime": 7.6892,
      "eval_samples_per_second": 487.695,
      "eval_steps_per_second": 7.673,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 118.92008209228516,
      "learning_rate": 3.8888888888888884e-05,
      "loss": 0.3698,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.4227418899536133,
      "eval_runtime": 7.705,
      "eval_samples_per_second": 486.695,
      "eval_steps_per_second": 7.657,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 73.51454162597656,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.2757,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8813333333333333,
      "eval_loss": 0.375572144985199,
      "eval_runtime": 7.8254,
      "eval_samples_per_second": 479.211,
      "eval_steps_per_second": 7.54,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 224.3568572998047,
      "learning_rate": 2.333333333333333e-05,
      "loss": 0.222,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.3762975335121155,
      "eval_runtime": 7.8971,
      "eval_samples_per_second": 474.86,
      "eval_steps_per_second": 7.471,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 108.19109344482422,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.1803,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8842666666666666,
      "eval_loss": 0.37285175919532776,
      "eval_runtime": 7.9865,
      "eval_samples_per_second": 469.545,
      "eval_steps_per_second": 7.388,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 252.81727600097656,
      "learning_rate": 7.777777777777777e-06,
      "loss": 0.1557,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.37149932980537415,
      "eval_runtime": 7.8659,
      "eval_samples_per_second": 476.739,
      "eval_steps_per_second": 7.501,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 6.5411433916416e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}