trainer_state.json

{
  "best_metric": 0.8784,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0597/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": 71.50728607177734,
      "learning_rate": 6.222222222222222e-05,
      "loss": 3.4288,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5946666666666667,
      "eval_loss": 2.2085440158843994,
      "eval_runtime": 7.7959,
      "eval_samples_per_second": 481.022,
      "eval_steps_per_second": 7.568,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 221.3895263671875,
      "learning_rate": 5.444444444444444e-05,
      "loss": 1.2214,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7994666666666667,
      "eval_loss": 0.7490626573562622,
      "eval_runtime": 7.8766,
      "eval_samples_per_second": 476.092,
      "eval_steps_per_second": 7.491,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 139.1968231201172,
      "learning_rate": 4.666666666666666e-05,
      "loss": 0.677,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8368,
      "eval_loss": 0.5769738554954529,
      "eval_runtime": 8.1744,
      "eval_samples_per_second": 458.75,
      "eval_steps_per_second": 7.218,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 255.44309997558594,
      "learning_rate": 3.8888888888888884e-05,
      "loss": 0.5234,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.49693796038627625,
      "eval_runtime": 8.2548,
      "eval_samples_per_second": 454.281,
      "eval_steps_per_second": 7.147,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 96.42994689941406,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.4352,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.45131856203079224,
      "eval_runtime": 8.6149,
      "eval_samples_per_second": 435.292,
      "eval_steps_per_second": 6.849,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 90.75151062011719,
      "learning_rate": 2.333333333333333e-05,
      "loss": 0.383,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.44937005639076233,
      "eval_runtime": 8.5996,
      "eval_samples_per_second": 436.067,
      "eval_steps_per_second": 6.861,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 135.86485290527344,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.3429,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.44058045744895935,
      "eval_runtime": 7.9861,
      "eval_samples_per_second": 469.563,
      "eval_steps_per_second": 7.388,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 163.18663024902344,
      "learning_rate": 7.777777777777777e-06,
      "loss": 0.3173,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8784,
      "eval_loss": 0.4281339943408966,
      "eval_runtime": 7.8736,
      "eval_samples_per_second": 476.273,
      "eval_steps_per_second": 7.493,
      "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
}