trainer_state.json

{
  "best_metric": 0.8674666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0610/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 71.65089416503906,
      "learning_rate": 3e-05,
      "loss": 3.8611,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2728,
      "eval_loss": 3.76068115234375,
      "eval_runtime": 15.4087,
      "eval_samples_per_second": 243.369,
      "eval_steps_per_second": 3.829,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 127.42249298095703,
      "learning_rate": 3e-05,
      "loss": 3.3175,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5133333333333333,
      "eval_loss": 2.658383846282959,
      "eval_runtime": 15.4493,
      "eval_samples_per_second": 242.73,
      "eval_steps_per_second": 3.819,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 193.4296875,
      "learning_rate": 3e-05,
      "loss": 1.7499,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7186666666666667,
      "eval_loss": 1.1630604267120361,
      "eval_runtime": 15.4855,
      "eval_samples_per_second": 242.163,
      "eval_steps_per_second": 3.81,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 326.1895446777344,
      "learning_rate": 3e-05,
      "loss": 0.9577,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7864,
      "eval_loss": 0.8000806570053101,
      "eval_runtime": 15.5737,
      "eval_samples_per_second": 240.791,
      "eval_steps_per_second": 3.788,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 440.9952392578125,
      "learning_rate": 3e-05,
      "loss": 0.7037,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8117333333333333,
      "eval_loss": 0.6625063419342041,
      "eval_runtime": 15.4575,
      "eval_samples_per_second": 242.6,
      "eval_steps_per_second": 3.817,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 148.85496520996094,
      "learning_rate": 3e-05,
      "loss": 0.5743,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8296,
      "eval_loss": 0.5775328278541565,
      "eval_runtime": 14.743,
      "eval_samples_per_second": 254.357,
      "eval_steps_per_second": 4.002,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 317.4260559082031,
      "learning_rate": 3e-05,
      "loss": 0.4972,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.5170561075210571,
      "eval_runtime": 15.4806,
      "eval_samples_per_second": 242.239,
      "eval_steps_per_second": 3.811,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 162.48443603515625,
      "learning_rate": 3e-05,
      "loss": 0.4285,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8533333333333334,
      "eval_loss": 0.4888891279697418,
      "eval_runtime": 15.4171,
      "eval_samples_per_second": 243.236,
      "eval_steps_per_second": 3.827,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 219.98931884765625,
      "learning_rate": 3e-05,
      "loss": 0.3825,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.45401498675346375,
      "eval_runtime": 15.635,
      "eval_samples_per_second": 239.847,
      "eval_steps_per_second": 3.774,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.3587863155968e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}