trainer_state.json

{
  "best_metric": 0.9149333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0427/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": 27.23236846923828,
      "learning_rate": 0.0004849231551964771,
      "loss": 1.2239,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.48369935154914856,
      "eval_runtime": 15.4191,
      "eval_samples_per_second": 243.205,
      "eval_steps_per_second": 3.826,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.52349090576172,
      "learning_rate": 0.0004415111107797445,
      "loss": 0.3427,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.4092682898044586,
      "eval_runtime": 14.439,
      "eval_samples_per_second": 259.714,
      "eval_steps_per_second": 4.086,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 28.211688995361328,
      "learning_rate": 0.000375,
      "loss": 0.1928,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8928,
      "eval_loss": 0.37289291620254517,
      "eval_runtime": 14.5623,
      "eval_samples_per_second": 257.515,
      "eval_steps_per_second": 4.052,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 23.27015495300293,
      "learning_rate": 0.00029341204441673266,
      "loss": 0.1049,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8928,
      "eval_loss": 0.41024231910705566,
      "eval_runtime": 14.4087,
      "eval_samples_per_second": 260.26,
      "eval_steps_per_second": 4.095,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 28.04730796813965,
      "learning_rate": 0.00020658795558326743,
      "loss": 0.0476,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9088,
      "eval_loss": 0.4005545675754547,
      "eval_runtime": 14.5651,
      "eval_samples_per_second": 257.464,
      "eval_steps_per_second": 4.051,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 30.91367530822754,
      "learning_rate": 0.00012500000000000006,
      "loss": 0.0306,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9072,
      "eval_loss": 0.4021916687488556,
      "eval_runtime": 14.3347,
      "eval_samples_per_second": 261.603,
      "eval_steps_per_second": 4.116,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 29.948665618896484,
      "learning_rate": 5.848888922025553e-05,
      "loss": 0.0123,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.912,
      "eval_loss": 0.41911691427230835,
      "eval_runtime": 14.3437,
      "eval_samples_per_second": 261.439,
      "eval_steps_per_second": 4.113,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 45.57271957397461,
      "learning_rate": 1.5076844803522921e-05,
      "loss": 0.0172,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9085333333333333,
      "eval_loss": 0.4156717360019684,
      "eval_runtime": 14.3233,
      "eval_samples_per_second": 261.811,
      "eval_steps_per_second": 4.119,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 22.00727653503418,
      "learning_rate": 0.0,
      "loss": 0.0108,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9149333333333334,
      "eval_loss": 0.3868911564350128,
      "eval_runtime": 14.3285,
      "eval_samples_per_second": 261.716,
      "eval_steps_per_second": 4.118,
      "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
}