trainer_state.json

{
  "best_metric": 0.8834666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0615/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": 96.22808837890625,
      "learning_rate": 5e-05,
      "loss": 3.7039,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4896,
      "eval_loss": 3.1459245681762695,
      "eval_runtime": 14.7616,
      "eval_samples_per_second": 254.037,
      "eval_steps_per_second": 3.997,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 57.610904693603516,
      "learning_rate": 5e-05,
      "loss": 1.809,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7522666666666666,
      "eval_loss": 0.9642399549484253,
      "eval_runtime": 14.4469,
      "eval_samples_per_second": 259.571,
      "eval_steps_per_second": 4.084,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 116.99556732177734,
      "learning_rate": 5e-05,
      "loss": 0.7964,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.824,
      "eval_loss": 0.6303656101226807,
      "eval_runtime": 14.4329,
      "eval_samples_per_second": 259.823,
      "eval_steps_per_second": 4.088,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 42.228004455566406,
      "learning_rate": 5e-05,
      "loss": 0.5661,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5027812123298645,
      "eval_runtime": 14.5014,
      "eval_samples_per_second": 258.595,
      "eval_steps_per_second": 4.069,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 68.7126693725586,
      "learning_rate": 5e-05,
      "loss": 0.4283,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8741333333333333,
      "eval_loss": 0.41580426692962646,
      "eval_runtime": 14.5612,
      "eval_samples_per_second": 257.533,
      "eval_steps_per_second": 4.052,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 207.101318359375,
      "learning_rate": 5e-05,
      "loss": 0.3704,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8736,
      "eval_loss": 0.4041142463684082,
      "eval_runtime": 14.4493,
      "eval_samples_per_second": 259.528,
      "eval_steps_per_second": 4.083,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 55.651214599609375,
      "learning_rate": 5e-05,
      "loss": 0.2976,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8834666666666666,
      "eval_loss": 0.38896965980529785,
      "eval_runtime": 14.5341,
      "eval_samples_per_second": 258.013,
      "eval_steps_per_second": 4.059,
      "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
}