trainer_state.json

{
  "best_metric": 0.8666666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0437/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": 85.53441619873047,
      "learning_rate": 3e-05,
      "loss": 3.8564,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.32293333333333335,
      "eval_loss": 3.747537851333618,
      "eval_runtime": 7.5935,
      "eval_samples_per_second": 493.843,
      "eval_steps_per_second": 7.77,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 173.04058837890625,
      "learning_rate": 3e-05,
      "loss": 3.1467,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5632,
      "eval_loss": 2.251485586166382,
      "eval_runtime": 7.613,
      "eval_samples_per_second": 492.578,
      "eval_steps_per_second": 7.75,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 275.196044921875,
      "learning_rate": 3e-05,
      "loss": 1.5929,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7162666666666667,
      "eval_loss": 1.1360286474227905,
      "eval_runtime": 7.9747,
      "eval_samples_per_second": 470.238,
      "eval_steps_per_second": 7.398,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 169.09400939941406,
      "learning_rate": 3e-05,
      "loss": 0.9567,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7786666666666666,
      "eval_loss": 0.7654730677604675,
      "eval_runtime": 7.569,
      "eval_samples_per_second": 495.44,
      "eval_steps_per_second": 7.795,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 415.43585205078125,
      "learning_rate": 3e-05,
      "loss": 0.7278,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8136,
      "eval_loss": 0.6191505193710327,
      "eval_runtime": 8.2029,
      "eval_samples_per_second": 457.158,
      "eval_steps_per_second": 7.193,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 188.81103515625,
      "learning_rate": 3e-05,
      "loss": 0.5949,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8232,
      "eval_loss": 0.5625346302986145,
      "eval_runtime": 7.7,
      "eval_samples_per_second": 487.015,
      "eval_steps_per_second": 7.662,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 256.2063293457031,
      "learning_rate": 3e-05,
      "loss": 0.5059,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.48416635394096375,
      "eval_runtime": 7.6377,
      "eval_samples_per_second": 490.988,
      "eval_steps_per_second": 7.725,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 415.6106872558594,
      "learning_rate": 3e-05,
      "loss": 0.4579,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8576,
      "eval_loss": 0.4530128240585327,
      "eval_runtime": 7.6789,
      "eval_samples_per_second": 488.351,
      "eval_steps_per_second": 7.683,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 184.34814453125,
      "learning_rate": 3e-05,
      "loss": 0.4016,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.43871086835861206,
      "eval_runtime": 7.9982,
      "eval_samples_per_second": 468.858,
      "eval_steps_per_second": 7.377,
      "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
}