trainer_state.json

{
  "best_metric": 0.9014778325123153,
  "best_model_checkpoint": "dinov2-small-imagenet1k-1-layer-finetuned-noh/checkpoint-66",
  "epoch": 2.9333333333333336,
  "eval_steps": 500,
  "global_step": 66,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 223.80079650878906,
      "learning_rate": 4.745762711864407e-05,
      "loss": 0.8188,
      "step": 10
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 90.21881103515625,
      "learning_rate": 3.898305084745763e-05,
      "loss": 0.6437,
      "step": 20
    },
    {
      "epoch": 0.9777777777777777,
      "eval_accuracy": 0.7619047619047619,
      "eval_loss": 0.47078460454940796,
      "eval_runtime": 425.2287,
      "eval_samples_per_second": 1.432,
      "eval_steps_per_second": 0.092,
      "step": 22
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 154.34889221191406,
      "learning_rate": 3.050847457627119e-05,
      "loss": 0.6102,
      "step": 30
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 38.342926025390625,
      "learning_rate": 2.2033898305084748e-05,
      "loss": 0.5057,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8850574712643678,
      "eval_loss": 0.33653634786605835,
      "eval_runtime": 178.3005,
      "eval_samples_per_second": 3.416,
      "eval_steps_per_second": 0.219,
      "step": 45
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 36.97127914428711,
      "learning_rate": 1.3559322033898305e-05,
      "loss": 0.4065,
      "step": 50
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 55.42195129394531,
      "learning_rate": 5.084745762711865e-06,
      "loss": 0.4411,
      "step": 60
    },
    {
      "epoch": 2.9333333333333336,
      "eval_accuracy": 0.9014778325123153,
      "eval_loss": 0.3210342824459076,
      "eval_runtime": 175.5054,
      "eval_samples_per_second": 3.47,
      "eval_steps_per_second": 0.222,
      "step": 66
    },
    {
      "epoch": 2.9333333333333336,
      "step": 66,
      "total_flos": 8.411145538115174e+16,
      "train_loss": 0.5577786402268843,
      "train_runtime": 2627.8256,
      "train_samples_per_second": 1.643,
      "train_steps_per_second": 0.025
    }
  ],
  "logging_steps": 10,
  "max_steps": 66,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 8.411145538115174e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}