checkpoint-1424/trainer_state.json

{
  "best_global_step": 1424,
  "best_metric": 0.23910000920295715,
  "best_model_checkpoint": "siglip2-finetune-full/checkpoint-1424",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1424,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.702247191011236,
      "grad_norm": 16.442720413208008,
      "learning_rate": 1.3449781659388646e-06,
      "loss": 0.6162,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9363852034091907,
      "eval_loss": 0.2713288962841034,
      "eval_model_preparation_time": 0.0023,
      "eval_runtime": 335.7294,
      "eval_samples_per_second": 67.799,
      "eval_steps_per_second": 8.477,
      "step": 712
    },
    {
      "epoch": 1.404494382022472,
      "grad_norm": 16.62710189819336,
      "learning_rate": 6.171761280931587e-07,
      "loss": 0.519,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9426676039012389,
      "eval_loss": 0.23910000920295715,
      "eval_model_preparation_time": 0.0023,
      "eval_runtime": 338.2117,
      "eval_samples_per_second": 67.301,
      "eval_steps_per_second": 8.415,
      "step": 1424
    }
  ],
  "logging_steps": 500,
  "max_steps": 1424,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 3.8128851917694075e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}