{
  "iteration": 200000,
  "best_val_loss": 5.296863555908203,
  "early_stopping_counter": 17,
  "learning_rate": 6e-06,
  "batch_size": 12,
  "block_size": 1024,
  "model_args": {
    "n_layer": 12,
    "n_head": 12,
    "n_embd": 768,
    "block_size": 1024,
    "bias": false,
    "vocab_size": 4096,
    "dropout": 0.0
  }
}