{
  "model_config": {
    "n_layer": 24,
    "n_head": 16,
    "n_embd": 1024,
    "block_size": 1024,
    "bias": false,
    "dropout": 0.0,
    "input_vocab_size": 12096,
    "output_vocab_size": 12096
  },
  "model_type": "coarse",
  "parameter_count": 327861248,
  "needs_tokenizer": true,
  "best_val_loss": 2.9014446608225506,
  "torch_dtype": "bfloat16"
}