{
  "architectures": [
    "Qwen2ForCausalLM"
  ],
  "model_type": "qwen2",
  "vocab_size": 8000,
  "hidden_size": 256,
  "intermediate_size": 684,
  "num_hidden_layers": 8,
  "num_attention_heads": 8,
  "num_key_value_heads": 2,
  "max_position_embeddings": 2048,
  "rope_theta": 10000.0,
  "rms_norm_eps": 1e-06,
  "tie_word_embeddings": true,
  "torch_dtype": "float32",
  "transformers_version": "4.35.0",
  "use_cache": false
}