{
  "architectures": [
    "Gemma2ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 2,
  "eos_token_id": 1,
  "head_dim": 256,
  "hidden_act": "gelu_pytorch_tanh",
  "hidden_size": 2304,
  "initializer_range": 0.02,
  "intermediate_size": 9216,
  "max_position_embeddings": 8192,
  "model_type": "gemma2",
  "num_attention_heads": 8,
  "num_hidden_layers": 36,
  "num_key_value_heads": 4,
  "pad_token_id": 0,
  "prefix_lm": false,
  "quantization_config": {
    "quant_method": "llama",
    "quantization_scheme": "llama",
    "llama_quantization_version": 2
  },
  "rms_norm_eps": 1e-06,
  "rope_theta": 10000.0,
  "sliding_window": 4096,
  "sliding_window_pattern": 4,
  "tie_word_embeddings": true,
  "vocab_size": 256000
}