{
  "alibi": false,
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 128000,
  "dtype": "float16",
  "eos_token_id": 128001,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "immediate_clear_past_key_value": false,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "long_sequence_init_args": {},
  "long_sequence_strategy_name": null,
  "long_sequence_strategy_type": null,
  "max_position_embeddings": 8192,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "pad_token_id": 0,
  "paddlenlp_version": "3.0.0b1.post20240906",
  "pretraining_tp": 1,
  "quantization_config": {
    "quant_type": "a8w8_fp8"
  },
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_scaling_factor": 1.0,
  "rope_scaling_type": null,
  "rope_theta": 500000.0,
  "seq_length": 2048,
  "tensor_parallel_degree": -1,
  "tie_word_embeddings": false,
  "use_fast_layer_norm": false,
  "use_flash_attention_for_generation": false,
  "use_last_token_for_generation": false,
  "use_long_sequence_strategies": false,
  "vocab_size": 128256
}