nluai/test_format

1{2  "_name_or_path": "vilm/vinallama-7b-chat",3  "architectures": [4    "LlamaForCausalLM"5  ],6  "attention_bias": false,7  "attention_dropout": 0.0,8  "bos_token_id": 1,9  "eos_token_id": 46303,10  "hidden_act": "silu",11  "hidden_size": 4096,12  "initializer_range": 0.02,13  "intermediate_size": 11008,14  "max_position_embeddings": 4096,15  "mlp_bias": false,16  "model_type": "llama",17  "num_attention_heads": 32,18  "num_hidden_layers": 32,19  "num_key_value_heads": 32,20  "pad_token_id": 0,21  "pretraining_tp": 1,22  "quantization_config": {23    "_load_in_4bit": true,24    "_load_in_8bit": false,25    "bnb_4bit_compute_dtype": "float16",26    "bnb_4bit_quant_storage": "uint8",27    "bnb_4bit_quant_type": "nf4",28    "bnb_4bit_use_double_quant": true,29    "llm_int8_enable_fp32_cpu_offload": false,30    "llm_int8_has_fp16_weight": false,31    "llm_int8_skip_modules": null,32    "llm_int8_threshold": 6.0,33    "load_in_4bit": true,34    "load_in_8bit": false,35    "quant_method": "bitsandbytes"36  },37  "rms_norm_eps": 1e-05,38  "rope_scaling": null,39  "rope_theta": 10000.0,40  "tie_word_embeddings": false,41  "torch_dtype": "float32",42  "transformers_version": "4.43.3",43  "use_cache": false,44  "vocab_size": 4630545}46