https://docs.llamaindex.ai/en/stable/module_guides/models/llms.html
可以按参数和评分来选择模型：
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
llama2-chat-7b 4bit (huggingface)
Llama2 seems to be quite chatty, which makes parsing structured outputs difficult. Fine-tuning and prompt engineering likely required for better performance on structured outputs.
Mistral-7B-instruct-v0.1 4bit (huggingface)
Mistral seems slightly more reliable for structured outputs compared to Llama2. Likely with some prompt engineering, it may do better.