大模型Llama2部署，基于text-generation-webui、Llama2-Chinese

支持windows、linux，显存要求10G左右

一、使用text-generation-webui部署

参考安装教程：傻瓜式！一键部署llama2+chatglm2，集成所有环境和微调功能，本地化界面操作！

Github地址：GitHub - oobabooga/text-generation-webui: A Gradio web UI for Large Language Models. Supports transformers, GPTQ, llama.cpp (ggml/gguf), Llama models.

模型下载地址：meta-llama/Llama-2-13b-chat-hf at main

遇到的问题：

Traceback (most recent call last):

File "D:\oobabooga_windows\text-generation-webui\server.py", line 12, in <module>

import gradio as gr

ModuleNotFoundError: No module named 'gradio'

--配置python环境变量，路径：D:\oobabooga_windows\installer_files\env（自带python环境），使用python -m pip install gradio安装环境

ModuleNotFoundError: No module named 'markdown'

--使用python -m pip install markdown安装环境

--又缺少module，直接install requirements文件

ValueError: When localhost is not accessible, a shareable link must be created. Please set share=True or check your proxy settings to allow access to localhost.

--关闭代理，重新运行

To create a public link, set `share=True` in `launch()`.

二、使用Llama2-Chinese部署

下载FlagAlpha/Llama2-Chinese-13b-Chat模型库：meta-llama/Llama-2-13b-chat-hf at main

cd D:Llama2-Chinese

Llama2-Chinese只支持4bit的数据模型，可正常运行：

python examples/chat_gradio.py --model_name_or_path D:\oobabooga_windows\text-generation-webui\models\Llama-2-7b-chat-hf

python examples/chat_gradio.py --model_name_or_path D:\model\Llama2-Chinese-7b-Chat