源码地址

https://github.com/QwenLM/Qwen

说明文档

https://github.com/QwenLM/Qwen/blob/main/README_CN.md

Qwen本地部署

• 项目介绍：Qwen 是一个全能的语言模型系列，包含各种参数量的模型，如 Qwen（基础预训练语言模型，即基座模型）和 Qwen-Chat（聊天模型，该模型采用人类对齐技术进行微调）。项目地址：QwenLM/Qwen: The official repo of Qwen (通义千问) chat & pretrained large language model proposed by Alibaba Cloud. (github.com)
• 项目环境要求：

python 3.8 and above

pytorch 1.12 and above, 2.0 and above are recommended

transformers 4.32 and above

CUDA 11.4 and above are recommended (this is for GPU users, flash-attention users, etc.)

• 准备环境与工具（本机）：python3.10、 CUDA12.2、 pycharm（可能需要）
• 本地部署

打开Anaconda prompt

1. 虚拟环境创建：conda create -n qwen python=3.10
2. 激活虚拟环境：conda activate qwen
3. 进入项目所在文件夹，如：

D:

D:\liu\Project\Qwen-main(进入上述网址将项目下载到本地)

1. 安装对应版本的pytorch：pip install torch==2.2.0 torchvision==0.17.0 torchaudio==2.2.0 --index-url https://download.pytorch.org/whl/cu121（与cuda对应）
2. 安装依赖包：pip install -r requirements.txt -i http://mirrors.aliyun.com/pypi/simple --trusted-host mirrors.aliyun.com（阿里源加速下载）
3. 输入pip install modelscope安装魔塔依赖包
4. 在本地创建一个python文件，如main.py

1. 在文件中输入以下代码将模型下载至本地（默认下载于C盘.Cache文件夹中）

from modelscope import snapshot_download

from transformers import AutoModelForCausalLM, AutoTokenizer

# Downloading model checkpoint to a local dir model_dir

# model_dir = snapshot_download('qwen/Qwen-7B')

# model_dir = snapshot_download('qwen/Qwen-7B-Chat')

# model_dir = snapshot_download('qwen/Qwen-14B')

model_dir = snapshot_download('qwen/Qwen-14B-Chat')

# Loading local checkpoints

# trust_remote_code is still set as True since we still load codes from local dir instead of transformers

tokenizer = AutoTokenizer.from_pretrained(model_dir, trust_remote_code=True)

model = AutoModelForCausalLM.from_pretrained(

    model_dir,

    device_map="auto",

    trust_remote_code=True

).eval()

模型下载过程如下：

1. 在上述代码后加入以下代码进行对话

response, history = model.chat(tokenizer, "你好", history=None)

print(response)

response, history = model.chat(tokenizer, "浙江的省会在哪里？", history=history)

print(response)

response, history = model.chat(tokenizer, "它有什么好玩的景点", history=history)

print(response)