huggingface下载模型的方法

下载方式一

```
# 查看服务器操作系统，根据操作系统选择git-lfs下载方式
uname -a
# 下载 git-lfs ubuntu版
sudo apt-get install git-lfs
# 下载 git-lfs centos版
sudo yum install git-lfs
# 初始化git-lfs
git lfs install
# 下载预训练权重
git clone https://huggingface.co/Tele-AI/Telechat-7B
```

下载方式二

```
from transformers import AutoModel, AutoTokenizer
import os, sys

os.environ['HF_ENDPOINT'] = 'https://hf-mirror.com'

# 以bert-base-uncased为例，你可以根据需要更换为其他模型
model_name = "Tele-AI/telechat-7B"

save_dir = model_name

# 下载并缓存模型和分词器(tokenizer)
model = AutoModel.from_pretrained(model_name, cache_dir=save_dir)
tokenizer = AutoTokenizer.from_pretrained(model_name, cache_dir=save_dir)

```

下载方式三

```
import os
os.environ['HF_ENDPOINT'] = 'https://hf-mirror.com'

from huggingface_hub import snapshot_download

snapshot_download(repo_id="Tele-AI/telechat-7B", 
local_dir_use_symlinks = False,
local_dir = "./Tele-AI/telechat-7B")

```