简体中文

我在服务器部署了RAGFlow，但是但我在上传大文件时，原本服务器的占用较大，结果突然下去了

欸？打开网页一看

我想这不对吧

后面找到了某篇文章：网站链接

据文章所说，是因为部署环境因为网络无法访问 huggingface.co ，导致本地模型无法下载，只需要在启动镜像前，在.env文件中，使用镜像地址就行：

# Uncomment the following line if you have limited access to huggingface.co:
# 把下面这行注释关闭
HF_ENDPOINT=https://hf-mirror.com

# Optimizations for MacOS
# Uncomment the following line if your OS is MacOS:
# MACOS=1

取消掉注释即可

然后重新启动容器

sudo docker compose -f docker-compose.yml up -d

English

I deployed RAGFlow on the server. However, when I uploaded a large file, the original server's memory usage suddenly decreased.

Oh? Open the web page.

Isn't this not right?

After finding an article: Website link.

According to the article, this is because the deployment environment cannot access huggingface.co (I think it's reasonable). To solve this issue, you need to uncomment the following line in the .env file when starting the container:

# Uncomment the following line if you have limited access to huggingface.co:
# 把下面这行注释关闭
HF_ENDPOINT=https://hf-mirror.com

# Optimizations for MacOS
# Uncomment the following line if your OS is MacOS:
# MACOS=1

Commenting out the above line resolves the issue.

Then, restart the container using:

sudo docker compose -f docker-compose.yml up -d