[BUG] Omnisafe cannot run on a Mac M1 device

[u21h2]

Required prerequisites

• [X] I have read the documentation https://omnisafe.readthedocs.io.
• [X] I have searched the Issue Tracker and Discussions that this hasn't already been reported. (+1 or comment there if it has.)
• [X] Consider asking first in a Discussion.

What version of OmniSafe are you using?

0.2.2

System information

3.9.13 (main, Aug 25 2022, 18:29:29) [Clang 12.0.0 ] darwin 0.2.2

Problem description

Admittedly, this is a very nice library. However, Omnisafe has the following bug when i run the following command omnisafe train --algo PPO --total-steps 1024 --vector-env-nums 1 --custom-cfgs algo_cfgs:update_cycle --custom-cfgs 512 --device cpu.

Note: My device is MacBook Pro M1 Pro.

Reproducible example code

The Python snippets:

Command lines:

Extra dependencies:

Steps to reproduce:

1. 2. 3.

Traceback

No response

Expected behavior

No response

Additional context

No response

[zmsn-2077]

We have not tested on the Mac platform before, but it seems necessary to support Mac and Windows platforms, we will start testing on the Mac platform as soon as possible and give a solution.

[zmsn-2077]

Thank you for providing your invaluable feedback. We would like to inform you that the OmniSafe repo has undergone an intense refactoring process over the past two weeks. The crux of OmniSafe is PyTorch and Safety-Gymnasium, with the former enjoying extensive support from its user community on Mac M-series chips. Our team has conducted an exhaustive examination of the Safety-Gymnasium's compatibility with Mac, and we are pleased to report that OmniSafe can now be trained on the Mac's M1 chip. Our detailed results were obtained using the macOS Ventura 13.3.1 and Apple M1 Max platform for testing.

If you meet the following error,

RuntimeError: No ffmpeg exe can be found. Install ffmpeg on your system, or set the IMAGEIO_FFMPEG_EXE environment variable. you can paste the following code into gymnasium/utils/save_video.py after from gymnasium import logger

import os
os.environ["IMAGEIO_FFMPEG_EXE"] = "/usr/bin/ffmpeg"

For more details about this error, you can refer to https://github.com/Zulko/moviepy/issues/1158

In addition to that, following OmniSafe's installation instructions, you can train your SafeRL model on MacOS.