instadeepai / instadeepai/Mava

[BUG] Issue due to newest gymnasium==1.1.0

Open
#1,163 0 comments 0 reactions 0 assignees View on GitHub
bug
Dominant language
Python
Stars
935
Forks
123
Avg merge
6d 8h
Merged PRs (30d)
9

Description

### Describe the bug
The latest working version of Mava with Gymnasium is `1.0.0,` while version `1.1.0` has errors.

### To Reproduce
Steps to reproduce the behavior:
1. Simply create new conda env and do all the installations.
2. Run `pip list | grep gymnasium`, the expected installed version will be 1.1.0.
3. Run PPO sebulba system and you'll get the 'expected behavior' below.
4. To fix, do `pip install gymnasium==1.0.0` and rerun PPO sebulba.

### Expected behavior
In the `make_env` function for Sebulba environments, we encountered an error from the `async_multiagent_worker`, as shown in the screenshot.
![Image](https://github.com/user-attachments/assets/a7d832ce-7f60-4462-8f67-b2f27cd8d82f)

### Possible Solution
Fix the bug given by the newer version or freeze the working version.

Contributor guide

Open the contributing guide

Research direction

Start with the make_env function for Sebulba environments and trace the async_multiagent_worker error while running the PPO Sebulba system. Compare behavior with gymnasium 1.0.0 and 1.1.0; done means the reported error is resolved or the working dependency version is explicitly constrained.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.