edbeeching / godot_rl_agents

An Open Source package that allows video game creators, AI researchers and hobbyists the opportunity to learn complex behaviors for their Non Player Characters or agents
MIT License
942 stars 69 forks source link

MacOS M1 support for rllib #76

Closed visuallization closed 1 year ago

visuallization commented 1 year ago

Example command: gdrl --trainer=rllib --env=gdrl --env_path=absolute/path/to/JumperHard.app --num_workers=10 --experiment=JumperHard --viz --speedup=8 --batched_sampling=True --num_gpus=0

visuallization commented 1 year ago

I also realized that samplefactory is also working on m1 macs if you set --device=cpu

gdrl --trainer=sf --env=gdrl --env_path=absolute/path/to/JumperHard.app --num_workers=10 --experiment=JumperHard --viz --speedup=8 --batched_sampling=True --device=cpu

Which actually makes sense because it only has troubles with cuda support.

edbeeching commented 1 year ago

Thanks, merged.