rlworkgroup / garage

A toolkit for reproducible reinforcement learning research.
MIT License
1.84k stars 309 forks source link

Update max resource usage docs for new EnvSpec #2247

Closed jimzers closed 3 years ago

jimzers commented 3 years ago

Changed env.spec.max_episode_length to env.spec.max_episode_steps. This deals with changes to OpenAI Gym's EnvSpec for gym version 0.17.2.

jimzers commented 3 years ago

Ignore... Turns out this change only applies to OpenAI Gym's EnvSpec. Looked into .unwrapped.spec.max_episode_steps but could not figure things out.

Upon further inspection it seems that there is problem with the sampler... Will open an issue. Sorry!