A problem with the ppo.py runtime.

ScheiklP / sofa_zoo

Reinforcement learning scripts for sofa_env environments.

MIT License

5 stars 5 forks source link

A problem with the ppo.py runtime. #7

Closed wjyustl closed 2 months ago

wjyustl commented 3 months ago

Hi, @ScheiklP When I run ppo.py in "rope_threading", the program breaks when "total_timesteps" reaches 50000 with the following error: I guess the problem may have something to do with the render of Xservice. Have you ever encountered this problem? Can you give me some advice?

ScheiklP commented 3 months ago

Hi @wjyustl, no sorry, I have never seen this error before. Could you try using a non-vectorized environment? Sometimes errors get lost, when you have a multiprocessing setup. You can test this by using the DummyVecEnv and setting the number of environments to 1.