Zhendong-Wang / Diffusion-Policies-for-Offline-RL

Apache License 2.0
253 stars 36 forks source link

About the training effect #11

Closed HenryZhang-git closed 1 year ago

HenryZhang-git commented 1 year ago

Hi!Wang I have run the code python main.py --env_name walker2d-medium-expert-v2 --device 0 --ms online --lr_decay which you mentioned in 'README.md'.And then I get a file named 'debug.log'.When I use the tensorboard to visualize the data, I get some figures. image image image image

Are these Loss Functions a normal result of training? I'm looking forward to your reply. Thank you very much!