-
Notifications
You must be signed in to change notification settings - Fork 129
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CSE 276F Submission] PushT-v1 #378
Conversation
@StoneT2000 , could I request a review on this PR please? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nice work, would love to get this merged in, just see the comments. ppo works fast as well.
@StoneT2000 , I've made the changes. Let me know if there are any more issues |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
only some small issues with comments left but i will fix that myself. This looks good to merge
PushT from diffusion policy digital twin
Created PandaStick agent
Fully batched calculation for T block intersection area with goal T area, zero explicit loops
Support for ur5e robot is a todo for true twinness
Wristcam attached to PandaStick is a todo for visual based RL/IL
ppo args (consistent learning):
python ppo.py --env_id="PushT-v1" --exp-name="final_run" --num_envs=1024 --update_epochs=8 --num_minibatches=32 --total_timesteps=55_000_000 --eval_freq=8 --num-steps=100 --num_eval_steps=100 --gamma=0.99