Skip to content
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

Unexpectedly bad results for billiards.py variant #14

Open
DomNomNom opened this issue Feb 8, 2020 · 5 comments
Open

Unexpectedly bad results for billiards.py variant #14

DomNomNom opened this issue Feb 8, 2020 · 5 comments

Comments

@DomNomNom
Copy link
Contributor

DomNomNom commented Feb 8, 2020

In billiards.py, I've done some changes to the starting condition of the cue ball and the target position. These changes are available here.

What I observed was that in almost every iteration, the result got worse than before not better.
i.e. Loss seems to be increasing:
Loss is increasing

The ending stable state is not that surprising (no gradient) but the initial trend is concerning me.

@yuanming-hu
Copy link
Member

The billiards example is not implemented with the time-of-impact fix (see our paper) so I'm not surprised that it's making inverse progress. I'll fix this in a few days.

@DomNomNom
Copy link
Contributor Author

DomNomNom commented Feb 10, 2020 via email

@yuanming-hu
Copy link
Member

Hi @DomNomNom ,

I added the TOI fix to billiards.py. Gradient quality is significantly improved. Could you have a try? (PS: please also upgrade taichi to 0.4.6+). Thanks.

@DomNomNom
Copy link
Contributor Author

Hi @yuanming-hu

I've updated taichi and merged your changes, however I'm still seeing similar symptoms:
BadLoss2

Here's what I'm running: https://github.com/DomNomNom/difftaichi/blob/5326ebfc1c8bd5bfe56b3ba5b68ad891b158aa20/examples/billiards.py
Note: after updating, pip3 list | grep taichi gave me taichi-nightly-cuda-10-1 0.5.0.

@DomNomNom
Copy link
Contributor Author

Any thoughts on this?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants