Fix differential equation tutorial not converging#592
Merged
penelopeysm merged 3 commits intomainfrom May 9, 2025
Merged
Conversation
penelopeysm
force-pushed
the
py/fix-differential-equations
branch
from
026a22c to
67d9439
Compare
penelopeysm
force-pushed
the
py/fix-differential-equations
branch
from
c1b084f to
20d3031
Compare
Contributor
|
Preview the changes: https://turinglang.org/docs/pr-previews/592 |
Member
|
Categorically, I have no objection to the updates here. Maybe you could add a sentence on prior selection, it can be something like "we are choosing prior close to the ground truth". I think it might make it more clear what we are trying to demonstrate here. This is not related to the changes made in this PR, just something crosses my mind. |
Member
Author
|
Thanks @sunxd3! I added a note about the priors |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #588 by reducing the amount of random noise in the synthetic data, and also making the priors in the model more strongly peaked on the ground truth. (I suppose that this is a bit of a cheat and ideally we'd sample more, but the docs already take a long time to run.)
Before: https://turinglang.org/docs/tutorials/bayesian-differential-equations/
After: https://turinglang.org/docs/pr-previews/592/tutorials/bayesian-differential-equations/