Fix bug with multiple minibatch variables #7408
Merged
Conversation
This file contains 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
ricardoV94
force-pushed
the
fix_multiple_minibatch_bug
branch
from
b47c935
to
f1e3d9c
Compare
Fixes bug in VI with multiple Minibatch variables, which occurred due to separate calls to model.logp (from model.datalogp and model.varlogp) that create distinct clones of the RandomIntegersRV underlying minibatch slicing. `compile_pymc` would not set any updates in this case
ricardoV94
force-pushed
the
fix_multiple_minibatch_bug
branch
from
f1e3d9c
to
10f3aef
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #7408 +/- ##
==========================================
- Coverage 92.19% 92.18% -0.02%
==========================================
Files 103 103
Lines 17212 17261 +49
==========================================
+ Hits 15869 15912 +43
- Misses 1343 1349 +6
|
| total_size=len(y), | ||
| ) | ||
| mean_field = pm.fit(10_000, obj_optimizer=pm.adam(learning_rate=0.01), progressbar=False) | ||
| np.testing.assert_allclose(mean_field.mean.get_value(), true_weights, rtol=1e-1) |
There was a problem hiding this comment.
Does this test need to run the whole model and check parameter recovery? It should be enough to compile the function and check that minibatch_feature and minibatch_y change after each loss function execution right?
There was a problem hiding this comment.
This should run pretty fast, with minibatch of 1. I think it's a useful integration test, we didn't have any linear regression minibatch test in the codebase.
Besides VI has a very complex logic leading to building the function that I rather treat as a black box
jessegrabowski
approved these changes
Jul 10, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Description
Reported in https://discourse.pymc.io/t/verifying-that-minibatch-is-actually-randomly-sampling/14308
The bug occurred due to separate calls to model.logp (from model.datalogp and model.varlogp) that create distinct clones of the RandomIntegersRV underlying minibatch slicing.
compile_pymcwould not set any updates in this caseRelated Issue
Checklist
Type of change
📚 Documentation preview 📚: https://pymc--7408.org.readthedocs.build/en/7408/