Add ADAM and RMSProp optimizers #2041
Open
Conversation
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
Contributor
Memory benchmark result| Test Name | %Δ | Master (MB) | PR (MB) | Δ (MB) | Time PR (s) | Time Master (s) |
| -------------------------------------- | ------------ | ------------------ | ------------------ | ------------ | ------------------ | ------------------ |
test_objective_jac_w7x | 2.08 % | 3.952e+03 | 4.034e+03 | 82.14 | 47.33 | 43.93 |
test_proximal_jac_w7x_with_eq_update | 1.70 % | 6.613e+03 | 6.726e+03 | 112.66 | 198.04 | 198.83 |
test_proximal_freeb_jac | -0.08 % | 1.323e+04 | 1.322e+04 | -10.12 | 106.29 | 106.91 |
test_proximal_freeb_jac_blocked | -0.06 % | 7.550e+03 | 7.545e+03 | -4.70 | 93.96 | 94.68 |
test_proximal_freeb_jac_batched | 0.17 % | 7.561e+03 | 7.575e+03 | 13.02 | 93.96 | 94.27 |
test_proximal_jac_ripple | -0.27 % | 3.579e+03 | 3.570e+03 | -9.60 | 71.90 | 73.01 |
test_proximal_jac_ripple_bounce1d | -2.68 % | 3.765e+03 | 3.664e+03 | -101.03 | 84.17 | 85.57 |
test_eq_solve | -0.09 % | 2.028e+03 | 2.026e+03 | -1.75 | 101.17 | 101.58 |For the memory plots, go to the summary of |
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #2041 +/- ##
=======================================
Coverage 95.75% 95.76%
=======================================
Files 102 102
Lines 28344 28374 +30
=======================================
+ Hits 27142 27171 +29
- Misses 1202 1203 +1
🚀 New features to boost your workflow:
|
YigitElma
requested review from
a team,
ddudt,
dpanici,
f0uriest,
rahulgaur104 and
unalmis
and removed request for
a team
YigitElma
commented
Dec 18, 2025
desc/optimize/stochastic.py
Outdated
| Update rule for 'rmsprop': | ||
| v_{k} = beta*v_{k-1} + (1-beta)*grad(x_{k})^2 | ||
| x_{k+1} = x_{k} - alpha * grad(x_{k}) / (sqrt(v_{k}) + epsilon) |
Collaborator
Author
There was a problem hiding this comment.
Some literature uses different Greek letters for alpha and beta, but I used these to make api slightly more compact.
YigitElma
commented
Dec 18, 2025
Collaborator
Author
There was a problem hiding this comment.
Sorry for the unrelated change, noticed this while fixing the doc errors
dpanici
reviewed
Dec 19, 2025
| ValueError, | ||
| "x_scale should be one of 'auto' or array-like, got {}".format(x_scale), | ||
| ) | ||
| if isinstance(x_scale, str): |
Collaborator
There was a problem hiding this comment.
should say in docstring that 'x_scale="auto"' is essentially just means no scaling as we just set xscale to 1 in that case.
dpanici
requested changes
Dec 19, 2025
Collaborator
dpanici
left a comment
dpanici
left a comment
There was a problem hiding this comment.
just the small docstring fix, should be explitict that x_scale='"auto"` does no scaling here
dpanici
approved these changes
Dec 22, 2025
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.
x_scaleis now used with SGD methods too