Repository navigation
Fix PACE metrics for current NumPy and allow TensorFlow 2.21 - #43
Merged
Merged
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
PACE fitting fails with NumPy 2.4+ when its metrics code passes a one-element array to
float(). Extract that scalar explicitly and add regression coverage for flat and column regularization inputs. Allow the TensorFlow 2.21 series with a<2.22ceiling, and update the installation documentation and CPU compatibility workflow.Validation: pacemaker's full suite passed 294 tests on Python 3.12.15 / TensorFlow 2.21.0 / NumPy 2.5.4, and both Cu–I and Ethanol CLI workflows completed. On an NVIDIA H200 NVL (driver 595.58.03), all nine PACE fitting tests passed with 185 fitting calls whose energies, forces, and gradients were verified on GPU:0. The same nine tests and 185 GPU fitting calls passed on an NVIDIA A100 80 GB PCIe (driver 575.57.08), after prepending the environment CUDA library paths. Existing GRACE model/gradient/export checks passed 137 tests (two expected failures); the two new scalar regression cases also passed.
This is coordinated with the pacemaker refresh based on
ICAMS/python-ace:feature/build-speedup. The backend fix must land before pacemaker CI can run against the latest public upstream. GPU validation covers fitting correctness; it does not establish long-running training performance or memory stability.The TensorFlow dependency bound is package-wide. The GPU validation above covers PACE compatibility fitting; the broader GRACE checks in this refresh were on CPU. Full GRACE GPU training and long-running stability remain outside this validation.
Coordinated pacemaker PR: ICAMS/python-ace#106.