Rebase on main + review fixes#3
Open
orionarcher wants to merge 66 commits intothomasloux:feat/support-plumedfrom
Open
Rebase on main + review fixes#3orionarcher wants to merge 66 commits intothomasloux:feat/support-plumedfrom
orionarcher wants to merge 66 commits intothomasloux:feat/support-plumedfrom
Conversation
Co-authored-by: Orion Cohen <27712051+orionarcher@users.noreply.github.com> Co-authored-by: orionarcher <orioncohen@gmail.com>
Co-authored-by: Daniel Zuegner <dzuegner@microsoft.com>
…TorchSim#466) Signed-off-by: Rhys Goodall <rhys.goodall@outlook.com>
Co-authored-by: Orion Cohen <27712051+orionarcher@users.noreply.github.com>
Signed-off-by: Rhys Goodall <rhys.goodall@outlook.com>
Co-authored-by: Abhijeet Gangan <abhijeetgangan@g.ucla.edu>
… angle_tolerance in symmetry pipeline (TorchSim#527)
…m different the integrators (TorchSim#521)
Co-authored-by: Stefano Falletta <49149059+falletta@users.noreply.github.com>
PlumedModel.forward() now returns "biased_energy" and "biased_forces" keys in addition to the combined "energy" and "forces". This lets users inspect the PLUMED bias contribution independently and integrates with the extras system via store_model_extras().
PLUMED virial corrections are not applied to the stress tensor, so emit a warning at construction time if the wrapped model has compute_stress=True.
PLUMED bias forces are computed via numpy and detached from the autograd graph, so they are incompatible with differentiable-force workflows.
…r.py Import _PLUMED_LENGTH_FACTOR and _PLUMED_ENERGY_FACTOR from plumed.py instead of redefining them. No circular import risk since plumed.py only imports from plumed_builder lazily inside functions.
These are the PLUMED bias corrections only, not the total biased values. The total is in "energy" and "forces" (unbiased + correction). Users can recover unbiased values via energy - plumed_bias_energy.
Clarify in the docstring that energy = model_energy + energy_bias and forces = model_forces + forces_bias.
orionarcher
force-pushed
the
orionarcher/rebase-plumed-pr
branch
from
4d3ed50 to
d1ac7cb
Compare
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
12 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.
Summary
tests/conftest.py,torch_sim/__init__.py,torch_sim/plumed.py)use_neighbor_listparam from tutorials, removed nonexistentStateDictimport, addedPlumedModelto__all__, added ty override for plumed.pyenergy_biasandforces_biasseparately so users can inspect the PLUMED correction (energy = model_energy + energy_bias)compute_stress=True(virial not applied) orretain_graph=True(bias forces detached from autograd)plumed.pyandplumed_builder.pyTest plan
🤖 Generated with Claude Code