Rewrite target and dataset info classes #216
Closed
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
Co-authored-by: Philip Loche <[email protected]>
Co-authored-by: Philip Loche <[email protected]>
* Write train output to hydra's output directory * Added evaluation function * Add usage example for cli interface * update train cli * Disable MacOS tests * Add cli skeleton for exporter --------- Co-authored-by: frostedoyster <[email protected]>
* Add gradient calculator * Temporary losses * Forces and stresses * Support multiple model outputs in SOAP-BPNN
…or-models into trainer-class
--------- Co-authored-by: frostedoyster <[email protected]> Co-authored-by: Filippo Bigi <[email protected]> Co-authored-by: Arslan Mazitov <[email protected]>
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Outdated suggestions cannot be applied.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.
Based on the discussion in #185 moving the creation of the DatasetInfo from the train function into a function I fall into a rabbit hole and ended up with rewrtiting the target and dataset dataclasses. The former classes
TargetInfoandDatasetInfonow have new methods forcopy,unionandupdate.Also, I introduced a new class
TargetInfoDictwhich is a special dictionary to work (only) withTargetInfoinstances. TheTargetInfoDicthas methods forunionintersectionanddifference. Basically, all of these methods were already used by some architectures but were implemented always by “hand”.One thing you will notice that I changed also the types for the
gradientsinTargetInfoand the atomic_types in theDatasetInfo. Both are now of typeset.I think these changes make our code cleaner and also easier for architecture developers to work with our Info objects. If we need more methods we can now add them easily.
These new classes also allow us to ship the targets and the target infos (TargetInfoDict) together when reading targets from a file with the
read_targetsfunction. Metatensor style :-)TODO
📚 Documentation preview 📚: https://metatensor-models--216.org.readthedocs.build/en/216/