Add export_valid_split: write the valid_ratio split as GPUMD-ready files - #4
Merged
Conversation
export_valid_split(data_file, valid_ratio, run_seed, output_dir) writes train.xyz + test.xyz reproducing exactly the split that train_nep(valid_ratio=..., run_seed=...) uses internally, with frames copied verbatim (raw text, untouched fields/precision) in input-file order - so the same data partition can be trained in GPUMD and the loss curves compared directly. The split draw is factored into data.valid_split_indices and both trainers now call it, so the trainers and the export can never disagree. Test: reference energies in energy_test.out from a valid_ratio run equal the exported test.xyz energies row for row; partition is verbatim and deterministic. Also restores train_sharded's 'from .data import' line that a refactor script had accidentally dropped (unreleased; DDP smoke test re-run and passing).
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
export_valid_split(data_file, valid_ratio, run_seed, output_dir) writes train.xyz + test.xyz reproducing exactly the split that train_nep(valid_ratio=..., run_seed=...) uses internally, with frames copied verbatim (raw text, untouched fields/precision) in input-file order - so the same data partition can be trained in GPUMD and the loss curves compared directly.
The split draw is factored into data.valid_split_indices and both trainers now call it, so the trainers and the export can never disagree. Test: reference energies in energy_test.out from a valid_ratio run equal the exported test.xyz energies row for row; partition is verbatim and deterministic.
Also restores train_sharded's 'from .data import' line that a refactor script had accidentally dropped (unreleased; DDP smoke test re-run and passing).