Add DeBERTa training and ONNX export pipeline for required phrases extraction - #5222
Open
Kaushik-Kumar-CEG wants to merge 1 commit into
Open
Conversation
Train the DeBERTa BIOES and CRF model with validated datasets and best-checkpoint selection. Add strict model export, reproducibility metadata, and focused offline tests. Signed-off-by: Kaushik Kumar <kaushikrjpm10@gmail.com>
Kaushik-Kumar-CEG
force-pushed
the
gsoc/train-deberta-model
branch
from
August 31, 2026 15:01
750d2ab to
4ccfe39
Compare
Kaushik-Kumar-CEG
marked this pull request as ready for review
August 31, 2026 15:01
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Part of #5137.
Adds the training and ONNX export scripts for the required phrase model. The model uses DeBERTa with BIOES labels and a CRF layer, and keeps the checkpoint with the best validation F1.
Also includes focused tests for dataset validation, token alignment, model loading, and export.
Historical result: 0.8445 test F1. A new full run with the updated script is pending.
The packaged version is in aboutcode-org/scancode-required-phrases#2.