Add DiffGRM generative recommender - #718
Open
hieuddo wants to merge 1 commit into
Open
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
This PR integrates
DiffGRM(Diffusion-based Generative Recommendation Model) intocornacthrough theNextItemRecommenderinterface.The official repository at commit
ad7b971c7e525e9fea6fb8e362a5c49dccb2473cdid not contain a root repository license, so we tried to implementDiffGRMwithout copying the code directly (ours isApache-2.0).Key changes
DiffGRMmodel using Cornac'sNextItemRecommenderandFeatureModalityinterfaces.Verification and reproducibility
The evaluation uses Amazon Sports 2014 5-core, chronological leave-one-out,
max_len=50, the frozen released Semantic-ID table, paper beam size 128, and seeds 2024-2026. Checkpoints are selected using the SID-level validation objective0.8 * NDCG@10 + 0.2 * Recall@10.The three-seed mean is
5.19%-7.29%below the paper across the four metrics. Seed 2025, which also has the highest declared validation score, is within0.46%-0.91%of every paper value. The paper likely used results from one seed instead of average of multiple runs (see https://github.com/hieuddo/cornac/blob/c0d200c560de6d3708b329395a6f327d61963e39/cornac/models/diffgrm/README.md#reproduction-path)This verification shows the Cornac implementation achieves reasonably comparable results.
Related Issues
Checklist:
README.md(if you are adding a new model).examples/README.md(if you are adding a new example).