Serving bundle — Semantic Recommendation & Ranking Engine

Everything the API needs at runtime, computed once from the training history: 23,109 movie vectors (bge-small-en-v1.5, 384-d), movie metadata, aggregate item statistics (popularity, Bayesian average rating, …), the warm and cold-start LightGBM LambdaMART rankers and the validated serving settings. Loaded by recsys.serving.Engine — code: https://github.com/Vanshika-0111/semantic-recsys

No individual ratings are included. The MovieLens licence does not allow redistributing the ratings, so the demo users of the local bundle are left out (0 here).

Ratings: MovieLens 32M — F. Maxwell Harper and Joseph A. Konstan. 2015. The MovieLens Datasets: History and Context. ACM TiiS 5(4). GroupLens Research does not endorse this project. Non-commercial use only.
Movie metadata / poster paths: TMDB (via a public Kaggle dump); not endorsed or certified by TMDB.

Provenance: split a99079d4a40a, code 8e192bf, bundle version 1.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support