hopit-ai

hopit-ai / Moda

Public

Breakdown of the components of State of the art Fashion Search

24
1
100% credibility
Found Apr 15, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

MODA is an open benchmark for evaluating complete fashion product search systems using H&M purchase data across lexical, dense, hybrid retrieval, and reranking components.

How It Works

1
🔍 Discover better fashion search

You find MODA, a free tool to test how well shopping searches work using real store data.

2
📥 Get the shopping catalog

Download lists of clothes and customer searches to use as your test playground.

3
🚀 Start the search helper

Turn on a simple background tool that lets you quickly find matching items.

4
🧠 Add smart matching tricks

Try different ways to match searches like keywords, smart guesses, or picture-like understanding.

5
📊 Run tests and compare

Watch as it tests hundreds of searches and shows which tricks find the right clothes best.

🏆 See your leaderboard

Enjoy clear charts proving what makes shopping searches fast and accurate for customers.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 24 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Moda?

Moda delivers an open benchmark for end-to-end fashion search pipelines, using 253k purchase-grounded queries against 105k H&M products. It provides a component-by-component breakdown into BM25 lexical matching, dense embeddings, hybrid fusion, NER attribute boosting, and cross-encoder reranking—showing gains like +105% nDCG over embedding-only baselines. Built in Python with Docker for OpenSearch and quick-start scripts to index your catalog CSV and run evals.

Why is it gaining traction?

Unlike proprietary tools from Algolia or Bloomreach, or embedding-only libraries like Marqo, Moda isolates each pipeline stage's impact with reproducible metrics and latency breakdowns (full pipeline at 62ms/query on Apple Silicon). Developers get leaderboards, ablation configs, and easy extension to your data—no cloud GPUs needed. The query provenance from real purchases hooks those chasing production-grade insights.

Who should use this?

E-commerce engineers tuning fashion/clothing search at scale, especially teams debating lexical vs. semantic retrieval or adding NER boosts. Product leads evaluating vendor claims against open baselines. Indie hackers prototyping hybrid search on apparel catalogs.

Verdict

Grab it for benchmarking your fashion search stack—docs and quickstarts are solid despite 19 stars and 1.0% credibility score signaling early maturity. Run the H&M repro first; extend to your CSV for real wins, but expect tweaks for prod scale.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.