A 213M fashion retrieval encoder (SigLIP2-base-384 backbone, fashion-vocabulary build).
Open weights.
Served with its calibrated recipe it becomes
MODA Pro Lite+ — the strongest open
system at ≤250M parameters on catalogue and title search. The recipe lives in that repository;
the weights live here, and Pro Lite+ pulls them at load time.
Results
MAP@10, full corpus, all ground-truth queries, one evaluator (pytrec_eval map_cut.10).
MODA is FashionSigLIP with its own serving recipe, shown for reference.
benchmark
MODA
Pro Lite (bare)
Pro Lite+ (with recipe)
KAGL
0.2887
0.3055
0.3201
Polyvore
0.3726
0.3952
0.4049
Atlas
0.1862
0.1814
0.1904
Fashion200K
0.1946
0.1758
0.1846
DeepFashion In-Shop
0.1642
0.0930
0.1026
DeepFashion Multimodal
0.0147
0.0118
0.0133
Pro Lite+ leads the ≤250M class on KAGL, Polyvore and Atlas — +10.9% over MODA on KAGL,
+8.7% on Polyvore, both significant under a paired bootstrap (10,000 resamples).
The recipe is worth +2.5% to +12.8% over the bare encoder on every benchmark, and costs
nothing at query time: the views are fused into a single vector before indexing.
Where this model is weak, stated plainly. Pro Lite is tuned for short catalogue titles.
On long natural-language descriptions it trails FashionSigLIP substantially — DeepFashion
In-Shop queries average 75 words, and Pro Lite+ scores 0.1026 there against MODA's 0.1642.
If your queries are descriptions rather than titles, use
MODA Duo, which routes per query.
768-d embeddings, cosine similarity, one vector per item. Index them in any vector database.
For the recipe that lifts these numbers to the Pro Lite+ column, use
moda-pro-lite-plus.
Evaluation
All figures are full corpus, all ground-truth queries, MAP@10 under one evaluator
(pytrec_eval map_cut.10), float32. Per-query results and confidence intervals are in the
repository.
