README.md

metadata

title: CrispEmbed
sdk: docker
app_port: 7860
pinned: false

CrispEmbed — Text, Image, Face & Math OCR Embedding Demo

Lightweight embedding inference via ggml. No Python runtime, no ONNX.

Text Embeddings: Dense, sparse (SPLADE/BGE-M3), ColBERT multi-vector, cross-encoder reranking. 10 encoder/decoder architectures, 58 models.

Math OCR: Six engines for math-image → LaTeX (printed + handwritten).

Image Embeddings: CLIP/SigLIP cross-modal text-image search.

Face: Detection (SCRFD/YuNet) + Recognition (ArcFace/AuraFace/SFace).

Powered by the CrispEmbed C++ engine. Models auto-download on first use.