Crate mif_embed 

Local sentence-embedding inference for the MIF (Modeled Information Format) ecosystem, via candle.

Embedder loads sentence-transformers/all-MiniLM-L6-v2 (384-dimensional, mean-pooled, L2-normalized sentence embeddings) from the Hugging Face Hub on first use, caching the model files under the platform cache directory so later runs are offline. Inference runs on CPU only.

Structs

Embedder

Loads sentence-transformers/all-MiniLM-L6-v2 once and computes normalized sentence embeddings from it.

Enums

EmbedError

Errors from loading the embedding model or running inference.

Constants

EMBEDDING_DIM

Output embedding dimensionality of sentence-transformers/all-MiniLM-L6-v2.