Module candle 

GTE-small embedder using Candle.

Uses thenlper/gte-small model for text embeddings:

• 384 dimensions
• 512 max tokens
• BERT architecture

Structs

CandleEmbedder

GTE-small embedder using Candle.

Constants

EMBEDDING_DIM 🔒

Embedding dimension for gte-small.

MAX_TOKENS 🔒

Maximum sequence length.

MODEL_ID 🔒

Model identifier on HuggingFace Hub.