Quick start

Use the pretrained weight as embedding

Fine tune the transformer pretrained weight by contrastive learning

Query search by faiss

Rerank to enhance the performance

Langchain example for RAG