BartAmin/Fine-tuned-RAG
I developed a fine-tuned retrieval head for RAG that learns to more reliably retrieve relevant passages by transforming the query embeddings before retrieval. It is trained on synthetically generated question-chunk pairs from the corpus, and benchmarked against standard top-K cosine similarity on the isaacus/legal-rag-bench legal corpus.
GitHub repository with 17 stars and 3 forks.
Language: Python