Productrank_bm25

rank-bm25

A Python library implementing the BM25 retrieval algorithm, supporting several BM25 variants and used to build sparse retrieval indexes. Lowercasing and tokenization preprocessing must be performed separately.

Timeline (last 90 days)

Oct 15

The script installs only four dependencies: anthropic, openai, rank-bm25, and numpy; it requires the environment variables ANTHROPIC_API_KEY and OPENAI_API_KEY.

Unverified50%

All Facts (1)

Unverified

The script installs only four dependencies: anthropic, openai, rank-bm25, and numpy; it requires the environment variables ANTHROPIC_API_KEY and OPENAI_API_KEY.

50%

Source Articles

Testing Anthropic Contextual Retrieval with 80 Lines of Code: How Much Does Chunk Prefixing Actually Improve Recall?Oct 15