We present CLAP NQ, a benchmark Long-form Question Answering dataset for the full RAG pipeline. CLAP NQ includes long answers with grounded gold passages from Natural Questions (NQ) and a corpus to perform either retrieval, generation, or the full RAG pipeline. The CLAP NQ answers are concise, 3x smaller than the full passage, and cohesive, with multiple pieces of the passage that are not contiguous.
This is the annotated data for the generation portion of the RAG pipeline.
For more… See the full description on the dataset page:
https://huggingface.co/datasets/deralili/clapnq.
