This dataset was generated using YourBench (v0.6.0), an open-source framework for generating domain-specific benchmarks from document collections.
ingestion: Read raw source documents, convert them to normalized markdown and save for downstream steps
summarization: Perform hierarchical summarization: chunk-level LLM summaries followed by combine-stage reduction
chunking: Split texts into token-based single-hop and multi-hop… See the full description on the dataset page:
https://huggingface.co/datasets/LunaticMuch/finos-cdm-cdm-java-distribution.
