Dataset · databricks
OfficeQA Pro v2 document reasoning benchmark
databricks/officeqa-pro-v2
Dataset for evaluating models and agents on US Federal report documents.
What's inside
Contains English question-answer pairs for evaluating end-to-end logical reasoning over real documents. Covers US financial reporting spanning two centuries (1793-2024). The dataset size is less than 1K examples.
What you can do with it
- Evaluating LLM and agent document reasoning capabilities
- Testing retrieval-augmented generation (RAG) systems
- Verifying information extraction quality from historical reports
- Tasks
- Question answering, Text generation, Text retrieval
- Languages
- English
- Size
- up to 1K rows
- Downloads this month
- 25.2K
- Likes
- 25
- Access
- on request on Hugging Face
How to load it
from datasets import load_dataset
ds = load_dataset("databricks/officeqa-pro-v2")