CoQA is a large-scale dataset for building Conversational Question Answering systems.
Our dataset contains 127k questions with answers, obtained from 8k conversations about text passages from seven diverse domains. The questions are conversational, and the answers are free-form text with their corresponding evidence highlighted in the passage.
Supported Tasks and Leaderboards
More Information Needed
Languages… See the full description on the dataset page: https://huggingface.co/datasets/stanfordnlp/coqa.
