Datasets:
Modalities:
Text
Formats:
json
Sub-tasks:
document-retrieval
Languages:
English
Size:
100K - 1M
Tags:
conversational-search
conversational-retrieval
information-retrieval
multi-turn
query-rewriting
benchmark
License:
Commit History
3 days ago
-
Each record: history + question (tier and passages belong to the question); generated subset removed; card with a worked example
Verified signaturecf29032
4 days ago
-
One benchmark config with train / dev / test splits (eval_set names each set); data samples in the card
Verified signature0439dc1 -
ConvRevise: benchmark, development and training questions, the three collections as JSONL, dataset card
Verified signature4723cba