Dataset · THUIR
Large Chinese text ranking benchmark
THUIR/T2Ranking
Large-scale Chinese dataset for training and evaluating passage search and ranking models.
What's inside
The dataset is a large Chinese benchmark for passage ranking in information retrieval. It contains between 1 and 10 million examples, which allow evaluating how accurately a system builds a list of results by relevance to a search query.
What you can do with it
- Training and evaluating re-rankers for search systems and RAG in Chinese
- Evaluating semantic similarity and text classification models
- Tasks
- Text retrieval, Text classification, Sentence similarity
- Languages
- Chinese
- Size
- 1M–10M rows
- Downloads this month
- 4.6K
- Likes
- 36
How to load it
from datasets import load_dataset
ds = load_dataset("THUIR/T2Ranking")