Datasets:
Tasks:
Text Retrieval
Modalities:
Text
Formats:
json
Sub-tasks:
document-retrieval
Languages:
Bengali
Size:
100K - 1M
metadata
language:
- bn
multilinguality:
- monolingual
task_categories:
- text-retrieval
task_ids:
- document-retrieval
configs:
- config_name: AIR-Bench_24.05
data_files:
- split: corpus_default
path: AIR-Bench_24.05/default/corpus.jsonl
- split: queries_default_dev
path: AIR-Bench_24.05/default/dev_queries.jsonl
- split: queries_default_test
path: AIR-Bench_24.05/default/test_queries.jsonl
- split: qrels_default_dev
path: AIR-Bench_24.05/default/dev_qrels.jsonl
Available Versions:
- AIR-Bench_24.05
- Task / Domain / Language: qa / wiki / bn
- Available Datasets (Dataset Name: Splits):
- default: dev, test