---
language:
- ja
task_categories:
- text-retrieval
dataset_info:
  features:
  - name: query
    dtype: string
  - name: positives
    sequence: string
  - name: negatives
    sequence: string
  - name: bm25_negatives
    sequence: string
  splits:
  - name: train
    num_bytes: 6710513973
    num_examples: 391061
  download_size: 3304340359
  dataset_size: 6710513973
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

[Under Construction]

This is a repository containing all the queries from the Japanese part of the MMarco dataset, the multilingual version of the MSMarco dataset.

For each query, there are matching hard negatives:
- 25 of them retrieved by the multilingual e5 base model.
- Up to 10 of them retrieved by the basic implementation of BM25 from Japanese in the Anserini library.