README.md

---
annotations_creators:
- human-annotated
language:
- ara
- ben
- eng
- fin
- ind
- jpn
- kor
- rus
- swa
- tel
- tha
license: cc-by-sa-3.0
multilinguality: multilingual
task_categories:
- text-retrieval
task_ids: []
dataset_info:
- config_name: arabic-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 1245714994
    num_examples: 2106586
  download_size: 562700591
  dataset_size: 1245714994
- config_name: arabic-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 10793262
    num_examples: 374523
  - name: dev
    num_bytes: 87088
    num_examples: 3115
  - name: test
    num_bytes: 36208
    num_examples: 1257
  download_size: 3781951
  dataset_size: 10916558
- config_name: arabic-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 831496
    num_examples: 12377
  - name: dev
    num_bytes: 207878
    num_examples: 3115
  - name: test
    num_bytes: 74696
    num_examples: 1081
  download_size: 621343
  dataset_size: 1114070
- config_name: bengali-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 313718344
    num_examples: 304059
  download_size: 113214290
  dataset_size: 313718344
- config_name: bengali-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 1410063
    num_examples: 51861
  - name: dev
    num_bytes: 11764
    num_examples: 443
  - name: test
    num_bytes: 3548
    num_examples: 130
  download_size: 479618
  dataset_size: 1425375
- config_name: bengali-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 233828
    num_examples: 1713
  - name: dev
    num_bytes: 60547
    num_examples: 440
  - name: test
    num_bytes: 16464
    num_examples: 111
  download_size: 135025
  dataset_size: 310839
- config_name: english-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 13953017759
    num_examples: 32907100
  download_size: 8012208332
  dataset_size: 13953017759
- config_name: english-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 3163945
    num_examples: 108401
  - name: dev
    num_bytes: 24984
    num_examples: 878
  - name: test
    num_bytes: 26591
    num_examples: 935
  download_size: 1201824
  dataset_size: 3215520
- config_name: english-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 180772
    num_examples: 3547
  - name: dev
    num_bytes: 45396
    num_examples: 878
  - name: test
    num_bytes: 39063
    num_examples: 744
  download_size: 179373
  dataset_size: 265231
- config_name: finnish-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 749089306
    num_examples: 1908757
  download_size: 446059094
  dataset_size: 749089306
- config_name: finnish-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 5510842
    num_examples: 197338
  - name: dev
    num_bytes: 47540
    num_examples: 1738
  - name: test
    num_bytes: 39723
    num_examples: 1451
  download_size: 1870069
  dataset_size: 5598105
- config_name: finnish-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 336371
    num_examples: 6561
  - name: dev
    num_bytes: 89446
    num_examples: 1738
  - name: test
    num_bytes: 63723
    num_examples: 1254
  download_size: 335034
  dataset_size: 489540
- config_name: indonesian-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 545994305
    num_examples: 1469399
  download_size: 276886119
  dataset_size: 545994305
- config_name: indonesian-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 4169295
    num_examples: 148491
  - name: dev
    num_bytes: 33405
    num_examples: 1224
  - name: test
    num_bytes: 26807
    num_examples: 961
  download_size: 1438860
  dataset_size: 4229507
- config_name: indonesian-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 241258
    num_examples: 4902
  - name: dev
    num_bytes: 60988
    num_examples: 1224
  - name: test
    num_bytes: 43451
    num_examples: 829
  download_size: 221700
  dataset_size: 345697
- config_name: japanese-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: dev
    num_bytes: 387783.4788097246
    num_examples: 903
  - name: test
    num_bytes: 377906.3802353905
    num_examples: 880
  - name: train
    num_bytes: 3006085074
    num_examples: 7000027
  download_size: 1692453219
  dataset_size: 3006850763.859045
- config_name: japanese-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 3198715
    num_examples: 112408
  - name: dev
    num_bytes: 25703
    num_examples: 928
  - name: test
    num_bytes: 25819
    num_examples: 923
  download_size: 1127195
  dataset_size: 3250237
- config_name: japanese-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 236537
    num_examples: 3697
  - name: dev
    num_bytes: 60007
    num_examples: 928
  - name: test
    num_bytes: 47149
    num_examples: 720
  download_size: 210465
  dataset_size: 343693
- config_name: korean-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 637416518
    num_examples: 1496126
  download_size: 364890926
  dataset_size: 637416518
- config_name: korean-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 1083048
    num_examples: 39485
  - name: dev
    num_bytes: 8121
    num_examples: 307
  - name: test
    num_bytes: 13482
    num_examples: 492
  download_size: 386327
  dataset_size: 1104651
- config_name: korean-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 86921
    num_examples: 1295
  - name: dev
    num_bytes: 20056
    num_examples: 303
  - name: test
    num_bytes: 28967
    num_examples: 421
  download_size: 87049
  dataset_size: 135944
- config_name: russian-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 5965881322
    num_examples: 9597504
  download_size: 2799247131
  dataset_size: 5965881322
- config_name: russian-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 4701206
    num_examples: 163899
  - name: dev
    num_bytes: 38306
    num_examples: 1375
  - name: test
    num_bytes: 33048
    num_examples: 1168
  download_size: 1735174
  dataset_size: 4772560
- config_name: russian-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 502545
    num_examples: 5366
  - name: dev
    num_bytes: 128310
    num_examples: 1375
  - name: test
    num_bytes: 97183
    num_examples: 995
  download_size: 413272
  dataset_size: 728038
- config_name: swahili-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 34142833
    num_examples: 136689
  download_size: 17501035
  dataset_size: 34142833
- config_name: swahili-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 1689502
    num_examples: 63390
  - name: dev
    num_bytes: 16381
    num_examples: 623
  - name: test
    num_bytes: 19984
    num_examples: 743
  download_size: 549763
  dataset_size: 1725867
- config_name: swahili-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 105403
    num_examples: 2072
  - name: dev
    num_bytes: 26676
    num_examples: 526
  - name: test
    num_bytes: 36137
    num_examples: 670
  download_size: 100853
  dataset_size: 168216
- config_name: telugu-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 563646332
    num_examples: 548224
  download_size: 156110911
  dataset_size: 563646332
- config_name: telugu-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 3249925
    num_examples: 117793
  - name: dev
    num_bytes: 26450
    num_examples: 983
  - name: test
    num_bytes: 18434
    num_examples: 677
  download_size: 964408
  dataset_size: 3294809
- config_name: telugu-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 446743
    num_examples: 3880
  - name: dev
    num_bytes: 112859
    num_examples: 983
  - name: test
    num_bytes: 73365
    num_examples: 646
  download_size: 253930
  dataset_size: 632967
- config_name: thai-corpus
  features:
  - name: _id
    dtype: string
  - name: title
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: dev
    num_bytes: 574630351
    num_examples: 568855
  - name: test
    num_bytes: 574630351
    num_examples: 568855
  - name: train
    num_bytes: 574630351
    num_examples: 568855
  download_size: 845293048
  dataset_size: 1723891053
- config_name: thai-qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int64
  splits:
  - name: train
    num_bytes: 2775220
    num_examples: 100472
  - name: dev
    num_bytes: 22050
    num_examples: 817
  - name: test
    num_bytes: 37576
    num_examples: 1368
  download_size: 934865
  dataset_size: 2834846
- config_name: thai-queries
  features:
  - name: _id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 443382
    num_examples: 3319
  - name: dev
    num_bytes: 108804
    num_examples: 807
  - name: test
    num_bytes: 158612
    num_examples: 1190
  download_size: 320820
  dataset_size: 710798
configs:
- config_name: arabic-corpus
  data_files:
  - split: train
    path: arabic-corpus/train-*
- config_name: arabic-qrels
  data_files:
  - split: train
    path: arabic-qrels/train-*
  - split: dev
    path: arabic-qrels/dev-*
  - split: test
    path: arabic-qrels/test-*
- config_name: arabic-queries
  data_files:
  - split: train
    path: arabic-queries/train-*
  - split: dev
    path: arabic-queries/dev-*
  - split: test
    path: arabic-queries/test-*
- config_name: bengali-corpus
  data_files:
  - split: train
    path: bengali-corpus/train-*
- config_name: bengali-qrels
  data_files:
  - split: train
    path: bengali-qrels/train-*
  - split: dev
    path: bengali-qrels/dev-*
  - split: test
    path: bengali-qrels/test-*
- config_name: bengali-queries
  data_files:
  - split: train
    path: bengali-queries/train-*
  - split: dev
    path: bengali-queries/dev-*
  - split: test
    path: bengali-queries/test-*
- config_name: english-corpus
  data_files:
  - split: train
    path: english-corpus/train-*
- config_name: english-qrels
  data_files:
  - split: train
    path: english-qrels/train-*
  - split: dev
    path: english-qrels/dev-*
  - split: test
    path: english-qrels/test-*
- config_name: english-queries
  data_files:
  - split: train
    path: english-queries/train-*
  - split: dev
    path: english-queries/dev-*
  - split: test
    path: english-queries/test-*
- config_name: finnish-corpus
  data_files:
  - split: train
    path: finnish-corpus/train-*
- config_name: finnish-qrels
  data_files:
  - split: train
    path: finnish-qrels/train-*
  - split: dev
    path: finnish-qrels/dev-*
  - split: test
    path: finnish-qrels/test-*
- config_name: finnish-queries
  data_files:
  - split: train
    path: finnish-queries/train-*
  - split: dev
    path: finnish-queries/dev-*
  - split: test
    path: finnish-queries/test-*
- config_name: indonesian-corpus
  data_files:
  - split: train
    path: indonesian-corpus/train-*
- config_name: indonesian-qrels
  data_files:
  - split: train
    path: indonesian-qrels/train-*
  - split: dev
    path: indonesian-qrels/dev-*
  - split: test
    path: indonesian-qrels/test-*
- config_name: indonesian-queries
  data_files:
  - split: train
    path: indonesian-queries/train-*
  - split: dev
    path: indonesian-queries/dev-*
  - split: test
    path: indonesian-queries/test-*
- config_name: japanese-corpus
  data_files:
  - split: train
    path: japanese-corpus/train-*
  - split: dev
    path: japanese-corpus/dev-*
  - split: test
    path: japanese-corpus/test-*
- config_name: japanese-qrels
  data_files:
  - split: train
    path: japanese-qrels/train-*
  - split: dev
    path: japanese-qrels/dev-*
  - split: test
    path: japanese-qrels/test-*
- config_name: japanese-queries
  data_files:
  - split: train
    path: japanese-queries/train-*
  - split: dev
    path: japanese-queries/dev-*
  - split: test
    path: japanese-queries/test-*
- config_name: korean-corpus
  data_files:
  - split: train
    path: korean-corpus/train-*
- config_name: korean-qrels
  data_files:
  - split: train
    path: korean-qrels/train-*
  - split: dev
    path: korean-qrels/dev-*
  - split: test
    path: korean-qrels/test-*
- config_name: korean-queries
  data_files:
  - split: train
    path: korean-queries/train-*
  - split: dev
    path: korean-queries/dev-*
  - split: test
    path: korean-queries/test-*
- config_name: russian-corpus
  data_files:
  - split: train
    path: russian-corpus/train-*
- config_name: russian-qrels
  data_files:
  - split: train
    path: russian-qrels/train-*
  - split: dev
    path: russian-qrels/dev-*
  - split: test
    path: russian-qrels/test-*
- config_name: russian-queries
  data_files:
  - split: train
    path: russian-queries/train-*
  - split: dev
    path: russian-queries/dev-*
  - split: test
    path: russian-queries/test-*
- config_name: swahili-corpus
  data_files:
  - split: train
    path: swahili-corpus/train-*
- config_name: swahili-qrels
  data_files:
  - split: train
    path: swahili-qrels/train-*
  - split: dev
    path: swahili-qrels/dev-*
  - split: test
    path: swahili-qrels/test-*
- config_name: swahili-queries
  data_files:
  - split: train
    path: swahili-queries/train-*
  - split: dev
    path: swahili-queries/dev-*
  - split: test
    path: swahili-queries/test-*
- config_name: telugu-corpus
  data_files:
  - split: train
    path: telugu-corpus/train-*
- config_name: telugu-qrels
  data_files:
  - split: train
    path: telugu-qrels/train-*
  - split: dev
    path: telugu-qrels/dev-*
  - split: test
    path: telugu-qrels/test-*
- config_name: telugu-queries
  data_files:
  - split: train
    path: telugu-queries/train-*
  - split: dev
    path: telugu-queries/dev-*
  - split: test
    path: telugu-queries/test-*
- config_name: thai-corpus
  data_files:
  - split: train
    path: thai-corpus/train-*
  - split: dev
    path: thai-corpus/dev-*
  - split: test
    path: thai-corpus/test-*
- config_name: thai-qrels
  data_files:
  - split: train
    path: thai-qrels/train-*
  - split: dev
    path: thai-qrels/dev-*
  - split: test
    path: thai-qrels/test-*
- config_name: thai-queries
  data_files:
  - split: train
    path: thai-queries/train-*
  - split: dev
    path: thai-queries/dev-*
  - split: test
    path: thai-queries/test-*
tags:
- mteb
- text
---
<!-- adapted from https://github.com/huggingface/huggingface_hub/blob/v0.30.2/src/huggingface_hub/templates/datasetcard_template.md -->

<div align="center" style="padding: 40px 20px; background-color: white; border-radius: 12px; box-shadow: 0 2px 10px rgba(0, 0, 0, 0.05); max-width: 600px; margin: 0 auto;">
  <h1 style="font-size: 3.5rem; color: #1a1a1a; margin: 0 0 20px 0; letter-spacing: 2px; font-weight: 700;">MrTidyRetrieval</h1>
  <div style="font-size: 1.5rem; color: #4a4a4a; margin-bottom: 5px; font-weight: 300;">An <a href="https://github.com/embeddings-benchmark/mteb" style="color: #2c5282; font-weight: 600; text-decoration: none;" onmouseover="this.style.textDecoration='underline'" onmouseout="this.style.textDecoration='none'">MTEB</a> dataset</div>
  <div style="font-size: 0.9rem; color: #2c5282; margin-top: 10px;">Massive Text Embedding Benchmark</div>
</div>

Mr. TyDi is a multi-lingual benchmark dataset built on TyDi, covering eleven typologically diverse languages. It is designed for monolingual retrieval, specifically to evaluate ranking with learned dense representations.

|               |                                             |
|---------------|---------------------------------------------|
| Task category | t2t                              |
| Domains       | Encyclopaedic, Written                               |
| Reference     | https://huggingface.co/datasets/castorini/mr-tydi |


## How to evaluate on this task

You can evaluate an embedding model on this dataset using the following code:

```python
import mteb

task = mteb.get_tasks(["MrTidyRetrieval"])
evaluator = mteb.MTEB(task)

model = mteb.get_model(YOUR_MODEL)
evaluator.run(model)
```

<!-- Datasets want link to arxiv in readme to autolink dataset with paper -->
To learn more about how to run models on `mteb` task check out the [GitHub repitory](https://github.com/embeddings-benchmark/mteb). 

## Citation

If you use this dataset, please cite the dataset as well as [mteb](https://github.com/embeddings-benchmark/mteb), as this dataset likely includes additional processing as a part of the [MMTEB Contribution](https://github.com/embeddings-benchmark/mteb/tree/main/docs/mmteb).

```bibtex

@article{mrtydi,
  author = {Xinyu Zhang and Xueguang Ma and Peng Shi and Jimmy Lin},
  journal = {arXiv:2108.08787},
  title = {{Mr. TyDi}: A Multi-lingual Benchmark for Dense Retrieval},
  year = {2021},
}


@article{enevoldsen2025mmtebmassivemultilingualtext,
  title={MMTEB: Massive Multilingual Text Embedding Benchmark},
  author={Kenneth Enevoldsen and Isaac Chung and Imene Kerboua and Márton Kardos and Ashwin Mathur and David Stap and Jay Gala and Wissam Siblini and Dominik Krzemiński and Genta Indra Winata and Saba Sturua and Saiteja Utpala and Mathieu Ciancone and Marion Schaeffer and Gabriel Sequeira and Diganta Misra and Shreeya Dhakal and Jonathan Rystrøm and Roman Solomatin and Ömer Çağatan and Akash Kundu and Martin Bernstorff and Shitao Xiao and Akshita Sukhlecha and Bhavish Pahwa and Rafał Poświata and Kranthi Kiran GV and Shawon Ashraf and Daniel Auras and Björn Plüster and Jan Philipp Harries and Loïc Magne and Isabelle Mohr and Mariya Hendriksen and Dawei Zhu and Hippolyte Gisserot-Boukhlef and Tom Aarsen and Jan Kostkan and Konrad Wojtasik and Taemin Lee and Marek Šuppa and Crystina Zhang and Roberta Rocca and Mohammed Hamdy and Andrianos Michail and John Yang and Manuel Faysse and Aleksei Vatolin and Nandan Thakur and Manan Dey and Dipam Vasani and Pranjal Chitale and Simone Tedeschi and Nguyen Tai and Artem Snegirev and Michael Günther and Mengzhou Xia and Weijia Shi and Xing Han Lù and Jordan Clive and Gayatri Krishnakumar and Anna Maksimova and Silvan Wehrli and Maria Tikhonova and Henil Panchal and Aleksandr Abramov and Malte Ostendorff and Zheng Liu and Simon Clematide and Lester James Miranda and Alena Fenogenova and Guangyu Song and Ruqiya Bin Safi and Wen-Ding Li and Alessia Borghini and Federico Cassano and Hongjin Su and Jimmy Lin and Howard Yen and Lasse Hansen and Sara Hooker and Chenghao Xiao and Vaibhav Adlakha and Orion Weller and Siva Reddy and Niklas Muennighoff},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2502.13595},
  year={2025},
  url={https://arxiv.org/abs/2502.13595},
  doi = {10.48550/arXiv.2502.13595},
}

@article{muennighoff2022mteb,
  author = {Muennighoff, Niklas and Tazi, Nouamane and Magne, Lo{\"\i}c and Reimers, Nils},
  title = {MTEB: Massive Text Embedding Benchmark},
  publisher = {arXiv},
  journal={arXiv preprint arXiv:2210.07316},
  year = {2022}
  url = {https://arxiv.org/abs/2210.07316},
  doi = {10.48550/ARXIV.2210.07316},
}
```

# Dataset Statistics
<details>
  <summary> Dataset Statistics</summary>

The following code contains the descriptive statistics from the task. These can also be obtained using:

```python
import mteb

task = mteb.get_task("MrTidyRetrieval")

desc_stats = task.metadata.descriptive_stats
```

```json
{
    "test": {
        "num_samples": 58051987,
        "number_of_characters": 20129435459,
        "num_documents": 58043326,
        "min_document_length": 2,
        "average_document_length": 346.79462777512094,
        "max_document_length": 61659,
        "unique_documents": 58043326,
        "num_queries": 8661,
        "min_query_length": 6,
        "average_query_length": 37.15783396836393,
        "max_query_length": 144,
        "unique_queries": 8661,
        "none_queries": 0,
        "num_relevant_docs": 10105,
        "min_relevant_docs_per_query": 1,
        "average_relevant_docs_per_query": 1.1667243967209329,
        "max_relevant_docs_per_query": 3,
        "unique_relevant_docs": 8926,
        "num_instructions": null,
        "min_instruction_length": null,
        "average_instruction_length": null,
        "max_instruction_length": null,
        "unique_instructions": null,
        "num_top_ranked": null,
        "min_top_ranked_per_query": null,
        "average_top_ranked_per_query": null,
        "max_top_ranked_per_query": null
    }
}
```

</details>

---
*This dataset card was automatically generated using [MTEB](https://github.com/embeddings-benchmark/mteb)*