Hugging Face's logo

Datasets:
hakari-bench
/
NanoMIRACL

Modalities:
Text
Formats:
parquet
Languages:
Arabic
Bengali
German
Size:
100K - 1M
Tags:
information-retrieval
retrieval
nano
bm25
hakari-bench
dense-retrieval
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
NanoMIRACL / README.md
hotchpotch's picture
hotchpotch
Improve Nano dataset README provenance and overview
ea97b8c verified
preview code
|
Raw
History Blame Contribute Delete
20.7 kB
metadata 
configs:
  - config_name: corpus
    data_files:
      - split: ar
        path: corpus/ar-00000-of-00001.parquet
      - split: bn
        path: corpus/bn-00000-of-00001.parquet
      - split: de
        path: corpus/de-00000-of-00001.parquet
      - split: en
        path: corpus/en-00000-of-00001.parquet
      - split: es
        path: corpus/es-00000-of-00001.parquet
      - split: fa
        path: corpus/fa-00000-of-00001.parquet
      - split: fi
        path: corpus/fi-00000-of-00001.parquet
      - split: fr
        path: corpus/fr-00000-of-00001.parquet
      - split: hi
        path: corpus/hi-00000-of-00001.parquet
      - split: id
        path: corpus/id-00000-of-00001.parquet
      - split: ja
        path: corpus/ja-00000-of-00001.parquet
      - split: ko
        path: corpus/ko-00000-of-00001.parquet
      - split: ru
        path: corpus/ru-00000-of-00001.parquet
      - split: sw
        path: corpus/sw-00000-of-00001.parquet
      - split: te
        path: corpus/te-00000-of-00001.parquet
      - split: th
        path: corpus/th-00000-of-00001.parquet
      - split: yo
        path: corpus/yo-00000-of-00001.parquet
      - split: zh
        path: corpus/zh-00000-of-00001.parquet
  - config_name: queries
    data_files:
      - split: ar
        path: queries/ar-00000-of-00001.parquet
      - split: bn
        path: queries/bn-00000-of-00001.parquet
      - split: de
        path: queries/de-00000-of-00001.parquet
      - split: en
        path: queries/en-00000-of-00001.parquet
      - split: es
        path: queries/es-00000-of-00001.parquet
      - split: fa
        path: queries/fa-00000-of-00001.parquet
      - split: fi
        path: queries/fi-00000-of-00001.parquet
      - split: fr
        path: queries/fr-00000-of-00001.parquet
      - split: hi
        path: queries/hi-00000-of-00001.parquet
      - split: id
        path: queries/id-00000-of-00001.parquet
      - split: ja
        path: queries/ja-00000-of-00001.parquet
      - split: ko
        path: queries/ko-00000-of-00001.parquet
      - split: ru
        path: queries/ru-00000-of-00001.parquet
      - split: sw
        path: queries/sw-00000-of-00001.parquet
      - split: te
        path: queries/te-00000-of-00001.parquet
      - split: th
        path: queries/th-00000-of-00001.parquet
      - split: yo
        path: queries/yo-00000-of-00001.parquet
      - split: zh
        path: queries/zh-00000-of-00001.parquet
    default: true
  - config_name: qrels
    data_files:
      - split: ar
        path: qrels/ar-00000-of-00001.parquet
      - split: bn
        path: qrels/bn-00000-of-00001.parquet
      - split: de
        path: qrels/de-00000-of-00001.parquet
      - split: en
        path: qrels/en-00000-of-00001.parquet
      - split: es
        path: qrels/es-00000-of-00001.parquet
      - split: fa
        path: qrels/fa-00000-of-00001.parquet
      - split: fi
        path: qrels/fi-00000-of-00001.parquet
      - split: fr
        path: qrels/fr-00000-of-00001.parquet
      - split: hi
        path: qrels/hi-00000-of-00001.parquet
      - split: id
        path: qrels/id-00000-of-00001.parquet
      - split: ja
        path: qrels/ja-00000-of-00001.parquet
      - split: ko
        path: qrels/ko-00000-of-00001.parquet
      - split: ru
        path: qrels/ru-00000-of-00001.parquet
      - split: sw
        path: qrels/sw-00000-of-00001.parquet
      - split: te
        path: qrels/te-00000-of-00001.parquet
      - split: th
        path: qrels/th-00000-of-00001.parquet
      - split: yo
        path: qrels/yo-00000-of-00001.parquet
      - split: zh
        path: qrels/zh-00000-of-00001.parquet
  - config_name: bm25
    data_files:
      - split: ar
        path: bm25/ar-00000-of-00001.parquet
      - split: bn
        path: bm25/bn-00000-of-00001.parquet
      - split: de
        path: bm25/de-00000-of-00001.parquet
      - split: en
        path: bm25/en-00000-of-00001.parquet
      - split: es
        path: bm25/es-00000-of-00001.parquet
      - split: fa
        path: bm25/fa-00000-of-00001.parquet
      - split: fi
        path: bm25/fi-00000-of-00001.parquet
      - split: fr
        path: bm25/fr-00000-of-00001.parquet
      - split: hi
        path: bm25/hi-00000-of-00001.parquet
      - split: id
        path: bm25/id-00000-of-00001.parquet
      - split: ja
        path: bm25/ja-00000-of-00001.parquet
      - split: ko
        path: bm25/ko-00000-of-00001.parquet
      - split: ru
        path: bm25/ru-00000-of-00001.parquet
      - split: sw
        path: bm25/sw-00000-of-00001.parquet
      - split: te
        path: bm25/te-00000-of-00001.parquet
      - split: th
        path: bm25/th-00000-of-00001.parquet
      - split: yo
        path: bm25/yo-00000-of-00001.parquet
      - split: zh
        path: bm25/zh-00000-of-00001.parquet
  - config_name: harrier_oss_v1_270m
    data_files:
      - split: ar
        path: harrier_oss_v1_270m/ar-00000-of-00001.parquet
      - split: bn
        path: harrier_oss_v1_270m/bn-00000-of-00001.parquet
      - split: de
        path: harrier_oss_v1_270m/de-00000-of-00001.parquet
      - split: en
        path: harrier_oss_v1_270m/en-00000-of-00001.parquet
      - split: es
        path: harrier_oss_v1_270m/es-00000-of-00001.parquet
      - split: fa
        path: harrier_oss_v1_270m/fa-00000-of-00001.parquet
      - split: fi
        path: harrier_oss_v1_270m/fi-00000-of-00001.parquet
      - split: fr
        path: harrier_oss_v1_270m/fr-00000-of-00001.parquet
      - split: hi
        path: harrier_oss_v1_270m/hi-00000-of-00001.parquet
      - split: id
        path: harrier_oss_v1_270m/id-00000-of-00001.parquet
      - split: ja
        path: harrier_oss_v1_270m/ja-00000-of-00001.parquet
      - split: ko
        path: harrier_oss_v1_270m/ko-00000-of-00001.parquet
      - split: ru
        path: harrier_oss_v1_270m/ru-00000-of-00001.parquet
      - split: sw
        path: harrier_oss_v1_270m/sw-00000-of-00001.parquet
      - split: te
        path: harrier_oss_v1_270m/te-00000-of-00001.parquet
      - split: th
        path: harrier_oss_v1_270m/th-00000-of-00001.parquet
      - split: yo
        path: harrier_oss_v1_270m/yo-00000-of-00001.parquet
      - split: zh
        path: harrier_oss_v1_270m/zh-00000-of-00001.parquet
  - config_name: reranking_hybrid
    data_files:
      - split: ar
        path: reranking_hybrid/ar-00000-of-00001.parquet
      - split: bn
        path: reranking_hybrid/bn-00000-of-00001.parquet
      - split: de
        path: reranking_hybrid/de-00000-of-00001.parquet
      - split: en
        path: reranking_hybrid/en-00000-of-00001.parquet
      - split: es
        path: reranking_hybrid/es-00000-of-00001.parquet
      - split: fa
        path: reranking_hybrid/fa-00000-of-00001.parquet
      - split: fi
        path: reranking_hybrid/fi-00000-of-00001.parquet
      - split: fr
        path: reranking_hybrid/fr-00000-of-00001.parquet
      - split: hi
        path: reranking_hybrid/hi-00000-of-00001.parquet
      - split: id
        path: reranking_hybrid/id-00000-of-00001.parquet
      - split: ja
        path: reranking_hybrid/ja-00000-of-00001.parquet
      - split: ko
        path: reranking_hybrid/ko-00000-of-00001.parquet
      - split: ru
        path: reranking_hybrid/ru-00000-of-00001.parquet
      - split: sw
        path: reranking_hybrid/sw-00000-of-00001.parquet
      - split: te
        path: reranking_hybrid/te-00000-of-00001.parquet
      - split: th
        path: reranking_hybrid/th-00000-of-00001.parquet
      - split: yo
        path: reranking_hybrid/yo-00000-of-00001.parquet
      - split: zh
        path: reranking_hybrid/zh-00000-of-00001.parquet
language:
  - ar
  - bn
  - de
  - en
  - es
  - fa
  - fi
  - fr
  - hi
  - id
  - ja
  - ko
  - ru
  - sw
  - te
  - th
  - yo
  - zh
tags:
  - information-retrieval
  - retrieval
  - nano
  - bm25
  - hakari-bench
  - dense-retrieval
  - reranking
dataset_info:
  - config_name: bm25
    features:
      - name: query-id
        dtype: string
      - name: corpus-ids
        list: string
    splits:
      - name: ar
        num_bytes: 1266611
        num_examples: 200
      - name: bn
        num_bytes: 1177519
        num_examples: 200
      - name: de
        num_bytes: 1298603
        num_examples: 200
      - name: en
        num_bytes: 1357970
        num_examples: 200
      - name: es
        num_bytes: 1288433
        num_examples: 200
      - name: fa
        num_bytes: 1262738
        num_examples: 200
      - name: fi
        num_bytes: 1218233
        num_examples: 200
      - name: fr
        num_bytes: 1309162
        num_examples: 200
      - name: hi
        num_bytes: 1208415
        num_examples: 200
      - name: id
        num_bytes: 1218449
        num_examples: 200
      - name: ja
        num_bytes: 1269008
        num_examples: 200
      - name: ko
        num_bytes: 1218171
        num_examples: 200
      - name: ru
        num_bytes: 1292152
        num_examples: 200
      - name: sw
        num_bytes: 1108061
        num_examples: 200
      - name: te
        num_bytes: 1161605
        num_examples: 200
      - name: th
        num_bytes: 1189537
        num_examples: 200
      - name: yo
        num_bytes: 657193
        num_examples: 119
      - name: zh
        num_bytes: 1263280
        num_examples: 200
    download_size: 21789444
    dataset_size: 21765140
  - config_name: corpus
    features:
      - name: _id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: ar
        num_bytes: 7174282
        num_examples: 10000
      - name: bn
        num_bytes: 12035077
        num_examples: 10000
      - name: de
        num_bytes: 4819401
        num_examples: 10000
      - name: en
        num_bytes: 4904869
        num_examples: 10000
      - name: es
        num_bytes: 4794294
        num_examples: 10000
      - name: fa
        num_bytes: 5768565
        num_examples: 10000
      - name: fi
        num_bytes: 4233118
        num_examples: 10000
      - name: fr
        num_bytes: 4151659
        num_examples: 10000
      - name: hi
        num_bytes: 10878177
        num_examples: 10000
      - name: id
        num_bytes: 4336884
        num_examples: 10000
      - name: ja
        num_bytes: 4982205
        num_examples: 10000
      - name: ko
        num_bytes: 4938423
        num_examples: 10000
      - name: ru
        num_bytes: 7807671
        num_examples: 10000
      - name: sw
        num_bytes: 2937176
        num_examples: 10000
      - name: te
        num_bytes: 10843968
        num_examples: 10000
      - name: th
        num_bytes: 11582202
        num_examples: 10000
      - name: yo
        num_bytes: 2330437
        num_examples: 10000
      - name: zh
        num_bytes: 3801244
        num_examples: 10000
    download_size: 59148256
    dataset_size: 112319652
  - config_name: harrier_oss_v1_270m
    features:
      - name: query-id
        dtype: string
      - name: corpus-ids
        list: string
    splits:
      - name: ar
        num_bytes: 1273629
        num_examples: 200
      - name: bn
        num_bytes: 1176065
        num_examples: 200
      - name: de
        num_bytes: 1301028
        num_examples: 200
      - name: en
        num_bytes: 1363247
        num_examples: 200
      - name: es
        num_bytes: 1294926
        num_examples: 200
      - name: fa
        num_bytes: 1262210
        num_examples: 200
      - name: fi
        num_bytes: 1218495
        num_examples: 200
      - name: fr
        num_bytes: 1311982
        num_examples: 200
      - name: hi
        num_bytes: 1207316
        num_examples: 200
      - name: id
        num_bytes: 1224728
        num_examples: 200
      - name: ja
        num_bytes: 1272296
        num_examples: 200
      - name: ko
        num_bytes: 1217697
        num_examples: 200
      - name: ru
        num_bytes: 1293929
        num_examples: 200
      - name: sw
        num_bytes: 1110175
        num_examples: 200
      - name: te
        num_bytes: 1159682
        num_examples: 200
      - name: th
        num_bytes: 1196182
        num_examples: 200
      - name: yo
        num_bytes: 655592
        num_examples: 119
      - name: zh
        num_bytes: 1268069
        num_examples: 200
    download_size: 21831454
    dataset_size: 21807248
  - config_name: qrels
    features:
      - name: query-id
        dtype: string
      - name: corpus-id
        dtype: string
    splits:
      - name: ar
        num_bytes: 7267
        num_examples: 386
      - name: bn
        num_bytes: 7436
        num_examples: 407
      - name: de
        num_bytes: 13650
        num_examples: 538
      - name: en
        num_bytes: 11324
        num_examples: 560
      - name: es
        num_bytes: 23740
        num_examples: 934
      - name: fa
        num_bytes: 10577
        num_examples: 427
      - name: fi
        num_bytes: 6081
        num_examples: 328
      - name: fr
        num_bytes: 10612
        num_examples: 417
      - name: hi
        num_bytes: 9654
        num_examples: 410
      - name: id
        num_bytes: 12133
        num_examples: 654
      - name: ja
        num_bytes: 7131
        num_examples: 373
      - name: ko
        num_bytes: 9572
        num_examples: 508
      - name: ru
        num_bytes: 10750
        num_examples: 555
      - name: sw
        num_bytes: 7188
        num_examples: 405
      - name: te
        num_bytes: 3855
        num_examples: 211
      - name: th
        num_bytes: 6309
        num_examples: 343
      - name: yo
        num_bytes: 3143
        num_examples: 144
      - name: zh
        num_bytes: 11666
        num_examples: 471
    download_size: 122826
    dataset_size: 172088
  - config_name: queries
    features:
      - name: _id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: ar
        num_bytes: 13271
        num_examples: 200
      - name: bn
        num_bytes: 27235
        num_examples: 200
      - name: de
        num_bytes: 12632
        num_examples: 200
      - name: en
        num_bytes: 10199
        num_examples: 200
      - name: es
        num_bytes: 13486
        num_examples: 200
      - name: fa
        num_bytes: 17958
        num_examples: 200
      - name: fi
        num_bytes: 9852
        num_examples: 200
      - name: fr
        num_bytes: 12361
        num_examples: 200
      - name: hi
        num_bytes: 31659
        num_examples: 200
      - name: id
        num_bytes: 9871
        num_examples: 200
      - name: ja
        num_bytes: 12505
        num_examples: 200
      - name: ko
        num_bytes: 13045
        num_examples: 200
      - name: ru
        num_bytes: 18950
        num_examples: 200
      - name: sw
        num_bytes: 9853
        num_examples: 200
      - name: te
        num_bytes: 22938
        num_examples: 200
      - name: th
        num_bytes: 27216
        num_examples: 200
      - name: yo
        num_bytes: 6697
        num_examples: 119
      - name: zh
        num_bytes: 9826
        num_examples: 200
    download_size: 183744
    dataset_size: 279554
  - config_name: reranking_hybrid
    features:
      - name: query-id
        dtype: string
      - name: corpus-ids
        list: string
    splits:
      - name: ar
        num_bytes: 253968
        num_examples: 200
      - name: bn
        num_bytes: 236250
        num_examples: 200
      - name: de
        num_bytes: 261001
        num_examples: 200
      - name: en
        num_bytes: 272500
        num_examples: 200
      - name: es
        num_bytes: 259586
        num_examples: 200
      - name: fa
        num_bytes: 253821
        num_examples: 200
      - name: fi
        num_bytes: 244186
        num_examples: 200
      - name: fr
        num_bytes: 263592
        num_examples: 200
      - name: hi
        num_bytes: 242989
        num_examples: 200
      - name: id
        num_bytes: 242548
        num_examples: 200
      - name: ja
        num_bytes: 254284
        num_examples: 200
      - name: ko
        num_bytes: 244095
        num_examples: 200
      - name: ru
        num_bytes: 259305
        num_examples: 200
      - name: sw
        num_bytes: 223021
        num_examples: 200
      - name: te
        num_bytes: 234272
        num_examples: 200
      - name: th
        num_bytes: 239389
        num_examples: 200
      - name: yo
        num_bytes: 132452
        num_examples: 119
      - name: zh
        num_bytes: 254917
        num_examples: 200
    download_size: 4393763
    dataset_size: 4372176

NanoMIRACL

This dataset is a Nano-style retrieval dataset for HAKARI-bench.

NanoMIRACL is a compact multilingual benchmark derived from MIRACL. Each language split evaluates monolingual retrieval of Wikipedia passages for natural-language questions.

This rebuild uses hotchpotch/miracl-hf-unified dev queries at source revision 21ad00eb467639e927b5badb7c49f4947c6c24ca. For each sampled query, it preserves all source positive passages and expands the split-local corpus to 10,000 documents using source negatives as hard-negative candidates plus deterministic random fill.

Usage 

from datasets import load_dataset

dataset_id = "hakari-bench/NanoMIRACL"
split = "ar"

queries = load_dataset(dataset_id, "queries", split=split)
corpus = load_dataset(dataset_id, "corpus", split=split)
qrels = load_dataset(dataset_id, "qrels", split=split)
reranking_candidates = load_dataset(dataset_id, "reranking_hybrid", split=split)

Data Layout

This dataset uses six Hugging Face Datasets configs:

  • corpus: documents with _id and text
  • queries: queries with _id and text
  • qrels: positive relevance labels with query-id and corpus-id
  • bm25: BM25 candidate lists with query-id and corpus-ids
  • harrier_oss_v1_270m: dense candidate lists from microsoft/harrier-oss-v1-270m
  • reranking_hybrid: RRF candidate lists built from bm25 and harrier_oss_v1_270m

Each config has the same Nano split names.

Candidate Construction

  • bm25: local BM25 top-500 with automatic language-aware tokenization. The resolved tokenizer is shown in the Candidate Quality table, for example wordseg@ja.
  • harrier_oss_v1_270m: dense top-500 from microsoft/harrier-oss-v1-270m. In tables this is shown as Dense; Dense means microsoft/harrier-oss-v1-270m with the web_search_query prompt for queries and cosine similarity over normalized embeddings.
  • reranking_hybrid: RRF over bm25 and harrier_oss_v1_270m using rrf_k=100, keeping the RRF top-100.

Safeguard means rank 101 is appended only when RRF top-100 contains no qrels-positive document.

Split Statistics

Length statistics are character counts computed with len(str(text)).

Nano split Queries Corpus Qrels Query chars avg Query chars p50 Query chars p75 Doc chars avg Doc chars p50 Doc chars p75
ar 200 10000 386 30.1 27.0 35.0 392.3 276.0 520.0
bn 200 10000 407 47.2 43.5 58.0 446.2 359.0 597.0
de 200 10000 538 45.4 43.0 52.0 457.2 372.0 612.0
en 200 10000 560 39.9 37.5 48.0 471.8 392.0 645.0
es 200 10000 934 47.6 47.0 55.0 453.2 370.0 597.0
fa 200 10000 427 40.0 38.0 46.0 310.7 213.0 413.0
fi 200 10000 328 37.2 33.5 46.2 393.6 326.0 522.0
fr 200 10000 417 43.3 43.0 51.0 385.3 296.0 513.0
hi 200 10000 410 54.8 52.0 65.0 419.3 320.0 559.0
id 200 10000 654 38.3 37.0 45.0 416.5 316.0 567.0
ja 200 10000 373 17.5 17.0 21.0 173.4 135.0 231.0
ko 200 10000 508 21.7 19.0 22.0 205.3 163.0 273.0
ru 200 10000 555 45.5 41.0 53.5 423.3 313.0 569.0
sw 200 10000 405 38.3 37.0 44.0 278.0 204.0 338.0
te 200 10000 211 38.4 36.5 43.0 409.0 295.0 486.0
th 200 10000 343 43.6 40.0 50.2 409.9 346.0 538.0
yo 119 10000 144 37.7 36.0 42.0 176.7 81.0 187.0
zh 200 10000 471 10.9 11.0 12.0 133.4 105.0 176.0

Candidate Quality

nDCG@10 and Recall@100 are computed from the included candidate rankings against the included qrels, then reported as 0-100 scores such as 52.45. Recall@100 uses only the top 100 candidates; an optional rank-101 safeguard positive is not counted in Recall@100.

Dense means microsoft/harrier-oss-v1-270m with the web_search_query prompt and cosine similarity.

Nano split BM25 tokenizer BM25 nDCG@10 Dense nDCG@10 Hybrid nDCG@10 BM25 Recall@100 Dense Recall@100 Hybrid Recall@100 Hybrid candidates Safeguard positives
Mean - 57.15 76.27 69.00 93.32 94.02 99.26 - 19
ar stemmer@arabic 63.52 82.23 75.14 97.40 96.58 99.83 100 0
bn whitespace 50.33 76.61 65.37 95.17 95.57 99.75 100 0
de stemmer@german 51.72 73.89 64.18 91.78 93.10 98.27 100-101 1
en english_porter_stop 67.74 77.21 74.74 99.38 95.00 99.75 100 0
es stemmer@spanish 68.61 77.93 74.78 98.06 93.31 100.00 100 0
fa regex 57.88 64.76 63.34 96.22 88.99 99.30 100-101 1
fi stemmer@finnish 77.34 86.34 83.32 98.58 95.42 100.00 100 0
fr stemmer@french 46.58 68.28 58.96 97.96 91.35 99.88 100 0
hi stemmer@hindi 30.37 68.47 51.74 70.18 91.99 96.00 100-101 7
id stemmer@indonesian 67.73 70.76 71.71 98.01 95.04 99.83 100 0
ja wordseg@ja 66.01 77.45 72.23 97.33 91.78 100.00 100 0
ko wordseg@ko 49.94 69.10 70.26 95.74 92.02 98.12 100-101 3
ru stemmer@russian 58.87 76.93 68.16 89.50 93.20 98.43 100-101 3
sw regex 58.52 78.72 72.92 94.89 94.95 99.75 100 0
te whitespace 52.92 87.20 69.53 87.92 92.50 98.25 100-101 3
th wordseg@th 62.29 81.01 72.96 95.58 95.25 99.50 100-101 1
yo regex 58.16 84.16 76.51 92.44 97.48 100.00 100 0
zh wordseg@zh 40.22 71.91 56.19 83.62 98.90 100.00 100 0

Hybrid Safeguard Summary

  • Safeguard positives: 19
  • Rows limited by corpus size: 0
  • Metadata file: reranking_hybrid_metadata.json

Source Links

License

NanoMIRACL is a derived dataset. Users must comply with the licenses, terms, and attribution requirements of the upstream source datasets.