NanoRARb / README.md
hotchpotch's picture
Update dataset README for reranking_hybrid candidates
017849a verified
|
Raw
History Blame Contribute Delete
25.1 kB
---
configs:
- config_name: corpus
data_files:
- split: NanoARCChallenge
path: corpus/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: corpus/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: corpus/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: corpus/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: corpus/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: corpus/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: corpus/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: corpus/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: corpus/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: corpus/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: corpus/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: corpus/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: corpus/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: corpus/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: corpus/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: corpus/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: corpus/NanoWinoGrande-00000-of-00001.parquet
- config_name: queries
data_files:
- split: NanoARCChallenge
path: queries/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: queries/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: queries/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: queries/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: queries/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: queries/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: queries/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: queries/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: queries/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: queries/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: queries/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: queries/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: queries/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: queries/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: queries/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: queries/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: queries/NanoWinoGrande-00000-of-00001.parquet
default: true
- config_name: qrels
data_files:
- split: NanoARCChallenge
path: qrels/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: qrels/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: qrels/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: qrels/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: qrels/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: qrels/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: qrels/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: qrels/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: qrels/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: qrels/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: qrels/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: qrels/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: qrels/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: qrels/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: qrels/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: qrels/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: qrels/NanoWinoGrande-00000-of-00001.parquet
- config_name: bm25
data_files:
- split: NanoARCChallenge
path: bm25/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: bm25/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: bm25/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: bm25/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: bm25/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: bm25/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: bm25/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: bm25/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: bm25/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: bm25/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: bm25/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: bm25/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: bm25/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: bm25/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: bm25/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: bm25/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: bm25/NanoWinoGrande-00000-of-00001.parquet
- config_name: harrier_oss_v1_270m
data_files:
- split: NanoARCChallenge
path: harrier_oss_v1_270m/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: harrier_oss_v1_270m/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: harrier_oss_v1_270m/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: harrier_oss_v1_270m/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: harrier_oss_v1_270m/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: harrier_oss_v1_270m/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: harrier_oss_v1_270m/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: harrier_oss_v1_270m/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: harrier_oss_v1_270m/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: harrier_oss_v1_270m/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: harrier_oss_v1_270m/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: harrier_oss_v1_270m/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: harrier_oss_v1_270m/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: harrier_oss_v1_270m/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: harrier_oss_v1_270m/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: harrier_oss_v1_270m/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: harrier_oss_v1_270m/NanoWinoGrande-00000-of-00001.parquet
- config_name: reranking_hybrid
data_files:
- split: NanoARCChallenge
path: reranking_hybrid/NanoARCChallenge-00000-of-00001.parquet
- split: NanoAlphaNLI
path: reranking_hybrid/NanoAlphaNLI-00000-of-00001.parquet
- split: NanoHellaSwag
path: reranking_hybrid/NanoHellaSwag-00000-of-00001.parquet
- split: NanoPIQA
path: reranking_hybrid/NanoPIQA-00000-of-00001.parquet
- split: NanoQuail
path: reranking_hybrid/NanoQuail-00000-of-00001.parquet
- split: NanoRARbCode
path: reranking_hybrid/NanoRARbCode-00000-of-00001.parquet
- split: NanoRARbMath
path: reranking_hybrid/NanoRARbMath-00000-of-00001.parquet
- split: NanoSIQA
path: reranking_hybrid/NanoSIQA-00000-of-00001.parquet
- split: NanoSpartQA
path: reranking_hybrid/NanoSpartQA-00000-of-00001.parquet
- split: NanoTempReasonL1
path: reranking_hybrid/NanoTempReasonL1-00000-of-00001.parquet
- split: NanoTempReasonL2Context
path: reranking_hybrid/NanoTempReasonL2Context-00000-of-00001.parquet
- split: NanoTempReasonL2Fact
path: reranking_hybrid/NanoTempReasonL2Fact-00000-of-00001.parquet
- split: NanoTempReasonL2Pure
path: reranking_hybrid/NanoTempReasonL2Pure-00000-of-00001.parquet
- split: NanoTempReasonL3Context
path: reranking_hybrid/NanoTempReasonL3Context-00000-of-00001.parquet
- split: NanoTempReasonL3Fact
path: reranking_hybrid/NanoTempReasonL3Fact-00000-of-00001.parquet
- split: NanoTempReasonL3Pure
path: reranking_hybrid/NanoTempReasonL3Pure-00000-of-00001.parquet
- split: NanoWinoGrande
path: reranking_hybrid/NanoWinoGrande-00000-of-00001.parquet
language:
- en
tags:
- information-retrieval
- retrieval
- nano
- bm25
- dense-retrieval
- reranking
- hakari-bench
dataset_info:
- config_name: bm25
features:
- name: query-id
dtype: string
- name: corpus-ids
list: string
splits:
- name: NanoARCChallenge
num_bytes: 2397141
num_examples: 200
- name: NanoAlphaNLI
num_bytes: 1592820
num_examples: 200
- name: NanoHellaSwag
num_bytes: 1992265
num_examples: 200
- name: NanoPIQA
num_bytes: 1495378
num_examples: 200
- name: NanoQuail
num_bytes: 1585660
num_examples: 200
- name: NanoRARbCode
num_bytes: 1655666
num_examples: 200
- name: NanoRARbMath
num_bytes: 1566078
num_examples: 200
- name: NanoSIQA
num_bytes: 1493669
num_examples: 200
- name: NanoSpartQA
num_bytes: 1720733
num_examples: 200
- name: NanoTempReasonL1
num_bytes: 1394908
num_examples: 200
- name: NanoTempReasonL2Context
num_bytes: 1388048
num_examples: 200
- name: NanoTempReasonL2Fact
num_bytes: 1386249
num_examples: 200
- name: NanoTempReasonL2Pure
num_bytes: 1400672
num_examples: 200
- name: NanoTempReasonL3Context
num_bytes: 1388349
num_examples: 200
- name: NanoTempReasonL3Fact
num_bytes: 1386720
num_examples: 200
- name: NanoTempReasonL3Pure
num_bytes: 1396035
num_examples: 200
- name: NanoWinoGrande
num_bytes: 1502057
num_examples: 200
download_size: 26744275
dataset_size: 26742448
- config_name: corpus
features:
- name: _id
dtype: string
- name: text
dtype: string
splits:
- name: NanoARCChallenge
num_bytes: 550089
num_examples: 9350
- name: NanoAlphaNLI
num_bytes: 637339
num_examples: 10000
- name: NanoHellaSwag
num_bytes: 860409
num_examples: 10000
- name: NanoPIQA
num_bytes: 1169438
num_examples: 10000
- name: NanoQuail
num_bytes: 449092
num_examples: 10000
- name: NanoRARbCode
num_bytes: 2756968
num_examples: 10000
- name: NanoRARbMath
num_bytes: 5023939
num_examples: 10000
- name: NanoSIQA
num_bytes: 403986
num_examples: 10000
- name: NanoSpartQA
num_bytes: 113191
num_examples: 1592
- name: NanoTempReasonL1
num_bytes: 268865
num_examples: 10000
- name: NanoTempReasonL2Context
num_bytes: 380065
num_examples: 10000
- name: NanoTempReasonL2Fact
num_bytes: 380065
num_examples: 10000
- name: NanoTempReasonL2Pure
num_bytes: 380065
num_examples: 10000
- name: NanoTempReasonL3Context
num_bytes: 379782
num_examples: 10000
- name: NanoTempReasonL3Fact
num_bytes: 379782
num_examples: 10000
- name: NanoTempReasonL3Pure
num_bytes: 379782
num_examples: 10000
- name: NanoWinoGrande
num_bytes: 134837
num_examples: 5095
download_size: 7668654
dataset_size: 14647694
- config_name: harrier_oss_v1_270m
features:
- name: query-id
dtype: string
- name: corpus-ids
list: string
splits:
- name: NanoARCChallenge
num_bytes: 2393245
num_examples: 200
- name: NanoAlphaNLI
num_bytes: 1591628
num_examples: 200
- name: NanoHellaSwag
num_bytes: 1992262
num_examples: 200
- name: NanoPIQA
num_bytes: 1491910
num_examples: 200
- name: NanoQuail
num_bytes: 1580730
num_examples: 200
- name: NanoRARbCode
num_bytes: 1510235
num_examples: 200
- name: NanoRARbMath
num_bytes: 1501950
num_examples: 200
- name: NanoSIQA
num_bytes: 1493251
num_examples: 200
- name: NanoSpartQA
num_bytes: 1737541
num_examples: 200
- name: NanoTempReasonL1
num_bytes: 1391940
num_examples: 200
- name: NanoTempReasonL2Context
num_bytes: 1387736
num_examples: 200
- name: NanoTempReasonL2Fact
num_bytes: 1388225
num_examples: 200
- name: NanoTempReasonL2Pure
num_bytes: 1388790
num_examples: 200
- name: NanoTempReasonL3Context
num_bytes: 1388367
num_examples: 200
- name: NanoTempReasonL3Fact
num_bytes: 1388985
num_examples: 200
- name: NanoTempReasonL3Pure
num_bytes: 1388855
num_examples: 200
- name: NanoWinoGrande
num_bytes: 1480614
num_examples: 200
download_size: 26498499
dataset_size: 26496264
- config_name: qrels
features:
- name: query-id
dtype: string
- name: corpus-id
dtype: string
splits:
- name: NanoARCChallenge
num_bytes: 9063
num_examples: 200
- name: NanoAlphaNLI
num_bytes: 5835
num_examples: 200
- name: NanoHellaSwag
num_bytes: 7463
num_examples: 200
- name: NanoPIQA
num_bytes: 5435
num_examples: 200
- name: NanoQuail
num_bytes: 5810
num_examples: 200
- name: NanoRARbCode
num_bytes: 8980
num_examples: 200
- name: NanoRARbMath
num_bytes: 5380
num_examples: 200
- name: NanoSIQA
num_bytes: 5450
num_examples: 200
- name: NanoSpartQA
num_bytes: 12383
num_examples: 384
- name: NanoTempReasonL1
num_bytes: 4979
num_examples: 200
- name: NanoTempReasonL2Context
num_bytes: 4975
num_examples: 200
- name: NanoTempReasonL2Fact
num_bytes: 4975
num_examples: 200
- name: NanoTempReasonL2Pure
num_bytes: 4975
num_examples: 200
- name: NanoTempReasonL3Context
num_bytes: 5004
num_examples: 200
- name: NanoTempReasonL3Fact
num_bytes: 5004
num_examples: 200
- name: NanoTempReasonL3Pure
num_bytes: 5004
num_examples: 200
- name: NanoWinoGrande
num_bytes: 5343
num_examples: 200
download_size: 60606
dataset_size: 106058
- config_name: queries
features:
- name: _id
dtype: string
- name: text
dtype: string
splits:
- name: NanoARCChallenge
num_bytes: 30628
num_examples: 200
- name: NanoAlphaNLI
num_bytes: 24448
num_examples: 200
- name: NanoHellaSwag
num_bytes: 27426
num_examples: 200
- name: NanoPIQA
num_bytes: 11068
num_examples: 200
- name: NanoQuail
num_bytes: 366441
num_examples: 200
- name: NanoRARbCode
num_bytes: 99388
num_examples: 200
- name: NanoRARbMath
num_bytes: 43755
num_examples: 200
- name: NanoSIQA
num_bytes: 28878
num_examples: 200
- name: NanoSpartQA
num_bytes: 135064
num_examples: 200
- name: NanoTempReasonL1
num_bytes: 13265
num_examples: 200
- name: NanoTempReasonL2Context
num_bytes: 5788304
num_examples: 200
- name: NanoTempReasonL2Fact
num_bytes: 354146
num_examples: 200
- name: NanoTempReasonL2Pure
num_bytes: 13905
num_examples: 200
- name: NanoTempReasonL3Context
num_bytes: 6401050
num_examples: 200
- name: NanoTempReasonL3Fact
num_bytes: 401694
num_examples: 200
- name: NanoTempReasonL3Pure
num_bytes: 16394
num_examples: 200
- name: NanoWinoGrande
num_bytes: 25885
num_examples: 200
download_size: 6772280
dataset_size: 13781739
- config_name: reranking_hybrid
features:
- name: query-id
dtype: string
- name: corpus-ids
list: string
splits:
- name: NanoARCChallenge
num_bytes: 485686
num_examples: 200
- name: NanoAlphaNLI
num_bytes: 321131
num_examples: 200
- name: NanoHellaSwag
num_bytes: 403060
num_examples: 200
- name: NanoPIQA
num_bytes: 301849
num_examples: 200
- name: NanoQuail
num_bytes: 315624
num_examples: 200
- name: NanoRARbCode
num_bytes: 318915
num_examples: 200
- name: NanoRARbMath
num_bytes: 298826
num_examples: 200
- name: NanoSIQA
num_bytes: 303240
num_examples: 200
- name: NanoSpartQA
num_bytes: 345385
num_examples: 200
- name: NanoTempReasonL1
num_bytes: 282785
num_examples: 200
- name: NanoTempReasonL2Context
num_bytes: 278517
num_examples: 200
- name: NanoTempReasonL2Fact
num_bytes: 277515
num_examples: 200
- name: NanoTempReasonL2Pure
num_bytes: 281848
num_examples: 200
- name: NanoTempReasonL3Context
num_bytes: 278630
num_examples: 200
- name: NanoTempReasonL3Fact
num_bytes: 277450
num_examples: 200
- name: NanoTempReasonL3Pure
num_bytes: 280858
num_examples: 200
- name: NanoWinoGrande
num_bytes: 300624
num_examples: 200
download_size: 5350649
dataset_size: 5351943
---
# NanoRARb
This dataset is a Nano-style retrieval dataset for [HAKARI-bench](https://github.com/hakari-bench/hakari-bench).
NanoRARb contains 17 Nano retrieval splits derived from RAR-b. Each split keeps up to 200 eligible queries and up to 10000 corpus documents, with exact duplicate query and document text removed where the generator records that policy.
## Usage
```python
from datasets import load_dataset
dataset_id = "hakari-bench/NanoRARb"
split = "NanoARCChallenge"
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 |
|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|
| NanoARCChallenge | 200 | 9350 | 200 | 126.7 | 108.5 | 171.2 | 30.9 | 30.0 | 43.0 |
| NanoAlphaNLI | 200 | 10000 | 200 | 103.8 | 102.5 | 119.0 | 43.8 | 42.0 | 51.0 |
| NanoHellaSwag | 200 | 10000 | 200 | 114.7 | 115.5 | 148.2 | 62.2 | 55.0 | 77.0 |
| NanoPIQA | 200 | 10000 | 200 | 37.9 | 37.0 | 46.0 | 98.0 | 65.0 | 124.0 |
| NanoQuail | 200 | 10000 | 200 | 1813.8 | 1823.5 | 1861.0 | 25.0 | 22.0 | 33.0 |
| NanoRARbCode | 200 | 10000 | 200 | 470.1 | 430.5 | 542.0 | 256.0 | 244.0 | 333.0 |
| NanoRARbMath | 200 | 10000 | 200 | 201.3 | 166.5 | 256.8 | 481.3 | 386.5 | 603.0 |
| NanoSIQA | 200 | 10000 | 200 | 126.9 | 124.0 | 137.2 | 21.5 | 19.0 | 27.0 |
| NanoSpartQA | 200 | 1592 | 384 | 654.9 | 621.0 | 727.5 | 49.8 | 55.0 | 62.0 |
| NanoTempReasonL1 | 200 | 10000 | 200 | 49.9 | 51.0 | 52.0 | 9.0 | 9.0 | 9.0 |
| NanoTempReasonL2Context | 200 | 10000 | 200 | 28755.2 | 6562.0 | 75421.0 | 19.9 | 17.0 | 23.0 |
| NanoTempReasonL2Fact | 200 | 10000 | 200 | 1744.4 | 672.0 | 4822.0 | 19.9 | 17.0 | 23.0 |
| NanoTempReasonL2Pure | 200 | 10000 | 200 | 53.0 | 53.0 | 61.0 | 19.9 | 17.0 | 23.0 |
| NanoTempReasonL3Context | 200 | 10000 | 200 | 31804.1 | 7032.0 | 75374.0 | 19.9 | 17.0 | 23.0 |
| NanoTempReasonL3Fact | 200 | 10000 | 200 | 1981.1 | 836.0 | 4775.0 | 19.9 | 17.0 | 23.0 |
| NanoTempReasonL3Pure | 200 | 10000 | 200 | 65.1 | 62.0 | 76.0 | 19.9 | 17.0 | 23.0 |
| NanoWinoGrande | 200 | 5095 | 200 | 112.0 | 110.0 | 122.2 | 7.7 | 7.0 | 9.0 |
## 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 | - | 15.36 | 24.69 | 23.10 | 46.01 | 65.99 | 64.45 | - | 1170 |
| NanoARCChallenge | english_porter_stop | 3.86 | 11.13 | 6.42 | 22.50 | 36.00 | 35.50 | 100-101 | 129 |
| NanoAlphaNLI | english_porter_stop | 32.88 | 58.98 | 47.77 | 67.50 | 91.50 | 89.50 | 100-101 | 21 |
| NanoHellaSwag | english_porter_stop | 13.93 | 12.53 | 15.51 | 52.50 | 52.50 | 59.50 | 100-101 | 81 |
| NanoPIQA | english_porter_stop | 24.43 | 40.17 | 37.41 | 59.50 | 68.00 | 67.50 | 100-101 | 65 |
| NanoQuail | english_porter_stop | 5.22 | 11.74 | 9.82 | 28.50 | 46.00 | 46.50 | 100-101 | 107 |
| NanoRARbCode | regex | 13.18 | 11.73 | 17.73 | 44.50 | 43.50 | 57.50 | 100-101 | 85 |
| NanoRARbMath | english_porter_stop | 61.47 | 78.18 | 73.50 | 94.50 | 94.00 | 97.50 | 100-101 | 5 |
| NanoSIQA | english_porter_stop | 2.39 | 6.18 | 4.05 | 18.50 | 38.50 | 33.50 | 100-101 | 133 |
| NanoSpartQA | english_porter_stop | 18.88 | 26.34 | 34.19 | 60.67 | 54.83 | 62.17 | 100-101 | 37 |
| NanoTempReasonL1 | english_porter_stop | 1.25 | 4.88 | 1.29 | 3.50 | 64.50 | 37.00 | 100-101 | 126 |
| NanoTempReasonL2Context | english_porter_stop | 11.14 | 21.71 | 20.49 | 46.00 | 79.50 | 77.00 | 100-101 | 46 |
| NanoTempReasonL2Fact | english_porter_stop | 6.15 | 30.05 | 25.13 | 72.00 | 89.50 | 92.50 | 100-101 | 15 |
| NanoTempReasonL2Pure | english_porter_stop | 0.00 | 4.83 | 0.33 | 0.50 | 48.50 | 33.00 | 100-101 | 134 |
| NanoTempReasonL3Context | english_porter_stop | 9.45 | 19.26 | 16.68 | 36.00 | 76.00 | 73.00 | 100-101 | 54 |
| NanoTempReasonL3Fact | english_porter_stop | 5.47 | 25.49 | 19.81 | 66.00 | 87.00 | 92.50 | 100-101 | 15 |
| NanoTempReasonL3Pure | english_porter_stop | 0.74 | 7.07 | 2.38 | 9.50 | 54.00 | 41.50 | 100-101 | 117 |
| NanoWinoGrande | english_porter_stop | 50.67 | 49.46 | 60.20 | 100.00 | 98.00 | 100.00 | 100 | 0 |
## Hybrid Safeguard Summary
- Safeguard positives: 1170
- Rows limited by corpus size: 0
- Metadata file: `reranking_hybrid_metadata.json`
## Source Links
- Source benchmark: `RAR-b`
- `mteb/ARCChallenge`: https://huggingface.co/datasets/mteb/ARCChallenge
- `mteb/AlphaNLI`: https://huggingface.co/datasets/mteb/AlphaNLI
- `mteb/HellaSwag`: https://huggingface.co/datasets/mteb/HellaSwag
- `mteb/PIQA`: https://huggingface.co/datasets/mteb/PIQA
- `mteb/Quail`: https://huggingface.co/datasets/mteb/Quail
- `mteb/RARbCode`: https://huggingface.co/datasets/mteb/RARbCode
- `mteb/RARbMath`: https://huggingface.co/datasets/mteb/RARbMath
- `mteb/SIQA`: https://huggingface.co/datasets/mteb/SIQA
- `mteb/SpartQA`: https://huggingface.co/datasets/mteb/SpartQA
- `mteb/TempReasonL1`: https://huggingface.co/datasets/mteb/TempReasonL1
- `mteb/TempReasonL2Context`: https://huggingface.co/datasets/mteb/TempReasonL2Context
- `mteb/TempReasonL2Fact`: https://huggingface.co/datasets/mteb/TempReasonL2Fact
- `mteb/TempReasonL2Pure`: https://huggingface.co/datasets/mteb/TempReasonL2Pure
- `mteb/TempReasonL3Context`: https://huggingface.co/datasets/mteb/TempReasonL3Context
- `mteb/TempReasonL3Fact`: https://huggingface.co/datasets/mteb/TempReasonL3Fact
- `mteb/TempReasonL3Pure`: https://huggingface.co/datasets/mteb/TempReasonL3Pure
- `mteb/WinoGrande`: https://huggingface.co/datasets/mteb/WinoGrande
## License
NanoRARb is a derived dataset. Users must comply with the licenses, terms, and attribution requirements of the upstream datasets and benchmarks.