Datasets:
mteb
/

Modalities:
Text
Formats:
json
Languages:
English
Libraries:
Datasets
pandas
msmarco-v2 / README.md
Muennighoff's picture
Update README.md
10181e1 verified
---
language:
- en
multilinguality:
- monolingual
task_categories:
- text-retrieval
source_datasets:
- msmarco-v2
task_ids:
- document-retrieval
config_names:
- corpus
tags:
- text-retrieval
dataset_info:
- config_name: default
features:
- name: query-id
dtype: string
- name: corpus-id
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 9631462
num_examples: 284212
- name: dev
num_bytes: 136961
num_examples: 4009
- name: dev2
num_bytes: 150735
num_examples: 4411
- config_name: corpus
features:
- name: _id
dtype: string
- name: title
dtype: string
- name: text
dtype: string
splits:
- name: corpus
num_bytes: 50691069190
num_examples: 138364198
- config_name: queries
features:
- name: _id
dtype: string
- name: text
dtype: string
splits:
- name: queries
num_bytes: 13379527
num_examples: 285328
configs:
- config_name: default
data_files:
- split: train
path: qrels/train.jsonl
- split: dev
path: qrels/dev.jsonl
- split: dev2
path: qrels/dev2.jsonl
- config_name: corpus
data_files:
- split: corpus
path: corpus.jsonl.gz
- config_name: queries
data_files:
- split: queries
path: queries.jsonl
---