metadata
language:
- en
dataset_info:
- config_name: eli5
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 25298207
num_examples: 25000
download_size: 15538771
dataset_size: 25298207
- config_name: fever
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 461884567
num_examples: 71257
download_size: 279121262
dataset_size: 461884567
- config_name: fiqa
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 74970232
num_examples: 5500
download_size: 43692330
dataset_size: 74970232
- config_name: hotpot_qa
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 466370386
num_examples: 68657
download_size: 286508306
dataset_size: 466370386
- config_name: msmarco_passage
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 1469115771
num_examples: 400782
download_size: 836954087
dataset_size: 1469115771
- config_name: nq
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 508591032
num_examples: 76923
download_size: 312556583
dataset_size: 508591032
- config_name: quora_duplicate
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 46431262
num_examples: 60000
download_size: 26717034
dataset_size: 46431262
- config_name: squad
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 1394419158
num_examples: 74377
download_size: 752527010
dataset_size: 1394419158
- config_name: synthetic
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
- name: instruction
dtype: string
splits:
- name: train
num_bytes: 450333569
num_examples: 224791
download_size: 260055908
dataset_size: 450333569
- config_name: triviaqa
features:
- name: query
dtype: string
- name: positive
sequence: string
- name: negative
sequence: string
splits:
- name: train
num_bytes: 375282874
num_examples: 52738
download_size: 231064553
dataset_size: 375282874
configs:
- config_name: eli5
data_files:
- split: train
path: eli5/train-*
- config_name: fever
data_files:
- split: train
path: fever/train-*
- config_name: fiqa
data_files:
- split: train
path: fiqa/train-*
- config_name: hotpot_qa
data_files:
- split: train
path: hotpot_qa/train-*
- config_name: msmarco_passage
data_files:
- split: train
path: msmarco_passage/train-*
- config_name: nq
data_files:
- split: train
path: nq/train-*
- config_name: quora_duplicate
data_files:
- split: train
path: quora_duplicate/train-*
- config_name: squad
data_files:
- split: train
path: squad/train-*
- config_name: synthetic
data_files:
- split: train
path: synthetic/train-*
- config_name: triviaqa
data_files:
- split: train
path: triviaqa/train-*
Training data for e5-R-mistral-7b. Code is here.