metadata
configs:
- config_name: fineweb
kind: doc
data_files:
- split: train
path: fineweb/*
- config_name: cc-news
kind: doc
data_files:
- split: train
path: cc-news/*
- config_name: vocabulary
kind: doc
data_files:
- split: train
path: vocabulary/*
- config_name: bert-tokens
kind: doc
data_files:
- split: train
path: bert-tokens/*
- config_name: amazon-qa
kind: query
data_files:
- split: train
path: amazon-qa/*
- config_name: msmarco
kind: query
data_files:
- split: train
path: msmarco/*
- config_name: pubmedqa
kind: query
data_files:
- split: train
path: pubmedqa/*
- config_name: trivia-qa
kind: query
data_files:
- split: train
path: trivia-qa/*
- config_name: lotte
kind: query
data_files:
- split: train
path: lotte/*
distil-data
A reproduction of the (unsupervised) text data used to train MongoDB/mdbr-leaf-ir