name: youtube_subtitles
fullname: YouTube Subtitles
lang: en
category: informal
description: The YouTube Subtitles dataset is a parallel corpus of text gathered from
human generated closed- captions on YouTube. In addition to providing mul- tilingual
data, Youtube Subtitles is also a source of educational content, popular culture,
and natural dialog.
license: MIT License
homepage: https://youtube.com
version: 1.0.0
num_docs: 150749
num_docs_before_processing: 0
num_segments: 713475
num_sents: 16074289
num_words: 303286377
size_in_bytes: 1725812988
num_bytes_before_processing: 0
size_in_human_bytes: 1.61 GiB
data_files_modified: '2022-02-23 12:35:04'
info_updated: '2022-02-26 03:06:09'
data_files:
train: youtube_subtitles-train.parquet
test: youtube_subtitles-test.parquet
valid: youtube_subtitles-valid.parquet
meta_files:
train: null
test: null
valid: null
features:
columns:
id: id
text: text
data:
id: int
text: str
This repository has been archived by the owner on May 9, 2024. It is now read-only.