Dataset Viewer
Auto-converted to Parquet Duplicate
id
string
shard
string
transcript
string
translation_en
string
transcript_lang
string
locale_lang
string
length
float64
audio_bitrate
float64
2be6650879fb9bfd00a476b446de80da
0000
[{"start":0,"duration":2000,"text":"USA"}]
null
ab
uz
437
145.584
004baeb9a4b5fe6bf0da38feffa64790
0000
[{"start":654,"duration":3626,"text":"[musiek]"},{"start":1280,"duration":3000,"text":"you"},{"start":5239,"duration":2020,"text":">> [musiek]"}]
[{"start":654,"duration":3626,"text":"[music]"},{"start":1280,"duration":3000,"text":"you"},{"start":5239,"duration":2020,"text":"[music]"}]
af
id
22
125.867
02984dc793e4acea4b93342fc54fea8d
0000
[{"start":480,"duration":5640,"text":"Geloof is nie altyd veilig nie. Wat","words":[{"t":0,"w":"Geloof"},{"t":520,"w":"is"},{"t":719,"w":"nie"},{"t":839,"w":"altyd"},{"t":1520,"w":"veilig"},{"t":1919,"w":"nie."},{"t":2840,"w":"Wat"}]},{"start":3520,"duration":4960,"text":"beteken dit? Dit beteken dat ons dink","words":...
[{"start":480,"duration":5640,"text":"Faith is not always safe. What does"},{"start":3520,"duration":4960,"text":"this mean? This means that we"},{"start":6120,"duration":5160,"text":"sometimes think that faith should always bring moonlight and"},{"start":8480,"duration":4760,"text":"roses into our lives. And that's...
af
ig
67
92.583
0312bfc94408cc2f7d43c343502677fb
0000
[{"start":1160,"duration":6159,"text":"de question","words":[{"t":0,"w":"de"},{"t":400,"w":"question"}]},{"start":4319,"duration":4681,"text":"n²+ 15+","words":[{"t":0,"w":"n²+"},{"t":801,"w":"15+"}]},{"start":7319,"duration":7480,"text":"56+"},{"start":9000,"duration":10680,"text":"7+ 8 ilut factorette expression pol ...
[{"start":1160,"duration":6159,"text":"de question"},{"start":4319,"duration":4681,"text":"n²+ 15+"},{"start":7319,"duration":7480,"text":"56+"},{"start":9000,"duration":10680,"text":"7+ 8 ilut factorette expression pol mais"},{"start":14799,"duration":4881,"text":"on va bien s on va par"},{"start":20920,"duration":732...
af
cy
63
142.867
0674b53137ed67678211c3c196f4ee13
0000
"[{\"start\":399,\"duration\":5521,\"text\":\"Goeiemiddag kollegas en leeders weereens\",\"words\":[(...TRUNCATED)
null
af
nl
5,883
118.37
06c6679d4fed54982356a058b44a939c
0000
"[{\"start\":3480,\"duration\":8840,\"text\":\"Laat die woorde van my mond en die\",\"words\":[{\"t\(...TRUNCATED)
null
af
af
851
108.174
0e5dd687705158272232e4917cc7a72f
0000
"[{\"start\":680,\"duration\":6119,\"text\":\"Shalom to everyone. In this PowerPoint\",\"words\":[{\(...TRUNCATED)
"[{\"start\":680,\"duration\":6119,\"text\":\"Hello everyone. In this PowerPoint\"},{\"start\":3439(...TRUNCATED)
af
th
2,181
137.559
0e7905147f787ef9b7c209df781d5950
0000
"[{\"start\":2120,\"duration\":4560,\"text\":\"Goeiedag.\"},{\"start\":4080,\"duration\":5320,\"text(...TRUNCATED)
"[{\"start\":2120,\"duration\":4560,\"text\":\"Good day.\"},{\"start\":4080,\"duration\":5320,\"text(...TRUNCATED)
af
sw
1,639
108.97
1575ab8bb7019eaa80f65585d5b5353e
0000
"[{\"start\":520,\"duration\":3960,\"text\":\"So vriende, hoe moet ons mekaar leer en\",\"words\":[{(...TRUNCATED)
"[{\"start\":520,\"duration\":3960,\"text\":\"So friends, how should we teach and admonish one anoth(...TRUNCATED)
af
nl
205
131.133
158081cbabb597002ff9551611f063d0
0000
"[{\"start\":280,\"duration\":7200,\"text\":\"Here, in hierdie oggend erken ons\",\"words\":[{\"t\":(...TRUNCATED)
null
af
af
2,329
97.251
End of preview. Expand in Data Studio

YODAS v3

Paper

YODAS v3 is a large web-crawled dataset containing over 1.1 million hours of audio that were originally released under a CC-BY-3.0 license. The dataset contains audio in over 100 languages. YODAS v3 can be used for a variety of multi-modal tasks, including Automatic Speech Recognition, Text-to-Speech, and Audio Representation Learning. We crawl a distinct set of videos from the v1 and v2 versions of YODAS, to guarantee that there are no overlaps in the data.

For each audio file, we also release the following metadata:

  • segment-level transcripts (if available)
  • English translations (if available)
  • language IDs
  • audio length
  • effective bandwith
  • effective channel count

Approximately 75% of the data has transcripts. Around 60% of the non-English data has English translations. All transcripts and translations come with sentence-level timestamps. 95% of the transcripts also have word-level timestamps.

Language IDs are provided at the file-level, and may not be entirely accurate on individual sentences. The language IDs were obtained in 2 ways:

  1. the language ID provided with the transcript, if available
  2. the provided locale of the audio from the upload

While we originally reported the ID distribution from (2) in our paper, we found that the IDs from (1) are likely more accurate, if available. See the Analysis section below for more details. We therefore organize the upload directory by the language ID detected from (1), but include both IDs in the metadata. Any audio files without a transcript are put in the unk directory.

Structure

Analysis

Citation

Downloads last month
361