Hugging Face's logo

Datasets:
NagaYu
/
downstep-bench

Tasks:
Token Classification
Modalities:
Tabular
Text
Formats:
json
Languages:
Japanese
Size:
10K - 100K
Tags:
japanese
pitch-accent
prosody
tts
openjtalk
compounds
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
downstep-bench / test.jsonl

Commit History

Downstep compound pitch-accent benchmark: 89,539 compounds, listed/unlisted/hard
010ee9b
verified

NagaYu commited on