thennal/whisper-medium-ml
Automatic Speech Recognition
•
Updated
•
501
•
5
Datasets:
Tasks:
Automatic Speech Recognition
Languages:
Malayalam
Multilinguality:
monolingual
Size Categories:
1K<n<10K
Language Creators:
crowdsourced
Annotations Creators:
crowdsourced
Tags:
License:
The dataset viewer is taking too long to fetch the data. Try to refresh this page.
Server-side error
Error code: ClientConnectionError
SMC Malayalam Speech Corpus
Malayalam Speech Corpus (MSC) is a repository of curated speech samples collected using MSC web application, released by Swathanthra Malayalam Computing. The official blog post and source data can be found at https://blog.smc.org.in/malayalam-speech-corpus/.
Dataset Summary
The first version of Malayalam Speech Corpus contains 1541 speech samples from 75 contributors amounting to 1:38:16 hours of speech. It has 482 unique sentences, 1400 unique words, 553 unique syllables and 48 unique phonemes.
- Downloads last month
- 31
Homepage:
blog.smc.org.in
Size of downloaded dataset files:
423 MB
Size of the auto-converted Parquet files:
423 MB
Number of rows:
1,541
Models trained or fine-tuned on thennal/msc