Skip to content

carlfm01/my-speech-datasets

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 

Repository files navigation

About

I'm training a Mozilla's DeepSpeech model using public domain data, I'll be releasing the data I collect also under public domain.

About Data collection methodology

Automatically aligned the text with the Windows speech recognition, then as validation of the alignment used a Mozilla's DeepSpeech model using a few different language models. The first model to start the validation was trained on voxforge Spanish data and on top of the ones that scored the highest confidence with the Windows Speech.

Languages

Initially I'm collecting only Spanish, eventually I'll add more languages.

100h of clean speech from a single speaker LJSpeech format!

Issues

Please if you find wrong transcription open an issue or send a PR to fix them.

License

Public domain.