Spaces:
Running
on
Zero
Apply for community grant: Academic project (gpu)
Our project is trying to bring speech recognition to one of the main languages of Paraguay, the Guaraní language. Guaraní is spoken by 87% of the Paraguayan population, yet it is considered a low resource language because of spanish predomination in Paraguayan education, business documents and government forms. Making Guaraní mostly a spoken language; in this form there is ton of available resources. With this whisper implementation we intend to expand the amount of written resources for the guaraní language, with the hope that this will result in larger language models based on it. Help us keep alive a Language that is spoken by 6 million people worldwide including Argentina, Brazil, Paraguay and Bolivia; making it the most spoken indigenous language of the Americas !! This project is also going to be published as part of our Computer Science Engineering Undergraduate work.
Thanks, I managed to make it work by updating PEFT