robmsmt / KerasDeepSpeech

A Keras CTC implementation of Baidu's DeepSpeech for model experimentation
GNU Affero General Public License v3.0
242 stars 79 forks source link

Language model for bangla #13

Open Arif45 opened 5 years ago

Arif45 commented 5 years ago

I want to implement it for both bangla isolated and continuous speech. Where i can find language model for banlga if not available how can i make language model?

robmsmt commented 5 years ago

Hello, I am not sure where to find a language model for banlga but it should be quite straight forward to make one yourself from existing text. This repo lists some examples and i'd also keep an eye on Mozilla's Common voice.

To create the language model I would recommend either KenLM or MitLM if your corpus is small i.e. less than 1000 sentences.