[D] Want to build an ASR model that trains in the Urdu language and detects Urdu labels. Later translates it into English and create a transcript file.

​

Hello, I am working on a speech recognition project which gets Urdu speech and translates it into English making a transcripted file of the audio (In English).

I know there are tons of libraries that would make my work easy but I want to build it from scratch. I have datasets that I could train my model on it but I am here confused about what steps should I take if I go start building a model from scratch? …

asr language machinelearning