Teaching your own model based on 5 words


I am a beginner and I would like to train a model for my language, it is very rare.

First I would like to understand the whole process of training a model based on 5 words.

I have the words and audio files in .wav format

I have followed the installation steps:


what should i do next?

The DeepSpeech Playbook walks you through training.

5 words? Tough!
-Add data: Find similar stuff, change your words, make fake audio.
-Pick a model: Tiny for 5 words, adapt a bigger one.
-Train smart: Start small, track progress, split data.
-It’s hard: Aim for basic recognition, keep improving.
-Find help: Look for low-resource language tools.
Go preserve your unique language!