Deepspeech using a GPU

Virtual environment

We will need to create another virtual environment for deepspeech-gpu

Install deepspeech-gpu

Model and audio files

If you've been following along you can use the same model and audio files from the Deepspeech basics article

If not you can install them like so:

Installing CUDA and cuDNN

This goes without saying but make sure you have an Nvidia GPU and the proprietary drivers installed

You will need both of these libraries in order to run inference

If you already have CUDA 10 installed then great you can move on

If not I have an easy way to do so using conda

Run inference

We can now transcribe the audio file

Tagged in deepspeech