This script is based on the deepspeech model created by Mozilla.
In order to use this script, you need to create one folder called data/. Move any file (audio) you like to this folder.
Download the deepspeech-0.1.1-models.tar.gz from https://github.com/mozilla/DeepSpeech/releases and move the files to models/ folder.
Install the dependencies from the requirements.txt.
In the code change the path_to_audio to point to the file you want to transcribe and hit run in the IDE of your choice.