Simple speech-to-text application powered by OpenAI Whisper.
You need to download audio recognition model data file, for example, small. Note: do not use *.en data files in order to support multiple languages. When Romashki Audio Notebook is launched, you need to specify the path to the model file in the application settings.