sherpa

Hint

During speech recognition, it does not need to access the Internet. Everyting is processed locally on your device.

k2-fsa/sherpa use PyTorch for neural network computation.

https://k2-fsa.github.io/icefall/model-export/export-with-torch-jit-script.html>`_ for how to export models.

In the following, we describe how to use k2-fsa/sherpa.

• Installation
  • From pre-compiled wheels
  • From source
  • Check your installation
  • Where to get help
• Pre-trained models
  • Offline CTC models
  • Offline transducer models
  • Online transducer models