Icefall for dummies tutorial

This tutorial walks you step by step about how to create a simple ASR (Automatic Speech Recognition) system with Next-gen Kaldi.

We use the yesno dataset for demonstration. We select it out of two reasons:

• It is quite tiny, containing only about 12 minutes of data

• The training can be finished within 20 seconds on CPU.

That also means you don’t need a GPU to run this tutorial.

Let’s get started!

Please follow items below sequentially.

Note

The Data Preparation runs only on Linux and on macOS. All other parts run on Linux, macOS, and Windows.

Help from the community is appreciated to port the Data Preparation to Windows.

• Environment setup
  • Create a virtual environment
  • Install dependencies
  • Install icefall
  • For the more curious
• Data Preparation
  • For the more curious
  • A quick look to the generated files
• Training
  • For the more curious
• Decoding
  • For the more curious
• Model Export
  • Export the model parameters via model.state_dict()
  • Export via torch.jit.script()
  • Export via torch.onnx.export()
  • For the more curious