Benchmarking Quantized Mobile Speech Recognition Models with PyTorch Lightning and Grid

PyTorch Lightning enables you to rapidly train models while not worrying about boilerplate. While this makes training easier, in practice models are not trained for the sake of training models but rather for deploying to production applications. In this fourth and final part of the tutorial, we summarize our findings from the first three parts (Training a baseline model, Background on Quantization

[satojkovic]

quantization, speedup