Knowledge-Distillation-Toolkit
Knowledge-Distillation-Toolkit copied to clipboard
How to save the quantized model properly for wav2vec.
I want to directly load the quantized model. So how to save the wav2vec model properly.
I haven't tried to save and load quantized model. Maybe this thread (https://discuss.pytorch.org/t/how-do-i-save-and-load-quantization-model/64188/4) will help?
You may also save the original model and quantize it after loading.