How do I save and load quantization model

Be sure you do the whole post training preparation process (by running layer fusion, torch.quantization.prepare() and torch.quantization.convert() ) before loading the state_dict.

3 Likes