Hi there! Just wondering how to access the weight parameter of a quantized NN model which is loaded using torch.jit.load().
I already read this post (Acces to weights after quantization · Issue #33013 · pytorch/pytorch · GitHub), but I’d like to know when the model is saved by using torch.jit.save()
I tried to use the following code; however, the results show only scale and zero_point, not weight and bias.
q = torch.jit.load(‘xxxxxxx.pt’)
q.state_dict()