We encountred an exception when trying to retrieve "int_weight" from the model

[solitary-1]

We have successfully trained the model of "LFC_1W1A" in bnn_pynq example. When we tried to retrieve "int_weight" from the model, we encountered an exception which stopped us from doing so.

Here is the code we used to print out "int_weight": def forward(self, x): for mod in self.features: if isinstance(mod, QuantLinear): print(mod.int_weight()) # where we encountered the exception We want to know if there is a problem with our code or our method of getting the quantized weights. Thanks in advance.

[volcacius]

Hello,

What the error is saying is that basically quantization is not enabled. Are you running training using brevitas_bnn_pynq_train --network LFC_1W1A --experiments /path/to/experiments ?

Alessandro

[solitary-1]

Hello,

What the error is saying is that basically quantization is not enabled. Are you running training using brevitas_bnn_pynq_train --network LFC_1W1A --experiments /path/to/experiments ?

Alessandro

Follow your steps, but there was an error in running the line of command.Look forward to your reply.

[volcacius]

Fixed by #200 .