This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
hello, I have some question when I try QAT in edgeai-torchvision.
after I run a pretrained model with phase=training and quantize=True, get a model like below
and the following code is
1. why we need to excute extra QAT? I want to know the effect.
2.why the code set the quantize to False, if this, the model will be original, and the onnx saved Indeed.
Hello,
The topic expert looking into this topic is currently out-of-office and is expected to be back in 2-3 days. So please expect delay in response to this query.
Thanks.
Have a look at the documentation on Quantization to understand more about QAT and PTQ.
QAT is optional - you also have the option to use PTQ that is part of TIDL.