Deconvolution onnx sample
Hello
what it means?
Should I use during train upsample + conv, and then during export to onnx replace them to ConvTranspose?
Can you attach some onnx example with right operators?
If Deconv consist Upsample, why we are using cpu to inference separate Upsample operator instead of npu?
UPD
Sorry, I opened discussion in KL720 topic instead of KL520 by mistake, can someone move it to the right place?
The discussion has been closed due to inactivity. To continue with the topic, please feel free to post a new discussion.
Comments
Hi Denis,
It means the operator deconvolution could be used in KL520 and our compiler will use upsampling + conv to implement the operator. So you don't need to do other operation with the layer deconvolution.