search

yhhhli / APoT_Quantization

PyTorch implementation for the APoT quantization (ICLR 2020)
258 stars 51 forks source link

Bottleneck Doubts #3

Closed dronefreak closed 4 years ago

dronefreak commented 4 years ago

This is more like a doubt and less like an issue. I am new to this field, so I am unable to understand, why the QuantConv() layers have not been implemented in the Bottleneck class. Only the BasicBlock witnesses the usage of the quantized convolutional layers. Is there a specific reason for that? Does it affect the accuracy too much?

yhhhli commented 4 years ago

Hello, BottleNeck block is the components of ResNet-50 or higher, (like ResNet-152). For ResNet-18 and 34, we only need to quantize the Basic block. I will provide codes and checkpoints later for ResNet-50. Please check for our updates!