Closed imyhxy closed 2 years ago
@ttyio Could you give some advice?
@imyhxy , could you share with us ONNX file to debug? thanks
Hello @imyhxy , yes you are right horizontal fusion is disabled on QAT network due to some issue, we have tracked this as feature request internally. Sorry for the inconvenience.
@ttyio Thanks, wait for the upgrade of TensorRT.🌞
@ttyio What was the bug id? Has this been fixed?
Closing due to >14 days without activity. Please feel free to reopen if the issue still exists. Thanks
Description
Hi there, recently I was using
pytorch_quantization
to quantize my detection model, and I found out that there is a different processing step between native TensorRT INT8 PTQ and the QDQ method.When there are two convolution layers with the same kernel size operate on the same input tensor, the native TensorRT INT8 PTQ would merge those convolution layers into single one convolution layer. Like the following pattern:
Compare to above, the QDQ method won't triger this tensor method even thought I use the same
TensorQuantize
node for both branch, the output is following:Environment
TensorRT Version: 8.2.0.6 NVIDIA GPU: T4 NVIDIA Driver Version: 460.32.03 CUDA Version: 11.5 CUDNN Version: 8.3 Operating System: Ubuntu 18.04 Python Version (if applicable): 3.8 Tensorflow Version (if applicable): PyTorch Version (if applicable): 1.9.0 Baremetal or Container (if so, version):
Relevant Files
Steps To Reproduce