Kevin Chen

Is there more error provided in the log? Alternatively, you can try adding the `-DBUILD_LIBRARY_ONLY=1` flag to your CMAKE build command to see if that fixes it.

This is a known limitation, we are working on this feature and will be available in a future release.

We don't officially support the Windows build of this repository, error seems to becoming from Protobuf. What version do you have installed?

The ONNX parser calls `initLibNvInferPlugins()` internally. This function registers the builtin TensorRT plugins. This error may be coming from a memory leak in the InstanceNormalization plugin itself. Which version of...

You can build the optimization profiles with a range of dimensions using the min/max/opt profiles. You can see [this link](https://docs.nvidia.com/deeplearning/tensorrt/developer-guide/index.html#opt_profiles) for more info.

How are you building the library? I've just tested the 7.2.1 branch with protobuf 3.6.1 and it compiled fine for me. ``` build$ protoc --version libprotoc 3.6.1 cmake .. -DTENSORRT_ROOT=$TRT7...

Looking into it - it seems like all we have to do is update the base image here https://github.com/NVIDIA/TensorRT/blob/main/docker/ubuntu-20.04-aarch64.Dockerfile#L19 to 11.6.2. I'm verifying it on my end - can you...

INT32 padding has been supported for the past few releases, closing this issue. If the issue persists, please open a new issue.