Zero Zeng

@azhurkevich Can you help here ^ ^ many thanks!

refer to https://github.com/zerollzeng/tiny-tensorrt/blob/d267b9cc9be5408a3c5a1b4c57e0be2d24a7de1d/src/Trt.cpp#L286

I'm not an expert on TF-TRT, but TRT doesn't support IndexLookup layer, onnx doesn't support this layer too?

Also from your log, I don't think TensorRT gets involved here. so it would be better to ask in the Tensorflow repo.

I believe we only supported Arm architecture since 8.0 GA (https://developer.nvidia.com/nvidia-tensorrt-8x-download#trt801https://developer.nvidia.com/nvidia-tensorrt-8x-download#trt801)

can you upgrade to a newer version of CUDA and CUDNN?

> Or one-to-one correspondence, use TensorRT 8.x GA to build the 8.x version of TensorRT OSS I think we only guarantee this will work, or you might run into unexpected...

@ttyio Looks like we can improve it ^ ^

Can you share the onnx file here?

I think it's better to ask in the pytorch repo.