Description
Hello dear NVIDIA developer community!
I am trying to run all layers of YOLOv3 on NVDLA using tkDNN. The leaky ReLU, route, upsample, and yolo layers are implemented by using TensorRT plugins with CUDA code and they can run on GPU without a problem. However, they are not compatible with NVDLA. What should I do in order to run the custom plugin layers on NVDLA, if possible? If its not possible, is changing these layers into multiple NVDLA compatible layers the only solution?
Best regards
Environment
TensorRT Version: 7
GPU Type: Jetson NX iGPU
Nvidia Driver Version:
CUDA Version: 10.2
CUDNN Version: 8
Operating System + Version: Ubuntu 18.04
Baremetal or Container (if container which image + tag): Docker image: nvcr.io/nvidia/l4t-base r32.4.3