I installed OpenVINO dependencies and converted the model to OpenVINO format. I have OpenCL device available:
$ clinfo -l
Platform #0: Intel(R) OpenCL HD Graphics
`-- Device #0: Intel(R) Graphics [0xa7a0]
Trying to run yolo predict on a GPU, e.g.
yolo predict model=openvino_model source='samples/*.jpg' device=gpu
yolo predict model=openvino_model source='samples/*.jpg' device=0
always results in Invalid CUDA device requested. It runs fine on the CPU, though.
What parameter should I use for yolo to find and use OpenCL device #0?