RuntimeError: Didn't find engine for operation quantized::conv2d_prepack NoQEngine

Hi,

I’m trying to run a quantized faster rcnn model on the CPU of the Jetson Nano but am having this runtime error. I’m using Pytorch to build the model and have seen this issue posted here, however I am using qnnpack as the quantization engine. Is there any way to resolve this?

Upon further debugging I found this error message:

terminate called after throwing an instance of ‘c10::Error’

what(): quantized engine QNNPACK is not supported

Exception raised from setQEngine at /media/nvidia/NVME/pytorch/pytorch-v1.9.0/aten/src/ATen/Context.cpp:181 (most recent call first):

frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits, std::allocator >) + 0xa0 (0x7f4e3ab300 in /home/harry/.local/lib/python3.6/site-packages/torch/lib/libc10.so)

frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&) + 0xb4 (0x7f4e3a76b4 in /home/harry/.local/lib/python3.6/site-packages/torch/lib/libc10.so)

frame #2: at::Context::setQEngine(c10::QEngine) + 0x138 (0x7f64c94940 in /home/harry/.local/lib/python3.6/site-packages/torch/lib/libtorch_cpu.so)

frame #3: THPModule_setQEngine(_object*, _object*) + 0x94 (0x7f697f6364 in /home/harry/.local/lib/python3.6/site-packages/torch/lib/libtorch_python.so)

frame #5: python3() [0x52ba70]

frame #7: python3() [0x529978]

frame #9: python3() [0x5f4d34]

frame #11: python3() [0x5a7228]

frame #12: python3() [0x582308]

frame #16: python3() [0x52b108]

frame #21: __libc_start_main + 0xe0 (0x7f82579720 in /lib/aarch64-linux-gnu/libc.so.6)

frame #22: python3() [0x420e94]

Hi,

This is a non-supported error.

quantized engine QNNPACK is not supported

Have you run the model on another platform with the same PyTorch version and script successfully?
If not, could you give it a try?

More, do you use the package from the below topic?

Thanks.

Hi,

I did originally install the whl file from that post. I have been able to run the model on other platforms just not this one so far. I am working on installing Pytorch from source with qnnpack enabled.

Thanks
Harry

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

Hi,

Have you built the QNNPACK-enabled PyTorch successfully?

Thanks.