Installing paddlepaddle_gpu on Orin NX

Hello all,

I’m trying to get paddlepaddle_gpu working on Jetson Orin NX.

  1. Installing paddlepaddle_gpu using paddlepaddle_gpu-2.5.2-cp38-cp38m-linux_aarch64.whl
  2. Installing paddleocr - pip3 install paddleocr
  3. Testing with following code:
    import paddle


, but I’m getting:
Running verify PaddlePaddle program …
Traceback (most recent call last):
File “”, line 4, in
File “/home/user/.local/lib/python3.8/site-packages/paddle/utils/”, line 249, in run_check
run_static_single(use_cuda, use_xpu)
File “/home/user/.local/lib/python3.8/site-packages/paddle/utils/”, line 134, in run_static_single
param_grads = paddle.static.append_backward(
File “”, line 2, in append_backward
File “/home/user/.local/lib/python3.8/site-packages/paddle/fluid/”, line 25, in impl
return wrapped_func(*args, **kwargs)
File “/home/user/.local/lib/python3.8/site-packages/paddle/fluid/”, line 473, in impl
return func(*args, **kwargs)
File “/home/user/.local/lib/python3.8/site-packages/paddle/fluid/”, line 2179, in append_backward
File “/home/user/.local/lib/python3.8/site-packages/paddle/fluid/”, line 1756, in append_backward_vars
RuntimeError: (NotFound) The kernel sum_grad is not registered.
[Hint: Expected iter != kernels
.end(), but received iter == kernels
.end().] (at /workspace/compile-arm/2.5/2.5.2/Paddle/paddle/phi/core/
[operator < reduce_sum_grad > error]

Jetson info:
Package: nvidia-jetpack
Version: 5.1.1-b56
Architecture: arm64
Maintainer: NVIDIA Corporation
Installed-Size: 194
Depends: nvidia-jetpack-runtime (= 5.1.1-b56), nvidia-jetpack-dev (= 5.1.1-b56)
Homepage: Jetson - Embedded AI Computing Platform | NVIDIA Developer
Priority: standard
Section: metapackages
Filename: pool/main/n/nvidia-jetpack/nvidia-jetpack_5.1.1-b56_arm64.deb
Size: 29304
SHA256: 7b6c8c6cb16028dcd141144b6b0bbaa762616d0a47aafa3c3b720cb02b2c8430
SHA1: 387e4e47133c4235666176032af0f2ec86461dbb
MD5sum: 0a8692031bf35cc46f7a498e2937bda9
Description: NVIDIA Jetpack Meta Package
Description-md5: ad1462289bdbc54909ae109d1d32c0a8

Any ideas?


In general, the package is compiled with dGPU.
Please check with the package owner to see if they support the Jetson environment first.


Actually even having this error detection works.


Do you mean further issues when inference?
You already installed their package for Jetson and it’s expected to work correctly.


This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.