The GPU is not working. onnxruntime-gpu does not seem to be activated in any case.

[thuongvovan]

Describe the bug
The GPU is not working. onnxruntime-gpu does not seem to be activated in any case.

To Reproduce
Steps to reproduce the behavior:

Here is the English translation of your text:

---

I am running the rembg library in Docker and have tried different scenarios, but it seems the GPU is still not working. onnxruntime-gpu does not appear to be called in any case. Nothing happens.

The Docker image nvidia/cuda:11.8.0-cudnn8-runtime-ubi8 has CUDA 11 and cuDNN 8, which are compatible with the default onnxruntime-gpu==1.18.1 version.

The Docker image nvidia/cuda:12.3.2-cudnn9-runtime-ubuntu22.04 has CUDA 12 and cuDNN 9, which are compatible with the default onnxruntime-gpu==1.18.1 version distributed in the Azure DevOps Feed.

Strangely, even when using an image without integrated CUDA or cuDNN, nothing happens. Usually, if using an environment that doesn't support onnxruntime-gpu, there would be an error log like the one below:

*************** EP Error ***************
EP Error /onnxruntime_src/onnxruntime/python/onnxruntime_pybind_state.cc:456 void onnxruntime::python::RegisterTensorRTPluginsAsCustomOps(onnxruntime::python::PySessionOptions&, const ProviderOptions&) Please install TensorRT libraries as mentioned in the GPU requirements page, make sure they're in the PATH or LD_LIBRARY_PATH, and that your GPU is supported.
when using ['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'AzureExecutionProvider', 'CPUExecutionProvider']
Falling back to ['CUDAExecutionProvider', 'CPUExecutionProvider'] and retrying.

---

Please check again!!

Rembg version:
v2.0.58

[github-actions]

This issue is stale because it has been open for 30 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.

[sasachen]

how to fix?