Jetson Orin Agx torch.cuda.is_available() is False

[sm1045]

Good afternoon. I am trying to run whisper into a Jetson Orin Agx. I complete the configuratión of the device according to the requirements suggested for nvidia. When I try to make a translation throws me this error:

__RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU._

Can anyone helps me plz?

p/d: sorry for my bad written English

[sm1045]

CUDA compiler seems to be ok. But, I can´t achieve that whisper could use it as a device.

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Sun_Oct_23_22:16:07_PDT_2022
Cuda compilation tools, release 11.4, V11.4.315
Build cuda_11.4.r11.4/compiler.31964100_0

[sm1045]

Also, I leave here more information.

Python version: 3.8.10 (default, Mar 13 2023, 10:26:41) [GCC 9.4.0] (64-bit runtime) Python platform: Linux-5.10.104-tegra-aarch64-with-glibc2.29 Is CUDA available: False CUDA runtime version: 11.4.315 CUDA_MODULE_LOADING set to: N/A GPU models and configuration: Could not collect Nvidia driver version: Could not collect

[sm1045]

Downgrading torch version and adding env var related to CUDA was the solution

[iamwavecut]

Please, describe the versions and ENV vars.