Torch.cuda not true

user53799 March 27, 2023, 5:05am 1

Hardware Platform (Jetson / GPU) NX
*
• JetPack Version (valid for Jetson only) 4.4.1
• TensorRT Version 7.1.3

I am using nvcr.io/nvidia/l4t-pytorch:r35.1.0-pth1.11-py3
docker image.
but torch.cuda.is_available() False.

Please tell me how to use this docker image correctly?
Shall I install jetpack5.0 on NX first? But I need JP4.4 for another app.

Topic		Replies	Views
Torch.cuda.is_available() return false on docker image "dustynv/jetson-inference:r32.5.0" Jetson Nano pytorch	3	671	June 14, 2023
Jetson Orin Nx, Pytorch 2.1.2 on Jetpack 6.0 DP, torch.cuda.is_available() is false Jetson Orin NX pytorch	8	2115	January 30, 2024
Docker Image for Jetson AGX Orin for pytorch and cuda Jetson AGX Orin docker , pytorch	2	97	October 22, 2024
Trouble trying to install torch in Docker container on JP6.0-dp Jetson Orin NX docker , pytorch	2	417	March 21, 2024
L4t cuda torch cuda False Jetson Orin Nano cuda	2	144	June 18, 2024
Docker image for jetpack 4.4 with cuda driver,tensorrt library installed Jetson TX2 docker	2	377	October 18, 2021
My jetson nano board returns 'False' to torch.cuda.is_available() in local directory Jetson Nano cuda	6	12533	August 31, 2021
How to get cuda working in a docker container for pytorch applications Jetson AGX Orin docker , pytorch	3	809	May 30, 2024
GPU access inside Docker in Jetson Xavier Jetson Xavier NX cuda , docker	2	2014	March 7, 2022
PyTorch doesn't see Cuda device Jetson AGX Xavier pytorch	5	3474	March 23, 2022