Please provide the following information when requesting support. • Hardware (T4/V100/Xavier/Nano/etc) • Network Type :Yolo_v4 • TLT Version Configuration of the TAO Toolkit Instance task_group: [‘model’, ‘dataset’, ‘deploy’] format_version: 3.0 toolkit_version: 5.5.0 published_date: 08/26/20…

Please reinstall driver and reboot. Thanks. Refer to TA0 v3.21.08 - pycuda._driver.LogicError: cuInit failed: system not yet initialized - #3 by Morganh .

I tried reinstalling driver and rebooting as mentioned but still having the same error. Also tried what was said in the post : $ python * *>>>import pycuda* *>>> import pycuda.driver as cuda* *>>> cuda.init()* * that showed me nothing. and $ nvidia-smi $ dpkg -l |grep cuda ii cuda-cccl-12-1 …

[image] valentin1234: showed me nothing. Showing nothing will be correct result. Do you mean there is not error now when you run >>> import pycuda.driver as cuda >>> cuda.init()

At first i had ModuleNotFoundError: No module named ‘pycuda’ so i did pip install pycuda And now, i’ve no errors when running : >>> import pycuda.driver as cuda >>> cuda.init() But still get errors on the notebook

[image] valentin1234: But still get errors on the notebook What are the errors? Can you share the full logs?

When running : # If you use your own dataset, you will need to run the code below to generate the best anchor shape !tao model yolo_v4 kmeans \ -l $DATA_DOWNLOAD_DIR/kitti_split/training/label \ -i $DATA_DOWNLOAD_DIR/kitti_split/training/image \ -n 9 \ -x 960 \ -y 544 \ # -e…

Please run ! tao model yolo_v4 run /bin/bash Then inside the docker, run below. # nvidia-smi # python # >>> import pycuda.driver as cuda >>> cuda.init()

Here is what i get : 2025-02-17 10:35:21,125 [TAO Toolkit] [INFO] root 160: Registry: ['nvcr.io'] 2025-02-17 10:35:21,198 [TAO Toolkit] [INFO] nvidia_tao_cli.components.instance_handler.local_instance 360: Running command in container: nvcr.io/nvidia/tao/tao-toolkit:5.0.0-tf1.15.5 2025-02-17 10:35:…

No CUDA-capable device is detected

Accelerated Computing Intelligent Video Analytics TAO Toolkit

Morganh February 17, 2025, 9:41am 10

It is not related to TAO.
Please check if the hint from Nvida Container Toolkit: Failed to initialize NVML: Unknown Error - #3 by SimonBirrell can help you.

Topic		Replies	Views
No CUDA-capable device is detected - yolov4 TAO Toolkit	9	441	August 16, 2024
TA0 v3.21.08 - pycuda._driver.LogicError: cuInit failed: system not yet initialized TAO Toolkit cuda , tao	10	1135	April 29, 2024
No CUDA-capable device is detected on tao detectnet_v2 dataset convert TAO Toolkit pycuda , omniverse_extension	12	6530	December 15, 2021
pycuda._driver.Error: cuInit failed: unknown error TAO Toolkit yolo , pycuda , tao	7	3910	January 5, 2022
TAO 4.0 AutoML - the provided PTX was compiled with an unsupported toolchain TAO Toolkit	5	845	July 17, 2023
Can't run tao 3.0 w/ RTX A6000 TAO Toolkit	14	1466	December 17, 2021
Problem with tlt file mounting TAO Toolkit	28	3019	December 23, 2021
TAO Toolkit 5.5.0 - cuInit failed: no CUDA-capable device is detected TAO Toolkit cuda	5	417	January 14, 2025
Docker instantiation failed when running tao ssd TAO Toolkit	16	1220	December 4, 2021
Unable to train yolov4 with Tao succesfully TAO Toolkit	5	701	April 28, 2023

No CUDA-capable device is detected

Related topics