qengineering / jetson-nano-image Goto Github PK

Do you have any tips to upgrate python3.6 to python3.8 and still use the pre-installed library?

I just tried to remove python 3.6 and replace it with python 3.8, then reinstall library for 3.8 version, but I got a lot of error and opencv 4.5.5 didn't work.

Hi Qengineers!
First of all: Thanks a lot for that awsome image!! Before, I had huge problems with installing Yolo-dependencies because of missing python 3.7. Now, its perfectly working. ;)
But: It came out another issue with preinstalled opencv. For my project I want to visualize a webcamstream on a hdmi-connected display. Using the cv2.imwrite function always outputs the following error message:

qt.qpa.xcb: could not connect to display 
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "/home/jetson/.local/lib/python3.8/site-packages/cv2/qt/plugins" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

Available platform plugins are: xcb.

Aborted (core dumped)

It seems like WITH_QT variable was disabled on installation. Is there a possibility to enable qt support afterwards? Or does somebody has another solution for me?

Thanks a lot!
Chris

It would be nice to include a step by step guide to reproduce the build of the image.

In that way others can try to fix possible issues with some installs.

When I try to list the physical GPU using the Tensorflow using the command "tf.config.list_physical_devices(device_type='GPU')". it gives me the following error .

I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set
I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1
E tensorflow/stream_executor/cuda/cuda_driver.cc:328] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is
I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (nano): /proc/driver/nvidia/version does not exist

I use Jetson nano 4G B01

Hello, just flashed the image and I am unable to import tensorflow with python or python3?

Hi all,

Thanks so much for providing an image file, but currently the Gdrive link is broken. Can you please update? This is so useful to have.

The broken link is this one: https://drive.google.com/file/d/1753wRCFUV7U6vJa3H5xC1cq5MIcGKwFm/view?usp=sharing

Thanks!

The instructions in the README say that the password is jetson, but what's the username? I'm trying to use your image on a headless jetson nano and to SSH into it (hoping there's an SSH server running out of the box) I also need the username. I tried to connect via serial but it didn't work.

Hi, I downloaded 7zip file (about 14GB) and unzipped.

however the img file is not 21GB but 57.7GB (62,021,173,248 bytes), so I couldn't flash the image on my SD card (32GB).

Is the actual image size 57.7GB? Or is there something I missed?

Hi!
Please can you share it in a different repo? Mega is frustrating.

I'm trying to run a project that requires torch>=1.10, but when I try to install it I get "No matching distribution found for torch>=1.10 (from lightweight-gan)". Could you please tell me how I can install torch?
Maybe you could also include it in your next release
The full log for reference:

(lightweight-gan) jetson@nano:~/lightweight-gan$ pip3 install lightweight-gan
Collecting lightweight-gan
  Using cached https://files.pythonhosted.org/packages/6b/ea/1ed30e2285ef18fd62f4bfe38183377c9a3cc352967e45e961d7179773f1/lightweight_gan-0.23.2-py3-none-any.whl
Collecting retry (from lightweight-gan)
  Using cached https://files.pythonhosted.org/packages/4b/0d/53aea75710af4528a25ed6837d71d117602b01946b307a3912cb3cfcbcba/retry-0.9.2-py2.py3-none-any.whl
Collecting kornia>=0.5.4 (from lightweight-gan)
  Using cached https://files.pythonhosted.org/packages/50/33/89ebc68f8712c29dc451e79708451192b55e788ddfddfc1abbcab936e374/kornia-0.6.5-py2.py3-none-any.whl
Collecting torch>=1.10 (from lightweight-gan)
  Could not find a version that satisfies the requirement torch>=1.10 (from lightweight-gan) (from versions: )
No matching distribution found for torch>=1.10 (from lightweight-gan)

Hi Sir,

Any possibility of you working and releasing ANPR on jetson nano ?

Hello,
Thank you for providing the JetsonNano.img file but it looks like the file has some issues that need attention. I'm getting this error when I use Etcher. I'm also getting an error when I want to extract the image.

.

Jetson-Nano-image just gets flash screen never completes boot

How to proceed?

Thank you very much for the image!
I used balenaEtcher to flash this image to a 128GB SD Card, but I can't boot my Jetson Nano B01 4GB, it shows:

[1.175793] tegradc tegradc.1: dpd enable lookup fail:-19
[1.323835] imx219 7-0010: imx219_board_setup: error during i2c read probe (-121)
[1.323901] imx219 7-0010: board setup failed
[1.347757] imx219 8-0010: imx219_board_setup: error during i2c read probe (-121)
[1.347815] imx219 8-0010: board setup failed
cp: not writing throught dangling symlink 'etc/resolve.conf'
[2.404920] Kernel panic - not syncing:Attempted to kill init! exitcode=0x00000000
[2.404920]
[2.414104] CPU:1 Pid: 1 Comm: sh Not tainted 4.9.253-tegra #1
[2.420047] Hardware name: NVIDIA Jetson Nano Developer Kit(DT)
[2.426075] Call trace:
……
……
……
[2.575663] Rebooting in 5 seconds..

and reboot endlessly...

Could you give me some suggestions about it?
It would be nice if you provide the md5sum of this image, then I can see if I had trouble downloading the image.

Hi!
I installed current image on SD catd.
I tried to run docker:

sudo systemctl start docker.service
Job for docker.service failed because the control process exited with error code.
See "systemctl status docker.service" and "journalctl -xe" for details.

But trouble not in Docker, trouble in kernel:

root@nano:~# cd /lib/modules/4.9.253-tegra/kernel/fs
root@nano:/lib/modules/4.9.253-tegra/kernel/fs# ls -a
.  ..  binfmt_misc.ko  btrfs  cifs  fuse  nfs_common  nfsd  overlayfs
root@nano:/lib/modules/4.9.253-tegra/kernel/fs# cd overlayfs/
root@nano:/lib/modules/4.9.253-tegra/kernel/fs/overlayfs# ls -al
total 100
drwxr-xr-x 2 root root  4096 aug  2  2021 .
drwxr-xr-x 8 root root  4096 aug  2  2021 ..
-rw-r--r-- 1 root root 92760 jul 26  2021 overlay.ko
root@nano:/lib/modules/4.9.253-tegra/kernel/fs/overlayfs# insmod ./overlay.ko 
insmod: ERROR: could not insert module ./overlay.ko: Invalid module format
root@nano:/lib/modules/4.9.253-tegra/kernel/fs/overlayfs# insmod overlay.ko 
insmod: ERROR: could not insert module overlay.ko: Invalid module format
root@nano:/lib/modules/4.9.253-tegra/kernel/fs/overlayfs# 


 uname -a
Linux nano 4.9.337-tegra #1 SMP PREEMPT Thu Jun 8 21:19:14 PDT 2023 aarch64 aarch64 aarch64 GNU/Linux

So, question: where I can get overlay.ko files and to fix trouble with Kernel?

Hi, Team!

I have such problem https://forums.developer.nvidia.com/t/why-it-doesnt-see-cuda-docker-container/264897

Could we solve them or I should create my own docker image and mount all needful libraries by myself?

Currently running jtop at any point makes it so that the system ends up crashing when the machine is rebooted

Here are some screenshots that might be useful for you guys

My plan right now is that I'm just going to reinstall everything from scratch and make sure that I don't run jtop, but I would like to use jtop in the future since it is a really nice utility

Overall, I ❤️ the work you guys have put into this!

Hi great work. Could you please provide a Dockerfile to be able creating a Docker Image ?

Hi. Just installed this image and checked the version of OpenCV under Python3.
It shows 4.1.1, not 4.5.3.
Python (2.7) shows OpenCV 4.5.3 though.
How can I use CV2 4.5.3 with Python3?
Thanks

Hello,

can the figure be used for the Jetson Nano 2GB? Would it be possible that the pre-installed frameworks like OpenCV, Pytorch etc. are also updated?

Can be set during the installation or before that the CPU runs at 2.0?

I was using this Nano image for OpenCV+CUDA support and it was a huge relief, seeing all the packages pre-installed. I use the EdiMax EW-7811un Wi-Fi adapter. I tried installing the Wi-Fi drivers for the same, using the following guide : https://learn.sparkfun.com/tutorials/adding-wifi-to-the-nvidia-jetson/all , and understood that the Wi-Fi drivers weren't installing and connection to Wi-Fi was not possible still. But, it worked on the regular JetPack 4.6.3 version(the stock NVIDIA Image), without all the DNN frameworks. What could be the possible issue?

I have copied the ubuntu 20.04 image with JP 4.6 from QEngineering. Its of great help.
But I am stuck with capturing video and images.
I have tried CSI-Camera/Simple_camera.py
It works great with python whilst python3 throws errors it says cannot open camera.
Can you pls help me. Iam using YOLOV8 inference/detection with custom model.

File downloading is canceled by network error.
Can you please upload the image file to anywhere else?

Since my project only uses pytorch and my sd card is 32 GB, so I'd like to delete the tensorflow and other libraries that I dont need.
My question is: how to delete those libraries safely?

Tensortrt does not work. I am trying to run simple code in python3 -
import tensorrt as trt
trt.version

Also have issues in installing onnx and onnxsim modules with this image.

Can someone help me on this.

Thank you very much for the image file.
It's very much appreciated.

Unfortunately, I cannot flash it because balena etcher sais that my media is 2.1MB smaller than the image file.
(I am using a 32GB SD Card)
How do you suggest me to shrink the image file?

Thanks a lot!

Hello. Yesterday, I downloaded the clone file and successfully installed it on the SD card. I checked that the opencv, pytorch, and torchvision files are installed using the import and print functions. However, when I import and check for CUDA enable with print output, it does not appear to be enabled. Is it like this in the clone file, or do I need to follow the steps mentioned in this link: https://github.com/Qengineering/Install-OpenCV-Jetson-Nano and reinstall? I'm new to Jetson Nano, can you help me with this?

Thanks for the image, I've tried it and works amazing!

I think it would be useful to a lot of people (including myself) if the kernel in this image included support for the PPS device tree. Here is a tutorial on how to do this: link. I would do this on my own Nano, but I don't have an Ubuntu 18.04 machine to run the Jetson SDK and flash the Nano, so I'd like for it to come included in the SD image.

Would this be possible? I'd greatly appreciate it!

From some initial research, there may be a conflict unpacking .7z files on OS-X. I tried creating images in 3 different ways, with the native unzip as described on the Nvidia imaging page, with the Unarchiver, and with a homebrew install of 7zip. The JetsonNano.7z unpacks into a .img file (which is required for both Imager and Etcher, as neither support .7z), and after several attempts in various ways (using both imager, etcher, and cmd line instructions ) the same thing happens at every boot attempt, as follows:

[   1.177435] tegradc tegradc.1: dpd enable lookup fail:-19
[   1.385966] imx219 7-0010: imx219_board_setup: error during i2c read probe (-121)
[   1.386098] imx219 7-0010: board setup failed
[   1.410380] imx219 8-0010: imx219_board_setup: error during i2c read probe (-121)
[   1.410457] imx219 8-0010: board setup failed
mount: /mnt: wrong fs type, bad option, bad superblock on /dev/mcblk0p1, missing codepage or helper program, or other error.
bash: cannot set terminal process group (-1): Inappropriate ioctl for device
bash: no job control in this shell
bash-4.4#

ls shows standard directories and in the /dev/ folder shows mmcblk0 through mmcblk0p9

Is there a mounting procedure that can correct this from here?

I’m trying to run a TensorRT Tiny YOLOv4 model on a 4GB Nano development board.

I’m running the Nano headless, so should have enough memory :

021-12-14 07:54:04.568367: F tensorflow/core/kernels/image/resize_bilinear_op_gpu.cu.cc:447] Non-OK-status: GpuLaunchKernel(kernel, config.block_count, config.thread_per_block, 0, d.stream(), config.virtual_thread_count, images.data(), height_scale, width_scale, batch, in_height, in_width, channels, out_height, out_width, output.data()) status: Internal: too many resources requested for launch
Fatal Python error: Aborted