Just to let you know, my beautiful wife has been cutting my hair for nearly 40 years!! Keep up the great tutorials!! I have to say I agree with LisaM; awesome teacher!!
Your a brave man to cut your own hair, Kudos. Oh yeah that is really awesome, it's actually picking out the correct objects. You've really put a lot of work on this, each one draws you in more and more. Looking forward to next weeks video.
Hi Paul, Another day, another lesson. Got the webcam and Pi cam working, the Pi cam image is cropped so it does not show the whole image, not sure why.
hello, the compiler was shouting at me for writing confident..but when i changed to confidence the program worked. The recognition is not perfect but i think it due to the 'googlenet' or something
Thanks this video and other videos. They’re really helpful . And I'm trying to implement kalman filter in 3 dimension(x,y,z position). I have gps coordinate and I wnat to predict next position with kalman filter. how can I do this problem's math model? Can you help me about this topic? Any examples about that or learn more about this topic?
Hi Paul, Since PyTorch 1.1.0 and torchvision 0.3.0 appear to be broken now, Dusty at NVIDIA suggested that I install PyTorch 1.4.0 and torch 0.5.0 which I successfully accomplished and now I can import both torch and torchvision into python3. However, since I did not have the “The Model Downloader” in jetson-inference build PyTorch 1.1.0 for python3.6, the "make -j$(nproc)" executed after cmake ../ ends differently than when using PyTorch 1.1.0. My make ends with the following: [ 93%] Built target jetson-inference-python-27 [ 94%] Built target homography-camera [ 95%] Built target trt-bench [100%] Built target jetson-inference-python-36 Where as yours ends with "[100%] Built target camer-capture" and does not include "Built target homography-camera" and "Built target trt-bench". I assume that this is because “The Model Downloader” is looking for PyTorch 1.1.0 and I have manually installed PyTorch 1.4.0 to correct the PyTorch 1.1.0 and torch 0.3.0 compatibility issue? Regards, Tom C
You are out of my pay grade. On jetpack 4.3, I just followed Dusty's original instructions on Hello AI world, and it all just worked. Really do not understand all the different dependencies, so if that does not work, dont know enough to be helpful.
@@paulmcwhorter Yes, Dusty was mystified as to why torchvision 0.3.0 was failing to import into Python3, when installing PyTorch 1.1.0 from "The Model Downloader", and thought that it might have to do with some unexpected update to torchvision 0.3.0 making it incompatible with PyTorch 1.1.0. He indicated that he had already moved on to Jetpack 4.4 and therefore would not be able to duplicate my issue with PyTorch 1.1.0 and torchvision 0.3.0 to be able troubleshoot it. I have asked him how "The Model Downloader" is affected by not installing PyTorch 1.1.0 for Python3 in relation to the build function and I have yet to hear back from him. Bottom line here is that anyone who wants to follow your tutorial will get the same error that I encountered when trying to import torchvision into python3 if they following your instructions for "The Model Downloader" build, leaving us dead in the water as to installing the "NVIDIA Object Detection and Inference Tools". Hopefully Dusty might come up with a solution for the "The Model Downloader" build issue with PyTorch. If not, future tutorial installers might have to move on to Jetpack 4.4 in order to be able to successfully install the "NVIDIA Object Detection and Inference Tools". Regards, Tom C
And the problem with jetpack 4.4 is that is breaks the face recognizer library, and brings it down to a few frames per second, where it runs at over 10 fps on jetpack 4.3. We need some stable build to work on. It is a really big investment for me to make a tutorial series, and if it becomes obsolete because the core os behave differently a month or so later, that leads to a platform you can not educate on or build on.
@@paulmcwhorter I hear you Paul, but right now anyone new to this tutorial, like me, is obviously going to have a problem with importing torchvision into python3 so they are locked out from moving ahead with future tutorials if future tutorials use torchvision. Here is what Dusty had to say about the PyTorch installation: "Glad you were able to get it working with PyTorch 1.4 and torchvision 0.5 - you don’t have to change anything in the ‘Model Downloader’, because that actually doesn’t download PyTorch models (only ONNX, caffe, and UFF). And even if it did download PyTorch models, those aren’t typically dependent on the PyTorch version if saved correctly. However as far as the ‘PyTorch Installer’ tool goes, you should skip that, because it will just want to install PyTorch 1.1 and torchvision 0.3 again. It sounds like I need to change that tool to only install PyTorch 1.4 and torchvision 0.5 when on JetPack 4.3. Thanks for letting me know about it." Sounds like you should add a warning not to install PyTorch 1.1.0 during "The Model Downloader" and install PyTorch 1.4.0 and torchvision 0.5.0 manually before running "The Model Downloader"? Regards, Tom Coyle
Update: Dusty at NVIDA has updated the install-pytorch.sh script to use PyTorch 1.4 / torchvision 0.5 on JetPack 4.2 and JetPack 4.3. See commit 8a7acf to jetson-inference master branch so there will be no need for a warning about the PyTorch install during "The Model Downloader" execution. Regards, Tom Coyle
Hi Paul, Thanks for making such a detailed video tutorial series. I am facing an issue while running jetson-inference. *When I am running my python code in code-oss terminal. It is running endless with lot of processing messages but does not produce detection windows(display).* I tried re-installing again but nothing helped.
Hi Paul, as i do not have a webCam, so i have to use '/dev/video0'. However, this does not work and i replace the '/dev/video0' with just a '0'. So, cam = jetson.utils.gstCamera(1920,1080, '0'), just that the video would be inverted and i have not idea how to add in the flip parameter to the above. Any advice would be good ?
Paul, I'm able to view my IP Cam via rtsp in opencv "cam=cv2.VideoCapture("rtsp link"). Any idea how to get rtsp working in this lesson with "cam =jetson.utils.gstCamera"? Thanks in Advance
Hi Paul, Apparently ldconfig did not link torch and torch vision modules to python. I captured a CLI output from when I hit the OK button for the PyTorch installation for Python 3.6 to the return of the CLI bash line. I believe that PyTorch did not install correctly based on the CLI output of the PyTorch installation. Apparently I have experienced some "cockpit error" and need to go back over my installed software and see what is missing. Regards, Tom Coyle
I went back over jetson nano tutorials 36-38 and found the missing software installs and updated my software installation. I am now able to import torch into python3, however attempting to import torchvision results in this error: >>> import torchvision Traceback (most recent call last): File "", line 1, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/__init__.py", line 1, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/__init__.py", line 11, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/detection/__init__.py", line 1, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/detection/faster_rcnn.py", line 7, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/ops/__init__.py", line 1, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/ops/boxes.py", line 2, in File "", line 971, in _find_and_load File "", line 955, in _find_and_load_unlocked File "", line 656, in _load_unlocked File "", line 626, in _load_backward_compatible File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/_C.py", line 8, in File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/_C.py", line 7, in __bootstrap__ TypeError: exec_module() missing 1 required positional argument: 'module' >>> Do you have any idea what could be causing the torchvision import problem based on this error message? Regards, Tom Coyle
I think it would be best to start over with a new flash of jetpack 4.3, and just start over. The commands to install pytorch and torchvision from the earlier videos should work.
@@paulmcwhorter Thanks for the quick response and suggestion. That is what I did to get to the point of being able to import torch, but still not importing torchvision.
Dusty at Nvidia has recommended I do this: sudo apt-get install libjpeg-dev zlib1g-dev $ git clone --branch v0.3.0 github.com/pytorch/vision torchvision $ cd torchvision $ sudo python3 setup.py install $ cd ../ # attempting to load torchvision from build dir will result in import error $ pip3 install 'pillow
I was trying to install the packages you did on the first menu but I lost internet connection so I didn't get half of them. Any ideas on how to redo the download so that I can have the same packages as you? Thanks in advance
I hit enter too soon and did not get all the options checked. If I just cmake again it does not bring up the option box. Is there a way to go back and check all required boxes?
When i run the cmake ../ command it ipen a window there is not models or library instead it shows to install pytorch that's all ,do need to install jetpack 4.4
Intellisense for jetson.inference and jetson.utils appears to be broken. Tried to use the general techniques you used in Lesson 48 (fixing intellisense with cv2) but so far no luck. Hope we can get inselllisense back. I'm also using jetpack 4.4 Great lesson!
Hi Quaternion, I sent the following to Paul concerning tutorial 48: "Hi Paul, Problem here. In my version of code-oss (1.32.3), "Preferences: Open Settings (JSON)" looks like this: { "workbench.colorTheme" : "Default Light+" , "editor.fontSize" : 16 } There is no Python path in "Open Settings (JSON)", unless you added a Python path to "Open Settings (JSON) sometime in the past. If that is not the case, why is my "Open Settings (JSON)" different from yours and how should I proceed?" I am running Jetpack 4.3 and had an issue with the PyTorch install that you can read about on this tutorial's comments. When compiling code involving jestson-inference and jetson-utilities the auto complete seemed to work reliability without Paul's update. Regards, Tom C
This works fine when I have hdmi cable plugged into nano, but fails to run when using remote desktop interface. Error: MESA LOADER failed to open swrast Failed to load driver:swrast Anyone has any idea on this issue?
Hi, skipped ahead and installed jetpack4.4 and found a bug as follows: with jetson.inference.imageNet("googlenet"). bug is in jetson-inference/plugins/FlattenConcat.cpp namespace { const char* FLATTENCONCAT_PLUGIN_VERSION{"1"}; const char* FLATTENCONCAT_PLUGIN_NAME{"FlattenConcat_TRT"}; --> change to _FlattenConcat_TRT } */ make clean make j4 make install appears to work. thanks to rjknight123 who also found it in a different context.
@@paulmcwhorter Nothing wrong with your face Paul, I'm just saying it doesn't need to be so large when the real information is on the screen behind your face.
Jump right in all you other Linux gurus! Once your done the sudo apt-get update, isn't a sudo apt-get upgrade necessary. Doesn't the "update" just update the catalog and determine what needs to be "upgraded"? Maybe I've read the documentation wrong?
I think that's true, but if you are installing new things from apt without upgrading your old stuff, don't you want to update the catalog and not run an upgrade? I am not a linux guru, but I think that's probably correct.
![[LIVE] : ONE ลุมพินี 87 วันนี้!! คู่เอก "ก้องชัย vs โชคปรีชา"](http://i.ytimg.com/vi/YN6e7HVnekw/mqdefault.jpg)
I stumbled upon this him bout a month ago... I have to say, he's the best teacher ever.... God bless this man.
Thanks!
years later And he still does the best tutorial videos keep it up :D
Paul, this lesson was one of the best! Thank you for all of your hard work.
Just to let you know, my beautiful wife has been cutting my hair for nearly 40 years!! Keep up the great tutorials!! I have to say I agree with LisaM; awesome teacher!!
Successfully run in Jetson Xavier AGX. Thank you for all of your hard work.
Best Jetson Nano Videos Ever!!!
Lessons keeps getting interesting. So much stuff to learn!
Glad to hear it!
Your a brave man to cut your own hair, Kudos. Oh yeah that is really awesome, it's actually picking out the correct objects. You've really put a lot of work on this, each one draws you in more and more. Looking forward to next weeks video.
I knew there would be something better than OpenCV! :) GREAT job here.
Works like a SpaceX rocket in NX, fantstic lesson Paul
Once again, A wonderful lesson
amend my advice to read "sudo make install" of course
Hi Paul,
Another day, another lesson. Got the webcam and Pi cam working, the Pi cam image is cropped so it does not show the whole image, not sure why.
hello, the compiler was shouting at me for writing confident..but when i changed to confidence the program worked. The recognition is not perfect but i think it due to the 'googlenet' or something
Thanks this video and other videos. They’re really helpful . And I'm trying to implement kalman filter in 3 dimension(x,y,z position). I have gps coordinate and I wnat to predict next position with kalman filter. how can I do this problem's math model? Can you help me about this topic? Any examples about that or learn more about this topic?
What happened to the example at 26:48 ? Switched to a new file after installing the new things.
Great lesson! Thank you so much.
Hi Paul,
Since PyTorch 1.1.0 and torchvision 0.3.0 appear to be broken now, Dusty at NVIDIA suggested that I install PyTorch 1.4.0 and torch 0.5.0 which I successfully accomplished and now I can import both torch and torchvision into python3. However, since I did not have the “The Model Downloader” in jetson-inference build PyTorch 1.1.0 for python3.6, the "make -j$(nproc)" executed after cmake ../ ends differently than when using PyTorch 1.1.0. My make ends with the following:
[ 93%] Built target jetson-inference-python-27
[ 94%] Built target homography-camera
[ 95%] Built target trt-bench
[100%] Built target jetson-inference-python-36
Where as yours ends with "[100%] Built target camer-capture" and does not include "Built target homography-camera"
and "Built target trt-bench". I assume that this is because “The Model Downloader” is looking for PyTorch 1.1.0 and I have manually installed PyTorch 1.4.0 to correct the PyTorch 1.1.0 and torch 0.3.0 compatibility issue?
Regards,
Tom C
You are out of my pay grade. On jetpack 4.3, I just followed Dusty's original instructions on Hello AI world, and it all just worked. Really do not understand all the different dependencies, so if that does not work, dont know enough to be helpful.
@@paulmcwhorter Yes, Dusty was mystified as to why torchvision 0.3.0 was failing to import into Python3, when installing PyTorch 1.1.0 from "The Model Downloader", and thought that it might have to do with some unexpected update to torchvision 0.3.0 making it incompatible with PyTorch 1.1.0.
He indicated that he had already moved on to Jetpack 4.4 and therefore would not be able to duplicate my issue with PyTorch 1.1.0 and torchvision 0.3.0 to be able troubleshoot it. I have asked him how "The Model Downloader" is affected by not installing PyTorch 1.1.0 for Python3 in relation to the build function and I have yet to hear back from him.
Bottom line here is that anyone who wants to follow your tutorial will get the same error that I encountered when trying to import torchvision into python3 if they following your instructions for "The Model Downloader" build, leaving us dead in the water as to installing the "NVIDIA Object Detection and Inference Tools". Hopefully Dusty might come up with a solution for the "The Model Downloader" build issue with PyTorch. If not, future tutorial installers might have to move on to Jetpack 4.4 in order to be able to successfully install the "NVIDIA Object Detection and Inference Tools".
Regards, Tom C
And the problem with jetpack 4.4 is that is breaks the face recognizer library, and brings it down to a few frames per second, where it runs at over 10 fps on jetpack 4.3. We need some stable build to work on. It is a really big investment for me to make a tutorial series, and if it becomes obsolete because the core os behave differently a month or so later, that leads to a platform you can not educate on or build on.
@@paulmcwhorter I hear you Paul, but right now anyone new to this tutorial, like me, is obviously going to have a problem with importing torchvision into python3 so they are locked out from moving ahead with future tutorials if future tutorials use torchvision.
Here is what Dusty had to say about the PyTorch installation:
"Glad you were able to get it working with PyTorch 1.4 and torchvision 0.5 - you don’t have to change anything in the ‘Model Downloader’, because that actually doesn’t download PyTorch models (only ONNX, caffe, and UFF). And even if it did download PyTorch models, those aren’t typically dependent on the PyTorch version if saved correctly.
However as far as the ‘PyTorch Installer’ tool goes, you should skip that, because it will just want to install PyTorch 1.1 and torchvision 0.3 again. It sounds like I need to change that tool to only install PyTorch 1.4 and torchvision 0.5 when on JetPack 4.3. Thanks for letting me know about it."
Sounds like you should add a warning not to install PyTorch 1.1.0 during "The Model Downloader" and install PyTorch 1.4.0 and torchvision 0.5.0 manually before running "The Model Downloader"?
Regards, Tom Coyle
Update: Dusty at NVIDA has updated the install-pytorch.sh script to use PyTorch 1.4 / torchvision 0.5 on JetPack 4.2 and JetPack 4.3. See commit 8a7acf to jetson-inference master branch so there will be no need for a warning about the PyTorch install during "The Model Downloader" execution.
Regards, Tom Coyle
It's been back to the barracks cuts since May
Hi Paul, Thanks for making such a detailed video tutorial series. I am facing an issue while running jetson-inference. *When I am running my python code in code-oss terminal. It is running endless with lot of processing messages but does not produce detection windows(display).* I tried re-installing again but nothing helped.
Cool, I will be up to speed by the end of this video. Time to start messing with arduino.
Hi Paul, as i do not have a webCam, so i have to use '/dev/video0'. However, this does not work and i replace the '/dev/video0' with just a '0'. So, cam = jetson.utils.gstCamera(1920,1080, '0'), just that the video would be inverted and i have not idea how to add in the flip parameter to the above. Any advice would be good ?
How to specify a resolution using a CSI camera?
Paul, I'm able to view my IP Cam via rtsp in opencv "cam=cv2.VideoCapture("rtsp link"). Any idea how to get rtsp working in this lesson with "cam =jetson.utils.gstCamera"? Thanks in Advance
Hi Paul,
Apparently ldconfig did not link torch and torch vision modules to python.
I captured a CLI output from when I hit the OK button for the PyTorch installation for Python 3.6 to the return of the CLI bash line.
I believe that PyTorch did not install correctly based on the CLI output of the PyTorch installation. Apparently I have experienced some "cockpit error" and need to go back over my installed software and see what is missing.
Regards,
Tom Coyle
I went back over jetson nano tutorials 36-38 and found the missing software installs and updated my software installation.
I am now able to import torch into python3, however attempting to import torchvision results in this error:
>>> import torchvision
Traceback (most recent call last):
File "", line 1, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/__init__.py", line 1, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/__init__.py", line 11, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/detection/__init__.py", line 1, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/models/detection/faster_rcnn.py", line 7, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/ops/__init__.py", line 1, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/ops/boxes.py", line 2, in
File "", line 971, in _find_and_load
File "", line 955, in _find_and_load_unlocked
File "", line 656, in _load_unlocked
File "", line 626, in _load_backward_compatible
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/_C.py", line 8, in
File "/usr/local/lib/python3.6/dist-packages/torchvision-0.3.0-py3.6-linux-aarch64.egg/torchvision/_C.py", line 7, in __bootstrap__
TypeError: exec_module() missing 1 required positional argument: 'module'
>>>
Do you have any idea what could be causing the torchvision import problem based on this error message?
Regards,
Tom Coyle
I think it would be best to start over with a new flash of jetpack 4.3, and just start over. The commands to install pytorch and torchvision from the earlier videos should work.
@@paulmcwhorter Thanks for the quick response and suggestion. That is what I did to get to the point of being able to import torch, but still not importing torchvision.
Dusty at Nvidia has recommended I do this:
sudo apt-get install libjpeg-dev zlib1g-dev
$ git clone --branch v0.3.0 github.com/pytorch/vision torchvision
$ cd torchvision
$ sudo python3 setup.py install
$ cd ../ # attempting to load torchvision from build dir will result in import error
$ pip3 install 'pillow
Unfortunately uninstalling torchvision and then reinstalling per dusty's commands did not work and I am still getting the same error as before.
Paul, installing object detection steps are same for NX as well ? or different ?
should be the same
Jetson Utils gstCamera Only Support All CSI (IMX ...) or RaspberryPI CAM?
Hello Paul. How do you propperly terminate this program? I kill it, but I think something's left behind.
I was trying to install the packages you did on the first menu but I lost internet connection so I didn't get half of them. Any ideas on how to redo the download so that I can have the same packages as you? Thanks in advance
I hit enter too soon and did not get all the options checked. If I just cmake again it does not bring up the option box. Is there a way to go back and check all required boxes?
paul i am having issue with v4l-utils please help
In 2023: No module named "torchvision", any suggestion? My JetPack is 4.6
thanks for a grate lesson.
does error (Could not register plugin creator: ::FlattenConcat_TRT)
have something to do with jetpack 4.4?
how can i call my own trained model in the code ?
and thanks ur videos are very helpful
Any way to make intellisense work on these new libraries?
When i run the cmake ../ command it ipen a window there is not models or library instead it shows to install pytorch that's all ,do need to install jetpack 4.4
If import torch or torchvision does not work go to home and type
$ cd jetson-inference/build
$ ./install-pytorch.sh
and if it is not working you can use chmod +x install-pytorch.sh after that u can use ./install-pytorch.sh
Thank you! Turns out I did not select it before pressing the OK button.
@@alpertemizel9892 Thank you!
Intellisense for jetson.inference and jetson.utils appears to be broken. Tried to use the general techniques you used in Lesson 48 (fixing intellisense with cv2) but so far no luck. Hope we can get inselllisense back. I'm also using jetpack 4.4 Great lesson!
I was briefly on jetpack 4.4 on the nano, and then retroed back to 4.3. I had many problems on 4.4 including very poor fps on a number of programs.
If you figure it out, post an update if you don't mind.
Hi Quaternion, I sent the following to Paul concerning tutorial 48:
"Hi Paul, Problem here. In my version of code-oss (1.32.3), "Preferences: Open Settings (JSON)" looks like this:
{
"workbench.colorTheme" : "Default Light+" ,
"editor.fontSize" : 16
}
There is no Python path in "Open Settings (JSON)", unless you added a Python path to "Open Settings (JSON) sometime in the past.
If that is not the case, why is my "Open Settings (JSON)" different from yours and how should I proceed?"
I am running Jetpack 4.3 and had an issue with the PyTorch install that you can read about on this tutorial's comments. When compiling code involving jestson-inference and jetson-utilities the auto complete seemed to work reliability without Paul's update.
Regards,
Tom C
@@paulmcwhorter ah.. I installed everything twice thinking I had done something wrong installing 4.4. Back on 4.3 now too.. what a pain!
awesome stuff
it seems that this setup doesn't work on jetpack 4.6
Have any one could setup correctly?
Ha. I had my wife cut my hair with the dog clippers. But made it to a professional shop this week.
This works fine when I have hdmi cable plugged into nano, but fails to run when using remote desktop interface.
Error:
MESA LOADER failed to open swrast
Failed to load driver:swrast
Anyone has any idea on this issue?
Thank you!
Hi, skipped ahead and installed jetpack4.4 and found a bug as follows:
with jetson.inference.imageNet("googlenet").
bug is in jetson-inference/plugins/FlattenConcat.cpp
namespace
{
const char* FLATTENCONCAT_PLUGIN_VERSION{"1"};
const char* FLATTENCONCAT_PLUGIN_NAME{"FlattenConcat_TRT"}; --> change to _FlattenConcat_TRT
}
*/
make clean
make j4
make install
appears to work.
thanks to rjknight123 who also found it in a different context.
Good video, but it sure would be better if he kept his face image reduced.
Actually some people like my face, saying it brings a sense of calm in a crazy world spinning out of control
@@paulmcwhorter Nothing wrong with your face Paul, I'm just saying it doesn't need to be so large when the real information is on the screen behind your face.
My better half gives me a #2 buzz about once a month!
I prefer a #3 on the side, and then the top with sciscors.
@@paulmcwhorter For me, nothing on the top! A lot like my avatar.
nice hair cut. the easier way is to shave it all at zero on the clippers. ;)
Jump right in all you other Linux gurus! Once your done the sudo apt-get update, isn't a sudo apt-get upgrade necessary. Doesn't the "update" just update the catalog and determine what needs to be "upgraded"? Maybe I've read the documentation wrong?
I think that's true, but if you are installing new things from apt without upgrading your old stuff, don't you want to update the catalog and not run an upgrade? I am not a linux guru, but I think that's probably correct.
anyone facing issues in installing v4l-utils??
That me