Failed install pytorch in jetson nano

Autonomous Machines Jetson & Embedded Systems Jetson Nano
,
1

Hi, i am trying to using Jetson Nano and watching Nvidia tutorial about training object detection with ssd model but get stuck when installed pytorch using this as instructions https://github.com/dusty-nv/jetson-inference/blob/master/docs/pytorch-transfer-learning.md#installing-pytorch

I am already reinstall os with newest version and repeat the step but still get same result, can anyone help me?

3

Hi @TrickyCV, what is the error you are getting, and which version of JetPack-L4T are you using?

You could also try running the jetson-inference docker container which already has PyTorch installed in it: https://github.com/dusty-nv/jetson-inference/blob/master/docs/aux-docker.md

5

error
error1600×900 321 KB

Here the error msg, and the version of jetpack is 4.6.1-b110

this after i do sudo update & upgrade and installing jetson-inference after reinstall os

6

Hmm, ok interesting - can you try running pip3 install setuptools first?

7

error2
error21600×900 271 KB

Still get error msg, after do pip3 install

8

Hmm, not sure why this script is still making errors for you, sorry about that - next please try pip3 install 'Cython<3'

You can also use the jetson-inference docker container which has PyTorch, or see the instructions for manually installing PyTorch here:

9

Hi, @dusty_nv , thanks for solution and answer so quickly. And how to load the onnx custom model into script object detecting script?

10

No problem @TrickyCV - first, make sure that detectnet/detectnet.py can load your custom ssd-mobilenet.onnx model from the command-line like here:

https://github.com/dusty-nv/jetson-inference/blob/master/docs/pytorch-ssd.md#processing-images-with-tensorrt

Then, you can create your own version of detectnet.py that loads a custom model using the parameterized API instead:

1 Like
11

Thanks man

13

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.