[Bug]: Google Colab script for exporting YOLO-NAS is incompatible with Python 3.12

[kingp0dd]

Checklist

• I have updated to the latest available Frigate version.
• I have cleared the cache of my browser.
• I have tried a different browser to see if it is related to my browser.
• I have tried reproducing the issue in incognito mode to rule out problems with any third party extensions or plugins I have installed.
• I have asked the AI at https://docs.frigate.video about my issue.

Describe the problem you are having

The build fails due to incomp with Python 3.12. I think the last version that'll work is 3.10

Steps to reproduce

...

Version

N/A

In which browser(s) are you experiencing the issue with?

No response

Frigate config file

N/A

docker-compose file or Docker CLI command

N/A

Relevant Frigate log output

N/A

Relevant go2rtc log output

N/A

Operating system

Home Assistant OS

Install method

Home Assistant Add-on

Network connection

Wired

Camera make and model

N/A

Screenshots of the Frigate UI's System metrics pages

N/A

Any other information that may be helpful

No response

[inkeep-ai-assistant[bot]]

The YOLO-NAS notebook has known compatibility issues with Python versions(1)(2). The community has identified that Python 3.11 is required for super-gradients to install properly(2).

To resolve the Python 3.12 incompatibility, you can use this updated approach(2):

1. Disconnect and delete your runtime in Google Colab (Runtime > Disconnect and delete runtime)(2)

2. Use the following command in the first cell(2):

! pip install git+https://github.com/Deci-AI/super-gradients.git

(2)

3. Add a new code cell after the first one with this code to fix URL references(2):

! sed -i 's/sghub.deci.ai/sg-hub-nv.s3.amazonaws.com/' /usr/local/lib/python3.11/dist-packages/super_gradients/training/pretrained_models.py
! sed -i 's/sghub.deci.ai/sg-hub-nv.s3.amazonaws.com/' /usr/local/lib/python3.11/dist-packages/super_gradients/training/utils/checkpoint_utils.py

(2)

If you're using Python 3.12, make sure you aren't using an older instance - older instances were 3.11, newer ones are 3.12(2). You may need to blow away your instance and re-do it(2).

Once you successfully generate the YOLO-NAS model, you can use it with Frigate's ONNX detector using this configuration(3):

detectors:
  onnx:
    type: onnx

model:
  model_type: yolonas
  width: 320
  height: 320
  input_pixel_format: bgr
  input_tensor: nchw
  path: /config/yolo_nas_s.onnx
  labelmap_path: /labelmap/coco-80.txt

(3)

Note that the input_tensor: nchw parameter is required for YOLO-NAS models(1).

📚 Sources:

1. [Detector Support]: Yolo-NAS notebook from documentation fails to run with 404 error
2. [Support]: Notebook to generate the YOLO-NAS Model not working
3. Supported Hardware

---

Have another question? Just tag @inkeep.