NotFoundError: /lib/x86_64-linux-gnu/libm.so.6: version `GLIBC_2.29' not found

General
1

When I follow the Readme file to train your own Yolov8 network and I set everything up, including the docker image provided on github, I get this error, and the only solution seems to be replacing the first line ARG base_image=pytorch/pytorch:1.7.1-cuda11.0-cudnn8-runtime with ARG base_image=pytorch/pytorch:2.4.0-cuda11.8-cudnn9-runtime or at least some of these newer images. But then, when it comes to quantization of the model, it raises a warning that it’s unable to find any GPU (My previous topic).

Any ideas?

2

Hi @JanDev,
The easiest path to have an error free environment is to use the Suite docker from our develooper zone. have you tried using it?

1 Like
3

Thank you so much. I wanted to look at the documentation shortly after the release, but the website was done, so I didn’t read the documentation before.
If I can recommend something, I would change the Readme file on GitHub so that fewer people would get lost.
Thank you very much once again! :blush: