Any presets other than medium quality resulted in stalled ray process #4172
Unanswered
Minhao-Zhang
asked this question in
Q&A
Replies: 2 comments 2 replies
-
Likewise, I installed AutoGluon using pip and started seeing "Running the sub-fit in a ray process to avoid memory leakage" when I would set the presets to anything other than "medium_quality". I even just left it on for a whole day, and no progress. |
Beta Was this translation helpful? Give feedback.
0 replies
-
Thanks for reporting this issue. Can you share the package versions by running following codes:
|
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Environment
I am using the docker image nvcr.io/nvidia/pytorch:24.04-py3 and installed AutoGluon using pip.
I have a RTX 2060 (notebook) GPU.
Issue
Whenever I tried to train a model using TabularPredictor, fit will run smoothly with the preset 'medium_quality' with CPU usage shown below. The training finished in half an hour or so.
However, if I try to use any other preset, I will get the terminal output of
and nothing happens. The CPU usage is very low and I do not know why.
I have tried with a time_limit and without one, I have tried using GPU or not using GPU, I have even tried with a fresh docker container, but nothing seems to help. If you could give me a suggestion or point me to a documentation, that would be a great help.
Beta Was this translation helpful? Give feedback.
All reactions