Cuda OOM #2236
ShamshadAhmedShorthillsAI
started this conversation in
General
Cuda OOM
#2236
Replies: 1 comment 2 replies
-
|
Hey @ShamshadAhmedShorthillsAI. Which project are you referring to? |
Beta Was this translation helpful? Give feedback.
2 replies
-
|
@mturoci H2o LLM studio |
Beta Was this translation helpful? Give feedback.
-
|
I see. In that case you want to post in https://github.com/h2oai/h2o-llmstudio/discussions. cc @pascal-pfeiffer |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I have created an experiment with nvidia gpu machine with 24 GB of memory and the experiment ran successfully. Now when I am trying to run that same experiment(and keeping every config same as previous experiment). I am getting OOM. I am using mistral-7b-v0.1 base model. Any solution for that?
Beta Was this translation helpful? Give feedback.
All reactions