lmstudio-ai / lmstudio-ai/lmstudio-bug-tracker

VRAM not releasing, RTX 4090 / Linux / 570.86.16 Driver / 12.8 CUDA

Open
#511 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
No language data
Stars
152
Forks
38
PR merge metrics
No merged PRs in 30d

Description

Which version of LM Studio?
Example: LM Studio 0.3.12

Which operating system?
Pop! OS
kernel 6.9.3-76060903-generic
CPU 9950X

What is the bug?
GPU memory not releasing with no models loaded.
Workaround - system reboot

Screenshots
If applicable, add screenshots to help explain your problem.
Image

Logs
main.log

To Reproduce
Not sure how to reproduce - after ejecting a model it doesn't free the memory.
This doesn't happen all the time though.
Trying to interrupt a stalled response may put the code in a bad state.
If it helps, I had Mistral Small 24b loaded, flash attention on.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the attached main.log and the reported environment: Pop! OS, kernel 6.9.3-76060903-generic, RTX 4090, driver 570.86.16, and CUDA 12.8. Try reproducing the issue by ejecting Mistral Small 24b with flash attention enabled, including interrupting a stalled response; done means GPU memory is released without requiring a reboot.

Written by the indexing model from the issue text.

Assessment

Tech stack
linux
Domain
desktop, performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.