[bug]: Image finishes generating, and then OOM's and models that are well under 24gigs OOM's after generating the image. even models as small as 6gigs.

[sgt-Script]

Is there an existing issue for this problem?

• I have searched the existing issues

Operating system

Linux

GPU vendor

Nvidia (CUDA)

GPU model

Tesla P40

GPU VRAM

24G

Version number

5.7.0rc1

Browser

chrome, firefox, edge

Python dependencies

Local System
accelerate

1.0.1

compel

2.0.2

cuda

12.4

diffusers

0.31.0

numpy

1.26.3

opencv

4.9.0.80

onnx

1.16.1

pillow

11.0.0

python

3.11.10

torch

2.4.1+cu124

torchvision

0.19.1+cu124

transformers

4.46.3

xformers

Not Installed

What happened

i will select a model, Invoke will start the generation fine, i can see the ram being taken up by nvtop. once the image is done, and it before it places it in the gallery, it OOM's. even with smaller mdoels no bigger than 6Gigs. This started with most models a few weeks ago and i cant figure out how to correct it. partial loading was enabled as well but i never used it before.

invoke error.docx

What you expected to happen

image should be placed in gallery

How to reproduce the problem

start the app, chose a model, start generating, when finished it OOMS

Additional context

No response

Discord username

No response