[Bug]: Keeps using up all my harddrive

[Soularis10]

Is there an existing issue for this?

• [X] I have searched the existing issues and checked the recent builds/commits

What happened?

Previously before I installed controlnet I could render images normally, but after installing my harddrive usage keeps working at 100%. With my memory slowly increasing. And when I close my stable diffusion, the harddrive usage instantly drops to 0%

So now I can't render any images without my memory being fully used up and crashing my computer. I even tried deleting stable diffusion, reinstalling with no controlnet and it still does not work.

Steps to reproduce the problem

1. Run Stable Diffusion with ControlNet (Runs smoothly)
2. Close Stable DIffusion (Since my memory keeps increasing)

What should have happened?

It should have ran controlnet normally without using too much of my harddrive

Version or Commit where the problem happens

Stable Diffusion 2.1?

What Python version are you running on ?

Python 3.10.x

What platforms do you use to access the UI ?

Windows

What device are you running WebUI on?

Nvidia GPUs (GTX 16 below)

Cross attention optimization

xformers

What browsers do you use to access the UI ?

Google Chrome

Command Line Arguments

--xformers --lowvram

List of extensions

Controlnet Except for a few checkpoints, VAE and LoRA models

Console logs

To create a public link, set `share=True` in `launch()`.
Startup time: 14.0s (import torch: 3.9s, import gradio: 2.1s, import ldm: 0.9s, other imports: 1.9s, setup codeformer: 0.2s, load scripts: 3.3s, create ui: 1.0s, gradio launch: 0.5s).
Creating model from config: D:\stable-diffusion-webui\configs\v1-inference.yaml
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.

Additional information

I suspect it is more of a stable diffusion issue, since when I delete the controlnet in the extension folder and venv folder. even after deleting the whole stable diffusion. I still encountered the same problem.

PS: I have no experience in coding at all, Please send help :(

Processor Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz 2.60 GHz Installed RAM 16.0 GB GPU: NVIDIA Geforce GTX 1070 (8.8GB of VRAM) System type 64-bit operating system, x64-based processor

I am running on a gaming laptop

[missionfloyd]

Stable Diffusion 2.1?

It's at the top of the console output and the bottom of the UI.

--xformers --lowvram

With 8GB of VRAM, you probably don't need --lowvram or --medvram.

How big is C:? Can you put it there instead?

Or you can get another SSD. They've gotten pretty affordable. 1TB can be had for around $50.

[Soularis10]

Stable Diffusion 2.1?

I think my version is v1.4.0

--xformers --lowvram With 8GB of VRAM, you probably don't need --lowvram or --medvram.

This is how much my VRAM I have

How big is C:? Can you put it there instead?

My C drive is quite small so I don't have space to put it there

Or you can get another SSD. They've gotten pretty affordable. 1TB can be had for around $50.

I was thinking of getting a 1TB SSD to have faster file transfer. I have no idea why it suddenly consumed so much VRAM out of nowhere when previously 3 weeks ago it was fine.

[w-e-w]

Version or Commit where the problem happens Stable Diffusion 2.1?

please do not confuse stable diffusion with stable diffusion webui

[highnrgappalachian]

Next up: teaching people to not confuse the name of the repo's creator with the repo's name.

[ClashSAN]

--lowvram uses more regular ram, if you only use --upcast-sampling --xformers the ram usage shouldn't exceed the amount you have.

[Soularis10]

Hi sorry for the late reply, Update: I reinstalled Git, Python and Stable diffusion itself and it worked for a few days, after that turned corrupted with file and directory is corrupted or unreadable. I even used a empty 4TB SSD to store the stable diffusion

Stable Diffusion 2.1?

I am abit confused at how to find the version but I just follow this video to download stable diffusion

• https://www.youtube.com/watch?v=kqXpAKVQDNU&list=WL&index=53&t=246s

--lowvram uses more regular ram, if you only use --upcast-sampling --xformers the ram usage shouldn't exceed the amount you have.

By right, So I have no idea why this happens.

[w-e-w]

@Soularis10

I am abit confused at how to find the version but I just follow this video to download stable diffusion

bottom of the ui or in termainal at the beginning during lunch

[Soularis10]

@w-e-w @missionfloyd Thanks a lot for the help, here is the one that I am using

[ClashSAN]

How are the results now when using those arguments?

[Soularis10]

How are the results now when using those arguments?

Thanks a lot, After removing my --lowvram and changing it to --upcast-sampling --xformers. It is working now. But I will monitor for another day to see if the problem happens again before closing this thread.

[Soularis10]

Thanks everyone for all the help, it is still working after a few days. I think the reason why it was causing this issue was that by using --lowvram when I did not need to. Caused more memory to be allocated to my regular ram and overloading it. Since I have not a lot of regular ram.