Choose GPU

[giale-uno]

Hi guy!
I'm trying to use A1111 deforum with my second GPU (nvidia rtx 3080), instead of the internal basic gpu of my laptop.
I found a guide online which says to add a text line to "webui-user.bat" file, but it doesn't work for me, any help?

[Sakura-Luna]

Post your hardware specs and edited webui-user.

[Sakura-Luna]

WebUI does not support intel integrated graphics, so you don't need to specify it. WebUI will only use the CPU or Nvidia GPU on your computer.

[giale-uno]

Ok.. The fact is that when A1111 is working i can see from my task manager that the NVIDIA GPU is running very low, almost at 0%.

It seems that A1111 doesn't use it... So what i have to do? Should i remove or edit that additional text that i added to webui-user.bat?
If yes, can you please write me exactely what i have to write in there?

many thanks! It's quite important to resolve it

[Sakura-Luna]

This is indeed a bit strange, the WebUI should take up more VRAM by default. Can you see the VRAM usage increase when you generate an image?

[JanardhanS-17]

Hi @giale-uno - Were you able to fix this issue and use dedicated GPU to generate image?
Could you please let me know the fix you used ?

[bubakzlesa]

The task manager has multiple graphs representing different categories of GPU usage. If you switch to the CUDA category (top right graph), you will see that the GPU is indeed being used correctly for stable diffusion.

[killacan]

You can try specifying the GPU in the command line Arguments. for example, when I want to use my other GPU I change the line set COMMANDLINE_ARGS= to set COMMANDLINE_ARGS= --device-id=1 and I don't have the line set CUDA_VISIBLE_DEVICES=1

The device-id is going to be the GPU you are using, so try different numbers until it works. I believe it starts from 0 and goes up.

Edit: removed --xformers from the line because it does not apply here.

[killacan]

Also the --xformers is separate, so not needed for selecting GPU, its just what I have in that line.

[Sakura-Luna]

PyTorch does not support intel, so there is no need to use it.

[killacan]

ok but if Automatic1111 is running and working, and the GPU is not being used, it means that the wrong device is being used, so selecting the device might resolve the issue. If it does not resolve the issue then we try other stuff until something works.

[killacan]

Also gpu-z will give you more detailed information about your GPU usage. Task Manager does a terrible job with this for Stable Diffusion.

[JanardhanS-17]

Hi Killacan,
I tried this above method but it did not work -
set COMMANDLINE_ARGS=--device-id=1 --xformers
I still get below error -
raise RuntimeError(
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check

[JanardhanS-17]

Hi All,
Can anyone please help me on how to make use of Dedicated GPU to generate images using this code -
I get below error -
raise RuntimeError(
RuntimeError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check

I tried using CPU alone but it take more time - I used below command for that -
set COMMANDLINE_ARGS=--use-cpu all --precision full --no-half --skip-torch-cuda-test

[Se1toku]

The only thing I know is that if you add "--skip-torch-cuda-test", then 1111 will run on your CPU

[killacan]

what GPU do you have?

[Kotori05]

i have a similar problem:" Invalid device id: str"
i have 2gpu, when i used "CUDA_VISIBLE_DEVICES=0 python ./launch.py --device-id=0 --xformers " to start webui, it work well.
but when i used "CUDA_VISIBLE_DEVICES=1 python ./launch.py --device-id=1 --xformers " to start webui.
i get errors " Invalid device id: str"

my server:
nvidia-smi
Sun Apr 7 19:29:38 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.65 Driver Version: 551.86 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 3090 Ti On | 00000000:01:00.0 Off | Off |
| 0% 32C P5 30W / 480W | 11937MiB / 24564MiB | 44% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
| 1 NVIDIA GeForce RTX 3090 On | 00000000:04:00.0 Off | N/A |
| 0% 40C P8 7W / 350W | 2MiB / 24576MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| No running processes found |
+-----------------------------------------------------------------------------------------+

nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Mar_28_02:18:24_PDT_2024
Cuda compilation tools, release 12.4, V12.4.131
Build cuda_12.4.r12.4/compiler.34097967_0

when i choose device1(3090) to run webui i get errors:
Invalid device id: str
Traceback (most recent call last):
File "/mnt/e/WSL/stable-diffusion-webui/modules/errors.py", line 98, in run
code()
File "/mnt/e/WSL/stable-diffusion-webui/modules/devices.py", line 106, in enable_tf32
if cuda_no_autocast():
File "/mnt/e/WSL/stable-diffusion-webui/modules/devices.py", line 30, in cuda_no_autocast
torch.cuda.get_device_capability(device_id) == (7, 5)
File "/home/kotori/anaconda3/envs/sd/lib/python3.10/site-packages/torch/cuda/init.py", line 439, in get_device_capability
prop = get_device_properties(device)
File "/home/kotori/anaconda3/envs/sd/lib/python3.10/site-packages/torch/cuda/init.py", line 456, in get_device_properties
raise AssertionError("Invalid device id")
AssertionError: Invalid device id

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/mnt/e/WSL/stable-diffusion-webui/./launch.py", line 48, in
main()
File "/mnt/e/WSL/stable-diffusion-webui/./launch.py", line 44, in main
start()
File "/mnt/e/WSL/stable-diffusion-webui/modules/launch_utils.py", line 465, in start
import webui
File "/mnt/e/WSL/stable-diffusion-webui/webui.py", line 13, in
initialize.imports()
File "/mnt/e/WSL/stable-diffusion-webui/modules/initialize.py", line 36, in imports
shared_init.initialize()
File "/mnt/e/WSL/stable-diffusion-webui/modules/shared_init.py", line 17, in initialize
from modules import options, shared_options
File "/mnt/e/WSL/stable-diffusion-webui/modules/shared_options.py", line 4, in
from modules import localization, ui_components, shared_items, shared, interrogate, shared_gradio_themes, util, sd_emphasis
File "/mnt/e/WSL/stable-diffusion-webui/modules/interrogate.py", line 13, in
from modules import devices, paths, shared, lowvram, modelloader, errors, torch_utils
File "/mnt/e/WSL/stable-diffusion-webui/modules/devices.py", line 113, in
errors.run(enable_tf32, "Enabling TF32")
File "/mnt/e/WSL/stable-diffusion-webui/modules/errors.py", line 100, in run
display(task, e)
File "/mnt/e/WSL/stable-diffusion-webui/modules/errors.py", line 68, in display
te = traceback.TracebackException.from_exception(e)
File "/home/kotori/anaconda3/envs/sd/lib/python3.10/traceback.py", line 572, in from_exception
return cls(type(exc), exc, exc.traceback, *args, **kwargs)
AttributeError: 'str' object has no attribute 'traceback'

i need help, appreciate.