[ ] The issue exists after disabling all extensions
[ ] The issue exists on a clean installation of webui
[X] The issue is caused by an extension, but I believe it is caused by a bug in the webui
[X] The issue exists in the current version of the webui
[X] The issue has not been reported before recently
[ ] The issue has been reported before but has not been fixed yet
What happened?
WEBUI Reactor is using the full CPU instead of GPU, it seems to be taking longer than automatic1111 webui. No use of CUDA and consistent VRAM usage when Reactor is applied during image generation.
Steps to reproduce the problem
Start webui.bat
Use txt2img or img2img
Enable Reactor and add a single image with codeformers. On Settings, select CUDA
Run
What should have happened?
WebUI should go through VRAM or use CUDA I believe.
What browsers do you use to access the UI ?
Google Chrome
Sysinfo
(No Sysinfo)
(laptop) i7-11800H, nvidia 3070 8gb, 64gb RAM, Windows 10 64-bit
Console logs
venv "C:\Git\stable-diffusion-webui-forge\venv\Scripts\Python.exe"
Python 3.10.6 (tags/v3.10.6:9c7b4bd, Aug 1 2022, 21:53:49) [MSC v.1932 64 bit (AMD64)]
Version: f0.0.10-latest-58-ge6263135
Commit hash: e62631350a408edf3b1f0a9dd45a43f9f0e95ead
Launching Web UI with arguments: --ckpt-dir E:\Stable-diffusion
Total VRAM 8192 MB, total RAM 65352 MB
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 3070 Laptop GPU : native
VAE dtype: torch.bfloat16
Using pytorch cross attention
ControlNet preprocessor location: C:\Git\stable-diffusion-webui-forge\models\ControlNetPreprocessor
21:07:18 - ReActor - STATUS - Running v0.6.1 on Device: CUDA
Loading weights [84835fc746] from E:\Stable-diffusion\photo_v5.safetensors
2024-02-06 21:07:21,447 - ControlNet - INFO - ControlNet UI callback registered.
Running on local URL: http://127.0.0.1:7860
To create a public link, set `share=True` in `launch()`.
model_type EPS
UNet ADM Dimension 0
Using pytorch attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using pytorch attention in VAE
Startup time: 51.3s (prepare environment: 9.9s, import torch: 5.6s, import gradio: 1.5s, setup paths: 1.2s, initialize shared: 0.2s, other imports: 1.1s, list SD models: 12.7s, load scripts: 14.3s, create ui: 2.5s, gradio launch: 1.3s, app_started_callback: 0.8s).
extra {'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_l.logit_scale'}
Loading VAE weights specified in settings: C:\Git\stable-diffusion-webui\models\VAE\vae-ft-mse-840000-ema-pruned.vae.safetensors
To load target model SD1ClipModel
Begin to load 1 model
Moving model(s) has taken 0.17 seconds
Model loaded in 9.5s (load weights from disk: 3.4s, forge load real models: 2.0s, load VAE: 2.9s, load textual inversion embeddings: 0.7s, calculate empty prompt: 0.5s).
Automatic Memory Management: 52 Modules in 0.01 seconds.
[] []
2024-02-06 21:25:42,576 - ControlNet - INFO - Current ControlNet ControlNetPatcher: C:\Git\stable-diffusion-webui-forge\models\ControlNet\models\control_v11p_sd15_lineart_fp16.safetensors
To load target model AutoencoderKL
Begin to load 1 model
2024-02-06 21:25:44,620 - ControlNet - INFO - ControlNet Method lineart_coarse patched.
To load target model BaseModel
To load target model ControlNet
Begin to load 2 models
Moving model(s) has taken 4.21 seconds
100%|████████████████████████████████████████████████████████████████████████████████████| 7/7 [00:02<00:00, 2.40it/s]
21:25:59 - ReActor - STATUS - Working: source face index [0], target face index [0]██████| 7/7 [00:01<00:00, 3.10it/s]
21:25:59 - ReActor - STATUS - Analyzing Source Image...
21:26:03 - ReActor - STATUS - Analyzing Target Image...
21:26:04 - ReActor - STATUS - Detecting Source Face, Index = 0
21:26:04 - ReActor - STATUS - Detected: -39- y.o. Male
21:26:06 - ReActor - STATUS - Detecting Target Face, Index = 0
21:26:06 - ReActor - STATUS - Detected: -39- y.o. Male
21:26:06 - ReActor - STATUS - Swapping Source into Target
21:26:08 - ReActor - STATUS - Correcting Face Mask
21:26:08 - ReActor - STATUS - Restoring the face with CodeFormer
Total progress: 100%|████████████████████████████████████████████████████████████████████| 7/7 [00:24<00:00, 3.57s/it]
2024-02-06 21:27:22,487 - ControlNet - INFO - ControlNet Input Mode: InputMode.SIMPLE████| 7/7 [00:24<00:00, 3.10it/s]
Checklist
What happened?
WEBUI Reactor is using the full CPU instead of GPU, it seems to be taking longer than automatic1111 webui. No use of CUDA and consistent VRAM usage when Reactor is applied during image generation.
Steps to reproduce the problem
What should have happened?
WebUI should go through VRAM or use CUDA I believe.
What browsers do you use to access the UI ?
Google Chrome
Sysinfo
(No Sysinfo) (laptop) i7-11800H, nvidia 3070 8gb, 64gb RAM, Windows 10 64-bit
Console logs
Additional information
No response