[Bug]: In Ubuntu 23.10, Forge is much slower than A1111 with XL models. With 1.5 models there is no substantial difference.

[epictetito]

Checklist

• [ ] The issue exists after disabling all extensions
• [X] The issue exists on a clean installation of webui
• [ ] The issue is caused by an extension, but I believe it is caused by a bug in the webui
• [ ] 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?

Ubuntu 23.10 Once installed and with no apparent errors, Forge is much slower than A1111 with XL models creating the same images with the same models. With 1.5 models there is no substantial difference.

Steps to reproduce the problem

Simply starting Forge with ./webui.sh

What should have happened?

Forge is supposed to "make development easier, optimize resource management, and speed up inference."

What browsers do you use to access the UI ?

Brave

Sysinfo

sysinfo-2024-02-17-12-20.json

Console logs

./webui.sh 

################################################################
Install script for stable-diffusion + Web UI
Tested on Debian 11 (Bullseye), Fedora 34+ and openSUSE Leap 15.4 or newer.                                                            
################################################################

################################################################
Running on fluque user
################################################################

################################################################
Repo already cloned, using it as install directory
################################################################

################################################################
python venv already activate or run without venv: /home/fluque/forge/webui/venv
################################################################

################################################################
Launching launch.py...
################################################################
Using TCMalloc: libtcmalloc_minimal.so.4
libtcmalloc_minimal.so.4 is not linked with libpthreadand will trigger undefined symbol: ptthread_Key_Create error
Using TCMalloc: libtcmalloc.so.4
libtcmalloc.so.4 is not linked with libpthreadand will trigger undefined symbol: ptthread_Key_Create error
Python 3.11.6 (main, Oct  8 2023, 05:06:43) [GCC 13.2.0]
Version: f0.0.9-latest-51-g5bea443d
Commit hash: 5bea443d94f3a85f819cb8541c1bba0aac208d83
Launching Web UI with arguments: --xformers --autolaunch
Total VRAM 12042 MB, total RAM 31879 MB
xformers version: 0.0.23.post1
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 3060 : native
VAE dtype: torch.bfloat16
Using xformers cross attention
ControlNet preprocessor location: /home/fluque/forge/webui/models/ControlNetPreprocessor
Loading weights [676f0d60c8] from /home/fluque/forge/webui/models/Stable-diffusion/dreamshaperXL_turboDpmppSDE.safetensors
2024-02-17 13:23:04,178 - ControlNet - INFO - ControlNet UI callback registered.
Running on local URL:  http://127.0.0.1:7861

To create a public link, set `share=True` in `launch()`.
model_type EPS
UNet ADM Dimension 2816
Se está abriendo en una sesión de navegador existente.
Startup time: 8.6s (prepare environment: 1.9s, import torch: 3.4s, import gradio: 0.7s, setup paths: 0.7s, other imports: 0.6s, load scripts: 0.6s, create ui: 0.3s, gradio launch: 0.4s).
Using xformers attention in VAE
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
Using xformers attention in VAE
extra {'cond_stage_model.clip_l.transformer.text_model.embeddings.position_ids', 'cond_stage_model.clip_l.logit_scale', 'cond_stage_model.clip_l.text_projection', 'cond_stage_model.clip_g.transformer.text_model.embeddings.position_ids'}
left over keys: dict_keys(['conditioner.embedders.0.logit_scale', 'conditioner.embedders.0.text_projection'])
To load target model SDXLClipModel
Begin to load 1 model
Moving model(s) has taken 0.23 seconds
Model loaded in 4.1s (load weights from disk: 0.6s, forge load real models: 2.6s, load VAE: 0.2s, calculate empty prompt: 0.6s).

Additional information

No response

[BenDes21]

Also on Windows for me ( rtx 4080 12gb )

[huangwu5717]

me to(centos 40hx 8gb )

[zgz757183190]

Running on local URL: http://127.0.0.1:7861

How to change 127.0.0.1 to 0.0.0, so that everyone can access it?