Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

SwarmUI Master Tutorial - Use Stable Diffusion 3.5 Large and FLUX model with Amazing Performance and more (Patreon)

Published:
2024-12-03 22:55:00
Edited:
2025-04-12 15:02:34
Imported:
2026-06

Content

In this public tutorial post I will share all the necessary stuff regarding how to use Stable Diffusion 3.5 Large, FLUX Dev, SD 1.5, SDXL, Stable Diffusion 3 and on your computer and also on the cloud (RunPod + Massed Compute).

=======

Patreon exclusive posts index to find our scripts easily, Patreon scripts updates history to see which updates arrived to which scripts and amazing Patreon special generative scripts list that you can use in any of your task.

Join discord to get help, chat, discuss and also tell me your discord username to get your special rank : SECourses Discord

Please also Star, Watch and Fork our Stable Diffusion & Generative AI  GitHub repository and join our Reddit subreddit and follow me on LinkedIn (my real profile)

=======

Amazing Ultra Important Tutorials with Chapters and Manually Written Subtitles / Captions

Updated Playlist : https://www.youtube.com/playlist?list=PL_pbwdIyffsku5Rfv2CJZSiFdGXtkkdiB

FLUX Tools Outpainting, Inpainting (Fill), Redux, Depth & Canny Ultimate Tutorial Guide with SwarmUI

https://youtu.be/hewDdVJEqOQ

Stable Diffusion 3.5 Large How To Use Tutorial With Best Configuration and Comparison With FLUX DEV

https://youtu.be/-zOKhoO9a5s

FLUX Full Fine-Tuning / DreamBooth Tutorial That Shows A Lot Info Regarding SwarmUI Latest

https://youtu.be/FvpWy1x5etM

Full FLUX Tutorial - FLUX Beats Midjourney for Real

https://youtu.be/bupRePUOA18

Main Windows SwarmUI Tutorial (Watch To Learn How to Use)

How to install and use. You have to watch this to learn how to use SwarmUI

Has 70 chapters and manually fixed captions

https://youtu.be/HKX8_F1Er_w

Cloud Tutorial (Massed Compute - RunPod - Kaggle)

If you don't have a powerful GPU or you want to use powerful GPU this is the tutorial you need

48 GB A6000 GPU is only 31 cents per hour on Massed Compute with our special coupon

https://youtu.be/XFUZof6Skkw

Free Kaggle Account Notebook for GPU-Poor

Installs latest version of SwarmUI on a free Kaggle account

Works with Dual T4 GPU at the same time

Supports SD 1.5, SDXL, SD3, FLUX, FLUX Tools, SD 3.5 Large, Mochi Tex to Video and Stable Cascade and more

Download from here : https://www.patreon.com/posts/106650931

18 February 2025 Update

Hunyuan text-to-video support arrived to SwarmUI

Public info link that has full information for how to manually install and execute from the SwarmUI developer below

https://github.com/mcmonkeyprojects/SwarmUI/blob/master/docs/Video%20Model%20Support.md#hunyuan-video

SwarmUI Unified Downloader that we maintain (recommended)

 https://www.patreon.com/posts/114517862

Unified downloader has so many features including auto download GGUF models (20 steps) and Fast-GGUF (6 steps) models and configuration screenshots for Hunyuan

4 December 2024 Update

Full tutorial for FLUX Tools published : https://youtu.be/hewDdVJEqOQ

Outpainting, Inpainting, Redux Style Transfer + Re-Imagine + Combine Multiple Images, Depth and Canny

28 November 2024 Update

Tutorial : https://youtu.be/hewDdVJEqOQ

Updated out-painting best config : Link

To fix mask line you need to do second inpainting pass on mask line with regular FLUX dev model config : Link

New FLUX Tools added to the SwarmUI

I recommend using our updated unified downloader to auto download (Windows, RunPod, Massed Compute) : SwarmUI unified downloader (auto downloads you do nothing)

Download flux1-redux-dev.safetensors into :

SwarmUI/Models/style_models

Download flux1-fill-dev.safetensors , flux1-canny-dev.safetensors , flux1-depth-dev.safetensors into

SwarmUI/Models/diffusion_models

Download flux1-canny-dev-lora.safetensors , flux1-depth-dev-lora.safetensors into

SwarmUI/Models/Lora

7 November 2024 Update

Newest SOTA Text-to Video model genmo/mochi-1-preview now supported by SwarmUI

You can download FP8 scaled unified model from this repo

Put the downloaded model into \SwarmUI\Models\Stable-Diffusion

Alternatively use our updated SwarmUI unified downloader

Mochi 1 Tutorial Prompts : GitHub Gist

26 October 2024 Update

Full tutorial video published and if you also upvote this reddit post I appreciate that

https://youtu.be/-zOKhoO9a5s

Stable Diffusion 3.5 Large published

https://stability.ai/news/introducing-stable-diffusion-3-5

Update your SwarmUI to the latest - make sure updated worked

For Stable Diffusion 3.5 Large I have compared 62 prompts and working Samplers + Schedulers

You can download all full grids as a zip : click here

For realism : DPM++ 2S Ancestral + SGM Uniform - This takes 2x time compared to Eular

For stylization like anime : Eular + Beta

Put FP16 and FP8 FLUX models into : \SwarmUI\Models\diffusion_models

Put FP16 and FP8 Stable Diffusion 3.5 Large into : \SwarmUI\Models\Stable-Diffusion

Put FP8 or FP16 T5 XXL and Clip-L and Clip-G into : \SwarmUI\Models\clip

For FLUX Dev best sampler is iPNDM + DDIM Uniform

Ultra fast and robust automatic models downloader for Windows, RunPod and Massed Compute : click to download

https://www.patreon.com/posts/114517862

For manually download click this link to go to downloads repo

T5 XXL Rename to Skip Download : t5xxl_enconly.safetensors

Test results

As FP8 version of T5 XXL it will download t5xxl_fp8_e4m3fn_scaled.safetensors which is supposed to be more closers to FP16 version

Here comparison (excellent quality) : FP8_vs_FP8_Scaled_vs_FP16_T5_XXL_With_FLUX.jpg

FP8 version of Stable Diffusion Large 3.5 contains Text Encoders FP8 versions as well it is from ComfyUI

Here comparison i think FP8 scaled works better : FP8_vs_FP16_SD_3_5_Large.jpg

We also added newest experimental scaled FP8 version of FLUX Dev

I think scaled FP8 FLUX DEV better than FP8 raw : FP8_vs_FP8_Scaled_vs_FP16_FLUX_DEV.jpg

FLUX vs SD 3.5 Large Comparison: FLUX_ vs_SD35.jpg

18 September 2024 Update

install-linux.sh file updated and now auto installs ComfyUI manager extension

Just wait backends to fully load it takes some time

You can check progress from Server > Logs > Debug menu

3 August 2024 Update

SwarmUI now supports newest black-forest-labs FLUX models

Full tutorial here : >> https://youtu.be/bupRePUOA18

black-forest-labs has the original developers that developed first Stable Diffusion models

I am recording a tutorial video

You can find automatic downloaders for FLUX models at below post for both Windows, RunPod and Massed Compute

1-Click FLUX models downloader scripts for Windows, RunPod and Massed Compute are in below post including FP16 T5 Downloader file

https://www.patreon.com/posts/109289967

The above downloader scripts have resume capability, skip if fully downloaded capability and will auto retry 100 times if an error occurs and continue from where ever left automatically and downloads from Hugging Face directly

Or you can instructions you can see at this image : How To Download And Use FLUX Models On Swarm UI.png

FLUX models official repo : https://huggingface.co/black-forest-labs

FLUX.1-schnell is timestep distilled - turbo model - 4 steps or more

FLUX.1-dev is guidance distilled - best weights public model - 20 steps or more

Non-distilled model is only available via API

FLUX works with only CFG 1 - thus not supporting negative prompts at the moment

FP8 Model works same as FP16-dev model if you are not manually using FP16 sampling (requires 28 GB) : https://huggingface.co/Kijai/flux-fp8/tree/main

FP8 vs FP16 vs Turbo vs FP16 T5 model detailed comparison (16 different tests) : https://imgsli.com/MjgzODYy/7/6

FP16 T5 Text Encoder can be downloaded from : https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main

21 June 2024 Update

Stable Swarm UI migrated into SwarmUI : https://github.com/mcmonkeyprojects/SwarmUI

Links are updated

Update Error Fix

fatal: detected dubious ownership in repository at 'G:/swarm_ui_tutorial/SwarmUI'

If you get above error run below config (fix according to your install path)

git config --global --add safe.directory G:/swarm_ui_tutorial/SwarmUI

General Info

To download and install StableSwarmUI : StableSwarmUI

Very important documents : docs

Very important features : Features

Full prompting syntax : Prompt%20Syntax.md

Download very best new upscalers : released-sorted-by-new

All upscalers as a single zip file : all_models

Helaman upscalers with thumbnails : helaman

SD3 Download model : stable-diffusion-3-sd3

SD3 Official Page : stable-diffusion-3-medium

SDXL model RealVis XL 4 : RealVisXL_V4.0.safetensors

SD 1.5 model : Hyper Realism V3

I have converted Fooocus 275 SDXL styles into StableSwarmUI presets with SD3 configuration

The presets also have updated thumbnail / preview image with SD3

Moreover, scripts to generate and update these presets automatically added to the zip file in below post as well

You can download all presets from here : download 275 presets

Official Stable Swarm UI Discord server to contact its developer : https://discord.gg/q2y38cqjNw

Yolo Face segment into yolov8 folder : https://huggingface.co/Bingsu/adetailer/resolve/main/face_yolov9c.pt

Cloud tutorials are in production as well for Massed Compute, RunPod and hopefully a Kaggle

Massed Compute Instructions

Tutorial : https://youtu.be/XFUZof6Skkw

Please register Massed Compute from here : https://vm.massedcompute.com/signup?linkId=lp_034338&sourceId=secourses&tenantId=massed-compute

After registering and loading balance go to this link : https://vm.massedcompute.com/deploy

Select RTX A6000, if not available RTX A6000 [ALT Config] (same VRAM, GPU, Disk etc only RAM is 24 GB instead of 48)

Then make the selections as below and deploy your machine

This Massed Compute virtual machine image we have as SECourses has pre-installed SwarmUI with very best SD 1.5, SDXL and SD3 models

You can also increase your GPU quantity to utilize all of them at once with SwarmUI

Don't forget to run SwarmUI (1-click button) update before start using it

If our image gets broken reply to this post and let me know - we will update it ASAP

Also don't forget to apply Coupon code as SECourses to get A6000 GPU as 31 cents per hour (on RunPod cheapest 69 cents per hour)

How To Connect SwarmUI from your PC that is Running on Massed Compute - Preffered Cloudflare Way

First open a terminal and execute below commands to install cloudflared

wget https://github.com/cloudflare/cloudflared/releases/download/2024.8.2/cloudflared-linux-amd64.deb

sudo dpkg -i cloudflared-linux-amd64.deb

Then run SwarmUI to update to update it latest, then close the started SwarmUI terminal

Then open a new terminal and execute below commands and it will start SwarmUI and will give you a cloudflare link like (my-pills-sailing-pad-netherlands.trycloudflare .com) that you can connect

cd /home/Ubuntu/apps/StableSwarmUI/

./launch-linux.sh --launch_mode none --cloudflared-path cloudflared

How To Connect SwarmUI from your PC that is Running on Massed Compute - I dont recommend 1 GB Bandwith

Get your token from here for free : https://dashboard.ngrok.com/get-started/your-authtoken

sudo snap install ngrok

ngrok authtoken replace_with_your_ngrok_token

ngrok http 7801

RunPod Instructions

Follow this tutorial step by step to install on RunPod machine : https://youtu.be/XFUZof6Skkw

Please use this link to register : https://runpod.io?ref=1aka98lq - (it gives me credits thus i can do more tutorials)

RunPod massive tutorial includes runpodctl for fast download to PC : https://youtu.be/QN1vdGhjcRc

RunPod permanent network storage system tutorial : https://youtu.be/8Qf4x3-DFf4

How to download models very fast to RunPod and how to upload files, folders from RunPod to Hugging Face ultra fast detailed tutorial : https://youtu.be/X5WVZ0NMaTg

Choose below template and any GPU

RunPod Pytorch 2.2.0

runpod/pytorch:2.2.0-py3.10-cuda12.1.1-devel-ubuntu22.04

Download install-linux.sh and upload into workspace

chmod +x install-linux.sh

./install-linux.sh

It will auto install and start the SwarmUI

Then connect from 7801 port

If you want to fix No module named 'onnxruntime' error which is 100% caused by SwarmUI itself while the app is running execute below commands after uploading install-linux_fix_step2.sh file

chmod +x install-linux_fix_step2.sh

./install-linux_fix_step2.sh

After executing and installing missing libraries, disable your back-end and enable again for it to become working

If you turn off your pod and start again execute below command to start again

cd SwarmUI

cd launchtools

wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh

chmod +x dotnet-install.sh

./dotnet-install.sh --channel 8.0 --runtime aspnetcore

./dotnet-install.sh --channel 8.0

cd ..

./launch-linux.sh --launch_mode none --host 0.0.0.0

SwarmUI Windows Requirements

https://dotnet.microsoft.com/en-us/download/dotnet/8.0

Git : https://git-scm.com/downloads

How to install Git, Python, CUDA, cuDNN, C++ Tools : https://youtu.be/DrhUHnYfwC0

Error and Solutions

If you get below errors

RuntimeError: CUDA error: operation not supported

CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.

Add this parameter to the back-end and try : --disable-cuda-malloc

RunPod Cloudflared

wget https://github.com/cloudflare/cloudflared/releases/download/2024.8.2/cloudflared-linux-amd64.deb

dpkg -i cloudflared-linux-amd64.deb

git clone https://github.com/mcmonkeyprojects/SwarmUI

cd SwarmUI

cd launchtools

wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh

chmod +x dotnet-install.sh

./dotnet-install.sh --channel 8.0 --runtime aspnetcore

./dotnet-install.sh --channel 8.0

cd ..

./launch-linux.sh --launch_mode none --cloudflared-path cloudflared

Files

Comments

Steve

Benefits/drawbacks?

Lou

This is really cool. Looking forward to the tutorial. Thanks.

Lou

Yeah the UI is outstanding. Playing around with it and learning as I go. I think StableSwarm has huge potential. Super easy to install, looks great and there is a lot under the hood to figure out.

Neil Francese

I see the SD3 devs recommend and released a sample ComfyUI workflow. They also have a few different text_encoders options. Are you planning to do any experimenting with their flow and the various text encoder options? You do such a great job evaluating and finding ideal settings so I was just curious :)

reaper557

I am excited for this tutorial! A super one-two punch, as I've haven't used the SwarmUI yet either.

Steve

also looking for the option to point the model folder to look in the A1111 folder for checkpoints instead of the root. anyone find that? EDIT: ANSWER is at 14:24 in the video he put out... SERVER>SERVER CONFIGURATION>PATHS

Lou

Deleted my response because SEcourses has a much simpler answer 😋

Lou

Yeah your right I totally forgot. Great video by the way. Thanks!

Manpreet Singh

Is it possible to have swarmUI running on one machine, say a Mac and using the resources (GPUs and CPU) on another machine (say Ubuntu)?

Siva

Thanks for the styles , but it is lacking in details too much, I tried to 2x upscale , still coming out very bland. Its fast ,i should accept it. But not worth as of now. WE need loras/checkpoint to make it good. SDXL Still wins

Eyepatch_101

Please, if you can post a Kaggle or colab notebook it would be great, as I don't have a GPU intensive machine....

Manpreet Singh

Any plans for comparing SD3 with other models (SDXL, SD1.5) for dreambooth training?

Steve

FYI Civitai banned SD3 because of their licensing..."as written the creator license seemed to imply that Stability had edit and takedown power over not just any resource trained on SD3 but also on any resources that used its outputs in their datasets." useful to know if you are posting images created with it or training on it.

Ec Jep

This tutorial is the best one I've seen and watched so far on SD! Wow - just so many details. I have also discovered the incredible power of in comfyui. Nothing could get close to Adetailer until I discovered this feature. I am now generating incredible high quality pictures using this feature. Adetailer was the one item left that was keeping me on A1111 and now I think I am a convert to SwarmUI. The features you describe are so easy to use. Thanks for all your effort on this video as it must have taken quite a long time to put togehter!

Ec Jep

Do you think kohya will support training on either SD3 or the flux models? Currently my training stopped at SDXL.

Carl de Jager

First of all, thanks for all the great tutorials. I've just tried to get this running on Runpod. Everything works up until I have to connect the http connection. It doesn't appear as a button on my pod. And opening 0.0.0.0: 7801 from my browser also doesn't connect. I'm on a Mac.

davide carta

Bye. I joined your patreon yesterday and plan to follow many of your wonderful guides. I am starting with this one to use flux with SwarmUI on kaggle (I already use flux with forge on lightining.ai and with lora trained models on replicate). i am not an expert so i need a little help on a couple of things: I tried yesterday and managed to get it all started on kaggle. Today however I noticed that I have to reinstall everything every time because obviously the temp folder is deleted. That's fine. But I thought there was no need to reinstall swarmUI on kaggle/working, because it stays. But when I try to start it up it tells me I don't have permissions. So I reinstalled SwarmUI as well. But there's a problem: the backends don't start. They just stay loading. Can the permissions issue be resolved? And why do the backends remain loading? Edit: I have this error when start SwarmUI: 22:04:53.478 [Warning] [ComfyUI-0/STDERR] Traceback (most recent call last): 22:04:53.479 [Warning] [ComfyUI-0/STDERR] File "/kaggle/working/SwarmUI/dlbackend/ComfyUI/main.py", line 87, in 22:04:53.479 [Warning] [ComfyUI-0/STDERR] import comfy.utils 22:04:53.479 [Warning] [ComfyUI-0/STDERR] File "/kaggle/working/SwarmUI/dlbackend/ComfyUI/comfy/utils.py", line 20, in 22:04:53.480 [Warning] [ComfyUI-0/STDERR] import torch 22:04:53.480 [Warning] [ComfyUI-0/STDERR] File "/kaggle/working/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/__init__.py", line 290, in 22:04:53.480 [Warning] [ComfyUI-0/STDERR] from torch._C import * # noqa: F403 22:04:53.480 [Warning] [ComfyUI-0/STDERR] ImportError: /kaggle/working/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/lib/../../nvidia/cusparse/lib/libcusparse.so.12: undefined symbol: __nvJitLinkComplete_12_4, version libnvJitLink.so.12 ps. Sorry, I'm italian I'm don't speak english very well. Deepl is my translator ;)

Mykola Kotliar

strange problem. I downloaded your caddy notebook. I did everything according to the instructions. Please help who can. But SwarmUI does not generate on kagli with any model. Attempts were with models in the temporary folder, and the model loaded during installation. The process hangs on loading the model and that's it. == SwarmUI logs 2024-09-02 16:37 == 16:37:34.184 [Info] User local requested 1 image with model 'OfficialStableDiffusion/sd_xl_base_1.0.safetensors'... In debug logs there is constantly waiting for server [Debug] [Load ComfyUI Self-Starting #0] ComfyUI-0 port 7821 waiting for server...

Ahmet Inceelli

Hello, I get an error below as install swarm_ui yordam giris noktası, D:\swarm_ui\SwarmUl\dibackend\comfy\python_embeded\Lib\site-pa kages\torchaudio\lib\libtorchaudio.pyd dinamik baglantı kitaplıgında bulunamadı.

Summer Bliss

Hi, I am using RunPod with SwarmUI and from there ComfyUI, please can you help me install ComfyUI manager?

Summer Bliss

Hi, everything was working for me in Runpod using Swarm with Comfy based on your instructions, until today suddenly ComfyUI is not showing any nodes and the 'Manager' button is missing from the floating buttons part. Please help, thanks a lot. (I have been using the Pytorch 2.1 template in Runpod, should I use 2.4 or 2.5 ? )

Summer Bliss

Yea I reinstalled and it works fine now, thanks! By the way, I manually deleted the 'SwarmUI' folder before installing again using your command lines. Was that the correct way? I mean was there a 'proper' way to uninstall SwarmUI first? Thx

4401 4401

hi, have followed instructions to load swarm on several runpods in the past but today I'm getting this error after the restart. 08:50:09.786 [Init] Starting webserver on http://0.0.0.0:7801 08:50:09.817 [Init] Swarm is up to date! You have version 0.9.2.3, and 0.9.2-Beta is the latest. 08:50:09.825 [Init] Program is running. 08:50:10.547 [Init] Self-Start ComfyUI-0 on port 7821 is loading... 08:50:12.437 [Warning] [ComfyUI-0/STDERR] Traceback (most recent call last): 08:50:12.437 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/main.py", line 90, in 08:50:12.437 [Warning] [ComfyUI-0/STDERR] import execution 08:50:12.437 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/execution.py", line 13, in 08:50:12.437 [Warning] [ComfyUI-0/STDERR] import nodes 08:50:12.437 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/nodes.py", line 21, in 08:50:12.438 [Warning] [ComfyUI-0/STDERR] import comfy.diffusers_load 08:50:12.438 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/diffusers_load.py", line 3, in 08:50:12.438 [Warning] [ComfyUI-0/STDERR] import comfy.sd 08:50:12.438 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/sd.py", line 5, in 08:50:12.438 [Warning] [ComfyUI-0/STDERR] from comfy import model_management 08:50:12.438 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 143, in 08:50:12.438 [Warning] [ComfyUI-0/STDERR] total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) 08:50:12.438 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 112, in get_torch_device 08:50:12.438 [Warning] [ComfyUI-0/STDERR] return torch.device(torch.cuda.current_device()) 08:50:12.438 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 878, in current_device 08:50:12.438 [Warning] [ComfyUI-0/STDERR] _lazy_init() 08:50:12.439 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 314, in _lazy_init 08:50:12.439 [Warning] [ComfyUI-0/STDERR] torch._C._cuda_init() 08:50:12.439 [Warning] [ComfyUI-0/STDERR] RuntimeError: CUDA driver initialization failed, you might not have a CUDA gpu. 08:50:12.645 [Info] Self-Start ComfyUI-0 unexpectedly exited (if something failed, change setting `LogLevel` to `Debug` to see why!) 08:50:12.645 [Info] Self-Start ComfyUI-0 had errors before shutdown: [ComfyUI-0/STDERR] Traceback (most recent call last): [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/main.py", line 90, in [ComfyUI-0/STDERR] import execution [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/execution.py", line 13, in [ComfyUI-0/STDERR] import nodes [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/nodes.py", line 21, in [ComfyUI-0/STDERR] import comfy.diffusers_load [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/diffusers_load.py", line 3, in [ComfyUI-0/STDERR] import comfy.sd [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/sd.py", line 5, in [ComfyUI-0/STDERR] from comfy import model_management [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 143, in [ComfyUI-0/STDERR] total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 112, in get_torch_device [ComfyUI-0/STDERR] return torch.device(torch.cuda.current_device()) [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 878, in current_device [ComfyUI-0/STDERR] _lazy_init() [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 314, in _lazy_init [ComfyUI-0/STDERR] torch._C._cuda_init() [ComfyUI-0/STDERR] RuntimeError: CUDA driver initialization failed, you might not have a CUDA gpu.

4401 4401

I'm running the 2.1 torch and 4xa6000, khoya worked fine but now cant get swarm to install to test training

Myk

I thought SwarmUI would be great...but it's not. Namely, i cant tell it where i want the generations to be placed(SAVED). Also, i have THOUSANDS of loras/embeddings/models that have spaces in them, and they wont load and crashes. I MEAN WOW... Why do i have to rename thousands of files when ever other UI can use them just fine no matter how they are named. PLUS, when it was loading the folders i placed to maunally point to them, the program was pulling from and compairing my database to one from Civitai!!! thats a serious issue when i didn't tell it to do this. I DONT want my database compromised and it looks like that this UI was doing it. WHY was it verifying my models/loras/embeddings?

Myk

Thanks! i will do that!

Julio Ibanez

Any good reasons to crop images instead of using buckets? My training images have all sort of aspect ratios with relevant information I would like to keep.

Julio Ibanez

For me, bucketing seems to work just fine with my RTX 3090, and now I’m attempting to train Dreambooth/Flux with regularization images. However, the process is very time-consuming (about 3 days with only 4 repetitions). You mentioned that it’s possible to resume training at any point by selecting the latest trained model. Are there any specific considerations I should keep in mind if I do this? For example, should I use the exact same settings, training images, and regularization images? I’m wondering if, when resuming the process with the same training data, it might repeat the initial training steps with the same resources, potentially leading to overfitting

Julio Ibanez

BUG and solution for training images. There is a current bug with training images as the output regularization images generated is incorrectly named. The path generated by default is named "Repeats number"_"Class prompt name" Which is incorrect. To solve this issue, the folder needs to be renamed manually after generated to "Repeats number_"Instance prompt name"+"space"+"Class prompt name" Otherwise it will not utilise the Regularization images utilized despite they are copied over the output folder. Also related to the topic of resuming a training process, I found there is a UI button "Save training state" and "Save training state at end of training" I have never used. Have you? I am resuming my training with these enabled just to see because my system seems to be crashing after several hours and usually when checking my computer screen (despite is set to full performance and never to go to sleep) However the problem seems to be related memory management and potentially this could solve it: PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True I will let you know if this did the trick. for now it seems is working.

Nabil Boulezaz

hi how do i use my modules ? do i need to download again ? i have stabled dufission it has modules and webui has same modules im running out of space

Nabil Boulezaz

I still didn't get it do i use model root and put the path to my flux or stable diffusion folder?

Robert Arsene

Is there a way you could share SwarmUI Presets? Like save all the settings you have where you generate this cool images and share it with us so we can load it in SwarmUI?

Neil Rhodes

Seems like the lora links are broken even on hugging face! :(

kurlu

Thank you Furkan for this fantastic FLUX Inpaint/Outpaint video. Just one question. Your Flux Guidance Scale is up to 30, but mine is max. 10. I am using 0.9.3.2. They changed this? Whats the difference?

kurlu

AHH!!!! 🤦🏽🤦🏽 I feel a little stupid now.

kurlu

i played now a lot with the inpainting and redux. what i am not able to accomplish is inserting an object. i generated a picture and modified to my liking with inpainting. now i want to insert a picture of an object with inpainting. and the inserted object picture should mach the style of the base picture and also be positioned in a certain place.

kurlu

just had a look on this In-Context-LoRA. 🤯 will take some time to wrap my head around this... or i am patient and wait for your tutorial.

Jakis Ktos

Great are these tutorials, thank you very much. I have a problem with other checkpoints Flux downloaded from Civitai. The ones you provided work normally, the others don't want to. Can you give some hints which flux checkpoints will be correct in SwarmUI? Sorry for my English, this is an automatic translation, I don't know English :(

Bayram TATAR

Hello Mr. Furkan.. I want to join your discord but it says the invitation has expired. how can I join. I am your monthly member.

Fred

Hello! I am new at all this and I'm running in a bit of an issue: I am trying to run SwarmUI_Kaggle_v13.ipynb as per instructed in this video: https://www.youtube.com/watch?v=XFUZof6Skkw At the 27:50 mark, it is instructed to ''go to the link you got from step 3 and click visit site'' However, I see in the comment section that step 3 is no longer required as Kaggle now supports Gradio live share. I do not know what Gradio live share is (I am a total beginner), is there any way someone can explain to me how to replace step 3 with Gradio live share so that I can open SwarmUI on Kaggle? Thanks and have a good day!

Fred

Awesome! Thank you so much

4401 4401

anyone having problems installing Swarm on runpod today? It worked like 2 days ago but been trying to start a new pod all day and the backend wont load. Have tried several times with different templates and different graphics cards and same issue with backend not loading each time? 20:06:47.372 [Init] Self-Start ComfyUI-0 on port 7821 is loading... 20:06:50.331 [Warning] [ComfyUI-0/STDERR] Traceback (most recent call last): 20:06:50.331 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/main.py", line 136, in 20:06:50.331 [Warning] [ComfyUI-0/STDERR] import execution 20:06:50.332 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/execution.py", line 13, in 20:06:50.332 [Warning] [ComfyUI-0/STDERR] import nodes 20:06:50.332 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/nodes.py", line 22, in 20:06:50.332 [Warning] [ComfyUI-0/STDERR] import comfy.diffusers_load 20:06:50.332 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/diffusers_load.py", line 3, in 20:06:50.332 [Warning] [ComfyUI-0/STDERR] import comfy.sd 20:06:50.332 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/sd.py", line 6, in 20:06:50.333 [Warning] [ComfyUI-0/STDERR] from comfy import model_management 20:06:50.333 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 168, in 20:06:50.333 [Warning] [ComfyUI-0/STDERR] total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) 20:06:50.333 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 131, in get_torch_device 20:06:50.333 [Warning] [ComfyUI-0/STDERR] return torch.device(torch.cuda.current_device()) 20:06:50.333 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 971, in current_device 20:06:50.333 [Warning] [ComfyUI-0/STDERR] _lazy_init() 20:06:50.334 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 319, in _lazy_init 20:06:50.334 [Warning] [ComfyUI-0/STDERR] torch._C._cuda_init() 20:06:50.334 [Warning] [ComfyUI-0/STDERR] RuntimeError: CUDA unknown error - this may be due to an incorrectly set up environment, e.g. changing env variable CUDA_VISIBLE_DEVICES after program start. Setting the available devices to be zero. 20:06:50.626 [Info] Self-Start ComfyUI-0 unexpectedly exited (if something failed, change setting `LogLevel` to `Debug` to see why!)

문아인💋(moon ain) (edited)

Comment edits

2026-09-07 10:00 root@7c34ee6cf8cf:/workspace# cd SwarmUI cd launchtools wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh chmod +x dotnet-install.sh ./dotnet-install.sh --channel 8.0 --runtime aspnetcore ./dotnet-install.sh --channel 8.0 cd .. ./launch-linux.sh --launch_mode none --host 0.0.0.0 --2025-02-20 13:28:57-- https://dot.net/v1/dotnet-install.sh Resolving dot.net (dot.net)... 20.76.201.171, 20.112.250.133, 20.236.44.162, ... Connecting to dot.net (dot.net)|20.76.201.171|:443... connected. HTTP request sent, awaiting response... 301 Moved Permanently Location: https://builds.dotnet.microsoft.com/dotnet/scripts/v1/dotnet-install.sh [following] --2025-02-20 13:28:58-- https://builds.dotnet.microsoft.com/dotnet/scripts/v1/dotnet-install.sh Resolving builds.dotnet.microsoft.com (builds.dotnet.microsoft.com)... 13.107.246.69, 2620:1ec:bdf::69 Connecting to builds.dotnet.microsoft.com (builds.dotnet.microsoft.com)|13.107.246.69|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 65133 (64K) [application/octet-stream] Saving to: ‘dotnet-install.sh’ dotnet-install.sh 100%[======================>] 63.61K --.-KB/s in 0.007s 2025-02-20 13:28:58 (8.80 MB/s) - ‘dotnet-install.sh’ saved [65133/65133] dotnet-install: ASP.NET Core Runtime with version '8.0.13' is already installed. dotnet-install: .NET Core SDK with version '8.0.406' is already installed. 13:28:59.934 [Init] === SwarmUI v0.9.5.0 Starting at 2025-02-20 13:28:59 === 13:28:59.976 [Init] Prepping extension: SwarmUI.Builtin_ScorersExtension.ScorersExtension... 13:28:59.976 [Init] Prepping extension: SwarmUI.Builtin_ImageBatchToolExtension.ImageBatchToolExtension... 13:28:59.976 [Init] Prepping extension: SwarmUI.Builtin_GridGeneratorExtension.GridGeneratorExtension... 13:28:59.976 [Init] Prepping extension: SwarmUI.Builtin_DynamicThresholding.DynamicThresholdingExtension... 13:28:59.977 [Init] Prepping extension: SwarmUI.Builtin_ComfyUIBackend.ComfyUIBackendExtension... 13:28:59.977 [Init] Prepping extension: SwarmUI.Builtin_AutoWebUIExtension.AutoWebUIBackendExtension... 13:29:00.016 [Init] Parsing command line... 13:29:00.016 [Init] Loading settings file... 13:29:00.034 [Init] Re-saving settings file... 13:29:00.055 [Init] Applying command line settings... 13:29:00.083 [Init] SwarmUI was installed 2025-02-20 (0 days ago) with version 0.9.5.0 13:29:00.099 [Init] Prepping options... 13:29:00.125 [Init] CPU Cores: 16 | RAM: 503.52 GiB total, 494.47 GiB available, 0 B virtual, -540654923776 B swap 13:29:00.125 [Init] GPU 0: NVIDIA GeForce RTX 4090 | Temp 34C | Util 0% GPU, 0% Memory | VRAM 23.99 GiB total, 23.54 GiB free, 2.00 MiB used 13:29:00.125 [Init] GPU 1: NVIDIA GeForce RTX 4090 | Temp 41C | Util 0% GPU, 0% Memory | VRAM 23.99 GiB total, 23.54 GiB free, 2.00 MiB used 13:29:00.125 [Init] Will use GPU accelerations specific to NVIDIA GeForce RTX 40xx series and newer. 13:29:00.233 [Init] Loading models list... 13:29:00.264 [Init] Loading backends... 13:29:00.264 [Init] Loading backends from file... 13:29:00.267 [Init] Prepping API... 13:29:00.267 [Init] Initializing backend #0 - ComfyUI Self-Starting... 13:29:00.269 [Init] Prepping webserver... 13:29:00.271 [Init] Backend request handler loop ready... 13:29:00.389 [Init] Scan for web extensions... 13:29:00.390 [Init] Readying extensions for launch... 13:29:00.390 [Init] Launching server... 13:29:00.391 [Init] Starting webserver on http://0.0.0.0:7801 13:29:00.401 [Init] Current git commit marked as date 2025-02-19 22:49:36 13:29:00.403 [Init] Swarm is up to date! You have version 0.9.5.0, and 0.9.5-Beta is the latest. 13:29:00.428 [Init] Program is running. 13:29:01.441 [Init] Self-Start ComfyUI-0 on port 7821 is loading... 13:29:04.291 [Warning] [ComfyUI-0/STDERR] Traceback (most recent call last): 13:29:04.291 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/main.py", line 136, in 13:29:04.291 [Warning] [ComfyUI-0/STDERR] import execution 13:29:04.291 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/execution.py", line 13, in 13:29:04.291 [Warning] [ComfyUI-0/STDERR] import nodes 13:29:04.291 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/nodes.py", line 22, in 13:29:04.291 [Warning] [ComfyUI-0/STDERR] import comfy.diffusers_load 13:29:04.291 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/diffusers_load.py", line 3, in 13:29:04.291 [Warning] [ComfyUI-0/STDERR] import comfy.sd 13:29:04.292 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/sd.py", line 6, in 13:29:04.292 [Warning] [ComfyUI-0/STDERR] from comfy import model_management 13:29:04.292 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 168, in 13:29:04.292 [Warning] [ComfyUI-0/STDERR] total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) 13:29:04.292 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 131, in get_torch_device 13:29:04.292 [Warning] [ComfyUI-0/STDERR] return torch.device(torch.cuda.current_device()) 13:29:04.292 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 971, in current_device 13:29:04.292 [Warning] [ComfyUI-0/STDERR] _lazy_init() 13:29:04.292 [Warning] [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 319, in _lazy_init 13:29:04.292 [Warning] [ComfyUI-0/STDERR] torch._C._cuda_init() 13:29:04.292 [Warning] [ComfyUI-0/STDERR] RuntimeError: CUDA unknown error - this may be due to an incorrectly set up environment, e.g. changing env variable CUDA_VISIBLE_DEVICES after program start. Setting the available devices to be zero. 13:29:04.592 [Info] Self-Start ComfyUI-0 unexpectedly exited (if something failed, change setting `LogLevel` to `Debug` to see why!) 13:29:04.592 [Info] Self-Start ComfyUI-0 had errors before shutdown: [ComfyUI-0/STDERR] Adding extra search path checkpoints /workspace/SwarmUI/Models/Stable-Diffusion [ComfyUI-0/STDERR] Adding extra search path vae /workspace/SwarmUI/Models/VAE [ComfyUI-0/STDERR] Adding extra search path loras /workspace/SwarmUI/Models/Lora [ComfyUI-0/STDERR] Adding extra search path loras /workspace/SwarmUI/Models/LyCORIS [ComfyUI-0/STDERR] Adding extra search path upscale_models /workspace/SwarmUI/Models/ESRGAN [ComfyUI-0/STDERR] Adding extra search path upscale_models /workspace/SwarmUI/Models/RealESRGAN [ComfyUI-0/STDERR] Adding extra search path upscale_models /workspace/SwarmUI/Models/SwinIR [ComfyUI-0/STDERR] Adding extra search path upscale_models /workspace/SwarmUI/Models/upscale-models [ComfyUI-0/STDERR] Adding extra search path upscale_models /workspace/SwarmUI/Models/upscale_models [ComfyUI-0/STDERR] Adding extra search path embeddings /workspace/SwarmUI/Models/Embeddings [ComfyUI-0/STDERR] Adding extra search path embeddings /workspace/SwarmUI/Models/embeddings [ComfyUI-0/STDERR] Adding extra search path hypernetworks /workspace/SwarmUI/Models/hypernetworks [ComfyUI-0/STDERR] Adding extra search path controlnet /workspace/SwarmUI/Models/controlnet [ComfyUI-0/STDERR] Adding extra search path controlnet /workspace/SwarmUI/Models/ControlNet [ComfyUI-0/STDERR] Adding extra search path clip /workspace/SwarmUI/Models/clip [ComfyUI-0/STDERR] Adding extra search path clip /workspace/SwarmUI/Models/CLIP [ComfyUI-0/STDERR] Adding extra search path clip_vision /workspace/SwarmUI/Models/clip_vision [ComfyUI-0/STDERR] Adding extra search path unet /workspace/SwarmUI/Models/unet [ComfyUI-0/STDERR] Adding extra search path diffusion_models /workspace/SwarmUI/Models/diffusion_models [ComfyUI-0/STDERR] Adding extra search path gligen /workspace/SwarmUI/Models/gligen [ComfyUI-0/STDERR] Adding extra search path ipadapter /workspace/SwarmUI/Models/ipadapter [ComfyUI-0/STDERR] Adding extra search path yolov8 /workspace/SwarmUI/Models/yolov8 [ComfyUI-0/STDERR] Adding extra search path tensorrt /workspace/SwarmUI/Models/tensorrt [ComfyUI-0/STDERR] Adding extra search path clipseg /workspace/SwarmUI/Models/clipseg [ComfyUI-0/STDERR] Adding extra search path style_models /workspace/SwarmUI/Models/style_models [ComfyUI-0/STDERR] Adding extra search path custom_nodes /workspace/SwarmUI/src/BuiltinExtensions/ComfyUIBackend/DLNodes [ComfyUI-0/STDERR] Adding extra search path custom_nodes /workspace/SwarmUI/src/BuiltinExtensions/ComfyUIBackend/ExtraNodes [ComfyUI-0/STDERR] Checkpoint files will always be loaded safely. [ComfyUI-0/STDERR] Traceback (most recent call last): [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/main.py", line 136, in [ComfyUI-0/STDERR] import execution [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/execution.py", line 13, in [ComfyUI-0/STDERR] import nodes [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/nodes.py", line 22, in [ComfyUI-0/STDERR] import comfy.diffusers_load [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/diffusers_load.py", line 3, in [ComfyUI-0/STDERR] import comfy.sd [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/sd.py", line 6, in [ComfyUI-0/STDERR] from comfy import model_management [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 168, in [ComfyUI-0/STDERR] total_vram = get_total_memory(get_torch_device()) / (1024 * 1024) [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/comfy/model_management.py", line 131, in get_torch_device [ComfyUI-0/STDERR] return torch.device(torch.cuda.current_device()) [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 971, in current_device [ComfyUI-0/STDERR] _lazy_init() [ComfyUI-0/STDERR] File "/workspace/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.10/site-packages/torch/cuda/__init__.py", line 319, in _lazy_init [ComfyUI-0/STDERR] torch._C._cuda_init() [ComfyUI-0/STDERR] RuntimeError: CUDA unknown error - this may be due to an incorrectly set up environment, e.g. changing env variable CUDA_VISIBLE_DEVICES after program start. Setting the available devices to be zero. 13:29:42.554 [Warning] Comfy backend 0 has failed to load value set repeatedly. Is it stuck loading very slowly, or has it internally failed? Ignoring errors of HttpRequestException: Connection refused (localhost:7821) 13:29:45.637 [Info] Creating new admin session 'local' for 100.64.0.34 (forwarded-for: 59.29.112.169, 172.71.110.218) 13:30:06.735 [Warning] User local requested edit of backend 0. 13:30:06.741 [Info] ComfyUI backend 0 shutting down... 13:30:09.191 [Warning] User local requested restart of backend all. 13:30:10.434 [Info] Saving backends... HELP ME PLZ..
2025-02-20 13:32

JC

The Flux links on the "28 November 2024 Update" are dead, are these still needed?

anmine42

Hi! I often use your method of inpainting with segment:yolo in swarmui. However, since the last update of swarm I get an error in swarm when including the yolo9c segementation: ComfyUI errored: { "error": { "type": "invalid_prompt", "message": "Cannot execute because node SwarmMaskBlur does not exist.", "details": "Node ID '#102'", "extra_info": {} }, "node_errors": {} } segment:yolo-face_yolov9c.pt-1,0.7,0.5 Edit: A fresh install of swarmui seems to have solved the issue...