DaSiWa ComfyUI Installer
One binary. Walk away. Come back to a working ComfyUI.
A professional-grade installer for [ComfyUI](https://github.com/comfyanonymous/ComfyUI) built on a zero-conflict, fully isolated architecture. No Python knowledge required. No admin rights needed. No system files touched.
Unified One-Page Installer
The installer now presents the full setup on a single local page. Folder selection, install mode, GPU and CUDA target, SageAttention, FFmpeg, optional downloads, and the final install plan are all visible at once, with no separate pages or script chain to follow.
That unified layout is deliberate:
Choose the ComfyUI folder once.
Pick the install mode once.
Review hardware and version overrides in the same view.
Toggle SageAttention, FFmpeg, and optional downloads without leaving the page.
Inspect the live JSON plan before starting the install.
Use Extra Settings for in-memory overrides without mutating the embedded defaults.
What it does
You run the standalone installer binary. It opens a local web UI, asks you a handful of questions upfront, then handles everything else unattended:
Downloads and configures a fully portable Python 3.12 environment isolated inside your ComfyUI folder
Clones ComfyUI at the latest stable release tag or a specific version you choose
Detects your GPU and installs the exact right PyTorch build for your hardware
Optionally installs SageAttention, FFmpeg, and a curated set of custom nodes
Creates a ready-to-launch
run_comfyuistarter in your ComfyUI folderOn subsequent runs, detects what's already installed and asks whether to update, refresh, or leave it alone
Quick Install
The release binaries are the simplest path for most users. Download the file for your OS, put it in the folder where you want the installer state to live, and run it directly:
Linux:
./dasiwa-installer-linux-amd64Windows:
dasiwa-installer-windows-amd64.exe
./dasiwa-installer-linux-amd64On Windows, double-click dasiwa-installer-windows-amd64.exe.
If you want a direct download from the latest GitHub release:
Bash:
curl -L -o dasiwa-installer-linux-amd64 \
https://raw.githubusercontent.com/darksidewalker/dasiwa-comfyui-installer/main/dasiwa-installer-linux-amd64chmod +x dasiwa-installer-linux-amd64Powershell:
$u='https://raw.githubusercontent.com/darksidewalker/dasiwa-comfyui-installer/main/dasiwa-installer-windows-amd64.exe'; $o='dasiwa-installer-windows-amd64.exe'; if (Get-Command curl.exe -ErrorAction SilentlyContinue) { curl.exe -fL --retry 5 --retry-delay 2 -o $o $u } else { Start-BitsTransfer -Source $u -Destination $o }The app opens a local browser page and runs the native Go install engine. The UI, default config, placeholder assets, README, and license are embedded in the binary, so users do not need Python scripts, shell scripts, PowerShell scripts, config.json, a node-list text file, or a cloned copy of this repository next to the executable. On first run it creates a local .dasiwa/ bootstrap directory for uv, the managed Python runtime, and cache files, then keeps all ComfyUI packages inside ComfyUI/venv/.
ā¶ļø After installation
Open the ComfyUI folder that was created and launch:
Windows ā double-click
run_comfyui.batLinux ā run
./run_comfyui.sh
Your browser will open automatically. The first launch may take a minute while ComfyUI pulls its frontend assets.
ā Do not run as Administrator ā it is not needed and can cause permission issues.
š ļø Requirements
GPU: NVIDIA (GTX 10-series or newer), AMD (RX 6000+) or Intel Arc
Disk space: ~20 GB free for a full install including models
Internet: Required during installation
Python / Admin rights: Not needed ā everything is self-contained
Git is installed automatically on Windows if missing.
āļø Disclaimer
Free community tool, provided as-is. The author is not responsible for system errors, data loss, or hardware issues. AI workloads are demanding ā make sure your cooling is adequate. Third-party software downloaded by this installer (ComfyUI, PyTorch, etc.) is subject to its own respective licenses.
Description
Evergreen installer
FAQ
Comments (138)
where is the workflows folder?
Standard comfyui folder .../ComfyUI/user/default/workflows/
@darksidewalkerĀ there's no user/ on my install. just installed it.
@ambersaiĀ ComfyUI creates the user folder after 1st launch
@darksidewalkerĀ The man the myth the legend, comes out with another amazing release!!
@darksidewalkerĀ I also cannot find an excutable or .py to actually launch the backend with. I apologize, I'm very used to the Pixaroma portable EZ Installer which adds .bats for all these launch options and addons.
Edit: It looks like they are in a comfyUI/.ci folder. Never seen one done like this but if you're installing this and can't find these files, that's where I found them.
@ambersaiĀ I located a path error on Windows, I'll prepare an update
@darksidewalkerĀ oh good its not me!
@ambersaiĀ Sorry for the hassle, W11 is sometimes ... a mess with dependencies.
I updated and it should be fixed, just delete the old install and download the new install.bat/sh
@darksidewalkerĀ confirmed working! thank you!!!
If you're having trouble finding the files to run this install, just check comfyUI/.ci and they should be in there.
Sorry for the hassle, W11 is sometimes ... a mess with dependencies.
I updated and it should be fixed, just delete the old install and download the new install.bat/sh
When I try to run the Nvidia bats I get the following error:
H:\StableDiffusionLocal\ComfyUI-Dasiwa\ComfyUI\.ci\windows_nvidia_base_files>.\python_embeded\python.exe -s ComfyUI\main.py --windows-standalone-buildThe system cannot find the path specified.H:\StableDiffusionLocal\ComfyUI-Dasiwa\ComfyUI\.ci\windows_nvidia_base_files>echo If you see this and ComfyUI did not start try updating your Nvidia Drivers to the latest. If you get a c10.dll error you need to install vc redist that you can find: https://aka.ms/vc14/vc_redist.x64.exeI haven't edited or moved this file in any way, and don't see why it wouldn't be able to find the filepath that installed it.
I have updated my Nvidia drivers and reinstalled the runtime listed at the URL, and I am still getting the same error.
Sorry for the hassle, W11 is sometimes ... a mess with dependencies.
I updated and it should be fixed, just delete the old installation and download the new install.bat/sh
Ok, darksidewalker has already fixed it, just use the new script
Not needed with latest update anymore.
Do not run the installer.bat "as admin" this will do crazy things like install it in "System32" by default. If it needs admin rights it will prompt for it.
The crash was due to W11 faking python as MSStore url if it was not installed.
@darksidewalkerĀ Thanks :) I've downloaded it 20 minutes ago from Github and still got the old installer.bat. I removed the guide to not cause any damage. Great work as always, thank you for your workflows, models and now this <3
I'm using the 1070 and got this error:
torch.AcceleratorError: CUDA error: no kernel image is available for execution on the device
Search for `cudaErrorNoKernelImageForDevice' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
Well ... I did not think of supporting 10y old hardware ... this is not implemented yet. Maybe on a next release.
@eptgames I added Nividia 10xx (pascal) support, but could not test. Maybe have a try now :)
It should install the most stable python and torch version for that hardware, but do not expect it to run everything flawless with comfy, because that's really dated.
@darksidewalkerĀ Oh! Thank you for your trouble. I'll try it later.
Thanks for the installer! I have a question: How could I install sage attention and triton to this comfyui installation?
The same as you would do it on any other comfyui installation.
Ultra noob here; I placed the .bat on my ComfyUI installation folder (not where the "ComfyUI.exe" is located). So once I installed all the stuff requested on the bat, do I just simply open ComfyUI as always?
As I do not know how you start comfy normally, I cannot tell, but the installer places a run_comfyui to start it. There are many ways to start comfyui.
Hello. I installed and ran it with your workflow, and with DaSiWa WAN 2.2 i2v FastFidelity C-SVI-22 I get some problems: nodes 146, 201, 389, 152 are not working - status (disconnected). Reinstalled, updated, nothing helps, and I'm not an expert. At the same time, workflow itself starts and even gives results, but because of the idle node 152, I can't even write a prompt prompt in "Sampler #02" and further, only "Sampler #01" works
You should update to the latest workflow version, also this is the page for my comfyui installer, not the workflow itself.
@darksidewalkerĀ I used your latest worlflow. And it works with comfyUI installed via Matrix, but for some reason it has errors with the comfyUI that I install using your installer. There are no questions about workflow itself
@aurau673Ā C-SVI-22 is old, not my latest. That's why I wrote this. Maybe you also used an outdated installer. I don't know.
@darksidewalkerĀ Sorry, I just specified the wrong workflow version. I used C-SVI-27. And the installer downloaded and installed yesterday from this page. And now I tried downloading and installing it again, and I got the same thing. At the same time, it don't show such errors with your other workflows, only with C-SVI. Well, I guess I'll have to use another ComfyUI. Thanks
@aurau673Ā Maybe you did not start comfyui correctly and it did not load the custom nodes with dependencies. I can assure you, it is working, tested yesterday.
Are you a god or something?
Hahaha š thx, just an ordinary ancient dark vampire sorcerer š
@darksidewalker will not install it quits at: Python 3.12 not found at C:\Program Files\Python312\python.exe
my python is where it most people install it at: AppData\Local\Programs\Python\Python312
I implemented a new version with a other python logic for W11, maybe try now.
But if you would have installed python system-wide, it would have worked. Just saying.
At risk of sounding like an idiot, I am not getting the run_comfyui.bat after using the install.bat. I did put Python312 where it asked and it did create a ComfyUI folder but no run within that folder for the next steps.
Should I just like run it all again from scratch?
Maybe a bug, any other errors?
@darksidewalkerĀ Not that I can tell, the install.bat seems to run and complete finr, I tried deleting and doing it again from the start making sure my antivirus or any other programs weren't trying to quarantine specific files, but it just isn't giving me the run for Comfy.
I reworked the tool yesterday, changing some logic. Maybe try the new installer if this works. Tested it on a W11 and it worked just fine.
Can you please make this tool so it installs portable Python and keeps everything under one folder? This seems to be the issue everyone is having. Many do not have python installed on their computers and want to use self contained ComfyUI installs. Thank you! You are amazing!
Well, it is, the runner was just not updated. Did the update and mentioned in description/readme.
@DarksidewalkerĀ could you please tell me what i'm doing wrong? i have no idea what i'm doing, but as far as i can tell, neither runner, either from the zip file or the powershell command, do anything to prevent uv and other things from going into users/appdata, i found out the hard way by bricking your install when i randomly came across these folders on my own; i can't find any mention of any update, or mention in any readme, or any description, and nothing in the script is even attempting to create these folders/installs locally
I wish i had found this guy sooner. Tried lots of workflows and became an unwilling python pro before breaking everything again. Now i make booby vids.
I feel that so hard man.
darksidewalker truly the GOAT
Does not detect my GPU. I'm on linux mint with a 5060 ti and it thinks I'm using Intel integrated. Time to try the downgrade option I guess.
Any log?
I updated the detect logic a bit, may you try again?
@darksidewalkerĀ Now it detects my gpu correctly, nice!
What is the intended way to edit or update the installed custom nodes? I had assumed that the install script would do this, but I can't find a way to make a simple custom_nodes.txt edit stick since the install script replaces it with the one urls.custom_nodes in the config file points to, and if there's a separate custom node install/upgrade script I can't find one.
Could of course override urls.custom_nodes with my own, but that seems excessive for just a couple of additions.
You are right, I changed the download logic and it will always overwrite since there is only an overwrite option for the config.json atm. You could just install your custom nodes after the installation with git or manager, till I implement an overwrite option.
@darksidewalkerĀ Cool, thanks. This tool is invaluable when comfy occasionally and inevitably blows itself up on update and has to be nuked and reinstalled. Great work!
@ashurbanipalĀ yeah for me too, that is why I made it, reinstall everything is a chore without it
čæäøŖcomfy UIäøęÆęMacē³»ē»ēm4 maxåļ¼å®č£ ę¶gpuéä»ä¹ļ¼
Read on front page, please. No MAC support. Nvidia, AMD and Intel works.
Am I doing something wrong with the installation or do I have to manually install all the missing nodes and models? I'm able to get all the missing nodes through the comfy missing nodes installer but not for the models. I can get the diffusion models from Civitai but I don't know where to get all the other models.
Yeah, I am wondering this too.
This installer is for my workflows, if everything works correct there are no missing nodes.
Models are not automatic downloaded, since they depend on the workflows and are also described there in detail (with urls).
@darksidewalkerĀ I'm stupid. I didn't know there were further instructions in the workflow itself. Thank you.
@CannedPsychoĀ Same... Thank you!
I love it! Thanks so much! Its so much faster! Thanks for all of your great work! Insane! :-*
With ur installer there is no Python Embeded Folder. Sage Intention is kinda there but can not be used, should be installed is the message :D is there a solution?
Sorry I can not help with sageattention.
There is no python embedded, because UV directly installs all of python portable into the venv.
@darksidewalkerĀ okay but thank you anyway! You are the best!
I will try this installer.
Edit: the installer works well by itself. Getting all the necessary files together outside of the main install can be time consuming, but the end results are worth it.
installed perfectly on the ai 395. question, how to get rid of the warning about missin RTX nodes: comfyui_nvidia_rtx_nodes (1) which i oobviously cant install? other then that - great stuff!
Don't use a workflow where RTX nodes are in or delete the RTX nodes.
installed fine without sage att2.. when I run the batch file "run_comfyui" I get a 127.0.0.1 refused to connect. Advice? None of the usual fixes for firewall or permissions /ports seems to work.
Comfyui is already running and it cannot open the port again.
@darksidewalkerĀ no, this is not the case. comfy ui is definitely not already running. I wanted to let you know that running the script resulted in my entire drive in read only mode and its been a disaster. really messed up my environments and I would recommend pulling this down for a look.
@memyselfandeĀ that is nonsense, the script can not do that. It is just the normal start line for comfyui
Where is the download for the littledemonv2 safe tensor?
On my page
@darksidewalkerĀ hey sorry do you have the direct link i cant find it
@baronvonewmanĀ civitai.com/models/2151205
@darksidewalkerĀ thank you!!
Please Help. Is comfy Manager in a different spot in this workflow it doesnt appear in the same spot. its unable to download the node that holds all the checkpoints. maybe i downloaded something wrong i have been trying for a week before commenting.
My installer installs the up2date manager, not the legacy one. The button is called extensions or you use the option in comfyui to enable the legacy one
Tried your installer. It ask if i'd like to downgrade GiT from 2.53.2 to 2.53.0. If I say no it closes out. If I say yes it wants me to terminate MSYS2 and then closes out without moving forward.
Should be fixed
sage attention did not work for me automatically on windows 11 for rtx 3080 10gb. I resolved it by running these commands
$env:UV_SKIP_WHEEL_FILENAME_CHECK=1
& "driveletter:\yourpath\ComfyUI\venv\Scripts\uv.exe" pip install --force-reinstall --no-cache --python "driveletter:\yourpath\ComfyUI\venv\Scripts\python.exe" https://github.com/woct0rdho/SageAttention/releases/download/v2.2.0-windows.post4/sageattention-2.2.0+cu128torch2.9.0andhigher.post4-cp39-abi3-win_amd64.whl
& "driveletter:\yourpath\ComfyUI\venv\Scripts\uv.exe" pip install --no-cache --python "driveletter:\yourpath\ComfyUI\venv\Scripts\python.exe" "triton-windows==3.6.0.post26"
basically bypassed the installers that are provided which do not work and used different ones that dont require the whole cuda setup either and worked like a charm ;)
Nice, just fixed it. the installer now also works to fetch the correct wheel.
@DarksidewalkerĀ Hmm, I'm on ubuntu, and my sageattention failed also (I just did a fresh use of your installer; sageatt was working in my old folder setup, so I'll take a look and see if I can fix it myself) -- thanks for the installer though, clean and fresh, feels great
@skipwestcottĀ Do you have an error, why it did not work?
@DarksidewalkerĀ I hadn't noticed at the time using the installer, but I just get ' from sageattention import sageattn
ModuleNotFoundError: No module named 'sageattention'
Pretty basic, but no, I didn't see the installer show an error; can it be 'repaired' somehow, or am I on the hook to finish this up manually?
Actually, now that I'm thinking about it, I'm not sure of even the name of the Venv your runner starts up, so I'm not 100% on how I would add it just yet ...
@skipwestcottĀ the error just says that there is no sageattentions installed.
Normally the installer should work if anyhow possible, otherwise there is only manual install possible.
@DarksidewalkerĀ yeah, I have a SageAttention folder from the install, just isn't detected.
I'm fairly new to this and installed everything, but your workflow, which I understand uses your nodes, didn't install any and as its an offline instance, how would one install the missing nodes? I looked at some comments and it seemed for most it worked out the gate.
This is not an workflow. This installer will install comfyui, adding custom_node for my workflows and additional components automatically.
At the end you will have an ComfyUI folder. If you used this, there should be no missing nodes.
Made a good 2 hour attempt to make this work on my 5090, but no bueno. Sure I can get it to work on cu128 without sage attention, but why would I want to run slow? Went back to my python 3.11.9 with CU130 and Pytorch 2.11... and Sage 2.2. Keep up the good fight, I appreciate your models and lora work.
I don't know what you're mean by run slow.
anyway to make this run in a cloud gpu? I tried this ver's comfyui, but it doesn't run the server and instead runs into errors. I tried skipping to python ./main.py to no avail. However, the github ver. does work
The servers may have no UV. I don't know. This installer is made for local installs.
With your recent Omniforge workflow, when I load it in this portable ComfyUI, I get:
Installation Required
Install RequiredStringFunction|pysssss
Install RequiredDaSiWa_ResolutionScaleCalculator
in subgraph 'Settings and Backend'
Install RequiredDaSiWa_NodeStatusSwitch
in subgraph 'Settings and Backend'
Install RequiredFrameInterpolationModelLoader
in subgraph 'Settings and Backend'
Install RequiredFrameInterpolate
in subgraph 'Settings and Backend'
After reading description and comments, it seems it should have been installed automatically no? Let me know if I'm doing something wrong or if I should proceed with downloading/installing them manually
Comfyui is outdated
How can i make it use cu130 for optimized CUDA operations... it defaults to pytorch version 2.10.0+cu128
Write a config (howto is on the github). Normally it defaults to patorch 2.11+ and cu130 atm, except choices need something else.
Can confirm that this installer works on Windows 11 with RTX 5090. SageAttention and NAG both work. Generated with DaSiWa's workflow and checkpoints (both safetensors and gguf). Great work on the installer, very smooth!
I'm getting an error about this video node ComfyUI-LTXVideo nodes
I install it from the extension manager but the it says there's an install error and the error flag persists. Is there a solution for this?
(sorry for the repost, thought this page was more appropriate but can't find the original comment)
The LTX custom nodes installs a wrong kornia version, they have to fix it or you have to downgrade kornia to like 0.8.2 or earlier
No input found for flattened id [146:389] slot [0]
This is not an installer error. š¤·
seems like something broke cuda and sage with latest update, even refresh or clean reainstall finish with fail - AssertionError: Torch not compiled with CUDA enabled
Press any key to continue . . .
There is only one solution... Switch to Linux š
Should be fixed
I got an error, but I can not find a solution:
(I am pretty new to this. I tried a few things. But nothing worked.)
[*] Installing Torch for NVIDIA (NVIDIA GEFORCE RTX 3060)...
Using Python 3.12.13 environment at: venv
x No solution found when resolving dependencies:
`-> Because there is no version of torch==2.12.0+cu128 and you require torch==2.12.0+cu128, we can conclude that
your requirements are unsatisfiable.
hint: torch was found on https://download.pytorch.org/whl/cu128, but not at the requested version (torch==2.12.0+cu128). A compatible version may be available on a subsequent index (e.g., https://pypi.org/simple). By default, uv will only consider versions that are published on the first index that contains a given package, to avoid dependency confusion attacks. If all indexes are equally trusted, use --index-strategy unsafe-best-match to consider all versions from all indexes, regardless of the order in which they were defined.[-] CRITICAL phase failed: torch install: Command '['uv', 'pip', 'install', 'torch==2.12.0+cu128', 'torchvision', 'torchaudio', '--extra-index-url', 'https://download.pytorch.org/whl/cu128']' returned non-zero exit status 1.
[-] Installation aborted: Command '['uv', 'pip', 'install', 'torch==2.12.0+cu128', 'torchvision', 'torchaudio', '--extra-index-url', 'https://download.pytorch.org/whl/cu128']' returned non-zero exit status 1.
Please try again
@DarksidewalkerĀ It worked. Really a big thanks to you. (I also looked at the other comments and your response time is amazing. Keep up the good work :) )
I'm having an issue with the auto installer and a fresh Install of Windows 11 with a 5080.
hint: setuptools was found on https://download.pytorch.org/whl/cu130, but not at the requested version (setuptools==81.0.0). A compatible version may be available on a subsequent index (e.g., https://pypi.org/simple). By default, uv will only consider versions that are published on the first index that contains a given package, to avoid dependency confusion attacks. If all indexes are equally trusted, use --index-strategy unsafe-best-match to consider all versions from all indexes, regardless of the order in which they were defined.[-] Phase failed: priority package enforcement: Command '['uv', 'pip', 'install', '--upgrade', 'kornia==0.8.2', 'setuptools==81.0.0', 'triton==3.7.0', 'torch==2.12.0+cu130', '--extra-index-url', 'https://download.pytorch.org/whl/cu130']' returned non-zero exit status 1.
Any help boss?
Try again please
It went further! But now has a triton error.
Using Python 3.12.13 environment at: venv
Ć No solution found when resolving dependencies:
ā°āā¶ Because triton==3.7.0 has no wheels with a matching platform tag (e.g., win_amd64) and you require
triton==3.7.0, we can conclude that your requirements are unsatisfiable.
hint: Wheels are available for triton (v3.7.0) on the following platforms: manylinux_2_27_aarch64, manylinux_2_27_x86_64, manylinux_2_28_aarch64, manylinux_2_28_x86_64[-] Phase failed: priority package enforcement: Command '['uv', 'pip', 'install', '--upgrade', 'kornia==0.8.2', 'setuptools==81.0.0', 'triton==3.7.0', 'torch==2.12.0+cu130', '--extra-index-url', 'https://download.pytorch.org/whl/cu130', '--index-strategy', 'unsafe-best-match']' returned non-zero exit status 1.
[!] Continuing with remaining phases...
@DarksidewalkerĀ Sorry I forgot to add you to the reply
Try again :)
@DarksidewalkerĀ So far so good Ty Sir!
Thank you, this helps a lot!
I am currently using the easy Comfy installer (Tavris1) and had been getting a lot of errors in the past week, so in desperation I went to Comfy Manager and clicked 'UPDATE ALL'.....oh FFS what a mess that made?!
All my Dasiwa WAN22 and LTX workflows had white bars across the video options that said 'DISCONNECTED' and I had no idea how to fix them.
After a few desperate hours of trying to save my beloved workflows I discovered an 'experimental' feature where you could roll back to a previous backup of the instance (which luckily it did itself prior to clicking 'update all').
Should I use this portable comfy edition instead? Just be very careful with updating and addressing the warnings you get on startup about this or that being outdated?
I genuinely thought I would not be able to get the Dasiwa WAN / LTX workflows going again at one point.
Well my installer is no harm, it just creates an portable folder, nothing is changed on the OS. If it does not work, just throw the folder away.
@DarksidewalkerĀ Thanks my friend. I appreciate the advice. I'm installing your version now as it looks like it updates itself automatically and does NOT break your incredible workflows and models like the other Comfy version did after updating all the components. Even better, it looks like if you release a new model for WAN or LTX, it even updates that as well?
There is so much complexity to your workflows 'under the hood', so when you get 3-6 white fields that all say 'DISCONNECTED' - an amateur AI enthusiast like me has no idea where to begin to try to fix them.
AI help just told me to 'reconnect the wires' - but unless you know what you're doing, you would have no chance at fixing this.
I like how you keep a simple interface on top, and have everything else hidden in the background.
@DarksidewalkerĀ - dumb question time (sorry).
I thought the idea of using your all in one Comfy installation was that it would go out and fetch all the required models for your workflows (that I included in the download)?
But everything is empty and showing multiple errors for all the missing models.
Does that still have to be done manually? There is no model manager that fetches the exact models that your workflows require?
I thought one of the benefits is that this Comfy would automatically update the models and workflows as you introduced new versions? I may have misunderstood.
@maxdee007909Ā it is not designed to update and download the models, this would be TB data
A few weeks ago I used the installer and worked just fine. Then I messed up and had to reinstall, now it doesnt work anymore. Some problem with the Dasiwa_ResolutionScaleCalculator. This part is blank now in the SETTINGS tab.
Value not in list
mode: 16 not in ['Standard', 'WAN/LTX (Div32)', 'LTX 2-Stage (Div64)', 'CUSTOM']
Value 0 smaller than min of 1
custom_aspect_height
Failed to convert an input value to a INT value
custom_aspect_width, 9:16 - Social, invalid literal for int() with base 10: '9:16 - Social'
Maybe your browser cache messed it up, clean it, refresh, and download the workflow again
@DarksidewalkerĀ Thanks lad, I did that and it is working again now. Cheers.
I seem to be getting an error when starting the installer:
"Creating venv with Python 3.12... Native installer exited with error: exec: "uv": executable file not found in %PATH%"
I am not sure what executable file it is looking for? I'm also not sure where it is looking for it. I have tried my ComfyUI installation folder and the folder with that executable, but it isn't working.
Any tips?
Might be a bug, on what OS you are using trying?
@DarksidewalkerĀ Windows 11. I know, I knowā¦ the real solution is to use Linux ;). But alasā¦
@JadziaDaxxx42179Ā I will look into it, I did not extensively tested on W11 for v2.0, so this might just be a bug
yeah, had the same issue as well
Should work now
@DarksidewalkerĀ ty
@DarksidewalkerĀ Thanks! It works!
This is really high-quality. Every time I've re-downloaded and used it, it's a meaningfully better experience. Thanks a lot for your hard work.
For me it skip downloading additional files with 0 errors.
Says like:
DaSiWa WAN 2.2 i2v FastFidelity C-AiO-18.json already exists; skipping.
DaSiWa WAN 2.2 i2v FastFidelity C-SVI-10.json already exists; skipping.
DaSiWa LTX OmniForge C-LTX23-10.json already exists; skipping.
DaSiWa - Hybrid - DeepDream - C-WLTX-10.json already exists; skipping.
Downloading T5 Encoder (1/3)...
Downloading T5 Encoder (2/3)...
Downloading T5 Encoder (3/3)...
and then just keep going and folder where it should be is empty
I made the dl more robust, but I did not find any issues. You could update the app and try again.
Running run_comfyui.bat it can't connect to: 127.0.0.1:8188
Any idea?
Something is already running on that port.
This like your models is f'n awesome!
Thank you!
I'm getting a 404 error when attempting to curl the windows .exe file and cannot see it in the repo. I used the command you have in the read me. Are there any issues currently with the repo or is it something on my end?
Should work now, or go and dl from github.
@DarksidewalkerĀ Worked! Thanks for the quick response!