Comfyui automatic1111 not working. Step 3: Download models. First, open ComfyUI navigate to " Manager " and click " Update All " to update ComfyUI and the nodes. 3. Apr 6, 2024 · update ruff to 0. I can confirm the embeddings are there, and that they do work. Python 3. Once in a while, it DOES work, but for some reason ComfyUI got attention recently because the developer works for StabilityAI and was able to be the first to get SDXL running. Jul 19, 2023 · We’re building the MEGAZORD of image generation power. already used both the 700 pruned model and the kohya pruned model as well. Has no effect if no latent_keyframes are passed in, or no batch_indeces are Once installed, hit apply and reload under the tab Installed. In this guide, we'll show you how to use the SDXL v1. Here’s a thorough comparison to aid in your decision-making. I'm sure there's an easy way to activate a venv in python, but when I google the issue I get a milliion different ways of doing it that aren't seemingly applicable to this codebase. The AI gives more attention to what comes first in each chunk. I use to work with Auto1111 on a daily basis but recently, everything went sideways, extension , inpaint, controlnet, starting to exhibit random issues and crashes. I don't know what causing it, but textual inversions that worked yesterday, today stops working. It's not about being stubborn, it's just that how I use Automatic1111 there is nothing I would automate and thus no reason to change. AUTOMATIC1111 stable-diffusion-webui v1. Add some content to the following directories: C:\stable-diffusion-webui\embeddings. If you use a rectangular image, the IP Adapter preprocessor will crop it from the center to a square, so you may get a cropped-off face. Members Online Up to 7fps now using SD-Hyper 1 Step LORA - perfect for little Wizards. I've del'd the ComfyUI directory & git cloned it back in. I upped Python to 11. Launch ComfyUI by running python main. Use the same resolution for generation as for the original image. Next. Not quite. If the problems persist, then most probably, you've messed something in the bat Jan 6, 2024 · ComfyUI surpasses Aotomatic1111 in versatility and capability, offering a comprehensive suite for AI art generation. Sep 27, 2023 · This is the not so elegant way and only worked once for me: Clone ComfyUI 2. Soft inpainting seamlessly adds new content that blends with the original image. I have a 4090 and it takes about the same time as SDXL to generate an image. And use automatic1111 for sd 1. 6 is assumed to installed and working properly. The folders insightface and insightface-0. Mar 12, 2023 · I rebuilt the venv for the new folder and the newest update does not put any activate scripts, bat, ps1, nadda, into the scripts folder. pt files to *. Enable a Controlnet. I use JSON. That is an incredible increase in load times, especially if you are loading 10 checkpoints to do block merges lmao. As a lots of people said there is already a standalone app named Comfy UI. Like, yeah you can drag a workflow into the window and sure it's fast but even though I'm sure it's "flexible" it feels like pulling teeth to work with. Notably, it does not expect column names in the first row and both count and aliases are technically optional, although count is always included in the default data. I did two things to have ComfyUI glom on to Automatic1111 (write up: AI Art with ComfyUI and Stable Diffusion SDXL — Day Zero Basics For an Automatic1111 User | by Eric Richards | Jul, 2023 | Medium ) Editing the ComfyUI configuration file to add the base directory of Automatic1111 for all the models and embeddings and Python 3. If set to 0. Contribute to Coyote-A/ultimate-upscale-for-automatic1111 development by creating an account on GitHub. " Does anybody have idea how to fix it and make ConfyUI working again with isntalled WAS-NS, please? I'm not familiar with Python etc. В этом видео разберемся что такое ComfyUi и зачем он нужен. Remove what you tried to do: Delete all folders inside venv\lib\site-packages\ named: torch torchvision torchaudio. ) Set the Mask Blur to 40. Also, if this is new and exciting to you, feel free to post, but don't spam all your work. If --upcast-sampling works as a fix with your card, you should have 2x speed (fp16) compared to running in full precisi I installed the ControlNet extension in the Extension Tabs from the Mikubill Github, I downloaded the scribble model from Hugging face put it into extension/controlNet/models. Advantages of ComfyUI: ComfyUI VS AUTOMATIC1111. ai, demonstrating its robustness and reliability in daily use. So from that aspect, they'll never give the same results unless you set A1111 to use the CPU for the seed. Do you use AUTOMATIC1111’s Stable Diffusion webui? Do you want to add the power of node-based AI workflows to the tools you already use? We’re committing STABLE INCEPTION today on Building Dreams. If you have another Stable Diffusion UI you might be able to reuse the dependencies. Step 3: Download a checkpoint model. Troubleshooting. Just earlier today it was working fine. Installation is straightforward, supporting a wide range of hardware configurations, including CPU-only modes. refiner 0. Its making the best of both worlds: Automatic1111 and ComfyUI. Ticked Enable under ControlNet loaded in an image, inverted colors because it has white backgrounds. 10:8188. AUTO1111 and ComfyUI unite with the Another easy trick is just drag the input bubble to blank space and release and it will give you a list of recommended fitting input nodes. Turn on Soft Inpainting by checking the check box next to it. Don't use other versions unless you are looking for trouble. scanner. We will go with the default setting. Then run: venv\scripts\activate. Automatic1111 is still popular and does a lot of things ComfyUI can't. After reloading close the windows console running automatic1111. Aug 8, 2023 · Ideally this should work but it didn’t for me which is when instead of spending hours troubleshooting I found a better way. I'm finding that I have no idea how to make this work with the inpainting workflow I am used to using in Automatic1111. By creating and connecting nodes that perform different parts of the process, you can run Stable Diffusion. 10:7862, previously 10. Hey guys, When I click on the Textual Inversion tab in AUTOMATIC1111, it gives me the following message: Nothing here. Much more efficient and effective than Automatic1111, for me, which I also still use. ComfyUI lives in its own directory. If you're watching this, you've probably run into the SDXL GPU challenge. Not familiar with the vocabulary, but it seems like it's not adding in more detail from the prompt as it's de-noising. Had to roll back to a 531 version of Studio. 1 is assumed to be installed and working properly stable-diffusion-xl models are assumed to present Git is assumed to be installed and working properly. 0 on 8GB VRAM? Automatic1111 & ComfyUi. example” file and rename it to simply “extra_models_path. /webui. ↑ Node setup 1: Generates image and then upscales it with USDU (Save portrait to your PC and then drag and drop it into you ComfyUI interface and replace prompt with your's, press "Queue Prompt") ↑ Node setup 2: Upscales any custom image Some of the example combinations like having multiple prompts working on multi areas of a composition is something that Comfy excels at and would prove much more labor intensive in Automatic1111. Note that this is Automatic1111. Jul 19, 2023 · AUTO1111 and ComfyUI unite with the sd-webui-comfyui extension from ModelSurge. 2. Well, You should share a paths to where we should put extra models , this workflow is intimidating if you did not created it yourself, you know whats where and thats fine but be mindful that others have no idea, so i tried it, doesnt work and im not a noob with nodes, i just dont see where i need to point to face detection model and segment Automatic1111 Textual Inversion stop working. I highly recommend to move it up: CompfyUi. I've rebooted. 0 with Torch 2. But yes just take the workflow image or final picture PNG and drag it into comfyUI and it recreates the workflow. This works BUT I keep getting erratic RAM (not VRAM) usage; and I regularly hit 16gigs of RAM use and end up swapping to my SSD. Video tutorial on how to use ComfyUI, a powerful and modular Stable Diffusion GUI and backend, is here. WAS-NS doesn't use any YAML that I am aware of. Create the folder ComfyUI > models > instantid. Nvidia GPU RTX 4070. I am not training models or anything like that so I'm not missing that feature. The speed on AUTOMATIC1111 is quite different. Select Controlnet Control Type "All" so you can have access to a weird May 2, 2024 · Upgrading to IP Adapter V2: 1. Installing ComfyUI ReActor problem I realise this may not be the place to seek such answers but I’m looking to cast as wide a net as possible, and so far this community has been the go-to spot for technical answers with Automatic1111, so I figure maybe I’ll find some answers to a ComfyUI question too. py; Note: Remember to add your models, VAE, LoRAs etc. Stable Diffusion v1. Oct 22, 2023 · Insightface seems to be not installed properly. AUTO1111 and ComfyUI unite with the sd-webui-comfyui extension from ModelSurge. Anything but Automatic1111. Dec 19, 2023 · In ComfyUI, every node represents a different part of the Stable Diffusion process. В чем его Feb 14, 2024 · Takes about 3 minutes to generate an image on an RTX 3070 though. Can't seem to get SDXL working when using LoRAs on Automatic1111. 4 & upped PIP to 23. That's a very long time. A1111 released a developmental branch of Web-UI this morning that allows the choice of . But I still get the same message when running PYTHON MAIN. Prompt: beautiful landscape scenery glass bottle with a galaxy inside cute fennec fox snow HDR sunset Sampling method: Euler a Sampling steps: 1 Size: 512 x 512 CFG Scale: 1 Step 1: Generate some face images, or find an existing one to use. Attached is a screen cap: top window has the directory where the checkpoint files are located; second window, the search paths I For instance (word:1. It also seems like ComfyUI is way too intense on using heavier weights on (words:1. Then use the manager to install any missing nodes. Enter your OpenAI API key. "easynegative") I will get just black image with several colored blobs. Should you use ComfyUI instead of AUTOMATIC1111? Here’s a comparison. py, then delete venv folder and let it redownload everything next time you run it. Aug 29, 2023 · How to install stable diffusion SDXL? How to install and use ComfyUI?Don't do that. Download the InstandID IP-Adpater model. File "D:\ComfyUI_windows_portable\python_embeded\lib\site-packages\yaml\scanner. checkpoint image generation -> hires fix -> refiner. step 30. In Open WebUI, navigate to Settings > Images. 6. Download the zip file from github, extract into a new folder, copy the models folder from your old Automatic1111 folder, + copy and replace the web. Note that image size options will depend on the selected model: DALL·E 2: Supports 256x256, 512x512, or 1024x1024 images. create an image in 30 steps -> hires fix -> change from hires fix step 5 to refiner -> genaration complete. To make sure you get the right ones, according to Pytorch, what you do is: Activate the venv: (open a command prompt, and cd to the webui root). My folders for Stable Diffusion have gotten extremely huge. In fact, this extension lets you do something I'm not sure is even possible in Comfy: defining regions using prompt keywords. 0 will not have any effect during the duration of this Timestep Keyframe's effect, and will increase sampling speed by not doing any work. 1) in ComfyUI is much stronger than (word:1. On my freshly restarted Apple M1, SDXL Turbo takes 71 seconds to generate a 512×512 image with 1 step with ComfyUI. At loading stage in the console it prints that all Deoldify not working in Automatic1111 Question - Help Share Add a Comment. The benefits of using ComfyUI are: Lightweight: it runs fast. 0. I recommend using 512x512 square here. yaml”. Select the SDXL checkpoint that you want to use. Mar 14, 2023 · Also in the extra_model_paths. Sometimes when I'm having a plethora of problems, the solution is having a new clean sheet Automatic1111. ok so your checkpoint folder and vae are probably empty in the main comfy ui portable folder. Aug 6, 2023 · ComfyUI: Harder to learn, node based interface very fast generations, generating anywhere from 5-10x faster than AUTOMATIC1111. Feb 7, 2024 · ComfyUI_windows_portable\ComfyUI\models\upscale_models. 5 model (for example), the embeddings list will be populated again. New Branch of A1111 supports SDXL Refiner as HiRes Fix. Tested on my 3050 4gig with 16gig RAM and it works! ComfyUI’s feature updating is more related to the commercial company StabilityAI's commercial promotion of StabilityAI's own products, like SVD, and so on. ЧАСТЬ 1 Начинаем новую серию видео про ComfyUI. DALL·E 3: Supports 1024x1024, 1792x1024, or 1024x1792 images. hires fix step 10. ComfyUI VS AUTOMATIC1111. . 3 get the nix-autobahn script and place it in the ComfyUI folder: GitHub - Lassulus/nix-autobahn (for some reason nix-ld did not work) So just switch to comfyui and use a predefined workflow until automatic1111 is fixed. I do all my own static image work in Automatic1111. I remember it working fine before, but it seems to have gotten buggy at some point. Other repos do things different and scripts may add or remove features from this list. 20. All the workflows I am seeing for animation are done in ComfyUI so I use that for animation work now, but I still stick to Automatic1111 for everything else. Open pose simply doesnt work. Without inversions all works well. Download ComfyUI SDXL Workflow. Collaborate outside of code Extension for Automatic1111 and ComfyUI to automatically create masks for Background/Hair/Body/Face train embeddings on 8GB (also reports of 6GB working) Extras tab with: GFPGAN, neural network that fixes faces; CodeFormer, face restoration tool as an alternative to GFPGAN; RealESRGAN, neural network upscaler; ESRGAN, neural network upscaler with a lot of third party models; SwinIR and Swin2SR , neural network upscalers Apr 18, 2024 · On the ComfyUI Manager menu, click Update All to update all custom nodes and ComfyUI iteself. I have about a decade of blender node experience, so I figured that this would be a perfect match for me. In Auto1111, SD processes the prompts in chunks of 75 tokens. 400 GB's at this point and i would like to break things up by atleast taking all the models and placing them on another drive. PY AUTOMATIC1111 stable-diffusion-webui. Install the ComfyUI dependencies. works for me. dll files in stable-diffusion-webui\venv\Lib\site-packages\torch\lib with the ones from cudnn-windows-x86_64-8. There are some custom nodes/extensions to make generation between the two interfaces compatible. Step 1: Install 7-Zip. I went into Automatic1111, it crashed while I was switching models (safetenors). 163_cuda11-archive\bin. Pros of Using ComfyUI: I was facing similar issues when i first started using ComfyUI, try adjusting CFG scale to 5 and if your prompts are big like in A1111, add a token merging node. 6 64-bit. 5 since they have different extensions and models to allow me to use one set of settings with SDXL and a different set of setting with SD v1. If I use EasyNegative for example, it works, I just don't see any of the others. Sort by: Best. This extension is obsolete. I think the noise is also generated differently where A1111 uses GPU by default and ComfyUI uses CPU by default, which makes using the same seed give different results. This is due to the larger size of the SDXL Turbo model. Open comment sort options ComfyUI AnyNode now lets Local LLMs code nodes Txt2img. Putting the *. 9. Get a good quality headshot, square format, showing just the face. Hello, I tried various LoRAs trained on SDXL 1. D:\ComfyUI_windows_portable>pause 请按任意键继续. Automatic1111 is not slower in implementing features. I am really sorry for the confusion 😕 Nov 26, 2023 · Automatic1111からComfyUIへの移行で手間取った点を備忘録として残します。少しずつ移行して行く予定なのでその都度更新します。 ComfyUIのダウンロードや基本的なGUIの説明は以下サイトが参考になります。 【Stable Diffusion】ComfyUIとは?インストール方法と基本的な使い方について | イクログ (ikuriblog For this, we simply need to copy the “extra_models_path. In order to improve faces even more, you can try the FaceDetailer node from the ComfyUI-Impact-Pack plugin. By combining various nodes in ComfyUI, you can create a workflow for generating images in Stable Diffusion. Easy to share: Each file is a reproducible workflow. Multiplies the attention to x by 1. AUTOMATIC1111 has long been regarded as the default choice for Stable Diffusion GUI. Go to the preprocessing tab; Select "Use BLIP for caption" Preprocess images; What should have happened? In the end, it turned out Vlad enabled by default some optimization that wasn't enabled by default in Automatic1111. 🟦null_latent_kf_strength: strength to assign to latents that are unaccounted for in the passed in latent_keyframes. A1111 does a lot behind the scenes with prompts, while ComfyUI Doesn't, making it more sensitive to the Prompt length , sampler shouldn't affect but i always use Euler normal , try it out. Nov 30, 2023 · Enter txt2img settings On the txt2img page of AUTOMATIC1111, select the sd_xl_turbo_1. So, SDXL Turbo is still slower. It's possible, I suppose, that there's something ComfyUI is using which A1111 hasn't yet incorporated, like when pytorch 2. Automatic1111 is fully Apr 18, 2024 · Soft Inpainting. StabilityAI is a commercial and for-profit company, devoting for offline and open source models. Do you use AUTOMATIC11 We're building the MEGAZORD of image generation power. Feb 24, 2024 · In ComfyUI, there are nodes that cover every aspect of image creation in Stable Diffusion. ComfyUI is the least user-friendly thing I've ever seen in my life. bat file. 5 seconds now (with no Loras on a 3090) there is the 10-20 min wait to convert each model, but it is worth it to do your favorites. pt files in the embeddings directory of Automatic1111 (stable-diffusion-webui\embeddings) does not work. ScannerError: mapping values are not allowed here in "D:\ComfyUI_windows_portable\ComfyUI\extra_model_paths. But here's the thing: This rule isn't about the whole prompt, but for each chunk. Apr 10, 2024 · It seems their implementation for loading is pretty outdated, they took a patch from Automatic1111 and tried it out. x, embeddings that are created with 1. Also I click enable and also added the anotation files. Members Online IC-Light - For manipulating the illumination of images, GitHub repo and ComfyUI node by kijai (only SD1. The program needs 16gb of regular RAM to run smoothly. Win 11 System: #Change Paths as desired 0 Command Prompt: Install and run with:. A1111 is REALLY unstable compared to ComfyUI. neither has any influence on my model. If you have 8gb RAM, consider making an 8gb page file/swap file Nov 30, 2023 · ComfyUI. i've run into issue that textural inversions are not showing in the list. 1. Feb 19, 2023 · Part of what my PR does on that matter, just made it into a recent commit. ADMIN MOD. Jul 18, 2023 · I just installed ComfyUI, no problems, everything seems to be running good, with one exception. So, you’ll find nodes to load a checkpoint model, take prompt inputs, save the output image, and more. News. The numbering system follows the tag API docs of Danbooru: Apr 15, 2024 · ComfyUI and Automatic1111 Stable Diffusion WebUI (Automatic1111 WebUI) are two open-source applications that enable you to generate images… 6 min read · Mar 21, 2024 Prompting Pixels bmemac. 0 on automatic1111, but about 80% of the time I do, I get this error: RuntimeError: The size of tensor a (1024) must match the size of tensor b (2048) at non-singleton dimension 1. what you need is to either copy and drop them there or use a " A symbolic link (symlink) " in windows to basically shortcut the windows to link it to your other SD folders mainly either the vlads, easydiffusion or Automatic1111 directory. Clean install of Automatic1111 (not in Windows user folder) No Stability Matrix. ckpts during HiRes Fix. 5. Installation is complicated and annoying to setup, most people would have to watch YT tutorials just to get A1111 installed properly. Selected the Preprocessor and Model. "(x)": emphasis. 4 #109 opened Jun 29, Jan 28, 2023 · Not sure if this is the same thing you are having. for example. When hitting refresh, this message pops in cmd: CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. I have integrated the code into Automatic1111 img2img pipeline and the webUI now has Image CFG Scale for instruct-pix2pix models built into the img2img interface. 1) in A1111. With ComfyUI you just download the portable zip file, unzip it and get ComfyUI running instantly, even a kid can get ComfyUI installed. The TensorRT unet stuff recently released for Automatic1111 is pretty cool (not sure if it is out for ComfyUI yet?) Speeds up generation x2, I can make an SDXL image image in 6. Now ComfyUI doesn't work. We all know that prompt order matters - what you put at the beginning of a prompt is given more attention by the AI than what goes at the end. py", line 577, in fetch_value raise ScannerError(None, None, yaml. 10. For now all you have to do is: Step 1: make these changes to launch. You should check in NVidia settings, under 3D settings, and see if the relatively new option for "Prefer Sysmem Offload" is toggled on (or device default). Flexible: very configurable. Download the InstantID ControlNet model. Before running automatic1111 again with the extension installed, you need to open cmd as admin, cd to the stable-diffusion-webui folder, and run: git apply --ignore-whitespace extensions/sd-webui-latent-couple/0001 Nov 22, 2023 · the python process launched for comfyui is not properly terminated and remains running after exiting automatic1111 which causes issues and errors if trying to relaunch automatic1111 while the origi Jan 25, 2023 · DEPRECATED: This extension is no longer required. This post hopes to bridge the gap by providing the following bare-bone inpainting examples with detailed instructions in ComfyUI. And I got it working. 📌 Links: I will be getting an intel i5 12th generation, NVIDIA GTX 4060 8GB laptop, 16 GB RAM ( LPDDR4 ), for AI related work. You can create your own workflows but it’s not necessary since there are already so many good ComfyUI workflows out there. I wen't from an average 30 MB/s or less from the NAS, to an average of around 199MB/s from the NAS. ui bat file. So you can install it and run it and every other program on your hard disk will stay exactly the same. So I usually use AUTOMATIC1111 on my rendering machine (3060 12G, 16gig RAM, Win10) and decided to install ComfyUI to try SDXL. 5 won't be visible in the list: As soon as I load a 1. yaml there is now a Comfyui section to put im guessing models from another comfyui models folder. Renaming the *. What I'm showing here is an extension for Automatic1111's webui which aim to improve native ui/ux. 1 Create a virtual Python environment (I did not try a nix-shell or flake so far) 2. neither the open pose editor can generate a picture that works with the open pose Mar 20, 2024 · ComfyUI vs AUTOMATIC1111. to the corresponding Comfy folders, as discussed in ComfyUI manual installation. Put it in the newly created instantid folder. It should work now (except for sendto inpaint) as overrides. AUTOMATIC1111 is the de facto GUI for Stable Diffusion. Keep the denoising strength at 1. I asked in WAS-NS Github and author answered me "This is a problem with a YAML file loading. I was trying to split the SDXL and SD v1. Feb 25, 2023 · open pose doesn't work neither on automatic1111 nor comfyUI. Feb 23, 2023 · A friend of mine for example is doing this on a GTX 960 (what a madman) and he's experiencing up to 3 times the speed when doing inference in ComfyUI over Automatic's. 0 wasn't yet supported in A1111. 0_fp16 model from the Stable Diffusion Checkpoint dropdown menu. 5 takes 41 seconds with 20 steps. Does not work in webui 1. Seems like if you select a model that is based on SD 2. Now, if you don't have the relevant nodes installed and you are getting a missing node error, then there are two ways to install nodes:-. ` Feb 23, 2024 · 6. sh {your_arguments*} *For many AMD GPUs, you must add --precision full --no-half or --upcast-sampling arguments to avoid NaN errors or crashing. catboxanon edited this page Mar 4, 2024 · 21 revisions. 4 or 1. You should see something like this: This is rather hidden at the bottom of the screen by default. dist-info in C:\Automatic1111\webui\venv\Lib\site-packages are present. Star. Here's the guide to running SDXL with ComfyUI. Drag the image to be inpainted on to the Controlnet image panel. If I add any of it to negative prompt (ex. Cutting-edge workflows. Once we have the file ready, we edit it and if we have not modified the default configuration of Automatic1111 we only need to modify the following line: base_path: /{path to stable diffusion's folder}/stable ComfyUI uses the CPU for seeding, A1111 uses the GPU. The resources for inpainting workflow are scarce and riddled with errors. Put it in the folder ComfyUI > models > controlnet. Installing ComfyUI on Windows. The process is quite straight forward, you navigate to the Extensions tab in Automatic1111. I also just learned of something called TensorRT, that might give around another %50 speed boost, which I just moved my ComfyUI machine to my IoT VLAN 10. It works. Select OpenAI as your image generation backend. Lets you use two different positive prompts. Aug 5, 2023 · Was working before adding the file and path arguments. Assuming you have launched the WebUI by running webui-user. I've seen quite a few comments about people not being able to run stable diffusion XL 1 Nov 19, 2022 · ValueError: The following model_kwargs are not used by the model: ['encoder_hidden_states', 'encoder_attention_mask'] (note: typos in the generate arguments will also show up in this list. Yet, with the emergence of ComfyUI, users find themselves at a crossroad, questioning the better choice. bat. Choose the DALL·E model you wish to use. Instead of building a workflow from scratch, we’ll be using a pre-built workflow designed for running SDXL in ComfyUI. Aug 19, 2023 · The automatic1111 Face restore option that uses CodeFormer or GFPGAN is not present in ComfyUI, however, you’ll notice that it produces better faces anyway. Jul 30, 2023 · Who Says You Can't Run SDXL 1. I, on the other hand, are on a RTX 3090 TI and inference for me is 4 to 6 times slower than in Automatic's. 2 Source the environment 2. I want a checkbox that says "upscale" or whatever that I can turn on and off. ComfyUI vs Automatic1111 Feb 3, 2024 · This doesn't helps - ComfyUI still not working. Transparent: The data flow is in front of you. ComfyUI also uses xformers by default, which is non-deterministic. 5 I see a lot of mis-information about how various prompt features work, so I dug up the parser and wrote up notes from the code itself, to help reduce some confusion. yaml. Restarted WebUi. current buggy image generation process. Updating ComfyUI on Windows. I've adjusted the steps and CFP without success. I have read that ComfyUI overall provides more performance benefits especially in ram/vram usage compared to automatic1111. 0 base and refiner models with AUTOMATIC1111's Stable Diffusion WebUI. Step 4: Start ComfyUI. I want a slider for how many images I want in a ComfyUI erratic RAM (not VRAM) usage. 7. The program is tested to work on Python 3. (If you don’t see this option, you need to update your A1111. Method 1 (Manual): Working great for my use case. It is not relates to the standalone app ComfyUI. If it work for more than two days, its a fucking miracle just there. 2) and just gives weird results. But I haven't heard of anything like that currently. Use the same resolution for inpainting as for the original image. The tool is compatible with stability. Steps to reproduce the problem. A lot of newcomers to ComfyUI are coming from much simpler interfaces like AUTOMATIC1111, InvokeAI, or SD. 0-RC. Step 2: replace the . yaml", line 9, column 12. Step 2: Download the standalone version of ComfyUI. May 9, 2023 · I'm an Automatic1111 user but was attracted to ComfyUI because of it's node based approach. 5 for the moment) Jan 20, 2024 · Inpainting in ComfyUI has not been as easy and intuitive as in AUTOMATIC1111. AUTOMATIC1111. It has --listen and --port but since the move, Auto1111 works and Koyha works, but Comfy has been unreachable. I can do 1024x1024 on a 8g 2080s, but i had to set --medvram. Multiple aliases need to be comma separated as well, but encased in string quotes to not break the CSV parsing. Dec 31, 2023 · In ComfyUI they go in a directory names "embeddings", they are called by their name in the prompt and it works well. 2. I followed the directions for pointing ComfyUI to my A1111 checkpoint models, and it doesn't appear to locate them. Alternative to local installation. Jun 6, 2023 · Textural inversions not working #11064. I found I was having a driver issue with my 3080. Update pytorch lightning utilities ( #15310) Add Size as an XYZ Grid option ( #15354) Use HF_ENDPOINT variable for HuggingFace domain with default ( #15443) re-add update_file_entry ( #15446) This discussion was created from the release 1. I'm still able to generate images with AUTOMATIC1111 which resemble what I was getting with ComfyUI, so it does seem to be a problem with ComfyUI specifically. Nov 16, 2023 · Plan and track work Discussions. bin as I read somewhere does not work. Textural inversions not working. #11064. Some samplers also throw some erorrs, but again not an issue for me. jk hw sc ky xt kj xd nx ok ha