- 26
- 80 489
FoxtonAI
Приєднався 5 лют 2024
A journey of AI discovery & knowledge
Roop Unleashed Tutorial - Image + Video Face Swap AI
Roop Unleashed Tutorial for both IMAGE face swap and VIDEO face swap functions, to achieve the best results by using the Roop Unleashed best settings.
In this Roop Unleashed full tutorial, I’ll show you a number of example results at the end of each section, so you can see a real results comparison, with both types of face swap.
The great thing about using Roop Unleashed, compared to other face swapping tools, is that everything is really quick and easy to do.
So sit back and enjoy this Roop Unleashed tutorial, to learn how to image and video face swap, the easy and best way.
LINKS:
How to Install Roop Unleashed in 1-Click:
ua-cam.com/video/8vjR4Zvi0-8/v-deo.html
NVIDIA GPU CUDA List:
developer.nvidia.com/cuda-gpus
CHAPTER TIMESTAMPS:
00:00 Roop Unleashed Tutorial - How to Use Roop for Deepfakes
00:42 GPU or CPU Settings - How to Run Roop on GPU or CPU
02:03 Roop Unleashed Face Swap: Image to Image
06:31 Roop Unleashed Face Swap: Image to Video
COPYRIGHT NOTICE:
Copyright-free sample video clips courtesy of Pexels.
All other images, audio and video created and owned by FoxtonAI.
In this Roop Unleashed full tutorial, I’ll show you a number of example results at the end of each section, so you can see a real results comparison, with both types of face swap.
The great thing about using Roop Unleashed, compared to other face swapping tools, is that everything is really quick and easy to do.
So sit back and enjoy this Roop Unleashed tutorial, to learn how to image and video face swap, the easy and best way.
LINKS:
How to Install Roop Unleashed in 1-Click:
ua-cam.com/video/8vjR4Zvi0-8/v-deo.html
NVIDIA GPU CUDA List:
developer.nvidia.com/cuda-gpus
CHAPTER TIMESTAMPS:
00:00 Roop Unleashed Tutorial - How to Use Roop for Deepfakes
00:42 GPU or CPU Settings - How to Run Roop on GPU or CPU
02:03 Roop Unleashed Face Swap: Image to Image
06:31 Roop Unleashed Face Swap: Image to Video
COPYRIGHT NOTICE:
Copyright-free sample video clips courtesy of Pexels.
All other images, audio and video created and owned by FoxtonAI.
Переглядів: 359
Відео
Napkin AI Tutorial - Text to Professional Visuals Instantly and Free
Переглядів 73914 днів тому
Napkin AI tutorial and demo of all the key features, using a variety of examples, to get you up and running super quick. The Napkin AI tool is one of those actually useful AI tools. If you’re like me, once you experience it, you’ll end up using it frequently, for so many different everyday personal and work related tasks. Napkin AI takes any text that you throw at it, and produces a whole varie...
Best Flux Models Comparison Test - Image Quality vs Speed using Forge UI
Переглядів 1 тис.14 днів тому
Flux Forge UI Tutorial - Flux models comparison of 4 Schnell and 4 Dev Flux models, tested for generation times and image quality. If you’ve wondering about the actual, like-for-like image quality differences, and generation times, of the most popular Flux AI models, then this video is for you. In this video, I’ve selected the 8 best Flux models, 4 Schnell models and 4 Dev models, generated a w...
How to Install Roop Unleashed in 1-Click - Image + Video Face Swap AI
Переглядів 3,1 тис.21 день тому
Learn how to install Roop Unleashed on a Windows PC with the 1-Click installer. Roop Unleashed, is still one of the most popular face-swapping tools around. It’s completely free, can be installed locally on a PC, is frequently updated, and has an easy-to-use interface. It’s a solid choice, for UNCENSORED deepfakes without training, for both images and videos. It used to be an absolute time-cons...
How to Pick the Best Flux Models for YOU and Use in Forge UI
Переглядів 3,4 тис.Місяць тому
Flux Forge UI Tutorial covering my current pick of the best Flux models available, and the fundamental understanding and considerations, to choose and use the right model, for YOUR specific needs and hardware specs. The focus of this video, is aimed squarely at helping you understand the factors to consider, for choosing the best Flux models for YOU, so that you can save time, and find the swee...
Ultimate Fooocus Inpaint Tutorial - Step by Step Image Transformation
Переглядів 2,3 тис.Місяць тому
Detailed step-by-step Fooocus inpainting tutorial, covering all 4 inpaint methods, and how to produce the best results for each of them. Inpainting, is an incredibly powerful function, that allows you to modify or enhance an existing image. You can selectively edit specific areas of an image, and let Fooocus AI fill in the gaps in a realistic and coherent way. I’ll be walking through worked exa...
Install Stable Diffusion Forge UI - Quick & Easy 1-Click Method
Переглядів 2,6 тис.Місяць тому
The quickest and easiest way to install the Stable Diffusion WebUI Forge. The SD Forge UI is built on top of the Automatic1111 UI, and offers significant advantages, especially in terms of image generation speed and added functionality. For Automatic1111 users, the interfaces are incredibly similar, so switching to the Forge UI is seamless, with a very low learning curve. With the recent releas...
Install LivePortrait in 1-Click - Quick & Easy Method for Windows Web UI
Переглядів 6512 місяці тому
NEW - the super quick and easy way to install LivePortrait locally on a Windows PC, using the official one-click installer for Windows. LivePortrait has seen some rapid updates in the last few weeks, including video to video, humans mode, animals mode, a whole host of tweaks and improvements, and all importantly, the long awaited one-click installer for Windows, using the Gradio Web UI in your ...
8K Fooocus Upscale Comparison and Tutorial - All Methods Tested
Переглядів 6982 місяці тому
Comparison test of both Fooocus upscale image methods, to find out which method produces the best result. Fooocus has two main ways to upscale images. The “Fast 2x” upscale option, and the “Standard 2x” upscale option. Each of these options uses different back-end upscale processes, and each produces very different results. On top of this, there are some key settings that you can apply to each ...
Ultimate Fooocus Face Swap Comparison and Tutorial - All Methods Tested
Переглядів 9 тис.2 місяці тому
Comparison test of all 4 main Fooocus face swap methods, to finally find out which method produces the best results. If you’re using Stable Diffusion, then Fooocus is probably the easiest of all the UIs, to do high quality face swapping with. However, Fooocus offers 4 main ways to face swap, which is just too much choice for me. I tend to use Fooocus for certain tasks, as it let’s me just get i...
Add Custom Image Aspect Ratios and Resolutions in Fooocus
Переглядів 1,3 тис.2 місяці тому
The quick and easy way, to add ANY custom image aspect ratio and custom resolution to Fooocus AI. The most popular aspect ratio that’s missing, from the default install of Fooocus, is the 1920px by 1080px resolution, for a 16:9 widescreen aspect ratio. I’ll also run through, how to specify WHICH aspect ratio is auto selected as default, when you launch Fooocus AI. I’ll be running through everyt...
Install Fooocus in 2 Easy Steps - Latest Version + Dark Theme
Переглядів 1,1 тис.2 місяці тому
The quickest and easiest 2 step method, to install the latest version of Fooocus, plus enable a dark theme for it. Fooocus is still the most intuitive and straightforward UI, for generating high quality images with Stable Diffusion, quickly and effortlessly. So, if you’re looking for a no-nonsense Stable Diffusion UI, with the usual complexities taken care of for you, and still produce outstand...
8K Upscale with Ultimate SD Upscale in Stable Diffusion Automatic1111
Переглядів 2,9 тис.2 місяці тому
How to upscale a 512px image to 8k resolution using the Stable Diffusion Ultimate SD Upscale extension. The Ultimate extension is a newer and improved version of the SD Upscale extension that comes bundled as default with Automatic1111. Upscaling images with Ultimate SD Upscale is still one of the easiest and most popular ways to produce super high quality upscaled images. I’ll be walking throu...
LivePortrait: Ultimate Install Tutorial for Windows Web UI
Переглядів 4,8 тис.3 місяці тому
Complete, recommended, step-by-step Live Portrait installation tutorial for Windows Web UI (Gradio UI). This tutorial contains every step...an ultimate install guide that even non-technical beginners can follow along with. Once in a while, an AI Tool pops up, that delivers BEYOND expectations. The newly released LivePortrait AI Face Animator, is one of those tools. It ticks all the boxes - simp...
Fix Faces with ADetailer in Stable Diffusion Automatic1111
Переглядів 1,7 тис.3 місяці тому
Walk through of the steps to fix and improve faces using the Stable Diffusion ADetailer extension. You’ll need to make sure that you have ADetailer and the best face models installed. If you don’t, then take a look at my ADetailer Install video. I’ll leave a link to that video in the description below for you. I’ll be walking through everything on my Windows 11 PC, using a standard local instal...
Install ADetailer in Stable Diffusion Automatic1111
Переглядів 1,1 тис.3 місяці тому
Install ADetailer in Stable Diffusion Automatic1111
Merge Checkpoints in Stable Diffusion Automatic1111
Переглядів 6 тис.6 місяців тому
Merge Checkpoints in Stable Diffusion Automatic1111
Install LoRA Models in Stable Diffusion Automatic1111
Переглядів 2,6 тис.7 місяців тому
Install LoRA Models in Stable Diffusion Automatic1111
Install Checkpoint Models in Stable Diffusion Automatic1111
Переглядів 4,2 тис.7 місяців тому
Install Checkpoint Models in Stable Diffusion Automatic1111
Enable Xformers in Stable Diffusion Automatic1111
Переглядів 6 тис.7 місяців тому
Enable Xformers in Stable Diffusion Automatic1111
How to Install ControlNet 1.1 in Stable Diffusion Automatic1111
Переглядів 3,2 тис.7 місяців тому
How to Install ControlNet 1.1 in Stable Diffusion Automatic1111
Enable Dark Mode in Stable Diffusion Automatic1111
Переглядів 9707 місяців тому
Enable Dark Mode in Stable Diffusion Automatic1111
How to Uninstall Stable Diffusion Automatic1111
Переглядів 8 тис.7 місяців тому
How to Uninstall Stable Diffusion Automatic1111
How to Update Automatic1111 Stable Diffusion
Переглядів 4,5 тис.7 місяців тому
How to Update Automatic1111 Stable Diffusion
How to Install Stable Diffusion Automatic1111
Переглядів 5 тис.7 місяців тому
How to Install Stable Diffusion Automatic1111
How to Install Stable Diffusion Web UI Automatic1111 in 1 Click
Переглядів 4,4 тис.7 місяців тому
How to Install Stable Diffusion Web UI Automatic1111 in 1 Click
How to reduce the head size of the swap head?
Man, it takes ages to process without GPU. 😑 I'm broke AF.
I feel you! All the AI tools really need an NVIDIA GPU for ease and performance to run locally.
Thank you master. Is it possible to convert from two different sources to two different targets separately? video, i mean.
Never tried that...but doing them one step at a time would achieve the same result...if I understand your question correctly.
Hi thank you for making an excellent tutorial, first time to actually use these kinds of tools. Installed smoothly and made image to image samples flawlessly, my question is, is this censored or uncensored? Hopefully also you could show us other alternatives apart from roop that you think is better or same. Thank you!!
Glad you liked the video. Roop is uncensored. I'm currently testing all the deepfake tools for a side by side comparison video.
please show us how to use it with obs studio for video calling
Thanks for your suggestion. I'll try to take a look at doing that.
Slider'ı sikeyim ya. Yukarıda formül var oradan açıklasana hacı! Yarım saat sağa kaydırınca ne olur, sola kaydırınca ne olur. Bla bla bla...
Legend, thank you. Super clear and to the point!
Thanks for your comment. Glad you liked the video.
Hey, Great video! Apprecite your videos so much, my knowledge on stable-diffusion as a whole has expanded so much! Could you please make a video showing how to use Roop? Thanks! Keep up the great work 👍
Thank you for your kind words. Glad the videos have helped you. It's a good suggestion. Let me put together a Roop Unleashed tutorial video and get up in the next couple of days.
Here you go...Roop Unleashed Tutorial video just uploaded: ua-cam.com/video/WaAf-ylSFD4/v-deo.html Hope it helps.
Is there some tutorial how to install these checkpoints, what does it mean, how to train..w/e is a training? I get feeling that with each tutorial video I found I know less. I understand where to put vae, lora, but for checkpoints I don´t know where to save them.
Hello. Yes, it can all be a bit confusing at first. I have a number of starter videos for Automatic1111 on my channel. I think you'll find them very helpful and make things a lot clearer for you. This is a video I did specifically to answer your question on Checkpoints: Install Checkpoint Models in Stable Diffusion Automatic1111: ua-cam.com/video/pO1ztt9AHo0/v-deo.html Hope that helps.
I couldn't run the original Flux.1 Dev which is 22GB+ with 16GB of RAM (24GB VRAM, RTX A5000), I had to up to 32GB RAM...
That's strange. I run it fine (although it takes a while) on my 8GB GPU, 16GB RAM and default virtual memory allocation.
Can we do these steps on the phone or iPad as well?
Unfortunately, at the moment, you can only create or edit a napkin on your PC or laptop, but you can view your napkins on all devices.
@@FoxtonAI 😔
Thanks a lot!
You're welcome. Glad you liked the video.
Hello! Do you know if there is any way to cancel updates for stable diffusion Automatic 1111 at startup? I'm tired of constantly waiting for the program to run with the risk that one day my models will stop working
Hello! If you have Automatic1111 set to update automatically on launch, then you can just edit your "webui-user.bat" file in Notepad, remove the "git pull" syntax, and then save the file. You basically need to do the opposite of this previous video of mine: ua-cam.com/video/sGbK3M-giqE/v-deo.html Hope that helps.
@@FoxtonAI I LOVE YOU!!! THANK YOU!)))
Sorry for the late question, I just found your excellent tutorial. According to your hope, if image quality is the most important aspect, with a 24GB RTX3090 video card, 64GB RAM, Win 11, what configuration is best? For example, can I find a better one? flux1-dev-fp8; 30 steps; ae.safetensors; clip_l.safetensors; t5xxl_fp16.safetensors
Hello. If I was lucky enough to have your hardware spec, and I wanted the very best quality images, and not bothered about generation time, then I would use: Dev FP16, 24+ steps, ae, clip_l, t5xxl_fp16. Changing the Dev FP16 with the Dev FP8 would give very similar results, but a lot faster. You may be interested in my follow up Flux video, which compares different Flux model image results (FP8, GUFF Q8, NF4 and the GUFF Q4): ua-cam.com/video/lEa2PzSdoxs/v-deo.html Hope that helps.
@@FoxtonAI Thank you very much!
Thx for this....why after restarting my computer, the swap process always fails with a notification : Error in execution: Non-zero status code returned while running Gemm node. Name:'Gemm_568' Status Message: CUDA error cudaErrorNoKernelImageForDevice:no kernel image is available for execution on the device.
Unfortunately, I have never seen an error like that before.
thanks for sorting that folder path thing out I would've spent forever trying to fix that
Glad that helped you...the good news is that my fix has now been implemented in the updated version of the installer...they must have watched my video!
Hi, could you make a video tutorial where you explain how to create checkpoints from scratch? I would like to create one of my own :)
Thanks for that suggestion. I may take a look at doing that.
@@FoxtonAI thx a lot :)
Hey, actually, when I open up SD webUI using Stability Matrix, it just shows an error saying connection errored out. Could you help me with it please?
That's strange. Did you manage to get it sorted?
@@FoxtonAI Nope, could you help me?
Conda environment creation failed! What should I do to correct it? I have tried reinstalling it too but it doesn’t work
Has a conda folder been created at: roop-unleashed\installer_files\conda
@FoxtonAI yes it is there. I am also facing this issue
You may want to check that your conda PATH is set correctly. In the search box on your taskbar, type in "edit the system environment variables", then open it. Click the "Environment Variables" button. In the top box, select "Path", then click the "Edit" button. You should see a conda PATH in there at "roop-unleashed\conda\condabin". If you can't see this, or it is different, then edit or add your correct full PATH accordingly. Sometimes, if you've tried installing Roop Unleashed with several methods, different PATHS get setup multiple times and your latest install reads the wrong one. So, if you're doing a new install using my 1-click method, then first delete all the PATHS for Roop Unleashed in the Environment Variables, then delete the whole roop-unleashed folder on your hard drive, then reinstall Roop Unleashed using my video. This will then give you a fresh, clean install, and set all PATHS, including conda PATH correctly. Hope that helps.
How to download this file?
Which file do you want to download?
@@FoxtonAI the upscale extention
@@ragulp7160 You can find the link in the video description: github.com/Coyote-A/ultimate-upscale-for-automatic1111
are there any hardware requirements? I do everything you do and it doesn't open for me
What hardware do you have?
brother i need some help i installed "roop-unleashed" on my own pc... now tell me how can i access "roop-unleashed" on my phone, when i run it on my pc..... please help
Not sure, I just use it on my PC.
You can probably access using the ip and port it shows in your PC browser but if you're running on PC and it doesn't work try using google remote desktop to access your PC in your phone.
Hello. I’m having trouble finding some of the Python files in my files app. I don’t have anywhere where I can access the PIP packages or cache, is it already gone? I don’t have appdata, local, or windowsapps in my files app. Any help will be greatly appreciated!
If they're not there, then you don't need to worry about them.
This helps so much! love the tutorial^^
Thank you very much. Glad you liked the video.
Great video my friend. keep going
Thank you. Glad you liked the video.
Its saids (Miniconda not found) please help me
Hello. That's strange...everything should auto-install. Try deleting it and reinstalling, in case anything was corrupted during download/install.
Does the following happen to anyone? : - That after about thirty minutes and at 100% of the work, the video only lasts two seconds ? - That when starting the Face swap in the CMD, from 1% of the work to 100% of the work, it is done in a few seconds and there is no final result ?? Thank you!!
Never had that before.
Im still confused what is the difference between checkpoint models and loras and all the other different options!
It can all be a bit confusing at first. Checkpoints are the big models that make images on their own. LoRAs are "mini models" that plug into a checkpoint and change their outputs. They let checkpoints make styles, characters and concepts that the base checkpoint they're used on doesn't know or didn't know very well.
@@FoxtonAI Thank you for explaining! :)
Is there a version of roop unleashed for macOS and where can I find the Link from also is it easy to install it. Thanks. John E
Hello John. Take a look at the GitHub page, at the install section. I think there are instructions there for Mac.
@@FoxtonAI Thank you for your fast response to my message, Unfortunatelly I know about Github as everything on their homepage is always so difficult to follow. Thanks anyway.
i tried both versions of Forge the newest and the one that you recommend. both say are not supported by my device....I have windows 11 and the new surface pro. Any advice? I am desperate? I also cannot install comfy ui
Unfortunately, it's difficult to nail down the issue without hands-on specifics. I do know that previous Surface Pro machines had an Intel SSD installed, which were notorious for issues like this. If you have another drive, you may want to try installing on that, to see if this resolves your issue.
@@FoxtonAIWell nope...apparently it's my graphics card ...it simply sucks...and there won't be anything I can do. I do not have an Nvidia card. Any other UI I can install flux on with my crappy laptop?
That's a shame. You're probably best off using an online site to use Flux.
It doesn't open automatically for me when everything is installed. How can I open it??
Double-click the "windows_run.bat" file. See video section: 04:37 How to Open Roop Unleashed After Installation.
@@FoxtonAIThanks a lot!!! I apologize for my clumsiness. :_(
is it possible to change the position of the model?
If I understand your question correctly...not with inpainting. You would need to change the pose using ControlNet.
@@FoxtonAI I solved the problem. Thank you :)
how to use source Video instead of image ???
Next to the "Source Image" tab, is a "Source Video" tab. Make sure you have the latest version of LivePortrait installed. Older versions do not have the source video option.
@@FoxtonAI Thank you Sir.
This is all so unintuitive, specifically because it basically has limitless possibilities. I realize there are many different checkpoints that are all great to use. They are all huge files. 1.99gb to 6.46gb for just ONE checkpoint. if you download them, you end up with over 100gb (dreamshaper, aniverse, SD 1.5, etc.) of checkpoints installed on your SD webui... my question is: is it more feasible to just browse civitai, and just focus on one project, then search for checkpoints, loras, and embeddings, etc., download whichever you fancy, then install and experiment once loaded on the webui, then delete the checkpoint when finished (saving storage space)? it just doesn't seem feasible to save each and every checkpoint unless you have the storage space adequate for such a program.
Hello. Many people have excess space on their drives, so they don't mind downloading numerous checkpoints. But if space is limited, what you suggest is a good idea, to only download those you need for a project and then delete any you no longer use.
Hello master, I would like a video about audio generation, voice cloning and voice modification.
Hello. I do a lot with audio, so that's maybe something I'll do. Thanks for your suggestion.
miniconda hook not found plss help me
Hello. That's strange...everything should auto-install. Try deleting it and reinstalling, in case anything was corrupted during download/install.
@@FoxtonAI after reinstall conda environment creation failed
@@mr.masktech3671 browse through the folders and find miniconda. Install it manually.
Hey, master. I’ve watched all of your videos, but there’s still something I don’t fully understand. I’m using two 3080 Ti graphics cards, and I was wondering if you could make a video showing how to install roop-unleashed using GPUs? I would really appreciate it!
Glad you liked the videos. For Roop Unleashed, go into the Settings, and then make sure "cuda" is selected to use your GPU. Thanks for your suggestion...I think I will.
I really like this Stable Matrix. Thanks for the tutorial.
Thank you. Glad you like it and found the video helpful.
Here's something. I tried installing this on a new installation of Windows 10. Before doing it, you have to install some version of Microsoft C++ distributable. Your one click method with Stability Matrix, checks for it and downloads it. The batch file in this method does not check for it and it will create an error. Now most people who have played any type of video game or used any other graphics program will probably already have those C++ files. But in my case, on a new install of the OS, they won't be there and will create an error.
Thanks for your useful clarification.
Another great tute! VERY well done. When working on percentages ( 50/50 to 60/40, etc ), would one be able to edit/modify the new checkpoint or must a new one be created?
Thank you. Glad you liked the video. Yes, you can then merge your new checkpoint.
@@FoxtonAI Apologies, I meant in reference to 4:16. What is I wanted to change it to be less or more. Would I have to create a new merge or can I just use the one I had created and edit the percentages and save again,. Thank you!
You can just adjust the percentages, give it a new custom name and hit the Merge button.
@@FoxtonAI Thank you so much! :)
HI. I like the way you teach, you;'e quite precise and you give it thought. Thank you. Curious, I have Forge UI for Flux and have downloaded many models for it. New to Fooocus and I intend to install it locally. I see that it also needs models. Would Fooocus need to have a separate folder for its models or is there a way to share them between the two. I'm trying to watch my Hard Drive Space. Cheers.
Thanks for your kind words. Glad the video was helpful. Yes you can share the models. The easiest way is to download the checkpoint models to your Forge UI folder (Forge\webui\models\Stable-diffusion). Then in Fooocus, open in Notepad, the config.txt file (Fooocus\Fooocus\config.txt). At the top of that file, just change the "path_checkpoints" location to your Forge UI folder. Hope that helps. These Fooocus videos may be of help to you: Fooocus: Install in 2 Steps - Latest Version + Dark Theme: ua-cam.com/video/0oOUwVrKAC0/v-deo.html Add Custom Image Aspect Ratios and Resolutions in Fooocus: ua-cam.com/video/4HzUrBx0c44/v-deo.html
@@FoxtonAI You are awesome! Really good tutes. They are very relevant and eclectic. Thank you! Subscribed for certain.
Just use stability matrix..
thank you
You're welcome. Glad you liked the video.
Can the model and text encoder mix match? ex Dev Guff Q8 + T5 fp16 default vice versa. Also, I was told to use Dev guff or FP8 when using loras even though I have a 24gb vram. To confirm that is what you were explaining in this section: 5:56 -7:00
Hello. Yes, you can choose to use the T5 FP8 or T5 FP16 text encoder with models. Personally, I often use the Dev FP8 model with the T5 FP16. LoRAs have problems with the NF4 models, so using GUFF or FP8 with LoRAs was the correct advice given to you. Hope that clarifies.
@@FoxtonAI Thank you. Not sure why but my FP8 is creating black image on comfy and all files are in respective folders. My FP16 dev works.
Strange about the FP8 black images...maybe it's a ComfyUI technicality somewhere. Glad the FP16 is now working well.
Does forge work well with pony models?
Should not be a problem.
Wait…. Where In the first example does the meta data get copied from, also - will the button auto change to load parameters? Never seen this before
The original image was generated previously. If you go to your output images folder, there will be a log.html file in there. Open that, and you'll see all of the images that you generated for that day, together with the metadata and a copy button. Copy and paste that into the text prompt box, and the button will change smartly to "Load Parameters", press it to populate the parameters. By the way...you can also save metadata directly to the image, by selecting that option in your settings. Hope that helps.
@@FoxtonAI ahh thanks, makes sense , I was running on colab so need to save that prior, by opening it before i quit session.
@@FoxtonAI This is something I had no idea... Looking at my folders now with all that metadata. Perhaps the best piece of knowledge thus far
no module 'xformers'. Processing without... no module 'xformers'. Processing without... No module 'xformers'. Proceeding without it. This is the message just under the message loading modules :(
You need to make sure you have done a standard install of A1111 using a recent version. That would come with Xformers included, then you would just use the method in this video to actually enable it. Hope that helps.
Maybe its because I use ComfyUI, but my Dev FP8 with everything only peaks around 11GB vram when running. By your math it should require way more. I was even able to generate fairly quickly on a 8GB card too.
ComfyUI and Forge handle things slightly differently. Your numbers with the Dev FP8 sound about right, since it's an 11GB model-only size. I presume the text encoders are loaded and processed before the model. My times with Dev FP8 are generally 1-3 mins, depending on image dimensions. In the video, the offload to RAM and longer generation times came with using the Dev FP16 model. Thanks for confirming your numbers in ComfyUI.
working
Glad the video was helpful.
Thanks!
Glad you liked the video.
How to fix This script relies on Miniconda which can not be silently installed under a path with spaces.
Hello. That's correct, see video section "Extract Roop Unleashed Installer File", or take a look at this exact question and solution on the Roop Unleashed GitHub page: github.com/C0untFloyd/roop-unleashed/issues/227 Hope that helps.
You need to remove all the spaces in the names of your install folder and above folders for it to work.