Відео

I feel you! All the AI tools really need an NVIDIA GPU for ease and performance to run locally.

Never tried that...but doing them one step at a time would achieve the same result...if I understand your question correctly.

Glad you liked the video. Roop is uncensored. I'm currently testing all the deepfake tools for a side by side comparison video.

Thanks for your suggestion. I'll try to take a look at doing that.

Thanks for your comment. Glad you liked the video.

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.

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.

That's strange. I run it fine (although it takes a while) on my 8GB GPU, 16GB RAM and default virtual memory allocation.

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 😔

You're welcome. Glad you liked the video.

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!)))

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!

Unfortunately, I have never seen an error like that before.

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!

Thanks for that suggestion. I may take a look at doing that.

@@FoxtonAI thx a lot :)

That's strange. Did you manage to get it sorted?

@@FoxtonAI Nope, could you help me?

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.

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

What hardware do you have?

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.

If they're not there, then you don't need to worry about them.

Thank you very much. Glad you liked the video.

Thank you. Glad you liked the video.

Hello. That's strange...everything should auto-install. Try deleting it and reinstalling, in case anything was corrupted during download/install.

Never had that before.

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! :)

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.

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.

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. :_(

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 :)

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.

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. I do a lot with audio, so that's maybe something I'll do. Thanks for your suggestion.

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.

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.

Thank you. Glad you like it and found the video helpful.

Thanks for your useful clarification.

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! :)

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..

You're welcome. Glad you liked the video.

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.

Should not be a problem.

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

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.

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.

Glad the video was helpful.

Glad you liked the video.

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.