Відео

Hello, yes, you will put any keywords in the top box. However, the workflow to use Flux model changes quite rapidly. If you are using the fp8 simplified version, then there will be only one clip text encode (normal default one, same as sd1.5, sdxl one). For vram, it defaults to normal vram mode.

Yes, requires quite a bit of trial and error though. Will be better to get lora specific to the art style.

@CodeCraftersCorner I managed to create. The quality is better than xl

Thank you! Glad it was helpful!

Thank you! For me personally, i am using these: sampler: euler, scheduler: sgm_uniform, steps: 20 for (finding the composition), 40 (final image), cfg: 4.5 - 5.5 to make the model follow the prompt.

Thanks for the feedback! For me personally, i am using these: sampler: euler, scheduler: sgm_uniform, steps: 20 for (finding the composition), 40 (final image), cfg: 4.5 - 5.5 to make the model follow the prompt.

Thank you! Glad to help!

Thanks for sharing. I do the same. for steps, I do 20 and check the composition and then 40 for the final output. CFG: from 4.5 - 5.5 depending on the complexity of the prompt.

Hello is a diffusion based model. So we can expect same improvements as with sd1.5, sdxl models.

Hello, I'll try. The sampler dictates the noisy image will change to a detail one. The model sampling shift how this happens.

I, too, share the same opinion. It took me nearly 6 hours (with breaks in between) to generate all these images, and it’s just not practical to keep generating so many when adjusting the prompt. Thanks for sharing your thoughts!

@@CodeCraftersCorner btw thanks for trying these out, I personally simply try to prompt things that didn't or did work with earlier models while swapping samplers and schedulers, so it is nice to see some sort of structured overview too.

Thanks you! That's was fast!

Thank you for sharing!

Glad it was helpful!

Thank you very much! I appreciate that!

Hello, usually the error message should be in the terminal (CMD). It will tell you if there is anything missing.

@@CodeCraftersCorner python.exe has been quit the cmd only has `Using split attention in VAE` no error code left

Not for now. Pyramid Flow uses its own model.

@@CodeCraftersCorner I'm getting added token errors. Also had to change the name of the config.json file because it couldn't find it until I changed the name. I'm assuming I'll get more errors but I'm stuck at the added tokens error

Thank you! No, not yet but you will find more videos on the channel about Stable Diffusion and ComfyUI.

Thanks! Yes, I am doing the large version but it's taking longer

Hello! Yes, I am doing the large version but it's taking longer. I will post the results once completed.

Thanks for sharing! Will try it.

Thank you too

Try this settings: Model Sampling SD3 shift on 2.00, steps on 4, cfg on 1,sampler - euler, scheduler-beta. Time for generation with amazing results about 15 seconds.

@@pawelthe1606 thanks and unfortunately now i'm able to use cpu interference , os either turbo sd35 or waiting for a similar lightning 4step lora , then i stuck at cfg 1 , so i ask about what best scheduler and sampler for realistic output

Yes, I planned to experiment with realistic pictures too.

No, please do it in the same video 😜

Thanks a lot!

Thank you very much! Thanks for the support!

Hello, yes. You can use the images from the GitHub repository (link in description). You can drag and drop the image onto ComfyUI to load the workflow.

Thanks for sharing.

Sure, you can. It's just another option.

The attached UI isn't necessary, as the native ComfyUI is designed for creativity and experimentation. While it might seem challenging at first, everyone starts somewhere and gets more comfortable over time.

Looks so...is the new SD model suitable for T-shirt graphics?

Will do a comparison!

Yes!

@@CodeCraftersCorner I can send you info as soon are ready

Thanks for sharing!

Thanks! Good point! Will check.

Thank you!

Thank for sharing! I will try.

@@CodeCraftersCorner I mean... it is there for some reason for sure. I just don't know why...

no way, you're even explaining that to. damn, talk about comprehensive. but cant install manager through it? u didnt mention.

i had to come back because Pinokia looks like its going to install comfyui manager automatically yay. if so Pinokio seems like the best tool ever for non coders or people who dont want to follow many links and go through a million steps, however, i have hit a snag with my graphics cards last and latest driver still being to old. it gave me the link to get the pytorch script to install referencing something called cuda. it didnt work with the basic install either. i guess pinokio automatically installs the gpu version :O damn, i might have to do it the first way you explained after all, i guess at least it would have installed all dependancies so it should be quicker now

Yes, pinokio is one click installer but if the maintainers do not update the installation file, there's not much we can do. It's good to know how to install manually in that case.

i love your arrows and boxes to direct attention, may i ask how u do that incase i one day make tutorial videos for helpful things like you

Thank you!

Oh, I used a windows utility tool called zoomit. It's free, you can get it from microsoft website.

@@CodeCraftersCorner wow thanks. that will be so handy. thanks for the timely response

Great! This is a custom node made by an independent developer. I do not think it will become part of ComfyUI Core just yet. So, your knowledge of ComfyUI will still be useful.

Sorry to hear that!

Does Forge have workflows you can search through like Flow? I like different 1 click workflows without dealing with nodes if possible. If Forge has that, then perfect.

@@StrikerTVFang I'm not kicking this video at all, it just looks like Forge to me.

Thanks for sharing your thoughts! Yes, Forge has its own advantage while this is still ComfyUI.

@@GrantLylick huh? No I was genuinely asking a question because I was thinking about using Forge lol

@@StrikerTVFang I have both installed. Forge is pretty good.

Thanks for sharing your thoughts! Yes, for people who like to be creative and tinker, the nodes are better. Yes, the queue panel can be collapsed. I actually made a video on the UI explaining the different elements. It can be found on the channel called: October 2024 ComfyUI updates.

Thank you!

Maybe try to update comfy and see what happen. Make sure you got proper diffusers for large model also

Try to update you ComfyUI through the terminal again. This implementation is part of the ComfyUI Core, so if you have enough memory, it should work. Also, you can try to redownload the checkpoints in case one got corrupted.