00:04 Introduction to Flux and its creators 02:24 Robin Romach's significant contributions to Flux development 07:19 Using Comfy org or Kaji fp8 lower threshold versions for resource efficiency and convenience 09:46 Managing model loading and workflow rearrangement in Flux 15:25 Schell model for fast image generation with resource requirements 17:44 Flux guidance node for image generation 22:45 Adjusting parameter values in Flux can affect image focus. 25:22 Flux has advanced text and layout generation capabilities 30:16 Flux image processing allows for easy style transfer and image enhancement 32:42 Flux offers high image quality and strong prompt understanding.
Excellent video, I think that, together with Matteo's video, are the two best videos about Flux on YT. I look forward to more Flux videos. I send you a big hug from Argentina!
Thank you for this nice and clear tutorial ! But I think two links are missing in the descriptionfor those two files that you show at 6:57. -The first one for the t5xxl_fp16.safetensors: -The second one for the t5xxl_fp8_e4m3fn.safetensors: And I will have a question . At 9:16 which file do you drag into the ComfyUI interface to have this default setting? In advance, thanks for the support :)
Those two files are CLIP models, and they should be placed in the CLIP folder, which is the folder shown in the top-right corner of the screen. If you're asking how to download them, I mentioned the download link at 6:37 in the video, and you can find it here: huggingface.co/comfyanonymous/flux_text_encoders/tree/main. At 9:16, you just need to drag the image or JSON file with the workflow into ComfyUI. Hope this helps!
The one I had before was ComfyUI's betaui, but ComfyUI is all about this new UI after the 15 August update, and you can select its location via the settings menu
I'm using the latest version of ComfyUI. Since August 15th, you can customize the ComfyUI menu layout. You can choose to place it at the bottom or top in the settings.
00:04 Introduction to Flux and its creators
02:24 Robin Romach's significant contributions to Flux development
07:19 Using Comfy org or Kaji fp8 lower threshold versions for resource efficiency and convenience
09:46 Managing model loading and workflow rearrangement in Flux
15:25 Schell model for fast image generation with resource requirements
17:44 Flux guidance node for image generation
22:45 Adjusting parameter values in Flux can affect image focus.
25:22 Flux has advanced text and layout generation capabilities
30:16 Flux image processing allows for easy style transfer and image enhancement
32:42 Flux offers high image quality and strong prompt understanding.
good job~
What an incredible video to get started! Should be on the top of every Flux learning playlist. Thanks!
wow~~Thank you so much~
Excellent summation of something that has changed a lot over the last few weeks. Super clear and informative - thank you!
wow~thanks for liking
| one of the most excelellent videos about to start incl. deep dive into comfy and flux |
Excellent video, I think that, together with Matteo's video, are the two best videos about Flux on YT. I look forward to more Flux videos. I send you a big hug from Argentina!
Thanks for the like, Matteo is a very respectable developer and I have benefited a lot from his code and sharing!
Can you please share the video you mentioned about "Matteo's video?" Thank you.
This video like was listening a prophet and gather all knowledge once... thank you!
wow~ thanks for liking it!
@@SweetValberryCould you please help me to generate a concept if i pay you?
@@Roman-GreenBoulevard Thanks for asking! I just do this in my spare time, so my time and energy are limited.
@@SweetValberry thats fine im not in hurry pls leave me a mail where i can message you
Fantastic Flux User Manual.🙋🏼♂
😁😁
amazing, thanks for this great tutorial!!! and congrats for all the knowledge you have achieved
Thank you so much! I’m really glad you enjoyed the tutorial.
don't forget guys recently they have released different quantization, even for cards as low as 4GB so make sure to check those out
Yes, I talk about it in this video, you can check it out.th-cam.com/video/TAMTQubMhxw/w-d-xo.html
Thank you for this really well explained introduction to FLUX.
Thanks! Glad it was helpful.
Really awesome tutorial 👏👏👏 thanks a lot
😄😄
Thank you for this nice and clear tutorial !
But I think two links are missing in the descriptionfor those two files that you show at 6:57.
-The first one for the t5xxl_fp16.safetensors:
-The second one for the t5xxl_fp8_e4m3fn.safetensors:
And I will have a question . At 9:16 which file do you drag into the ComfyUI interface to have this default setting?
In advance, thanks for the support :)
Those two files are CLIP models, and they should be placed in the CLIP folder, which is the folder shown in the top-right corner of the screen. If you're asking how to download them, I mentioned the download link at 6:37 in the video, and you can find it here: huggingface.co/comfyanonymous/flux_text_encoders/tree/main.
At 9:16, you just need to drag the image or JSON file with the workflow into ComfyUI.
Hope this helps!
@@SweetValberry Thank you !! :)
Great video! I don't have the "FaceDetailer" node, how can I do to?
"FaceDetailer" belongs to the "ComfyUI-Impact-Pack" custom node, you can download it here: github.com/ltdrdata/ComfyUI-Impact-Pack
Thanks, how did you make your ComfyUI Menu go to the bottom taskbar?
The one I had before was ComfyUI's betaui, but ComfyUI is all about this new UI after the 15 August update, and you can select its location via the settings menu
Thanks for this, top job! ♥♥♥
Thanks for liking it~
Can you use Reactor with flux comfyui? I tried but nothing seems to work as it did with 1.5. Thanks
Yes, there is no problem with the ReActor, are you encountering any errors when using it?
I don't see the comfy combined models anywhere? Please advise. thanks
Do you mean the ComfyOrg version?It can be downloaded here: huggingface.co/Comfy-Org/flux1-dev/tree/main
which comfy UI is that? is there an addon that puts the UX in the bottom?
I'm using the latest version of ComfyUI. Since August 15th, you can customize the ComfyUI menu layout. You can choose to place it at the bottom or top in the settings.
@@SweetValberry that was it, thank you.
Great communicator, look forward to the update in a few weeks. Things are moving fast, well, except my GPU.
Ksampler Efficient mode is missing from my version. Everything updated :/
Do you mean the node ‘KSampler(Efficient)’? It belongs to this custom node: github.com/jags111/efficiency-nodes-comfyui
@@SweetValberry Yes, these ones, thank you!
Good stuff here!
😄
Will you ever be able to get Flux.V1 Pro on ComfyUI? Or is dev the best it will always be?
You can only use API with API token to use the Flux Pro, for me, I think dev is the best model we can use locally.
Great job !!!
😄
can i use a 3090 instead of a 4090?
If you can use 4090, it's obviously a better choice.
Bravo 👌
😄😄
why does mine crash every time i try a generation with my 4090?
Are you talking about ComfyUI crashing or PC crashing? Could it be a memory issue?
@@SweetValberry Comfy crashes. i have 32GB of Ram. I just installed Nobora on another ssd ill try that soon its probably stupid windows
what a hazzle, I use Forge in Pinokio :-D
This guy cooks 🎉
😃
Brillant
😄
很利害
👏👏👏👏👏