You have to add hugging face cli in the project. Then you have to login into hugging face using cli and you hugging face token. Then you should be able to download the file from hugging face
I showed how to use lora for flux in this video: ruclips.net/video/cvcrv21AHO8/видео.html You have to download and keep the file in lora folder. Also check if there are unwanted character in lora's name.
Personnaly i'll stay on foocus, this fork is cool but too much demanding on ressources, and frankly, not really giving THAT MUCH better results. Need lot of optimisations for fast generations. An d more loras, etc.
@@Aininja94 Thanks for answering brother, I appreciate your videos! but it hurts me that it doesn't work because ruined doesn't work on amd either, apparently
I am kind of working on an application myself. I am not an expert like this guys. But i am also looking for workaround for flux in that. Let's hope my channel gives me some income so i can go full time into project.
@@Aininja94 Do you have any discord to share information? I'm also learning, but I'm a little stuck when it comes to generating a profile face of the same person starting from a frontal face, if you can make a video about that it would be great!
I am a introvert person first google and second RUclips is the only thing I started using. I don't understand discord because never tried. But if you are looking for similar face generation. Then I have made 3 videos for it. It does the job for consistent face. ruclips.net/p/PLPFN04Wspxqvx68xxXlpFpgmyKy91ig1H
i get this error : from .version import InvalidVersion, Version, _TrimmedRelease ImportError: cannot import name '_TrimmedRelease' from 'packaging.version' (/usr/local/lib/python3.10/dist-packages/packaging/version.py) Get LoRA keywords for lcm-lora-sdxl.safetensors (SDXL 1.0 - LORA) what im doing wrong ?
Google colab link to start:
colab.research.google.com/drive/1vvXTwDUJoIT4trsiJtdBrPKOjjFu0X74?usp=sharing
Hi buddy, can you make for us a video about creating and training a flux lora model but on google colab please?
I am 25% done making video on flux lora but on personal computer. Making it on colab, I think will have to be a seperate video.
@@Aininja94 Alr, we'll wait for ya buddy, I liked your explanations!
Nice job. Thank you very much !
thanks bro
Hope you cleared the hugging face download issue, I thought I should be explaining in a short video
how to load flux model on huggingface, when i load model on colab it shows error like this: "Username/Password Authentication Failed"
You have to add hugging face cli in the project. Then you have to login into hugging face using cli and you hugging face token.
Then you should be able to download the file from hugging face
how can i find the seed in ruined fooocus , no seed is generated after the photo is getting generated
In the right section, click the info tab. You will see the image info of the current generated image. It has the seed id
does this method work for SD 1.5 models?
Are you trying to run SD 1.5? I don't think it will work in RFooocus. But a1111 Or Forge UI will work on colab
@Aininja94 do you have a video explaining the process?
I made this video on forge UI where I just tested SD1. 5, SDXL & FLUX ruclips.net/video/awE2C2R9u6M/видео.html
@@Aininja94 thanks dude
It would have been better if you had given the link to the notebook bro😇
New to all this things. Right now making my first git commit 🥲
@Aininja94 ok 👍 good luck🤞
How do I take two photos and make people appear together talking and hugging with the same face? In this case, there will be two different faces.
I made a video on combine character using AI, that should work
use Kling Ai.
Bro Please share I want this notebook please I can't do it it's too long please share thank you in advance 😊😊 🙏🙏🙏
Will add a file for colab and let you know
Click this link:
colab.research.google.com/drive/1vvXTwDUJoIT4trsiJtdBrPKOjjFu0X74?usp=sharing
@Aininja94 thank you bro 🙏😊
Show us the actress model lora to put in flux preset
I showed how to use lora for flux in this video: ruclips.net/video/cvcrv21AHO8/видео.html
You have to download and keep the file in lora folder. Also check if there are unwanted character in lora's name.
we need flux v1 no focus
Personnaly i'll stay on foocus, this fork is cool but too much demanding on ressources, and frankly, not really giving THAT MUCH better results. Need lot of optimisations for fast generations. An d more loras, etc.
Yes if SDXL is getting the things done.
For flux code needs to be optimised. Else forge UI / swarm UI could be a choice.
Could flux1d be installed in fooocus 2.5.5 and have it work? taking into account that I have amd and was able to get focus to work?
No, the original fooocus will not support flux. It's specifically for SDXL models
@@Aininja94 Thanks for answering brother, I appreciate your videos! but it hurts me that it doesn't work because ruined doesn't work on amd either, apparently
I am kind of working on an application myself. I am not an expert like this guys. But i am also looking for workaround for flux in that.
Let's hope my channel gives me some income so i can go full time into project.
@@Aininja94 Do you have any discord to share information? I'm also learning, but I'm a little stuck when it comes to generating a profile face of the same person starting from a frontal face, if you can make a video about that it would be great!
I am a introvert person first google and second RUclips is the only thing I started using. I don't understand discord because never tried.
But if you are looking for similar face generation. Then I have made 3 videos for it. It does the job for consistent face. ruclips.net/p/PLPFN04Wspxqvx68xxXlpFpgmyKy91ig1H
i get this error :
from .version import InvalidVersion, Version, _TrimmedRelease
ImportError: cannot import name '_TrimmedRelease' from 'packaging.version' (/usr/local/lib/python3.10/dist-packages/packaging/version.py)
Get LoRA keywords for lcm-lora-sdxl.safetensors (SDXL 1.0 - LORA)
what im doing wrong ?
Please post the error image of terminal in GitHub issue section.
Also the lora file name.
Did this error came while you were trying flux model?