Тёмный

SegMoE - The Stable Diffusion Mixture of Experts for Image Generation! 

Nerdy Rodent
Подписаться 51 тыс.
Просмотров 12 тыс.
50% 1

Mixture of experts. Seems hot for AI text generation... but what if you had a mixture of experts for IMAGE generation? Oh. Segmind just did that. Welcome to SegMoE - the mixture of experts for SDXL, SDXL Turbo and Stable Diffusion 1.5.
Want to support the channel?
/ nerdyrodent
== Links ==
huggingface.co...
github.com/seg...
github.com/seg...
== More Stable Diffusion Stuff! ==
Faster Stable Diffusions with the LCM LoRA - • LCM LoRA = Speedy Stab...
How do I create an animated SD avatar? - • Create your own animat...
Installing Anaconda for MS Windows Beginners - • Anaconda - Python Inst...
Add anything to your AI art in seconds - • 3 Amazing and Fun Upda...
Video-to-Video AI using AnimateDiff - • How To Use AnimateDiff...
One image Gets You a Consistent Character in ANY pose - • Reposer = Consistent S...

Опубликовано:

 

4 окт 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 43   
@Mediiiicc
@Mediiiicc 8 месяцев назад
Need one of those experts to specialize in "hands" lol
@sherpya
@sherpya 8 месяцев назад
one expert per finger 😂
@greengoblin9567
@greengoblin9567 7 месяцев назад
Per cohk
@MarcSpctr
@MarcSpctr 8 месяцев назад
Finally a finetuned model for hands and legs can be used as expert, and maybe some model which can understand stuff like ON, ABOVE, UNDER, INSIDE, etc.
@MrGTAmodsgerman
@MrGTAmodsgerman 8 месяцев назад
What you wanna generate with "inside" in relation to body parts?
@worthstream
@worthstream 8 месяцев назад
This will be a game changer as soon as it's somewhat optimized. Expecially if they do manage to release a finetuning framework. Using prompts to compute gating functions is an ok starting point, but a (relatively) quick fine tune of that can make the difference.
@elihusolano5993
@elihusolano5993 8 месяцев назад
Hope you have a speedy recovery. Thanks for the great content.
@paulpardee
@paulpardee 8 месяцев назад
early days, as you say... I don't think this really gives the concept a fair shake. You have models that are better at one thing than others, but all the models currently out today are generalists who just happen to be slightly better at text or prompt adherence, or counting... An expert model would be focused on just text or just counting and those don't exist as far as I know. I'd love to see models built for this that have markup built in to tell Moe what they specialize in so it could direct that work to them... It'd be even better if you could have a standard library of models and Moe would dynamically load the best ones based on your prompt.
@ritpop
@ritpop 8 месяцев назад
I don't comment a lot but your content is great. Hope you get better soon.
@c0nsumption
@c0nsumption 8 месяцев назад
Fn love that you’re always willing to get dirty when there no community support bud. Thanks for the hard work 🙏🏽
@kariannecrysler640
@kariannecrysler640 8 месяцев назад
So few comments! I’m not used to that lol. Hope you’re good my nerdy friend ✌️💕🤘🥰 🐭
@NerdyRodent
@NerdyRodent 8 месяцев назад
Will be soon!
@kariannecrysler640
@kariannecrysler640 8 месяцев назад
@@NerdyRodent very happy to hear that 😁
@blacksage81
@blacksage81 8 месяцев назад
I feel like these researchers skipped a whole breakthrough by skipping Qlora, and the myriad of quantization flavors we could have played with and went straight to moe, when nearly all the models are just finetunes of the sd base. Its odd.
@ImAlecPonce
@ImAlecPonce 8 месяцев назад
looks so cool!! I only have 16 gig vram though
@SandyGoneByeBye
@SandyGoneByeBye 8 месяцев назад
hoping you're feeling back to full rodent normal soon
@stephantual
@stephantual 8 месяцев назад
Thanks that was fun :) 🤠
@fast_harmonic_psychedelic
@fast_harmonic_psychedelic 8 месяцев назад
i guess its a little better but CLIP training with partiprompts would be just as good
@elihusolano5993
@elihusolano5993 8 месяцев назад
can this new MoE be applied to Loras?
@AC-zv3fx
@AC-zv3fx 8 месяцев назад
I thought those experts must be trained with the model, so it can know what model to choose
@AC-zv3fx
@AC-zv3fx 8 месяцев назад
I wonder if it is possible to create MOE of Pony diffusion, AnimagineXL 3, realistic model and a model that is based on illustrations or traditional paintings.
@nickolaygr3371
@nickolaygr3371 7 месяцев назад
its like computer processors evolution
@aimademerich
@aimademerich 8 месяцев назад
This is phenomenal!!
@yahiiia9269
@yahiiia9269 8 месяцев назад
Could you theoretically use multiple LCM Turbo models?
@poipoi300
@poipoi300 7 месяцев назад
Wonder if we could truly consider this MoE. Haven't read the code, but I suspect all this does is amplify bias, probably akin to LCM but instead it's distributed.
@fast_harmonic_psychedelic
@fast_harmonic_psychedelic 8 месяцев назад
theyre all general models, none of the constituents are experts on any particular thing lol
@fast_harmonic_psychedelic
@fast_harmonic_psychedelic 8 месяцев назад
the whole MOE paradigm seems to me to be theoretically dubious lol
@Alice_Fumo
@Alice_Fumo 8 месяцев назад
Ok, but why?
@LouisGedo
@LouisGedo 8 месяцев назад
👋
@mattkupka1702
@mattkupka1702 8 месяцев назад
How was this much different than a checkpoint merge
@DoorknobHead
@DoorknobHead 8 месяцев назад
___m_/ o o \_m___ 0:46 Can someone take the Segmoe Ferret to the vet and get that ringworm removed from it's neck? Thanx, in advance.
@kallamamran
@kallamamran 8 месяцев назад
Isn't this just the same as merged models?
@JavierGarcia-td8ut
@JavierGarcia-td8ut 8 месяцев назад
in the SDXL I think you are using too low CFG setting... maybe?
@sadshed4585
@sadshed4585 7 месяцев назад
what cuda do you have? my torch is not saying cuda is available
@NerdyRodent
@NerdyRodent 7 месяцев назад
I use 12.3 locally
@oquletz
@oquletz 8 месяцев назад
i don't realy understand what is this. this is a tool to merge models? does it work for sd 1.5?
@aimademerich
@aimademerich 8 месяцев назад
Wow this whole time I thought your voice was AI, get well soon
@AliasArketer
@AliasArketer 8 месяцев назад
I boggle at what has been done, I boggle at what may yet BE done. We're in territory that we can't show grandparents and convince them it isn't magic anymore than other silly daftards can be convinced it isn't copy-paste.
@renovacio5847
@renovacio5847 8 месяцев назад
By by Chat GPT4 😂.. i was using it because the image generation.. but know..
@Guytron95
@Guytron95 8 месяцев назад
groovy. Too bad they didn't include image-to-image but still groovy.
@erics7004
@erics7004 8 месяцев назад
Me, with 4gb vram GPU 😢😢
@bilybob-c4p
@bilybob-c4p 3 месяца назад
why didn't this take off? Use a face, hands, background etc... expert and get way better images
@raymond_luxury_yacht
@raymond_luxury_yacht 8 месяцев назад
24gb humblebrag
Далее
PhotoMaker - better than IPAdapter?
12:51
Просмотров 41 тыс.
3D Gaussian Splatting! - Computerphile
17:40
Просмотров 138 тыс.
Mcdonalds cups and ball trick 🤯🥤 #shorts
00:25
Просмотров 150 тыс.
AI vs Artists - The Biggest Art Heist in History
44:23
Просмотров 348 тыс.
AI Video Tools Are Exploding. These Are the Best
23:13
Просмотров 173 тыс.
How AI 'Understands' Images (CLIP) - Computerphile
18:05
There Is Something Hiding Inside Earth
11:35
Просмотров 2,6 млн
10 AI Animation Tools You Won’t Believe are Free
16:02
Adobe is horrible. So I tried the alternative
25:30