Pixart Sigma - Low VRAM & SDXL+ Resolutions in ComfyUI for Free!
Вставка
- Опубліковано 4 жов 2024
- Pixart Sigma was released recently, and while the main repo takes a little tweaking to run nicely on 24GB VRAM, ComfyUI comes to the rescue making it easy to run in just 6 GB!
Like with ELLA, T5 encodings replace CLIP leading to increased prompt adherence. Inthis video I show you how to get going for FREE in ComfyUI, and compare SDXL vs Pixart Sigma.
All this without one of those SD3 “no commercial use” licenses!
Want to support the channel? Get PixArt workflows and more!
/ nerdyrodent
Links
github.com/Pix...
github.com/cit...
huggingface.co...
github.com/Pix...
huggingface.co...
== More Stable Diffusion Stuff! ==
Installing Anaconda for MS Windows Beginners - • Anaconda - Python Inst...
Installing ComfyUI - • Install Stable Diffusi...
ComfyUI Workflow Creation Essentials For Beginners - • ComfyUI Workflow Creat...
Faster Stable Diffusions with the LCM LoRA - • LCM LoRA = Speedy Stab...
Make an Animated, Talking Avatar - • Create your own animat...
One Image Gets You a Consistent Character in ANY pose - • Reposer = Consistent S... - Наука та технологія
Wow. I'm impressed. The prompt understanding is something else - quite amazing!
Yup, it’s pretty good!
Thx for sharing. The demo looks promising and the model has a good understanding of prompts.
Yup! It’s great for prompt understanding and so ideal for generating an initial image 👍🏽
Thanks buddy! Especially caught your words that PixArt generates more different faces than SD. These identical smiling cute ladies with almond-shaped features are already starting to get annoying, and it has not even been a year. I have installed PixArt at my place, I'm studying it now, but the first results are impressive.
Catching up on some stuff! Thanks Nerdy! 😊
Ah, so that's why I couldn't load the model in comfy! Thanks - you cleared up where I was going wrong with the custom node install.
Glad I could help!
I wish we could convert any fine-tune sdxl model to T5 encoder like juggernaut that would be amazing
Since Stable diffusion 3 weight will probably never be released (or at least delayed until stability AI has something much better) its very nice to see new open models emerging!
All the stability ai stuff is research use only now anyway… unless you pay their subscription 😐
came for pixart, stayed for the outro.
wait.. i guess.. I left after the outro?
well, at least it got you a like, then !
Hey Nerdy. I respect your opinion and you haven't said anything about SD3. Please let us know what you think of it and the fact it doesn't seem to be open source anymore.
my focus is on things with open licensing, so with the SD stuff all being for research only, non-commercial use now it’s less of a focus. Great if you’re ok with paying the 20 dollars a month though!
@@NerdyRodent It's just sad that they seem to have changed their philosophy on people being able to use their product for free. Blender for example still continues to be free to the user. If the usage was reasonably priced then it wouldn't be such a problem but when you experiment and fiddle around with prompts like i and many others do it will be very expensive.
Better to test native Sdxl the models like zavychromal are trained by users in one direction they prefer. Zavy does better with real ppl and models i think.
Thanks for the great content! What I wasn't sure about was if I already have comfyui Install up and running, can I just skip the preparation step and move directly to step two? Thanks!
Yup, the video assumes you’ve got ComfyUI installed already 👍🏽
Hey can u please make a video on how to add high res fix script to pixart sigma I found a sd1.5 pixart workflow thar works on cpu its called Abominal spaghetti workflow I saw the sd1.5 ksampler but I want to add it to thar but the efficiency ksampler doesn't have the start step and end step nodes so I can't add it to the workflow this hrf helps alot with low res images
Like this one 😉 ControlNet Union ProMax, Pixart Sigma 900M & More New Stuff!
ua-cam.com/video/JgYzCEzHDrc/v-deo.html
do you have a workflow which has pixart and controlnet and ipadapter?
This uses SDXL components but isnt actually sdxl right? Like we cant use sdxl control nets or anything
Yes, you need to use Pixart control nets
@@NerdyRodent Is there Pixart control nets workflow avail ?
i am trying to do some experiments , will it work with either ipadapter or faceid , have anyone tried it
what linux distro are you using / can recommend?
Ubuntu LTS 😀
@@NerdyRodent nice, thanks for the quick reply
This new model is great. Thanks Rodent!
.git directory in pixart_sigma_sdxlvae_T5_diffusers have extra 20 GB. Can be deleted.
Hoping this comes to forge
6min left
I'd guess they are samey on sdxl because you are using a community model. I think it would be fair if you used the base model, at least in addition.
There is, of course, the actual prompt understanding component to consider as well
But I see you use CPU to run, not GPU to run.
Yes. Offloading the T5 node to run on the CPU saves loads of VRAM 😃
@@NerdyRodent I don't quite understand, in what cases can we run GPU?
👋
👋
I feel super sad when I see something very cool and is Cuda
dependent, fuck AMD
Linux + CUDA is where it's at for AI at the moment :/
@@NerdyRodent ye.. :/
I use ROCm on Linux and I'm Ok with it.
@@San4itos RX580 here so I don't know how to use Rocm properly and think HIP isn't available I even try to use Zluda in a new version and message related to HIP shows up