A whole bunch of updates went into ComfyUI recently, and with them we get a selection of new samplers such as EulerCFG++ and DEIS, as well as the new GITS scheduler. See them all in action, then try it yourself at home!
Want to support the channel?
https://www.patreon.com/NerdyRodent
* DEIS, GITS, iPNDM - https://github.com/zju-pi/diff-sampler
* CFG++ - https://arxiv.org/abs/2406.08070
== Learn More Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
If you've been hunting for an AI art model with a reasonable licence, then perhaps Lumina, Pixart or Hunyuan may be of interest? Each one has it's own little quirks and today they're pitted against each other in another battle of the prompts!
Want to support the channel?
https://www.patreon.com/NerdyRodent
Lumina Next - https://github.com/Alpha-VLLM/Lumina-T2X
Hunyuan - https://github.com/Tencent/HunyuanDiT
Hunyuan Video - https://youtu.be/oDK0-KesWQo
Pixart Sigma - https://github.com/PixArt-alpha/PixArt-sigma
Pixart Sigma Video - https://youtu.be/TQduyxvzX4Q
Music created using udio.
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Time for a prompt showdown! Try these at home using your favourite model (such as Stable Diffusion 3) and see how they compare to both Pixart Sigma and HunYuan DiT. A wide range of styles are covered, and the prompts are listed below for your copy-and-paste ease :)
Want to support the channel?
https://www.patreon.com/NerdyRodent
https://www.patreon.com/posts/prompt-showdown-106258102
Update: HunYuan v1.1 is out now too - and it's even better!
https://huggingface.co/Tencent-Hunyuan/HunyuanDiT-v1.1
= Prompt Showdown! =
Negative Prompt:
many hands, really wobbly, distorted and blurry fingers and hands.
Positive Prompts:
1:
2: A woman sleeping on the grass.
3: An avocado chair
4: Anime art style blue rabbit flying through the air wearing a red cape and goggles
5: Vector-art style business logo, SVG, simple, plain, psychic rodent emoji.
6: A fantasy-art style rodent mage, level 12
7: Chibi rodent scientist discovering where he left his pencil
8: A giraffe engineer, epic Shōnen manga style, high detail Seinen
9: Kemono style illustration of a rabbit doctor wearing a blue surgeon's gown walking down the path of an old cemetery. Green eyes, brown fur, fine ears, solemn feel. misty path. dignified, austere. fog, gloom. haunted vibe.
10: Cthulhu stands over a kitten, oil painting style. A classical artwork, vintage, old. The scene is set as the beast towers over the tiny, but dreadful, all-black kitten. The bright summer sky is in high contrast to the dark, evil shadows that lurk beneath. The style is reminiscent of Jacques Stella and Pompeo Batoni. In the background is an ancient fiery temple of doom, hewn from the very rock face itself by the kitten. The image quality is astounding.
11: A cubist art style kangaroo druid in the forest, cubism, quality artwork, soft pastel shades
12: In the Chinese ink painting style, a little deer stands leisurely in the lush bamboo forest. The morning light shines through the bamboo leaves, casting mottled light and shadows. The little deer quietly drinks the clear stream water, and the bamboo leaves are swaying in the wind. The whole picture exudes a sense of tranquillity and harmony.
13: A stone statue of a deer relaxing on top of a colourful bed in an old, Victorian house. Professional photo, bokeh.
14: A childish doodle of a bad kitten
15: Paper-cut art style tiger emerging from a bunch of flowers, bold colours, depth, shading, 3d effect, pop-up
16: A ghostly face is peering through the window into my house from outside! The ghost looks very scary but may be wearing a mask, and a sense of evil can be felt. charcoal art, shading, sketch.
17: 在一个充满创意的画廊里，一幅引人注目的画作展示了一只时髦的啮齿动物。画作采用了印象派风格，色彩斑斓，笔触细腻，充满生动的光影效果。画中的小鼠戴着复古圆框眼镜，穿着时尚的格子衬衫和牛仔裤，手持一杯咖啡，悠闲地站在一片繁茂的花园中。背景是柔和的色彩斑点和模糊的树影，营造出宁静与优雅的氛围，突显了小鼠的独特魅力。
== More AI Things! ==
* Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Easy Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
Just a quick video to show some of the language understanding capabilities of two image generators. This time it's HunYuan DiT vs the base SDXL 1.0 model. Some people wondered how the Chinese vs English language negative prompts would impact the SDXL generation, and so here is the answer! Also included are a few other languages, so if you're got your score card from the original video, you can now do some updates ;)
Just to handle two things at once, I've also included the "extended version" for one of the pieces of outro music I've made on Udio!
Full video: https://youtu.be/oDK0-KesWQo
Udio song link: https://www.udio.com/songs/hmeqhhQiEniVK6rUapCrFs
Want to support the channel?
https://www.patreon.com/NerdyRodent
#shorts #sdxl #ai
Overall, in HunYuan-DiT's tests, it scores 59.0% vs 56.7% for Stable Diffusion 3, making it objectively better. More than 50 professional evaluators performed the evaluations, so it must be true... or is it? For you to be able to judge for yourself, I pit HunYuan-DiT against the weaker (but freely available) SDXL in a battle for AI supremacy!
Want to support the channel?
https://www.patreon.com/NerdyRodent
https://www.patreon.com/posts/hunyuan-dit-than-105823648
GitHub Repo:
https://github.com/Tencent/HunyuanDiT
== More Nerdy Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with Hyper SDXL - https://youtu.be/LAQYZWbmkwA
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
Combine an LLM with Stable Diffusion and you can Omost generate the image you ask for!
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://github.com/lllyasviel/Omost
https://www.patreon.com/posts/your-image-is-105386033
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Khoj is your AI "second brain". Get answers to your questions, whether they be online or in your own notes. Use local or hosted LLMs. Self-host locally or use our cloud instance. Access from Obsidian, Emacs, Desktop app, Web or Whatsapp. Make agents, do automations and more! RAG made easy :)
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://github.com/khoj-ai/khoj
https://khoj.dev/
== More Links! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Chapters
0:00 Khoj introduction
10:13 Khoj installation
Faces. Most of us have at least 1, and they're great fun to play with. This AI stuff seems to be zooming ahead really quickly, so I figured why not put a bunch of face things together, both old and new, and see how quickly it can generate? Thus, Refacer was born. Powered by SDXL, you can pop in any face and change it to whatever style takes your fancy! Realistic to anime, painting to realistic, cartoon to pixel art - it's all down to your choice of prompt and model :)
Want to support the channel?
https://www.patreon.com/NerdyRodent
GitHub:
https://github.com/nerdyrodent/AVeryComfyNerd
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Photomaker - https://youtu.be/ZTck128jfFY
* Hyper SDXL - https://youtu.be/LAQYZWbmkwA
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Reposer = A Consistent Character In ANY Pose - https://youtu.be/SacK9tMVNUA
Yes. There's nothing new about speaking to an AI and having it speak back, however, when the responses are this fast then things start to become a bit more fun - especially with an AI as unhinged as this one...
dnhkng GlaDOS is free, open-source goodness, meaning you can download and run on your own computer in the comfort of your own home. Everything runs locally, so no API keys to mess around with either. Yay!
NB. This is not a valve product, and yes - they probably need a better name for the project ;)
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://github.com/dnhkng/GlaDOS
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
== Contents ==
0:00 - dnhkng GlaDOS Intro
4:43 - dnhkng GlaDOS Install
The new Hyper-SD models are FREE and there are THREE ComfyUI workflows to play with! Use the amazing 1-step unet, or speed up existing models by using the LoRAs.
Better than LCM? Take a look for yourself and see!
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://huggingface.co/ByteDance/Hyper-SD
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI for Beginners - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflows for Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* Make A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
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 workflows and more!
https://www.patreon.com/NerdyRodent
Links
https://github.com/PixArt-alpha/PixArt-sigma
https://github.com/city96/ComfyUI_ExtraModels
https://huggingface.co/spaces/PixArt-alpha/PixArt-Sigma
https://github.com/PixArt-alpha/PixArt-alpha/blob/master/asset/docs/pixart_comfyui.md
https://huggingface.co/PixArt-alpha/PixArt-Sigma/blob/main/PixArt-Sigma-XL-2-1024-MS.pth
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI - https://youtu.be/2r3uM_b3zA8
* ComfyUI Workflow Creation Essentials For Beginners - https://youtu.be/VM9snsuoqBc
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* One Image Gets You a Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
UDIO vs SUNO - here we go! Generating music from text has never been easier, and boy is this a lot of fun :) Enter your own lyrics, use AI generated ones, extend, remix and more!
No install!
Want to support the channel?
https://www.patreon.com/NerdyRodent
Link:
https://www.udio.com/
== More Stable Diffusion Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI - https://youtu.be/2r3uM_b3zA8
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* One Image Gets You a Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Diffusion models have made incredible strides in text-to-image generation, but they still struggle with dense, complex prompts that involve multiple objects, detailed attributes, and intricate relationships.
Enter ELLA - the Efficient Large Language Model Adapter that's poised to revolutionize how diffusion models handle sophisticated prompts. This ingenious adapter equips text-to-image models with the power of large language models, without the need to train either the U-Net or the language model itself.
In this video, I dive into using ELLA in ComfyUI, and explore how it tackles the limitations of current text encoders like CLIP. Prepare to be amazed as Ishowcase the superior performance of ELLA compared to CLIP conditioning, unlocking a new level of sophistication in text-to-image generation. If you've ever struggled to craft the perfect prompt, this video is a must-watch!
(* oops! Was playing with scheduling in the video and so forgot to switch back to sgm unfiform, but it really doesn’t make much difference. It also turns out they are no longer going to release the SDXL weights.)
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://ella-diffusion.github.io/
https://github.com/TencentQQGYLab/ELLA
https://github.com/ExponentialML/ComfyUI_ELLA
https://huggingface.co/QQGYLab/ELLA/blob/main/ella-sd1.5-tsc-t5xl.safetensors
https://huggingface.co/google/flan-t5-xl/tree/main
https://huggingface.co/Kijai/flan-t5-xl-encoder-only-bf16/tree/main
== More Stable Diffusion Related Stuff! ==
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Installing ComfyUI - https://youtu.be/2r3uM_b3zA8
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Make an Animated, Talking Avatar - https://youtu.be/Z7TLukqckR0
* One Image Gets You a Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Inpainting with the latest BrushNet models for AI generation using stable diffusion is lots of fun! Get great results quickly thanks to both random and segmentation mask models - all for free on your own computer! Installation is quick and easy, so why not give it a go today?
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://github.com/TencentARC/BrushNet
https://huggingface.co/spaces/TencentARC/BrushNet
https://pytorch.org/
https://huggingface.co/spaces/abhishek/StableSAM/blob/main/sam_vit_h_4b8939.pth
Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
https://pixabay.com/photos/teddy-bear-toy-plush-stuffed-toy-1835598/
== More Stable Diffusion Stuff! ==
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* Video-to-Video AI using AnimateDiff - https://youtu.be/kJp8JzA2aVU
* How to make Consistent Characters From 1 Picture In ComfyUI - https://youtu.be/SacK9tMVNUA
0:00 Inpainting Introduction
1:19 BrushNet 0-Install Option
1:37 BrushNet Local Install Option
7:26 BrushNet Examples
12:03 ComfyUI Inpainting Example
Using python & AI for animation images is always something I find fun, and there's a new kid on the block - AniPortrait! Using the power of Stable Diffusion, this repo can create animated avatars with fairly minimal effort. There have been quite a few research advances over the years, and this brings us a step closer. Want to install some of the latest, cutting edge, AI research at home? Well now you can, as this will run on consumer hardware too 😃 This step-by-step tutorial will guide you through getting it running locally, meaning you too can be on the cutting edge! Crack that anaconda terminal open once again to begin your journey…
Want to support the channel?
https://www.patreon.com/NerdyRodent
AniPortrait - https://github.com/Zejun-Yang/AniPortrait
AniPortrait paper - https://arxiv.org/abs/2403.17694
Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
RVC WebUI - https://youtu.be/g4td-n8-QZk
The new IP Composition Adapter model is a great companion to any Stable Diffusion workflow. Just provide a single image, and the power of artificial intelligence will analyse the very composition itself - ready for you use!
Check out some of things you can do with it :)
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://huggingface.co/ostris/ip-composition-adapter
== More Stable Diffusion Stuff! ==
* Faster Stable Diffusions with LCM LoRA - https://youtu.be/zrxd95Mxz24
* SD Generated Avatar Animation - https://youtu.be/Z7TLukqckR0
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Video-to-Video AI using AnimateDiff - https://youtu.be/kJp8JzA2aVU
* One image = A Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Visual style prompting aims to produce a diverse range of images while maintaining specific style elements and nuances. During the denoising process, they keep the query from original features while swapping the key and value with those from reference features in the late self-attention layers.
Their approach allows for the visual style prompting without any fine-tuning, ensuring that generated images maintain a faithful style.
My personal favourite so far - and yes, it works in ComfyUI too ;)
Want to support the channel?
https://www.patreon.com/NerdyRodent
Links:
https://github.com/naver-ai/Visual-Style-Prompting
https://github.com/ExponentialML/ComfyUI_VisualStylePrompting - WIP
== More Stable Diffusion Stuff! ==
* Install ComfyUI - https://youtu.be/2r3uM_b3zA8
* Make Image QUICKLY with an LCM LoRA! - https://youtu.be/zrxd95Mxz24
* How do I create an animated SD avatar? - https://youtu.be/Z7TLukqckR0
* Video-to-Video AI using AnimateDiff - https://youtu.be/kJp8JzA2aVU
* Consistent Characters in ANY pose with ONE Image! - https://youtu.be/SacK9tMVNUA
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
Stable Diffusion normally doesn't make transparent PNGs, but now you can thanks to Layer Diffuse!
Available for both Forge WebUI and ComfyUI, Layer Diffuse does much more than just create single images without backgrounds too.
Links:
https://github.com/huchenlei/ComfyUI-layerdiffuse
https://github.com/layerdiffusion/sd-forge-layerdiffuse
== More Stable Diffusion Stuff! ==
* Faster Stable Diffusions with the LCM LoRA - https://youtu.be/zrxd95Mxz24
* How do I create an animated SD avatar? - https://youtu.be/Z7TLukqckR0
* Installing Anaconda for MS Windows Beginners - https://youtu.be/OjOn0Q_U8cY
* Add anything to your AI art in seconds - https://youtu.be/riLmjBlywcg
* Video-to-Video AI using AnimateDiff - https://youtu.be/kJp8JzA2aVU
* One image Gets You a Consistent Character in ANY pose - https://youtu.be/SacK9tMVNUA
Want to support the channel?
https://www.patreon.com/NerdyRodent
Conectar
