Video - https://youtu.be/SIVA8fPfPU4
Even more text-to-video and image-to-video fun this week, thanks to SkyReels-V1 models from SkyworkAI!
Thanks to Kijai, there are smaller models available so you won't need 24+GB VRAM to run these, though it will certainly help! Even th...
2025-02-22 17:18:57 +0000 UTC View PostVideo - https://youtu.be/5Wae2H3ures
There have been a few new models recently, so I figured I'd take a look at what they have to offer!
1. Probably the best of them is Lumina 2.0 which features some impressive prompt comprehension in a 2B model. The styles are a bit hit and miss, but at least it has some ;)
2025-02-16 13:43:43 +0000 UTC
View Post
Video - https://youtu.be/6FBnKIjqT04
Open Source music generation at home takes another step forward thanks to the change of license for YuE, which is now Apache 2.0!
Like Suno or Udio, you can provide a set of lyrics along with some tags in order to create some AI music.
Text prompts aren't your only control though, as you can also upload vocal or instrumental audio clips (30...
2025-02-08 18:56:23 +0000 UTC View Post
Video - https://youtu.be/T7nufrDmc9A
DeepSeek. It's all the rage at the moment, so I figured I'd give it a spin in ComfyUI to see how good it was at prompt enhancing for image generation!
The following models were used to compare against:
DeepSeek-R1 (14B)
Gemma2 (9B)
Llama 3.2 (3B)
Olmo (7B)
Phi4 (14B)
2025-01-31 20:29:41 +0000 UTC View PostVideo - https://youtu.be/2w_vlTFyntI
EasyAnimate have released their new v5.1 models giving us a variety of control methods such as pose, scribble, depth, trajectory and even camera control. Even better, it all looks to be under an Apache 2.0 license, making it the latest truly ...
2025-01-25 18:09:41 +0000 UTC View PostVideo - https://youtu.be/fjUdzKEIypk
ComfyUI now supports some of the Nvidia Cosmos models, which means image to video fun at home, for free, on your own computer :)
Ok, so it's not as good as the pay-for websites, but it is still a lot of fun with plenty of potential in this first release.
2025-01-19 15:51:01 +0000 UTC
View Post
Video - https://youtu.be/beiVSzsH2R0
Who wants to generate images or video faster? Me, for one! Thanks to these new cache nodes for ComfyUI, those much needed speed boosts are available - even for those of use with older cards!
Things can go even faster with torch.compile,...
2025-01-11 21:50:53 +0000 UTC View Post
Video - https://youtu.be/bHqJ6mo1P-4
SillyTavern is a user interface you can install on your computer (and Android phones) that allows you to interact with LLMs and other backend APIs.
Create your own tavern filled with characters of your choice. Give them all personalities, avatars and even indivi...
2025-01-05 21:12:52 +0000 UTC View PostVideo - https://youtu.be/vwYItu2OaYk
I'm back once again with a quick look back into 2024 and a relaxing, easy-to-use, fast, low VRAM image animation node called DepthFlow.
Turn your 2D images into stunning 2.5D parallax animations using Depthflow in ComfyUI.
An open source Immersity alternative.
I'd also like to wish everyone a very Happy New Year. 2025 here we come!
2024-12-31 14:45:31 +0000 UTC
View Post
Yes, I'm not here for the holiday season :)
Being that time of year I am currently away with family. However, I just wanted to take a moment to say a big thank you to everyone, and to wish you all a wonderful new year!
I'm sure we'll see even more amazing stuff in 2025, but what would your New Year's Wish be for AI? Something totally new? More image to video? Virtual human avatars?
Video - https://youtu.be/dASMuSCCNSU
Ever wanted to use one LoRA for part of your image, and another totally different LoRA (or model) for another?
Well, if you've seen the recent ComfyUI blog post about some new nodes for adding LoRAs to conditioning areas, along with various other nodes too, then you'll know this is now a lot easier to do!
The blog quite is quite long a...
2024-12-14 13:56:10 +0000 UTC View PostThe licence only allows certain territories to use the model, so I've removed the video and there are no outputs shown here. Next time I'll need to make sure I read the licence properly!
Mistake 1:
The license for the Tencent Hunyuan model only allows usage in certain territories, and I’ve now realized that the license excludes the United Kingdom and South Korea, which I initially missed. As a result, I’ve removed the video and the model files to c...
2024-12-06 17:04:40 +0000 UTC View PostVideo - https://youtu.be/6AHL0XZKbqg
Loads of new things in the Fluxtapoz repo now including
InstantX based Regional Prompting
RF-Edit
Perturbed Attention Guidance (PAG) and Smoothed Energy Guidance (SEG)
(Example RF-Edit before and afte...
2024-12-01 15:53:26 +0000 UTC View Post
Video - https://youtu.be/1AlmHTFWjLk
Black Forest Labs have released a bunch of new models for us to play with, so lets get to it! Depth, Canny, Redux and Fill - plus depth & canny LoRA versions too!
The ComfyUI examples page has the links and locations for each of their ...
2024-11-22 00:40:46 +0000 UTC View PostVideo - https://youtu.be/n2QZEtBU89o
In-Context loras help keep things... in context!
If you get Flux to generate a single, large image and prompt for various parts of that image (such as top-left, bottom-right, etc), then you can crop the final image into a number of smaller (but hopefully consistent) images.
I...
2024-11-16 16:39:17 +0000 UTC View PostVideo - https://youtu.be/mwiz1PNDWGg
OmniGen is a multimodal model which can follow instructions in English. No need for IPAdapters, Depth or Canny controlNets, etc as OmniGen does it all. Ask it questions about the picture and get it to change any aspect you like!
One thing OmniGen does really well is create consistent characters, and even better still you can easily have up t...
2024-11-10 21:54:27 +0000 UTC View Post
Detail Daemon is an easy way to add details to Flux, SDXL or SD1.5 generations!
Video Guide: https://youtu.be/dG9HfgngsqM
https://github.com/Jonseed/ComfyUI-Detail-Daemon
2 samplers,...
2024-10-31 23:48:03 +0000 UTC View Post
Video - https://youtu.be/WbLlYCc4ZnU
Fed up of having to add loads of loras to change the style of Flux.1? Welcome to PixelWave!
Example of a slightly less "fluxy" chin & nose... but best of all it's an actual painting style!
Video coming soon!
Stability AI: https://stability.ai/news/introducing-stable-diffusion-3-5
ComfyUI Examples: https://comfyanonymous.github.io/ComfyUI_examples/sd3/
Video - https://youtu.be/-qVDWE0A1ek
Loads of changes in this one.
With the recent release of the AliMama beta inpainting controlnet, I figured I'd try it out, and now you can too!
The new Input Image section let's you easily pick from a variety of image-related options:
Enable IMG2IMG will use the image latent instead of an empty one
IMG Meg...
2024-10-17 12:11:51 +0000 UTC View PostVideo Guide: https://youtu.be/qMNJ0dCoT_I
This is one for those who don't like noodles, tweaking things - or even moving around in the ComfyUI interface!
Everything you need for generating is all in once place, with shortcuts keys for the first-time use settings. This leaves you to focus (foooocus?) on prompting, with just the required options for:
Prompting
Image size
...
2024-10-12 13:07:48 +0000 UTC View PostVideo - https://youtu.be/Sc6HbNjUlgI
1. Pick a scheduler and steps (always 8 for an initial hyper lora image)
2. Connect it (optionally via split sigmas) to the debug node to view it's values
3. Add a CustomScheduler node & set the steps
4. Copy the values from your chosen scheduler into the custom one
5. Tweak!
Some v...
2024-10-05 15:01:58 +0000 UTC View Post
In the Enhanced Edition I'm also trying to just how far you can push the upscale. As soon as it hits anything close to 4000 then everything goes all blocky, so somewhere more around 3k is probably the maximum at the moment.
This version is somewhat experimental, but my eyes eventually just start seeing lines everywhere 0_0
The different types of noise also seem to have an impact on the stripes, with pyramid noise getting rid of most of them... also leaving some colourful noise in ...
2024-09-29 18:32:04 +0000 UTC View PostJust a very simple workflow for creating high resolution images with ComfyUI!
This is designed to make images with sides around 1k->2.5k (max), hopefully with less banding than usual. The notes in the workflow have all the links, and there is a copy below also:
Impact Pack = ID 2
Extra Models for ComfyUI = ID 53
ComfyUI-GGUF = ID 54
Well, this took a long time to test! During testing found a few things, which I split into 3 sections.
Don't have the VRAM for a controlnet, or simply want a new way to do IMG2IMG? Then try ddim_uniform along with Hyper an...
2024-09-20 11:09:22 +0000 UTC View Post
video - https://youtu.be/PRiqTnP3no8
Thanks to the increased resolution, up-scaling takes much longer than generating the initial image... if only there was a way to speed that up a bit?
Well, it looks like Hyper-SD also works great when up-scaling too, and combined with the Euler CFG PP sampler, a latent multipl...
2024-09-05 20:44:52 +0000 UTC View Post
video - https://youtu.be/JqEi82Q9Qfc
Hyper-SD have released some Flux.1 loras for you to reduce the number of steps required in Dev down to just 8! :)
If that isn't enough, Shakker & InstantX have partnered to give us all these goodies:
Controlnet Canny
Controlnet Depth
Controlnet Union
Controlnet Union Pro
LoRA Add Detail...
2024-09-01 12:51:44 +0000 UTC View PostVideo - https://youtu.be/g8W3xe5kRBQ
A whole different set of nodes for Flux.1 are available from XLabs to give a new sampler along with a whole bunch of different models for ControlNets, LoRAs and an initial ipadapter.
Canny & HED models allow you to control your generations using image outlines, whilst their depth model will work with depth maps. Use your own, or pre-process standard images as...
2024-08-23 13:54:13 +0000 UTC View PostLoads more unets have dropped - ideal for those with smol GPUs! There is even a Q2 :)
Can be 50% slower than the standard dev model FP8 weight type, but it's a small price to pay. Wonder how many times I can mention small?
Anyway, here are all the options at the moment:
https://huggingface.co/city96/FLUX.1-dev-gguf/tree/main
2024-08-18 11:25:58 +0000 UTC
View Post
Figured I'd try a different model, so this time it's llava!
Much the same idea as with using Florence to do the captioning, only this time I'm using LLM Party and Llava:
ollama pull llava-llama3
As with the Florence captioning you'll need to set 3 paths:
1. The directory with all your images
2. The full path to where you want to save the text files (e....
2024-08-17 18:54:35 +0000 UTC View Post