Fooocus
stable-fast
Fooocus | stable-fast | |
---|---|---|
34 | 11 | |
35,143 | 965 | |
- | - | |
9.8 | 9.4 | |
5 days ago | about 1 month ago | |
Python | Python | |
GNU General Public License v3.0 only | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
Fooocus
-
AI, but at what cost? The energy-inefficient AI era is already here
But we can come to a pretty realistic (although not as accurate) conclusion if we put our minds to it. I chose Fooocus for this example, which is the most straightforward (and I believe popular) stable diffusion GUI out there. Let's start simple:
-
How to Persist Data in Google Colab Using JuiceFS
# Install the JuiceFS client. !curl -sSL https://d.juicefs.com/install | sh - # Mount the JuiceFS file system. !juicefs mount rediss://:[email protected]/1 myjfs -d # Create the directory structure for Fooocus models in JuiceFS. !mkdir -p myjfs/models/{checkpoints,loras,embeddings,vae_approx,upscale_models,inpaint,controlnet,clip_vision} # Clone the Fooocus repository. !git clone https://github.com/lllyasviel/Fooocus.git
-
Stable Cascade
That looks very impressive unless the demo is cherrypicked, would be great if this could be implemented into a frontend like Fooocus https://github.com/lllyasviel/Fooocus
-
Stable Code 3B: Coding on the Edge
You might be thinking of Fooocus: https://github.com/lllyasviel/Fooocus
The Stable Diffusion web interface that got a lot of people's attention originally was Automatic1111: https://github.com/AUTOMATIC1111/stable-diffusion-webui
Fooocus is definitely more beginner friendly. It does a lot of the prompt engineering for you. Automatic1111 has a ton of plugins, most notably ControlNet which gives you fine grained control over the images, but there is a learning curve.
-
Ask HN: How are you using ChatGPT for yourself?
I just installed this last night on my laptop:
https://github.com/lllyasviel/Fooocus
Highly recommend:
>"Looking up from the deck of golden gate bridge at the towers and metal work, the towers rise and arch back in an ominous and foreboding manner. more artistic, like an alphonse mucha propaganda poster - slightly fish-eye feeling" -- https://i.imgur.com/vyNg79f.jpg
the local UI and 1.27.0.0.1 - https://i.imgur.com/wRwghuN.jpg
-
It took SEVEN MINUTES to do this using fooocus on a 1060 3g with 16 ram. Can I make it faster?
Why not use the Google Colab notebook while it's still a free option at: https://github.com/lllyasviel/Fooocus It's not bad. I've been using the Colab Fooocus notebook and A1111 on Sage with the 4 free hours of GPU time. The Colab has the Juggernaut model preloaded, but I combined some code from another notebook to add other models and loras.
- Could I use SDXL on a 4gb VRAM?
-
Looking for an open source Image generator with no limits
i'm trying to test the abilities of image generators and the risks that can comes with it. and i'm looking for an image generator that work locally and has no limits. i used the Fooocus project from github and the juggernut model and it's capable of generating nude pictures but not fully nude pictures. and it doesn't work will with bloody scenes. any recommendation for a better model
-
What is the licensing of SD models/frameworks?
I recently saw this video from Fireship and I started wondering about licensing of SD models and frameworks. Fireship shows Fooocus and advertises it as a cool solution. What I started wondering about is, Fooocus downloads a couple of models: Juggernaut XL, some control nets, some loras. What licensing is tied to all of this? One I am most insterested in is JuggernautXL, on civitai it's listed as having CreativeML Open RAIL++-M license but in the description there's remark: For business inquires, commercial licensing, custom models, and consultation contact me under [email protected] There's a lot of separate parts going on in AI frameworks and it's a bit unclear to understand if I can use this commercialy.
-
What AI is best for this kind of pictures?
For running locally w/o a lot of "hazzle" i would recommend using https://github.com/lllyasviel/Fooocus and the "Sticker" style, which is available within the UI (Advanced tab). If you need lot of text directly in you images, you may have difficulties with SD and other AI models. In this case LoRas could help. Example: https://civitai.com/posts/880523 (check details for used prompt) In this case i used the following LoRa (a LoRa is kind of specialized submodel for style, concept or person) .
stable-fast
-
Has anyone managed to get TensorRT working in ComfyUI on Windows?
Download (https://github.com/chengzeyi/stable-fast/releases) and install stable-fast binary, compiled according to your system: pip install stable_fast-0.0.13.post3+torch210cu118-cp310-cp310-win_amd64.whl
- Optimum-NVIDIA - 28x faster inference in just 1 line of code !?
- stable-fast for SD inference: Faster than AITemplate, On par with TensorRT
- [N] stable-fast for SD inference: Faster than AITemplate, On par with TensorRT
- Stable-fast for SD inference: Faster than AITemplate, On par with TensorRT
-
SDXL Turbo: A Real-Time Text-to-Image Generation Model
SDXL and ControlNet are already optimized, if thats what you mean: https://github.com/chengzeyi/stable-fast
(Note the links to various SD compilers).
But the whole field is moving so fast that people aren't even adopting the compilers at large.
-
Getting sub 100ms refresh rate on LCMs
> already compiling
Hmm, well if you mean torch.compile, y'all should still check out stable-fast, which is claiming ~16ms/iter on a 4090:
https://github.com/chengzeyi/stable-fast#rtx-4090-512x512-ba...
-
Generate images fast with SD 1.5 while typing on Gradio
Now combine this with an optimized SD implementation, like:
https://github.com/chengzeyi/stable-fast
Or AITemplate, and you are at 15FPS on a larger consumer GPU. 10 with a controlnet you can use for some motion consistency.
-
S-LoRA: Serving Concurrent LoRA Adapters
Since I am sending you down the rabbit hole anyway, you should check out sfast:
https://github.com/chengzeyi/stable-fast
It's, the most promising "fast" and flexible stable diffusion implementation akin to this paper or vLLM that I know of. It doesn't have as many caveats as other implementations, like AITemplate (which is basically Turing+ and linux only) or torch.compile (basically no support for changing inputs/loras).
-
🚀Announcing stable-fast v0.0.5: Speed Optimization for SDXL, Dynamic CUDA Graph
About 2 weeks ago, I released the stable-fast project, which is a lightweight inference performance optimization framework for HuggingFace Diffusers. It provides best performance while keeping the compilation dynamic and flexible, and supports ControlNet and LoRA seamlessly.
What are some alternatives?
ComfyUI-AIT
gpt-fast - Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.
stable-diffusion-webui-forge
TensorRT-LLM - TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM also contains components to create Python and C++ runtimes that execute those TensorRT engines.
ComfyUI - The most powerful and modular stable diffusion GUI, api and backend with a graph/nodes interface.
StableCascade - Official Code for Stable Cascade
PixArt-alpha - PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis
InvokeAI - InvokeAI is a leading creative engine for Stable Diffusion models, empowering professionals, artists, and enthusiasts to generate and create visual media using the latest AI-driven technologies. The solution offers an industry leading WebUI, supports terminal use through a CLI, and serves as the foundation for multiple commercial products.
ollama - Get up and running with Llama 3, Mistral, Gemma, and other large language models.
CushyStudio - 🛋 The AI and Generative Art platform for everyone
k-diffusion - Karras et al. (2022) diffusion models for PyTorch
social-ai - social ai image generation