Spaces:
Runtime error
Apply for community grant: Personal project (gpu)
We writing this request to apply for a GPU grant in support of our potential project, which is our goal is to develop an open source text to image based by utilizing the new Stable Cascade architecture that developed by StabilityAI.
Our goal is to make the first 2D (Anime) style in Stable Cascade architecture. It trained with 33K+ curated datasets from various sources. Our curation method is using CLIP+MLP approach and our watermark detector which is make our datasets is clean from "bad images" and watermark.
Our model (SomniumSC v1) has been downloaded by 200+ on CivitAI with 20 likes, and 217 downloads with 2 likes on Huggingface. We also always update our model with better approach, and currently we launch a new version which is SomniumSC v1.1. SomniumSC has better text rendering and object rendering than our SDXL based model AnysomniumXL v3.5.1
As a pioneer of the 2D (Anime) style in Stable Cascade architecture, we are optimistic to be the best model that available in Stable Cascade currently and gain more engagement toward this model. This model need more resource than Stable Diffusion 1.5, with minimum VRAM needed is 24GB and currently I can't managed to run it on CPU even on 24 core and 228GB setup, that's why we requesting this GPU Grant. We also want to share our model as the introduction to public on how good is Stable Cascade, and why they should reconsider to move on from Stable Diffusion 1.5 into Stable Cascade which is use less resource than SDXL.
I hope you can accept this GPU grant request. This GPU grant will be useful in AIArt community and could introduce novel approach on Text to Image diffusion based model
Thank you
