tiny-stable-diffusion
tiny-stable-diffusion copied to clipboard
Tiny optimized Stable-diffusion that can run on GPUs with just 1GB of VRAM. (Beta)
Is this project actively being maintained? Apologize, but it was a struggle to get this working on Windows with the instructions provided. Setting up SD on Windows has become easier,...
According to https://github.com/openai/whisper/discussions/47#discussioncomment-5558664 it is required to install CUDA and a specific pytorch version of it. The latest by now is cu118, cu119 not released yet https://download.pytorch.org/whl/cu119
required for CUDA to work, see https://github.com/Xilinx/Vitis-AI/issues/822#issuecomment-1151970403
It's really impressive work! Could you please let me know if there are training scripts available? I'd like to train it on my own dataset.
Thank you for your efforts. I'm curious to know if there are any codes or scripts for quantizing my own 2-bit stable diffusion models, rather than relying on the pre-existing...
I am very intrested with your work, do you have any samples generated by 2-bit sd models? And how fast is it compared with original one?