Fetching the paper…

QuEST: Low-bit Diffusion Model Quantization via Efficient Selective Finetuning · Around