1 paper
Basile Lewandowski, Simon Kurz, Aditya Shankar +3
Diffusion models are the go-to method for Text-to-Image generation, but their iterative denoising processes has high inference latency. Quantization reduces compute time by using l…