The inference speed of the int8 quantization version of SDXL is much slower than that of fp16 #3830

Sign in to view logs

Run time

Learn about OS pricing on GitHub Actions

Job	Run time
Authorization	0s
Upload log	0s
Vulnerability scan	0s
Start ci job	0s
	0s