Input type (torch.cuda.HalfTensor) and weight type (torch.FloatTensor) should be the same

Question

fabwa opened this issue 2 years ago · comments

Still getting this when trying to use GFPGAN.
Latest version as of now.

!!Runtime error (txt2img)!! 
 Input type (torch.cuda.HalfTensor) and weight type (torch.FloatTensor) should be the same
exiting...calling os._exit(0)

fabwa · Answer 1 · Thu Sep 01 2022 16:51:17 GMT+0800 (China Standard Time)

Seems to work now