encoder4editing
encoder4editing copied to clipboard
How to quantize the encoder model and convert to half precision (16 bit)