bertbobson commited on
Commit
f0dc99b
·
verified ·
1 Parent(s): 19e96f2

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -4,7 +4,7 @@ The FP8 weights were cast to FP32 with the FP8 scales, then downcast to BF16 bef
4
 
5
  For use with https://github.com/BobJohnson24/ComfyUI-INT8-Fast
6
 
7
- Speed 1.78x faster(2.03s/it) than FP8(3.62s/it) on my 3090, without compile.
8
 
9
  Quick comparison:
10
 
 
4
 
5
  For use with https://github.com/BobJohnson24/ComfyUI-INT8-Fast
6
 
7
+ Speed is 1.78x faster(2.03s/it) than FP8(3.62s/it) on my 3090, without compile.
8
 
9
  Quick comparison:
10