Add CPU fp8 support

Since norm layer need fp32, I only convert the linear operation layer(conv2d/linear)

And TE have some pytorch function not support bf16 amp in CPU. I add a condition to indicate if the autocast is for unet.

1 files changed, 1 insertions, 1 deletions

diff --git a/modules/processing.py b/modules/processing.py
index 40598f5c..2df8a7ea 100644
--- a/modules/processing.py
+++ b/modules/processing.py
@@ -865,7 +865,7 @@ def process_images_inner(p: StableDiffusionProcessing) -> Processed:
             if p.n_iter > 1:

                 shared.state.job = f"Batch {n+1} out of {p.n_iter}"

 

-            with devices.without_autocast() if devices.unet_needs_upcast else devices.autocast():

+            with devices.without_autocast() if devices.unet_needs_upcast else devices.autocast(unet=True):

                 samples_ddim = p.sample(conditioning=p.c, unconditional_conditioning=p.uc, seeds=p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, prompts=p.prompts)

 

             if getattr(samples_ddim, 'already_decoded', False):