Public/automatic
1
0
Fork 0
mirror of https://github.com/vladmandic/automatic synced 2026-09-19 17:24:32 +02:00
Code Issues Packages Projects Releases Wiki Activity

Flux change no-half-vae check order

Browse Source
This commit is contained in:
Disty0
2024-08-22 03:01:48 +03:00
parent 3165bbc4f8
commit 35d70b35ce
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
modules/model_flux.py
+3 -3
View File
@@ -37,9 +37,6 @@ def load_quanto_text_encoder_2(repo_path):
return text_encoder_2
def load_flux(checkpoint_info, diffusers_load_config):
if devices.dtype == torch.float16 and not shared.opts.no_half_vae:
shared.log.warning("FLUX VAE doesn't support FP16! Enabling no-half-vae")
shared.opts.no_half_vae = True
if "qint8" in checkpoint_info.name.lower() or "qint4" in checkpoint_info.name.lower():
shared.log.debug(f'Loading FLUX: model="{checkpoint_info.name}" quant=True')
from installer import install
@@ -64,4 +61,7 @@ def load_flux(checkpoint_info, diffusers_load_config):
else:
pipe = diffusers.FluxPipeline.from_pretrained(checkpoint_info.path, cache_dir=shared.opts.diffusers_dir, **diffusers_load_config)
shared.log.debug(f'Loading FLUX: model="{checkpoint_info.name}" quant=False')
if devices.dtype == torch.float16 and not shared.opts.no_half_vae:
shared.log.warning("FLUX VAE doesn't support FP16! Enabling no-half-vae")
shared.opts.no_half_vae = True
return pipe