ostris/ai-toolkit
1
0
Fork 0
mirror of https://github.com/ostris/ai-toolkit.git synced 2026-03-02 09:09:48 +00:00
Code Issues Packages Projects Releases Wiki Activity

Added ability to use civit ai url ar model name and built a model downloader and cache manager for it

Browse Source
This commit is contained in:
Jaret Burkett
2023-08-13 13:09:51 -06:00
parent 1487d13191
commit 355c80df07
4 changed files with 236 additions and 217 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
toolkit/stable_diffusion_model.py
View File

@@ -143,6 +143,13 @@ class StableDiffusion:
prediction_type=prediction_type,
steps_offset=1
)
model_path = self.model_config.name_or_path
if 'civitai.com' in self.model_config.name_or_path:
# load is a civit ai model, use the loader.
from toolkit.civitai import get_model_path_from_url
model_path = get_model_path_from_url(self.model_config.name_or_path)
if self.model_config.is_xl:
if self.custom_pipeline is not None:
pipln = self.custom_pipeline
@@ -150,17 +157,17 @@ class StableDiffusion:
pipln = CustomStableDiffusionXLPipeline
# see if path exists
if not os.path.exists(self.model_config.name_or_path):
if not os.path.exists(model_path):
# try to load with default diffusers
pipe = pipln.from_pretrained(
self.model_config.name_or_path,
model_path,
dtype=dtype,
scheduler_type='ddpm',
device=self.device_torch,
).to(self.device_torch)
else:
pipe = pipln.from_single_file(
self.model_config.name_or_path,
model_path,
dtype=dtype,
scheduler_type='ddpm',
device=self.device_torch,
@@ -180,10 +187,10 @@ class StableDiffusion:
pipln = CustomStableDiffusionPipeline
# see if path exists
if not os.path.exists(self.model_config.name_or_path):
if not os.path.exists(model_path):
# try to load with default diffusers
pipe = pipln.from_pretrained(
self.model_config.name_or_path,
model_path,
dtype=dtype,
scheduler_type='dpm',
device=self.device_torch,
@@ -193,7 +200,7 @@ class StableDiffusion:
).to(self.device_torch)
else:
pipe = pipln.from_single_file(
self.model_config.name_or_path,
model_path,
dtype=dtype,
scheduler_type='dpm',
device=self.device_torch,