xinyun/ComfyUI-KJNodes
1
0
Fork 0
mirror of https://git.datalinker.icu/kijai/ComfyUI-KJNodes.git synced 2025-12-10 13:24:44 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update nodes.py

Browse Source
This commit is contained in:
kijai 2024-05-09 02:56:55 +03:00
parent 928e5cc778
commit 46e91987be
1 changed files with 14 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
nodes/nodes.py
View File

@ -1646,7 +1646,7 @@ class LoadICLightUnet:
CATEGORY = "KJNodes/experimental" CATEGORY = "KJNodes/experimental"
def load(self, model, model_path): def load(self, model, model_path):
print("LoadICLightUnet: Checking ResAdapter path") print("LoadICLightUnet: Checking LoadICLightUnet path")
model_full_path = folder_paths.get_full_path("unet", model_path) model_full_path = folder_paths.get_full_path("unet", model_path)
if not os.path.exists(model_full_path): if not os.path.exists(model_full_path):
raise Exception("Invalid model path") raise Exception("Invalid model path")
@ -1657,6 +1657,7 @@ class LoadICLightUnet:
conv_layer = model_clone.model.diffusion_model.input_blocks[0][0] conv_layer = model_clone.model.diffusion_model.input_blocks[0][0]
print(f"Current number of input channels: {conv_layer.in_channels}") print(f"Current number of input channels: {conv_layer.in_channels}")
# Create a new Conv2d layer with 8 input channels # Create a new Conv2d layer with 8 input channels
new_conv_layer = torch.nn.Conv2d(8, conv_layer.out_channels, kernel_size=conv_layer.kernel_size, stride=conv_layer.stride, padding=conv_layer.padding) new_conv_layer = torch.nn.Conv2d(8, conv_layer.out_channels, kernel_size=conv_layer.kernel_size, stride=conv_layer.stride, padding=conv_layer.padding)
new_conv_layer.weight.zero_() new_conv_layer.weight.zero_()
@ -1668,7 +1669,19 @@ class LoadICLightUnet:
model_clone.model.diffusion_model.input_blocks[0][0] = new_conv_layer model_clone.model.diffusion_model.input_blocks[0][0] = new_conv_layer
# Verify the change # Verify the change
print(f"New number of input channels: {model_clone.model.diffusion_model.input_blocks[0][0].in_channels}") print(f"New number of input channels: {model_clone.model.diffusion_model.input_blocks[0][0].in_channels}")
# Monkey patch because I don't know what I'm doing
from comfy.model_base import IP2P
import types
base_model_instance = model_clone.model
# Dynamically add the extra_conds method from IP2P to the instance of BaseModel
def bound_extra_conds(self, **kwargs):
return IP2P.extra_conds(self, **kwargs)
base_model_instance.process_ip2p_image_in = lambda image: image
base_model_instance.extra_conds = types.MethodType(bound_extra_conds, base_model_instance)
# Some Proper patching
new_state_dict = load_torch_file(model_full_path) new_state_dict = load_torch_file(model_full_path)
prefix_to_remove = 'model.' prefix_to_remove = 'model.'
new_keys_dict = {key[len(prefix_to_remove):]: new_state_dict[key] for key in new_state_dict if key.startswith(prefix_to_remove)} new_keys_dict = {key[len(prefix_to_remove):]: new_state_dict[key] for key in new_state_dict if key.startswith(prefix_to_remove)}
@ -1677,11 +1690,9 @@ class LoadICLightUnet:
try: try:
for key in new_keys_dict: for key in new_keys_dict:
model_clone.add_patches({key: (new_keys_dict[key],)}, 1.0, 1.0) model_clone.add_patches({key: (new_keys_dict[key],)}, 1.0, 1.0)
#print(f"Added patch for: {key}")
except: except:
raise Exception("Could not patch model") raise Exception("Could not patch model")
print("LoadICLightUnet: Added LoadICLightUnet patches") print("LoadICLightUnet: Added LoadICLightUnet patches")
#model_clone.model.diffusion_model.in_channels = 8
return (model_clone, ) return (model_clone, )