The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Support loading checkpoints with the UNETLoader node.
comfyanonymousd7484ef30c97
diff --git a/comfy/sd.py b/comfy/sd.pyindex 4ba00a62c..e343e1fa8 100644--- a/comfy/sd.py+++ b/comfy/sd.py@@ -563,24 +563,32 @@ def load_checkpoint_guess_config(ckpt_path, output_vae=True, output_clip=True, oreturn (model_patcher, clip, vae, clipvision)-def load_unet_state_dict(sd): #load unet in diffusers format+def load_unet_state_dict(sd): #load unet in diffusers or regular format++ #Allow loading unets from checkpoint files+ checkpoint = False+ diffusion_model_prefix = model_detection.unet_prefix_from_state_dict(sd)+ temp_sd = comfy.utils.state_dict_prefix_replace(sd, {diffusion_model_prefix: ""}, filter_keys=True)+ if len(temp_sd) > 0:+ sd = temp_sd+ checkpoint = True+parameters = comfy.utils.calculate_parameters(sd)unet_dtype = model_management.unet_dtype(model_params=parameters)load_device = model_management.get_torch_device()- if 'transformer_blocks.0.attn.add_q_proj.weight' in sd: #MMDIT SD3+ if checkpoint or "input_blocks.0.0.weight" in sd or 'clf.1.weight' in sd: #ldm or stable cascade+ model_config = model_detection.model_config_from_unet(sd, "")+ if model_config is None:+ return None+ new_sd = sd+ elif 'transformer_blocks.0.attn.add_q_proj.weight' in sd: #MMDIT SD3new_sd = model_detection.convert_diffusers_mmdit(sd, "")if new_sd is None:return Nonemodel_config = model_detection.model_config_from_unet(new_sd, "")if model_config is None:return None- elif "input_blocks.0.0.weight" in sd or 'clf.1.weight' in sd: #ldm or stable cascade- model_config = model_detection.model_config_from_unet(sd, "")- if model_config is None:- return None- new_sd = sd-else: #diffusersmodel_config = model_detection.model_config_from_diffusers_unet(sd)if model_config is None: