Remove ckpt loading when loading lora clip

coreml_suite/converter.py

@@ -277,11 +277,11 @@ def convert(
     pipe_cls = MODEL_TYPE_TO_PIPE_CLS[model_type]
     ref_pipe = pipe_cls.from_single_file(ckpt_path)
 
-    for lora_path in lora_paths:
-        ref_pipe.load_lora_weights(lora_path)
-        ref_pipe.fuse_lora()
-
     if not os.path.exists(unet_out_path):
+        for lora_path in lora_paths:
+            ref_pipe.load_lora_weights(lora_path)
+            ref_pipe.fuse_lora()
+
         convert_unet(
             ref_pipe,
             model_type,

coreml_suite/nodes.py

@@ -11,7 +11,6 @@ from comfy.model_patcher import ModelPatcher
 from coreml_suite import COREML_NODE
 from comfy.sd import CLIP
 from coreml_suite import converter
-from coreml_suite.clip import CoreMLCLIP, SDClipModelCoreML
 from coreml_suite.converter import ModelType
 from coreml_suite.lcm.utils import add_lcm_model_options, lcm_patch, is_lcm
 from coreml_suite.logger import logger
@@ -165,6 +164,7 @@ class CoreMLModelAdapter:
         return (model_patcher,)
 
 
+<<<<<<< HEAD
 class COREML_LOAD_CLIP(CoreMLLoader):
     PACKAGE_DIRNAME = "clip"
     RETURN_TYPES = ("CLIP",)
@@ -275,7 +275,7 @@ class COREML_CONVERT(COREML_NODE):
             out_path=unet_out_path, out_name=out_name, submodule_name="unet"
         )
 
-        _, clip = load_lora(lora_stack, ckpt_name)
+        clip = load_lora(lora_stack, ckpt_name)
 
         return (CoreMLModel(unet_target_path, compute_unit, "compiled"), clip)
 
@@ -290,9 +290,9 @@ def load_lora(lora_params, ckpt_name):
         ckpt_name.copy() if isinstance(ckpt_name, (list, dict, set)) else ckpt_name
     )
 
-    def recursive_load_lora(lora_params, ckpt, clip):
+    def recursive_load_lora(lora_params, clip):
         if len(lora_params) == 0:
-            return ckpt, clip
+            return clip
 
         lora_name, strength_model, strength_clip = lora_params[0]
         if os.path.isabs(lora_name):
@@ -300,21 +300,21 @@ def load_lora(lora_params, ckpt_name):
         else:
             lora_path = folder_paths.get_full_path("loras", lora_name)
 
-        lora_model, lora_clip = sd.load_lora_for_models(
-            ckpt, clip, utils.load_torch_file(lora_path), strength_model, strength_clip
+        lora_clip = sd.load_lora_for_models(
+            None, clip, utils.load_torch_file(lora_path), strength_model, strength_clip
         )
 
         # Call the function again with the new lora_model and lora_clip and the remaining tuples
-        return recursive_load_lora(lora_params[1:], lora_model, lora_clip)
+        return recursive_load_lora(lora_params[1:], lora_clip)
 
     ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name)
-    ckpt, clip, _, _ = sd.load_checkpoint_guess_config(
+    _, clip, _, _ = sd.load_checkpoint_guess_config(
         ckpt_path,
         output_vae=False,
         output_clip=True,
         embedding_directory=folder_paths.get_folder_paths("embeddings"),
     )
 
-    lora_model, lora_clip = recursive_load_lora(lora_params, ckpt, clip)
+    lora_clip = recursive_load_lora(lora_params, clip)
 
-    return lora_model, lora_clip
+    return lora_clip