Load .yaml config if present

coreml_suite/converter.py

@@ -278,6 +278,7 @@ def convert(
     controlnet_support: bool = False,
     lora_weights: list[tuple[str | os.PathLike, float]] = None,
     attn_impl: str = AttentionImplementations.SPLIT_EINSUM.name,
+    config_path: str = None,
 ):
     if os.path.exists(unet_out_path):
         logger.info(f"Found existing model at {unet_out_path}! Skipping..")
@@ -290,7 +291,7 @@ def convert(
     model_type = ModelType.SD15
 
     pipe_cls = MODEL_TYPE_TO_PIPE_CLS[model_type]
-    ref_pipe = pipe_cls.from_single_file(ckpt_path)
+    ref_pipe = pipe_cls.from_single_file(ckpt_path, original_config_file=config_path)
 
     for i, lora_weight in enumerate(lora_weights or []):
         lora_path, strength = lora_weight

coreml_suite/nodes.py

@@ -265,9 +265,10 @@ class COREML_CONVERT(COREML_NODE):
         unet_out_path = converter.get_out_path("unet", f"{out_name}")
         ckpt_path = folder_paths.get_full_path("checkpoints", ckpt_name)
 
-        lora_paths = [
-            folder_paths.get_full_path("loras", lora[0]) for lora in lora_params
-        ]
+        config_filename = ckpt_name.split(".")[0] + ".yaml"
+        config_path = folder_paths.get_full_path("configs", config_filename)
+        if config_path:
+            logger.info(f"Using config file {config_path}")
 
         converter.convert(
             ckpt_path=ckpt_path,
@@ -277,6 +278,7 @@ class COREML_CONVERT(COREML_NODE):
             controlnet_support=controlnet_support,
             lora_weights=lora_weights,
             attn_impl=attention_implementation,
+            config_path=config_path,
         )
         unet_target_path = converter.compile_model(
             out_path=unet_out_path, out_name=out_name, submodule_name="unet"