fix bug

Author: billwuhao

Diff for: NotaGenNode.py

@@ -120,10 +120,11 @@ def inference_patch(self, model, period, composer, instrumentation,
         nota_model_path = os.path.join(self.nota_model_path, model)
         if self.model_cache is None:
             checkpoint = torch.load(nota_model_path, map_location=torch.device(self.device))
-            self.model_cache = checkpoint
-        else:
-            checkpoint = self.model_cache
-        nota_model.load_state_dict(checkpoint['model'])
+            NotaGenRun.model_cache = checkpoint
+            del checkpoint
+            torch.cuda.empty_cache()
+            
+        nota_model.load_state_dict(self.model_cache['model'])
         nota_model = nota_model.to(self.device)
         nota_model.eval()
 
@@ -338,9 +339,8 @@ def inference_patch(self, model, period, composer, instrumentation,
             if unload_model:
                 del patchilizer
                 del nota_model
-                del checkpoint
+                NotaGenRun.model_cache = None
                 torch.cuda.empty_cache()
-                self.model_cache = None
 
             return (
                 audio,
@@ -352,9 +352,9 @@ def inference_patch(self, model, period, composer, instrumentation,
             if unload_model:
                 del patchilizer
                 del nota_model
-                del checkpoint
+                NotaGenRun.model_cache = None
                 torch.cuda.empty_cache()
-                self.model_cache = None
+                
             print(f".abc and .xml was saved to {INTERLEAVED_OUTPUT_FOLDER} and {ORIGINAL_OUTPUT_FOLDER}")
             raise Exception("Conversion of .mp3 and .png failed, try again or check if MuseScore4 installation was successful.")
 

Diff for: pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "notagen-mw"
 description = "Symbolic Music Generation, NotaGen node for ComfyUI."
-version = "2.2.2"
+version = "2.2.3"
 license = {file = "LICENSE"}
 
 [project.urls]