Update controlnet_module.py

controlnet_module.py

@@ -3,8 +3,26 @@ from diffusers import StableDiffusionControlNetPipeline, ControlNetModel
 from controlnet_aux import OpenposeDetector
 from PIL import Image
 import random
-import cv2 #generiert Pose-Maske, geht auch mit matlibplot
+import cv2
 import numpy as np
+import gradio as gr
+
+class ControlNetProgressCallback:
+    def __init__(self, progress, total_steps):
+        self.progress = progress
+        self.total_steps = total_steps
+        self.current_step = 0
+    
+    def __call__(self, pipe, step_index, timestep, callback_kwargs):
+        self.current_step = step_index + 1
+        progress_percentage = self.current_step / self.total_steps
+        
+        # Fortschritt aktualisieren
+        if self.progress is not None:
+            self.progress(progress_percentage, desc=f"ControlNet: Schritt {self.current_step}/{self.total_steps}")
+        
+        print(f"ControlNet Fortschritt: {self.current_step}/{self.total_steps} ({progress_percentage:.1%})")
+        return callback_kwargs
 
 class ControlNetProcessor:
     def __init__(self, device="cuda", torch_dtype=torch.float32):
@@ -19,10 +37,8 @@ class ControlNetProcessor:
         if self.pose_detector is None:
             print("Loading Pose Detector...")
             try:
-                # OpenposeDetector ohne matplotlib Abhängigkeit
                 self.pose_detector = OpenposeDetector.from_pretrained(
                     "lllyasviel/ControlNet",
-                    #torch_dtype=self.torch_dtype
                 )
             except Exception as e:
                 print(f"Warnung: Pose-Detector konnte nicht geladen werden: {e}")
@@ -31,7 +47,6 @@ class ControlNetProcessor:
     def extract_pose_simple(self, image):
         """Einfache Pose-Extraktion ohne komplexe Abhängigkeiten"""
         try:
-            # Fallback: Einfache Kantenerkennung als Pose-Approximation
             img_array = np.array(image.convert("RGB"))
             edges = cv2.Canny(img_array, 100, 200)
             pose_image = Image.fromarray(edges).convert("RGB")
@@ -55,14 +70,17 @@ class ControlNetProcessor:
             return self.extract_pose_simple(image)
     
     def generate_with_controlnet(self, image, prompt, negative_prompt, 
-                               steps, guidance_scale, controlnet_strength):
-        """Generiert Bild mit ControlNet"""
+                               steps, guidance_scale, controlnet_strength, progress=None):
+        """Generiert Bild mit ControlNet und Fortschrittsanzeige"""
         try:
-            # Zuerst Pipeline laden um Fehler früh zu erkennen
+            # Pipeline laden
             pipe = self.load_controlnet_pipeline()
             
             # Pose extrahieren
             print("🔄 ControlNet: Extrahiere Pose...")
+            if progress:
+                progress(0.05, desc="ControlNet: Extrahiere Pose...")
+            
             pose_map = self.extract_pose(image)
             
             # Zufälliger Seed
@@ -70,8 +88,14 @@ class ControlNetProcessor:
             generator = torch.Generator(device=self.device).manual_seed(seed)
             print(f"ControlNet Seed: {seed}")
             
-            # ControlNet anwenden
+            # Progress Callback erstellen
+            callback = None
+            if progress is not None:
+                callback = ControlNetProgressCallback(progress, int(steps))
+            
             print("🔄 ControlNet: Wende Pose-Kontrolle an...")
+            
+            # ControlNet anwenden mit Callback
             result = pipe(
                 prompt=prompt,
                 image=pose_map,
@@ -82,15 +106,27 @@ class ControlNetProcessor:
                 controlnet_conditioning_scale=controlnet_strength,
                 height=512,
                 width=512,
-                output_type="pil"
+                output_type="pil",
+                callback_on_step_end=callback,
+                callback_on_step_end_tensor_inputs=[],
             )
             
+            # Debug-Ausgabe der tatsächlichen Steps
+            try:
+                scheduler = pipe.scheduler
+                if hasattr(scheduler, 'timesteps'):
+                    actual_steps = len(scheduler.timesteps)
+                    print(f"🎯 CONTROLNET TATSÄCHLICHE STEPS: {actual_steps} (von {steps} angefordert)")
+            except Exception as e:
+                print(f"⚠️  Konnte ControlNet Scheduler-Info nicht auslesen: {e}")
+            
             print("✅ ControlNet abgeschlossen!")
             return result.images[0]
             
         except Exception as e:
             print(f"❌ Fehler in ControlNet: {e}")
-            # Fallback: Originalbild zurückgeben
+            import traceback
+            traceback.print_exc()
             return image.convert("RGB").resize((512, 512))
 
     def load_controlnet_pipeline(self):