Update app.py

app.py

@@ -118,30 +118,32 @@ class TextToImageProgressCallback:
         return False
 
 class ImageToImageProgressCallback:
-    def __init__(self, progress, total_steps):
-        self.progress = progress
-        self.total_steps = total_steps
-        self.current_step = 0
-        self.actual_total_steps = None  # Neue Variable für tatsächliche Steps
-    
+    def __init__(self, progress, total_steps, strength):
+    self.progress = progress
+    self.total_steps = total_steps
+    self.current_step = 0
+    self.strength = strength
+    self.actual_total_steps = None
+
+
     def __call__(self, step, timestep, latents):
-        self.current_step = step + 1
-        
-        # ERGÄNZUNG: Tatsächliche Gesamtschritte aus dem Scheduler lesen
-        if self.actual_total_steps is None:
-            # Versuche, die tatsächliche Schrittanzahl zu ermitteln
-            try:
-                # Der Scheduler hat normalerweise config.num_train_timesteps oder ähnliches
-                self.actual_total_steps = self.total_steps
-                print(f"🎯 INTERNE STEP-AUSGABE: Scheduler führt {self.actual_total_steps} Schritte durch")
-            except:
-                self.actual_total_steps = self.total_steps
-                print(f"🎯 INTERNE STEP-AUSGABE: Verwende konfigurierte {self.total_steps} Schritte")
+    self.current_step = step + 1
+    
+    # Korrekte Berechnung der tatsächlichen Steps
+    if self.actual_total_steps is None:
+        # Bei Strength < 1.0 werden weniger Steps verwendet
+        if self.strength < 1.0:
+            self.actual_total_steps = int(self.total_steps * self.strength)
+        else:
+            self.actual_total_steps = self.total_steps
         
-        progress_percent = (step / self.total_steps) * 100
-        # AUSGABE ZEILE - Hier werden die Steps ausgegeben:
-        self.progress(progress_percent / 100, desc=f"Schritt {self.current_step}/{self.total_steps} - {progress_percent:.1f}%")
-        return False
+        print(f"🎯 INTERNE STEP-AUSGABE: Strength {self.strength} → {self.actual_total_steps} tatsächliche Denoising-Schritte")
+    
+    progress_percent = (step / self.actual_total_steps) * 100
+    self.progress(progress_percent / 100, desc=f"Schritt {self.current_step}/{self.actual_total_steps} - {progress_percent:.1f}%")
+    return False
+    
+    
 
 # === FUNKTIONEN ===
 def text_to_image(prompt, steps, guidance_scale, progress=gr.Progress()):
@@ -241,20 +243,23 @@ def img_to_image(image, prompt, neg_prompt, strength, steps, guidance_scale, fac
             print("Keine gültigen Koordinaten - keine Maske angewendet")
             mask = None
 
+
         # Detaillierte Debug-Informationen vor dem Pipeline-Aufruf
         print(f"⚙️  PIPELINE-KONFIGURATION:")
         print(f"   - Angefordert: {int(steps)} Steps")
-        print(f"   - Strength: {adj_strength}")
+        print(f"   - Strength: {adj_strength:.3f}")
         print(f"   - Scheduler: {pipe.scheduler.__class__.__name__}")
 
-        # Berechne die tatsächlichen Steps basierend auf Strength
-        if hasattr(pipe.scheduler, 'config') and hasattr(pipe.scheduler.config, 'num_train_timesteps'):
-            total_train_steps = pipe.scheduler.config.num_train_timesteps
-            actual_inference_steps = int(adj_strength * total_train_steps)
-            print(f"🎯 INTERNE STEP-AUSGABE: Strength {adj_strength} → {actual_inference_steps} tatsächliche Denoising-Schritte")
+        # Korrekte Berechnung der tatsächlichen Steps
+        actual_steps_from_strength = int(steps * adj_strength)
+        print(f"🎯 KORREKTE INTERNE STEP-AUSGABE: {int(steps)} Steps × Strength {adj_strength:.3f} = {actual_steps_from_strength} tatsächliche Schritte")
+
 
+
+        # Progress Callback hinzufügen
+        callback = ImageToImageProgressCallback(progress, int(steps), adj_strength)
         # Progress Callback hinzufügen
-        callback = ImageToImageProgressCallback(progress, int(steps))
+        #callback = ImageToImageProgressCallback(progress, int(steps))
 
         # --- PIPELINE-AUFRUF ---
         result = pipe(
@@ -270,20 +275,21 @@ def img_to_image(image, prompt, neg_prompt, strength, steps, guidance_scale, fac
             callback_steps=1
         )
 
+
         # ZUSÄTZLICHE AUSGABE: Tatsächliche Steps vom Scheduler
         try:
-            # Der Scheduler hat Informationen über die tatsächliche Schrittanzahl
             scheduler = pipe.scheduler
             print(f"🔧 SCHEDULER-INFO: {scheduler.__class__.__name__}")
-            print(f"📊 TATSÄCHLICHE STEP-KONFIGURATION: {int(steps)} Schritte mit Strength {adj_strength}")
-            
-            # Bei einigen Schedulern kann man die Timesteps sehen
+            print(f"📊 TATSÄCHLICHE STEP-KONFIGURATION: {int(steps)} Schritte mit Strength {adj_strength:.3f}")
+    
             if hasattr(scheduler, 'timesteps'):
-                actual_steps = len(scheduler.timesteps)
-                print(f"🎯 INTERNE STEP-AUSGABE: Scheduler verwendete {actual_steps} tatsächliche Denoising-Schritte")
-                
+            actual_steps = len(scheduler.timesteps)
+            print(f"🎯 BESTÄTIGTE INTERNE STEP-AUSGABE: Scheduler verwendete {actual_steps} tatsächliche Denoising-Schritte")
+        
         except Exception as e:
             print(f"⚠️  Konnte Scheduler-Info nicht auslesen: {e}")
+
+
         
         end_time = time.time()
         print(f"Bild transformiert in {end_time - start_time:.2f} Sekunden")
@@ -607,4 +613,4 @@ if __name__ == "__main__":
         server_port=7860,
         max_file_size="10MB",
         show_error=True
-    )
+    )