Update app.py

app.py

@@ -35,7 +35,7 @@ print("[INFO] Loading Florence-2-base model on CPU...")
 try:
     processor = AutoProcessor.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)
     model = AutoModelForCausalLM.from_pretrained(
-        "microsoft/Florence-2-base",
+        "microsoft/Florence-2-Large",
         trust_remote_code=True,
         attn_implementation="eager"
     ).to(DEVICE).eval()
@@ -95,13 +95,14 @@ def analyze_image(image: Image.Image, task: str = "<MORE_DETAILED_CAPTION>") ->
             generated_ids = model.generate(
                 input_ids=inputs["input_ids"],
                 pixel_values=inputs["pixel_values"],
-                max_new_tokens=1024,
+                max_new_tokens=1024,       # Maximum token length
                 num_beams=3,
-                do_sample=False
+                do_sample=False,
+                repetition_penalty=1.2     # Helps avoid repetitive outputs
             )
         
         # Decode and post-process
-        generated_text = processor.batch_decode(generated_ids, skip_special_tokens=False)[0]
+        generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
         result = processor.post_process_generation(
             generated_text,
             task=task,
@@ -112,7 +113,7 @@ def analyze_image(image: Image.Image, task: str = "<MORE_DETAILED_CAPTION>") ->
         
     except Exception as e:
         raise ValueError(f"Failed to analyze image: {e}")
-
+        
 # ===== API Endpoints =====
 @app.get("/")
 async def root():