Update app.py

app.py

@@ -50,18 +50,18 @@ def process_and_describe(payload: dict):
         # 1️⃣ Save & upload image
         local_tmp_path, hf_url, path_in_repo, size_bytes = save_and_upload_image(image_b64)
 
-        # 2️⃣ Prepare prompt
-        prompt = "Describe this image in detail."
+        # 2️⃣ Prepare prompt (optional, some models ignore this for basic image_to_text)
+        # prompt = "Describe this image in detail." 
         
         # Open the image using PIL for the InferenceClient
         image = Image.open(local_tmp_path)
 
         # 3️⃣ Call VLM using Hugging Face Inference Client
-        # The client automatically handles the API call and authentication
+        # Removed the problematic 'details' argument
         vlm_text = hf_client.image_to_text(
             image=image,
             model=HF_VLM_MODEL,
-            details=True, # Set details=True for more comprehensive output if available
+            # details=True # <-- REMOVED THIS LINE
         )
 
         return {