Update app.py

app.py

@@ -54,6 +54,67 @@ def extract_medicine_names(image):
         generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
     )[0]
     
+    # Remove <|im_end|> and any other special tokens that might appear in the output
+    output_text = output_text.replace("<|im_end|>", "").strip()
+    
+    return output_text
+
+# Create a singleton model and processor to avoid reloading for each request
+model_instance = None
+processor_instance = None
+
+def get_model_and_processor():
+    global model_instance, processor_instance
+    if model_instance is None or processor_instance is None:
+        model_instance, processor_instance = load_model()
+    return model_instance, processor_instance
+
+# Optimized extraction function that uses the singleton model
+def extract_medicine_names_optimized(image):
+    if image is None:
+        return "Please upload an image."
+    
+    model, processor = get_model_and_processor()
+    
+    # Prepare the message with the specific prompt for medicine extraction
+    messages = [
+        {
+            "role": "user",
+            "content": [
+                {
+                    "type": "image",
+                    "image": image,
+                },
+                {"type": "text", "text": "Extract and list ONLY the names of medicines/drugs from this prescription image. Output the medicine names as a numbered list without any additional information or descriptions."},
+            ],
+        }
+    ]
+    
+    # Prepare for inference
+    text = processor.apply_chat_template(
+        messages, tokenize=False, add_generation_prompt=True
+    )
+    image_inputs, video_inputs = process_vision_info(messages)
+    inputs = processor(
+        text=[text],
+        images=image_inputs,
+        videos=video_inputs,
+        padding=True,
+        return_tensors="pt",
+    )
+    
+    # Generate output
+    generated_ids = model.generate(**inputs, max_new_tokens=256)
+    generated_ids_trimmed = [
+        out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
+    ]
+    output_text = processor.batch_decode(
+        generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
+    )[0]
+    
+    # Remove <|im_end|> and any other special tokens that might appear in the output
+    output_text = output_text.replace("<|im_end|>", "").strip()
+    
     return output_text
 
 # Create Gradio interface
@@ -70,7 +131,7 @@ with gr.Blocks(title="Medicine Name Extractor") as app:
             output_text = gr.Textbox(label="Extracted Medicine Names", lines=10)
     
     extract_btn.click(
-        fn=extract_medicine_names,
+        fn=extract_medicine_names_optimized,
         inputs=input_image,
         outputs=output_text
     )