Update app.py

app.py

@@ -10,7 +10,7 @@ headers = {"Authorization": f"Bearer {HF_TOKEN}"}
 client = AsyncInferenceClient(api_url)
 
 
-system_message = """
+system_prompt = """
 ### Instruction:
 Refactor the provided Python code to improve its maintainability and efficiency and reduce complexity. Include the refactored code along with the comments on the changes made for improving the metrics.
 ### Input:
@@ -93,8 +93,7 @@ def analyze_sales_data(sales_records):
 # Stream text - stream tokens with InferenceClient from TGI
 async def predict(message, chatbot,  temperature=0.1, max_new_tokens=4096, top_p=0.6, repetition_penalty=1.15,):
     
-    if system_prompt != "":
-        input_prompt = f"{system_prompt}"
+    
     
     temperature = float(temperature)
     if temperature < 1e-2:
@@ -102,7 +101,7 @@ async def predict(message, chatbot,  temperature=0.1, max_new_tokens=4096, top_p
     top_p = float(top_p)
     
   
-    input_prompt = input_prompt + str(message) + " [/INST] "
+    input_prompt = system_prompt + str(message) + " [/INST] "
 
     partial_message = ""
     async for token in await client.text_generation(prompt=input_prompt,