Update app.py

response = llm.chat.completions.create(
model=GRANITE_MODEL,
messages=[
{"role": "user", "content": prompt}
],
max_tokens=512,
temperature=0.2,
top_p=0.9,
)
return response.choices[0].message["content"].strip()

app.py

@@ -101,15 +101,17 @@ def generate_answer(question, context):
             f"Question: {question}\nAnswer:"
         )
 
-    response = llm.text_generation(
-        prompt,
-        max_new_tokens=512,
+    response = llm.chat.completions.create(
+        model=GRANITE_MODEL,
+        messages=[
+            {"role": "user", "content": prompt}
+        ],
+        max_tokens=512,
         temperature=0.2,
         top_p=0.9,
-        repetition_penalty=1.05,
     )
 
-    return response.strip()
+    return response.choices[0].message["content"].strip()
 
 
 # ============================================================