Update app.py

app.py

@@ -94,11 +94,11 @@ async def get_answer(q: QueryM ):
     try:
         client = InferenceClient()
         generate_kwargs = dict(
-        max_new_tokens= 256,
+        max_new_tokens= int(q.tokens),
         do_sample=True,
-        top_p= 0.9,
-        top_k=10,
-        temperature=0.9,
+        top_p= q.topp,
+        top_k=int(q.topk),
+        temperature=q.temp,
         )
         inputs= text
         response = client.post(json={"inputs": inputs, "parameters": generate_kwargs}, model="mistralai/Mistral-7B-Instruct-v0.1")