make temperature and max_length parameters

main.py

@@ -35,18 +35,19 @@ token_size_limit = 512
 app = FastAPI()
 
 
+# { msg: string, temperature: float, max_length: number }
 @app.post('/reply')
 async def Reply(req: Request):
     request = await req.json()
-    msg = request['msg']
+    msg = request.get('msg')
     print(f'MSG: {msg}')
 
     input_ids = tokenizer(msg, return_tensors='pt').input_ids  # .to('cuda')
     output = model.generate(
         input_ids[:, -token_size_limit:],
         do_sample=True,
-        temperature=0.9,
-        max_length=100,
+        temperature=request.get('temperature', 0.9),
+        max_length=request.get('max_length', 100),
     )
     reply = tokenizer.batch_decode(output)[0]
     print(f'REPLY: {reply}')