Update app.py

app.py

@@ -22,18 +22,18 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 HF_TOKEN = os.getenv("HF_TOKEN")
 device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 
-model_name = "fluently-sets/FalconThink3-10B-IT"
+model_name = "ehristoforu/rufalcon3-3b-it"
 
 model = AutoModelForCausalLM.from_pretrained(
     model_name,
     torch_dtype=torch.float16,
     device_map="auto"
 )
-tokenizer = AutoTokenizer.from_pretrained("fluently-sets/FalconThink3-10B-IT")
+tokenizer = AutoTokenizer.from_pretrained("ehristoforu/rufalcon3-3b-it")
 
 
 
-@spaces.GPU(duration=90)
+@spaces.GPU(duration=60)
 def generate(
     message: str,
     chat_history: list[tuple[str, str]],