Upload app.py

app.py

@@ -12,7 +12,7 @@ DEFAULT_MAX_NEW_TOKENS = 1024
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 ACCESS_TOKEN = os.getenv("HF_TOKEN", "")
 
-model_id = "Qwen/Qwen2.5-7B-Instruct"
+model_id = "Qwen/Qwen2.5-14B-Instruct-GPTQ-Int8"
 model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto", token=ACCESS_TOKEN)
 tokenizer = AutoTokenizer.from_pretrained(model_id, token=ACCESS_TOKEN)
 tokenizer.use_default_system_prompt = False
@@ -104,8 +104,8 @@ chat_interface = gr.Interface(
             value=1.2,
         ),
     ],
-    title="Chat with the Model",
-    description="Provide system settings and a message to interact with the model.",
+    title="Model testing",
+    description="Provide system settings and a prompt to interact with the model.",
 )
 
 chat_interface.queue(max_size=20).launch(share = True)