Update used model

app.py

@@ -7,7 +7,7 @@ if True:
         subprocess.check_call([sys.executable, "-m", "pip", "install", package_name])
 
     # Example usage:
-    install_package("argilla==2.0.0rc1")
+    install_package("argilla==2.0.0s")
 
 import os
 from threading import Thread
@@ -26,10 +26,9 @@ DEFAULT_MAX_NEW_TOKENS = 1024
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "8192"))
 
 if torch.cuda.is_available():
-    model_id = "Qwen/Qwen2-0.5B-Instruct"
+    model_id = "davidberenstein1957/ultra-feedback-dutch-cleaned-hq-spin-geitje-7b-ultra-sft_iter2"
     model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
     tokenizer = AutoTokenizer.from_pretrained(model_id)
-
 style = "<style>.user-message,.system-message{display:flex;margin:10px}.user-message .message-content{background-color:#c2e3f7;color:#000}.system-message .message-content{background-color:#f5f5f5;color:#000}.message-content{padding:10px;border-radius:10px;max-width:70%;word-wrap:break-word}.container{display:flex;justify-content:space-between}.column{width:48%}</style>"
 
 client = rg.Argilla(api_url="https://davidberenstein1957-argilla-gradio.hf.space", api_key="owner.apikey")