Update app.py

app.py

@@ -33,14 +33,13 @@ repetition_penalty=1.7
 
 tokenizer = transformers.AutoTokenizer.from_pretrained(model_name)
 model = transformers.AutoModelForCausalLM.from_pretrained(model_name,
-    device_map="auto"
-#   torch_dtype=torch.bfloat16,
-#   load_in_4bit=True
+    device_map=device,
+    torch_dtype=torch.bfloat16
 )
-# model.eval()
+model.eval()
 
 class StarlingBot:
-    def __init__(self, system_prompt="The following dialogue is a conversation"):
+    def __init__(self, system_prompt="I am Starling-7B by Tonic-AI, I ready to do anything to help my user."):
         self.system_prompt = system_prompt
 
     def predict(self, user_message, assistant_message, system_prompt, do_sample, temperature=0.4, max_new_tokens=700, top_p=0.99, repetition_penalty=1.9):
@@ -75,7 +74,6 @@ iface = gr.Interface(
     fn=starling_bot.predict,
     title=title,
     description=description,
-#   examples=examples,
     inputs=[
         gr.Textbox(label="🌟🤩User Message", type="text", lines=5),
         gr.Textbox(label="💫🌠Starling Assistant Message or Instructions ", lines=2),
@@ -87,5 +85,5 @@ iface = gr.Interface(
         gr.Slider(label="Repetition penalty", value=1.9, minimum=1.0, maximum=2.0, step=0.05)
     ],
     outputs="text",
-#   theme="ParityError/Anime"
+    theme="ParityError/Anime"
 )