Update app.py

app.py

@@ -17,14 +17,16 @@ tokenizer = AutoTokenizer.from_pretrained(model_name)
 
 @spaces.GPU
 def generate(message: str, chat_history: list[tuple[str, str]], temperature=0.7, top_p=0.9, top_k=50, max_new_tokens=512, repetition_penalty=1.1) -> Iterator[str]:
-    """Generates text responses using Zurich model with streaming."""
+    messages = [
+        {"role": "system", "content": "You are a helpful assistant named Zurich, a 7 billion parameter Large Language Model, fine-tuned and trained by Ruben Roy. You have been trained with the GammaCorpus v2 dataset, a structured and filtered multi-turn conversation dataset created by Ruben Roy."}
+    ]
     
-    conversation = []
     for user, assistant in chat_history:
-        conversation.extend([{"role": "user", "content": user}, {"role": "assistant", "content": assistant}])
-    conversation.append({"role": "user", "content": message})
+        messages.append({"role": "user", "content": user})
+        messages.append({"role": "assistant", "content": assistant})
+    messages.append({"role": "user", "content": message})
     
-    input_ids = tokenizer.apply_chat_template(conversation, return_tensors="pt")
+    input_ids = tokenizer.apply_chat_template(messages, return_tensors="pt")
     
     if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
         input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]