change chat completion to text to text

app.py

@@ -13,29 +13,25 @@ def respond(
     hf_token: gr.OAuthToken,
 ):
     """
-    Sends a chat message to the Hugging Face Inference API using the provided token and parameters.
+    Sends a user input to the summarization model using text-to-text interface.
     """
     client = InferenceClient(
         token=hf_token.token,
         model="Bocklitz-Lab/lit2vec-tldr-bart-model"
     )
 
-    messages = [{"role": "system", "content": system_message}] + history
-    messages.append({"role": "user", "content": message})
+    # You can prepend the system message if needed
+    input_text = f"{system_message}\n\n{message}"
 
-    response = ""
-
-    for message_chunk in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    response = client.text_to_text(
+        input=input_text,
+        max_new_tokens=max_tokens,
         temperature=temperature,
-        top_p=top_p,
-    ):
-        if message_chunk.choices and message_chunk.choices[0].delta.content:
-            token = message_chunk.choices[0].delta.content
-            response += token
-            yield response
+        top_p=top_p
+    )
+
+    yield response
+
 
 # Define the Gradio interface
 chatbot = gr.ChatInterface(