modify instantiation

app.py

@@ -6,8 +6,8 @@ For more information on `huggingface_hub` Inference API support, please check th
 """
 # requires space hardware update to use large models (TODO)
 # client = InferenceClient("mistralai/Mistral-Large-Instruct-2407")
-client = InferenceClient("EleutherAI/gpt-neo-125M")
-
+# Note change in instatiation***
+text_generator = pipeline("text-generation", model="EleutherAI/gpt-neo-125M")
 
 def respond(message, history, system_message, max_tokens, temperature, top_p):
     # Construct the prompt with system message, history, and user input
@@ -15,7 +15,7 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
     prompt += f"\nUser: {message}\nAssistant:"
 
     # Generate a response using the model
-    response = client(prompt, max_length=max_tokens, temperature=temperature, top_p=top_p, do_sample=True)[0]['generated_text']
+    response = text_generator(prompt, max_length=max_tokens, temperature=temperature, top_p=top_p, do_sample=True)
     
     # Extract the assistant's response part (after "Assistant:")
     assistant_response = response.split("Assistant:", 1)[-1].strip()