app.py

import gradio as gr
from huggingface_hub import InferenceClient
import random

client = InferenceClient("google/gemma-3-27b-it")
#change LLM here

#def echo(message, history):
    #choices = ["Ask me later", "Not today, sweetie", "One day we'll know", "Indeed!", "Nahhhh the vibes off cuh", "Cro don't do it", "Oh skibidi", "Ts pmo icl", "All of this is fake", "YES GET IT GURL"]
    #value = random.choice(choices)
    #return value

def respond(message, history):
    messages = [{"role":  "system", "content": "be a happy and sassy single mom from boston who divorced her 3rd husband (thank god for that)"}]
    #change content value to change personality
    if history:
        messages.extend(history)
        #adds multiple value + types rather than just one
    messages.append({"role": "user", "content": message})
    response = ""
    for messages in client.chat_completion(messages, max_tokens=10000, stream=True):
        token=messages.choices[0].delta.content
        response+=token
        yield response
    
    #change max tokens to a greater number
    
    #return response['choices'][0]['message']['content'].strip()
                    
chatbot = gr.ChatInterface(respond, type = "messages")
        

chatbot.launch(debug=True)