app.py

import gradio as gr
import random
from huggingface_hub import InferenceClient

gr.themes.builder(
    primary_hue = colors.indigo,
    secondary_hue = colors.rose,
    neutral_hue = colors.gray
)

client = InferenceClient ("HuggingFaceH4/zephyr-7b-beta") #change the LLM

def respond(message, history):
    messages = [{"role" : "system", "content" : "You are a chatbot who helps with mental health"}] #change personality
    if history: 
        messages.extend(history)
        
    messages.append({"role" : "user", "content" : message})
    
    response = ""
    for message in client.chat_completion(
        messages, 
        max_tokens = 100, #change length
        stream =True
    ):
        token = message.choices[0].delta.content
        response += token
        print(type(message))
        print(response)
        yield response
    #print(response["choices"][0]["message"]["content"].strip())
    
    
def random_message(message, history):
    choices = ["try again later", "isn't looking good for you", "perhaps", "maybe", "doubtful", "yes", "no", "without a doubt", "most likely", "signs point to yes", "it is certain", "my sources say no", "reply hazy, try again"]
    random_choices = random.choice(choices)
    return random_choices

chatbot = gr.ChatInterface(respond, type = "messages")

chatbot.launch(debug=True)