app.py

import gradio as gr
import os
from huggingface_hub import InferenceClient

hf_token = os.getenv("hf_token")

client = InferenceClient(api_key=hf_token)

def get_response(user_input):
    messages = [
        { "role": "system", "content": "you are xylaria 1.4 senoa, developed by sk md saad amin. You give links for images if the user asks for images." },
        { "role": "user", "content": user_input }
    ]
    
    stream = client.chat.completions.create(
        model="Qwen/QwQ-32B-Preview", 
        messages=messages, 
        temperature=0.5,
        max_tokens=10240,
        top_p=0.7,
        stream=True
    )
    
    response = ""
    for chunk in stream:
        response += chunk.choices[0].delta.content
    return response

def chat_interface():
    with gr.Blocks() as demo:
        with gr.Row():
            with gr.Column(scale=0.8):
                input_textbox = gr.Textbox(
                    label="Type your message", 
                    placeholder="Type to Xylaria...",
                    lines=1,
                    max_lines=3,
                    interactive=True,
                    elem_id="user-input",
                    show_label=False
                )
            with gr.Column(scale=0.2):
                send_button = gr.Button("Send", elem_id="send-btn")

        chat_output = gr.Chatbot(
            elem_id="chat-box",
            label="Xylaria 1.4 Senoa Chatbot",
            show_label=False
        )
        
        def submit_input(user_input, chat_history):
            response = get_response(user_input)
            chat_history.append((user_input, response))
            return "", chat_history

        input_textbox.submit(submit_input, [input_textbox, chat_output], [input_textbox, chat_output])
        send_button.click(submit_input, [input_textbox, chat_output], [input_textbox, chat_output])
        
    return demo

demo = chat_interface()
demo.launch()