app.py

import gradio as gr
import torch
from transformers import pipeline

generate_text = pipeline(model="databricks/dolly-v2-3b", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
theme = gr.themes.Soft(
    primary_hue=gr.themes.Color("#ededed", "#fee2e2", "#fecaca", "#fca5a5", "#f87171", "#ef4444", "#dc2626", "#b91c1c", "#991b1b", "#7f1d1d", "#6c1e1e"),
    neutral_hue="red",
)
title = """<h1 align="center">Chat with awesome Dolly V2 3B model</h1><br><h2 align="center">🏃‍♂️💨Streaming with Transformers & Gradio💪</h2>"""
with gr.Blocks(theme=theme) as demo:
    gr.HTML(title)
    gr.HTML("This is run in cpu, so it might be a bit slow. Also, the model is great with language but might fail answering questions with numbers. If you want to skip waiting time, just clone the app.")
    Dolly = gr.Chatbot().style(height=500)
    msg = gr.Textbox()
    clear = gr.Button("Clear")

    def respond(message, chat_history):
        res = generate_text(message)
        a = (res[0]["generated_text"])
        b = str(a)
        bot_message = b
        chat_history.append((message, bot_message))
        return "", chat_history

    msg.submit(respond, [msg, Dolly], [msg, Dolly])
    clear.click(lambda: None, None, Dolly, queue=False)
demo.launch(debug=True)