app.py

import gradio as gr
from huggingface_hub import InferenceClient


def respond(
    message,
    history: list[dict[str, str]],
    max_tokens,
    temperature,
    top_p,
    hf_token: gr.OAuthToken,
):
    client = InferenceClient(
        token=hf_token.token,
        model="TheDrummer/Tiger-Gemma-9B-v3"
    )

    # Tiger-Gemma does NOT support system prompt, so skip it
    messages = []
    messages.extend(history)
    messages.append({"role": "user", "content": message})

    response = ""

    for chunk in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        choices = chunk.choices
        token = ""
        if len(choices) and choices[0].delta.content:
            token = choices[0].delta.content

        response += token
        yield response


chatbot = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p"),
    ],
    title="Tiger-Gemma 9B Chat 🐯",
)

with gr.Blocks() as demo:
    with gr.Sidebar():
        gr.LoginButton()
    chatbot.render()


if __name__ == "__main__":
    demo.launch()