app.py

from huggingface_hub import InferenceClient
import gradio as gr

client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.1")

val_image = gr.Image("/file=val_speaking_transparent.gif")

PLACEHOLDER = f"""
<div style="padding: 30px; text-align: center; display: flex; flex-direction: column; align-items: center;">
   <h1 style="font-size: 28px; margin-bottom: 2px; opacity: 0.55;">Hi Jennifer, welcome to Treasury and Finance</h1>
   <p style="font-size: 18px; margin-bottom: 2px; opacity: 0.65;">Ask me anything about working at here...</p>
    <p>You might find these links of interest:</p>
    <a href="https://www.dtf.vic.gov.au/funds-programs-and-policies/victorian-public-service-enterprise-agreement-2020">Read about our enterprise agreement</a>
    <a href="https://www.vic.gov.au/make-content-accessible">Read guidance to making accessible content</a>
    <a href="https://www.vic.gov.au/victorian-government-directory">Here's the Victorian Government directory</a>
</div>. 
"""

DESCRIPTION = """
    You might find these links of interest

    - [Read about our enterprise agreement](https://www.dtf.vic.gov.au/funds-programs-and-policies/victorian-public-service-enterprise-agreement-2020)
    - [Here's the Victorian Government directory](https://www.vic.gov.au/victorian-government-directory)
    - [Read guidance to making accessible content](https://www.vic.gov.au/make-content-accessible)
"""

TITLE = "Hi I'm Val the Voyager, welcome onboard!"


def format_prompt(message, history):
    prompt = "<s>"
    for user_prompt, bot_response in history:
        prompt += f"[INST] {user_prompt} [/INST]"
        prompt += f" {bot_response}</s> "
    prompt += f"[INST] {message} [/INST]"
    return prompt


def generate(
    prompt,
    history,
    temperature=0.9,
    max_new_tokens=256,
    top_p=0.95,
    repetition_penalty=1.0,
):
    temperature = float(temperature)
    if temperature < 1e-2:
        temperature = 1e-2
    top_p = float(top_p)

    generate_kwargs = dict(
        temperature=temperature,
        max_new_tokens=max_new_tokens,
        top_p=top_p,
        repetition_penalty=repetition_penalty,
        do_sample=True,
        seed=42,
    )

    formatted_prompt = format_prompt(prompt, history)

    stream = client.text_generation(
        formatted_prompt,
        **generate_kwargs,
        stream=True,
        details=True,
        return_full_text=False,
    )
    output = ""

    for response in stream:
        output += response.token.text
        yield output
    return output


additional_inputs = [
    gr.Slider(
        label="Temperature",
        value=0.9,
        minimum=0.0,
        maximum=1.0,
        step=0.05,
        interactive=True,
        info="Higher values produce more diverse outputs",
    ),
    gr.Slider(
        label="Max new tokens",
        value=256,
        minimum=0,
        maximum=1048,
        step=64,
        interactive=True,
        info="The maximum numbers of new tokens",
    ),
    gr.Slider(
        label="Top-p (nucleus sampling)",
        value=0.90,
        minimum=0.0,
        maximum=1,
        step=0.05,
        interactive=True,
        info="Higher values sample more low-probability tokens",
    ),
    gr.Slider(
        label="Repetition penalty",
        value=1.2,
        minimum=1.0,
        maximum=2.0,
        step=0.05,
        interactive=True,
        info="Penalize repeated tokens",
    ),
]

gr.ChatInterface(
    fn=generate,
    chatbot=gr.Chatbot(
        show_share_button=False,
        show_copy_button=True,
        likeable=True,
        layout="bubble",
        placeholder=PLACEHOLDER,
        # label=DESCRIPTION,
        # show_label=True,
    ),
    additional_inputs=additional_inputs,
    examples=[
        ["What should I do on my first day?"],
        ["Ask me what an acronym stands for"],
        ["How can I check my leave allowance?"],
        ["Where can I find a floor map of 1 Macarthur?"],
        ["How can I find out about DTF's Disability network?"],
    ],
    cache_examples=False,
    title=TITLE,
    # description=DESCRIPTION,
).launch(show_api=False)