app.py

from huggingface_hub import InferenceClient
import gradio as gr

client = InferenceClient(
    "HuggingFaceH4/zephyr-7b-alpha"
)


def format_prompt(message, history):
    system = "<|system|>When asked a question, answer only the question. Do no elaborate, or add on. Just answer the question in one to two sentences. You sentences should be at the 5th or 6th grade level.</s>\n"
    prompt = ""
    for user_prompt, bot_response in history:
        prompt += f"<|user|>\n{user_prompt}</s>\n"
        prompt += f"<|assistant|>\n{bot_response}</s>\n"
    prompt += f"<|user|>\n{message}</s>\n"
    return prompt

def generate(
    prompt, history, temperature=0.9, max_new_tokens=500, 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",
    )
]

css = """
  #mkd {
    height: 500px; 
    overflow: auto; 
    border: 1px solid #ccc; 
  }
"""

with gr.Blocks(css=css) as inf:
    gr.HTML("<h1><center>zephyr-7b-alpha<h1><center>")
    gr.HTML("<h3><center>In this demo, you can chat with <a href='https://huggingface.co/HuggingFaceH4/zephyr-7b-alpha'>zephyr-7b-alpha</a> model. 💬<h3><center>")
    gr.ChatInterface(
        generate,
        additional_inputs=additional_inputs,
        examples=[["Can squirrel swims?"], ["Write a poem about squirrel."]]
    )

inf.queue().launch()