app.py

import datetime
import os
import re
from io import StringIO

import gradio as gr
import pandas as pd
from huggingface_hub import upload_file
from text_generation import Client

from dialogues import DialogueTemplate
from share_btn import (community_icon_html, loading_icon_html, share_btn_css,
                       share_js)

HF_TOKEN = os.environ.get("HF_TOKEN", None)
API_TOKEN = os.environ.get("API_TOKEN", None)
DIALOGUES_DATASET = "HuggingFaceH4/starchat_playground_dialogues"

model2endpoint = {
    "starchat-alpha": "https://api-inference.huggingface.co/models/HuggingFaceH4/starcoderbase-finetuned-oasst1",
    "starchat-beta": "https://api-inference.huggingface.co/models/HuggingFaceH4/starchat-beta",
}
model_names = list(model2endpoint.keys())


def save_inputs_and_outputs(now, inputs, outputs, generate_kwargs, model):
    buffer = StringIO()
    timestamp = datetime.datetime.now().strftime("%Y-%m-%dT%H:%M:%S.%f")
    file_name = f"prompts_{timestamp}.jsonl"
    data = {"model": model, "inputs": inputs, "outputs": outputs, "generate_kwargs": generate_kwargs}
    pd.DataFrame([data]).to_json(buffer, orient="records", lines=True)

    # Push to Hub
    upload_file(
        path_in_repo=f"{now.date()}/{now.hour}/{file_name}",
        path_or_fileobj=buffer.getvalue().encode(),
        repo_id=DIALOGUES_DATASET,
        token=HF_TOKEN,
        repo_type="dataset",
    )

    # Clean and rerun
    buffer.close()


def get_total_inputs(inputs, chatbot, preprompt, user_name, assistant_name, sep):
    past = []
    for data in chatbot:
        user_data, model_data = data

        if not user_data.startswith(user_name):
            user_data = user_name + user_data
        if not model_data.startswith(sep + assistant_name):
            model_data = sep + assistant_name + model_data

        past.append(user_data + model_data.rstrip() + sep)

    if not inputs.startswith(user_name):
        inputs = user_name + inputs

    total_inputs = preprompt + "".join(past) + inputs + sep + assistant_name.rstrip()

    return total_inputs


def wrap_html_code(text):
    pattern = r"<.*?>"
    matches = re.findall(pattern, text)
    if len(matches) > 0:
        return f"```{text}```"
    else:
        return text


def has_no_history(chatbot, history):
    return not chatbot and not history


def generate(
    model_name,
    system_message,
    user_message,
    chatbot,
    history,
    temperature,
    top_k,
    top_p,
    max_new_tokens,
    repetition_penalty,
    do_save=True,
):
    client = Client(
        model2endpoint[model_name],
        headers={"Authorization": f"Bearer {API_TOKEN}"},
        timeout=60,
    )
    # Don't return meaningless message when the input is empty
    if not user_message:
        print("Empty input")

    history.append(user_message)

    past_messages = []
    for data in chatbot:
        user_data, model_data = data

        past_messages.extend(
            [{"role": "user", "content": user_data}, {"role": "assistant", "content": model_data.rstrip()}]
        )

    if len(past_messages) < 1:
        dialogue_template = DialogueTemplate(
            system=system_message, messages=[{"role": "user", "content": user_message}]
        )
        prompt = dialogue_template.get_inference_prompt()
    else:
        dialogue_template = DialogueTemplate(
            system=system_message, messages=past_messages + [{"role": "user", "content": user_message}]
        )
        prompt = dialogue_template.get_inference_prompt()

    generate_kwargs = {
        "temperature": temperature,
        "top_k": top_k,
        "top_p": top_p,
        "max_new_tokens": max_new_tokens,
    }

    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,
        truncate=4096,
        seed=42,
        stop_sequences=["<|end|>"],
    )

    stream = client.generate_stream(
        prompt,
        **generate_kwargs,
    )

    output = ""
    for idx, response in enumerate(stream):
        if response.token.special:
            continue
        output += response.token.text
        if idx == 0:
            history.append(" " + output)
        else:
            history[-1] = output

        chat = [
            (wrap_html_code(history[i].strip()), wrap_html_code(history[i + 1].strip()))
            for i in range(0, len(history) - 1, 2)
        ]

        # chat = [(history[i].strip(), history[i + 1].strip()) for i in range(0, len(history) - 1, 2)]

        yield chat, history, user_message, ""

    if HF_TOKEN and do_save:
        try:
            now = datetime.datetime.now()
            current_time = now.strftime("%Y-%m-%d %H:%M:%S")
            print(f"[{current_time}] Pushing prompt and completion to the Hub")
            save_inputs_and_outputs(now, prompt, output, generate_kwargs, model_name)
        except Exception as e:
            print(e)

    return chat, history, user_message, ""


examples = [
    "How can I write a Python function to generate the nth Fibonacci number?",
    "How do I get the current date using shell commands? Explain how it works.",
    "What's the meaning of life?",
    "Write a function in Javascript to reverse words in a given string.",
    "Give the following data {'Name':['Tom', 'Brad', 'Kyle', 'Jerry'], 'Age':[20, 21, 19, 18], 'Height' : [6.1, 5.9, 6.0, 6.1]}. Can you plot one graph with two subplots as columns. The first is a bar graph showing the height of each person. The second is a bargraph showing the age of each person? Draw the graph in seaborn talk mode.",
    "Create a regex to extract dates from logs",
    "How to decode JSON into a typescript object",
    "Write a list into a jsonlines file and save locally",
]


def clear_chat():
    return [], []


def delete_last_turn(chat, history):
    if chat and history:
        chat.pop(-1)
        history.pop(-1)
        history.pop(-1)
    return chat, history


def process_example(args):
    for [x, y] in generate(args):
        pass
    return [x, y]


title = """<h1 align="center">⭐ StarChat Playground 💬</h1>"""
custom_css = """
#banner-image {
    display: block;
    margin-left: auto;
    margin-right: auto;
}

#chat-message {
    font-size: 14px;
    min-height: 300px;
}
"""

with gr.Blocks(analytics_enabled=False, css=custom_css) as demo:
    gr.HTML(title)

    with gr.Row():
        with gr.Column():
            gr.Image("thumbnail.png", elem_id="banner-image", show_label=False)
        with gr.Column():
            gr.Markdown(
                """
            💻 This demo showcases a series of **[StarChat](https://huggingface.co/models?search=huggingfaceh4/starchat)** language models, which are fine-tuned versions of the StarCoder family to act as helpful coding assistants.  The base model has 16B parameters and was pretrained on one trillion tokens sourced from 80+ programming languages, GitHub issues, Git commits, and Jupyter notebooks (all permissively licensed).

            📝 For more details, check out our [blog post](https://huggingface.co/blog/starchat-alpha).

            ⚠️ **Intended Use**: this app and its [supporting models](https://huggingface.co/models?search=huggingfaceh4/starchat) are provided as educational tools to explain large language model fine-tuning; not to serve as replacement for human expertise.

            ⚠️ **Known Failure Modes**: the alpha and beta version of **StarChat** have not been aligned to human preferences with techniques like RLHF, so they can produce problematic outputs (especially when prompted to do so). Since the base model was pretrained on a large corpus of code, it may produce code snippets that are syntactically valid but semantically incorrect.  For example, it may produce code that does not compile or that produces incorrect results.  It may also produce code that is vulnerable to security exploits.  We have observed the model also has a tendency to produce false URLs which should be carefully inspected before clicking. For more details on the model's limitations in terms of factuality and biases, see the [model card](https://huggingface.co/HuggingFaceH4/starchat-alpha#bias-risks-and-limitations).

            ⚠️ **Data Collection**: by default, we are collecting the prompts entered in this app to further improve and evaluate the models. Do **NOT** share any personal or sensitive information while using the app! You can opt out of this data collection by removing the checkbox below.
    """
            )

    with gr.Row():
        do_save = gr.Checkbox(
            value=True,
            label="Store data",
            info="You agree to the storage of your prompt and generated text for research and development purposes:",
        )

    with gr.Row():
        selected_model = gr.Radio(choices=model_names, value=model_names[1], label="Select a model")

    with gr.Accordion(label="System Prompt", open=False, elem_id="parameters-accordion"):
        system_message = gr.Textbox(
            elem_id="system-message",
            placeholder="Below is a conversation between a human user and a helpful AI coding assistant.",
            show_label=False,
        )
    with gr.Row():
        with gr.Box():
            output = gr.Markdown()
            chatbot = gr.Chatbot(elem_id="chat-message", label="Chat")

    with gr.Row():
        with gr.Column(scale=3):
            user_message = gr.Textbox(placeholder="Enter your message here", show_label=False, elem_id="q-input")
            with gr.Row():
                send_button = gr.Button("Send", elem_id="send-btn", visible=True)

                # regenerate_button = gr.Button("Regenerate", elem_id="send-btn", visible=True)
                delete_turn_button = gr.Button("Delete last turn", elem_id="delete-btn", visible=True)

                clear_chat_button = gr.Button("Clear chat", elem_id="clear-btn", visible=True)

            with gr.Accordion(label="Parameters", open=False, elem_id="parameters-accordion"):
                temperature = gr.Slider(
                    label="Temperature",
                    value=0.2,
                    minimum=0.0,
                    maximum=1.0,
                    step=0.1,
                    interactive=True,
                    info="Higher values produce more diverse outputs",
                )
                top_k = gr.Slider(
                    label="Top-k",
                    value=50,
                    minimum=0.0,
                    maximum=100,
                    step=1,
                    interactive=True,
                    info="Sample from a shortlist of top-k tokens",
                )
                top_p = gr.Slider(
                    label="Top-p (nucleus sampling)",
                    value=0.95,
                    minimum=0.0,
                    maximum=1,
                    step=0.05,
                    interactive=True,
                    info="Higher values sample more low-probability tokens",
                )
                max_new_tokens = gr.Slider(
                    label="Max new tokens",
                    value=1024,
                    minimum=0,
                    maximum=2048,
                    step=4,
                    interactive=True,
                    info="The maximum numbers of new tokens",
                )
                repetition_penalty = gr.Slider(
                    label="Repetition Penalty",
                    value=1.2,
                    minimum=0.0,
                    maximum=10,
                    step=0.1,
                    interactive=True,
                    info="The parameter for repetition penalty. 1.0 means no penalty.",
                )
            # with gr.Group(elem_id="share-btn-container"):
            #     community_icon = gr.HTML(community_icon_html, visible=True)
            #     loading_icon = gr.HTML(loading_icon_html, visible=True)
            # share_button = gr.Button("Share to community", elem_id="share-btn", visible=True)
            with gr.Row():
                gr.Examples(
                    examples=examples,
                    inputs=[user_message],
                    cache_examples=False,
                    fn=process_example,
                    outputs=[output],
                )

    history = gr.State([])
    # To clear out "message" input textbox and use this to regenerate message
    last_user_message = gr.State("")

    user_message.submit(
        generate,
        inputs=[
            selected_model,
            system_message,
            user_message,
            chatbot,
            history,
            temperature,
            top_k,
            top_p,
            max_new_tokens,
            repetition_penalty,
            do_save,
        ],
        outputs=[chatbot, history, last_user_message, user_message],
    )

    send_button.click(
        generate,
        inputs=[
            selected_model,
            system_message,
            user_message,
            chatbot,
            history,
            temperature,
            top_k,
            top_p,
            max_new_tokens,
            repetition_penalty,
            do_save,
        ],
        outputs=[chatbot, history, last_user_message, user_message],
    )

    delete_turn_button.click(delete_last_turn, [chatbot, history], [chatbot, history])
    clear_chat_button.click(clear_chat, outputs=[chatbot, history])
    selected_model.change(clear_chat, outputs=[chatbot, history])
    # share_button.click(None, [], [], _js=share_js)

demo.queue(concurrency_count=16).launch(debug=True)