app.py

import os

import gradio as gr
from gradio_leaderboard import Leaderboard, ColumnFilter, SelectColumns
from apscheduler.schedulers.background import BackgroundScheduler
from huggingface_hub import HfApi

from src.leaderboard import get_leaderboard_df, DISPLAY_BY_DEFAULT, SEARCH_COLUMNS
from src.display.text_blocks import (
    TITLE,
    INTRODUCTION_TEXT,
    LLM_BENCHMARKS_TEXT,
    CITATION_BUTTON_LABEL,
    CITATION_BUTTON_TEXT,
)

REPO_ID = "taagarwa/coding-agent-leaderboard"
TOKEN = os.environ.get("HF_TOKEN")
API = HfApi(token=TOKEN)


def restart_space():
    API.restart_space(repo_id=REPO_ID)


LEADERBOARD_DF = get_leaderboard_df()


def init_leaderboard(dataframe):
    if dataframe is None or dataframe.empty:
        raise ValueError("Leaderboard DataFrame is empty or None.")
    return Leaderboard(
        value=dataframe,
        select_columns=SelectColumns(
            default_selection=DISPLAY_BY_DEFAULT,
            label="Select Columns to Display:",
        ),
        search_columns=SEARCH_COLUMNS,
        filter_columns=[
            ColumnFilter(label="Dataset", column="dataset", type="checkboxgroup"),
            ColumnFilter(label="Number of Parameters (B)", column="model_num_params", type="slider", min=0.5, max=150),
            ColumnFilter(label="Precision", column="precision", type="checkboxgroup"),
        ],
        interactive=False,
    )


demo = gr.Blocks()
with demo:
    gr.HTML(TITLE)
    gr.Markdown(INTRODUCTION_TEXT, elem_classes="markdown-text")

    with gr.Tabs(elem_classes="tab-buttons") as tabs:
        with gr.TabItem("🏅 Coding Agent Benchmark", elem_id="llm-benchmark-tab-table", id=0):
            leaderboard = init_leaderboard(LEADERBOARD_DF)
            gr.Markdown("\* `internal` refers to internal benchmarks performed by the model provider where the harness/environment were not made public")

        with gr.TabItem("📝 About", elem_id="llm-benchmark-tab-table", id=2):
            gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")

    with gr.Row():
        with gr.Accordion("📙 Citation", open=False):
            citation_button = gr.Textbox(
                value=CITATION_BUTTON_TEXT,
                label=CITATION_BUTTON_LABEL,
                lines=20,
                elem_id="citation-button",
                show_copy_button=True,
            )

scheduler = BackgroundScheduler()
scheduler.add_job(restart_space, "interval", seconds=1800)
scheduler.start()
demo.queue(default_concurrency_limit=40).launch()