import json import requests from datasets import load_dataset import gradio as gr from huggingface_hub import HfApi, hf_hub_download from huggingface_hub.repocard import metadata_load import pandas as pd from matchmaking import * from background_task import init_matchmaking from apscheduler.schedulers.background import BackgroundScheduler block = gr.Blocks() env = [ { "name": "Soccer", "global": None, }, ] matchmaking = Matchmaking() scheduler = BackgroundScheduler() scheduler.add_job(func=init_matchmaking, trigger="interval", seconds=60) scheduler.start() def update_elos(): matchmaking.read_history() matchmaking.compute_elo() matchmaking.save_elo_data() def get_env_data() -> pd.DataFrame: data = pd.read_csv(f"env_elos/elo.csv") # data = pd.DataFrame(columns=["user", "model", "elo", "games_played"]) return data with block: gr.Markdown(f""" # 🏆 The Deep Reinforcement Learning Course Leaderboard 🏆 This is the leaderboard of trained agents during the Deep Reinforcement Learning Course. A free course from beginner to expert. This is the Soccer environment leaderboard, use Ctrl+F to find your rank 🏆 We use an ELO rating to sort the models. You **can click on the model's name** to be redirected to its model card which includes documentation. 🤖 You want to try to train your agents? Sign up to the Hugging Face free Deep Reinforcement Learning Course 🤗 . You want to compare two agents? It's possible using this Spaces demo 👀 . 🔧 There is an **environment missing?** Please open an issue. """) with gr.Row(): refresh_data = gr.Button("Refresh") val = gr.Variable(value=[env["name"]]) refresh_data.click(get_env_data, inputs=[val], outputs=env["global"]) with gr.Row(): env["global"] = gr.components.DataFrame( get_env_data(), headers=["Ranking 🏆", "User 🤗", "Model id 🤖", "ELO 🚀", "Games played 🎮"], datatype=["number", "markdown", "markdown", "number", "number"] ) block.launch()