Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
import os | |
import re | |
import csv | |
import json | |
import time | |
import random | |
import asyncio | |
import discord | |
import logging | |
import os.path | |
import secrets | |
import gspread | |
import datetime | |
import requests | |
import threading | |
import gradio_client | |
import numpy as np | |
import pandas as pd | |
import gradio as gr | |
import plotly.graph_objects as go | |
from tabulate import tabulate | |
from requests import HTTPError | |
from gradio_client import Client | |
from discord import Color, Embed | |
from discord.ui import Button, View | |
from discord.ext import commands, tasks | |
from datetime import datetime, timedelta | |
from urllib.parse import urlparse, parse_qs | |
from apscheduler.executors.pool import ThreadPoolExecutor | |
from apscheduler.executors.asyncio import AsyncIOExecutor | |
from apscheduler.schedulers.asyncio import AsyncIOScheduler | |
from gspread_formatting.dataframe import format_with_dataframe | |
from apscheduler.schedulers.background import BackgroundScheduler | |
from gspread_dataframe import get_as_dataframe, set_with_dataframe | |
from huggingface_hub import HfApi, list_liked_repos, list_metrics, list_models | |
DISCORD_TOKEN = os.environ.get("DISCORD_TOKEN", None) | |
intents = discord.Intents.all() | |
bot = commands.Bot(command_prefix='!', intents=intents) | |
GRADIO_APP_URL = "https://huggingface.co/spaces/discord-community/LevelBot" | |
# Dictionary to store user IDs and their corresponding unique strings | |
user_tokens = {} | |
"""""" | |
XP_PER_MESSAGE = 10 # 100k messages = 1M exp = lvl 100 | |
"""""" | |
service_account = json.loads(os.environ.get('KEY')) | |
file_path = 'service_account.json' | |
with open(file_path, 'w') as json_file: | |
json.dump(service_account, json_file) | |
gspread_bot = gspread.service_account(filename='service_account.json') | |
worksheet = gspread_bot.open("levelbot").sheet1 | |
test_merge_worksheet = gspread_bot.open("test_merge").sheet1 | |
"""""" | |
bot_ids = [1136614989411655780, 1166392942387265536, 1158038249835610123, 1130774761031610388, 1155489509518098565, 1155169841276260546, 1152238037355474964, 1154395078735953930] | |
"""""" | |
api = HfApi() | |
"""""" | |
global_df = pd.DataFrame() | |
print(type(global_df)) | |
community_global_df = pd.DataFrame() | |
community_global_df_with_id = pd.DataFrame() | |
community_global_df_gradio = pd.DataFrame() | |
test_merge = pd.read_csv("https://docs.google.com/spreadsheets/d/1C8aLqgCqLYcMiIFf-P_Aosaa03C_WLIB_UyqvjSdWg8/export?format=csv&gid=0") | |
class DMButton(Button): | |
def __init__(self, label, style): | |
super().__init__(label=label, style=style) | |
async def callback(self, interaction: discord.Interaction): | |
# await interaction.user.send(self.message) # this is for DMs, but users may have DMs disabled | |
user_id = interaction.user.id | |
if int(user_id) in user_tokens: | |
del user_tokens[int(user_id)] # always delete all past tokens for a given user when creating new link | |
unique_string = generate_unique_string() | |
user_tokens[user_id] = unique_string | |
unique_link = f"<{GRADIO_APP_URL}?user_id={user_id}&token={unique_string}>" | |
message = f"Login link generated! To complete the verification process:\n- 1 Visit this link,\n- 2 click the 'π€Sign in with Hugging Face' button\n\n{unique_link}" | |
await interaction.response.send_message(message, ephemeral=True) | |
def generate_unique_string(length=6): | |
return secrets.token_hex(length // 2) | |
async def on_ready(): | |
try: | |
global global_df | |
await asyncio.sleep(1.1) | |
print(f'Logged in as {bot.user.name}') | |
print(f"XP_PER_MESSAGE: {XP_PER_MESSAGE}") | |
guild = bot.get_guild(879548962464493619) | |
await guild.chunk() # get all users into bot cache | |
channel = bot.get_channel(900125909984624713) | |
if channel: | |
try: | |
message = await channel.fetch_message(1271145797433557023) | |
if message: | |
button = DMButton(label="Verify Discord Account", style=discord.ButtonStyle.primary) | |
view = View(timeout=None) | |
view.add_item(button) | |
await message.edit(view=view) | |
print("message edited") | |
except discord.NotFound: | |
print(f"Message with ID 1271145797433557023 not found.") | |
except discord.HTTPException as e: | |
print(f"Failed to fetch message with ID 1271145797433557023: {e}") | |
"""import data from google sheets -> HF Space df (doesn't make API call this way, as it's read-only)""" | |
global_df = test_merge | |
print(f"csv successfully retrieved: \n {global_df}") | |
# updates both leaderboards | |
remove_huggingfolks.start() | |
print("------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"on_ready Error: {e}") | |
def update_google_sheet(): | |
"""save data from HF Space -> google sheets (makes 1 API call)""" | |
try: | |
print("Updating google sheets...") | |
print("------------------------------------------------------------------------") | |
name = "test_merge_worksheet" | |
set_with_dataframe(test_merge_worksheet, global_df) | |
timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S") | |
print("------------------------------------------------------------------------") | |
print(f"Google sheet {name} {test_merge_worksheet} successfully updated at {timestamp}! \n{global_df}") | |
print("------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"update_google_sheet Error: {e}") | |
#@tasks.loop(minutes=1) tasks.loop leads to heartbeat blocked issues (merging calculations too much with normal discord bot functions) | |
def update_hub_stats(): | |
try: | |
global global_df | |
print("Updating hub stats...") | |
print("------------------------------------------------------------------------") | |
for index, row in global_df.iterrows(): | |
user = row['hf_user_name'] | |
if pd.notna(user): | |
#print(f"user: {user}") | |
url = f"https://huggingface.co/api/users/{user}/overview" | |
#print(f"url: {url}") | |
response = requests.get(url) | |
#print(f"response: {response}") | |
if response.status_code == 200: | |
data = response.json() | |
#print(f"data: {data}") | |
likes = data["numLikes"] | |
models = data["numModels"] | |
datasets = data["numDatasets"] | |
spaces = data["numSpaces"] | |
discussions = data["numDiscussions"] | |
papers = data["numPapers"] | |
upvotes = data["numUpvotes"] | |
# recalculate level as well (no longer only depends on discord activity) | |
try: | |
global_df.loc[index, 'likes'] = likes | |
global_df.loc[index, 'models'] = models | |
global_df.loc[index, 'datasets'] = datasets | |
global_df.loc[index, 'spaces'] = spaces | |
global_df.loc[index, 'discussions'] = discussions | |
global_df.loc[index, 'papers'] = papers | |
global_df.loc[index, 'upvotes'] = upvotes | |
total_hub_activity = likes + models + datasets + spaces + discussions + papers + upvotes | |
total_hub_exp = total_hub_activity * 20 #2x better than discord | |
total_hub_exp_string = "L" + str(total_hub_exp) + "L" | |
# hub exp | |
global_df.loc[index, 'hub_exp'] = total_hub_exp_string | |
# total exp (discord + hub) | |
discord_exp = row['discord_exp'] | |
discord_exp_int = discord_exp.strip('L') | |
total_exp = int(discord_exp_int) + int(total_hub_exp) | |
total_exp_string = "L" + str(total_exp) + "L" | |
global_df.loc[index, 'total_exp'] = total_exp_string | |
# level | |
level = calculate_level(total_exp) | |
global_df.loc[index, 'discord_level'] = level | |
except Exception as e: | |
print(f"{e} error updating the dataframe") | |
except Exception as e: | |
print(f"Failed to parse data for user {user}. {e}") | |
timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S") | |
print("------------------------------------------------------------------------") | |
print(f"Hub stats successfully updated at {timestamp}! \n{global_df}") | |
print("------------------------------------------------------------------------") | |
thread_pool_executor = ThreadPoolExecutor(max_workers=2) | |
asyncio_executor = AsyncIOExecutor() | |
scheduler = BackgroundScheduler(executors={ | |
'default': thread_pool_executor | |
}) | |
scheduler.add_job(update_google_sheet, trigger='interval', minutes=1, max_instances=2, executor='default') | |
scheduler.add_job(update_hub_stats, trigger='interval', minutes=1.5, max_instances=2, executor='default') | |
scheduler.start() | |
#asyncio.get_event_loop().run_forever() | |
def calculate_level(xp): | |
return int(xp ** (1.0 / 3.0)) | |
def calculate_xp(level): | |
return (int(level ** 3)) | |
async def add_exp(member_id): | |
try: | |
"""Uses member_id to create new record or update old one [member_id_column] ... [member_exp_column]""" | |
await asyncio.sleep(0.1) | |
global global_df | |
guild = bot.get_guild(879548962464493619) | |
member = guild.get_member(member_id) # bot.get_user == User, guild.get_member == Member (not the same thing!) | |
lvl1 = guild.get_role(1171861537699397733) | |
lvl2 = guild.get_role(1171861595115245699) | |
lvl3 = guild.get_role(1171861626715115591) | |
lvl4 = guild.get_role(1171861657975259206) | |
lvl5 = guild.get_role(1171861686580412497) | |
lvl6 = guild.get_role(1171861900301172736) | |
lvl7 = guild.get_role(1171861936258941018) | |
lvl8 = guild.get_role(1171861968597024868) | |
lvl9 = guild.get_role(1171862009982242836) | |
lvl10 = guild.get_role(1164188093713223721) | |
lvl11 = guild.get_role(1171524944354607104) | |
lvl12 = guild.get_role(1171524990257082458) | |
lvl13 = guild.get_role(1171525021928263791) | |
lvl14 = guild.get_role(1171525062201966724) | |
lvl15 = guild.get_role(1171525098465918996) | |
lvl16 = guild.get_role(1176826165546201099) | |
lvl17 = guild.get_role(1176826221301092392) | |
lvl18 = guild.get_role(1176826260643659776) | |
lvl19 = guild.get_role(1176826288816791693) | |
lvl20 = guild.get_role(1176826319447801896) | |
lvl21 = guild.get_role(1195030831174008902) | |
lvl22 = guild.get_role(1195030883351150592) | |
lvl23 = guild.get_role(1196055555006009445) | |
lvl24 = guild.get_role(1196055640917938216) | |
lvl25 = guild.get_role(1196055712506318869) | |
lvl26 = guild.get_role(1196055775924195378) | |
lvl27 = guild.get_role(1196055837018435664) | |
lvl28 = guild.get_role(1196055908267081849) | |
lvl29 = guild.get_role(1196055970804150352) | |
lvl30 = guild.get_role(1196056027720847380) | |
lvl31 = guild.get_role(1206542603261186078) | |
lvl32 = guild.get_role(1206542673549205514) | |
lvl33 = guild.get_role(1206542690939048007) | |
lvl34 = guild.get_role(1206542707862806568) | |
lvl35 = guild.get_role(1206542723633512468) | |
lvl36 = guild.get_role(1206542738728681485) | |
lvl37 = guild.get_role(1206542754625101866) | |
lvl38 = guild.get_role(1206542771314364416) | |
lvl39 = guild.get_role(1206542785973321758) | |
lvl40 = guild.get_role(1206542802155208725) | |
lvl41 = guild.get_role(1206568953221218354) | |
lvl42 = guild.get_role(1206568979393413150) | |
lvl43 = guild.get_role(1206568997374394368) | |
lvl44 = guild.get_role(1206569014747463681) | |
lvl45 = guild.get_role(1206569031650385921) | |
lvl46 = guild.get_role(1206569047207182356) | |
lvl47 = guild.get_role(1206569062851805254) | |
lvl48 = guild.get_role(1206569077112315984) | |
lvl49 = guild.get_role(1206569091826057306) | |
lvl50 = guild.get_role(1206569107118493757) | |
lvls = { | |
1: lvl1, 2: lvl2, 3: lvl3, 4: lvl4, 5: lvl5, 6: lvl6, 7: lvl7, 8: lvl8, 9: lvl9, 10: lvl10, | |
11: lvl11, 12: lvl12, 13: lvl13, 14: lvl14, 15: lvl15, 16: lvl16, 17: lvl17, 18: lvl18, 19: lvl19, 20: lvl20, | |
21: lvl21, 22: lvl22, 23: lvl23, 24: lvl24, 25: lvl25, 26: lvl26, 27: lvl27, 28: lvl28, 29: lvl29, 30: lvl30, | |
31: lvl31, 32: lvl32, 33: lvl33, 34: lvl34, 35: lvl35, 36: lvl36, 37: lvl37, 38: lvl38, 39: lvl39, 40: lvl40, | |
41: lvl41, 42: lvl42, 43: lvl43, 44: lvl44, 45: lvl45, 46: lvl46, 47: lvl47, 48: lvl48, 49: lvl49, 50: lvl50, | |
} | |
member_found = False | |
print(f"Searching for member_id {member_id} in dataframe...") | |
# discord_user_id column | |
# iterate over items of first column (discord_user_id) | |
for index, cell_value in global_df.iloc[:, 0].items(): | |
# tldr; set_as_dataframe forces scientific notation which corrupts discord_user_id data. | |
# set_as_dataframe is still highly efficient (1 API call), so we format numerical data as strings, | |
# which results in efficient google sheet updating + data integrity | |
cell_value = str(cell_value) | |
if cell_value.startswith("L") and cell_value.endswith("L"): | |
cell_value_clipped = cell_value[1:-1] | |
if cell_value_clipped == str(member_id): # str(member_id) needed, it is int by default | |
print("test4") | |
# if found, update that row... | |
member_found = True | |
print(f"Record for {member} found at row {index + 1}, column 1") | |
# increment the old experience value (better not to replace outright) | |
old_xp = global_df.loc[index, 'discord_exp'] | |
# remove L (write, so we replace) | |
old_xp = str(old_xp) | |
if old_xp.startswith("L") and old_xp.endswith("L"): | |
old_xp = old_xp[1:-1] | |
# str -> int temporarily for adding | |
new_xp = int(old_xp) + XP_PER_MESSAGE | |
total_exp = global_df.loc[index, 'total_exp'] | |
hub_xp = global_df.loc[index, 'hub_exp'] | |
total_exp = str(total_exp) | |
hub_xp = str(hub_xp) | |
if total_exp.startswith("L") and total_exp.endswith("L"): | |
total_exp = total_exp[1:-1] | |
if hub_xp.startswith("L") and hub_xp.endswith("L"): | |
hub_xp = hub_xp[1:-1] | |
# set old level; use this for more accurate logging and jumping multiple levels at once (for example, verifying) | |
old_total_xp = int(total_exp) | |
#old_level = calculate_level(old_total_xp) | |
# check if hub exp not empty | |
if hub_xp.strip(): | |
total_exp = int(new_xp) + int(hub_xp) | |
else: | |
total_exp = int(new_xp) | |
# total v | |
current_level = calculate_level(total_exp) | |
# convert back to string + google sheet proofing | |
new_xp = str(new_xp) | |
if not new_xp.startswith("L") and not new_xp.endswith("L"): | |
new_xp = "L" + str(new_xp) + "L" | |
global_df.loc[index, 'discord_exp'] = new_xp # do not change column name | |
# after | |
total_exp = str(total_exp) | |
if not total_exp.startswith("L") and not total_exp.endswith("L"): | |
total_exp = "L" + str(total_exp) + "L" | |
# add back to dataframe in memory after checking redundantly; | |
if total_exp.startswith("L") and total_exp.endswith("L"): | |
print("test5") | |
global_df.loc[index, 'total_exp'] = total_exp # do not change column name | |
print(f"Record for {member} updated from {old_total_xp} to {global_df.loc[index, 'total_exp']} (+{XP_PER_MESSAGE}) ") | |
# level up | |
verified_role = guild.get_role(900063512829755413) | |
print(f"Current_level for {member}: {current_level}") | |
if current_level >= 2 and current_level <=50: | |
print("test6") | |
current_role = lvls[current_level] | |
if current_role not in member.roles: # if we need to level up / update role | |
print("test7") | |
# finding leaderboard rank + excluding huggingfolks (still need exclusion) | |
try: | |
print("Calculating rank...") | |
copy_df = global_df.copy() | |
copy_df['discord_user_id'] = copy_df['discord_user_id'].str.strip('L').astype(str) | |
copy_df['total_exp'] = copy_df['total_exp'].str.strip('L').astype(int) | |
row = copy_df[copy_df['discord_user_id'] == str(member_id)] | |
target_exp = row['total_exp'].values[0] | |
rank = (copy_df['total_exp'] > target_exp).sum() + 1 | |
print(f"The rank for {member} based on total_exp is: {rank}") | |
except Exception as e: | |
print(f"Discord ID {member} {member_id} not found in the DataFrame. {e}") | |
rank = "π€" | |
# if level 3 -> then send embed, remove some exp | |
if current_level >= 3: # could change to 4 maybe | |
if verified_role not in member.roles: | |
# L12345L -> `12345` -> 12345 | |
total_exp = total_exp[1:-1] | |
total_exp = int(total_exp) | |
if total_exp % 30 == 0: # staggers messages so we don't send one every time exp is earned | |
# claim exp (-30 for level 3, but +100 as bonus exp. This scales infinitely until the member verifies, | |
# so they can continue earning exp, it just won't translate to levels and the leaderboard. | |
# This way they can claim at any time and get a big boost in levels! | |
claim_exp = total_exp + 70 | |
# send embed | |
embed = Embed(color=Color.red()) | |
embed.set_author(name=f"{member}", icon_url=member.avatar.url if member.avatar else bot.user.avatar.url) | |
embed.title = f"β οΈYour account is not Verified! Unable to level up `π` -> `{current_level}` β" | |
msg = f'π€ Hey {member}! You can continue leveling up in the Hugging Face Discord server by Verifying your account, and claim `{claim_exp}` bonus exp points!' | |
embed.description = f"{msg}" | |
verification_link = "https://discord.com/channels/879548962464493619/900125909984624713" | |
embed.add_field(name="Verify Here:", value=verification_link, inline=True) | |
u_1 = "π Earn exp for activity on Discord and HF and climb the β leaderboard !" | |
u_2 = "π Feature your content in weekly news and increase its visibility!" | |
u_3 = "π Early access to Beta features!" | |
u_4 = "π‘οΈ Secure your progress, and restore if needed!" | |
embed.add_field(name="You can Unlock:", value=f"{u_1}\n{u_2}\n{u_3}\n{u_4}", inline=True) | |
embed.set_image(url='https://cdn.discordapp.com/attachments/1150399343912833024/1205537451242688573/download_1.png?ex=65d8bb3e&is=65c6463e&hm=042fe7dd3521887db0bd48eeb846de1cc7c75194f9e95215c23512ff61ea3475&') | |
lunar = bot.get_user(811235357663297546) | |
await member.send(embed=embed) | |
await lunar.send(embed=embed) | |
print("Sent verification cap embed to {member}") | |
print("------------------------------------------------------------------------") | |
return | |
# increment the old level value (better to replace outright) | |
# only increment level column if you are lvl2 or 3+ with verified role (this may make some members not appear) | |
global_df.loc[index, 'discord_level'] = current_level # do not change column name | |
# remove all level roles then add new role | |
current_level_roles = [role for level, role in lvls.items() if role in member.roles] | |
if current_level_roles: | |
print(f"current_level_roles for {member}: {current_level_roles}") | |
for removable_role in current_level_roles: | |
await member.remove_roles(removable_role) | |
print(f"Removed {removable_role} from {member}") | |
removable_role_name = removable_role.name | |
else: | |
removable_role_name = "π" | |
await member.add_roles(current_role) | |
print(f"Level Up! Gave {member} {current_role}") | |
if current_level == 2: # special embed with opt-out instructions for discord newcomers | |
embed = Embed(color=Color.blue()) | |
embed.set_author(name=f"{member}", icon_url=member.avatar.url if member.avatar else bot.user.avatar.url) | |
embed.title = f"Level Up! `{removable_role_name}` -> `{current_level}`" | |
msg = f'π€ Congrats {member}! You just leveled up in the Hugging Face Discord server. To opt out of these notifications, you can right click -> "Block" me!' | |
embed.description = f"{msg}." | |
embed.add_field(name="Leaderboard Ranking:", value=f"π **{rank}**\n\nhttps://discord.com/channels/879548962464493619/1197143964994773023", inline=True) | |
msg3 = "- Posting\n- Reacting / being reacted to\n- Being active on the Hugging Face Hub (verify to link your Hub + Discord accounts!)" | |
embed.add_field(name="How to Level Up:", value=msg3, inline=True) | |
verification_link = "https://discord.com/channels/879548962464493619/900125909984624713" | |
embed.add_field(name="Verify Here:", value=verification_link, inline=True) | |
lunar = bot.get_user(811235357663297546) | |
await member.send(embed=embed) | |
await lunar.send(embed=embed) | |
print(f"Sent levelup embed to {member}") | |
return | |
if current_role in member.roles: | |
# send embed | |
embed = Embed(color=Color.blue()) | |
embed.set_author(name=f"{member}", icon_url=member.avatar.url if member.avatar else bot.user.avatar.url) | |
embed.title = f"Level Up! `{removable_role_name}` -> `{current_level}`" | |
msg = f'π€ Congrats {member}! You just leveled up in the Hugging Face Discord server' | |
embed.description = f"{msg}." | |
embed.add_field(name="Leaderboard Ranking:", value=f"π **{rank}**\n\nhttps://discord.com/channels/879548962464493619/1197143964994773023", inline=True) | |
msg3 = "- Posting\n- Reacting / being reacted to\n- Being active on the Hugging Face Hub (verify to link your Hub + Discord accounts!)" | |
embed.add_field(name="How to Level Up:", value=msg3, inline=True) | |
verification_link = "https://discord.com/channels/879548962464493619/900125909984624713" | |
embed.add_field(name="Verify Here:", value=verification_link, inline=True) | |
lunar = bot.get_user(811235357663297546) | |
await member.send(embed=embed) | |
await lunar.send(embed=embed) | |
print(f"Sent levelup embed to {member}") | |
print("------------------------------------------------------------------------") | |
if not member_found: # this only checks if discord_user_id (with L) is present in discord_user_id column, | |
# if not, create new record | |
print(f"Creating new record for {member}") | |
xp = 10 # define somewhere else? | |
current_level = calculate_level(xp) | |
xp = str(xp) | |
if not xp.startswith("L") and not xp.endswith("L"): | |
xp = "L" + str(xp) + "L" | |
member_id = str(member_id) | |
if not member_id.startswith("L") and not member_id.endswith("L"): | |
member_id = "L" + str(member_id) + "L" | |
member_name = str(member.name) | |
hf_user_name = "n/a" | |
hub_exp = "L0L" | |
total_exp = xp | |
verified_date = "n/a" | |
# need to initialize these when creating new record | |
likes = 0 | |
models = 0 | |
datasets = 0 | |
spaces = 0 | |
discussions = 0 | |
papers = 0 | |
upvotes = 0 | |
row_data = [member_id, member_name, xp, current_level, hf_user_name, hub_exp, total_exp, verified_date, likes, models, datasets, spaces, discussions, papers, upvotes] | |
global_df.loc[len(global_df.index)] = row_data | |
print("------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"add_exp Error: {e}") | |
async def on_message(message): | |
org_link = "https://huggingface.co/organizations/discord-community/share/wPKRAHYbAlaEaCxUxcqVyaaaeZcYagDvqc" | |
invite_message = "Click to join our community org on the HF Hub!" | |
try: | |
if message.author.id not in bot_ids: # could change to if author does not have bot role (roleid) | |
print(f"adding exp from message {message.author}") | |
await asyncio.sleep(0.1) | |
await add_exp(message.author.id) | |
# add check for verification | |
if message.content.find("!help") != -1: | |
await message.channel.send( | |
"To verify your π€ account, generate a login link in the verification channel! " | |
) | |
await bot.process_commands(message) | |
except Exception as e: | |
print(f"on_message Error: {e}") | |
async def on_reaction_add(reaction, user): | |
try: | |
if user.id not in bot_ids: | |
print(f"adding exp from react {user}") | |
await asyncio.sleep(0.1) | |
if not isinstance(reaction.message.channel, discord.DMChannel) and reaction.message.author.id != user.id: # can't earn exp from reacting in bot DMs, which is harder to track, and can't earn while self-reacting, which is abuseable | |
await add_exp(reaction.message.author.id) | |
await asyncio.sleep(0.1) | |
except Exception as e: | |
print(f"on_reaction_add Error: {e}") | |
def format_hyperlink(username): | |
return f'<a target="_blank" href="https://huggingface.co/{username}" style="color: var(--link-text-color);">{username}</a>' | |
async def remove_huggingfolks(): | |
try: | |
guild = bot.get_guild(879548962464493619) | |
role = discord.utils.get(guild.roles, id=897376942817419265) | |
await guild.chunk() | |
members_with_role = [member.id for member in guild.members if role in member.roles] | |
top_num_rows = await asyncio.to_thread(_remove_huggingfolks_sync, members_with_role) | |
channel = bot.get_channel(1197143964994773023) | |
message = await channel.fetch_message(1197148293164187678) | |
new_table = tabulate(top_num_rows, headers=["Name", "Level", "Experience", "Rank"], tablefmt="plain") | |
await message.edit(content=f"Updated Leaderboard:\n```\n{new_table}\n```") | |
print("Updated discord leaderboard!") | |
print("------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"remove_huggingfolks Error: {e}") | |
def _remove_huggingfolks_sync(members_with_role): | |
global community_global_df | |
global community_global_df_with_id | |
global community_global_df_gradio | |
community_global_df = global_df.copy() | |
community_global_df['discord_user_id'] = community_global_df['discord_user_id'].str.strip('L').astype(str) | |
for member_id in members_with_role: | |
community_global_df = community_global_df[community_global_df['discord_user_id'] != str(member_id)] | |
community_global_df_with_id = community_global_df.copy() | |
reorder = ['discord_user_id', 'hf_user_name', 'discord_exp', 'discord_level', 'discord_user_name', 'hub_exp', 'total_exp', 'verified_date'] | |
community_global_df = community_global_df[reorder] | |
community_global_df.drop(community_global_df.columns[0], axis=1, inplace=True) | |
community_global_df.drop(community_global_df.columns[1], axis=1, inplace=True) | |
community_global_df.drop(community_global_df.columns[2], axis=1, inplace=True) | |
community_global_df.drop(community_global_df.columns[2], axis=1, inplace=True) | |
community_global_df.drop(community_global_df.columns[3], axis=1, inplace=True) | |
# drop rows (records) where hf_user_name does not exist (we display only verified users) | |
community_global_df = community_global_df.dropna(subset=['hf_user_name']) | |
community_global_df['total_exp'] = community_global_df['total_exp'].str.strip('L').astype(int) | |
community_global_df['total_exp'] = pd.to_numeric(community_global_df['total_exp'], errors='coerce').fillna(0).astype(int) | |
community_global_df = community_global_df.nlargest(len(community_global_df), 'total_exp') | |
community_global_df_gradio = community_global_df.copy() | |
community_global_df_gradio['hf_user_name'] = community_global_df_gradio['hf_user_name'].apply(format_hyperlink) | |
top_num = 30 | |
top_num_exp = community_global_df.head(top_num).copy() | |
top_num_exp['Rank'] = ['π₯', 'π₯', 'π₯'] + [''] * (top_num - 3) | |
top_num_rows = top_num_exp.values.tolist() | |
return top_num_rows | |
async def xp_help(ctx): | |
try: | |
help_message = "How to earn Discord / Hub exp: Post messages, react, Like, discuss, create repos and papers" | |
await ctx.author.send(help_message) | |
except Exception as e: | |
print(f"xp_help Error: {e}") | |
DISCORD_TOKEN = os.environ.get("DISCORD_TOKEN", None) | |
def run_bot(): | |
bot.run(DISCORD_TOKEN) | |
threading.Thread(target=run_bot).start() | |
def get_data(): | |
try: | |
return community_global_df_gradio | |
except Exception as e: | |
print(f"get_data Error: {e}") | |
def get_data2(): | |
try: | |
display_data = { | |
'π€ Hub (+30 exp)': ['Creating Repos', 'Papers', 'Likes/Upvotes', 'Discussions'], | |
'Discord (+10 exp)': ['Posting messages', 'Reacting', '', ''] | |
} | |
display_df = pd.DataFrame(display_data) | |
return display_df | |
except Exception as e: | |
print(f"get_data2 Error: {e}") | |
async def on_member_join(member): | |
try: | |
# Check if the member already has a record | |
global global_df | |
member_id = f"L{member.id}L" # Ensure consistent formatting with the database | |
if member_id not in global_df['discord_user_id'].values: | |
# Create a new row for the user | |
print(f"Creating new record for {member}") | |
xp = 0 # Start with zero XP | |
current_level = calculate_level(xp) | |
xp = f"L{xp}L" | |
member_name = member.name | |
hf_user_name = "n/a" | |
hub_exp = "L0L" | |
total_exp = xp | |
verified_date = "n/a" | |
# Initialize additional columns for the new row | |
likes = 0 | |
models = 0 | |
datasets = 0 | |
spaces = 0 | |
discussions = 0 | |
papers = 0 | |
upvotes = 0 | |
row_data = [ | |
member_id, member_name, xp, current_level, hf_user_name, hub_exp, | |
total_exp, verified_date, likes, models, datasets, spaces, discussions, | |
papers, upvotes | |
] | |
global_df.loc[len(global_df.index)] = row_data # Add the new row | |
print(f"Added new user {member.name} with ID {member.id} to the database.") | |
else: | |
print(f"User {member.name} with ID {member.id} already exists in the database.") | |
except Exception as e: | |
print(f"Error adding new member {member}: {e}") | |
#------------------------------------------------------------------------------------------------------------------------------- | |
# year->get all dates | |
def get_dates_in_year(year): | |
start_date = datetime(year, 1, 1) | |
end_date = datetime(year + 1, 1, 1) | |
delta = end_date - start_date | |
return [start_date + timedelta(days=i) for i in range(delta.days)] | |
# HF API->get activity | |
def get_user_likes(username): | |
url = f"https://huggingface.co/api/users/{username}/likes" | |
response = requests.get(url) | |
if response.status_code == 200: | |
return response.json() | |
else: | |
print(f"Failed to fetch data: {response.status_code}") | |
return [] | |
# data->get dates | |
def parse_activity_dates(activity_data): | |
activity_dates = [] | |
for item in activity_data: | |
timestamp = item.get('createdAt') | |
if timestamp: | |
date = datetime.fromisoformat(timestamp[:-1]).date() | |
activity_dates.append(date) | |
return activity_dates | |
def create_plot(username, year): | |
dates_in_year = get_dates_in_year(year) | |
dates_in_year = [d.date() for d in dates_in_year] | |
activity_data = get_user_likes(username) | |
activity_dates = parse_activity_dates(activity_data) | |
activity_count = {date: 0 for date in dates_in_year} | |
for date in activity_dates: | |
if date in activity_count: | |
activity_count[date] += 1 | |
total_activities = sum(activity_count.values()) | |
# activity->color | |
def get_color(activity_count): | |
if activity_count == 0: | |
return 'white' | |
elif activity_count < 5: | |
return 'lightgreen' | |
elif activity_count < 10: | |
return 'green' | |
else: | |
return 'darkgreen' | |
# data->Plotly | |
num_weeks = len(dates_in_year) // 7 + (1 if len(dates_in_year) % 7 != 0 else 0) | |
z = [[None for _ in range(num_weeks)] for _ in range(7)] | |
hover_texts = [["" for _ in range(num_weeks)] for _ in range(7)] | |
for idx, date in enumerate(dates_in_year): | |
week = idx // 7 | |
day = idx % 7 | |
z[day][week] = activity_count[date] | |
hover_texts[day][week] = f"{date}: {activity_count[date]} activities" | |
# month labels | |
month_labels = ["JAN", "FEB", "MAR", "APR", "MAY", "JUN", "JUL", "AUG", "SEP", "OCT", "NOV", "DEC"] | |
month_positions = [0] # JAN | |
for month in range(2, 13): | |
first_day_of_month = datetime(year, month, 1).date() | |
month_position = (first_day_of_month - datetime(year, 1, 1).date()).days // 7 | |
month_positions.append(month_position) | |
# heatmap | |
fig = go.Figure(data=go.Heatmap( | |
z=z, | |
x=list(range(num_weeks)), | |
y=list(range(7)), | |
hoverinfo='text', | |
text=hover_texts, | |
showscale=False, | |
colorscale=[ | |
[0, 'white'], | |
[0.25, 'lightgreen'], | |
[0.5, 'green'], | |
[1, 'darkgreen'] | |
], | |
zmin=0, | |
zmax=20 | |
)) | |
# lock the plot size and ensure cells are square | |
cell_size = 20 | |
fig.update_layout( | |
title=f'{username} Activity in {year} | Total Contributions: {total_activities}', | |
xaxis_nticks=52, | |
yaxis_nticks=7, | |
xaxis_title=' ', | |
yaxis_title=' ', | |
yaxis=dict( | |
tickmode='array', | |
tickvals=list(range(7)), | |
ticktext=['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'], | |
scaleanchor="x", # This ensures that the cells are square | |
scaleratio=1 | |
), | |
xaxis=dict( | |
tickmode='array', | |
tickvals=month_positions, | |
ticktext=month_labels, | |
scaleanchor="y", | |
scaleratio=1, | |
fixedrange=True, # disable zoom and pan | |
), | |
autosize=False, | |
width=num_weeks * cell_size, | |
height=14 * cell_size, # halve height of each cell | |
margin=dict( | |
l=20, | |
r=20, | |
b=20, | |
t=50, | |
pad=2 | |
) | |
) | |
fig.update_layout( | |
xaxis=dict(fixedrange=True), # disable zoom and pan | |
yaxis=dict(fixedrange=True) # disable zoom and pan | |
) | |
return fig | |
#------------------------------------------------------------------------------------------------------------------------------- | |
def verify_button(profile: gr.OAuthProfile | None, request: gr.Request) -> str: | |
url_str = str(request.url) | |
query_params = parse_qs(urlparse(url_str).query) | |
user_id = query_params.get('user_id', [None])[0] | |
print(f"user_id={user_id}") | |
token = query_params.get('token', [None])[0] | |
print(f"token={token}") | |
print(f"||| token:{token}||| user_id:{user_id}||| profile:{profile}||| user_tokens:{user_tokens}") | |
if user_id is None or token is None: | |
return "# β Invalid link. Generate a new link [here](https://discord.com/channels/879548962464493619/900125909984624713) !" | |
if int(user_id) not in user_tokens or user_tokens[int(user_id)] != token: | |
return "# β Invalid or expired link. Generate a new link [here](https://discord.com/channels/879548962464493619/900125909984624713) !" | |
if profile is None: | |
return f"# β Not logged in with Hugging Face yet." | |
# check if hf_user_name in dataframe: | |
if profile.username in global_df['hf_user_name'].values: | |
return (f"# β The HF account {profile.username} is already verified! To change discord accounts or HF accounts, contact @lunarflu or [email protected]") | |
# check if discord_user_id in dataframe: | |
altered_member_id = "L" + str(user_id) + "L" | |
if altered_member_id in global_df['discord_user_id'].values: | |
hf_user_name = global_df.loc[global_df['discord_user_id'] == altered_member_id, 'hf_user_name'].iloc[0] | |
if pd.isnull(hf_user_name) or hf_user_name == 'n/a': | |
# empty (no link created yet between discord_user_id and hf_user_name) so we can update | |
verified_date = datetime.now().strftime("%m/%d/%Y, %H:%M:%S") | |
global_df.loc[global_df['discord_user_id'] == altered_member_id, 'hf_user_name'] = profile.username | |
global_df.loc[global_df['discord_user_id'] == altered_member_id, 'verified_date'] = verified_date | |
org_link = "https://huggingface.co/organizations/discord-community/share/wPKRAHYbAlaEaCxUxcqVyaaaeZcYagDvqc" | |
invite_message = "Click to join our community org on the HF Hub!" | |
return (f"# β Verification successful! [Discord ID {user_id} <---> {profile.username}] π€\n{invite_message}\n{org_link} ") | |
print(f"# β Verification successful! [Discord ID {user_id} <---> {profile.username}] π€") | |
else: | |
return (f"# β The Discord account {user_id} is already verified! To change discord accounts or HF accounts, contact @lunarflu or [email protected]") | |
print(f"# β The Discord account {user_id} is already verified! To change discord accounts or HF accounts, contact @lunarflu or [email protected]") | |
# Remove the token after successful verification | |
del user_tokens[int(user_id)] | |
demo = gr.Blocks() | |
with demo: | |
try: | |
dataframe1 = pd.read_csv("https://docs.google.com/spreadsheets/d/1C8aLqgCqLYcMiIFf-P_Aosaa03C_WLIB_UyqvjSdWg8/export?format=csv&gid=0") # required for timing | |
print(dataframe1) | |
level_counts = dataframe1['discord_level'].value_counts().sort_index() | |
dataframe2 = pd.DataFrame({ | |
'Levels': level_counts.index, | |
'Members': level_counts.values | |
}) | |
print(dataframe2) | |
TITLE = """<h1 align="center" id="space-title">π€ Hugging Face Level Leaderboard</h1>""" | |
gr.HTML(TITLE) | |
with gr.Tabs(elem_classes="tab-buttons") as tabs: | |
#------------------------------------------------------------------------------ | |
with gr.TabItem("β Discord Verification", elem_id="verify-tab", id=2): | |
login_button = gr.LoginButton() | |
m1 = gr.Markdown() | |
demo.load(verify_button, inputs=None, outputs=m1) | |
def check_login_status(): | |
try: | |
return login_button.get_session().get("oauth_info", None) | |
except AttributeError: | |
return None | |
def check_login_wrapper(): | |
session = check_login_status() | |
if session is None: | |
return "Not logged in." | |
else: | |
return f"Logged in as {session.get('username', 'Unknown')}" | |
login_button.click(check_login_wrapper, inputs=None, outputs=m1) | |
#------------------------------------------------------------------------------ | |
with gr.TabItem("π Level leaderboard", elem_id="level-table", id=0): | |
#gr.Markdown("# π Experience Leaderboard") | |
with gr.Row(): | |
with gr.Column(): | |
gr.DataFrame(get_data, every=5, height=500, datatype=["markdown"], interactive=False, col_count=(3, "fixed"), column_widths=["100px","100px","100px"]) | |
with gr.Column(): | |
gr.BarPlot( | |
value=dataframe2, | |
x="Levels", | |
y="Members", | |
title="Level Distribution", | |
height=450, | |
width=450, | |
interactive=False | |
) | |
with gr.Row(): | |
gr.Markdown("# π How to earn Experience!") | |
with gr.Row(): | |
gr.DataFrame(get_data2, every=5, interactive=False) | |
#------------------------------------------------------------------------------ | |
with gr.TabItem("π Activity Heatmap", elem_id="activity-heatmap", id=1): | |
with gr.Row(): | |
username_input = gr.Textbox(label="Enter Username") | |
with gr.Row(): | |
year_buttons = [gr.Button(str(year)) for year in range(2021, 2025)] | |
with gr.Row(): | |
plot = gr.Plot(label="Activity Heatmap") | |
# buttons for updating the plot | |
for button in year_buttons: | |
button.click( | |
fn=create_plot, | |
inputs=[username_input, gr.State(int(button.value))], | |
outputs=plot | |
) | |
#with gr.TabItem("π Hub-only leaderboard", elem_id="hub-table", id=2): | |
except Exception as e: | |
print(f"gradio demo Error: {e}") | |
demo.queue().launch() |