Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
import asyncio | |
import csv | |
import datetime | |
import json | |
import logging | |
import os | |
import os.path | |
import random | |
import re | |
import threading | |
import time | |
import discord | |
import gradio as gr | |
import gradio_client | |
import gspread | |
import numpy as np | |
import pandas as pd | |
import requests | |
from apscheduler.executors.pool import ThreadPoolExecutor | |
from apscheduler.schedulers.background import BackgroundScheduler | |
from discord import Color, Embed | |
from discord.ext import commands | |
from gradio_client import Client | |
from gspread_dataframe import get_as_dataframe, set_with_dataframe | |
from gspread_formatting.dataframe import format_with_dataframe | |
from huggingface_hub import HfApi, list_liked_repos, list_metrics, list_models | |
from tabulate import tabulate | |
from datetime import datetime | |
DISCORD_TOKEN = os.environ.get("DISCORD_TOKEN", None) | |
intents = discord.Intents.all() | |
bot = commands.Bot(command_prefix='!', intents=intents) | |
#logger = logging.getLogger(__name__) | |
#logging.basicConfig(level=logging.DEBUG) | |
"""""" | |
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 | |
worksheet2 = gspread_bot.open("hf_discord_verified_users_test").sheet1 | |
"""""" | |
bot_ids = [1136614989411655780, 1166392942387265536, 1158038249835610123, 1130774761031610388, 1155489509518098565, 1155169841276260546, 1152238037355474964, 1154395078735953930] | |
"""""" | |
api = HfApi() | |
"""""" | |
#csv_file = 'data.csv' | |
global_df = pd.DataFrame() | |
print(type(global_df)) | |
community_global_df = pd.DataFrame() | |
data = pd.read_csv("https://docs.google.com/spreadsheets/d/1hQSsIg1Y9WdBF_CdNM1L1rUUREoxKqRTe3_ILo-WK8w/export?format=csv&gid=0") | |
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}") | |
# testing sheet -> read -> paste sheet | |
"""import data from google sheets -> HF Space .csv (doesn't make API call this way, as it's read-only)""" | |
data['discord_user_id'] = data['discord_user_id'].astype(str) | |
global_df = data | |
print(f"csv successfully retrieved: \n {global_df}") | |
#data.to_csv(csv_file, index=False) | |
await remove_huggingfolks() | |
print(f"------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"on_message Error: {e}") | |
def update_google_sheet(): | |
"""save data from HF Space -> google sheets (makes 1 API call)""" | |
# run this in separate thread? | |
try: | |
print("Updating...") | |
name = "levelbot" | |
#worksheet = gspread_bot.open(name).sheet1 | |
global_df['discord_user_id'] = global_df['discord_user_id'].astype(str) | |
# could just do this / format | |
set_with_dataframe(worksheet, global_df) | |
# very important, otherwise gspread will force format IDs in scientific notation by default (making them unusable) | |
worksheet.format(f"A1:A{len(global_df.index)+1}", { | |
"numberFormat": { | |
"type": 'TEXT', | |
}, | |
}) | |
timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S") | |
print(f"Google sheet {name} successfully updated at {timestamp}! \n{global_df}") | |
print(f"------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"on_message Error: {e}") | |
executor = ThreadPoolExecutor(max_workers=1) | |
scheduler = BackgroundScheduler(executors={'default': executor}) | |
scheduler.add_job(update_google_sheet, trigger='interval', minutes=2, max_instances=1) | |
scheduler.start() | |
""" | |
scheduler = BackgroundScheduler() | |
scheduler.add_job(update_google_sheet, "interval", seconds=60) | |
scheduler.start() | |
""" | |
def calculate_level(xp): | |
return int(xp ** (1.0 / 3.0)) | |
def calculate_xp(level): | |
return (int(level ** 3)) | |
processed_users = set() | |
async def periodic_api_test(): # needs rewrite, can do same thing and interact with csv -> push to google sheets | |
try: | |
await asyncio.sleep(0.1) | |
column_values_3 = worksheet2.col_values(3) | |
column_values_8 = worksheet2.col_values(8) | |
for i, user in enumerate(column_values_3): | |
if user in processed_users: | |
continue | |
url = f"https://huggingface.co/api/users/{user}/overview" | |
response = requests.get(url) | |
if response.status_code == 200: | |
data = response.json() | |
likes = data["numLikes"] | |
models = data["numModels"] | |
datasets = data["numDatasets"] | |
spaces = data["numSpaces"] | |
discussions = data["numDiscussions"] | |
papers = data["numPapers"] | |
upvotes = data["numUpvotes"] | |
worksheet2.update(values=[[likes, models, datasets, spaces, discussions, papers, upvotes]], | |
range_name=f'G{i+1}:M{i+1}') | |
processed_users.add(user) | |
else: | |
print(f"Failed to retrieve data for user {user}. Status code: {response.status_code}") | |
except Exception as e: | |
print(f"periodic_api_test Error: {e}") | |
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) | |
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, | |
} | |
# xp = int64 | |
# member_name = string | |
# discord_user_id = string | |
member_found = False | |
for index, cell_value in global_df.iloc[:, 0].items(): | |
if cell_value == str(member_id): | |
# 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'] | |
new_xp = old_xp + XP_PER_MESSAGE | |
global_df.loc[index, 'discord_exp'] = new_xp # do not change column name | |
print(f"Record for {member} updated from {old_xp} to {global_df.loc[index, 'discord_exp']} (+{XP_PER_MESSAGE}) ") | |
# level up | |
current_level = calculate_level(new_xp) | |
print(f"Current_level for {member}: {current_level}") | |
if current_level >= 2 and current_level <=30: | |
current_role = lvls[current_level] | |
if current_role not in member.roles: # if we leveled up | |
await member.add_roles(current_role) | |
print(f"Level Up! Gave {member} {current_role}") | |
await member.remove_roles(lvls[current_level-1]) | |
print(f"Removed {lvls[current_level-1]} from {member}") | |
#print(f"{member} Level up! {current_level-1} -> {current_level}!") | |
if current_role in member.roles: # needs update; reference exp reward for verification | |
#await member.send(f"Level up! {current_level-1} -> {current_level}!") | |
#if member_id == 811235357663297546: | |
row = global_df[global_df['discord_user_id'] == str(member_id)] # does this need to be string? | |
if not row.empty: | |
target_exp = row['discord_exp'].values[0] | |
rank = (global_df['discord_exp'] > target_exp).sum() + 1 | |
print(f"The rank for discord_id {member_id} based on discord_exp is: {rank}") | |
else: | |
print(f"Discord ID {member_id} not found in the DataFrame.") | |
# 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! `{current_level-1}` -> `{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) | |
# can use ^ to track changes in leaderboard ranking and display in levelup message when climbing | |
# (green red arrow thingies) | |
# could also pull ranks above and below yourself (if they exist, i.e.) | |
# 1 nateraw | |
# 2 [you are here] | |
# 3 osanseviero | |
# also, ahead by how much exp, how many levels maybe | |
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) | |
#You can verify your account to earn 100 points! To verify, do A. | |
print(f"------------------------------------------------------------------------") | |
if not member_found: | |
# if not, create new record | |
print(f"creating new record for {member}") | |
xp = 10 # define somewhere else? | |
member_id = str(member_id) | |
member_name = str(member.name) | |
current_level = calculate_level(xp) | |
row_data = [member_id, member_name, xp, current_level] | |
global_df.loc[len(global_df.index)] = row_data | |
# initial role assignment | |
if current_level == 1: | |
if lvl1 not in member.roles: | |
await member.add_roles(lvl1) | |
print(f"Gave {member} {lvl1}") # can log this better | |
print(f"------------------------------------------------------------------------") | |
except Exception as e: | |
print(f"add_exp Error: {e}") | |
async def on_message(message): | |
try: | |
if message.author.id not in bot_ids: # could change to if author does not have bot role (roleid) | |
if "!help_xp" not in message.content: | |
print(f"adding exp from message {message.author}") | |
await asyncio.sleep(0.1) | |
await add_exp(message.author.id) | |
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.id}") | |
await asyncio.sleep(0.1) | |
await add_exp(user.id) | |
await asyncio.sleep(0.1) | |
await add_exp(reaction.message.author.id) | |
except Exception as e: | |
print(f"on_reaction_add Error: {e}") | |
# non-huggingfolks data | |
async def remove_huggingfolks(): | |
try: | |
# remove huggingfolks | |
global community_global_df | |
community_global_df = global_df.copy() | |
guild = bot.get_guild(879548962464493619) | |
role = discord.utils.get(guild.roles, id=897376942817419265) | |
members_with_role = [member.id for member in guild.members if role in member.roles] | |
for member_id in members_with_role: | |
community_global_df = community_global_df[community_global_df.iloc[:, 0] != str(member_id)] | |
print(community_global_df) | |
community_global_df['discord_exp'] = pd.to_numeric(community_global_df['discord_exp']) | |
top_30_exp = community_global_df.nlargest(30, 'discord_exp') | |
top_30_exp.drop(top_30_exp.columns[0], axis=1, inplace=True) | |
top_30_exp['D'] = ['π₯','π₯','π₯'] | |
top_30_rows = top_30_exp.values.tolist() | |
print(top_30_rows) | |
channel = bot.get_channel(1197143964994773023) | |
message = await channel.fetch_message(1197148293164187678) | |
# put into message / leaderboard | |
new_table = tabulate(top_30_rows, headers=["Name", "Experience", "Level", "Rank"], tablefmt="plain") | |
await message.edit(content=f"Updated Leaderboard:\n```\n{new_table}\n```") | |
except Exception as e: | |
print(f"remove_huggingfolks Error: {e}") | |
async def update_leaderboard(): # needs rewrite | |
try: | |
# split into clear sections, discord bot only and gspread only | |
if ctx.author.id == 811235357663297546: | |
worksheet = gspread_bot.open("levelbot").sheet1 | |
# get from global_df | |
# update second global df | |
# use both in gradio and discord to present? | |
names_list = worksheet.col_values(2)[1:] | |
levels_list = worksheet.col_values(4)[1:] | |
exp_list = worksheet.col_values(3)[1:] | |
# for 3 lists | |
combined_list = [list(sublist) for sublist in zip(names_list, levels_list, exp_list)] | |
combined_list = [[name, int(level), int(exp)] for name, level, exp in combined_list] | |
combined_list = sorted(combined_list, key=lambda x: x[1], reverse=True) | |
print(combined_list) | |
top_results = combined_list[:num_results] | |
#print(top_results) | |
""" | |
# get position, then find that value in updated_names_list | |
levels_list = list(map(int, levels_list)) | |
data_pairs = list(zip(names_list, levels_list)) | |
sorted_data_pairs = sorted(data_pairs, key=lambda x: x[1], reverse=True) | |
top_data_pairs = sorted_data_pairs[:num_results] | |
""" | |
top_results = [r for r in top_results if r[0] not in members_with_role] | |
for name, level, xp in top_results: | |
print(f"Name: {name}, Level: {level}, Exp: {xp}") | |
def xp_required_to_next_level(current_level, current_xp): | |
level_floor_xp = calculate_xp(current_level) | |
level_ceiling_xp = calculate_xp(current_level+1) | |
xp_to_level_up = level_ceiling_xp - current_xp | |
return (xp_to_level_up) | |
except Exception as e: | |
print(f"update_leaderboard Error: {e}") | |
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"on_message Error: {e}") | |
async def count_users_with_role(ctx, role_id): | |
role = discord.utils.get(ctx.guild.roles, id=int(role_id)) | |
count = sum(1 for member in ctx.guild.members if role in member.roles) | |
await ctx.send(f"Number of users with the role '{role.name}': {count}") | |
# embeds with user pfps? | |
# name, pfp, time in server.... | |
# discord_level column | |
# pick 10 highest | |
# update | |
# weekly do different count | |
# count number of messages per user for every channel (total messages) | |
# fix sheet if necessary | |
# might need cell location data to pull both level and username at same time | |
# add emojis for some color | |
# check if members are still in the server | |
"""""" | |
DISCORD_TOKEN = os.environ.get("DISCORD_TOKEN", None) | |
def run_bot(): | |
bot.run(DISCORD_TOKEN) | |
threading.Thread(target=run_bot).start() | |
URL = "https://docs.google.com/spreadsheets/d/1hQSsIg1Y9WdBF_CdNM1L1rUUREoxKqRTe3_ILo-WK8w/edit#gid=0" | |
csv_url = URL.replace('/edit#gid=', '/export?format=csv&gid=') | |
def get_data(): | |
try: | |
data = pd.read_csv(csv_url) | |
first_3_columns = data.iloc[:, 1:4] | |
sorted = first_3_columns.sort_values(by='discord_exp', ascending=False) | |
sorted.to_csv('first_3_columns.csv', index=False) | |
return sorted | |
except Exception as e: | |
print(f"on_message Error: {e}") | |
# csv | |
# read into pandas dataframe1 | |
# read levels column and create pandas dataframe2 with first column containing levels from 2-max found in dataframe1 | |
# create second column in dataframe2 for number of each level found in dataframe1 levels column | |
demo = gr.Blocks() | |
with demo: | |
try: | |
dataframe1 = pd.read_csv(csv_url) | |
column_values_unique = sorted(dataframe1.iloc[:, 3].unique()) | |
dataframe2 = pd.DataFrame({'Levels': column_values_unique}) | |
counts = {} | |
for value in data.iloc[:, 3]: | |
counts[value] = counts.get(value, 0) + 1 | |
dataframe2['Members'] = dataframe2['Levels'].map(counts) | |
print("\nDataframe 2:") | |
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("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, 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.TabItem("π Members of the Week", elem_id="week-table", id=1): | |
#with gr.TabItem("π Hub-only leaderboard", elem_id="hub-table", id=2): | |
except Exception as e: | |
print(f"on_message Error: {e}") | |
demo.queue().launch() | |