Upload 7 files

src/assets/text_content.py

@@ -4,7 +4,7 @@ INTRODUCTION_TEXT = """
 <h6 align="center">
 The CLEM Leaderboard aims to track, rank and evaluate current cLLMs (chat-optimized Large Language Models) with the suggested pronounciation “clems”. 
 
-The benchmarking approach is described in [Clembench: Using Game Play to Evaluate Chat-Optimized Language Models as Conversational Agents](https://arxiv.org/abs/2305.13455).
+The benchmarking approach is described in [Clembench: Using Game Play to Evaluate Chat-Optimized Language Models as Conversational Agents](https://aclanthology.org/2023.emnlp-main.689.pdf).
 
 Source code for benchmarking "clems" is available here: [Clembench](https://github.com/clembench/clembench)
 

src/leaderboard_utils.py

@@ -3,12 +3,7 @@ import pandas as pd
 import requests, json
 from io import StringIO
 
-from datetime import datetime
-
-
 def get_github_data():
-    current_time = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
-    print(f"LOADING GITHUB DATAAA.... at time = {current_time}")
     '''
     Get data from csv files on Github
     Args:
@@ -18,7 +13,7 @@ def get_github_data():
         all_dfs: list of dataframes for previous versions + latest version including columns for all games
         all_vnames: list of the names for the previous versions + latest version (For Details and Versions Tab Dropdown)
     '''
-    uname = "kushal-10"
+    uname = "clembench"
     repo = "clembench-runs"
     json_url = f"https://raw.githubusercontent.com/{uname}/{repo}/main/benchmark_runs.json"
     resp = requests.get(json_url)

src/plot_utils.py

@@ -46,8 +46,7 @@ def plotly_plot(df:pd.DataFrame, LIST:list, ALL:list, NAMES:list, LEGEND:list):
         xaxis_title='% Played',
         yaxis_title='Quality Score',
         title='Overview of benchmark results',
-        height=1000,
-        width=1000
+        height=1000
     )
 
     fig.update_xaxes(range=[-5, 105])