inti commit

.gitignore

@@ -0,0 +1,6 @@
+
+*.pyc
+ZeroEval-main/.DS_Store
+ZeroEval-main/result_dirs/.DS_Store
+ZeroEval-main/result_dirs/zebra-grid/.DS_Store
+.DS_Store

README.md

@@ -1,13 +1,60 @@
 ---
-title: ZebraLogicBench Leaderboard
-emoji: 📉
+title: Zebra Logic Bench
+emoji: 🦓
 colorFrom: blue
-colorTo: red
+colorTo: yellow
 sdk: gradio
-sdk_version: 4.37.2
+sdk_version: 4.19.2
 app_file: app.py
-pinned: false
-license: apache-2.0
+pinned: true
+fullWidth: true
+hf_oauth: true
+api: false
+tags: 
+    - leaderboard
+datasets: 
+    - allenai/ZebraLogicBench
+    - allenai/ZebraLogicBench-private
+models: 
+    - Qwen/Qwen2-72B-Instruct
+    - Qwen/Qwen1.5-72B-Chat
+    - Qwen/Qwen1.5-7B-Chat
+    - meta-llama/Meta-Llama-3-8B-Instruct
+    - meta-llama/Meta-Llama-3-70B-Instruct
+    - meta-llama/Llama-2-13b-chat-hf
+    - meta-llama/Llama-2-70b-chat-hf
+    - meta-llama/Llama-2-7b-chat-hf
+    - mistralai/Mistral-7B-Instruct-v0.1
+    - mistralai/Mistral-7B-Instruct-v0.2
+    - mistralai/Mixtral-8x7B-Instruct-v0.1
+    - microsoft/Phi-3-medium-128k-instruct
+    - microsoft/Phi-3-mini-128k-instruct
+    - NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO
+    - NousResearch/Hermes-2-Theta-Llama-3-8B
+    - 01-ai/Yi-1.5-34B-Chat
+    - 01-ai/Yi-1.5-9B-Chat
+    - 01-ai/Yi-1.5-6B-Chat
+    - google/gemma-7b-it
+    - google/gemma-2b-it
+    - allenai/tulu-2-dpo-70b
+    - HuggingFaceH4/zephyr-7b-beta
+    - Nexusflow/Starling-LM-7B-beta
+    - databricks/dbrx-instruct
+    - princeton-nlp/Llama-3-Instruct-8B-SimPO
+    - chujiezheng/Llama-3-Instruct-8B-SimPO-ExPO
+    - chujiezheng/Starling-LM-7B-beta-ExPO
+    - ZhangShenao/SELM-Zephyr-7B-iter-3
+    - deepseek-ai/DeepSeek-V2-Chat
+    - m-a-p/neo_7b_instruct_v0.1
+    - 01-ai/Yi-34B-chat
+    - lmsys/vicuna-13b-v1.5
+    - HuggingFaceH4/zephyr-7b-gemma-v0.1
+    - deepseek-ai/DeepSeek-Coder-V2
+    - THUDM/glm-4-9b-chat
+    - chujiezheng/neo_7b_instruct_v0.1-ExPO
+    - ZhangShenao/SELM-Llama-3-8B-Instruct-iter-3
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+Paper: arxiv.org/abs/2406.04770

ZeroEval-main/result_dirs/zebra-grid.summary.json

@@ -0,0 +1,321 @@
+[
+  {
+    "Model": "claude-3-5-sonnet-20240620",
+    "Mode": "greedy",
+    "Puzzle Acc": "33.40",
+    "Cell Acc": "54.34",
+    "No answer": "0.00",
+    "Easy Puzzle Acc": "87.50",
+    "Hard Puzzle Acc": "12.36",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1141.94"
+  },
+  {
+    "Model": "claude-3-5-sonnet-20240620",
+    "Mode": "sampling",
+    "Puzzle Acc": "33.40",
+    "Cell Acc": "53.01",
+    "No answer": "0.10",
+    "Easy Puzzle Acc": "88.21",
+    "Hard Puzzle Acc": "12.08",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1153.83"
+  },
+  {
+    "Model": "gpt-4o-2024-05-13",
+    "Mode": "sampling",
+    "Puzzle Acc": "30.80",
+    "Cell Acc": "46.19",
+    "No answer": "6.60",
+    "Easy Puzzle Acc": "81.07",
+    "Hard Puzzle Acc": "11.25",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1549.74"
+  },
+  {
+    "Model": "gpt-4-turbo-2024-04-09",
+    "Mode": "greedy",
+    "Puzzle Acc": "28.40",
+    "Cell Acc": "47.90",
+    "No answer": "0.10",
+    "Easy Puzzle Acc": "80.71",
+    "Hard Puzzle Acc": "8.06",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1148.46"
+  },
+  {
+    "Model": "gpt-4o-2024-05-13",
+    "Mode": "greedy",
+    "Puzzle Acc": "28.20",
+    "Cell Acc": "38.72",
+    "No answer": "19.30",
+    "Easy Puzzle Acc": "77.86",
+    "Hard Puzzle Acc": "8.89",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1643.51"
+  },
+  {
+    "Model": "gpt-4-0314",
+    "Mode": "greedy",
+    "Puzzle Acc": "27.10",
+    "Cell Acc": "47.43",
+    "No answer": "0.20",
+    "Easy Puzzle Acc": "77.14",
+    "Hard Puzzle Acc": "7.64",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1203.17"
+  },
+  {
+    "Model": "claude-3-opus-20240229",
+    "Mode": "greedy",
+    "Puzzle Acc": "27.00",
+    "Cell Acc": "48.91",
+    "No answer": "0.00",
+    "Easy Puzzle Acc": "78.21",
+    "Hard Puzzle Acc": "7.08",
+    "Total Puzzles": 1000,
+    "Reason Lens": "855.72"
+  },
+  {
+    "Model": "gpt-4-turbo-2024-04-09",
+    "Mode": "sampling",
+    "Puzzle Acc": "26.40",
+    "Cell Acc": "47.93",
+    "No answer": "0.00",
+    "Easy Puzzle Acc": "74.29",
+    "Hard Puzzle Acc": "7.78",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1165.90"
+  },
+  {
+    "Model": "deepseek-chat",
+    "Mode": "greedy",
+    "Puzzle Acc": "22.70",
+    "Cell Acc": "42.46",
+    "No answer": "5.20",
+    "Easy Puzzle Acc": "68.57",
+    "Hard Puzzle Acc": "4.86",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1260.23"
+  },
+  {
+    "Model": "Qwen2-72B-Instruct",
+    "Mode": "greedy",
+    "Puzzle Acc": "21.40",
+    "Cell Acc": "38.32",
+    "No answer": "10.20",
+    "Easy Puzzle Acc": "63.93",
+    "Hard Puzzle Acc": "4.86",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1813.82"
+  },
+  {
+    "Model": "deepseek-coder",
+    "Mode": "greedy",
+    "Puzzle Acc": "21.10",
+    "Cell Acc": "41.58",
+    "No answer": "4.90",
+    "Easy Puzzle Acc": "64.64",
+    "Hard Puzzle Acc": "4.17",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1324.55"
+  },
+  {
+    "Model": "gemini-1.5-pro",
+    "Mode": "sampling",
+    "Puzzle Acc": "19.70",
+    "Cell Acc": "45.24",
+    "No answer": "0.40",
+    "Easy Puzzle Acc": "60.00",
+    "Hard Puzzle Acc": "4.03",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1356.77"
+  },
+  {
+    "Model": "gemini-1.5-flash",
+    "Mode": "greedy",
+    "Puzzle Acc": "19.40",
+    "Cell Acc": "31.77",
+    "No answer": "22.70",
+    "Easy Puzzle Acc": "59.29",
+    "Hard Puzzle Acc": "3.89",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1538.18"
+  },
+  {
+    "Model": "gemini-1.5-pro",
+    "Mode": "greedy",
+    "Puzzle Acc": "19.40",
+    "Cell Acc": "44.59",
+    "No answer": "0.80",
+    "Easy Puzzle Acc": "55.71",
+    "Hard Puzzle Acc": "5.28",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1336.17"
+  },
+  {
+    "Model": "yi-large-preview",
+    "Mode": "greedy",
+    "Puzzle Acc": "18.90",
+    "Cell Acc": "42.61",
+    "No answer": "1.40",
+    "Easy Puzzle Acc": "58.93",
+    "Hard Puzzle Acc": "3.33",
+    "Total Puzzles": 1000,
+    "Reason Lens": "833.36"
+  },
+  {
+    "Model": "yi-large",
+    "Mode": "greedy",
+    "Puzzle Acc": "18.80",
+    "Cell Acc": "39.83",
+    "No answer": "1.80",
+    "Easy Puzzle Acc": "58.21",
+    "Hard Puzzle Acc": "3.47",
+    "Total Puzzles": 1000,
+    "Reason Lens": "757.01"
+  },
+  {
+    "Model": "claude-3-sonnet-20240229",
+    "Mode": "greedy",
+    "Puzzle Acc": "18.70",
+    "Cell Acc": "43.66",
+    "No answer": "0.00",
+    "Easy Puzzle Acc": "58.93",
+    "Hard Puzzle Acc": "3.06",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1095.37"
+  },
+  {
+    "Model": "Qwen2-72B-Instruct",
+    "Mode": "sampling",
+    "Puzzle Acc": "18.70",
+    "Cell Acc": "40.57",
+    "No answer": "3.20",
+    "Easy Puzzle Acc": "57.50",
+    "Hard Puzzle Acc": "3.61",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1894.72"
+  },
+  {
+    "Model": "gemini-1.5-flash",
+    "Mode": "sampling",
+    "Puzzle Acc": "18.40",
+    "Cell Acc": "36.03",
+    "No answer": "12.80",
+    "Easy Puzzle Acc": "57.86",
+    "Hard Puzzle Acc": "3.06",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1713.03"
+  },
+  {
+    "Model": "Meta-Llama-3-70B-Instruct",
+    "Mode": "greedy",
+    "Puzzle Acc": "16.80",
+    "Cell Acc": "42.31",
+    "No answer": "0.20",
+    "Easy Puzzle Acc": "52.86",
+    "Hard Puzzle Acc": "2.78",
+    "Total Puzzles": 1000,
+    "Reason Lens": "809.95"
+  },
+  {
+    "Model": "gemma-2-27b-it@nvidia",
+    "Mode": "greedy",
+    "Puzzle Acc": "16.30",
+    "Cell Acc": "41.18",
+    "No answer": "1.10",
+    "Easy Puzzle Acc": "50.71",
+    "Hard Puzzle Acc": "2.92",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1014.56"
+  },
+  {
+    "Model": "claude-3-haiku-20240307",
+    "Mode": "greedy",
+    "Puzzle Acc": "14.30",
+    "Cell Acc": "37.87",
+    "No answer": "0.10",
+    "Easy Puzzle Acc": "47.86",
+    "Hard Puzzle Acc": "1.25",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1015.06"
+  },
+  {
+    "Model": "reka-core-20240501",
+    "Mode": "greedy",
+    "Puzzle Acc": "13.00",
+    "Cell Acc": "33.88",
+    "No answer": "4.00",
+    "Easy Puzzle Acc": "43.21",
+    "Hard Puzzle Acc": "1.25",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1078.29"
+  },
+  {
+    "Model": "gemma-2-9b-it",
+    "Mode": "greedy",
+    "Puzzle Acc": "12.90",
+    "Cell Acc": "37.07",
+    "No answer": "0.50",
+    "Easy Puzzle Acc": "42.14",
+    "Hard Puzzle Acc": "1.53",
+    "Total Puzzles": 1000,
+    "Reason Lens": "859.14"
+  },
+  {
+    "Model": "gemma-2-9b-it@nvidia",
+    "Mode": "greedy",
+    "Puzzle Acc": "12.80",
+    "Cell Acc": "36.79",
+    "No answer": "0.00",
+    "Easy Puzzle Acc": "41.79",
+    "Hard Puzzle Acc": "1.53",
+    "Total Puzzles": 1000,
+    "Reason Lens": "849.84"
+  },
+  {
+    "Model": "Meta-Llama-3-8B-Instruct",
+    "Mode": "greedy",
+    "Puzzle Acc": "11.90",
+    "Cell Acc": "23.70",
+    "No answer": "29.20",
+    "Easy Puzzle Acc": "40.71",
+    "Hard Puzzle Acc": "0.69",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1216.40"
+  },
+  {
+    "Model": "gpt-3.5-turbo-0125",
+    "Mode": "greedy",
+    "Puzzle Acc": "10.10",
+    "Cell Acc": "33.06",
+    "No answer": "0.10",
+    "Easy Puzzle Acc": "33.57",
+    "Hard Puzzle Acc": "0.97",
+    "Total Puzzles": 1000,
+    "Reason Lens": "820.66"
+  },
+  {
+    "Model": "reka-flash-20240226",
+    "Mode": "greedy",
+    "Puzzle Acc": "9.30",
+    "Cell Acc": "25.67",
+    "No answer": "18.70",
+    "Easy Puzzle Acc": "30.71",
+    "Hard Puzzle Acc": "0.97",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1074.80"
+  },
+  {
+    "Model": "Qwen2-7B-Instruct",
+    "Mode": "greedy",
+    "Puzzle Acc": "8.40",
+    "Cell Acc": "22.06",
+    "No answer": "24.40",
+    "Easy Puzzle Acc": "29.29",
+    "Hard Puzzle Acc": "0.28",
+    "Total Puzzles": 1000,
+    "Reason Lens": "1473.23"
+  }
+]

_about_us.md

@@ -0,0 +1,16 @@
+## About Us
+
+### Team 
+
+We are from [AllenAI](https://allenai.org/) (AI2), a non-profit research organization.  
+
+[Bill Yuchen Lin](https://yuchenlin.xyz/), [Ronan Le Bras](https://rlebras.github.io/), and [Yejin Choi](https://homes.cs.washington.edu/~yejin/).
+ 
+ 
+### Contact
+
+Please contact us in the following ways:
+- Github Issues/PRs for adding a new model: [https://github.com/allenai/WildBench](https://github.com/allenai/WildBench)
+- HF Discussions for general questions about the leaderboard: [https://huggingface.co/spaces/allenai/WildBench/discussions](https://huggingface.co/spaces/allenai/WildBench/discussions)
+- Other questions: Please contact Yuchen with email: yuchenl[at]allenai[dot]org
+ 

_header.md

@@ -0,0 +1,5 @@
+<br/>
+
+# 🦁 WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild
+[📑 Paper](https://allenai.github.io/WildBench/WildBench_paper.pdf) | [💻 GitHub](https://github.com/allenai/WildBench) | [🤗 HuggingFace](https://huggingface.co/collections/allenai/wildbench-65e8f2fa9c1260a85a933627) | [🐦 X](https://x.com/billyuchenlin/status/1795746137875554531) | [💬 Discussion](https://huggingface.co/spaces/allenai/WildBench/discussions) | ⚙️ **Version**: **V2** | **# Models**: {model_num} | Updated: **{LAST_UPDATED}**
+ 

_metrics.md

@@ -0,0 +1 @@
+## 

app.py

@@ -0,0 +1,164 @@
+"""A gradio app that renders a static leaderboard. This is used for Hugging Face Space."""
+import ast
+import argparse
+import glob
+import pickle
+import plotly
+import gradio as gr
+import numpy as np
+import pandas as pd
+import gradio as gr
+import pandas as pd
+from pathlib import Path
+import json
+from constants import *
+from datetime import datetime, timezone 
+# from datasets import Dataset, load_dataset, concatenate_datasets
+import os, uuid 
+from utils_display import model_info
+from constants import column_names,  LEADERBOARD_REMARKS, DEFAULT_K, LEADERBOARD_REMARKS_MAIN
+import pytz
+from data_utils import post_processing
+
+# get the last updated time from the elo_ranks.all.jsonl file
+LAST_UPDATED = None 
+# with open("_intro.md", "r") as f:
+#     INTRO_MD = f.read()
+INTRO_MD = ""
+with open("_about_us.md", "r") as f:
+    ABOUT_MD = f.read()
+
+with open("_header.md", "r") as f:
+    HEADER_MD = f.read()
+
+with open("_metrics.md", "r") as f:
+    METRICS_MD = f.read()
+ 
+original_df = None  
+# available_models = [] # to be filled in later
+available_models = list(model_info.keys()) 
+ 
+def _tab_leaderboard():
+    global original_df, available_models, gpt4t_dfs, haiku_dfs, llama_dfs, score_df
+    with gr.TabItem("📊 Main", elem_id="od-benchmark-tab-table-ablation", id=0, elem_classes="subtab"): 
+        default_main_df = original_df.copy() 
+        default_main_df.insert(0, "", range(1, 1 + len(default_main_df)))
+        default_main_df_no_task = default_main_df.copy()
+        # default_main_df_no_task = hide_task_column(default_main_df)
+        # default_main_df_no_task = rerank(default_main_df_no_task, rank_column=WB_ELO_COLUMN)
+        # default_main_df_no_task = rerank(default_main_df_no_task, rank_column=HYBRID_AVG_COLUMN)
+        with gr.Row():
+            # with gr.Column(scale=5):
+                # gr.Markdown(LEADERBOARD_REMARKS_MAIN, elem_classes="markdown-text-small top-left-LP") 
+                # with gr.Row():
+                    # with gr.Column(scale=2):
+                        # md = gr.Markdown(" ### 👀 More presentation options ⬇️", elem_classes="markdown-text")
+            
+                    # with gr.Column(scale=3):
+            # with gr.Column(scale=2): 
+                # gr.Markdown(f"""**__🪧  Default options:__**  K={DEFAULT_K}; Hybrid-Macro; for best corr w/ LMSYS Elo.""", elem_classes="markdown-text")
+            
+            
+                # gr.Markdown(LENGTH_MARGIN_DESC_MD, elem_classes="markdown-text-tiny no_margin")
+            with gr.Column(scale=5):
+                with gr.Accordion("💬 Metric explanations", open=False, elem_classes="accordion-label"):
+                    gr.Markdown(LEADERBOARD_REMARKS_MAIN, elem_classes="markdown-text-small no_margin") 
+                rank_column_radio = gr.Radio(["🆚+💯 Hybrid", "🆚 Reward-Mix (Pairwise)", "💯 Score (Individual)", "🌟 WB Elo (beta)" ], show_label=False, elem_id="rank-column-radio", 
+                                             value="🌟 WB Elo (beta)"
+                                            # value="🆚+💯 Hybrid"
+                                             )
+            with gr.Column(scale=2): 
+                with gr.Row():
+                    checkbox_show_task_categorized = gr.Checkbox(label="🆚 by Task Type", elem_id="show-task-categorized", value=False)
+                    show_open_source_model_only = gr.Checkbox(label="🔑 Open Models", elem_id="show-open-source-models", value=False)
+        # with gr.Row():
+        #     with gr.Column(scale=2):
+                
+        leaderboard_table = gr.components.Dataframe(
+            value=default_main_df_no_task,
+            datatype= ["number", "markdown", "markdown", "number"],
+            # max_rows=None,
+            height=6000,
+            elem_id="leaderboard-table",
+            interactive=False,
+            visible=True,
+            column_widths=[50, 260,120, 120, 120, 130,100,100,110,100],
+            wrap=True
+            # min_width=60,
+        ) 
+        # checkbox_show_task_categorized.change(fn=length_margin_change, inputs=[length_margin_choices, gr.Text("main", visible=False), checkbox_show_task_categorized, show_open_source_model_only, rank_column_radio], outputs=[leaderboard_table])
+        # show_open_source_model_only.change(fn=length_margin_change, inputs=[length_margin_choices, gr.Text("main", visible=False), checkbox_show_task_categorized, show_open_source_model_only, rank_column_radio], outputs=[leaderboard_table])
+        # rank_column_radio.change(fn=length_margin_change, inputs=[length_margin_choices, gr.Text("main", visible=False), checkbox_show_task_categorized, show_open_source_model_only, rank_column_radio], outputs=[leaderboard_table])
+ 
+
+ 
+def _tab_submit():
+    pass
+
+
+def build_demo():
+    global original_df, available_models, gpt4t_dfs, haiku_dfs, llama_dfs
+
+    with gr.Blocks(theme=gr.themes.Soft(), css=css, js=js_light) as demo:
+        gr.HTML(BANNER, elem_id="banner")
+        # convert LAST_UPDATED to the PDT time 
+        LAST_UPDATED = datetime.now(pytz.timezone('US/Pacific')).strftime("%Y-%m-%d %H:%M:%S")
+        # header_md_text = HEADER_MD.replace("{model_num}", str(len(original_df["-1"]))).replace("{LAST_UPDATED}", str(LAST_UPDATED))
+        # gr.Markdown(header_md_text, elem_classes="markdown-text") 
+
+        with gr.Tabs(elem_classes="tab-buttons") as tabs: 
+            with gr.TabItem("🏅 Leaderboard", elem_id="od-benchmark-tab-table", id=0):
+                _tab_leaderboard() 
+
+            with gr.TabItem("🚀 Submit Your Results", elem_id="od-benchmark-tab-table", id=3):
+                _tab_submit() 
+
+            with gr.TabItem("📮 About Us", elem_id="od-benchmark-tab-table", id=4):
+                gr.Markdown(ABOUT_MD, elem_classes="markdown-text")
+        
+        with gr.Row():
+            with gr.Accordion("📙 Citation", open=False, elem_classes="accordion-label"):
+                gr.Textbox(
+                    value=CITATION_TEXT, 
+                    lines=7,
+                    label="Copy the BibTeX snippet to cite this source",
+                    elem_id="citation-button",
+                    show_copy_button=True)
+                # ).style(show_copy_button=True)
+
+    return demo 
+
+
+
+def data_load(result_file):
+    global original_df
+    print(f"Loading {result_file}")
+    column_names_main = column_names.copy()
+    # column_names_main.update({})
+    main_ordered_columns = ORDERED_COLUMN_NAMES 
+    click_url = True 
+    # read json file from the result_file 
+    with open(result_file, "r") as f:
+        data = json.load(f)
+    # floatify the data, if possible
+    for d in data:
+        for k, v in d.items():
+            try:
+                d[k] = float(v)
+            except:
+                pass
+    original_df = pd.DataFrame(data)
+    original_df = post_processing(original_df, column_names_main, ordered_columns=main_ordered_columns, click_url=click_url, rank_column=RANKING_COLUMN)
+    # print(original_df.columns) 
+    
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--share", action="store_true")
+    parser.add_argument("--result_file", help="Path to results table", default="ZeroEval-main/result_dirs/zebra-grid.summary.json")
+    
+    args = parser.parse_args()
+    data_load(args.result_file)    
+    print(original_df)
+    demo = build_demo()
+    demo.launch(share=args.share, height=3000, width="100%")

constants.py

@@ -0,0 +1,283 @@
+from pathlib import Path
+from collections import OrderedDict
+
+DEFAULT_K = "∞"
+# DEFAULT_K = "1500"
+
+banner_url = "https://allenai.github.io/WildBench/gray_banner.png" # the same repo here.
+BANNER = f'<div style="display: flex; justify-content: flex-start;"><img src="{banner_url}" alt="Banner" style="width: 40vw; min-width: 300px; max-width: 800px;"> </div>'
+
+TITLE = "<html> <head> <style> h1 {text-align: center;} </style> </head> <body> <h1> 🦁 AI2 WildBench Leaderboard </b> </body> </html>"
+ 
+WINRATE_HEATMAP = "<div><img src='https://github.com/WildEval/WildBench-Leaderboard/blob/main/gradio/pairwise_win_fractions.png?raw=true' style='width:100%;'></div>"
+
+CITATION_TEXT = """@misc{lin2024wildbench,
+    title={WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild},
+    author={Bill Yuchen Lin and Yuntian Deng and Khyathi Chandu and Faeze Brahman and Abhilasha Ravichander and Valentina Pyatkin and Nouha Dziri and Ronan Le Bras and Yejin Choi},
+    year={2024},
+    eprint={2406.04770},
+    archivePrefix={arXiv},
+    primaryClass={cs.CL},
+    url={https://arxiv.org/abs/2406.04770}
+}
+"""
+
+# make column_names as an ordered dict
+ 
+
+
+column_names = OrderedDict({
+    "Model": "Model",   
+    "Mode": "Mode",
+    "Puzzle Acc": "Puzzle Acc",
+    "Cell Acc": "Cell Acc",
+    "No answer": "No answer",
+    "Easy Puzzle Acc": "Easy Puzzle Acc",
+    "Hard Puzzle Acc": "Hard Puzzle Acc",
+    # "Total Puzzles": "Total Puzzles",
+    # "Reason Lens": "Reason Lens",
+})
+
+
+
+LEADERBOARD_REMARKS = """**WB Reward**: for each comparison (A vs B), a reward for A is **+/-1** if A is **much better/worse** than B, and **+/-0.5** if A is **slightly better/worse** than B; when there is a **Tie**, the reward is **0**.  
+"""
+
+# **WB Reward**: for each pairwise comparison, a reward for A is **+/-1** if A is **much better/worse** than B, and **+/-0.5** if A is **slightly better/worse** than B; 0 for a **Tie**.
+# The baseline models are GPT4-Turbo, Haiku, and Llama2-70B, and Mix is the average of the three. 
+# **WB Score** individually scores each model based on checklists.
+# Evaluator is GPT-4-Turbo.
+LEADERBOARD_REMARKS_MAIN = """ 
+"""
+  
+RANKING_COLUMN = "Puzzle Acc"
+
+ORDERED_COLUMN_NAMES = [
+    "Model", 
+    "Mode",
+    "Puzzle Acc",
+    "Easy Puzzle Acc",
+    "Hard Puzzle Acc",
+    "Cell Acc",
+    "No answer", 
+]
+
+ 
+js_light = """
+function refresh() {
+    const url = new URL(window.location);
+
+    if (url.searchParams.get('__theme') !== 'light') {
+        url.searchParams.set('__theme', 'light');
+        window.location.href = url.href;
+    }
+
+    // Find the fieldset with the given id
+    const fieldset = document.getElementById("rank-column-radio");
+
+    // Create a new span element with the text "Rank by:"
+    const rankBySpan = document.createElement("span");
+    rankBySpan.textContent = "Rank by: ";
+    rankBySpan.style.fontWeight = "bold"; // Optional: make the text bold
+     rankBySpan.style.fontSize = "19px"; // Larger font size
+    rankBySpan.style.paddingRight = "18px"; // Add padding on the right
+
+    // Wrap the span and the labels in a flex container
+    const flexContainer = document.createElement("div");
+    flexContainer.style.display = "flex";
+    flexContainer.style.alignItems = "center";
+
+    // Insert the rankBySpan at the beginning of the flex container
+    flexContainer.appendChild(rankBySpan);
+
+    // Move all existing labels into the flex container
+    while (fieldset.firstChild) {
+        flexContainer.appendChild(fieldset.firstChild);
+    }
+
+    // Append the flex container back to the fieldset
+    fieldset.appendChild(flexContainer);
+}
+"""
+
+js_code = """
+function scroll_top() {
+    console.log("Hello from Gradio!");  
+    const bubbles = document.querySelectorAll('.bubble-wrap');
+    bubbles.forEach((bubble, index) => {
+        setTimeout(() => {
+            bubble.scrollTop = 0;
+        }, index * 100); // Delay of 100ms between each iteration
+    });
+    
+} 
+"""
+
+
+TASK_TYPE_STR = "**Tasks**: Info seeking (**InfoSek**), Creative Writing (**CrtWrt**), Coding&Debugging (**Code**), Reasoning (**Reason**), Editing (**Edit**), **Math**, Planning (**Plan**), Brainstorming (**Brnstrm**), Role playing (**RolPly**), Advice seeking (**AdvSek**), Data Analysis (**DataAna**)"
+
+css = """
+
+ 
+
+code {
+    font-size: large;
+}
+footer {visibility: hidden}
+.top-left-LP{
+    margin-top: 6px;
+    margin-left: 5px;
+}
+.no_margin{
+    margin-top: 0px;
+    margin-left: 0px;
+    margin-right: 0px;
+    margin-bottom: 0px;
+    padding-top: 0px;
+    padding-left: 0px;
+    padding-right: 0px;
+    padding-bottom: 0px;
+}
+.markdown-text{font-size: 14pt}
+.markdown-text-tiny{font-size: 10pt}
+.markdown-text-small{font-size: 13pt}
+.markdown-text-tiny{font-size: 12pt}
+.markdown-text-tiny-red{
+    font-size: 12pt;
+    color: red;
+    background-color: yellow;
+    font-color: red;
+    font-weight: bold;
+}
+th {
+  text-align: center;
+  font-size: 17px; /* Adjust the font size as needed */
+}
+td {
+  font-size: 15px; /* Adjust the font size as needed */
+  text-align: center;
+}
+
+.sample_button{
+    border: 2px solid #000000;
+    border-radius: 10px;
+    padding: 10px;
+    font-size: 17pt;
+    font-weight: bold;
+    margin: 5px;
+    background-color: #D8BFD8;
+}
+
+.chat-common{
+    height: auto;
+    max-height: 400px;
+    min-height: 100px; 
+}
+.chat-specific{
+    height: auto;
+    max-height: 600px;
+    min-height: 200px; 
+}
+#od-benchmark-tab-table-button{
+    font-size: 15pt;
+    font-weight: bold;
+} 
+
+.btn_boderline{
+    border: 1px solid #000000;
+    border-radius: 5px;
+    padding: 5px;
+    margin: 5px;
+    font-size: 15pt;
+    font-weight: bold; 
+}
+
+.btn_boderline_next{
+    border: 0.1px solid #000000;
+    border-radius: 5px;
+    padding: 5px;
+    margin: 5px;
+    font-size: 15pt;
+    font-weight: bold; 
+}
+
+.btn_boderline_gray{
+    border: 0.5px solid gray;
+    border-radius: 5px;
+    padding: 5px;
+    margin: 5px;
+    font-size: 15pt;
+    font-weight: italic; 
+}
+.btn_boderline_selected{
+    border: 2px solid purple;
+    background-color: #f2f2f2;
+    border-radius: 5px;
+    padding: 5px;
+    margin: 5px;
+    font-size: 15pt;
+    font-weight: bold;  
+}
+.accordion-label button span{
+    font-size: 14pt;
+    font-weight: bold;
+} 
+
+#show-task-categorized span{
+    font-size: 13pt;
+    font-weight: bold;
+}
+
+#show-open-source-models span{
+    font-size: 13pt;
+    font-weight: bold;
+}
+
+#select-models span{
+    font-size: 10pt;
+}
+
+#select-tasks span{
+    font-size: 10pt;
+}
+
+
+.markdown-text-details{
+    margin: 10px;
+    padding: 10px;
+}
+
+
+button.selected[role="tab"][aria-selected="true"] {
+    font-size: 18px; /* or any other size you prefer */
+    font-weight: bold;
+}
+
+#od-benchmark-tab-table-ablation-button {
+    font-size: larger; /* Adjust the font size as needed */
+}
+
+
+.plotly-plot{
+    height: auto;
+    max-height: 600px;
+    min-height: 600px; 
+}
+
+#length-margin-radio{
+    font-size: 10pt;
+    # padding: 0px;
+    # margin: 1px;
+}
+
+#show-task-categorized{
+    font-size: 12pt; 
+    font-decoration: bold;
+}
+
+#show-open-source-models{
+    font-size: 12pt; 
+    font-decoration: bold;
+}
+"""
+ 

data_utils.py

@@ -0,0 +1,46 @@
+from datasets import load_dataset, Dataset
+import os 
+from datasets import load_dataset
+from datasets.utils.logging import disable_progress_bar
+from constants import column_names, RANKING_COLUMN, ORDERED_COLUMN_NAMES
+from utils_display import make_clickable_model
+
+import random 
+disable_progress_bar()
+import math 
+import json 
+from tqdm import tqdm
+import numpy as np
+
+id_to_data = None 
+model_len_info = None 
+bench_data = None 
+eval_results = None 
+score_eval_results = None  
+
+# Formats the columns
+def formatter(x):
+    if type(x) is str:
+        x = x
+    else: 
+        x = round(x, 1)
+    return x
+ 
+
+def post_processing(df, column_names, rank_column=RANKING_COLUMN, ordered_columns=ORDERED_COLUMN_NAMES, click_url=True):
+    for col in df.columns:
+        if col == "Model" and click_url:
+            df[col] = df[col].apply(lambda x: x.replace(x, make_clickable_model(x)))
+        else:
+            df[col] = df[col].apply(formatter) # For numerical values 
+        if "Elo" in col:
+            df[col] = df[col].replace('-', np.nan).astype(float)
+
+
+    df.rename(columns=column_names, inplace=True)
+    list_columns = [col for col in ordered_columns if col in df.columns]
+    df = df[list_columns]
+    if rank_column in df.columns:
+        df.sort_values(by=rank_column, inplace=True, ascending=False)
+    return df
+  

model_info.json

@@ -0,0 +1,65 @@
+{
+    "Qwen2-72B-Instruct": {"pretty_name": "Qwen2-72B-Instruct", "hf_model_id": "Qwen/Qwen2-72B-Instruct"},
+    "Qwen2-7B-Instruct": {"pretty_name": "Qwen2-7B-Instruct", "hf_model_id": "Qwen/Qwen2-7B-Instruct"},
+    "Qwen1.5-72B-Chat-greedy": {"pretty_name": "Qwen1.5-72B-Chat", "hf_model_id": "Qwen/Qwen1.5-72B-Chat"},
+    "Qwen1.5-7B-Chat": {"pretty_name": "Qwen1.5-7B-Chat", "hf_model_id": "Qwen/Qwen1.5-7B-Chat"},
+    "Meta-Llama-3-8B-Instruct": {"pretty_name": "Llama-3-8B-Instruct", "hf_model_id": "meta-llama/Meta-Llama-3-8B-Instruct"},
+    "Meta-Llama-3-70B-Instruct": {"pretty_name": "Llama-3-70B-Instruct", "hf_model_id": "meta-llama/Meta-Llama-3-70B-Instruct"},
+    "Llama-2-13b-chat-hf": {"pretty_name": "Llama-2-13B-chat", "hf_model_id": "meta-llama/Llama-2-13b-chat-hf"},
+    "Llama-2-70b-chat-hf": {"pretty_name": "Llama-2-70B-chat", "hf_model_id": "meta-llama/Llama-2-70b-chat-hf"},
+    "Llama-2-7b-chat-hf": {"pretty_name": "Llama-2-7B-chat", "hf_model_id": "meta-llama/Llama-2-7b-chat-hf"},
+    "Mistral-7B-Instruct-v0.1": {"pretty_name": "Mistral-7B-Instruct", "hf_model_id": "mistralai/Mistral-7B-Instruct-v0.1"},
+    "Mistral-7B-Instruct-v0.2": {"pretty_name": "Mistral-7B-Instruct-v0.2", "hf_model_id": "mistralai/Mistral-7B-Instruct-v0.2"},
+    "Mixtral-8x7B-Instruct-v0.1": {"pretty_name": "Mixtral-8x7B-Instruct", "hf_model_id": "mistralai/Mixtral-8x7B-Instruct-v0.1"},
+    "command-r": {"pretty_name": "Command-R", "hf_model_id": "https://cohere.com/command"},
+    "command-r-plus": {"pretty_name": "Command-R-Plus", "hf_model_id": "https://cohere.com/command"},
+    "Phi-3-medium-128k-instruct": {"pretty_name": "Phi-3-medium-128k", "hf_model_id": "microsoft/Phi-3-medium-128k-instruct"},
+    "Phi-3-mini-128k-instruct": {"pretty_name": "Phi-3-mini-128k", "hf_model_id": "microsoft/Phi-3-mini-128k-instruct"},
+    "Nous-Hermes-2-Mixtral-8x7B-DPO": {"pretty_name": "Hermes-2-Mixtral-8x7B-DPO", "hf_model_id": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"},
+    "Hermes-2-Theta-Llama-3-8B": {"pretty_name": "Hermes-2-Theta-Llama-3-8B", "hf_model_id": "NousResearch/Hermes-2-Theta-Llama-3-8B"},    
+    "yi-large": {"pretty_name": "Yi-Large", "hf_model_id": "https://platform.01.ai/"},
+    "yi-large-preview": {"pretty_name": "Yi-Large-Preview", "hf_model_id": "https://platform.01.ai/"},
+    "Yi-1.5-34B-Chat": {"pretty_name": "Yi-1.5-34B-Chat", "hf_model_id": "01-ai/Yi-1.5-34B-Chat"},
+    "Yi-1.5-9B-Chat": {"pretty_name": "Yi-1.5-9B-Chat", "hf_model_id": "01-ai/Yi-1.5-9B-Chat"},
+    "Yi-1.5-6B-Chat": {"pretty_name": "Yi-1.5-6B-Chat", "hf_model_id": "01-ai/Yi-1.5-6B-Chat"}, 
+    "reka-flash-20240226": {"pretty_name": "Reka Flash", "hf_model_id": "https://www.reka.ai/"},
+    "reka-core-20240501": {"pretty_name": "Reka Core", "hf_model_id": "https://www.reka.ai/"},
+    "reka-edge": {"pretty_name": "Reka Edge", "hf_model_id": "https://www.reka.ai/"},
+    "gemini-1.5-pro": {"pretty_name": "Gemini 1.5 Pro", "hf_model_id": "https://blog.google/technology/ai/google-gemini-ai/"},
+    "gemini-1.5-flash": {"pretty_name": "Gemini 1.5 Flash", "hf_model_id": "https://blog.google/technology/ai/google-gemini-ai/"},
+    "gemma-7b-it": {"pretty_name": "Gemma-7B-it", "hf_model_id": "google/gemma-7b-it"},
+    "gemma-2b-it": {"pretty_name": "Gemma-2B-it", "hf_model_id": "google/gemma-2b-it"},
+    "gpt-3.5-turbo-0125": {"pretty_name": "gpt-3.5-turbo-0125", "hf_model_id": "https://platform.openai.com/"},
+    "gpt-4-0125-preview": {"pretty_name": "gpt-4-0125-preview", "hf_model_id": "https://platform.openai.com/"},
+    "gpt-4o-2024-05-13": {"pretty_name": "gpt-4o-2024-05-13", "hf_model_id": "https://platform.openai.com/"},
+    "gpt-4-turbo-2024-04-09": {"pretty_name": "gpt-4-turbo-2024-04-09", "hf_model_id": "https://platform.openai.com/"},
+    "gpt-4-0314": {"pretty_name": "gpt-4-0314", "hf_model_id": "https://platform.openai.com/"},
+    "tulu-2-dpo-70b": {"pretty_name": "Tulu-2-dpo-70b", "hf_model_id": "allenai/tulu-2-dpo-70b"},
+    "zephyr-7b-beta": {"pretty_name": "Zephyr-7b-beta", "hf_model_id": "HuggingFaceH4/zephyr-7b-beta"},
+    "mistral-large-2402": {"pretty_name": "Mistral-Large", "hf_model_id": "https://mistral.ai/news/mistral-large/"},
+    "claude-3-haiku-20240307": {"pretty_name": "Claude 3 Haiku", "hf_model_id": "https://www.anthropic.com/claude"},
+    "claude-3-opus-20240229": {"pretty_name": "Claude 3 Opus", "hf_model_id": "https://www.anthropic.com/claude"},
+    "claude-3-sonnet-20240229": {"pretty_name": "Claude 3 Sonnet", "hf_model_id": "https://www.anthropic.com/claude"},
+    "claude-3-5-sonnet-20240620": {"pretty_name": "Claude 3.5 Sonnet", "hf_model_id": "https://www.anthropic.com/claude"},
+    "Starling-LM-7B-beta": {"pretty_name": "StarlingLM-7B-beta", "hf_model_id": "Nexusflow/Starling-LM-7B-beta"},
+    "dbrx-instruct": {"pretty_name": "DBRX Instruct", "hf_model_id": "databricks/dbrx-instruct"},
+    "Llama-3-Instruct-8B-SimPO": {"pretty_name": "Llama3-Inst-8B-SimPO", "hf_model_id": "princeton-nlp/Llama-3-Instruct-8B-SimPO"},
+    "Llama-3-Instruct-8B-SimPO-ExPO": {"pretty_name": "Llama3-Inst-8B-SimPO-ExPO", "hf_model_id": "chujiezheng/Llama-3-Instruct-8B-SimPO-ExPO"},
+    "Starling-LM-7B-beta-ExPO": {"pretty_name": "Starling-LM-7B-beta-ExPO", "hf_model_id": "chujiezheng/Starling-LM-7B-beta-ExPO"},
+    "SELM-Zephyr-7B-iter-3": {"pretty_name": "SELM (Zephyr-7B-iter3)", "hf_model_id": "ZhangShenao/SELM-Zephyr-7B-iter-3"},
+    "deepseekv2-chat": {"pretty_name": "DeepSeek-V2-Chat", "hf_model_id": "deepseek-ai/DeepSeek-V2-Chat"},
+    "deepseek-coder-v2": {"pretty_name": "DeepSeek-Coder-V2-Inst", "hf_model_id": "deepseek-ai/DeepSeek-Coder-V2-Instruct"},
+    "deepseek-chat": {"pretty_name": "DeepSeek-V2-Chat", "hf_model_id": "https://platform.deepseek.com/api-docs/api/deepseek-api/", "open": true},
+    "deepseek-coder": {"pretty_name": "DeepSeek-Coder-V2", "hf_model_id": "https://platform.deepseek.com/api-docs/api/deepseek-api/", "open": true},
+    "gemma-2-27b-it@nvidia": {"pretty_name": "Gemma-2-27B-it", "hf_model_id": "https://huggingface.co/google/gemma-2-27b-it"},
+    "gemma-2-9b-it@nvidia": {"pretty_name": "Gemma-2-9B-it", "hf_model_id": "https://huggingface.co/google/gemma-2-9b-it"},
+    "neo_7b_instruct_v0.1": {"pretty_name": "Neo-7B-Instruct", "hf_model_id": "m-a-p/neo_7b_instruct_v0.1"},
+    "Yi-34B-Chat": {"pretty_name": "Yi-34B-Chat", "hf_model_id": "01-ai/Yi-34B-chat"},
+    "vicuna-13b-v1.5": {"pretty_name": "Vicuna-13b-v1.5", "hf_model_id": "lmsys/vicuna-13b-v1.5"},
+    "zephyr-7b-gemma-v0.1": {"pretty_name": "Zephyr-7b-Gemma", "hf_model_id": "HuggingFaceH4/zephyr-7b-gemma-v0.1"},
+    "glm-4-9b-chat": {"pretty_name": "GLM-4-9B-Chat", "hf_model_id": "THUDM/glm-4-9b-chat"},
+    "neo_7b_instruct_v0.1-ExPO": {"pretty_name": "Neo-7B-Instruct-ExPO", "hf_model_id": "chujiezheng/neo_7b_instruct_v0.1-ExPO"},
+    "SELM-Llama-3-8B-Instruct-iter-3": {"pretty_name": "SELM (Llama3-8B-Inst-iter3)", "hf_model_id": "ZhangShenao/SELM-Llama-3-8B-Instruct-iter-3"},
+    "nemotron-4-340b-instruct": {"pretty_name": "Nemotron-4-340B-Instruct", "hf_model_id": "nvidia/Nemotron-4-340B-Instruct"},
+    "Llama-3-8B-Magpie-Align-v0.1": {"pretty_name": "Magpie-8B-Align-v0.1", "hf_model_id": "Magpie-Align/Llama-3-8B-Magpie-Align-v0.1"}
+}

requirements.txt

@@ -0,0 +1,4 @@
+gradio[oauth]==4.19.2
+datasets
+toolz==0.12.1
+plotly

style.css

@@ -0,0 +1,27 @@
+body { 
+	font-family: -apple-system, BlinkMacSystemFont, "Arial", sans-serif;
+}
+
+h1 {
+	font-size: 16px;
+	margin-top: 0;
+}
+
+p {
+	color: rgb(107, 114, 128);
+	font-size: 15px;
+	margin-bottom: 10px;
+	margin-top: 5px;
+}
+
+.card {
+	max-width: 620px;
+	margin: 0 auto;
+	padding: 16px;
+	border: 1px solid lightgray;
+	border-radius: 16px;
+}
+
+.card p:last-child {
+	margin-bottom: 0;
+}

themes.py

@@ -0,0 +1,45 @@
+from __future__ import annotations
+from typing import Iterable
+import gradio as gr
+from gradio.themes.base import Base
+from gradio.themes.utils import colors, fonts, sizes
+import time
+
+class Seafoam(Base):
+    def __init__(
+        self,
+        *,
+        primary_hue: colors.Color | str = colors.blue,
+        secondary_hue: colors.Color | str = colors.gray,
+        neutral_hue: colors.Color | str = colors.gray,
+        spacing_size: sizes.Size | str = sizes.spacing_md,
+        radius_size: sizes.Size | str = sizes.radius_md,
+        text_size: sizes.Size | str = sizes.text_lg,
+        font: fonts.Font
+        | str
+        | Iterable[fonts.Font | str] = (
+            fonts.GoogleFont("Quicksand"),
+            "ui-sans-serif",
+            "sans-serif",
+        ),
+        font_mono: fonts.Font
+        | str
+        | Iterable[fonts.Font | str] = (
+            fonts.GoogleFont("IBM Plex Mono"),
+            "ui-monospace",
+            "monospace",
+        ),
+    ):
+        super().__init__(
+            primary_hue=primary_hue,
+            secondary_hue=secondary_hue,
+            neutral_hue=neutral_hue,
+            spacing_size=spacing_size,
+            radius_size=radius_size,
+            text_size=text_size,
+            font=font,
+            font_mono=font_mono,
+        )
+
+
+seafoam = Seafoam()

update_data.sh

@@ -0,0 +1,40 @@
+TARGET_DIR="ZeroEval-main"
+
+rm -r $TARGET_DIR
+# Download the ZIP file
+curl -L -o zeroeval.zip https://github.com/yuchenlin/ZeroEval/archive/refs/heads/main.zip
+unzip zeroeval.zip
+rm zeroeval.zip
+
+#!/bin/bash
+
+# Define the target directory and the exception folder 
+EXCEPTION_FOLDER="result_dirs"
+
+# Ensure the target directory exists
+if [ -d "$TARGET_DIR" ]; then
+  # Loop through each item in the target directory
+  for item in "$TARGET_DIR"/*; do
+    # Check if it is not the exception folder
+    if [ "$(basename "$item")" != "$EXCEPTION_FOLDER" ]; then
+      # Remove the item (file or directory)
+      rm -rf "$item"
+      echo "Removed: $item"
+    fi
+  done
+else
+  echo "Target directory does not exist: $TARGET_DIR"
+fi
+
+# only keep the result_dirs/zebra-grid under result_dirs folder; remove all other sub-folders under result_dirs
+# Remove all subdirectories in result_dirs except zebra-grid
+find "$TARGET_DIR/result_dirs" -maxdepth 1 -type d ! -name 'zebra-grid' ! -name 'result_dirs' -exec rm -rf {} +
+
+rm -rf $TARGET_DIR/.github
+rm -rf $TARGET_DIR/.gitignore
+
+
+# tables
+
+
+# bash update_table.sh

utils_display.py

@@ -0,0 +1,34 @@
+import json 
+
+with open("model_info.json", "r") as f:
+    model_info = json.load(f)
+
+def make_clickable_model(model_name):
+    global model_info
+    modified_model_name = model_name
+    if model_name in model_info:
+        if model_info[model_name]["hf_model_id"].startswith("http"):
+            link = model_info[model_name]["hf_model_id"]
+            modified_model_name = f'🔒 <a target="_blank" href="{link}" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">{model_info[model_name]["pretty_name"]}</a>'
+        else:
+            link = f"https://huggingface.co/{model_info[model_name]['hf_model_id']}"
+            modified_model_name = f'🔑 <a target="_blank" href="{link}" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">{model_info[model_name]["pretty_name"]}</a>'
+            if "Neo-7B" in modified_model_name:
+                # models that are fully open source
+                modified_model_name = modified_model_name.replace("🔑", "💎🔑") 
+    
+    if "🚨</a>" in modified_model_name:
+        modified_model_name = modified_model_name.replace(' 🚨</a>', '</a> 🚨')
+    # if model_name in ["gpt-4-turbo-2024-04-09", "Llama-2-70b-chat-hf", "claude-3-haiku-20240307"]:
+    #     modified_model_name = modified_model_name.replace('style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;"', 'style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted; font-weight: bold; background-color: var(--link-background-color);"')
+    return modified_model_name
+    
+
+def styled_error(error):
+    return f"<p style='color: red; font-size: 20px; text-align: center;'>{error}</p>"
+
+def styled_warning(warn):
+    return f"<p style='color: orange; font-size: 20px; text-align: center;'>{warn}</p>"
+
+def styled_message(message):
+    return f"<p style='color: green; font-size: 20px; text-align: center;'>{message}</p>"