weichiang's picture
update 20240409
bf1a5a5
raw
history blame
17.9 kB
"""A gradio app that renders a static leaderboard. This is used for Hugging Face Space."""
import ast
import argparse
import glob
import pickle
import gradio as gr
import numpy as np
import pandas as pd
# notebook_url = "https://colab.research.google.com/drive/1RAWb22-PFNI-X1gPVzc927SGUdfr6nsR?usp=sharing"
notebook_url = "https://colab.research.google.com/drive/1KdwokPjirkTmpO_P1WByFNFiqxWQquwH#scrollTo=o_CpbkGEbhrK"
basic_component_values = [None] * 6
leader_component_values = [None] * 5
def make_default_md(arena_df, elo_results):
total_votes = sum(arena_df["num_battles"]) // 2
total_models = len(arena_df)
leaderboard_md = f"""
# πŸ† LMSYS Chatbot Arena Leaderboard
| [Vote](https://chat.lmsys.org) | [Blog](https://lmsys.org/blog/2023-05-03-arena/) | [GitHub](https://github.com/lm-sys/FastChat) | [Paper](https://arxiv.org/abs/2306.05685) | [Dataset](https://github.com/lm-sys/FastChat/blob/main/docs/dataset_release.md) | [Twitter](https://twitter.com/lmsysorg) | [Discord](https://discord.gg/HSWAKCrnFx) |
LMSYS [Chatbot Arena](https://lmsys.org/blog/2023-05-03-arena/) is a crowdsourced open platform for LLM evals.
We've collected over **500,000** human preference votes to rank LLMs with the Elo ranking system.
"""
return leaderboard_md
def make_arena_leaderboard_md(arena_df):
total_votes = sum(arena_df["num_battles"]) // 2
total_models = len(arena_df)
leaderboard_md = f"""
Total #models: **{total_models}**. Total #votes: **{total_votes}**. Last updated: April 9, 2024.
Contribute your vote πŸ—³οΈ at [chat.lmsys.org](https://chat.lmsys.org)! Find more analysis in the [notebook]({notebook_url}).
"""
return leaderboard_md
def make_full_leaderboard_md(elo_results):
leaderboard_md = f"""
Three benchmarks are displayed: **Arena Elo**, **MT-Bench** and **MMLU**.
- [Chatbot Arena](https://chat.lmsys.org/?arena) - a crowdsourced, randomized battle platform. We use 500K+ user votes to compute Elo ratings.
- [MT-Bench](https://arxiv.org/abs/2306.05685): a set of challenging multi-turn questions. We use GPT-4 to grade the model responses.
- [MMLU](https://arxiv.org/abs/2009.03300) (5-shot): a test to measure a model's multitask accuracy on 57 tasks.
πŸ’» Code: The MT-bench scores (single-answer grading on a scale of 10) are computed by [fastchat.llm_judge](https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge).
The MMLU scores are mostly computed by [InstructEval](https://github.com/declare-lab/instruct-eval).
Higher values are better for all benchmarks. Empty cells mean not available.
"""
return leaderboard_md
def make_leaderboard_md_live(elo_results):
leaderboard_md = f"""
# Leaderboard
Last updated: {elo_results["last_updated_datetime"]}
{elo_results["leaderboard_table"]}
"""
return leaderboard_md
def update_elo_components(max_num_files, elo_results_file):
log_files = get_log_files(max_num_files)
# Leaderboard
if elo_results_file is None: # Do live update
battles = clean_battle_data(log_files)
elo_results = report_elo_analysis_results(battles)
leader_component_values[0] = make_leaderboard_md_live(elo_results)
leader_component_values[1] = elo_results["win_fraction_heatmap"]
leader_component_values[2] = elo_results["battle_count_heatmap"]
leader_component_values[3] = elo_results["bootstrap_elo_rating"]
leader_component_values[4] = elo_results["average_win_rate_bar"]
# Basic stats
basic_stats = report_basic_stats(log_files)
md0 = f"Last updated: {basic_stats['last_updated_datetime']}"
md1 = "### Action Histogram\n"
md1 += basic_stats["action_hist_md"] + "\n"
md2 = "### Anony. Vote Histogram\n"
md2 += basic_stats["anony_vote_hist_md"] + "\n"
md3 = "### Model Call Histogram\n"
md3 += basic_stats["model_hist_md"] + "\n"
md4 = "### Model Call (Last 24 Hours)\n"
md4 += basic_stats["num_chats_last_24_hours"] + "\n"
basic_component_values[0] = md0
basic_component_values[1] = basic_stats["chat_dates_bar"]
basic_component_values[2] = md1
basic_component_values[3] = md2
basic_component_values[4] = md3
basic_component_values[5] = md4
def update_worker(max_num_files, interval, elo_results_file):
while True:
tic = time.time()
update_elo_components(max_num_files, elo_results_file)
durtaion = time.time() - tic
print(f"update duration: {durtaion:.2f} s")
time.sleep(max(interval - durtaion, 0))
def load_demo(url_params, request: gr.Request):
logger.info(f"load_demo. ip: {request.client.host}. params: {url_params}")
return basic_component_values + leader_component_values
def model_hyperlink(model_name, link):
return f'<a target="_blank" href="{link}" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">{model_name}</a>'
def load_leaderboard_table_csv(filename, add_hyperlink=True):
lines = open(filename).readlines()
heads = [v.strip() for v in lines[0].split(",")]
rows = []
for i in range(1, len(lines)):
row = [v.strip() for v in lines[i].split(",")]
for j in range(len(heads)):
item = {}
for h, v in zip(heads, row):
if h == "Arena Elo rating":
if v != "-":
v = int(ast.literal_eval(v))
else:
v = np.nan
elif h == "MMLU":
if v != "-":
v = round(ast.literal_eval(v) * 100, 1)
else:
v = np.nan
elif h == "MT-bench (win rate %)":
if v != "-":
v = round(ast.literal_eval(v[:-1]), 1)
else:
v = np.nan
elif h == "MT-bench (score)":
if v != "-":
v = round(ast.literal_eval(v), 2)
else:
v = np.nan
item[h] = v
if add_hyperlink:
item["Model"] = model_hyperlink(item["Model"], item["Link"])
rows.append(item)
return rows
def build_basic_stats_tab():
empty = "Loading ..."
basic_component_values[:] = [empty, None, empty, empty, empty, empty]
md0 = gr.Markdown(empty)
gr.Markdown("#### Figure 1: Number of model calls and votes")
plot_1 = gr.Plot(show_label=False)
with gr.Row():
with gr.Column():
md1 = gr.Markdown(empty)
with gr.Column():
md2 = gr.Markdown(empty)
with gr.Row():
with gr.Column():
md3 = gr.Markdown(empty)
with gr.Column():
md4 = gr.Markdown(empty)
return [md0, plot_1, md1, md2, md3, md4]
def get_full_table(arena_df, model_table_df):
values = []
for i in range(len(model_table_df)):
row = []
model_key = model_table_df.iloc[i]["key"]
model_name = model_table_df.iloc[i]["Model"]
# model display name
row.append(model_name)
if model_key in arena_df.index:
idx = arena_df.index.get_loc(model_key)
row.append(round(arena_df.iloc[idx]["rating"]))
else:
row.append(np.nan)
row.append(model_table_df.iloc[i]["MT-bench (score)"])
row.append(model_table_df.iloc[i]["MMLU"])
# Organization
row.append(model_table_df.iloc[i]["Organization"])
# license
row.append(model_table_df.iloc[i]["License"])
values.append(row)
values.sort(key=lambda x: -x[1] if not np.isnan(x[1]) else 1e9)
return values
def get_arena_table(arena_df, model_table_df):
# sort by rating
arena_df = arena_df.sort_values(by=["final_ranking", "rating"], ascending=[True, False])
values = []
for i in range(len(arena_df)):
row = []
model_key = arena_df.index[i]
model_name = model_table_df[model_table_df["key"] == model_key]["Model"].values[
0
]
# rank
ranking = arena_df.iloc[i].get("final_ranking") or i+1
row.append(ranking)
# model display name
row.append(model_name)
# elo rating
row.append(round(arena_df.iloc[i]["rating"]))
upper_diff = round(
arena_df.iloc[i]["rating_q975"] - arena_df.iloc[i]["rating"]
)
lower_diff = round(
arena_df.iloc[i]["rating"] - arena_df.iloc[i]["rating_q025"]
)
row.append(f"+{upper_diff}/-{lower_diff}")
# num battles
row.append(round(arena_df.iloc[i]["num_battles"]))
# Organization
row.append(
model_table_df[model_table_df["key"] == model_key]["Organization"].values[0]
)
# license
row.append(
model_table_df[model_table_df["key"] == model_key]["License"].values[0]
)
cutoff_date = model_table_df[model_table_df["key"] == model_key]["Knowledge cutoff date"].values[0]
if cutoff_date == "-":
row.append("Unknown")
else:
row.append(cutoff_date)
values.append(row)
return values
def build_leaderboard_tab(elo_results_file, leaderboard_table_file, show_plot=False):
if elo_results_file is None: # Do live update
default_md = "Loading ..."
p1 = p2 = p3 = p4 = None
else:
with open(elo_results_file, "rb") as fin:
elo_results = pickle.load(fin)
if "full" in elo_results:
elo_results = elo_results["full"]
p1 = elo_results["win_fraction_heatmap"]
p2 = elo_results["battle_count_heatmap"]
p3 = elo_results["bootstrap_elo_rating"]
p4 = elo_results["average_win_rate_bar"]
arena_df = elo_results["leaderboard_table_df"]
default_md = make_default_md(arena_df, elo_results)
md_1 = gr.Markdown(default_md, elem_id="leaderboard_markdown")
if leaderboard_table_file:
data = load_leaderboard_table_csv(leaderboard_table_file)
model_table_df = pd.DataFrame(data)
with gr.Tabs() as tabs:
# arena table
arena_table_vals = get_arena_table(arena_df, model_table_df)
with gr.Tab("Arena Elo", id=0):
md = make_arena_leaderboard_md(arena_df)
gr.Markdown(md, elem_id="leaderboard_markdown")
gr.Dataframe(
headers=[
"Rank",
"πŸ€– Model",
"⭐ Arena Elo",
"πŸ“Š 95% CI",
"πŸ—³οΈ Votes",
"Organization",
"License",
"Knowledge Cutoff",
],
datatype=[
"str",
"markdown",
"number",
"str",
"number",
"str",
"str",
"str",
],
value=arena_table_vals,
elem_id="arena_leaderboard_dataframe",
height=700,
column_widths=[50, 200, 120, 100, 100, 150, 150, 100],
wrap=True,
)
with gr.Tab("Full Leaderboard", id=1):
md = make_full_leaderboard_md(elo_results)
gr.Markdown(md, elem_id="leaderboard_markdown")
full_table_vals = get_full_table(arena_df, model_table_df)
gr.Dataframe(
headers=[
"πŸ€– Model",
"⭐ Arena Elo",
"πŸ“ˆ MT-bench",
"πŸ“š MMLU",
"Organization",
"License",
],
datatype=["markdown", "number", "number", "number", "str", "str"],
value=full_table_vals,
elem_id="full_leaderboard_dataframe",
column_widths=[200, 100, 100, 100, 150, 150],
height=700,
wrap=True,
)
if not show_plot:
gr.Markdown(
""" ## Visit our [HF space](https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard) for more analysis!
If you want to see more models, please help us [add them](https://github.com/lm-sys/FastChat/blob/main/docs/arena.md#how-to-add-a-new-model).
""",
elem_id="leaderboard_markdown",
)
else:
pass
gr.Markdown(
f"""Note: we take the 95% confidence interval into account when determining a model's ranking.
A model is ranked higher only if its lower bound of model score is higher than the upper bound of the other model's score.
See Figure 3 below for visualization of the confidence intervals.
""",
elem_id="leaderboard_markdown"
)
leader_component_values[:] = [default_md, p1, p2, p3, p4]
if show_plot:
gr.Markdown(
f"""## More Statistics for Chatbot Arena\n
Below are figures for more statistics. The code for generating them is also included in this [notebook]({notebook_url}).
You can find more discussions in this blog [post](https://lmsys.org/blog/2023-12-07-leaderboard/).
""",
elem_id="leaderboard_markdown"
)
with gr.Row():
with gr.Column():
gr.Markdown(
"#### Figure 1: Fraction of Model A Wins for All Non-tied A vs. B Battles"
)
plot_1 = gr.Plot(p1, show_label=False)
with gr.Column():
gr.Markdown(
"#### Figure 2: Battle Count for Each Combination of Models (without Ties)"
)
plot_2 = gr.Plot(p2, show_label=False)
with gr.Row():
with gr.Column():
gr.Markdown(
"#### Figure 3: Confidence Intervals on Model Strength (via Bootstrapping)"
)
plot_3 = gr.Plot(p3, show_label=False)
with gr.Column():
gr.Markdown(
"#### Figure 4: Average Win Rate Against All Other Models (Assuming Uniform Sampling and No Ties)"
)
plot_4 = gr.Plot(p4, show_label=False)
with gr.Accordion(
"πŸ“ Citation",
open=True,
):
citation_md = """
### Citation
Please cite the following paper if you find our leaderboard or dataset helpful.
```
@misc{chiang2024chatbot,
title={Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference},
author={Wei-Lin Chiang and Lianmin Zheng and Ying Sheng and Anastasios Nikolas Angelopoulos and Tianle Li and Dacheng Li and Hao Zhang and Banghua Zhu and Michael Jordan and Joseph E. Gonzalez and Ion Stoica},
year={2024},
eprint={2403.04132},
archivePrefix={arXiv},
primaryClass={cs.AI}
}
"""
gr.Markdown(citation_md, elem_id="leaderboard_markdown")
gr.Markdown(acknowledgment_md)
if show_plot:
return [md_1, plot_1, plot_2, plot_3, plot_4]
return [md_1]
block_css = """
#notice_markdown {
font-size: 104%
}
#notice_markdown th {
display: none;
}
#notice_markdown td {
padding-top: 6px;
padding-bottom: 6px;
}
#leaderboard_markdown {
font-size: 104%
}
#leaderboard_markdown td {
padding-top: 6px;
padding-bottom: 6px;
}
#leaderboard_dataframe td {
line-height: 0.1em;
}
footer {
display:none !important
}
.sponsor-image-about img {
margin: 0 20px;
margin-top: 20px;
height: 40px;
max-height: 100%;
width: auto;
float: left;
}
"""
acknowledgment_md = """
### Acknowledgment
We thank [Kaggle](https://www.kaggle.com/), [MBZUAI](https://mbzuai.ac.ae/), [a16z](https://www.a16z.com/), [Together AI](https://www.together.ai/), [Anyscale](https://www.anyscale.com/), [HuggingFace](https://huggingface.co/) for their generous [sponsorship](https://lmsys.org/donations/).
<div class="sponsor-image-about">
<img src="https://storage.googleapis.com/public-arena-asset/kaggle.png" alt="Kaggle">
<img src="https://storage.googleapis.com/public-arena-asset/mbzuai.jpeg" alt="MBZUAI">
<img src="https://storage.googleapis.com/public-arena-asset/a16z.jpeg" alt="a16z">
<img src="https://storage.googleapis.com/public-arena-asset/together.png" alt="Together AI">
<img src="https://storage.googleapis.com/public-arena-asset/anyscale.png" alt="AnyScale">
<img src="https://storage.googleapis.com/public-arena-asset/huggingface.png" alt="HuggingFace">
</div>
"""
def build_demo(elo_results_file, leaderboard_table_file):
text_size = gr.themes.sizes.text_lg
with gr.Blocks(
title="Chatbot Arena Leaderboard",
theme=gr.themes.Base(text_size=text_size),
css=block_css,
) as demo:
leader_components = build_leaderboard_tab(
elo_results_file, leaderboard_table_file, show_plot=True
)
return demo
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--share", action="store_true")
args = parser.parse_args()
elo_result_files = glob.glob("elo_results_*.pkl")
elo_result_files.sort(key=lambda x: int(x[12:-4]))
elo_result_file = elo_result_files[-1]
leaderboard_table_files = glob.glob("leaderboard_table_*.csv")
leaderboard_table_files.sort(key=lambda x: int(x[18:-4]))
leaderboard_table_file = leaderboard_table_files[-1]
demo = build_demo(elo_result_file, leaderboard_table_file)
demo.launch(share=args.share)