Spaces:
Running
Running
intro added with paper/artifacts
Browse files
app.py
CHANGED
@@ -4,10 +4,24 @@ import numpy as np
|
|
4 |
from gradio_leaderboard import Leaderboard, SelectColumns, ColumnFilter
|
5 |
|
6 |
# Define constants and enums
|
7 |
-
TITLE = "<h1>M-RewardBench Leaderboard</h1>"
|
8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
GOOGLE_SHEET_URL = "https://docs.google.com/spreadsheets/d/1qrD7plUdrBwAw7G6UeDVZAaV9ihxaNAcoiKwSaqotR4/export?gid=0&format=csv"
|
10 |
-
ABOUT_TEXT = """
|
|
|
|
|
|
|
|
|
|
|
11 |
|
12 |
|
13 |
class AutoEvalColumn:
|
@@ -111,6 +125,7 @@ def format_with_color(val, min_val=50, max_val=100):
|
|
111 |
return str(val)
|
112 |
|
113 |
demo = gr.Blocks(theme=gr.themes.Soft())
|
|
|
114 |
with demo:
|
115 |
gr.HTML(TITLE)
|
116 |
gr.Markdown(INTRODUCTION_TEXT)
|
@@ -153,6 +168,23 @@ with demo:
|
|
153 |
AutoEvalColumn.add_columns_from_df(df, numeric_cols)
|
154 |
|
155 |
leaderboard = init_leaderboard(df)
|
156 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
157 |
|
158 |
demo.launch(ssr_mode=False)
|
|
|
4 |
from gradio_leaderboard import Leaderboard, SelectColumns, ColumnFilter
|
5 |
|
6 |
# Define constants and enums
|
7 |
+
# TITLE = "<h1>M-RewardBench Leaderboard</h1>"
|
8 |
+
TITLE = '''<h1>
|
9 |
+
<span style="font-variant: small-caps;">M-RewardBench</span>: Evaluating Reward Models in Multilingual Settings
|
10 |
+
</h1>'''
|
11 |
+
INTRODUCTION_TEXT = '''
|
12 |
+
Evaluating the chat, safety, reasoning, and translation capabilities of Multilingual Reward Models.
|
13 |
+
|
14 |
+
π [Paper](https://arxiv.org/pdf/2410.15522.pdf) | π» [Code](https://github.com/for-ai/m-rewardbench) | π€ [Dataset](https://hf.co/datasets/C4AI-Community/multilingual-reward-bench) | π [arXiv](https://arxiv.org/abs/2410.15522) | π [Leaderboard](https://c4ai-community-m-rewardbench.hf.space/)
|
15 |
+
|
16 |
+
π https://m-rewardbench.github.io/'''
|
17 |
+
|
18 |
GOOGLE_SHEET_URL = "https://docs.google.com/spreadsheets/d/1qrD7plUdrBwAw7G6UeDVZAaV9ihxaNAcoiKwSaqotR4/export?gid=0&format=csv"
|
19 |
+
# ABOUT_TEXT = """
|
20 |
+
# <h1>
|
21 |
+
# <span style="font-variant: small-caps;">M-RewardBench</span>: Evaluating Reward Models in Multilingual Settings
|
22 |
+
# </h1>
|
23 |
+
|
24 |
+
# Welcome to M-RewardBench Leaderboard!"""
|
25 |
|
26 |
|
27 |
class AutoEvalColumn:
|
|
|
125 |
return str(val)
|
126 |
|
127 |
demo = gr.Blocks(theme=gr.themes.Soft())
|
128 |
+
|
129 |
with demo:
|
130 |
gr.HTML(TITLE)
|
131 |
gr.Markdown(INTRODUCTION_TEXT)
|
|
|
168 |
AutoEvalColumn.add_columns_from_df(df, numeric_cols)
|
169 |
|
170 |
leaderboard = init_leaderboard(df)
|
171 |
+
|
172 |
+
with gr.Row():
|
173 |
+
with gr.Accordion("π Citation", open=False):
|
174 |
+
citation_button = gr.Textbox(
|
175 |
+
value=r"""@misc{gureja2024mrewardbench,
|
176 |
+
title={M-RewardBench: Evaluating Reward Models in Multilingual Settings},
|
177 |
+
author={Srishti Gureja and Lester James V. Miranda and Shayekh Bin Islam and Rishabh Maheshwary and Drishti Sharma and Gusti Winata and Nathan Lambert and Sebastian Ruder and Sara Hooker and Marzieh Fadaee},
|
178 |
+
year={2024},
|
179 |
+
eprint={2410.15522},
|
180 |
+
archivePrefix={arXiv},
|
181 |
+
primaryClass={cs.CL},
|
182 |
+
url={https://arxiv.org/abs/2410.15522},
|
183 |
+
}""",
|
184 |
+
lines=7,
|
185 |
+
label="BibTeX",
|
186 |
+
elem_id="citation-button",
|
187 |
+
show_copy_button=True,
|
188 |
+
)
|
189 |
|
190 |
demo.launch(ssr_mode=False)
|