Spaces:
Running
Running
Kang Suhyun
commited on
Commit
•
5cbd838
1
Parent(s):
762adde
[#119] Update README and app with improved project description (#120)
Browse files* [#119] Update README and app with improved project description
This change updates the project description in both the README and the app to provide clearer information about Yanolja Arena.
* review
README.md
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
---
|
2 |
-
title: Arena
|
3 |
emoji: ⚔️
|
4 |
colorFrom: red
|
5 |
colorTo: purple
|
@@ -9,7 +9,9 @@ app_file: app.py
|
|
9 |
pinned: false
|
10 |
---
|
11 |
|
12 |
-
# [Arena](https://huggingface.co/spaces/yanolja/arena)
|
|
|
|
|
13 |
|
14 |
Get Involved: [Discuss and contribute on GitHub](https://github.com/yanolja/arena)
|
15 |
|
|
|
1 |
---
|
2 |
+
title: Yanolja Arena
|
3 |
emoji: ⚔️
|
4 |
colorFrom: red
|
5 |
colorTo: purple
|
|
|
9 |
pinned: false
|
10 |
---
|
11 |
|
12 |
+
# [Yanolja Arena](https://huggingface.co/spaces/yanolja/arena)
|
13 |
+
|
14 |
+
Yanolja Arena is an open-source project designed to identify the most effective LLMs for summarization and translation tasks. It employs a comparative approach, evaluating two randomly selected models simultaneously and utilizing the ELO rating system to rank their performance.
|
15 |
|
16 |
Get Involved: [Discuss and contribute on GitHub](https://github.com/yanolja/arena)
|
17 |
|
app.py
CHANGED
@@ -79,9 +79,28 @@ css = """
|
|
79 |
}
|
80 |
"""
|
81 |
|
82 |
-
with gr.Blocks(title="Arena", css=css) as app:
|
83 |
set_token(app)
|
84 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
85 |
with gr.Row():
|
86 |
category_radio = gr.Radio(
|
87 |
choices=[category.value for category in response.Category],
|
|
|
79 |
}
|
80 |
"""
|
81 |
|
82 |
+
with gr.Blocks(title="Yanolja Arena", css=css) as app:
|
83 |
set_token(app)
|
84 |
|
85 |
+
with gr.Row():
|
86 |
+
gr.HTML("""
|
87 |
+
<h1 style="text-align: center; font-size: 28px; margin-bottom: 16px">Yanolja Arena</h1>
|
88 |
+
<p style="text-align: center; font-size: 16px">Yanolja Arena helps find the best LLMs for summarizing and translating text. We compare two random models at a time and use an ELO rating system to score them.</p>
|
89 |
+
<p style="text-align: center; font-size: 16px">This is an open-source project. Check it out on <a href="https://github.com/yanolja/arena">GitHub</a>.</p>
|
90 |
+
""")
|
91 |
+
with gr.Accordion("How to Use", open=False):
|
92 |
+
gr.Markdown("""
|
93 |
+
1. **For Summaries:**
|
94 |
+
- Enter the text you want summarized into the prompt box.
|
95 |
+
|
96 |
+
2. **For Translations:**
|
97 |
+
- Choose the language you're translating from and to.
|
98 |
+
- Enter the text you want translated into the prompt box.
|
99 |
+
|
100 |
+
3. **Voting:**
|
101 |
+
- After you see both results, pick which one you think is better.
|
102 |
+
""")
|
103 |
+
|
104 |
with gr.Row():
|
105 |
category_radio = gr.Radio(
|
106 |
choices=[category.value for category in response.Category],
|