Spaces:
Running
Running
initial
Browse files- DESCRIPTION.md +1 -0
- README.md +5 -8
- app.py +1790 -0
- requirements.txt +4 -0
DESCRIPTION.md
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
SeaEval Leaderboard.
|
README.md
CHANGED
|
@@ -1,13 +1,10 @@
|
|
| 1 |
---
|
| 2 |
-
title: SeaEval Leaderboard
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: gradio
|
| 7 |
-
sdk_version: 4.
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
-
license: cc-by-nc-4.0
|
| 11 |
---
|
| 12 |
-
|
| 13 |
-
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
| 1 |
---
|
| 2 |
+
title: SeaEval Leaderboard
|
| 3 |
+
emoji: 🥇
|
| 4 |
+
colorFrom: blue
|
| 5 |
+
colorTo: indigo
|
| 6 |
sdk: gradio
|
| 7 |
+
sdk_version: 4.0.2
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
|
|
|
| 10 |
---
|
|
|
|
|
|
app.py
ADDED
|
@@ -0,0 +1,1790 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from functools import partial
|
| 2 |
+
import json
|
| 3 |
+
|
| 4 |
+
from datasets import load_dataset
|
| 5 |
+
import gradio as gr
|
| 6 |
+
from huggingface_hub import get_hf_file_metadata, HfApi, hf_hub_download, hf_hub_url
|
| 7 |
+
from huggingface_hub.repocard import metadata_load
|
| 8 |
+
import pandas as pd
|
| 9 |
+
|
| 10 |
+
TASKS = [
|
| 11 |
+
"BitextMining",
|
| 12 |
+
"Classification",
|
| 13 |
+
"Clustering",
|
| 14 |
+
"PairClassification",
|
| 15 |
+
"Reranking",
|
| 16 |
+
"Retrieval",
|
| 17 |
+
"STS",
|
| 18 |
+
"Summarization",
|
| 19 |
+
]
|
| 20 |
+
|
| 21 |
+
TASK_LIST_BITEXT_MINING = ['BUCC (de-en)', 'BUCC (fr-en)', 'BUCC (ru-en)', 'BUCC (zh-en)', 'Tatoeba (afr-eng)', 'Tatoeba (amh-eng)', 'Tatoeba (ang-eng)', 'Tatoeba (ara-eng)', 'Tatoeba (arq-eng)', 'Tatoeba (arz-eng)', 'Tatoeba (ast-eng)', 'Tatoeba (awa-eng)', 'Tatoeba (aze-eng)', 'Tatoeba (bel-eng)', 'Tatoeba (ben-eng)', 'Tatoeba (ber-eng)', 'Tatoeba (bos-eng)', 'Tatoeba (bre-eng)', 'Tatoeba (bul-eng)', 'Tatoeba (cat-eng)', 'Tatoeba (cbk-eng)', 'Tatoeba (ceb-eng)', 'Tatoeba (ces-eng)', 'Tatoeba (cha-eng)', 'Tatoeba (cmn-eng)', 'Tatoeba (cor-eng)', 'Tatoeba (csb-eng)', 'Tatoeba (cym-eng)', 'Tatoeba (dan-eng)', 'Tatoeba (deu-eng)', 'Tatoeba (dsb-eng)', 'Tatoeba (dtp-eng)', 'Tatoeba (ell-eng)', 'Tatoeba (epo-eng)', 'Tatoeba (est-eng)', 'Tatoeba (eus-eng)', 'Tatoeba (fao-eng)', 'Tatoeba (fin-eng)', 'Tatoeba (fra-eng)', 'Tatoeba (fry-eng)', 'Tatoeba (gla-eng)', 'Tatoeba (gle-eng)', 'Tatoeba (glg-eng)', 'Tatoeba (gsw-eng)', 'Tatoeba (heb-eng)', 'Tatoeba (hin-eng)', 'Tatoeba (hrv-eng)', 'Tatoeba (hsb-eng)', 'Tatoeba (hun-eng)', 'Tatoeba (hye-eng)', 'Tatoeba (ido-eng)', 'Tatoeba (ile-eng)', 'Tatoeba (ina-eng)', 'Tatoeba (ind-eng)', 'Tatoeba (isl-eng)', 'Tatoeba (ita-eng)', 'Tatoeba (jav-eng)', 'Tatoeba (jpn-eng)', 'Tatoeba (kab-eng)', 'Tatoeba (kat-eng)', 'Tatoeba (kaz-eng)', 'Tatoeba (khm-eng)', 'Tatoeba (kor-eng)', 'Tatoeba (kur-eng)', 'Tatoeba (kzj-eng)', 'Tatoeba (lat-eng)', 'Tatoeba (lfn-eng)', 'Tatoeba (lit-eng)', 'Tatoeba (lvs-eng)', 'Tatoeba (mal-eng)', 'Tatoeba (mar-eng)', 'Tatoeba (max-eng)', 'Tatoeba (mhr-eng)', 'Tatoeba (mkd-eng)', 'Tatoeba (mon-eng)', 'Tatoeba (nds-eng)', 'Tatoeba (nld-eng)', 'Tatoeba (nno-eng)', 'Tatoeba (nob-eng)', 'Tatoeba (nov-eng)', 'Tatoeba (oci-eng)', 'Tatoeba (orv-eng)', 'Tatoeba (pam-eng)', 'Tatoeba (pes-eng)', 'Tatoeba (pms-eng)', 'Tatoeba (pol-eng)', 'Tatoeba (por-eng)', 'Tatoeba (ron-eng)', 'Tatoeba (rus-eng)', 'Tatoeba (slk-eng)', 'Tatoeba (slv-eng)', 'Tatoeba (spa-eng)', 'Tatoeba (sqi-eng)', 'Tatoeba (srp-eng)', 'Tatoeba (swe-eng)', 'Tatoeba (swg-eng)', 'Tatoeba (swh-eng)', 'Tatoeba (tam-eng)', 'Tatoeba (tat-eng)', 'Tatoeba (tel-eng)', 'Tatoeba (tgl-eng)', 'Tatoeba (tha-eng)', 'Tatoeba (tuk-eng)', 'Tatoeba (tur-eng)', 'Tatoeba (tzl-eng)', 'Tatoeba (uig-eng)', 'Tatoeba (ukr-eng)', 'Tatoeba (urd-eng)', 'Tatoeba (uzb-eng)', 'Tatoeba (vie-eng)', 'Tatoeba (war-eng)', 'Tatoeba (wuu-eng)', 'Tatoeba (xho-eng)', 'Tatoeba (yid-eng)', 'Tatoeba (yue-eng)', 'Tatoeba (zsm-eng)']
|
| 22 |
+
TASK_LIST_BITEXT_MINING_OTHER = ["BornholmBitextMining"]
|
| 23 |
+
|
| 24 |
+
TASK_LIST_CLASSIFICATION = [
|
| 25 |
+
"AmazonCounterfactualClassification (en)",
|
| 26 |
+
"AmazonPolarityClassification",
|
| 27 |
+
"AmazonReviewsClassification (en)",
|
| 28 |
+
"Banking77Classification",
|
| 29 |
+
"EmotionClassification",
|
| 30 |
+
"ImdbClassification",
|
| 31 |
+
"MassiveIntentClassification (en)",
|
| 32 |
+
"MassiveScenarioClassification (en)",
|
| 33 |
+
"MTOPDomainClassification (en)",
|
| 34 |
+
"MTOPIntentClassification (en)",
|
| 35 |
+
"ToxicConversationsClassification",
|
| 36 |
+
"TweetSentimentExtractionClassification",
|
| 37 |
+
]
|
| 38 |
+
|
| 39 |
+
TASK_LIST_CLASSIFICATION_NORM = [x.replace(" (en)", "") for x in TASK_LIST_CLASSIFICATION]
|
| 40 |
+
|
| 41 |
+
TASK_LIST_CLASSIFICATION_DA = [
|
| 42 |
+
"AngryTweetsClassification",
|
| 43 |
+
"DanishPoliticalCommentsClassification",
|
| 44 |
+
"DKHateClassification",
|
| 45 |
+
"LccSentimentClassification",
|
| 46 |
+
"MassiveIntentClassification (da)",
|
| 47 |
+
"MassiveScenarioClassification (da)",
|
| 48 |
+
"NordicLangClassification",
|
| 49 |
+
"ScalaDaClassification",
|
| 50 |
+
]
|
| 51 |
+
|
| 52 |
+
TASK_LIST_CLASSIFICATION_NB = [
|
| 53 |
+
"NoRecClassification",
|
| 54 |
+
"NordicLangClassification",
|
| 55 |
+
"NorwegianParliament",
|
| 56 |
+
"MassiveIntentClassification (nb)",
|
| 57 |
+
"MassiveScenarioClassification (nb)",
|
| 58 |
+
"ScalaNbClassification",
|
| 59 |
+
]
|
| 60 |
+
|
| 61 |
+
TASK_LIST_CLASSIFICATION_PL = [
|
| 62 |
+
"AllegroReviews",
|
| 63 |
+
"CBD",
|
| 64 |
+
"MassiveIntentClassification (pl)",
|
| 65 |
+
"MassiveScenarioClassification (pl)",
|
| 66 |
+
"PAC",
|
| 67 |
+
"PolEmo2.0-IN",
|
| 68 |
+
"PolEmo2.0-OUT",
|
| 69 |
+
]
|
| 70 |
+
|
| 71 |
+
TASK_LIST_CLASSIFICATION_SV = [
|
| 72 |
+
"DalajClassification",
|
| 73 |
+
"MassiveIntentClassification (sv)",
|
| 74 |
+
"MassiveScenarioClassification (sv)",
|
| 75 |
+
"NordicLangClassification",
|
| 76 |
+
"ScalaSvClassification",
|
| 77 |
+
"SweRecClassification",
|
| 78 |
+
]
|
| 79 |
+
|
| 80 |
+
TASK_LIST_CLASSIFICATION_ZH = [
|
| 81 |
+
"AmazonReviewsClassification (zh)",
|
| 82 |
+
"IFlyTek",
|
| 83 |
+
"JDReview",
|
| 84 |
+
"MassiveIntentClassification (zh-CN)",
|
| 85 |
+
"MassiveScenarioClassification (zh-CN)",
|
| 86 |
+
"MultilingualSentiment",
|
| 87 |
+
"OnlineShopping",
|
| 88 |
+
"TNews",
|
| 89 |
+
"Waimai",
|
| 90 |
+
]
|
| 91 |
+
|
| 92 |
+
TASK_LIST_CLASSIFICATION_OTHER = ['AmazonCounterfactualClassification (de)', 'AmazonCounterfactualClassification (ja)', 'AmazonReviewsClassification (de)', 'AmazonReviewsClassification (es)', 'AmazonReviewsClassification (fr)', 'AmazonReviewsClassification (ja)', 'AmazonReviewsClassification (zh)', 'MTOPDomainClassification (de)', 'MTOPDomainClassification (es)', 'MTOPDomainClassification (fr)', 'MTOPDomainClassification (hi)', 'MTOPDomainClassification (th)', 'MTOPIntentClassification (de)', 'MTOPIntentClassification (es)', 'MTOPIntentClassification (fr)', 'MTOPIntentClassification (hi)', 'MTOPIntentClassification (th)', 'MassiveIntentClassification (af)', 'MassiveIntentClassification (am)', 'MassiveIntentClassification (ar)', 'MassiveIntentClassification (az)', 'MassiveIntentClassification (bn)', 'MassiveIntentClassification (cy)', 'MassiveIntentClassification (de)', 'MassiveIntentClassification (el)', 'MassiveIntentClassification (es)', 'MassiveIntentClassification (fa)', 'MassiveIntentClassification (fi)', 'MassiveIntentClassification (fr)', 'MassiveIntentClassification (he)', 'MassiveIntentClassification (hi)', 'MassiveIntentClassification (hu)', 'MassiveIntentClassification (hy)', 'MassiveIntentClassification (id)', 'MassiveIntentClassification (is)', 'MassiveIntentClassification (it)', 'MassiveIntentClassification (ja)', 'MassiveIntentClassification (jv)', 'MassiveIntentClassification (ka)', 'MassiveIntentClassification (km)', 'MassiveIntentClassification (kn)', 'MassiveIntentClassification (ko)', 'MassiveIntentClassification (lv)', 'MassiveIntentClassification (ml)', 'MassiveIntentClassification (mn)', 'MassiveIntentClassification (ms)', 'MassiveIntentClassification (my)', 'MassiveIntentClassification (nl)', 'MassiveIntentClassification (pt)', 'MassiveIntentClassification (ro)', 'MassiveIntentClassification (ru)', 'MassiveIntentClassification (sl)', 'MassiveIntentClassification (sq)', 'MassiveIntentClassification (sw)', 'MassiveIntentClassification (ta)', 'MassiveIntentClassification (te)', 'MassiveIntentClassification (th)', 'MassiveIntentClassification (tl)', 'MassiveIntentClassification (tr)', 'MassiveIntentClassification (ur)', 'MassiveIntentClassification (vi)', 'MassiveIntentClassification (zh-TW)', 'MassiveScenarioClassification (af)', 'MassiveScenarioClassification (am)', 'MassiveScenarioClassification (ar)', 'MassiveScenarioClassification (az)', 'MassiveScenarioClassification (bn)', 'MassiveScenarioClassification (cy)', 'MassiveScenarioClassification (de)', 'MassiveScenarioClassification (el)', 'MassiveScenarioClassification (es)', 'MassiveScenarioClassification (fa)', 'MassiveScenarioClassification (fi)', 'MassiveScenarioClassification (fr)', 'MassiveScenarioClassification (he)', 'MassiveScenarioClassification (hi)', 'MassiveScenarioClassification (hu)', 'MassiveScenarioClassification (hy)', 'MassiveScenarioClassification (id)', 'MassiveScenarioClassification (is)', 'MassiveScenarioClassification (it)', 'MassiveScenarioClassification (ja)', 'MassiveScenarioClassification (jv)', 'MassiveScenarioClassification (ka)', 'MassiveScenarioClassification (km)', 'MassiveScenarioClassification (kn)', 'MassiveScenarioClassification (ko)', 'MassiveScenarioClassification (lv)', 'MassiveScenarioClassification (ml)', 'MassiveScenarioClassification (mn)', 'MassiveScenarioClassification (ms)', 'MassiveScenarioClassification (my)', 'MassiveScenarioClassification (nl)', 'MassiveScenarioClassification (pt)', 'MassiveScenarioClassification (ro)', 'MassiveScenarioClassification (ru)', 'MassiveScenarioClassification (sl)', 'MassiveScenarioClassification (sq)', 'MassiveScenarioClassification (sw)', 'MassiveScenarioClassification (ta)', 'MassiveScenarioClassification (te)', 'MassiveScenarioClassification (th)', 'MassiveScenarioClassification (tl)', 'MassiveScenarioClassification (tr)', 'MassiveScenarioClassification (ur)', 'MassiveScenarioClassification (vi)', 'MassiveScenarioClassification (zh-TW)']
|
| 93 |
+
|
| 94 |
+
TASK_LIST_CLUSTERING = [
|
| 95 |
+
"ArxivClusteringP2P",
|
| 96 |
+
"ArxivClusteringS2S",
|
| 97 |
+
"BiorxivClusteringP2P",
|
| 98 |
+
"BiorxivClusteringS2S",
|
| 99 |
+
"MedrxivClusteringP2P",
|
| 100 |
+
"MedrxivClusteringS2S",
|
| 101 |
+
"RedditClustering",
|
| 102 |
+
"RedditClusteringP2P",
|
| 103 |
+
"StackExchangeClustering",
|
| 104 |
+
"StackExchangeClusteringP2P",
|
| 105 |
+
"TwentyNewsgroupsClustering",
|
| 106 |
+
]
|
| 107 |
+
|
| 108 |
+
|
| 109 |
+
TASK_LIST_CLUSTERING_DE = [
|
| 110 |
+
"BlurbsClusteringP2P",
|
| 111 |
+
"BlurbsClusteringS2S",
|
| 112 |
+
"TenKGnadClusteringP2P",
|
| 113 |
+
"TenKGnadClusteringS2S",
|
| 114 |
+
]
|
| 115 |
+
|
| 116 |
+
TASK_LIST_CLUSTERING_PL = [
|
| 117 |
+
"8TagsClustering",
|
| 118 |
+
]
|
| 119 |
+
|
| 120 |
+
TASK_LIST_CLUSTERING_ZH = [
|
| 121 |
+
"CLSClusteringP2P",
|
| 122 |
+
"CLSClusteringS2S",
|
| 123 |
+
"ThuNewsClusteringP2P",
|
| 124 |
+
"ThuNewsClusteringS2S",
|
| 125 |
+
]
|
| 126 |
+
|
| 127 |
+
TASK_LIST_PAIR_CLASSIFICATION = [
|
| 128 |
+
"SprintDuplicateQuestions",
|
| 129 |
+
"TwitterSemEval2015",
|
| 130 |
+
"TwitterURLCorpus",
|
| 131 |
+
]
|
| 132 |
+
|
| 133 |
+
TASK_LIST_PAIR_CLASSIFICATION_PL = [
|
| 134 |
+
"CDSC-E",
|
| 135 |
+
"PPC",
|
| 136 |
+
"PSC",
|
| 137 |
+
"SICK-E-PL",
|
| 138 |
+
]
|
| 139 |
+
|
| 140 |
+
TASK_LIST_PAIR_CLASSIFICATION_ZH = [
|
| 141 |
+
"Cmnli",
|
| 142 |
+
"Ocnli",
|
| 143 |
+
]
|
| 144 |
+
|
| 145 |
+
TASK_LIST_RERANKING = [
|
| 146 |
+
"AskUbuntuDupQuestions",
|
| 147 |
+
"MindSmallReranking",
|
| 148 |
+
"SciDocsRR",
|
| 149 |
+
"StackOverflowDupQuestions",
|
| 150 |
+
]
|
| 151 |
+
|
| 152 |
+
TASK_LIST_RERANKING_ZH = [
|
| 153 |
+
"CMedQAv1",
|
| 154 |
+
"CMedQAv2",
|
| 155 |
+
"MMarcoReranking",
|
| 156 |
+
"T2Reranking",
|
| 157 |
+
]
|
| 158 |
+
|
| 159 |
+
TASK_LIST_RETRIEVAL = [
|
| 160 |
+
"ArguAna",
|
| 161 |
+
"ClimateFEVER",
|
| 162 |
+
"CQADupstackRetrieval",
|
| 163 |
+
"DBPedia",
|
| 164 |
+
"FEVER",
|
| 165 |
+
"FiQA2018",
|
| 166 |
+
"HotpotQA",
|
| 167 |
+
"MSMARCO",
|
| 168 |
+
"NFCorpus",
|
| 169 |
+
"NQ",
|
| 170 |
+
"QuoraRetrieval",
|
| 171 |
+
"SCIDOCS",
|
| 172 |
+
"SciFact",
|
| 173 |
+
"Touche2020",
|
| 174 |
+
"TRECCOVID",
|
| 175 |
+
]
|
| 176 |
+
|
| 177 |
+
TASK_LIST_RETRIEVAL_PL = [
|
| 178 |
+
"ArguAna-PL",
|
| 179 |
+
"DBPedia-PL",
|
| 180 |
+
"FiQA-PL",
|
| 181 |
+
"HotpotQA-PL",
|
| 182 |
+
"MSMARCO-PL",
|
| 183 |
+
"NFCorpus-PL",
|
| 184 |
+
"NQ-PL",
|
| 185 |
+
"Quora-PL",
|
| 186 |
+
"SCIDOCS-PL",
|
| 187 |
+
"SciFact-PL",
|
| 188 |
+
"TRECCOVID-PL",
|
| 189 |
+
]
|
| 190 |
+
|
| 191 |
+
TASK_LIST_RETRIEVAL_ZH = [
|
| 192 |
+
"CmedqaRetrieval",
|
| 193 |
+
"CovidRetrieval",
|
| 194 |
+
"DuRetrieval",
|
| 195 |
+
"EcomRetrieval",
|
| 196 |
+
"MedicalRetrieval",
|
| 197 |
+
"MMarcoRetrieval",
|
| 198 |
+
"T2Retrieval",
|
| 199 |
+
"VideoRetrieval",
|
| 200 |
+
]
|
| 201 |
+
|
| 202 |
+
TASK_LIST_RETRIEVAL_NORM = TASK_LIST_RETRIEVAL + [
|
| 203 |
+
"CQADupstackAndroidRetrieval",
|
| 204 |
+
"CQADupstackEnglishRetrieval",
|
| 205 |
+
"CQADupstackGamingRetrieval",
|
| 206 |
+
"CQADupstackGisRetrieval",
|
| 207 |
+
"CQADupstackMathematicaRetrieval",
|
| 208 |
+
"CQADupstackPhysicsRetrieval",
|
| 209 |
+
"CQADupstackProgrammersRetrieval",
|
| 210 |
+
"CQADupstackStatsRetrieval",
|
| 211 |
+
"CQADupstackTexRetrieval",
|
| 212 |
+
"CQADupstackUnixRetrieval",
|
| 213 |
+
"CQADupstackWebmastersRetrieval",
|
| 214 |
+
"CQADupstackWordpressRetrieval"
|
| 215 |
+
]
|
| 216 |
+
|
| 217 |
+
TASK_LIST_STS = [
|
| 218 |
+
"BIOSSES",
|
| 219 |
+
"SICK-R",
|
| 220 |
+
"STS12",
|
| 221 |
+
"STS13",
|
| 222 |
+
"STS14",
|
| 223 |
+
"STS15",
|
| 224 |
+
"STS16",
|
| 225 |
+
"STS17 (en-en)",
|
| 226 |
+
"STS22 (en)",
|
| 227 |
+
"STSBenchmark",
|
| 228 |
+
]
|
| 229 |
+
|
| 230 |
+
TASK_LIST_STS_PL = [
|
| 231 |
+
"CDSC-R",
|
| 232 |
+
"SICK-R-PL",
|
| 233 |
+
"STS22 (pl)",
|
| 234 |
+
]
|
| 235 |
+
|
| 236 |
+
TASK_LIST_STS_ZH = [
|
| 237 |
+
"AFQMC",
|
| 238 |
+
"ATEC",
|
| 239 |
+
"BQ",
|
| 240 |
+
"LCQMC",
|
| 241 |
+
"PAWSX",
|
| 242 |
+
"QBQTC",
|
| 243 |
+
"STS22 (zh)",
|
| 244 |
+
"STSB",
|
| 245 |
+
]
|
| 246 |
+
|
| 247 |
+
TASK_LIST_STS_OTHER = ["STS17 (ar-ar)", "STS17 (en-ar)", "STS17 (en-de)", "STS17 (en-tr)", "STS17 (es-en)", "STS17 (es-es)", "STS17 (fr-en)", "STS17 (it-en)", "STS17 (ko-ko)", "STS17 (nl-en)", "STS22 (ar)", "STS22 (de)", "STS22 (de-en)", "STS22 (de-fr)", "STS22 (de-pl)", "STS22 (es)", "STS22 (es-en)", "STS22 (es-it)", "STS22 (fr)", "STS22 (fr-pl)", "STS22 (it)", "STS22 (pl)", "STS22 (pl-en)", "STS22 (ru)", "STS22 (tr)", "STS22 (zh-en)", "STSBenchmark",]
|
| 248 |
+
TASK_LIST_STS_NORM = [x.replace(" (en)", "").replace(" (en-en)", "") for x in TASK_LIST_STS]
|
| 249 |
+
|
| 250 |
+
TASK_LIST_SUMMARIZATION = ["SummEval",]
|
| 251 |
+
|
| 252 |
+
TASK_LIST_EN = TASK_LIST_CLASSIFICATION + TASK_LIST_CLUSTERING + TASK_LIST_PAIR_CLASSIFICATION + TASK_LIST_RERANKING + TASK_LIST_RETRIEVAL + TASK_LIST_STS + TASK_LIST_SUMMARIZATION
|
| 253 |
+
TASK_LIST_PL = TASK_LIST_CLASSIFICATION_PL + TASK_LIST_CLUSTERING_PL + TASK_LIST_PAIR_CLASSIFICATION_PL + TASK_LIST_RETRIEVAL_PL + TASK_LIST_STS_PL
|
| 254 |
+
TASK_LIST_ZH = TASK_LIST_CLASSIFICATION_ZH + TASK_LIST_CLUSTERING_ZH + TASK_LIST_PAIR_CLASSIFICATION_ZH + TASK_LIST_RERANKING_ZH + TASK_LIST_RETRIEVAL_ZH + TASK_LIST_STS_ZH
|
| 255 |
+
|
| 256 |
+
TASK_TO_METRIC = {
|
| 257 |
+
"BitextMining": "f1",
|
| 258 |
+
"Clustering": "v_measure",
|
| 259 |
+
"Classification": "accuracy",
|
| 260 |
+
"PairClassification": "cos_sim_ap",
|
| 261 |
+
"Reranking": "map",
|
| 262 |
+
"Retrieval": "ndcg_at_10",
|
| 263 |
+
"STS": "cos_sim_spearman",
|
| 264 |
+
"Summarization": "cos_sim_spearman",
|
| 265 |
+
}
|
| 266 |
+
|
| 267 |
+
def make_clickable_model(model_name, link=None):
|
| 268 |
+
if link is None:
|
| 269 |
+
link = "https://huggingface.co/" + model_name
|
| 270 |
+
# Remove user from model name
|
| 271 |
+
return (
|
| 272 |
+
f'<a target="_blank" style="text-decoration: underline" href="{link}">{model_name.split("/")[-1]}</a>'
|
| 273 |
+
)
|
| 274 |
+
|
| 275 |
+
# Models without metadata, thus we cannot fetch their results naturally
|
| 276 |
+
EXTERNAL_MODELS = [
|
| 277 |
+
"all-MiniLM-L12-v2",
|
| 278 |
+
"all-MiniLM-L6-v2",
|
| 279 |
+
"all-mpnet-base-v2",
|
| 280 |
+
"allenai-specter",
|
| 281 |
+
"Baichuan-text-embedding",
|
| 282 |
+
"bert-base-swedish-cased",
|
| 283 |
+
"bert-base-uncased",
|
| 284 |
+
"bge-base-zh-v1.5",
|
| 285 |
+
"bge-large-zh-v1.5",
|
| 286 |
+
"bge-large-zh-noinstruct",
|
| 287 |
+
"bge-small-zh-v1.5",
|
| 288 |
+
"contriever-base-msmarco",
|
| 289 |
+
"cross-en-de-roberta-sentence-transformer",
|
| 290 |
+
"dfm-encoder-large-v1",
|
| 291 |
+
"dfm-sentence-encoder-large-1",
|
| 292 |
+
"distiluse-base-multilingual-cased-v2",
|
| 293 |
+
"DanskBERT",
|
| 294 |
+
"e5-base",
|
| 295 |
+
"e5-large",
|
| 296 |
+
"e5-small",
|
| 297 |
+
"electra-small-nordic",
|
| 298 |
+
"electra-small-swedish-cased-discriminator",
|
| 299 |
+
"gbert-base",
|
| 300 |
+
"gbert-large",
|
| 301 |
+
"gelectra-base",
|
| 302 |
+
"gelectra-large",
|
| 303 |
+
"gottbert-base",
|
| 304 |
+
"glove.6B.300d",
|
| 305 |
+
"gtr-t5-base",
|
| 306 |
+
"gtr-t5-large",
|
| 307 |
+
"gtr-t5-xl",
|
| 308 |
+
"gtr-t5-xxl",
|
| 309 |
+
"herbert-base-retrieval-v2",
|
| 310 |
+
"komninos",
|
| 311 |
+
"luotuo-bert-medium",
|
| 312 |
+
"LASER2",
|
| 313 |
+
"LaBSE",
|
| 314 |
+
"m3e-base",
|
| 315 |
+
"m3e-large",
|
| 316 |
+
"msmarco-bert-co-condensor",
|
| 317 |
+
"multilingual-e5-base",
|
| 318 |
+
"multilingual-e5-large",
|
| 319 |
+
"multilingual-e5-small",
|
| 320 |
+
"nb-bert-base",
|
| 321 |
+
"nb-bert-large",
|
| 322 |
+
"norbert3-base",
|
| 323 |
+
"norbert3-large",
|
| 324 |
+
"paraphrase-multilingual-MiniLM-L12-v2",
|
| 325 |
+
"paraphrase-multilingual-mpnet-base-v2",
|
| 326 |
+
"sentence-bert-swedish-cased",
|
| 327 |
+
"sentence-t5-base",
|
| 328 |
+
"sentence-t5-large",
|
| 329 |
+
"sentence-t5-xl",
|
| 330 |
+
"sentence-t5-xxl",
|
| 331 |
+
"sup-simcse-bert-base-uncased",
|
| 332 |
+
"st-polish-paraphrase-from-distilroberta",
|
| 333 |
+
"st-polish-paraphrase-from-mpnet",
|
| 334 |
+
"text2vec-base-chinese",
|
| 335 |
+
"text2vec-large-chinese",
|
| 336 |
+
"text-embedding-ada-002",
|
| 337 |
+
"text-similarity-ada-001",
|
| 338 |
+
"text-similarity-babbage-001",
|
| 339 |
+
"text-similarity-curie-001",
|
| 340 |
+
"text-similarity-davinci-001",
|
| 341 |
+
"text-search-ada-doc-001",
|
| 342 |
+
"text-search-ada-001",
|
| 343 |
+
"text-search-babbage-001",
|
| 344 |
+
"text-search-curie-001",
|
| 345 |
+
"text-search-davinci-001",
|
| 346 |
+
"titan-embed-text-v1",
|
| 347 |
+
"unsup-simcse-bert-base-uncased",
|
| 348 |
+
"use-cmlm-multilingual",
|
| 349 |
+
"voyage-lite-01-instruct",
|
| 350 |
+
"xlm-roberta-base",
|
| 351 |
+
"xlm-roberta-large",
|
| 352 |
+
]
|
| 353 |
+
|
| 354 |
+
EXTERNAL_MODEL_TO_LINK = {
|
| 355 |
+
"allenai-specter": "https://huggingface.co/sentence-transformers/allenai-specter",
|
| 356 |
+
"allenai-specter": "https://huggingface.co/sentence-transformers/allenai-specter",
|
| 357 |
+
"all-MiniLM-L12-v2": "https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2",
|
| 358 |
+
"all-MiniLM-L6-v2": "https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2",
|
| 359 |
+
"all-mpnet-base-v2": "https://huggingface.co/sentence-transformers/all-mpnet-base-v2",
|
| 360 |
+
"Baichuan-text-embedding": "https://platform.baichuan-ai.com/docs/text-Embedding",
|
| 361 |
+
"bert-base-swedish-cased": "https://huggingface.co/KB/bert-base-swedish-cased",
|
| 362 |
+
"bert-base-uncased": "https://huggingface.co/bert-base-uncased",
|
| 363 |
+
"bge-base-zh-v1.5": "https://huggingface.co/BAAI/bge-base-zh-v1.5",
|
| 364 |
+
"bge-large-zh-v1.5": "https://huggingface.co/BAAI/bge-large-zh-v1.5",
|
| 365 |
+
"bge-large-zh-noinstruct": "https://huggingface.co/BAAI/bge-large-zh-noinstruct",
|
| 366 |
+
"bge-small-zh-v1.5": "https://huggingface.co/BAAI/bge-small-zh-v1.5",
|
| 367 |
+
"contriever-base-msmarco": "https://huggingface.co/nthakur/contriever-base-msmarco",
|
| 368 |
+
"cross-en-de-roberta-sentence-transformer": "https://huggingface.co/T-Systems-onsite/cross-en-de-roberta-sentence-transformer",
|
| 369 |
+
"DanskBERT": "https://huggingface.co/vesteinn/DanskBERT",
|
| 370 |
+
"distiluse-base-multilingual-cased-v2": "https://huggingface.co/sentence-transformers/distiluse-base-multilingual-cased-v2",
|
| 371 |
+
"dfm-encoder-large-v1": "https://huggingface.co/chcaa/dfm-encoder-large-v1",
|
| 372 |
+
"dfm-sentence-encoder-large-1": "https://huggingface.co/chcaa/dfm-encoder-large-v1",
|
| 373 |
+
"e5-base": "https://huggingface.co/intfloat/e5-base",
|
| 374 |
+
"e5-large": "https://huggingface.co/intfloat/e5-large",
|
| 375 |
+
"e5-small": "https://huggingface.co/intfloat/e5-small",
|
| 376 |
+
"electra-small-nordic": "https://huggingface.co/jonfd/electra-small-nordic",
|
| 377 |
+
"electra-small-swedish-cased-discriminator": "https://huggingface.co/KBLab/electra-small-swedish-cased-discriminator",
|
| 378 |
+
"gbert-base": "https://huggingface.co/deepset/gbert-base",
|
| 379 |
+
"gbert-large": "https://huggingface.co/deepset/gbert-large",
|
| 380 |
+
"gelectra-base": "https://huggingface.co/deepset/gelectra-base",
|
| 381 |
+
"gelectra-large": "https://huggingface.co/deepset/gelectra-large",
|
| 382 |
+
"glove.6B.300d": "https://huggingface.co/sentence-transformers/average_word_embeddings_glove.6B.300d",
|
| 383 |
+
"gottbert-base": "https://huggingface.co/uklfr/gottbert-base",
|
| 384 |
+
"gtr-t5-base": "https://huggingface.co/sentence-transformers/gtr-t5-base",
|
| 385 |
+
"gtr-t5-large": "https://huggingface.co/sentence-transformers/gtr-t5-large",
|
| 386 |
+
"gtr-t5-xl": "https://huggingface.co/sentence-transformers/gtr-t5-xl",
|
| 387 |
+
"gtr-t5-xxl": "https://huggingface.co/sentence-transformers/gtr-t5-xxl",
|
| 388 |
+
"herbert-base-retrieval-v2": "https://huggingface.co/ipipan/herbert-base-retrieval-v2",
|
| 389 |
+
"komninos": "https://huggingface.co/sentence-transformers/average_word_embeddings_komninos",
|
| 390 |
+
"luotuo-bert-medium": "https://huggingface.co/silk-road/luotuo-bert-medium",
|
| 391 |
+
"LASER2": "https://github.com/facebookresearch/LASER",
|
| 392 |
+
"LaBSE": "https://huggingface.co/sentence-transformers/LaBSE",
|
| 393 |
+
"m3e-base": "https://huggingface.co/moka-ai/m3e-base",
|
| 394 |
+
"m3e-large": "https://huggingface.co/moka-ai/m3e-large",
|
| 395 |
+
"msmarco-bert-co-condensor": "https://huggingface.co/sentence-transformers/msmarco-bert-co-condensor",
|
| 396 |
+
"multilingual-e5-base": "https://huggingface.co/intfloat/multilingual-e5-base",
|
| 397 |
+
"multilingual-e5-large": "https://huggingface.co/intfloat/multilingual-e5-large",
|
| 398 |
+
"multilingual-e5-small": "https://huggingface.co/intfloat/multilingual-e5-small",
|
| 399 |
+
"nb-bert-base": "https://huggingface.co/NbAiLab/nb-bert-base",
|
| 400 |
+
"nb-bert-large": "https://huggingface.co/NbAiLab/nb-bert-large",
|
| 401 |
+
"norbert3-base": "https://huggingface.co/ltg/norbert3-base",
|
| 402 |
+
"norbert3-large": "https://huggingface.co/ltg/norbert3-large",
|
| 403 |
+
"paraphrase-multilingual-mpnet-base-v2": "https://huggingface.co/sentence-transformers/paraphrase-multilingual-mpnet-base-v2",
|
| 404 |
+
"paraphrase-multilingual-MiniLM-L12-v2": "https://huggingface.co/sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
|
| 405 |
+
"sentence-bert-swedish-cased": "https://huggingface.co/KBLab/sentence-bert-swedish-cased",
|
| 406 |
+
"sentence-t5-base": "https://huggingface.co/sentence-transformers/sentence-t5-base",
|
| 407 |
+
"sentence-t5-large": "https://huggingface.co/sentence-transformers/sentence-t5-large",
|
| 408 |
+
"sentence-t5-xl": "https://huggingface.co/sentence-transformers/sentence-t5-xl",
|
| 409 |
+
"sentence-t5-xxl": "https://huggingface.co/sentence-transformers/sentence-t5-xxl",
|
| 410 |
+
"sup-simcse-bert-base-uncased": "https://huggingface.co/princeton-nlp/sup-simcse-bert-base-uncased",
|
| 411 |
+
"st-polish-paraphrase-from-distilroberta": "https://huggingface.co/sdadas/st-polish-paraphrase-from-distilroberta",
|
| 412 |
+
"st-polish-paraphrase-from-mpnet": "https://huggingface.co/sdadas/st-polish-paraphrase-from-mpnet",
|
| 413 |
+
"text2vec-base-chinese": "https://huggingface.co/shibing624/text2vec-base-chinese",
|
| 414 |
+
"text2vec-large-chinese": "https://huggingface.co/GanymedeNil/text2vec-large-chinese",
|
| 415 |
+
"text-embedding-ada-002": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 416 |
+
"text-similarity-ada-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 417 |
+
"text-similarity-babbage-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 418 |
+
"text-similarity-curie-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 419 |
+
"text-similarity-davinci-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 420 |
+
"text-search-ada-doc-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 421 |
+
"text-search-ada-query-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 422 |
+
"text-search-ada-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 423 |
+
"text-search-curie-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 424 |
+
"text-search-babbage-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 425 |
+
"text-search-davinci-001": "https://beta.openai.com/docs/guides/embeddings/types-of-embedding-models",
|
| 426 |
+
"titan-embed-text-v1": "https://docs.aws.amazon.com/bedrock/latest/userguide/embeddings.html",
|
| 427 |
+
"unsup-simcse-bert-base-uncased": "https://huggingface.co/princeton-nlp/unsup-simcse-bert-base-uncased",
|
| 428 |
+
"use-cmlm-multilingual": "https://huggingface.co/sentence-transformers/use-cmlm-multilingual",
|
| 429 |
+
"voyage-lite-01-instruct": "https://docs.voyageai.com/embeddings/",
|
| 430 |
+
"xlm-roberta-base": "https://huggingface.co/xlm-roberta-base",
|
| 431 |
+
"xlm-roberta-large": "https://huggingface.co/xlm-roberta-large",
|
| 432 |
+
}
|
| 433 |
+
|
| 434 |
+
EXTERNAL_MODEL_TO_DIM = {
|
| 435 |
+
"all-MiniLM-L12-v2": 384,
|
| 436 |
+
"all-MiniLM-L6-v2": 384,
|
| 437 |
+
"all-mpnet-base-v2": 768,
|
| 438 |
+
"allenai-specter": 768,
|
| 439 |
+
"Baichuan-text-embedding": 1024,
|
| 440 |
+
"bert-base-swedish-cased": 768,
|
| 441 |
+
"bert-base-uncased": 768,
|
| 442 |
+
"bge-base-zh-v1.5": 768,
|
| 443 |
+
"bge-large-zh-v1.5": 1024,
|
| 444 |
+
"bge-large-zh-noinstruct": 1024,
|
| 445 |
+
"bge-small-zh-v1.5": 512,
|
| 446 |
+
"contriever-base-msmarco": 768,
|
| 447 |
+
"cross-en-de-roberta-sentence-transformer": 768,
|
| 448 |
+
"DanskBERT": 768,
|
| 449 |
+
"distiluse-base-multilingual-cased-v2": 512,
|
| 450 |
+
"dfm-encoder-large-v1": 1024,
|
| 451 |
+
"dfm-sentence-encoder-large-1": 1024,
|
| 452 |
+
"e5-base": 768,
|
| 453 |
+
"e5-small": 384,
|
| 454 |
+
"e5-large": 1024,
|
| 455 |
+
"electra-small-nordic": 256,
|
| 456 |
+
"electra-small-swedish-cased-discriminator": 256,
|
| 457 |
+
"luotuo-bert-medium": 768,
|
| 458 |
+
"LASER2": 1024,
|
| 459 |
+
"LaBSE": 768,
|
| 460 |
+
"gbert-base": 768,
|
| 461 |
+
"gbert-large": 1024,
|
| 462 |
+
"gelectra-base": 768,
|
| 463 |
+
"gelectra-large": 1024,
|
| 464 |
+
"glove.6B.300d": 300,
|
| 465 |
+
"gottbert-base": 768,
|
| 466 |
+
"gtr-t5-base": 768,
|
| 467 |
+
"gtr-t5-large": 768,
|
| 468 |
+
"gtr-t5-xl": 768,
|
| 469 |
+
"gtr-t5-xxl": 768,
|
| 470 |
+
"herbert-base-retrieval-v2": 768,
|
| 471 |
+
"komninos": 300,
|
| 472 |
+
"m3e-base": 768,
|
| 473 |
+
"m3e-large": 768,
|
| 474 |
+
"msmarco-bert-co-condensor": 768,
|
| 475 |
+
"multilingual-e5-base": 768,
|
| 476 |
+
"multilingual-e5-small": 384,
|
| 477 |
+
"multilingual-e5-large": 1024,
|
| 478 |
+
"nb-bert-base": 768,
|
| 479 |
+
"nb-bert-large": 1024,
|
| 480 |
+
"norbert3-base": 768,
|
| 481 |
+
"norbert3-large": 1024,
|
| 482 |
+
"paraphrase-multilingual-MiniLM-L12-v2": 384,
|
| 483 |
+
"paraphrase-multilingual-mpnet-base-v2": 768,
|
| 484 |
+
"sentence-bert-swedish-cased": 768,
|
| 485 |
+
"sentence-t5-base": 768,
|
| 486 |
+
"sentence-t5-large": 768,
|
| 487 |
+
"sentence-t5-xl": 768,
|
| 488 |
+
"sentence-t5-xxl": 768,
|
| 489 |
+
"sup-simcse-bert-base-uncased": 768,
|
| 490 |
+
"st-polish-paraphrase-from-distilroberta": 768,
|
| 491 |
+
"st-polish-paraphrase-from-mpnet": 768,
|
| 492 |
+
"text2vec-base-chinese": 768,
|
| 493 |
+
"text2vec-large-chinese": 1024,
|
| 494 |
+
"text-embedding-ada-002": 1536,
|
| 495 |
+
"text-similarity-ada-001": 1024,
|
| 496 |
+
"text-similarity-babbage-001": 2048,
|
| 497 |
+
"text-similarity-curie-001": 4096,
|
| 498 |
+
"text-similarity-davinci-001": 12288,
|
| 499 |
+
"text-search-ada-doc-001": 1024,
|
| 500 |
+
"text-search-ada-query-001": 1024,
|
| 501 |
+
"text-search-ada-001": 1024,
|
| 502 |
+
"text-search-babbage-001": 2048,
|
| 503 |
+
"text-search-curie-001": 4096,
|
| 504 |
+
"text-search-davinci-001": 12288,
|
| 505 |
+
"titan-embed-text-v1": 1536,
|
| 506 |
+
"unsup-simcse-bert-base-uncased": 768,
|
| 507 |
+
"use-cmlm-multilingual": 768,
|
| 508 |
+
"voyage-lite-01-instruct": 1024,
|
| 509 |
+
"xlm-roberta-base": 768,
|
| 510 |
+
"xlm-roberta-large": 1024,
|
| 511 |
+
}
|
| 512 |
+
|
| 513 |
+
EXTERNAL_MODEL_TO_SEQLEN = {
|
| 514 |
+
"all-MiniLM-L12-v2": 512,
|
| 515 |
+
"all-MiniLM-L6-v2": 512,
|
| 516 |
+
"all-mpnet-base-v2": 514,
|
| 517 |
+
"allenai-specter": 512,
|
| 518 |
+
"Baichuan-text-embedding": 512,
|
| 519 |
+
"bert-base-swedish-cased": 512,
|
| 520 |
+
"bert-base-uncased": 512,
|
| 521 |
+
"bge-base-zh-v1.5": 512,
|
| 522 |
+
"bge-large-zh-v1.5": 512,
|
| 523 |
+
"bge-large-zh-noinstruct": 512,
|
| 524 |
+
"bge-small-zh-v1.5": 512,
|
| 525 |
+
"contriever-base-msmarco": 512,
|
| 526 |
+
"cross-en-de-roberta-sentence-transformer": 514,
|
| 527 |
+
"DanskBERT": 514,
|
| 528 |
+
"dfm-encoder-large-v1": 512,
|
| 529 |
+
"dfm-sentence-encoder-large-1": 512,
|
| 530 |
+
"distiluse-base-multilingual-cased-v2": 512,
|
| 531 |
+
"e5-base": 512,
|
| 532 |
+
"e5-large": 512,
|
| 533 |
+
"e5-small": 512,
|
| 534 |
+
"electra-small-nordic": 512,
|
| 535 |
+
"electra-small-swedish-cased-discriminator": 512,
|
| 536 |
+
"gbert-base": 512,
|
| 537 |
+
"gbert-large": 512,
|
| 538 |
+
"gelectra-base": 512,
|
| 539 |
+
"gelectra-large": 512,
|
| 540 |
+
"gottbert-base": 512,
|
| 541 |
+
"glove.6B.300d": "N/A",
|
| 542 |
+
"gtr-t5-base": 512,
|
| 543 |
+
"gtr-t5-large": 512,
|
| 544 |
+
"gtr-t5-xl": 512,
|
| 545 |
+
"gtr-t5-xxl": 512,
|
| 546 |
+
"herbert-base-retrieval-v2": 514,
|
| 547 |
+
"komninos": "N/A",
|
| 548 |
+
"luotuo-bert-medium": 512,
|
| 549 |
+
"LASER2": "N/A",
|
| 550 |
+
"LaBSE": 512,
|
| 551 |
+
"m3e-base": 512,
|
| 552 |
+
"m3e-large": 512,
|
| 553 |
+
"msmarco-bert-co-condensor": 512,
|
| 554 |
+
"multilingual-e5-base": 514,
|
| 555 |
+
"multilingual-e5-large": 514,
|
| 556 |
+
"multilingual-e5-small": 512,
|
| 557 |
+
"nb-bert-base": 512,
|
| 558 |
+
"nb-bert-large": 512,
|
| 559 |
+
"norbert3-base": 512,
|
| 560 |
+
"norbert3-large": 512,
|
| 561 |
+
"paraphrase-multilingual-MiniLM-L12-v2": 512,
|
| 562 |
+
"paraphrase-multilingual-mpnet-base-v2": 514,
|
| 563 |
+
"sentence-bert-swedish-cased": 512,
|
| 564 |
+
"sentence-t5-base": 512,
|
| 565 |
+
"sentence-t5-large": 512,
|
| 566 |
+
"sentence-t5-xl": 512,
|
| 567 |
+
"sentence-t5-xxl": 512,
|
| 568 |
+
"sup-simcse-bert-base-uncased": 512,
|
| 569 |
+
"st-polish-paraphrase-from-distilroberta": 514,
|
| 570 |
+
"st-polish-paraphrase-from-mpnet": 514,
|
| 571 |
+
"text2vec-base-chinese": 512,
|
| 572 |
+
"text2vec-large-chinese": 512,
|
| 573 |
+
"text-embedding-ada-002": 8191,
|
| 574 |
+
"text-similarity-ada-001": 2046,
|
| 575 |
+
"text-similarity-babbage-001": 2046,
|
| 576 |
+
"text-similarity-curie-001": 2046,
|
| 577 |
+
"text-similarity-davinci-001": 2046,
|
| 578 |
+
"text-search-ada-doc-001": 2046,
|
| 579 |
+
"text-search-ada-query-001": 2046,
|
| 580 |
+
"text-search-ada-001": 2046,
|
| 581 |
+
"text-search-babbage-001": 2046,
|
| 582 |
+
"text-search-curie-001": 2046,
|
| 583 |
+
"text-search-davinci-001": 2046,
|
| 584 |
+
"titan-embed-text-v1": 8000,
|
| 585 |
+
"use-cmlm-multilingual": 512,
|
| 586 |
+
"unsup-simcse-bert-base-uncased": 512,
|
| 587 |
+
"voyage-lite-01-instruct": 4096,
|
| 588 |
+
"xlm-roberta-base": 514,
|
| 589 |
+
"xlm-roberta-large": 514,
|
| 590 |
+
}
|
| 591 |
+
|
| 592 |
+
EXTERNAL_MODEL_TO_SIZE = {
|
| 593 |
+
"allenai-specter": 0.44,
|
| 594 |
+
"all-MiniLM-L12-v2": 0.13,
|
| 595 |
+
"all-MiniLM-L6-v2": 0.09,
|
| 596 |
+
"all-mpnet-base-v2": 0.44,
|
| 597 |
+
"bert-base-uncased": 0.44,
|
| 598 |
+
"bert-base-swedish-cased": 0.50,
|
| 599 |
+
"bge-base-zh-v1.5": 0.41,
|
| 600 |
+
"bge-large-zh-v1.5": 1.30,
|
| 601 |
+
"bge-large-zh-noinstruct": 1.30,
|
| 602 |
+
"bge-small-zh-v1.5": 0.10,
|
| 603 |
+
"cross-en-de-roberta-sentence-transformer": 1.11,
|
| 604 |
+
"contriever-base-msmarco": 0.44,
|
| 605 |
+
"DanskBERT": 0.50,
|
| 606 |
+
"distiluse-base-multilingual-cased-v2": 0.54,
|
| 607 |
+
"dfm-encoder-large-v1": 1.42,
|
| 608 |
+
"dfm-sentence-encoder-large-1": 1.63,
|
| 609 |
+
"e5-base": 0.44,
|
| 610 |
+
"e5-small": 0.13,
|
| 611 |
+
"e5-large": 1.34,
|
| 612 |
+
"electra-small-nordic": 0.09,
|
| 613 |
+
"electra-small-swedish-cased-discriminator": 0.06,
|
| 614 |
+
"gbert-base": 0.44,
|
| 615 |
+
"gbert-large": 1.35,
|
| 616 |
+
"gelectra-base": 0.44,
|
| 617 |
+
"gelectra-large": 1.34,
|
| 618 |
+
"glove.6B.300d": 0.48,
|
| 619 |
+
"gottbert-base": 0.51,
|
| 620 |
+
"gtr-t5-base": 0.22,
|
| 621 |
+
"gtr-t5-large": 0.67,
|
| 622 |
+
"gtr-t5-xl": 2.48,
|
| 623 |
+
"gtr-t5-xxl": 9.73,
|
| 624 |
+
"herbert-base-retrieval-v2": 0.50,
|
| 625 |
+
"komninos": 0.27,
|
| 626 |
+
"luotuo-bert-medium": 1.31,
|
| 627 |
+
"LASER2": 0.17,
|
| 628 |
+
"LaBSE": 1.88,
|
| 629 |
+
"m3e-base": 0.41,
|
| 630 |
+
"m3e-large": 0.41,
|
| 631 |
+
"msmarco-bert-co-condensor": 0.44,
|
| 632 |
+
"multilingual-e5-base": 1.11,
|
| 633 |
+
"multilingual-e5-small": 0.47,
|
| 634 |
+
"multilingual-e5-large": 2.24,
|
| 635 |
+
"nb-bert-base": 0.71,
|
| 636 |
+
"nb-bert-large": 1.42,
|
| 637 |
+
"norbert3-base": 0.52,
|
| 638 |
+
"norbert3-large": 1.47,
|
| 639 |
+
"paraphrase-multilingual-mpnet-base-v2": 1.11,
|
| 640 |
+
"paraphrase-multilingual-MiniLM-L12-v2": 0.47,
|
| 641 |
+
"sentence-bert-swedish-cased": 0.50,
|
| 642 |
+
"sentence-t5-base": 0.22,
|
| 643 |
+
"sentence-t5-large": 0.67,
|
| 644 |
+
"sentence-t5-xl": 2.48,
|
| 645 |
+
"sentence-t5-xxl": 9.73,
|
| 646 |
+
"sup-simcse-bert-base-uncased": 0.44,
|
| 647 |
+
"st-polish-paraphrase-from-distilroberta": 0.50,
|
| 648 |
+
"st-polish-paraphrase-from-mpnet": 0.50,
|
| 649 |
+
"text2vec-base-chinese": 0.41,
|
| 650 |
+
"text2vec-large-chinese": 1.30,
|
| 651 |
+
"unsup-simcse-bert-base-uncased": 0.44,
|
| 652 |
+
"use-cmlm-multilingual": 1.89,
|
| 653 |
+
"xlm-roberta-base": 1.12,
|
| 654 |
+
"xlm-roberta-large": 2.24,
|
| 655 |
+
}
|
| 656 |
+
|
| 657 |
+
MODELS_TO_SKIP = {
|
| 658 |
+
"baseplate/instructor-large-1", # Duplicate
|
| 659 |
+
"radames/e5-large", # Duplicate
|
| 660 |
+
"gentlebowl/instructor-large-safetensors", # Duplicate
|
| 661 |
+
"Consensus/instructor-base", # Duplicate
|
| 662 |
+
"GovCompete/instructor-xl", # Duplicate
|
| 663 |
+
"GovCompete/e5-large-v2", # Duplicate
|
| 664 |
+
"t12e/instructor-base", # Duplicate
|
| 665 |
+
"michaelfeil/ct2fast-e5-large-v2",
|
| 666 |
+
"michaelfeil/ct2fast-e5-large",
|
| 667 |
+
"michaelfeil/ct2fast-e5-small-v2",
|
| 668 |
+
"newsrx/instructor-xl-newsrx",
|
| 669 |
+
"newsrx/instructor-large-newsrx",
|
| 670 |
+
"fresha/e5-large-v2-endpoint",
|
| 671 |
+
"ggrn/e5-small-v2",
|
| 672 |
+
"michaelfeil/ct2fast-e5-small",
|
| 673 |
+
"jncraton/e5-small-v2-ct2-int8",
|
| 674 |
+
"anttip/ct2fast-e5-small-v2-hfie",
|
| 675 |
+
"newsrx/instructor-large",
|
| 676 |
+
"newsrx/instructor-xl",
|
| 677 |
+
"dmlls/all-mpnet-base-v2",
|
| 678 |
+
"cgldo/semanticClone",
|
| 679 |
+
"Malmuk1/e5-large-v2_Sharded",
|
| 680 |
+
"jncraton/gte-small-ct2-int8",
|
| 681 |
+
"Einas/einas_ashkar",
|
| 682 |
+
"gruber/e5-small-v2-ggml",
|
| 683 |
+
"jncraton/bge-small-en-ct2-int8",
|
| 684 |
+
"vectoriseai/bge-small-en",
|
| 685 |
+
"recipe/embeddings",
|
| 686 |
+
"dhairya0907/thenlper-get-large",
|
| 687 |
+
"Narsil/bge-base-en",
|
| 688 |
+
"kozistr/fused-large-en",
|
| 689 |
+
"sionic-ai/sionic-ai-v2", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
|
| 690 |
+
"sionic-ai/sionic-ai-v1", # Wait for https://huggingface.co/sionic-ai/sionic-ai-v2/discussions/1
|
| 691 |
+
"BAAI/bge-large-en", # Deprecated in favor of v1.5
|
| 692 |
+
"BAAI/bge-base-en", # Deprecated in favor of v1.5
|
| 693 |
+
"BAAI/bge-small-en", # Deprecated in favor of v1.5
|
| 694 |
+
"d0rj/e5-large-en-ru",
|
| 695 |
+
"d0rj/e5-base-en-ru",
|
| 696 |
+
"d0rj/e5-small-en-ru",
|
| 697 |
+
"aident-ai/bge-base-en-onnx",
|
| 698 |
+
"barisaydin/bge-base-en",
|
| 699 |
+
"barisaydin/gte-large",
|
| 700 |
+
"barisaydin/gte-base",
|
| 701 |
+
"barisaydin/gte-small",
|
| 702 |
+
"barisaydin/bge-small-en",
|
| 703 |
+
"odunola/e5-base-v2",
|
| 704 |
+
"goldenrooster/multilingual-e5-large",
|
| 705 |
+
"davidpeer/gte-small",
|
| 706 |
+
"barisaydin/bge-large-en",
|
| 707 |
+
"jamesgpt1/english-large-v1",
|
| 708 |
+
"vectoriseai/bge-large-en-v1.5",
|
| 709 |
+
"vectoriseai/bge-base-en-v1.5",
|
| 710 |
+
"vectoriseai/instructor-large",
|
| 711 |
+
"vectoriseai/instructor-base",
|
| 712 |
+
"vectoriseai/gte-large",
|
| 713 |
+
"vectoriseai/gte-base",
|
| 714 |
+
"vectoriseai/e5-large-v2",
|
| 715 |
+
"vectoriseai/bge-small-en-v1.5",
|
| 716 |
+
"vectoriseai/e5-base-v2",
|
| 717 |
+
"vectoriseai/e5-large",
|
| 718 |
+
"vectoriseai/multilingual-e5-large",
|
| 719 |
+
"vectoriseai/gte-small",
|
| 720 |
+
"vectoriseai/ember-v1",
|
| 721 |
+
"vectoriseai/e5-base",
|
| 722 |
+
"vectoriseai/e5-small-v2",
|
| 723 |
+
"michaelfeil/ct2fast-bge-large-en-v1.5",
|
| 724 |
+
"michaelfeil/ct2fast-bge-large-en-v1.5",
|
| 725 |
+
"michaelfeil/ct2fast-bge-base-en-v1.5",
|
| 726 |
+
"michaelfeil/ct2fast-gte-large",
|
| 727 |
+
"michaelfeil/ct2fast-gte-base",
|
| 728 |
+
"michaelfeil/ct2fast-bge-small-en-v1.5",
|
| 729 |
+
"rizki/bgr-tf",
|
| 730 |
+
"ef-zulla/e5-multi-sml-torch",
|
| 731 |
+
"cherubhao/yogamodel",
|
| 732 |
+
"morgendigital/multilingual-e5-large-quantized",
|
| 733 |
+
"jncraton/gte-tiny-ct2-int8",
|
| 734 |
+
"Research2NLP/electrical_stella",
|
| 735 |
+
"Intel/bge-base-en-v1.5-sts-int8-static",
|
| 736 |
+
"Intel/bge-base-en-v1.5-sts-int8-dynamic",
|
| 737 |
+
"Intel/bge-base-en-v1.5-sst2",
|
| 738 |
+
"Intel/bge-base-en-v1.5-sst2-int8-static",
|
| 739 |
+
"Intel/bge-base-en-v1.5-sst2-int8-dynamic",
|
| 740 |
+
"Intel/bge-small-en-v1.5-sst2",
|
| 741 |
+
"Intel/bge-small-en-v1.5-sst2-int8-dynamic",
|
| 742 |
+
"Intel/bge-small-en-v1.5-sst2-int8-static",
|
| 743 |
+
"binqiangliu/EmbeddingModlebgelargeENv1.5",
|
| 744 |
+
"DecisionOptimizationSystem/DeepFeatEmbeddingLargeContext",
|
| 745 |
+
"woody72/multilingual-e5-base",
|
| 746 |
+
"Severian/embed",
|
| 747 |
+
"Frazic/udever-bloom-3b-sentence",
|
| 748 |
+
"jamesgpt1/zzz",
|
| 749 |
+
}
|
| 750 |
+
|
| 751 |
+
EXTERNAL_MODEL_RESULTS = {model: {k: {v: []} for k, v in TASK_TO_METRIC.items()} for model in EXTERNAL_MODELS}
|
| 752 |
+
|
| 753 |
+
def add_lang(examples):
|
| 754 |
+
if not(examples["eval_language"]):
|
| 755 |
+
examples["mteb_dataset_name_with_lang"] = examples["mteb_dataset_name"]
|
| 756 |
+
else:
|
| 757 |
+
examples["mteb_dataset_name_with_lang"] = examples["mteb_dataset_name"] + f' ({examples["eval_language"]})'
|
| 758 |
+
return examples
|
| 759 |
+
|
| 760 |
+
def add_task(examples):
|
| 761 |
+
# Could be added to the dataset loading script instead
|
| 762 |
+
if examples["mteb_dataset_name"] in TASK_LIST_CLASSIFICATION_NORM + TASK_LIST_CLASSIFICATION_DA + TASK_LIST_CLASSIFICATION_NB + TASK_LIST_CLASSIFICATION_PL + TASK_LIST_CLASSIFICATION_SV + TASK_LIST_CLASSIFICATION_ZH:
|
| 763 |
+
examples["mteb_task"] = "Classification"
|
| 764 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_CLUSTERING + TASK_LIST_CLUSTERING_DE + TASK_LIST_CLUSTERING_PL + TASK_LIST_CLUSTERING_ZH:
|
| 765 |
+
examples["mteb_task"] = "Clustering"
|
| 766 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_PAIR_CLASSIFICATION + TASK_LIST_PAIR_CLASSIFICATION_PL + TASK_LIST_PAIR_CLASSIFICATION_ZH:
|
| 767 |
+
examples["mteb_task"] = "PairClassification"
|
| 768 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_RERANKING + TASK_LIST_RERANKING_ZH:
|
| 769 |
+
examples["mteb_task"] = "Reranking"
|
| 770 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_RETRIEVAL_NORM + TASK_LIST_RETRIEVAL_PL + TASK_LIST_RETRIEVAL_ZH:
|
| 771 |
+
examples["mteb_task"] = "Retrieval"
|
| 772 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_STS_NORM + TASK_LIST_STS_PL + TASK_LIST_STS_ZH:
|
| 773 |
+
examples["mteb_task"] = "STS"
|
| 774 |
+
elif examples["mteb_dataset_name"] in TASK_LIST_SUMMARIZATION:
|
| 775 |
+
examples["mteb_task"] = "Summarization"
|
| 776 |
+
elif examples["mteb_dataset_name"] in [x.split(" ")[0] for x in TASK_LIST_BITEXT_MINING + TASK_LIST_BITEXT_MINING_OTHER]:
|
| 777 |
+
examples["mteb_task"] = "BitextMining"
|
| 778 |
+
else:
|
| 779 |
+
print("WARNING: Task not found for dataset", examples["mteb_dataset_name"])
|
| 780 |
+
examples["mteb_task"] = "Unknown"
|
| 781 |
+
return examples
|
| 782 |
+
|
| 783 |
+
for model in EXTERNAL_MODELS:
|
| 784 |
+
ds = load_dataset("mteb/results", model)
|
| 785 |
+
# For local debugging:
|
| 786 |
+
#, download_mode='force_redownload', verification_mode="no_checks")
|
| 787 |
+
ds = ds.map(add_lang)
|
| 788 |
+
ds = ds.map(add_task)
|
| 789 |
+
base_dict = {"Model": make_clickable_model(model, link=EXTERNAL_MODEL_TO_LINK.get(model, "https://huggingface.co/spaces/mteb/leaderboard"))}
|
| 790 |
+
# For now only one metric per task - Could add more metrics lateron
|
| 791 |
+
for task, metric in TASK_TO_METRIC.items():
|
| 792 |
+
ds_dict = ds.filter(lambda x: (x["mteb_task"] == task) and (x["metric"] == metric))["test"].to_dict()
|
| 793 |
+
ds_dict = {k: round(v, 2) for k, v in zip(ds_dict["mteb_dataset_name_with_lang"], ds_dict["score"])}
|
| 794 |
+
EXTERNAL_MODEL_RESULTS[model][task][metric].append({**base_dict, **ds_dict})
|
| 795 |
+
|
| 796 |
+
def get_dim_seq_size(model):
|
| 797 |
+
filenames = [sib.rfilename for sib in model.siblings]
|
| 798 |
+
dim, seq, size = "", "", ""
|
| 799 |
+
if "1_Pooling/config.json" in filenames:
|
| 800 |
+
st_config_path = hf_hub_download(model.modelId, filename="1_Pooling/config.json")
|
| 801 |
+
dim = json.load(open(st_config_path)).get("word_embedding_dimension", "")
|
| 802 |
+
elif "2_Pooling/config.json" in filenames:
|
| 803 |
+
st_config_path = hf_hub_download(model.modelId, filename="2_Pooling/config.json")
|
| 804 |
+
dim = json.load(open(st_config_path)).get("word_embedding_dimension", "")
|
| 805 |
+
if "config.json" in filenames:
|
| 806 |
+
config_path = hf_hub_download(model.modelId, filename="config.json")
|
| 807 |
+
config = json.load(open(config_path))
|
| 808 |
+
if not dim:
|
| 809 |
+
dim = config.get("hidden_dim", config.get("hidden_size", config.get("d_model", "")))
|
| 810 |
+
seq = config.get("n_positions", config.get("max_position_embeddings", config.get("n_ctx", config.get("seq_length", ""))))
|
| 811 |
+
# Get model file size without downloading
|
| 812 |
+
if "pytorch_model.bin" in filenames:
|
| 813 |
+
url = hf_hub_url(model.modelId, filename="pytorch_model.bin")
|
| 814 |
+
meta = get_hf_file_metadata(url)
|
| 815 |
+
size = round(meta.size / 1e9, 2)
|
| 816 |
+
elif "pytorch_model.bin.index.json" in filenames:
|
| 817 |
+
index_path = hf_hub_download(model.modelId, filename="pytorch_model.bin.index.json")
|
| 818 |
+
"""
|
| 819 |
+
{
|
| 820 |
+
"metadata": {
|
| 821 |
+
"total_size": 28272820224
|
| 822 |
+
},....
|
| 823 |
+
"""
|
| 824 |
+
size = json.load(open(index_path))
|
| 825 |
+
if ("metadata" in size) and ("total_size" in size["metadata"]):
|
| 826 |
+
size = round(size["metadata"]["total_size"] / 1e9, 2)
|
| 827 |
+
elif "model.safetensors" in filenames:
|
| 828 |
+
url = hf_hub_url(model.modelId, filename="model.safetensors")
|
| 829 |
+
meta = get_hf_file_metadata(url)
|
| 830 |
+
size = round(meta.size / 1e9, 2)
|
| 831 |
+
return dim, seq, size
|
| 832 |
+
|
| 833 |
+
def make_datasets_clickable(df):
|
| 834 |
+
"""Does not work"""
|
| 835 |
+
if "BornholmBitextMining" in df.columns:
|
| 836 |
+
link = "https://huggingface.co/datasets/strombergnlp/bornholmsk_parallel"
|
| 837 |
+
df = df.rename(
|
| 838 |
+
columns={f'BornholmBitextMining': '<a target="_blank" style="text-decoration: underline" href="{link}">BornholmBitextMining</a>',})
|
| 839 |
+
return df
|
| 840 |
+
|
| 841 |
+
def add_rank(df):
|
| 842 |
+
cols_to_rank = [col for col in df.columns if col not in ["Model", "Model Size (GB)", "Embedding Dimensions", "Sequence Length"]]
|
| 843 |
+
if len(cols_to_rank) == 1:
|
| 844 |
+
df.sort_values(cols_to_rank[0], ascending=False, inplace=True)
|
| 845 |
+
else:
|
| 846 |
+
df.insert(1, "Average", df[cols_to_rank].mean(axis=1, skipna=False))
|
| 847 |
+
df.sort_values("Average", ascending=False, inplace=True)
|
| 848 |
+
df.insert(0, "Rank", list(range(1, len(df) + 1)))
|
| 849 |
+
df = df.round(2)
|
| 850 |
+
# Fill NaN after averaging
|
| 851 |
+
df.fillna("", inplace=True)
|
| 852 |
+
return df
|
| 853 |
+
|
| 854 |
+
def get_mteb_data(tasks=["Clustering"], langs=[], datasets=[], fillna=True, add_emb_dim=False, task_to_metric=TASK_TO_METRIC, rank=True):
|
| 855 |
+
api = HfApi()
|
| 856 |
+
models = api.list_models(filter="mteb")
|
| 857 |
+
# Initialize list to models that we cannot fetch metadata from
|
| 858 |
+
df_list = []
|
| 859 |
+
for model in EXTERNAL_MODEL_RESULTS:
|
| 860 |
+
results_list = [res for task in tasks for res in EXTERNAL_MODEL_RESULTS[model][task][task_to_metric[task]]]
|
| 861 |
+
if len(datasets) > 0:
|
| 862 |
+
res = {k: v for d in results_list for k, v in d.items() if (k == "Model") or any([x in k for x in datasets])}
|
| 863 |
+
elif langs:
|
| 864 |
+
# Would be cleaner to rely on an extra language column instead
|
| 865 |
+
langs_format = [f"({lang})" for lang in langs]
|
| 866 |
+
res = {k: v for d in results_list for k, v in d.items() if any([k.split(" ")[-1] in (k, x) for x in langs_format])}
|
| 867 |
+
else:
|
| 868 |
+
res = {k: v for d in results_list for k, v in d.items()}
|
| 869 |
+
# Model & at least one result
|
| 870 |
+
if len(res) > 1:
|
| 871 |
+
if add_emb_dim:
|
| 872 |
+
res["Model Size (GB)"] = EXTERNAL_MODEL_TO_SIZE.get(model, "")
|
| 873 |
+
res["Embedding Dimensions"] = EXTERNAL_MODEL_TO_DIM.get(model, "")
|
| 874 |
+
res["Sequence Length"] = EXTERNAL_MODEL_TO_SEQLEN.get(model, "")
|
| 875 |
+
df_list.append(res)
|
| 876 |
+
|
| 877 |
+
for model in models:
|
| 878 |
+
if model.modelId in MODELS_TO_SKIP: continue
|
| 879 |
+
print("MODEL", model)
|
| 880 |
+
readme_path = hf_hub_download(model.modelId, filename="README.md")
|
| 881 |
+
meta = metadata_load(readme_path)
|
| 882 |
+
# meta['model-index'][0]["results"] is list of elements like:
|
| 883 |
+
# {
|
| 884 |
+
# "task": {"type": "Classification"},
|
| 885 |
+
# "dataset": {
|
| 886 |
+
# "type": "mteb/amazon_massive_intent",
|
| 887 |
+
# "name": "MTEB MassiveIntentClassification (nb)",
|
| 888 |
+
# "config": "nb",
|
| 889 |
+
# "split": "test",
|
| 890 |
+
# },
|
| 891 |
+
# "metrics": [
|
| 892 |
+
# {"type": "accuracy", "value": 39.81506388702084},
|
| 893 |
+
# {"type": "f1", "value": 38.809586587791664},
|
| 894 |
+
# ],
|
| 895 |
+
# },
|
| 896 |
+
# Use "get" instead of dict indexing to skip incompat metadata instead of erroring out
|
| 897 |
+
if len(datasets) > 0:
|
| 898 |
+
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks) and any([x in sub_res.get("dataset", {}).get("name", "") for x in datasets])]
|
| 899 |
+
elif langs:
|
| 900 |
+
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks) and (sub_res.get("dataset", {}).get("config", "default") in ("default", *langs))]
|
| 901 |
+
else:
|
| 902 |
+
task_results = [sub_res for sub_res in meta["model-index"][0]["results"] if (sub_res.get("task", {}).get("type", "") in tasks)]
|
| 903 |
+
out = [{res["dataset"]["name"].replace("MTEB ", ""): [round(score["value"], 2) for score in res["metrics"] if score["type"] == task_to_metric.get(res["task"]["type"])][0]} for res in task_results]
|
| 904 |
+
out = {k: v for d in out for k, v in d.items()}
|
| 905 |
+
out["Model"] = make_clickable_model(model.modelId)
|
| 906 |
+
# Model & at least one result
|
| 907 |
+
if len(out) > 1:
|
| 908 |
+
if add_emb_dim:
|
| 909 |
+
out["Embedding Dimensions"], out["Sequence Length"], out["Model Size (GB)"] = get_dim_seq_size(model)
|
| 910 |
+
df_list.append(out)
|
| 911 |
+
df = pd.DataFrame(df_list)
|
| 912 |
+
# If there are any models that are the same, merge them
|
| 913 |
+
# E.g. if out["Model"] has the same value in two places, merge & take whichever one is not NaN else just take the first one
|
| 914 |
+
df = df.groupby("Model", as_index=False).first()
|
| 915 |
+
# Put 'Model' column first
|
| 916 |
+
cols = sorted(list(df.columns))
|
| 917 |
+
cols.insert(0, cols.pop(cols.index("Model")))
|
| 918 |
+
df = df[cols]
|
| 919 |
+
if rank:
|
| 920 |
+
df = add_rank(df)
|
| 921 |
+
if fillna:
|
| 922 |
+
df.fillna("", inplace=True)
|
| 923 |
+
return df
|
| 924 |
+
|
| 925 |
+
def get_mteb_average():
|
| 926 |
+
global DATA_OVERALL, DATA_CLASSIFICATION_EN, DATA_CLUSTERING, DATA_PAIR_CLASSIFICATION, DATA_RERANKING, DATA_RETRIEVAL, DATA_STS_EN, DATA_SUMMARIZATION
|
| 927 |
+
|
| 928 |
+
DATA_OVERALL = get_mteb_data(
|
| 929 |
+
tasks=[
|
| 930 |
+
"Classification",
|
| 931 |
+
"Clustering",
|
| 932 |
+
"PairClassification",
|
| 933 |
+
"Reranking",
|
| 934 |
+
"Retrieval",
|
| 935 |
+
"STS",
|
| 936 |
+
"Summarization",
|
| 937 |
+
],
|
| 938 |
+
datasets=TASK_LIST_CLASSIFICATION + TASK_LIST_CLUSTERING + TASK_LIST_PAIR_CLASSIFICATION + TASK_LIST_RERANKING + TASK_LIST_RETRIEVAL + TASK_LIST_STS + TASK_LIST_SUMMARIZATION,
|
| 939 |
+
fillna=False,
|
| 940 |
+
add_emb_dim=True,
|
| 941 |
+
rank=False,
|
| 942 |
+
)
|
| 943 |
+
# Debugging:
|
| 944 |
+
# DATA_OVERALL.to_csv("overall.csv")
|
| 945 |
+
|
| 946 |
+
DATA_OVERALL.insert(1, f"Average ({len(TASK_LIST_EN)} datasets)", DATA_OVERALL[TASK_LIST_EN].mean(axis=1, skipna=False))
|
| 947 |
+
DATA_OVERALL.insert(2, f"Classification Average ({len(TASK_LIST_CLASSIFICATION)} datasets)", DATA_OVERALL[TASK_LIST_CLASSIFICATION].mean(axis=1, skipna=False))
|
| 948 |
+
DATA_OVERALL.insert(3, f"Clustering Average ({len(TASK_LIST_CLUSTERING)} datasets)", DATA_OVERALL[TASK_LIST_CLUSTERING].mean(axis=1, skipna=False))
|
| 949 |
+
DATA_OVERALL.insert(4, f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION)} datasets)", DATA_OVERALL[TASK_LIST_PAIR_CLASSIFICATION].mean(axis=1, skipna=False))
|
| 950 |
+
DATA_OVERALL.insert(5, f"Reranking Average ({len(TASK_LIST_RERANKING)} datasets)", DATA_OVERALL[TASK_LIST_RERANKING].mean(axis=1, skipna=False))
|
| 951 |
+
DATA_OVERALL.insert(6, f"Retrieval Average ({len(TASK_LIST_RETRIEVAL)} datasets)", DATA_OVERALL[TASK_LIST_RETRIEVAL].mean(axis=1, skipna=False))
|
| 952 |
+
DATA_OVERALL.insert(7, f"STS Average ({len(TASK_LIST_STS)} datasets)", DATA_OVERALL[TASK_LIST_STS].mean(axis=1, skipna=False))
|
| 953 |
+
DATA_OVERALL.insert(8, f"Summarization Average ({len(TASK_LIST_SUMMARIZATION)} dataset)", DATA_OVERALL[TASK_LIST_SUMMARIZATION].mean(axis=1, skipna=False))
|
| 954 |
+
DATA_OVERALL.sort_values(f"Average ({len(TASK_LIST_EN)} datasets)", ascending=False, inplace=True)
|
| 955 |
+
# Start ranking from 1
|
| 956 |
+
DATA_OVERALL.insert(0, "Rank", list(range(1, len(DATA_OVERALL) + 1)))
|
| 957 |
+
|
| 958 |
+
DATA_OVERALL = DATA_OVERALL.round(2)
|
| 959 |
+
|
| 960 |
+
DATA_CLASSIFICATION_EN = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_CLASSIFICATION])
|
| 961 |
+
# Only keep rows with at least one score in addition to the "Model" & rank column
|
| 962 |
+
DATA_CLASSIFICATION_EN = DATA_CLASSIFICATION_EN[DATA_CLASSIFICATION_EN.iloc[:, 2:].ne("").any(axis=1)]
|
| 963 |
+
|
| 964 |
+
DATA_CLUSTERING = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_CLUSTERING])
|
| 965 |
+
DATA_CLUSTERING = DATA_CLUSTERING[DATA_CLUSTERING.iloc[:, 2:].ne("").any(axis=1)]
|
| 966 |
+
|
| 967 |
+
DATA_PAIR_CLASSIFICATION = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_PAIR_CLASSIFICATION])
|
| 968 |
+
DATA_PAIR_CLASSIFICATION = DATA_PAIR_CLASSIFICATION[DATA_PAIR_CLASSIFICATION.iloc[:, 2:].ne("").any(axis=1)]
|
| 969 |
+
|
| 970 |
+
DATA_RERANKING = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_RERANKING])
|
| 971 |
+
DATA_RERANKING = DATA_RERANKING[DATA_RERANKING.iloc[:, 2:].ne("").any(axis=1)]
|
| 972 |
+
|
| 973 |
+
DATA_RETRIEVAL = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_RETRIEVAL])
|
| 974 |
+
DATA_RETRIEVAL = DATA_RETRIEVAL[DATA_RETRIEVAL.iloc[:, 2:].ne("").any(axis=1)]
|
| 975 |
+
|
| 976 |
+
DATA_STS_EN = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_STS])
|
| 977 |
+
DATA_STS_EN = DATA_STS_EN[DATA_STS_EN.iloc[:, 2:].ne("").any(axis=1)]
|
| 978 |
+
|
| 979 |
+
DATA_SUMMARIZATION = add_rank(DATA_OVERALL[["Model"] + TASK_LIST_SUMMARIZATION])
|
| 980 |
+
DATA_SUMMARIZATION = DATA_SUMMARIZATION[DATA_SUMMARIZATION.iloc[:, 1:].ne("").any(axis=1)]
|
| 981 |
+
|
| 982 |
+
# Fill NaN after averaging
|
| 983 |
+
DATA_OVERALL.fillna("", inplace=True)
|
| 984 |
+
|
| 985 |
+
DATA_OVERALL = DATA_OVERALL[["Rank", "Model", "Model Size (GB)", "Embedding Dimensions", "Sequence Length", f"Average ({len(TASK_LIST_EN)} datasets)", f"Classification Average ({len(TASK_LIST_CLASSIFICATION)} datasets)", f"Clustering Average ({len(TASK_LIST_CLUSTERING)} datasets)", f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION)} datasets)", f"Reranking Average ({len(TASK_LIST_RERANKING)} datasets)", f"Retrieval Average ({len(TASK_LIST_RETRIEVAL)} datasets)", f"STS Average ({len(TASK_LIST_STS)} datasets)", f"Summarization Average ({len(TASK_LIST_SUMMARIZATION)} dataset)"]]
|
| 986 |
+
DATA_OVERALL = DATA_OVERALL[DATA_OVERALL.iloc[:, 5:].ne("").any(axis=1)]
|
| 987 |
+
|
| 988 |
+
return DATA_OVERALL
|
| 989 |
+
|
| 990 |
+
def get_mteb_average_zh():
|
| 991 |
+
global DATA_OVERALL_ZH, DATA_CLASSIFICATION_ZH, DATA_CLUSTERING_ZH, DATA_PAIR_CLASSIFICATION_ZH, DATA_RERANKING_ZH, DATA_RETRIEVAL_ZH, DATA_STS_ZH
|
| 992 |
+
DATA_OVERALL_ZH = get_mteb_data(
|
| 993 |
+
tasks=[
|
| 994 |
+
"Classification",
|
| 995 |
+
"Clustering",
|
| 996 |
+
"PairClassification",
|
| 997 |
+
"Reranking",
|
| 998 |
+
"Retrieval",
|
| 999 |
+
"STS",
|
| 1000 |
+
],
|
| 1001 |
+
datasets=TASK_LIST_CLASSIFICATION_ZH + TASK_LIST_CLUSTERING_ZH + TASK_LIST_PAIR_CLASSIFICATION_ZH + TASK_LIST_RERANKING_ZH + TASK_LIST_RETRIEVAL_ZH + TASK_LIST_STS_ZH,
|
| 1002 |
+
fillna=False,
|
| 1003 |
+
add_emb_dim=True,
|
| 1004 |
+
rank=False,
|
| 1005 |
+
)
|
| 1006 |
+
# Debugging:
|
| 1007 |
+
# DATA_OVERALL_ZH.to_csv("overall.csv")
|
| 1008 |
+
|
| 1009 |
+
DATA_OVERALL_ZH.insert(1, f"Average ({len(TASK_LIST_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_ZH].mean(axis=1, skipna=False))
|
| 1010 |
+
DATA_OVERALL_ZH.insert(2, f"Classification Average ({len(TASK_LIST_CLASSIFICATION_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_CLASSIFICATION_ZH].mean(axis=1, skipna=False))
|
| 1011 |
+
DATA_OVERALL_ZH.insert(3, f"Clustering Average ({len(TASK_LIST_CLUSTERING_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_CLUSTERING_ZH].mean(axis=1, skipna=False))
|
| 1012 |
+
DATA_OVERALL_ZH.insert(4, f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_PAIR_CLASSIFICATION_ZH].mean(axis=1, skipna=False))
|
| 1013 |
+
DATA_OVERALL_ZH.insert(5, f"Reranking Average ({len(TASK_LIST_RERANKING_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_RERANKING_ZH].mean(axis=1, skipna=False))
|
| 1014 |
+
DATA_OVERALL_ZH.insert(6, f"Retrieval Average ({len(TASK_LIST_RETRIEVAL_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_RETRIEVAL_ZH].mean(axis=1, skipna=False))
|
| 1015 |
+
DATA_OVERALL_ZH.insert(7, f"STS Average ({len(TASK_LIST_STS_ZH)} datasets)", DATA_OVERALL_ZH[TASK_LIST_STS_ZH].mean(axis=1, skipna=False))
|
| 1016 |
+
DATA_OVERALL_ZH.sort_values(f"Average ({len(TASK_LIST_ZH)} datasets)", ascending=False, inplace=True)
|
| 1017 |
+
# Start ranking from 1
|
| 1018 |
+
DATA_OVERALL_ZH.insert(0, "Rank", list(range(1, len(DATA_OVERALL_ZH) + 1)))
|
| 1019 |
+
|
| 1020 |
+
DATA_OVERALL_ZH = DATA_OVERALL_ZH.round(2)
|
| 1021 |
+
|
| 1022 |
+
DATA_CLASSIFICATION_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_CLASSIFICATION_ZH])
|
| 1023 |
+
# Only keep rows with at least one score in addition to the "Model" & rank column
|
| 1024 |
+
DATA_CLASSIFICATION_ZH = DATA_CLASSIFICATION_ZH[DATA_CLASSIFICATION_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1025 |
+
|
| 1026 |
+
DATA_CLUSTERING_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_CLUSTERING_ZH])
|
| 1027 |
+
DATA_CLUSTERING_ZH = DATA_CLUSTERING_ZH[DATA_CLUSTERING_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1028 |
+
|
| 1029 |
+
DATA_PAIR_CLASSIFICATION_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_PAIR_CLASSIFICATION_ZH])
|
| 1030 |
+
DATA_PAIR_CLASSIFICATION_ZH = DATA_PAIR_CLASSIFICATION_ZH[DATA_PAIR_CLASSIFICATION_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1031 |
+
|
| 1032 |
+
DATA_RERANKING_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_RERANKING_ZH])
|
| 1033 |
+
DATA_RERANKING_ZH = DATA_RERANKING_ZH[DATA_RERANKING_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1034 |
+
|
| 1035 |
+
DATA_RETRIEVAL_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_RETRIEVAL_ZH])
|
| 1036 |
+
DATA_RETRIEVAL_ZH = DATA_RETRIEVAL_ZH[DATA_RETRIEVAL_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1037 |
+
|
| 1038 |
+
DATA_STS_ZH = add_rank(DATA_OVERALL_ZH[["Model"] + TASK_LIST_STS_ZH])
|
| 1039 |
+
DATA_STS_ZH = DATA_STS_ZH[DATA_STS_ZH.iloc[:, 2:].ne("").any(axis=1)]
|
| 1040 |
+
|
| 1041 |
+
# Fill NaN after averaging
|
| 1042 |
+
DATA_OVERALL_ZH.fillna("", inplace=True)
|
| 1043 |
+
|
| 1044 |
+
DATA_OVERALL_ZH = DATA_OVERALL_ZH[["Rank", "Model", "Model Size (GB)", "Embedding Dimensions", "Sequence Length", f"Average ({len(TASK_LIST_ZH)} datasets)", f"Classification Average ({len(TASK_LIST_CLASSIFICATION_ZH)} datasets)", f"Clustering Average ({len(TASK_LIST_CLUSTERING_ZH)} datasets)", f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION_ZH)} datasets)", f"Reranking Average ({len(TASK_LIST_RERANKING_ZH)} datasets)", f"Retrieval Average ({len(TASK_LIST_RETRIEVAL_ZH)} datasets)", f"STS Average ({len(TASK_LIST_STS_ZH)} datasets)"]]
|
| 1045 |
+
DATA_OVERALL_ZH = DATA_OVERALL_ZH[DATA_OVERALL_ZH.iloc[:, 5:].ne("").any(axis=1)]
|
| 1046 |
+
|
| 1047 |
+
return DATA_OVERALL_ZH
|
| 1048 |
+
|
| 1049 |
+
def get_mteb_average_pl():
|
| 1050 |
+
global DATA_OVERALL_PL, DATA_CLASSIFICATION_PL, DATA_CLUSTERING_PL, DATA_PAIR_CLASSIFICATION_PL, DATA_RETRIEVAL_PL, DATA_STS_PL
|
| 1051 |
+
DATA_OVERALL_PL = get_mteb_data(
|
| 1052 |
+
tasks=[
|
| 1053 |
+
"Classification",
|
| 1054 |
+
"Clustering",
|
| 1055 |
+
"PairClassification",
|
| 1056 |
+
"Retrieval",
|
| 1057 |
+
"STS",
|
| 1058 |
+
],
|
| 1059 |
+
datasets=TASK_LIST_CLASSIFICATION_PL + TASK_LIST_CLUSTERING_PL + TASK_LIST_PAIR_CLASSIFICATION_PL + TASK_LIST_RETRIEVAL_PL + TASK_LIST_STS_PL,
|
| 1060 |
+
fillna=False,
|
| 1061 |
+
add_emb_dim=True,
|
| 1062 |
+
rank=False,
|
| 1063 |
+
)
|
| 1064 |
+
# Debugging:
|
| 1065 |
+
# DATA_OVERALL_PL.to_csv("overall.csv")
|
| 1066 |
+
|
| 1067 |
+
DATA_OVERALL_PL.insert(1, f"Average ({len(TASK_LIST_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_PL].mean(axis=1, skipna=False))
|
| 1068 |
+
DATA_OVERALL_PL.insert(2, f"Classification Average ({len(TASK_LIST_CLASSIFICATION_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_CLASSIFICATION_PL].mean(axis=1, skipna=False))
|
| 1069 |
+
DATA_OVERALL_PL.insert(3, f"Clustering Average ({len(TASK_LIST_CLUSTERING_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_CLUSTERING_PL].mean(axis=1, skipna=False))
|
| 1070 |
+
DATA_OVERALL_PL.insert(4, f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_PAIR_CLASSIFICATION_PL].mean(axis=1, skipna=False))
|
| 1071 |
+
DATA_OVERALL_PL.insert(5, f"Retrieval Average ({len(TASK_LIST_RETRIEVAL_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_RETRIEVAL_PL].mean(axis=1, skipna=False))
|
| 1072 |
+
DATA_OVERALL_PL.insert(6, f"STS Average ({len(TASK_LIST_STS_PL)} datasets)", DATA_OVERALL_PL[TASK_LIST_STS_PL].mean(axis=1, skipna=False))
|
| 1073 |
+
DATA_OVERALL_PL.sort_values(f"Average ({len(TASK_LIST_PL)} datasets)", ascending=False, inplace=True)
|
| 1074 |
+
# Start ranking from 1
|
| 1075 |
+
DATA_OVERALL_PL.insert(0, "Rank", list(range(1, len(DATA_OVERALL_PL) + 1)))
|
| 1076 |
+
|
| 1077 |
+
DATA_OVERALL_PL = DATA_OVERALL_PL.round(2)
|
| 1078 |
+
|
| 1079 |
+
DATA_CLASSIFICATION_PL = add_rank(DATA_OVERALL_PL[["Model"] + TASK_LIST_CLASSIFICATION_PL])
|
| 1080 |
+
# Only keep rows with at least one score in addition to the "Model" & rank column
|
| 1081 |
+
DATA_CLASSIFICATION_PL = DATA_CLASSIFICATION_PL[DATA_CLASSIFICATION_PL.iloc[:, 2:].ne("").any(axis=1)]
|
| 1082 |
+
|
| 1083 |
+
DATA_CLUSTERING_PL = add_rank(DATA_OVERALL_PL[["Model"] + TASK_LIST_CLUSTERING_PL])
|
| 1084 |
+
DATA_CLUSTERING_PL = DATA_CLUSTERING_PL[DATA_CLUSTERING_PL.iloc[:, 2:].ne("").any(axis=1)]
|
| 1085 |
+
|
| 1086 |
+
DATA_PAIR_CLASSIFICATION_PL = add_rank(DATA_OVERALL_PL[["Model"] + TASK_LIST_PAIR_CLASSIFICATION_PL])
|
| 1087 |
+
DATA_PAIR_CLASSIFICATION_PL = DATA_PAIR_CLASSIFICATION_PL[DATA_PAIR_CLASSIFICATION_PL.iloc[:, 2:].ne("").any(axis=1)]
|
| 1088 |
+
|
| 1089 |
+
DATA_RETRIEVAL_PL = add_rank(DATA_OVERALL_PL[["Model"] + TASK_LIST_RETRIEVAL_PL])
|
| 1090 |
+
DATA_RETRIEVAL_PL = DATA_RETRIEVAL_PL[DATA_RETRIEVAL_PL.iloc[:, 2:].ne("").any(axis=1)]
|
| 1091 |
+
|
| 1092 |
+
DATA_STS_PL = add_rank(DATA_OVERALL_PL[["Model"] + TASK_LIST_STS_PL])
|
| 1093 |
+
DATA_STS_PL = DATA_STS_PL[DATA_STS_PL.iloc[:, 2:].ne("").any(axis=1)]
|
| 1094 |
+
|
| 1095 |
+
# Fill NaN after averaging
|
| 1096 |
+
DATA_OVERALL_PL.fillna("", inplace=True)
|
| 1097 |
+
|
| 1098 |
+
DATA_OVERALL_PL = DATA_OVERALL_PL[["Rank", "Model", "Model Size (GB)", "Embedding Dimensions", "Sequence Length", f"Average ({len(TASK_LIST_PL)} datasets)", f"Classification Average ({len(TASK_LIST_CLASSIFICATION_PL)} datasets)", f"Clustering Average ({len(TASK_LIST_CLUSTERING_PL)} datasets)", f"Pair Classification Average ({len(TASK_LIST_PAIR_CLASSIFICATION_PL)} datasets)", f"Retrieval Average ({len(TASK_LIST_RETRIEVAL_PL)} datasets)", f"STS Average ({len(TASK_LIST_STS_PL)} datasets)"]]
|
| 1099 |
+
DATA_OVERALL_PL = DATA_OVERALL_PL[DATA_OVERALL_PL.iloc[:, 5:].ne("").any(axis=1)]
|
| 1100 |
+
|
| 1101 |
+
return DATA_OVERALL_PL
|
| 1102 |
+
|
| 1103 |
+
get_mteb_average()
|
| 1104 |
+
|
| 1105 |
+
|
| 1106 |
+
get_mteb_average_pl()
|
| 1107 |
+
get_mteb_average_zh()
|
| 1108 |
+
DATA_BITEXT_MINING = get_mteb_data(["BitextMining"], [], TASK_LIST_BITEXT_MINING)
|
| 1109 |
+
DATA_BITEXT_MINING_OTHER = get_mteb_data(["BitextMining"], [], TASK_LIST_BITEXT_MINING_OTHER)
|
| 1110 |
+
DATA_CLASSIFICATION_DA = get_mteb_data(["Classification"], [], TASK_LIST_CLASSIFICATION_DA)
|
| 1111 |
+
DATA_CLASSIFICATION_NB = get_mteb_data(["Classification"], [], TASK_LIST_CLASSIFICATION_NB)
|
| 1112 |
+
DATA_CLASSIFICATION_SV = get_mteb_data(["Classification"], [], TASK_LIST_CLASSIFICATION_SV)
|
| 1113 |
+
DATA_CLASSIFICATION_OTHER = get_mteb_data(["Classification"], [], TASK_LIST_CLASSIFICATION_OTHER)
|
| 1114 |
+
DATA_CLUSTERING_DE = get_mteb_data(["Clustering"], [], TASK_LIST_CLUSTERING_DE)
|
| 1115 |
+
DATA_STS_OTHER = get_mteb_data(["STS"], [], TASK_LIST_STS_OTHER)
|
| 1116 |
+
|
| 1117 |
+
# Exact, add all non-nan integer values for every dataset
|
| 1118 |
+
NUM_SCORES = 0
|
| 1119 |
+
DATASETS = []
|
| 1120 |
+
MODELS = []
|
| 1121 |
+
# LANGUAGES = []
|
| 1122 |
+
for d in [
|
| 1123 |
+
DATA_BITEXT_MINING,
|
| 1124 |
+
DATA_BITEXT_MINING_OTHER,
|
| 1125 |
+
DATA_CLASSIFICATION_EN,
|
| 1126 |
+
DATA_CLASSIFICATION_DA,
|
| 1127 |
+
DATA_CLASSIFICATION_NB,
|
| 1128 |
+
DATA_CLASSIFICATION_PL,
|
| 1129 |
+
DATA_CLASSIFICATION_SV,
|
| 1130 |
+
DATA_CLASSIFICATION_ZH,
|
| 1131 |
+
DATA_CLASSIFICATION_OTHER,
|
| 1132 |
+
DATA_CLUSTERING,
|
| 1133 |
+
DATA_CLUSTERING_DE,
|
| 1134 |
+
DATA_CLUSTERING_PL,
|
| 1135 |
+
DATA_CLUSTERING_ZH,
|
| 1136 |
+
DATA_PAIR_CLASSIFICATION,
|
| 1137 |
+
DATA_PAIR_CLASSIFICATION_PL,
|
| 1138 |
+
DATA_PAIR_CLASSIFICATION_ZH,
|
| 1139 |
+
DATA_RERANKING,
|
| 1140 |
+
DATA_RERANKING_ZH,
|
| 1141 |
+
DATA_RETRIEVAL,
|
| 1142 |
+
DATA_RETRIEVAL_PL,
|
| 1143 |
+
DATA_RETRIEVAL_ZH,
|
| 1144 |
+
DATA_STS_EN,
|
| 1145 |
+
DATA_STS_PL,
|
| 1146 |
+
DATA_STS_ZH,
|
| 1147 |
+
DATA_STS_OTHER,
|
| 1148 |
+
DATA_SUMMARIZATION,
|
| 1149 |
+
]:
|
| 1150 |
+
# NUM_SCORES += d.iloc[:, 1:].apply(lambda x: sum([1 for y in x if isinstance(y, float) and not np.isnan(y)]), axis=1).sum()
|
| 1151 |
+
cols_to_ignore = 3 if "Average" in d.columns else 2
|
| 1152 |
+
# Count number of scores including only non-nan floats & excluding the rank column
|
| 1153 |
+
NUM_SCORES += d.iloc[:, cols_to_ignore:].notna().sum().sum()
|
| 1154 |
+
# Exclude rank & model name column (first two); Do not count different language versions as different datasets
|
| 1155 |
+
DATASETS += [i.split(" ")[0] for i in d.columns[cols_to_ignore:]]
|
| 1156 |
+
# LANGUAGES += [i.split(" ")[-1] for i in d.columns[cols_to_ignore:]]
|
| 1157 |
+
MODELS += d["Model"].tolist()
|
| 1158 |
+
|
| 1159 |
+
NUM_DATASETS = len(set(DATASETS))
|
| 1160 |
+
# NUM_LANGUAGES = len(set(LANGUAGES))
|
| 1161 |
+
NUM_MODELS = len(set(MODELS))
|
| 1162 |
+
|
| 1163 |
+
|
| 1164 |
+
|
| 1165 |
+
|
| 1166 |
+
|
| 1167 |
+
block = gr.Blocks()
|
| 1168 |
+
with block:
|
| 1169 |
+
gr.Markdown(f"""
|
| 1170 |
+
SeaEval Leaderboard. To submit, refer to the <a href="https://seaeval.github.io/" target="_blank" style="text-decoration: underline">SeaEval Website</a> Refer to the [SeaEval paper](https://arxiv.org/abs/2309.04766) for details on metrics, tasks and models.
|
| 1171 |
+
|
| 1172 |
+
- **Total Datasets**: 31
|
| 1173 |
+
- **Total Languages**: 8
|
| 1174 |
+
- **Total Models**: 5
|
| 1175 |
+
""")
|
| 1176 |
+
with gr.Tabs():
|
| 1177 |
+
|
| 1178 |
+
|
| 1179 |
+
# dataset 1: cross-mmlu
|
| 1180 |
+
with gr.TabItem("Cross-MMLU"):
|
| 1181 |
+
with gr.Row():
|
| 1182 |
+
gr.Markdown("""
|
| 1183 |
+
**Overall Cross-MMLU Leaderboard** 🔮
|
| 1184 |
+
|
| 1185 |
+
- **Metric:** Cross-Lingual Consistency, Accuracy, AC3
|
| 1186 |
+
- **Languages:** English, Chinese, Malay, Indonesian, Spanish, Vietnamese, Filipino
|
| 1187 |
+
""")
|
| 1188 |
+
|
| 1189 |
+
with gr.TabItem("Zero-Shot"):
|
| 1190 |
+
|
| 1191 |
+
|
| 1192 |
+
with gr.TabItem("Overall"):
|
| 1193 |
+
|
| 1194 |
+
with gr.Row():
|
| 1195 |
+
data_bitext_mining = gr.components.Dataframe(
|
| 1196 |
+
DATA_BITEXT_MINING,
|
| 1197 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_BITEXT_MINING.columns),
|
| 1198 |
+
type="pandas",
|
| 1199 |
+
)
|
| 1200 |
+
with gr.Row():
|
| 1201 |
+
data_run_bitext_mining = gr.Button("Refresh")
|
| 1202 |
+
data_run_bitext_mining.click(
|
| 1203 |
+
partial(get_mteb_data, tasks=["BitextMining"], datasets=TASK_LIST_BITEXT_MINING),
|
| 1204 |
+
outputs=data_bitext_mining,
|
| 1205 |
+
)
|
| 1206 |
+
|
| 1207 |
+
with gr.TabItem("Detailed Consistency"):
|
| 1208 |
+
|
| 1209 |
+
with gr.Row():
|
| 1210 |
+
data_bitext_mining = gr.components.Dataframe(
|
| 1211 |
+
DATA_BITEXT_MINING,
|
| 1212 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_BITEXT_MINING.columns),
|
| 1213 |
+
type="pandas",
|
| 1214 |
+
)
|
| 1215 |
+
with gr.Row():
|
| 1216 |
+
data_run_bitext_mining = gr.Button("Refresh")
|
| 1217 |
+
data_run_bitext_mining.click(
|
| 1218 |
+
partial(get_mteb_data, tasks=["BitextMining"], datasets=TASK_LIST_BITEXT_MINING),
|
| 1219 |
+
outputs=data_bitext_mining,
|
| 1220 |
+
)
|
| 1221 |
+
|
| 1222 |
+
with gr.TabItem("Five-Shot"):
|
| 1223 |
+
|
| 1224 |
+
with gr.TabItem("Overall"):
|
| 1225 |
+
|
| 1226 |
+
with gr.Row():
|
| 1227 |
+
data_bitext_mining = gr.components.Dataframe(
|
| 1228 |
+
DATA_BITEXT_MINING,
|
| 1229 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_BITEXT_MINING.columns),
|
| 1230 |
+
type="pandas",
|
| 1231 |
+
)
|
| 1232 |
+
with gr.Row():
|
| 1233 |
+
data_run_bitext_mining = gr.Button("Refresh")
|
| 1234 |
+
data_run_bitext_mining.click(
|
| 1235 |
+
partial(get_mteb_data, tasks=["BitextMining"], datasets=TASK_LIST_BITEXT_MINING),
|
| 1236 |
+
outputs=data_bitext_mining,
|
| 1237 |
+
)
|
| 1238 |
+
|
| 1239 |
+
with gr.TabItem("Detailed Consistency"):
|
| 1240 |
+
|
| 1241 |
+
with gr.Row():
|
| 1242 |
+
data_bitext_mining = gr.components.Dataframe(
|
| 1243 |
+
DATA_BITEXT_MINING,
|
| 1244 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_BITEXT_MINING.columns),
|
| 1245 |
+
type="pandas",
|
| 1246 |
+
)
|
| 1247 |
+
with gr.Row():
|
| 1248 |
+
data_run_bitext_mining = gr.Button("Refresh")
|
| 1249 |
+
data_run_bitext_mining.click(
|
| 1250 |
+
partial(get_mteb_data, tasks=["BitextMining"], datasets=TASK_LIST_BITEXT_MINING),
|
| 1251 |
+
outputs=data_bitext_mining,
|
| 1252 |
+
)
|
| 1253 |
+
|
| 1254 |
+
|
| 1255 |
+
|
| 1256 |
+
with gr.TabItem("Classification"):
|
| 1257 |
+
with gr.TabItem("English"):
|
| 1258 |
+
with gr.Row():
|
| 1259 |
+
gr.Markdown("""
|
| 1260 |
+
**Classification English Leaderboard** ❤️
|
| 1261 |
+
|
| 1262 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1263 |
+
- **Languages:** English
|
| 1264 |
+
""")
|
| 1265 |
+
with gr.Row():
|
| 1266 |
+
data_classification_en = gr.components.Dataframe(
|
| 1267 |
+
DATA_CLASSIFICATION_EN,
|
| 1268 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_EN.columns),
|
| 1269 |
+
type="pandas",
|
| 1270 |
+
)
|
| 1271 |
+
with gr.Row():
|
| 1272 |
+
data_run_classification_en = gr.Button("Refresh")
|
| 1273 |
+
data_run_classification_en.click(
|
| 1274 |
+
partial(get_mteb_data, tasks=["Classification"], langs=["en"]),
|
| 1275 |
+
outputs=data_classification_en,
|
| 1276 |
+
)
|
| 1277 |
+
with gr.TabItem("Chinese"):
|
| 1278 |
+
with gr.Row():
|
| 1279 |
+
gr.Markdown("""
|
| 1280 |
+
**Classification Chinese Leaderboard** 🧡🇨🇳
|
| 1281 |
+
|
| 1282 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1283 |
+
- **Languages:** Chinese
|
| 1284 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1285 |
+
""")
|
| 1286 |
+
with gr.Row():
|
| 1287 |
+
data_classification_zh = gr.components.Dataframe(
|
| 1288 |
+
DATA_CLASSIFICATION_ZH,
|
| 1289 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_ZH.columns),
|
| 1290 |
+
type="pandas",
|
| 1291 |
+
)
|
| 1292 |
+
with gr.Row():
|
| 1293 |
+
data_run_classification_zh = gr.Button("Refresh")
|
| 1294 |
+
data_run_classification_zh.click(
|
| 1295 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_ZH),
|
| 1296 |
+
outputs=data_classification_zh,
|
| 1297 |
+
)
|
| 1298 |
+
with gr.TabItem("Danish"):
|
| 1299 |
+
with gr.Row():
|
| 1300 |
+
gr.Markdown("""
|
| 1301 |
+
**Classification Danish Leaderboard** 🤍🇩🇰
|
| 1302 |
+
|
| 1303 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1304 |
+
- **Languages:** Danish
|
| 1305 |
+
- **Credits:** [Kenneth Enevoldsen](https://github.com/KennethEnevoldsen), [scandinavian-embedding-benchmark](https://kennethenevoldsen.github.io/scandinavian-embedding-benchmark/)
|
| 1306 |
+
""")
|
| 1307 |
+
with gr.Row():
|
| 1308 |
+
data_classification_da = gr.components.Dataframe(
|
| 1309 |
+
DATA_CLASSIFICATION_DA,
|
| 1310 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_DA.columns),
|
| 1311 |
+
type="pandas",
|
| 1312 |
+
)
|
| 1313 |
+
with gr.Row():
|
| 1314 |
+
data_run_classification_da = gr.Button("Refresh")
|
| 1315 |
+
data_run_classification_da.click(
|
| 1316 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_DA),
|
| 1317 |
+
outputs=data_run_classification_da,
|
| 1318 |
+
)
|
| 1319 |
+
with gr.TabItem("Norwegian"):
|
| 1320 |
+
with gr.Row():
|
| 1321 |
+
gr.Markdown("""
|
| 1322 |
+
**Classification Norwegian Leaderboard** 💙🇳🇴
|
| 1323 |
+
|
| 1324 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1325 |
+
- **Languages:** Norwegian Bokmål
|
| 1326 |
+
- **Credits:** [Kenneth Enevoldsen](https://github.com/KennethEnevoldsen), [scandinavian-embedding-benchmark](https://kennethenevoldsen.github.io/scandinavian-embedding-benchmark/)
|
| 1327 |
+
""")
|
| 1328 |
+
with gr.Row():
|
| 1329 |
+
data_classification_nb = gr.components.Dataframe(
|
| 1330 |
+
DATA_CLASSIFICATION_NB,
|
| 1331 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_NB.columns),
|
| 1332 |
+
type="pandas",
|
| 1333 |
+
)
|
| 1334 |
+
with gr.Row():
|
| 1335 |
+
data_run_classification_nb = gr.Button("Refresh")
|
| 1336 |
+
data_run_classification_nb.click(
|
| 1337 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_NB),
|
| 1338 |
+
outputs=data_classification_nb,
|
| 1339 |
+
)
|
| 1340 |
+
with gr.TabItem("Polish"):
|
| 1341 |
+
with gr.Row():
|
| 1342 |
+
gr.Markdown("""
|
| 1343 |
+
**Classification Polish Leaderboard** 🤍🇵🇱
|
| 1344 |
+
|
| 1345 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1346 |
+
- **Languages:** Polish
|
| 1347 |
+
- **Credits:** [Rafał Poświata](https://github.com/rafalposwiata)
|
| 1348 |
+
""")
|
| 1349 |
+
with gr.Row():
|
| 1350 |
+
data_classification_pl = gr.components.Dataframe(
|
| 1351 |
+
DATA_CLASSIFICATION_PL,
|
| 1352 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_PL.columns),
|
| 1353 |
+
type="pandas",
|
| 1354 |
+
)
|
| 1355 |
+
with gr.Row():
|
| 1356 |
+
data_run_classification_pl = gr.Button("Refresh")
|
| 1357 |
+
data_run_classification_pl.click(
|
| 1358 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_PL),
|
| 1359 |
+
outputs=data_classification_pl,
|
| 1360 |
+
)
|
| 1361 |
+
with gr.TabItem("Swedish"):
|
| 1362 |
+
with gr.Row():
|
| 1363 |
+
gr.Markdown("""
|
| 1364 |
+
**Classification Swedish Leaderboard** 💛🇸🇪
|
| 1365 |
+
|
| 1366 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1367 |
+
- **Languages:** Swedish
|
| 1368 |
+
- **Credits:** [Kenneth Enevoldsen](https://github.com/KennethEnevoldsen), [scandinavian-embedding-benchmark](https://kennethenevoldsen.github.io/scandinavian-embedding-benchmark/)
|
| 1369 |
+
""")
|
| 1370 |
+
with gr.Row():
|
| 1371 |
+
data_classification_sv = gr.components.Dataframe(
|
| 1372 |
+
DATA_CLASSIFICATION_SV,
|
| 1373 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_SV.columns),
|
| 1374 |
+
type="pandas",
|
| 1375 |
+
)
|
| 1376 |
+
with gr.Row():
|
| 1377 |
+
data_run_classification_sv = gr.Button("Refresh")
|
| 1378 |
+
data_run_classification_sv.click(
|
| 1379 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_SV),
|
| 1380 |
+
outputs=data_classification_sv,
|
| 1381 |
+
)
|
| 1382 |
+
with gr.TabItem("Other"):
|
| 1383 |
+
with gr.Row():
|
| 1384 |
+
gr.Markdown("""
|
| 1385 |
+
**Classification Other Languages Leaderboard** 💜💚💙
|
| 1386 |
+
|
| 1387 |
+
- **Metric:** [Accuracy](https://huggingface.co/spaces/evaluate-metric/accuracy)
|
| 1388 |
+
- **Languages:** 47 (Only languages not included in the other tabs)
|
| 1389 |
+
""")
|
| 1390 |
+
with gr.Row():
|
| 1391 |
+
data_classification = gr.components.Dataframe(
|
| 1392 |
+
DATA_CLASSIFICATION_OTHER,
|
| 1393 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLASSIFICATION_OTHER) * 10,
|
| 1394 |
+
type="pandas",
|
| 1395 |
+
)
|
| 1396 |
+
with gr.Row():
|
| 1397 |
+
data_run_classification = gr.Button("Refresh")
|
| 1398 |
+
data_run_classification.click(
|
| 1399 |
+
partial(get_mteb_data, tasks=["Classification"], datasets=TASK_LIST_CLASSIFICATION_OTHER),
|
| 1400 |
+
outputs=data_classification,
|
| 1401 |
+
)
|
| 1402 |
+
with gr.TabItem("Clustering"):
|
| 1403 |
+
with gr.TabItem("English"):
|
| 1404 |
+
with gr.Row():
|
| 1405 |
+
gr.Markdown("""
|
| 1406 |
+
**Clustering Leaderboard** ✨
|
| 1407 |
+
|
| 1408 |
+
- **Metric:** Validity Measure (v_measure)
|
| 1409 |
+
- **Languages:** English
|
| 1410 |
+
""")
|
| 1411 |
+
with gr.Row():
|
| 1412 |
+
data_clustering = gr.components.Dataframe(
|
| 1413 |
+
DATA_CLUSTERING,
|
| 1414 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLUSTERING.columns),
|
| 1415 |
+
type="pandas",
|
| 1416 |
+
)
|
| 1417 |
+
with gr.Row():
|
| 1418 |
+
data_run_clustering_en = gr.Button("Refresh")
|
| 1419 |
+
data_run_clustering_en.click(
|
| 1420 |
+
partial(get_mteb_data, tasks=["Clustering"], datasets=TASK_LIST_CLUSTERING),
|
| 1421 |
+
outputs=data_clustering,
|
| 1422 |
+
)
|
| 1423 |
+
with gr.TabItem("Chinese"):
|
| 1424 |
+
with gr.Row():
|
| 1425 |
+
gr.Markdown("""
|
| 1426 |
+
**Clustering Chinese Leaderboard** ✨🇨🇳
|
| 1427 |
+
|
| 1428 |
+
- **Metric:** Validity Measure (v_measure)
|
| 1429 |
+
- **Languages:** Chinese
|
| 1430 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1431 |
+
""")
|
| 1432 |
+
with gr.Row():
|
| 1433 |
+
data_clustering_zh = gr.components.Dataframe(
|
| 1434 |
+
DATA_CLUSTERING_ZH,
|
| 1435 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLUSTERING_ZH.columns),
|
| 1436 |
+
type="pandas",
|
| 1437 |
+
)
|
| 1438 |
+
with gr.Row():
|
| 1439 |
+
data_run_clustering_zh = gr.Button("Refresh")
|
| 1440 |
+
data_run_clustering_zh.click(
|
| 1441 |
+
partial(get_mteb_data, tasks=["Clustering"], datasets=TASK_LIST_CLUSTERING_ZH),
|
| 1442 |
+
outputs=data_clustering_zh,
|
| 1443 |
+
)
|
| 1444 |
+
with gr.TabItem("German"):
|
| 1445 |
+
with gr.Row():
|
| 1446 |
+
gr.Markdown("""
|
| 1447 |
+
**Clustering German Leaderboard** ✨🇩🇪
|
| 1448 |
+
|
| 1449 |
+
- **Metric:** Validity Measure (v_measure)
|
| 1450 |
+
- **Languages:** German
|
| 1451 |
+
- **Credits:** [Silvan](https://github.com/slvnwhrl)
|
| 1452 |
+
""")
|
| 1453 |
+
with gr.Row():
|
| 1454 |
+
data_clustering_de = gr.components.Dataframe(
|
| 1455 |
+
DATA_CLUSTERING_DE,
|
| 1456 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLUSTERING_DE.columns) * 2,
|
| 1457 |
+
type="pandas",
|
| 1458 |
+
)
|
| 1459 |
+
with gr.Row():
|
| 1460 |
+
data_run_clustering_de = gr.Button("Refresh")
|
| 1461 |
+
data_run_clustering_de.click(
|
| 1462 |
+
partial(get_mteb_data, tasks=["Clustering"], datasets=TASK_LIST_CLUSTERING_DE),
|
| 1463 |
+
outputs=data_clustering_de,
|
| 1464 |
+
)
|
| 1465 |
+
with gr.TabItem("Polish"):
|
| 1466 |
+
with gr.Row():
|
| 1467 |
+
gr.Markdown("""
|
| 1468 |
+
**Clustering Polish Leaderboard** ✨🇵🇱
|
| 1469 |
+
|
| 1470 |
+
- **Metric:** Validity Measure (v_measure)
|
| 1471 |
+
- **Languages:** Polish
|
| 1472 |
+
- **Credits:** [Rafał Poświata](https://github.com/rafalposwiata)
|
| 1473 |
+
""")
|
| 1474 |
+
with gr.Row():
|
| 1475 |
+
data_clustering_pl = gr.components.Dataframe(
|
| 1476 |
+
DATA_CLUSTERING_PL,
|
| 1477 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_CLUSTERING_PL.columns) * 2,
|
| 1478 |
+
type="pandas",
|
| 1479 |
+
)
|
| 1480 |
+
with gr.Row():
|
| 1481 |
+
data_run_clustering_pl = gr.Button("Refresh")
|
| 1482 |
+
data_run_clustering_pl.click(
|
| 1483 |
+
partial(get_mteb_data, tasks=["Clustering"], datasets=TASK_LIST_CLUSTERING_PL),
|
| 1484 |
+
outputs=data_clustering_pl,
|
| 1485 |
+
)
|
| 1486 |
+
with gr.TabItem("Pair Classification"):
|
| 1487 |
+
with gr.TabItem("English"):
|
| 1488 |
+
with gr.Row():
|
| 1489 |
+
gr.Markdown("""
|
| 1490 |
+
**Pair Classification English Leaderboard** 🎭
|
| 1491 |
+
|
| 1492 |
+
- **Metric:** Average Precision based on Cosine Similarities (cos_sim_ap)
|
| 1493 |
+
- **Languages:** English
|
| 1494 |
+
""")
|
| 1495 |
+
with gr.Row():
|
| 1496 |
+
data_pair_classification = gr.components.Dataframe(
|
| 1497 |
+
DATA_PAIR_CLASSIFICATION,
|
| 1498 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_PAIR_CLASSIFICATION.columns),
|
| 1499 |
+
type="pandas",
|
| 1500 |
+
)
|
| 1501 |
+
with gr.Row():
|
| 1502 |
+
data_run_pair_classification = gr.Button("Refresh")
|
| 1503 |
+
data_run_pair_classification.click(
|
| 1504 |
+
partial(get_mteb_data, tasks=["PairClassification"], datasets=TASK_LIST_PAIR_CLASSIFICATION),
|
| 1505 |
+
outputs=data_pair_classification,
|
| 1506 |
+
)
|
| 1507 |
+
with gr.TabItem("Chinese"):
|
| 1508 |
+
with gr.Row():
|
| 1509 |
+
gr.Markdown("""
|
| 1510 |
+
**Pair Classification Chinese Leaderboard** 🎭🇨🇳
|
| 1511 |
+
|
| 1512 |
+
- **Metric:** Average Precision based on Cosine Similarities (cos_sim_ap)
|
| 1513 |
+
- **Languages:** Chinese
|
| 1514 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1515 |
+
""")
|
| 1516 |
+
with gr.Row():
|
| 1517 |
+
data_pair_classification_zh = gr.components.Dataframe(
|
| 1518 |
+
DATA_PAIR_CLASSIFICATION_ZH,
|
| 1519 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_PAIR_CLASSIFICATION_ZH.columns),
|
| 1520 |
+
type="pandas",
|
| 1521 |
+
)
|
| 1522 |
+
with gr.Row():
|
| 1523 |
+
data_run_pair_classification_zh = gr.Button("Refresh")
|
| 1524 |
+
data_run_pair_classification_zh.click(
|
| 1525 |
+
partial(get_mteb_data, tasks=["PairClassification"], datasets=TASK_LIST_PAIR_CLASSIFICATION_ZH),
|
| 1526 |
+
outputs=data_pair_classification_zh,
|
| 1527 |
+
)
|
| 1528 |
+
with gr.TabItem("Polish"):
|
| 1529 |
+
with gr.Row():
|
| 1530 |
+
gr.Markdown("""
|
| 1531 |
+
**Pair Classification Polish Leaderboard** 🎭🇵🇱
|
| 1532 |
+
|
| 1533 |
+
- **Metric:** Average Precision based on Cosine Similarities (cos_sim_ap)
|
| 1534 |
+
- **Languages:** Polish
|
| 1535 |
+
- **Credits:** [Rafał Poświata](https://github.com/rafalposwiata)
|
| 1536 |
+
""")
|
| 1537 |
+
with gr.Row():
|
| 1538 |
+
data_pair_classification_pl = gr.components.Dataframe(
|
| 1539 |
+
DATA_PAIR_CLASSIFICATION_PL,
|
| 1540 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_PAIR_CLASSIFICATION_PL.columns),
|
| 1541 |
+
type="pandas",
|
| 1542 |
+
)
|
| 1543 |
+
with gr.Row():
|
| 1544 |
+
data_run_pair_classification_pl = gr.Button("Refresh")
|
| 1545 |
+
data_run_pair_classification_pl.click(
|
| 1546 |
+
partial(get_mteb_data, tasks=["PairClassification"], datasets=TASK_LIST_PAIR_CLASSIFICATION_PL),
|
| 1547 |
+
outputs=data_pair_classification_pl,
|
| 1548 |
+
)
|
| 1549 |
+
with gr.TabItem("Reranking"):
|
| 1550 |
+
with gr.TabItem("English"):
|
| 1551 |
+
with gr.Row():
|
| 1552 |
+
gr.Markdown("""
|
| 1553 |
+
**Reranking English Leaderboard** 🥈
|
| 1554 |
+
|
| 1555 |
+
- **Metric:** Mean Average Precision (MAP)
|
| 1556 |
+
- **Languages:** English
|
| 1557 |
+
""")
|
| 1558 |
+
with gr.Row():
|
| 1559 |
+
data_reranking = gr.components.Dataframe(
|
| 1560 |
+
DATA_RERANKING,
|
| 1561 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_RERANKING.columns),
|
| 1562 |
+
type="pandas",
|
| 1563 |
+
)
|
| 1564 |
+
with gr.Row():
|
| 1565 |
+
data_run_reranking = gr.Button("Refresh")
|
| 1566 |
+
data_run_reranking.click(
|
| 1567 |
+
partial(get_mteb_data, tasks=["Reranking"], datasets=TASK_LIST_RERANKING),
|
| 1568 |
+
outputs=data_reranking,
|
| 1569 |
+
)
|
| 1570 |
+
with gr.TabItem("Chinese"):
|
| 1571 |
+
with gr.Row():
|
| 1572 |
+
gr.Markdown("""
|
| 1573 |
+
**Reranking Chinese Leaderboard** 🥈🇨🇳
|
| 1574 |
+
|
| 1575 |
+
- **Metric:** Mean Average Precision (MAP)
|
| 1576 |
+
- **Languages:** Chinese
|
| 1577 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1578 |
+
""")
|
| 1579 |
+
with gr.Row():
|
| 1580 |
+
data_reranking_zh = gr.components.Dataframe(
|
| 1581 |
+
DATA_RERANKING_ZH,
|
| 1582 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_RERANKING_ZH.columns),
|
| 1583 |
+
type="pandas",
|
| 1584 |
+
)
|
| 1585 |
+
with gr.Row():
|
| 1586 |
+
data_run_reranking_zh = gr.Button("Refresh")
|
| 1587 |
+
data_run_reranking_zh.click(
|
| 1588 |
+
partial(get_mteb_data, tasks=["Reranking"], datasets=TASK_LIST_RERANKING_ZH),
|
| 1589 |
+
outputs=data_reranking_zh,
|
| 1590 |
+
)
|
| 1591 |
+
with gr.TabItem("Retrieval"):
|
| 1592 |
+
with gr.TabItem("English"):
|
| 1593 |
+
with gr.Row():
|
| 1594 |
+
gr.Markdown("""
|
| 1595 |
+
**Retrieval English Leaderboard** 🔎
|
| 1596 |
+
|
| 1597 |
+
- **Metric:** Normalized Discounted Cumulative Gain @ k (ndcg_at_10)
|
| 1598 |
+
- **Languages:** English
|
| 1599 |
+
""")
|
| 1600 |
+
with gr.Row():
|
| 1601 |
+
data_retrieval = gr.components.Dataframe(
|
| 1602 |
+
DATA_RETRIEVAL,
|
| 1603 |
+
# Add support for more columns than existing as a buffer for CQADupstack & other Retrieval tasks (e.g. MSMARCOv2)
|
| 1604 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_RETRIEVAL.columns) * 2,
|
| 1605 |
+
type="pandas",
|
| 1606 |
+
)
|
| 1607 |
+
with gr.Row():
|
| 1608 |
+
data_run_retrieval = gr.Button("Refresh")
|
| 1609 |
+
data_run_retrieval.click(
|
| 1610 |
+
partial(get_mteb_data, tasks=["Retrieval"], datasets=TASK_LIST_RETRIEVAL),
|
| 1611 |
+
outputs=data_retrieval,
|
| 1612 |
+
)
|
| 1613 |
+
with gr.TabItem("Chinese"):
|
| 1614 |
+
with gr.Row():
|
| 1615 |
+
gr.Markdown("""
|
| 1616 |
+
**Retrieval Chinese Leaderboard** 🔎🇨🇳
|
| 1617 |
+
|
| 1618 |
+
- **Metric:** Normalized Discounted Cumulative Gain @ k (ndcg_at_10)
|
| 1619 |
+
- **Languages:** Chinese
|
| 1620 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1621 |
+
""")
|
| 1622 |
+
with gr.Row():
|
| 1623 |
+
data_retrieval_zh = gr.components.Dataframe(
|
| 1624 |
+
DATA_RETRIEVAL_ZH,
|
| 1625 |
+
# Add support for more columns than existing as a buffer for CQADupstack & other Retrieval tasks (e.g. MSMARCOv2)
|
| 1626 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_RETRIEVAL_ZH.columns) * 2,
|
| 1627 |
+
type="pandas",
|
| 1628 |
+
)
|
| 1629 |
+
with gr.Row():
|
| 1630 |
+
data_run_retrieval_zh = gr.Button("Refresh")
|
| 1631 |
+
data_run_retrieval_zh.click(
|
| 1632 |
+
partial(get_mteb_data, tasks=["Retrieval"], datasets=TASK_LIST_RETRIEVAL_ZH),
|
| 1633 |
+
outputs=data_retrieval_zh,
|
| 1634 |
+
)
|
| 1635 |
+
with gr.TabItem("Polish"):
|
| 1636 |
+
with gr.Row():
|
| 1637 |
+
gr.Markdown("""
|
| 1638 |
+
**Retrieval Polish Leaderboard** 🔎🇵🇱
|
| 1639 |
+
|
| 1640 |
+
- **Metric:** Normalized Discounted Cumulative Gain @ k (ndcg_at_10)
|
| 1641 |
+
- **Languages:** Polish
|
| 1642 |
+
- **Credits:** [Konrad Wojtasik](https://github.com/kwojtasi) & [BEIR-PL](https://arxiv.org/abs/2305.19840)
|
| 1643 |
+
""")
|
| 1644 |
+
with gr.Row():
|
| 1645 |
+
data_retrieval_pl = gr.components.Dataframe(
|
| 1646 |
+
DATA_RETRIEVAL_PL,
|
| 1647 |
+
# Add support for more columns than existing as a buffer for CQADupstack & other Retrieval tasks (e.g. MSMARCOv2)
|
| 1648 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_RETRIEVAL_PL.columns) * 2,
|
| 1649 |
+
type="pandas",
|
| 1650 |
+
)
|
| 1651 |
+
with gr.Row():
|
| 1652 |
+
data_run_retrieval_pl = gr.Button("Refresh")
|
| 1653 |
+
data_run_retrieval_pl.click(
|
| 1654 |
+
partial(get_mteb_data, tasks=["Retrieval"], datasets=TASK_LIST_RETRIEVAL_PL),
|
| 1655 |
+
outputs=data_retrieval_pl,
|
| 1656 |
+
)
|
| 1657 |
+
with gr.TabItem("STS"):
|
| 1658 |
+
with gr.TabItem("English"):
|
| 1659 |
+
with gr.Row():
|
| 1660 |
+
gr.Markdown("""
|
| 1661 |
+
**STS English Leaderboard** 🤖
|
| 1662 |
+
|
| 1663 |
+
- **Metric:** Spearman correlation based on cosine similarity
|
| 1664 |
+
- **Languages:** English
|
| 1665 |
+
""")
|
| 1666 |
+
with gr.Row():
|
| 1667 |
+
data_sts_en = gr.components.Dataframe(
|
| 1668 |
+
DATA_STS_EN,
|
| 1669 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_STS_EN.columns),
|
| 1670 |
+
type="pandas",
|
| 1671 |
+
)
|
| 1672 |
+
with gr.Row():
|
| 1673 |
+
data_run_sts_en = gr.Button("Refresh")
|
| 1674 |
+
data_run_sts_en.click(
|
| 1675 |
+
partial(get_mteb_data, tasks=["STS"], datasets=TASK_LIST_STS),
|
| 1676 |
+
outputs=data_sts_en,
|
| 1677 |
+
)
|
| 1678 |
+
with gr.TabItem("Chinese"):
|
| 1679 |
+
with gr.Row():
|
| 1680 |
+
gr.Markdown("""
|
| 1681 |
+
**STS Chinese Leaderboard** 🤖🇨🇳
|
| 1682 |
+
|
| 1683 |
+
- **Metric:** Spearman correlation based on cosine similarity
|
| 1684 |
+
- **Languages:** Chinese
|
| 1685 |
+
- **Credits:** [FlagEmbedding](https://github.com/FlagOpen/FlagEmbedding)
|
| 1686 |
+
""")
|
| 1687 |
+
with gr.Row():
|
| 1688 |
+
data_sts_zh = gr.components.Dataframe(
|
| 1689 |
+
DATA_STS_ZH,
|
| 1690 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_STS_ZH.columns),
|
| 1691 |
+
type="pandas",
|
| 1692 |
+
)
|
| 1693 |
+
with gr.Row():
|
| 1694 |
+
data_run_sts_zh = gr.Button("Refresh")
|
| 1695 |
+
data_run_sts_zh.click(
|
| 1696 |
+
partial(get_mteb_data, tasks=["STS"], datasets=TASK_LIST_STS_ZH),
|
| 1697 |
+
outputs=data_sts_zh,
|
| 1698 |
+
)
|
| 1699 |
+
with gr.TabItem("Polish"):
|
| 1700 |
+
with gr.Row():
|
| 1701 |
+
gr.Markdown("""
|
| 1702 |
+
**STS Polish Leaderboard** 🤖🇵🇱
|
| 1703 |
+
|
| 1704 |
+
- **Metric:** Spearman correlation based on cosine similarity
|
| 1705 |
+
- **Languages:** Polish
|
| 1706 |
+
- **Credits:** [Rafał Poświata](https://github.com/rafalposwiata)
|
| 1707 |
+
""")
|
| 1708 |
+
with gr.Row():
|
| 1709 |
+
data_sts_pl = gr.components.Dataframe(
|
| 1710 |
+
DATA_STS_PL,
|
| 1711 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_STS_PL.columns),
|
| 1712 |
+
type="pandas",
|
| 1713 |
+
)
|
| 1714 |
+
with gr.Row():
|
| 1715 |
+
data_run_sts_pl = gr.Button("Refresh")
|
| 1716 |
+
data_run_sts_pl.click(
|
| 1717 |
+
partial(get_mteb_data, tasks=["STS"], datasets=TASK_LIST_STS_PL),
|
| 1718 |
+
outputs=data_sts_pl,
|
| 1719 |
+
)
|
| 1720 |
+
with gr.TabItem("Other"):
|
| 1721 |
+
with gr.Row():
|
| 1722 |
+
gr.Markdown("""
|
| 1723 |
+
**STS Other Leaderboard** 👽
|
| 1724 |
+
|
| 1725 |
+
- **Metric:** Spearman correlation based on cosine similarity
|
| 1726 |
+
- **Languages:** Arabic, Chinese, Dutch, English, French, German, Italian, Korean, Polish, Russian, Spanish (Only language combos not included in the other tabs)
|
| 1727 |
+
""")
|
| 1728 |
+
with gr.Row():
|
| 1729 |
+
data_sts_other = gr.components.Dataframe(
|
| 1730 |
+
DATA_STS_OTHER,
|
| 1731 |
+
datatype=["number", "markdown"] + ["number"] * len(DATA_STS_OTHER.columns) * 2,
|
| 1732 |
+
type="pandas",
|
| 1733 |
+
)
|
| 1734 |
+
with gr.Row():
|
| 1735 |
+
data_run_sts_other = gr.Button("Refresh")
|
| 1736 |
+
data_run_sts_other.click(
|
| 1737 |
+
partial(get_mteb_data, tasks=["STS"], datasets=TASK_LIST_STS_OTHER),
|
| 1738 |
+
outputs=data_sts_other,
|
| 1739 |
+
)
|
| 1740 |
+
with gr.TabItem("Summarization"):
|
| 1741 |
+
with gr.Row():
|
| 1742 |
+
gr.Markdown("""
|
| 1743 |
+
**Summarization Leaderboard** 📜
|
| 1744 |
+
|
| 1745 |
+
- **Metric:** Spearman correlation based on cosine similarity
|
| 1746 |
+
- **Languages:** English
|
| 1747 |
+
""")
|
| 1748 |
+
with gr.Row():
|
| 1749 |
+
data_summarization = gr.components.Dataframe(
|
| 1750 |
+
DATA_SUMMARIZATION,
|
| 1751 |
+
datatype=["number", "markdown"] + ["number"] * 2,
|
| 1752 |
+
type="pandas",
|
| 1753 |
+
)
|
| 1754 |
+
with gr.Row():
|
| 1755 |
+
data_run = gr.Button("Refresh")
|
| 1756 |
+
data_run.click(
|
| 1757 |
+
partial(get_mteb_data, tasks=["Summarization"]),
|
| 1758 |
+
outputs=data_summarization,
|
| 1759 |
+
)
|
| 1760 |
+
gr.Markdown(r"""
|
| 1761 |
+
|
| 1762 |
+
If this work is useful to you, please consider citing:
|
| 1763 |
+
|
| 1764 |
+
```bibtex
|
| 1765 |
+
@article{SeaEval2023,
|
| 1766 |
+
title={SeaEval for Multilingual Foundation Models: From Cross-Lingual Alignment to Cultural Reasoning},
|
| 1767 |
+
author={Wang, Bin and Liu, Zhengyuan and Huang, Xin and Jiao, Fangkai and Ding, Yang and Aw, Ai Ti and Chen, Nancy F.},
|
| 1768 |
+
journal={arXiv preprint arXiv:2309.04766},
|
| 1769 |
+
year={2023}
|
| 1770 |
+
}
|
| 1771 |
+
```
|
| 1772 |
+
""")
|
| 1773 |
+
# Running the functions on page load in addition to when the button is clicked
|
| 1774 |
+
# This is optional - If deactivated the data loaded at "Build time" is shown like for Overall tab
|
| 1775 |
+
"""
|
| 1776 |
+
block.load(get_mteb_data, inputs=[task_bitext_mining], outputs=data_bitext_mining)
|
| 1777 |
+
"""
|
| 1778 |
+
|
| 1779 |
+
block.queue(max_size=10)
|
| 1780 |
+
block.launch(server_name="0.0.0.0", share=True)
|
| 1781 |
+
|
| 1782 |
+
|
| 1783 |
+
# Possible changes:
|
| 1784 |
+
# Could add graphs / other visual content
|
| 1785 |
+
# Could add verification marks
|
| 1786 |
+
|
| 1787 |
+
# Sources:
|
| 1788 |
+
# https://huggingface.co/spaces/gradio/leaderboard
|
| 1789 |
+
# https://huggingface.co/spaces/huggingface-projects/Deep-Reinforcement-Learning-Leaderboard
|
| 1790 |
+
# https://getemoji.com/
|
requirements.txt
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
gradio
|
| 2 |
+
datasets
|
| 3 |
+
pandas
|
| 4 |
+
huggingface_hub
|