Add chatbot_arena_20240814.csv

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+chatbot_arena_20240814.csv filter=lfs diff=lfs merge=lfs -text

README.md

@@ -16,3 +16,5 @@ license: apache-2.0
 [Evalica](https://github.com/dustalov/evalica) is a library for pairwise comparisons as described in paper
 Reliable, Reproducible, and Really Fast Leaderboards with Evalica
 ([arXiv](https://arxiv.org/abs/2412.11314)).
+
+Chatbot Arena dataset `chatbot_arena_20240814.csv` was derived from the [clean_battle_20240814_public.json](https://storage.googleapis.com/arena_external_data/public/clean_battle_20240814_public.json) dataset available from <https://lmarena.ai/>.

app.py

@@ -277,6 +277,8 @@ def main() -> None:
             ["llmfao.csv", "Average Win Rate", False, True, 100],
             ["llmfao.csv", "Bradley-Terry (1952)", False, True, 100],
             ["llmfao.csv", "Elo (1960)", False, True, 100],
+            ["llmfao.csv", "Elo (1960)", False, True, 100],
+            ["chatbot_arena_20240814.csv", "Bradley-Terry (1952)", False, False, 0],
         ],
         title="Evalica: Turn Your Side-by-Side Comparisons into Ranking!",
         description="""

chatbot_arena_20240814.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ad4480918fe15df6a60860ee6e498cac730f5585588981f9ac8ee889f4e6e82
+size 74002919