Spaces:

MERaLiON
/

SeaEval_Leaderboard

Running

binwang commited on Feb 27, 2024

Commit

0ecf0cf

1 Parent(s): c2e2d08

new results

Files changed (2) hide show

all_results.json CHANGED Viewed

The diff for this file is too large to render. See raw diff

app.py CHANGED Viewed

@@ -170,7 +170,7 @@ def get_data_cross_mmlu_language(eval_mode='zero_shot', fillna=True, rank=True):
     df = df[cols]
     if rank:
-        df = add_rank(df, compute_average=True)
     if fillna:
         df.fillna("", inplace=True)
@@ -309,7 +309,7 @@ def get_data_cross_logiqa_language(eval_mode='zero_shot', fillna=True, rank=True
     df = df[cols]
     if rank:
-        df = add_rank(df, compute_average=True)
     if fillna:
         df.fillna("", inplace=True)

     df = df[cols]
     if rank:
+        df = add_rank(df, compute_average=False)
     if fillna:
         df.fillna("", inplace=True)
     df = df[cols]
     if rank:
+        df = add_rank(df, compute_average=False)
     if fillna:
         df.fillna("", inplace=True)