Spaces:
Running
Running
orionweller
commited on
Commit
•
404b92c
1
Parent(s):
3219cef
fixed update
Browse files- refresh.py +10 -7
refresh.py
CHANGED
@@ -323,11 +323,17 @@ def get_mteb_data(tasks=["Clustering"], langs=[], datasets=[], fillna=True, add_
|
|
323 |
df['MLSUMClusteringS2S (fr)'] = df['MLSUMClusteringS2S (fr)'].fillna(df['MLSUMClusteringS2S'])
|
324 |
datasets.remove('MLSUMClusteringS2S')
|
325 |
if ('PawsXPairClassification (fr)' in datasets) and ('PawsX (fr)' in cols):
|
326 |
-
|
327 |
-
|
328 |
-
else:
|
329 |
df['PawsXPairClassification (fr)'] = df['PawsX (fr)']
|
|
|
|
|
|
|
330 |
datasets.remove('PawsX (fr)')
|
|
|
|
|
|
|
|
|
331 |
# Filter invalid columns
|
332 |
cols = [col for col in cols if col in base_columns + datasets]
|
333 |
i = 0
|
@@ -356,10 +362,7 @@ def get_mteb_average(task_dict: dict):
|
|
356 |
)
|
357 |
# Debugging:
|
358 |
# DATA_OVERALL.to_csv("overall.csv")
|
359 |
-
|
360 |
-
DATA_OVERALL.insert(1, f"Average ({len(all_tasks)} datasets)", DATA_OVERALL[all_tasks].mean(axis=1, skipna=False))
|
361 |
-
except Exception as e:
|
362 |
-
breakpoint()
|
363 |
for i, (task_category, task_category_list) in enumerate(task_dict.items()):
|
364 |
DATA_OVERALL.insert(i+2, f"{task_category} Average ({len(task_category_list)} datasets)", DATA_OVERALL[task_category_list].mean(axis=1, skipna=False))
|
365 |
DATA_OVERALL.sort_values(f"Average ({len(all_tasks)} datasets)", ascending=False, inplace=True)
|
|
|
323 |
df['MLSUMClusteringS2S (fr)'] = df['MLSUMClusteringS2S (fr)'].fillna(df['MLSUMClusteringS2S'])
|
324 |
datasets.remove('MLSUMClusteringS2S')
|
325 |
if ('PawsXPairClassification (fr)' in datasets) and ('PawsX (fr)' in cols):
|
326 |
+
# for the first bit no model has it, hence no column for it. We can remove this in a month or so
|
327 |
+
if "PawsXPairClassification (fr)" not in cols:
|
|
|
328 |
df['PawsXPairClassification (fr)'] = df['PawsX (fr)']
|
329 |
+
else:
|
330 |
+
df['PawsXPairClassification (fr)'] = df['PawsXPairClassification (fr)'].fillna(df['PawsX (fr)'])
|
331 |
+
# make all the columns the same
|
332 |
datasets.remove('PawsX (fr)')
|
333 |
+
cols.remove('PawsX (fr)')
|
334 |
+
df.drop(columns=['PawsX (fr)'], inplace=True)
|
335 |
+
cols.append('PawsXPairClassification (fr)')
|
336 |
+
|
337 |
# Filter invalid columns
|
338 |
cols = [col for col in cols if col in base_columns + datasets]
|
339 |
i = 0
|
|
|
362 |
)
|
363 |
# Debugging:
|
364 |
# DATA_OVERALL.to_csv("overall.csv")
|
365 |
+
DATA_OVERALL.insert(1, f"Average ({len(all_tasks)} datasets)", DATA_OVERALL[all_tasks].mean(axis=1, skipna=False))
|
|
|
|
|
|
|
366 |
for i, (task_category, task_category_list) in enumerate(task_dict.items()):
|
367 |
DATA_OVERALL.insert(i+2, f"{task_category} Average ({len(task_category_list)} datasets)", DATA_OVERALL[task_category_list].mean(axis=1, skipna=False))
|
368 |
DATA_OVERALL.sort_values(f"Average ({len(all_tasks)} datasets)", ascending=False, inplace=True)
|