Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix scoring bug in get_summary_df #792

Merged
merged 8 commits into from
Oct 14, 2024
Merged

Fix scoring bug in get_summary_df #792

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
2d1ac6f
Merge pull request #777 from mlcommons/dev
priyakasimbeg Jul 30, 2024
bdece3b
Merge pull request #781 from mlcommons/dev
priyakasimbeg Aug 29, 2024
a23b5ea
Merge pull request #784 from mlcommons/dev
priyakasimbeg Sep 5, 2024
d09fab5
fixed :bug in summary_df
Niccolo-Ajroldi Oct 11, 2024
57656d8
fixed :bug in summary_df
Niccolo-Ajroldi Oct 11, 2024
223b90b
fixed :bug in summary_df
Niccolo-Ajroldi Oct 11, 2024
076c1a7
fixed :bug in summary_df
Niccolo-Ajroldi Oct 11, 2024
d0fa888
removed pdf
Niccolo-Ajroldi Oct 12, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions scoring/score_submissions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,9 +88,15 @@ def get_summary_df(workload, workload_df, include_test_split=False):
summary_df['time to best eval on val (s)'] = workload_df.apply(
lambda x: x['accumulated_submission_time'][x['index best eval on val']],
axis=1)
summary_df['time to target on val (s)'] = summary_df.apply(
lambda x: x['time to best eval on val (s)']
if x['val target reached'] else np.inf,
workload_df['val target reached'] = workload_df[validation_metric].apply(
lambda x: target_op(x, validation_target)).apply(np.any)
workload_df['index to target on val'] = workload_df.apply(
lambda x: np.argmax(target_op(x[validation_metric], validation_target))
if x['val target reached'] else np.nan,
axis=1)
summary_df['time to target on val (s)'] = workload_df.apply(
lambda x: x['accumulated_submission_time'][int(x[
'index to target on val'])] if x['val target reached'] else np.inf,
axis=1)

# test metrics
Expand Down
Loading