Skip to content

2d buckets estimation fix #13377

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

Merged
merged 1 commit into from
May 1, 2025
Merged

2d buckets estimation fix #13377

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
6 changes: 3 additions & 3 deletions scripts/speech_recognition/estimate_duration_bins_2d.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -264,7 +264,7 @@ def _estimate_token_buckets(max_bucket_duration, start_idx, end_idx, corr_subbuc
# Here, we'd end up making token subbuckets for: [5, 20, 40]
# where [20, 40] bucket will have 4 times more subbuckets (as we combined 4 buckets into 1) than usual bucket in that settings.

for i, (duration_bin, bin_idx) in enumerate(zip(duration_bins, bin_indexes[1:])):
for i, (duration_bin, binidx) in enumerate(zip(duration_bins, bin_indexes[1:])):
if (i != len(duration_bins) - 1 and duration_bins[i + 1] == duration_bin) or (
i == len(duration_bins) - 1 and max_duration == duration_bin
):
Expand All @@ -276,7 +276,7 @@ def _estimate_token_buckets(max_bucket_duration, start_idx, end_idx, corr_subbuc
end_idx=binidx,
corr_subbuckets=num_subbuckets * skipped_buckets,
)
start_idx = bin_idx
start_idx = binidx
skipped_buckets = 1

# Estimate an extra 2D bin set for global max duration.
Expand All @@ -285,7 +285,7 @@ def _estimate_token_buckets(max_bucket_duration, start_idx, end_idx, corr_subbuc
_estimate_token_buckets(
max_bucket_duration=max_duration,
start_idx=start_idx,
end_idx=binidx,
end_idx=len(sizes),
corr_subbuckets=num_subbuckets * skipped_buckets,
)
return bins, tps_thresholds
Expand Down
Loading