Skip to content

perf: cache redundant sum() calls in repetition filters #924

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
HYLcool merged 1 commit into from
Mar 6, 2026
+6 −8
Merged

perf: cache redundant sum() calls in repetition filters #924

Show file tree
Hide file tree
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
9 changes: 4 additions & 5 deletions data_juicer/ops/filter/character_repetition_filter.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,16 +62,15 @@ def compute_stats_batched(self, samples):
samples_stats[idx][StatsKeys.char_rep_ratio] = 0.0
continue

freq_char_ngrams = sorted(list(freq_char_ngrams.values()), reverse=True)
num_no_rep_char_ngrams = len([el for el in freq_char_ngrams if el == 1])
freq_char_ngrams = sorted(freq_char_ngrams.values(), reverse=True)
num_no_rep_char_ngrams = freq_char_ngrams.count(1)
num_rep_char_ngrams = min(
int(np.sqrt(len(freq_char_ngrams))),
len(freq_char_ngrams) - num_no_rep_char_ngrams,
)
total = sum(freq_char_ngrams)
samples_stats[idx][StatsKeys.char_rep_ratio] = (
(sum(freq_char_ngrams[:num_rep_char_ngrams]) / sum(freq_char_ngrams))
if sum(freq_char_ngrams) != 0
else 0.0
(sum(freq_char_ngrams[:num_rep_char_ngrams]) / total) if total != 0 else 0.0
)

return samples
Expand Down
5 changes: 2 additions & 3 deletions data_juicer/ops/filter/word_repetition_filter.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,9 +103,8 @@ def compute_stats_batched(self, samples, context=False):

freq_word_ngrams = list(freq_word_ngrams.values())
rep_more_than_one = [freq for freq in freq_word_ngrams if freq > 1]
samples_stats[idx][StatsKeys.word_rep_ratio] = (
(sum(rep_more_than_one) / sum(freq_word_ngrams)) if sum(freq_word_ngrams) != 0 else 0.0
)
total = sum(freq_word_ngrams)
samples_stats[idx][StatsKeys.word_rep_ratio] = (sum(rep_more_than_one) / total) if total != 0 else 0.0

return samples

Expand Down