Skip to content

DRY sampler improvements #6053

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
oobabooga merged 7 commits into from
Jun 13, 2024
Merged

DRY sampler improvements #6053

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
7c03e4a
Improve DRY sampler performance 1 of 2 (simple data type changes).
belladoreai May 22, 2024
3c51b95
Fix DRY crash on very repetitive inputs.
belladoreai May 25, 2024
62e76ed
Rename variable.
belladoreai May 26, 2024
918ff94
Change DRY max match length.
belladoreai May 26, 2024
686e7a8
Comment.
belladoreai May 28, 2024
ed43c5e
Merge branch 'dev' into belladoreai-dev-dry-optimization2
oobabooga Jun 13, 2024
b556ee2
lint
oobabooga Jun 13, 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
Prev Previous commit
Next Next commit
Comment.
  • Loading branch information
@belladoreai
belladoreai committed May 28, 2024
commit 686e7a8344f1aa9d27da3055906119f55f085d25
5 changes: 1 addition & 4 deletions modules/sampler_hijack.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -234,13 +234,10 @@ def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> to
input_ids = input_ids[:, -self._range:]

for input_ids_row, scores_row in zip(input_ids, scores):

# Use normal Python data types for improved performance
input_ids = input_ids_row.tolist()

# Raw integer must be extracted here to check for set membership.
last_token = input_ids[-1]

last_token = input_ids[-1]
if last_token in self.sequence_breakers:
continue

Expand Down