riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 02:02:21 +06:00
Code Issues Actions 33 Packages Projects Releases Wiki Activity

change to apply pad_to_multiple_of to labels (#13949)

Browse Source
This commit is contained in:
Jungwoo Park 2021-10-11 22:35:20 +09:00 committed by GitHub
parent dca6796876
commit 6e4c8f683c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 7 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/transformers/data/data_collator.py
View File

@ -534,6 +534,13 @@ class DataCollatorForSeq2Seq:
# same length to return tensors.
if labels is not None:
max_label_length = max(len(l) for l in labels)
if self.pad_to_multiple_of is not None:
max_label_length = (
(max_label_length + self.pad_to_multiple_of - 1)
// self.pad_to_multiple_of
* self.pad_to_multiple_of
)
padding_side = self.tokenizer.padding_side
for feature in features:
remainder = [self.label_pad_token_id] * (max_label_length - len(feature["labels"]))