Skip to content

[trainer] make generate work with multigpu #8716

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
stas00 merged 2 commits into from
Nov 23, 2020
Merged

[trainer] make generate work with multigpu #8716

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
784159a
make generate work with multigpu
stas00 Nov 22, 2020
91abcdd
better fix - thanks @sgugger
stas00 Nov 23, 2020
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
3 changes: 2 additions & 1 deletion examples/seq2seq/seq2seq_trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -189,7 +189,8 @@ def prediction_step(
}

if self.args.predict_with_generate and not self.args.prediction_loss_only:
generated_tokens = model.generate(
_model = model.module if hasattr(model, "module") else model
generated_tokens = _model.generate(
inputs["input_ids"],
attention_mask=inputs["attention_mask"],
**gen_kwargs,
Expand Down
9 changes: 1 addition & 8 deletions examples/seq2seq/test_finetune_trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,13 +4,7 @@

from transformers import BertTokenizer, EncoderDecoderModel
from transformers.file_utils import is_datasets_available
from transformers.testing_utils import (
TestCasePlus,
execute_subprocess_async,
get_gpu_count,
require_torch_non_multi_gpu_but_fix_me,
slow,
)
from transformers.testing_utils import TestCasePlus, execute_subprocess_async, get_gpu_count, slow
from transformers.trainer_callback import TrainerState
from transformers.trainer_utils import set_seed

Expand Down Expand Up @@ -52,7 +46,6 @@ def test_finetune_trainer_slow(self):
assert "test_results.json" in contents

@slow
@require_torch_non_multi_gpu_but_fix_me
def test_finetune_bert2bert(self):
if not is_datasets_available():
return
Expand Down