Skip to content

fix the model card issue as use_cuda_amp is no more available #26731

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
LysandreJik merged 1 commit into from
Oct 11, 2023
Merged

fix the model card issue as use_cuda_amp is no more available #26731

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 src/transformers/modelcard.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -895,10 +895,10 @@ def extract_hyperparameters_from_trainer(trainer):
hyperparameters["num_epochs"] = trainer.args.num_train_epochs

if trainer.args.fp16:
if trainer.use_cuda_amp:
hyperparameters["mixed_precision_training"] = "Native AMP"
elif trainer.use_apex:
if trainer.use_apex:
hyperparameters["mixed_precision_training"] = f"Apex, opt level {trainer.args.fp16_opt_level}"
else:
hyperparameters["mixed_precision_training"] = "Native AMP"
Copy link
Collaborator

@ydshieh ydshieh Oct 11, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't know well this part. Could you share why use_cuda_amp is no longer available from trainer? Otherwise LGTM. Thank you!

Copy link
Contributor

@muellerzr muellerzr Oct 11, 2023 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

SharedDDP was deprecated (and now removed) in favor of using Accelerate natively: #25702


if trainer.args.label_smoothing_factor != 0.0:
hyperparameters["label_smoothing_factor"] = trainer.args.label_smoothing_factor
Expand Down