Skip to content
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.

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

Add configuration mixed_precision for finetuning #23

Merged
merged 3 commits into from
Jan 5, 2024
Merged

Conversation

carsonwang
Copy link
Contributor

@carsonwang carsonwang commented Jan 3, 2024

-Add configuration mixed_precision for finetuning, which has a great impact on finetuning performance.
-Update some accelerate env values from "True"/"False" to "true"/"false" to be consistent.

@@ -22,6 +22,7 @@ Training:
learning_rate: 1.0e-05
lr_scheduler: linear
weight_decay: 0.0
mixed_precision: bf16
Copy link
Contributor

Choose a reason for hiding this comment

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

add a log_step option in General part?
This option can be used to control number of training logs for DefaultTrainer

Copy link
Contributor Author

Choose a reason for hiding this comment

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

We should add it. But let's do it in a separate PR.

@harborn harborn merged commit cd05f0e into main Jan 5, 2024
6 checks passed
carsonwang pushed a commit to carsonwang/llm-on-ray that referenced this pull request Jan 9, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants