[core] Import tensorflow inside relevant methods in trainer_utils
#26106
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The documentation is not available anymore as the PR was closed or merged. |
amyeroberts
approved these changes
Sep 12, 2023
parambharat
pushed a commit
to parambharat/transformers
that referenced
this pull request
Sep 26, 2023
…huggingface#26106) import tensorflow inside relevant methods in trainer_utils
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
Imports tensorflow inside relevant methods in
trainer_utils.When Deepspeed and accelerate are available in your env, it will try to import transformers all the way down to https://github.com/huggingface/transformers/blob/main/src/transformers/trainer_utils.py#L49
Leading to considerable slowdowns in import time
This PR imports tensorflow inside the relevant methods so that it will never get globally imported by transformers anymore.
If this patch gets applied, there is no speedup, however this can be fixed by lazy loading TF in accelerate itself:
cc @amyeroberts @patrickvonplaten @pacman100 @muellerzr @SunMarc