2/n inter batch parallelism #9047
Merged
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
for more information, see https://pre-commit.ci
…rchLightning/pytorch-lightning into 1/n_inter_batch_parallelism
…lightning into 0/n_inter_batch_parallelism
tchaton
requested review from
awaelchli,
Borda,
carmocca,
justusschock,
kaushikb11,
SeanNaren and
williamFalcon
as code owners
Codecov Report
@@ Coverage Diff @@
## master #9047 +/- ##
======================================
Coverage 89% 89%
======================================
Files 175 175
Lines 14463 14497 +34
======================================
+ Hits 12812 12844 +32
- Misses 1651 1653 +2 |
for more information, see https://pre-commit.ci
for more information, see https://pre-commit.ci
kaushikb11
reviewed
Aug 23, 2021
carmocca
reviewed
Aug 23, 2021
for more information, see https://pre-commit.ci
carmocca
approved these changes
Aug 23, 2021
for more information, see https://pre-commit.ci
9 tasks
for more information, see https://pre-commit.ci
justusschock
approved these changes
Aug 23, 2021
aazzolini
reviewed
Aug 27, 2021
| @@ -280,6 +280,8 @@ def _training_step( | |||
| training_step_output = self.trainer.accelerator.training_step(step_kwargs) | |||
| self.trainer.accelerator.post_training_step() | |||
|
|
|||
| del step_kwargs | |||
| @@ -91,8 +91,9 @@ def advance( | |||
| if batch is None: | |||
| raise StopIteration | |||
|
|
|||
| with self.trainer.profiler.profile("evaluation_batch_to_device"): | |||
| batch = self.trainer.accelerator.batch_to_device(batch, dataloader_idx=dataloader_idx) | |||
| if not self.trainer.data_connector.evaluation_data_fetcher.store_on_device: | |||
| @@ -47,22 +47,27 @@ class SimpleDataFetcher(AbstractDataFetcher): | |||
| def fetching_function(self): | |||
| while True: | |||
| try: | |||
| yield next(self.dataloader_iter), False | |||
| return next(self.dataloader_iter), False | |||
There was a problem hiding this comment.
Why this change?
I am having a hard time to undertsand why we have a while loop that will only run for one iteration. It's not clear why this hasn't been replaced by an if statement, or why we didn't just remove the while loop here.
Also it seems that this function now became just semantically equivalent to implement next in a regular iterator, why do we still need this fecthing_function here?
|
@aazzolini I also noticed some of the things you mention here and opened #9058 to address them. Thanks for reviewing! |
This was referenced Jan 12, 2022
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?
This PR adds DataFetcher to
data_connectorand loops.Parts #8867
Parts #8316
Does your PR introduce any breaking changes? If yes, please list them.
Before submitting
PR review
Anyone in the community is welcome to review the PR.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:
Did you have fun?
Make sure you had fun coding 🙃