Skip to content

Deprecate log_gpu_memory off the Trainer constructor #9124

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

Closed
wants to merge 3 commits into from
Closed

Deprecate log_gpu_memory off the Trainer constructor #9124

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b8f9301
Move logging GPU metrics out of the core trainer
ananthsub Aug 26, 2021
2b7858c
go for direct deprecation instead
ananthsub Aug 26, 2021
0bba67b
Update logger_connector.py
ananthsub Aug 26, 2021
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
5 changes: 4 additions & 1 deletion CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,7 +147,10 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
- Deprecated `DataModule` properties: `train_transforms`, `val_transforms`, `test_transforms`, `size`, `dims` ([#8851](https://github.com/PyTorchLightning/pytorch-lightning/pull/8851))


- Deprecated `prepare_data_per_node` flag on Trainer and set it as a property of `DataHooks`, accessible in the `LightningModule` and `LightningDataModule` [#8958](https://github.com/PyTorchLightning/pytorch-lightning/pull/8958)
- Deprecated `prepare_data_per_node` flag on Trainer and set it as a property of `DataHooks`, accessible in the `LightningModule` and `LightningDataModule` ([#8958](https://github.com/PyTorchLightning/pytorch-lightning/pull/8958))


- Deprecated `log_gpu_memory` flag on the Trainer in favor of passing the `GPUStatsMonitor` callback to the Trainer ([#9124](https://github.com/PyTorchLightning/pytorch-lightning/pull/9124/))


### Removed
Expand Down
6 changes: 6 additions & 0 deletions pytorch_lightning/trainer/connectors/logger_connector/logger_connector.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,11 +24,17 @@
from pytorch_lightning.utilities.apply_func import apply_to_collection, move_data_to_device
from pytorch_lightning.utilities.metrics import metrics_to_scalars
from pytorch_lightning.utilities.types import _EVALUATE_OUTPUT
from pytorch_lightning.utilities.warnings import rank_zero_deprecation


class LoggerConnector:
def __init__(self, trainer: "pl.Trainer", log_gpu_memory: Optional[str] = None) -> None:
self.trainer = trainer
if log_gpu_memory is not None:
Copy link
Contributor

@tchaton tchaton Aug 26, 2021

Choose a reason for hiding this comment

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

Do you plan adding log_gpu_memory to the callback in another PR ?

Copy link
Contributor

@tchaton tchaton Aug 26, 2021

Choose a reason for hiding this comment

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

If yes, it should be done the other way around.

rank_zero_deprecation(
"Setting `log_gpu_memory` with the trainer flag is deprecated and will be removed in v1.7.0! "
"Please monitor GPU stats with the `GPUStatsMonitor` callback directly instead."
)
self.log_gpu_memory = log_gpu_memory
self.eval_loop_results = []
self._val_log_step: int = 0
Expand Down
4 changes: 4 additions & 0 deletions pytorch_lightning/trainer/trainer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -236,6 +236,10 @@ def __init__(

log_gpu_memory: None, 'min_max', 'all'. Might slow performance

.. deprecated:: v1.5
Deprecated in v1.5.0 and will be removed in v1.7.0
Please use the ``GPUStatsMonitor`` callback directly instead.

log_every_n_steps: How often to log within steps (defaults to every 50 steps).

prepare_data_per_node: If True, each LOCAL_RANK=0 will call prepare data.
Expand Down
7 changes: 7 additions & 0 deletions tests/deprecated_api/test_remove_1-7.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,3 +87,10 @@ def test_v1_7_0_trainer_prepare_data_per_node(tmpdir):
match="Setting `prepare_data_per_node` with the trainer flag is deprecated and will be removed in v1.7.0!"
):
_ = Trainer(prepare_data_per_node=False)


def test_v1_7_0_trainer_log_gpu_memory(tmpdir):
with pytest.deprecated_call(
match="Setting `log_gpu_memory` with the trainer flag is deprecated and will be removed in v1.7.0!"
):
_ = Trainer(log_gpu_memory="min_max")