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.

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

Jump to bottom

Weekly Patch Release v.1.2.7 [full merge, no squash] #6850

Merged
merged 18 commits into from
Apr 7, 2021
Merged

Weekly Patch Release v.1.2.7 [full merge, no squash] #6850

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
51e9591
update readme by v1.2.x (#6728)
Borda Mar 29, 2021
553c3e5
[bugfix] Add support for omegaconf and tpu (#6741)
tchaton Mar 30, 2021
34e7282
[docs] Update Bolts link (#6743)
akihironitta Mar 30, 2021
709b695
Update logic for checking TPUs availability (#6767)
kaushikb11 Mar 31, 2021
8c01064
resolve bug (#6781)
tchaton Apr 1, 2021
5c9dbc3
Fix validation progress counter with check_val_every_n_epoch > 1 (#5952)
sailordiary Apr 2, 2021
591a087
Remove extinct parameters from lightning_module.rst (#6801)
elimohl Apr 2, 2021
2ceed8a
Update TPU docs for installation (#6794)
kaushikb11 Apr 3, 2021
17969f3
fix boolean check on iterable dataset when len not defined (#6828)
awaelchli Apr 5, 2021
cd997d6
Sanitize `None` params during pruning (#6836)
karthikprasad Apr 5, 2021
c92f84a
Fix `unfreeze_and_add_param_group` expects `modules` rather than `mod…
sadiqj Apr 5, 2021
31b2d2b
Enforce an epoch scheduler interval when using SWA (#6588)
mibaumgartner Apr 6, 2021
215a9c9
Fix DPP + SyncBN (#6838)
BloodAxe Apr 6, 2021
3643954
[Fix] TPU Training Type Plugin (#6816)
kaushikb11 Apr 6, 2021
716ade5
Fix support for symlink save_dir in TensorBoardLogger (#6730)
ethanwharris Apr 6, 2021
653232c
Fixed missing arguments in `lr_find` call (#6784)
THasthika Apr 6, 2021
c8c5d05
Update Changelog & version
kaushikb11 Apr 6, 2021
86f6d08
Fix TPU tests for checkpoint
kaushikb11 Apr 6, 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
Prev Previous commit
Next Next commit
Fix DPP + SyncBN (#6838) 
* Fix DPP + SyncBN

Ensure that model is already on correct GPU before applying SyncBN conversion

* Fix order of SyncBN for ddp_spawn
  • Loading branch information
@BloodAxe @kaushikb11
BloodAxe authored and kaushikb11 committed Apr 6, 2021
commit 215a9c9c89df01b7370b55d0269ac94a04aacb55
6 changes: 3 additions & 3 deletions pytorch_lightning/plugins/training_type/ddp.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -257,12 +257,12 @@ def pre_dispatch(self):
self.dist.rank = self.global_rank
self.dist.device = self.root_device

if self.sync_batchnorm:
self.model = self.configure_sync_batchnorm(self.model)

# move the model to the correct device
self.model_to_device()

if self.sync_batchnorm:
self.model = self.configure_sync_batchnorm(self.model)

self.configure_ddp()

self.barrier()
Expand Down
6 changes: 3 additions & 3 deletions pytorch_lightning/plugins/training_type/ddp_spawn.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -148,12 +148,12 @@ def new_process(self, process_idx, trainer, mp_queue):
self.dist.rank = self.global_rank
self.dist.device = self.root_device

if self.sync_batchnorm:
self.model = self.configure_sync_batchnorm(self.model)

# move the model to the correct device
self.model_to_device()

if self.sync_batchnorm:
self.model = self.configure_sync_batchnorm(self.model)

self.configure_ddp()

self.barrier()
Expand Down
3 changes: 0 additions & 3 deletions tests/trainer/test_dataloaders.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -721,8 +721,6 @@ def __len__(self):
assert has_len(dataloader)
assert has_iterable_dataset(dataloader)
trainer = Trainer(default_root_dir=tmpdir, max_steps=3)
with pytest.warns(UserWarning, match='Your `IterableDataset` has `__len__` defined.'):
trainer.validate(model, val_dataloaders=[dataloader])
with pytest.warns(UserWarning, match='Your `IterableDataset` has `__len__` defined.'):
trainer.fit(model, train_dataloader=dataloader, val_dataloaders=[dataloader])
with pytest.warns(UserWarning, match='Your `IterableDataset` has `__len__` defined.'):
Expand All @@ -735,7 +733,6 @@ def __len__(self):
assert not has_len(dataloader)
assert has_iterable_dataset(dataloader)
trainer = Trainer(default_root_dir=tmpdir, max_steps=3)
trainer.validate(model, val_dataloaders=dataloader)
trainer.fit(model, train_dataloader=dataloader, val_dataloaders=[dataloader])
trainer.test(model, test_dataloaders=dataloader)
trainer.predict(model, dataloaders=dataloader)
Expand Down