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

rename variables nb -> num #567

Merged
merged 5 commits into from
Dec 4, 2019
Merged

rename variables nb -> num #567

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
717bbba
rename nb -> num
Borda Dec 2, 2019
156172f
flake8
Borda Dec 2, 2019
fdc40f8
batch_nb, epoch_nb, gpu_nb, split_nb
Borda Dec 2, 2019
1cfdc48
add _num deprecations
Borda Dec 2, 2019
30a3872
Merge branch 'master' into switch_nb-num
williamFalcon Dec 4, 2019
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
20 changes: 10 additions & 10 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -96,15 +96,15 @@ To use lightning do 2 things:
def forward(self, x):
return torch.relu(self.l1(x.view(x.size(0), -1)))

def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
# REQUIRED
x, y = batch
y_hat = self.forward(x)
loss = F.cross_entropy(y_hat, y)
tensorboard_logs = {'train_loss': loss}
return {'loss': loss, 'log': tensorboard_logs}

def validation_step(self, batch, batch_nb):
def validation_step(self, batch, batch_idx):
# OPTIONAL
x, y = batch
y_hat = self.forward(x)
Expand Down Expand Up @@ -154,16 +154,16 @@ use something other than tensorboard).
Here are more advanced examples
```python
# train on cpu using only 10% of the data (for demo purposes)
trainer = Trainer(max_nb_epochs=1, train_percent_check=0.1)
trainer = Trainer(max_num_epochs=1, train_percent_check=0.1)

# train on 4 gpus (lightning chooses GPUs for you)
# trainer = Trainer(max_nb_epochs=1, gpus=4, distributed_backend='ddp')
# trainer = Trainer(max_num_epochs=1, gpus=4, distributed_backend='ddp')

# train on 4 gpus (you choose GPUs)
# trainer = Trainer(max_nb_epochs=1, gpus=[0, 1, 3, 7], distributed_backend='ddp')
# trainer = Trainer(max_num_epochs=1, gpus=[0, 1, 3, 7], distributed_backend='ddp')

# train on 32 gpus across 4 nodes (make sure to submit appropriate SLURM job)
# trainer = Trainer(max_nb_epochs=1, gpus=8, nb_gpu_nodes=4, distributed_backend='ddp')
# trainer = Trainer(max_num_epochs=1, gpus=8, num_gpu_nodes=4, distributed_backend='ddp')

# train (1 epoch only here for demo)
trainer.fit(model)
Expand All @@ -187,10 +187,10 @@ You define the blue parts using the LightningModule interface:

```python
# what to do in the training loop
def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):

# what to do in the validation loop
def validation_step(self, batch, batch_nb):
def validation_step(self, batch, batch_idx):

# how to aggregate validation_step outputs
def validation_end(self, outputs):
Expand All @@ -205,7 +205,7 @@ def test_dataloader():

```python
# define what happens for training here
def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
x, y = batch

# define your own forward and loss calculation
Expand All @@ -232,7 +232,7 @@ def training_step(self, batch, batch_nb):

```python
# define what happens for validation here
def validation_step(self, batch, batch_nb):
def validation_step(self, batch, batch_idx):
x, y = batch

# or as basic as a CNN classification
Expand Down
4 changes: 2 additions & 2 deletions docs/source/intro.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ class BERT(pl.LightningModule):
elif model_name == 'my_cool_version':
self.net = MyCoolVersion()

def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
if self.task == 'standard_bert':
# do standard bert training with self.net...
# return loss
Expand All @@ -35,7 +35,7 @@ class CoolerNotBERT(pl.LightningModule):
def __init__(self):
self.net = ...

def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
# do some other cool task
# return loss
```
Expand Down
6 changes: 3 additions & 3 deletions pl_examples/domain_templates/gan.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,12 +90,12 @@ def forward(self, z):
def adversarial_loss(self, y_hat, y):
return F.binary_cross_entropy(y_hat, y)

def training_step(self, batch, batch_nb, optimizer_i):
def training_step(self, batch, batch_idx, optimizer_idx):
imgs, _ = batch
self.last_imgs = imgs

# train generator
if optimizer_i == 0:
if optimizer_idx == 0:
# sample noise
z = torch.randn(imgs.shape[0], self.hparams.latent_dim)

Expand Down Expand Up @@ -125,7 +125,7 @@ def training_step(self, batch, batch_nb, optimizer_i):
return output

# train discriminator
if optimizer_i == 1:
if optimizer_idx == 1:
# Measure discriminator's ability to classify real from generated samples

# how well can it label as real?
Expand Down
2 changes: 1 addition & 1 deletion pl_examples/full_examples/imagenet/imagenet_example.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -234,7 +234,7 @@ def main(hparams):
trainer = pl.Trainer(
default_save_path=hparams.save_path,
gpus=hparams.gpus,
max_nb_epochs=hparams.epochs,
max_num_epochs=hparams.epochs,
Copy link
Contributor

Choose a reason for hiding this comment

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

add backward compatibility

distributed_backend=hparams.distributed_backend,
use_amp=hparams.use_16bit
)
Expand Down
2 changes: 1 addition & 1 deletion pl_examples/multi_node_examples/multi_node_ddp2_demo.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ def main(hparams):
# ------------------------
trainer = Trainer(
gpus=2,
nb_gpu_nodes=2,
num_nodes=2,
distributed_backend='ddp2'
)

Expand Down
2 changes: 1 addition & 1 deletion pl_examples/multi_node_examples/multi_node_ddp_demo.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ def main(hparams):
# ------------------------
trainer = Trainer(
gpus=2,
nb_gpu_nodes=2,
num_nodes=2,
distributed_backend='ddp'
)

Expand Down
6 changes: 3 additions & 3 deletions pytorch_lightning/core/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,13 +34,13 @@ def __init__(self):
def forward(self, x):
return torch.relu(self.l1(x.view(x.size(0), -1)))

def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
# REQUIRED
x, y = batch
y_hat = self.forward(x)
return {'loss': F.cross_entropy(y_hat, y)}

def validation_step(self, batch, batch_nb):
def validation_step(self, batch, batch_idx):
# OPTIONAL
x, y = batch
y_hat = self.forward(x)
Expand All @@ -51,7 +51,7 @@ def validation_end(self, outputs):
avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()
return {'avg_val_loss': avg_loss}

def test_step(self, batch, batch_nb):
def test_step(self, batch, batch_idx):
# OPTIONAL
x, y = batch
y_hat = self.forward(x)
Expand Down
64 changes: 32 additions & 32 deletions pytorch_lightning/core/lightning.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,7 +109,7 @@ def training_step(self, *args, **kwargs):
"""return loss, dict with metrics for tqdm

:param batch: The output of your dataloader. A tensor, tuple or list
:param int batch_nb: Integer displaying which batch this is
:param int batch_idx: Integer displaying which batch this is
:return: dict with loss key and optional log, progress keys
if implementing training_step, return whatever you need in that step:
- loss -> tensor scalar [REQUIRED]
Expand All @@ -124,7 +124,7 @@ def training_step(self, *args, **kwargs):

.. code-block:: python

def training_step(self, batch, batch_nb):
def training_step(self, batch, batch_idx):
x, y, z = batch

# implement your own
Expand All @@ -150,7 +150,7 @@ def training_step(self, batch, batch_nb):
.. code-block:: python

# Multiple optimizers (ie: GANs)
def training_step(self, batch, batch_nb, optimizer_idx):
def training_step(self, batch, batch_idx, optimizer_idx):
if optimizer_idx == 0:
# do training_step with encoder
if optimizer_idx == 1:
Expand All @@ -163,7 +163,7 @@ def training_step(self, batch, batch_nb, optimizer_idx):
.. code-block:: python

# Truncated back-propagation through time
def training_step(self, batch, batch_nb, hiddens):
def training_step(self, batch, batch_idx, hiddens):
# hiddens are the hiddens from the previous truncated backprop step

You can also return a -1 instead of a dict to stop the current loop. This is useful
Expand Down Expand Up @@ -192,9 +192,9 @@ def training_end(self, *args, **kwargs):
.. code-block:: python

# WITHOUT training_end
# if used in DP or DDP2, this batch is 1/nb_gpus large
def training_step(self, batch, batch_nb):
# batch is 1/nb_gpus big
# if used in DP or DDP2, this batch is 1/num_gpus large
def training_step(self, batch, batch_idx):
# batch is 1/num_gpus big
x, y = batch

out = self.forward(x)
Expand All @@ -204,8 +204,8 @@ def training_step(self, batch, batch_nb):

# --------------
# with training_end to do softmax over the full batch
def training_step(self, batch, batch_nb):
# batch is 1/nb_gpus big
def training_step(self, batch, batch_idx):
# batch is 1/num_gpus big
x, y = batch

out = self.forward(x)
Expand All @@ -225,7 +225,7 @@ def training_end(self, outputs):
.. code-block:: python

# Multiple optimizers (ie: GANs)
def training_step(self, batch, batch_nb, optimizer_idx):
def training_step(self, batch, batch_idx, optimizer_idx):
if optimizer_idx == 0:
# do training_step with encoder
if optimizer_idx == 1:
Expand All @@ -237,7 +237,7 @@ def training_step(self, batch, batch_nb, optimizer_idx):
.. code-block:: python

# Truncated back-propagation through time
def training_step(self, batch, batch_nb, hiddens):
def training_step(self, batch, batch_idx, hiddens):
# hiddens are the hiddens from the previous truncated backprop step

You can also return a -1 instead of a dict to stop the current loop. This is useful if you want to
Expand All @@ -249,17 +249,17 @@ def validation_step(self, *args, **kwargs):
"""return whatever outputs will need to be aggregated in validation_end

:param batch: The output of your dataloader. A tensor, tuple or list
:param int batch_nb: Integer displaying which batch this is
:param int batch_idx: Integer displaying which batch this is
:param int dataloader_idx: Integer displaying which dataloader this is (only if multiple val datasets used)
:return dict: Dict or OrderedDict - passed to the validation_end step

.. code-block:: python

# if you have one val dataloader:
def validation_step(self, batch, batch_nb)
def validation_step(self, batch, batch_idx)

# if you have multiple val dataloaders:
def validation_step(self, batch, batch_nb, dataloader_idxdx)
def validation_step(self, batch, batch_idx, dataloader_idxdx)

If you don't need to validate you don't need to implement this method.
In this step you'd normally generate examples or calculate anything of interest such as accuracy.
Expand All @@ -275,7 +275,7 @@ def validation_step(self, batch, batch_nb, dataloader_idxdx)
.. code-block:: python

# CASE 1: A single validation dataset
def validation_step(self, batch, batch_nb):
def validation_step(self, batch, batch_idx):
x, y = batch

# implement your own
Expand Down Expand Up @@ -307,7 +307,7 @@ def validation_step(self, batch, batch_nb):
.. code-block:: python

# CASE 2: multiple validation datasets
def validation_step(self, batch, batch_nb, dataset_idx):
def validation_step(self, batch, batch_idx, dataset_idx):
# dataset_idx tells you which dataset this is.

The `dataset_idx` corresponds to the order of datasets returned in `val_dataloader`.
Expand All @@ -318,17 +318,17 @@ def test_step(self, *args, **kwargs):
"""return whatever outputs will need to be aggregated in test_end

:param batch: The output of your dataloader. A tensor, tuple or list
:param int batch_nb: Integer displaying which batch this is
:param int batch_idx: Integer displaying which batch this is
:param int dataloader_idx: Integer displaying which dataloader this is (only if multiple test datasets used)
:return dict: Dict or OrderedDict with metrics to display in progress bar. All keys must be tensors.

.. code-block:: python

# if you have one test dataloader:
def test_step(self, batch, batch_nb)
def test_step(self, batch, batch_idx)

# if you have multiple test dataloaders:
def test_step(self, batch, batch_nb, dataloader_idxdx)
def test_step(self, batch, batch_idx, dataloader_idxdx)


**OPTIONAL**
Expand All @@ -348,7 +348,7 @@ def test_step(self, batch, batch_nb, dataloader_idxdx)
.. code-block:: python

# CASE 1: A single test dataset
def test_step(self, batch, batch_nb):
def test_step(self, batch, batch_idx):
x, y = batch

# implement your own
Expand All @@ -375,7 +375,7 @@ def test_step(self, batch, batch_nb):
.. code-block:: python

# CASE 2: multiple test datasets
def test_step(self, batch, batch_nb, dataset_idx):
def test_step(self, batch, batch_idx, dataset_idx):
# dataset_idx tells you which dataset this is.


Expand Down Expand Up @@ -694,13 +694,13 @@ def configure_optimizers(self):
"""
raise NotImplementedError

def optimizer_step(self, epoch_nb, batch_nb, optimizer, optimizer_i, second_order_closure=None):
def optimizer_step(self, epoch_idx, batch_idx, optimizer, optimizer_idx, second_order_closure=None):
"""Do something instead of the standard optimizer behavior

:param int epoch_nb:
:param int batch_nb:
:param int epoch_idx:
:param int batch_idx:
:param optimizer:
:param optimizer_i:
:param optimizer_idx:
:param second_order_closure: closure for second order methods
:return:

Expand All @@ -712,21 +712,21 @@ def optimizer_step(self, epoch_nb, batch_nb, optimizer, optimizer_i, second_orde
.. code-block:: python

# DEFAULT
def optimizer_step(self, current_epoch, batch_nb, optimizer, optimizer_i, second_order_closure=None):
def optimizer_step(self, current_epoch, batch_idx, optimizer, optimizer_idx, second_order_closure=None):
optimizer.step()
optimizer.zero_grad()

# Alternating schedule for optimizer steps (ie: GANs)
def optimizer_step(self, current_epoch, batch_nb, optimizer, optimizer_i, second_order_closure=None):
def optimizer_step(self, current_epoch, batch_idx, optimizer, optimizer_idx, second_order_closure=None):
# update generator opt every 2 steps
if optimizer_i == 0:
if batch_nb % 2 == 0 :
if optimizer_idx == 0:
if batch_idx % 2 == 0 :
optimizer.step()
optimizer.zero_grad()

# update discriminator opt every 4 steps
if optimizer_i == 1:
if batch_nb % 4 == 0 :
if optimizer_idx == 1:
if batch_idx % 4 == 0 :
optimizer.step()
optimizer.zero_grad()

Expand All @@ -739,7 +739,7 @@ def optimizer_step(self, current_epoch, batch_nb, optimizer, optimizer_i, second
.. code-block:: python

# learning rate warm-up
def optimizer_step(self, current_epoch, batch_nb, optimizer, optimizer_i, second_order_closure=None):
def optimizer_step(self, current_epoch, batch_idx, optimizer, optimizer_idx, second_order_closure=None):
# warm up lr
if self.trainer.global_step < 500:
lr_scale = min(1., float(self.trainer.global_step + 1) / 500.)
Expand Down
Loading