Introduction to Pytorch Lightning¶
Author: PL team
License: CC BY-SA
Generated: 2021-07-26T23:14:44.105855
In this notebook, we’ll go over the basics of lightning by preparing models to train on the MNIST Handwritten Digits dataset.
Open in 
Give us a ⭐ on Github | Check out the documentation | Join us on Slack
Setup¶
This notebook requires some packages besides pytorch-lightning.
[1]:
! pip install --quiet "pytorch-lightning>=1.3" "torchvision" "torchmetrics>=0.3" "torch>=1.6, <1.9"
WARNING: Value for scheme.platlib does not match. Please report this to <https://github.com/pypa/pip/issues/10151>
distutils: /usr/local/lib/python3.9/dist-packages
sysconfig: /usr/lib/python3.9/site-packages
WARNING: Value for scheme.purelib does not match. Please report this to <https://github.com/pypa/pip/issues/10151>
distutils: /usr/local/lib/python3.9/dist-packages
sysconfig: /usr/lib/python3.9/site-packages
WARNING: Value for scheme.headers does not match. Please report this to <https://github.com/pypa/pip/issues/10151>
distutils: /usr/local/include/python3.9/UNKNOWN
sysconfig: /usr/include/python3.9/UNKNOWN
WARNING: Value for scheme.scripts does not match. Please report this to <https://github.com/pypa/pip/issues/10151>
distutils: /usr/local/bin
sysconfig: /usr/bin
WARNING: Value for scheme.data does not match. Please report this to <https://github.com/pypa/pip/issues/10151>
distutils: /usr/local
sysconfig: /usr
WARNING: Additional context:
user = False
home = None
root = None
prefix = None
[2]:
import os
import torch
from pytorch_lightning import LightningModule, Trainer
from pytorch_lightning.metrics.functional import accuracy
from torch import nn
from torch.nn import functional as F
from torch.utils.data import DataLoader, random_split
from torchvision import transforms
from torchvision.datasets import MNIST
PATH_DATASETS = os.environ.get('PATH_DATASETS', '.')
AVAIL_GPUS = min(1, torch.cuda.device_count())
BATCH_SIZE = 256 if AVAIL_GPUS else 64
Simplest example¶
Here’s the simplest most minimal example with just a training loop (no validation, no testing).
Keep in Mind - A LightningModule is a PyTorch nn.Module - it just has a few more helpful features.
[3]:
class MNISTModel(LightningModule):
def __init__(self):
super().__init__()
self.l1 = torch.nn.Linear(28 * 28, 10)
def forward(self, x):
return torch.relu(self.l1(x.view(x.size(0), -1)))
def training_step(self, batch, batch_nb):
x, y = batch
loss = F.cross_entropy(self(x), y)
return loss
def configure_optimizers(self):
return torch.optim.Adam(self.parameters(), lr=0.02)
By using the Trainer you automatically get: 1. Tensorboard logging 2. Model checkpointing 3. Training and validation loop 4. early-stopping
[4]:
# Init our model
mnist_model = MNISTModel()
# Init DataLoader from MNIST Dataset
train_ds = MNIST(PATH_DATASETS, train=True, download=True, transform=transforms.ToTensor())
train_loader = DataLoader(train_ds, batch_size=BATCH_SIZE)
# Initialize a trainer
trainer = Trainer(
gpus=AVAIL_GPUS,
max_epochs=3,
progress_bar_refresh_rate=20,
)
# Train the model ⚡
trainer.fit(mnist_model, train_loader)
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]
| Name | Type | Params
--------------------------------
0 | l1 | Linear | 7.9 K
--------------------------------
7.9 K Trainable params
0 Non-trainable params
7.9 K Total params
0.031 Total estimated model params size (MB)
/home/AzDevOps_azpcontainer/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/data_loading.py:102: UserWarning: The dataloader, train dataloader, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 12 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.
rank_zero_warn(
A more complete MNIST Lightning Module Example¶
That wasn’t so hard was it?
Now that we’ve got our feet wet, let’s dive in a bit deeper and write a more complete LightningModule for MNIST…
This time, we’ll bake in all the dataset specific pieces directly in the LightningModule. This way, we can avoid writing extra code at the beginning of our script every time we want to run it.
Note what the following built-in functions are doing:¶
-
This is where we can download the dataset. We point to our desired dataset and ask torchvision’s
MNISTdataset class to download if the dataset isn’t found there.Note we do not make any state assignments in this function (i.e.
self.something = ...)
setup(stage) ⚙️
Loads in data from file and prepares PyTorch tensor datasets for each split (train, val, test).
Setup expects a ‘stage’ arg which is used to separate logic for ‘fit’ and ‘test’.
If you don’t mind loading all your datasets at once, you can set up a condition to allow for both ‘fit’ related setup and ‘test’ related setup to run whenever
Noneis passed tostage(or ignore it altogether and exclude any conditionals).Note this runs across all GPUs and it *is* safe to make state assignments here
-
train_dataloader(),val_dataloader(), andtest_dataloader()all return PyTorchDataLoaderinstances that are created by wrapping their respective datasets that we prepared insetup()
[5]:
class LitMNIST(LightningModule):
def __init__(self, data_dir=PATH_DATASETS, hidden_size=64, learning_rate=2e-4):
super().__init__()
# Set our init args as class attributes
self.data_dir = data_dir
self.hidden_size = hidden_size
self.learning_rate = learning_rate
# Hardcode some dataset specific attributes
self.num_classes = 10
self.dims = (1, 28, 28)
channels, width, height = self.dims
self.transform = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, )),
])
# Define PyTorch model
self.model = nn.Sequential(
nn.Flatten(),
nn.Linear(channels * width * height, hidden_size),
nn.ReLU(),
nn.Dropout(0.1),
nn.Linear(hidden_size, hidden_size),
nn.ReLU(),
nn.Dropout(0.1),
nn.Linear(hidden_size, self.num_classes),
)
def forward(self, x):
x = self.model(x)
return F.log_softmax(x, dim=1)
def training_step(self, batch, batch_idx):
x, y = batch
logits = self(x)
loss = F.nll_loss(logits, y)
return loss
def validation_step(self, batch, batch_idx):
x, y = batch
logits = self(x)
loss = F.nll_loss(logits, y)
preds = torch.argmax(logits, dim=1)
acc = accuracy(preds, y)
# Calling self.log will surface up scalars for you in TensorBoard
self.log('val_loss', loss, prog_bar=True)
self.log('val_acc', acc, prog_bar=True)
return loss
def test_step(self, batch, batch_idx):
# Here we just reuse the validation_step for testing
return self.validation_step(batch, batch_idx)
def configure_optimizers(self):
optimizer = torch.optim.Adam(self.parameters(), lr=self.learning_rate)
return optimizer
####################
# DATA RELATED HOOKS
####################
def prepare_data(self):
# download
MNIST(self.data_dir, train=True, download=True)
MNIST(self.data_dir, train=False, download=True)
def setup(self, stage=None):
# Assign train/val datasets for use in dataloaders
if stage == 'fit' or stage is None:
mnist_full = MNIST(self.data_dir, train=True, transform=self.transform)
self.mnist_train, self.mnist_val = random_split(mnist_full, [55000, 5000])
# Assign test dataset for use in dataloader(s)
if stage == 'test' or stage is None:
self.mnist_test = MNIST(self.data_dir, train=False, transform=self.transform)
def train_dataloader(self):
return DataLoader(self.mnist_train, batch_size=BATCH_SIZE)
def val_dataloader(self):
return DataLoader(self.mnist_val, batch_size=BATCH_SIZE)
def test_dataloader(self):
return DataLoader(self.mnist_test, batch_size=BATCH_SIZE)
[6]:
model = LitMNIST()
trainer = Trainer(
gpus=AVAIL_GPUS,
max_epochs=3,
progress_bar_refresh_rate=20,
)
trainer.fit(model)
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]
| Name | Type | Params
-------------------------------------
0 | model | Sequential | 55.1 K
-------------------------------------
55.1 K Trainable params
0 Non-trainable params
55.1 K Total params
0.220 Total estimated model params size (MB)
/home/AzDevOps_azpcontainer/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/data_loading.py:102: UserWarning: The dataloader, val dataloader 0, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 12 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.
rank_zero_warn(
/home/AzDevOps_azpcontainer/.local/lib/python3.9/site-packages/deprecate/deprecation.py:115: LightningDeprecationWarning: The `accuracy` was deprecated since v1.3.0 in favor of `torchmetrics.functional.classification.accuracy.accuracy`. It will be removed in v1.5.0.
stream(template_mgs % msg_args)
/home/AzDevOps_azpcontainer/.local/lib/python3.9/site-packages/pytorch_lightning/callbacks/model_checkpoint.py:610: LightningDeprecationWarning: Relying on `self.log('val_loss', ...)` to set the ModelCheckpoint monitor is deprecated in v1.2 and will be removed in v1.4. Please, create your own `mc = ModelCheckpoint(monitor='your_monitor')` and use it as `Trainer(callbacks=[mc])`.
warning_cache.deprecation(
Testing¶
To test a model, call trainer.test(model).
Or, if you’ve just trained a model, you can just call trainer.test() and Lightning will automatically test using the best saved checkpoint (conditioned on val_loss).
[7]:
trainer.test()
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]
/home/AzDevOps_azpcontainer/.local/lib/python3.9/site-packages/pytorch_lightning/trainer/data_loading.py:102: UserWarning: The dataloader, test dataloader 0, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 12 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.
rank_zero_warn(
--------------------------------------------------------------------------------
DATALOADER:0 TEST RESULTS
{'val_acc': 0.9236999750137329, 'val_loss': 0.2604999840259552}
--------------------------------------------------------------------------------
[7]:
[{'val_loss': 0.2604999840259552, 'val_acc': 0.9236999750137329}]
Bonus Tip¶
You can keep calling trainer.fit(model) as many times as you’d like to continue training
[8]:
trainer.fit(model)
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1]
| Name | Type | Params
-------------------------------------
0 | model | Sequential | 55.1 K
-------------------------------------
55.1 K Trainable params
0 Non-trainable params
55.1 K Total params
0.220 Total estimated model params size (MB)
In Colab, you can use the TensorBoard magic function to view the logs that Lightning has created for you!
[9]:
# Start tensorboard.
%load_ext tensorboard
%tensorboard --logdir lightning_logs/
Congratulations - Time to Join the Community!¶
Congratulations on completing this notebook tutorial! If you enjoyed this and would like to join the Lightning movement, you can do so in the following ways!
Star Lightning on GitHub¶
The easiest way to help our community is just by starring the GitHub repos! This helps raise awareness of the cool tools we’re building.
Join our Slack!¶
The best way to keep up to date on the latest advancements is to join our community! Make sure to introduce yourself and share your interests in #general channel
Contributions !¶
The best way to contribute to our community is to become a code contributor! At any time you can go to Lightning or Bolt GitHub Issues page and filter for “good first issue”.
You can also contribute your own notebooks with useful examples !
Great thanks from the entire Pytorch Lightning Team for your interest !¶
