beginner/introyt/trainingyt

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

Introduction || Tensors || Autograd || Building Models || TensorBoard Support || Training Models || Model Understanding

Training with PyTorch

Created On: Nov 30, 2021 | Last Updated: May 31, 2023 | Last Verified: Nov 05, 2024

Follow along with the video below or on youtube.

Introduction

In past videos, we’ve discussed and demonstrated:

• Building models with the neural network layers and functions of the torch.nn module

• The mechanics of automated gradient computation, which is central to gradient-based model training

• Using TensorBoard to visualize training progress and other activities

In this video, we’ll be adding some new tools to your inventory:

• We’ll get familiar with the dataset and dataloader abstractions, and how they ease the process of feeding data to your model during a training loop

• We’ll discuss specific loss functions and when to use them

• We’ll look at PyTorch optimizers, which implement algorithms to adjust model weights based on the outcome of a loss function

Finally, we’ll pull all of these together and see a full PyTorch training loop in action.

Dataset and DataLoader

The Dataset and DataLoader classes encapsulate the process of pulling your data from storage and exposing it to your training loop in batches.

The Dataset is responsible for accessing and processing single instances of data.

The DataLoader pulls instances of data from the Dataset (either automatically or with a sampler that you define), collects them in batches, and returns them for consumption by your training loop. The DataLoader works with all kinds of datasets, regardless of the type of data they contain.

For this tutorial, we’ll be using the Fashion-MNIST dataset provided by TorchVision. We use torchvision.transforms.Normalize() to zero-center and normalize the distribution of the image tile content, and download both training and validation data splits.

import torch
import torchvision
import torchvision.transforms as transforms

# PyTorch TensorBoard support
from torch.utils.tensorboard import SummaryWriter
from datetime import datetime


transform = transforms.Compose(
    [transforms.ToTensor(),
    transforms.Normalize((0.5,), (0.5,))])

# Create datasets for training & validation, download if necessary
training_set = torchvision.datasets.FashionMNIST('./data', train=True, transform=transform, download=True)
validation_set = torchvision.datasets.FashionMNIST('./data', train=False, transform=transform, download=True)

# Create data loaders for our datasets; shuffle for training, not for validation
training_loader = torch.utils.data.DataLoader(training_set, batch_size=4, shuffle=True)
validation_loader = torch.utils.data.DataLoader(validation_set, batch_size=4, shuffle=False)

# Class labels
classes = ('T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',
        'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle Boot')

# Report split sizes
print('Training set has {} instances'.format(len(training_set)))
print('Validation set has {} instances'.format(len(validation_set)))

  0%|          | 0.00/26.4M [00:00<?, ?B/s]
  0%|          | 65.5k/26.4M [00:00<01:13, 360kB/s]
  1%|          | 229k/26.4M [00:00<00:38, 674kB/s]
  3%|▎         | 918k/26.4M [00:00<00:12, 2.08MB/s]
 14%|█▍        | 3.67M/26.4M [00:00<00:03, 7.18MB/s]
 37%|███▋      | 9.83M/26.4M [00:00<00:00, 16.7MB/s]
 60%|█████▉    | 15.8M/26.4M [00:01<00:00, 22.0MB/s]
 82%|████████▏ | 21.7M/26.4M [00:01<00:00, 25.3MB/s]
100%|██████████| 26.4M/26.4M [00:01<00:00, 19.1MB/s]

  0%|          | 0.00/29.5k [00:00<?, ?B/s]
100%|██████████| 29.5k/29.5k [00:00<00:00, 323kB/s]

  0%|          | 0.00/4.42M [00:00<?, ?B/s]
  1%|▏         | 65.5k/4.42M [00:00<00:12, 360kB/s]
  4%|▍         | 197k/4.42M [00:00<00:07, 575kB/s]
 19%|█▊        | 819k/4.42M [00:00<00:01, 1.88MB/s]
 74%|███████▍  | 3.28M/4.42M [00:00<00:00, 6.49MB/s]
100%|██████████| 4.42M/4.42M [00:00<00:00, 6.06MB/s]

  0%|          | 0.00/5.15k [00:00<?, ?B/s]
100%|██████████| 5.15k/5.15k [00:00<00:00, 52.8MB/s]
Training set has 60000 instances
Validation set has 10000 instances

As always, let’s visualize the data as a sanity check:

import matplotlib.pyplot as plt
import numpy as np

# Helper function for inline image display
def matplotlib_imshow(img, one_channel=False):
    if one_channel:
        img = img.mean(dim=0)
    img = img / 2 + 0.5     # unnormalize
    npimg = img.numpy()
    if one_channel:
        plt.imshow(npimg, cmap="Greys")
    else:
        plt.imshow(np.transpose(npimg, (1, 2, 0)))

dataiter = iter(training_loader)
images, labels = next(dataiter)

# Create a grid from the images and show them
img_grid = torchvision.utils.make_grid(images)
matplotlib_imshow(img_grid, one_channel=True)
print('  '.join(classes[labels[j]] for j in range(4)))

Coat  Sneaker  Dress  Dress

The Model

The model we’ll use in this example is a variant of LeNet-5 - it should be familiar if you’ve watched the previous videos in this series.

import torch.nn as nn
import torch.nn.functional as F

# PyTorch models inherit from torch.nn.Module
class GarmentClassifier(nn.Module):
    def __init__(self):
        super(GarmentClassifier, self).__init__()
        self.conv1 = nn.Conv2d(1, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16 * 4 * 4, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)

    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = x.view(-1, 16 * 4 * 4)
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x


model = GarmentClassifier()

Loss Function

For this example, we’ll be using a cross-entropy loss. For demonstration purposes, we’ll create batches of dummy output and label values, run them through the loss function, and examine the result.

loss_fn = torch.nn.CrossEntropyLoss()

# NB: Loss functions expect data in batches, so we're creating batches of 4
# Represents the model's confidence in each of the 10 classes for a given input
dummy_outputs = torch.rand(4, 10)
# Represents the correct class among the 10 being tested
dummy_labels = torch.tensor([1, 5, 3, 7])

print(dummy_outputs)
print(dummy_labels)

loss = loss_fn(dummy_outputs, dummy_labels)
print('Total loss for this batch: {}'.format(loss.item()))

tensor([[0.4271, 0.6384, 0.3452, 0.7453, 0.4921, 0.3817, 0.5209, 0.0885, 0.9970,
         0.9892],
        [0.8800, 0.9021, 0.6677, 0.9624, 0.0349, 0.5844, 0.1598, 0.6857, 0.1892,
         0.1251],
        [0.2240, 0.7840, 0.8938, 0.4519, 0.1508, 0.3038, 0.8100, 0.1413, 0.4139,
         0.4835],
        [0.9473, 0.6924, 0.9306, 0.2292, 0.6798, 0.8082, 0.4483, 0.2938, 0.5292,
         0.7377]])
tensor([1, 5, 3, 7])
Total loss for this batch: 2.3937389850616455

Optimizer

For this example, we’ll be using simple stochastic gradient descent with momentum.

It can be instructive to try some variations on this optimization scheme:

• Learning rate determines the size of the steps the optimizer takes. What does a different learning rate do to the your training results, in terms of accuracy and convergence time?

• Momentum nudges the optimizer in the direction of strongest gradient over multiple steps. What does changing this value do to your results?

• Try some different optimization algorithms, such as averaged SGD, Adagrad, or Adam. How do your results differ?

# Optimizers specified in the torch.optim package
optimizer = torch.optim.SGD(model.parameters(), lr=0.001, momentum=0.9)

The Training Loop

Below, we have a function that performs one training epoch. It enumerates data from the DataLoader, and on each pass of the loop does the following:

• Gets a batch of training data from the DataLoader

• Zeros the optimizer’s gradients

• Performs an inference - that is, gets predictions from the model for an input batch

• Calculates the loss for that set of predictions vs. the labels on the dataset

• Calculates the backward gradients over the learning weights

• Tells the optimizer to perform one learning step - that is, adjust the model’s learning weights based on the observed gradients for this batch, according to the optimization algorithm we chose

• It reports on the loss for every 1000 batches.

• Finally, it reports the average per-batch loss for the last 1000 batches, for comparison with a validation run

def train_one_epoch(epoch_index, tb_writer):
    running_loss = 0.
    last_loss = 0.

    # Here, we use enumerate(training_loader) instead of
    # iter(training_loader) so that we can track the batch
    # index and do some intra-epoch reporting
    for i, data in enumerate(training_loader):
        # Every data instance is an input + label pair
        inputs, labels = data

        # Zero your gradients for every batch!
        optimizer.zero_grad()

        # Make predictions for this batch
        outputs = model(inputs)

        # Compute the loss and its gradients
        loss = loss_fn(outputs, labels)
        loss.backward()

        # Adjust learning weights
        optimizer.step()

        # Gather data and report
        running_loss += loss.item()
        if i % 1000 == 999:
            last_loss = running_loss / 1000 # loss per batch
            print('  batch {} loss: {}'.format(i + 1, last_loss))
            tb_x = epoch_index * len(training_loader) + i + 1
            tb_writer.add_scalar('Loss/train', last_loss, tb_x)
            running_loss = 0.

    return last_loss

Per-Epoch Activity

There are a couple of things we’ll want to do once per epoch:

• Perform validation by checking our relative loss on a set of data that was not used for training, and report this

• Save a copy of the model

Here, we’ll do our reporting in TensorBoard. This will require going to the command line to start TensorBoard, and opening it in another browser tab.

# Initializing in a separate cell so we can easily add more epochs to the same run
timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
writer = SummaryWriter('runs/fashion_trainer_{}'.format(timestamp))
epoch_number = 0

EPOCHS = 5

best_vloss = 1_000_000.

for epoch in range(EPOCHS):
    print('EPOCH {}:'.format(epoch_number + 1))

    # Make sure gradient tracking is on, and do a pass over the data
    model.train(True)
    avg_loss = train_one_epoch(epoch_number, writer)


    running_vloss = 0.0
    # Set the model to evaluation mode, disabling dropout and using population
    # statistics for batch normalization.
    model.eval()

    # Disable gradient computation and reduce memory consumption.
    with torch.no_grad():
        for i, vdata in enumerate(validation_loader):
            vinputs, vlabels = vdata
            voutputs = model(vinputs)
            vloss = loss_fn(voutputs, vlabels)
            running_vloss += vloss

    avg_vloss = running_vloss / (i + 1)
    print('LOSS train {} valid {}'.format(avg_loss, avg_vloss))

    # Log the running loss averaged per batch
    # for both training and validation
    writer.add_scalars('Training vs. Validation Loss',
                    { 'Training' : avg_loss, 'Validation' : avg_vloss },
                    epoch_number + 1)
    writer.flush()

    # Track best performance, and save the model's state
    if avg_vloss < best_vloss:
        best_vloss = avg_vloss
        model_path = 'model_{}_{}'.format(timestamp, epoch_number)
        torch.save(model.state_dict(), model_path)

    epoch_number += 1

EPOCH 1:
  batch 1000 loss: 1.784976828455925
  batch 2000 loss: 0.8263223152449355
  batch 3000 loss: 0.6916487076813355
  batch 4000 loss: 0.6096141297444702
  batch 5000 loss: 0.569686681009829
  batch 6000 loss: 0.5379622910583858
  batch 7000 loss: 0.5362045865270775
  batch 8000 loss: 0.5039491483343299
  batch 9000 loss: 0.4627635038787266
  batch 10000 loss: 0.46716841775504875
  batch 11000 loss: 0.42577900271408725
  batch 12000 loss: 0.4293474618118489
  batch 13000 loss: 0.41360994360793846
  batch 14000 loss: 0.4013340015343856
  batch 15000 loss: 0.394765554038866
LOSS train 0.394765554038866 valid 0.422852098941803
EPOCH 2:
  batch 1000 loss: 0.38278596836034556
  batch 2000 loss: 0.38892397124494893
  batch 3000 loss: 0.3666358025862137
  batch 4000 loss: 0.3774275084128603
  batch 5000 loss: 0.3842072506857658
  batch 6000 loss: 0.3682632900656899
  batch 7000 loss: 0.3571386761087924
  batch 8000 loss: 0.3769625986551691
  batch 9000 loss: 0.34585123410601226
  batch 10000 loss: 0.349092586322964
  batch 11000 loss: 0.332665092885989
  batch 12000 loss: 0.33922079248694353
  batch 13000 loss: 0.3508776856112818
  batch 14000 loss: 0.3217826355858124
  batch 15000 loss: 0.3316082783928578
LOSS train 0.3316082783928578 valid 0.3533305525779724
EPOCH 3:
  batch 1000 loss: 0.3101922840481275
  batch 2000 loss: 0.32035058495154956
  batch 3000 loss: 0.32936868814887565
  batch 4000 loss: 0.3129736438259133
  batch 5000 loss: 0.3316463240184603
  batch 6000 loss: 0.33540282097583984
  batch 7000 loss: 0.30903379454429525
  batch 8000 loss: 0.3482079836736375
  batch 9000 loss: 0.29847264505916976
  batch 10000 loss: 0.3172011704890465
  batch 11000 loss: 0.3049195747375343
  batch 12000 loss: 0.297806606514172
  batch 13000 loss: 0.30823753245500846
  batch 14000 loss: 0.32453707323713754
  batch 15000 loss: 0.30613208849704826
LOSS train 0.30613208849704826 valid 0.3252071142196655
EPOCH 4:
  batch 1000 loss: 0.3054257575027441
  batch 2000 loss: 0.298001502304327
  batch 3000 loss: 0.3014335927081065
  batch 4000 loss: 0.29491765104489603
  batch 5000 loss: 0.2821206657881139
  batch 6000 loss: 0.2905257716884571
  batch 7000 loss: 0.2931791045982227
  batch 8000 loss: 0.308601964775251
  batch 9000 loss: 0.304633750305773
  batch 10000 loss: 0.2945715122923066
  batch 11000 loss: 0.2919909181561852
  batch 12000 loss: 0.3059344952407264
  batch 13000 loss: 0.28693745871676946
  batch 14000 loss: 0.26961352903507757
  batch 15000 loss: 0.2736693370296089
LOSS train 0.2736693370296089 valid 0.3376033306121826
EPOCH 5:
  batch 1000 loss: 0.26626920011397304
  batch 2000 loss: 0.2704936440042184
  batch 3000 loss: 0.2808960044778214
  batch 4000 loss: 0.27183746823063937
  batch 5000 loss: 0.27171938311338456
  batch 6000 loss: 0.27300490664795507
  batch 7000 loss: 0.2686056153687168
  batch 8000 loss: 0.26906107153145875
  batch 9000 loss: 0.28043227519997527
  batch 10000 loss: 0.2656501787790912
  batch 11000 loss: 0.27493628575908213
  batch 12000 loss: 0.28932940587129635
  batch 13000 loss: 0.27727131318160353
  batch 14000 loss: 0.28784957085908175
  batch 15000 loss: 0.28634940097837536
LOSS train 0.28634940097837536 valid 0.3038865923881531

To load a saved version of the model:

saved_model = GarmentClassifier()
saved_model.load_state_dict(torch.load(PATH))

Once you’ve loaded the model, it’s ready for whatever you need it for - more training, inference, or analysis.

Note that if your model has constructor parameters that affect model structure, you’ll need to provide them and configure the model identically to the state in which it was saved.

Other Resources

• Docs on the data utilities, including Dataset and DataLoader, at pytorch.org

• A note on the use of pinned memory for GPU training

• Documentation on the datasets available in TorchVision, TorchText, and TorchAudio

• Documentation on the loss functions available in PyTorch

• Documentation on the torch.optim package, which includes optimizers and related tools, such as learning rate scheduling

• A detailed tutorial on saving and loading models

• The Tutorials section of pytorch.org contains tutorials on a broad variety of training tasks, including classification in different domains, generative adversarial networks, reinforcement learning, and more