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%|          | 164k/26.4M [00:00<00:56, 465kB/s]
  3%|▎         | 688k/26.4M [00:00<00:16, 1.57MB/s]
 11%|█         | 2.82M/26.4M [00:00<00:04, 5.55MB/s]
 28%|██▊       | 7.27M/26.4M [00:00<00:01, 12.3MB/s]
 49%|████▊     | 12.8M/26.4M [00:01<00:00, 18.5MB/s]
 71%|███████   | 18.8M/26.4M [00:01<00:00, 23.1MB/s]
 94%|█████████▍| 24.8M/26.4M [00:01<00:00, 26.1MB/s]
100%|██████████| 26.4M/26.4M [00:01<00:00, 18.0MB/s]

  0%|          | 0.00/29.5k [00:00<?, ?B/s]
100%|██████████| 29.5k/29.5k [00:00<00:00, 326kB/s]

  0%|          | 0.00/4.42M [00:00<?, ?B/s]
  1%|▏         | 65.5k/4.42M [00:00<00:12, 355kB/s]
  5%|▌         | 229k/4.42M [00:00<00:06, 669kB/s]
 19%|█▉        | 852k/4.42M [00:00<00:01, 1.91MB/s]
 79%|███████▊  | 3.47M/4.42M [00:00<00:00, 6.76MB/s]
100%|██████████| 4.42M/4.42M [00:00<00:00, 5.97MB/s]

  0%|          | 0.00/5.15k [00:00<?, ?B/s]
100%|██████████| 5.15k/5.15k [00:00<00:00, 51.3MB/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)))

Dress  Bag  Sneaker  Trouser

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([[0.2615, 0.0680, 0.6463, 0.9937, 0.3504, 0.8821, 0.2427, 0.1258, 0.0305,
         0.0822],
        [0.0944, 0.4151, 0.5707, 0.0673, 0.4398, 0.1267, 0.8395, 0.4250, 0.9845,
         0.7354],
        [0.1820, 0.2709, 0.7457, 0.0565, 0.8342, 0.0041, 0.3880, 0.5764, 0.5443,
         0.1876],
        [0.3566, 0.1779, 0.9328, 0.2806, 0.2315, 0.8483, 0.2551, 0.7624, 0.0552,
         0.9620]])
tensor([1, 5, 3, 7])
Total loss for this batch: 2.5246968269348145

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: 1.8637439629882575
  batch 2000 loss: 0.9264033259702846
  batch 3000 loss: 0.7004553715707734
  batch 4000 loss: 0.6298066498190165
  batch 5000 loss: 0.5766402523722499
  batch 6000 loss: 0.5501439673268469
  batch 7000 loss: 0.5175673807188869
  batch 8000 loss: 0.5097704701682669
  batch 9000 loss: 0.4745651287258952
  batch 10000 loss: 0.4611728880517767
  batch 11000 loss: 0.43946013386838606
  batch 12000 loss: 0.4464166041428325
  batch 13000 loss: 0.43270455794787266
  batch 14000 loss: 0.4093883478245698
  batch 15000 loss: 0.41265470427821854
LOSS train 0.41265470427821854 valid 0.4659363329410553
EPOCH 2:
  batch 1000 loss: 0.39029235641355625
  batch 2000 loss: 0.4026143496113946
  batch 3000 loss: 0.38099835417387656
  batch 4000 loss: 0.39694612299313303
  batch 5000 loss: 0.3830260536352289
  batch 6000 loss: 0.3749699195966532
  batch 7000 loss: 0.3587064477120002
  batch 8000 loss: 0.3500676108513144
  batch 9000 loss: 0.37422816658200464
  batch 10000 loss: 0.34804187821148663
  batch 11000 loss: 0.3531418056777038
  batch 12000 loss: 0.3531652553510066
  batch 13000 loss: 0.3621450867006788
  batch 14000 loss: 0.36119725092485894
  batch 15000 loss: 0.3381529517005838
LOSS train 0.3381529517005838 valid 0.40801024436950684
EPOCH 3:
  batch 1000 loss: 0.3441478884675453
  batch 2000 loss: 0.3291099863762647
  batch 3000 loss: 0.33744279164131513
  batch 4000 loss: 0.3248490593174647
  batch 5000 loss: 0.31116376548158586
  batch 6000 loss: 0.3203691777046479
  batch 7000 loss: 0.3257730000029842
  batch 8000 loss: 0.3269687126284116
  batch 9000 loss: 0.323592563013035
  batch 10000 loss: 0.31044773180341145
  batch 11000 loss: 0.3265057227806246
  batch 12000 loss: 0.3132079298844765
  batch 13000 loss: 0.31345399576899946
  batch 14000 loss: 0.31480225349571267
  batch 15000 loss: 0.31691658809121875
LOSS train 0.31691658809121875 valid 0.3407316505908966
EPOCH 4:
  batch 1000 loss: 0.2950099313746032
  batch 2000 loss: 0.29936983725682603
  batch 3000 loss: 0.293884304475534
  batch 4000 loss: 0.28607402624508177
  batch 5000 loss: 0.29341298843484176
  batch 6000 loss: 0.28971555939753124
  batch 7000 loss: 0.30780316224124543
  batch 8000 loss: 0.3036485634235214
  batch 9000 loss: 0.2864140385771243
  batch 10000 loss: 0.2769247607487196
  batch 11000 loss: 0.29243525436209167
  batch 12000 loss: 0.292175021926676
  batch 13000 loss: 0.2860977888671914
  batch 14000 loss: 0.3126203253843705
  batch 15000 loss: 0.2975845007579919
LOSS train 0.2975845007579919 valid 0.31715095043182373
EPOCH 5:
  batch 1000 loss: 0.2591698765270667
  batch 2000 loss: 0.27572174825237017
  batch 3000 loss: 0.2641295134468091
  batch 4000 loss: 0.2653548823763576
  batch 5000 loss: 0.2840304220055168
  batch 6000 loss: 0.27457090629893355
  batch 7000 loss: 0.2623442139832114
  batch 8000 loss: 0.28994368762509837
  batch 9000 loss: 0.26480666732649116
  batch 10000 loss: 0.309013732829153
  batch 11000 loss: 0.281807249325444
  batch 12000 loss: 0.2799575989842451
  batch 13000 loss: 0.2642146244214964
  batch 14000 loss: 0.27759422915820325
  batch 15000 loss: 0.2773170724936281
LOSS train 0.2773170724936281 valid 0.3220783472061157

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