Computer Vision 101 – Fashion MNIST

0369

This is a basic computer vision example using PyTorch.
What you’ll find here :

  1. What is computer vision
  2. A simple classification example using a deep learning computer vision model

Without further ado lets begin… (before jumping right into the code please understand what computer vision is)

What is Computer Vision

Computer vision is a field of computer science that focuses on building intelligent systems that can process, analyse and and derive information from visual data like images, videos. Lately people are achieving marvellous things using 3D point cloud information…

FASHION MNIST CLASSIFICATION

Link to the dataset : Fashion Mnist
You dont need to download the dataset manually, they are included as part of pytorch
Its better if you use jupyter-notebook as the code in this blog is a step by step process with data visualisation in between for better understanding. 

<span>import</span> <span>torch</span>
<span>import</span> <span>torchvision</span>
<span>import</span> <span>numpy</span> <span>as</span> <span>np</span>
<span>import</span> <span>matplotlib.pyplot</span> <span>as</span> <span>plt</span>
<span>import</span> <span>torch.nn</span> <span>as</span> <span>nn</span>
<span>import</span> <span>torch.nn.functional</span> <span>as</span> <span>F</span>
<span>from</span> <span>torchvision.datasets</span> <span>import</span> <span>FashionMNIST</span>
<span>from</span> <span>torchvision.transforms</span> <span>import</span> <span>ToTensor</span>
<span>from</span> <span>torchvision.utils</span> <span>import</span> <span>make_grid</span>
<span>from</span> <span>torch.utils.data.dataloader</span> <span>import</span> <span>DataLoader</span>
<span>from</span> <span>torch.utils.data</span> <span>import</span> <span>random_split</span>
<span>import</span> <span>torch</span>
<span>import</span> <span>torchvision</span>
<span>import</span> <span>numpy</span> <span>as</span> <span>np</span>
<span>import</span> <span>matplotlib.pyplot</span> <span>as</span> <span>plt</span>
<span>import</span> <span>torch.nn</span> <span>as</span> <span>nn</span>
<span>import</span> <span>torch.nn.functional</span> <span>as</span> <span>F</span>
<span>from</span> <span>torchvision.datasets</span> <span>import</span> <span>FashionMNIST</span>
<span>from</span> <span>torchvision.transforms</span> <span>import</span> <span>ToTensor</span>
<span>from</span> <span>torchvision.utils</span> <span>import</span> <span>make_grid</span>
<span>from</span> <span>torch.utils.data.dataloader</span> <span>import</span> <span>DataLoader</span>
<span>from</span> <span>torch.utils.data</span> <span>import</span> <span>random_split</span>
import torch
import torchvision
import numpy as np
import matplotlib.pyplot as plt
import torch.nn as nn
import torch.nn.functional as F
from torchvision.datasets import FashionMNIST
from torchvision.transforms import ToTensor
from torchvision.utils import make_grid
from torch.utils.data.dataloader import DataLoader
from torch.utils.data import random_split

Enter fullscreen mode Exit fullscreen mode

Lets download our test and train dataset, the data that is used for training is stored in dataset and the test data is stored in test_dataset 

<span>dataset</span> <span>=</span> <span>FashionMNIST</span><span>(</span><span>root</span><span>=</span><span>'</span><span>data/</span><span>'</span><span>,</span> <span>download</span><span>=</span><span>True</span><span>,</span> <span>transform</span><span>=</span><span>ToTensor</span><span>())</span>
<span>test_dataset</span> <span>=</span> <span>FashionMNIST</span><span>(</span><span>root</span><span>=</span><span>'</span><span>data/</span><span>'</span><span>,</span> <span>train</span><span>=</span><span>False</span><span>,</span> <span>transform</span><span>=</span><span>ToTensor</span><span>())</span>
<span>dataset</span> <span>=</span> <span>FashionMNIST</span><span>(</span><span>root</span><span>=</span><span>'</span><span>data/</span><span>'</span><span>,</span> <span>download</span><span>=</span><span>True</span><span>,</span> <span>transform</span><span>=</span><span>ToTensor</span><span>())</span>
<span>test_dataset</span> <span>=</span> <span>FashionMNIST</span><span>(</span><span>root</span><span>=</span><span>'</span><span>data/</span><span>'</span><span>,</span> <span>train</span><span>=</span><span>False</span><span>,</span> <span>transform</span><span>=</span><span>ToTensor</span><span>())</span>
dataset = FashionMNIST(root='data/', download=True, transform=ToTensor())
test_dataset = FashionMNIST(root='data/', train=False, transform=ToTensor())

Enter fullscreen mode Exit fullscreen mode

The training process usually involves taking a small subset of training data for validation(this subset is not used in the training process). We have a subset of 10000 datapoints for validating our model. 

<span>val_size</span> <span>=</span> <span>10000</span>
<span>train_size</span> <span>=</span> <span>len</span><span>(</span><span>dataset</span><span>)</span> <span>-</span> <span>val_size</span>
<span>train_ds</span><span>,</span> <span>val_ds</span> <span>=</span> <span>random_split</span><span>(</span><span>dataset</span><span>,</span> <span>[</span><span>train_size</span><span>,</span> <span>val_size</span><span>])</span>
<span>print</span><span>(</span><span>"</span><span>Training dataset size : </span><span>"</span><span>,</span> <span>len</span><span>(</span><span>train_ds</span><span>))</span> 
<span>print</span><span>(</span><span>"</span><span>Validation dataset size : </span><span>"</span><span>,</span> <span>len</span><span>(</span><span>val_ds</span><span>))</span>
<span>val_size</span> <span>=</span> <span>10000</span>
<span>train_size</span> <span>=</span> <span>len</span><span>(</span><span>dataset</span><span>)</span> <span>-</span> <span>val_size</span>
<span>train_ds</span><span>,</span> <span>val_ds</span> <span>=</span> <span>random_split</span><span>(</span><span>dataset</span><span>,</span> <span>[</span><span>train_size</span><span>,</span> <span>val_size</span><span>])</span>
<span>print</span><span>(</span><span>"</span><span>Training dataset size : </span><span>"</span><span>,</span> <span>len</span><span>(</span><span>train_ds</span><span>))</span> 
<span>print</span><span>(</span><span>"</span><span>Validation dataset size : </span><span>"</span><span>,</span> <span>len</span><span>(</span><span>val_ds</span><span>))</span>
val_size = 10000
train_size = len(dataset) - val_size
train_ds, val_ds = random_split(dataset, [train_size, val_size])
print("Training dataset size : ", len(train_ds)) 
print("Validation dataset size : ", len(val_ds))

Enter fullscreen mode Exit fullscreen mode

Lets define our batch size which specifies how many data points we are planning to use in each iteration during the training process. 

<span>batch_size</span><span>=</span><span>128</span>
<span>batch_size</span><span>=</span><span>128</span>
batch_size=128

Enter fullscreen mode Exit fullscreen mode

DataLoader is a utility provided by pytorch that helps in creating an iterable for our dataset. The arguments shuffle is used to shuffle the data, num_workers is used for transferring data to RAM or whatever memory you are using and pin_memory is used for implicit CPU-to-CPU copy which helps to make the processing much more efficient. 

<span>train_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>train_ds</span><span>,</span> <span>batch_size</span><span>,</span> <span>shuffle</span><span>=</span><span>True</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
<span>val_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>val_ds</span><span>,</span> <span>batch_size</span><span>*</span><span>2</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
<span>test_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>test_dataset</span><span>,</span> <span>batch_size</span><span>*</span><span>2</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
<span>train_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>train_ds</span><span>,</span> <span>batch_size</span><span>,</span> <span>shuffle</span><span>=</span><span>True</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
<span>val_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>val_ds</span><span>,</span> <span>batch_size</span><span>*</span><span>2</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
<span>test_loader</span> <span>=</span> <span>DataLoader</span><span>(</span><span>test_dataset</span><span>,</span> <span>batch_size</span><span>*</span><span>2</span><span>,</span> <span>num_workers</span><span>=</span><span>4</span><span>,</span> <span>pin_memory</span><span>=</span><span>True</span><span>)</span>
train_loader = DataLoader(train_ds, batch_size, shuffle=True, num_workers=4, pin_memory=True)
val_loader = DataLoader(val_ds, batch_size*2, num_workers=4, pin_memory=True)
test_loader = DataLoader(test_dataset, batch_size*2, num_workers=4, pin_memory=True)

Enter fullscreen mode Exit fullscreen mode

Lets Visualise out training dataset, the function below in jupyter notebook displays a grid of 126(our batch_size) images from the training subset. 

<span>for</span> <span>images</span><span>,</span> <span>_</span> <span>in</span> <span>train_loader</span><span>:</span>
    <span>print</span><span>(</span><span>'</span><span>images.shape:</span><span>'</span><span>,</span> <span>images</span><span>.</span><span>shape</span><span>)</span>
    <span>plt</span><span>.</span><span>figure</span><span>(</span><span>figsize</span><span>=</span><span>(</span><span>16</span><span>,</span><span>8</span><span>))</span>
    <span>plt</span><span>.</span><span>axis</span><span>(</span><span>'</span><span>off</span><span>'</span><span>)</span>
    <span>plt</span><span>.</span><span>imshow</span><span>(</span><span>make_grid</span><span>(</span><span>images</span><span>,</span> <span>nrow</span><span>=</span><span>16</span><span>).</span><span>permute</span><span>((</span><span>1</span><span>,</span> <span>2</span><span>,</span> <span>0</span><span>)))</span>
    <span>break</span>
<span>for</span> <span>images</span><span>,</span> <span>_</span> <span>in</span> <span>train_loader</span><span>:</span>
    <span>print</span><span>(</span><span>'</span><span>images.shape:</span><span>'</span><span>,</span> <span>images</span><span>.</span><span>shape</span><span>)</span>
    <span>plt</span><span>.</span><span>figure</span><span>(</span><span>figsize</span><span>=</span><span>(</span><span>16</span><span>,</span><span>8</span><span>))</span>
    <span>plt</span><span>.</span><span>axis</span><span>(</span><span>'</span><span>off</span><span>'</span><span>)</span>
    <span>plt</span><span>.</span><span>imshow</span><span>(</span><span>make_grid</span><span>(</span><span>images</span><span>,</span> <span>nrow</span><span>=</span><span>16</span><span>).</span><span>permute</span><span>((</span><span>1</span><span>,</span> <span>2</span><span>,</span> <span>0</span><span>)))</span>
    <span>break</span>
for images, _ in train_loader:
    print('images.shape:', images.shape)
    plt.figure(figsize=(16,8))
    plt.axis('off')
    plt.imshow(make_grid(images, nrow=16).permute((1, 2, 0)))
    break

Enter fullscreen mode Exit fullscreen mode

图片[1]-Computer Vision 101 - Fashion MNIST - 拾光赋-拾光赋
The accuracy function compares our model output to the actual desired output and returns an accuracy score, this helps to monitor our training process. 

<span>def</span> <span>accuracy</span><span>(</span><span>outputs</span><span>,</span> <span>labels</span><span>):</span>
    <span>_</span><span>,</span> <span>preds</span> <span>=</span> <span>torch</span><span>.</span><span>max</span><span>(</span><span>outputs</span><span>,</span> <span>dim</span><span>=</span><span>1</span><span>)</span>
    <span>return</span> <span>torch</span><span>.</span><span>tensor</span><span>(</span><span>torch</span><span>.</span><span>sum</span><span>(</span><span>preds</span> <span>==</span> <span>labels</span><span>).</span><span>item</span><span>()</span> <span>/</span> <span>len</span><span>(</span><span>preds</span><span>))</span>
<span>def</span> <span>accuracy</span><span>(</span><span>outputs</span><span>,</span> <span>labels</span><span>):</span>
    <span>_</span><span>,</span> <span>preds</span> <span>=</span> <span>torch</span><span>.</span><span>max</span><span>(</span><span>outputs</span><span>,</span> <span>dim</span><span>=</span><span>1</span><span>)</span>
    <span>return</span> <span>torch</span><span>.</span><span>tensor</span><span>(</span><span>torch</span><span>.</span><span>sum</span><span>(</span><span>preds</span> <span>==</span> <span>labels</span><span>).</span><span>item</span><span>()</span> <span>/</span> <span>len</span><span>(</span><span>preds</span><span>))</span>
def accuracy(outputs, labels):
    _, preds = torch.max(outputs, dim=1)
    return torch.tensor(torch.sum(preds == labels).item() / len(preds))

Enter fullscreen mode Exit fullscreen mode

And finally…. lets define our D.L model

  • init : contructor where the neural network layers are defined, here one output layer and 2 hidden layers are used which are made of linear layer

  • forward : This is where we have defined the model architecture,initially we are flattening our tensor as it has to pass through linear layers.

  • training_step : This is where the actual training happens, here we pass the data as input to a function called self which is part of the base class nn.Module which return the predictions as tensors.

  • validation_step : This is a very similar function to training_step , the only difference is that this is used for validating the training, this is where we use the accuracy function to evaluate our current state of model.

  • validation_epoch_end and epoch_end : These are used to display the state of our model at the end of the training process i.e after all epochs have completed. 

<span>class</span> <span>MnistModel</span><span>(</span><span>nn</span><span>.</span><span>Module</span><span>):</span>
    <span>"""</span><span>Feedfoward neural network with 1 hidden layer</span><span>"""</span>
    <span>def</span> <span>__init__</span><span>(</span><span>self</span><span>,</span> <span>in_size</span><span>,</span> <span>out_size</span><span>):</span>
        <span>super</span><span>().</span><span>__init__</span><span>()</span>
        <span># hidden layer </span>        <span>self</span><span>.</span><span>linear1</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>in_size</span><span>,</span> <span>16</span><span>)</span>
        <span># hidden layer 2 </span>        <span>self</span><span>.</span><span>linear2</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>16</span><span>,</span> <span>32</span><span>)</span>
        <span># output layer </span>        <span>self</span><span>.</span><span>linear3</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>32</span><span>,</span> <span>out_size</span><span>)</span>
    <span>def</span> <span>forward</span><span>(</span><span>self</span><span>,</span> <span>xb</span><span>):</span>
        <span># Flatten the image tensors </span>        <span>out</span> <span>=</span> <span>xb</span><span>.</span><span>view</span><span>(</span><span>xb</span><span>.</span><span>size</span><span>(</span><span>0</span><span>),</span> <span>-</span><span>1</span><span>)</span>
        <span># Get intermediate outputs using hidden layer 1 </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear1</span><span>(</span><span>out</span><span>)</span>
        <span># Apply activation function </span>        <span>out</span> <span>=</span> <span>F</span><span>.</span><span>relu</span><span>(</span><span>out</span><span>)</span>
        <span># Get intermediate outputs using hidden layer 2 </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear2</span><span>(</span><span>out</span><span>)</span>
        <span># Apply activation function </span>        <span>out</span> <span>=</span> <span>F</span><span>.</span><span>relu</span><span>(</span><span>out</span><span>)</span>
        <span># Get predictions using output layer </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear3</span><span>(</span><span>out</span><span>)</span>
        <span>return</span> <span>out</span>
    <span>def</span> <span>training_step</span><span>(</span><span>self</span><span>,</span> <span>batch</span><span>):</span>
        <span>images</span><span>,</span> <span>labels</span> <span>=</span> <span>batch</span> 
        <span>out</span> <span>=</span> <span>self</span><span>(</span><span>images</span><span>)</span>                  <span># Generate predictions </span>        <span>loss</span> <span>=</span> <span>F</span><span>.</span><span>cross_entropy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span> <span># Calculate loss </span>        <span>return</span> <span>loss</span>
    <span>def</span> <span>validation_step</span><span>(</span><span>self</span><span>,</span> <span>batch</span><span>):</span>
        <span>images</span><span>,</span> <span>labels</span> <span>=</span> <span>batch</span> 
        <span>out</span> <span>=</span> <span>self</span><span>(</span><span>images</span><span>)</span>                    <span># Generate predictions </span>        <span>loss</span> <span>=</span> <span>F</span><span>.</span><span>cross_entropy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span>   <span># Calculate loss </span>        <span>acc</span> <span>=</span> <span>accuracy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span>           <span># Calculate accuracy </span>        <span>return</span> <span>{</span><span>'</span><span>val_loss</span><span>'</span><span>:</span> <span>loss</span><span>,</span> <span>'</span><span>val_acc</span><span>'</span><span>:</span> <span>acc</span><span>}</span>
    <span>def</span> <span>validation_epoch_end</span><span>(</span><span>self</span><span>,</span> <span>outputs</span><span>):</span>
        <span>batch_losses</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>outputs</span><span>]</span>
        <span>epoch_loss</span> <span>=</span> <span>torch</span><span>.</span><span>stack</span><span>(</span><span>batch_losses</span><span>).</span><span>mean</span><span>()</span>   <span># Combine losses </span>        <span>batch_accs</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>outputs</span><span>]</span>
        <span>epoch_acc</span> <span>=</span> <span>torch</span><span>.</span><span>stack</span><span>(</span><span>batch_accs</span><span>).</span><span>mean</span><span>()</span>      <span># Combine accuracies </span>        <span>return</span> <span>{</span><span>'</span><span>val_loss</span><span>'</span><span>:</span> <span>epoch_loss</span><span>.</span><span>item</span><span>(),</span> <span>'</span><span>val_acc</span><span>'</span><span>:</span> <span>epoch_acc</span><span>.</span><span>item</span><span>()}</span>
    <span>def</span> <span>epoch_end</span><span>(</span><span>self</span><span>,</span> <span>epoch</span><span>,</span> <span>result</span><span>):</span>
        <span>print</span><span>(</span><span>"</span><span>Epoch [{}], val_loss: {:.4f}, val_acc: {:.4f}</span><span>"</span><span>.</span><span>format</span><span>(</span><span>epoch</span><span>,</span> <span>result</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>],</span> <span>result</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]))</span>
<span>class</span> <span>MnistModel</span><span>(</span><span>nn</span><span>.</span><span>Module</span><span>):</span>
    <span>"""</span><span>Feedfoward neural network with 1 hidden layer</span><span>"""</span>
    <span>def</span> <span>__init__</span><span>(</span><span>self</span><span>,</span> <span>in_size</span><span>,</span> <span>out_size</span><span>):</span>
        <span>super</span><span>().</span><span>__init__</span><span>()</span>
        <span># hidden layer </span>        <span>self</span><span>.</span><span>linear1</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>in_size</span><span>,</span> <span>16</span><span>)</span>
        <span># hidden layer 2 </span>        <span>self</span><span>.</span><span>linear2</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>16</span><span>,</span> <span>32</span><span>)</span>
        <span># output layer </span>        <span>self</span><span>.</span><span>linear3</span> <span>=</span> <span>nn</span><span>.</span><span>Linear</span><span>(</span><span>32</span><span>,</span> <span>out_size</span><span>)</span>

    <span>def</span> <span>forward</span><span>(</span><span>self</span><span>,</span> <span>xb</span><span>):</span>
        <span># Flatten the image tensors </span>        <span>out</span> <span>=</span> <span>xb</span><span>.</span><span>view</span><span>(</span><span>xb</span><span>.</span><span>size</span><span>(</span><span>0</span><span>),</span> <span>-</span><span>1</span><span>)</span>
        <span># Get intermediate outputs using hidden layer 1 </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear1</span><span>(</span><span>out</span><span>)</span>
        <span># Apply activation function </span>        <span>out</span> <span>=</span> <span>F</span><span>.</span><span>relu</span><span>(</span><span>out</span><span>)</span>
        <span># Get intermediate outputs using hidden layer 2 </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear2</span><span>(</span><span>out</span><span>)</span>
        <span># Apply activation function </span>        <span>out</span> <span>=</span> <span>F</span><span>.</span><span>relu</span><span>(</span><span>out</span><span>)</span>
        <span># Get predictions using output layer </span>        <span>out</span> <span>=</span> <span>self</span><span>.</span><span>linear3</span><span>(</span><span>out</span><span>)</span>
        <span>return</span> <span>out</span>

    <span>def</span> <span>training_step</span><span>(</span><span>self</span><span>,</span> <span>batch</span><span>):</span>
        <span>images</span><span>,</span> <span>labels</span> <span>=</span> <span>batch</span> 
        <span>out</span> <span>=</span> <span>self</span><span>(</span><span>images</span><span>)</span>                  <span># Generate predictions </span>        <span>loss</span> <span>=</span> <span>F</span><span>.</span><span>cross_entropy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span> <span># Calculate loss </span>        <span>return</span> <span>loss</span>

    <span>def</span> <span>validation_step</span><span>(</span><span>self</span><span>,</span> <span>batch</span><span>):</span>
        <span>images</span><span>,</span> <span>labels</span> <span>=</span> <span>batch</span> 
        <span>out</span> <span>=</span> <span>self</span><span>(</span><span>images</span><span>)</span>                    <span># Generate predictions </span>        <span>loss</span> <span>=</span> <span>F</span><span>.</span><span>cross_entropy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span>   <span># Calculate loss </span>        <span>acc</span> <span>=</span> <span>accuracy</span><span>(</span><span>out</span><span>,</span> <span>labels</span><span>)</span>           <span># Calculate accuracy </span>        <span>return</span> <span>{</span><span>'</span><span>val_loss</span><span>'</span><span>:</span> <span>loss</span><span>,</span> <span>'</span><span>val_acc</span><span>'</span><span>:</span> <span>acc</span><span>}</span>

    <span>def</span> <span>validation_epoch_end</span><span>(</span><span>self</span><span>,</span> <span>outputs</span><span>):</span>
        <span>batch_losses</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>outputs</span><span>]</span>
        <span>epoch_loss</span> <span>=</span> <span>torch</span><span>.</span><span>stack</span><span>(</span><span>batch_losses</span><span>).</span><span>mean</span><span>()</span>   <span># Combine losses </span>        <span>batch_accs</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>outputs</span><span>]</span>
        <span>epoch_acc</span> <span>=</span> <span>torch</span><span>.</span><span>stack</span><span>(</span><span>batch_accs</span><span>).</span><span>mean</span><span>()</span>      <span># Combine accuracies </span>        <span>return</span> <span>{</span><span>'</span><span>val_loss</span><span>'</span><span>:</span> <span>epoch_loss</span><span>.</span><span>item</span><span>(),</span> <span>'</span><span>val_acc</span><span>'</span><span>:</span> <span>epoch_acc</span><span>.</span><span>item</span><span>()}</span>

    <span>def</span> <span>epoch_end</span><span>(</span><span>self</span><span>,</span> <span>epoch</span><span>,</span> <span>result</span><span>):</span>
        <span>print</span><span>(</span><span>"</span><span>Epoch [{}], val_loss: {:.4f}, val_acc: {:.4f}</span><span>"</span><span>.</span><span>format</span><span>(</span><span>epoch</span><span>,</span> <span>result</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>],</span> <span>result</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]))</span>
class MnistModel(nn.Module):
    """Feedfoward neural network with 1 hidden layer"""
    def __init__(self, in_size, out_size):
        super().__init__()
        # hidden layer         self.linear1 = nn.Linear(in_size, 16)
        # hidden layer 2         self.linear2 = nn.Linear(16, 32)
        # output layer         self.linear3 = nn.Linear(32, out_size)

    def forward(self, xb):
        # Flatten the image tensors         out = xb.view(xb.size(0), -1)
        # Get intermediate outputs using hidden layer 1         out = self.linear1(out)
        # Apply activation function         out = F.relu(out)
        # Get intermediate outputs using hidden layer 2         out = self.linear2(out)
        # Apply activation function         out = F.relu(out)
        # Get predictions using output layer         out = self.linear3(out)
        return out

    def training_step(self, batch):
        images, labels = batch 
        out = self(images)                  # Generate predictions         loss = F.cross_entropy(out, labels) # Calculate loss         return loss

    def validation_step(self, batch):
        images, labels = batch 
        out = self(images)                    # Generate predictions         loss = F.cross_entropy(out, labels)   # Calculate loss         acc = accuracy(out, labels)           # Calculate accuracy         return {'val_loss': loss, 'val_acc': acc}

    def validation_epoch_end(self, outputs):
        batch_losses = [x['val_loss'] for x in outputs]
        epoch_loss = torch.stack(batch_losses).mean()   # Combine losses         batch_accs = [x['val_acc'] for x in outputs]
        epoch_acc = torch.stack(batch_accs).mean()      # Combine accuracies         return {'val_loss': epoch_loss.item(), 'val_acc': epoch_acc.item()}

    def epoch_end(self, epoch, result):
        print("Epoch [{}], val_loss: {:.4f}, val_acc: {:.4f}".format(epoch, result['val_loss'], result['val_acc']))

Enter fullscreen mode Exit fullscreen mode

CPU or GPU, you can use anything to train the model, this function automatically detects your available device and assigns it for training. 

<span>def</span> <span>get_default_device</span><span>():</span>
    <span>"""</span><span>Pick GPU if available, else CPU</span><span>"""</span>
    <span>if</span> <span>torch</span><span>.</span><span>cuda</span><span>.</span><span>is_available</span><span>():</span>
        <span>return</span> <span>torch</span><span>.</span><span>device</span><span>(</span><span>'</span><span>cuda</span><span>'</span><span>)</span>
    <span>else</span><span>:</span>
        <span>return</span> <span>torch</span><span>.</span><span>device</span><span>(</span><span>'</span><span>cpu</span><span>'</span><span>)</span>
<span>device</span> <span>=</span> <span>get_default_device</span><span>()</span>
<span>def</span> <span>get_default_device</span><span>():</span>
    <span>"""</span><span>Pick GPU if available, else CPU</span><span>"""</span>
    <span>if</span> <span>torch</span><span>.</span><span>cuda</span><span>.</span><span>is_available</span><span>():</span>
        <span>return</span> <span>torch</span><span>.</span><span>device</span><span>(</span><span>'</span><span>cuda</span><span>'</span><span>)</span>
    <span>else</span><span>:</span>
        <span>return</span> <span>torch</span><span>.</span><span>device</span><span>(</span><span>'</span><span>cpu</span><span>'</span><span>)</span>
<span>device</span> <span>=</span> <span>get_default_device</span><span>()</span>
def get_default_device():
    """Pick GPU if available, else CPU"""
    if torch.cuda.is_available():
        return torch.device('cuda')
    else:
        return torch.device('cpu')
device = get_default_device()

Enter fullscreen mode Exit fullscreen mode

Now let’s move the data to the chosen device, to understand this better if you are using GPU/CPU you cannot store the complete data into the the available memory(RAM/GPU-memory), so we are going to transfer it as batches at the time of training, this is an efficient way of training. 

<span>def</span> <span>to_device</span><span>(</span><span>data</span><span>,</span> <span>device</span><span>):</span>
    <span>"""</span><span>Move tensor(s) to chosen device</span><span>"""</span>
    <span>if</span> <span>isinstance</span><span>(</span><span>data</span><span>,</span> <span>(</span><span>list</span><span>,</span><span>tuple</span><span>)):</span>
        <span>return</span> <span>[</span><span>to_device</span><span>(</span><span>x</span><span>,</span> <span>device</span><span>)</span> <span>for</span> <span>x</span> <span>in</span> <span>data</span><span>]</span>
    <span>return</span> <span>data</span><span>.</span><span>to</span><span>(</span><span>device</span><span>,</span> <span>non_blocking</span><span>=</span><span>True</span><span>)</span>
<span>def</span> <span>to_device</span><span>(</span><span>data</span><span>,</span> <span>device</span><span>):</span>
    <span>"""</span><span>Move tensor(s) to chosen device</span><span>"""</span>
    <span>if</span> <span>isinstance</span><span>(</span><span>data</span><span>,</span> <span>(</span><span>list</span><span>,</span><span>tuple</span><span>)):</span>
        <span>return</span> <span>[</span><span>to_device</span><span>(</span><span>x</span><span>,</span> <span>device</span><span>)</span> <span>for</span> <span>x</span> <span>in</span> <span>data</span><span>]</span>
    <span>return</span> <span>data</span><span>.</span><span>to</span><span>(</span><span>device</span><span>,</span> <span>non_blocking</span><span>=</span><span>True</span><span>)</span>
def to_device(data, device):
    """Move tensor(s) to chosen device"""
    if isinstance(data, (list,tuple)):
        return [to_device(x, device) for x in data]
    return data.to(device, non_blocking=True)

Enter fullscreen mode Exit fullscreen mode

Now let’s define a class to handle all the data movement 

<span>class</span> <span>DeviceDataLoader</span><span>():</span>
    <span>"""</span><span>Wrap a dataloader to move data to a device</span><span>"""</span>
    <span>def</span> <span>__init__</span><span>(</span><span>self</span><span>,</span> <span>dl</span><span>,</span> <span>device</span><span>):</span>
        <span>self</span><span>.</span><span>dl</span> <span>=</span> <span>dl</span>
        <span>self</span><span>.</span><span>device</span> <span>=</span> <span>device</span>
    <span>def</span> <span>__iter__</span><span>(</span><span>self</span><span>):</span>
        <span>"""</span><span>Yield a batch of data after moving it to device</span><span>"""</span>
        <span>for</span> <span>b</span> <span>in</span> <span>self</span><span>.</span><span>dl</span><span>:</span> 
            <span>yield</span> <span>to_device</span><span>(</span><span>b</span><span>,</span> <span>self</span><span>.</span><span>device</span><span>)</span>
    <span>def</span> <span>__len__</span><span>(</span><span>self</span><span>):</span>
        <span>"""</span><span>Number of batches</span><span>"""</span>
        <span>return</span> <span>len</span><span>(</span><span>self</span><span>.</span><span>dl</span><span>)</span>
<span>class</span> <span>DeviceDataLoader</span><span>():</span>
    <span>"""</span><span>Wrap a dataloader to move data to a device</span><span>"""</span>
    <span>def</span> <span>__init__</span><span>(</span><span>self</span><span>,</span> <span>dl</span><span>,</span> <span>device</span><span>):</span>
        <span>self</span><span>.</span><span>dl</span> <span>=</span> <span>dl</span>
        <span>self</span><span>.</span><span>device</span> <span>=</span> <span>device</span>

    <span>def</span> <span>__iter__</span><span>(</span><span>self</span><span>):</span>
        <span>"""</span><span>Yield a batch of data after moving it to device</span><span>"""</span>
        <span>for</span> <span>b</span> <span>in</span> <span>self</span><span>.</span><span>dl</span><span>:</span> 
            <span>yield</span> <span>to_device</span><span>(</span><span>b</span><span>,</span> <span>self</span><span>.</span><span>device</span><span>)</span>

    <span>def</span> <span>__len__</span><span>(</span><span>self</span><span>):</span>
        <span>"""</span><span>Number of batches</span><span>"""</span>
        <span>return</span> <span>len</span><span>(</span><span>self</span><span>.</span><span>dl</span><span>)</span>
class DeviceDataLoader():
    """Wrap a dataloader to move data to a device"""
    def __init__(self, dl, device):
        self.dl = dl
        self.device = device

    def __iter__(self):
        """Yield a batch of data after moving it to device"""
        for b in self.dl: 
            yield to_device(b, self.device)

    def __len__(self):
        """Number of batches"""
        return len(self.dl)

Enter fullscreen mode Exit fullscreen mode 

<span>train_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>train_loader</span><span>,</span> <span>device</span><span>)</span>
<span>val_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>val_loader</span><span>,</span> <span>device</span><span>)</span>
<span>test_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>test_loader</span><span>,</span> <span>device</span><span>)</span>
<span>train_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>train_loader</span><span>,</span> <span>device</span><span>)</span>
<span>val_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>val_loader</span><span>,</span> <span>device</span><span>)</span>
<span>test_loader</span> <span>=</span> <span>DeviceDataLoader</span><span>(</span><span>test_loader</span><span>,</span> <span>device</span><span>)</span>
train_loader = DeviceDataLoader(train_loader, device)
val_loader = DeviceDataLoader(val_loader, device)
test_loader = DeviceDataLoader(test_loader, device)

Enter fullscreen mode Exit fullscreen mode

Now we are defining two functions to explicitly handle the training and validation of the model 

<span>def</span> <span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>):</span>
    <span>outputs</span> <span>=</span> <span>[</span><span>model</span><span>.</span><span>validation_step</span><span>(</span><span>batch</span><span>)</span> <span>for</span> <span>batch</span> <span>in</span> <span>val_loader</span><span>]</span>
    <span>return</span> <span>model</span><span>.</span><span>validation_epoch_end</span><span>(</span><span>outputs</span><span>)</span>
<span>def</span> <span>fit</span><span>(</span><span>epochs</span><span>,</span> <span>lr</span><span>,</span> <span>model</span><span>,</span> <span>train_loader</span><span>,</span> <span>val_loader</span><span>,</span> <span>opt_func</span><span>=</span><span>torch</span><span>.</span><span>optim</span><span>.</span><span>SGD</span><span>):</span>
    <span>history</span> <span>=</span> <span>[]</span>
    <span>optimizer</span> <span>=</span> <span>opt_func</span><span>(</span><span>model</span><span>.</span><span>parameters</span><span>(),</span> <span>lr</span><span>)</span>
    <span>for</span> <span>epoch</span> <span>in</span> <span>range</span><span>(</span><span>epochs</span><span>):</span>
        <span># Training Phase </span>        <span>for</span> <span>batch</span> <span>in</span> <span>train_loader</span><span>:</span>
            <span>loss</span> <span>=</span> <span>model</span><span>.</span><span>training_step</span><span>(</span><span>batch</span><span>)</span>
            <span>loss</span><span>.</span><span>backward</span><span>()</span>
            <span>optimizer</span><span>.</span><span>step</span><span>()</span>
            <span>optimizer</span><span>.</span><span>zero_grad</span><span>()</span>
        <span># Validation phase </span>        <span>result</span> <span>=</span> <span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>)</span>
        <span>model</span><span>.</span><span>epoch_end</span><span>(</span><span>epoch</span><span>,</span> <span>result</span><span>)</span>
        <span>history</span><span>.</span><span>append</span><span>(</span><span>result</span><span>)</span>
    <span>return</span> <span>history</span>
<span>def</span> <span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>):</span>
    <span>outputs</span> <span>=</span> <span>[</span><span>model</span><span>.</span><span>validation_step</span><span>(</span><span>batch</span><span>)</span> <span>for</span> <span>batch</span> <span>in</span> <span>val_loader</span><span>]</span>
    <span>return</span> <span>model</span><span>.</span><span>validation_epoch_end</span><span>(</span><span>outputs</span><span>)</span>

<span>def</span> <span>fit</span><span>(</span><span>epochs</span><span>,</span> <span>lr</span><span>,</span> <span>model</span><span>,</span> <span>train_loader</span><span>,</span> <span>val_loader</span><span>,</span> <span>opt_func</span><span>=</span><span>torch</span><span>.</span><span>optim</span><span>.</span><span>SGD</span><span>):</span>
    <span>history</span> <span>=</span> <span>[]</span>
    <span>optimizer</span> <span>=</span> <span>opt_func</span><span>(</span><span>model</span><span>.</span><span>parameters</span><span>(),</span> <span>lr</span><span>)</span>
    <span>for</span> <span>epoch</span> <span>in</span> <span>range</span><span>(</span><span>epochs</span><span>):</span>
        <span># Training Phase </span>        <span>for</span> <span>batch</span> <span>in</span> <span>train_loader</span><span>:</span>
            <span>loss</span> <span>=</span> <span>model</span><span>.</span><span>training_step</span><span>(</span><span>batch</span><span>)</span>
            <span>loss</span><span>.</span><span>backward</span><span>()</span>
            <span>optimizer</span><span>.</span><span>step</span><span>()</span>
            <span>optimizer</span><span>.</span><span>zero_grad</span><span>()</span>
        <span># Validation phase </span>        <span>result</span> <span>=</span> <span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>)</span>
        <span>model</span><span>.</span><span>epoch_end</span><span>(</span><span>epoch</span><span>,</span> <span>result</span><span>)</span>
        <span>history</span><span>.</span><span>append</span><span>(</span><span>result</span><span>)</span>
    <span>return</span> <span>history</span>
def evaluate(model, val_loader):
    outputs = [model.validation_step(batch) for batch in val_loader]
    return model.validation_epoch_end(outputs)

def fit(epochs, lr, model, train_loader, val_loader, opt_func=torch.optim.SGD):
    history = []
    optimizer = opt_func(model.parameters(), lr)
    for epoch in range(epochs):
        # Training Phase         for batch in train_loader:
            loss = model.training_step(batch)
            loss.backward()
            optimizer.step()
            optimizer.zero_grad()
        # Validation phase         result = evaluate(model, val_loader)
        model.epoch_end(epoch, result)
        history.append(result)
    return history

Enter fullscreen mode Exit fullscreen mode

Each of our data point is of dimension 28*28 = 784 and we have 10 distinct labels(classes) which we need to classify 

<span>input_size</span> <span>=</span> <span>784</span>
<span>num_classes</span> <span>=</span> <span>10</span>
<span>input_size</span> <span>=</span> <span>784</span>
<span>num_classes</span> <span>=</span> <span>10</span>
input_size = 784
num_classes = 10

Enter fullscreen mode Exit fullscreen mode

Now let’s initialise our model and move it inside GPU/CPU memory 

<span>model</span> <span>=</span> <span>MnistModel</span><span>(</span><span>input_size</span><span>,</span> <span>out_size</span><span>=</span><span>num_classes</span><span>)</span>
<span>to_device</span><span>(</span><span>model</span><span>,</span> <span>device</span><span>)</span>
<span>model</span> <span>=</span> <span>MnistModel</span><span>(</span><span>input_size</span><span>,</span> <span>out_size</span><span>=</span><span>num_classes</span><span>)</span>
<span>to_device</span><span>(</span><span>model</span><span>,</span> <span>device</span><span>)</span>
model = MnistModel(input_size, out_size=num_classes)
to_device(model, device)

Enter fullscreen mode Exit fullscreen mode

Initially evaluate the model without any training that is the model now has random weights and biases 

<span>history</span> <span>=</span> <span>[</span><span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>)]</span>
<span>print</span><span>(</span><span>"</span><span>Before Training : </span><span>"</span><span>,</span><span>history</span><span>)</span>
<span>history</span> <span>=</span> <span>[</span><span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>val_loader</span><span>)]</span>
<span>print</span><span>(</span><span>"</span><span>Before Training : </span><span>"</span><span>,</span><span>history</span><span>)</span>
history = [evaluate(model, val_loader)]
print("Before Training : ",history)

Enter fullscreen mode Exit fullscreen mode

Now lets train it for 10 epochs and learning rate is 0.5 

<span>history</span> <span>+=</span> <span>fit</span><span>(</span><span>10</span><span>,</span> <span>0.5</span><span>,</span> <span>model</span><span>,</span> <span>train_loader</span><span>,</span> <span>val_loader</span><span>)</span>
<span>history</span> <span>+=</span> <span>fit</span><span>(</span><span>10</span><span>,</span> <span>0.5</span><span>,</span> <span>model</span><span>,</span> <span>train_loader</span><span>,</span> <span>val_loader</span><span>)</span>
history += fit(10, 0.5, model, train_loader, val_loader)

Enter fullscreen mode Exit fullscreen mode

and our training is done……..

To visualise the loss and accuracy the code given below could be very helpful, run this is jupyter notebook 

<span>losses</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>history</span><span>]</span>
<span>plt</span><span>.</span><span>plot</span><span>(</span><span>losses</span><span>,</span> <span>'</span><span>-x</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>xlabel</span><span>(</span><span>'</span><span>epoch</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>ylabel</span><span>(</span><span>'</span><span>loss</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>title</span><span>(</span><span>'</span><span>Loss vs. No. of epochs</span><span>'</span><span>);</span>
<span>losses</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_loss</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>history</span><span>]</span>
<span>plt</span><span>.</span><span>plot</span><span>(</span><span>losses</span><span>,</span> <span>'</span><span>-x</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>xlabel</span><span>(</span><span>'</span><span>epoch</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>ylabel</span><span>(</span><span>'</span><span>loss</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>title</span><span>(</span><span>'</span><span>Loss vs. No. of epochs</span><span>'</span><span>);</span>
losses = [x['val_loss'] for x in history]
plt.plot(losses, '-x')
plt.xlabel('epoch')
plt.ylabel('loss')
plt.title('Loss vs. No. of epochs');

Enter fullscreen mode Exit fullscreen mode 

<span>accuracies</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>history</span><span>]</span>
<span>plt</span><span>.</span><span>plot</span><span>(</span><span>accuracies</span><span>,</span> <span>'</span><span>-x</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>xlabel</span><span>(</span><span>'</span><span>epoch</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>ylabel</span><span>(</span><span>'</span><span>accuracy</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>title</span><span>(</span><span>'</span><span>Accuracy vs. No. of epochs</span><span>'</span><span>);</span>
<span>accuracies</span> <span>=</span> <span>[</span><span>x</span><span>[</span><span>'</span><span>val_acc</span><span>'</span><span>]</span> <span>for</span> <span>x</span> <span>in</span> <span>history</span><span>]</span>
<span>plt</span><span>.</span><span>plot</span><span>(</span><span>accuracies</span><span>,</span> <span>'</span><span>-x</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>xlabel</span><span>(</span><span>'</span><span>epoch</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>ylabel</span><span>(</span><span>'</span><span>accuracy</span><span>'</span><span>)</span>
<span>plt</span><span>.</span><span>title</span><span>(</span><span>'</span><span>Accuracy vs. No. of epochs</span><span>'</span><span>);</span>
accuracies = [x['val_acc'] for x in history]
plt.plot(accuracies, '-x')
plt.xlabel('epoch')
plt.ylabel('accuracy')
plt.title('Accuracy vs. No. of epochs');

Enter fullscreen mode Exit fullscreen mode

Lets do some predictions/classification with our trained model 

<span>def</span> <span>predict_image</span><span>(</span><span>img</span><span>,</span> <span>model</span><span>):</span>
    <span>xb</span> <span>=</span> <span>to_device</span><span>(</span><span>img</span><span>.</span><span>unsqueeze</span><span>(</span><span>0</span><span>),</span> <span>device</span><span>)</span>
    <span>yb</span> <span>=</span> <span>model</span><span>(</span><span>xb</span><span>)</span>
    <span>_</span><span>,</span> <span>preds</span>  <span>=</span> <span>torch</span><span>.</span><span>max</span><span>(</span><span>yb</span><span>,</span> <span>dim</span><span>=</span><span>1</span><span>)</span>
    <span>return</span> <span>preds</span><span>[</span><span>0</span><span>].</span><span>item</span><span>()</span>
<span>def</span> <span>predict_image</span><span>(</span><span>img</span><span>,</span> <span>model</span><span>):</span>
    <span>xb</span> <span>=</span> <span>to_device</span><span>(</span><span>img</span><span>.</span><span>unsqueeze</span><span>(</span><span>0</span><span>),</span> <span>device</span><span>)</span>
    <span>yb</span> <span>=</span> <span>model</span><span>(</span><span>xb</span><span>)</span>
    <span>_</span><span>,</span> <span>preds</span>  <span>=</span> <span>torch</span><span>.</span><span>max</span><span>(</span><span>yb</span><span>,</span> <span>dim</span><span>=</span><span>1</span><span>)</span>
    <span>return</span> <span>preds</span><span>[</span><span>0</span><span>].</span><span>item</span><span>()</span>
def predict_image(img, model):
    xb = to_device(img.unsqueeze(0), device)
    yb = model(xb)
    _, preds  = torch.max(yb, dim=1)
    return preds[0].item()

Enter fullscreen mode Exit fullscreen mode

We are taking one image from the test_data subset to test our model 

<span>img</span><span>,</span> <span>label</span> <span>=</span> <span>test_dataset</span><span>[</span><span>0</span><span>]</span>
<span>plt</span><span>.</span><span>imshow</span><span>(</span><span>img</span><span>[</span><span>0</span><span>],</span> <span>cmap</span><span>=</span><span>'</span><span>gray</span><span>'</span><span>)</span>
<span>print</span><span>(</span><span>'</span><span>Label:</span><span>'</span><span>,</span> <span>dataset</span><span>.</span><span>classes</span><span>[</span><span>label</span><span>],</span> <span>'</span><span>, Predicted:</span><span>'</span><span>,</span> <span>dataset</span><span>.</span><span>classes</span><span>[</span><span>predict_image</span><span>(</span><span>img</span><span>,</span> <span>model</span><span>)])</span>
<span>img</span><span>,</span> <span>label</span> <span>=</span> <span>test_dataset</span><span>[</span><span>0</span><span>]</span>
<span>plt</span><span>.</span><span>imshow</span><span>(</span><span>img</span><span>[</span><span>0</span><span>],</span> <span>cmap</span><span>=</span><span>'</span><span>gray</span><span>'</span><span>)</span>
<span>print</span><span>(</span><span>'</span><span>Label:</span><span>'</span><span>,</span> <span>dataset</span><span>.</span><span>classes</span><span>[</span><span>label</span><span>],</span> <span>'</span><span>, Predicted:</span><span>'</span><span>,</span> <span>dataset</span><span>.</span><span>classes</span><span>[</span><span>predict_image</span><span>(</span><span>img</span><span>,</span> <span>model</span><span>)])</span>
img, label = test_dataset[0]
plt.imshow(img[0], cmap='gray')
print('Label:', dataset.classes[label], ', Predicted:', dataset.classes[predict_image(img, model)])

Enter fullscreen mode Exit fullscreen mode

Lets look at the fruit of our work 

<span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>test_loader</span><span>)</span>
<span>evaluate</span><span>(</span><span>model</span><span>,</span> <span>test_loader</span><span>)</span>
evaluate(model, test_loader)

Enter fullscreen mode Exit fullscreen mode

This is what I got

Hope this blog helped you, by the way this my first blog so do provide valuable feedback, thanks !!!!

原文链接：Computer Vision 101 – Fashion MNIST

展开阅读全文
© 版权声明
文章版权声明 1、本网站名称：拾光赋
2、本站永久网址：https://www.blogs.ink
3、本网站的文章部分内容可能来源于网络，仅供大家学习与参考，如有侵权，请联系站长QQ：805375623进行删除处理。
4、本站一切资源不代表本站立场，并不代表本站赞同其观点和对其真实性负责。
5、本站一律禁止以任何方式发布或转载任何违法的相关信息，访客发现请向站长举报
6、本站资源大多存储在云盘，如发现链接失效，请联系我们我们会第一时间更新。

THE END
Python（EN）
# python# AI# pytorch# computervision
喜欢就支持一下吧
点赞9 分享
Sometimes a winner is just a dreamer that never gives up.
有时候，成功者只是坚持梦想不放弃的人
kity的头像-拾光赋
相关推荐
评论 抢沙发

请登录后发表评论

    暂无评论内容

文章目录
kity的头像-拾光赋
文章目录
今日剩余 78.6%
本周剩余 96.9%
本月剩余 47.7%
本年剩余 79.4%
最近评论
Vito的头像-拾光赋

Vito5个月前2

你就是我心中的那首忐忑，总是让我惊心动魄
Jason.Meng的头像-拾光赋钻石会员

Jason.Meng徽章-人气大使-拾光赋5个月前2

总是会在人生的不同阶段反复爱上大佬的文章~
每日一言
I lose, I lose, but I never give up.
我输过,我败过,但我从未放弃过