How to Print Model Summary in PyTorch: Simple Guide

To print a model summary in PyTorch, use the summary function from the torchsummary package by passing your model and input size. This shows layer details, output shapes, and parameter counts in a neat table.

Syntax

The basic syntax to print a model summary in PyTorch is:

• summary(model, input_size): Prints the model architecture summary.
• model: Your PyTorch model instance.
• input_size: Tuple representing the shape of a single input example (channels, height, width).

This function displays each layer's name, output shape, and number of parameters.

python

from torchsummary import summary

summary(model, input_size=(channels, height, width))

Example

This example shows how to print the summary of a simple convolutional neural network using torchsummary. It demonstrates the layer types, output shapes, and parameter counts.

python

import torch
import torch.nn as nn
from torchsummary import summary

class SimpleCNN(nn.Module):
    def __init__(self):
        super(SimpleCNN, self).__init__()
        self.conv1 = nn.Conv2d(3, 16, kernel_size=3, padding=1)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(16, 32, kernel_size=3, padding=1)
        self.fc1 = nn.Linear(32 * 8 * 8, 10)  # assuming input 3x32x32

    def forward(self, x):
        x = self.pool(torch.relu(self.conv1(x)))
        x = self.pool(torch.relu(self.conv2(x)))
        x = x.view(-1, 32 * 8 * 8)
        x = self.fc1(x)
        return x

model = SimpleCNN()
summary(model, input_size=(3, 32, 32))

Output

----------------------------------------------------------------
        Layer (type)               Output Shape         Param #
================================================================
            Conv2d-1          [1, 16, 32, 32]             448
         MaxPool2d-2          [1, 16, 16, 16]               0
            Conv2d-3          [1, 32, 16, 16]           4,640
         MaxPool2d-4            [1, 32, 8, 8]               0
            Linear-5                   [1, 10]           20,490
================================================================
Total params: 25,578
Trainable params: 25,578
Non-trainable params: 0
----------------------------------------------------------------

Common Pitfalls

Common mistakes when printing model summaries in PyTorch include:

• Not installing torchsummary before use. Install it via pip install torchsummary.
• Passing incorrect input_size shape. It must exclude batch size and be (channels, height, width).
• Trying to use summary on models that require multiple inputs without adjustment.
• Forgetting to move the model to the correct device (CPU/GPU) before summary if needed.

python

import torch
from torchsummary import summary

# Wrong input size example
# summary(model, input_size=(32, 32, 3))  # Incorrect order

# Correct input size
# summary(model, input_size=(3, 32, 32))

Quick Reference

Tips for printing model summaries in PyTorch:

• Always install torchsummary first.
• Use summary(model, input_size=(channels, height, width)).
• Input size excludes batch dimension.
• Works best for single input tensor models.
• Shows layer names, output shapes, and parameter counts.

Key Takeaways

Use torchsummary's summary() to print PyTorch model architecture clearly.

Pass the model and input size as (channels, height, width) without batch size.

Install torchsummary via pip before using summary().

Check input size order carefully to avoid errors.

Summary helps understand layer outputs and parameter counts quickly.