Zeroing gradients in PyTorch

Introduction

Zeroing gradients clears old gradient values before calculating new ones. This helps the model learn correctly without mixing old and new information.

Before starting a new training step in a neural network.

When updating model weights during backpropagation.

To avoid accumulating gradients from multiple batches unintentionally.

When training with mini-batches to ensure fresh gradient calculations.

Syntax

PyTorch

optimizer.zero_grad()

This command sets all gradients stored in the optimizer to zero.

It is usually called before loss.backward() in the training loop.

Examples

Clears gradients before computing new ones in a training step.

PyTorch

optimizer.zero_grad()

Alternative way to clear gradients by setting them to None.

PyTorch

for param in model.parameters():
    param.grad = None

Sample Model

This code shows how zeroing gradients works in a simple training step. It clears old gradients, computes new ones, updates weights, and clears gradients again.

PyTorch

import torch
import torch.nn as nn
import torch.optim as optim

# Simple linear model
model = nn.Linear(2, 1)
optimizer = optim.SGD(model.parameters(), lr=0.1)

# Dummy input and target
inputs = torch.tensor([[1.0, 2.0]])
target = torch.tensor([[1.0]])

# Forward pass
output = model(inputs)

# Compute mean squared error loss
loss_fn = nn.MSELoss()
loss = loss_fn(output, target)

# Zero gradients before backward pass
optimizer.zero_grad()

# Backward pass to compute gradients
loss.backward()

# Print gradients before optimizer step
grads_before = [param.grad.clone() for param in model.parameters()]

# Update weights
optimizer.step()

# Zero gradients again for next step
optimizer.zero_grad()

gradients_after = [param.grad for param in model.parameters()]

print("Gradients before optimizer step:")
for g in grads_before:
    print(g)

print("\nGradients after zero_grad call:")
for g in gradients_after:
    print(g)

Important Notes

Always zero gradients before calling loss.backward() to avoid mixing gradients from multiple steps.

Using param.grad = None can be more memory efficient than zeroing with zeros.

Forgetting to zero gradients can cause your model to learn incorrectly.

Summary

Zeroing gradients clears old gradient values before new calculations.

Use optimizer.zero_grad() before loss.backward() in training loops.

This ensures correct and fresh gradient updates for model learning.

Practice

1. Why do we need to zero gradients in PyTorch before calling loss.backward() during training?

easy

A. To initialize the model parameters randomly

B. To save memory by deleting the model weights

C. To speed up the training by skipping backward pass

D. To clear old gradients so new gradients are calculated correctly

Solution

1. Step 1: Understand gradient accumulation in PyTorch

   PyTorch accumulates gradients by default, so old gradients stay unless cleared.

2. Step 2: Purpose of zeroing gradients before backward pass

   Zeroing gradients ensures only the current batch's gradients are used for updating weights.

3. Final Answer:

   To clear old gradients so new gradients are calculated correctly -> Option D

4. Quick Check:

   Zero gradients = fresh gradient calculation [OK]

Hint: Always zero gradients before backward pass to avoid accumulation [OK]

Common Mistakes:

• Skipping zero_grad() causes wrong gradient sums
• Confusing zeroing gradients with resetting weights
• Thinking zero_grad() speeds training by skipping steps

2. Which of the following is the correct way to zero gradients for an optimizer named optimizer in PyTorch?

easy

A. optimizer.clear_grad()

B. optimizer.zero_gradients()

C. optimizer.zero_grad()

D. optimizer.reset_grad()

Solution

1. Step 1: Recall PyTorch optimizer API

   The correct method to zero gradients is zero_grad() on the optimizer object.

2. Step 2: Check method names for zeroing gradients

   Only zero_grad() is valid; others are incorrect or do not exist.

3. Final Answer:

   optimizer.zero_grad() -> Option C

4. Quick Check:

   Use zero_grad() method to clear gradients [OK]

Hint: Remember method name ends with '_grad', not '_gradients' [OK]

Common Mistakes:

• Using zero_gradients() which does not exist
• Confusing with clear_grad() or reset_grad() methods
• Calling zero_grad on model instead of optimizer

3. What will be the output of the following PyTorch code snippet?

import torch
x = torch.tensor([2.0], requires_grad=True)
optimizer = torch.optim.SGD([x], lr=0.1)
optimizer.zero_grad()
y = x * 3
loss = y ** 2
loss.backward()
print(x.grad.item())

medium

A. 36.0

B. 12.0

C. 18.0

D. 6.0

Solution

1. Step 1: Calculate loss and gradients

   y = 3 * x = 6.0; loss = y^2 = 36.0; gradient of loss wrt x is d(loss)/dx = 2*y * dy/dx = 2*6*3 = 36.

2. Step 2: Check printed gradient value

   x.grad will hold 36.0 after backward, but since x.grad is printed directly, it shows 36.0.

3. Final Answer:

   36.0 -> Option A

4. Quick Check:

   Gradient = 2*y*3 = 36.0 [OK]

Hint: Use chain rule: d(loss)/dx = 2*y * dy/dx [OK]

Common Mistakes:

• Calculating gradient as 12.0 by missing chain rule
• Confusing loss value with gradient value
• Forgetting to call zero_grad() before backward

4. Consider this training loop snippet:

for data, target in dataloader:
    output = model(data)
    loss = loss_fn(output, target)
    loss.backward()
    optimizer.step()
    optimizer.zero_grad()

What is the problem here?

medium

A. No problem, this is the correct order

B. Gradients are zeroed after optimizer.step(), causing stale gradients in next iteration

C. loss.backward() should be called after optimizer.step()

D. optimizer.zero_grad() should be called before loss.backward() to clear old gradients

Solution

1. Step 1: Understand correct order of zeroing gradients

   Gradients must be zeroed before calling loss.backward() to avoid accumulation.

2. Step 2: Identify error in given code

   Here, zero_grad() is called after optimizer.step(), so old gradients accumulate during backward pass.

3. Final Answer:

   Gradients are zeroed after optimizer.step(), causing stale gradients in next iteration -> Option B

4. Quick Check:

   Zero gradients before backward pass [OK]

Hint: Call zero_grad() before backward() to reset gradients [OK]

Common Mistakes:

• Calling zero_grad() after optimizer.step() instead of before backward()
• Swapping order of backward() and step()
• Assuming zero_grad() is optional

5. You want to accumulate gradients over 4 mini-batches before updating model weights to simulate a larger batch size. Which code snippet correctly implements this using zeroing gradients in PyTorch?

hard

A. optimizer.zero_grad()
for i, (data, target) in enumerate(dataloader):
output = model(data)
loss = loss_fn(output, target) / 4
loss.backward()
if (i + 1) % 4 == 0:
optimizer.step()
optimizer.zero_grad()

B. for i, (data, target) in enumerate(dataloader):
optimizer.zero_grad()
output = model(data)
loss = loss_fn(output, target)
loss.backward()
if (i + 1) % 4 == 0:
optimizer.step()

C. for i, (data, target) in enumerate(dataloader):
output = model(data)
loss = loss_fn(output, target)
loss.backward()
optimizer.step()
optimizer.zero_grad()

D. optimizer.zero_grad()
for data, target in dataloader:
output = model(data)
loss = loss_fn(output, target)
loss.backward()
optimizer.step()
optimizer.zero_grad()

Solution

1. Step 1: Understand gradient accumulation

   To accumulate gradients over multiple batches, zero_grad() is called once before loop, then backward() accumulates gradients.

2. Step 2: Check correct update and zeroing frequency

   After 4 batches, optimizer.step() updates weights, then zero_grad() clears gradients for next accumulation cycle.

3. Final Answer:

   Code snippet C correctly accumulates gradients and updates every 4 batches -> Option A

4. Quick Check:

   Zero once, accumulate 4 backward calls, then step and zero [OK]

Hint: Zero gradients once, step and zero every N batches to accumulate [OK]

Common Mistakes:

• Zeroing gradients inside loop every batch prevents accumulation
• Not dividing loss by accumulation steps causing large gradients
• Calling optimizer.step() every batch defeats accumulation purpose