Gradient accumulation and zeroing in PyTorch - ML Experiment: Train & Evaluate

Experiment - Gradient accumulation and zeroing

Problem:You are training a neural network on a small GPU that cannot fit a large batch size. Currently, you use a batch size of 16, but the model trains slowly and the gradients are reset after every batch. This limits the effective batch size and slows learning.

Current Metrics:Training loss decreases slowly, validation accuracy reaches about 70% after 10 epochs.

Issue:The model trains slowly because the batch size is small. Gradients are zeroed after every batch, so the model cannot accumulate gradients over multiple batches to simulate a larger batch size.

Your Task

Implement gradient accumulation to simulate a larger batch size of 64 by accumulating gradients over 4 batches before updating weights. Ensure gradients are zeroed correctly to avoid incorrect updates.

Do not change the batch size from 16.

Do not change the model architecture.

Use PyTorch for implementation.

Hint 1

Hint 2

Hint 3

Solution

PyTorch

import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import DataLoader, TensorDataset

# Simple dataset
X = torch.randn(1000, 10)
y = (X.sum(dim=1) > 0).long()
dataset = TensorDataset(X, y)
dataloader = DataLoader(dataset, batch_size=16, shuffle=True)

# Simple model
class SimpleNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.fc = nn.Linear(10, 2)
    def forward(self, x):
        return self.fc(x)

model = SimpleNet()
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=0.1)

accumulation_steps = 4  # accumulate gradients over 4 batches

for epoch in range(10):
    running_loss = 0.0
    optimizer.zero_grad()  # zero gradients before starting epoch
    for i, (inputs, labels) in enumerate(dataloader):
        outputs = model(inputs)
        loss = criterion(outputs, labels)
        loss = loss / accumulation_steps  # normalize loss
        loss.backward()  # accumulate gradients

        if (i + 1) % accumulation_steps == 0:
            optimizer.step()  # update weights
            optimizer.zero_grad()  # zero gradients after update

        running_loss += loss.item() * accumulation_steps

    print(f"Epoch {epoch+1}, Loss: {running_loss / len(dataloader):.4f}")

Added gradient accumulation by dividing loss by accumulation_steps and calling loss.backward() multiple times before optimizer.step().

Moved optimizer.zero_grad() to run only after optimizer.step() to avoid zeroing gradients too early.

Used a counter to update weights every 4 batches to simulate batch size of 64.

Results Interpretation

Before: Batch size 16, gradients zeroed every batch, slow training, validation accuracy ~70%.

After: Effective batch size 64 via gradient accumulation, gradients zeroed only after weight update, faster training, validation accuracy ~78%.

Gradient accumulation allows simulating larger batch sizes on limited hardware by accumulating gradients over multiple smaller batches before updating weights. Proper zeroing of gradients is essential to avoid incorrect updates.

Bonus Experiment

Try adding gradient clipping after accumulation to prevent exploding gradients and observe its effect on training stability.

💡 Hint

Use torch.nn.utils.clip_grad_norm_ after loss.backward() and before optimizer.step() to clip gradients.

Practice

1. What is the main purpose of zeroing gradients in PyTorch during training?

easy

A. To clear old gradients before computing new ones

B. To increase the learning rate automatically

C. To save the model weights to disk

D. To initialize the model parameters

Solution

1. Step 1: Understand gradient accumulation process

   Gradients accumulate by default in PyTorch, so old gradients must be cleared before new backward passes.

2. Step 2: Purpose of zeroing gradients

   Zeroing gradients resets them to zero to avoid mixing old and new gradient information.

3. Final Answer:

   To clear old gradients before computing new ones -> Option A

4. Quick Check:

   Zeroing gradients = clear old gradients [OK]

Hint: Zero gradients before backward to avoid mixing old info [OK]

Common Mistakes:

• Thinking zeroing resets model weights
• Confusing zeroing with saving model
• Believing zeroing changes learning rate

2. Which of the following is the correct way to zero gradients of a model named model in PyTorch?

easy

A. model.zero_grad()

B. model.clear_grad()

C. model.reset_gradients()

D. model.zero_gradients()

Solution

1. Step 1: Recall PyTorch API for zeroing gradients

   The official method to zero gradients is zero_grad() on the model or optimizer.

2. Step 2: Check method names

   Only model.zero_grad() is a valid PyTorch method; others are incorrect or do not exist.

3. Final Answer:

   model.zero_grad() -> Option A

4. Quick Check:

   Correct method name = zero_grad() [OK]

Hint: Use model.zero_grad() to clear gradients [OK]

Common Mistakes:

• Using non-existent methods like clear_grad()
• Confusing zero_grad with reset_gradients
• Adding extra letters to method names

3. Consider this PyTorch code snippet:

optimizer.zero_grad()
for i, batch in enumerate(data_loader):
    outputs = model(batch)
    loss = loss_fn(outputs, targets)
    loss.backward()
    if (i + 1) % 4 == 0:
        optimizer.step()
        optimizer.zero_grad()

What is the effect of this code on gradient updates?

medium

A. Updates model every batch without accumulation

B. Updates model only once after all batches

C. Does not update model at all

D. Accumulates gradients over 4 batches before updating

Solution

1. Step 1: Analyze the loop and optimizer calls

   Gradients are zeroed before loop and after every 4 batches. Backward is called every batch, so gradients accumulate for 4 batches.

2. Step 2: Understand optimizer.step() timing

   Optimizer updates weights every 4 batches, using accumulated gradients, then zeroes gradients to start fresh.

3. Final Answer:

   Accumulates gradients over 4 batches before updating -> Option D

4. Quick Check:

   Step every 4 batches = gradient accumulation [OK]

Hint: Step optimizer after several backward calls to accumulate [OK]

Common Mistakes:

• Assuming update happens every batch
• Forgetting gradients accumulate without zeroing
• Thinking zero_grad inside loop clears gradients too early

4. Identify the error in this PyTorch training loop:

for batch in data_loader:
    optimizer.zero_grad()
    outputs = model(batch)
    loss = loss_fn(outputs, targets)
    loss.backward()
    optimizer.step()
    loss.backward()

What is wrong here?

medium

A. Not calling optimizer.zero_grad() before backward

B. Missing optimizer.step() call

C. Calling loss.backward() twice without zeroing gradients

D. Using loss.backward() before computing loss

Solution

1. Step 1: Review order of backward and optimizer calls

   Loss.backward() is called twice without zeroing gradients in between, causing gradient accumulation unintentionally.

2. Step 2: Understand consequences of double backward

   Second backward adds gradients again, which is usually a mistake unless intentional accumulation is desired.

3. Final Answer:

   Calling loss.backward() twice without zeroing gradients -> Option C

4. Quick Check:

   Double backward without zero_grad causes wrong gradients [OK]

Hint: Call backward once per batch unless accumulating intentionally [OK]

Common Mistakes:

• Calling backward multiple times without zeroing
• Forgetting zero_grad before backward
• Misplacing optimizer.step() call

5. You want to train a large model on a GPU with limited memory by accumulating gradients over 8 mini-batches before updating weights. Which sequence correctly implements this in PyTorch?

hard

A. Call optimizer.zero_grad() every batch, call loss.backward() every batch, call optimizer.step() every batch

B. Call optimizer.zero_grad() once before loop, call loss.backward() every batch, call optimizer.step() and zero_grad() every 8 batches

C. Call optimizer.zero_grad() every 8 batches, call loss.backward() every 8 batches, call optimizer.step() every batch

D. Call optimizer.zero_grad() once before loop, call loss.backward() every 8 batches, call optimizer.step() every batch

Solution

1. Step 1: Understand gradient accumulation logic

   Gradients must be zeroed once before accumulation starts, then loss.backward() called every batch to accumulate gradients.

2. Step 2: Timing optimizer step and zero_grad

   After accumulating gradients over 8 batches, call optimizer.step() to update weights, then zero_grad() to reset gradients for next accumulation.

3. Final Answer:

   Call optimizer.zero_grad() once before loop, call loss.backward() every batch, call optimizer.step() and zero_grad() every 8 batches -> Option B

4. Quick Check:

   Zero once, backward every batch, step and zero every 8 batches [OK]

Hint: Zero once, backward each batch, step and zero after N batches [OK]

Common Mistakes:

• Zeroing gradients every batch breaks accumulation
• Calling backward only every N batches misses gradients
• Stepping optimizer too often or too rarely