Why Multi-head attention in PyTorch? - Purpose & Use Cases

The Big Idea

What if your model could listen to every part of a sentence at once and truly understand it?

The Scenario

Imagine trying to understand a long conversation by focusing on just one word at a time. You might miss important connections between different parts of the conversation.

The Problem

Manually tracking relationships between all words in a sentence is slow and confusing. It's easy to miss key details or misunderstand the meaning because you can only focus on one thing at once.

The Solution

Multi-head attention lets the model look at many parts of the sentence at the same time. It learns different ways to connect words, so it understands the full meaning better and faster.

Before vs After

✗ Before

attention = softmax(Q @ K.T) @ V

✓ After

outputs = [softmax(Q_i @ K_i.T) @ V_i for i in heads]
multi_head_output = concat(outputs) @ W_o

What It Enables

It enables models to understand complex language by focusing on multiple relationships simultaneously, improving tasks like translation and summarization.

Real Life Example

When you use a voice assistant, multi-head attention helps it understand your request by considering different words and their connections all at once, making replies smarter and more accurate.

Key Takeaways

Manual focus on single word relations is slow and limited.

Multi-head attention looks at many word connections at once.

This leads to better understanding and faster learning in language tasks.

Practice

1. What is the main purpose of multi-head attention in PyTorch's MultiheadAttention layer?

easy

A. To reduce the size of the input data

B. To perform simple addition of input vectors

C. To increase the number of layers in the model

D. To allow the model to focus on different parts of the input simultaneously

Solution

1. Step 1: Understand the concept of multi-head attention

   Multi-head attention splits the input into multiple parts (heads) to focus on different information simultaneously.

2. Step 2: Relate to PyTorch's MultiheadAttention layer

   This layer implements this idea to improve model understanding by attending to various input features at once.

3. Final Answer:

   To allow the model to focus on different parts of the input simultaneously -> Option D

4. Quick Check:

   Multi-head attention = focus on multiple input parts [OK]

Hint: Multi-head means multiple focus points at once [OK]

Common Mistakes:

• Thinking it reduces input size
• Confusing it with adding layers
• Assuming it just adds vectors

2. Which of the following is the correct way to create a multi-head attention layer with 8 heads and embedding dimension 64 in PyTorch?

easy

A. torch.nn.MultiheadAttention(8, 64)

B. torch.nn.MultiheadAttention(embed_dim=64, num_heads=8)

C. torch.nn.MultiheadAttention(num_heads=64, embed_dim=8)

D. torch.nn.MultiheadAttention(64)

Solution

1. Step 1: Check PyTorch MultiheadAttention parameter order

   The constructor requires embed_dim first, then num_heads as keyword arguments.

2. Step 2: Verify correct syntax

   torch.nn.MultiheadAttention(embed_dim=64, num_heads=8) correctly uses embed_dim=64 and num_heads=8. torch.nn.MultiheadAttention(num_heads=64, embed_dim=8) swaps parameters incorrectly, A and D use positional arguments wrongly.

3. Final Answer:

   torch.nn.MultiheadAttention(embed_dim=64, num_heads=8) -> Option B

4. Quick Check:

   embed_dim first, num_heads second [OK]

Hint: Remember: embed_dim first, num_heads second in constructor [OK]

Common Mistakes:

• Swapping embed_dim and num_heads
• Using positional args incorrectly
• Missing one of the required parameters

3. Given the following PyTorch code, what is the shape of the output tensor attn_output?

import torch
mha = torch.nn.MultiheadAttention(embed_dim=16, num_heads=4)
query = torch.rand(5, 3, 16)  # (sequence_length, batch_size, embed_dim)
key = torch.rand(5, 3, 16)
value = torch.rand(5, 3, 16)
attn_output, attn_weights = mha(query, key, value)

medium

A. (5, 3, 16)

B. (3, 5, 16)

C. (5, 16, 3)

D. (3, 16, 5)

Solution

1. Step 1: Understand input shapes for MultiheadAttention

   Inputs are (sequence_length, batch_size, embed_dim). Here, query shape is (5, 3, 16).

2. Step 2: Check output shape of attn_output

   Output shape matches query shape: (5, 3, 16).

3. Final Answer:

   (5, 3, 16) -> Option A

4. Quick Check:

   Output shape = query shape = (5, 3, 16) [OK]

Hint: Output shape matches query input shape [OK]

Common Mistakes:

• Mixing batch and sequence dimensions
• Confusing embed_dim position
• Assuming output shape changes with heads

4. Identify the error in this PyTorch multi-head attention code snippet:

import torch
mha = torch.nn.MultiheadAttention(embed_dim=32, num_heads=4)
query = torch.rand(10, 2, 32)
key = torch.rand(10, 2, 16)
value = torch.rand(10, 2, 32)
output, weights = mha(query, key, value)

medium

A. The embed_dim of key does not match embed_dim of query and value

B. The batch size is incorrect

C. The sequence length is too large

D. The number of heads is too high

Solution

1. Step 1: Check embed_dim consistency

   MultiheadAttention requires query, key, and value to have the same embedding dimension. Here, key has embed_dim 16, while query and value have 32.

2. Step 2: Identify the error cause

   This mismatch causes a runtime error because the attention mechanism cannot align different embedding sizes.

3. Final Answer:

   The embed_dim of key does not match embed_dim of query and value -> Option A

4. Quick Check:

   Embed dims must match for query, key, value [OK]

Hint: Check all embed_dims match before calling attention [OK]

Common Mistakes:

• Ignoring embed_dim mismatch
• Assuming batch size causes error
• Thinking sequence length is the problem

5. You want to apply multi-head attention to a batch of sentences with varying lengths. Which PyTorch feature helps you handle padding tokens correctly during attention computation?

hard

A. Increasing the number of heads to cover padding

B. Setting batch_first=True in MultiheadAttention

C. Using the key_padding_mask argument in MultiheadAttention

D. Manually removing padding tokens before attention

Solution

1. Step 1: Understand padding in variable-length sequences

   Sentences have different lengths, so padding tokens are added to make batch tensors uniform.

2. Step 2: Use key_padding_mask to ignore padding

   PyTorch's MultiheadAttention accepts key_padding_mask to mask out padded positions during attention, preventing them from affecting results.

3. Final Answer:

   Using the key_padding_mask argument in MultiheadAttention -> Option C

4. Quick Check:

   Mask padding tokens with key_padding_mask [OK]

Hint: Use key_padding_mask to ignore padded tokens [OK]

Common Mistakes:

• Removing padding manually (inefficient)
• Confusing batch_first with padding handling
• Increasing heads does not fix padding