Self-attention mechanism in PyTorch

Introduction

Self-attention helps a model focus on important parts of the input when making decisions. It looks at all parts of the input and decides which parts matter most for each piece.

Understanding relationships between words in a sentence for translation.

Finding important features in an image for recognition.

Analyzing time series data where past and future points influence each other.

Improving chatbot responses by focusing on relevant conversation parts.

Syntax

PyTorch

import torch
import torch.nn.functional as F

def self_attention(query, key, value):
    scores = torch.matmul(query, key.transpose(-2, -1)) / (query.size(-1) ** 0.5)
    weights = F.softmax(scores, dim=-1)
    output = torch.matmul(weights, value)
    return output, weights

query, key, and value are tensors representing the input in different ways.

The scores measure how much each part should pay attention to others.

Examples

Simple example where query, key, and value are single vectors.

PyTorch

query = torch.tensor([[1., 0., 1.]])
key = torch.tensor([[1., 0., 1.]])
value = torch.tensor([[0., 1., 0.]])
output, weights = self_attention(query, key, value)

Batch of 2 sequences, each with 3 items and 4 features.

PyTorch

query = torch.randn(2, 3, 4)  # batch=2, seq_len=3, features=4
key = torch.randn(2, 3, 4)
value = torch.randn(2, 3, 4)
output, weights = self_attention(query, key, value)

Sample Model

This program computes self-attention on a small example with 3 items in a sequence. It prints the attention weights and the output vectors.

PyTorch

import torch
import torch.nn.functional as F

def self_attention(query, key, value):
    scores = torch.matmul(query, key.transpose(-2, -1)) / (query.size(-1) ** 0.5)
    weights = F.softmax(scores, dim=-1)
    output = torch.matmul(weights, value)
    return output, weights

# Example input: batch size 1, sequence length 3, feature size 4
query = torch.tensor([[[1., 0., 1., 0.],
                       [0., 2., 0., 2.],
                       [1., 1., 1., 1.]]])
key = query.clone()
value = torch.tensor([[[0., 1., 0., 1.],
                       [1., 0., 1., 0.],
                       [0., 0., 1., 1.]]])

output, weights = self_attention(query, key, value)

print("Attention weights:", weights)
print("Output:", output)

Important Notes

Self-attention uses the same input for query, key, and value to find relationships within the input.

Dividing by the square root of the feature size helps keep the scores stable.

Softmax turns scores into probabilities that sum to 1, showing importance.

Summary

Self-attention helps models focus on important parts of input data.

It compares each part to every other part using query, key, and value.

Outputs are weighted sums showing combined important information.

Practice

1. What is the main purpose of the self-attention mechanism in models like Transformers?

easy

A. To convert input data into images

B. To increase the size of the input data

C. To reduce the number of layers in the model

D. To help the model focus on important parts of the input data

Solution

1. Step 1: Understand self-attention's role

   Self-attention helps the model decide which parts of the input are important by comparing each part to others.

2. Step 2: Identify the main goal

   The main goal is to focus on important parts, not to change input size or model layers.

3. Final Answer:

   To help the model focus on important parts of the input data -> Option D

4. Quick Check:

   Self-attention = focus on important parts [OK]

Hint: Self-attention means focusing on important input parts [OK]

Common Mistakes:

• Thinking self-attention changes input size
• Confusing self-attention with model depth
• Assuming it converts data formats

2. Which of the following is the correct PyTorch code snippet to compute scaled dot-product attention scores given query Q and key K tensors?

easy

A. scores = torch.matmul(Q, K.transpose(-2, -1)) / math.sqrt(Q.size(-1))

B. scores = torch.matmul(K, Q.transpose(-2, -1)) * math.sqrt(Q.size(-1))

C. scores = torch.matmul(Q, K) / Q.size(0)

D. scores = torch.matmul(Q.transpose(-2, -1), K) / math.sqrt(K.size(-1))

Solution

1. Step 1: Recall scaled dot-product attention formula

   Attention scores = (Q x K^T) / sqrt(d_k), where d_k is the dimension of keys.

2. Step 2: Match formula with code

   scores = torch.matmul(Q, K.transpose(-2, -1)) / math.sqrt(Q.size(-1)) correctly transposes K, multiplies Q and K^T, and divides by sqrt of last dimension size.

3. Final Answer:

   scores = torch.matmul(Q, K.transpose(-2, -1)) / math.sqrt(Q.size(-1)) -> Option A

4. Quick Check:

   Scaled dot-product = QxK^T / sqrt(d_k) [OK]

Hint: Transpose keys and divide by sqrt of key dimension [OK]

Common Mistakes:

• Multiplying K by Q instead of Q by K^T
• Forgetting to transpose keys
• Using multiplication instead of division by sqrt

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

import torch
Q = torch.randn(2, 4, 8)  # batch=2, seq_len=4, dim=8
K = torch.randn(2, 4, 8)
V = torch.randn(2, 4, 8)

scores = torch.matmul(Q, K.transpose(-2, -1)) / (8 ** 0.5)
weights = torch.nn.functional.softmax(scores, dim=-1)
attn_output = torch.matmul(weights, V)

medium

A. (4, 2, 8)

B. (2, 4, 8)

C. (2, 8, 4)

D. (4, 8, 2)

Solution

1. Step 1: Analyze input tensor shapes

   Q, K, V have shape (batch=2, seq_len=4, dim=8).

2. Step 2: Understand matrix multiplications

   Scores shape: (2,4,4), weights shape: (2,4,4), multiplying weights (2,4,4) by V (2,4,8) results in (2,4,8).

3. Final Answer:

   (2, 4, 8) -> Option B

4. Quick Check:

   Output shape = (batch, seq_len, dim) = (2,4,8) [OK]

Hint: Output shape matches Q and V batch and seq_len dims [OK]

Common Mistakes:

• Confusing batch and sequence dimensions
• Mixing up transpose effects on shape
• Assuming output shape changes dimension size

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

import torch
Q = torch.randn(1, 3, 5)
K = torch.randn(1, 3, 5)
V = torch.randn(1, 3, 5)

scores = torch.matmul(Q, K) / (5 ** 0.5)
weights = torch.nn.functional.softmax(scores, dim=-1)
output = torch.matmul(weights, V)

medium

A. Q and V should have the same shape

B. Softmax should be applied on dim=0 instead of dim=-1

C. K should be transposed before multiplication with Q

D. Division by sqrt(5) is incorrect

Solution

1. Step 1: Check matrix multiplication for scores

   Q shape is (1,3,5), K shape is (1,3,5). To multiply Q by K^T, K must be transposed on last two dims.

2. Step 2: Identify missing transpose

   Code multiplies Q by K directly, causing shape mismatch or wrong attention scores.

3. Final Answer:

   K should be transposed before multiplication with Q -> Option C

4. Quick Check:

   QxK^T needed, missing transpose causes error [OK]

Hint: Always transpose keys before multiplying with queries [OK]

Common Mistakes:

• Skipping transpose of K
• Applying softmax on wrong dimension
• Confusing shapes of Q and V

5. You want to implement multi-head self-attention in PyTorch with 4 heads. If your input tensor has shape (batch=2, seq_len=6, dim=32), what should be the dimension size of each head's query, key, and value vectors?

hard

A. 8

B. 4

C. 16

D. 32

Solution

1. Step 1: Understand multi-head attention dimension split

   Total dimension 32 is split evenly across 4 heads, so each head gets 32 / 4 = 8 dimensions.

2. Step 2: Confirm dimension per head

   Each head's query, key, and value vectors have dimension 8 to keep total dimension consistent.

3. Final Answer:

   8 -> Option A

4. Quick Check:

   32 total dim / 4 heads = 8 per head [OK]

Hint: Divide total dim by number of heads for each head size [OK]

Common Mistakes:

• Using total dimension for each head
• Dividing by wrong number of heads
• Confusing batch size with dimension