Model optimization (pruning, quantization) in Computer Vision - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is model pruning in machine learning?

Model pruning is a technique that removes less important parts of a neural network, like some connections or neurons, to make the model smaller and faster without losing much accuracy.

beginner

Explain quantization in the context of neural networks.

Quantization reduces the precision of the numbers used in a model, for example changing 32-bit floats to 8-bit integers, which makes the model smaller and faster to run, especially on devices with limited resources.

intermediate

How does pruning help improve model performance?

Pruning removes unnecessary parts of the model, which reduces its size and speeds up predictions, making it easier to run on devices with less memory or slower processors.

intermediate

What is a common trade-off when applying quantization to a model?

The trade-off is between model size and speed versus accuracy. Quantization makes the model smaller and faster but can slightly reduce its accuracy due to lower number precision.

advanced

Name two common types of pruning used in model optimization.

Two common types are: 1) Weight pruning, which removes individual connections with small weights, and 2) Structured pruning, which removes entire neurons or filters to simplify the model structure.

What does pruning mainly remove from a neural network?

A Training data samples

B Output layers

C Input features

D Less important connections or neurons

Pruning removes less important connections or neurons to reduce model size and speed up inference.

Quantization typically changes model numbers from:

A Integers to floats

B 32-bit floats to 8-bit integers

C 8-bit integers to 32-bit floats

D Strings to numbers

Quantization reduces precision, often converting 32-bit floating-point numbers to 8-bit integers.

Which is a benefit of model pruning?

A Speeds up model inference

B Requires more memory

C Adds more layers

D Increases model size

Pruning reduces model size and speeds up inference by removing unnecessary parts.

What is a possible downside of quantization?

A Model becomes slower

B Model uses more memory

C Model accuracy may slightly decrease

D Model requires more training data

Quantization can slightly reduce accuracy due to lower number precision.

Structured pruning removes:

A Entire neurons or filters

B Individual weights only

C Training samples

D Input features

Structured pruning removes whole neurons or filters to simplify the model.

Describe how pruning and quantization help optimize a computer vision model for deployment on mobile devices.

Think about how smaller and faster models help on phones.

Explain the trade-offs involved when applying pruning and quantization to a neural network.

Consider what you lose and gain with these techniques.

Practice

1. What is the main goal of model pruning in computer vision?

easy

A. To remove less important parts of the model to reduce size

B. To increase the number of layers in the model

C. To add more training data for better accuracy

D. To convert the model to a different programming language

Solution

1. Step 1: Understand pruning concept

   Pruning means removing parts of the model that contribute less to its output.

2. Step 2: Identify pruning goal

   The goal is to reduce model size and speed up inference by cutting unnecessary parts.

3. Final Answer:

   To remove less important parts of the model to reduce size -> Option A

4. Quick Check:

   Pruning = Remove less important parts [OK]

Hint: Pruning cuts unneeded parts to shrink model size [OK]

Common Mistakes:

• Thinking pruning adds layers instead of removing
• Confusing pruning with data augmentation
• Believing pruning changes programming language

2. Which of the following is the correct way to apply quantization in TensorFlow Lite?

easy

A. model = tf.lite.TFLiteConverter.from_keras_model(model).convert()

B. converter.optimizations = [tf.lite.Optimize.DEFAULT]

C. model.compile(optimizer='adam', loss='mse')

D. model.fit(x_train, y_train, epochs=10)

Solution

1. Step 1: Identify quantization syntax

   In TensorFlow Lite, quantization is enabled by setting converter.optimizations to Optimize.DEFAULT.

2. Step 2: Check other options

   model = tf.lite.TFLiteConverter.from_keras_model(model).convert() converts model but does not enable quantization. Options B and C are training commands, not quantization.

3. Final Answer:

   converter.optimizations = [tf.lite.Optimize.DEFAULT] -> Option B

4. Quick Check:

   Quantization flag = converter.optimizations [OK]

Hint: Quantization needs converter.optimizations set to Optimize.DEFAULT [OK]

Common Mistakes:

• Confusing model conversion with quantization
• Using training commands instead of conversion flags
• Missing the optimization setting for quantization

3. Given this PyTorch pruning code snippet, what will be the output size of the model's first linear layer weights after pruning 20% of connections?

import torch
import torch.nn.utils.prune as prune

model = torch.nn.Sequential(
    torch.nn.Linear(100, 50),
    torch.nn.ReLU()
)
prune.l1_unstructured(model[0], name='weight', amount=0.2)
pruned_weights = model[0].weight
print((pruned_weights != 0).sum().item())

medium

A. 8000

B. 5000

C. 10000

D. 4000

Solution

1. Step 1: Calculate total weights

   The first linear layer has 100 inputs and 50 outputs, so total weights = 100 * 50 = 5000.

2. Step 2: Calculate remaining weights after pruning

   Pruning 20% removes 20% of weights, so remaining weights = 80% of 5000 = 4000.

3. Step 3: Understand pruning method

   PyTorch's l1_unstructured pruning does not remove weights but masks them, so the weight tensor size remains 5000, but the number of non-zero weights is 4000.

4. Step 4: Check print output

   The print statement counts non-zero weights, so output is 4000.

5. Final Answer:

   4000 -> Option D

6. Quick Check:

   5000 * 0.8 = 4000 [OK]

Hint: Remaining weights = total * (1 - pruning amount) [OK]

Common Mistakes:

• Calculating total weights incorrectly
• Using pruning amount as remaining instead of removed
• Confusing layer input/output dimensions

4. You tried to quantize a model but got an error: AttributeError: 'TFLiteConverter' object has no attribute 'optimizations'. What is the likely cause?

medium

A. Quantization requires training the model again

B. Model is too large to quantize

C. Using an outdated TensorFlow version without quantization support

D. The model has no weights to quantize

Solution

1. Step 1: Understand the error

   The error says the converter object lacks 'optimizations' attribute, meaning the TensorFlow version is old.

2. Step 2: Identify cause

   Older TensorFlow versions do not support the 'optimizations' attribute needed for quantization.

3. Final Answer:

   Using an outdated TensorFlow version without quantization support -> Option C

4. Quick Check:

   Missing attribute = outdated TensorFlow [OK]

Hint: Check TensorFlow version supports quantization features [OK]

Common Mistakes:

• Assuming model size causes attribute error
• Thinking quantization needs retraining always
• Believing model without weights causes this error

5. You want to deploy a computer vision model on a mobile device with limited memory and CPU. Which combination of optimization techniques is best to reduce model size and speed up inference without much accuracy loss?

hard

A. Apply pruning to remove unimportant weights, then quantize weights to 8-bit integers

B. Only increase model layers to improve accuracy

C. Use full precision weights and no pruning for best accuracy

D. Train longer without any model size changes

Solution

1. Step 1: Understand device constraints

   Mobile devices have limited memory and CPU, so model size and speed matter.

2. Step 2: Choose optimization techniques

   Pruning removes unnecessary weights reducing size; quantization reduces number precision speeding inference.

3. Step 3: Combine pruning and quantization

   Using both together reduces size and speeds up model with minimal accuracy loss.

4. Final Answer:

   Apply pruning to remove unimportant weights, then quantize weights to 8-bit integers -> Option A

5. Quick Check:

   Pruning + quantization = smaller, faster model [OK]

Hint: Combine pruning and quantization for efficient mobile models [OK]

Common Mistakes:

• Only increasing layers without optimization
• Ignoring quantization benefits
• Assuming full precision is always best for deployment