Defining a model class in PyTorch - Model Pipeline Trace

Model Pipeline - Defining a model class

This pipeline shows how raw data is prepared and passed through a custom model class in PyTorch. The model learns to make predictions by adjusting its internal settings during training.

Data Flow - 6 Stages

1Data in

1000 rows x 10 columns→Raw input features representing 10 measurements per example→1000 rows x 10 columns

[[0.5, 1.2, 0.3, ..., 0.7], [1.1, 0.4, 0.9, ..., 0.2], ...]

↓

2Preprocessing

1000 rows x 10 columns→Normalize features to have mean 0 and std 1→1000 rows x 10 columns

[[-0.1, 0.5, -0.7, ..., 0.0], [0.8, -0.3, 0.4, ..., -0.5], ...]

↓

3Feature Engineering

1000 rows x 10 columns→No additional features added, pass as is→1000 rows x 10 columns

[[-0.1, 0.5, -0.7, ..., 0.0], [0.8, -0.3, 0.4, ..., -0.5], ...]

↓

4Model Trains

1000 rows x 10 columns→Pass through custom PyTorch model class with 2 linear layers and ReLU→1000 rows x 3 columns

[[0.2, 0.5, 0.3], [0.7, 0.1, 0.2], ...]

↓

5Metrics Improve

1000 rows x 3 columns→Calculate loss and accuracy, update model weights→N/A

Loss decreases from 1.2 to 0.3, accuracy increases from 40% to 85%

↓

6Prediction

1 row x 10 columns→Model outputs class scores for single example→1 row x 3 columns

[[0.4, 0.3, 0.3]]

Training Trace - Epoch by Epoch

Loss
1.2 |*****
0.9 |****
0.7 |***
0.5 |**
0.3 |*

Epoch	Loss ↓	Accuracy ↑	Observation
1	1.2	0.40	Model starts with high loss and low accuracy
2	0.9	0.55	Loss decreases, accuracy improves
3	0.7	0.65	Model learns useful patterns
4	0.5	0.75	Loss continues to drop, accuracy rises
5	0.3	0.85	Model converges with good accuracy

Prediction Trace - 5 Layers

Layer 1: Input Layer

Layer 2: First Linear Layer

Layer 3: ReLU Activation

Layer 4: Second Linear Layer

Layer 5: Softmax (optional)

Model Quiz - 3 Questions

What happens to the data shape after the first linear layer?

A It changes from 10 features to 3 features

B It stays the same with 10 features

C It changes from 10 features to 5 features

D It doubles to 20 features

Key Insight

Defining a model class in PyTorch allows you to control how data flows through layers. The model learns by adjusting weights to reduce loss and improve accuracy. Activation functions like ReLU help the model learn complex patterns by adding non-linearity.