Model Pipeline - Saving and loading models
This pipeline shows how a machine learning model is trained, saved to a file, and later loaded back to make predictions without retraining.
0
0
Saving and loading models in ML Python - Model Pipeline Trace
Data Flow - 6 Stages
1Data input
1000 rows x 10 columns→Load dataset with 10 features per example→1000 rows x 10 columns
[[5.1, 3.5, 1.4, ..., 0.2], [4.9, 3.0, 1.4, ..., 0.2], ...]
↓
2Train/test split
1000 rows x 10 columns→Split data into 800 training and 200 testing rows→Train: 800 rows x 10 columns, Test: 200 rows x 10 columns
Train sample: [5.1, 3.5, 1.4, ..., 0.2], Test sample: [6.7, 3.1, 4.7, ..., 1.5]
↓
3Model training
800 rows x 10 columns→Train model on training data→Trained model object
Model weights updated after training
↓
4Model saving
Trained model object→Save model to file 'model.pkl'→File 'model.pkl' saved on disk
Binary file storing model parameters
↓
5Model loading
File 'model.pkl'→Load model from file→Loaded model object identical to trained model
Model ready for prediction without retraining
↓
6Prediction
200 rows x 10 columns→Use loaded model to predict on test data→200 rows x 1 column (predicted labels)
[0, 1, 0, 2, 1, ...]
Training Trace - Epoch by Epoch
Loss
1.0 |****
0.8 |****
0.6 |****
0.4 |****
0.2 |****
0.0 +----
1 2 3 4 5 Epochs| Epoch | Loss ↓ | Accuracy ↑ | Observation |
|---|---|---|---|
| 1 | 0.85 | 0.60 | Model starts learning with moderate accuracy |
| 2 | 0.65 | 0.75 | Loss decreases and accuracy improves |
| 3 | 0.50 | 0.82 | Model continues to improve |
| 4 | 0.40 | 0.87 | Good convergence observed |
| 5 | 0.35 | 0.90 | Training stabilizes with high accuracy |
Prediction Trace - 3 Layers
Layer 1: Input test sample
Layer 2: Model loaded from file
Layer 3: Prediction
Model Quiz - 3 Questions
Test your understanding
Why do we save a trained model to a file?
Key Insight
Saving and loading models allows us to keep the trained knowledge and use it anytime without repeating the training process, saving time and resources.