How sarthakdem18/hrm works
This is an end-to-end MNIST digit-classification project: training loads and preprocesses data, trains and evaluates a CNN, and saves a model; prediction accepts an image or demonstrates predictions on MNIST test samples. The README also describes an interactive notebook and project/viva documentation. Training and inference relationships below are grounded in the excerpts; notebook behavior is documented but its source was not sampled.
Main entry points: train.py Training pipeline · predict.py Prediction entry point · outputs Generated plots · digit_cnn.keras Saved CNN model
Components 17
Training
- Data loading and preprocessing
handwritten_digit_recognition/train.py - Sample digit plot
handwritten_digit_recognition/train.py - CNN architecture
handwritten_digit_recognition/train.py - Training pipeline
handwritten_digit_recognition/train.py - Saved CNN model
handwritten_digit_recognition/model/digit_cnn.keras
Evaluation
- Test evaluation
handwritten_digit_recognition/train.py - Training curves
handwritten_digit_recognition/train.py - Evaluation visualizations
handwritten_digit_recognition/train.py - Generated plots
handwritten_digit_recognition/outputs
Inference
- Prediction entry point
handwritten_digit_recognition/predict.py - Image preprocessing
handwritten_digit_recognition/predict.py - Model loading
handwritten_digit_recognition/predict.py - Digit prediction
handwritten_digit_recognition/predict.py
Learning materials
- Interactive notebook
handwritten_digit_recognition/notebooks/handwritten_digit_recognition.ipynb - Project documentation
handwritten_digit_recognition/documentation/project_documentation.md
Other components
- User
- MNIST dataset
Connections 21
- User → Training pipeline: runs training
README.md - Training pipeline → Data loading and preprocessing: loads and preprocesses
handwritten_digit_recognition/train.py - Data loading and preprocessing → MNIST dataset: loads dataset
handwritten_digit_recognition/train.py - Data loading and preprocessing → Sample digit plot: creates sample plot
handwritten_digit_recognition/train.py - Training pipeline → CNN architecture: builds model
handwritten_digit_recognition/train.py - Training pipeline → Saved CNN model: saves model
handwritten_digit_recognition/train.py - Training pipeline → Training curves: plots training history
handwritten_digit_recognition/train.py - Training pipeline → Test evaluation: evaluates model
handwritten_digit_recognition/train.py - Test evaluation → Evaluation visualizations: plots results
handwritten_digit_recognition/train.py - Sample digit plot → Generated plots: writes image
handwritten_digit_recognition/train.py - Training curves → Generated plots: writes curves
handwritten_digit_recognition/train.py - Evaluation visualizations → Generated plots: writes plots
handwritten_digit_recognition/train.py - User → Prediction entry point: runs prediction
README.md - Prediction entry point → MNIST dataset: loads demo samples
handwritten_digit_recognition/predict.py - Prediction entry point → Model loading: loads model
handwritten_digit_recognition/predict.py - Model loading → Saved CNN model: reads model
handwritten_digit_recognition/predict.py - Prediction entry point → Digit prediction: predicts samples
handwritten_digit_recognition/predict.py - Digit prediction → Image preprocessing: preprocesses image
handwritten_digit_recognition/predict.py - Digit prediction → Model loading: loads when needed
handwritten_digit_recognition/predict.py - User → Interactive notebook: opens notebook
README.md - User → Project documentation: consults guide
README.md