This repository contains code and resources for building a Handwritten Digits Classifier using PyTorch. The goal of this project is to prototype a system for optical character recognition (OCR) on handwritten characters, specifically focusing on the MNIST database of handwritten digits.
As a machine learning engineer, you have been tasked with providing a proof of concept for the OCR system. In this project, you will preprocess the MNIST dataset, build and train a neural network using PyTorch, and fine-tune the model for optimal performance.
To get started with the project, follow these steps:
-
Set up the environment:
- Ensure you have Python 3.x installed.
-
Explore the code and resources in the repository.
-
Open the Jupyter Notebook:
jupyter notebook -
Open the provided Jupyter Notebook
Handwritten_Digits_Classifier.ipynb. -
Follow the instructions in the notebook to preprocess the dataset, build the neural network, and train the model.
-
Experiment with different hyperparameters, architectures, or techniques to improve the performance of the model.
-
load the model checkpoint and test the model on the test dataset.
- The MNIST database of handwritten digits.
- PyTorch, an open-source machine learning library used for building and training neural networks.
Special thanks to UDACITY & AWS for providing the opportunity to work on this project.