A ML and DS project on fashion products classification based on MNIST fashion dataset.
The goal of this project is to classify fashion products based on the MNIST fashion dataset. The dataset contains 10 classes of fashion products.
The dataset used for this project is the Fashion MNIST
The models folder contains the trained model. The notebooks folder contains the exploratory data analysis and model training notebooks. The evaluate_model.py
script is used to evaluate the model on the test set and save the output in the output.txt
file.
Fashion-Classification
│
├── dataset
│ └── fashion-mnist_test.csv
│
├── models
│ └── model_v2.pkl
│
├── notebooks
│ ├── model_training.ipynb
│ └── exploratory_data_analysis.ipynb
│
├── evaluate_model.py
├── model_summary.txt
├── output.txt
├── .gitignore
├── LICENSE
├── README.md
└── requirements.txt
- fastai
- scikit-learn
The model used for this project is a ResNet34 model with custom head layer which is trained and dine-tuned on the Fashion MNIST dataset. The model was trained for 5 epochs with a batch size of 32. The model achieved an accuracy of 89.9% on the test set.
- Clone the repository and navigate to the project directory.
git clone https://github.com/abir0/Fashion-Classification.git
cd Fashion-Classification
- Create a virtual environment and activate it.
python -m venv venv
venv\Scripts\activate
- Install the dependencies
pip install -r requirements.txt
- Run the
evaluate_model.py
script
python evaluate_model.py
- The output will be saved in the
output.txt
file.