The jupyter notebook Training of a Convolutional Neural Network for the MNIST classification problem.ipynb
, into the src
folder contains the training and testing of a convolutional neural network for the MNIST classification problem. The notebook is self-contained and can be run on Google Colab.
Access the colab notebook here.
Take in consideration that the training of the tuner can take a long time.
In case you want to test the notebook without the need of execute the tuner, you can use the keras_tuner_dir
, which contains the results of the tuner.
Also, in case you are interested in the TensorBoard results, they are saved in the logs
folder.
The result model is saved in the model.h5
file.
A brief analysis of the jupyter notebook is available in the cnn_results.pdf
file.