Using PyTorch to create and train a two layer convolutional neural net for MNIST dataset classification
Dataset is autodownloaded from PyTorch datasets when file is run
Trained model is saved as testModel.ckpt -
1st iteration: Model Accuracy on the 10000 test images: 98.84 %
2nd iteration after small edits: Model Accuracy on the 10000 test images: 99.01 %