Implement MNIST experiment

[HerrCooker]

    self.layers = nn.Sequential(
        nn.Flatten(),
        nn.Linear(28*28, 20),
        nn.ReLU(),
        nn.Linear(20, 10)
    )

https://colab.research.google.com/github/csc-training/intro-to-dl/blob/master/day1/02-pytorch-mnist-mlp.ipynb#scrollTo=YIHDyz7abU6F

Why not load dataset already as batches?

• we want the dataloader to handle sample to batch conversion

  • this way, the batchSize can be dynamically changed

• test if test dataset is corrupted

How to initialize weights?

• xavier initialization

Bug in loss vs validationLoss