Example configurations to locally test DLP

Classification problem

The website already populates a sample architecture (the one with Linear(4, 10), ReLU(), Linear(10, 3), Softmax()). Use that

Default Dataset: IRIS

Loss: CELoss()

Problem Type: Classification

Epochs: Any positive integer

Regression problem

Default Dataset: California Housing

Loss: MSELoss()

Problem Type: Regression

Architecture: Linear(8, 10), ReLU(), Linear(10, 1) (note that there is no softmax)

Image Classification problem

Using the website's default values, you can choose any of the default dataset.

To check file upload training, you first have to click the upload button and wait for the backend to recieve it (wait for the message "dropzone has finished working" in your python shell). Then you hit the training button.