Using vectors as batches

[RoshenS]

The following changes are to be implemented from #2 :

1. Obtain a sample from the function to model as a vector instead of a single number.

2. Use the vectors as batches for training.

Note that the MLP will still process each element of the vector as input data.

Using vectors as batches increased the time taken per epoch with the loss function seeing a greater decrease per epoch.

[fbraeu90]

1) You have to be careful how you use the single vector as in- and output. Each point should still be processed separately but the vector dimension should actually the batch size. Are you considering this? 2) Please be careful with the terminology. CNN --> Convolutional Neural Network is a network type in which you use convolution layers. This type is mainly used for structured inputs like images. Here, you are using fully connected layers (in pytorch often referred to as Linear) that are called MLPs --> Multilayer Perceptrons

[RoshenS]

i. I was not considering this. In that case the modifications to the code needed are minimal.

ii. Noted on that. I have edited the relevant issues and graphs.

Thanks!