Regression results about the RTDL models.

[linkedlist771]

Hi, you did a great implementation of the tab-transformer. However, when I use your example notebook to do the simple regression for the Sin(x), neither the baseline model or the FTTransformer give the good results. I have no idea about this and want to know why.

Here is the link

[Yura52]

In the cell where the training happens, you have warnings that correctly point to the issue. To fix this, you have to make y_all a one-dimensional array. Namely, instead of this line:

y_all = np.sin(X_all).astype('float32').reshape(-1, 1)

you should do this:

y_all = np.sin(X_all).astype('float32').squeeze(1)