eschibli

`output_chunk_shift` is supported for all global forecasting models as of [0.28](https://github.com/unit8co/darts/pull/2176). I believe this includes all of the neural network-based models except possibly the RNN cell (for which it doesn't...

I'm not intimately familiar with every model implemented in Darts, but for the case of D-Linear, as well as most models I can think of, only the dataset should depend...

Hello Giacomo, sorry I've been (and still am) very busy this week and don't have much time. What exactly are you overwriting? `TrainingDataset`?

RINorm is only scale invariant the variance of the sample is much greater than `rin.eps`. `eps` is added to the sample variance for numerical stability before scaling to avoid numerical...

The norm is exactly invertible at all scales, but doesn't produce unit variance for inputs with very low scales, because it doesn't scale by the sample std, it scales by...

@andrewwarrington re: your first point, I think you are right. The weights are initialized at 1, and under normal circumstances should probably not be small or negative. As currently implemented...

Sure I'll see what I can do next week. Thanks Dennis.

What would be required to get this merged? I find it very useful and am working to extend it to time series, which I think would be beneficial to the...