A dataset class for generating non-overlapping sliding windows from a time series dataset. This is useful for training models that require f

Use a mlp layer to embedding the time-series.

Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`TimeMoeDecoderLayer`] Args: config: TimeMoeCo

Multi-headed attention from 'Attention Is All You Need' paper. Modified to use sliding window attention: Longformer and "Generating Long Sequ

A dataset that generates windows of time series data with pack technique.