Greg DeVosNouri

unable to replicate ETTM1 results

7

Hello, I find the results of this paper to be very impressive. I tried replicating the ETTM1 out 48 results running the command found in the readme. Do you have...

Query Selector Model

2

Fixes #1120. ### Summary This model is a direct adapted from the author's [repo](https://github.com/moraieu/query-selector ). There are also a few more minor documentation fixes mostly recommended by PyCharm. ### Other...

Three new alternatives to layer norm

10

Fixes #1113. minor typos. ### Summary ### Other Information

I've seen pretty good results using [RevIN](https://github.com/ts-kim/RevIN) with larger datasets. I'm also waiting on a license.

Add the [Query Selector](https://github.com/moraieu/query-selector) model. I have a working non-probabilistic version working. Working on making it probabilistic.

DLinear

1

Add the [Dlinear ](https://github.com/cure-lab/DLinear) model from cure-lab. I have a working non-probabilistic version working. Working on making it probabilistic.

SCINet

1

Add the [SCINet ](https://github.com/cure-lab/SCINet) model. I have a working non-probabilistic version working. Working on making it probabilistic.

**Is your feature request related to a current problem? Please describe.** - ScaleNorm from "Transformers without tears: Improving the normalization of self-attention." - RMSNorm from "Root mean square layer normalization."

PyData talk

1

**Is your feature request related to a current problem? Please describe.** This recent PyData talk goes though time series features engineering and shoes how to implement them in darts. Do...

beta-NLL

3

**Is your feature request related to a current problem? Please describe.** beta-NLL from the paper ["On the Pitfalls of Heteroscedastic Uncertainty Estimation with Probabilistic Neural Networks"](https://arxiv.org/abs/2203.09168) **Describe proposed solution** We...