https://github.com/nixtla logo
a

Akmal Soliev

07/13/2023, 3:23 PM
Hey @Kin Gtz. Olivares any updates on this issue? https://github.com/Nixtla/neuralforecast/issues/566
a

Akmal Soliev

07/13/2023, 3:29 PM
I mean this one:
Copy code
Regarding the RuntimeError: maximum size for tensor at dimension 2 is 280, but the size is 336

For the moment we are yet to include the exogenous partial predictions, as the current version of NBEATSx is not performing the projections on the exogenous basis, rather its taking the exogenous data as inputs of the MLPs.
We will get to work on this feature soon.
looking at the example it doesn't have exo?
k

Kin Gtz. Olivares

07/13/2023, 3:37 PM
No for the moment, we have not implemented the decomposition against exogenous variables in NeuralForecast yet
a

Akmal Soliev

07/13/2023, 3:38 PM
Thanks!
k

Kin Gtz. Olivares

07/13/2023, 3:38 PM
Here is the original paper code: https://github.com/cchallu/nbeatsx
👍 1
a

Akmal Soliev

07/13/2023, 3:40 PM
Any idea on what modifications can be done to add support for exo?
k

Kin Gtz. Olivares

07/13/2023, 3:45 PM
a

Akmal Soliev

07/13/2023, 3:46 PM
Wanna collaborate on this feature?
k

Kin Gtz. Olivares

07/13/2023, 3:47 PM
If want it in NF, yes. I think it can be useful for many people. Do you want to try?
a

Akmal Soliev

07/13/2023, 3:48 PM
Sure, I'll start by reading your paper, maybe we can setup a chat next week or so?
k

Kin Gtz. Olivares

07/13/2023, 3:49 PM
Sounds good. Here is the NBEATSx paper: https://arxiv.org/abs/2104.05522
a

Akmal Soliev

07/13/2023, 3:49 PM
Too slow, I'm already reading it 😄
k

Kin Gtz. Olivares

07/13/2023, 3:50 PM
🙂
a

Akmal Soliev

07/13/2023, 3:50 PM
Okay talk soon
m

Max (Nixtla)

07/13/2023, 6:31 PM
Nice! Thanks @Akmal Soliev.
❤️ 1
4 Views