Kashif Rasul

thanks will review soon!

@ShaneOss thanks mate! I was thinking of upgrading to lighting instead of pytorch-lightning... what do you think?

no worries, this week is choka for me but will have a look soon

@ShaneOss so we realized that if we require `gluonts[torch]` as the top level requirement, it pulls in the working versions of lighting and pytorch-lightning so closing this PR for now

@hanlaoshi dlinear is an inherently univariate model so there is no need to group the time series etc. just pass the time series individually and predict individually like other univariate...

oh i see now you are not using those imports... so the issue seems to be in the dataset preparation helpers you have rather than dlinear... can you confirm that...

yes they are... the issue is in the `split` function i see so even before the model is called... i will try to debug

it doesnt make sense for dlinear to be multivariate, you would need to rehape the context_length, multivariate_dim tensor to context_length*multivariate_dim vector and then the nn.linear would be potentially huge: ```...

so the evaluator returns all these metrics so just use those I would say instead of re-writing these functions

ok have a look at the implementation here: https://github.com/huggingface/evaluate/pull/509