nicocheh comments

Results 9 comments of


                                            nicocheh

TemporalFusionTransformer predicting with mode "raw" results in RuntimeError: Sizes of tensors must match except in dimension 0

Any solution on this? I get the same thing...

Softplus transformation really slow in group normalizer

Actually any transformation in the target normalizer (using group or encoder normalizer) makes the process really slow... Any thoughts on this? Is it really necessary/important to use a target_normalizer?

Softplus transformation really slow in group normalizer

Is it normal that it slows down the process really much? I am forecasting positive integers, so i would really like to use softplus or something like softplus, relu or...

RAM consumption of TimeSeriesDataset

@jdb78 Using smaller datasets, i realized that the RAM consumption gets really high and after that it goes down. Do you have any clue of which operation of TimeSeriesDataset can...

Can i get nbinom parameters with DeepAR model?

3- How can i save and load the model to use it later?

Littérature on EncoderNormalizer

Interested on this as well

tf.math.betainc is missing gradients

Any news or solution to this? @csuter @jesnie

"RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn" when using my loss-function

I am having the same issue with Single target + multiple GPUs with TFT with QuantileLoss(). Any news on this? @jdb78 thank you for this awesome project!

"RuntimeError: element 0 of tensors does not require grad and does not have a grad_fn" when using my loss-function

> @jdb78, I saw several cases in issues but not try out a solution. I use 1GPU, 1 target and got the same error. sorry I cannot provide a colab,...