Hi all, I am getting the following error when passing keyword args for the

crossvalidation

method of the

NeuralForecast()

object: How can I specify refit=False?

I also had this problem..then I found that in the newer version of neuralforecast the refit argument is there..so I uninstalled and installed the package neuralforecast and the issue got resolved

I just updated it but now when I try to train the models it complains of missing values when calling the

fit

function. There were missing values before already but it was working. The update was 1.6.4-pyhd8ed1ab_0 --> 1.7.1-pyhd8ed1ab_0

Missing values can crash the training or produce NaNs in the predictions. If you want to keep them you can add a column called available_mask with zeros in the places where you have NaNs

Ok then, I am backfilling them in the new version