A walk with fastai2 - Tabular - Lesson 4, TabNet and Time Series

sound is very low

Can you explain how metric accuracy worked in case of time series?

I don't understand what do you mean by splits here...i mean let's assume i have a data like :
date;target;temp;humidity;pressure;
day1;15.2;28.0; 80.0;34.0
day2;17.0;29.0;85.0;34.0

the target is my y....are the splits here 2(the days) or 3 (the predictors)? and how do i specify that assuming i am starting with a pandas dataframe?

thanks