Regarding forecasting for many stations at a time

[EnugulaVishnucharan]

Hi authors
I am working on time series data with exogenous features and found your paper TimeXer quite interesting .

1. I have weather data of 500 stations (exogenous and endogenous features). With your TimeXer setting , I believe we can forecast for a single station at a time. I want to forecast at a time for all the stations . Your paper claim that forecasting is done for 3850 stations (large scale meteorology data) . Is the forecasting done for each station at a time or all the stations at once .I went through your git but could not find . Please let me know if I am missing any file which has this information.
   Is there a way possible to forecast all the stations at a time using exogenous features.
2. Please let me know if I can get pretrained model of TimeXer so that I can finetune the model on my data.
3. In the visualization used is it possible to indicate the forecasting data with timestamp on x axis.