What does the weight option do in glmboost in R




While learning about gradient boosting in R using the glmboost function, I came across the below syntax:

What does the option weights specify.The package says that it is an optional vector of weights used for fitting process,does this mean that we give more weights to some parameters?
If this is so,on what basis do we select these variables.Also,is there any specific weights that we apply or do we derive the weights somehow?