I have a flask of silica gel connected to a flask that contains a bit of water, so when I heat the silica gel the water absorbed gets evaporated and I want it to move to the other flask, but it always condenses on the roof of the first flask, so is there any way to force the vapour to move flasks?
To prevent the condensation, you need to create the conditions under which the vapor does not become saturated. The way to do that is to heat up the whole transfer line (the roof of the first flask and all the tubes that follow) - the higher the temperature, the higher the saturated vapor pressure. Heated transfer lines are often employed in commercial equipment when there's need to get condensible gases from one setup (e.g. thermogravimetric) to the other (e.g. mass spectrometer).
In your flask, the water will always go from the drops of liquid water to the silica, and not in the opposite direction. If you want to remove the water from the silica, the only way to do it is to heat the hydrated silica in the air oder under a vacuum, and quickly close the hot silica when dry.