Additional training of existing models possible?

[shpshp]

Is there a way to take existing model and make some additional fine training (on a much smaller input dataset compared to the initial volume) to make it more accurate on the specific material? Thanks in advance!