Pruning

hi all,
is there any tool that can do pruning on a given network,
in order to make the network “smaller” so that the inference process using the final engine (that is created using TensorRT) will run faster?
thanks,
yair