ntopics	The number of events.	N_E
nsubtopics	The number of slots.	N_S
topic_stick	Stickiness parameter. (See Cheung et al., (2013) NAACL paper)
topic_smooth	Event smoothing.	\delta for P^E, P^E_{init}
subtopic_smooth	Slot smoothing.	\delta for P^S, P^S
head_smooth	Head smoothing.	\delta for P^H_{lemm}
arg_smooth	Argument smoothing.	\delta for P^A_{lemm}
covar_smooth	Covariance matrix smoothing.	\beta
ndims	Number of dimensions in DS vectors	k
niters	Maximum number of iterations trained.
method	Training algorithm.
rseed	Random seed used to initialize the model.
perturb	Amount of perturbation in the initialization.
ds_model	The distributional semantic model that is used.