Derive an integer hash from all args, for use as a random seed.

Return a pair of datasets corresponding to a random split of the given dataset, with n datapoints in the first dataset and the rest in the la

Perform one update step, given a list of (x, y) tuples for all environments. Admits an optional list of unlabeled minibatche

Terms being computed: * Li = Loss(xi, yi, params) * Gi = Grad(Li, params) * Lj = Loss(xj, yj, Optimizer(