Mixture model using pm.Mixture or writing in parts

Questions version agnostic
6

Apologies, another quick question. Is there some nuance if using the tensor object when assigning weights using a distribution?
I.e., something like,

    w = [pm.Dirichlet(f'w{i}', a=np.array([2, 2])) for i in range(2)]
    weights = pt.empty((len(y), 2))
    weights = pt.set_subtensor(weights[:t], w[0])
    weights = pt.set_subtensor(weights[t:], w[1])

leads to a weights w0=w1 of zero or one. Or should the tensor not be used here?