Evidence accumulation (decision-making) models, pseudo-likelihoods, and pymc3

I just came across this discussion (GSoC 2018 Proposal: HDDM & ABC) and wonder if HDDM (and other evidence accumulation models such as LBA) have been implemented with pymc3.
I also wonder if a pseudo-likelihood for a simulated model (obtained with probability density approximation; e.g., Holmes, 2015 J Math Psych) could be passed to pymc3 samplers easily (for example Differential Evolution MCMC), and if anybody has ever written some example codes.
Thanks for the help,