not the whole model obviously since it just came out. but people have been successful in replicating the core RL principle behind it.