In the last few days? No, that would be impossible; no one has the resources to train a base model that quickly. But there are definitely a lot of people working on it.

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