The amount of computing resources it's going to take to retrain the model is enormous. So the most of us will have to wait for a big company to publish or leak their weights until we get to use anything written in the paper.