augmenting a simple model with new training data

[rumi1359]

Lets say I've trained a ExactGPModel called model with some train_x and train_y tensors. Say I want to augment train_x with a new value: train_x = torch.cat((train_x, torch.tensor([x]))).
Question: I could reinitialize the model and train but is there a more efficient way?

[wjmaddox]

This can be accomplished with new_model = model.get_fantasy_model(new_x, new_y).