The last stock predictor could only predict the following day based on the last n days.
Now it can predict any number of days ahead.
Since when you try to predict more days ahead some may overlap, this plot is made by taking the mean of all the overlapping days. Hence, it’s not very accurate, but gives an idea of how the model is able to generalise.
You can find the code on:
The ideal is not to make a generalization based on all the predictions, but to rely on only one prediction for the following chosen days.