This repository will show how to use Bayesian Inference for language model-like tasks, like next word prediction. The theory behind Bayesian Inference is that, unlike frequentist methods, we can use knowledge of our existing dataset (a prior) to inform our predictions. We'll see in this section how this task can be applied to next word prediction.
jschulberg/Next-Word-Generation
Use Bayesian Inference for language model-like tasks, like next word prediction.
Jupyter Notebook