AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Engram Technology for Statistical Language Modelling
Over the years, statistical language modelling was minted by this approach. They are still the work horse to day of most speech processing systems. Although they're trained on many, more than 50 million words, it might be 50 billion words or even approaching trillions of words. Their simplicity means that they're fast use. In practice, when you're trying to look up ar probability, they're very fast. There are approaches to getting more accurate statistical models at predicting that next word that are more accurate, but typically they're very time consuming and so hard to use in practice.