Generative AI

BERT

An encoder trained by masking words — still the workhorse for classification and search.

BERT hides a fraction of the tokens and trains the model to fill them in, using context from both directions. That produces representations well suited to understanding rather than generation.

For classification, entity extraction and search relevance, a fine-tuned encoder of this kind is usually cheaper, faster and more accurate than prompting a large generative model.

Also in Generative AI

LLM GPT GANs RLHF QLoRA Agents

Apply

Begin the first module

Become AI native, it is the real deal today, and if it is not for you, you have lost nothing but learnt a new skill.

Apply