AI engineering glossary · LLM training alignment
What is adapter?
Adapter (also called Adapter layer): A small trainable module inserted into a frozen pretrained network, typically a down-projection, non-linearity and up-projection with a residual connection. Only adapters are trained and stored per task, and they are swapped in as needed.
Adapter: an example
A single base model keeps one adapter file each for Hindi support and Tamil support, each only a few megabytes.
Related terms
More LLM training alignment terms
- Advantage function
- Alignment
- Alignment tax
- Alpaca
- Base model
- Bradley-Terry model
- Causal language modelling
- Chat template
Where you learn it
Adapter is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.