Training & post-training
Model training
The process of adjusting model parameters using examples and an objective.
Definition
Training repeatedly measures model outputs against an objective and updates parameters to reduce error or increase a desired reward.
Why it matters
Training choices determine capability, inherited bias, data provenance and the controls required before a model enters a bank.
Related concepts
- Model parameters
The learned numerical values that determine a model's behaviour.
- Pretraining
Initial large-scale training that gives a foundation model broad capability.
- Fine-tuning
Additional training that adapts a model to a task, domain or behaviour.