AI use cases

Adapt an existing base model using a smaller trainable component instead of updating most of the model's weights.

From Is Fine-Tuning Still Needed? LLMs, RAG, & LoRA by IBM Technology