AI Terminology Course
AI Terminology
/
Intermediate

Fine-Tuning

Definition

The process of taking a pre-trained foundational AI model and continuing its training on a smaller, highly specific dataset to adapt it for a specialized task or tone.

Explain Like I'm New

The AI already went to college and got a general degree (Pre-training). Fine-tuning is sending the AI to a 2-week bootcamp to learn how to be a specialized lawyer or a specific company's brand ambassador.

Real World Example

You take the open-source LLaMA model and Fine-Tune it on 10,000 examples of your company's support emails. The model stops talking like a generic AI and starts using your exact corporate catchphrases and tone.

Common Use Cases

  • •Tone matching
  • •Specialized coding tasks
  • •Medical/Legal AI

Interview Questions

basic

  • Is Fine-Tuning a good way to teach an AI new facts (like your company's updated vacation policy)?

intermediate

  • What is 'Pre-Training' compared to 'Fine-Tuning'?

Flash Cards

Question

Good for teaching facts?

Click to reveal answer
Answer

NO. Fine-tuning is for teaching an AI *how* to behave (tone, format, style). It is terrible at memorizing specific facts. Use RAG for teaching facts.

Question

Pre-Training vs Fine-Tuning?

Click to reveal answer
Answer

Pre-training is the initial phase where the model reads the entire internet for 6 months on supercomputers to learn basic English. Fine-tuning happens afterward, takes a few hours, and adjusts the model for a specific job.