Training
Fine-tuning
Further training a pre-trained AI model on your own data to specialise it for a specific task or style.
01 ——
In plain English
Fine-tuning takes a general-purpose model (like GPT or Llama) and continues training it on a smaller, task-specific dataset. The result is a model that behaves differently — it may reply in your brand's tone, follow a specific format, or be better at a narrow domain like legal or medical text.
When to fine-tune:
- You need consistent output style or format
- You want the model to "know" proprietary knowledge deeply, not just be told it
- Prompt engineering alone isn't reliable enough
When NOT to fine-tune:
- You just need the model to know a few facts — use RAG instead
- You lack enough labelled examples (typically need hundreds to thousands)
- Cost is a concern — fine-tuned models are more expensive to host
02 ——
Related terms
LLM
Large Language Model — the type of AI behind tools like ChatGPT and Claude, trained to understand and generate text.
RLHF
Reinforcement Learning from Human Feedback — the training technique that teaches AI models to be helpful, harmless, and honest.
LoRA
Low-Rank Adaptation — a cheap way to fine-tune large AI models by training a small set of extra weights instead of the whole model.
Distillation
Training a smaller, cheaper AI model to mimic the outputs of a larger, more capable one — preserving most of the quality at a fraction of the cost.
Training Data
The dataset an AI model learns from — its quality, diversity, and biases directly shape what the model can do and how well it does it.
Back to glossaryLast reviewed May 2026