Léxico / TÉCNICA
Fine-Tuning
The process of further training a pre-trained AI model on a specific dataset to improve its performance for a particular task or domain.
Fine-tuning takes a general-purpose model and specializes it using your own data. Instead of training from scratch, you build on top of what the model already knows.
When to Fine-Tune
- You need consistent output in a specific style or format
- The model struggles with domain-specific terminology
- Prompt engineering alone isn’t getting the results you need
Trade-Offs
- Pro: Better accuracy for your use case with shorter prompts
- Pro: Can reduce token usage by removing lengthy instructions
- Con: Requires curated training data and compute resources
- Con: The fine-tuned model can drift from the base model’s general abilities