loading
loading
Additional training on specific data to adapt a pre-trained model to your use case.
Fine-tuning adjusts the weights of an existing model on a curated dataset of examples, teaching it to follow your preferred format, tone, or domain vocabulary. It's more expensive than prompt engineering but can produce more consistent behavior for narrow, repetitive tasks. Hugging Face provides tooling and hosting for open-source fine-tuned models.
Plainly
Think of Fine-tuning as the brain part that guesses or decides. Additional training on specific data to adapt a pre-trained model to your use case.
In practice
Use it when model choice, prompts, latency, cost, or quality affect the product result. In practice, define the owner, input, output, and failure mode before you rely on it.