A model must adopt a strict house style. What usually fits fine-tuning?
Why A is the answer
Consistent style and format
Fine-tuning suits stable behavior, and live data suits tools or retrieval.
In detail
- Stable style and format suit fine-tuning.
- Live data suits tools or retrieval.
Use case
An illustrative scenario, not a real incident.
A publisher needs a strict style across all drafts, which suits fine-tuning.
See it in different ways
flowchart TD
A{"Is the need new knowledge?"} -->|Yes| B["RAG"]
A -->|"No, it is behavior or style"| C{"Is the budget limited?"}
C -->|Yes| D["LoRA or other PEFT"]
C -->|No| E["Full fine-tuning"]stateDiagram-v2 [*] --> Need Need --> Retrieval: facts that change Need --> FineTuning: stable style or format FineTuning --> Lora: limited GPU budget FineTuning --> FullTuning: budget allows it Lora --> EarlyStopping: small dataset
mindmap
root(("Adapting a model"))
n1["Fine-tuning"]
n2["Style"]
n3["Format"]
n4["Retrieval"]
n5["Current facts"]Key point: Fine-tuning suits stable behavior.
Why the other options are wrong
- B. Facts that are updated every hour
- Hourly facts suit retrieval.
- C. The customer's current account balance
- Live balances suit tools.
- D. Today's stock of each warehouse
- Live stock suits tools.