2026/Free online book · Technical Deep Dives
Fine-Tune, or Don't
A Practical Decision Process for Customizing AI Models
Access
Free
Chapters
17
Read time
176 min
Alpesh Nakrani
VP of Growth at ViitorCloud · former CTO & COO
Fine-tuning is reached for too early and dismissed too fast. This deep dive is a decision procedure: when prompting, retrieval, or routing solves it cheaper, and the narrow cases where changing the weights is genuinely the right tool.
Fine-tuning is the first instinct and usually the wrong one. A sober procedure for deciding when weights should change.
This edition is free to read onsite. Each chapter has its own URL, so readers can bookmark, share, and return to the exact section they need.
Table of contents
FMFront Matter: Fine-Tune, or Don'tA Practical Decision Process for Customizing AI Models5 minINTIntroduction: The Wrong Operation, Performed WellA team I will call the support team, the details are composited from several real projects, but the shape is exact, had a problem that sounded like a fine-tuning problem. They ran tier-one customer support for a fast-moving SaaS product.9 min01The Support Bot That Knew the Old ProductWorking claim: Most "we should fine-tune" decisions are made before anyone has named what they are trying to change.9 min02What Fine-Tuning Actually ChangesWorking claim: A fine-tune does not "teach the model your data" in the way the phrase implies. It nudges a probability distribution toward the patterns your examples demonstrate.9 min03Five False DiagnosesThis chapter turns five false diagnoses into a concrete operating problem for the fine tuning or not book.11 min04The Customization Menu and a Decision TreeWorking claim: Fine-tuning is one item on a menu of nine customization techniques, most of which are cheaper, faster, and more reversible.8 min05Format, Behavior, and the Shape of a Repeated TaskWorking claim: This is where fine-tuning shines.9 min06House Style, Domain Phrasing, and Tool DisciplineWorking claim: Three behaviors are worth training that teams often try to prompt forever: a genuine house style, fluent domain phrasing, and disciplined tool use.8 min07Specializing Small Models and Distilling DownWorking claim: The most economically compelling fine-tune is not making a big model smarter, it is making a small model good enough. A large frontier model used as a generalist on a narrow, high-volume task is paying for capability it does not use.9 min08Demonstrations, Corrections, and PreferencesWorking claim: A fine-tuning dataset is the model update, compiled.9 min09Labels, Disagreement, and CoverageThis chapter turns labels, disagreement, and coverage into a concrete operating problem for the fine tuning or not book.9 min10Contamination, Leakage, and the Splits That Save YouWorking claim: The most common way a fine-tune lies to you is through the splits. If training and test data overlap, your eval reports a number that production will not honor, and you will ship a model you believe is good.8 min11Synthetic Data: When It Helps, When It PoisonsWorking claim: Synthetic training data, examples generated by a model rather than collected from reality, is the most powerful and most dangerous tool in the data toolkit.9 min12SFT, LoRA, and QLoRA in Practical TermsWorking claim: You do not need the math to make the decision.8 min13Preference Tuning, DPO, and DistillationWorking claim: When the target has no single right answer but a clear better-and-worse, demonstrations are the wrong tool and preference tuning is the right one.9 min14What to Fine-Tune: Generators, Retrievers, and RoutersWorking claim: "Fine-tune the model" almost always means "fine-tune the generator, " and that is frequently the lowest-impact place to spend a training run.8 min15Baselines, Regression Walls, and the Release GateWorking claim: A fine-tune you cannot evaluate against its alternatives is a fine-tune you cannot justify.9 min16Versioning, Lineage, Drift, and RetirementWorking claim: Training a model is the beginning of the operational burden, not the end. A fine-tuned model is a thing you now version, reproduce, monitor, retrain, retire, and answer for, to auditors, to regulators, to the on-call engineer at 3 a.9 min17Ten Playbooks for the Decision MeetingWorking claim: Everything in this book reduces to a decision a team makes in a room: should we fine-tune this, and if so, how? This chapter is ten such decisions, worked.12 minAAppendix A: Back MatterGlossary, implementation checklist, and source register for the book.9 min