Fine-Tuning Dataset Size Estimator

Estimate how many training examples you need for your fine-tuning task.

Estimate a minimum and recommended number of fine-tuning examples based on task type, how strong the baseline model already is, and your target accuracy, with practical data-collection guidance for each scenario. It runs free in your browser on Gera Tools, with nothing uploaded.

Last updated Source: Gera Tools

How accurate is this estimate?

It is a planning heuristic, not a guarantee. Real requirements depend on data quality, task difficulty, and label noise. Use the number to size your effort, then validate with a small pilot run and a held-out eval set.

Fine-tuning dataset size estimator

“How many examples do I need?” is the first question of every fine-tuning project, and the honest answer is “it depends.” This tool turns that into a concrete starting number by weighing the three factors that matter most: what kind of task it is, how good the base model already is, and how high you need accuracy to go.

How it works

Each task type carries a base example count reflecting its difficulty. Teaching a fixed output format or adapting a writing style needs far fewer examples than teaching new classification boundaries, domain-specific reasoning, or novel factual knowledge. The estimator scales that base by two multipliers:

  1. Baseline capability — a strong base model that already handles the task reasonably well needs fewer examples to nudge toward your target. A weak baseline requires more examples to move the needle at all.
  2. Target accuracy — the last few percentage points of accuracy are disproportionately expensive in data terms. Moving from 70% to 80% costs far less data than moving from 92% to 97%.

The output is a minimum (safe starting point for a pilot run) and a recommended ceiling (what to plan toward if the minimum falls short).

Task-type guidance

Different task families have genuinely different data requirements:

Task typeWhy data needs differ
Format / style adaptationThe model already understands the content; you are just changing the surface pattern
Binary or small-set classificationClear boundaries; consistent labelling is the main challenge
Multi-class classificationMore classes multiply the examples needed per boundary
Structured extraction (JSON, tables)Needs many input formats represented to generalise
Domain knowledge injectionHardest — the base model lacks the underlying facts
Instruction followingModerate; usually a few hundred diverse examples work

Worked example

Suppose you want a model to extract JSON-structured entities from legal contracts, and the base model is decent at general extraction but has never seen legal language. A good target accuracy is 90%. The estimator would return something like: minimum 400 examples, recommended 800–1,200. You would collect 400 varied contracts, label them carefully, fine-tune, evaluate on a 100-example held-out set, and add more data only if accuracy lags.

Practical guidance

  • Start at the minimum, not the maximum. Collect the smaller number, run a pilot fine-tune, and measure on a held-out set before labelling more. Over-collecting wastes labelling budget.
  • Quality beats quantity. A few hundred clean, diverse, correctly labelled examples usually outperform thousands of noisy ones. Deduplicate inputs and check that labels are consistent.
  • Match the eval to the goal. Your held-out evaluation set should look exactly like real production inputs, not the training distribution. A misleading eval produces a misleading accuracy number.
  • Reserve a test split before you start labelling. Setting aside test data after seeing what the model struggles with can inflate reported accuracy.
  • Revisit the baseline slider as the model improves. After a successful pilot run, the base model is now your new baseline. Update the estimate before planning the next data collection wave.