dragon_prompts: Prompts from a run's data files

In dragonfarm: Fine-Tune Small Language Models with LoRA from R

Prompts from a run's data files

Description

The user turns of a run's training or held-out rows, for feeding dragon_synthesize_pairs() or dragon_judge(). Works for prompt/response and preference-pair runs alike.

Usage

dragon_prompts(run, split = c("eval", "train"), n = Inf, seed = 42)

Arguments

run	A dragon_run or run directory.
split	"eval" for the held-out rows, "train" for the rest.
n	Maximum number of prompts. Inf for all.
seed	Seed used when sampling down to n.

Value

A character vector.

dragonfarm documentation built on Oct. 12, 2026, 5:07 p.m.