workflow-fine-tuning-gemini

What problem does it solve?

This Skill automates the creation of supervised fine-tuning datasets by orchestrating question-answer generation, iterative review, and dataset synthesis so teams do not have to manually assemble and validate large QA-driven training sets.

Core Features & Use Cases

• Orchestrates answers-questions and answers-questions-revisor to produce and refine an initial answer base with up to five correction loops.
• Synthesizes final training examples with dataset-synthesizer and dataset-synthesizer-revisor, enforcing review loops and producing a validated dataset file path.
• Safety and governance: requires interaction logs for risk assessment and seeks explicit user authorization before proceeding past correction limits.
• Use case: produce a high-quality SFT dataset from a questions file, project docs, and interaction logs for fine-tuning Gemini-family models.

Quick Start

Run the workflow-fine-tuning-gemini skill to generate a validated fine-tuning dataset from the project title, questions file, logs directory, documentation directory, and desired number of output lines.