# /ml-finetune - ML Model Fine-Tuning & Dataset Pipeline

> JSONL dataset formatter, LoRA fine-tuning scripts, and model evaluation.

Canonical: https://www.agentickit.co/commands/ml-finetune
Last updated: 2026-11-18

Prepares training datasets, validates JSONL formats, configures fine-tuning jobs (LoRA / OpenAI), and benchmarks model output accuracy.

**Usage:** `Use when customizing open-source LLMs on proprietary domain data.`

## Pipeline
1. Clean, deduplicate, and format training conversation pairs into JSONL
2. Split dataset into training (80%) and validation (20%) sets
3. Create fine-tuning submission script with hyperparameter configuration
4. Generate evaluation benchmark suite to measure accuracy improvements

## Output files
- `scripts/format-dataset.py`
- `scripts/finetune.py`
- `data/train.jsonl`

**Claude Code frontmatter:** `argument-hint: [model, dataset, and eval target]` · `model: inherit`

## Related
- [ml-engineer](https://www.agentickit.co/agents/ml-engineer)
- [ai-engineer](https://www.agentickit.co/agents/ai-engineer)
- [ml-training](https://www.agentickit.co/skills/ml-training)
- [llm-integration](https://www.agentickit.co/skills/llm-integration)
