---
sdk: gradio
title: PEFT-Bench Leaderboard
emoji: 📊
colorFrom: indigo
colorTo: purple
---

# PEFT-Bench Leaderboard

**PEFT-Bench** tracks reproducible benchmark scores for parameter-efficient
fine-tuning methods (LoRA, QLoRA, DoRA, VeRA, …) across a suite of NLP tasks.
Scores populate automatically from the template repo → `peft-bench-eval`.

* **Metrics store:** [`Mdrnfox/peft-bench-metrics`](https://huggingface.co/datasets/Mdrnfox/peft-bench-metrics)
* **How to submit a run:** coming soon in **`peft-bench-eval`**

_This board is brand new—rows will appear as soon as the first evaluation
completes._