# Quantization with Unified Adaptive Distillation to enable multi-LoRA based one-for-all Generative Vision Models on edge

arXiv: https://arxiv.org/abs/2603.29535 · HTML: https://mlsanity.com/paper/2603.29535/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 21 references that are in the map.
- First posted on arXiv 31 Mar 2026.

## Builds on (most cited first, 19 of 19)

- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [Palette: Image-to-Image Diffusion Models](https://mlsanity.com/paper/2111.05826/) (2021)
- [AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition](https://mlsanity.com/paper/2205.13535/) (2022)
- [QA-LoRA: Quantization-Aware Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2309.14717/) (2023)
- [Dual Diffusion for Unified Image Generation and Understanding](https://mlsanity.com/paper/2501.00289/) (2024)
- [A Survey on LoRA of Large Language Models](https://mlsanity.com/paper/2407.11046/) (2024)
- [Conv-Adapter: Exploring Parameter Efficient Transfer Learning for ConvNets](https://mlsanity.com/paper/2208.07463/) (2022)
- [Deep Learning-based Image and Video Inpainting: A Survey](https://mlsanity.com/paper/2401.03395/) (2024)
- [MobileDiffusion: Instant Text-to-Image Generation on Mobile Devices](https://mlsanity.com/paper/2311.16567/) (2023)
- [One Diffusion to Generate Them All](https://mlsanity.com/paper/2411.16318/) (2024)
- [DiffI2I: Efficient Diffusion Model for Image-to-Image Translation](https://mlsanity.com/paper/2308.13767/) (2023)
- [Parameter-Efficient Fine-Tuning for Pre-Trained Vision Models: A Survey and Benchmark](https://mlsanity.com/paper/2402.02242/) (2024)
- [UniVG: A Generalist Diffusion Model for Unified Image Generation and Editing](https://mlsanity.com/paper/2503.12652/) (2025)
- [Portrait Diffusion: Training-free Face Stylization with Chain-of-Painting](https://mlsanity.com/paper/2312.02212/) (2023)
- [Diffusion Models on the Edge: Challenges, Optimizations, and Applications](https://mlsanity.com/paper/2504.15298/) (2025)
- [Lifelong Learning with Task-Specific Adaptation: Addressing the Stability-Plasticity Dilemma](https://mlsanity.com/paper/2503.06213/) (2025)
- [Multi LoRA Meets Vision: Merging multiple adapters to create a multi task model](https://mlsanity.com/paper/2411.14064/) (2024)

## Abstract (arXiv)

Generative Artificial Intelligence (GenAI) features such as image editing, object removal, and prompt-guided image transformation are increasingly integrated into mobile applications. However, deploying Large Vision Models (LVMs) for such tasks on resource-constrained devices remains challenging due to their high memory and compute requirements. While Low-Rank Adapters (LoRAs) enable parameter-efficient task adaptation, existing Mobile deployment pipelines typically compile separate model binaries for each LoRA + a copy of the foundation model, resulting in redundant storage and increased runtime overhead. In this work, we present a unified framework for enabling multi-task GenAI inference on edge devices using a single shared model. Our key idea is to treat LoRA weights as runtime inputs rather than embedding them into the compiled model graph, allowing dynamic task switching at runtime without recompilation. Then, to support efficient on-device execution, we introduce QUAD (Quantization with Unified Adaptive Distillation), a quantizationaware training strategy that aligns multiple LoRA adapters under a shared quantization profile. We implement the proposed system with a lightweight runtime stack compatible with mobile NPUs and evaluate it across multiple chipsets. Experimental results demonstrate up to 6x and 4x reduction in memory footprint and latency improvements, respectively, while maintaining high visual quality across multiple GenAI tasks.