提出QArtSR方法，实现扩散模型超分的低比特量化，2比特仍保持高精度。
QArtSR: Quantization via Reverse-Module and Timestep-Retraining in One-Step Diffusion based Image Super-Resolution
- 通过反向模块与时间步重训练，优化量化过程中的信息损失。
- 4比特量化性能接近全精度模型，2比特仍显著优于现有方法。
- 适合追求模型轻量化的图像超分辨率研究者与应用开发者。
一阶段扩散图像超分辨率（OSDSR）模型表现日益优越。尽管其去噪步骤缩减至一步且可量化至8位以降低开销，但仍有潜力进一步压缩至更低比特。为探索更高效的量化方案，本文提出一种名为QArtSR的新方法，结合反向模块与时间步重训练策略。首先分析时间步值对量化模型性能的影响；随后引入时间步重训练量化（TRQ）和逐模块反向量化（RPQ）机制，并采用模块损失与图像损失联合更新所有量化模块。仅微调量化适配组件参数，不修改原始权重。为确保模块充分优化，增加扩展端到端训练阶段。实验表明，4比特与2比特量化下QArtSR均显著优于当前主流方法；4比特性能接近全精度模型。代码将开源于https://github.com/libozhu03/QArtSR。
原文摘要 · Abstract (English)
One-step diffusion-based image super-resolution (OSDSR) models are showing increasingly superior performance nowadays. However, although their denoising steps are reduced to one and they can be quantized to 8-bit to reduce the costs further, there is still significant potential for OSDSR to quantize to lower bits. To explore more possibilities of quantized OSDSR, we propose an efficient method, Quantization via reverse-module and timestep-retraining for OSDSR, named QArtSR. Firstly, we investigate the influence of timestep value on the performance of quantized models. Then, we propose Timestep Retraining Quantization (TRQ) and Reversed Per-module Quantization (RPQ) strategies to calibrate the quantized model. Meanwhile, we adopt the module and image losses to update all quantized modules. We only update the parameters in quantization finetuning components, excluding the original weights. To ensure that all modules are fully finetuned, we add extended end-to-end training after per-module stage. Our 4-bit and 2-bit quantization experimental results indicate that QArtSR obtains superior effects against the recent leading comparison methods. The performance of 4-bit QArtSR is close to the full-precision one. Our code will be released at https://github.com/libozhu03/QArtSR.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。