Accurate Compression of Text-to-Image Diffusion Models via Vector Quantization

• URL: http://arxiv.org/abs/2409.00492v1
• Date: Sat, 31 Aug 2024 16:09:20 GMT
• Title: Accurate Compression of Text-to-Image Diffusion Models via Vector Quantization
• Authors: Vage Egiazarian, Denis Kuznedelev, Anton Voronov, Ruslan Svirschevski, Michael Goin, Daniil Pavlov, Dan Alistarh, Dmitry Baranchuk,
• Abstract summary: State-of-the-art text-to-image models are becoming less accessible in practice. Post-training quantization (PTQ) tackles this issue by compressing the pretrained model weights into lower-bit representations. This work demonstrates that more versatile vector quantization (VQ) may achieve higher compression rates for large-scale text-to-image diffusion models.
• Score: 33.20136645196318
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Text-to-image diffusion models have emerged as a powerful framework for high-quality image generation given textual prompts. Their success has driven the rapid development of production-grade diffusion models that consistently increase in size and already contain billions of parameters. As a result, state-of-the-art text-to-image models are becoming less accessible in practice, especially in resource-limited environments. Post-training quantization (PTQ) tackles this issue by compressing the pretrained model weights into lower-bit representations. Recent diffusion quantization techniques primarily rely on uniform scalar quantization, providing decent performance for the models compressed to 4 bits. This work demonstrates that more versatile vector quantization (VQ) may achieve higher compression rates for large-scale text-to-image diffusion models. Specifically, we tailor vector-based PTQ methods to recent billion-scale text-to-image models (SDXL and SDXL-Turbo), and show that the diffusion models of 2B+ parameters compressed to around 3 bits using VQ exhibit the similar image quality and textual alignment as previous 4-bit compression techniques.

Related papers

• Fast constrained sampling in pre-trained diffusion models [77.21486516041391]
  Diffusion models have dominated the field of large, generative image models. We propose an algorithm for fast-constrained sampling in large pre-trained diffusion models.
  arXiv  Detail & Related papers  (2024-10-24T14:52:38Z)
• Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis [62.06970466554273]
  We present Meissonic, which non-autoregressive masked image modeling (MIM) text-to-image elevates to a level comparable with state-of-the-art diffusion models like SDXL. We leverage high-quality training data, integrate micro-conditions informed by human preference scores, and employ feature compression layers to further enhance image fidelity and resolution. Our model not only matches but often exceeds the performance of existing models like SDXL in generating high-quality, high-resolution images.
  arXiv  Detail & Related papers  (2024-10-10T17:59:17Z)
• Zero-Shot Image Compression with Diffusion-Based Posterior Sampling [34.50287066865267]
  This work addresses the gap by harnessing the image prior learned by existing pre-trained diffusion models for solving the task of lossy image compression. Our method, PSC (Posterior Sampling-based Compression), utilizes zero-shot diffusion-based posterior samplers. PSC achieves competitive results compared to established methods, paving the way for further exploration of pre-trained diffusion models and posterior samplers for image compression.
  arXiv  Detail & Related papers  (2024-07-13T14:24:22Z)
• 2DQuant: Low-bit Post-Training Quantization for Image Super-Resolution [83.09117439860607]
  Low-bit quantization has become widespread for compressing image super-resolution (SR) models for edge deployment. It is notorious that low-bit quantization degrades the accuracy of SR models compared to their full-precision (FP) counterparts. We present a dual-stage low-bit post-training quantization (PTQ) method for image super-resolution, namely 2DQuant, which achieves efficient and accurate SR under low-bit quantization.
  arXiv  Detail & Related papers  (2024-06-10T06:06:11Z)
• Post-training Quantization for Text-to-Image Diffusion Models with Progressive Calibration and Activation Relaxing [49.800746112114375]
  We propose a novel post-training quantization method (Progressive and Relaxing) for text-to-image diffusion models. We are the first to achieve quantization for Stable Diffusion XL while maintaining the performance.
  arXiv  Detail & Related papers  (2023-11-10T09:10:09Z)
• Effective Quantization for Diffusion Models on CPUs [5.419875683826296]
  Quantization, a technique employed to compress deep learning models for enhanced efficiency, presents challenges when applied to diffusion models. We introduce a novel approach to quantize the diffusion models by leveraging both quantization-aware training and distillation. Our results show the quantized models can maintain the high image quality while demonstrating the inference efficiency on CPUs.
  arXiv  Detail & Related papers  (2023-11-02T13:14:01Z)
• Q-Diffusion: Quantizing Diffusion Models [52.978047249670276]
  Post-training quantization (PTQ) is considered a go-to compression method for other tasks. We propose a novel PTQ method specifically tailored towards the unique multi-timestep pipeline and model architecture. We show that our proposed method is able to quantize full-precision unconditional diffusion models into 4-bit while maintaining comparable performance.
  arXiv  Detail & Related papers  (2023-02-08T19:38:59Z)
• Image Compression with Product Quantized Masked Image Modeling [44.15706119017024]
  Recent neural compression methods have been based on the popular hyperprior framework. It relies on Scalar Quantization and offers a very strong compression performance. This contrasts from recent advances in image generation and representation learning, where Vector Quantization is more commonly employed.
  arXiv  Detail & Related papers  (2022-12-14T17:50:39Z)
• Extreme Generative Image Compression by Learning Text Embedding from Diffusion Models [13.894251782142584]
  We propose a generative image compression method that demonstrates the potential of saving an image as a short text embedding. Our method outperforms other state-of-the-art deep learning methods in terms of both perceptual quality and diversity.
  arXiv  Detail & Related papers  (2022-11-14T22:54:19Z)
• Quantization Guided JPEG Artifact Correction [69.04777875711646]
  We develop a novel architecture for artifact correction using the JPEG files quantization matrix. This allows our single model to achieve state-of-the-art performance over models trained for specific quality settings.
  arXiv  Detail & Related papers  (2020-04-17T00:10:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.