Diverse Sample Generation: Pushing the Limit of Data-free Quantization

by   Haotong Qin, et al.

Recently, generative data-free quantization emerges as a practical approach that compresses the neural network to low bit-width without access to real data. It generates data to quantize the network by utilizing the batch normalization (BN) statistics of its full-precision counterpart. However, our study shows that in practice, the synthetic data completely constrained by BN statistics suffers severe homogenization at distribution and sample level, which causes serious accuracy degradation of the quantized network. This paper presents a generic Diverse Sample Generation (DSG) scheme for the generative data-free post-training quantization and quantization-aware training, to mitigate the detrimental homogenization. In our DSG, we first slack the statistics alignment for features in the BN layer to relax the distribution constraint. Then we strengthen the loss impact of the specific BN layer for different samples and inhibit the correlation among samples in the generation process, to diversify samples from the statistical and spatial perspective, respectively. Extensive experiments show that for large-scale image classification tasks, our DSG can consistently outperform existing data-free quantization methods on various neural architectures, especially under ultra-low bit-width (e.g., 22 diversifying caused by our DSG brings a general gain in various quantization methods, demonstrating diversity is an important property of high-quality synthetic data for data-free quantization.



There are no comments yet.


page 1

page 12


Diversifying Sample Generation for Accurate Data-Free Quantization

Quantization has emerged as one of the most prevalent approaches to comp...

ClusterQ: Semantic Feature Distribution Alignment for Data-Free Quantization

Network quantization has emerged as a promising method for model compres...

Data-Free Quantization with Accurate Activation Clipping and Adaptive Batch Normalization

Data-free quantization is a task that compresses the neural network to l...

Qimera: Data-free Quantization with Synthetic Boundary Supporting Samples

Model quantization is known as a promising method to compress deep neura...

Arbitrary Bit-width Network: A Joint Layer-Wise Quantization and Adaptive Inference Approach

Conventional model quantization methods use a fixed quantization scheme ...

Learning in School: Multi-teacher Knowledge Inversion for Data-Free Quantization

User data confidentiality protection is becoming a rising challenge in t...

Fine-grained Data Distribution Alignment for Post-Training Quantization

While post-training quantization receives popularity mostly due to its e...

Code Repositories


This project is the official implementation of our paper Diverse Sample Generation: Pushing the Limit of Data-free Quantization.

view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.