The explosion of data available in life sciences is fueling an increasing demand for expressive models and computational methods. Graph transformation is a model for dynamic systems with a large variety of applications. We introduce a novel method of the graph transformation model construction, combining generative and dynamical viewpoints to give a fully automated data-driven model inference method. The method takes the input dynamical properties, given as a "snapshot" of the dynamics encoded by explicit transitions, and constructs a compatible model. The obtained model is guaranteed to be minimal, thus framing the approach as model compression (from a set of transitions into a set of rules). The compression is permissive to a lossy case, where the constructed model is allowed to exhibit behavior outside of the input transitions, thus suggesting a completion of the input dynamics. The task of graph transformation model inference is naturally highly challenging due to the combinatorics involved. We tackle the exponential explosion by proposing a heuristically minimal translation of the task into a well-established problem, set cover, for which highly optimized solutions exist. We further showcase how our results relate to Kolmogorov complexity expressed in terms of graph transformation.
Knowledge distillation (KD) is a promising yet challenging model compression technique that transfers rich learning representations from a well-performing but cumbersome teacher model to a compact student model. Previous methods for image super-resolution (SR) mostly compare the feature maps directly or after standardizing the dimensions with basic algebraic operations (e.g. average, dot-product). However, the intrinsic semantic differences among feature maps are overlooked, which are caused by the disparate expressive capacity between the networks. This work presents MiPKD, a multi-granularity mixture of prior KD framework, to facilitate efficient SR model through the feature mixture in a unified latent space and stochastic network block mixture. Extensive experiments demonstrate the effectiveness of the proposed MiPKD method.
Structured data, prevalent in tables, databases, and knowledge graphs, poses a significant challenge in its representation. With the advent of large language models (LLMs), there has been a shift towards linearization-based methods, which process structured data as sequential token streams, diverging from approaches that explicitly model structure, often as a graph. Crucially, there remains a gap in our understanding of how these linearization-based methods handle structured data, which is inherently non-linear. This work investigates the linear handling of structured data in encoder-decoder language models, specifically T5. Our findings reveal the model's ability to mimic human-designed processes such as schema linking and syntax prediction, indicating a deep, meaningful learning of structure beyond simple token sequencing. We also uncover insights into the model's internal mechanisms, including the ego-centric nature of structure node encodings and the potential for model compression due to modality fusion redundancy. Overall, this work sheds light on the inner workings of linearization-based methods and could potentially provide guidance for future research.
Post-training quantization (PTQ) is an efficient model compression technique that quantizes a pretrained full-precision model using only a small calibration set of unlabeled samples without retraining. PTQ methods for convolutional neural networks (CNNs) provide quantization results comparable to full-precision counterparts. Directly applying them to vision transformers (ViTs), however, incurs severe performance degradation, mainly due to the differences in architectures between CNNs and ViTs. In particular, the distribution of activations for each channel vary drastically according to input instances, making PTQ methods for CNNs inappropriate for ViTs. To address this, we introduce instance-aware group quantization for ViTs (IGQ-ViT). To this end, we propose to split the channels of activation maps into multiple groups dynamically for each input instance, such that activations within each group share similar statistical properties. We also extend our scheme to quantize softmax attentions across tokens. In addition, the number of groups for each layer is adjusted to minimize the discrepancies between predictions from quantized and full-precision models, under a bit-operation (BOP) constraint. We show extensive experimental results on image classification, object detection, and instance segmentation, with various transformer architectures, demonstrating the effectiveness of our approach.
This paper introduces LLM-Streamline, a pioneer work on layer pruning for large language models (LLMs). It is based on the observation that different layers have varying impacts on hidden states, enabling the identification of less important layers to be pruned.LLM-Streamline comprises two parts: layer pruning, which removes consecutive layers with the lowest importance based on target sparsity, and layer replacement, a novel module that trains a lightweight network to replace the pruned layers to mitigate performance loss. Additionally, a new metric called stability is proposed to address the limitations of the widely used accuracy metric in evaluating model compression. Experiments show that LLM-Streamline outperforms both previous and concurrent state-of-the-art pruning methods in terms of both performance and training efficiency.Our code is available at https://github.com/RUCKBReasoning/LLM-Streamline
The rise of Modular Deep Learning showcases its potential in various Natural Language Processing applications. Parameter-efficient fine-tuning (PEFT) modularity has been shown to work for various use cases, from domain adaptation to multilingual setups. However, all this work covers the case where the modular components are trained and deployed within one single Pre-trained Language Model (PLM). This model-specific setup is a substantial limitation on the very modularity that modular architectures are trying to achieve. We ask whether current modular approaches are transferable between models and whether we can transfer the modules from more robust and larger PLMs to smaller ones. In this work, we aim to fill this gap via a lens of Knowledge Distillation, commonly used for model compression, and present an extremely straightforward approach to transferring pre-trained, task-specific PEFT modules between same-family PLMs. Moreover, we propose a method that allows the transfer of modules between incompatible PLMs without any change in the inference complexity. The experiments on Named Entity Recognition, Natural Language Inference, and Paraphrase Identification tasks over multiple languages and PEFT methods showcase the initial potential of transferable modularity.
Few-shot model compression aims to compress a large model into a more compact one with only a tiny training set (even without labels). Block-level pruning has recently emerged as a leading technique in achieving high accuracy and low latency in few-shot CNN compression. But, few-shot compression for Vision Transformers (ViT) remains largely unexplored, which presents a new challenge. In particular, the issue of sparse compression exists in traditional CNN few-shot methods, which can only produce very few compressed models of different model sizes. This paper proposes a novel framework for few-shot ViT compression named DC-ViT. Instead of dropping the entire block, DC-ViT selectively eliminates the attention module while retaining and reusing portions of the MLP module. DC-ViT enables dense compression, which outputs numerous compressed models that densely populate the range of model complexity. DC-ViT outperforms state-of-the-art few-shot compression methods by a significant margin of 10 percentage points, along with lower latency in the compression of ViT and its variants.
To reduce the inference cost of large language models, model compression is increasingly used to create smaller scalable models. However, little is known about their robustness to minority subgroups defined by the labels and attributes of a dataset. In this paper, we investigate the effects of 18 different compression methods and settings on the subgroup robustness of BERT language models. We show that worst-group performance does not depend on model size alone, but also on the compression method used. Additionally, we find that model compression does not always worsen the performance on minority subgroups. Altogether, our analysis serves to further research into the subgroup robustness of model compression.
Model compression has gained significant popularity as a means to alleviate the computational and memory demands of machine learning models. Each compression technique leverages unique features to reduce the size of neural networks. Although intuitively combining different techniques may enhance compression effectiveness, we find that the order in which they are combined significantly influences performance. To identify the optimal sequence for compressing neural networks, we propose the Order of Compression, a systematic and optimal sequence to apply multiple compression techniques in the most effective order. We start by building the foundations of the orders between any two compression approaches and then demonstrate inserting additional compression between any two compressions will not break the order of the two compression approaches. Based on the foundations, an optimal order is obtained with topological sorting. Validated on image-based regression and classification networks across different datasets, our proposed Order of Compression significantly reduces computational costs by up to 859 times on ResNet34, with negligible accuracy loss (-0.09% for CIFAR10) compared to the baseline model. We believe our simple yet effective exploration of the order of compression will shed light on the practice of model compression.
This paper introduces TinySaver, an early-exit-like dynamic model compression approach which employs tiny models to substitute large models adaptively. Distinct from traditional compression techniques, dynamic methods like TinySaver can leverage the difficulty differences to allow certain inputs to complete their inference processes early, thereby conserving computational resources. Most existing early exit designs are implemented by attaching additional network branches to the model's backbone. Our study, however, reveals that completely independent tiny models can replace a substantial portion of the larger models' job with minimal impact on performance. Employing them as the first exit can remarkably enhance computational efficiency. By searching and employing the most appropriate tiny model as the computational saver for a given large model, the proposed approaches work as a novel and generic method to model compression. This finding will help the research community in exploring new compression methods to address the escalating computational demands posed by rapidly evolving AI models. Our evaluation of this approach in ImageNet-1k classification demonstrates its potential to reduce the number of compute operations by up to 90\%, with only negligible losses in performance, across various modern vision models.
Recent advances in EEG-based BCI technologies have revealed the potential of brain-to-robot collaboration through the integration of sensing, computing, communication, and control. In this paper, we present BRIEDGE as an end-to-end system for multi-brain to multi-robot interaction through an EEG-adaptive neural network and an encoding-decoding communication framework, as illustrated in Fig.1. As depicted, the edge mobile server or edge portable server will collect EEG data from the users and utilize the EEG-adaptive neural network to identify the users' intentions. The encoding-decoding communication framework then encodes the EEG-based semantic information and decodes it into commands in the process of data transmission. To better extract the joint features of heterogeneous EEG data as well as enhance classification accuracy, BRIEDGE introduces an informer-based ProbSparse self-attention mechanism. Meanwhile, parallel and secure transmissions for multi-user multi-task scenarios under physical channels are addressed by dynamic autoencoder and autodecoder communications. From mobile computing and edge AI perspectives, model compression schemes composed of pruning, weight sharing, and quantization are also used to deploy lightweight EEG-adaptive models running on both transmitter and receiver sides. Based on the effectiveness of these components, a code map representing various commands enables multiple users to control multiple intelligent agents concurrently. Our experiments in comparison with state-of-the-art works show that BRIEDGE achieves the best classification accuracy of heterogeneous EEG data, and more stable performance under noisy environments.
Recently, the number of parameters in DNNs has explosively increased, as exemplified by LLMs (Large Language Models), making inference on small-scale computers more difficult. Model compression technology is, therefore, essential for integration into products. In this paper, we propose a method of quantization-aware training. We introduce a novel normalization (Layer-Batch Normalization) that is independent of the mini-batch size and does not require any additional computation cost during inference. Then, we quantize the weights by the scaled round-clip function with the weight standardization. We also quantize activation functions using the same function and apply surrogate gradients to train the model with both quantized weights and the quantized activation functions. We call this method Magic for the age of Quantised DNNs (MaQD). Experimental results show that our quantization method can be achieved with minimal accuracy degradation.
The industrial Internet of Things (IIoT) under Industry 4.0 heralds an era of interconnected smart devices where data-driven insights and machine learning (ML) fuse to revolutionize manufacturing. A noteworthy development in IIoT is the integration of federated learning (FL), which addresses data privacy and security among devices. FL enables edge sensors, also known as peripheral intelligence units (PIUs) to learn and adapt using their data locally, without explicit sharing of confidential data, to facilitate a collaborative yet confidential learning process. However, the lower memory footprint and computational power of PIUs inherently require deep neural network (DNN) models that have a very compact size. Model compression techniques such as pruning can be used to reduce the size of DNN models by removing unnecessary connections that have little impact on the model's performance, thus making the models more suitable for the limited resources of PIUs. Targeting the notion of compact yet robust DNN models, we propose the integration of iterative magnitude pruning (IMP) of the DNN model being trained in an over-the-air FL (OTA-FL) environment for IIoT. We provide a tutorial overview and also present a case study of the effectiveness of IMP in OTA-FL for an IIoT environment. Finally, we present future directions for enhancing and optimizing these deep compression techniques further, aiming to push the boundaries of IIoT capabilities in acquiring compact yet robust and high-performing DNN models.
Progress in machine learning (ML) has been fueled by scaling neural network models. This scaling has been enabled by ever more heroic feats of engineering, necessary for accommodating ML approaches that require high bandwidth communication between devices working in parallel. In this work, we propose a co-designed modular architecture and training approach for ML models, dubbed DIstributed PAth COmposition (DiPaCo). During training, DiPaCo distributes computation by paths through a set of shared modules. Together with a Local-SGD inspired optimization (DiLoCo) that keeps modules in sync with drastically reduced communication, Our approach facilitates training across poorly connected and heterogeneous workers, with a design that ensures robustness to worker failures and preemptions. At inference time, only a single path needs to be executed for each input, without the need for any model compression. We consider this approach as a first prototype towards a new paradigm of large-scale learning, one that is less synchronous and more modular. Our experiments on the widely used C4 benchmark show that, for the same amount of training steps but less wall-clock time, DiPaCo exceeds the performance of a 1 billion-parameter dense transformer language model by choosing one of 256 possible paths, each with a size of 150 million parameters.
Recently, the scale of transformers has grown rapidly, which introduces considerable challenges in terms of training overhead and inference efficiency in the scope of task adaptation. Existing works, namely Parameter-Efficient Fine-Tuning (PEFT) and model compression, have separately investigated the challenges. However, PEFT cannot guarantee the inference efficiency of the original backbone, especially for large-scale models. Model compression requires significant training costs for structure searching and re-training. Consequently, a simple combination of them cannot guarantee accomplishing both training efficiency and inference efficiency with minimal costs. In this paper, we propose a novel Parallel Yielding Re-Activation (PYRA) method for such a challenge of training-inference efficient task adaptation. PYRA first utilizes parallel yielding adaptive weights to comprehensively perceive the data distribution in downstream tasks. A re-activation strategy for token modulation is then applied for tokens to be merged, leading to calibrated token features. Extensive experiments demonstrate that PYRA outperforms all competing methods under both low compression rate and high compression rate, demonstrating its effectiveness and superiority in maintaining both training efficiency and inference efficiency for large-scale foundation models. Our code is available at https://github.com/THU-MIG/PYRA.
As deep learning (DL) models are increasingly being integrated into our everyday lives, ensuring their safety by making them robust against adversarial attacks has become increasingly critical. DL models have been found to be susceptible to adversarial attacks which can be achieved by introducing small, targeted perturbations to disrupt the input data. Adversarial training has been presented as a mitigation strategy which can result in more robust models. This adversarial robustness comes with additional computational costs required to design adversarial attacks during training. The two objectives -- adversarial robustness and computational efficiency -- then appear to be in conflict of each other. In this work, we explore the effects of two different model compression methods -- structured weight pruning and quantization -- on adversarial robustness. We specifically explore the effects of fine-tuning on compressed models, and present the trade-off between standard fine-tuning and adversarial fine-tuning. Our results show that compression does not inherently lead to loss in model robustness and adversarial fine-tuning of a compressed model can yield large improvement to the robustness performance of models. We present experiments on two benchmark datasets showing that adversarial fine-tuning of compressed models can achieve robustness performance comparable to adversarially trained models, while also improving computational efficiency.
The advancements in Large Language Models (LLMs) have been hindered by their substantial sizes, which necessitate LLM compression methods for practical deployment. Singular Value Decomposition (SVD) offers a promising solution for LLM compression. However, state-of-the-art SVD-based LLM compression methods have two key limitations: truncating smaller singular values may lead to higher compression loss, and the lack of update on the compressed weight after SVD truncation. In this work, we propose SVD-LLM, a new SVD-based LLM compression method that addresses the limitations of existing methods. SVD-LLM incorporates a truncation-aware data whitening strategy to ensure a direct mapping between singular values and compression loss. Moreover, SVD-LLM adopts a layer-wise closed-form model parameter update strategy to compensate for accuracy degradation under high compression ratios. We evaluate SVD-LLM on a total of 10 datasets and eight models from three different LLM families at four different scales. Our results demonstrate the superiority of SVD-LLM over state-of-the-arts, especially at high model compression ratios.
When training deep neural networks, the phenomenon of $\textit{dying neurons}$ $\unicode{x2013}$units that become inactive or saturated, output zero during training$\unicode{x2013}$ has traditionally been viewed as undesirable, linked with optimization challenges, and contributing to plasticity loss in continual learning scenarios. In this paper, we reassess this phenomenon, focusing on sparsity and pruning. By systematically exploring the impact of various hyperparameter configurations on dying neurons, we unveil their potential to facilitate simple yet effective structured pruning algorithms. We introduce $\textit{Demon Pruning}$ (DemP), a method that controls the proliferation of dead neurons, dynamically leading to network sparsity. Achieved through a combination of noise injection on active units and a one-cycled schedule regularization strategy, DemP stands out for its simplicity and broad applicability. Experiments on CIFAR10 and ImageNet datasets demonstrate that DemP surpasses existing structured pruning techniques, showcasing superior accuracy-sparsity tradeoffs and training speedups. These findings suggest a novel perspective on dying neurons as a valuable resource for efficient model compression and optimization.
Unsupervised semantic hashing has emerged as an indispensable technique for fast image search, which aims to convert images into binary hash codes without relying on labels. Recent advancements in the field demonstrate that employing large-scale backbones (e.g., ViT) in unsupervised semantic hashing models can yield substantial improvements. However, the inference delay has become increasingly difficult to overlook. Knowledge distillation provides a means for practical model compression to alleviate this delay. Nevertheless, the prevailing knowledge distillation approaches are not explicitly designed for semantic hashing. They ignore the unique search paradigm of semantic hashing, the inherent necessities of the distillation process, and the property of hash codes. In this paper, we propose an innovative Bit-mask Robust Contrastive knowledge Distillation (BRCD) method, specifically devised for the distillation of semantic hashing models. To ensure the effectiveness of two kinds of search paradigms in the context of semantic hashing, BRCD first aligns the semantic spaces between the teacher and student models through a contrastive knowledge distillation objective. Additionally, to eliminate noisy augmentations and ensure robust optimization, a cluster-based method within the knowledge distillation process is introduced. Furthermore, through a bit-level analysis, we uncover the presence of redundancy bits resulting from the bit independence property. To mitigate these effects, we introduce a bit mask mechanism in our knowledge distillation objective. Finally, extensive experiments not only showcase the noteworthy performance of our BRCD method in comparison to other knowledge distillation methods but also substantiate the generality of our methods across diverse semantic hashing models and backbones. The code for BRCD is available at https://github.com/hly1998/BRCD.
Sparsification-based pruning has been an important category in model compression. Existing methods commonly set sparsity-inducing penalty terms to suppress the importance of dropped weights, which is regarded as the suppressed sparsification paradigm. However, this paradigm inactivates the dropped parts of networks causing capacity damage before pruning, thereby leading to performance degradation. To alleviate this issue, we first study and reveal the relative sparsity effect in emerging stimulative training and then propose a structured pruning framework, named STP, based on an enhanced sparsification paradigm which maintains the magnitude of dropped weights and enhances the expressivity of kept weights by self-distillation. Besides, to find an optimal architecture for the pruned network, we propose a multi-dimension architecture space and a knowledge distillation-guided exploration strategy. To reduce the huge capacity gap of distillation, we propose a subnet mutating expansion technique. Extensive experiments on various benchmarks indicate the effectiveness of STP. Specifically, without fine-tuning, our method consistently achieves superior performance at different budgets, especially under extremely aggressive pruning scenarios, e.g., remaining 95.11% Top-1 accuracy (72.43% in 76.15%) while reducing 85% FLOPs for ResNet-50 on ImageNet. Codes will be released soon.