Decentralized distributed learning is the key to enabling large-scale machine learning (training) on edge devices utilizing private user-generated local data, without relying on the cloud. However, the practical realization of such on-device training is limited by the communication and compute bottleneck. In this paper, we propose and show the convergence of low precision decentralized training that aims to reduce the computational complexity and communication cost of decentralized training. Many feedback-based compression techniques have been proposed in the literature to reduce communication costs. To the best of our knowledge, there is no work that applies and shows compute efficient training techniques such as quantization, pruning, etc., for peer-to-peer decentralized learning setups. Since real-world applications have a significant skew in the data distribution, we design "Range-EvoNorm" as the normalization activation layer which is better suited for low precision training over non-IID data. Moreover, we show that the proposed low precision training can be used in synergy with other communication compression methods decreasing the communication cost further. Our experiments indicate that 8-bit decentralized training has minimal accuracy loss compared to its full precision counterpart even with non-IID data. However, when low precision training is accompanied by communication compression through sparsification we observe a 1-2% drop in accuracy. The proposed low precision decentralized training decreases computational complexity, memory usage, and communication cost by 4x and compute energy by a factor of ~20x, while trading off less than a $1\%$ accuracy for both IID and non-IID data. In particular, with higher skew values, we observe an increase in accuracy (by ~ 0.5%) with low precision training, indicating the regularization effect of the quantization.
Learning to synthesize data has emerged as a promising direction in zero-shot quantization (ZSQ), which represents neural networks by low-bit integer without accessing any of the real data. In this paper, we observe an interesting phenomenon of intra-class heterogeneity in real data and show that existing methods fail to retain this property in their synthetic images, which causes a limited performance increase. To address this issue, we propose a novel zero-shot quantization method referred to as IntraQ. First, we propose a local object reinforcement that locates the target objects at different scales and positions of the synthetic images. Second, we introduce a marginal distance constraint to form class-related features distributed in a coarse area. Lastly, we devise a soft inception loss which injects a soft prior label to prevent the synthetic images from being overfitting to a fixed object. Our IntraQ is demonstrated to well retain the intra-class heterogeneity in the synthetic images and also observed to perform state-of-the-art. For example, compared to the advanced ZSQ, our IntraQ obtains 9.17\% increase of the top-1 accuracy on ImageNet when all layers of MobileNetV1 are quantized to 4-bit. Code is at https://github.com/zysxmu/IntraQ.
Modern convolutional neural networks (CNNs) are known to be overconfident in terms of their calibration on unseen input data. That is to say, they are more confident than they are accurate. This is undesirable if the probabilities predicted are to be used for downstream decision making. When considering accuracy, CNNs are also surprisingly robust to compression techniques, such as quantization, which aim to reduce computational and memory costs. We show that this robustness can be partially explained by the calibration behavior of modern CNNs, and may be improved with overconfidence. This is due to an intuitive result: low confidence predictions are more likely to change post-quantization, whilst being less accurate. High confidence predictions will be more accurate, but more difficult to change. Thus, a minimal drop in post-quantization accuracy is incurred. This presents a potential conflict in neural network design: worse calibration from overconfidence may lead to better robustness to quantization. We perform experiments applying post-training quantization to a variety of CNNs, on the CIFAR-100 and ImageNet datasets.
In low-latency or mobile applications, lower computation complexity, lower memory footprint and better energy efficiency are desired. Many prior works address this need by removing redundant parameters. Parameter quantization replaces floating-point arithmetic with lower precision fixed-point arithmetic, further reducing complexity. Typical training of quantized weight neural networks starts from fully quantized weights. Quantization creates random noise. As a way to compensate for this noise, during training, we propose to quantize some weights while keeping others in floating-point precision. A deep neural network has many layers. To arrive at a fully quantized weight network, we start from one quantized layer and then quantize more and more layers. We show that the order of layer quantization affects accuracies. Order count is large for deep neural networks. A sensitivity pre-training is proposed to guide the layer quantization order. Recent work in weight binarization replaces weight-input matrix multiplication with additions. We apply the proposed iterative training to weight binarization. Our experiments cover fully connected and convolutional networks on MNIST, CIFAR-10 and ImageNet datasets. We show empirically that, starting from partial binary weights instead of from fully binary ones, training reaches fully binary weight networks with better accuracies for larger and deeper networks. Layer binarization in the forward order results in better accuracies. Guided layer binarization can further improve that. The improvements come at a cost of longer training time.
Transformer-based language models are applied to a wide range of applications in natural language processing. However, they are inefficient and difficult to deploy. In recent years, many compression algorithms have been proposed to increase the implementation efficiency of large Transformer-based models on target hardware. In this work we present a new method for training sparse pre-trained Transformer language models by integrating weight pruning and model distillation. These sparse pre-trained models can be used to transfer learning for a wide range of tasks while maintaining their sparsity pattern. We demonstrate our method with three known architectures to create sparse pre-trained BERT-Base, BERT-Large and DistilBERT. We show how the compressed sparse pre-trained models we trained transfer their knowledge to five different downstream natural language tasks with minimal accuracy loss. Moreover, we show how to further compress the sparse models' weights to 8bit precision using quantization-aware training. For example, with our sparse pre-trained BERT-Large fine-tuned on SQuADv1.1 and quantized to 8bit we achieve a compression ratio of $40$X for the encoder with less than $1\%$ accuracy loss. To the best of our knowledge, our results show the best compression-to-accuracy ratio for BERT-Base, BERT-Large, and DistilBERT.
Model quantization has emerged as an indispensable technique to accelerate deep learning inference. While researchers continue to push the frontier of quantization algorithms, existing quantization work is often unreproducible and undeployable. This is because researchers do not choose consistent training pipelines and ignore the requirements for hardware deployments. In this work, we propose Model Quantization Benchmark (MQBench), a first attempt to evaluate, analyze, and benchmark the reproducibility and deployability for model quantization algorithms. We choose multiple different platforms for real-world deployments, including CPU, GPU, ASIC, DSP, and evaluate extensive state-of-the-art quantization algorithms under a unified training pipeline. MQBench acts like a bridge to connect the algorithm and the hardware. We conduct a comprehensive analysis and find considerable intuitive or counter-intuitive insights. By aligning the training settings, we find existing algorithms have about the same performance on the conventional academic track. While for the hardware-deployable quantization, there is a huge accuracy gap which remains unsettled. Surprisingly, no existing algorithm wins every challenge in MQBench, and we hope this work could inspire future research directions.
Multi-modal cues, including spatial information, facial expression and voiceprint, are introduced to the speech separation and speaker extraction tasks to serve as complementary information to achieve better performance. However, the introduction of these cues brings about an increasing number of parameters and model complexity, which makes it harder to deploy these models on resource-constrained devices. In this paper, we alleviate the aforementioned problem by proposing a Lightweight Multi-modal framework for Speaker Extraction (LiMuSE). We propose to use GC-equipped TCN, which incorporates Group Communication (GC) and Temporal Convolutional Network (TCN) in the Context Codec module, the audio block and the fusion block. The experiments on the MC_GRID dataset demonstrate that LiMuSE achieves on par or better performance with a much smaller number of parameters and less model complexity. We further investigate the impacts of the quantization of LiMuSE. Our code and dataset are provided.
Model quantization is known as a promising method to compress deep neural networks, especially for inferences on lightweight mobile or edge devices. However, model quantization usually requires access to the original training data to maintain the accuracy of the full-precision models, which is often infeasible in real-world scenarios for security and privacy issues. A popular approach to perform quantization without access to the original data is to use synthetically generated samples, based on batch-normalization statistics or adversarial learning. However, the drawback of such approaches is that they primarily rely on random noise input to the generator to attain diversity of the synthetic samples. We find that this is often insufficient to capture the distribution of the original data, especially around the decision boundaries. To this end, we propose Qimera, a method that uses superposed latent embeddings to generate synthetic boundary supporting samples. For the superposed embeddings to better reflect the original distribution, we also propose using an additional disentanglement mapping layer and extracting information from the full-precision model. The experimental results show that Qimera achieves state-of-the-art performances for various settings on data-free quantization. Code is available at https://github.com/iamkanghyunchoi/qimera.
Vast requirement of computation power of Deep Neural Networks is a major hurdle to their real world applications. Many recent Application Specific Integrated Circuit (ASIC) chips feature dedicated hardware support for Neural Network Acceleration. However, as ASICs take multiple years to develop, they are inevitably out-paced by the latest development in Neural Architecture Research. For example, Transformer Networks do not have native support on many popular chips, and hence are difficult to deploy. In this paper, we propose Arch-Net, a family of Neural Networks made up of only operators efficiently supported across most architectures of ASICs. When a Arch-Net is produced, less common network constructs, like Layer Normalization and Embedding Layers, are eliminated in a progressive manner through label-free Blockwise Model Distillation, while performing sub-eight bit quantization at the same time to maximize performance. Empirical results on machine translation and image classification tasks confirm that we can transform latest developed Neural Architectures into fast running and as-accurate Arch-Net, ready for deployment on multiple mass-produced ASIC chips. The code will be available at https://github.com/megvii-research/Arch-Net.
Distributed learning paradigms such as federated learning often involve transmission of model updates, or gradients, over a network, thereby avoiding transmission of private data. However, it is possible for sensitive information about the training data to be revealed from such gradients. Prior works have demonstrated that labels can be revealed analytically from the last layer of certain models (e.g., ResNet), or they can be reconstructed jointly with model inputs by using Gradients Matching [Zhu et al'19] with additional knowledge about the current state of the model. In this work, we propose a method to discover the set of labels of training samples from only the gradient of the last layer and the id to label mapping. Our method is applicable to a wide variety of model architectures across multiple domains. We demonstrate the effectiveness of our method for model training in two domains - image classification, and automatic speech recognition. Furthermore, we show that existing reconstruction techniques improve their efficacy when used in conjunction with our method. Conversely, we demonstrate that gradient quantization and sparsification can significantly reduce the success of the attack.
In recent years, image recognition applications have developed rapidly. A large number of studies and techniques have emerged in different fields, such as face recognition, pedestrian and vehicle re-identification, landmark retrieval, and product recognition. In this paper, we propose a practical lightweight image recognition system, named PP-ShiTu, consisting of the following 3 modules, mainbody detection, feature extraction and vector search. We introduce popular strategies including metric learning, deep hash, knowledge distillation and model quantization to improve accuracy and inference speed. With strategies above, PP-ShiTu works well in different scenarios with a set of models trained on a mixed dataset. Experiments on different datasets and benchmarks show that the system is widely effective in different domains of image recognition. All the above mentioned models are open-sourced and the code is available in the GitHub repository PaddleClas on PaddlePaddle.
Most state-of-the-art Graph Neural Networks (GNNs) can be defined as a form of graph convolution which can be realized by message passing between direct neighbors or beyond. To scale such GNNs to large graphs, various neighbor-, layer-, or subgraph-sampling techniques are proposed to alleviate the "neighbor explosion" problem by considering only a small subset of messages passed to the nodes in a mini-batch. However, sampling-based methods are difficult to apply to GNNs that utilize many-hops-away or global context each layer, show unstable performance for different tasks and datasets, and do not speed up model inference. We propose a principled and fundamentally different approach, VQ-GNN, a universal framework to scale up any convolution-based GNNs using Vector Quantization (VQ) without compromising the performance. In contrast to sampling-based techniques, our approach can effectively preserve all the messages passed to a mini-batch of nodes by learning and updating a small number of quantized reference vectors of global node representations, using VQ within each GNN layer. Our framework avoids the "neighbor explosion" problem of GNNs using quantized representations combined with a low-rank version of the graph convolution matrix. We show that such a compact low-rank version of the gigantic convolution matrix is sufficient both theoretically and experimentally. In company with VQ, we design a novel approximated message passing algorithm and a nontrivial back-propagation rule for our framework. Experiments on various types of GNN backbones demonstrate the scalability and competitive performance of our framework on large-graph node classification and link prediction benchmarks.
Outlier detection (OD) is a key learning task for finding rare and deviant data samples, with many time-critical applications such as fraud detection and intrusion detection. In this work, we propose TOD, the first tensor-based system for efficient and scalable outlier detection on distributed multi-GPU machines. A key idea behind TOD is decomposing complex OD applications into a small collection of basic tensor algebra operators. This decomposition enables TOD to accelerate OD computations by leveraging recent advances in deep learning infrastructure in both hardware and software. Moreover, to deploy memory-intensive OD applications on modern GPUs with limited on-device memory, we introduce two key techniques. First, provable quantization speeds up OD computations and reduces its memory footprint by automatically performing specific floating-point operations in lower precision while provably guaranteeing no accuracy loss. Second, to exploit the aggregated compute resources and memory capacity of multiple GPUs, we introduce automatic batching, which decomposes OD computations into small batches for both sequential execution on a single GPU and parallel execution on multiple GPUs. TOD supports a diverse set of OD algorithms. Extensive evaluation on 11 real and 3 synthetic OD datasets shows that TOD is on average 10.9x faster than the leading CPU-based OD system PyOD (with a maximum speedup of 38.9x), and can handle much larger datasets than existing GPU-based OD systems. In addition, TOD allows easy integration of new OD operators, enabling fast prototyping of emerging and yet-to-discovered OD algorithms.
Quantization is a popular technique that $transforms$ the parameter representation of a neural network from floating-point numbers into lower-precision ones ($e.g.$, 8-bit integers). It reduces the memory footprint and the computational cost at inference, facilitating the deployment of resource-hungry models. However, the parameter perturbations caused by this transformation result in $behavioral$ $disparities$ between the model before and after quantization. For example, a quantized model can misclassify some test-time samples that are otherwise classified correctly. It is not known whether such differences lead to a new security vulnerability. We hypothesize that an adversary may control this disparity to introduce specific behaviors that activate upon quantization. To study this hypothesis, we weaponize quantization-aware training and propose a new training framework to implement adversarial quantization outcomes. Following this framework, we present three attacks we carry out with quantization: (i) an indiscriminate attack for significant accuracy loss; (ii) a targeted attack against specific samples; and (iii) a backdoor attack for controlling the model with an input trigger. We further show that a single compromised model defeats multiple quantization schemes, including robust quantization techniques. Moreover, in a federated learning scenario, we demonstrate that a set of malicious participants who conspire can inject our quantization-activated backdoor. Lastly, we discuss potential counter-measures and show that only re-training consistently removes the attack artifacts. Our code is available at https://github.com/Secure-AI-Systems-Group/Qu-ANTI-zation
Graph signals arise in various applications, ranging from sensor networks to social media data. The high-dimensional nature of these signals implies that they often need to be compressed in order to be stored and transmitted. The common framework for graph signal compression is based on sampling, resulting in a set of continuous-amplitude samples, which in turn have to be quantized into a finite bit representation. In this work we study the joint design of graph signal sampling along with quantization, for graph signal compression. We focus on bandlimited graph signals, and show that the compression problem can be represented as a task-based quantization setup, in which the task is to recover the spectrum of the signal. Based on this equivalence, we propose a joint design of the sampling and recovery mechanisms for a fixed quantization mapping, and present an iterative algorithm for dividing the available bit budget among the discretized samples. Furthermore, we show how the proposed approach can be realized using graph filters combining elements corresponding the neighbouring nodes of the graph, thus facilitating distributed implementation at reduced complexity. Our numerical evaluations on both synthetic and real world data shows that the joint sampling and quantization method yields a compact finite bit representation of high-dimensional graph signals, which allows reconstruction of the original signal with accuracy within a small gap of that achievable with infinite resolution quantizers.
Parameterized Quantum Circuits (PQC) are promising towards quantum advantage on near-term quantum hardware. However, due to the large quantum noises (errors), the performance of PQC models has a severe degradation on real quantum devices. Take Quantum Neural Network (QNN) as an example, the accuracy gap between noise-free simulation and noisy results on IBMQ-Yorktown for MNIST-4 classification is over 60%. Existing noise mitigation methods are general ones without leveraging unique characteristics of PQC; on the other hand, existing PQC work does not consider noise effect. To this end, we present QuantumNAT, a PQC-specific framework to perform noise-aware optimizations in both training and inference stages to improve robustness. We experimentally observe that the effect of quantum noise to PQC measurement outcome is a linear map from noise-free outcome with a scaling and a shift factor. Motivated by that, we propose post-measurement normalization to mitigate the feature distribution differences between noise-free and noisy scenarios. Furthermore, to improve the robustness against noise, we propose noise injection to the training process by inserting quantum error gates to PQC according to realistic noise models of quantum hardware. Finally, post-measurement quantization is introduced to quantize the measurement outcomes to discrete values, achieving the denoising effect. Extensive experiments on 8 classification tasks using 6 quantum devices demonstrate that QuantumNAT improves accuracy by up to 43%, and achieves over 94% 2-class, 80% 4-class, and 34% 10-class classification accuracy measured on real quantum computers. The code for construction and noise-aware training of PQC is available in the TorchQuantum library.
Deep learning has been recently applied to physical layer processing in digital communication systems in order to improve end-to-end performance. In this work, we introduce a novel deep learning solution for soft bit quantization across wideband channels. Our method is trained end-to-end with quantization- and entropy-aware augmentations to the loss function and is used at inference in conjunction with source coding to achieve near-optimal compression gains over wideband channels. To efficiently train our method, we prove and verify that a fixed feature space quantization scheme is sufficient for efficient learning. When tested on channel distributions never seen during training, the proposed method achieves a compression gain of up to $10 \%$ in the high SNR regime versus previous state-of-the-art methods. To encourage reproducible research, our implementation is publicly available at https://github.com/utcsilab/wideband-llr-deep.
In this paper, we propose a novel Reinforcement Learning (RL) framework for problems with continuous action spaces: Action Quantization from Demonstrations (AQuaDem). The proposed approach consists in learning a discretization of continuous action spaces from human demonstrations. This discretization returns a set of plausible actions (in light of the demonstrations) for each input state, thus capturing the priors of the demonstrator and their multimodal behavior. By discretizing the action space, any discrete action deep RL technique can be readily applied to the continuous control problem. Experiments show that the proposed approach outperforms state-of-the-art methods such as SAC in the RL setup, and GAIL in the Imitation Learning setup. We provide a website with interactive videos: https://google-research.github.io/aquadem/ and make the code available: https://github.com/google-research/google-research/tree/master/aquadem.
Quantization and pruning are core techniques used to reduce the inference costs of deep neural networks. State-of-the-art quantization techniques are currently applied to both the weights and activations; however, pruning is most often applied to only the weights of the network. In this work, we jointly apply novel uniform quantization and unstructured pruning methods to both the weights and activations of deep neural networks during training. Using our methods, we empirically evaluate the currently accepted prune-then-quantize paradigm across a wide range of computer vision tasks and observe a non-commutative nature when applied to both the weights and activations of deep neural networks. Informed by these observations, we articulate the non-commutativity hypothesis: for a given deep neural network being trained for a specific task, there exists an exact training schedule in which quantization and pruning can be introduced to optimize network performance. We identify that this optimal ordering not only exists, but also varies across discriminative and generative tasks. Using the optimal training schedule within our training framework, we demonstrate increased performance per memory footprint over existing solutions.
In the low-bit quantization field, training Binary Neural Networks (BNNs) is the extreme solution to ease the deployment of deep models on resource-constrained devices, having the lowest storage cost and significantly cheaper bit-wise operations compared to 32-bit floating-point counterparts. In this paper, we introduce Sub-bit Neural Networks (SNNs), a new type of binary quantization design tailored to compress and accelerate BNNs. SNNs are inspired by an empirical observation, showing that binary kernels learnt at convolutional layers of a BNN model are likely to be distributed over kernel subsets. As a result, unlike existing methods that binarize weights one by one, SNNs are trained with a kernel-aware optimization framework, which exploits binary quantization in the fine-grained convolutional kernel space. Specifically, our method includes a random sampling step generating layer-specific subsets of the kernel space, and a refinement step learning to adjust these subsets of binary kernels via optimization. Experiments on visual recognition benchmarks and the hardware deployment on FPGA validate the great potentials of SNNs. For instance, on ImageNet, SNNs of ResNet-18/ResNet-34 with 0.56-bit weights achieve 3.13/3.33 times runtime speed-up and 1.8 times compression over conventional BNNs with moderate drops in recognition accuracy. Promising results are also obtained when applying SNNs to binarize both weights and activations. Our code is available at https://github.com/yikaiw/SNN.