Quantization replaces floating point arithmetic with integer arithmetic in deep neural network models, providing more efficient on-device inference with less power and memory. In this work, we propose a framework for formally verifying properties of quantized neural networks. Our baseline technique is based on integer linear programming which guarantees both soundness and completeness. We then show how efficiency can be improved by utilizing gradient-based heuristic search methods and also bound-propagation techniques. We evaluate our approach on perception networks quantized with PyTorch. Our results show that we can verify quantized networks with better scalability and efficiency than the previous state of the art.
End-to-end automatic speech recognition (ASR) models have seen revolutionary quality gains with the recent development of large-scale universal speech models (USM). However, deploying these massive USMs is extremely expensive due to the enormous memory usage and computational cost. Therefore, model compression is an important research topic to fit USM-based ASR under budget in real-world scenarios. In this study, we propose a USM fine-tuning approach for ASR, with a low-bit quantization and N:M structured sparsity aware paradigm on the model weights, reducing the model complexity from parameter precision and matrix topology perspectives. We conducted extensive experiments with a 2-billion parameter USM on a large-scale voice search dataset to evaluate our proposed method. A series of ablation studies validate the effectiveness of up to int4 quantization and 2:4 sparsity. However, a single compression technique fails to recover the performance well under extreme setups including int2 quantization and 1:4 sparsity. By contrast, our proposed method can compress the model to have 9.4% of the size, at the cost of only 7.3% relative word error rate (WER) regressions. We also provided in-depth analyses on the results and discussions on the limitations and potential solutions, which would be valuable for future studies.
Style transfer for out-of-domain (OOD) singing voice synthesis (SVS) focuses on generating high-quality singing voices with unseen styles (such as timbre, emotion, pronunciation, and articulation skills) derived from reference singing voice samples. However, the endeavor to model the intricate nuances of singing voice styles is an arduous task, as singing voices possess a remarkable degree of expressiveness. Moreover, existing SVS methods encounter a decline in the quality of synthesized singing voices in OOD scenarios, as they rest upon the assumption that the target vocal attributes are discernible during the training phase. To overcome these challenges, we propose StyleSinger, the first singing voice synthesis model for zero-shot style transfer of out-of-domain reference singing voice samples. StyleSinger incorporates two critical approaches for enhanced effectiveness: 1) the Residual Style Adaptor (RSA) which employs a residual quantization module to capture diverse style characteristics in singing voices, and 2) the Uncertainty Modeling Layer Normalization (UMLN) to perturb the style attributes within the content representation during the training phase and thus improve the model generalization. Our extensive evaluations in zero-shot style transfer undeniably establish that StyleSinger outperforms baseline models in both audio quality and similarity to the reference singing voice samples. Access to singing voice samples can be found at https://aaronz345.github.io/StyleSingerDemo/.
While transformer models have been highly successful, they are computationally inefficient. We observe that for each layer, the full width of the layer may be needed only for a small subset of tokens inside a batch and that the "effective" width needed to process a token can vary from layer to layer. Motivated by this observation, we introduce the Adaptive Computation Module (ACM), a generic module that dynamically adapts its computational load to match the estimated difficulty of the input on a per-token basis. An ACM consists of a sequence of learners that progressively refine the output of their preceding counterparts. An additional gating mechanism determines the optimal number of learners to execute for each token. We also propose a distillation technique to replace any pre-trained model with an "ACMized" variant. Our evaluation of transformer models in computer vision and speech recognition demonstrates that substituting layers with ACMs significantly reduces inference costs without degrading the downstream accuracy for a wide interval of user-defined budgets.
Transformer-based large language models (e.g., BERT and GPT) achieve great success, and fine-tuning, which tunes a pre-trained model on a task-specific dataset, is the standard practice to utilize these models for downstream tasks. However, Transformer fine-tuning has long running time and high memory consumption due to the large size of the models. We propose the SPT system to fine-tune Transformer-based models efficiently by introducing sparsity. We observe that the memory consumption of Transformer mainly comes from storing attention weights for multi-head attention (MHA), and the majority of running time is spent on feed-forward network (FFN). Thus, we design the sparse MHA module, which computes and stores only large attention weights to reduce memory consumption, and the routed FFN module, which dynamically activates a subset of model parameters for each token to reduce computation cost. We implement SPT on PyTorch and customize CUDA kernels to run sparse MHA and routed FFN efficiently. Specifically, we use product quantization to identify the large attention weights and compute attention via sparse matrix multiplication for sparse MHA. For routed FFN, we batch the tokens according to their activated model parameters for efficient computation. We conduct extensive experiments to evaluate SPT on various model configurations. The results show that SPT consistently outperforms well-optimized baselines, reducing the peak memory consumption by up to 50% and accelerating fine-tuning by up to 2.2x.
Although neural networks have made remarkable advancements in various applications, they require substantial computational and memory resources. Network quantization is a powerful technique to compress neural networks, allowing for more efficient and scalable AI deployments. Recently, Re-parameterization has emerged as a promising technique to enhance model performance while simultaneously alleviating the computational burden in various computer vision tasks. However, the accuracy drops significantly when applying quantization on the re-parameterized networks. We identify that the primary challenge arises from the large variation in weight distribution across the original branches. To address this issue, we propose a coarse & fine weight splitting (CFWS) method to reduce quantization error of weight, and develop an improved KL metric to determine optimal quantization scales for activation. To the best of our knowledge, our approach is the first work that enables post-training quantization applicable on re-parameterized networks. For example, the quantized RepVGG-A1 model exhibits a mere 0.3% accuracy loss. The code is in https://github.com/NeonHo/Coarse-Fine-Weight-Split.git
Image prefiltering with just noticeable distortion (JND) improves coding efficiency in a visual lossless way by filtering the perceptually redundant information prior to compression. However, real JND cannot be well modeled with inaccurate masking equations in traditional approaches or image-level subject tests in deep learning approaches. Thus, this paper proposes a fine-grained JND prefiltering dataset guided by image quality assessment for accurate block-level JND modeling. The dataset is constructed from decoded images to include coding effects and is also perceptually enhanced with block overlap and edge preservation. Furthermore, based on this dataset, we propose a lightweight JND prefiltering network, IQNet, which can be applied directly to different quantization cases with the same model and only needs 3K parameters. The experimental results show that the proposed approach to Versatile Video Coding could yield maximum/average bitrate savings of 41\%/15\% and 53\%/19\% for all-intra and low-delay P configurations, respectively, with negligible subjective quality loss. Our method demonstrates higher perceptual quality and a model size that is an order of magnitude smaller than previous deep learning methods.
This study examines 4-bit quantization methods like GPTQ in large language models (LLMs), highlighting GPTQ's overfitting and limited enhancement in Zero-Shot tasks. While prior works merely focusing on zero-shot measurement, we extend task scope to more generative categories such as code generation and abstractive summarization, in which we found that INT4 quantization can significantly underperform. However, simply shifting to higher precision formats like FP6 has been particularly challenging, thus overlooked, due to poor performance caused by the lack of sophisticated integration and system acceleration strategies on current AI hardware. Our results show that FP6, even with a coarse-grain quantization scheme, performs robustly across various algorithms and tasks, demonstrating its superiority in accuracy and versatility. Notably, with the FP6 quantization, \codestar-15B model performs comparably to its FP16 counterpart in code generation, and for smaller models like the 406M it closely matches their baselines in summarization. Neither can be achieved by INT4. To better accommodate various AI hardware and achieve the best system performance, we propose a novel 4+2 design for FP6 to achieve similar latency to the state-of-the-art INT4 fine-grain quantization. With our design, FP6 can become a promising solution to the current 4-bit quantization methods used in LLMs.
Visual Place Recognition (VPR) is a critical task for performing global re-localization in visual perception systems. It requires the ability to accurately recognize a previously visited location under variations such as illumination, occlusion, appearance and viewpoint. In the case of robotic systems and augmented reality, the target devices for deployment are battery powered edge devices. Therefore whilst the accuracy of VPR methods is important so too is memory consumption and latency. Recently new works have focused on the recall@1 metric as a performance measure with limited focus on resource utilization. This has resulted in methods that use deep learning models too large to deploy on low powered edge devices. We hypothesize that these large models are highly over-parameterized and can be optimized to satisfy the constraints of a low powered embedded system whilst maintaining high recall performance. Our work studies the impact of compact convolutional network architecture design in combination with full-precision and mixed-precision post-training quantization on VPR performance. Importantly we not only measure performance via the recall@1 score but also measure memory consumption and latency. We characterize the design implications on memory, latency and recall scores and provide a number of design recommendations for VPR systems under these resource limitations.
Compressing large neural networks with minimal performance loss is crucial to enabling their deployment on edge devices. (Cho et al., 2022) proposed a weight quantization method that uses an attention-based clustering algorithm called differentiable $k$-means (DKM). Despite achieving state-of-the-art results, DKM's performance is constrained by its heavy memory dependency. We propose an implicit, differentiable $k$-means algorithm (IDKM), which eliminates the major memory restriction of DKM. Let $t$ be the number of $k$-means iterations, $m$ be the number of weight-vectors, and $b$ be the number of bits per cluster address. IDKM reduces the overall memory complexity of a single $k$-means layer from $\mathcal{O}(t \cdot m \cdot 2^b)$ to $\mathcal{O}( m \cdot 2^b)$. We also introduce a variant, IDKM with Jacobian-Free-Backpropagation (IDKM-JFB), for which the time complexity of the gradient calculation is independent of $t$ as well. We provide a proof of concept of our methods by showing that, under the same settings, IDKM achieves comparable performance to DKM with less compute time and less memory. We also use IDKM and IDKM-JFB to quantize a large neural network, Resnet18, on hardware where DKM cannot train at all.
Post-training quantization (PTQ) has played a key role in compressing large language models (LLMs) with ultra-low costs. However, existing PTQ methods only focus on handling the outliers within one layer or one block, which ignores the dependency of blocks and leads to severe performance degradation in low-bit settings. In this paper, we propose CBQ, a cross-block reconstruction-based PTQ method for LLMs. CBQ employs a cross-block dependency using a homologous reconstruction scheme, establishing long-range dependencies across multiple blocks to minimize error accumulation. Furthermore, CBQ incorporates a coarse-to-fine preprocessing (CFP) strategy for suppressing weight and activation outliers, coupled with an adaptive LoRA-Rounding technique for precise weight quantization. These innovations enable CBQ to not only handle extreme outliers effectively but also improve overall quantization accuracy. Extensive experiments show that CBQ achieves superior low-bit quantization (W4A4, W4A8, W2A16) and outperforms existing state-of-the-art methods across various LLMs and datasets. Notably, CBQ quantizes the 4-bit LLAMA1-65B model within only 4.3 hours on a single GPU, achieving a commendable tradeoff between performance and quantization efficiency.
Bio-inspired Spiking Neural Networks (SNN) are now demonstrating comparable accuracy to intricate convolutional neural networks (CNN), all while delivering remarkable energy and latency efficiency when deployed on neuromorphic hardware. In particular, ANN-to-SNN conversion has recently gained significant traction in developing deep SNNs with close to state-of-the-art (SOTA) test accuracy on complex image recognition tasks. However, advanced ANN-to-SNN conversion approaches demonstrate that for lossless conversion, the number of SNN time steps must equal the number of quantization steps in the ANN activation function. Reducing the number of time steps significantly increases the conversion error. Moreover, the spiking activity of the SNN, which dominates the compute energy in neuromorphic chips, does not reduce proportionally with the number of time steps. To mitigate the accuracy concern, we propose a novel ANN-to-SNN conversion framework, that incurs an exponentially lower number of time steps compared to that required in the SOTA conversion approaches. Our framework modifies the SNN integrate-and-fire (IF) neuron model with identical complexity and shifts the bias term of each batch normalization (BN) layer in the trained ANN. To mitigate the spiking activity concern, we propose training the source ANN with a fine-grained L1 regularizer with surrogate gradients that encourages high spike sparsity in the converted SNN. Our proposed framework thus yields lossless SNNs with ultra-low latency, ultra-low compute energy, thanks to the ultra-low timesteps and high spike sparsity, and ultra-high test accuracy, for example, 73.30% with only 4 time steps on the ImageNet dataset.
Unsupervised semantic segmentation (USS) aims to discover and recognize meaningful categories without any labels. For a successful USS, two key abilities are required: 1) information compression and 2) clustering capability. Previous methods have relied on feature dimension reduction for information compression, however, this approach may hinder the process of clustering. In this paper, we propose a novel USS framework called Expand-and-Quantize Unsupervised Semantic Segmentation (EQUSS), which combines the benefits of high-dimensional spaces for better clustering and product quantization for effective information compression. Our extensive experiments demonstrate that EQUSS achieves state-of-the-art results on three standard benchmarks. In addition, we analyze the entropy of USS features, which is the first step towards understanding USS from the perspective of information theory.
The design of sparse linear arrays has proven instrumental in the implementation of cost-effective and efficient automotive radar systems for high-resolution imaging. This paper investigates the impact of coarse quantization on measurements obtained from such arrays. To recover azimuth angles from quantized measurements, we leverage the low-rank properties of the constructed Hankel matrix. In particular, by addressing the one-bit Hankel matrix completion problem through a developed singular value thresholding algorithm, our proposed approach accurately estimates the azimuth angles of interest. We provide comprehensive insights into recovery performance and the required number of one-bit samples. The effectiveness of our proposed scheme is underscored by numerical results, demonstrating successful reconstruction using only one-bit data.
Data privacy is of great concern in cloud machine-learning service platforms, when sensitive data are exposed to service providers. While private computing environments (e.g., secure enclaves), and cryptographic approaches (e.g., homomorphic encryption) provide strong privacy protection, their computing performance still falls short compared to cloud GPUs. To achieve privacy protection with high computing performance, we propose Delta, a new private training and inference framework, with comparable model performance as non-private centralized training. Delta features two asymmetric data flows: the main information-sensitive flow and the residual flow. The main part flows into a small model while the residuals are offloaded to a large model. Specifically, Delta embeds the information-sensitive representations into a low-dimensional space while pushing the information-insensitive part into high-dimension residuals. To ensure privacy protection, the low-dimensional information-sensitive part is secured and fed to a small model in a private environment. On the other hand, the residual part is sent to fast cloud GPUs, and processed by a large model. To further enhance privacy and reduce the communication cost, Delta applies a random binary quantization technique along with a DP-based technique to the residuals before sharing them with the public platform. We theoretically show that Delta guarantees differential privacy in the public environment and greatly reduces the complexity in the private environment. We conduct empirical analyses on CIFAR-10, CIFAR-100 and ImageNet datasets and ResNet-18 and ResNet-34, showing that Delta achieves strong privacy protection, fast training, and inference without significantly compromising the model utility.
Large Language Models (LLMs) trained with self-supervision on vast corpora of web text fit to the social biases of that text. Without intervention, these social biases persist in the model's predictions in downstream tasks, leading to representational harm. Many strategies have been proposed to mitigate the effects of inappropriate social biases learned during pretraining. Simultaneously, methods for model compression have become increasingly popular to reduce the computational burden of LLMs. Despite the popularity and need for both approaches, little work has been done to explore the interplay between these two. We perform a carefully controlled study of the impact of model compression via quantization and knowledge distillation on measures of social bias in LLMs. Longer pretraining and larger models led to higher social bias, and quantization showed a regularizer effect with its best trade-off around 20% of the original pretraining time.
We propose a new model architecture specifically suited for text-to-speech (TTS) models. We combine WavLM, a pre-trained self-supervised learning (SSL) speech model, and the BEST-RQ vector quantization framework. We assess the extent to which the more task-agnostic WavLM, coupled with the superior suitability of the simplistic BEST-RQ framework for a wider array of downstream tasks, yields favorable outcomes. Experiments on the LibriSpeech dataset with SUPERB benchmarking assert that the proposed model significantly underperforms. We speculate the underlying reason for this performance is related to the difference between featurizing raw audio waveforms and spectrograms with a quantizer. We discuss the limitations of this approach to better guide future advancements in TTS.
Offline meta-reinforcement learning (meta-RL) methods, which adapt to unseen target tasks with prior experience, are essential in robot control tasks. Current methods typically utilize task contexts and skills as prior experience, where task contexts are related to the information within each task and skills represent a set of temporally extended actions for solving subtasks. However, these methods still suffer from limited performance when adapting to unseen target tasks, mainly because the learned prior experience lacks generalization, i.e., they are unable to extract effective prior experience from meta-training tasks by exploration and learning of continuous latent spaces. We propose a framework called decoupled meta-reinforcement learning (DCMRL), which (1) contrastively restricts the learning of task contexts through pulling in similar task contexts within the same task and pushing away different task contexts of different tasks, and (2) utilizes a Gaussian quantization variational autoencoder (GQ-VAE) for clustering the Gaussian distributions of the task contexts and skills respectively, and decoupling the exploration and learning processes of their spaces. These cluster centers which serve as representative and discrete distributions of task context and skill are stored in task context codebook and skill codebook, respectively. DCMRL can acquire generalizable prior experience and achieve effective adaptation to unseen target tasks during the meta-testing phase. Experiments in the navigation and robot manipulation continuous control tasks show that DCMRL is more effective than previous meta-RL methods with more generalizable prior experience.
We develop an automated pipeline to streamline neural architecture codesign for fast, real-time Bragg peak analysis in high-energy diffraction microscopy. Traditional approaches, notably pseudo-Voigt fitting, demand significant computational resources, prompting interest in deep learning models for more efficient solutions. Our method employs neural architecture search and AutoML to enhance these models, including hardware costs, leading to the discovery of more hardware-efficient neural architectures. Our results match the performance, while achieving a 13$\times$ reduction in bit operations compared to the previous state-of-the-art. We show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains.
In the realm of deep neural network deployment, low-bit quantization presents a promising avenue for enhancing computational efficiency. However, it often hinges on the availability of training data to mitigate quantization errors, a significant challenge when data availability is scarce or restricted due to privacy or copyright concerns. Addressing this, we introduce GenQ, a novel approach employing an advanced Generative AI model to generate photorealistic, high-resolution synthetic data, overcoming the limitations of traditional methods that struggle to accurately mimic complex objects in extensive datasets like ImageNet. Our methodology is underscored by two robust filtering mechanisms designed to ensure the synthetic data closely aligns with the intrinsic characteristics of the actual training data. In case of limited data availability, the actual data is used to guide the synthetic data generation process, enhancing fidelity through the inversion of learnable token embeddings. Through rigorous experimentation, GenQ establishes new benchmarks in data-free and data-scarce quantization, significantly outperforming existing methods in accuracy and efficiency, thereby setting a new standard for quantization in low data regimes. Code is released at \url{https://github.com/Intelligent-Computing-Lab-Yale/GenQ}.