arXiv subjects

Yun Liang

Publications and source records attributed to Yun Liang.

At least 37 records · Page 2Linked to original sources

arXiv2025

Open-Source Acceleration of Stable-Diffusion.cpp Deployable on All Devices

Stable diffusion plays a crucial role in generating high-quality images. However, image generation is time-consuming and memory-intensive. To address this, stable-diffusion.cpp (Sdcpp) emerges as an efficient inference framework to accelerate the diffusion models. Although it is lightweight, the current implementation of ggml_conv_2d operator in Sdcpp is suboptimal, exhibiting both high inference latency and massive memory usage. To address this, in this work, we present an optimized version of Sdcpp leveraging the Winograd algorithm to accelerate 2D convolution operations, which is the primary bottleneck in the pipeline. By analyzing both dependent and independent computation graphs, we exploit the device's locality and parallelism to achieve substantial performance improvements. Our framework delivers correct end-to-end results across various stable diffusion models, including SDv1.4, v1.5, v2.1, SDXL, and SDXL-Turbo. Our evaluation results demonstrate a speedup up to 2.76x for individual convolutional layers and an inference speedup up to 4.79x for the overall image generation process, compared with the original Sdcpp on M1 pro. Homepage: https://github.com/SealAILab/stable-diffusion-cpp

cs.CV↗

arXiv2024

Scattering halos in strongly interacting Feshbach molecular Bose-Einstein condensates

We investigate the scattering halos resulting from collisions between discrete momentum components in the time-of-flight expansion of interaction-tunable $^6\rm Li_2$ molecular Bose-Einstein condensates. A key highlight of this study is the observation of the influence of interactions on the collisional scattering process. We measure the production of scattering halos at different interaction levels by varying the number of particles and the scattering length, and quantitatively assess the applicability of perturbation theory. To delve into a general theory of scattering halos, we introduce a scattering factor and obtain a universal relation between it and the halo ratio. Furthermore, we simulate the formation of scattering halos under non-perturbative conditions and analyze the discrepancies between simulation results and experiments through a return pulse experiment. This study enhances our understanding of the physical mechanisms underlying scattering processes in many-body systems and provides new perspectives for further theoretical research.

cond-mat.quant-gas↗

arXiv2024

The width of $f_{0}(980)$ in isospin-symmetry-breaking decays

The scalar meson $f_{0}(980)$ has long posed a perplexing puzzle within the realm of light hadron physics. Conventionally, its mass and width in normal decay processes have been estimated as $M=990\pm20$~MeV/$c^2$ and $Γ=40-100$~MeV, respectively. Theoretical explanations regarding the internal structure of $f_{0}(980)$ range from it being a conventional quark-antiquark meson to a tetraquark state, a $K\overline{K}$ molecule, or even a quark-antiquark gluon hybrid. However, a definitive consensus has remained elusive over a considerable duration. Recent observations by the BESIII experiment have unveiled anomalously narrow widths of $f_{0}(980)$ in five independent isospin-symmetry-breaking decay channels. Harnessing these experimental findings, we performed a simultaneous fit to the $ππ$ invariant mass distributions, resulting in a refined determination of the mass and width in isospin-symmetry-breaking decays as $M=990.0\pm0.4(\text{stat})\pm0.1(\text{syst})$~MeV/$c^2$ and $Γ=11.4\pm1.1(\text{stat})\pm0.9(\text{syst})$~MeV, respectively. Here, the first errors are statistical and the second are systematic. Furthermore, by employing the parameterized Flatté formula to fit the same $ππ$ invariant mass distributions, we ascertained the values of the two coupling constants, $g_{fππ}$ and $g_{fK\overline{K}}$, as $g_{fππ}=0.46\pm0.03$ and $g_{fK\overline{K}}=1.24\pm0.32$, respectively. Based on the joint confidence regions of $g_{fππ}$ and $g_{fK\overline{K}}$, we draw the conclusion that the experimental data exhibit a propensity to favor the $K\overline{K}$ molecule model and the quark-antiquark ($q\bar{q}$) model, while offering relatively less support for the tetraquarks ($q^{2}\bar{q}^{2}$) model and the quark-antiquark gluon ($q\bar{q}g$) hybrid model.

hep-ph↗

arXiv2024

Collisional scattering of strongly interacting D-band Feshbach molecules in optical lattices

The excited bands in optical lattices manifest an important tool for studying quantum simulation and many-body physics, making it crucial to measure high-band scattering dynamics under strong interactions. This work investigates both experimentally and theoretically the collisional scattering of $^{6}\rm Li_2$ molecular Bose-Einstein condensate in the $D$ band of a one-dimensional optical lattice, with interaction strength directly tunable via magnetic Feshbach resonance. We find a clear dependence of the $D$-band lifetimes on the interaction strength within the strongly interacting regime, which arises from the fact that the scattering cross-section is proportional to the square of the scattering length. The maximum lifetime versus lattice depth is measured to reveal the effects of interactions. We also investigate the scattering channels of $D$-band molecules under different interaction levels and develop a reliable two-body scattering rate equation. This work provides insight into the interplay between interaction and the collisional scattering of high-band bosons in optical lattices, paving the way for research into strong correlation effects in high-band lattice systems.

cond-mat.quant-gas↗

arXiv2024

Theseus: Exploring Efficient Wafer-Scale Chip Design for Large Language Models

The emergence of the large language model~(LLM) poses an exponential growth of demand for computation throughput, memory capacity, and communication bandwidth. Such a demand growth has significantly surpassed the improvement of corresponding chip designs. With the advancement of fabrication and integration technologies, designers have been developing Wafer-Scale Chips~(WSCs) to scale up and exploit the limits of computation density, memory capacity, and communication bandwidth at the level of a single chip. Existing solutions have demonstrated the significant advantages of WSCs over traditional designs, showing potential to effectively support LLM workloads. Despite the benefits, exploring the early-stage design space of WSCs for LLMs is a crucial yet challenging task due to the enormous and complicated design space, time-consuming evaluation methods, and inefficient exploration strategies. To address these challenges, we propose Theseus, an efficient WSC design space exploration framework for LLMs. We construct the design space of WSCs with various constraints considering the unique characteristics of WSCs. We propose efficient evaluation methodologies for large-scale NoC-based WSCs and introduce multi-fidelity Bayesian optimization to efficiently explore the design space. Evaluation results demonstrate the efficiency of Theseus that the searched Pareto optimal results outperform GPU cluster and existing WSC designs by up to 62.8\%/73.7\% in performance and 38.6\%/42.4\% in power consumption for LLM training, while improving up to 23.2$\times$ and 15.7$\times$ for the performance and power of inference tasks. Furthermore, we conduct case studies to address the design tradeoffs in WSCs and provide insights to facilitate WSC designs for LLMs.

cs.AR↗

arXiv2024

MCUBERT: Memory-Efficient BERT Inference on Commodity Microcontrollers

In this paper, we propose MCUBERT to enable language models like BERT on tiny microcontroller units (MCUs) through network and scheduling co-optimization. We observe the embedding table contributes to the major storage bottleneck for tiny BERT models. Hence, at the network level, we propose an MCU-aware two-stage neural architecture search algorithm based on clustered low-rank approximation for embedding compression. To reduce the inference memory requirements, we further propose a novel fine-grained MCU-friendly scheduling strategy. Through careful computation tiling and re-ordering as well as kernel design, we drastically increase the input sequence lengths supported on MCUs without any latency or accuracy penalty. MCUBERT reduces the parameter size of BERT-tiny and BERT-mini by 5.7$\times$ and 3.0$\times$ and the execution memory by 3.5$\times$ and 4.3$\times$, respectively. MCUBERT also achieves 1.5$\times$ latency reduction. For the first time, MCUBERT enables lightweight BERT models on commodity MCUs and processing more than 512 tokens with less than 256KB of memory.

cs.LG↗

arXiv2024

Memory Matching is not Enough: Jointly Improving Memory Matching and Decoding for Video Object Segmentation

Memory-based video object segmentation methods model multiple objects over long temporal-spatial spans by establishing memory bank, which achieve the remarkable performance. However, they struggle to overcome the false matching and are prone to lose critical information, resulting in confusion among different objects. In this paper, we propose an effective approach which jointly improving the matching and decoding stages to alleviate the false matching issue.For the memory matching stage, we present a cost aware mechanism that suppresses the slight errors for short-term memory and a shunted cross-scale matching for long-term memory which establish a wide filed matching spaces for various object scales. For the readout decoding stage, we implement a compensatory mechanism aims at recovering the essential information where missing at the matching stage. Our approach achieves the outstanding performance in several popular benchmarks (i.e., DAVIS 2016&2017 Val (92.4%&88.1%), and DAVIS 2017 Test (83.9%)), and achieves 84.8%&84.6% on YouTubeVOS 2018&2019 Val.

cs.CV↗

arXiv2024

ToCoAD: Two-Stage Contrastive Learning for Industrial Anomaly Detection

Current unsupervised anomaly detection approaches perform well on public datasets but struggle with specific anomaly types due to the domain gap between pre-trained feature extractors and target-specific domains. To tackle this issue, this paper presents a two-stage training strategy, called \textbf{ToCoAD}. In the first stage, a discriminative network is trained by using synthetic anomalies in a self-supervised learning manner. This network is then utilized in the second stage to provide a negative feature guide, aiding in the training of the feature extractor through bootstrap contrastive learning. This approach enables the model to progressively learn the distribution of anomalies specific to industrial datasets, effectively enhancing its generalizability to various types of anomalies. Extensive experiments are conducted to demonstrate the effectiveness of our proposed two-stage training strategy, and our model produces competitive performance, achieving pixel-level AUROC scores of 98.21\%, 98.43\% and 97.70\% on MVTec AD, VisA and BTAD respectively.

cs.CV↗

arXiv2024

vMCU: Coordinated Memory Management and Kernel Optimization for DNN Inference on MCUs

IoT devices based on microcontroller units (MCU) provide ultra-low power consumption and ubiquitous computation for near-sensor deep learning models (DNN). However, the memory of MCU is usually 2-3 orders of magnitude smaller than mobile devices, which makes it challenging to map DNNs onto MCUs. Previous work separates memory management and kernel implementation for MCU and relies on coarse-grained memory management techniques such as inplace update to reduce memory consumption. In this paper, we propose to coordinate memory management and kernel optimization for DNN inference on MCUs to enable fine-grained memory management. The key idea is to virtualize the limited memory of MCU as a large memory pool. Each kernel divides the memory pool into kernel-specific segments and handles segment load and store while computing DNN layers. Memory consumption can be reduced because using the fine-grained segment-level memory control, we can overlap the memory footprint of different tensors without the need to materialize them at the same time. Following this idea, we implement \ours{} for DNN inference on MCU. Evaluation for single layers on ARM Cortex-M4 and Cortex-M7 processors shows that \ours{} can reduce from $12.0\%$ to $49.5\%$ RAM usage and from $20.6\%$ to $53.0\%$ energy consumption compared to state-of-the-art work. For full DNN evaluation, \ours{} can reduce the memory bottleneck by $61.5\%$, enabling more models to be deployed on low-end MCUs.

cs.AR↗

arXiv2024

The Dawn of AI-Native EDA: Opportunities and Challenges of Large Circuit Models

Within the Electronic Design Automation (EDA) domain, AI-driven solutions have emerged as formidable tools, yet they typically augment rather than redefine existing methodologies. These solutions often repurpose deep learning models from other domains, such as vision, text, and graph analytics, applying them to circuit design without tailoring to the unique complexities of electronic circuits. Such an AI4EDA approach falls short of achieving a holistic design synthesis and understanding, overlooking the intricate interplay of electrical, logical, and physical facets of circuit data. This paper argues for a paradigm shift from AI4EDA towards AI-native EDA, integrating AI at the core of the design process. Pivotal to this vision is the development of a multimodal circuit representation learning technique, poised to provide a comprehensive understanding by harmonizing and extracting insights from varied data sources, such as functional specifications, RTL designs, circuit netlists, and physical layouts. We champion the creation of large circuit models (LCMs) that are inherently multimodal, crafted to decode and express the rich semantics and structures of circuit data, thus fostering more resilient, efficient, and inventive design methodologies. Embracing this AI-native philosophy, we foresee a trajectory that transcends the current innovation plateau in EDA, igniting a profound shift-left in electronic design methodology. The envisioned advancements herald not just an evolution of existing EDA tools but a revolution, giving rise to novel instruments of design tools that promise to radically enhance design productivity and inaugurate a new epoch where the optimization of circuit performance, power, and area (PPA) is achieved not incrementally, but through leaps that redefine the benchmarks of electronic systems' capabilities.

cs.AR↗

arXiv2024

Monad: Towards Cost-effective Specialization for Chiplet-based Spatial Accelerators

Advanced packaging offers a new design paradigm in the post-Moore era, where many small chiplets can be assembled into a large system. Based on heterogeneous integration, a chiplet-based accelerator can be highly specialized for a specific workload, demonstrating extreme efficiency and cost reduction. To fully leverage this potential, it is critical to explore both the architectural design space for individual chiplets and different integration options to assemble these chiplets, which have yet to be fully exploited by existing proposals. This paper proposes Monad, a cost-aware specialization approach for chiplet-based spatial accelerators that explores the tradeoffs between PPA and fabrication costs. To evaluate a specialized system, we introduce a modeling framework considering the non-uniformity in dataflow, pipelining, and communications when executing multiple tensor workloads on different chiplets. We propose to combine the architecture and integration design space by uniformly encoding the design aspects for both spaces and exploring them with a systematic ML-based approach. The experiments demonstrate that Monad can achieve an average of 16% and 30% EDP reduction compared with the state-of-the-art chiplet-based accelerators, Simba and NN-Baton, respectively.

cs.AR↗

arXiv2024

Causal Intervention for Subject-Deconfounded Facial Action Unit Recognition

Subject-invariant facial action unit (AU) recognition remains challenging for the reason that the data distribution varies among subjects. In this paper, we propose a causal inference framework for subject-invariant facial action unit recognition. To illustrate the causal effect existing in AU recognition task, we formulate the causalities among facial images, subjects, latent AU semantic relations, and estimated AU occurrence probabilities via a structural causal model. By constructing such a causal diagram, we clarify the causal effect among variables and propose a plug-in causal intervention module, CIS, to deconfound the confounder \emph{Subject} in the causal diagram. Extensive experiments conducted on two commonly used AU benchmark datasets, BP4D and DISFA, show the effectiveness of our CIS, and the model with CIS inserted, CISNet, has achieved state-of-the-art performance.

cs.CV↗

arXiv2024

Temporal Talbot interferometer of strongly interacting molecular Bose-Einstein condensate

Talbot interferometer, as a periodic reproduction of momentum distribution in the time domain, finds significant applications in multiple research. The inter-particle interactions during the diffraction and interference process introduce numerous many-body physics problems, leading to unconventional interference characteristics. This work investigates both experimentally and theoretically the influence of interaction in a Talbot interferometer with a $^{6}\rm Li_2$ molecular Bose-Einstein condensate in a one-dimensional optical lattice, with interaction strength directly tunable via magnetic Feshbach resonance. A clear dependence of the period and amplitude of signal revivals on the interaction strength can be observed. While interactions increase the decay rate of the signal and advance the revivals, we find that over a wide range of interactions, the Talbot interferometer remains highly effective over a certain evolutionary timescale, including the case of fractional Talbot interference. This work provides insight into the interplay between interaction and the coherence properties of a temporal Talbot interference in optical lattices, paving the way for research into quantum interference in strongly interacting systems.

cond-mat.quant-gas↗

arXiv2024

AAT: Adapting Audio Transformer for Various Acoustics Recognition Tasks

Recently, Transformers have been introduced into the field of acoustics recognition. They are pre-trained on large-scale datasets using methods such as supervised learning and semi-supervised learning, demonstrating robust generality--It fine-tunes easily to downstream tasks and shows more robust performance. However, the predominant fine-tuning method currently used is still full fine-tuning, which involves updating all parameters during training. This not only incurs significant memory usage and time costs but also compromises the model's generality. Other fine-tuning methods either struggle to address this issue or fail to achieve matching performance. Therefore, we conducted a comprehensive analysis of existing fine-tuning methods and proposed an efficient fine-tuning approach based on Adapter tuning, namely AAT. The core idea is to freeze the audio Transformer model and insert extra learnable Adapters, efficiently acquiring downstream task knowledge without compromising the model's original generality. Extensive experiments have shown that our method achieves performance comparable to or even superior to full fine-tuning while optimizing only 7.118% of the parameters. It also demonstrates superiority over other fine-tuning methods.

cs.SD↗

arXiv2023

Structure-Aware Parametric Representations for Time-Resolved Light Transport

Time-resolved illumination provides rich spatio-temporal information for applications such as accurate depth sensing or hidden geometry reconstruction, becoming a useful asset for prototyping and as input for data-driven approaches. However, time-resolved illumination measurements are high-dimensional and have a low signal-to-noise ratio, hampering their applicability in real scenarios. We propose a novel method to compactly represent time-resolved illumination using mixtures of exponentially-modified Gaussians that are robust to noise and preserve structural information. Our method yields representations two orders of magnitude smaller than discretized data, providing consistent results in applications such as hidden scene reconstruction and depth estimation, and quantitative improvements over previous approaches.

eess.SP↗

arXiv2023

Memory-aware Scheduling for Complex Wired Networks with Iterative Graph Optimization

Memory-aware network scheduling is becoming increasingly important for deep neural network (DNN) inference on resource-constrained devices. However, due to the complex cell-level and network-level topologies, memory-aware scheduling becomes very challenging. While previous algorithms all suffer from poor scalability, in this paper, we propose an efficient memory-aware scheduling framework based on iterative computation graph optimization. Our framework features an iterative graph fusion algorithm that simplifies the computation graph while preserving the scheduling optimality. We further propose an integer linear programming formulation together with topology-aware variable pruning to schedule the simplified graph efficiently. We evaluate our method against prior-art algorithms on different networks and demonstrate that our method outperforms existing techniques in all the benchmarks, reducing the peak memory footprint by 13.4%, and achieving better scalability for networks with complex network-level topologies.

cs.LG↗

arXiv2023

OpenPARF: An Open-Source Placement and Routing Framework for Large-Scale Heterogeneous FPGAs with Deep Learning Toolkit

This paper proposes OpenPARF, an open-source placement and routing framework for large-scale FPGA designs. OpenPARF is implemented with the deep learning toolkit PyTorch and supports massive parallelization on GPU. The framework proposes a novel asymmetric multi-electrostatic field system to solve FPGA placement. It considers fine-grained routing resources inside configurable logic blocks (CLBs) for FPGA routing and supports large-scale irregular routing resource graphs. Experimental results on ISPD 2016 and ISPD 2017 FPGA contest benchmarks and industrial benchmarks demonstrate that OpenPARF can achieve 0.4-12.7% improvement in routed wirelength and more than $2\times$ speedup in placement. We believe that OpenPARF can pave the road for developing FPGA physical design engines and stimulate further research on related topics.

cs.AR↗

arXiv2023

Proteus: Simulating the Performance of Distributed DNN Training

DNN models are becoming increasingly larger to achieve unprecedented accuracy, and the accompanying increased computation and memory requirements necessitate the employment of massive clusters and elaborate parallelization strategies to accelerate DNN training. In order to better optimize the performance and analyze the cost, it is indispensable to model the training throughput of distributed DNN training. However, complex parallelization strategies and the resulting complex runtime behaviors make it challenging to construct an accurate performance model. In this paper, we present Proteus, the first standalone simulator to model the performance of complex parallelization strategies through simulation execution. Proteus first models complex parallelization strategies with a unified representation named Strategy Tree. Then, it compiles the strategy tree into a distributed execution graph and simulates the complex runtime behaviors, comp-comm overlap and bandwidth sharing, with a Hierarchical Topo-Aware Executor (HTAE). We finally evaluate Proteus across a wide variety of DNNs on three hardware configurations. Experimental results show that Proteus achieves $3.0\%$ average prediction error and preserves order for training throughput of various parallelization strategies. Compared to state-of-the-art approaches, Proteus reduces prediction error by up to $133.8\%$.

cs.DC↗