Huaizheng Zhang

Huaizheng Zhang, Yuanming Li, Wencong Xiao et al. · berkeley

New architecture GPUs like A100 are now equipped with multi-instance GPU (MIG) technology, which allows the GPU to be partitioned into multiple small, isolated instances. This technology provides more flexibility for users to support both deep learning training and inference workloads, but efficiently utilizing it can still be challenging. The vision of this paper is to provide a more comprehensive and practical benchmark study for MIG in order to eliminate the need for tedious manual benchmarking and tuning efforts. To achieve this vision, the paper presents MIGPerf, an open-source tool that streamlines the benchmark study for MIG. Using MIGPerf, the authors conduct a series of experiments, including deep learning training and inference characterization on MIG, GPU sharing characterization, and framework compatibility with MIG. The results of these experiments provide new insights and guidance for users to effectively employ MIG, and lay the foundation for further research on the orchestration of hybrid training and inference workloads on MIGs. The code and results are released on https://github.com/MLSysOps/MIGProfiler. This work is still in progress and more results will be published soon.

Yizheng Huang, Huaizheng Zhang, Yuanming Li et al. · berkeley

The success of today's AI applications requires not only model training (Model-centric) but also data engineering (Data-centric). In data-centric AI, active learning (AL) plays a vital role, but current AL tools 1) require users to manually select AL strategies, and 2) can not perform AL tasks efficiently. To this end, this paper presents an automatic and efficient MLOps system for AL, named ALaaS (Active-Learning-as-a-Service). Specifically, 1) ALaaS implements an AL agent, including a performance predictor and a workflow controller, to decide the most suitable AL strategies given users' datasets and budgets. We call this a predictive-based successive halving early-stop (PSHEA) procedure. 2) ALaaS adopts a server-client architecture to support an AL pipeline and implements stage-level parallelism for high efficiency. Meanwhile, caching and batching techniques are employed to further accelerate the AL process. In addition to efficiency, ALaaS ensures accessibility with the help of the design philosophy of configuration-as-a-service. Extensive experiments show that ALaaS outperforms all other baselines in terms of latency and throughput. Also, guided by the AL agent, ALaaS can automatically select and run AL strategies for non-expert users under different datasets and budgets. Our code is available at \url{https://github.com/MLSysOps/Active-Learning-as-a-Service}.

Lei Zhang, Guanyu Gao, Huaizheng Zhang

Data drift is a thorny challenge when deploying person re-identification (ReID) models into real-world devices, where the data distribution is significantly different from that of the training environment and keeps changing. To tackle this issue, we propose a federated spatial-temporal incremental learning approach, named FedSTIL, which leverages both lifelong learning and federated learning to continuously optimize models deployed on many distributed edge clients. Unlike previous efforts, FedSTIL aims to mine spatial-temporal correlations among the knowledge learnt from different edge clients. Specifically, the edge clients first periodically extract general representations of drifted data to optimize their local models. Then, the learnt knowledge from edge clients will be aggregated by centralized parameter server, where the knowledge will be selectively and attentively distilled from spatial- and temporal-dimension with carefully designed mechanisms. Finally, the distilled informative spatial-temporal knowledge will be sent back to correlated edge clients to further improve the recognition accuracy of each edge client with a lifelong learning method. Extensive experiments on a mixture of five real-world datasets demonstrate that our method outperforms others by nearly 4% in Rank-1 accuracy, while reducing communication cost by 62%. All implementation codes are publicly available on https://github.com/MSNLAB/Federated-Lifelong-Person-ReID

Chenhao Ye, Huaizheng Zhang, Mingcong Han et al.

Modern LLM reinforcement learning (RL) workloads require a highly efficient weight transfer system to scale training across heterogeneous computational resources. However, existing weight transfer approaches either fail to provide flexibility for dynamically scaling clusters or incur fundamental data movement overhead, resulting in poor performance. We introduce Reference-Oriented Storage (ROS), a new storage abstraction for RL weight transfer that exploits the highly replicated model weights in place. ROS presents the illusion that certain versions of the model weights are stored and can be fetched on demand. Underneath, ROS does not physically store any copies of the weights; instead, it tracks the workers that hold these weights on GPUs for inference. Upon request, ROS directly uses them to serve reads. We build TensorHub, a production-quality system that extends the ROS idea with topology-optimized transfer, strong consistency, and fault tolerance. Evaluation shows that TensorHub fully saturates RDMA bandwidth and adapts to three distinct rollout workloads with minimal engineering effort. Specifically, TensorHub reduces total GPU stall time by up to 6.7x for standalone rollouts, accelerates weight update for elastic rollout by 4.8x, and cuts cross-datacenter rollout stall time by 19x. TensorHub has been deployed in production to support cutting-edge RL training.

Huaizheng Zhang, Yizheng Huang, Yuanming Li

We introduce DataCI, a comprehensive open-source platform designed specifically for data-centric AI in dynamic streaming data settings. DataCI provides 1) an infrastructure with rich APIs for seamless streaming dataset management, data-centric pipeline development and evaluation on streaming scenarios, 2) an carefully designed versioning control function to track the pipeline lineage, and 3) an intuitive graphical interface for a better interactive user experience. Preliminary studies and demonstrations attest to the easy-to-use and effectiveness of DataCI, highlighting its potential to revolutionize the practice of data-centric AI in streaming data contexts.

Mingjia Shi, Yuhao Zhou, Kai Wang et al.

Classical federated learning (FL) enables training machine learning models without sharing data for privacy preservation, but heterogeneous data characteristic degrades the performance of the localized model. Personalized FL (PFL) addresses this by synthesizing personalized models from a global model via training on local data. Such a global model may overlook the specific information that the clients have been sampled. In this paper, we propose a novel scheme to inject personalized prior knowledge into the global model in each client, which attempts to mitigate the introduced incomplete information problem in PFL. At the heart of our proposed approach is a framework, the PFL with Bregman Divergence (pFedBreD), decoupling the personalized prior from the local objective function regularized by Bregman divergence for greater adaptability in personalized scenarios. We also relax the mirror descent (RMD) to extract the prior explicitly to provide optional strategies. Additionally, our pFedBreD is backed up by a convergence analysis. Sufficient experiments demonstrate that our method reaches the state-of-the-art performances on 5 datasets and outperforms other methods by up to 3.5% across 8 benchmarks. Extensive analyses verify the robustness and necessity of proposed designs.

Huaizheng Zhang, Yuanming Li, Qiming Ai et al.

Combining \underline{v}ideo streaming and online \underline{r}etailing (V2R) has been a growing trend recently. In this paper, we provide practitioners and researchers in multimedia with a cloud-based platform named Hysia for easy development and deployment of V2R applications. The system consists of: 1) a back-end infrastructure providing optimized V2R related services including data engine, model repository, model serving and content matching; and 2) an application layer which enables rapid V2R application prototyping. Hysia addresses industry and academic needs in large-scale multimedia by: 1) seamlessly integrating state-of-the-art libraries including NVIDIA video SDK, Facebook faiss, and gRPC; 2) efficiently utilizing GPU computation; and 3) allowing developers to bind new models easily to meet the rapidly changing deep learning (DL) techniques. On top of that, we implement an orchestrator for further optimizing DL model serving performance. Hysia has been released as an open source project on GitHub, and attracted considerable attention. We have published Hysia to DockerHub as an official image for seamless integration and deployment in current cloud environments.

Huaizheng Zhang, Yuanming Li, Yizheng Huang et al.

MLModelCI provides multimedia researchers and developers with a one-stop platform for efficient machine learning (ML) services. The system leverages DevOps techniques to optimize, test, and manage models. It also containerizes and deploys these optimized and validated models as cloud services (MLaaS). In its essence, MLModelCI serves as a housekeeper to help users publish models. The models are first automatically converted to optimized formats for production purpose and then profiled under different settings (e.g., batch size and hardware). The profiling information can be used as guidelines for balancing the trade-off between performance and cost of MLaaS. Finally, the system dockerizes the models for ease of deployment to cloud environments. A key feature of MLModelCI is the implementation of a controller, which allows elastic evaluation which only utilizes idle workers while maintaining online service quality. Our system bridges the gap between current ML training and serving systems and thus free developers from manual and tedious work often associated with service deployment. We release the platform as an open-source project on GitHub under Apache 2.0 license, with the aim that it will facilitate and streamline more large-scale ML applications and research projects.

Huaizheng Zhang, Han Hu, Guanyu Gao et al.

Motivated by the prowess of deep learning (DL) based techniques in prediction, generalization, and representation learning, we develop a novel framework called DeepQoE to predict video quality of experience (QoE). The end-to-end framework first uses a combination of DL techniques (e.g., word embeddings) to extract generalized features. Next, these features are combined and fed into a neural network for representation learning. Such representations serve as inputs for classification or regression tasks. Evaluating the performance of DeepQoE with two datasets, we show that for the small dataset, the accuracy of all shallow learning algorithm is improved by using the representation derived from DeepQoE. For the large dataset, our DeepQoE framework achieves significant performance improvement in comparison to the best baseline method (90.94% vs. 82.84%). Moreover, DeepQoE, also released as an open source tool, provides video QoE research much-needed flexibility in fitting different datasets, extracting generalized features, and learning representations.

Yizheng Huang, Huaizheng Zhang, Yonggang Wen et al.

MLOps is about taking experimental ML models to production, i.e., serving the models to actual users. Unfortunately, existing ML serving systems do not adequately handle the dynamic environments in which online data diverges from offline training data, resulting in tedious model updating and deployment works. This paper implements a lightweight MLOps plugin, termed ModelCI-e (continuous integration and evolution), to address the issue. Specifically, it embraces continual learning (CL) and ML deployment techniques, providing end-to-end supports for model updating and validation without serving engine customization. ModelCI-e includes 1) a model factory that allows CL researchers to prototype and benchmark CL models with ease, 2) a CL backend to automate and orchestrate the model updating efficiently, and 3) a web interface for an ML team to manage CL service collaboratively. Our preliminary results demonstrate the usability of ModelCI-e, and indicate that eliminating the interference between model updating and inference workloads is crucial for higher system efficiency.

Yuanming Li, Huaizheng Zhang, Shanshan Jiang et al.

AI engineering has emerged as a crucial discipline to democratize deep neural network (DNN) models among software developers with a diverse background. In particular, altering these DNN models in the deployment stage posits a tremendous challenge. In this research, we propose and develop a low-code solution, ModelPS (an acronym for "Model Photoshop"), to enable and empower collaborative DNN model editing and intelligent model serving. The ModelPS solution embodies two transformative features: 1) a user-friendly web interface for a developer team to share and edit DNN models pictorially, in a low-code fashion, and 2) a model genie engine in the backend to aid developers in customizing model editing configurations for given deployment requirements or constraints. Our case studies with a wide range of deep learning (DL) models show that the system can tremendously reduce both development and communication overheads with improved productivity.

Huaizheng Zhang, Meng Shen, Yizheng Huang et al.

DNN-based video analytics have empowered many new applications (e.g., automated retail). Meanwhile, the proliferation of fog devices provides developers with more design options to improve performance and save cost. To the best of our knowledge, this paper presents the first serverless system that takes full advantage of the client-fog-cloud synergy to better serve the DNN-based video analytics. Specifically, the system aims to achieve two goals: 1) Provide the optimal analytics results under the constraints of lower bandwidth usage and shorter round-trip time (RTT) by judiciously managing the computational and bandwidth resources deployed in the client, fog, and cloud environment. 2) Free developers from tedious administration and operation tasks, including DNN deployment, cloud and fog's resource management. To this end, we implement a holistic cloud-fog system referred to as VPaaS (Video-Platform-as-a-Service). VPaaS adopts serverless computing to enable developers to build a video analytics pipeline by simply programming a set of functions (e.g., model inference), which are then orchestrated to process videos through carefully designed modules. To save bandwidth and reduce RTT, VPaaS provides a new video streaming protocol that only sends low-quality video to the cloud. The state-of-the-art (SOTA) DNNs deployed at the cloud can identify regions of video frames that need further processing at the fog ends. At the fog ends, misidentified labels in these regions can be corrected using a light-weight DNN model. To address the data drift issues, we incorporate limited human feedback into the system to verify the results and adopt incremental learning to improve our system continuously. The evaluation demonstrates that VPaaS is superior to several SOTA systems: it maintains high accuracy while reducing bandwidth usage by up to 21%, RTT by up to 62.5%, and cloud monetary cost by up to 50%.

Huaizheng Zhang, Yizheng Huang, Yonggang Wen et al.

Deep learning (DL) models have become core modules for many applications. However, deploying these models without careful performance benchmarking that considers both hardware and software's impact often leads to poor service and costly operational expenditure. To facilitate DL models' deployment, we implement an automatic and comprehensive benchmark system for DL developers. To accomplish benchmark-related tasks, the developers only need to prepare a configuration file consisting of a few lines of code. Our system, deployed to a leader server in DL clusters, will dispatch users' benchmark jobs to follower workers. Next, the corresponding requests, workload, and even models can be generated automatically by the system to conduct DL serving benchmarks. Finally, developers can leverage many analysis tools and models in our system to gain insights into the trade-offs of different system configurations. In addition, a two-tier scheduler is incorporated to avoid unnecessary interference and improve average job compilation time by up to 1.43x (equivalent of 30\% reduction). Our system design follows the best practice in DL clusters operations to expedite day-to-day DL service evaluation efforts by the developers. We conduct many benchmark experiments to provide in-depth and comprehensive evaluations. We believe these results are of great values as guidelines for DL service configuration and resource allocation.

Huaizheng Zhang, Yong Luo, Qiming Ai et al.

Given the massive market of advertising and the sharply increasing online multimedia content (such as videos), it is now fashionable to promote advertisements (ads) together with the multimedia content. It is exhausted to find relevant ads to match the provided content manually, and hence, some automatic advertising techniques are developed. Since ads are usually hard to understand only according to its visual appearance due to the contained visual metaphor, some other modalities, such as the contained texts, should be exploited for understanding. To further improve user experience, it is necessary to understand both the topic and sentiment of the ads. This motivates us to develop a novel deep multimodal multitask framework to integrate multiple modalities to achieve effective topic and sentiment prediction simultaneously for ads understanding. In particular, our model first extracts multimodal information from ads and learn high-level and comparable representations. The visual metaphor of the ad is decoded in an unsupervised manner. The obtained representations are then fed into the proposed hierarchical multimodal attention modules to learn task-specific representations for final prediction. A multitask loss function is also designed to train both the topic and sentiment prediction models jointly in an end-to-end manner. We conduct extensive experiments on the latest and large advertisement dataset and achieve state-of-the-art performance for both prediction tasks. The obtained results could be utilized as a benchmark for ads understanding.

Guanyu Gao, Linsen Dong, Huaizheng Zhang et al.

Adaptive bitrate (ABR) streaming is the de facto solution for achieving smooth viewing experiences under unstable network conditions. However, most of the existing rate adaptation approaches for ABR are content-agnostic, without considering the semantic information of the video content. Nevertheless, semantic information largely determines the informativeness and interestingness of the video content, and consequently affects the QoE for video streaming. One common case is that the user may expect higher quality for the parts of video content that are more interesting or informative so as to reduce video distortion and information loss, given that the overall bitrate budgets are limited. This creates two main challenges for such a problem: First, how to determine which parts of the video content are more interesting? Second, how to allocate bitrate budgets for different parts of the video content with different significances? To address these challenges, we propose a Content-of-Interest (CoI) based rate adaptation scheme for ABR. We first design a deep learning approach for recognizing the interestingness of the video content, and then design a Deep Q-Network (DQN) approach for rate adaptation by incorporating video interestingness information. The experimental results show that our method can recognize video interestingness precisely, and the bitrate allocation for ABR can be aligned with the interestingness of video content while not compromising the performances on objective QoE metrics.

Yong Luo, Huaizheng Zhang, Yongjie Wang et al.

Recruitment of appropriate people for certain positions is critical for any companies or organizations. Manually screening to select appropriate candidates from large amounts of resumes can be exhausted and time-consuming. However, there is no public tool that can be directly used for automatic resume quality assessment (RQA). This motivates us to develop a method for automatic RQA. Since there is also no public dataset for model training and evaluation, we build a dataset for RQA by collecting around 10K resumes, which are provided by a private resume management company. By investigating the dataset, we identify some factors or features that could be useful to discriminate good resumes from bad ones, e.g., the consistency between different parts of a resume. Then a neural-network model is designed to predict the quality of each resume, where some text processing techniques are incorporated. To deal with the label deficiency issue in the dataset, we propose several variants of the model by either utilizing the pair/triplet-based loss, or introducing some semi-supervised learning technique to make use of the abundant unlabeled data. Both the presented baseline model and its variants are general and easy to implement. Various popular criteria including the receiver operating characteristic (ROC) curve, F-measure and ranking-based average precision (AP) are adopted for model evaluation. We compare the different variants with our baseline model. Since there is no public algorithm for RQA, we further compare our results with those obtained from a website that can score a resume. Experimental results in terms of different criteria demonstrate the effectiveness of the proposed method. We foresee that our approach would transform the way of future human resources management.