PyTorchKR Papers

최신 논문

총 3,273편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2503.01877공식 코드 · Papers with Code

Starjob: Dataset for LLM-Driven Job Shop Scheduling

Large Language Models (LLMs) have shown remarkable capabilities across various domains, but their potential for solving combinatorial optimization problems remains largely unexplored. In this paper, we investigate the applicability of LLMs to the Job Shop Scheduling Problem (JSSP), a classic challenge in combinatorial optimization that requires efficient job allocation to machines to minimize makespan. To this end, we introduce Starjob, the first supervised dataset for JSSP, comprising 130k instances specifically designed for training LLMs. Leveraging this dataset, we fine-tune the LLaMA 8B 4-bit quantized model with the LoRA method to develop an end-to-end scheduling approach. Our evaluation on standard benchmarks demonstrates that the proposed LLM-based method not only surpasses traditional Priority Dispatching Rules (PDRs) but also achieves notable improvements over state-of-the-art neural approaches like L2D, with an average improvement of 15.36% on DMU and 7.85% on Taillard benchmarks. These results highlight the untapped potential of LLMs in tackling combinatorial optimization problems, paving the way for future advancements in this area.

arXiv:2503.01324

MAB-Based Channel Scheduling for Asynchronous Federated Learning in Non-Stationary Environments

Federated learning enables distributed model training across clients under central coordination without raw data exchange. However, in wireless implementations, frequent parameter updates between the server and clients create significant communication overhead. While existing research assumes known channel state information (CSI) or stationary distributions, practical wireless channels exhibit non-stationary characteristics due to channel fading, user mobility, and hostile attacks. The unavailability of CSI and time-varying statistics can cause unpredictable transmission failures, exacerbating client staleness and affecting model convergence. To address these challenges, we propose an asynchronous federated learning scheduling framework for non-stationary channel environments to reduce staleness while promoting fair and efficient communication and aggregation.We focus on two channel scenarios: extremely non-stationary and piecewise stationary. Age of Information (AoI) quantifies client staleness under non-stationary conditions. Through a rigorous convergence analysis, we explore how AoI and per-round client participation affect learning performance. The scheduling problem is modeled within a multi-armed bandit (MAB) framework, and we derive the theoretical lower bounds on AoI regret. Based on these findings, we develop scheduling strategies for both scenarios using the GLR-CUCB and M-exp3 algorithms, also deriving their respective upper bounds on AoI regret. To address imbalanced client updates, we introduce an adaptive allocation strategy that incorporates marginal utility and fairness. Simulations demonstrate that our algorithm reduces AoI regret growth, accelerates federated learning convergence, and promotes fairer aggregation.

arXiv:2503.00537

Scalable Reinforcement Learning for Virtual Machine Scheduling

Recent advancements in reinforcement learning (RL) have shown promise for optimizing virtual machine scheduling (VMS) in small-scale clusters. The utilization of RL to large-scale cloud computing scenarios remains notably constrained. This paper introduces a scalable RL framework, called Cluster Value Decomposition Reinforcement Learning (CVD-RL), to surmount the scalability hurdles inherent in large-scale VMS. The CVD-RL framework innovatively combines a decomposition operator with a look-ahead operator to adeptly manage representation complexities, while complemented by a Top-$k$ filter operator that refines exploration efficiency. Different from existing approaches limited to clusters of $10$ or fewer physical machines (PMs), CVD-RL extends its applicability to environments encompassing up to $50$ PMs. Furthermore, the CVD-RL framework demonstrates generalization capabilities that surpass contemporary SOTA methodologies across a variety of scenarios in empirical studies. This breakthrough not only showcases the framework's exceptional scalability and performance but also represents a significant leap in the application of RL for VMS within complex, large-scale cloud infrastructures. The code is available at https://anonymous.4open.science/r/marl4sche-D0FE.

arXiv:2503.00569

Communication-Efficient Device Scheduling for Federated Learning Using Lyapunov Optimization

Federated learning (FL) is a useful tool that enables the training of machine learning models over distributed data without having to collect data centrally. When deploying FL in constrained wireless environments, however, intermittent connectivity of devices, heterogeneous connection quality, and non-i.i.d. data can severely slow convergence. In this paper, we consider FL with arbitrary device participation probabilities for each round and show that by weighing each device's update by the reciprocal of their per-round participation probability, we can guarantee convergence to a stationary point. Our bound applies to non-convex loss functions and non-i.i.d. datasets and recovers state-of-the-art convergence rates for both full and uniform partial participation, including linear speedup, with only a single-sided learning rate. Then, using the derived convergence bound, we develop a new online client selection and power allocation algorithm that utilizes the Lyapunov drift-plus-penalty framework to opportunistically minimize a function of the convergence bound and the average communication time under a transmit power constraint. We use optimization over manifold techniques to obtain a solution to the minimization problem. Thanks to the Lyapunov framework, one key feature of the algorithm is that knowledge of the channel distribution is not required and only the instantaneous channel state information needs to be known. Using the CIFAR-10 dataset with varying levels of data heterogeneity, we show through simulations that the communication time can be significantly decreased using our algorithm compared to uniformly random participation, especially for heterogeneous channel conditions.

arXiv:2503.01325

Towards net-zero manufacturing: carbon-aware scheduling for GHG emissions reduction

Detailed scheduling has traditionally been optimized for the reduction of makespan and manufacturing costs. However, growing awareness of environmental concerns and increasingly stringent regulations are pushing manufacturing towards reducing the carbon footprint of its operations. Scope 2 emissions, which are the indirect emissions related to the production and consumption of grid electricity, are in fact estimated to be responsible for more than one-third of the global GHG emissions. In this context, carbon-aware scheduling can serve as a powerful way to reduce manufacturing's carbon footprint by considering the time-dependent carbon intensity of the grid and the availability of on-site renewable electricity. This study introduces a carbon-aware permutation flow-shop scheduling model designed to reduce scope 2 emissions. The model is formulated as a mixed-integer linear problem, taking into account the forecasted grid generation mix and available on-site renewable electricity, along with the set of jobs to be scheduled and their corresponding power requirements. The objective is to find an optimal day-ahead schedule that minimizes scope 2 emissions. The problem is addressed using a dedicated memetic algorithm, combining evolutionary strategy and local search. Results from computational experiments confirm that by considering the dynamic carbon intensity of the grid and on-site renewable electricity availability, substantial reductions in carbon emissions can be achieved.

arXiv:2503.00588

An Improved NSGA-II with local search for multi-objective energy-efficient flowshop scheduling problem

There has been an increasing concern to reduce the energy consumption in manufacturing and other industries. Energy consumption in manufacturing industries is directly related to efficient schedules. The contribution of this paper includes: i) a permutation flowshop scheduling problem (PFLSP) mathematical model by considering energy consumed by each machine in the system. ii) an improved non-dominated sorted genetic algorithm with Taguchi method with further incorporating local search (NSGA-II_LS) is proposed for the multi-objective PFLSP model. iii) solved 90 benchmarks problems of Taillard (1993) for the minimisation of flowtime (FT) and energy consumption (EC). The performance of the proposed NSGA_LS algorithm is evaluated on the benchmark problems selected from the published literature Li et. al, (2018). From these results, it is noted that the proposed algorithm performed better on both the objectives i.e., FT and EC minimization in 5 out of 9 cases. On FT objective our algorithm performed better in 8 out of 9 cases and on EC objective 5 out of 9 cases. Overall, the proposed algorithm achieved 47% and 15.44% average improvement in FT and EC minimization respectively on the benchmark problems. From the results of 90 benchmark problems, it is observed that average difference in FT and EC between two solutions is decreasing as the problem size increases from 5 machines to 10 machines with an exception in one case. Further, it is observed that the performance of the proposed algorithm is better as the problem size increases in both jobs and machines. These results can act as standard solutions for further research.

arXiv:2503.00323공식 코드 · Papers with Code

FLStore: Efficient Federated Learning Storage for non-training workloads

Federated Learning (FL) is an approach for privacy-preserving Machine Learning (ML), enabling model training across multiple clients without centralized data collection. With an aggregator server coordinating training, aggregating model updates, and storing metadata across rounds. In addition to training, a substantial part of FL systems are the non-training workloads such as scheduling, personalization, clustering, debugging, and incentivization. Most existing systems rely on the aggregator to handle non-training workloads and use cloud services for data storage. This results in high latency and increased costs as non-training workloads rely on large volumes of metadata, including weight parameters from client updates, hyperparameters, and aggregated updates across rounds, making the situation even worse. We propose FLStore, a serverless framework for efficient FL non-training workloads and storage. FLStore unifies the data and compute planes on a serverless cache, enabling locality-aware execution via tailored caching policies to reduce latency and costs. Per our evaluations, compared to cloud object store based aggregator server FLStore reduces per request average latency by 71% and costs by 92.45%, with peak improvements of 99.7% and 98.8%, respectively. Compared to an in-memory cloud cache based aggregator server, FLStore reduces average latency by 64.6% and costs by 98.83%, with peak improvements of 98.8% and 99.6%, respectively. FLStore integrates seamlessly with existing FL frameworks with minimal modifications, while also being fault-tolerant and highly scalable.

arXiv:2502.20576공식 코드 · Papers with Code

OmniRouter: Budget and Performance Controllable Multi-LLM Routing

Large language models (LLMs) deliver superior performance but require substantial computational resources and operate with relatively low efficiency, while smaller models can efficiently handle simpler tasks with fewer resources. LLM routing is a crucial paradigm that dynamically selects the most suitable large language models from a pool of candidates to process diverse inputs, ensuring optimal resource utilization while maintaining response quality. Existing routing frameworks typically model this as a locally optimal decision-making problem, selecting the presumed best-fit LLM for each query individually, which overlook global budget constraints, resulting in ineffective resource allocation. To tackle this problem, we introduce OmniRouter, a fundamentally controllable routing framework for multi-LLM serving. Instead of making per-query greedy choices, OmniRouter models the routing task as a constrained optimization problem, assigning models that minimize total cost while ensuring the required performance level. Specifically, a hybrid retrieval-augmented predictor is designed to predict the capabilities and costs of LLMs and a constrained optimizer is employed to control globally optimal query-model allocation. Experiments show that OmniRouter achieves up to 6.30% improvement in response accuracy while simultaneously reducing computational costs by at least 10.15% compared to competitive router baselines. The code and the dataset are available at https://github.com/agiresearch/OmniRouter.

arXiv:2502.19913공식 코드 · Papers with Code

SkipPipe: Partial and Reordered Pipelining Framework for Training LLMs in Heterogeneous Networks

Data and pipeline parallelism are ubiquitous for training of Large Language Models (LLM) on distributed nodes. Driven by the need for cost-effective training, recent work explores efficient communication arrangement for end to end training. Motivated by LLM's resistance to layer skipping and layer reordering, in this paper, we explore stage (several consecutive layers) skipping in pipeline training, and challenge the conventional practice of sequential pipeline execution. We derive convergence and throughput constraints (guidelines) for pipelining with skipping and swapping pipeline stages. Based on these constraints, we propose SkipPipe, the first partial pipeline framework to reduce the end-to-end training time for LLMs while preserving the convergence. The core of SkipPipe is a path scheduling algorithm that optimizes the paths for individual microbatches and reduces idle time (due to microbatch collisions) on the distributed nodes, complying with the given stage skipping ratio. We extensively evaluate SkipPipe on LLaMa models from 500M to 8B parameters on up to 20 nodes. Our results show that SkipPipe reduces training iteration time by up to $55\%$ compared to full pipeline. Our partial pipeline training also improves resistance to layer omission during inference, experiencing a drop in perplexity of only $7\%$ when running only half the model. Our code is available at https://github.com/gensyn-ai/skippipe.

arXiv:2502.18773

Research on Edge Computing and Cloud Collaborative Resource Scheduling Optimization Based on Deep Reinforcement Learning

This study addresses the challenge of resource scheduling optimization in edge-cloud collaborative computing using deep reinforcement learning (DRL). The proposed DRL-based approach improves task processing efficiency, reduces overall processing time, enhances resource utilization, and effectively controls task migrations. Experimental results demonstrate the superiority of DRL over traditional scheduling algorithms, particularly in managing complex task allocation, dynamic workloads, and multiple resource constraints. Despite its advantages, further improvements are needed to enhance learning efficiency, reduce training time, and address convergence issues. Future research should focus on increasing the algorithm's fault tolerance to handle more complex and uncertain scheduling scenarios, thereby advancing the intelligence and efficiency of edge-cloud computing systems.

arXiv:2502.17826

Feedback-Free Resource Scheduling for Flexible Multi-BS Cooperation in FD-RAN

Flexible cooperation among base stations (BSs) is critical to improve resource utilization efficiency and meet personalized user demands. However, its practical implementation is hindered by the current radio access network (RAN), which relies on the coupling of uplink and downlink transmissions and channel state information feedback with inherent issues such as overheads and delays. To overcome these limitations, we consider the fully-decoupled RAN (FD-RAN), in which uplink and downlink functionalities are independent, and feedback-free MIMO transmission is adopted at the physical layer. To further deliver flexible cooperation in FD-RAN, we study feedback-free downlink multi-BS resource scheduling under the practical scheduling process. The problem is considered based on network load conditions. In heavy-load states where it is impossible for all user demands to be met, an optimal greedy algorithm is proposed, maximizing the weighted sum of user demand satisfaction rates. In light-load states where at least one solution exists to satisfy all user demands, an optimal two-stage resource allocation algorithm is designed to further minimize network energy consumption by leveraging the flexibility of cooperation. Extensive simulations validate the superiority of proposed algorithms in performance and running time, and highlight the potential for realizing flexible cooperation in practice.

arXiv:2502.17408

Joint Holographic Beamforming and User Scheduling with Individual QoS Constraints

Reconfigurable holographic surfaces (RHS) have emerged as a transformative material technology, enabling dynamic control of electromagnetic waves to generate versatile holographic beam patterns. This paper addresses the problem of joint hybrid holographic beamforming and user scheduling under per-user minimum quality-of-service (QoS) constraints, a critical challenge in resource-constrained networks. However, such a problem results in mixed-integer non-convex optimization, making it difficult to identify feasible solutions efficiently. To overcome this challenge, we propose a novel iterative optimization framework that jointly solves the problem to maximize the RHS-assisted network sum-rate, efficiently managing holographic beamforming patterns, dynamically scheduling users, and ensuring the minimum QoS requirements for each scheduled user. The proposed framework relies on zero-forcing digital beamforming, gradient-ascent-based holographic beamformer optimization, and a greedy user selection principle. Our extensive simulation results validate the effectiveness of the proposed scheme, demonstrating their superior performance compared to the benchmark algorithms in terms of sum-rate performance, while meeting the minimum per-user QoS constraints

arXiv:2502.17013

Latency-Aware Resource Allocation for Integrated Communications, Computation, and Sensing in Cell-Free mMIMO Systems

In this paper, we investigate a cell-free massive multiple-input and multiple-output (MIMO)-enabled integration communication, computation, and sensing (ICCS) system, aiming to minimize the maximum computation latency to guarantee the stringent sensing requirements. We consider a two-tier offloading framework, where each multi-antenna terminal can optionally offload its local tasks to either multiple mobile-edge servers for distributed computation or the cloud server for centralized computation while satisfying the sensing requirements and power constraint. The above offloading problem is formulated as a mixed-integer programming and non-convex problem, which can be decomposed into three sub-problems, namely, distributed offloading decision, beamforming design, and execution scheduling mechanism. First, the continuous relaxation and penalty-based techniques are applied to tackle the distributed offloading strategy. Then, the weighted minimum mean square error (WMMSE) and successive convex approximation (SCA)-based lower bound are utilized to design the integrated communication and sensing (ISAC) beamforming. Finally, the other resources can be judiciously scheduled to minimize the maximum latency. A rigorous convergence analysis and numerical results substantiate the effectiveness of our method. Furthermore, simulation results demonstrate that multi-point cooperation in cell-free massive MIMO-enabled ICCS significantly reduces overall computation latency, in comparison to the benchmark schemes.

arXiv:2502.17366

Distributed Coordination for Heterogeneous Non-Terrestrial Networks

To guarantee global coverage and ubiquitous connectivity, the Non-terrestrial Network (NTN) technology has been regarded as a key enabling technology in the Six Generation (6G) network, which consists of the unmanned aerial vehicle (UAV), high-altitude platform (HAP), and satellite. It is noted that the unique characteristics of various NTN platforms directly impact the design and implementation of NTNs, which results in highly dynamic and heterogeneous networks. Even within the same tier, such as the space tier, the NTNs are developed based on different platforms including Low Earth Orbit (LEO), Medium Earth Orbit (MEO), and Geostationary Earth Orbit (GEO). Therefore, distributed coordination among heterogeneous NTNs remains an important challenge. Although distributed learning framework finds a wide range of applications by leveraging rich distributed data and computation resources. The explicit and systematic analysis of the individual layers' challenges, and corresponding distributed coordination solutions in heterogeneous NTNs has not been proposed yet. In this article, we first summarize the unique characteristics of each NTN platform, and analyze the corresponding impact on the design and implementation of the NTN. We then identify the communication challenges of heterogeneous NTNs in individual layers, where the potential coordinated solutions are identified. We further illustrate the multi-agent deep reinforcement learning (MADRL) algorithms tailored for coordinated solutions in heterogeneous NTNs. Last but not least, we present a case study of the user scheduling optimization problem in heterogeneous UAVs-based cellular networks, where the multi-agent deep deterministic policy gradient (MADDPG) technique is developed to validate the effectiveness of distributed coordination in heterogeneous NTNs.

arXiv:2502.15018

Using tournaments to calculate AUROC for zero-shot classification with LLMs

Large language models perform surprisingly well on many zero-shot classification tasks, but are difficult to fairly compare to supervised classifiers due to the lack of a modifiable decision boundary. In this work, we propose and evaluate a method that converts binary classification tasks into pairwise comparison tasks, obtaining relative rankings from LLMs. Repeated pairwise comparisons can be used to score instances using the Elo rating system (used in chess and other competitions), inducing a confidence ordering over instances in a dataset. We evaluate scheduling algorithms for their ability to minimize comparisons, and show that our proposed algorithm leads to improved classification performance, while also providing more information than traditional zero-shot classification.

arXiv:2502.17260

Robust Federated Learning in Unreliable Wireless Networks: A Client Selection Approach

Federated learning (FL) has emerged as a promising distributed learning paradigm for training deep neural networks (DNNs) at the wireless edge, but its performance can be severely hindered by unreliable wireless transmission and inherent data heterogeneity among clients. Existing solutions primarily address these challenges by incorporating wireless resource optimization strategies, often focusing on uplink resource allocation across clients under the assumption of homogeneous client-server network standards. However, these approaches overlooked the fact that mobile clients may connect to the server via diverse network standards (e.g., 4G, 5G, Wi-Fi) with customized configurations, limiting the flexibility of server-side modifications and restricting applicability in real-world commercial networks. This paper presents a novel theoretical analysis about how transmission failures in unreliable networks distort the effective label distributions of local samples, causing deviations from the global data distribution and introducing convergence bias in FL. Our analysis reveals that a carefully designed client selection strategy can mitigate biases induced by network unreliability and data heterogeneity. Motivated by this insight, we propose FedCote, a client selection approach that optimizes client selection probabilities without relying on wireless resource scheduling. Experimental results demonstrate the robustness of FedCote in DNN-based classification tasks under unreliable networks with frequent transmission failures.

arXiv:2502.16239

Separated Contrastive Learning for Matching in Cross-domain Recommendation with Curriculum Scheduling

Cross-domain recommendation (CDR) is a task that aims to improve the recommendation performance in a target domain by leveraging the information from source domains. Contrastive learning methods have been widely adopted among intra-domain (intra-CL) and inter-domain (inter-CL) users/items for their representation learning and knowledge transfer during the matching stage of CDR. However, we observe that directly employing contrastive learning on mixed-up intra-CL and inter-CL tasks ignores the difficulty of learning from inter-domain over learning from intra-domain, and thus could cause severe training instability. Therefore, this instability deteriorates the representation learning process and hurts the quality of generated embeddings. To this end, we propose a novel framework named SCCDR built up on a separated intra-CL and inter-CL paradigm and a stop-gradient operation to handle the drawback. Specifically, SCCDR comprises two specialized curriculum stages: intra-inter separation and inter-domain curriculum scheduling. The former stage explicitly uses two distinct contrastive views for the intra-CL task in the source and target domains, respectively. Meanwhile, the latter stage deliberately tackles the inter-CL tasks with a curriculum scheduling strategy that derives effective curricula by accounting for the difficulty of negative samples anchored by overlapping users. Empirical experiments on various open-source datasets and an offline proprietary industrial dataset extracted from a real-world recommender system, and an online A/B test verify that SCCDR achieves state-of-the-art performance over multiple baselines.

arXiv:2502.16796

MobileSteward: Integrating Multiple App-Oriented Agents with Self-Evolution to Automate Cross-App Instructions

Mobile phone agents can assist people in automating daily tasks on their phones, which have emerged as a pivotal research spotlight. However, existing procedure-oriented agents struggle with cross-app instructions, due to the following challenges: (1) complex task relationships, (2) diverse app environment, and (3) error propagation and information loss in multi-step execution. Drawing inspiration from object-oriented programming principles, we recognize that object-oriented solutions is more suitable for cross-app instruction. To address these challenges, we propose a self-evolving multi-agent framework named MobileSteward, which integrates multiple app-oriented StaffAgents coordinated by a centralized StewardAgent. We design three specialized modules in MobileSteward: (1) Dynamic Recruitment generates a scheduling graph guided by information flow to explicitly associate tasks among apps. (2) Assigned Execution assigns the task to app-oriented StaffAgents, each equipped with app-specialized expertise to address the diversity between apps. (3) Adjusted Evaluation conducts evaluation to provide reflection tips or deliver key information, which alleviates error propagation and information loss during multi-step execution. To continuously improve the performance of MobileSteward, we develop a Memory-based Self-evolution mechanism, which summarizes the experience from successful execution, to improve the performance of MobileSteward. We establish the first English Cross-APP Benchmark (CAPBench) in the real-world environment to evaluate the agents' capabilities of solving complex cross-app instructions. Experimental results demonstrate that MobileSteward achieves the best performance compared to both single-agent and multi-agent frameworks, highlighting the superiority of MobileSteward in better handling user instructions with diverse complexity.

arXiv:2502.15936

Space-O-RAN: Enabling Intelligent, Open, and Interoperable Non Terrestrial Networks in 6G

Satellite networks are rapidly evolving, yet most \glspl{ntn} remain isolated from terrestrial orchestration frameworks. Their control architectures are typically monolithic and static, limiting their adaptability to dynamic traffic, topology changes, and mission requirements. These constraints lead to inefficient spectrum use and underutilized network capacity. Although \gls{ai} promises automation, its deployment in orbit is limited by computing, energy, and connectivity limitations. This paper introduces Space-O-RAN, a distributed control architecture that extends Open RAN principles into satellite constellations through hierarchical, closed-loop control. Lightweight \glspl{dapp} operate onboard satellites, enabling real-time functions like scheduling and beam steering without relying on persistent ground access. Cluster-level coordination is managed via \glspl{spaceric}, which leverage low-latency \glspl{isl} for autonomous decisions in orbit. Strategic tasks, including AI training and policy updates, are transferred to terrestrial platforms \glspl{smo} using digital twins and feeder links. A key enabler is the dynamic mapping of the O-RAN interfaces to satellite links, supporting adaptive signaling under varying conditions. Simulations using the Starlink topology validate the latency bounds that inform this architectural split, demonstrating both feasibility and scalability for autonomous satellite RAN operations.

arXiv:2502.15791공식 코드 · Papers with Code

Learning-Guided Rolling Horizon Optimization for Long-Horizon Flexible Job-Shop Scheduling

Long-horizon combinatorial optimization problems (COPs), such as the Flexible Job-Shop Scheduling Problem (FJSP), often involve complex, interdependent decisions over extended time frames, posing significant challenges for existing solvers. While Rolling Horizon Optimization (RHO) addresses this by decomposing problems into overlapping shorter-horizon subproblems, such overlap often involves redundant computations. In this paper, we present L-RHO, the first learning-guided RHO framework for COPs. L-RHO employs a neural network to intelligently fix variables that in hindsight did not need to be re-optimized, resulting in smaller and thus easier-to-solve subproblems. For FJSP, this means identifying operations with unchanged machine assignments between consecutive subproblems. Applied to FJSP, L-RHO accelerates RHO by up to 54% while significantly improving solution quality, outperforming other heuristic and learning-based baselines. We also provide in-depth discussions and verify the desirable adaptability and generalization of L-RHO across numerous FJSP variates, distributions, online scenarios and benchmark instances. Moreover, we provide a theoretical analysis to elucidate the conditions under which learning is beneficial.