Search

In China, there has long been controversy over the “Poetic Translation” and “Prose Translation. Those who insist on the “Poetic translation” style think that Shakespeare’s plays are poetic, so the Chinese translations should adopt the form of “poem”, as a result of which the translations can approach the original works and the “Real Shakespeare” in terms of language form and style, etc. while the “Prose Translation” fails to achieve this. The “Poetic Translation” sprang from the transplantation of “Blank Verse” starting with the distinguished translator of Sun Dayu, followed by Lin Tongji, Bian Zhilin and Fang Ping etc. What is the effect of the“Poetic Translation” Does it pursue “form” or resemble in appearance while losing the real expressions. The paper aims at searching for it.

The high computational and memory requirements of generative large language models (LLMs) make it challenging to serve them quickly and cheaply. This paper introduces SpecInfer, an LLM serving system that accelerates generative LLM inference with speculative inference and token tree verification. A key insight behind SpecInfer is to combine various collectively boost-tuned small language models to jointly predict the LLM's outputs; the predictions are organized as a token tree, whose nodes each represent a candidate token sequence. The correctness of all candidate token sequences represented by a token tree is verified by the LLM in parallel using a novel tree-based parallel decoding mechanism. SpecInfer uses an LLM as a token tree verifier instead of an incremental decoder, which significantly reduces the end-to-end latency and computational requirement for serving generative LLMs while provably preserving model quality.

The strong demand for efficient and performant deployment of Deep Learning (DL) applications prompts the rapid development of a rich DL ecosystem. To keep up with this fast advancement, it is crucial for modern DL frameworks to efficiently integrate a variety of optimized tensor algebra libraries and runtimes as their backends and generate the fastest possible executable using these backends. However, current DL frameworks require significant manual effort and expertise to integrate every new backend while failing to unleash its full potential. Given the fast-evolving nature of the DL ecosystem, this manual approach often slows down continuous innovations across different layers; it prevents hardware vendors from the fast deployment of their cutting-edge libraries, DL framework developers must repeatedly adjust their hand-coded rules to accommodate new versions of libraries, and machine learning practitioners need to wait for the integration of new technologies and often encounter unsatisfactory performance. In this paper, we propose Collage, a DL framework that offers seamless integration of DL backends. Collage provides an expressive backend registration interface that allows users to precisely specify the capability of various backends. By leveraging the specifications of available backends, Collage automatically searches for an optimized backend placement strategy for a given workload and execution environment. Our evaluation shows that Collage outperforms the best existing framework for each hardware by $1.26\times$, $1.43\times$, $1.40\times$ on average on NVIDIA's RTX 2070 GPU, V100 GPU, and Intel's Xeon 8259CL CPU, respectively. Collage has been open-sourced and deployed in Apache TVM., Comment: Published in PACT 22

Boosting the runtime performance of deep neural networks (DNNs) is critical due to their wide adoption in real-world tasks. Existing approaches to optimizing the tensor algebra expression of a DNN only consider expressions representable by a fixed set of predefined operators, missing possible optimization opportunities between general expressions. We propose OLLIE, the first derivation-based tensor program optimizer. OLLIE optimizes tensor programs by leveraging transformations between general tensor algebra expressions, enabling a significantly larger expression search space that includes those supported by prior work as special cases. OLLIE uses a hybrid derivation-based optimizer that effectively combines explorative and guided derivations to quickly discover highly optimized expressions. Evaluation on seven DNNs shows that OLLIE can outperform existing optimizers by up to 2.73$\times$ (1.46$\times$ on average) on an A100 GPU and up to 2.68$\times$ (1.51$\times$) on a V100 GPU, respectively.

As more practical and scalable quantum computers emerge, much attention has been focused on realizing quantum supremacy in machine learning. Existing quantum ML methods either (1) embed a classical model into a target Hamiltonian to enable quantum optimization or (2) represent a quantum model using variational quantum circuits and apply classical gradient-based optimization. The former method leverages the power of quantum optimization but only supports simple ML models, while the latter provides flexibility in model design but relies on gradient calculation, resulting in barren plateau (i.e., gradient vanishing) and frequent classical-quantum interactions. To address the limitations of existing quantum ML methods, we introduce Quark, a gradient-free quantum learning framework that optimizes quantum ML models using quantum optimization. Quark does not rely on gradient computation and therefore avoids barren plateau and frequent classical-quantum interactions. In addition, Quark can support more general ML models than prior quantum ML methods and achieves a dataset-size-independent optimization complexity. Theoretically, we prove that Quark can outperform classical gradient-based methods by reducing model query complexity for highly non-convex problems; empirically, evaluations on the Edge Detection and Tiny-MNIST tasks show that Quark can support complex ML models and significantly reduce the number of measurements needed for discovering near-optimal weights for these tasks., Comment: under review

Existing quantum compilers optimize quantum circuits by applying circuit transformations designed by experts. This approach requires significant manual effort to design and implement circuit transformations for different quantum devices, which use different gate sets, and can miss optimizations that are hard to find manually. We propose Quartz, a quantum circuit superoptimizer that automatically generates and verifies circuit transformations for arbitrary quantum gate sets. For a given gate set, Quartz generates candidate circuit transformations by systematically exploring small circuits and verifies the discovered transformations using an automated theorem prover. To optimize a quantum circuit, Quartz uses a cost-based backtracking search that applies the verified transformations to the circuit. Our evaluation on three popular gate sets shows that Quartz can effectively generate and verify transformations for different gate sets. The generated transformations cover manually designed transformations used by existing optimizers and also include new transformations. Quartz is therefore able to optimize a broad range of circuits for diverse gate sets, outperforming or matching the performance of hand-tuned circuit optimizers., Comment: 28 pages. Extended version of the paper presented in PLDI 2022. Typos corrected and artifact reference updated

DNN models across many domains continue to grow in size, resulting in high resource requirements for effective training, and unpalatable (and often unaffordable) costs for organizations and research labs across scales. This paper aims to significantly reduce training costs with effective use of preemptible instances, i.e., those that can be obtained at a much cheaper price while idle, but may be preempted whenever requested by priority users. Doing so, however, requires new forms of resiliency and efficiency to cope with the possibility of frequent preemptions - a failure model that is drastically different from the occasional failures in normal cluster settings that existing checkpointing techniques target. We present Bamboo, a distributed system that tackles these challenges by introducing redundant computations into the training pipeline, i.e., whereby one node performs computations over not only its own layers but also over some layers in its neighbor. Our key insight is that training large models often requires pipeline parallelism where "pipeline bubbles" naturally exist. Bamboo carefully fills redundant computations into these bubbles, providing resilience at a low cost. Across a variety of widely used DNN models, Bamboo outperforms traditional checkpointing by 3.7x in training throughput, and reduces costs by 2.4x compared to a setting where on-demand instances are used., Comment: 17 pages

Considering the dynamic changes of unmanned surface vehicle (USV) in berthing tasks, the planning and control modes are divided into two phases: the remote phase and the terminal phase. According to the main influencing factors of the two phases, an improved artificial potential field method is proposed to complete autonomous berthing trajectory planning based on the analysis of environment constraint, berth point constraint and USV’s dynamics constraint. Combining with the dynamic characteristics and control objectives at different phases of berthing and analyzing the fuzzy rule regulation strategy of USV’s heading and speed control, an improved adaptive fuzzy PID control method is proposed to solve the control problem of USV, which is influenced by weak maneuver, large disturbance, limited water area and strong shore effect. Finally, the comparative test of berthing simulation verifies the superiority of the proposed control method. The autonomous berthing field experiment is completed based on the "Dolphin-I" small USV. It verifies the validity and feasibility of the proposed autonomous berthing method.

Existing quantum compilers focus on mapping a logical quantum circuit to a quantum device and its native quantum gates. Only simple circuit identities are used to optimize the quantum circuit during the compilation process. This approach misses more complex circuit identities, which could be used to optimize the quantum circuit further. We propose Quanto, the first quantum optimizer that automatically generates circuit identities. Quanto takes as input a gate set and generates provably correct circuit identities for the gate set. Quanto's automatic generation of circuit identities includes single-qubit and two-qubit gates, which leads to a new database of circuit identities, some of which are novel to the best of our knowledge. In addition to the generation of new circuit identities, Quanto's optimizer applies such circuit identities to quantum circuits and finds optimized quantum circuits that have not been discovered by other quantum compilers, including IBM Qiskit and Cambridge Quantum Computing Tket. Quanto's database of circuit identities could be applied to improve existing quantum compilers and Quanto can be used to generate identity databases for new gate sets.

Outlier detection (OD) is a key learning task for finding rare and deviant data samples, with many time-critical applications such as fraud detection and intrusion detection. In this work, we propose TOD, the first tensor-based system for efficient and scalable outlier detection on distributed multi-GPU machines. A key idea behind TOD is decomposing complex OD applications into a small collection of basic tensor algebra operators. This decomposition enables TOD to accelerate OD computations by leveraging recent advances in deep learning infrastructure in both hardware and software. Moreover, to deploy memory-intensive OD applications on modern GPUs with limited on-device memory, we introduce two key techniques. First, provable quantization speeds up OD computations and reduces its memory footprint by automatically performing specific floating-point operations in lower precision while provably guaranteeing no accuracy loss. Second, to exploit the aggregated compute resources and memory capacity of multiple GPUs, we introduce automatic batching, which decomposes OD computations into small batches for both sequential execution on a single GPU and parallel execution on multiple GPUs. TOD supports a diverse set of OD algorithms. Extensive evaluation on 11 real and 3 synthetic OD datasets shows that TOD is on average 10.9x faster than the leading CPU-based OD system PyOD (with a maximum speedup of 38.9x), and can handle much larger datasets than existing GPU-based OD systems. In addition, TOD allows easy integration of new OD operators, enabling fast prototyping of emerging and yet-to-discovered OD algorithms., Comment: Code available at https://github.com/yzhao062/pytod

Graph Neural Networks (GNNs) are based on repeated aggregations of information across nodes' neighbors in a graph. However, because common neighbors are shared between different nodes, this leads to repeated and inefficient computations. We propose Hierarchically Aggregated computation Graphs (HAGs), a new GNN graph representation that explicitly avoids redundancy by managing intermediate aggregation results hierarchically, eliminating repeated computations and unnecessary data transfers in GNN training and inference. We introduce an accurate cost function to quantitatively evaluate the runtime performance of different HAGs and use a novel HAG search algorithm to find optimized HAGs. Experiments show that the HAG representation significantly outperforms the standard GNN graph representation by increasing the end-to-end training throughput by up to 2.8x and reducing the aggregations and data transfers in GNN training by up to 6.3x and 5.6x, while maintaining the original model accuracy., Comment: 12 pages, 4 figures

The past few years have witnessed growth in the computational requirements for training deep convolutional neural networks. Current approaches parallelize training onto multiple devices by applying a single parallelization strategy (e.g., data or model parallelism) to all layers in a network. Although easy to reason about, these approaches result in suboptimal runtime performance in large-scale distributed training, since different layers in a network may prefer different parallelization strategies. In this paper, we propose layer-wise parallelism that allows each layer in a network to use an individual parallelization strategy. We jointly optimize how each layer is parallelized by solving a graph search problem. Our evaluation shows that layer-wise parallelism outperforms state-of-the-art approaches by increasing training throughput, reducing communication costs, achieving better scalability to multiple GPUs, while maintaining original network accuracy.

Management system for energy (EMS) to be built as a rounded and effective management system of the document in the organization can be improved continuously through the energy saving monitoring, internal audit, management review etc. Currently that system still hasn’t been widely established. The problem is that the energy service organization doesn’t know how to help the enterprise to establish and run EMS. Through the concept and frame analysis of EMS, we propose a method how to establish and run EMS. The method is to identify all energy factors firstly, analysis the energy factor to be controlled prioritily, then draw up to set up the management project and make the system document. All the energy factors in the system's document are required to be able measurable. EMS is run by means of recording and investigating each energy factor in the energy using process, and carry out the purpose of the energy target.

We numerically investigate the broadband perfect infrared absorption by tuning epsilon-near-zero (ENZ) and epsilon-near-pole (ENP) resonances of multilayer indium tin oxide nanowires (ITO NWs). The monolayer ITO NWs array shows intensive absorption at ENZ and ENP wavelengths for p polarization, while only at the ENP wavelength for s polarization. Moreover, the ENP resonances are almost omnidirectional and the ENZ resonances are angularly dependent. Therefore, the absorption bandwidth is broader for p polarization than that for s polarization when polarized waves are incident obliquely. The ENZ resonances can be tuned by altering the doping concentration and volume filling factor of ITO NWs. However, the ENP resonances only can be tuned by changing the doping concentration of ITO NWs, and volume filling factor impacts little on the ENP resonances. Based on the strong absorption properties of each layer at their own ENP and ENZ resonances, the tuned absorption of the bilayer ITO NWs with the different doping concentrations can be broader and stronger. Furthermore, multilayer ITO NWs can achieve broadband perfect absorption by controlling the doping concentration, volume filling factor, and length of the NWs in each layer. This study has the potential to apply to applications requiring efficient absorption and energy conversion.

This paper proposes an assembly structure incorporating two-dimensional silicon gratings and germanium nanorods applied to photovoltaic absorbers. The absorptance of the assembly structure is numerically investigated using the finite-difference time-domain method. The results demonstrate that such a structure can greatly improve the absorptance and conversion efficiency compared to the gratings or nanowires in the 300-1100 nm wavelength region. The average spectral absorptance of such a structure reaches up to 0.983, even closes in to unity in some wave regions, which is mainly attributed to the guided mode resonance and Fabry-Perot resonance identified by analyzing the electromagnetic field and power dissipation. The effects of different diameters and lengths of the nanorod component, as well as the widths and depths of the grating component, on the absorptance are further examined. It is found that the absorptance of the assembly structure is insensitive to the incident angle of less than 30° for both TM and TE waves. The photovoltaic absorbers with such a structure can yield an ideal conversion efficiency as high as 47.9%, which shows great potential for applying the assembly structure to photovoltaic absorbers.