We propose an adapter based multi-domain Transformer based language model (LM) for Transformer ASR. The model consists of a big size common LM and small size adapters. The model can perform multi-domain adaptation with only the small size adapters and its related layers. The proposed model can reuse the full fine-tuned LM which is fine-tuned using all layers of an original model. The proposed LM can be expanded to new domains by adding about 2% of parameters for a first domain and 13% parameters for after second domain. The proposed model is also effective in reducing the model maintenance cost because it is possible to omit the costly and time-consuming common LM pre-training process. Using proposed adapter based approach, we observed that a general LM with adapter can outperform a dedicated music domain LM in terms of word error rate (WER).
In recent years, many efforts have been made to complete knowledge graphs (KGs) by various graph embedding methods, most of which only focus on static KGs (SKGs) without considering the time dependency of facts. However, KGs in reality are dynamic and there exists correlations between facts with different timestamps. Due to the sparsity of temporal KGs (TKGs), SKG embedding methods cannot be directly applied to TKGs. And existing methods of TKG embedding suffer from two issues: (1) they follow the pattern of SKG embedding where all facts need to be retrained when a new timestamp appears; (2) they don't provide a general way to transplant SKG embedding methods to TKGs and therefore lack extensibility. In this paper, we propose a novel Recursive Temporal Fact Embedding Framework (RTFE) to transplant translation-based or graph neural network-based SKG embedding methods to TKGs. In the recursive way, timestamp parameters provide a good starting point for the next future timestamp. And existing SKG embedding models can be used as components. Experiments on TKGs show that our proposed framework (1) outperforms the state-of-the-art baseline model in the entity prediction task on fact datasets; (2) achieves similar performance compared with the state-of-the-art baseline model in relation prediction task on fact datasets; and (3) shows performance in the entity prediction task on event datasets.
Overparameterization has been shown to benefit both the optimization and generalization of neural networks, but large networks are resource hungry at both training and test time. Network pruning can reduce test-time resource requirements, but is typically applied to trained networks and therefore cannot avoid the expensive training process. We aim to prune networks at initialization, thereby saving resources at training time as well. Specifically, we argue that efficient training requires preserving the gradient flow through the network. This leads to a simple but effective pruning criterion we term Gradient Signal Preservation (GraSP). We empirically investigate the effectiveness of the proposed method with extensive experiments on CIFAR-10, CIFAR-100, Tiny-ImageNet and ImageNet, using VGGNet and ResNet architectures. Our method can prune 80% of the weights of a VGG-16 network on ImageNet at initialization, with only a 1.6% drop in top-1 accuracy. Moreover, our method achieves significantly better performance than the baseline at extreme sparsity levels.
We present an approach to construct appropriate and efficient emulators for Hamiltonian flow maps. Intended future applications are long-term tracing of fast charged particles in accelerators and magnetic plasma confinement configurations. The method is based on multi-output Gaussian process regression on scattered training data. To obtain long-term stability the symplectic property is enforced via the choice of the matrix-valued covariance function. Based on earlier work on spline interpolation we observe derivatives of the generating function of a canonical transformation. A product kernel produces an accurate implicit method, whereas a sum kernel results in a fast explicit method from this approach. Both correspond to a symplectic Euler method in terms of numerical integration. These methods are applied to the pendulum and the H\'enon-Heiles system and results compared to an symmetric regression with orthogonal polynomials. In the limit of small mapping times, the Hamiltonian function can be identified with a part of the generating function and thereby learned from observed time-series data of the system's evolution. Besides comparable performance of implicit kernel and spectral regression for symplectic maps, we demonstrate a substantial increase in performance for learning the Hamiltonian function compared to existing approaches.
This paper describes the design, implementation, and evaluation of Otak, a system that allows two non-colluding cloud providers to run machine learning (ML) inference without knowing the inputs to inference. Prior work for this problem mostly relies on advanced cryptography such as two-party secure computation (2PC) protocols that provide rigorous guarantees but suffer from high resource overhead. Otak improves efficiency via a new 2PC protocol that (i) tailors recent primitives such as function and homomorphic secret sharing to ML inference, and (ii) uses trusted hardware in a limited capacity to bootstrap the protocol. At the same time, Otak reduces trust assumptions on trusted hardware by running a small code inside the hardware, restricting its use to a preprocessing step, and distributing trust over heterogeneous trusted hardware platforms from different vendors. An implementation and evaluation of Otak demonstrates that its CPU and network overhead converted to a dollar amount is 5.4$-$385$\times$ lower than state-of-the-art 2PC-based works. Besides, Otak's trusted computing base (code inside trusted hardware) is only 1,300 lines of code, which is 14.6$-$29.2$\times$ lower than the code-size in prior trusted hardware-based works.
Dialogue policy learning for task-oriented dialogue systems has enjoyed great progress recently mostly through employing reinforcement learning methods. However, these approaches have become very sophisticated. It is time to re-evaluate it. Are we really making progress developing dialogue agents only based on reinforcement learning? We demonstrate how (1)~traditional supervised learning together with (2)~a simulator-free adversarial learning method can be used to achieve performance comparable to state-of-the-art RL-based methods. First, we introduce a simple dialogue action decoder to predict the appropriate actions. Then, the traditional multi-label classification solution for dialogue policy learning is extended by adding dense layers to improve the dialogue agent performance. Finally, we employ the Gumbel-Softmax estimator to alternatively train the dialogue agent and the dialogue reward model without using reinforcement learning. Based on our extensive experimentation, we can conclude the proposed methods can achieve more stable and higher performance with fewer efforts, such as the domain knowledge required to design a user simulator and the intractable parameter tuning in reinforcement learning. Our main goal is not to beat reinforcement learning with supervised learning, but to demonstrate the value of rethinking the role of reinforcement learning and supervised learning in optimizing task-oriented dialogue systems.
Research challenges encountered across science, engineering, and economics can frequently be formulated as optimization tasks. In chemistry and materials science, recent growth in laboratory digitization and automation has sparked interest in optimization-guided autonomous discovery and closed-loop experimentation. Experiment planning strategies based on off-the-shelf optimization algorithms can be employed in fully autonomous research platforms to achieve desired experimentation goals with the minimum number of trials. However, the experiment planning strategy that is most suitable to a scientific discovery task is a priori unknown while rigorous comparisons of different strategies are highly time and resource demanding. As optimization algorithms are typically benchmarked on low-dimensional synthetic functions, it is unclear how their performance would translate to noisy, higher-dimensional experimental tasks encountered in chemistry and materials science. We introduce Olympus, a software package that provides a consistent and easy-to-use framework for benchmarking optimization algorithms against realistic experiments emulated via probabilistic deep-learning models. Olympus includes a collection of experimentally derived benchmark sets from chemistry and materials science and a suite of experiment planning strategies that can be easily accessed via a user-friendly python interface. Furthermore, Olympus facilitates the integration, testing, and sharing of custom algorithms and user-defined datasets. In brief, Olympus mitigates the barriers associated with benchmarking optimization algorithms on realistic experimental scenarios, promoting data sharing and the creation of a standard framework for evaluating the performance of experiment planning strategies
Cloud data center capacities are over-provisioned to handle demand peaks and hardware failures which leads to low resources' utilization. One way to improve resource utilization and thus reduce the total cost of ownership is to offer unused resources (referred to as ephemeral resources) at a lower price. However, reselling resources needs to meet the expectations of its customers in terms of Quality of Service. The goal is so to maximize the amount of reclaimed resources while avoiding SLA penalties. To achieve that, cloud providers have to estimate their future utilization to provide availability guarantees. The prediction should consider a safety margin for resources to react to unpredictable workloads. The challenge is to find the safety margin that provides the best trade-off between the amount of resources to reclaim and the risk of SLA violations. Most state-of-the-art solutions consider a fixed safety margin for all types of metrics (e.g., CPU, RAM). However, a unique fixed margin does not consider various workloads variations over time which may lead to SLA violations or/and poor utilization. In order to tackle these challenges, we propose ReLeaSER, a Reinforcement Learning strategy for optimizing the ephemeral resources' utilization in the cloud. ReLeaSER dynamically tunes the safety margin at the host-level for each resource metric. The strategy learns from past prediction errors (that caused SLA violations). Our solution reduces significantly the SLA violation penalties on average by 2.7x and up to 3.4x. It also improves considerably the CPs' potential savings by 27.6% on average and up to 43.6%.
Modern approaches to text to speech require the entire input character sequence to be processed before any audio is synthesised. This latency limits the suitability of such models for time-sensitive tasks like simultaneous interpretation. Interleaving the action of reading a character with that of synthesising audio reduces this latency. However, the order of this sequence of interleaved actions varies across sentences, which raises the question of how the actions should be chosen. We propose a reinforcement learning based framework to train an agent to make this decision. We compare our performance against that of deterministic, rule-based systems. Our results demonstrate that our agent successfully balances the trade-off between the latency of audio generation and the quality of synthesised audio. More broadly, we show that neural sequence-to-sequence models can be adapted to run in an incremental manner.
Supply and demand are two fundamental concepts of sellers and customers. Predicting demand accurately is critical for organizations in order to be able to make plans. In this paper, we propose a new approach for demand prediction on an e-commerce web site. The proposed model differs from earlier models in several ways. The business model used in the e-commerce web site, for which the model is implemented, includes many sellers that sell the same product at the same time at different prices where the company operates a market place model. The demand prediction for such a model should consider the price of the same product sold by competing sellers along the features of these sellers. In this study we first applied different regression algorithms for specific set of products of one department of a company that is one of the most popular online e-commerce companies in Turkey. Then we used stacked generalization or also known as stacking ensemble learning to predict demand. Finally, all the approaches are evaluated on a real world data set obtained from the e-commerce company. The experimental results show that some of the machine learning methods do produce almost as good results as the stacked generalization method.