In this paper, graph theory is used to calculate the contributions of individual generators and loads to line flows and the real power transfer between individual generators and loads that are significant to transmission open access. Related lemmas are proved which present necessary conditions required by the method. Based on ac load flow solution a novel method is suggested which can decide downstream and upstream power flow tracing paths very fast and calculate the contribution factors of generations and loads to the line flows efficiently. The power transfer between generators and loads can also be determined. The suggested method is suitable for both active and reactive power tracings of real power systems. Index Terms-Graph theory application, power flow tracing, transmission open access. Ping Wei received her M.Eng. degree in electrical engineering from Southeast University, China. She is now a Ph.D. student, Department of EEE, the University of Hong Kong. Her research area is power system operation, power market, and operations research applications in power systems.

We present Deep Voice 3, a fully-convolutional attention-based neural textto-speech (TTS) system. Deep Voice 3 matches state-of-the-art neural speech synthesis systems in naturalness while training an order of magnitude faster. We scale Deep Voice 3 to dataset sizes unprecedented for TTS, training on more than eight hundred hours of audio from over two thousand speakers. In addition, we identify common error modes of attention-based speech synthesis networks, demonstrate how to mitigate them, and compare several different waveform synthesis methods. We also describe how to scale inference to ten million queries per day on a single GPU server.

In this work, we propose a new solution for parallel wave generation by WaveNet. In contrast to parallel WaveNet (van den Oord et al., 2018), we distill a Gaussian inverse autoregressive flow from the autoregressive WaveNet by minimizing a regularized KL divergence between their highly-peaked output distributions. Our method computes the KL divergence in closed-form, which simplifies the training algorithm and provides very efficient distillation. In addition, we introduce the first text-to-wave neural architecture for speech synthesis, which is fully convolutional and enables fast end-to-end training from scratch. It significantly outperforms the previous pipeline that connects a text-to-spectrogram model to a separately trained WaveNet (Ping et al., 2018). We also successfully distill a parallel waveform synthesizer conditioned on the hidden representation in this end-to-end model. 2 * These authors contributed equally to this work. Our method is named after the musical instrument clarinet, whose sound resembles human voice.2 Audio samples are in https://clarinet-demo.github.io/

Recent work on training neural retrievers foropen-domain question answering (OpenQA) has employed both supervised and unsupervised approaches. However, it remains unclear how unsupervised and supervised methods can be used most effectively for neural retrievers. In this work, we systematically study retriever pre-training. We first propose an approach of unsupervised pre-training with the Inverse Cloze Task and masked salient spans, followed by supervised finetuning using question-context pairs. This approach leads to absolute gains of 2+ points over the previous best result in the top-20 retrieval accuracy on Natural Questions and TriviaQA datasets. We next explore two approaches for end-toend training of the reader and retriever components in OpenQA models, which differ in the manner the reader ingests the retrieved documents. Our experiments demonstrate the effectiveness of these approaches as we obtain state-of-the-art results. On the Natural Questions dataset, we obtain a top-20 retrieval accuracy of 84%, an improvement of 5 points over the recent DPR model. We also showcase good results on answer extraction, outperforming recent models such as REALM and RAG by 3+ points. Our code is available at: https: //github.com/NVIDIA/Megatron-LM. * This work was done during an internship at NVIDIA.

Abstract. Hierarchical classifications are used pervasively by humans as a means to organize their data and knowledge about the world. One of their main advantages is that natural language labels, used to describe their contents, are easily understood by human users. However, at the same time, this is also one of their main disadvantages as these same labels are ambiguous and very hard to be reasoned about by software agents. This fact creates an insuperable hindrance for classifications to being embedded in the Semantic Web infrastructure. This paper presents an approach to converting classifications into lightweight ontologies, and it makes the following contributions: (i) it identifies the main NLP problems related to the conversion process and shows how they are different from the classical problems of NLP; (ii) it proposes heuristic solutions to these problems, which are especially effective in this domain; and (iii) it evaluates the proposed solutions by testing them on DMoz data.