Recent progress in natural language processing has been driven by advances in both model architecture and model pretraining. Transformer architectures have facilitated building higher-capacity models and pretraining has made it possible to effectively utilize this capacity for a wide variety of tasks. Transformers is an open-source library with the goal of opening up these advances to the wider machine learning community. The library consists of carefully engineered stateof-the art Transformer architectures under a unified API. Backing this library is a curated collection of pretrained models made by and available for the community. Transformers is designed to be extensible by researchers, simple for practitioners, and fast and robust in industrial deployments. The library is available at https://github.com/ huggingface/transformers.

No abstract

Much effort has been devoted to evaluate whether multi-task learning can be leveraged to learn rich representations that can be used in various Natural Language Processing (NLP) down-stream applications. However, there is still a lack of understanding of the settings in which multi-task learning has a significant effect. In this work, we introduce a hierarchical model trained in a multi-task learning setup on a set of carefully selected semantic tasks. The model is trained in a hierarchical fashion to introduce an inductive bias by supervising a set of low level tasks at the bottom layers of the model and more complex tasks at the top layers of the model. This model achieves state-of-the-art results on a number of tasks, namely Named Entity Recognition, Entity Mention Detection and Relation Extraction without hand-engineered features or external NLP tools like syntactic parsers. The hierarchical training supervision induces a set of shared semantic representations at lower layers of the model. We show that as we move from the bottom to the top layers of the model, the hidden states of the layers tend to represent more complex semantic information.

In this work we present a deep convolutional neural network using 3D convolutions for Gait Recognition in multiple views capturing spatio-temporal features. A special input format, consisting of the gray-scale image and optical flow enhance color invaranice. The approach is evaluated on three different datasets, including variances in clothing, walking speeds and the view angle. In contrast to most state-of-the-art Gait Recognition systems the used neural network is able to generalize gait features across multiple large view angle changes. The results show a comparable to better performance in comparison with previous approaches, especially for large view differences.

Large-scale pretrained language models define state of the art in natural language processing, achieving outstanding performance on a variety of tasks. We study how these architectures can be applied and adapted for natural language generation, comparing a number of architectural and training schemes. We focus in particular on open-domain dialog as a typical high entropy generation task, presenting and comparing different architectures for adapting pretrained models with state of the art results.

Although it is well established that regions of premotor cortex (PMC) are active during action observation, it remains controversial whether they play a causal role in action understanding. In the experiment reported here, we used off-line continuous theta-burst stimulation (cTBS) to investigate this question. Participants received cTBS over the hand and lip areas of left PMC, in separate sessions, before completing a pantomime-recognition task in which half of the trials contained pantomimed hand actions, and half contained pantomimed mouth actions. The results reveal a double dissociation: Participants were less accurate in recognizing pantomimed hand actions after receiving cTBS over the hand area than over the lip area and less accurate in recognizing pantomimed mouth actions after receiving cTBS over the lip area than over the hand area. This finding constrains theories of action understanding by showing that somatotopically organized regions of PMC contribute causally to action understanding and, thus, that the mechanisms underpinning action understanding and action performance overlap.

No abstract

No abstract