Texar: A Modularized, Versatile, and Extensible Toolkit for Text Generation

Abstract

We introduce Texar, an open-source toolkit aiming to support the broad set of text generation tasks that transform any inputs into natural language, such as machine translation, summarization, dialog, content manipulation, and so forth. With the design goals of modularity, versatility, and extensibility in mind, Texar extracts common patterns underlying the diverse tasks and methodologies, creates a library of highly reusable modules, and allows arbitrary model architectures and algorithmic paradigms. In Texar, model architecture, inference, and learning processes are properly decomposed. Modules at a high concept level can be freely assembled and plugged in/swapped out. The toolkit also supports a rich set of large-scale pretrained models. Texar is thus particularly suitable for researchers and practitioners to do fast prototyping and experimentation. The versatile toolkit also fosters technique sharing across different text generation tasks. Texar supports both TensorFlow and PyTorch, and is released under Apache License 2.0 at https://www.texar.io.

arXiv commentACL 2019 demo, expanded version

References in corpus (23)

• Sequence to Sequence Learning with Neural Networks
• Categorical Reparameterization with Gumbel-Softmax
• Recurrent Neural Network Regularization
• Sequence Level Training with Recurrent Neural Networks
• End-To-End Memory Networks
• DeepMind Control Suite
• MaskGAN: Better Text Generation via Filling in the______
• Professor Forcing: A New Algorithm for Training Recurrent Networks
• Hybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation
• An Actor-Critic Algorithm for Sequence Prediction
• Unsupervised Machine Translation Using Monolingual Corpora Only
• Adversarial Feature Matching for Text Generation
• Get To The Point: Summarization with Pointer-Generator Networks
• On Unifying Deep Generative Models
• Learning End-to-End Goal-Oriented Dialog
• Massive Exploration of Neural Machine Translation Architectures
• XNMT: The eXtensible Neural Machine Translation Toolkit
• Challenges in Data-to-Document Generation
• Deep Generative Models with Learnable Knowledge Constraints
• Z-Forcing: Training Stochastic Recurrent Networks
• Nematus: a Toolkit for Neural Machine Translation
• Connecting the Dots Between MLE and RL for Sequence Prediction
• Data-to-Text Generation with Style Imitation

Cited by in corpus (9)

• Hybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation
• Toward Controlled Generation of Text
• Connecting the Dots Between MLE and RL for Sequence Prediction
• Contextual Text Style Transfer
• LeafNATS: An Open-Source Toolkit and Live Demo System for Neural Abstractive Text Summarization
• MixingBoard: a Knowledgeable Stylized Integrated Text Generation Platform
• DialogueBERT: A Self-Supervised Learning based Dialogue Pre-training Encoder
• Controllable and Diverse Text Generation in E-commerce
• GTAE: Graph-Transformer based Auto-Encoders for Linguistic-Constrained Text Style Transfer