Texygen: A Benchmarking Platform for Text Generation Models

TL;DR

Texygen is a comprehensive benchmarking platform that standardizes evaluation and sharing of open-domain text generation models, enhancing reproducibility and research collaboration.

Contribution

It provides an integrated platform with models and metrics, streamlining evaluation and fostering reproducibility in text generation research.

Findings

Implemented multiple text generation models.

Covered diverse metrics for evaluation.

Facilitates research standardization and sharing.

Abstract

We introduce Texygen, a benchmarking platform to support research on open-domain text generation models. Texygen has not only implemented a majority of text generation models, but also covered a set of metrics that evaluate the diversity, the quality and the consistency of the generated texts. The Texygen platform could help standardize the research on text generation and facilitate the sharing of fine-tuned open-source implementations among researchers for their work. As a consequence, this would help in improving the reproductivity and reliability of future research work in text generation.