all AI news
High Quality Rather than High Model Probability: Minimum Bayes Risk Decoding with Neural Metrics. (arXiv:2111.09388v3 [cs.CL] UPDATED)
April 27, 2022, 1:12 a.m. | Markus Freitag, David Grangier, Qijun Tan, Bowen Liang
cs.LG updates on arXiv.org arxiv.org
In Neural Machine Translation, it is typically assumed that the sentence with
the highest estimated probability should also be the translation with the
highest quality as measured by humans. In this work, we question this
assumption and show that model estimates and translation quality only vaguely
correlate. We apply Minimum Bayes Risk (MBR) decoding on unbiased samples to
optimize diverse automated metrics of translation quality as an alternative
inference strategy to beam search. Instead of targeting the hypotheses with the …
More from arxiv.org / cs.LG updates on arXiv.org
Jobs in AI, ML, Big Data
Senior ML Researcher - 3D Geometry Processing | 3D Shape Generation | 3D Mesh Data
@ Promaton | Europe
Data Scientist
@ Motive | India - Remote
Senior Perception Engineer
@ NVIDIA | US, CA, Santa Clara
Business Data Analyst, Finance and Treasury Data Repositories, Senior Associate
@ State Street | Krakow, Poland
Junior AI Engineer (Internship)
@ Sony | SEU - Italy - Roma
Manager, Data Science 3
@ PayPal | USA - Pennsylvania - Virtual