Posted by on
Tags: , , , , , , , , , , , , , , , ,
Categories: Uncategorized

Facebook AI Research, together with Google’s DeepMind, University of Washington, and New York University, today introduced SuperGLUE, a series of benchmark tasks to measure the performance of modern, high performance language-understanding AI.

SuperGLUE was made on the premise that deep learning models for conversational AI have “hit a ceiling” and need greater challenges. It uses Google’s BERT as a model performance baseline. Considered state of the art in many regards in 2018, BERT’s performance has been surpassed by a number of models this year such as Microsoft’s MT-DNN, Google’s XLNet, and Facebook’s RoBERTa, all of which were are based in part on BERT and achieve performance above a human baseline average.

Read more here:

https://venturebeat.com/2019/08/14/ai-researchers-launch-superglue-a-rigorous-benchmark-for-language-understanding/

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.