3 ms·
SuperGLUE, a New Benchmark for NLP
- alok-g 7y agoSee also: [1] Excerpts: "Within one year of release, several NLP models have already surpassed human baseline performance on the GLUE benchmark." "According to Facebook AI, humans can obtain 100% accuracy on COPA while Google's BERT achieved only 74%, signifying there is a lot of room for NLP improvement." [1] https://www.zdnet.com/article/facebook-deepmind-nyu-and-uw-build-new-benchmarking-platform-for-ai-language-understanding/ https://www.zdnet.com/article/facebook-deepmind-nyu-and-uw-b...