Microsoft posted on its research blog about advancements it is making with the Turing Natural Language Generation (T-NLG) model. Similar to other deep learning language models like BERT and others, this one is more powerful in that it is the largest model ever published at 17 billion parameters.
Microsoft wrote that T-NLG "outperforms the state of the art on a variety of language modeling benchmarks and also excels when applied to numerous practical tasks, including summarization and question answering." "T-NLG is a Transformer-based generative language model, which means it can generate words to complete open-ended textual tasks. In addition to completing an unfinished sentence, it can generate direct answers to questions and summaries of input documents," they added.
This is being added to Bing, the blog post reads "Our work is actively being integrated into multiple Microsoft products including Bing, Office, and Xbox."
In fact, Frédéric Dubut from the Bing team said on Twitter "Super proud of the Turing team for their public announcement and release! Some of them work just a few doors down the hallway from me... Guess what's going to power @Bing web ranking soon."
Super proud of the Turing team for their public announcement and release! Some of them work just a few doors down the hallway from me... Guess what's going to power @Bing web ranking soon 😎 #DeepLearning #NLP #SEO https://t.co/EUumIwReP8
— Frédéric Dubut (@CoperniX) February 11, 2020
How does it compare? Look at this chart:
Move over BERT, here comes Turing. Read more in detail over here.
Forum discussion at Twitter.
