Microsoft cognitive open source Microsoft toolkit deep learning kit, to strengthen the learning

Heavy | Microsoft open source Microsoft Cognitive Toolkit deep learning toolkit, adding reinforcement learning elements – Sohu technology Author: Allison Linn Blog Microsoft from the heart of machines: Wu Pan, Li Yazhou compiled in today, Microsoft announced the release of the updated version of Microsoft Cognitive Toolkit, which is a system for deep learning, can be used to speed up the development of CPU and NVIDIA GPU on the voice and image recognition and search relevance etc.. This toolkit, known as CNTK, was first developed by Microsoft computer scientists who want to do their research more quickly and efficiently. It was soon beyond the field of voice and evolved into a product, including some international leading home appliance manufacturers and Microsoft’s flagship product group (flagship product groups), customers rely on it to carry out a variety of deep learning tasks. A key architect Frank Seide Microsoft Artificial Intelligence and Research Microsoft Cognitive Toolkit and chief scientist said: "we made it can be used in products of things from a research tool. "Frank Seide the latest version of the toolkit has now been released to GitHub via an open source certificate, which includes support for Python and C++ programming languages. Researchers can also use this new version to develop an artificial intelligence system called learning (reinforcement). Finally, the performance of the toolkit is also better than previous versions. It is also faster than other toolkits, especially when you need to run large data sets across multiple machines. In order to develop consumer products and professional products, this large-scale deployment is necessary for deep learning across multiple GPU. This is also the key to accelerating research breakthroughs. Last week, Microsoft Artificial Intelligence Research announced in recognition of the dialogue has reached the level of human and. The team will be able to achieve this milestone behind the huge speed thanks to the Microsoft Cognitive Toolkit. The team that developed the Microsoft toolkit says its ability to work across multiple servers is a key advantage over other depth learning kits. Better performance and accuracy can be achieved when the Microsoft toolkit is used to solve larger data sets. Microsoft Cognitive Toolkit has a built-in algorithm to minimize the degradation of this computation (degradation of computation). "Use Min相关的主题文章: