In the news | The Register
Microsoft, OpenAI method could make training large neural networks cheaper
Cost of tuning hyperparameters using μTransfer was 7% of what it would be to pre-train GPT-3. Companies scaling up their neural network models could cut expensive training costs by employing a technique developed by researchers at Microsoft and OpenAI.
In the news | Wall Street Journal
Forget 5G. Let’s Talk About 6G.
Computer scientists and engineers say the next generation will make cloud computing and the mobile internet globally ubiquitous. As wireless carriers jostle to see who can win the race to deliver 5G services to customers, planning is already well under…
In the news | TechRadar
Microsoft, OpenAI may have solved a fundamental AI bottleneck
Microsoft and Open AI have developed a new method for optimizing massive AI models that are too expensive to train multiple times, such as GPT-3. A blog post published by Microsoft Research describes a technique called µ-Parametrization (or µP), which…
µTransfer: A technique for hyperparameter tuning of enormous neural networks
| Edward Hu, Greg Yang, and Jianfeng Gao
Great scientific achievements cannot be made by trial and error alone. Every launch in the space program is underpinned by centuries of fundamental research in aerodynamics, propulsion, and celestial bodies. In the same way, when it comes to building large-scale…
她们非同凡响的“另一面”
编者按：“人类经常少年老成、青年迷惑、中年喜欢将他人的成绩与自己相比较，因此觉得受挫，好不容易活到老年还是一个没有成长的笨孩子。我们一直粗糙地活着，而人的一生，便也这样过往了，”女作家三毛曾这样说过。 这些描述是否与你的人生轨迹有些许重合？我们都不约而同地走上了同样的路，忘记了最初的梦想，失去了曾经的兴趣，与那个不一样的自己渐行渐远。不过，在微软亚洲研究院有这样一群女性，她们在看似重复的科学研究和...
微软亚洲研究院多模态模型NÜWA：以自然语言创造视觉内容
编者按：此前我们曾提出了一个问题：从文字脚本生成创意视频一共分几步？微软亚洲研究院的开放领域视频生成预训练模型给出了答案：只需一步。现在，我们追问：除了文字生成视频之外，还有哪些途径可以生成视频？我们能否使用自然语言对视觉内容进行编辑？微软亚洲研究院最新推出的多模态模型 NÜWA，不仅让视觉内容创造多了一条路，甚至还让 Windows 经典桌面有了更多的打开方式。 人类对于信息的感知有五种途径，包...
Since COVID-19 first broke out, nearly 280 million people worldwide have contracted the disease, and more the 5.4 million people have died as a result, bringing huge losses and injuries to the global economy and to social life. In contrast,…
Explainability
ML models can be a “black box” that make decisions in ways that we don’t understand. Model Explainability is a feature that helps practitioners' probe how a model makes decisions so they can have more confidence in the results.
In the news | EPFL News
Pass me the pastries, please
A new global study by researchers from EPFL, the University of Fribourg, and Microsoft Research has found there was an overall surge in high calorie foods such as pastries, bread and pies during the first wave of the COVID-19 pandemic,…