The talk will focus on Deep Learning, NLP, Large Scale Training, Universal Language Representation, as well as research opportunities at Microsoft.
Saksham Singhal is currently working as an Applied Scientist at Microsoft on projects across their entire deep learning life-cycle. This includes universal language representation model, question-answering models and other large scale models empowering various scenarios in Microsoft Office and Bing. Apart from this, his work encompasses improving training performance, run-time optimized inference, distributed training, and other similar aspects. They also work on optimizing the models for serving at a very tight latency constraints on a diverse set of hardware including CPUs, GPUs and FPGAs.
Saksham is an alumnus of the Master’s in Computational Data Science Program (2016-17). He will also discuss his role/work at Microsoft and share the cutting edge work happening in the industry and specifically in his team.