Xiaoxia Wu

About

Dr. Xiaoxia Wu is an accomplished researcher currently involved in the DeepSpeed project at Microsoft. Her work focuses on developing innovative methods to minimize the time and budget required for training or inferencing large-scale neural networks, with a particular emphasis on compression and natural language processing.

As a highly dedicated and passionate professional, Dr. Wu’s research interests span a wide range of topics, including large-scale optimization and machine learning in a broader sense. In her research, Dr. Wu consistently strives to make contributions to her field. Her expertise and commitment have not only made her a valuable asset to the DeepSpeed project, but they also continue to push the boundaries of machine learning and optimization techniques. For further information about her work and the DeepSpeed project, a visit to deepspeed.ai (opens in new tab) is highly recommended.

Earning her Ph.D. from The University of Texas at Austin, Dr. Wu demonstrated proficiency in efficient and robust methods for hyperparameter tuning, such as adaptive gradient descent and batch normalization. Her doctoral studies benefited greatly from the guidance of Rachel Ward and the informal co-advising of Léon Bottou.