BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Balancing Quality and Efficiency in Future AI Systems - Shiqiang W
 ang - IBM
DTSTART:20241129T110000Z
DTEND:20241129T120000Z
UID:TALK224551@talks.cam.ac.uk
CONTACT:Sally Matthews
DESCRIPTION:Quality and efficiency are both essential in AI systems as dat
 a sources become more diverse and model sizes grow. In this talk\, I will 
 present techniques to address challenges in data and model quality as well
  as their efficiency\, which are essential for building high-performing an
 d sustainable AI systems. I will first introduce a toolkit for enhancing t
 he quality of datasets\, which can be used in a broad range of learning ta
 sks including the training or fine tuning large language models (LLMs)\, l
 aying the groundwork for model training with good data. Then\, considering
  the specific challenge where data is distributed unevenly across sources 
 with varying sizes\, quality\, and availability\, such as in the case of f
 ederated learning\, I will introduce the FedAU algorithm. This algorithm d
 ynamically adjusts aggregation weights in the model training process based
  on the availability of data sources\, to prevent model bias and improve t
 raining convergence. Afterwards\, I will introduce techniques to make both
  training and inference more efficient\, focusing on a framework that opti
 mizes model selection from a zoo of LLMs to minimize energy usage while ma
 intaining model performance guarantees. Together\, these approaches form a
  blueprint for future AI systems that are capable of learning effectively 
 from a vast amount of data at diverse sources and delivering high quality 
 models while enhancing resource efficiency in real-world applications.\n\n
 Bio: Shiqiang Wang is a Staff Research Scientist at IBM T. J. Watson Resea
 rch Center\, NY\, USA. He received his Ph.D. from Imperial College London\
 , United Kingdom\, in 2015. His research focuses on the intersection of di
 stributed computing\, machine learning\, networking\, and optimization\, c
 urrently emphasizing on quality and efficiency aspects related to distribu
 ted data and models\, which has a broad range of applications including di
 stributed data analytics\, efficient model training and inference\, edge-b
 ased artificial intelligence (Edge AI)\, and large language models (LLMs).
  He received the IEEE Communications Society (ComSoc) Leonard G. Abraham P
 rize in 2021\, IEEE ComSoc Best Young Professional Award in Industry in 20
 21\, IBM Outstanding Technical Achievement Awards (OTAA) in 2019\, 2021\, 
 2022\, and 2023\, multiple Invention Achievement Awards from IBM since 201
 6. For more details\, please visit his homepage at https://shiqiang.wang/
LOCATION:Computer Lab\, LT1
END:VEVENT
END:VCALENDAR