Neural Architecture Search (NAS) has emerged as a revolutionary approach to automating the intricate design of Artificial Neural Networks (ANNs). Over the past decade, NAS has captivated the attention and enthusiasm of both the Computational Intelligence and Machine Learning communities.

Traditionally, NAS algorithms are grouped into three main categories based on their search methodologies:

    1. Reinforcement Learning
    2. Gradient-based Techniques
    3. Evolutionary Computation

Yet, the landscape of NAS is continually evolving. Modern NAS innovations often defy these clear-cut classifications. Some algorithms exhibit characteristics spanning multiple categories, while others introduce groundbreaking architectural encodings that challenge conventional search methods.

The dynamism of this field is undeniable. Contemporary NAS techniques are pushing boundaries by enabling the design of complex Deep Neural Networks within stringent time constraints. This progress is driven by:

    • Streamlined Search Spaces: Efficiently defined to focus on optimal solutions.
    • GPU Optimisation: Leveraging the power of GPUs for accelerated computations.
    • Approximation Techniques: Using intelligent approximating functions to enhance performance.

Notably, while some NAS approaches excel in specific domains, others produce adaptable architectures, empowering users to effortlessly transition between diverse applications.

This workshop aims to delve into the contemporary trends of NAS, explore its successes and challenges, and contemplate its future directions and broader impact within the AI community, industry, and society.

Date: 25 June, 2024
Time: 10:30 am – 12:30 pm
Venue: Lotus Junior 4D

Website: ANASTA

Zexuan Zhu (Shenzhen University, China)
Liang Feng (Chongqing University, China)
Yaqing Hou (Dalian University of Technology, China)
Ran Cheng (Southern University of Science and Technology, China)
Zhihui Zhan (South China University of Technology, China)

Computational Intelligence and Machine Learning have emerged as the driving forces behind advancements in the field of information technology. Rapid progress in related domains not only offers fresh perspectives for scientific inquiry but also exerts profound impacts across a myriad of practical applications. This workshop “Advances in Computational Intelligence and Machine Learning: Applications and Innovations” aims to explore the latest developments in Computational Intelligence and Machine Learning, showcasing their practical applications and innovations across various domains.
This workshop will encompass the following aspects:

    • Theory and Methods: We will delve into the latest theories and methods within the realms of Computational Intelligence and Machine Learning. This includes, but is not limited to, Deep Learning, Reinforcement Learning, Evolutionary Algorithms, Fuzzy Logic, Genetic Algorithms, and more. Discussions will revolve around the principles, advantages, and applications of these methodologies across diverse problem domains.
    • Practical Applications: Emphasis will be placed on the application of Computational Intelligence and Machine Learning in real-world scenarios. This encompasses Natural Language Processing, Computer Vision, Medical Diagnostics, and Financial Forecasting, among other fields. We will explore how these applications are reshaping our lives and work processes, as well as discuss their potential impacts and challenges.
    • Innovation and Future Trends: Discussions will center around the innovations and future trends in Computational Intelligence and Machine Learning. This includes emerging algorithms, novel application domains, and new research directions. We will explore how these innovations drive technological advancements and societal developments, while also deliberating on potential future directions and challenges.

Through this workshop, we aim to provide a platform for participants to exchange ideas and learn from one another, fostering collaboration and innovation within the realms of Computational Intelligence and Machine Learning. By facilitating dialogue and knowledge-sharing, we seek to propel advancements in these fields, addressing the increasingly complex and diverse technological and societal challenges of our time.

Date: 27 June, 2024
Time: 3:30 pm – 6:00 pm
Venue: Melati Ballroom 4101AB-4102

Website: ACIML

Ong Yew Soon (School of Computer Science and Engineering, Nanyang Technological University, Singapore)
Huanhuan Chen (School of Computer Science and Technology, University of Science and Technology of China, China)
Shan He (School of Computer Science, University of Birmingham, UK)
Bing Xue (School of Engineering and Computer Science, Victoria University of Wellington, New Zealand)

Artificial Intelligence (AI) is revolutionising the transportation sector including maritime, land, and aerospace industries, introducing unprecedented advancements for the next stage of transport connectivity and further economic growth. This workshop delves into the innovative applications of AI focusing on maritime operations, port traffic management, urban mobility, electric and autonomous transportation systems including aerospace. Emphasising the intersection of AI, machine learning models, and predictive analytics, the symposium highlights their potential to revolutionise transportation by reducing emissions and enhancing safety. Together we will discover how AI-driven insights can refine vehicle and vessel operations, paving the way for sustainable and safer transportation solutions. Don’t miss the opportunity to be part of the next stage of transport connectivity and economic growth.

Date: 27 June, 2024
Time: 10:30 am – 3:30 pm
Venue: Melati Ballroom 4101AB-4102

Website: AI-DIIT

Xiuju FU (Institute of High Performance Computing, A*STAR, Singapore)
Ran YAN (Nanyang Technological University, Singapore)
Zhiyong CUI (Beihang University, China)
Shuo FENG (Tsinghua University, China)
Yonggang LUO (AI Lab, Chongqing Changan Automobile, China)
Abhinav SAXENA (GE Aerospace Research, USA)
Xiao Feng YIN (Institute of High Performance Computing, A*STAR, Singapore)
Yunhui LIN (Institute of High Performance Computing, A*STAR, Singapore)
Hongwei WANG (Institute of High Performance Computing, A*STAR, Singapore)
Ping Chong CHUA (Institute of High Performance Computing, A*STAR, Singapore)

Recent advanced AI technologies, especially the large language models (LLMs) like GPTs, have significantly advanced the field of natural language processing (NLP) and led to the development of various LLM-based applications. One potential application is as communication interfaces in human-in-the-loop education systems, where the model serves as a mediator among the teacher, students and the machine capabilities including its own. This approach has several benefits, including the ability to personalise interactions, allow unprecedented flexibility and adaptivity for human-AI collaboration and improve the user experience. However, several challenges still exist in implementing this approach, including the need for more robust models, designing effective user interfaces, and ensuring ethical considerations are addressed.

Date: 25 June, 2024
Time: 10:30 am – 2:30 pm
Venue: Melati Junior 4011

Website: AI4EDU
Call for Paper: AI4EDU CFP

Qingsong Wen (Squirrel AI, USA)
Joleen Liang (Squirrel AI, China)
Richard Tong (Squirrel AI, USA)
Yu Lu (Beijing Normal University, China)
Liu Guimei (A*STAR, Singapore)
Xiangen Hu (The University of Memphis, Tennessee, USA)
Robby Robson (Eduworks, USA)

Energy is the cornerstone of societal prosperity and its production, delivery, and management are essential to mankind. As the dynamic energy field continues to evolve, it’s crucial that we broaden our perspective to include control and optimisation within intricate energy systems. The escalating energy needs of developing economies are undeniable. However, considering our planet’s finite resources and the significant climate impact from the power sector, it’s essential to devise strategies that maintain environmental integrity and champion sustainability. Integrating AI into the energy business requires trust, explainability, and fairness. By making AI systems trustworthy, transparent, and fair, we can use AI to solve difficult energy production, distribution, and management problems while protecting the environment and promoting sustainability.

Date: 27 June, 2024
Time: 3:30 pm – 6:00 pm
Venue: Lotus Junior 4D

Website: AI For Enegry

Zita Vale (Instituto Superior de Engenharia do Porto, Portugal)
G. Kumar Venaygamoorthy (Clemson University, USA)
João Soares (Instituto Superior de Engenharia do Porto, Portugal)

Pressing environmental challenges, such as climate change, deforestation, pollution, and loss of biodiversity, require urgent global and local attention. Addressing these issues demands multidisciplinary collaborative efforts to develop and implement large-scale sustainable solutions, ensuring the health and resilience of our planet. The machine learning (ML), or artificial intelligence (AI) community wishes to take action on solving these environmental issues but is often uncertain about the most effective intervention with maximum impacts. This workshop seeks to highlight potential environment-related research using ML/AI tools and illustrate the invaluable role ML/AI can play in reducing greenhouse gas emissions, large-scale estimation of carbon stock and biodiversity, smart and green transportation systems, effective optimisation of natural resources such as water and fisheries, ensuring sustainable agricultural practices, and resilience adaptation/mitigation practices to the reality of the changing climate. Many of these actions not only present opportunities for substantial real-world impact but also pose intriguing challenges for academic research.

Date: 27 June, 2024
Time: 10:30 am – 3:30 pm
Venue: Lotus Junior 4E

Website: AI4EI
Call for Paper: AI4EI CFP

Khoa D Doan (VinUniversity & CEI, Vietnam)
Huong T (Helen) Nguyen (University of Illinois at Urbana-Champaign, USA)
Nitesh Chawla (University of Notre Dame/Lucy Family Institute for Data and Society, USA)
Alexandre D’Aspremont (Ecole Normale Supérieure, France)
Karina Gin Yew-Hoong (National University of Singapore, Singapore)
Erick G. Sperandio Nascimento (Surrey Institute for People-Centred AI, University of Surrey, UK)
Harry Nguyen (University College Cork, Ireland)
Kok-seng Wong (VinUniversity & CEI, Vietnam)
Doanh N Nguyen (VinUniversity & CEI, Vietnam)
Alex A. Bandeira Santos (SENAI CIMATEC, Brazil)
Arshdeep Singh (Centre for Vision, Speech and Signal Processing, University of Surrey, UK)

Blockchain and Artificial Intelligence (AI) have individually transformed industries, and their convergence is poised to be the next frontier of innovation. This workshop aims to explore the multifaceted relationship between blockchain and AI, offering a platform for researchers, practitioners, and enthusiasts to discuss the latest developments, challenges, and opportunities at the intersection of these cutting-edge technologies. This workshop will provide a comprehensive overview of the synergies between blockchain and AI, fostering collaboration and innovation in these exciting fields. Participants will have the opportunity to network, share insights, and contribute to the advancement of this intersection of technologies.

Date: 25 June, 2024
Time: 10:30 am – 2:30 pm
Venue: Melati Junior 4111

Website: BCAI 2024
Call for Paper: BCAI 2024 CFP

Goh Siow Mong Rick (A*STAR, Singapore)
Liu Yong (A*STAR, Singapore)
Wei Qingsong (A*STAR, Singapore)
Toyoda Kentaroh (A*STAR, Singapore)
Yonggang Wen (Nanyang Technological University, Singapore)
David Lo (Singapore Management University, Singapore)
Liu Yang (Nanyang Technological University, Singapore)
Wei Wang (Hong Kong University of Science and Technology, Hong Kong, China)
Jin Zhang (Southern University of Science and Technology, Shenzhen, China)
Xiaofan Liu (City University of Hong Kong, AI-Powered Financial Technologies Ltd, Hong Kong, China)

From autonomous vehicles and 3D printing through to smart cities and digital twins, the gap between our physical and digital worlds is growing ever smaller. At the interface of these two worlds is Data-Centric Engineering, a rapidly emerging new branch of science for the 21st century which combines the power and insight available from large-scale data sources with the tools and technology that shape our real-world environment. Bridging the gap between the digital and physical realms requires the development of new machine learning algorithms that are capable of processing large volumes of data from satellites, cell phones, distributed sensors, etc. and making robust and transparent decisions that improve our daily lives and protect our natural environment. As Data-Centric Engineering is an emerging discipline it needs to define its boundaries, determine the techniques of importance, and nurture a diverse community of ECRs to take these ideas forward. The focus of this workshop is to bring the engineering and machine learning communities together to define Data-Centric Engineering, through keynote talks and contributed sessions, panel discussions, poster presentations and networking.

Date: 27 June, 2024
Time: 10:30 am – 3:30 pm
Venue: Lotus Junior 4D

Website: Crafting Data-Centric Engineering

Adam Sobey (The Alan Turing Institute, UK)
Gabin Kayumbi (The Alan Turing Institute, UK)
Jiangyan Shao (Wuhan University of Technology, China)
William Cooper (The Alan Turing Institute, UK)
Zack Xuereb Conti (The Alan Turing Institute, UK)
Lawrence Bull (University of Cambridge, UK)

In our rapidly changing world, the ability to learn and adapt is paramount. This capability enables us to navigate complex challenges, leveraging our past experiences to avoid previous pitfalls and apply knowledge in novel contexts. Remarkably, the human capacity for selecting and generalising relevant experiences to new problems stands as a testament to our cognitive sophistication.

Within the context of computational intelligence, several core learning technologies in neural and cognitive systems, fuzzy systems, probabilistic and possibilistic reasoning, have shown promise in emulating the generalisation capabilities of human learning, with many now used routinely to enhance our daily lives. Recently, in contrast to traditional machine learning approaches, Transfer Learning, which uses data from related source tasks to augment learning in a new (target) task, has attracted extensive attention and demonstrated great success in a wide range of real-world applications, including computer vision, natural language processing, speech recognition, etc.

In spite of several advances in computational intelligence, it is noted that the attempts to emulate such cognitive capabilities in problem solvers, especially those of an evolutionary nature, have received far less attention. In fact, most existing evolutionary algorithms (EAs) remain ill-equipped to exploit the potentially rich sources of knowledge that may be embedded in previous searches. With this, and the observation that any practically useful industrial system is likely to face a large number of (possibly repetitive) problems over a lifetime, it is contended that novel research advances in both the theory and application of Transfer Learning, especially for Optimisation, are primed to bring about a new wave in the real-world impact of intelligent systems.

The main goal of this workshop is to promote the research on crafting novel algorithm designs as well as theoretical analysis towards “intelligent” evolutionary computation, which possesses the transfer capabilities that evolve along with the problems solved. Further, this workshop also aims at providing a forum for academic and industrial researchers to explore future directions of research and promote evolutionary computation techniques to a wider audience in the society of computer science and engineering.

Date: 25 June, 2024
Time: 10:30 am – 12:30 pm
Venue: Lotus Junior 4D

Website: ETLOSI

Zexuan Zhu (Shenzhen University, China)
Liang Feng (Chongqing University, China)
Yaqing Hou (Dalian University of Technology, China)
Ran Cheng (Southern University of Science and Technology, China)
Zhihui Zhan (South China University of Technology, China)

Foundation models have garnered significant attention from both research and industry communities. It holds profound potential for healthcare applications. These models, which can be trained using vast amounts of unlabeled data, including medical images, medical reports, and Electronic Medical Records (EMR), without requiring explicit data annotation, hold immense potential for healthcare AI. Their ability to be trained with diverse data from multiple sources offers numerous benefits, including reduced reliance on data annotation for fine-tuning and the capacity for robust generalisation in real-world applications across various cohorts, ethnicities, and devices. This workshop aims to explore and advance the application of foundation models, especially multimodal foundation model including medical image, medical reports, genetic, etc. It will also look into foundation models from the lens of robustness, privacy preserving, explainable, federated, and distributed perspectives in healthcare.

Date: 25 June, 2024
Time: 10:30 am – 2:30 pm
Venue: Lotus Junior 4E

Website: FM4H

Yong Liu (A*STAR, Singapore)
Goh Siow Mong Rick (A*STAR, Singapore)
Xinxing Xu (A*STAR, Singapore)
Dacheng Tao (The University of Sydney, Australia)
Dinggang Shen (ShanghaiTech University, China)
Li Shuo (Case Western Reserve University, USA)

Evolutionary computation is a family of algorithms modeled on natural evolution. It applies mechanisms like mutation, crossover, and selection to a population of candidate solutions, allowing them to evolve over generations. This approach excels in exploring vast search spaces and addressing complex optimisation tasks, making it a powerful tool in fields like engineering and artificial intelligence.

Traditionally, the category of evolutionary computation encompasses several sub-categories and techniques. Here is a list of some of the main categories within this field:

    1. Genetic Algorithms (GA)
    2. Evolution Strategies (ES)
    3. Genetic Programming (GP)
    4. Swarm Intelligence Algorithms
    5. Differential Evolution (DE)
    6. Ant Colony Optimisation (ACO)
    7. Artificial Immune Systems (AIS), etc.

The dynamism and advanced research fields of evolutionary computation are vast and diverse, reflecting the ever-evolving nature of this field. Here is a list of some of the current dynamic and advanced research areas within evolutionary computation:

    • Hybridisation and Integration with Other Methods
    • Multi-objective Optimisation
    • Large-scale Optimisation
    • Constrained Optimisation
    • Evolutionary Multitasking and Transfer Optimisation
    • Evolutionary Deep Learning
    • Evolutionary Theory
    • Real-world Applications, etc.

These are just a few examples of the dynamism and advanced research fields within evolutionary computation. As the field continues to evolve, new challenges and opportunities will arise, driving further advancements in these and other areas. This workshop aims to delve into the contemporary trends of evolutionary computation techniques, explore its successes and challenges, and contemplate its future directions and broader impact within the AI community, industry, and society.

Date: 25 June, 2024
Time: 1:30 pm – 2:30 pm
Venue: Lotus Junior 4D

Website: FECIA

Zexuan Zhu (Shenzhen University, China)
Liang Feng (Chongqing University, China)
Yaqing Hou (Dalian University of Technology, China)
Ran Cheng (Southern University of Science and Technology, China)
Zhihui Zhan (South China University of Technology, China)

Workshop Abstract:

This workshop is a comprehensive deep-dive into accelerating and deploying large language models (LLMs). We start by empowering participants to optimize GPU kernels using Triton, ensuring cross-platform compatibility. Next, we explore DirectML-based quantization techniques, enabling the efficient execution of LLMs even on Windows laptops. Finally, we introduce a declarative paradigm for RAG (Retrieval-Augmented Generation) and LLM chaining using JamAI Base, simplifying the implementation of complex AI applications. This workshop blends theoretical insights with practical examples, equipping attendees with the skills and knowledge to accelerate and deploy their AI projects.

Workshop Objectives:

  1. Triton Mastery: Participants will gain proficiency in writing optimized GPU kernels for any GPU using Triton, enhancing their AI model performance and hardware versatility.
  2. Quantized LLM Deployment: Attendees will learn how to quantize LLMs for execution on DirectML, unlocking efficient LLM deployment on a wider range of devices, including Windows laptops.
  3. Declarative RAG and LLM Chaining: Participants will discover how to leverage the JamAI Base framework to simplify the creation and management of RAG pipelines and LLM chains, streamlining the development of AI applications.

Workshop Outline:

Part 1: A Practitioner’s Guide to Triton (How to write GPU kernel for any GPU)

  • Introduction to Triton: Why Triton, its advantages, and use cases in AI model acceleration. Comparison with ROCm and torch.compile.
  • Triton Environment Setup: Step-by-step instructions on setting up the Triton environment and debugging tools.
  • Triton Programming Model: Understanding block-based computation, vectorization, and memory access patterns.
  • Hands-On Examples:
    • Copying a tensor
    • Greyscaling an image
    • Matrix multiplication (naive and optimized)
    • Advanced optimizations (faster matmul)
  • Benchmarking and Profiling: How to benchmark Triton kernels against PyTorch, auto-tuning, and using Nsight Compute for profiling.

Part 2: Quantizing and Running LLMs on ONNX / DirectML (How to run LLM on Windows Laptop / Mobile)

  • Introduction to DirectML: Why DirectML is the fastest way to run LLMs on Windows machines (GPU), and how it outperforms llama.cpp.
  • Quantization Techniques: Overview of quantization methods (e.g., int4 AWQ) for LLMs.
  • Implementation: Step-by-step guide on quantizing and running LLMs (Llama-3, Phi-3, Mistral, etc.) on DirectML.
  • Performance Considerations: Tips for optimizing quantized LLM performance on DirectML.

Part 3: Declarative RAG and LLM Chaining with JamAI Base

  • The Declarative Paradigm: Introduction to the concept of defining the “what” instead of the “how” in AI application development.
  • JamAI Base Overview: Features and architecture of the JamAI Base platform, its embedded database (SQLite), vector database (LanceDB), and managed memory capabilities.
  • Hands-On with JamAI Base: Guided exercises to create and deploy RAG pipelines and LLM chains using the intuitive UI and REST API of JamAI Base.
  • Case Studies: Real-world examples demonstrating the power of the declarative paradigm for building complex AI applications.

Target Audience:

This workshop is designed for data scientists, machine learning engineers, developers, researchers, and AI practitioners interested in optimizing and deploying AI models. The workshop welcomes participants with varying levels of expertise, from those with basic Python knowledge and a familiarity with GPU concepts to experienced AI professionals seeking advanced acceleration techniques.

Workshop Materials:

  • Comprehensive slide decks for each part.
  • Jupyter notebooks with code examples for hands-on sessions.
  • Links to relevant resources and documentation.
  • Access to a virtual machine or cloud environment (if necessary).

Expected Outcomes:

By the end of this workshop, participants will:

  • Be proficient in using Triton for GPU kernel optimization.
  • Have the skills to quantize and run LLMs on DirectML.
  • Understand and apply the declarative paradigm for RAG and LLM chaining.
  • Gain practical experience using JamAI Base for rapid AI application development.
  • Be equipped with the knowledge and tools to accelerate and deploy their AI projects efficiently.

Additional Information:

  • The workshop will be interactive, with dedicated Q&A sessions and discussions.
  • We encourage participants to bring their laptops (Windows/MacOS) to follow along with the hands-on exercises.
  • We are committed to fostering a collaborative and inclusive learning environment.

We are confident that this workshop will provide a valuable and enriching experience for all attendees. We look forward to your participation in IEEE CAI 2024!

Date: 25 June, 2024
Time: 2:30 pm to 6:00 pm
Venue: Lotus Junior 4D


Tan Tun Jian, Embedded LLM
Tan Pin Siang, Embedded LLM
Tan Jia Huei, Embedded LLM
Cheong Ye Hur, Embedded LLM

The explosive success of large language models (LLMs) has captured the attention of the deep learning research community, and their powerful capabilities have demonstrated remarkable performance in various fields, such as natural language processing and computer vision. For example, since the ChatGPT of OpenAI has been online, it has accumulated over 10 billion users. However, LLMs face significant efficiency challenges, especially when deploying large neural networks with billions of parameters on edge devices such as personal computers and smartphones, which is impractical. In addition, LLMs have shown strong reliability issues (e.g., adversarial vulnerability, biased predictions, hallucinations, etc.), further impeding their applications in practice. These challenges have provided a strong impetus for model compression/robustness research in the era of LLMs, leading to the development of resource-efficient and robustness-enhancement methods. The research theme has brought challenges and opportunities to the deep learning community, requiring experts in deep learning, natural language processing, and computer vision to adopt interdisciplinary approaches to uncover the underlying mechanisms and develop practical solutions. By studying and addressing the efficiency and reliability challenges associated with LLMs, we could better enable practitioners to construct practical LLMs across various domains. The workshop will bring together researchers and practitioners from the machine learning communities to explore the latest advances and challenges in building practical learning methods, with a special focus on the efficiency and reliability of LLMs. The workshop consists of invited talks by leading experts in the field and contributed talks and poster sessions featuring the latest research.

Date: 27 June, 2024
Time: 3:30 pm – 6:00 pm
Venue: Melati Ballroom 4003-4, 4103-4

Website: Practical-DL
Call for Paper: Practical-DL CFP

Haotong Qin (Beihang University, China)
Ruihao Gong (SenseTime Research)
Jiakai Wang (Zhongguancyn Laboratory, China)
Qing Guo (A*STAR, Singapore)
Daquan Zhou (ByteDance, USA)
Olivera Kotevska (Oak Ridge National Laboratory, USA)
Aishan Liu (Beihang University, China)
Zhen Dong (UC Berkeley, USA)
Shanghang Zhang (Peking University, China)
Huanrui Yang (UC Berkeley, USA)
Xianglong Liu (Beihang University, China)
Dacheng Tao (Nanyang Technological University, Singapore)

In recent years, applications of machine learning for science and technology have been growing under the name Scientific Machine Learning (SciML). Techniques of this field are expected to accelerate essential processes in industry, such as physical simulations and the discovery of new drugs. Although more and more researchers and industrial companies are paying attention to this research area, there are still few opportunities to communicate with each other. This workshop aims to introduce the latest results in this field and to provide opportunities to meet, interact and start collaboration. This workshop also aims to present the latest research in scientific machine learning to clarify the benefits of the techniques in this field to the audience, provide a platform for the research groups in this area to meet, interact and start closer collaboration.

Date: 25 June, 2024
Time: 2:30 pm – 6:00 pm
Venue: Melati Junior 4011

Website: SMLIA2024

Takaharu Yaguchi (Kobe University, Japan)
Naonori Ueda (RIKEN, Japan)
Mizuka Komatsu (Kobe University, Japan)
Yuhan Chen (Kobe University, Japan)
Baige Xu (Kobe University, Japan)

Tensor Networks (TNs), with deep roots in quantum physics, chemistry, and applied mathematics, have demonstrated exceptional performance in handling high-dimensional data, generating multiway structured data, optimising neural network structures and etc. Recently, it is further developed as a potential driving force in advances of machine learning (ML) and artificial intelligence (AI), particularly in crucial areas such as quantum machine learning, trustworthy machine learning, and interpretable machine learning. This workshop aims to bring together researchers from multidisciplinary and discuss not only the fundamental challenges in TNs such as TN structure optimisation, efficient algorithms, and robustness, but also their extended application in addressing key challenges within machine learning, spanning domains such as efficiency, interpretability, reliability and etc. These efforts not only advance a deeper understanding of machine learning models and enhance their reliability but also ensure they meet the demand for resource efficiency, thereby increasing trustworthiness. Furthermore, our workshop is expected to unleash the potential of TNs to open new directions for future research in quantum-inspired machine learning.

Date: 25 June, 2024
Time: 2:30 pm – 6:00 pm
Venue: Lotus Junior 4E

Website: TMME CAI 2024

Qibin Zhao (RIKEN AIP, Japan)
Guoxu Zhou (Guangdong University of Technology, China)
Cesar F. Caiafa (University of Buenos Aires, Argentina)
Tatsuya Yokota (Nagoya Institute of Technology, Japan)
Yannan Chen (South China Normal University, China)
Yuning Qiu (RIKEN, Japan)
Andong Wang (RIKEN AIP, Japan)

This workshop will include an introduction to Gen AI followed by and a hands-on lab session on using Foundation Models and Generative AI.

During the workshop, the participants will be given a good understanding of the concepts and build a couple of custom generative AI applications leveraging powerful foundation models and other AI services available in AWS.

Participants can expect to build:
1) A custom app that integrates to Bedrock models using APIs. We will make custom code, dependencies, configuration, and instructions available for the workshop. Participants are required to bring their own laptops for the lab session.
2) A custom app that uses the RAG (Retrieval-augmented generation) approach and leverages additional services like knowledge base, vector db, etc. in addition to the foundation models

Date: 26 June, 2024
Time: 2:30 pm – 6:00 pm
Venue: Melati Junior 4111

Priscilla Amalraj (IEEE,USA)
Ashok Sivathapandian (IEEE, USA)
Vincent Oh (AWS)

Workshop 1: How to Train Foundation Language Models? 

Training a foundation language model is a complex process that involves many distinct pieces.

In this talk important aspects regarding identifying data sources, constructing pretraining datasets, and choosing a model architecture to consider when developing foundation models will be discussed.

Workshop 2: How to Train Foundation Language Models? 

In recent years, AI has made remarkable advancements, but with these advancements come a plethora of challenges that developers must navigate. This workshop aims to delve into some of the challenges facing AI engineers and scientists today, offering insights and strategies to overcome them.

We will focus on exploring the challenges of LLM data engineering, its potential impact for industries and how they can be optimized with high-performance computing.

We will also discuss how we can effectively leverage and apply pre-trained foundational models to existing applications. Most importantly, effective strategies to optimize LLM data engineering and accelerating model training and inference with high-performance computing. Finally, the potential of small language models in the era of LLMs.

Workshop 3: Hands-on session – ASPIRE 2A+ Explore Singapore’s Latest Supercomputer Built for AI

Explore ASPIRE 2A+, Singapore’s latest supercomputer built for running AI, Machine Learning, and Large Language Model workloads. Equipped with the latest Nvidia DGX H100 SuperPOD, ASPIRE2A+ offers accelerated infrastructure and scalable performance for the most demanding AI workloads.

Led by Nvidia’s expert trainer, this session will showcase some of ASPIRE 2A+ capabilities and its ready-to-use, fully supported software (Nvidia base command software stack) that speeds developer success including the capabilities.

Optimise Llama2 model using TensorRT-LLMMotivation:

Large language models (LLMs) offer incredible new capabilities, expanding the frontier of what is possible with AI. However, their large size and unique execution characteristics can make them difficult to use in cost-effective ways. Optimise LLM model using TensorRT-LLM, calibrate for lower precision with high accuracy, and deploy LLM at-scale. Increase inference performance at lower energy usage.

Pre-registration is required for this workshop (limited to 40 pax).

Date: 25 June, 2024
Time: 2:30 pm – 6:00 pm
Venue: Melati Junior 4111