Mixture of Experts: Advanced Architecture, Training, and Scaling
Chapter 1: Foundations of Sparse Expert Models
Conditional Computation Principles
The Sparse MoE Approach
Contrasting Dense vs. Sparse Activation
Mathematical Formulation of Basic MoE Layers
Chapter 2: Advanced MoE Architectures
Designing Effective Gating Networks
Hierarchical MoE Structures
Router Architectures: Linear, Non-Linear, Attention-Based
Expert Capacity and Sizing Considerations
Stabilization Techniques for Routers
Hands-on Practical: Implementing Custom Gating Mechanisms
Chapter 3: Training Dynamics and Optimization
The Load Balancing Problem in MoE
Auxiliary Loss Functions for Load Balancing
Router Optimization Strategies
Handling Dropped Tokens
Expert Specialization Collapse and Prevention
Impact of Optimizer Choice and Hyperparameters
Hands-on Practical: Implementing and Tuning Load Balancing Losses
Chapter 4: Scaling MoE Models: Distributed Training
Challenges in Distributed MoE Training
Expert Parallelism: Distributing Experts Across Devices
Integrating Expert Parallelism with Data Parallelism
All-to-All Communication Patterns
Pipeline Parallelism for MoE Models
Communication Optimization Techniques (e.g., Overlapping)
Frameworks and Libraries for Distributed MoE (e.g., DeepSpeed, Tutel)
Practice: Configuring Distributed MoE Training
Chapter 5: Inference Optimization and Deployment
Inference Challenges with Sparse Models
Batching Strategies for MoE Inference
Model Compression Techniques for MoE
Hardware Acceleration Considerations
Router Caching and Optimization
Deployment Patterns for Large Sparse Models
Hands-on Practical: Profiling MoE Inference
Handling Dropped Tokens
Was this section helpful?
- Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer, Noam Shazeer, Azalia Mirhoseini, Krzysztof Maziarz, Andy Davis, Quoc Le, Geoffrey Hinton, Jeff Dean, 2017 arXiv preprint arXiv:1701.06538 DOI: 10.48550/arXiv.1701.06538 - Introduces the foundational sparsely-gated MoE layer, Top-K routing, and the auxiliary loss for load balancing, which are central to understanding token dropping.
- Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity, William Fedus, Barret Zoph, and Noam Shazeer, 2021 Journal of Machine Learning Research, Vol. 23 DOI: 10.48550/arXiv.2101.03961 - Details the
capacity_factorand practical considerations for training and scaling MoE models, directly addressing the management of expert capacity and token dropping. - GLaM: Efficient Scaling of Language Models with Mixture-of-Experts, Nan Du, Yanping Huang, Andrew M. Dai, Simon Tong, Dmitry Lepikhin, Yuanzhong Xu, Maxim Krikun, Yanqi Zhou, Adams Wei Yu, Orhan Firat, Barret Zoph, Liam Fedus, Maarten Bosma, Zongwei Zhou, Tao Wang, Yu Emma Wang, Kellie Webster, Marie Pellat, Kevin Robinson, Kathleen Meier-Hellstern, Toju Duke, Lucas Dixon, Kun Zhang, Quoc V Le, Yonghui Wu, Zhifeng Chen, Claire Cui, 2022 ICML 2022 DOI: 10.48550/arXiv.2112.06905 - Presents an MoE architecture focusing on efficiency and scaling, addressing how resource allocation and capacity limits impact token processing and potential dropping.