REVIEW 22 cited by
FastMoE: A Fast Mixture-of-Expert Training System
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
FastMoE: A Fast Mixture-of-Expert Training System
read the original abstract
Mixture-of-Expert (MoE) presents a strong potential in enlarging the size of language model to trillions of parameters. However, training trillion-scale MoE requires algorithm and system co-design for a well-tuned high performance distributed training system. Unfortunately, the only existing platform that meets the requirements strongly depends on Google's hardware (TPU) and software (Mesh Tensorflow) stack, and is not open and available to the public, especially GPU and PyTorch communities. In this paper, we present FastMoE, a distributed MoE training system based on PyTorch with common accelerators. The system provides a hierarchical interface for both flexible model design and easy adaption to different applications, such as Transformer-XL and Megatron-LM. Different from direct implementation of MoE models using PyTorch, the training speed is highly optimized in FastMoE by sophisticated high-performance acceleration skills. The system supports placing different experts on multiple GPUs across multiple nodes, enabling enlarging the number of experts linearly against the number of GPUs. The source of FastMoE is available at https://github.com/laekov/fastmoe under Apache-2 license.
Forward citations
Cited by 22 Pith papers
-
BrainFIBRE: A Foundation Model via Information Decomposition for Brain Microstructure
BrainFIBRE pretrains a five-expert Mixture-of-Experts model on NODDI-derived microstructural maps and outperforms prior deep models on age, sex, cerebrovascular, neurodegenerative, and cognitive prediction.
-
Language-Assisted Super-Resolution from Real-World Low-Resolution Patches
LA-SR redefines unpaired super-resolution in language space by projecting images into a semantically rich representation and applying vision-language model guided losses to handle real-world degradations extracted fro...
-
ViBE: Co-Optimizing Workload Skew and Hardware Variability for MoE Serving
ViBE co-optimizes expert placement with measured GPU performance variability in MoE inference to cut execution-time imbalance, delivering 14% better SLO attainment and up to 45% lower P90 TTFT.
-
MoE-Prefill: Zero Redundancy Overheads in MoE Prefill Serving
MoE-Prefill achieves 1.35-1.59x higher throughput for prefill-only MoE serving by using asynchronous expert parallelism to overlap weight AllGather with computation and prefix-aware routing with true-FLOPs tracking.
-
From Tensor Buffer to Distributed Memory Hierarchy: A Survey of KV Cache Management for LLM Serving
KV-cache serving systems concentrate into five archetypes under a four-axis taxonomy, with ownership explaining residual distributed design variance and seven measurement gaps blocking next steps.
-
Relax Within, Balance Across: Geometry-Guided Load Balancing for Vision-Language Mixture-of-Experts
The standard mixed image-text load-balancing loss lets modality errors cancel at one token mix; ReBA balances each modality separately and per image, reducing load across resolutions and tiles with comparable accuracy.
-
On the Design of Mixture-of-Experts for Dynamic Gaussian Splatting
MoDE (joint canonical deformation experts) and MoE-GS (independent experts plus volume-aware pixel routing) improve dynamic Gaussian Splatting by composing complementary motion priors under different integration constraints.
-
On the Design of Mixture-of-Experts for Dynamic Gaussian Splatting
Two MoE integration strategies (joint canonical MoDE vs. independent-then-route MoE-GS) improve dynamic Gaussian Splatting by composing complementary deformation priors.
-
EMO: Frustratingly Easy Progressive Training of Extendable MoE
EMO progressively expands the expert pool in MoE models during training to match fixed-expert performance with improved wall-clock efficiency.
-
Fast MoE Inference via Predictive Prefetching and Expert Replication
Dynamic replication of predicted overloaded experts in MoE models achieves near-100% GPU utilization and up to 3x faster inference while retaining 90-95% of baseline performance.
-
Hierarchical Mixture-of-Experts with Two-Stage Optimization
Hi-MoE uses two-level hierarchical routing objectives to enforce group-level balance while promoting within-group specialization, yielding better perplexity and expert utilization than prior MoE baselines in NLP and v...
-
MoE-Prefill: Zero Redundancy Overheads in MoE Prefill Serving
ZeRO-Prefill achieves 1.35-1.59x higher throughput for MoE prefill serving by replacing per-layer activation AllToAll with overlapped asynchronous weight AllGather and prefix-aware routing.
-
AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs
AGoQ delivers up to 52% lower memory use and 1.34x faster training for 8B-32B LLaMA models by using near-4-bit adaptive activations and 8-bit gradients while preserving pretraining convergence and downstream accuracy.
-
TabICL: A Tabular Foundation Model for In-Context Learning on Large Data
TabICL scales in-context learning to large tabular data via column-then-row attention for row embeddings followed by a transformer, matching TabPFNv2 speed and performance while outperforming it and CatBoost on datase...
-
DeepSpeed Ulysses: System Optimizations for Enabling Training of Extreme Long Sequence Transformer Models
DeepSpeed-Ulysses keeps communication volume constant for sequence-parallel attention when sequence length and device count scale together, delivering 2.5x faster training on 4x longer sequences than prior SOTA.
-
BrainFIBRE: A Foundation Model via Information Decomposition for Brain Microstructure
BrainFIBRE presents a foundation model for brain microstructure that applies self-supervised partial information decomposition on NODDI maps to disentangle unique, synergistic, and redundant information and reports st...
-
Language-Assisted Super-Resolution from Real-World Low-Resolution Patches
LA-SR extracts real LR patches from depth-varying regions in single images and uses vision-language models with linguistic content and quality losses for unpaired super-resolution.
-
AGoQ: Activation and Gradient Quantization for Memory-Efficient Distributed Training of LLMs
AGoQ cuts LLM training memory by up to 52% and speeds it up by 1.34x using tailored 4-bit activations and 8-bit gradients with special communication, matching baseline accuracy on LLaMA models.
-
EMO: Frustratingly Easy Progressive Training of Extendable MoE
EMO progressively expands the expert pool in MoE models using scaling-law-derived token budgets per stage, matching fixed-expert performance while cutting wall-clock time and GPU cost.
-
From System 1 to System 2: A Survey of Reasoning Large Language Models
The survey organizes the shift of LLMs toward deliberate System 2 reasoning, covering model construction techniques, performance on math and coding benchmarks, and future research directions.
-
A Survey of Large Language Models
This survey reviews the background, key techniques, and evaluation methods for large language models, emphasizing emergent abilities that appear at large scales.
-
A Comprehensive Overview of Large Language Models
A survey paper providing an overview of Large Language Models, their background, and recent advances in the field.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.