Related papers: RollMux: Phase-Level Multiplexing for Disaggregate…

Multi-Task Learning (MTL) aims to boost predictive performance by sharing information across related tasks, yet conventional methods often suffer from negative transfer when unrelated or noisy tasks are forced to share representations. We…

Agentic reinforcement learning (RL) has emerged as a transformative workload in cloud clusters, enabling large language models (LLMs) to solve complex problems through interactions with real world. However, unlike traditional RL, agentic RL…

Deep learning recommendation models (DLRM) rely on large embedding tables to manage categorical sparse features. Expanding such embedding tables can significantly enhance model performance, but at the cost of increased GPU/CPU/memory usage.…

Looped computation shows promise in improving the reasoning-oriented performance of LLMs by scaling test-time compute. However, existing approaches typically require either training recurrent models from scratch or applying disruptive…

Federated Learning (FL) enables collaborative model training across multiple clients while preserving data privacy. Traditional FL methods often use a global model to fit all clients, assuming that clients' data are independent and…

Large-scale ML training jobs are frequently interrupted by hardware and software anomalies, failures, and management events. Existing solutions like checkpoint-restart or runtime reconfiguration suffer from long downtimes and degraded…

An efficient MCMC algorithm is presented to cluster the nodes of a network such that nodes with similar role in the network are clustered together. This is known as block-modelling or block-clustering. The model is the stochastic blockmodel…

Dramatic increases in the capabilities of neural network models in recent years are driven by scaling model size, training data, and corresponding computational resources. To develop the exceedingly large networks required in modern…

Disaggregated memory architectures provide benefits to applications beyond traditional scale out environments, such as independent scaling of compute and memory resources. They also provide an independent failure model, where computations…

Training deep reinforcement learning (DRL) models usually requires high computation costs. Therefore, compressing DRL models possesses immense potential for training acceleration and model deployment. However, existing methods that generate…

Training large language models (LLMs) for pretraining or adapting to new tasks and domains has become increasingly critical as their applications expand. However, as the model and the data sizes grow, the training process presents…

Modern Deep Learning (DL) models have grown to sizes requiring massive clusters of specialized, high-end nodes to train. Designing such clusters to maximize both performance and utilization--to amortize their steep cost--is a challenging…

Reinforcement learning with verifiable rewards (RLVR) has proven effective in eliciting complex reasoning in large language models (LLMs). However, standard RLVR training often leads to excessively verbose processes (in reasoning tasks) and…

Thanks to their remarkable flexibility, diffusion models and flow models have emerged as promising candidates for policy representation. However, efficient reinforcement learning (RL) upon these policies remains a challenge due to the lack…

Recent advances in reinforcement-learning research have demonstrated impressive results in building algorithms that can out-perform humans in complex tasks. Nevertheless, creating reinforcement-learning systems that can build abstractions…

Reinforcement learning (RL) plays a central role in large language model (LLM) post-training. Among existing approaches, Group Relative Policy Optimization (GRPO) is widely used, especially for RL with verifiable rewards (RLVR) fine-tuning.…

The recent progress made in large language models (LLMs) has brought tremendous application prospects to the world. The growing model size demands LLM training on multiple GPUs, while data parallelism is the most popular distributed…

Multimodal Large Language Models (MLLMs) have achieved remarkable advances by integrating text, image, and audio understanding within a unified architecture. However, existing distributed training frameworks remain fundamentally data-blind:…

With the rapid advancement of large language models (LLMs), reinforcement learning (RL) has emerged as a pivotal methodology for enhancing the reasoning capabilities of LLMs. Unlike traditional pre-training approaches, RL encompasses…

Modern cloud platforms increasingly host large-scale deep learning (DL) workloads, demanding high-throughput, low-latency GPU scheduling. However, the growing heterogeneity of GPU clusters and limited visibility into application…