9 papers
ScaleAcross Explorer: Exploring Communication Optimization for Scale-Across AI Model Training
Minghao Li, Alicia Golden, Samuel Hsia +14
The rapid scaling of large language model training requires distributing GPU resources across multiple data center buildings and regions. We refer to such paradigm as "scale-across…
Training LLMs with Fault Tolerant HSDP on 100,000 GPUs
Omkar Salpekar, Rohan Varma, Kenny Yu +20
Large-scale training systems typically use synchronous training, requiring all GPUs to be healthy simultaneously. In our experience training on O(100K) GPUs, synchronous training r…
The Llama 4 Herd: Architecture, Training, Evaluation, and Deployment Notes
Redacted by arXiv
This document consolidates publicly reported technical details about Metas Llama 4 model family. It summarizes (i) released variants (Scout and Maverick) and the broader herd conte…
Collective Communication for 100k+ GPUs
Min Si, Pavan Balaji, Yongzhou Chen +36
The increasing scale of large language models (LLMs) necessitates highly efficient collective communication frameworks, particularly as training workloads extend to hundreds of tho…
LLäMmlein: Transparent, Compact and Competitive German-Only Language Models from Scratch
Jan Pfister, Julia Wunderle, Andreas Hotho
We create two German-only decoder models, LLäMmlein 120M and 1B, transparently from scratch and publish them, along with the training data, for the German NLP research community t…
HALoS: Hierarchical Asynchronous Local SGD over Slow Networks for Geo-Distributed Large Language Model Training
Geon-Woo Kim, Junbo Li, Shashidhar Gandham +5
Training large language models (LLMs) increasingly relies on geographically distributed accelerators, causing prohibitive communication costs across regions and uneven utilization…