7 papers
Decoupled DiLoCo for Resilient Distributed Pre-training
Arthur Douillard, Keith Rush, Yani Donchev +14
Modern large-scale language model pre-training relies heavily on the single program multiple data (SPMD) paradigm, which requires tight coupling across accelerators. Due to this co…
VaultGemma: A Differentially Private Gemma Model
Amer Sinha, Thomas Mesnard, Ryan McKenna +18
We introduce VaultGemma 1B, a 1 billion parameter model within the Gemma family, fully trained with differential privacy. Pretrained on the identical data mixture used for the Gemm…
Communication-Efficient Language Model Training Scales Reliably and Robustly: Scaling Laws for DiLoCo
Zachary Charles, Gabriel Teston, Lucio Dery +5
As we scale to more massive machine learning models, the frequent synchronization demands inherent in data-parallel approaches create significant slowdowns, posing a critical chall…
Scaling Laws for Differentially Private Language Models
Ryan McKenna, Yangsibo Huang, Amer Sinha +9
Scaling laws have emerged as important components of large language model (LLM) training as they can predict performance gains through scale, and provide guidance on important hype…
Streaming DiLoCo with overlapping communication: Towards a Distributed Free Lunch
Arthur Douillard, Yanislav Donchev, Keith Rush +11
Training of large language models (LLMs) is typically distributed across a large number of accelerators to reduce training time. Since internal states and parameter gradients need…
DrJAX: Scalable and Differentiable MapReduce Primitives in JAX
Keith Rush, Zachary Charles, Zachary Garrett +2
We present DrJAX, a JAX-based library designed to support large-scale distributed and parallel machine learning algorithms that use MapReduce-style operations. DrJAX leverages JAX'…