8 papers
SkillRater: Untangling Capabilities in Multimodal Data
Naveen Sahi, Jeremy Dohmann, Armen Aghajanyan +1
Data curation methods typically assign samples a single quality score. We argue this scalar framing is fundamentally limited: when training requires multiple distinct capabilities,…
Improving MoE Compute Efficiency by Composing Weight and Data Sparsity
Maciej Kilian, Oleg Mkrtchyan, Luke Zettlemoyer +2
Mixture-of-Experts layers achieve compute efficiency through weight sparsity: each token activates only a subset of experts. Data sparsity, where each expert processes only a subse…
CoSMoEs: Compact Sparse Mixture of Experts
Patrick Huber, Akshat Shrivastava, Ernie Chang +3
Sparse Mixture of Expert (MoE) models are popular foundational architectures at large scale, however, under-explored at smaller sizes. Here, we show how to enable Compact Sparse Mi…
LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding
Mostafa Elhoushi, Akshat Shrivastava, Diana Liskovich +10
We present LayerSkip, an end-to-end solution to speed-up inference of large language models (LLMs). First, during training we apply layer dropout, with low dropout rates for earlie…
PrE-Text: Training Language Models on Private Federated Data in the Age of LLMs
Charlie Hou, Akshat Shrivastava, Hongyuan Zhan +5
On-device training is currently the most common approach for training machine learning (ML) models on private, distributed user data. Despite this, on-device training has several d…
CoDi: Conversational Distillation for Grounded Question Answering
Patrick Huber, Arash Einolghozati, Rylan Conway +6
Distilling conversational skills into Small Language Models (SLMs) with approximately 1 billion parameters presents significant challenges. Firstly, SLMs have limited capacity in t…