6 papers
Improving Data and Reward Design for Scientific Reasoning in Large Language Models
Zijie Chen, Zhenghao Lin, Xiao Liu +3
Solving open-ended science questions remains challenging for large language models, particularly due to inherently unreliable supervision and evaluation. The bottleneck lies in the…
MSign: An Optimizer Preventing Training Instability in Large Language Models via Stable Rank Restoration
Lianhai Ren, Yucheng Ding, Xiao Liu +2
Training instability remains a critical challenge in large language model (LLM) pretraining, often manifesting as sudden gradient explosions that waste significant computational re…
Sigma-MoE-Tiny Technical Report
Qingguo Hu, Zhenghao Lin, Ziyue Yang +12
Mixture-of-Experts (MoE) has emerged as a promising paradigm for foundation models due to its efficient and powerful scalability. In this work, we present Sigma-MoE-Tiny, an MoE la…
SIGMA: An AI-Empowered Training Stack on Early-Life Hardware
Lei Qu, Lianhai Ren, Peng Cheng +12
An increasing variety of AI accelerators is being considered for large-scale training. However, enabling large-scale training on early-life AI accelerators faces three core challen…
Beyond Sunk Costs: Boosting LLM Pre-training Efficiency via Orthogonal Growth of Mixture-of-Experts
Ruizhe Wang, Yucheng Ding, Xiao Liu +5
As the computational demands for pre-training Large Language Models (LLMs) continue to surge, the need for efficient training paradigms becomes critical. Despite the vast resources…
Training Matryoshka Mixture-of-Experts for Elastic Inference-Time Expert Utilization
Yaoxiang Wang, Qingguo Hu, Yucheng Ding +6
Mixture-of-Experts (MoE) has emerged as a promising paradigm for efficiently scaling large language models without a proportional increase in computational cost. However, the stand…