NEURIPS 2025 "long sequence training" Papers
2 papers found
StarTrail: Concentric Ring Sequence Parallelism for Efficient Near-Infinite-Context Transformer Model Training
Ziming Liu, Shaoyu Wang, Shenggan Cheng et al.
NEURIPS 2025posterarXiv:2407.00611
2
citations
StreamBP: Memory-Efficient Exact Backpropagation for Long Sequence Training of LLMs
Qijun Luo, Mengqi Li, Lei Zhao et al.
NEURIPS 2025posterarXiv:2506.03077
1
citations