Guardat en:
Dades bibliogràfiques
Autors principals: Wang, Jinbo, Wang, Mingze, Zhou, Zhanpeng, Yan, Junchi, E, Weinan, Wu, Lei
Format: Preprint
Publicat: 2025
Matèries:
Accés en línia:https://arxiv.org/abs/2502.19002
Etiquetes: Afegir etiqueta
Sense etiquetes, Sigues el primer a etiquetar aquest registre!
Taula de continguts:
  • Transformers consist of diverse building blocks, such as embedding layers, normalization layers, self-attention mechanisms, and point-wise feedforward networks. Thus, understanding the differences and interactions among these blocks is important. In this paper, we uncover a clear Sharpness Disparity across these blocks, which emerges early in training and intriguingly persists throughout the training process. Motivated by this finding, we propose Blockwise Learning Rate (LR), a strategy that tailors the LR to each block's sharpness, accelerating large language model (LLM) pre-training. By integrating Blockwise LR into AdamW, we consistently achieve lower terminal loss and nearly $2\times$ speedup compared to vanilla AdamW. We demonstrate this acceleration across GPT-2 and LLaMA, with model sizes ranging from 0.12B to 2B and datasets of OpenWebText, MiniPile, and C4. Finally, we incorporate Blockwise LR into Adam-mini (Zhang et al., 2024), a recently proposed memory-efficient variant of Adam, achieving a combined $2\times$ speedup and $2\times$ memory saving. These results underscore the potential of exploiting the sharpness disparity to improve LLM training.