Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning.

Mengzhou Xia, Tianyu Gao, Zhiyuan Zeng, Danqi Chen

Browse the full ICLR paper archive.