PaDeLLM-NER: Parallel Decoding in Large Language Models for Named Entity Recognition Jinghui Lu1, Ziwei Y ang 1, Y anjie Wang

Neural Information Processing Systems 

The main cause of high latency in LLMs is the sequential decoding process, which autoregressively generates all labels and mentions for NER, significantly increase the sequence length.

Similar Docs  Excel Report  more

TitleSimilaritySource
None found