BriefGPT.xyz
Oct, 2023
分治与统治：复杂自然语言处理任务的多变压器架构
Divide et Impera: Multi-Transformer Architectures for Complex NLP-Tasks
HTML
PDF
Solveig Helland, Elena Gavagnin, Alexandre de Spindler
TL;DR
采用细分子任务和多模型联合的方法，简化了微调数据集的编制，增加了整体可控性，并在减少性别偏见的复杂任务中展示了比单一模型更好的性能。
Abstract
The growing capabilities of
transformer models
pave the way for solving increasingly complex NLP tasks. A key to supporting application-specific requirements is the ability to fine-tune. However, compiling a
fine-tuning
→