UniMax: Fairer and More Effective Language Sampling for Large-Scale Multilingual Pretraining.
ICLR 2023（2023）
关键词
Keywords: multilingual,pretraining,language models,language sampling,language distribution,low-resource languages,overfitting
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要