sliceout_training_transformers_cnns_faster_while_using_less_memory

SliceOut: Training Transformers and CNNs faster while using less memory

https://arxiv.org/abs/2007.10909

Transformer, CNN, Memory, Memory 절약, Dropout, Regulization

sliceout_training_transformers_cnns_faster_while_using_less_memory.txt · 마지막으로 수정됨: 2024/03/23 02:38 저자 127.0.0.1