ShannonAI/Self_Explaining_Structures_Improve_NLP_Models

CUDA out of memory

Opened this issue · 2 comments

Tried to allocate 20.00 MiB (GPU 0; 1.96 GiB total capacity; 1.24 GiB already allocated; 2.88 MiB free; 1.25 GiB reserved in total by PyTorch)

maybe you can use batch accumulate to decrease the batch size.

Tried to allocate 20.00 MiB (GPU 0; 1.96 GiB total capacity; 1.24 GiB already allocated; 2.88 MiB free; 1.25 GiB reserved in total by PyTorch)

你是在window上跑的嘛