CUDA memory issue

[mokby]

When training for about 10 epoch ,error occured, batch size is 2

CUDA out of memory. Tried to allocate 11.47 GiB (GPU 0; 11.99 GiB total capacity; 11.80 GiB already allocated; 0 bytes free; 23.34 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

When test, same error occured, batch size is 1

CUDA out of memory. Tried to allocate 11.47 GiB (GPU 0; 11.99 GiB total capacity; 11.80 GiB already allocated; 0 bytes free; 23.34 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

[mokby]

Solved, dataset problem.