CUDA problem on training

[OliverVarnce]

Hello dears!

I'm trying to execute train.py but have error
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 498.00 MiB (GPU 0; 10.92 GiB total capacity; 9.22 GiB already allocated; 401.81 MiB free; 9.78 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

changing batch size and PYTORCH_CUDA_ALLOC_CONF do not helps me.

Can you give me your things?

[planet-miemie]

将batchsize和input-size调小，或者gpu配置高一点就好了