search

GoGoDuck912 / Self-Correction-Human-Parsing

An out-of-box human parsing representation extractor.
MIT License
1.03k stars 237 forks source link

CUDA problem on training #80

Open OliverVarnce opened 1 year ago

OliverVarnce commented 1 year ago

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 commented 1 year ago

将batchsize和input-size调小，或者gpu配置高一点就好了