RuntimeError: CUDA out of memory. Tried to allocate 126.00 MiB (GPU 0; 23.88 GiB total capacity; 23.02 GiB already allocated; 14.88 MiB free; 23.25 GiB reserved in total by PyTorch)
First I thought this is just because of the large batch size or large size of base model.
But the problem is that the error occurs sometimes and doesn't occur sometimes.
Even I didn't get this kind of error one time and got error later with the same image size, same batch size, same base model architecture.
I got an error as follows.
RuntimeError: CUDA out of memory. Tried to allocate 126.00 MiB (GPU 0; 23.88 GiB total capacity; 23.02 GiB already allocated; 14.88 MiB free; 23.25 GiB reserved in total by PyTorch)
First I thought this is just because of the large batch size or large size of base model. But the problem is that the error occurs sometimes and doesn't occur sometimes. Even I didn't get this kind of error one time and got error later with the same image size, same batch size, same base model architecture.
Thanks in advance.