"RuntimeError: CUDA out of memory. ~" in PyTorch (Windows) with CUDA 11.0 & NVIDIA RTX 3090
This error is related to the number of workers.
To solve this issue, reduce the number of workers in DataLoader
# Reference:
discuss.pytorch.org/t/cuda-out-of-memory-while-increase-num-workers-in-dataloader/101877
'LANGUAGE > PyTorch' 카테고리의 다른 글
| How to set Random Seed for PyTorch? (0) | 2021.05.17 |
|---|---|
| BiLSTM (0) | 2019.08.01 |
| PyTorch Examples (0) | 2019.08.01 |
| Random Seed Initialization for Reproducibility ... (0) | 2019.07.15 |
| Difference between model.eval() and torch.no_grad() (0) | 2019.07.15 |