클라우드 GPU
·
경진대회
https://www.runpod.io/ RunPod - The Cloud Built for AIDevelop, train, and scale AI models in one cloud. Spin up on-demand GPUs with GPU Cloud, scale ML inference with Serverless.www.runpod.iohttps://vast.ai/ Rent GPUs | Vast.aiReduce your cloud compute costs by 3-5X with the best cloud GPU rentals. Vast.ai's simple search interface allows fair comparison of GPU rentals from all providers.vast.ai..
결과 재현성을 위한 SEED 고정
·
경진대회
배경 dacon 대회에서 재현성 확인을 위해 같은 코드로 결과를 다시 내어 제출했는데, 결과가 계속 달라져서 원인을 찾아보니 data loader에도 seed를 설정해줘야 한다는 것이다. 아래 공식문서를 살펴봐도 된다. https://pytorch.org/docs/stable/notes/randomness.html Reproducibility — PyTorch 2.5 documentationReproducibility Completely reproducible results are not guaranteed across PyTorch releases, individual commits, or different platforms. Furthermore, results may not be reproducib..