← Glossary
Glossary
CUDA
NVIDIA’s platform for running general-purpose computation on the GPU instead of the CPU — what PyTorch/TensorFlow use under the hood to train and run models fast. Most "out of memory" and device-mismatch errors are CUDA-level.
Related articles
Fix CUDA Error: Device-Side Assert Triggered in PyTorchFix No module named 'faiss' and FAISS ntotal != k ErrorsFAISS nprobe, IndexIVFPQ and GPU Memory Errors FixedHow Neural Networks Work: Forward Pass & Backprop in NumPyHugging Face No Space Left on Device (Errno 28) FixFix Hugging Face CUDA Out of Memory at Model LoadFix HuggingFace Tokenizer Padding and pad_token ErrorsFix PyTorch 'Expected All Tensors to Be on the Same Device'Recommender Systems That Automate Campaign Execution: NLP and Collaborative Filtering in Practicetorch.cuda.OutOfMemoryError: How to Fix GPU Out of Memory in PyTorchPyTorch: Expected Scalar Type Long but Found Float
Related tools