Deep Learning Section 030
Datasets and DataLoaders
Getting data into the model quickly and correctly — the part that quietly decides how fast your training runs.
10 of 10 lessons published Three reading levels on every lesson
Lessons in order
Work top to bottom. Each lesson assumes the one above it.
- Writing your own Dataset
- What DataLoader actually does
- Writing a collate_fn
- num_workers, prefetching and the Windows spawn trap
- Samplers and weighted sampling
- IterableDataset for streams and huge files
- Image transforms with torchvision v2
- Datasets that do not fit in memory
- Surviving corrupt and missing files
- Finding out whether the GPU is waiting for data