pytorch_template

A documented template for a supervised deep learning project in PyTorch. Use datasets.py to define custom datasets and dataloaders, models.py to define architectures, and fill in the hyperparameters in pytorch-training-loop.ipynb to run through the training loop, with timing and validation metrics computed at each epoch.