Examples ======== We provide several examples using `ignite` to display how it helps to write compact and full-featured training loops in several lines of code: MNIST example ------------- Basic neural network training on MNIST dataset with/without :mod:`ignite.contrib` module: - `MNIST with ignite.contrib TQDM/Tensorboard/Visdom loggers `_ - `MNIST with native TQDM/Tensorboard/Visdom logging `_ These examples are ported from `pytorch/examples `_. Distributed examples -------------------- Training a ResNet on CIFAR10 in various configurations: 1) single gpu 2) single node multiple gpus 3) multiple nodes and multiple gpus 4) single or multiple TPUs - `CIFAR10 `_ - This example displays usage of :doc:`distributed` helper module. Other examples -------------- - `DCGAN `_ - plain Deep Convolution Generative Adversarial Networks training - `Reinforcement Learning `_ - Actor/Critic and Reinforce methods on Cart-Pole task - `Fast Neural Style `_ - Artistic style transfer implementation. These examples are ported from `pytorch/examples `_. Notebooks --------- - `Text Classification using Convolutional Neural Networks `_ - `Variational Auto Encoders `_ - `Training Cycle-GAN on Horses to Zebras with Nvidia/Apex `_ - `Another training Cycle-GAN on Horses to Zebras with Native Torch CUDA AMP `_ - `Finetuning EfficientNet-B0 on CIFAR100 `_ - `Convolutional Neural Networks for Classifying Fashion-MNIST Dataset `_ - `Hyperparameters tuning with Ax `_ - `Basic example of LR finder on MNIST `_ - `Benchmark mixed precision training on Cifar100: torch.cuda.amp vs nvidia/apex `_ - `MNIST training on a single TPU `_ All notebooks can be opened on Google Colab with a link: .. code-block:: text https://colab.research.google.com/github/pytorch/ignite/blob/master/examples/notebooks/ Reproducible trainings ---------------------- Inspired by `torchvision/references `_, we provide several reproducible baselines for vision tasks: - `ImageNet `_ - `Pascal VOC2012 `_ Features: - Distributed training with mixed precision by `nvidia/apex `_ - Experiments tracking with `MLflow `_ or `Polyaxon `_ or `ClearML `_