Basic MNIST Example with Ignite and ignite.contrib module

ported from pytorch-examples

Basic neural network training with Ignite and various built-in loggers from ignite.contrib:

• TQDM progress bar
• Tensorboard
• Visdom

Usage:

Requirements:

• torchvision: pip install torchvision
• tqdm: pip install tqdm

Logging with TQDM progress bar

Run the example:

python mnist_with_tqdm_logger.py

Logging with TensorboardX or torch.utils.tensorboard

Example with training and validation monitoring using Tensorboard.

Requirements:

• torchvision: pip install torchvision
• Optionally TensorboardX: pip install tensorboardX
• Tensorboard: pip install tensorboard

Optionally, user can install pynvml package on Python 3 and log GPU information: used memory, utilization.

Usage:

Run the example:

python mnist_with_tensorboard_logger.py --log_dir=/tmp/tensorboard_logs

Start tensorboard:

tensorboard --logdir=/tmp/tensorboard_logs/

Logging with Visdom

Example with training and validation monitoring using Visdom

Requirements:

• torchvision: pip install torchvision
• Visdom: pip install visdom

Usage:

Run the example:

python mnist_with_visdom_logger.py