Skip to main content

PyTorch Ignite TensorboardLogger

The cifar_ignite.py example script integrates ClearML into code that uses PyTorch Ignite.

The example script does the following:

  • Trains a neural network on the CIFAR-10 dataset for image classification.
  • Creates a ClearML Task named image classification CIFAR10 in the examples project.
  • Calls the Task.connect method to track task configuration.
  • Uses ignite's TensorboardLogger and attaches handlers to it. See TensorboardLogger.

ClearML's automatic logging captures information and outputs logged with TensorboardLogger.

Hyperparameters​

Parameters are explicitly reported to ClearML using the Task.connect method.

params = {'number_of_epochs': 20, 'batch_size': 64, 'dropout': 0.25, 'base_lr': 0.001, 'momentum': 0.9, 'loss_report': 100}
params = task.connect(params)  # enabling configuration override by clearml

The hyperparameter configurations can be viewed in the WebApp in the task's CONFIGURATION tab.

Hyperparameters Hyperparameters

Ignite TensorboardLogger​

TensorboardLogger is a handler to log metrics, parameters, and gradients when training a model. When ClearML is integrated into a script which uses TensorboardLogger, all information logged through the handler is automatically captured by ClearML.

Scalars​

ClearML automatically captures scalars logged through TensorboardLogger.

View the scalars in the task's page in the ClearML Web UI, in SCALARS.

Task scalars Task scalars

Model Snapshots​

ClearML automatically captures the model logged with Torch, and saves it as an artifact.

View saved snapshots in the task's ARTIFACTS tab.

Task models Task models

To view the model, in the ARTIFACTS tab, click the model name (or download it).

Model details Model details

Debug Samples​

ClearML automatically tracks images logged to TensorboardLogger. They appear in DEBUG SAMPLES.

Debug Samples Debug Samples

Ignite ClearMLLogger​

PyTorch Ignite also supports a dedicated ClearMLLogger handler to log metrics, text, model / optimizer parameters, plots, and model checkpoints during training and validation.

For more information, see the PyTorch Ignite ClearMLLogger example.