This page gives a run-down of many features FluxTraining.jl brings to the table. Most features are implemented as callbacks and using them is as simple as passing the callback when constructing a Learner and training with fit!:
cb = CoolFeature🕶️Callback()
learner = Learner(model, lossfn; callbacks=[cb], data=(trainiter, validiter))
fit!(learner, nepochs)By default, Learner will track only the loss function. You can track other metric with the Metrics callback. See also Metric, AbstractMetric.
The Scheduler callback takes care of hyperparameter scheduling. See the Hyperparameter scheduling tutorial and also Scheduler, Schedule, HyperParameter.
For logging, use the logging callbacks:
They each can have multiple logging backends, but right now the only ones implemented in FluxTraining.jl are TensorBoardBackend and MLFlowBackend. See also LoggerBackend, log_to, and Loggables.Loggable.
There is also an external package Wandb.jl that implements a logging backend for Weights&Biases.
Use the Checkpointer callback to create model checkpoints after every epoch.
Use EarlyStopping to stop when a stopping criterion is met. Supports all criteria in EarlyStopping.jl.