- A training round or "epoch" typically visits each training example once.
- The following settings for MaxTrainingRounds can be given:
Automatic train for approximately 20 seconds (default) n train for the given number of rounds, where a round is a traversal of the entire dataset Quantity[r,"unit"] train for a specific number of seconds, minutes or hours
- With the default specification MaxTrainingRounds->Automatic, the number of training rounds to use is calculated based on the time taken to evaluate one batch of examples and will typically train for less than a minute on networks with fewer than 1000 weight parameters. For nets that operate on variable-length inputs, the chosen value can vary substantially.
- MaxTrainingRounds->Automatic will never train for less than 10 rounds or more than 10,000 rounds.
Introduced in 2016