Logging

Syllabus currently support logging through TensorBoard or Weights and Biases through the log_metrics function for curricula. This method collects logs and passes them along to the base Curriculum class, which performs the actual logging. As such, you must call super().log_metrics() in your custom curriculum’s log_metrics function to ensure that the logs are propagated and saved.

syllabus.core.curriculum_base.Curriculum.log_metrics(self, writer, logs: List[Dict], step: int | None = None, log_n_tasks: int = 1)

Log the task distribution to the provided writer.

Parameters:

• writer – Tensorboard summary writer or wandb object

• logs – Cumulative list of logs to write

• step – Global step number

• log_n_tasks – Maximum number of tasks to log, defaults to 1. Use -1 to log all tasks.

Returns:

Updated logs list