Skip to contents

Allow resume model training from a specific checkpoint

Usage

luz_callback_resume_from_checkpoint(
  path,
  ...,
  restore_model_state = TRUE,
  restore_records = FALSE,
  restore_optimizer_state = FALSE,
  restore_callbacks_state = FALSE
)

Arguments

path

Path to the checkpoint that you want to resume.

...

currently unused.

restore_model_state

Wether to restore the model state from the checkpoint.

restore_records

Wether to restore records from the checkpoint.

restore_optimizer_state

Wether to restore the optimizer state from the checkpoint.

restore_callbacks_state

Wether to restore the callbacks state from the checkpoint.

Note

Read the checkpointing article in the pkgdown website for more information.