nvalchemi.training.TrainingStrategy.save_checkpoint

TrainingStrategy.save_checkpoint(root_folder, *, checkpoint_index=-1)

Save this strategy as a restartable checkpoint.

Rather than pickling the strategy, this writes a spec-based checkpoint: model weights and architecture, optimizer and scheduler state, the training counters (step_count, epoch_count, batch_count, global_step_count), and the state of any CheckpointableHook are each serialized through their Pydantic specs, so a restart reconstructs the objects without executing arbitrary pickled code. The non-serializable pieces – training_fn and loss_target_assembler – are intentionally excluded and must be supplied again at load time.

Checkpoints are indexed within root_folder and tracked by a manifest. checkpoint_index=-1 (the default) auto-increments from the latest manifest entry, so repeated calls accumulate 0, 1, 2, ..., while an explicit index overwrites that slot in place.

Parameters:

• root_folder (Path | str) – Root directory for checkpoint files.

• checkpoint_index (int, optional) – Checkpoint index to write. -1 auto-increments from the latest manifest index, or starts at 0 when no manifest exists.

Returns:

The checkpoint index that was written.

Return type:

int

See also

restore_checkpoint

Restore saved state into this strategy instance.

Notes

See Checkpointing in the training guide for the four categories of state a checkpoint captures and the developer requirements for custom models, schedules, and hooks.