Reload a model from a Shared/Distributed checkpoint
| 489 | |
| 490 | // Reload a model from a Shared/Distributed checkpoint |
| 491 | bool checkpoint::reload_model(model* m) |
| 492 | { |
| 493 | lbann::utils::grid_manager grid_raii(m->get_comm()->get_trainer_grid()); |
| 494 | return open_latest_checkpoint( |
| 495 | *(m->get_comm()), |
| 496 | "Reloading Model " + m->get_name(), |
| 497 | get_active_trainer().get_name(), |
| 498 | get_active_training_algorithm().get_type(), |
| 499 | [m](persist& p_ref) { return m->load_from_checkpoint_shared(p_ref); }, |
| 500 | [m](persist& p_ref) { return m->load_from_checkpoint_distributed(p_ref); }); |
| 501 | } |
| 502 | |
| 503 | // Reload a model from a Shared/Distributed checkpoint |
| 504 | bool checkpoint::reload_trainer(trainer* t) |
nothing calls this directly
no test coverage detected