o „o™i ã@sDddlmZmZddlmZmZmZddlmZGdd„deƒZ dS)é)ÚABCÚabstractmethod)ÚAnyÚDictÚOptional)Ú_PATHc @s„eZdZdZeddeeefdede eddfdd„ƒZ edded e edeeeffd d „ƒZ ededdfd d „ƒZ ddd„Z dS)Ú CheckpointIOa~Interface to save/load checkpoints as they are saved through the ``Strategy``. .. warning:: This is an :ref:`experimental ` feature. Typically most plugins either use the Torch based IO Plugin; ``TorchCheckpointIO`` but may require particular handling depending on the plugin. In addition, you can pass a custom ``CheckpointIO`` by extending this class and passing it to the Trainer, i.e ``Trainer(plugins=[MyCustomCheckpointIO()])``. .. note:: For some plugins, it is not possible to use a custom checkpoint plugin as checkpointing logic is not modifiable. NÚ checkpointÚpathÚstorage_optionsÚreturncCódS)a'Save model/training states as a checkpoint file through state-dump and file-write. Args: checkpoint: dict containing model and trainer state path: write-target path storage_options: Optional parameters when saving the model/training states. N©)Úselfr r r rrú]/home/ubuntu/.local/lib/python3.10/site-packages/lightning/fabric/plugins/io/checkpoint_io.pyÚsave_checkpoint&ózCheckpointIO.save_checkpointÚ map_locationcCr )aKLoad checkpoint from a path when resuming or loading ckpt for test/validate/predict stages. Args: path: Path to checkpoint map_location: a function, :class:`torch.device`, string or a dict specifying how to remap storage locations. Returns: The loaded checkpoint. Nr)rr rrrrÚload_checkpoint1rzCheckpointIO.load_checkpointcCr )ziRemove checkpoint file from the filesystem. Args: path: Path to checkpoint Nr)rr rrrÚremove_checkpoint>rzCheckpointIO.remove_checkpointcCr )z.This method is called to teardown the process.Nr)rrrrÚteardownGrzCheckpointIO.teardown)N)r N)Ú__name__Ú __module__Ú __qualname__Ú__doc__rrÚstrrrrrrrrrrrrrs* & rN) ÚabcrrÚtypingrrrÚ lightning.fabric.utilities.typesrrrrrrÚs