cloud_io

Functions

atomic_save	Saves a checkpoint atomically, avoiding the creation of incomplete checkpoints.
get_filesystem	rtype AbstractFileSystem
load	Loads a checkpoint.

Utilities related to data saving/loading.

pytorch_lightning.utilities.cloud_io.atomic_save(checkpoint, filepath)

Saves a checkpoint atomically, avoiding the creation of incomplete checkpoints.

Parameters

• checkpoint (Dict[str, Any]) – The object to save. Built to be used with the dump_checkpoint method, but can deal with anything which torch.save accepts.

• filepath (Union[str, Path]) – The path to which the checkpoint will be saved. This points to the file that the checkpoint will be stored in.

Return type

None

pytorch_lightning.utilities.cloud_io.load(path_or_url, map_location=None)

Loads a checkpoint.

Parameters

• path_or_url (Union[IO, str, Path]) – Path or URL of the checkpoint.

• map_location (Union[str, Callable, device, Dict[Union[str, device], Union[str, device]], None]) – a function, torch.device, string or a dict specifying how to remap storage locations.

Return type

Any