Bases: object
Configuration to start training from weights of a previously saved checkpoint (weights only, no optimizer/scheduler state).
- Parameters:
-
-
load: bool = False
-
alias: Literal['best', 'latest'] = 'latest'
-
strict: bool = True
-
partial_load: bool = False
-
run_path: str = ''
-
callbacks: list
-
__init__(
- load=False,
- alias='latest',
- strict=True,
- partial_load=False,
- run_path='',
- callbacks=<factory>,
)
- Parameters:
-
- Return type:
None
