types
Type hints, enums and protocols for the Bitfount libraries.
Classes
DistributedModelProtocol
class DistributedModelProtocol(*args, **kwargs):
Federated Model structural type.
This protocol should be implemented by classes that inherit from either
BitfountModel
, or both of _BaseModel
and DistributedModelMixIn
.
Ancestors
Methods
def apply_weight_updates( self, weight_updates: Sequence[_Weights],) ‑> Mapping[str, bitfount.types._TensorLike]:
Defined in DistributedModelMixIn.
def deserialize(self, filename: Union[str, os.PathLike]) ‑> None:
Defined in _BaseModel.
def deserialize_params( self, serialized_weights: _SerializedWeights,) ‑> Mapping[str, bitfount.types._TensorLike]:
Defined in DistributedModelMixIn.
def diff_params( self, old_params: _Weights, new_params: _Weights,) ‑> Mapping[str, bitfount.types._TensorLike]:
Defined in DistributedModelMixIn.
def evaluate( self, test_dl: Optional[_BitfountDataLoader] = None, pod_identifiers: Optional[List[str]] = None, **kwargs: Any,) ‑> Union[Tuple[np.ndarray, np.ndarray], Dict[str, float]]:
Defined in _BaseModel.
def fit( self, data: Optional[BaseSource] = None, metrics: Optional[Dict[str, Metric]] = None, pod_identifiers: Optional[List[str]] = None, **kwargs: Any,) ‑> Optional[Dict[str, str]]:
Defined in DistributedModelMixIn.
def get_param_states(self) ‑> Mapping[str, bitfount.types._TensorLike]:
Defined in DistributedModelMixIn.
def initialise_model( self, data: Optional[BaseSource] = None, context: Optional[ModelContext] = None,) ‑> None:
Defined in _BaseModel.
def log_(self, name: str, value: Any, **kwargs: Any) ‑> Any:
Defined in DistributedModelMixIn.
def reset_trainer(self) ‑> None:
Defined in DistributedModelMixIn.
def serialize(self, filename: Union[str, os.PathLike]) ‑> None:
Defined in _BaseModel.
def serialize_params(self, weights: _Weights) ‑> Mapping[str, List[float]]:
Defined in DistributedModelMixIn.
def set_model_training_iterations(self, iterations: int) ‑> None:
Defined in DistributedModelMixIn.
def set_pod_identifier(self, pod_identifier: str) ‑> None:
Defined in DistributedModelMixIn.
def tensor_precision(self) ‑> +T_DTYPE:
Defined in DistributedModelMixIn.
def update_params(self, new_model_params: _Weights) ‑> None:
Defined in DistributedModelMixIn.
Variables
- static
class_name : str
- static
datastructure : DataStructure
- static
epochs : Optional[int]
- static
fields_dict : ClassVar[T_FIELDS_DICT]
- static
nested_fields : ClassVar[T_NESTED_FIELDS]
- static
param_clipping : Optional[Dict[str, int]]
- static
pod_identifier : Optional[str]
- static
schema : BitfountSchema
- static
steps : Optional[int]
initialised : bool
- Defined in _BaseModel.