import_host_pipeline_ops

Obtains the custom TensorFlow operator in Rec SDK.

def import_host_pipeline_ops(so_pkg_name: str = LIBASC_OPS_SO) -> ModuleType

Parameter	Type	Mandatory/Optional	Description
so_pkg_name	string	Optional	Name of the Rec SDK TensorFlow .so package. Value range: [1, 100].

Success: A Python encapsulation module that contains the TensorFlow operator defined by Rec SDK in the .so file is returned.
Failure: RuntimeError: when unable to load the library or get the python wrappers.

from mx_rec.util.ops import import_host_pipeline_ops
host_pipeline_ops = import_host_pipeline_ops()

Operator	set_threshold
Function	Changes the feature access threshold.
Parameter Description	The first input parameter is the tf.Tensor at the upper layer in the tf context. emb_name: List[str]: feature table whose access threshold needs to be modified. ids_name: List[str]: only for compatibility.
Constraints	If the first input parameter value of set_threshold is 0, the corresponding embedding table does not accumulate features. (The access threshold remains unchanged, but the feature count is not accumulated and the historical value is used.)

Parent topic: Other APIs