Provides methods for neural network model-based training in the distributed processing mode.
More...
◆ DistributedStep1Local() [1/3]
Constructs the neural network training algorithm by copying input objects and parameters of another algorithm
- Parameters
-
context | Context to manage the algorithm |
other | An algorithm to be used as the source to initialize the input objects and parameters of the algorithm |
◆ DistributedStep1Local() [2/3]
Constructs the neural network training algorithm
- Parameters
-
context | Context to manage the algorithm |
cls | Data type to use in intermediate computations for the algorithm, Double.class or Float.class |
method | Computation method of the algorithm, TrainingMethod |
◆ DistributedStep1Local() [3/3]
Constructs neural network model-based training object with Float data type used for intermediate computations and default computation method
- Parameters
-
context | Context to manage the neural network training object |
◆ clone()
Returns the newly allocated neural network training algorithm with a copy of input objects and parameters of this neural network training algorithm
- Parameters
-
context | Context to manage the algorithm |
- Returns
- The newly allocated algorithm
◆ compute()
Runs the step 1 of the neural network training algorithm
- Returns
- Partial results of the neural network training algorithm
◆ finalizeCompute()
Computes the results of the neural network training algorithm
- Returns
- Results of the neural network training algorithm
◆ setPartialResult()
Registers user-allocated memory to store partial results of the neural network training algorithm
- Parameters
-
partialResult | Structure to store partial results of the neural network training algorithm |
◆ setResult()
Registers user-allocated memory to store the results of the neural network training algorithm
- Parameters
-
result | Structure to store the results of the neural network training algorithm |
◆ input
◆ parameter
Parameters of the algorithm
The documentation for this class was generated from the following file:
- neural_networks/training/DistributedStep1Local.java