Computes neural network training in the distributed processing mode on the master node.
More...
Constructs the neural network training algorithm by copying input objects and parameters of another neural network training algorithm
- Parameters
-
context | Context to manage algorithm |
other | An algorithm to be used as the source to initialize the input objects and parameters of the algorithm |
Constructs the neural network training algorithm
- Parameters
-
context | Context to manage algorithm |
cls | Data type to use in intermediate computations for the algorithm, Double.class or Float.class |
method | Computation method of the algorithm, TrainingMethod |
optimizationSolver | Optimization solver |
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 |
optimizationSolver | Optimization solver |
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 algorithm |
- Returns
- The newly allocated algorithm
Runs the neural network training algorithm on master
- Returns
- Partial results of the neural network training algorithm
Computes the results of the neural network training algorithm on master
- Returns
- Results of the neural network training algorithm
Registers user-allocated memory to store the results of the neural network training algorithm
- Returns
- Results of the neural network training algorithm
Initializes the neural network topology of the layers
- Parameters
-
dataSize | Size of the input data for the trainings |
topology | TrainingTopology of the layers |
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 |
Computation method for the algorithm
Parameters of the algorithm
The documentation for this class was generated from the following file:
- neural_networks/training/DistributedStep2Master.java