Class Model object for the prediction stage of neural network algorithm.
More...
◆ Model() [1/6]
◆ Model() [2/6]
Constructs model object for the prediction stage of neural network from the list of forward stages of the layers and the list of connections between the layers
- Parameters
-
[in] | forwardLayersForModel | List of forward stages of the layers |
[in] | nextLayersForModel | List of next layers for each layer with corresponding index |
- Deprecated:
- This item will be removed in a future release. Use Model::create instead.
◆ Model() [3/6]
Constructs model object for the prediction stage of neural network from the list of forward stages of the layers and the list of connections between the layers. And allocates storage for weights and biases of the forward layers is needed.
- Parameters
-
[in] | forwardLayersForModel | List of forward stages of the layers |
[in] | nextLayersForModel | List of next layers for each layer with corresponding index |
[in] | dummy | Data type to be used to allocate storage for weights and biases |
[in] | storeWeightsInTable | Flag. If true then the storage for weights and biases is allocated as a single piece of memory, otherwise weights and biases are allocated as separate tensors |
- Deprecated:
- This item will be removed in a future release. Use Model::create instead.
◆ Model() [4/6]
◆ Model() [5/6]
Constructs model object for the prediction stage of neural network from a collection of layer descriptors
- Parameters
-
[in] | topology | Collection of layer descriptors of every inserted layer |
- Deprecated:
- This item will be removed in a future release. Use Model::create instead.
◆ Model() [6/6]
Batch size that was used during the model allocation
◆ allocate()
Allocates the buffers needed for the prediction using neural network
- Parameters
-
[in] | sampleSize | Dimensionality of the batch for the input to the first layer |
[in] | parameter | Prediction model parameter |
- Returns
- Status of computations
◆ create() [1/4]
Constructs empty model for the prediction stage of neural network
- Parameters
-
[out] | stat | Status of the model construction |
- Returns
- Empty model for the prediction stage of neural network
◆ create() [2/4]
Constructs model object for the prediction stage of neural network from the list of forward stages of the layers and the list of connections between the layers
- Parameters
-
[in] | forwardLayersForModel | List of forward stages of the layers |
[in] | nextLayersForModel | List of next layers for each layer with corresponding index |
[out] | stat | Status of the model construction |
- Returns
- Model object for the prediction stage of neural network
◆ create() [3/4]
Constructs model object for the prediction stage of neural network from the list of forward stages of the layers and the list of connections between the layers. And allocates storage for weights and biases of the forward layers is needed.
- Parameters
-
[in] | forwardLayersForModel | List of forward stages of the layers |
[in] | nextLayersForModel | List of next layers for each layer with corresponding index |
[in] | storeWeightsInTable | Flag. If true then the storage for weights and biases is allocated as a single piece of memory, |
[out] | stat | Status of the model construction |
- Returns
- Model object for the prediction stage of neural network
◆ create() [4/4]
Constructs model object for the prediction stage of neural network from a collection of layer descriptors
- Parameters
-
[in] | topology | Collection of layer descriptors of every inserted layer |
[out] | stat | Status of the model construction |
- Returns
- Model object for the prediction stage of neural network
◆ getLayer()
const layers::forward::LayerIfacePtr getLayer |
( |
size_t |
index | ) |
const |
|
inline |
Returns the forward stage of a layer with certain index in the network
- Parameters
-
[in] | index | Index of the layer in the network |
- Returns
- Forward stage of a layer with certain index in the network
◆ getLayers()
const neural_networks::ForwardLayersPtr getLayers |
( |
| ) |
const |
|
inline |
Returns the list of forward stages of the layers
- Returns
- List of forward stages of the layers
◆ setLayers()
Sets list of forward stages of the layers and the list of connections between the layers
- Parameters
-
[in] | forwardLayers | List of forward stages of the layers |
[in] | nextLayers | List of next layers for each layer with corresponding index |
- Returns
- Status of computations
The documentation for this class was generated from the following file: