A neuron layer is the trainable unit in the neural network graph. These can chained together to allow the network to learn and represent complex relationships. More...
#include <NeuronLayer.h>
Public Member Functions | |
| virtual LayerType | GetType () const =0 |
| Returns the layer's type. More... | |
| virtual ObjectPtr< INeuronLayer > | InstanceLayer (const char *name, IGraphNode *input)=0 |
| Creates a new INeuronLayer with a different input, but weights shared with this layer. More... | |
| virtual void | SetRandomWeights ()=0 |
| Initialize the weights of the layer to random values. More... | |
| virtual void | Reset ()=0 |
| Resets the internal state. Should be called periodically on layers that need stepping (LSTM). The period to call this is defined by the sequence length used to train the layer. More... | |
| virtual void | Step ()=0 |
| Steps the internal state. Should be after execution of the layer and before new data is put into the layer. Only required to be called for layers that need stepping (LSTM). More... | |
 Public Member Functions inherited from SmartEngine::IGraphNode | |
| virtual const char * | GetName () const =0 |
| Returns the name of the graph node. More... | |
| virtual int | GetOutputDimension () const =0 |
| Returns the dimension of a single row of this node's output. The Output property's length will be a multiple of this value. More... | |
| virtual ObjectPtr< IMatrix > | Execute ()=0 |
| Synchronously execute the graph. The returned matrix is read-only. More... | |
| virtual ObjectPtr< IMatrix > | GetLastExecutionResult ()=0 |
| Get the results of last execution of the graph. The returned matrix is read-only. More... | |
| Public Member Functions inherited from SmartEngine::IObject | |
| virtual ObjectId | GetId () const =0 |
| Returns the ID of this object. More... | |
| virtual void | AddRef () const =0 |
| Increments the internal reference count on this object. It is not common to use this method directly. More... | |
| virtual void | Release () const =0 |
| Decrements the internal reference count on this object. It is not common to use this method directly. More... | |
| virtual int | GetRefCount () const =0 |
| Returns the number of references to this object. More... | |
| virtual void * | QueryInterface (ObjectClassId id)=0 |
| Queries the object for an interface and returns a pointer to that interface if found. More... | |
| void | operator= (IObject const &x)=delete |
| Public Member Functions inherited from SmartEngine::ISerializable | |
| virtual SerializationResult | Serialize (IMemoryBuffer *buffer)=0 |
| Write the contents of this object to a buffer. More... | |
| virtual SerializationResult | Deserialize (IMemoryBuffer *buffer)=0 |
| Fill this object with contents from a buffer. More... | |
Additional Inherited Members | |
| Public Attributes inherited from SmartEngine::IObject | |
| private | |
| __pad0__: IObject() {} IObject(IObject const&) = delete | |
Detailed Description
A neuron layer is the trainable unit in the neural network graph. These can chained together to allow the network to learn and represent complex relationships.
Member Function Documentation
◆ GetType()
|
pure virtual |
Returns the layer's type.
◆ InstanceLayer()
|
pure virtual |
Creates a new INeuronLayer with a different input, but weights shared with this layer.
- Parameters
-
name The name for the new layer. input The input into the new layer.
- Returns
◆ Reset()
|
pure virtual |
Resets the internal state. Should be called periodically on layers that need stepping (LSTM). The period to call this is defined by the sequence length used to train the layer.
◆ SetRandomWeights()
|
pure virtual |
Initialize the weights of the layer to random values.
◆ Step()
|
pure virtual |
Steps the internal state. Should be after execution of the layer and before new data is put into the layer. Only required to be called for layers that need stepping (LSTM).
