DeepCL
SNAPSHOT
Deep convolutional neural networks using OpenCL
|
CBatcher | Runs an epoch for a single set of already-loaded data |
CConvolutionalMaker | Use to create a convolutional layer |
CDropoutMaker | Use to create a dropout layer |
CFullyConnectedMaker | Use to create a fully-connected layer |
CGenericLoader | Use to load data from file, given the path to the images file |
CInputLayerMaker | Use to create an InputLayer, which can be added to a NeuralNet |
CLayer | A single layer within the neural net |
CNetdefToNet | Add layers to a NeuralNet object, based on a netdef-string |
CNetLearner | Runs multiple learning epochs using Batcher objects |
CNetLearnerOnDemand | Learns multiple epochs, for data that wont fit in memory |
CNetLearnerOnDemandv2 | Learns multiple epochs, for data that wont fit in memory |
CNeuralNet | NeuralNet: main container class for network layers |
CNormalizationLayerMaker | Use to add a NormalizationLayer to a NeuralNet |
COnDemandBatcher | Learns an entire epoch of training, for data that wont fit in memory |
COnDemandBatcherv2 | Learns an entire epoch of training, for data that wont fit in memory |
CPoolingMaker | Use to create a Max-Pooling layer |
CRandomPatchesMaker | Use to create a RandomPatches Layer |
CRandomTranslationsMaker | Use to create a RandomTranslations Layer |
CWeightsPersister | Use to read/write weights from a NeuralNet |