|
file | am-nnet-simple.cc [code] |
|
file | am-nnet-simple.h [code] |
|
file | attention-test.cc [code] |
|
file | attention.cc [code] |
|
file | attention.h [code] |
| This file contains the lower-level interface for self-attention.
|
|
file | convolution-test.cc [code] |
|
file | convolution.cc [code] |
|
file | convolution.h [code] |
| This file contains some fairly low-level utilities for implementing convolutional neural networks and related methods such as TDNNs, which are mostly used in nnet-convolutional-component.h.
|
|
file | decodable-online-looped.cc [code] |
|
file | decodable-online-looped.h [code] |
|
file | decodable-simple-looped.cc [code] |
|
file | decodable-simple-looped.h [code] |
|
file | discriminative-supervision.cc [code] |
|
file | discriminative-supervision.h [code] |
|
file | discriminative-training.cc [code] |
|
file | discriminative-training.h [code] |
|
file | natural-gradient-online-test.cc [code] |
|
file | natural-gradient-online.cc [code] |
|
file | natural-gradient-online.h [code] |
|
file | nnet-am-decodable-simple.cc [code] |
|
file | nnet-am-decodable-simple.h [code] |
|
file | nnet-analyze-test.cc [code] |
|
file | nnet-analyze.cc [code] |
|
file | nnet-analyze.h [code] |
| This file contains utilities for analyzing and checking computations, which are used in the optimization code.
|
|
file | nnet-attention-component.cc [code] |
|
file | nnet-attention-component.h [code] |
| Contains component(s) related to attention models.
|
|
file | nnet-batch-compute.cc [code] |
|
file | nnet-batch-compute.h [code] |
|
file | nnet-chain-diagnostics.cc [code] |
|
file | nnet-chain-diagnostics.h [code] |
|
file | nnet-chain-example.cc [code] |
|
file | nnet-chain-example.h [code] |
|
file | nnet-chain-training.cc [code] |
|
file | nnet-chain-training.h [code] |
|
file | nnet-combined-component.cc [code] |
|
file | nnet-combined-component.h [code] |
| You can view this as an overflow from nnet-simple-component.h.
|
|
file | nnet-common-test.cc [code] |
|
file | nnet-common.cc [code] |
|
file | nnet-common.h [code] |
|
file | nnet-compile-looped.cc [code] |
|
file | nnet-compile-looped.h [code] |
|
file | nnet-compile-test.cc [code] |
|
file | nnet-compile-utils-test.cc [code] |
|
file | nnet-compile-utils.cc [code] |
|
file | nnet-compile-utils.h [code] |
|
file | nnet-compile.cc [code] |
|
file | nnet-compile.h [code] |
|
file | nnet-component-itf.cc [code] |
|
file | nnet-component-itf.h [code] |
|
file | nnet-component-test.cc [code] |
|
file | nnet-computation-graph.cc [code] |
|
file | nnet-computation-graph.h [code] |
|
file | nnet-computation.cc [code] |
|
file | nnet-computation.h [code] |
| The two main classes defined in this header are struct ComputationRequest, which basically defines a request for a concrete computation that we want the network to do (e.g.
|
|
file | nnet-compute-test.cc [code] |
|
file | nnet-compute.cc [code] |
|
file | nnet-compute.h [code] |
|
file | nnet-convolutional-component-temp.h [code] |
|
file | nnet-convolutional-component.cc [code] |
|
file | nnet-convolutional-component.h [code] |
|
file | nnet-derivative-test.cc [code] |
|
file | nnet-descriptor-test.cc [code] |
|
file | nnet-descriptor.cc [code] |
|
file | nnet-descriptor.h [code] |
| This file contains class definitions for classes ForwardingDescriptor, SumDescriptor and Descriptor.
|
|
file | nnet-diagnostics.cc [code] |
|
file | nnet-diagnostics.h [code] |
|
file | nnet-discriminative-diagnostics.cc [code] |
|
file | nnet-discriminative-diagnostics.h [code] |
|
file | nnet-discriminative-example.cc [code] |
|
file | nnet-discriminative-example.h [code] |
|
file | nnet-discriminative-training.cc [code] |
|
file | nnet-discriminative-training.h [code] |
|
file | nnet-example-test.cc [code] |
|
file | nnet-example-utils.cc [code] |
|
file | nnet-example-utils.h [code] |
|
file | nnet-example.cc [code] |
|
file | nnet-example.h [code] |
|
file | nnet-general-component.cc [code] |
|
file | nnet-general-component.h [code] |
| This file contains declarations of components that are not "simple", meaning they care about the indexes they are operating on, don't return the kSimpleComponent flag in their Properties(), and may return a different number of outputs than inputs.
|
|
file | nnet-graph-test.cc [code] |
|
file | nnet-graph.cc [code] |
|
file | nnet-graph.h [code] |
| This file contains a few functions that treat the neural net as a graph on nodes: e.g.
|
|
file | nnet-nnet-test.cc [code] |
|
file | nnet-nnet.cc [code] |
|
file | nnet-nnet.h [code] |
|
file | nnet-normalize-component.cc [code] |
|
file | nnet-normalize-component.h [code] |
| This file contains declarations of components that in one way or another normalize their input: NormalizeComponent and BatchNormComponent.
|
|
file | nnet-optimize-test.cc [code] |
|
file | nnet-optimize-utils.cc [code] |
|
file | nnet-optimize-utils.h [code] |
|
file | nnet-optimize.cc [code] |
|
file | nnet-optimize.h [code] |
|
file | nnet-parse-test.cc [code] |
|
file | nnet-parse.cc [code] |
|
file | nnet-parse.h [code] |
|
file | nnet-simple-component.cc [code] |
|
file | nnet-simple-component.h [code] |
| This file contains declarations of components that are "simple", meaning they don't care about the indexes they are operating on, produce one output for one input, and return the kSimpleComponent flag in their Properties(): for example, tanh and affine components.
|
|
file | nnet-tdnn-component.cc [code] |
|
file | nnet-test-utils.cc [code] |
|
file | nnet-test-utils.h [code] |
| This file contains various routines that are useful in test code.
|
|
file | nnet-training.cc [code] |
|
file | nnet-training.h [code] |
|
file | nnet-utils-test.cc [code] |
|
file | nnet-utils.cc [code] |
|
file | nnet-utils.h [code] |
| This file contains some miscellaneous functions dealing with class Nnet.
|
|