File List

Here is a list of all files with brief descriptions:

abstract-file-backed-feature-extractor.H	Provides the abstract reranker::AbstractFileBackedFeatureExtractor class, which makes it easy for concrete subclasses to extract features based on lines of a file
basic-file-backed-feature-extractor.C	Provides the reranker::BasedFileBackedFeatureExtractor class
basic-file-backed-feature-extractor.H	Provides the reranker::BasedFileBackedFeatureExtractor class
candidate-set-iterator.H	Provides an interface and some implementations for iterating over CandidateSet instances
candidate-set-proto-reader-test.C	Test driver for reader of reranker::CandidateSet messages
candidate-set-proto-reader.C	Implementation for class that reads CandidateSetMessage instances and converts them to reranker::CandidateSet instances
candidate-set-proto-reader.H	Reads CandidateSetMessage instances and converts them to reranker::CandidateSet instances
candidate-set-proto-writer-test.C	Test driver for writer of reranker::CandidateSet messages
candidate-set-proto-writer.C	Implementation for serializer for reranker::CandidateSet instances to CandidateSetMessage instances
candidate-set-proto-writer.H	Serializer for reranker::CandidateSet instances to CandidateSetMessage instances
candidate-set-reader.H	Class for reading streams of training or test instances, where each training or test instance is a reranker::CandidateSet object
candidate-set-writer.H	Class for writing streams of training or test instances, where each training or test instance is a reranker::CandidateSet object
candidate-set.C	Implementation of reranker::CandidateSet
candidate-set.H	Class to hold a single training instance for a reranker, which is a set of examples, typically the n-best output of some input process, posibly including a gold-standard feature vector
candidate.C	Candidate and Candidate::Comparator factory implementation
candidate.H	Provides the reranker::Candidate class for representing a candidate hypothesis from an initial model
compile-features.C	Definition for executable that reads in reranker::CandidateSet instances and either outputs all symbolic feature names, or else reads in a symbol table from file and compiles all feature vectors in each reranker::CandidateSet instance using that supplied symbol table, serializing the reranker::CandidateSet instances to stdout
dot-product.H	Provides a dot product implementation of the reranker::KernelFunction interface
environment-impl.C	Implementation of the Environment class
environment-impl.H	Provides an environment for variables and their values, either primitive or Factory-constructible objects, or vectors thereof
environment-test.C	Test driver for the Environment class
environment.C	Contains the implementation of the static method to construct an empty Environment instance
environment.H	Provides an interface for an Environment, which contains a map from variables of a specific type (primitives, Factory-constructible objects, or vectors thereof) to their values
example-feature-extractor.H	Provides the reranker::ExampleFeatureExtractor class, which illustrates how to write a reranker::FeatureExtractor implementation
executive-feature-extractor-test.C	Provides a unit test for the reranker::ExecutiveFeatureExtractor class
executive-feature-extractor.C	Provides the implementation of the reranker::ExecutiveFeatureExtractor interface
executive-feature-extractor.H	Provides the reranker::ExecutiveFeatureExtractor class
extract-features.C	Definition for executable that reads in reranker::CandidateSet instances from a stream and then runs feature extractors on those instances using an reranker::ExecutiveFeatureExtractor
factory.C	Implementation of a generic dynamic object factory
factory.H	Provides a generic dynamic object factory
feature-extractor-test.C	Test driver for the reranker::FeatureExtractor and reranker::FeatureExtractorFactory classes
feature-extractor.C	Feature extractor interface and factory implementation
feature-extractor.H	Provides the reranker::FeatureExtractor interface
feature-vector-reader.H	De-serializer for reranker::FeatureVector instances from FeatureVecMessage instances
feature-vector-test.C	Test for the reranker::FeatureVector class
feature-vector-writer.H	Serializer for reranker::FeatureVector instances to FeatureVecMessage instances
feature-vector.H	Defines the reranker::FeatureVector class, which, as it happens, is used to store feature vectors
file-backed-loss-setter.C	Provides the implementation of the reranker::FileBackedLossSetter class
file-backed-loss-setter.H	Provides the reranker::FileBackedLossSetter class
file-backed-ngram-feature-extractor.C	Provides the reranker::FileBackedNgramFeatureExtractor class
file-backed-ngram-feature-extractor.H	Provides the reranker::FileBackedNgramFeatureExtractor class
hadoop-run.py	A python program which will train a reranking model on a Hadoop cluster using the Iterative Parameter Mixtures perceptron training algorithm
hadooputil.py	A set of utilities to help interface with Hadoop
interpreter-test.C	Test driver for the Interpreter class
interpreter.C	Interpreter implementation
interpreter.H	Provides an interpreter for assigning primitives and Factory-constructible objects to named variables, as well as vectors thereof
kernel-function.H	Provides the reranker::KernelFunction interface
mira-style-model.C	Provides the implementation of the reranker::MiraStyleModel reranker class
mira-style-model.H	Provides the reranker::MiraStyleModel reranker class
model-combine-shards.C	Combine the model shards output from the Model Merging Reducer
model-combine-symbols.C	Combine the model symbols
model-merge-reducer-main.C	MapReduce Reducer to merge the features from multiple models
model-merge-reducer.C	MapReduce Reducer classes for model merging
model-merge-reducer.H	Reducer classes for trainer
model-proto-reader.C	Implementation for de-serializer for reranker::Model instances from ModelMessage instances
model-proto-reader.H	Interface for de-serializer for reranker::Model instances from ModelMessage instances
model-proto-writer.C	Implementation of factory for serializers of reranker::Model instances to ModelMessage instances
model-proto-writer.H	Interface for serializer for reranker::Model instances to ModelMessage instances
model-reader.H	Provides the ModelReader class, which can create Model instances from a file
model.C	Model and model factory implementation
model.H	Reranker model interface
ngram-feature-extractor.C	Provides the implementation of the reranker::NgramFeatureExtractor class
ngram-feature-extractor.H	Provides the reranker::NgramFeatureExtractor class
perceptron-model-proto-reader.C	Implementation of de-serializer for reranker::PerceptronModel instances from ModelMessage instances
perceptron-model-proto-reader.H	De-serializer for reranker::PerceptronModel instances from ModelMessage instances
perceptron-model-proto-writer.C	Implementation for serializer for reranker::PerceptronModel instances to ModelMessage instances
perceptron-model-proto-writer.H	Serializer for reranker::PerceptronModel instances to ModelMessage instances
perceptron-model-test.C	Test driver for training a reranker::PerceptronModel
perceptron-model.C	Provides the reranker::PerceptronModel class implementation
perceptron-model.H	Provides the reranker::PerceptronModel reranker class
piped-model-evaluator.C	An executable that loads one or more development test files into memory and then continuously reads filenames from stdin, loading the Model instance from each file and evaluating on the development test data, printing the loss to stdout
pyutil.py
rank-feature-extractor.C	Provides the implementation of the reranker::RankFeatureExtractor class
rank-feature-extractor.H	Provides the reranker::RankFeatureExtractor class
run-model.C	Definition for executable that reads in reranker::CandidateSet instances from a stream and then optionally runs feature extractors on those instances using an reranker::ExecutiveFeatureExtractor and, finally, trains or tests a model on those instances
stream-init.H	Provides a generic dynamic object factory
stream-tokenizer-test.C	Test driver for the Tokenizer class
stream-tokenizer.C	Implementation of the StreamTokenizer class
stream-tokenizer.H	Provides the StreamTokenizer class
string-canonicalizer.C	Provides the implementation of the reranker::StringCanonicalizer class
string-canonicalizer.H	Provides the reranker::StringCanonicalizer class
symbol-table.C	Provides the implementation of the reranker::StaticSymbolTable, which holds an ever-growing symbol table
symbol-table.H	Provides the reranker::Symbols interface as well as the reranker::StaticSymbolTable implementation
symbolize-model.C	Definition for executable that reads in a model and a symbol table for that model and then writes that model back out to a file again, ensuring that the symbol table is used to decompile any non-symbolic feature uid’s in the model
tokenizer.H	Provides the Tokenizer class
training-time.H	Provides the reranker::Time class, which holds the three notions of training time: current epoch, current time step within the current epoch and absolute time step
training-vector-set.H	Provides the reranker::TrainingVectorSet class