From d0530fcd2e4b3663ba5dd6b17266a529f7791781 Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@etu.univ-amu.fr>
Date: Fri, 14 Dec 2018 15:09:17 +0100
Subject: [PATCH] Added a way to specify .fm file for each classifier, as
command line arguments
---
maca_common/include/ProgramParameters.hpp | 2 ++
maca_common/src/ProgramParameters.cpp | 1 +
trainer/src/macaon_train.cpp | 17 +++++++++++++++++
transition_machine/src/Classifier.cpp | 7 ++++++-
4 files changed, 26 insertions(+), 1 deletion(-)
diff --git a/maca_common/include/ProgramParameters.hpp b/maca_common/include/ProgramParameters.hpp
index 14d920f..a749f9d 100644
--- a/maca_common/include/ProgramParameters.hpp
+++ b/maca_common/include/ProgramParameters.hpp
@@ -6,6 +6,7 @@
#define PROGRAMPARAMETERS__H
#include <string>
+#include <map>
struct ProgramParameters
{
@@ -54,6 +55,7 @@ struct ProgramParameters
static int batchSize;
static std::string loss;
static std::string dicts;
+ static std::map<std::string,std::string> featureModelByClassifier;
private :
diff --git a/maca_common/src/ProgramParameters.cpp b/maca_common/src/ProgramParameters.cpp
index cd88256..41930ad 100644
--- a/maca_common/src/ProgramParameters.cpp
+++ b/maca_common/src/ProgramParameters.cpp
@@ -49,3 +49,4 @@ std::string ProgramParameters::sequenceDelimiter;
std::string ProgramParameters::classifierName;
int ProgramParameters::batchSize;
std::string ProgramParameters::loss;
+std::map<std::string,std::string> ProgramParameters::featureModelByClassifier;
diff --git a/trainer/src/macaon_train.cpp b/trainer/src/macaon_train.cpp
index ab8111c..1e3c79d 100644
--- a/trainer/src/macaon_train.cpp
+++ b/trainer/src/macaon_train.cpp
@@ -43,6 +43,8 @@ po::options_description getOptionsDescription()
("printEntropy", "Print mean entropy and standard deviation accross sequences")
("dicts", po::value<std::string>()->default_value(""),
"The .dict file describing all the dictionaries to be used in the experiement. By default the filename specified in the .tm file will be used")
+ ("featureModels", po::value<std::string>()->default_value(""),
+ "For each classifier, specify what .fm (feature model) file to use. By default the filename specified in the .cla file will be used. Example : --featureModel Parser=parser.fm,Tagger=tagger.fm")
("optimizer", po::value<std::string>()->default_value("amsgrad"),
"The learning algorithm to use : amsgrad | adam | sgd")
("loss", po::value<std::string>()->default_value("neglogsoftmax"),
@@ -273,6 +275,21 @@ int main(int argc, char * argv[])
ProgramParameters::dynamicProbability = vm["proba"].as<float>();
ProgramParameters::showFeatureRepresentation = vm["showFeatureRepresentation"].as<int>();
ProgramParameters::iterationSize = vm["iterationSize"].as<int>();
+ std::string featureModels = vm["featureModels"].as<std::string>();
+ if (!featureModels.empty())
+ {
+ auto byClassifiers = split(featureModels, ',');
+ for (auto & classifier : byClassifiers)
+ {
+ auto parts = split(classifier, '=');
+ if (parts.size() != 2)
+ {
+ fprintf(stderr, "ERROR (%s) : wrong format for argument of option featureModels. Aborting.\n", ERRINFO);
+ exit(1);
+ }
+ ProgramParameters::featureModelByClassifier[parts[0]] = parts[1];
+ }
+ }
if (ProgramParameters::nbTrain)
{
diff --git a/transition_machine/src/Classifier.cpp b/transition_machine/src/Classifier.cpp
index 88b1cd6..59d6d4a 100644
--- a/transition_machine/src/Classifier.cpp
+++ b/transition_machine/src/Classifier.cpp
@@ -55,7 +55,12 @@ Classifier::Classifier(const std::string & filename, bool trainMode)
if(fscanf(fd, "Feature Model : %s\n", buffer) != 1)
badFormatAndAbort(ERRINFO);
- fm.reset(new FeatureModel(ProgramParameters::expPath + buffer));
+ std::string fmFilename = ProgramParameters::expPath + buffer;
+
+ if (ProgramParameters::featureModelByClassifier.count(this->name))
+ fmFilename = ProgramParameters::featureModelByClassifier[this->name];
+
+ fm.reset(new FeatureModel(fmFilename));
if(fscanf(fd, "Action Set : %s\n", buffer) != 1)
badFormatAndAbort(ERRINFO);
--
GitLab