implement CalculateBleu for complete translations, add parameters --disable-bleu-feature and --real-bleu to replace dynamic sentence Bleu with real sentence Bleu

3 files changed, 67 insertions, 83 deletions

diff --git a/mira/Decoder.cpp b/mira/Decoder.cpp
index 1edc85150..ad71c9308 100755
--- a/mira/Decoder.cpp
+++ b/mira/Decoder.cpp
@@ -45,7 +45,7 @@ namespace Mira {
   }
 
   MosesDecoder::MosesDecoder(const string& inifile, int debuglevel, int argc, vector<string> decoder_params)
-		: m_manager(NULL) {
+	: m_manager(NULL) {
   	static int BASE_ARGC = 8;
 	  Parameter* params = new Parameter();
 	  char ** mosesargv = new char*[BASE_ARGC + argc];
@@ -96,6 +96,7 @@ namespace Mira {
                               vector< float>& bleuScores,
                               vector< float>& modelScores,
                               size_t numReturnedTranslations,
+                              bool realBleu,
                               bool distinct,
                               bool avgRefLength,
                               size_t rank,
@@ -110,13 +111,13 @@ namespace Mira {
     // run the decoder
     if (chartDecoding) {
     	return runChartDecoder(source, sentenceid, nBestSize, bleuObjectiveWeight, bleuScoreWeight,
-    			featureValues, bleuScores, modelScores, numReturnedTranslations, distinct, rank, epoch,
+    			featureValues, bleuScores, modelScores, numReturnedTranslations, realBleu, distinct, rank, epoch,
     			system);
     }
     else {
     	SearchAlgorithm search = staticData.GetSearchAlgorithm();
     	return runDecoder(source, sentenceid, nBestSize, bleuObjectiveWeight, bleuScoreWeight,
-    			featureValues, bleuScores, modelScores, numReturnedTranslations, distinct, rank, epoch,
+    			featureValues, bleuScores, modelScores, numReturnedTranslations, realBleu, distinct, rank, epoch,
     			search, system, filename);
     }
   }
@@ -130,6 +131,7 @@ namespace Mira {
   														vector< float>& bleuScores,
   														vector< float>& modelScores,
   														size_t numReturnedTranslations,
+  														bool realBleu,
   														bool distinct,
   														size_t rank,
   														size_t epoch,
@@ -160,26 +162,21 @@ namespace Mira {
     for (iter = nBestList.begin() ; iter != nBestList.end() ; ++iter) {
     	const Moses::TrellisPath &path = **iter;
     	featureValues.push_back(path.GetScoreBreakdown());
-    	float bleuScore = getBleuScore(featureValues.back());
-    	bleuScores.push_back(bleuScore);
+    	float bleuScore, dynBleuScore, realBleuScore;
+    	dynBleuScore = getBleuScore(featureValues.back());  
+    	realBleuScore = m_bleuScoreFeature->CalculateBleu(path.GetTargetPhrase());
+    	bleuScore = realBleu ? realBleuScore : dynBleuScore; 
+    	bleuScores.push_back(bleuScore);	
 
     	//std::cout << "Score breakdown: " << path.GetScoreBreakdown() << endl;
     	float scoreWithoutBleu = path.GetTotalScore() - (bleuObjectiveWeight * bleuScoreWeight * bleuScore);
     	modelScores.push_back(scoreWithoutBleu);
-
-    	Phrase bestPhrase = path.GetTargetPhrase();
-
+    	
     	if (iter != nBestList.begin())
-    		cerr << endl;
-    		cerr << "Rank " << rank << ", epoch " << epoch << ", \"";
-    		Phrase phrase = path.GetTargetPhrase();
-    		for (size_t pos = 0; pos < phrase.GetSize(); ++pos) {
-    		const Word &word = phrase.GetWord(pos);
-    		Word *newWord = new Word(word);
-    		cerr << *newWord;
-    	}
-
-    	cerr << "\", score: " << scoreWithoutBleu << ", Bleu: " << bleuScore << ", total: " << path.GetTotalScore();
+    		cerr << endl;  	
+    	cerr << "Rank " << rank << ", epoch " << epoch << ", \"" << path.GetTargetPhrase() << "\", score: " 
+    		 << scoreWithoutBleu << ", Bleu: " << bleuScore << ", total: " << path.GetTotalScore();
+    	cerr << " (d-bleu: " << dynBleuScore << ", r-bleu: " << realBleuScore << ") ";
 
     	// set bleu score to zero in the feature vector since we do not want to optimise its weight
     	setBleuScore(featureValues.back(), 0);
@@ -212,6 +209,7 @@ namespace Mira {
                               vector< float>& bleuScores,
                               vector< float>& modelScores,
                               size_t numReturnedTranslations,
+                              bool realBleu,
                               bool distinct,
                               size_t rank,
                               size_t epoch,
@@ -227,26 +225,21 @@ namespace Mira {
     for (iter = nBestList.begin() ; iter != nBestList.end() ; ++iter) {
     	const Moses::ChartTrellisPath &path = **iter;
     	featureValues.push_back(path.GetScoreBreakdown());
-    	float bleuScore = getBleuScore(featureValues.back());
+    	float bleuScore, dynBleuScore, realBleuScore;
+    	dynBleuScore = getBleuScore(featureValues.back());  
+    	realBleuScore = m_bleuScoreFeature->CalculateBleu(path.GetOutputPhrase());
+    	bleuScore = realBleu ? realBleuScore : dynBleuScore; 
     	bleuScores.push_back(bleuScore);
 
     	//std::cout << "Score breakdown: " << path.GetScoreBreakdown() << endl;
     	float scoreWithoutBleu = path.GetTotalScore() - (bleuObjectiveWeight * bleuScoreWeight * bleuScore);
     	modelScores.push_back(scoreWithoutBleu);
 
-    	Phrase bestPhrase = path.GetOutputPhrase();
-
     	if (iter != nBestList.begin())
-    		cerr << endl;
-    		cerr << "Rank " << rank << ", epoch " << epoch << ", \"";
-    		Phrase phrase = path.GetOutputPhrase();
-    		for (size_t pos = 0; pos < phrase.GetSize(); ++pos) {
-    		const Word &word = phrase.GetWord(pos);
-    		Word *newWord = new Word(word);
-    		cerr << *newWord;
-    	}
-
-    	cerr << "\", score: " << scoreWithoutBleu << ", Bleu: " << bleuScore << ", total: " << path.GetTotalScore();
+    	  cerr << endl;
+    	cerr << "Rank " << rank << ", epoch " << epoch << ", \"" << path.GetOutputPhrase() << "\", score: " 
+    		 << scoreWithoutBleu << ", Bleu: " << bleuScore << ", total: " << path.GetTotalScore();
+    	cerr << " (d-bleu: " << dynBleuScore << ", r-bleu: " << realBleuScore << ") ";
 
     	// set bleu score to zero in the feature vector since we do not want to optimise its weight
     	setBleuScore(featureValues.back(), 0);
@@ -267,7 +260,6 @@ namespace Mira {
         translations.push_back(translation);
     }
 
-//    cerr << "Rank " << rank << ", use cache: " << staticData.GetUseTransOptCache() << ", weights: " << staticData.GetAllWeights() << endl;
     return translations;
   }
 
@@ -389,10 +381,10 @@ namespace Mira {
   	return m_bleuScoreFeature->GetShortestRefIndex(ref_id);
   }
 
-  void MosesDecoder::setBleuParameters(bool sentenceBleu, bool scaleByInputLength, bool scaleByAvgInputLength,
+  void MosesDecoder::setBleuParameters(bool disable, bool sentenceBleu, bool scaleByInputLength, bool scaleByAvgInputLength,
 		  bool scaleByInverseLength, bool scaleByAvgInverseLength,
 		  float scaleByX, float historySmoothing, size_t scheme) {
-	  m_bleuScoreFeature->SetBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength,
+	  m_bleuScoreFeature->SetBleuParameters(disable, sentenceBleu, scaleByInputLength, scaleByAvgInputLength,
 			  scaleByInverseLength, scaleByAvgInverseLength,
 			  scaleByX, historySmoothing, scheme);
   }
diff --git a/mira/Decoder.h b/mira/Decoder.h
index e9f9aafe3..859f97094 100755
--- a/mira/Decoder.h
+++ b/mira/Decoder.h
@@ -59,6 +59,7 @@ class MosesDecoder {
                           std::vector< float>& bleuScores,
                           std::vector< float>& modelScores,
                           size_t numReturnedTranslations,
+                          bool realBleu,
                           bool distinct,
                           bool avgRefLength,
                           size_t rank,
@@ -73,6 +74,7 @@ class MosesDecoder {
                           std::vector< float>& bleuScores,
                           std::vector< float>& modelScores,
                           size_t numReturnedTranslations,
+                          bool realBleu,
                           bool distinct,
                           size_t rank,
                           size_t epoch,
@@ -88,6 +90,7 @@ class MosesDecoder {
                           std::vector< float>& bleuScores,
                           std::vector< float>& modelScores,
                           size_t numReturnedTranslations,
+                          bool realBleu,
                           bool distinct,
                           size_t rank,
                           size_t epoch,
@@ -110,7 +113,7 @@ class MosesDecoder {
     size_t getReferenceLength(size_t ref_id);
     size_t getClosestReferenceLength(size_t ref_id, int hypoLength);
     size_t getShortestReferenceIndex(size_t ref_id);
-    void setBleuParameters(bool sentenceBleu, bool scaleByInputLength, bool scaleByAvgInputLength,
+    void setBleuParameters(bool disable, bool sentenceBleu, bool scaleByInputLength, bool scaleByAvgInputLength,
     		bool scaleByInverseLength, bool scaleByAvgInverseLength,
     		float scaleByX, float historySmoothing, size_t scheme);
     void setAvgInputLength (float l) { m_bleuScoreFeature->SetAvgInputLength(l); }
@@ -129,8 +132,6 @@ class MosesDecoder {
     Moses::ChartManager *m_chartManager;
     Moses::Sentence *m_sentence;
     Moses::BleuScoreFeature *m_bleuScoreFeature;
-	
-
 };
 
 
diff --git a/mira/Main.cpp b/mira/Main.cpp
index be5a12e39..402ffff3f 100755
--- a/mira/Main.cpp
+++ b/mira/Main.cpp
@@ -128,8 +128,11 @@ int main(int argc, char** argv) {
   bool most_violated, all_violated, max_bleu_diff, one_against_all;
   bool feature_confidence, signed_counts, averageConfidenceCounts;
   float decay, core_r0, sparse_r0;
+  bool realBleu, disableBleuFeature;
   po::options_description desc("Allowed options");
   desc.add_options()
+    ("disable-bleu-feature", po::value<bool>(&disableBleuFeature)->default_value(false), "Disable the Bleu feature")
+	("real-bleu", po::value<bool>(&realBleu)->default_value(false), "Compute real sentence Bleu on complete translations") 
     ("bleu-weight", po::value<float>(&bleuWeight)->default_value(1.0), "Bleu weight used in decoder objective")
     ("bw-hope", po::value<float>(&bleuWeight_hope)->default_value(-1.0), "Bleu weight used in decoder objective for hope")
     ("bw-fear", po::value<float>(&bleuWeight_fear)->default_value(-1.0), "Bleu weight used in decoder objective for fear")
@@ -251,19 +254,6 @@ int main(int argc, char** argv) {
   
   const StaticData &staticData = StaticData::Instance();
 
-  cerr << "scale-all: " << scale_all << endl;
-  cerr << "scale-all-factor: " << scale_all_factor << endl;
-  cerr << "bleu weight: " << bleuWeight << endl;
-  cerr << "bleu weight hope: " << bleuWeight_hope << endl;
-  cerr << "bleu weight fear: " << bleuWeight_fear << endl;;
-  cerr << "bleu weight depends on lm: " << bleu_weight_lm << endl;
-  cerr << "by this factor: " << bleu_weight_lm_factor << endl;
-  cerr << "adjust dynamically: " << bleu_weight_lm_adjust << endl;
-  cerr << "l1-reg: " << l1_regularize << endl;
-  cerr << "l1-lambda: " << l1_lambda << endl;
-  cerr << "l2-reg: " << l2_regularize << endl;
-  cerr << "l2-lambda: " << l2_lambda << endl;
-
   bool trainWithMultipleFolds = false; 
   if (mosesConfigFilesFolds.size() > 0 || inputFilesFolds.size() > 0 || referenceFilesFolds.size() > 0) {
   	if (rank == 0)
@@ -409,7 +399,7 @@ int main(int argc, char** argv) {
 	string configFile = trainWithMultipleFolds? mosesConfigFilesFolds[myFold] : mosesConfigFile;
 	VERBOSE(1, "Rank " << rank << " reading config file from " << configFile << endl);
 	MosesDecoder* decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-	decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength,
+	decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength,
 			scaleByInverseLength, scaleByAvgInverseLength,
 			scaleByX, historySmoothing, bleu_smoothing_scheme);
 	SearchAlgorithm searchAlgorithm = staticData.GetSearchAlgorithm();
@@ -866,7 +856,7 @@ int main(int argc, char** argv) {
 				    delete decoder;
 				    StaticData::ClearDataStatic();
 				    decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-				    decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+				    decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
 				    decoder->setWeights(mosesWeights);
 				  }    
 					
@@ -874,7 +864,7 @@ int main(int argc, char** argv) {
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << hope_n << "best hope translations" << endl;
 					vector< vector<const Word*> > outputHope = decoder->getNBest(input, *sid, hope_n, 1.0, bleuWeight_hope,
 							featureValuesHope[batchPosition], bleuScoresHope[batchPosition], modelScoresHope[batchPosition],
-							1, distinctNbest, avgRefLength, rank, epoch, "");
+							1, realBleu, distinctNbest, avgRefLength, rank, epoch, "");
 					vector<const Word*> oracle = outputHope[0];
 					decoder->cleanup(chartDecoding);
 					ref_length = decoder->getClosestReferenceLength(*sid, oracle.size());
@@ -883,8 +873,6 @@ int main(int argc, char** argv) {
 					int oracleSize = (int)oracle.size();
 					cerr << endl;
 
-					//exit(0);
-
 					// count sparse features occurring in hope translation
 					featureValuesHope[batchPosition][0].IncrementSparseHopeFeatures();
 
@@ -918,14 +906,14 @@ int main(int argc, char** argv) {
                                             delete decoder;
 					    StaticData::ClearDataStatic();
                                             decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-                                            decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+                                            decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
                                             decoder->setWeights(mosesWeights);
                                           }
 
 						cerr << "Rank " << rank << ", epoch " << epoch << ", 1best wrt model score (debug or history)" << endl;
 						vector< vector<const Word*> > outputModel = decoder->getNBest(input, *sid, n, 0.0, bleuWeight,
 								featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
-								1, distinctNbest, avgRefLength, rank, epoch, "");
+								1, realBleu, distinctNbest, avgRefLength, rank, epoch, "");
 						bestModel = outputModel[0];
 						decoder->cleanup(chartDecoding);
 						cerr << endl;
@@ -940,14 +928,14 @@ int main(int argc, char** argv) {
 						delete decoder;
 					    StaticData::ClearDataStatic();
 					    decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-					    decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+					    decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
 					    decoder->setWeights(mosesWeights);
 					}
 
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << fear_n << "best fear translations" << endl;
 					vector< vector<const Word*> > outputFear = decoder->getNBest(input, *sid, fear_n, -1.0, bleuWeight_fear,
 							featureValuesFear[batchPosition], bleuScoresFear[batchPosition], modelScoresFear[batchPosition],
-							1,	distinctNbest, avgRefLength, rank, epoch, "");
+							1, realBleu, distinctNbest, avgRefLength, rank, epoch, "");
 					vector<const Word*> fear = outputFear[0];
 					decoder->cleanup(chartDecoding);
 					ref_length = decoder->getClosestReferenceLength(*sid, fear.size());
@@ -1035,7 +1023,7 @@ int main(int argc, char** argv) {
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << hope_n << "best hope translations" << endl;
 					vector< vector<const Word*> > outputHope = decoder->getNBest(input, *sid, hope_n, 1.0, bleuWeight_hope,
 							featureValuesHope[batchPosition], bleuScoresHope[batchPosition], modelScoresHope[batchPosition],
-							1, distinctNbest, avgRefLength, rank, epoch, "");
+							1, realBleu, distinctNbest, avgRefLength, rank, epoch, "");
 					vector<const Word*> oracle = outputHope[0];
 					decoder->cleanup(chartDecoding);
 					cerr << endl;
@@ -1048,7 +1036,7 @@ int main(int argc, char** argv) {
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << fear_n << "best wrt model score" << endl;
 					vector< vector<const Word*> > outputModel = decoder->getNBest(input, *sid, fear_n, 0.0, bleuWeight_fear,
 							featureValuesFear[batchPosition], bleuScoresFear[batchPosition], modelScoresFear[batchPosition],
-							1, distinctNbest, avgRefLength, rank, epoch, "");
+							realBleu, 1, distinctNbest, avgRefLength, rank, epoch, "");
 					bestModel = outputModel[0];
 					decoder->cleanup(chartDecoding);
 					cerr << endl;
@@ -1067,7 +1055,7 @@ int main(int argc, char** argv) {
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << rank_n << "best wrt model score" << endl;
 					vector< vector<const Word*> > outputModel = decoder->getNBest(input, *sid, rank_n, 0.0, bleuWeight,
 							featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
-							1,	distinctNbest, avgRefLength, rank, epoch, "");
+							1, realBleu, distinctNbest, avgRefLength, rank, epoch, "");
 					vector<const Word*> bestModel = outputModel[0];
 					decoder->cleanup(chartDecoding);
 					oneBests.push_back(bestModel);
@@ -1105,21 +1093,20 @@ int main(int argc, char** argv) {
 					  delete decoder;
 					  StaticData::ClearDataStatic();
 					  decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-					  decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+					  decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
 					  decoder->setWeights(mosesWeights);
 					}
 
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << n << "best hope translations" << endl;
 					size_t oraclePos = featureValues[batchPosition].size();
-					vector <vector<const Word*> > outputHope = decoder->getNBest(input, *sid, n, 1.0, bleuWeight_hope,
+					decoder->getNBest(input, *sid, n, 1.0, bleuWeight_hope,
 							featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
-							1,	distinctNbest, avgRefLength, rank, epoch, hope_nbest_filename.str());
-					vector<const Word*> oracle = outputHope[0];
+							0, realBleu, distinctNbest, avgRefLength, rank, epoch, hope_nbest_filename.str());
 					// needed for history
 					inputLengths.push_back(current_input_length);
 					ref_ids.push_back(*sid);
 					decoder->cleanup(chartDecoding);
-					ref_length = decoder->getClosestReferenceLength(*sid, oracle.size());
+					//ref_length = decoder->getClosestReferenceLength(*sid, oracle.size());
 					//float hope_length_ratio = (float)oracle.size()/ref_length;
 					cerr << endl;
 					
@@ -1132,18 +1119,25 @@ int main(int argc, char** argv) {
 					  delete decoder;
 					  StaticData::ClearDataStatic();
 					  decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-					  decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+					  decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
 					  decoder->setWeights(mosesWeights);
 					}
 
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << n << "best wrt model score" << endl;
+					if (historyBleu) {
 					vector< vector<const Word*> > outputModel = decoder->getNBest(input, *sid, n, 0.0, bleuWeight,
 							featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
-							1,	distinctNbest, avgRefLength, rank, epoch, model_nbest_filename.str());
-					vector<const Word*> bestModel = outputModel[0];
+							1, realBleu, distinctNbest, avgRefLength, rank, epoch, model_nbest_filename.str());
+						vector<const Word*> bestModel = outputModel[0];
+						oneBests.push_back(bestModel);
+					}
+					else {
+						decoder->getNBest(input, *sid, n, 0.0, bleuWeight,
+							featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
+							0, realBleu, distinctNbest, avgRefLength, rank, epoch, model_nbest_filename.str());
+					}
 					decoder->cleanup(chartDecoding);
-					oneBests.push_back(bestModel);
-					ref_length = decoder->getClosestReferenceLength(*sid, bestModel.size());
+					//ref_length = decoder->getClosestReferenceLength(*sid, bestModel.size());
 					//float model_length_ratio = (float)bestModel.size()/ref_length;
 					cerr << endl;
 
@@ -1152,21 +1146,17 @@ int main(int argc, char** argv) {
                                           delete decoder;
 					  StaticData::ClearDataStatic();
                                           decoder = new MosesDecoder(configFile, verbosity, decoder_params.size(), decoder_params);
-                                          decoder->setBleuParameters(sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
+                                          decoder->setBleuParameters(disableBleuFeature, sentenceBleu, scaleByInputLength, scaleByAvgInputLength, scaleByInverseLength, scaleByAvgInverseLength, scaleByX, historySmoothing, bleu_smoothing_scheme);
                                           decoder->setWeights(mosesWeights);
 					}
 					
 					cerr << "Rank " << rank << ", epoch " << epoch << ", " << n << "best fear translations" << endl;
-					vector< vector<const Word*> > outputFear = decoder->getNBest(input, *sid, n, -1.0, bleuWeight_fear,
+					decoder->getNBest(input, *sid, n, -1.0, bleuWeight_fear,
 							featureValues[batchPosition], bleuScores[batchPosition], modelScores[batchPosition],
-							1, distinctNbest, avgRefLength, rank, epoch, fear_nbest_filename.str());
-					vector<const Word*> fear = outputFear[0];
+							0, realBleu, distinctNbest, avgRefLength, rank, epoch, fear_nbest_filename.str());
 					decoder->cleanup(chartDecoding);
-					ref_length = decoder->getClosestReferenceLength(*sid, fear.size());
+					//ref_length = decoder->getClosestReferenceLength(*sid, fear.size());
 					//float fear_length_ratio = (float)fear.size()/ref_length;
-					for (size_t i = 0; i < fear.size(); ++i) {
-						delete fear[i];
-					}
 
 					examples_in_batch++;
 					
@@ -1851,8 +1841,8 @@ int main(int argc, char** argv) {
 					for (size_t i = 0; i < oneBests.size(); ++i) 
 						cerr << "Rank " << rank << ", epoch " << epoch << ", update history with 1best length: " << oneBests[i].size() << " ";
 					decoder->updateHistory(oneBests, inputLengths, ref_ids, rank, epoch);
-				}
-				deleteTranslations(oneBests);
+					deleteTranslations(oneBests);
+				}				
 			} // END TRANSLATE AND UPDATE BATCH
 
 			size_t mixing_base = mixingFrequency == 0 ? 0 : shard.size() / mixingFrequency;
@@ -2321,8 +2311,9 @@ void decodeHopeOrFear(size_t rank, size_t size, size_t decode, string filename,
 		if (decode == 1) factor = 1.0;
 		if (decode == 2) factor = -1.0;
 		cerr << "Rank " << rank << ", translating sentence " << sid << endl;
+		bool realBleu = false;
 		vector< vector<const Word*> > nbestOutput = decoder->getNBest(input, sid, n, factor, bleuWeight, dummyFeatureValues[0],
-				dummyBleuScores[0], dummyModelScores[0], n, true, false, rank, 0, "");
+				dummyBleuScores[0], dummyModelScores[0], n, realBleu, true, false, rank, 0, "");
 		cerr << endl;
 		decoder->cleanup(StaticData::Instance().GetSearchAlgorithm() == ChartDecoding);