moses-smt/mosesdecoder
1
1
Fork 0
mirror of https://github.com/moses-smt/mosesdecoder.git synced 2024-12-28 22:45:50 +03:00
Code Issues Packages Projects Releases Wiki Activity
ce516b770c
mosesdecoder/moses/Parameter.cpp

1263 lines
49 KiB
C++
Raw Blame History

// $Id$
/***********************************************************************
Moses - factored phrase-based language decoder
Copyright (C) 2006 University of Edinburgh
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
***********************************************************************/
#include <ctime>
#include <iostream>
#include <iterator>
#include <fstream>
#include <sstream>
#include <algorithm>
#include "Parameter.h"
#include "Util.h"
#include "InputFileStream.h"
#include "StaticData.h"
#include "UserMessage.h"
using namespace std;
namespace Moses
{
/** define allowed parameters */
Parameter::Parameter()
{
AddParam("beam-threshold", "b", "threshold for threshold pruning");
AddParam("config", "f", "location of the configuration file");
AddParam("continue-partial-translation", "cpt", "start from nonempty hypothesis");
AddParam("decoding-graph-backoff", "dpb", "only use subsequent decoding paths for unknown spans of given length");
AddParam("dlm-model", "Order, factor and vocabulary file for discriminative LM. Use * for filename to indicate unlimited vocabulary.");
AddParam("drop-unknown", "du", "drop unknown words instead of copying them");
AddParam("disable-discarding", "dd", "disable hypothesis discarding");
AddParam("factor-delimiter", "fd", "specify a different factor delimiter than the default");
AddParam("generation-file", "location and properties of the generation table");
AddParam("global-lexical-file", "gl", "discriminatively trained global lexical translation model file");
AddParam("glm-feature", "discriminatively trained global lexical translation feature, sparse producer");
AddParam("input-factors", "list of factors in the input");
AddParam("input-file", "i", "location of the input file to be translated");
AddParam("inputtype", "text (0), confusion network (1), word lattice (2) (default = 0)");
AddParam("labeled-n-best-list", "print out labels for each weight type in n-best list. default is true");
AddParam("lmodel-file", "location and properties of the language models");
AddParam("lmodel-dub", "dictionary upper bounds of language models");
AddParam("lmodel-oov-feature", "add language model oov feature, one per model");
AddParam("mapping", "description of decoding steps");
AddParam("max-partial-trans-opt", "maximum number of partial translation options per input span (during mapping steps)");
AddParam("max-trans-opt-per-coverage", "maximum number of translation options per input span (after applying mapping steps)");
AddParam("max-phrase-length", "maximum phrase length (default 20)");
AddParam("n-best-list", "file and size of n-best-list to be generated; specify - as the file in order to write to STDOUT");
AddParam("lattice-samples", "generate samples from lattice, in same format as nbest list. Uses the file and size arguments, as in n-best-list");
AddParam("n-best-factor", "factor to compute the maximum number of contenders (=factor*nbest-size). value 0 means infinity, i.e. no threshold. default is 0");
AddParam("print-all-derivations", "to print all derivations in search graph");
AddParam("output-factors", "list of factors in the output");
AddParam("phrase-drop-allowed", "da", "if present, allow dropping of source words"); //da = drop any (word); see -du for comparison
AddParam("report-all-factors", "report all factors in output, not just first");
AddParam("report-all-factors-in-n-best", "Report all factors in n-best-lists. Default is false");
#ifdef HAVE_SYNLM
AddParam("slmodel-file", "location of the syntactic language model file(s)");
AddParam("slmodel-factor", "factor to use with syntactic language model");
AddParam("slmodel-beam", "beam width to use with syntactic language model's parser");
#endif
AddParam("stack", "s", "maximum stack size for histogram pruning");
AddParam("stack-diversity", "sd", "minimum number of hypothesis of each coverage in stack (default 0)");
AddParam("threads","th", "number of threads to use in decoding (defaults to single-threaded)");
AddParam("translation-details", "T", "for each best hypothesis, report translation details to the given file");
AddParam("ttable-file", "location and properties of the translation tables");
AddParam("ttable-limit", "ttl", "maximum number of translation table entries per input phrase");
AddParam("translation-option-threshold", "tot", "threshold for translation options relative to best for input phrase");
AddParam("early-discarding-threshold", "edt", "threshold for constructing hypotheses based on estimate cost");
AddParam("verbose", "v", "verbosity level of the logging");
AddParam("references", "Reference file(s) - used for bleu score feature");
AddParam("output-factors", "list if factors in the output");
AddParam("cache-path", "?");
AddParam("distortion-limit", "dl", "distortion (reordering) limit in maximum number of words (0 = monotone, -1 = unlimited)");
AddParam("monotone-at-punctuation", "mp", "do not reorder over punctuation");
AddParam("distortion-file", "source factors (0 if table independent of source), target factors, location of the factorized/lexicalized reordering tables");
AddParam("distortion", "configurations for each factorized/lexicalized reordering model.");
AddParam("early-distortion-cost", "edc", "include estimate of distortion cost yet to be incurred in the score [Moore & Quirk 2007]. Default is no");
AddParam("xml-input", "xi", "allows markup of input with desired translations and probabilities. values can be 'pass-through' (default), 'inclusive', 'exclusive', 'ignore'");
AddParam("xml-brackets", "xb", "specify strings to be used as xml tags opening and closing, e.g. \"{{ }}\" (default \"< >\"). Avoid square brackets because of configuration file format. Valid only with text input mode" );
AddParam("minimum-bayes-risk", "mbr", "use miminum Bayes risk to determine best translation");
AddParam("lminimum-bayes-risk", "lmbr", "use lattice miminum Bayes risk to determine best translation");
AddParam("mira", "do mira training");
AddParam("consensus-decoding", "con", "use consensus decoding (De Nero et. al. 2009)");
AddParam("mbr-size", "number of translation candidates considered in MBR decoding (default 200)");
AddParam("mbr-scale", "scaling factor to convert log linear score probability in MBR decoding (default 1.0)");
AddParam("lmbr-thetas", "theta(s) for lattice mbr calculation");
AddParam("lmbr-pruning-factor", "average number of nodes/word wanted in pruned lattice");
AddParam("lmbr-p", "unigram precision value for lattice mbr");
AddParam("lmbr-r", "ngram precision decay value for lattice mbr");
AddParam("lmbr-map-weight", "weight given to map solution when doing lattice MBR (default 0)");
AddParam("lattice-hypo-set", "to use lattice as hypo set during lattice MBR");
AddParam("clean-lm-cache", "clean language model caches after N translations (default N=1)");
AddParam("use-persistent-cache", "cache translation options across sentences (default true)");
AddParam("persistent-cache-size", "maximum size of cache for translation options (default 10,000 input phrases)");
AddParam("recover-input-path", "r", "(conf net/word lattice only) - recover input path corresponding to the best translation");
AddParam("output-word-graph", "owg", "Output stack info as word graph. Takes filename, 0=only hypos in stack, 1=stack + nbest hypos");
AddParam("time-out", "seconds after which is interrupted (-1=no time-out, default is -1)");
AddParam("output-search-graph", "osg", "Output connected hypotheses of search into specified filename");
AddParam("output-search-graph-extended", "osgx", "Output connected hypotheses of search into specified filename, in extended format");
AddParam("unpruned-search-graph", "usg", "When outputting chart search graph, do not exclude dead ends. Note: stack pruning may have eliminated some hypotheses");
AddParam("output-search-graph-slf", "slf", "Output connected hypotheses of search into specified directory, one file per sentence, in HTK standard lattice format (SLF)");
AddParam("output-search-graph-hypergraph", "Output connected hypotheses of search into specified directory, one file per sentence, in a hypergraph format (see Kenneth Heafield's lazy hypergraph decoder)");
AddParam("include-lhs-in-search-graph", "lhssg", "When outputting chart search graph, include the label of the LHS of the rule (useful when using syntax)");
#ifdef HAVE_PROTOBUF
AddParam("output-search-graph-pb", "pb", "Write phrase lattice to protocol buffer objects in the specified path.");
#endif
AddParam("cube-pruning-pop-limit", "cbp", "How many hypotheses should be popped for each stack. (default = 1000)");
AddParam("cube-pruning-diversity", "cbd", "How many hypotheses should be created for each coverage. (default = 0)");
AddParam("search-algorithm", "Which search algorithm to use. 0=normal stack, 1=cube pruning, 2=cube growing. (default = 0)");
AddParam("constraint", "Location of the file with target sentences to produce constraining the search");
AddParam("description", "Source language, target language, description");
AddParam("max-chart-span", "maximum num. of source word chart rules can consume (default 10)");
AddParam("non-terminals", "list of non-term symbols, space separated");
AddParam("rule-limit", "a little like table limit. But for chart decoding rules. Default is DEFAULT_MAX_TRANS_OPT_SIZE");
AddParam("source-label-overlap", "What happens if a span already has a label. 0=add more. 1=replace. 2=discard. Default is 0");
AddParam("output-hypo-score", "Output the hypo score to stdout with the output string. For search error analysis. Default is false");
AddParam("unknown-lhs", "file containing target lhs of unknown words. 1 per line: LHS prob");
AddParam("phrase-pair-feature", "Source and target factors for phrase pair feature");
AddParam("phrase-boundary-source-feature", "Source factors for phrase boundary feature");
AddParam("phrase-boundary-target-feature", "Target factors for phrase boundary feature");
AddParam("phrase-length-feature", "Count features for source length, target length, both of each phrase");
AddParam("target-word-insertion-feature", "Count feature for each unaligned target word");
AddParam("source-word-deletion-feature", "Count feature for each unaligned source word");
AddParam("word-translation-feature", "Count feature for word translation according to word alignment");
AddParam("cube-pruning-lazy-scoring", "cbls", "Don't fully score a hypothesis until it is popped");
AddParam("parsing-algorithm", "Which parsing algorithm to use. 0=CYK+, 1=scope-3. (default = 0)");
AddParam("search-algorithm", "Which search algorithm to use. 0=normal stack, 1=cube pruning, 2=cube growing, 4=stack with batched lm requests (default = 0)");
AddParam("constraint", "Location of the file with target sentences to produce constraining the search");
AddParam("link-param-count", "Number of parameters on word links when using confusion networks or lattices (default = 1)");
AddParam("description", "Source language, target language, description");
AddParam("max-chart-span", "maximum num. of source word chart rules can consume (default 10)");
AddParam("non-terminals", "list of non-term symbols, space separated");
AddParam("rule-limit", "a little like table limit. But for chart decoding rules. Default is DEFAULT_MAX_TRANS_OPT_SIZE");
AddParam("source-label-overlap", "What happens if a span already has a label. 0=add more. 1=replace. 2=discard. Default is 0");
AddParam("output-hypo-score", "Output the hypo score to stdout with the output string. For search error analysis. Default is false");
AddParam("unknown-lhs", "file containing target lhs of unknown words. 1 per line: LHS prob");
AddParam("show-weights", "print feature weights and exit");
AddParam("start-translation-id", "Id of 1st input. Default = 0");
AddParam("output-unknowns", "Output the unknown (OOV) words to the given file, one line per sentence");
// Compact phrase table and reordering table.
AddParam("minlexr-memory", "Load lexical reordering table in minlexr format into memory");
AddParam("minphr-memory", "Load phrase table in minphr format into memory");
AddParam("print-alignment-info", "Output word-to-word alignment to standard out, separated from translation by |||. Word-to-word alignments are takne from the phrase table if any. Default is false");
AddParam("include-segmentation-in-n-best", "include phrasal segmentation in the n-best list. default is false");
AddParam("print-alignment-info-in-n-best", "Include word-to-word alignment in the n-best list. Word-to-word alignments are takne from the phrase table if any. Default is false");
AddParam("alignment-output-file", "print output word alignments into given file");
AddParam("sort-word-alignment", "Sort word alignments for more consistent display. 0=no sort (default), 1=target order");
AddParam("report-segmentation", "t", "report phrase segmentation in the output");
AddParam("link-param-count", "DEPRECATED. DO NOT USE. Number of parameters on word links when using confusion networks or lattices (default = 1)");
AddParam("weight-slm", "slm", "DEPRECATED. DO NOT USE. weight(s) for syntactic language model");
AddParam("weight-bl", "bl", "DEPRECATED. DO NOT USE. weight for bleu score feature");
AddParam("weight-d", "d", "DEPRECATED. DO NOT USE. weight(s) for distortion (reordering components)");
AddParam("weight-dlm", "dlm", "DEPRECATED. DO NOT USE. weight for discriminative LM feature function (on top of sparse weights)");
AddParam("weight-lr", "lr", "DEPRECATED. DO NOT USE. weight(s) for lexicalized reordering, if not included in weight-d");
AddParam("weight-generation", "g", "DEPRECATED. DO NOT USE. weight(s) for generation components");
AddParam("weight-i", "I", "DEPRECATED. DO NOT USE. weight(s) for word insertion - used for parameters from confusion network and lattice input links");
AddParam("weight-l", "lm", "DEPRECATED. DO NOT USE. weight(s) for language models");
AddParam("weight-lex", "lex", "DEPRECATED. DO NOT USE. weight for global lexical model");
AddParam("weight-glm", "glm", "DEPRECATED. DO NOT USE. weight for global lexical feature, sparse producer");
AddParam("weight-wt", "wt", "DEPRECATED. DO NOT USE. weight for word translation feature");
AddParam("weight-pp", "pp", "DEPRECATED. DO NOT USE. weight for phrase pair feature");
AddParam("weight-pb", "pb", "DEPRECATED. DO NOT USE. weight for phrase boundary feature");
AddParam("weight-t", "tm", "DEPRECATED. DO NOT USE. weights for translation model components");
AddParam("weight-w", "w", "DEPRECATED. DO NOT USE. weight for word penalty");
AddParam("weight-u", "u", "DEPRECATED. DO NOT USE. weight for unknown word penalty");
AddParam("weight-e", "e", "DEPRECATED. DO NOT USE. weight for word deletion");
AddParam("text-type", "DEPRECATED. DO NOT USE. should be one of dev/devtest/test, used for domain adaptation features");
AddParam("weight-file", "wf", "feature weights file. Do *not* put weights for 'core' features in here - they go in moses.ini");
AddParam("weight", "weights for ALL models, 1 per line 'WeightName value'. Weight names can be repeated");
AddParam("weight-overwrite", "special parameter for mert. All on 1 line. Overrides weights specified in 'weights' argument");
AddParam("input-scores", "2 numbers on 2 lines - [1] of scores on each edge of a confusion network or lattice input (default=1). [2] Number of 'real' word scores (0 or 1. default=0)");
AddParam("feature", "");
AddParam("print-id", "prefix translations with id. Default if false");
}
Parameter::~Parameter()
{
}
/** initialize a parameter, sub of constructor */
void Parameter::AddParam(const string &paramName, const string &description)
{
m_valid[paramName] = true;
m_description[paramName] = description;
}
/** initialize a parameter (including abbreviation), sub of constructor */
void Parameter::AddParam(const string &paramName, const string &abbrevName, const string &description)
{
m_valid[paramName] = true;
m_valid[abbrevName] = true;
m_abbreviation[paramName] = abbrevName;
m_fullname[abbrevName] = paramName;
m_description[paramName] = description;
}
/** print descriptions of all parameters */
void Parameter::Explain()
{
cerr << "Usage:" << endl;
for(PARAM_STRING::const_iterator iterParam = m_description.begin(); iterParam != m_description.end(); iterParam++) {
const string paramName = iterParam->first;
const string paramDescription = iterParam->second;
cerr << "\t-" << paramName;
PARAM_STRING::const_iterator iterAbbr = m_abbreviation.find( paramName );
if ( iterAbbr != m_abbreviation.end() )
cerr << " (" << iterAbbr->second << ")";
cerr << ": " << paramDescription << endl;
}
}
/** check whether an item on the command line is a switch or a value
* \param token token on the command line to checked **/
bool Parameter::isOption(const char* token)
{
if (! token) return false;
std::string tokenString(token);
size_t length = tokenString.size();
if (length > 0 && tokenString.substr(0,1) != "-") return false;
if (length > 1 && tokenString.substr(1,1).find_first_not_of("0123456789") == 0) return true;
return false;
}
/** load all parameters from the configuration file and the command line switches */
bool Parameter::LoadParam(const string &filePath)
{
const char *argv[] = {"executable", "-f", filePath.c_str() };
return LoadParam(3, (char**) argv);
}
/** load all parameters from the configuration file and the command line switches */
bool Parameter::LoadParam(int argc, char* argv[])
{
// config file (-f) arg mandatory
string configPath;
if ( (configPath = FindParam("-f", argc, argv)) == ""
&& (configPath = FindParam("-config", argc, argv)) == "") {
PrintCredit();
Explain();
cerr << endl;
UserMessage::Add("No configuration file was specified. Use -config or -f");
cerr << endl;
return false;
} else {
if (!ReadConfigFile(configPath)) {
UserMessage::Add("Could not read "+configPath);
return false;
}
}
// overwrite parameters with values from switches
for(PARAM_STRING::const_iterator iterParam = m_description.begin(); iterParam != m_description.end(); iterParam++) {
const string paramName = iterParam->first;
OverwriteParam("-" + paramName, paramName, argc, argv);
}
// ... also shortcuts
for(PARAM_STRING::const_iterator iterParam = m_abbreviation.begin(); iterParam != m_abbreviation.end(); iterParam++) {
const string paramName = iterParam->first;
const string paramShortName = iterParam->second;
OverwriteParam("-" + paramShortName, paramName, argc, argv);
}
// logging of parameters that were set in either config or switch
int verbose = 1;
if (m_setting.find("verbose") != m_setting.end() &&
m_setting["verbose"].size() > 0)
verbose = Scan<int>(m_setting["verbose"][0]);
if (verbose >= 1) { // only if verbose
TRACE_ERR( "Defined parameters (per moses.ini or switch):" << endl);
for(PARAM_MAP::const_iterator iterParam = m_setting.begin() ; iterParam != m_setting.end(); iterParam++) {
TRACE_ERR( "\t" << iterParam->first << ": ");
for ( size_t i = 0; i < iterParam->second.size(); i++ )
TRACE_ERR( iterParam->second[i] << " ");
TRACE_ERR( endl);
}
}
// convert old weights args to new format
if (!isParamSpecified("feature"))
ConvertWeightArgs();
CreateWeightsMap();
WeightOverwrite();
// check for illegal parameters
bool noErrorFlag = true;
for (int i = 0 ; i < argc ; i++) {
if (isOption(argv[i])) {
string paramSwitch = (string) argv[i];
string paramName = paramSwitch.substr(1);
if (m_valid.find(paramName) == m_valid.end()) {
UserMessage::Add("illegal switch: " + paramSwitch);
noErrorFlag = false;
}
}
}
// check if parameters make sense
return Validate() && noErrorFlag;
}
std::vector<float> &Parameter::GetWeights(const std::string &name)
{
std::vector<float> &ret = m_weights[name];
cerr << "WEIGHT " << name << "=";
for (size_t i = 0; i < ret.size(); ++i) {
cerr << ret[i] << ",";
}
cerr << endl;
return ret;
}
void Parameter::SetWeight(const std::string &name, size_t ind, float weight)
{
PARAM_VEC &newWeights = m_setting["weight"];
string line = name + SPrint(ind) + "= " + SPrint(weight);
newWeights.push_back(line);
}
void Parameter::SetWeight(const std::string &name, size_t ind, const vector<float> &weights)
{
PARAM_VEC &newWeights = m_setting["weight"];
string line = name + SPrint(ind) + "=";
for (size_t i = 0; i < weights.size(); ++i) {
line += " " + SPrint(weights[i]);
}
newWeights.push_back(line);
}
void Parameter::AddWeight(const std::string &name, size_t ind, const std::vector<float> &weights)
{
PARAM_VEC &newWeights = m_setting["weight"];
string sought = name + SPrint(ind) + "=";
for (size_t i = 0; i < newWeights.size(); ++i) {
string &line = newWeights[i];
if (line.find(sought) == 0) {
// found existing weight, most likely to be input weights. Append to this line
for (size_t i = 0; i < weights.size(); ++i) {
line += " " + SPrint(weights[i]);
}
return;
}
}
// nothing found. Just set
SetWeight(name, ind, weights);
}
void Parameter::ConvertWeightArgsSingleWeight(const string &oldWeightName, const string &newWeightName)
{
size_t ind = 0;
PARAM_MAP::iterator iterMap;
iterMap = m_setting.find(oldWeightName);
if (iterMap != m_setting.end())
{
const PARAM_VEC &weights = iterMap->second;
for (size_t i = 0; i < weights.size(); ++i)
{
SetWeight(newWeightName, ind, Scan<float>(weights[i]));
}
m_setting.erase(iterMap);
}
}
void Parameter::ConvertWeightArgsPhraseModel(const string &oldWeightName)
{
// process input weights 1st
if (isParamSpecified("weight-i")) {
vector<float> inputWeights = Scan<float>(m_setting["weight-i"]);
PARAM_VEC &numInputScores = m_setting["input-scores"];
if (inputWeights.size() == 1) {
CHECK(numInputScores.size() == 0);
numInputScores.push_back("1");
numInputScores.push_back("0");
}
else if (inputWeights.size() == 2) {
CHECK(numInputScores.size() == 0);
numInputScores.push_back("1");
numInputScores.push_back("1");
}
SetWeight("PhraseDictionaryBinary", 0, inputWeights);
}
// convert actually pt feature
VERBOSE(2,"Creating phrase table features" << endl);
size_t numInputScores = 0;
size_t numRealWordsInInput = 0;
map<string, size_t> ptIndices;
if (GetParam("input-scores").size()) {
numInputScores = Scan<size_t>(GetParam("input-scores")[0]);
}
if (GetParam("input-scores").size() > 1) {
numRealWordsInInput = Scan<size_t>(GetParam("input-scores")[1]);
}
// load phrase translation tables
if (GetParam("ttable-file").size() > 0) {
// weights
const vector<string> &translationVector = GetParam("ttable-file");
vector<size_t> maxTargetPhrase = Scan<size_t>(GetParam("ttable-limit"));
if(maxTargetPhrase.size() == 1 && translationVector.size() > 1) {
VERBOSE(1, "Using uniform ttable-limit of " << maxTargetPhrase[0] << " for all translation tables." << endl);
for(size_t i = 1; i < translationVector.size(); i++)
maxTargetPhrase.push_back(maxTargetPhrase[0]);
} else if(maxTargetPhrase.size() != 1 && maxTargetPhrase.size() < translationVector.size()) {
stringstream strme;
strme << "You specified " << translationVector.size() << " translation tables, but only " << maxTargetPhrase.size() << " ttable-limits.";
UserMessage::Add(strme.str());
return;
}
// MAIN LOOP
const PARAM_VEC &oldWeights = m_setting[oldWeightName];
size_t currOldInd = 0;
for(size_t currDict = 0 ; currDict < translationVector.size(); currDict++) {
stringstream ptLine;
vector<string> token = Tokenize(translationVector[currDict]);
if(currDict == 0 && token.size() == 4) {
UserMessage::Add("Phrase table specification in old 4-field format. No longer supported");
return;
}
CHECK(token.size() >= 5);
PhraseTableImplementation implementation = (PhraseTableImplementation) Scan<int>(token[0]);
string ptType;
switch (implementation)
{
case Memory:
ptType = "PhraseDictionaryMemory";
break;
case Binary:
ptType = "PhraseDictionaryBinary";
break;
case OnDisk:
ptType = "PhraseDictionaryOnDisk";
break;
case SCFG:
ptType = "PhraseDictionaryMemory";
break;
case Compact:
ptType = "PhraseDictionaryCompact";
break;
default:
break;
}
size_t ptInd;
if (ptIndices.find(ptType) == ptIndices.end()) {
ptIndices[ptType] = 0;
ptInd = 0;
}
else {
ptInd = ++ptIndices[ptType];
}
// weights
size_t numFFInd = (token.size() == 4) ? 2 : 3;
size_t numFF = Scan<size_t>(token[numFFInd]);
vector<float> weights(numFF);
for (size_t currFF = 0; currFF < numFF; ++currFF) {
CHECK(currOldInd < oldWeights.size());
float weight = Scan<float>(oldWeights[currOldInd]);
weights[currFF] = weight;
++currOldInd;
}
AddWeight(ptType, ptInd, weights);
// actual pt
ptLine << ptType << " ";
ptLine << "input-factor=" << token[1] << " ";
ptLine << "output-factor=" << token[2] << " ";
ptLine << "path=" << token[4] << " ";
//characteristics of the phrase table
vector<FactorType> input = Tokenize<FactorType>(token[1], ",")
,output = Tokenize<FactorType>(token[2], ",");
size_t numScoreComponent = Scan<size_t>(token[3]);
string filePath= token[4];
if(currDict==0) {
// only the 1st pt. THis is shit
// TODO. find what the assumptions made by confusion network about phrase table output which makes
// it only work with binary file. This is a hack
numScoreComponent += numInputScores + numRealWordsInInput;
}
ptLine << "num-features=" << numScoreComponent << " ";
ptLine << "num-input-features=" << (currDict==0 ? numInputScores + numRealWordsInInput : 0) << " ";
ptLine << "table-limit=" << maxTargetPhrase[currDict] << " ";
if (implementation == SuffixArray) {
ptLine << "target-path=" << token[5] << " ";
ptLine << "alignment-path=" << token[6] << " ";
}
AddFeature(ptLine.str());
} // for(size_t currDict = 0 ; currDict < translationVector.size(); currDict++) {
} // if (GetParam("ttable-file").size() > 0) {
m_setting.erase("weight-i");
m_setting.erase(oldWeightName);
m_setting.erase("ttable-file");
m_setting.erase("ttable-limit");
}
void Parameter::AddFeature(const std::string &line)
{
PARAM_VEC &features = m_setting["feature"];
features.push_back(line);
}
void Parameter::ConvertWeightArgsDistortion()
{
const string oldWeightName = "weight-d";
const string oldLexReordingName = "distortion-file";
// distortion / lex distortion
const PARAM_VEC &oldWeights = GetParam(oldWeightName);
if (oldWeights.size() > 0)
{
if (!isParamSpecified("search-algorithm") ||
(GetParam("search-algorithm").size() > 0
&& (Trim(GetParam("search-algorithm")[0]) == "0"
||Trim(GetParam("search-algorithm")[0]) == "1"
)
)
) {
// phrase-based. Add distance distortion to list of features
AddFeature("Distortion");
SetWeight("Distortion", 0, Scan<float>(oldWeights[0]));
}
// everything but the last is lex reordering model
size_t currOldInd = 1;
const PARAM_VEC &lextable = GetParam(oldLexReordingName);
for (size_t indTable = 0; indTable < lextable.size(); ++indTable) {
const string &line = lextable[indTable];
vector<string> toks = Tokenize(line);
size_t numFF = Scan<size_t>(toks[2]);
vector<float> weights(numFF);
for (size_t currFF = 0; currFF < numFF; ++currFF)
{
CHECK(currOldInd < oldWeights.size());
float weight = Scan<float>(oldWeights[currOldInd]);
weights[currFF] = weight;
++currOldInd;
}
SetWeight("LexicalReordering", indTable, weights);
stringstream strme;
strme << "LexicalReordering "
<< "type=" << toks[1] << " ";
vector<FactorType> factors = Tokenize<FactorType>(toks[0], "-");
CHECK(factors.size() == 2);
strme << "input-factor=" << factors[0]
<< " output-factor=" << factors[1] << " ";
strme << "num-features=" << toks[2] << " ";
strme << "path=" << toks[3];
AddFeature(strme.str());
}
}
m_setting.erase(oldWeightName);
m_setting.erase(oldLexReordingName);
}
void Parameter::ConvertWeightArgsLM()
{
const string oldWeightName = "weight-l";
const string oldFeatureName = "lmodel-file";
bool isChartDecoding = true;
if (!isParamSpecified("search-algorithm") ||
(GetParam("search-algorithm").size() > 0
&& (Trim(GetParam("search-algorithm")[0]) == "0"
||Trim(GetParam("search-algorithm")[0]) == "1"
)
)
) {
isChartDecoding = false;
}
vector<int> oovWeights;
if (isParamSpecified("lmodel-oov-feature")) {
oovWeights = Scan<int>(m_setting["lmodel-oov-feature"]);
}
size_t ind = 0;
PARAM_MAP::iterator iterMap;
iterMap = m_setting.find(oldWeightName);
if (iterMap != m_setting.end())
{
size_t currOldInd = 0;
const PARAM_VEC &weights = iterMap->second;
const PARAM_VEC &models = m_setting[oldFeatureName];
for (size_t lmIndex = 0; lmIndex < models.size(); ++lmIndex) {
const string &line = models[lmIndex];
vector<string> modelToks = Tokenize(line);
int lmType = Scan<int>(modelToks[0]);
string newFeatureName;
switch (lmType)
{
case 0:
newFeatureName = "SRILM";
break;
case 1:
newFeatureName = "IRSTLM";
break;
case 8:
case 9:
newFeatureName = "KENLM";
break;
default:
abort();
}
size_t numFF = 1;
if (oovWeights.size() > lmIndex)
numFF += oovWeights[lmIndex];
vector<float> weightsLM(numFF);
for (size_t currFF = 0; currFF < numFF; ++currFF)
{
CHECK(currOldInd < weights.size());
weightsLM[currFF] = Scan<float>(weights[currOldInd]);
if (isChartDecoding) {
weightsLM[currFF] = UntransformLMScore(weightsLM[currFF]);
}
++currOldInd;
}
SetWeight(newFeatureName, ind, weightsLM);
string featureLine = newFeatureName + " "
+ "factor=" + modelToks[1] + " " // factor
+ "order=" + modelToks[2] + " "; // order
if (lmType == 9) {
featureLine += "lazyken=1 ";
}
else if (lmType == 8) {
featureLine += "lazyken=0 ";
}
featureLine += "path=" + modelToks[3]; // file
AddFeature(featureLine);
} // for (size_t lmIndex = 0; lmIndex < models.size(); ++lmIndex) {
m_setting.erase(iterMap);
}
m_setting.erase(oldFeatureName);
}
void Parameter::ConvertWeightArgsGeneration(const std::string &oldWeightName, const std::string &newWeightName)
{
string oldFeatureName = "generation-file";
// distortion / lex distortion
PARAM_VEC &oldWeights = m_setting[oldWeightName];
if (oldWeights.size() > 0)
{
size_t currOldInd = 0;
PARAM_VEC &models = m_setting[oldFeatureName];
for (size_t indTable = 0; indTable < models.size(); ++indTable) {
string &line = models[indTable];
vector<string> modelToks = Tokenize(line);
size_t numFF = Scan<size_t>(modelToks[2]);
vector<float> weights(numFF);
for (size_t currFF = 0; currFF < numFF; ++currFF)
{
CHECK(currOldInd < oldWeights.size());
float weight = Scan<float>(oldWeights[currOldInd]);
weights[currFF] = weight;
++currOldInd;
}
SetWeight(newWeightName, indTable, weights);
stringstream strme;
strme << "Generation "
<< "input-factor=" << modelToks[0] << " "
<< "output-factor=" << modelToks[1] << " "
<< "num-features=" << modelToks[2] << " "
<< "path=" << modelToks[3];
AddFeature(strme.str());
}
}
m_setting.erase(oldWeightName);
m_setting.erase(oldFeatureName);
}
void Parameter::ConvertWeightArgsWordPenalty()
{
const std::string oldWeightName = "weight-w";
const std::string newWeightName = "WordPenalty";
bool isChartDecoding = true;
if (!isParamSpecified("search-algorithm") ||
(GetParam("search-algorithm").size() > 0
&& (Trim(GetParam("search-algorithm")[0]) == "0"
||Trim(GetParam("search-algorithm")[0]) == "1"
)
)
) {
isChartDecoding = false;
}
PARAM_MAP::iterator iterMap;
iterMap = m_setting.find(oldWeightName);
if (iterMap != m_setting.end())
{
const PARAM_VEC &weights = iterMap->second;
for (size_t i = 0; i < weights.size(); ++i)
{
float weight = Scan<float>(weights[i]);
if (isChartDecoding) {
weight *= 0.434294482;
}
SetWeight(newWeightName, i, weight);
}
m_setting.erase(iterMap);
}
}
void Parameter::ConvertWeightArgs()
{
// can't handle discr LM. must do it manually 'cos of bigram/n-gram split
CHECK( m_setting.count("weight-dlm") == 0);
// check that old & new format aren't mixed
if (m_setting.count("weight") &&
(m_setting.count("weight-i") || m_setting.count("weight-t") || m_setting.count("weight-w") ||
m_setting.count("weight-l") || m_setting.count("weight-u") || m_setting.count("weight-lex") ||
m_setting.count("weight-generation") || m_setting.count("weight-lr") || m_setting.count("weight-d")
))
{
cerr << "Do not mix old and new format for specify weights";
}
ConvertWeightArgsWordPenalty();
ConvertWeightArgsLM();
ConvertWeightArgsSingleWeight("weight-slm", "SyntacticLM");
ConvertWeightArgsSingleWeight("weight-u", "UnknownWordPenalty");
ConvertWeightArgsGeneration("weight-generation", "Generation");
ConvertWeightArgsDistortion();
// don't know or can't be bothered converting these weights
ConvertWeightArgsSingleWeight("weight-lr", "LexicalReordering");
ConvertWeightArgsSingleWeight("weight-bl", "BleuScoreFeature");
ConvertWeightArgsSingleWeight("weight-glm", "GlobalLexicalModel");
ConvertWeightArgsSingleWeight("weight-wt", "WordTranslationFeature");
ConvertWeightArgsSingleWeight("weight-pp", "PhrasePairFeature");
ConvertWeightArgsSingleWeight("weight-pb", "PhraseBoundaryFeature");
ConvertWeightArgsSingleWeight("weight-e", "WordDeletion"); // TODO Can't find real name
ConvertWeightArgsSingleWeight("weight-lex", "GlobalLexicalReordering"); // TODO Can't find real name
AddFeature("WordPenalty");
AddFeature("UnknownWordPenalty");
ConvertWeightArgsPhraseModel("weight-t");
}
void Parameter::CreateWeightsMap()
{
PARAM_VEC &vec = m_setting["weight"];
for (size_t i = 0; i < vec.size(); ++i)
{
const string &line = vec[i];
vector<string> toks = Tokenize(line);
CHECK(toks.size() >= 2);
string name = toks[0];
name = name.substr(0, name.size() - 1);
vector<float> weights(toks.size() - 1);
for (size_t i = 1; i < toks.size(); ++i) {
float weight = Scan<float>(toks[i]);
weights[i - 1] = weight;
}
m_weights[name] = weights;
}
}
void Parameter::WeightOverwrite()
{
PARAM_VEC &vec = m_setting["weight-overwrite"];
if (vec.size() == 0)
return;
// should only be 1 line
CHECK(vec.size() == 1);
string name("");
vector<float> weights;
vector<string> toks = Tokenize(vec[0]);
for (size_t i = 0; i < toks.size(); ++i)
{
const string &tok = toks[i];
if (tok.substr(tok.size() - 1, 1) == "=") {
// start of new feature
if (name != "") {
// save previous ff
m_weights[name] = weights;
weights.clear();
}
name = tok.substr(0, tok.size() - 1);
}
else {
// a weight for curr ff
float weight = Scan<float>(toks[i]);
weights.push_back(weight);
}
}
m_weights[name] = weights;
}
/** check that parameter settings make sense */
bool Parameter::Validate()
{
bool noErrorFlag = true;
PARAM_MAP::const_iterator iterParams;
for (iterParams = m_setting.begin(); iterParams != m_setting.end(); ++iterParams) {
const std::string &key = iterParams->first;
if (m_valid.find(key) == m_valid.end())
{
UserMessage::Add("Unknown parameter " + key);
noErrorFlag = false;
}
}
if (m_setting["lmodel-dub"].size() > 0) {
if (m_setting["lmodel-file"].size() != m_setting["lmodel-dub"].size()) {
stringstream errorMsg("");
errorMsg << "Config and parameters specify "
<< static_cast<int>(m_setting["lmodel-file"].size())
<< " language model files (lmodel-file), but "
<< static_cast<int>(m_setting["lmodel-dub"].size())
<< " LM upperbounds (lmodel-dub)"
<< endl;
UserMessage::Add(errorMsg.str());
noErrorFlag = false;
}
}
/*
const vector<float> &lmWeights = GetWeights("LM");
if (m_setting["lmodel-file"].size() * (m_setting.find("lmodel-oov-feature") != m_setting.end() ? 2 : 1)
!= lmWeights.size()) {
stringstream errorMsg("");
errorMsg << "Config and parameters specify "
<< static_cast<int>(m_setting["lmodel-file"].size())
<< " language model files (lmodel-file), but "
<< static_cast<int>(lmWeights.size())
<< " weights (weight-l)";
errorMsg << endl << "You might be giving '-lmodel-file TYPE FACTOR ORDER FILENAME' but you should be giving these four as a single argument, i.e. '-lmodel-file \"TYPE FACTOR ORDER FILENAME\"'";
errorMsg << endl << "You should also remember that each language model requires 2 weights, if and only if lmodel-oov-feature is on.";
UserMessage::Add(errorMsg.str());
noErrorFlag = false;
}
*/
// do files exist?
// input file
if (noErrorFlag && m_setting["input-file"].size() == 1) {
noErrorFlag = FileExists(m_setting["input-file"][0]);
if (!noErrorFlag) {
stringstream errorMsg("");
errorMsg << endl << "Input file " << m_setting["input-file"][0] << " does not exist";
UserMessage::Add(errorMsg.str());
}
}
// generation tables
if (noErrorFlag) {
std::vector<std::string> ext;
//raw tables in either un compressed or compressed form
ext.push_back("");
ext.push_back(".gz");
noErrorFlag = FilesExist("generation-file", 3, ext);
}
// distortion
if (noErrorFlag) {
std::vector<std::string> ext;
//raw tables in either un compressed or compressed form
ext.push_back("");
ext.push_back(".gz");
//prefix tree format
ext.push_back(".binlexr.idx");
//prefix tree format
ext.push_back(".minlexr");
noErrorFlag = FilesExist("distortion-file", 3, ext);
}
return noErrorFlag;
}
/** check whether a file exists */
bool Parameter::FilesExist(const string &paramName, int fieldNo, std::vector<std::string> const& extensions)
{
typedef std::vector<std::string> StringVec;
StringVec::const_iterator iter;
PARAM_MAP::const_iterator iterParam = m_setting.find(paramName);
if (iterParam == m_setting.end()) {
// no param. therefore nothing to check
return true;
}
const StringVec &pathVec = (*iterParam).second;
for (iter = pathVec.begin() ; iter != pathVec.end() ; ++iter) {
StringVec vec = Tokenize(*iter);
size_t tokenizeIndex;
if (fieldNo == -1)
tokenizeIndex = vec.size() - 1;
else
tokenizeIndex = static_cast<size_t>(fieldNo);
if (tokenizeIndex >= vec.size()) {
stringstream errorMsg("");
errorMsg << "Expected at least " << (tokenizeIndex+1) << " tokens per entry in '"
<< paramName << "', but only found "
<< vec.size();
UserMessage::Add(errorMsg.str());
return false;
}
const string &pathStr = vec[tokenizeIndex];
bool fileFound=0;
for(size_t i=0; i<extensions.size() && !fileFound; ++i) {
fileFound|=FileExists(pathStr + extensions[i]);
}
if(!fileFound) {
stringstream errorMsg("");
errorMsg << "File " << pathStr << " does not exist";
UserMessage::Add(errorMsg.str());
return false;
}
}
return true;
}
/** look for a switch in arg, update parameter */
// TODO arg parsing like this does not belong in the library, it belongs
// in moses-cmd
string Parameter::FindParam(const string &paramSwitch, int argc, char* argv[])
{
for (int i = 0 ; i < argc ; i++) {
if (string(argv[i]) == paramSwitch) {
if (i+1 < argc) {
return argv[i+1];
} else {
stringstream errorMsg("");
errorMsg << "Option " << paramSwitch << " requires a parameter!";
UserMessage::Add(errorMsg.str());
// TODO return some sort of error, not the empty string
}
}
}
return "";
}
/** update parameter settings with command line switches
* \param paramSwitch (potentially short) name of switch
* \param paramName full name of parameter
* \param argc number of arguments on command line
* \param argv values of paramters on command line */
void Parameter::OverwriteParam(const string &paramSwitch, const string &paramName, int argc, char* argv[])
{
int startPos = -1;
for (int i = 0 ; i < argc ; i++) {
if (string(argv[i]) == paramSwitch) {
startPos = i+1;
break;
}
}
if (startPos < 0)
return;
int index = 0;
m_setting[paramName]; // defines the parameter, important for boolean switches
while (startPos < argc && (!isOption(argv[startPos]))) {
if (m_setting[paramName].size() > (size_t)index)
m_setting[paramName][index] = argv[startPos];
else
m_setting[paramName].push_back(argv[startPos]);
index++;
startPos++;
}
}
/** read parameters from a configuration file */
bool Parameter::ReadConfigFile(const string &filePath )
{
InputFileStream inFile(filePath);
string line, paramName;
while(getline(inFile, line)) {
// comments
size_t comPos = line.find_first_of("#");
if (comPos != string::npos)
line = line.substr(0, comPos);
// trim leading and trailing spaces/tabs
line = Trim(line);
if (line.size() == 0) {
// blank line. do nothing.
}
else if (line[0]=='[') {
// new parameter
for (size_t currPos = 0 ; currPos < line.size() ; currPos++) {
if (line[currPos] == ']') {
paramName = line.substr(1, currPos - 1);
break;
}
}
} else {
// add value to parameter
m_setting[paramName].push_back(line);
}
}
return true;
}
struct Credit {
string name, contact, currentPursuits, areaResponsibility;
int sortId;
Credit(string name, string contact, string currentPursuits, string areaResponsibility) {
this->name = name ;
this->contact = contact ;
this->currentPursuits = currentPursuits ;
this->areaResponsibility = areaResponsibility;
this->sortId = rand() % 1000;
}
bool operator<(const Credit &other) const {
/*
if (areaResponsibility.size() != 0 && other.areaResponsibility.size() ==0)
return true;
if (areaResponsibility.size() == 0 && other.areaResponsibility.size() !=0)
return false;
return name < other.name;
*/
return sortId < other.sortId;
}
};
std::ostream& operator<<(std::ostream &os, const Credit &credit)
{
os << credit.name;
if (credit.contact != "")
os << "\t contact: " << credit.contact;
if (credit.currentPursuits != "")
os << " " << credit.currentPursuits;
if (credit.areaResponsibility != "")
os << " I'll answer question on: " << credit.areaResponsibility;
return os;
}
void Parameter::PrintCredit()
{
vector<Credit> everyone;
srand ( time(NULL) );
everyone.push_back(Credit("Nicola Bertoldi"
, "911"
, ""
, "scripts & other stuff"));
everyone.push_back(Credit("Ondrej Bojar"
, ""
, "czech this out!"
, ""));
everyone.push_back(Credit("Chris Callison-Burch"
, "anytime, anywhere"
, "international playboy"
, ""));
everyone.push_back(Credit("Alexandra Constantin"
, ""
, "eu sunt varza"
, ""));
everyone.push_back(Credit("Brooke Cowan"
, "brooke@csail.mit.edu"
, "if you're going to san francisco, be sure to wear a flower in your hair"
, ""));
everyone.push_back(Credit("Chris Dyer"
, "can't. i'll be out driving my mustang"
, "driving my mustang"
, ""));
everyone.push_back(Credit("Marcello Federico"
, "federico at itc at it"
, "Researcher at ITC-irst, Trento, Italy"
, "IRST language model"));
everyone.push_back(Credit("Evan Herbst"
, "Small college in upstate New York"
, ""
, ""));
everyone.push_back(Credit("Philipp Koehn"
, "only between 2 and 4am"
, ""
, "Nothing fazes this dude"));
everyone.push_back(Credit("Christine Moran"
, "weird building at MIT"
, ""
, ""));
everyone.push_back(Credit("Wade Shen"
, "via morse code"
, "buying another laptop"
, ""));
everyone.push_back(Credit("Richard Zens"
, "richard at aachen dot de"
, ""
, "ambiguous source input, confusion networks, confusing source code"));
everyone.push_back(Credit("Hieu Hoang", "http://www.hoang.co.uk/hieu/"
, "phd student at Edinburgh Uni. Original Moses developer"
, "general queries/ flames on Moses."));
sort(everyone.begin(), everyone.end());
cerr << "Moses - A beam search decoder for phrase-based statistical machine translation models" << endl
<< "Copyright (C) 2006 University of Edinburgh" << endl << endl
<< "This library is free software; you can redistribute it and/or" << endl
<< "modify it under the terms of the GNU Lesser General Public" << endl
<< "License as published by the Free Software Foundation; either" << endl
<< "version 2.1 of the License, or (at your option) any later version." << endl << endl
<< "This library is distributed in the hope that it will be useful," << endl
<< "but WITHOUT ANY WARRANTY; without even the implied warranty of" << endl
<< "MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU" << endl
<< "Lesser General Public License for more details." << endl << endl
<< "You should have received a copy of the GNU Lesser General Public" << endl
<< "License along with this library; if not, write to the Free Software" << endl
<< "Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA" << endl << endl
<< "***********************************************************************" << endl << endl
<< "Built on " << __DATE__ << " at " __TIME__ << endl << endl
<< "WHO'S FAULT IS THIS GODDAM SOFTWARE:" << endl;
ostream_iterator<Credit> out(cerr, "\n");
copy(everyone.begin(), everyone.end(), out);
cerr << endl << endl;
}
/** update parameter settings with command line switches
* \param paramName full name of parameter
* \param values inew values for paramName */
void Parameter::OverwriteParam(const string &paramName, PARAM_VEC values)
{
VERBOSE(2,"Overwriting parameter " << paramName);
m_setting[paramName]; // defines the parameter, important for boolean switches
if (m_setting[paramName].size() > 1){
VERBOSE(2," (the parameter had " << m_setting[paramName].size() << " previous values)");
CHECK(m_setting[paramName].size() == values.size());
}else{
VERBOSE(2," (the parameter does not have previous values)");
m_setting[paramName].resize(values.size());
}
VERBOSE(2," with the following values:");
int i=0;
for (PARAM_VEC::iterator iter = values.begin(); iter != values.end() ; iter++, i++){
m_setting[paramName][i] = *iter;
VERBOSE(2, " " << *iter);
}
VERBOSE(2, std::endl);
}
std::set<std::string> Parameter::GetWeightNames() const
{
std::set<std::string> ret;
std::map<std::string, std::vector<float> >::const_iterator iter;
for (iter = m_weights.begin(); iter != m_weights.end(); ++iter) {
const string &key = iter->first;
ret.insert(key);
}
return ret;
}
}
Reference in New Issue View Git Blame Copy Permalink