diff options
| author | Rico Sennrich <rico.sennrich@gmx.ch> | 2014-11-17 14:01:00 +0300 |
|---|---|---|
| committer | Rico Sennrich <rico.sennrich@gmx.ch> | 2014-11-17 14:01:00 +0300 |
| commit | 337ead1bbf4d4f1e6297a6c584a4718f8ba90173 (patch) | |
| tree | cf5849f5b58512203abd1e96a9f612d2ee329e97 | |
| parent | f757a6295a393a7edd4f6f300350d264a89adef7 (diff) | |
re-apply 31412f (osx compile)
| -rw-r--r-- | src/prepareNeuralLM.cpp | 4 | ||||
| -rw-r--r-- | src/trainNeuralNetwork.cpp | 8 |
2 files changed, 6 insertions, 6 deletions
diff --git a/src/prepareNeuralLM.cpp b/src/prepareNeuralLM.cpp index 13a534a..adedc72 100644 --- a/src/prepareNeuralLM.cpp +++ b/src/prepareNeuralLM.cpp @@ -240,7 +240,7 @@ void writeMmapNgrams(const string &input_filename, if (i %500000 == 0) { cerr<<"Shuffled "<<num_tokens-1<<" instances..."; } - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); for (int k=0;k<ngram_size;k++) { int temp_val = temp.at(i*ngram_size+k); temp.at(i*ngram_size+k) = @@ -263,7 +263,7 @@ void writeMmapNgrams(const string &input_filename, if (i %500000 == 0) { cerr<<"Shuffled "<<num_tokens-1<<" instances..."; } - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); for (int k=0;k<ngram_size;k++) { int temp_val = mMapVec->at(i*ngram_size+k); mMapVec->at(i*ngram_size+k) = diff --git a/src/trainNeuralNetwork.cpp b/src/trainNeuralNetwork.cpp index e231c20..a4cac12 100644 --- a/src/trainNeuralNetwork.cpp +++ b/src/trainNeuralNetwork.cpp @@ -312,7 +312,7 @@ int main(int argc, char** argv) if (i %500000 == 0) { cerr<<"Shuffled "<<training_data_size-1<<" instances..."; } - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); for (int k=0;k<myParam.ngram_size;k++) { int temp_val = training_data_flat_mmap->at(i*myParam.ngram_size+k); training_data_flat_mmap->at(i*myParam.ngram_size+k) = @@ -326,7 +326,7 @@ int main(int argc, char** argv) if (i %500000 == 0) { cerr<<"Shuffled "<<training_data_size-1<<" instances..."; } - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); for (int k=0;k<myParam.ngram_size;k++) { int temp_val = temp.at(i*myParam.ngram_size+k); temp.at(i*myParam.ngram_size+k) = @@ -348,7 +348,7 @@ int main(int argc, char** argv) if (i %500000 == 0) { cerr<<"Shuffled "<<training_data_size-1<<" instances..."; } - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); for (int k=0;k<myParam.ngram_size;k++) { int temp_val = training_data_flat_mmap->at(i*myParam.ngram_size+k); training_data_flat_mmap->at(i*myParam.ngram_size+k) = @@ -396,7 +396,7 @@ int main(int argc, char** argv) // Randomly shuffle training data to improve learning for (data_size_t i=training_data_size-1; i>0; i--) { - data_size_t j = uniform_int_distribution<data_size_t>(0, i-1)(rng); + data_size_t j = boost::random::uniform_int_distribution<data_size_t>(0, i-1)(rng); training_data.col(i).swap(training_data.col(j)); } } |