Skip to content
Snippets Groups Projects
Commit a4af59b6 authored by Franck Dary's avatar Franck Dary
Browse files

Removed permute from LSTM

parent 37ab76c9
No related branches found
No related tags found
No related merge requests found
......@@ -49,7 +49,7 @@ void Classifier::initNeuralNetwork(const std::string & topology)
}
},
{
std::regex("CNN\\((\\d+),(\\d+),(\\d+),(\\d+),\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\}\\,([+\\-]?\\d+)\\,([+\\-]?\\d+)\\)"),
std::regex("CNN\\(([+\\-]?\\d+),(\\d+),(\\d+),(\\d+),\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\}\\,([+\\-]?\\d+)\\,([+\\-]?\\d+)\\)"),
"CNN(unknownValueThreshold,leftBorder,rightBorder,nbStack,{columns},{focusedBuffer},{focusedStack},{focusedColumns},{maxNbElements},leftBorderRawInput, rightBorderRawInput) : CNN to capture context.",
[this,topology](auto sm)
{
......@@ -71,7 +71,7 @@ void Classifier::initNeuralNetwork(const std::string & topology)
}
},
{
std::regex("LSTM\\((\\d+),(\\d+),(\\d+),(\\d+),\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\}\\,([+\\-]?\\d+)\\,([+\\-]?\\d+)\\)"),
std::regex("LSTM\\(([+\\-]?\\d+),(\\d+),(\\d+),(\\d+),\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\},\\{(.*)\\}\\,([+\\-]?\\d+)\\,([+\\-]?\\d+)\\)"),
"LSTM(unknownValueThreshold,leftBorder,rightBorder,nbStack,{columns},{focusedBuffer},{focusedStack},{focusedColumns},{maxNbElements},leftBorderRawInput, rightBorderRawInput) : CNN to capture context.",
[this,topology](auto sm)
{
......
......@@ -6,6 +6,7 @@ LSTMNetworkImpl::LSTMNetworkImpl(int nbOutputs, int unknownValueThreshold, int l
constexpr int hiddenSize = 1024;
constexpr int contextLSTMSize = 512;
constexpr int focusedLSTMSize = 64;
constexpr int rawInputLSTMSize = 16;
setLeftBorder(leftBorder);
setRightBorder(rightBorder);
......@@ -16,21 +17,23 @@ LSTMNetworkImpl::LSTMNetworkImpl(int nbOutputs, int unknownValueThreshold, int l
if (leftWindowRawInput < 0 or rightWindowRawInput < 0)
rawInputSize = 0;
else
rawInputLSTM = register_module("rawInputLSTM", torch::nn::LSTM(torch::nn::LSTMOptions(embeddingsSize, focusedLSTMSize).batch_first(false).bidirectional(true)));
rawInputLSTM = register_module("rawInputLSTM", torch::nn::LSTM(torch::nn::LSTMOptions(embeddingsSize, rawInputLSTMSize).batch_first(true).bidirectional(true)));
int rawInputLSTMOutputSize = rawInputSize == 0 ? 0 : (rawInputLSTM->options.hidden_size() * (rawInputLSTM->options.bidirectional() ? 4 : 1));
int rawInputLSTMOutputSize = 0;
if (rawInputSize > 0)
rawInputLSTMOutputSize = (rawInputSize * rawInputLSTM->options.hidden_size() * (rawInputLSTM->options.bidirectional() ? 2 : 1));
wordEmbeddings = register_module("word_embeddings", torch::nn::Embedding(torch::nn::EmbeddingOptions(maxNbEmbeddings, embeddingsSize)));
embeddingsDropout = register_module("embeddings_dropout", torch::nn::Dropout(0.3));
lstmDropout = register_module("lstm_dropout", torch::nn::Dropout(0.3));
hiddenDropout = register_module("hidden_dropout", torch::nn::Dropout(0.3));
contextLSTM = register_module("contextLSTM", torch::nn::LSTM(torch::nn::LSTMOptions(columns.size()*embeddingsSize, contextLSTMSize).batch_first(false).bidirectional(true)));
contextLSTM = register_module("contextLSTM", torch::nn::LSTM(torch::nn::LSTMOptions(columns.size()*embeddingsSize, contextLSTMSize).batch_first(true).bidirectional(true)));
int totalLSTMOutputSize = contextLSTM->options.hidden_size() * (contextLSTM->options.bidirectional() ? 4 : 1) + rawInputLSTMOutputSize;
for (auto & col : focusedColumns)
{
lstms.emplace_back(register_module(fmt::format("LSTM_{}", col), torch::nn::LSTM(torch::nn::LSTMOptions(embeddingsSize, focusedLSTMSize).batch_first(false).bidirectional(true))));
lstms.emplace_back(register_module(fmt::format("LSTM_{}", col), torch::nn::LSTM(torch::nn::LSTMOptions(embeddingsSize, focusedLSTMSize).batch_first(true).bidirectional(true))));
totalLSTMOutputSize += lstms.back()->options.hidden_size() * (lstms.back()->options.bidirectional() ? 4 : 1) * (focusedBufferIndexes.size()+focusedStackIndexes.size());
}
......@@ -46,22 +49,18 @@ torch::Tensor LSTMNetworkImpl::forward(torch::Tensor input)
auto embeddings = embeddingsDropout(wordEmbeddings(input));
auto context = embeddings.narrow(1, rawInputSize, columns.size()*(1+leftBorder+rightBorder));
context = context.view({context.size(0), context.size(1)/(int)columns.size(), (int)columns.size()*(int)wordEmbeddings->options.embedding_dim()});
auto elementsEmbeddings = embeddings.narrow(1, rawInputSize+context.size(1), input.size(1)-(rawInputSize+context.size(1)));
context = context.permute({1,0,2});
std::vector<torch::Tensor> lstmOutputs;
if (rawInputSize != 0)
{
auto rawLetters = embeddings.narrow(1, 0, leftWindowRawInput+rightWindowRawInput+1).permute({1,0});
auto rawLetters = embeddings.narrow(1, 0, rawInputSize);
auto lstmOut = rawInputLSTM(rawLetters).output;
if (rawInputLSTM->options.bidirectional())
lstmOutputs.emplace_back(torch::cat({lstmOut[0],lstmOut[-1]}, 1));
else
lstmOutputs.emplace_back(lstmOut[-1]);
lstmOutputs.emplace_back(lstmOut.reshape({lstmOut.size(0), -1}));
}
auto curIndex = 0;
......@@ -70,22 +69,22 @@ torch::Tensor LSTMNetworkImpl::forward(torch::Tensor input)
long nbElements = maxNbElements[i];
for (unsigned int focused = 0; focused < focusedBufferIndexes.size()+focusedStackIndexes.size(); focused++)
{
auto lstmInput = elementsEmbeddings.narrow(1, curIndex, nbElements).permute({1,0,2});
auto lstmInput = elementsEmbeddings.narrow(1, curIndex, nbElements);
curIndex += nbElements;
auto lstmOut = lstms[i](lstmInput).output;
if (lstms[i]->options.bidirectional())
lstmOutputs.emplace_back(torch::cat({lstmOut[0],lstmOut[-1]}, 1));
lstmOutputs.emplace_back(torch::cat({lstmOut.narrow(1, 0, 1).squeeze(1),lstmOut.narrow(1, lstmOut.size(1)-1, 1).squeeze(1)}, 1));
else
lstmOutputs.emplace_back(lstmOut[-1]);
lstmOutputs.emplace_back(lstmOut.narrow(1, lstmOut.size(1)-1, 1).squeeze(1));
}
}
auto lstmOut = contextLSTM(context).output;
if (contextLSTM->options.bidirectional())
lstmOutputs.emplace_back(torch::cat({lstmOut[0],lstmOut[-1]}, 1));
else
lstmOutputs.emplace_back(lstmOut[-1]);
lstmOutputs.emplace_back(torch::cat({lstmOut.narrow(1, 0, 1).squeeze(1),lstmOut.narrow(1, lstmOut.size(1)-1, 1).squeeze(1)}, 1));
else
lstmOutputs.emplace_back(lstmOut.narrow(1, lstmOut.size(1)-1, 1).squeeze(1));
auto totalInput = lstmDropout(torch::cat(lstmOutputs, 1));
......
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment