// Resizes forward data to cope with an input image of the given width.
void LSTM::ResizeForward(const NetworkIO& input) {
int rounded_inputs = gate_weights_[CI].RoundInputs(na_);
source_.Resize(input, rounded_inputs);
which_fg_.ResizeNoInit(input.Width(), ns_);
if (IsTraining()) {
state_.ResizeFloat(input, ns_);
for (int w = 0; w < WT_COUNT; ++w) {
if (w == GFS && !Is2D()) continue;
node_values_[w].ResizeFloat(input, ns_);
}
}
}
// Helper computes min and mean best results in the output.
void LSTMRecognizer::OutputStats(const NetworkIO& outputs, float* min_output,
float* mean_output, float* sd) {
const int kOutputScale = MAX_INT8;
STATS stats(0, kOutputScale + 1);
for (int t = 0; t < outputs.Width(); ++t) {
int best_label = outputs.BestLabel(t, NULL);
if (best_label != null_char_ || t == 0) {
float best_output = outputs.f(t)[best_label];
stats.add(static_cast<int>(kOutputScale * best_output), 1);
}
}
*min_output = static_cast<float>(stats.min_bucket()) / kOutputScale;
*mean_output = stats.mean() / kOutputScale;
*sd = stats.sd() / kOutputScale;
}
// Converts the network output to a sequence of labels, with scores, using
// the simple character model (each position is a char, and the null_char_ is
// mainly intended for tail padding.)
void LSTMRecognizer::LabelsViaSimpleText(const NetworkIO& output,
GenericVector<int>* labels,
GenericVector<int>* xcoords) {
labels->truncate(0);
xcoords->truncate(0);
int width = output.Width();
for (int t = 0; t < width; ++t) {
float score = 0.0f;
int label = output.BestLabel(t, &score);
if (label != null_char_) {
labels->push_back(label);
xcoords->push_back(t);
}
}
xcoords->push_back(width);
}
// Runs forward propagation of activations on the input line.
// See NetworkCpp for a detailed discussion of the arguments.
void LSTM::Forward(bool debug, const NetworkIO& input,
const TransposedArray* input_transpose,
NetworkScratch* scratch, NetworkIO* output) {
input_map_ = input.stride_map();
input_width_ = input.Width();
if (softmax_ != NULL)
output->ResizeFloat(input, no_);
else if (type_ == NT_LSTM_SUMMARY)
output->ResizeXTo1(input, no_);
else
output->Resize(input, no_);
ResizeForward(input);
// Temporary storage of forward computation for each gate.
NetworkScratch::FloatVec temp_lines[WT_COUNT];
for (int i = 0; i < WT_COUNT; ++i) temp_lines[i].Init(ns_, scratch);
// Single timestep buffers for the current/recurrent output and state.
NetworkScratch::FloatVec curr_state, curr_output;
curr_state.Init(ns_, scratch);
ZeroVector<double>(ns_, curr_state);
curr_output.Init(ns_, scratch);
ZeroVector<double>(ns_, curr_output);
// Rotating buffers of width buf_width allow storage of the state and output
// for the other dimension, used only when working in true 2D mode. The width
// is enough to hold an entire strip of the major direction.
int buf_width = Is2D() ? input_map_.Size(FD_WIDTH) : 1;
GenericVector<NetworkScratch::FloatVec> states, outputs;
if (Is2D()) {
states.init_to_size(buf_width, NetworkScratch::FloatVec());
outputs.init_to_size(buf_width, NetworkScratch::FloatVec());
for (int i = 0; i < buf_width; ++i) {
states[i].Init(ns_, scratch);
ZeroVector<double>(ns_, states[i]);
outputs[i].Init(ns_, scratch);
ZeroVector<double>(ns_, outputs[i]);
}
}
// Used only if a softmax LSTM.
NetworkScratch::FloatVec softmax_output;
NetworkScratch::IO int_output;
if (softmax_ != NULL) {
softmax_output.Init(no_, scratch);
ZeroVector<double>(no_, softmax_output);
int rounded_softmax_inputs = gate_weights_[CI].RoundInputs(ns_);
if (input.int_mode())
int_output.Resize2d(true, 1, rounded_softmax_inputs, scratch);
softmax_->SetupForward(input, NULL);
}
NetworkScratch::FloatVec curr_input;
curr_input.Init(na_, scratch);
StrideMap::Index src_index(input_map_);
// Used only by NT_LSTM_SUMMARY.
StrideMap::Index dest_index(output->stride_map());
do {
int t = src_index.t();
// True if there is a valid old state for the 2nd dimension.
bool valid_2d = Is2D();
if (valid_2d) {
StrideMap::Index dim_index(src_index);
if (!dim_index.AddOffset(-1, FD_HEIGHT)) valid_2d = false;
}
// Index of the 2-D revolving buffers (outputs, states).
int mod_t = Modulo(t, buf_width); // Current timestep.
// Setup the padded input in source.
source_.CopyTimeStepGeneral(t, 0, ni_, input, t, 0);
if (softmax_ != NULL) {
source_.WriteTimeStepPart(t, ni_, nf_, softmax_output);
}
source_.WriteTimeStepPart(t, ni_ + nf_, ns_, curr_output);
if (Is2D())
source_.WriteTimeStepPart(t, ni_ + nf_ + ns_, ns_, outputs[mod_t]);
if (!source_.int_mode()) source_.ReadTimeStep(t, curr_input);
// Matrix multiply the inputs with the source.
PARALLEL_IF_OPENMP(GFS)
// It looks inefficient to create the threads on each t iteration, but the
// alternative of putting the parallel outside the t loop, a single around
// the t-loop and then tasks in place of the sections is a *lot* slower.
// Cell inputs.
if (source_.int_mode())
gate_weights_[CI].MatrixDotVector(source_.i(t), temp_lines[CI]);
else
gate_weights_[CI].MatrixDotVector(curr_input, temp_lines[CI]);
FuncInplace<GFunc>(ns_, temp_lines[CI]);
SECTION_IF_OPENMP
// Input Gates.
if (source_.int_mode())
gate_weights_[GI].MatrixDotVector(source_.i(t), temp_lines[GI]);
else
gate_weights_[GI].MatrixDotVector(curr_input, temp_lines[GI]);
FuncInplace<FFunc>(ns_, temp_lines[GI]);
SECTION_IF_OPENMP
// 1-D forget gates.
if (source_.int_mode())
gate_weights_[GF1].MatrixDotVector(source_.i(t), temp_lines[GF1]);
else
gate_weights_[GF1].MatrixDotVector(curr_input, temp_lines[GF1]);
FuncInplace<FFunc>(ns_, temp_lines[GF1]);
// 2-D forget gates.
if (Is2D()) {
if (source_.int_mode())
gate_weights_[GFS].MatrixDotVector(source_.i(t), temp_lines[GFS]);
else
gate_weights_[GFS].MatrixDotVector(curr_input, temp_lines[GFS]);
FuncInplace<FFunc>(ns_, temp_lines[GFS]);
}
SECTION_IF_OPENMP
// Output gates.
if (source_.int_mode())
gate_weights_[GO].MatrixDotVector(source_.i(t), temp_lines[GO]);
else
gate_weights_[GO].MatrixDotVector(curr_input, temp_lines[GO]);
FuncInplace<FFunc>(ns_, temp_lines[GO]);
END_PARALLEL_IF_OPENMP
// Apply forget gate to state.
MultiplyVectorsInPlace(ns_, temp_lines[GF1], curr_state);
if (Is2D()) {
// Max-pool the forget gates (in 2-d) instead of blindly adding.
inT8* which_fg_col = which_fg_[t];
memset(which_fg_col, 1, ns_ * sizeof(which_fg_col[0]));
if (valid_2d) {
const double* stepped_state = states[mod_t];
for (int i = 0; i < ns_; ++i) {
if (temp_lines[GF1][i] < temp_lines[GFS][i]) {
curr_state[i] = temp_lines[GFS][i] * stepped_state[i];
which_fg_col[i] = 2;
}
}
}
}
MultiplyAccumulate(ns_, temp_lines[CI], temp_lines[GI], curr_state);
// Clip curr_state to a sane range.
ClipVector<double>(ns_, -kStateClip, kStateClip, curr_state);
if (IsTraining()) {
// Save the gate node values.
node_values_[CI].WriteTimeStep(t, temp_lines[CI]);
node_values_[GI].WriteTimeStep(t, temp_lines[GI]);
node_values_[GF1].WriteTimeStep(t, temp_lines[GF1]);
node_values_[GO].WriteTimeStep(t, temp_lines[GO]);
if (Is2D()) node_values_[GFS].WriteTimeStep(t, temp_lines[GFS]);
}
FuncMultiply<HFunc>(curr_state, temp_lines[GO], ns_, curr_output);
if (IsTraining()) state_.WriteTimeStep(t, curr_state);
if (softmax_ != NULL) {
if (input.int_mode()) {
int_output->WriteTimeStepPart(0, 0, ns_, curr_output);
softmax_->ForwardTimeStep(NULL, int_output->i(0), t, softmax_output);
} else {
softmax_->ForwardTimeStep(curr_output, NULL, t, softmax_output);
}
output->WriteTimeStep(t, softmax_output);
if (type_ == NT_LSTM_SOFTMAX_ENCODED) {
CodeInBinary(no_, nf_, softmax_output);
}
} else if (type_ == NT_LSTM_SUMMARY) {
// Output only at the end of a row.
if (src_index.IsLast(FD_WIDTH)) {
output->WriteTimeStep(dest_index.t(), curr_output);
dest_index.Increment();
}
} else {
output->WriteTimeStep(t, curr_output);
}
// Save states for use by the 2nd dimension only if needed.
if (Is2D()) {
CopyVector(ns_, curr_state, states[mod_t]);
CopyVector(ns_, curr_output, outputs[mod_t]);
}
// Always zero the states at the end of every row, but only for the major
// direction. The 2-D state remains intact.
if (src_index.IsLast(FD_WIDTH)) {
ZeroVector<double>(ns_, curr_state);
ZeroVector<double>(ns_, curr_output);
}
} while (src_index.Increment());
#if DEBUG_DETAIL > 0
tprintf("Source:%s\n", name_.string());
source_.Print(10);
tprintf("State:%s\n", name_.string());
state_.Print(10);
tprintf("Output:%s\n", name_.string());
output->Print(10);
#endif
if (debug) DisplayForward(*output);
}
