From 83a89dcae60ff8d0567ad518608c06df822894b1 Mon Sep 17 00:00:00 2001
From: Reuben Morais <reuben.morais@gmail.com>
Date: Tue, 15 Oct 2019 12:43:03 +0200
Subject: [PATCH 1/2] Add debugging code to trie_load.cc
---
native_client/trie_load.cc | 42 +++++++++++++++++++++++++++++++++++++-
1 file changed, 41 insertions(+), 1 deletion(-)
diff --git a/native_client/trie_load.cc b/native_client/trie_load.cc
index eab97b4b..cd625c23 100644
--- a/native_client/trie_load.cc
+++ b/native_client/trie_load.cc
@@ -5,8 +5,13 @@
#include "ctcdecode/scorer.h"
#include "alphabet.h"
-using namespace std;
+#ifdef DEBUG
+#include <limits>
+#include <unordered_map>
+#include "ctcdecode/path_trie.h"
+#endif // DEBUG
+using namespace std;
int main(int argc, char** argv)
{
@@ -22,5 +27,40 @@ int main(int argc, char** argv)
return err;
}
Scorer scorer;
+
+#ifndef DEBUG
return scorer.init(0.0, 0.0, kenlm_path, trie_path, alphabet);
+#else
+ // Print some info about the FST
+ using FstType = fst::ConstFst<fst::StdArc>;
+
+ auto dict = scorer.dictionary.get();
+
+ struct state_info {
+ int range_min = std::numeric_limits<int>::max();
+ int range_max = std::numeric_limits<int>::min();
+ };
+
+ auto print_states_from = [&](int i) {
+ std::unordered_map<int, state_info> sinfo;
+ for (fst::ArcIterator<FstType> aiter(*dict, i); !aiter.Done(); aiter.Next()) {
+ const fst::StdArc& arc = aiter.Value();
+ sinfo[arc.nextstate].range_min = std::min(sinfo[arc.nextstate].range_min, arc.ilabel-1);
+ sinfo[arc.nextstate].range_max = std::max(sinfo[arc.nextstate].range_max, arc.ilabel-1);
+ }
+
+ for (auto it = sinfo.begin(); it != sinfo.end(); ++it) {
+ state_info s = it->second;
+ printf("%d -> state %d (chars 0x%X - 0x%X, '%c' - '%c')\n", i, it->first, (unsigned int)s.range_min, (unsigned int)s.range_max, (char)s.range_min, (char)s.range_max);
+ }
+ };
+
+ print_states_from(0);
+
+ // for (int i = 1; i < 10; ++i) {
+ // print_states_from(i);
+ // }
+#endif // DEBUG
+
+ return 0;
}
From 31d81740eeb0506339ec9a967f76fe7dba6a7b29 Mon Sep 17 00:00:00 2001
From: Reuben Morais <reuben.morais@gmail.com>
Date: Tue, 15 Oct 2019 12:46:59 +0200
Subject: [PATCH 2/2] Add debugging helpers to PathTrie
---
native_client/ctcdecode/Makefile | 4 ++--
native_client/ctcdecode/build_common.py | 2 +-
native_client/ctcdecode/path_trie.cpp | 28 +++++++++++++++++++++++++
native_client/ctcdecode/path_trie.h | 9 ++++++++
4 files changed, 40 insertions(+), 3 deletions(-)
diff --git a/native_client/ctcdecode/Makefile b/native_client/ctcdecode/Makefile
index c000fc9e..98f8f9fd 100644
--- a/native_client/ctcdecode/Makefile
+++ b/native_client/ctcdecode/Makefile
@@ -33,8 +33,8 @@ bindings: clean-keep-common
bindings-debug: clean-keep-common
pip install --quiet $(PYTHON_PACKAGES) wheel==0.31.0 setuptools==39.1.0
- AS=$(AS) CC=$(CC) CXX=$(CXX) LD=$(LD) CFLAGS="$(CFLAGS) $(CXXFLAGS)" LDFLAGS="$(LDFLAGS_NEEDED)" $(PYTHON_PATH) $(NUMPY_INCLUDE) python ./setup.py build_ext --debug --num_processes $(NUM_PROCESSES) $(PYTHON_PLATFORM_NAME) $(SETUP_FLAGS)
+ AS=$(AS) CC=$(CC) CXX=$(CXX) LD=$(LD) CFLAGS="$(CFLAGS) $(CXXFLAGS) -DDEBUG" LDFLAGS="$(LDFLAGS_NEEDED)" $(PYTHON_PATH) $(NUMPY_INCLUDE) python ./setup.py build_ext --debug --num_processes $(NUM_PROCESSES) $(PYTHON_PLATFORM_NAME) $(SETUP_FLAGS)
$(GENERATE_DEBUG_SYMS)
find temp_build -type f -name "*.o" -delete
- AS=$(AS) CC=$(CC) CXX=$(CXX) LD=$(LD) CFLAGS="$(CFLAGS) $(CXXFLAGS)" LDFLAGS="$(LDFLAGS_NEEDED)" $(PYTHON_PATH) $(NUMPY_INCLUDE) python ./setup.py bdist_wheel $(PYTHON_PLATFORM_NAME) $(SETUP_FLAGS)
+ AS=$(AS) CC=$(CC) CXX=$(CXX) LD=$(LD) CFLAGS="$(CFLAGS) $(CXXFLAGS) -DDEBUG" LDFLAGS="$(LDFLAGS_NEEDED)" $(PYTHON_PATH) $(NUMPY_INCLUDE) python ./setup.py bdist_wheel $(PYTHON_PLATFORM_NAME) $(SETUP_FLAGS)
rm -rf temp_build
diff --git a/native_client/ctcdecode/build_common.py b/native_client/ctcdecode/build_common.py
index 1965287e..1a4058b5 100644
--- a/native_client/ctcdecode/build_common.py
+++ b/native_client/ctcdecode/build_common.py
@@ -11,7 +11,7 @@ from multiprocessing.dummy import Pool
ARGS = ['-DKENLM_MAX_ORDER=6', '-std=c++11', '-Wno-unused-local-typedefs', '-Wno-sign-compare']
OPT_ARGS = ['-O3', '-DNDEBUG']
-DBG_ARGS = ['-O0', '-g', '-UNDEBUG']
+DBG_ARGS = ['-O0', '-g', '-UNDEBUG', '-DDEBUG']
INCLUDES = [
'..',
diff --git a/native_client/ctcdecode/path_trie.cpp b/native_client/ctcdecode/path_trie.cpp
index 64410d93..51f75ff3 100644
--- a/native_client/ctcdecode/path_trie.cpp
+++ b/native_client/ctcdecode/path_trie.cpp
@@ -174,3 +174,31 @@ void PathTrie::set_dictionary(PathTrie::FstType* dictionary) {
void PathTrie::set_matcher(std::shared_ptr<fst::SortedMatcher<FstType>> matcher) {
matcher_ = matcher;
}
+
+#ifdef DEBUG
+void PathTrie::vec(std::vector<PathTrie*>& out) {
+ if (parent != nullptr) {
+ parent->vec(out);
+ }
+ out.push_back(this);
+}
+
+void PathTrie::print(const Alphabet& a) {
+ std::vector<PathTrie*> chain;
+ vec(chain);
+ std::string tr;
+ printf("characters:\t ");
+ for (PathTrie* el : chain) {
+ printf("%X ", el->character);
+ if (el->character != ROOT_) {
+ tr.append(a.StringFromLabel(el->character));
+ }
+ }
+ printf("\ntimesteps:\t ");
+ for (PathTrie* el : chain) {
+ printf("%d ", el->timestep);
+ }
+ printf("\n");
+ printf("transcript:\t %s\n", tr.c_str());
+}
+#endif // DEBUG
\ No newline at end of file
diff --git a/native_client/ctcdecode/path_trie.h b/native_client/ctcdecode/path_trie.h
index 10a1b687..9b71f35b 100644
--- a/native_client/ctcdecode/path_trie.h
+++ b/native_client/ctcdecode/path_trie.h
@@ -9,6 +9,10 @@
#include "fst/fstlib.h"
+#ifdef DEBUG
+#include "alphabet.h"
+#endif
+
/* Trie tree for prefix storing and manipulating, with a dictionary in
* finite-state transducer for spelling correction.
*/
@@ -44,6 +48,11 @@ public:
// remove current path from root
void remove();
+#ifdef DEBUG
+ void vec(std::vector<PathTrie*>& out);
+ void print(const Alphabet& a);
+#endif // DEBUG
+
float log_prob_b_prev;
float log_prob_nb_prev;
float log_prob_b_cur;