Test int8 models (#107)

* Test int8 models

* Fix displaying help messages

* small fixes

* Fix jni test

@@ -35,7 +35,7 @@ fun main() {
 
     var objArray = WaveReader.readWave(
         assetManager = AssetManager(),
-        filename = "./sherpa-onnx-streaming-zipformer-en-2023-02-21/test_wavs/1089-134686-0001.wav",
+        filename = "./sherpa-onnx-streaming-zipformer-en-2023-02-21/test_wavs/0.wav",
     )
     var samples : FloatArray = objArray[0] as FloatArray
     var sampleRate : Int = objArray[1] as Int

@@ -25,6 +25,7 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 time $EXE \
@@ -37,6 +38,16 @@ time $EXE \
   $repo/test_wavs/1.wav \
   $repo/test_wavs/8k.wav
 
+time $EXE \
+  --tokens=$repo/tokens.txt \
+  --encoder=$repo/encoder-epoch-99-avg-1.int8.onnx \
+  --decoder=$repo/decoder-epoch-99-avg-1.int8.onnx \
+  --joiner=$repo/joiner-epoch-99-avg-1.int8.onnx \
+  --num-threads=2 \
+  $repo/test_wavs/0.wav \
+  $repo/test_wavs/1.wav \
+  $repo/test_wavs/8k.wav
+
 rm -rf $repo
 
 log "------------------------------------------------------------"
@@ -51,6 +62,7 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 time $EXE \
@@ -63,6 +75,16 @@ time $EXE \
   $repo/test_wavs/1.wav \
   $repo/test_wavs/8k.wav
 
+time $EXE \
+  --tokens=$repo/tokens.txt \
+  --encoder=$repo/encoder-epoch-99-avg-1.int8.onnx \
+  --decoder=$repo/decoder-epoch-99-avg-1.int8.onnx \
+  --joiner=$repo/joiner-epoch-99-avg-1.int8.onnx \
+  --num-threads=2 \
+  $repo/test_wavs/0.wav \
+  $repo/test_wavs/1.wav \
+  $repo/test_wavs/8k.wav
+
 rm -rf $repo
 
 log "------------------------------------------------------------"
@@ -77,6 +99,7 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 time $EXE \
@@ -89,4 +112,14 @@ time $EXE \
   $repo/test_wavs/2.wav \
   $repo/test_wavs/8k.wav
 
+time $EXE \
+  --tokens=$repo/tokens.txt \
+  --paraformer=$repo/model.int8.onnx \
+  --num-threads=2 \
+  --decoding-method=greedy_search \
+  $repo/test_wavs/0.wav \
+  $repo/test_wavs/1.wav \
+  $repo/test_wavs/2.wav \
+  $repo/test_wavs/8k.wav
+
 rm -rf $repo

@@ -25,12 +25,13 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 waves=(
-$repo/test_wavs/1089-134686-0001.wav
-$repo/test_wavs/1221-135766-0001.wav
-$repo/test_wavs/1221-135766-0002.wav
+$repo/test_wavs/0.wav
+$repo/test_wavs/1.wav
+$repo/test_wavs/8k.wav
 )
 
 for wave in ${waves[@]}; do
@@ -43,6 +44,16 @@ for wave in ${waves[@]}; do
   2
 done
 
+for wave in ${waves[@]}; do
+  time $EXE \
+  $repo/tokens.txt \
+  $repo/encoder-epoch-99-avg-1.int8.onnx \
+  $repo/decoder-epoch-99-avg-1.int8.onnx \
+  $repo/joiner-epoch-99-avg-1.int8.onnx \
+  $wave \
+  2
+done
+
 rm -rf $repo
 
 log "------------------------------------------------------------"
@@ -57,12 +68,13 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 waves=(
 $repo/test_wavs/0.wav
 $repo/test_wavs/1.wav
-$repo/test_wavs/2.wav
+$repo/test_wavs/8k.wav
 )
 
 for wave in ${waves[@]}; do
@@ -75,6 +87,16 @@ for wave in ${waves[@]}; do
   2
 done
 
+for wave in ${waves[@]}; do
+  time $EXE \
+  $repo/tokens.txt \
+  $repo/encoder-epoch-11-avg-1.int8.onnx \
+  $repo/decoder-epoch-11-avg-1.int8.onnx \
+  $repo/joiner-epoch-11-avg-1.int8.onnx \
+  $wave \
+  2
+done
+
 rm -rf $repo
 
 log "------------------------------------------------------------"
@@ -89,12 +111,13 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 waves=(
-$repo/test_wavs/1089-134686-0001.wav
-$repo/test_wavs/1221-135766-0001.wav
-$repo/test_wavs/1221-135766-0002.wav
+$repo/test_wavs/0.wav
+$repo/test_wavs/1.wav
+$repo/test_wavs/8k.wav
 )
 
 for wave in ${waves[@]}; do
@@ -107,10 +130,22 @@ for wave in ${waves[@]}; do
   2
 done
 
+# test int8
+#
+for wave in ${waves[@]}; do
+  time $EXE \
+  $repo/tokens.txt \
+  $repo/encoder-epoch-99-avg-1.int8.onnx \
+  $repo/decoder-epoch-99-avg-1.int8.onnx \
+  $repo/joiner-epoch-99-avg-1.int8.onnx \
+  $wave \
+  2
+done
+
 rm -rf $repo
 
 log "------------------------------------------------------------"
-log "Run streaming Zipformer transducer (Bilingual, Chinse + English)"
+log "Run streaming Zipformer transducer (Bilingual, Chinese + English)"
 log "------------------------------------------------------------"
 
 repo_url=https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
@@ -121,6 +156,7 @@ log "Download pretrained model and test-data from $repo_url"
 GIT_LFS_SKIP_SMUDGE=1 git clone $repo_url
 pushd $repo
 git lfs pull --include "*.onnx"
+ls -lh *.onnx
 popd
 
 waves=(
@@ -128,7 +164,7 @@ $repo/test_wavs/0.wav
 $repo/test_wavs/1.wav
 $repo/test_wavs/2.wav
 $repo/test_wavs/3.wav
-$repo/test_wavs/4.wav
+$repo/test_wavs/8k.wav
 )
 
 for wave in ${waves[@]}; do
@@ -141,6 +177,16 @@ for wave in ${waves[@]}; do
   2
 done
 
+for wave in ${waves[@]}; do
+  time $EXE \
+  $repo/tokens.txt \
+  $repo/encoder-epoch-99-avg-1.int8.onnx \
+  $repo/decoder-epoch-99-avg-1.int8.onnx \
+  $repo/joiner-epoch-99-avg-1.int8.onnx \
+  $wave \
+  2
+done
+
 # Decode a URL
 if [ $EXE == "sherpa-onnx-ffmpeg" ]; then
   time $EXE \
@@ -152,4 +198,14 @@ if [ $EXE == "sherpa-onnx-ffmpeg" ]; then
   2
 fi
 
+if [ $EXE == "sherpa-onnx-ffmpeg" ]; then
+  time $EXE \
+  $repo/tokens.txt \
+  $repo/encoder-epoch-99-avg-1.int8.onnx \
+  $repo/decoder-epoch-99-avg-1.int8.onnx \
+  $repo/joiner-epoch-99-avg-1.int8.onnx \
+  https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/resolve/main/test_wavs/4.wav \
+  2
+fi
+
 rm -rf $repo

@@ -46,3 +46,8 @@ run-sherpa-onnx-offline-paraformer.sh
 run-sherpa-onnx-offline-transducer.sh
 sherpa-onnx-paraformer-zh-2023-03-28
 run-offline-websocket-server-paraformer.sh
+run-*int8.sh
+a.sh
+run-offline-websocket-client-*.sh
+run-sherpa-onnx-*.sh
+sherpa-onnx-zipformer-en-2023-03-30

@@ -18,139 +18,13 @@
 #include <cstring>
 #include <fstream>
 #include <iomanip>
-#include <limits>
-#include <type_traits>
-#include <unordered_map>
 
 #include "sherpa-onnx/csrc/log.h"
-
-#ifdef _MSC_VER
-#define SHERPA_ONNX_STRTOLL(cur_cstr, end_cstr) \
-  _strtoi64(cur_cstr, end_cstr, 10);
-#else
-#define SHERPA_ONNX_STRTOLL(cur_cstr, end_cstr) strtoll(cur_cstr, end_cstr, 10);
-#endif
+#include "sherpa-onnx/csrc/macros.h"
+#include "sherpa-onnx/csrc/text-utils.h"
 
 namespace sherpa_onnx {
 
-/// Converts a string into an integer via strtoll and returns false if there was
-/// any kind of problem (i.e. the string was not an integer or contained extra
-/// non-whitespace junk, or the integer was too large to fit into the type it is
-/// being converted into).  Only sets *out if everything was OK and it returns
-/// true.
-template <class Int>
-bool ConvertStringToInteger(const std::string &str, Int *out) {
-  // copied from kaldi/src/util/text-util.h
-  static_assert(std::is_integral<Int>::value, "");
-  const char *this_str = str.c_str();
-  char *end = nullptr;
-  errno = 0;
-  int64_t i = SHERPA_ONNX_STRTOLL(this_str, &end);
-  if (end != this_str) {
-    while (isspace(*end)) ++end;
-  }
-  if (end == this_str || *end != '\0' || errno != 0) return false;
-  Int iInt = static_cast<Int>(i);
-  if (static_cast<int64_t>(iInt) != i ||
-      (i < 0 && !std::numeric_limits<Int>::is_signed)) {
-    return false;
-  }
-  *out = iInt;
-  return true;
-}
-
-// copied from kaldi/src/util/text-util.cc
-template <class T>
-class NumberIstream {
- public:
-  explicit NumberIstream(std::istream &i) : in_(i) {}
-
-  NumberIstream &operator>>(T &x) {
-    if (!in_.good()) return *this;
-    in_ >> x;
-    if (!in_.fail() && RemainderIsOnlySpaces()) return *this;
-    return ParseOnFail(&x);
-  }
-
- private:
-  std::istream &in_;
-
-  bool RemainderIsOnlySpaces() {
-    if (in_.tellg() != std::istream::pos_type(-1)) {
-      std::string rem;
-      in_ >> rem;
-
-      if (rem.find_first_not_of(' ') != std::string::npos) {
-        // there is not only spaces
-        return false;
-      }
-    }
-
-    in_.clear();
-    return true;
-  }
-
-  NumberIstream &ParseOnFail(T *x) {
-    std::string str;
-    in_.clear();
-    in_.seekg(0);
-    // If the stream is broken even before trying
-    // to read from it or if there are many tokens,
-    // it's pointless to try.
-    if (!(in_ >> str) || !RemainderIsOnlySpaces()) {
-      in_.setstate(std::ios_base::failbit);
-      return *this;
-    }
-
-    std::unordered_map<std::string, T> inf_nan_map;
-    // we'll keep just uppercase values.
-    inf_nan_map["INF"] = std::numeric_limits<T>::infinity();
-    inf_nan_map["+INF"] = std::numeric_limits<T>::infinity();
-    inf_nan_map["-INF"] = -std::numeric_limits<T>::infinity();
-    inf_nan_map["INFINITY"] = std::numeric_limits<T>::infinity();
-    inf_nan_map["+INFINITY"] = std::numeric_limits<T>::infinity();
-    inf_nan_map["-INFINITY"] = -std::numeric_limits<T>::infinity();
-    inf_nan_map["NAN"] = std::numeric_limits<T>::quiet_NaN();
-    inf_nan_map["+NAN"] = std::numeric_limits<T>::quiet_NaN();
-    inf_nan_map["-NAN"] = -std::numeric_limits<T>::quiet_NaN();
-    // MSVC
-    inf_nan_map["1.#INF"] = std::numeric_limits<T>::infinity();
-    inf_nan_map["-1.#INF"] = -std::numeric_limits<T>::infinity();
-    inf_nan_map["1.#QNAN"] = std::numeric_limits<T>::quiet_NaN();
-    inf_nan_map["-1.#QNAN"] = -std::numeric_limits<T>::quiet_NaN();
-
-    std::transform(str.begin(), str.end(), str.begin(), ::toupper);
-
-    if (inf_nan_map.find(str) != inf_nan_map.end()) {
-      *x = inf_nan_map[str];
-    } else {
-      in_.setstate(std::ios_base::failbit);
-    }
-
-    return *this;
-  }
-};
-
-/// ConvertStringToReal converts a string into either float or double
-/// and returns false if there was any kind of problem (i.e. the string
-/// was not a floating point number or contained extra non-whitespace junk).
-/// Be careful- this function will successfully read inf's or nan's.
-template <typename T>
-bool ConvertStringToReal(const std::string &str, T *out) {
-  std::istringstream iss(str);
-
-  NumberIstream<T> i(iss);
-
-  i >> *out;
-
-  if (iss.fail()) {
-    // Number conversion failed.
-    return false;
-  }
-
-  return true;
-}
-
 ParseOptions::ParseOptions(const std::string &prefix, ParseOptions *po)
     : print_args_(false), help_(false), usage_(""), argc_(0), argv_(nullptr) {
   if (po != nullptr && po->other_parser_ != nullptr) {
@@ -219,8 +93,8 @@ void ParseOptions::RegisterCommon(const std::string &name, T *ptr,
   std::string idx = name;
   NormalizeArgName(&idx);
   if (doc_map_.find(idx) != doc_map_.end()) {
-    SHERPA_ONNX_LOG(WARNING)
-        << "Registering option twice, ignoring second time: " << name;
+    SHERPA_ONNX_LOGE("Registering option twice, ignoring second time: %s",
+                     name.c_str());
   } else {
     this->RegisterSpecific(name, idx, ptr, doc, is_standard);
   }
@@ -289,12 +163,13 @@ void ParseOptions::RegisterSpecific(const std::string &name,
 
 void ParseOptions::DisableOption(const std::string &name) {
   if (argv_ != nullptr) {
-    SHERPA_ONNX_LOG(FATAL)
-        << "DisableOption must not be called after calling Read().";
+    SHERPA_ONNX_LOGE("DisableOption must not be called after calling Read().");
+    exit(-1);
   }
   if (doc_map_.erase(name) == 0) {
-    SHERPA_ONNX_LOG(FATAL) << "Option " << name
-                           << " was not registered so cannot be disabled: ";
+    SHERPA_ONNX_LOGE("Option %s was not registered so cannot be disabled: ",
+                     name.c_str());
+    exit(-1);
   }
   bool_map_.erase(name);
   int_map_.erase(name);
@@ -308,7 +183,8 @@ int ParseOptions::NumArgs() const { return positional_args_.size(); }
 
 std::string ParseOptions::GetArg(int i) const {
   if (i < 1 || i > static_cast<int>(positional_args_.size())) {
-    SHERPA_ONNX_LOG(FATAL) << "ParseOptions::GetArg, invalid index " << i;
+    SHERPA_ONNX_LOGE("ParseOptions::GetArg, invalid index %d", i);
+    exit(-1);
   }
 
   return positional_args_[i - 1];
@@ -460,7 +336,8 @@ int ParseOptions::Read(int argc, const char *const argv[]) {
       Trim(&value);
       if (!SetOption(key, value, has_equal_sign)) {
         PrintUsage(true);
-        SHERPA_ONNX_LOG(FATAL) << "Invalid option " << argv[i];
+        SHERPA_ONNX_LOGE("Invalid option %s", argv[i]);
+        exit(-1);
       }
     } else {
       break;
@@ -481,7 +358,7 @@ int ParseOptions::Read(int argc, const char *const argv[]) {
     std::ostringstream strm;
     for (int j = 0; j < argc; ++j) strm << Escape(argv[j]) << " ";
     strm << '\n';
-    SHERPA_ONNX_LOG(INFO) << strm.str();
+    SHERPA_ONNX_LOGE("%s", strm.str().c_str());
   }
   return i;
 }
@@ -522,7 +399,7 @@ void ParseOptions::PrintUsage(bool print_command_line /*=false*/) const {
     os << strm.str();
   }
 
-  SHERPA_ONNX_LOG(INFO) << os.str();
+  SHERPA_ONNX_LOGE("%s", os.str().c_str());
 }
 
 void ParseOptions::PrintConfig(std::ostream &os) const {
@@ -544,8 +421,9 @@ void ParseOptions::PrintConfig(std::ostream &os) const {
     } else if (string_map_.end() != string_map_.find(key)) {
       os << "'" << *string_map_.at(key) << "'";
     } else {
-      SHERPA_ONNX_LOG(FATAL)
-          << "PrintConfig: unrecognized option " << key << "[code error]";
+      SHERPA_ONNX_LOGE("PrintConfig: unrecognized option %s [code error]",
+                       key.c_str());
+      exit(-1);
     }
     os << '\n';
   }
@@ -555,7 +433,8 @@ void ParseOptions::PrintConfig(std::ostream &os) const {
 void ParseOptions::ReadConfigFile(const std::string &filename) {
   std::ifstream is(filename.c_str(), std::ifstream::in);
   if (!is.good()) {
-    SHERPA_ONNX_LOG(FATAL) << "Cannot open config file: " << filename;
+    SHERPA_ONNX_LOGE("Cannot open config file: %s", filename.c_str());
+    exit(-1);
   }
 
   std::string line, key, value;
@@ -572,12 +451,13 @@ void ParseOptions::ReadConfigFile(const std::string &filename) {
     if (line.length() == 0) continue;
 
     if (line.substr(0, 2) != "--") {
-      SHERPA_ONNX_LOG(FATAL)
-          << "Reading config file " << filename << ": line " << line_number
-          << " does not look like a line "
-          << "from a Kaldi command-line program's config file: should "
-          << "be of the form --x=y.  Note: config files intended to "
-          << "be sourced by shell scripts lack the '--'.";
+      SHERPA_ONNX_LOGE(
+          "Reading config file %s: line %d does not look like a line "
+          "from a sherpa-onnx command-line program's config file: should "
+          "be of the form --x=y.  Note: config files intended to "
+          "be sourced by shell scripts lack the '--'.",
+          filename.c_str(), line_number);
+      exit(-1);
     }
 
     // parse option
@@ -587,8 +467,9 @@ void ParseOptions::ReadConfigFile(const std::string &filename) {
     Trim(&value);
     if (!SetOption(key, value, has_equal_sign)) {
       PrintUsage(true);
-      SHERPA_ONNX_LOG(FATAL) << "Invalid option " << line << " in config file "
-                             << filename << ": line " << line_number;
+      SHERPA_ONNX_LOGE("Invalid option %s in config file %s: line %d",
+                       line.c_str(), filename.c_str(), line_number);
+      exit(-1);
     }
   }
 }
@@ -605,7 +486,8 @@ void ParseOptions::SplitLongArg(const std::string &in, std::string *key,
     *has_equal_sign = false;
   } else if (pos == 2) {  // we also don't allow empty keys: --=value
     PrintUsage(true);
-    SHERPA_ONNX_LOG(FATAL) << "Invalid option (no key): " << in;
+    SHERPA_ONNX_LOGE("Invalid option (no key): %s", in.c_str());
+    exit(-1);
   } else {                         // normal case: --option=value
     *key = in.substr(2, pos - 2);  // 2 because starts with --.
     *value = in.substr(pos + 1);
@@ -646,7 +528,8 @@ bool ParseOptions::SetOption(const std::string &key, const std::string &value,
                              bool has_equal_sign) {
   if (bool_map_.end() != bool_map_.find(key)) {
     if (has_equal_sign && value == "") {
-      SHERPA_ONNX_LOG(FATAL) << "Invalid option --" << key << "=";
+      SHERPA_ONNX_LOGE("Invalid option --%s=", key.c_str());
+      exit(-1);
     }
     *(bool_map_[key]) = ToBool(value);
   } else if (int_map_.end() != int_map_.find(key)) {
@@ -659,8 +542,9 @@ bool ParseOptions::SetOption(const std::string &key, const std::string &value,
     *(double_map_[key]) = ToDouble(value);
   } else if (string_map_.end() != string_map_.find(key)) {
     if (!has_equal_sign) {
-      SHERPA_ONNX_LOG(FATAL)
-          << "Invalid option --" << key << " (option format is --x=y).";
+      SHERPA_ONNX_LOGE("Invalid option --%s (option format is --x=y).",
+                       key.c_str());
+      exit(-1);
     }
     *(string_map_[key]) = value;
   } else {
@@ -683,37 +567,46 @@ bool ParseOptions::ToBool(std::string str) const {
   }
   // if it is neither true nor false:
   PrintUsage(true);
-  SHERPA_ONNX_LOG(FATAL)
-      << "Invalid format for boolean argument [expected true or false]: "
-      << str;
+  SHERPA_ONNX_LOGE(
+      "Invalid format for boolean argument [expected true or false]: %s",
+      str.c_str());
+  exit(-1);
   return false;  // never reached
 }
 
 int32_t ParseOptions::ToInt(const std::string &str) const {
   int32_t ret = 0;
-  if (!ConvertStringToInteger(str, &ret))
-    SHERPA_ONNX_LOG(FATAL) << "Invalid integer option \"" << str << "\"";
+  if (!ConvertStringToInteger(str, &ret)) {
+    SHERPA_ONNX_LOGE("Invalid integer option \"%s\"", str.c_str());
+    exit(-1);
+  }
   return ret;
 }
 
 uint32_t ParseOptions::ToUint(const std::string &str) const {
   uint32_t ret = 0;
-  if (!ConvertStringToInteger(str, &ret))
-    SHERPA_ONNX_LOG(FATAL) << "Invalid integer option \"" << str << "\"";
+  if (!ConvertStringToInteger(str, &ret)) {
+    SHERPA_ONNX_LOGE("Invalid integer option \"%s\"", str.c_str());
+    exit(-1);
+  }
   return ret;
 }
 
 float ParseOptions::ToFloat(const std::string &str) const {
   float ret;
-  if (!ConvertStringToReal(str, &ret))
-    SHERPA_ONNX_LOG(FATAL) << "Invalid floating-point option \"" << str << "\"";
+  if (!ConvertStringToReal(str, &ret)) {
+    SHERPA_ONNX_LOGE("Invalid floating-point option \"%s\"", str.c_str());
+    exit(-1);
+  }
   return ret;
 }
 
 double ParseOptions::ToDouble(const std::string &str) const {
   double ret;
-  if (!ConvertStringToReal(str, &ret))
-    SHERPA_ONNX_LOG(FATAL) << "Invalid floating-point option \"" << str << "\"";
+  if (!ConvertStringToReal(str, &ret)) {
+    SHERPA_ONNX_LOGE("Invalid floating-point option \"%s\"", str.c_str());
+    exit(-1);
+  }
   return ret;
 }
 

@@ -7,7 +7,11 @@
 
 #include <assert.h>
 
+#include <algorithm>
+#include <limits>
+#include <sstream>
 #include <string>
+#include <unordered_map>
 #include <vector>
 
 // This file is copied/modified from
@@ -15,6 +19,102 @@
 
 namespace sherpa_onnx {
 
+// copied from kaldi/src/util/text-util.cc
+template <class T>
+class NumberIstream {
+ public:
+  explicit NumberIstream(std::istream &i) : in_(i) {}
+
+  NumberIstream &operator>>(T &x) {
+    if (!in_.good()) return *this;
+    in_ >> x;
+    if (!in_.fail() && RemainderIsOnlySpaces()) return *this;
+    return ParseOnFail(&x);
+  }
+
+ private:
+  std::istream &in_;
+
+  bool RemainderIsOnlySpaces() {
+    if (in_.tellg() != std::istream::pos_type(-1)) {
+      std::string rem;
+      in_ >> rem;
+
+      if (rem.find_first_not_of(' ') != std::string::npos) {
+        // there is not only spaces
+        return false;
+      }
+    }
+
+    in_.clear();
+    return true;
+  }
+
+  NumberIstream &ParseOnFail(T *x) {
+    std::string str;
+    in_.clear();
+    in_.seekg(0);
+    // If the stream is broken even before trying
+    // to read from it or if there are many tokens,
+    // it's pointless to try.
+    if (!(in_ >> str) || !RemainderIsOnlySpaces()) {
+      in_.setstate(std::ios_base::failbit);
+      return *this;
+    }
+
+    std::unordered_map<std::string, T> inf_nan_map;
+    // we'll keep just uppercase values.
+    inf_nan_map["INF"] = std::numeric_limits<T>::infinity();
+    inf_nan_map["+INF"] = std::numeric_limits<T>::infinity();
+    inf_nan_map["-INF"] = -std::numeric_limits<T>::infinity();
+    inf_nan_map["INFINITY"] = std::numeric_limits<T>::infinity();
+    inf_nan_map["+INFINITY"] = std::numeric_limits<T>::infinity();
+    inf_nan_map["-INFINITY"] = -std::numeric_limits<T>::infinity();
+    inf_nan_map["NAN"] = std::numeric_limits<T>::quiet_NaN();
+    inf_nan_map["+NAN"] = std::numeric_limits<T>::quiet_NaN();
+    inf_nan_map["-NAN"] = -std::numeric_limits<T>::quiet_NaN();
+    // MSVC
+    inf_nan_map["1.#INF"] = std::numeric_limits<T>::infinity();
+    inf_nan_map["-1.#INF"] = -std::numeric_limits<T>::infinity();
+    inf_nan_map["1.#QNAN"] = std::numeric_limits<T>::quiet_NaN();
+    inf_nan_map["-1.#QNAN"] = -std::numeric_limits<T>::quiet_NaN();
+
+    std::transform(str.begin(), str.end(), str.begin(), ::toupper);
+
+    if (inf_nan_map.find(str) != inf_nan_map.end()) {
+      *x = inf_nan_map[str];
+    } else {
+      in_.setstate(std::ios_base::failbit);
+    }
+
+    return *this;
+  }
+};
+
+/// ConvertStringToReal converts a string into either float or double
+/// and returns false if there was any kind of problem (i.e. the string
+/// was not a floating point number or contained extra non-whitespace junk).
+/// Be careful- this function will successfully read inf's or nan's.
+template <typename T>
+bool ConvertStringToReal(const std::string &str, T *out) {
+  std::istringstream iss(str);
+
+  NumberIstream<T> i(iss);
+
+  i >> *out;
+
+  if (iss.fail()) {
+    // Number conversion failed.
+    return false;
+  }
+
+  return true;
+}
+
+template bool ConvertStringToReal<float>(const std::string &str, float *out);
+
+template bool ConvertStringToReal<double>(const std::string &str, double *out);
+
 void SplitStringToVector(const std::string &full, const char *delim,
                          bool omit_empty_strings,
                          std::vector<std::string> *out) {
@@ -43,7 +143,9 @@ bool SplitStringToFloats(const std::string &full, const char *delim,
   out->resize(split.size());
   for (size_t i = 0; i < split.size(); ++i) {
     // assume atof never fails
-    (*out)[i] = atof(split[i].c_str());
+    F f = 0;
+    if (!ConvertStringToReal(split[i], &f)) return false;
+    (*out)[i] = f;
   }
   return true;
 }

@@ -6,7 +6,9 @@
 #define SHERPA_ONNX_CSRC_TEXT_UTILS_H_
 #include <stdlib.h>
 
+#include <limits>
 #include <string>
+#include <type_traits>
 #include <vector>
 
 #ifdef _MSC_VER
@@ -21,6 +23,32 @@
 
 namespace sherpa_onnx {
 
+/// Converts a string into an integer via strtoll and returns false if there was
+/// any kind of problem (i.e. the string was not an integer or contained extra
+/// non-whitespace junk, or the integer was too large to fit into the type it is
+/// being converted into).  Only sets *out if everything was OK and it returns
+/// true.
+template <class Int>
+bool ConvertStringToInteger(const std::string &str, Int *out) {
+  // copied from kaldi/src/util/text-util.h
+  static_assert(std::is_integral<Int>::value, "");
+  const char *this_str = str.c_str();
+  char *end = nullptr;
+  errno = 0;
+  int64_t i = SHERPA_ONNX_STRTOLL(this_str, &end);
+  if (end != this_str) {
+    while (isspace(*end)) ++end;
+  }
+  if (end == this_str || *end != '\0' || errno != 0) return false;
+  Int iInt = static_cast<Int>(i);
+  if (static_cast<int64_t>(iInt) != i ||
+      (i < 0 && !std::numeric_limits<Int>::is_signed)) {
+    return false;
+  }
+  *out = iInt;
+  return true;
+}
+
 /// Split a string using any of the single character delimiters.
 /// If omit_empty_strings == true, the output will contain any
 /// nonempty strings after splitting on any of the
@@ -86,6 +114,10 @@ bool SplitStringToFloats(const std::string &full, const char *delim,
                          bool omit_empty_strings,  // typically false
                          std::vector<F> *out);
 
+// This is defined for F = float and double.
+template <typename T>
+bool ConvertStringToReal(const std::string &str, T *out);
+
 }  // namespace sherpa_onnx
 
 #endif  // SHERPA_ONNX_CSRC_TEXT_UTILS_H_