Release GIL to support multithreading in websocket servers. (#451)

python-api-examples/non_streaming_server.py

@@ -414,7 +414,7 @@ def get_args():
     parser.add_argument(
         "--max-batch-size",
         type=int,
-        default=25,
+        default=3,
         help="""Max batch size for computation. Note if there are not enough
         requests in the queue, it will wait for max_wait_ms time. After that,
         even if there are not enough requests, it still sends the
@@ -459,7 +459,7 @@ def get_args():
     parser.add_argument(
         "--max-active-connections",
         type=int,
-        default=500,
+        default=200,
         help="""Maximum number of active connections. The server will refuse
         to accept new connections once the current number of active connections
         equals to this limit.
@@ -533,6 +533,7 @@ def __init__(
         self.certificate = certificate
         self.http_server = HttpServer(doc_root)
 
+        self.nn_pool_size = nn_pool_size
         self.nn_pool = ThreadPoolExecutor(
             max_workers=nn_pool_size,
             thread_name_prefix="nn",
@@ -604,7 +605,9 @@ async def process_request(
     async def run(self, port: int):
         logging.info("started")
 
-        task = asyncio.create_task(self.stream_consumer_task())
+        tasks = []
+        for i in range(self.nn_pool_size):
+            tasks.append(asyncio.create_task(self.stream_consumer_task()))
 
         if self.certificate:
             logging.info(f"Using certificate: {self.certificate}")
@@ -636,7 +639,7 @@ async def run(self, port: int):
 
             await asyncio.Future()  # run forever
 
-        await task  # not reachable
+        await asyncio.gather(*tasks)  # not reachable
 
     async def recv_audio_samples(
         self,
@@ -722,6 +725,7 @@ async def stream_consumer_task(self):
                     batch.append(item)
             except asyncio.QueueEmpty:
                 pass
+
             stream_list = [b[0] for b in batch]
             future_list = [b[1] for b in batch]
 

python-api-examples/streaming_server.py

@@ -296,7 +296,7 @@ def get_args():
     parser.add_argument(
         "--max-batch-size",
         type=int,
-        default=50,
+        default=3,
         help="""Max batch size for computation. Note if there are not enough
         requests in the queue, it will wait for max_wait_ms time. After that,
         even if there are not enough requests, it still sends the
@@ -334,7 +334,7 @@ def get_args():
     parser.add_argument(
         "--max-active-connections",
         type=int,
-        default=500,
+        default=200,
         help="""Maximum number of active connections. The server will refuse
         to accept new connections once the current number of active connections
         equals to this limit.
@@ -478,6 +478,7 @@ def __init__(
         self.certificate = certificate
         self.http_server = HttpServer(doc_root)
 
+        self.nn_pool_size = nn_pool_size
         self.nn_pool = ThreadPoolExecutor(
             max_workers=nn_pool_size,
             thread_name_prefix="nn",
@@ -591,7 +592,9 @@ async def process_request(
         return status, header, response
 
     async def run(self, port: int):
-        task = asyncio.create_task(self.stream_consumer_task())
+        tasks = []
+        for i in range(self.nn_pool_size):
+            tasks.append(asyncio.create_task(self.stream_consumer_task()))
 
         if self.certificate:
             logging.info(f"Using certificate: {self.certificate}")
@@ -629,7 +632,7 @@ async def run(self, port: int):
 
             await asyncio.Future()  # run forever
 
-        await task  # not reachable
+        await asyncio.gather(*tasks)  # not reachable
 
     async def handle_connection(
         self,

sherpa-onnx/python/csrc/circular-buffer.cc

@@ -19,10 +19,12 @@ void PybindCircularBuffer(py::module *m) {
           [](PyClass &self, const std::vector<float> &samples) {
             self.Push(samples.data(), samples.size());
           },
-          py::arg("samples"))
-      .def("get", &PyClass::Get, py::arg("start_index"), py::arg("n"))
-      .def("pop", &PyClass::Pop, py::arg("n"))
-      .def("reset", &PyClass::Reset)
+          py::arg("samples"), py::call_guard<py::gil_scoped_release>())
+      .def("get", &PyClass::Get, py::arg("start_index"), py::arg("n"),
+           py::call_guard<py::gil_scoped_release>())
+      .def("pop", &PyClass::Pop, py::arg("n"),
+           py::call_guard<py::gil_scoped_release>())
+      .def("reset", &PyClass::Reset, py::call_guard<py::gil_scoped_release>())
       .def_property_readonly("size", &PyClass::Size)
       .def_property_readonly("head", &PyClass::Head)
       .def_property_readonly("tail", &PyClass::Tail);

sherpa-onnx/python/csrc/offline-recognizer.cc

@@ -41,19 +41,24 @@ void PybindOfflineRecognizer(py::module *m) {
   using PyClass = OfflineRecognizer;
   py::class_<PyClass>(*m, "OfflineRecognizer")
       .def(py::init<const OfflineRecognizerConfig &>(), py::arg("config"))
-      .def("create_stream",
-           [](const PyClass &self) { return self.CreateStream(); })
+      .def(
+          "create_stream",
+          [](const PyClass &self) { return self.CreateStream(); },
+          py::call_guard<py::gil_scoped_release>())
       .def(
           "create_stream",
           [](PyClass &self, const std::string &hotwords) {
             return self.CreateStream(hotwords);
           },
-          py::arg("hotwords"))
-      .def("decode_stream", &PyClass::DecodeStream)
-      .def("decode_streams",
-           [](const PyClass &self, std::vector<OfflineStream *> ss) {
-             self.DecodeStreams(ss.data(), ss.size());
-           });
+          py::arg("hotwords"), py::call_guard<py::gil_scoped_release>())
+      .def("decode_stream", &PyClass::DecodeStream,
+           py::call_guard<py::gil_scoped_release>())
+      .def(
+          "decode_streams",
+          [](const PyClass &self, std::vector<OfflineStream *> ss) {
+            self.DecodeStreams(ss.data(), ss.size());
+          },
+          py::call_guard<py::gil_scoped_release>());
 }
 
 }  // namespace sherpa_onnx

sherpa-onnx/python/csrc/offline-stream.cc

@@ -50,9 +50,20 @@ void PybindOfflineStream(py::module *m) {
       .def(
           "accept_waveform",
           [](PyClass &self, float sample_rate, py::array_t<float> waveform) {
+#if 0
+            auto report_gil_status = []() {
+              auto is_gil_held = false;
+              if (auto tstate = py::detail::get_thread_state_unchecked())
+                is_gil_held = (tstate == PyGILState_GetThisThreadState());
+
+              return is_gil_held ? "GIL held" : "GIL released";
+            };
+            std::cout << report_gil_status() << "\n";
+#endif
             self.AcceptWaveform(sample_rate, waveform.data(), waveform.size());
           },
-          py::arg("sample_rate"), py::arg("waveform"), kAcceptWaveformUsage)
+          py::arg("sample_rate"), py::arg("waveform"), kAcceptWaveformUsage,
+          py::call_guard<py::gil_scoped_release>())
       .def_property_readonly("result", &PyClass::GetResult);
 }
 

sherpa-onnx/python/csrc/offline-tts.cc

@@ -45,7 +45,7 @@ void PybindOfflineTts(py::module *m) {
   py::class_<PyClass>(*m, "OfflineTts")
       .def(py::init<const OfflineTtsConfig &>(), py::arg("config"))
       .def("generate", &PyClass::Generate, py::arg("text"), py::arg("sid") = 0,
-           py::arg("speed") = 1.0);
+           py::arg("speed") = 1.0, py::call_guard<py::gil_scoped_release>());
 }
 
 }  // namespace sherpa_onnx

sherpa-onnx/python/csrc/online-recognizer.cc

@@ -54,23 +54,31 @@ void PybindOnlineRecognizer(py::module *m) {
   using PyClass = OnlineRecognizer;
   py::class_<PyClass>(*m, "OnlineRecognizer")
       .def(py::init<const OnlineRecognizerConfig &>(), py::arg("config"))
-      .def("create_stream",
-           [](const PyClass &self) { return self.CreateStream(); })
+      .def(
+          "create_stream",
+          [](const PyClass &self) { return self.CreateStream(); },
+          py::call_guard<py::gil_scoped_release>())
       .def(
           "create_stream",
           [](PyClass &self, const std::string &hotwords) {
             return self.CreateStream(hotwords);
           },
-          py::arg("hotwords"))
-      .def("is_ready", &PyClass::IsReady)
-      .def("decode_stream", &PyClass::DecodeStream)
-      .def("decode_streams",
-           [](PyClass &self, std::vector<OnlineStream *> ss) {
-             self.DecodeStreams(ss.data(), ss.size());
-           })
-      .def("get_result", &PyClass::GetResult)
-      .def("is_endpoint", &PyClass::IsEndpoint)
-      .def("reset", &PyClass::Reset);
+          py::arg("hotwords"), py::call_guard<py::gil_scoped_release>())
+      .def("is_ready", &PyClass::IsReady,
+           py::call_guard<py::gil_scoped_release>())
+      .def("decode_stream", &PyClass::DecodeStream,
+           py::call_guard<py::gil_scoped_release>())
+      .def(
+          "decode_streams",
+          [](PyClass &self, std::vector<OnlineStream *> ss) {
+            self.DecodeStreams(ss.data(), ss.size());
+          },
+          py::call_guard<py::gil_scoped_release>())
+      .def("get_result", &PyClass::GetResult,
+           py::call_guard<py::gil_scoped_release>())
+      .def("is_endpoint", &PyClass::IsEndpoint,
+           py::call_guard<py::gil_scoped_release>())
+      .def("reset", &PyClass::Reset, py::call_guard<py::gil_scoped_release>());
 }
 
 }  // namespace sherpa_onnx

sherpa-onnx/python/csrc/online-stream.cc

@@ -28,8 +28,10 @@ void PybindOnlineStream(py::module *m) {
           [](PyClass &self, float sample_rate, py::array_t<float> waveform) {
             self.AcceptWaveform(sample_rate, waveform.data(), waveform.size());
           },
-          py::arg("sample_rate"), py::arg("waveform"), kAcceptWaveformUsage)
-      .def("input_finished", &PyClass::InputFinished);
+          py::arg("sample_rate"), py::arg("waveform"), kAcceptWaveformUsage,
+          py::call_guard<py::gil_scoped_release>())
+      .def("input_finished", &PyClass::InputFinished,
+           py::call_guard<py::gil_scoped_release>());
 }
 
 }  // namespace sherpa_onnx

sherpa-onnx/python/csrc/vad-model.cc

@@ -13,17 +13,21 @@ namespace sherpa_onnx {
 void PybindVadModel(py::module *m) {
   using PyClass = VadModel;
   py::class_<PyClass>(*m, "VadModel")
-      .def_static("create", &PyClass::Create, py::arg("config"))
-      .def("reset", &PyClass::Reset)
+      .def_static("create", &PyClass::Create, py::arg("config"),
+                  py::call_guard<py::gil_scoped_release>())
+      .def("reset", &PyClass::Reset, py::call_guard<py::gil_scoped_release>())
       .def(
           "is_speech",
           [](PyClass &self, const std::vector<float> &samples) -> bool {
             return self.IsSpeech(samples.data(), samples.size());
           },
-          py::arg("samples"))
-      .def("window_size", &PyClass::WindowSize)
-      .def("min_silence_duration_samples", &PyClass::MinSilenceDurationSamples)
-      .def("min_speech_duration_samples", &PyClass::MinSpeechDurationSamples);
+          py::arg("samples"), py::call_guard<py::gil_scoped_release>())
+      .def("window_size", &PyClass::WindowSize,
+           py::call_guard<py::gil_scoped_release>())
+      .def("min_silence_duration_samples", &PyClass::MinSilenceDurationSamples,
+           py::call_guard<py::gil_scoped_release>())
+      .def("min_speech_duration_samples", &PyClass::MinSpeechDurationSamples,
+           py::call_guard<py::gil_scoped_release>());
 }
 
 }  // namespace sherpa_onnx

sherpa-onnx/python/csrc/voice-activity-detector.cc

@@ -30,11 +30,12 @@ void PybindVoiceActivityDetector(py::module *m) {
           [](PyClass &self, const std::vector<float> &samples) {
             self.AcceptWaveform(samples.data(), samples.size());
           },
-          py::arg("samples"))
-      .def("empty", &PyClass::Empty)
-      .def("pop", &PyClass::Pop)
-      .def("is_speech_detected", &PyClass::IsSpeechDetected)
-      .def("reset", &PyClass::Reset)
+          py::arg("samples"), py::call_guard<py::gil_scoped_release>())
+      .def("empty", &PyClass::Empty, py::call_guard<py::gil_scoped_release>())
+      .def("pop", &PyClass::Pop, py::call_guard<py::gil_scoped_release>())
+      .def("is_speech_detected", &PyClass::IsSpeechDetected,
+           py::call_guard<py::gil_scoped_release>())
+      .def("reset", &PyClass::Reset, py::call_guard<py::gil_scoped_release>())
       .def_property_readonly("front", &PyClass::Front);
 }