High performance self-hosted photo and video management solution.
TypeScript backup-toolfluttergoogle-photosgoogle-photos-alternativejavascriptmobile-appnestjsnodejsphoto-galleryphotosphotos-managementself-hostedsveltesveltekittypescriptvideos
fix(ml): race when submitting to rknn execution queue (#31143)
Wenbin26a25f0c3abb
diff --git a/machine-learning/immich_ml/sessions/rknn/__init__.py b/machine-learning/immich_ml/sessions/rknn/__init__.pyindex 721fe5bbda..0d95e6cf94 100644--- a/machine-learning/immich_ml/sessions/rknn/__init__.py+++ b/machine-learning/immich_ml/sessions/rknn/__init__.py@@ -64,8 +64,7 @@ class RknnSession:run_options: Any = None,) -> list[NDArray[np.float32]]:input_data: list[NDArray[np.float32]] = [np.ascontiguousarray(v) for v in input_feed.values()]- self.rknnpool.put(input_data)- res = self.rknnpool.get()+ res = self.rknnpool.run(input_data)if res is None:raise RuntimeError("RKNN inference failed!")return resdiff --git a/machine-learning/immich_ml/sessions/rknn/rknnpool.py b/machine-learning/immich_ml/sessions/rknn/rknnpool.pyindex fd0af8bcc4..1caec1abba 100644--- a/machine-learning/immich_ml/sessions/rknn/rknnpool.py+++ b/machine-learning/immich_ml/sessions/rknn/rknnpool.py@@ -2,9 +2,9 @@# Following Apache License 2.0import logging-from concurrent.futures import Future, ThreadPoolExecutor+import threading+from concurrent.futures import ThreadPoolExecutorfrom pathlib import Path-from queue import Queuefrom typing import Callableimport numpy as np@@ -66,20 +66,18 @@ class RknnPoolExecutor:func: Callable[["RKNNLite", list[NDArray[np.float32]]], list[NDArray[np.float32]]],) -> None:self.tpes = tpes- self.queue: Queue[Future[list[NDArray[np.float32]]]] = Queue()self.rknn_pool = [init_rknn(model_path) for _ in range(tpes)]self.pool = ThreadPoolExecutor(max_workers=tpes)self.func = funcself.num = 0+ self.lock = threading.Lock()- def put(self, inputs: list[NDArray[np.float32]]) -> None:- self.queue.put(self.pool.submit(self.func, self.rknn_pool[self.num % self.tpes], inputs))- self.num += 1+ def run(self, inputs: list[NDArray[np.float32]]) -> list[NDArray[np.float32]]:+ with self.lock:+ idx = self.num % self.tpes+ self.num += 1- def get(self) -> list[NDArray[np.float32]] | None:- if self.queue.empty():- return None- fut = self.queue.get()+ fut = self.pool.submit(self.func, self.rknn_pool[idx], inputs)return fut.result()def release(self) -> None:diff --git a/machine-learning/test_main.py b/machine-learning/test_main.pyindex 9e3fe50cba..a50bea00b7 100644--- a/machine-learning/test_main.py+++ b/machine-learning/test_main.py@@ -564,7 +564,7 @@ class TestRknnSession:session.run(None, input_feed)- rknn_session.return_value.put.assert_called_once_with([input1, input2])+ rknn_session.return_value.run.assert_called_once_with([input1, input2])assert np_spy.call_count == 2np_spy.assert_has_calls([mock.call(input1), mock.call(input2)])