learning-loop-node 0.13.3__py3-none-any.whl → 0.13.5__py3-none-any.whl

learning_loop_node/detector/detector_logic.py

@@ -6,6 +6,7 @@ import numpy as np
 
 from ..data_classes import ImageMetadata, ModelInformation
 from ..globals import GLOBALS
+from .exceptions import NodeNeedsRestartError
 
 
 class DetectorLogic():
@@ -14,6 +15,8 @@ class DetectorLogic():
         self.model_format: str = model_format
         self._model_info: Optional[ModelInformation] = None
 
+        self._remaining_init_attempts: int = 2
+
     async def soft_reload(self):
         self._model_info = None
 
@@ -27,20 +30,22 @@ class DetectorLogic():
     def is_initialized(self) -> bool:
         return self._model_info is not None
 
-    def load_model(self):
+    def load_model_info_and_init_model(self):
         logging.info('Loading model from %s', GLOBALS.data_folder)
-        model_info = ModelInformation.load_from_disk(f'{GLOBALS.data_folder}/model')
-        if model_info is None:
+        self._model_info = ModelInformation.load_from_disk(f'{GLOBALS.data_folder}/model')
+        if self._model_info is None:
             logging.error('No model found')
             self._model_info = None
-            raise Exception('No model found')
         try:
-            self._model_info = model_info
             self.init()
             logging.info('Successfully loaded model %s', self._model_info)
+            self._remaining_init_attempts = 2
         except Exception:
+            self._remaining_init_attempts -= 1
             self._model_info = None
-            logging.error('Could not init model %s', model_info)
+            logging.error('Could not init model %s. Retries left: %s', self._model_info, self._remaining_init_attempts)
+            if self._remaining_init_attempts == 0:
+                raise NodeNeedsRestartError('Could not init model') from None
             raise
 
     @abstractmethod

learning_loop_node/detector/detector_node.py

@@ -3,6 +3,7 @@ import contextlib
 import os
 import shutil
 import subprocess
+import sys
 from dataclasses import asdict
 from datetime import datetime
 from threading import Thread
@@ -23,6 +24,7 @@ from ..globals import GLOBALS
 from ..helpers import environment_reader
 from ..node import Node
 from .detector_logic import DetectorLogic
+from .exceptions import NodeNeedsRestartError
 from .inbox_filter.relevance_filter import RelevanceFilter
 from .outbox import Outbox
 from .rest import about as rest_about
@@ -170,13 +172,13 @@ class DetectorNode(Node):
 
         # simulate startup
         await self.detector_logic.soft_reload()
-        self.detector_logic.load_model()
+        self.detector_logic.load_model_info_and_init_model()
         self.operation_mode = OperationMode.Idle
 
     async def on_startup(self) -> None:
         try:
             self.outbox.ensure_continuous_upload()
-            self.detector_logic.load_model()
+            self.detector_logic.load_model_info_and_init_model()
         except Exception:
             self.log.exception("error during 'startup'")
         self.operation_mode = OperationMode.Idle
@@ -200,8 +202,11 @@ class DetectorNode(Node):
         """The DetectorNode acts as a SocketIO server. This method sets up the server and defines the event handlers."""
         # pylint: disable=unused-argument
 
-        # Initialize the Socket.IO server
-        self.sio = socketio.AsyncServer(async_mode='asgi')
+        # Initialize the Socket.IO server with 20MB buffer size
+        self.sio = socketio.AsyncServer(
+            async_mode='asgi',
+            max_http_buffer_size=2e7,  # 20MB
+        )
         # Initialize and mount the ASGI app
         self.sio_app = socketio.ASGIApp(self.sio, socketio_path='/socket.io')
         self.mount('/ws', self.sio_app)
@@ -318,9 +323,12 @@ class DetectorNode(Node):
                 self.log.debug('not checking for updates; no target model selected')
                 return
 
-            current_version = self.detector_logic._model_info.version if self.detector_logic._model_info is not None else None  # pylint: disable=protected-access
+            if self.detector_logic.is_initialized:
+                current_version = self.detector_logic.model_info.version
+            else:
+                current_version = None
 
-            if not self.detector_logic.is_initialized or self.target_model.version != current_version:
+            if current_version != self.target_model.version:
                 self.log.info('Current model "%s" needs to be updated to %s',
                               current_version or "-", self.target_model.version)
 
@@ -346,7 +354,10 @@ class DetectorNode(Node):
                     self.log.info('Updated symlink for model to %s', os.readlink(model_symlink))
 
                     try:
-                        self.detector_logic.load_model()
+                        self.detector_logic.load_model_info_and_init_model()
+                    except NodeNeedsRestartError:
+                        self.log.error('Node needs restart')
+                        sys.exit(0)
                     except Exception:
                         self.log.exception('Could not load model, will retry download on next check')
                         shutil.rmtree(target_model_folder, ignore_errors=True)

learning_loop_node/detector/exceptions.py

@@ -0,0 +1,7 @@
+
+
+class NodeNeedsRestartError(Exception):
+    '''
+    NodeNeedsRestartError is raised when the node needs to be restarted.
+    This is e.g. the case when the GPU is not available anymore.
+    '''

{learning_loop_node-0.13.3.dist-info → learning_loop_node-0.13.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: learning-loop-node
-Version: 0.13.3
+Version: 0.13.5
 Summary: Python Library for Nodes which connect to the Zauberzeug Learning Loop
 Home-page: https://github.com/zauberzeug/learning_loop_node
 License: MIT

{learning_loop_node-0.13.3.dist-info → learning_loop_node-0.13.5.dist-info}/RECORD

@@ -11,8 +11,9 @@ learning_loop_node/data_classes/socket_response.py,sha256=tIdt-oYf6ULoJIDYQCecNM
 learning_loop_node/data_classes/training.py,sha256=FFPsr2AA7ynYz39MLZaFJ0sF_9Axll5HHbAA8nnirp0,5726
 learning_loop_node/data_exchanger.py,sha256=IG5ki3f3IsVuXbyw6q_gUIakgv-GMT6e9nhOhzjKgW4,9055
 learning_loop_node/detector/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-learning_loop_node/detector/detector_logic.py,sha256=FhGbu0mdF0tW0Gg8cr8xM6ZmZzEigGf0IcAWBjoxFrs,2191
-learning_loop_node/detector/detector_node.py,sha256=TMypE1Z0KnuPcJ9wLiflhdvd9oe0t86gAQyK7UvuDpA,25060
+learning_loop_node/detector/detector_logic.py,sha256=M6iGCj6OMscd460NUyqIfig-gNK6pNcnBZz-1iKBSLg,2506
+learning_loop_node/detector/detector_node.py,sha256=u_SWqwELmMKHJSC_R6KFVemteNBc1KaNu9WeCYVl9Js,25391
+learning_loop_node/detector/exceptions.py,sha256=C6KbNPlSbtfgDrZx2Hbhm7Suk9jVoR3fMRCO0CkrMsQ,196
 learning_loop_node/detector/inbox_filter/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 learning_loop_node/detector/inbox_filter/cam_observation_history.py,sha256=1PHgXRrhSQ34HSFw7mdX8ndRxHf_i1aP5nXXnrZxhAY,3312
 learning_loop_node/detector/inbox_filter/relevance_filter.py,sha256=NPEmrAtuGjIWCtHS0B3zDmnYWkhVFCLbd_7RUp08_AM,1372
@@ -96,6 +97,6 @@ learning_loop_node/trainer/test_executor.py,sha256=6BVGDN_6f5GEMMEvDLSG1yzMybSvg
 learning_loop_node/trainer/trainer_logic.py,sha256=eK-01qZzi10UjLMCQX8vy5eW2FoghPj3rzzDC-s3Si4,8792
 learning_loop_node/trainer/trainer_logic_generic.py,sha256=RQqon8JIVzxaNh0KdEe6tMxebsY0DgZllEohHR-AgqU,26846
 learning_loop_node/trainer/trainer_node.py,sha256=Dl4ZQAjjXQggibeBjvhXAoFClw1ZX2Kkt3v_fjrJnCI,4508
-learning_loop_node-0.13.3.dist-info/METADATA,sha256=zB6JXpeyUlyfBVDgEV8kABmdQ2QgqjaCn2U8wib5288,12761
-learning_loop_node-0.13.3.dist-info/WHEEL,sha256=WGfLGfLX43Ei_YORXSnT54hxFygu34kMpcQdmgmEwCQ,88
-learning_loop_node-0.13.3.dist-info/RECORD,,
+learning_loop_node-0.13.5.dist-info/METADATA,sha256=QiQQwfbMoEjN14j_IDkjuWccGrRa__pOIo1NKqMbW7s,12761
+learning_loop_node-0.13.5.dist-info/WHEEL,sha256=WGfLGfLX43Ei_YORXSnT54hxFygu34kMpcQdmgmEwCQ,88
+learning_loop_node-0.13.5.dist-info/RECORD,,