learning-loop-node 0.20.1__tar.gz → 0.21.0__tar.gz

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: learning_loop_node
-Version: 0.20.1
+Version: 0.21.0
 Summary: Python Library for Nodes which connect to the Zauberzeug Learning Loop
 Author-email: Zauberzeug GmbH <info@zauberzeug.com>
 License: MIT
@@ -140,6 +140,27 @@ The endpoint returns None if the upload was successful and an error message othe
 
 For both ways to upload an image, the tag `picked_by_system` is automatically added to the image metadata.
 
+### Node and model information
+
+The detector has a REST endpoint that reports what the node is currently doing and which model it has loaded:
+
+`curl http://localhost/about`
+
+The response contains:
+
+- `operation_mode`: the current operation mode (`startup`, `idle` or `detecting`)
+- `state`: the state of the node (e.g. `idle`, `online`, `detecting`)
+- `model_info`: information about the loaded model (id, organization, project, version, categories, resolution, model size), or `null` if no model is loaded yet
+- `target_model`: the model version the detector is trying to load, or `null` if none is set
+- `version_control`: the model versioning mode (`follow_loop`, `specific_version` or `pause`)
+
+The same information is available via the socketio event `about`:
+`sio.emit('about')`
+
+Use this endpoint if you need the model uuid or its categories. If you only need the version numbers, use `/model_version` (see below).
+
+All REST endpoints of the node are also listed in the automatically generated API documentation at `http://localhost/docs`.
+
 ### Changing the model versioning mode
 
 The detector can be configured to one of the following behaviors:
@@ -167,6 +188,22 @@ The model versioning configuration can also be changed via a socketio event:
 There is also a GET endpoint to fetch the current model versioning configuration:
 `sio.emit('get_model_version')` or `curl http://localhost/model_version`
 
+### Changing the operation mode
+
+The operation mode controls whether the detector may load a new model:
+
+- `startup`: used until the first model is loaded
+- `idle`: the detector checks for and performs model updates
+- `detecting`: the detector runs detections and does not load new models
+
+The mode can be read and changed via a REST endpoint. Example Usage:
+
+- Fetch the current operation mode: `curl http://localhost/operation_mode`
+- Allow model updates: `curl -X PUT -d "idle" http://localhost/operation_mode`
+- Block model updates: `curl -X PUT -d "detecting" http://localhost/operation_mode`
+
+The GET request returns the mode as plain text. The PUT request returns status code 422 if the given value is not one of the three modes above.
+
 ### Changing the outbox mode
 
 If the autoupload is set to `all` or `filtered` (selected) images and the corresponding detections are saved on HDD (the outbox). A background thread will upload the images and detections to the Learning Loop. The outbox is located in the `outbox` folder in the root directory of the node. The outbox can be cleared by deleting the files in the folder.

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/README.md

@@ -104,6 +104,27 @@ The endpoint returns None if the upload was successful and an error message othe
 
 For both ways to upload an image, the tag `picked_by_system` is automatically added to the image metadata.
 
+### Node and model information
+
+The detector has a REST endpoint that reports what the node is currently doing and which model it has loaded:
+
+`curl http://localhost/about`
+
+The response contains:
+
+- `operation_mode`: the current operation mode (`startup`, `idle` or `detecting`)
+- `state`: the state of the node (e.g. `idle`, `online`, `detecting`)
+- `model_info`: information about the loaded model (id, organization, project, version, categories, resolution, model size), or `null` if no model is loaded yet
+- `target_model`: the model version the detector is trying to load, or `null` if none is set
+- `version_control`: the model versioning mode (`follow_loop`, `specific_version` or `pause`)
+
+The same information is available via the socketio event `about`:
+`sio.emit('about')`
+
+Use this endpoint if you need the model uuid or its categories. If you only need the version numbers, use `/model_version` (see below).
+
+All REST endpoints of the node are also listed in the automatically generated API documentation at `http://localhost/docs`.
+
 ### Changing the model versioning mode
 
 The detector can be configured to one of the following behaviors:
@@ -131,6 +152,22 @@ The model versioning configuration can also be changed via a socketio event:
 There is also a GET endpoint to fetch the current model versioning configuration:
 `sio.emit('get_model_version')` or `curl http://localhost/model_version`
 
+### Changing the operation mode
+
+The operation mode controls whether the detector may load a new model:
+
+- `startup`: used until the first model is loaded
+- `idle`: the detector checks for and performs model updates
+- `detecting`: the detector runs detections and does not load new models
+
+The mode can be read and changed via a REST endpoint. Example Usage:
+
+- Fetch the current operation mode: `curl http://localhost/operation_mode`
+- Allow model updates: `curl -X PUT -d "idle" http://localhost/operation_mode`
+- Block model updates: `curl -X PUT -d "detecting" http://localhost/operation_mode`
+
+The GET request returns the mode as plain text. The PUT request returns status code 422 if the given value is not one of the three modes above.
+
 ### Changing the outbox mode
 
 If the autoupload is set to `all` or `filtered` (selected) images and the corresponding detections are saved on HDD (the outbox). A background thread will upload the images and detections to the Learning Loop. The outbox is located in the `outbox` folder in the root directory of the node. The outbox can be cleared by deleting the files in the folder.

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/learning_loop_node/data_classes/general.py

@@ -185,3 +185,4 @@ class DetectorStatus():
     target_model: Optional[str]
     errors: Dict
     operation_mode: str
+    node_version: Optional[str] = None

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/learning_loop_node/detector/detector_node.py

@@ -52,6 +52,7 @@ class DetectorNode(Node):
                  uuid: Optional[str] = None, use_backdoor_controls: bool = False) -> None:
         super().__init__(name, uuid=uuid, node_type='detector', needs_login=False, needs_sio=False)
         self._detector_factory = detector_factory
+        self.node_version: Optional[str] = os.environ.get('NODE_VERSION')
         self._detector: _DetectorState = _Initializing()
         self._exclusive_model_build: bool = os.environ.get('EXCLUSIVE_MODEL_BUILD', '0').lower() in ('1', 'true')
         self._remaining_init_attempts: int = 2
@@ -409,6 +410,7 @@ class DetectorNode(Node):
             current_model=current_model,
             target_model=target_model_version,
             model_format=self._detector_factory.model_format,
+            node_version=self.node_version,
         )
 
         self.log_status_on_change(status.state, status)

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/learning_loop_node/detector/rest/operation_mode.py

@@ -19,7 +19,7 @@ async def put_operation_mode(request: Request):
 
     Example Usage
 
-        curl -X PUT -d "check_for_updates" http://localhost/operation_mode
+        curl -X PUT -d "idle" http://localhost/operation_mode
         curl -X PUT -d "detecting" http://localhost/operation_mode
     '''
 

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/learning_loop_node.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: learning_loop_node
-Version: 0.20.1
+Version: 0.21.0
 Summary: Python Library for Nodes which connect to the Zauberzeug Learning Loop
 Author-email: Zauberzeug GmbH <info@zauberzeug.com>
 License: MIT
@@ -140,6 +140,27 @@ The endpoint returns None if the upload was successful and an error message othe
 
 For both ways to upload an image, the tag `picked_by_system` is automatically added to the image metadata.
 
+### Node and model information
+
+The detector has a REST endpoint that reports what the node is currently doing and which model it has loaded:
+
+`curl http://localhost/about`
+
+The response contains:
+
+- `operation_mode`: the current operation mode (`startup`, `idle` or `detecting`)
+- `state`: the state of the node (e.g. `idle`, `online`, `detecting`)
+- `model_info`: information about the loaded model (id, organization, project, version, categories, resolution, model size), or `null` if no model is loaded yet
+- `target_model`: the model version the detector is trying to load, or `null` if none is set
+- `version_control`: the model versioning mode (`follow_loop`, `specific_version` or `pause`)
+
+The same information is available via the socketio event `about`:
+`sio.emit('about')`
+
+Use this endpoint if you need the model uuid or its categories. If you only need the version numbers, use `/model_version` (see below).
+
+All REST endpoints of the node are also listed in the automatically generated API documentation at `http://localhost/docs`.
+
 ### Changing the model versioning mode
 
 The detector can be configured to one of the following behaviors:
@@ -167,6 +188,22 @@ The model versioning configuration can also be changed via a socketio event:
 There is also a GET endpoint to fetch the current model versioning configuration:
 `sio.emit('get_model_version')` or `curl http://localhost/model_version`
 
+### Changing the operation mode
+
+The operation mode controls whether the detector may load a new model:
+
+- `startup`: used until the first model is loaded
+- `idle`: the detector checks for and performs model updates
+- `detecting`: the detector runs detections and does not load new models
+
+The mode can be read and changed via a REST endpoint. Example Usage:
+
+- Fetch the current operation mode: `curl http://localhost/operation_mode`
+- Allow model updates: `curl -X PUT -d "idle" http://localhost/operation_mode`
+- Block model updates: `curl -X PUT -d "detecting" http://localhost/operation_mode`
+
+The GET request returns the mode as plain text. The PUT request returns status code 422 if the given value is not one of the three modes above.
+
 ### Changing the outbox mode
 
 If the autoupload is set to `all` or `filtered` (selected) images and the corresponding detections are saved on HDD (the outbox). A background thread will upload the images and detections to the Learning Loop. The outbox is located in the `outbox` folder in the root directory of the node. The outbox can be cleared by deleting the files in the folder.

{learning_loop_node-0.20.1 → learning_loop_node-0.21.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "learning_loop_node"
-version = "0.20.1"
+version = "0.21.0"
 description = "Python Library for Nodes which connect to the Zauberzeug Learning Loop"
 authors = [{ name = "Zauberzeug GmbH", email = "info@zauberzeug.com" }]
 license = { text = "MIT" }