fix: keep Version importable in slim installs

Author: cdeil

roboflow/core/version.py

@@ -27,17 +27,9 @@
     UNIVERSE_URL,
 )
 from roboflow.core.dataset import Dataset
-from roboflow.models.classification import ClassificationModel
-from roboflow.models.instance_segmentation import InstanceSegmentationModel
-from roboflow.models.keypoint_detection import KeypointDetectionModel
-from roboflow.models.object_detection import ObjectDetectionModel
-from roboflow.models.semantic_segmentation import SemanticSegmentationModel
-from roboflow.models.vlm import VLMModel
 from roboflow.util.annotations import amend_data_yaml
 from roboflow.util.general import extract_zip, write_line
-from roboflow.util.model_processor import package_custom_weights_interactive, validate_model_type_for_project
 from roboflow.util.train_recipe import fold_epochs_into_recipe
-from roboflow.util.versions import get_model_format, get_wrong_dependencies_versions
 
 if TYPE_CHECKING:
     import numpy as np
@@ -103,6 +95,8 @@ def __init__(
             if not has_model:
                 self.model = None
             elif self.type == TYPE_OBJECT_DETECTION:
+                from roboflow.models.object_detection import ObjectDetectionModel
+
                 self.model = ObjectDetectionModel(
                     self.__api_key,
                     self.id,
@@ -113,6 +107,8 @@ def __init__(
                     preprocessing=self.preprocessing,
                 )
             elif self.type == TYPE_CLASSICATION:
+                from roboflow.models.classification import ClassificationModel
+
                 self.model = ClassificationModel(
                     self.__api_key,
                     self.id,
@@ -123,6 +119,8 @@ def __init__(
                     preprocessing=self.preprocessing,
                 )
             elif self.type == TYPE_INSTANCE_SEGMENTATION:
+                from roboflow.models.instance_segmentation import InstanceSegmentationModel
+
                 self.model = InstanceSegmentationModel(
                     self.__api_key,
                     self.id,
@@ -131,10 +129,16 @@ def __init__(
                     local=local,
                 )
             elif self.type == TYPE_SEMANTIC_SEGMENTATION:
+                from roboflow.models.semantic_segmentation import SemanticSegmentationModel
+
                 self.model = SemanticSegmentationModel(self.__api_key, self.id)
             elif self.type == TYPE_KEYPOINT_DETECTION:
+                from roboflow.models.keypoint_detection import KeypointDetectionModel
+
                 self.model = KeypointDetectionModel(self.__api_key, self.id, version=version_without_workspace)
             elif self.type == TYPE_TEXT_IMAGE_PAIRS:
+                from roboflow.models.vlm import VLMModel
+
                 self.model = VLMModel(
                     self.__api_key,
                     self.id,
@@ -300,6 +304,8 @@ def create_training(self, speed=None, model_type=None, checkpoint=None, epochs=N
         self.__wait_if_generating()
 
         if model_type:
+            from roboflow.util.versions import get_model_format
+
             train_model_format = get_model_format(model_type)
             if train_model_format not in self.exports:
                 self.export(train_model_format)
@@ -486,6 +492,8 @@ def train(
 
         self.__wait_if_generating()
 
+        from roboflow.util.versions import get_model_format
+
         train_model_format = get_model_format(model_type)
         if train_model_format not in self.exports:
             self.export(train_model_format)
@@ -609,6 +617,8 @@ def live_plot(epochs, mAP, loss, title=""):
 
         if not getattr(self, "_model", None):
             if self.type == TYPE_OBJECT_DETECTION:
+                from roboflow.models.object_detection import ObjectDetectionModel
+
                 self.model = ObjectDetectionModel(
                     self.__api_key,
                     self.id,
@@ -618,6 +628,8 @@ def live_plot(epochs, mAP, loss, title=""):
                     preprocessing=self.preprocessing,
                 )
             elif self.type == TYPE_CLASSICATION:
+                from roboflow.models.classification import ClassificationModel
+
                 self.model = ClassificationModel(
                     self.__api_key,
                     self.id,
@@ -627,15 +639,21 @@ def live_plot(epochs, mAP, loss, title=""):
                     preprocessing=self.preprocessing,
                 )
             elif self.type == TYPE_INSTANCE_SEGMENTATION:
+                from roboflow.models.instance_segmentation import InstanceSegmentationModel
+
                 self.model = InstanceSegmentationModel(
                     self.__api_key,
                     self.id,
                     colors=self.colors,
                     preprocessing=self.preprocessing,
                 )
             elif self.type == TYPE_SEMANTIC_SEGMENTATION:
+                from roboflow.models.semantic_segmentation import SemanticSegmentationModel
+
                 self.model = SemanticSegmentationModel(self.__api_key, self.id)
             elif self.type == TYPE_KEYPOINT_DETECTION:
+                from roboflow.models.keypoint_detection import KeypointDetectionModel
+
                 self.model = KeypointDetectionModel(self.__api_key, self.id, version=self.version)
             else:
                 raise ValueError(f"Unsupported model type: {self.type}")
@@ -653,12 +671,16 @@ def deploy(self, model_type: str, model_path: str, filename: str = "weights/best
             model_path (str): File path to the model weights to be uploaded.
             filename (str, optional): The name of the weights file. Defaults to "weights/best.pt".
         """
+        from roboflow.util.model_processor import package_custom_weights_interactive
+
         bundle = package_custom_weights_interactive(model_type, model_path, filename, build_dir=model_path)
 
         self._validate_against_project_type(bundle.model_type)
         self._upload_zip(bundle.model_type, model_path, bundle.archive_path.name)
 
     def _validate_against_project_type(self, model_type: str) -> None:
+        from roboflow.util.model_processor import validate_model_type_for_project
+
         validate_model_type_for_project(model_type, self.type, self.project)
 
     def _upload_zip(self, model_type: str, model_path: str, model_file_name: str):
@@ -818,6 +840,8 @@ def data_yaml_callback(content: dict) -> dict:
                 content["train"] = location + content["train"].lstrip("..")
                 content["val"] = location + content["val"].lstrip("..")
             try:
+                from roboflow.util.versions import get_wrong_dependencies_versions
+
                 # get_wrong_dependencies_versions raises exception if ultralytics is not installed at all  # noqa: E501 // docs
                 if format == "yolov8" and not get_wrong_dependencies_versions(
                     dependencies_versions=[("ultralytics", "==", "8.0.196")]

tests/test_slim_compat.py

@@ -38,6 +38,11 @@ def test_import_rfapi(self):
 
         self.assertTrue(issubclass(RoboflowError, Exception))
 
+    def test_import_version(self):
+        from roboflow.core.version import Version
+
+        self.assertTrue(callable(Version))
+
     def test_import_cli(self):
         from roboflow.cli import app