Merge branch 'sdubinin/model-1448' of https://github.com/Labelbox/labelbox-python into sdubinin/model-1448

Author: dubininsergey

docs/conf.py

@@ -16,7 +16,7 @@
 project = 'Python SDK reference'
 copyright = '2024, Labelbox'
 author = 'Labelbox'
-release = '3.72.2'
+release = '3.73.0'
 
 # -- General configuration ---------------------------------------------------
 

libs/labelbox/CHANGELOG.md

@@ -1,3 +1,13 @@
+# Changelog
+# Version 3.73.0 (2024-06-20)
+## Added
+* Conversational data row checks (#1678)
+* UI ontology mode support (#1676)
+* Empty data row validation (#1667)
+
+## Fixed
+* Numpy semver locked to < 2.0.0 (#1681)
+
 # Changelog
 # Version 3.72.2 (2024-06-10)
 ## Added

libs/labelbox/pyproject.toml

@@ -1,10 +1,8 @@
 [project]
 name = "labelbox"
-version = "3.72.2"
+version = "3.73.0"
 description = "Labelbox Python API"
-authors = [
-    { name = "Labelbox", email = "engineering@labelbox.com" }
-]
+authors = [{ name = "Labelbox", email = "engineering@labelbox.com" }]
 dependencies = [
     "google-api-core>=1.22.1",
     "pydantic>=1.8",
@@ -16,7 +14,7 @@ dependencies = [
 ]
 readme = "README.md"
 requires-python = ">=3.8"
-classifiers=[
+classifiers = [
     # How mature is this project?
     "Development Status :: 5 - Production/Stable",
     # Indicate who your project is intended for
@@ -27,7 +25,7 @@ classifiers=[
     "Intended Audience :: Education",
     # Pick your license as you wish
     "License :: OSI Approved :: Apache Software License",
-     # Specify the Python versions you support here.
+    # Specify the Python versions you support here.
     "Programming Language :: Python :: 3",
     "Programming Language :: Python :: 3.8",
     "Programming Language :: Python :: 3.9",
@@ -47,7 +45,8 @@ Changelog = "https://github.com/Labelbox/labelbox-python/blob/develop/libs/label
 [project.optional-dependencies]
 data = [
     "shapely>=2.0.3",
-    "numpy>=1.24.4",
+    # numpy v2 breaks package since it only supports python >3.9
+    "numpy>=1.24.4, <2.0.0",
     "pillow>=10.2.0",
     "typeguard>=4.1.5",
     "imagesize>=1.4.1",
@@ -85,12 +84,12 @@ unit = "pytest tests/unit"
 # LABELBOX_TEST_ENVIRON="ephemeral" \
 # SERVICE_API_KEY=${SERVICE_API_KEY} \
 # LABELBOX_TEST_BASE_URL="http://host.docker.internal:8080" \
-integration = { cmd = "pytest tests/integration"  }
+integration = { cmd = "pytest tests/integration" }
 data = { cmd = "pytest tests/data" }
 yapf-lint = "yapf tests src -i --verbose --recursive --parallel --style \"google\""
 mypy-lint = "mypy src --pretty --show-error-codes --non-interactive --install-types"
 lint = { chain = ["yapf-lint", "mypy-lint"] }
-test = { chain = ["lint", "unit", "integration" ] }
+test = { chain = ["lint", "unit", "integration"] }
 
 [tool.hatch.metadata]
 allow-direct-references = true

libs/labelbox/src/labelbox/__init__.py

@@ -1,6 +1,6 @@
 name = "labelbox"
 
-__version__ = "3.72.2"
+__version__ = "3.73.0"
 
 from labelbox.client import Client
 from labelbox.schema.project import Project

libs/labelbox/src/labelbox/client.py

@@ -653,8 +653,7 @@ def delete_model_config(self, id: str) -> bool:
         params = {"id": id}
         result = self.execute(query, params)
         if not result:
-            raise labelbox.exceptions.ResourceNotFoundError(
-                Entity.ModelConfig, params)
+            raise labelbox.exceptions.ResourceNotFoundError(Entity.ModelConfig, params)
         return result['deleteModelConfig']['success']
 
     def create_dataset(self,
@@ -742,17 +741,85 @@ def create_project(self, **kwargs) -> Project:
         Raises:
             InvalidAttributeError: If the Project type does not contain
                 any of the attribute names given in kwargs.
-
-        NOTE: the following attributes are used only in chat model evaluation projects:
-            dataset_name_or_id, append_to_existing_dataset, data_row_count, editor_task_type
-            They are not used for general projects and not supported in this method
         """
-        #  The following arguments are not supported for general projects, only for chat model evaluation projects
-        kwargs.pop("dataset_name_or_id", None)
-        kwargs.pop("append_to_existing_dataset", None)
-        kwargs.pop("data_row_count", None)
-        kwargs.pop("editor_task_type", None)
-        return self._create_project(**kwargs)
+
+        auto_audit_percentage = kwargs.get("auto_audit_percentage")
+        auto_audit_number_of_labels = kwargs.get("auto_audit_number_of_labels")
+        if auto_audit_percentage is not None or auto_audit_number_of_labels is not None:
+            raise ValueError(
+                "quality_mode must be set instead of auto_audit_percentage or auto_audit_number_of_labels."
+            )
+
+        name = kwargs.get("name")
+        if name is None or not name.strip():
+            raise ValueError("project name must be a valid string.")
+
+        queue_mode = kwargs.get("queue_mode")
+        if queue_mode is QueueMode.Dataset:
+            raise ValueError(
+                "Dataset queue mode is deprecated. Please prefer Batch queue mode."
+            )
+        elif queue_mode is QueueMode.Batch:
+            logger.warning(
+                "Passing a queue mode of batch is redundant and will soon no longer be supported."
+            )
+
+        media_type = kwargs.get("media_type")
+        if media_type and MediaType.is_supported(media_type):
+            media_type_value = media_type.value
+        elif media_type:
+            raise TypeError(f"{media_type} is not a valid media type. Use"
+                            f" any of {MediaType.get_supported_members()}"
+                            " from MediaType. Example: MediaType.Image.")
+        else:
+            logger.warning(
+                "Creating a project without specifying media_type"
+                " through this method will soon no longer be supported.")
+            media_type_value = None
+
+        ontology_kind = kwargs.pop("ontology_kind", None)
+        if ontology_kind and OntologyKind.is_supported(ontology_kind):
+            editor_task_type_value = EditorTaskTypeMapper.to_editor_task_type(
+                ontology_kind, media_type).value
+        elif ontology_kind:
+            raise OntologyKind.get_ontology_kind_validation_error(ontology_kind)
+        else:
+            editor_task_type_value = None
+
+        quality_mode = kwargs.get("quality_mode")
+        if not quality_mode:
+            logger.info("Defaulting quality mode to Benchmark.")
+
+        data = kwargs
+        data.pop("quality_mode", None)
+        if quality_mode is None or quality_mode is QualityMode.Benchmark:
+            data[
+                "auto_audit_number_of_labels"] = BENCHMARK_AUTO_AUDIT_NUMBER_OF_LABELS
+            data["auto_audit_percentage"] = BENCHMARK_AUTO_AUDIT_PERCENTAGE
+        elif quality_mode is QualityMode.Consensus:
+            data[
+                "auto_audit_number_of_labels"] = CONSENSUS_AUTO_AUDIT_NUMBER_OF_LABELS
+            data["auto_audit_percentage"] = CONSENSUS_AUTO_AUDIT_PERCENTAGE
+        else:
+            raise ValueError(f"{quality_mode} is not a valid quality mode.")
+
+        params = {**data}
+        if media_type_value:
+            params["media_type"] = media_type_value
+        if editor_task_type_value:
+            params["editor_task_type"] = editor_task_type_value
+
+        extra_params = {
+            Field.String("dataset_name_or_id"):
+                params.pop("dataset_name_or_id", None),
+            Field.Boolean("append_to_existing_dataset"):
+                params.pop("append_to_existing_dataset", None),
+            Field.Int("data_row_count"):
+                params.pop("data_row_count", None),
+        }
+        extra_params = {k: v for k, v in extra_params.items() if v is not None}
+
+        return self._create(Entity.Project, params, extra_params)
 
     @overload
     def create_model_evaluation_project(self,
@@ -815,99 +882,13 @@ def create_model_evaluation_project(self,
             dataset_name_or_id = dataset_name
 
         kwargs["media_type"] = MediaType.Conversational
+        kwargs["ontology_kind"] = OntologyKind.ModelEvaluation
         kwargs["dataset_name_or_id"] = dataset_name_or_id
         kwargs["append_to_existing_dataset"] = append_to_existing_dataset
         kwargs["data_row_count"] = data_row_count
-        kwargs["editor_task_type"] = EditorTaskType.ModelChatEvaluation.value
-
-        return self._create_project(**kwargs)
-
-    def create_offline_model_evaluation_project(self, **kwargs) -> Project:
-        """
-        Creates a project for offline model evaluation.
-        Args:
-            **kwargs: Additional parameters to pass see the create_project method
-        Returns:
-            Project: The created project
-        """
-        kwargs[
-            "media_type"] = MediaType.Conversational  # Only Conversational is supported
-        kwargs[
-            "editor_task_type"] = EditorTaskType.OfflineModelChatEvaluation.value  # Special editor task type for offline model evaluation
-
-        # The following arguments are not supported for offline model evaluation
-        kwargs.pop("dataset_name_or_id", None)
-        kwargs.pop("append_to_existing_dataset", None)
-        kwargs.pop("data_row_count", None)
 
         return self.create_project(**kwargs)
 
-    def _create_project(self, **kwargs) -> Project:
-        auto_audit_percentage = kwargs.get("auto_audit_percentage")
-        auto_audit_number_of_labels = kwargs.get("auto_audit_number_of_labels")
-        if auto_audit_percentage is not None or auto_audit_number_of_labels is not None:
-            raise ValueError(
-                "quality_mode must be set instead of auto_audit_percentage or auto_audit_number_of_labels."
-            )
-
-        name = kwargs.get("name")
-        if name is None or not name.strip():
-            raise ValueError("project name must be a valid string.")
-
-        queue_mode = kwargs.get("queue_mode")
-        if queue_mode is QueueMode.Dataset:
-            raise ValueError(
-                "Dataset queue mode is deprecated. Please prefer Batch queue mode."
-            )
-        elif queue_mode is QueueMode.Batch:
-            logger.warning(
-                "Passing a queue mode of batch is redundant and will soon no longer be supported."
-            )
-
-        media_type = kwargs.get("media_type")
-        if media_type and MediaType.is_supported(media_type):
-            media_type_value = media_type.value
-        elif media_type:
-            raise TypeError(f"{media_type} is not a valid media type. Use"
-                            f" any of {MediaType.get_supported_members()}"
-                            " from MediaType. Example: MediaType.Image.")
-        else:
-            logger.warning(
-                "Creating a project without specifying media_type"
-                " through this method will soon no longer be supported.")
-            media_type_value = None
-
-        quality_mode = kwargs.get("quality_mode")
-        if not quality_mode:
-            logger.info("Defaulting quality mode to Benchmark.")
-
-        data = kwargs
-        data.pop("quality_mode", None)
-        if quality_mode is None or quality_mode is QualityMode.Benchmark:
-            data[
-                "auto_audit_number_of_labels"] = BENCHMARK_AUTO_AUDIT_NUMBER_OF_LABELS
-            data["auto_audit_percentage"] = BENCHMARK_AUTO_AUDIT_PERCENTAGE
-        elif quality_mode is QualityMode.Consensus:
-            data[
-                "auto_audit_number_of_labels"] = CONSENSUS_AUTO_AUDIT_NUMBER_OF_LABELS
-            data["auto_audit_percentage"] = CONSENSUS_AUTO_AUDIT_PERCENTAGE
-        else:
-            raise ValueError(f"{quality_mode} is not a valid quality mode.")
-
-        params = {**data}
-        if media_type_value:
-            params["media_type"] = media_type_value
-
-        extra_params = {
-            Field.String("dataset_name_or_id"):
-                params.pop("dataset_name_or_id", None),
-            Field.Boolean("append_to_existing_dataset"):
-                params.pop("append_to_existing_dataset", None),
-        }
-        extra_params = {k: v for k, v in extra_params.items() if v is not None}
-
-        return self._create(Entity.Project, params, extra_params)
-
     def get_roles(self) -> List[Role]:
         """
         Returns:

libs/labelbox/src/labelbox/schema/ontology_kind.py

@@ -26,28 +26,12 @@ def get_ontology_kind_validation_error(cls, ontology_kind):
 class EditorTaskType(Enum):
     ModelChatEvaluation = "MODEL_CHAT_EVALUATION"
     ResponseCreation = "RESPONSE_CREATION"
-    OfflineModelChatEvaluation = "OFFLINE_MODEL_CHAT_EVALUATION"
     Missing = None
 
     @classmethod
     def is_supported(cls, value):
         return isinstance(value, cls)
 
-    @classmethod
-    def _missing_(cls, name) -> 'EditorTaskType':
-        """Handle missing null new task types
-            Handle upper case names for compatibility with
-            the GraphQL"""
-
-        if name is None:
-            return cls.Missing
-
-        for name, member in cls.__members__.items():
-            if name == name.upper():
-                return member
-
-        return cls.Missing
-
 
 class EditorTaskTypeMapper:
 

libs/labelbox/src/labelbox/schema/project.py

@@ -121,7 +121,6 @@ class Project(DbObject, Updateable, Deletable):
     # Bind data_type and allowedMediaTYpe using the GraphQL type MediaType
     media_type = Field.Enum(MediaType, "media_type", "allowedMediaType")
     editor_task_type = Field.Enum(EditorTaskType, "editor_task_type")
-    data_row_count = Field.Int("data_row_count")
 
     # Relationships
     created_by = Relationship.ToOne("User", False, "created_by")