orca-sdk 0.0.101__tar.gz → 0.0.103__tar.gz

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/PKG-INFO

@@ -1,23 +1,25 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: orca_sdk
-Version: 0.0.101
+Version: 0.0.103
 Summary: SDK for interacting with Orca Services
 License: Apache-2.0
 Author: Orca DB Inc.
 Author-email: dev-rel@orcadb.ai
-Requires-Python: >=3.11,<4.0
+Requires-Python: >=3.11,<3.14
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Requires-Dist: datasets (>=3.1.0,<4.0.0)
-Requires-Dist: gradio (==5.13.0)
-Requires-Dist: httpx (>=0.20.0,<0.29.0)
+Requires-Dist: gradio (>=5.44.1,<6.0.0)
+Requires-Dist: httpx (>=0.28.1,<0.29.0)
+Requires-Dist: numpy (>=2.1.0,<3.0.0)
 Requires-Dist: pandas (>=2.2.3,<3.0.0)
 Requires-Dist: pyarrow (>=18.0.0,<19.0.0)
 Requires-Dist: python-dotenv (>=1.1.0,<2.0.0)
 Requires-Dist: scikit-learn (>=1.6.1,<2.0.0)
-Requires-Dist: torch (>=2.5.1,<3.0.0)
+Requires-Dist: torch (>=2.8.0,<3.0.0)
 Description-Content-Type: text/markdown
 
 <!--

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/_shared/metrics.py

@@ -238,7 +238,13 @@ def calculate_classification_metrics(
 
     accuracy = sklearn.metrics.accuracy_score(references, predictions)
     f1 = sklearn.metrics.f1_score(references, predictions, average=average)
-    loss = sklearn.metrics.log_loss(references, probabilities)
+    # Ensure sklearn sees the full class set corresponding to probability columns
+    # to avoid errors when y_true does not contain all classes.
+    loss = sklearn.metrics.log_loss(
+        references,
+        probabilities,
+        labels=list(range(probabilities.shape[1])),
+    )
 
     if num_classes_references == num_classes_predictions:
         # special case for binary classification: https://github.com/scikit-learn/scikit-learn/issues/20186

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/_shared/metrics_test.py

@@ -138,6 +138,25 @@ def test_roc_curve():
     assert np.all(np.diff(roc_curve["thresholds"]) >= 0)
 
 
+def test_log_loss_handles_missing_classes_in_y_true():
+    # y_true contains only a subset of classes, but predictions include an extra class column
+    y_true = np.array([0, 1, 0, 1])
+    y_score = np.array(
+        [
+            [0.7, 0.2, 0.1],
+            [0.1, 0.8, 0.1],
+            [0.6, 0.3, 0.1],
+            [0.2, 0.7, 0.1],
+        ]
+    )
+
+    metrics = calculate_classification_metrics(y_true, y_score)
+    expected_loss = sklearn.metrics.log_loss(y_true, y_score, labels=[0, 1, 2])
+
+    assert np.isfinite(metrics.loss)
+    assert np.allclose(metrics.loss, expected_loss)
+
+
 def test_precision_recall_curve_max_length():
     y_true = np.array([0, 1, 1, 0, 1])
     y_score = np.array([0.1, 0.9, 0.8, 0.6, 0.2])

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/_utils/auth.py

@@ -53,7 +53,7 @@ def _delete_org(org_id: str) -> None:
 def _authenticate_local_api(org_id: str = _DEFAULT_ORG_ID, api_key_name: str = "local") -> None:
     """Connect to the local API at http://localhost:1584/ and authenticate with a new API key"""
     _delete_api_key(org_id, api_key_name, if_not_exists="ignore")
-    OrcaCredentials.set_base_url("http://localhost:1584")
+    OrcaCredentials.set_api_url("http://localhost:1584")
     OrcaCredentials.set_api_key(_create_api_key(org_id, api_key_name))
     logging.info(f"Authenticated against local API at 'http://localhost:1584' with '{api_key_name}' API key")
 

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/_utils/prediction_result_ui.py

@@ -7,13 +7,13 @@ from typing import TYPE_CHECKING
 
 import gradio as gr
 
-from ..memoryset import LabeledMemoryLookup, ScoredMemoryLookup, LabeledMemoryset
+from ..memoryset import LabeledMemoryLookup, LabeledMemoryset, ScoredMemoryLookup
 
 if TYPE_CHECKING:
-    from ..telemetry import _Prediction
+    from ..telemetry import PredictionBase
 
 
-def inspect_prediction_result(prediction_result: _Prediction):
+def inspect_prediction_result(prediction_result: PredictionBase):
 
     def update_label(val: str, memory: LabeledMemoryLookup, progress=gr.Progress(track_tqdm=True)):
         progress(0)

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/classification_model_test.py

@@ -3,7 +3,7 @@ from uuid import uuid4
 
 import numpy as np
 import pytest
-from datasets.arrow_dataset import Dataset
+from datasets import Dataset
 
 from .classification_model import ClassificationMetrics, ClassificationModel
 from .conftest import skip_in_ci

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/client.py

@@ -36,6 +36,10 @@ class AddMemorySuggestion(TypedDict):
     label_name: str
 
 
+class AliveResponse(TypedDict):
+    ok: bool
+
+
 class ApiKeyMetadata(TypedDict):
     id: str
     org_id: str
@@ -201,6 +205,14 @@ class GetMemoriesRequest(TypedDict):
     memory_ids: list[str]
 
 
+class HealthyResponse(TypedDict):
+    ok: bool
+    checks: dict[str, bool]
+    durations: dict[str, int]
+    draining: bool
+    config: dict[str, str | float | int | bool | None]
+
+
 class InternalServerErrorResponse(TypedDict):
     status_code: NotRequired[int]
     message: str
@@ -282,6 +294,7 @@ class MemorysetClassPatternsMetrics(TypedDict):
     variance_spread: float
     mean_uniformity: float
     variance_uniformity: float
+    updated_at: str
 
 
 class MemorysetClusterAnalysisConfig(TypedDict):
@@ -299,6 +312,7 @@ class MemorysetClusterMetrics(TypedDict):
     cluster_metrics: list[ClusterMetrics]
     num_outliers: int
     num_clusters: int
+    updated_at: str
 
 
 class MemorysetConceptAnalysisConfig(TypedDict):
@@ -320,6 +334,8 @@ class MemorysetDuplicateAnalysisConfig(TypedDict):
 
 class MemorysetDuplicateMetrics(TypedDict):
     num_duplicates: int
+    num_potential_duplicates: int
+    updated_at: str
 
 
 class MemorysetLabelAnalysisConfig(TypedDict):
@@ -332,6 +348,8 @@ class MemorysetLabelMetrics(TypedDict):
     mean_neighbor_label_confidence: float
     mean_neighbor_label_entropy: float
     mean_neighbor_predicted_label_ambiguity: float
+    num_potential_mislabels: int
+    updated_at: str
 
 
 class MemorysetNeighborAnalysisConfig(TypedDict):
@@ -341,6 +359,7 @@ class MemorysetNeighborAnalysisConfig(TypedDict):
 
 class MemorysetNeighborMetrics(TypedDict):
     lookup_score_metrics: dict[str, LookupScoreMetrics]
+    updated_at: str
 
 
 class MemorysetProjectionAnalysisConfig(TypedDict):
@@ -349,7 +368,7 @@ class MemorysetProjectionAnalysisConfig(TypedDict):
 
 
 class MemorysetProjectionMetrics(TypedDict):
-    pass
+    updated_at: str
 
 
 class MemorysetUpdate(TypedDict):
@@ -357,6 +376,7 @@ class MemorysetUpdate(TypedDict):
     description: NotRequired[str | None]
     name: NotRequired[str]
     notes: NotRequired[str | None]
+    hidden: NotRequired[bool]
 
 
 class NotFoundErrorResponse(TypedDict):
@@ -475,6 +495,11 @@ class ROCCurve(TypedDict):
     true_positive_rates: list[float]
 
 
+class ReadyResponse(TypedDict):
+    ok: bool
+    draining: bool
+
+
 class RegressionEvaluationRequest(TypedDict):
     datasource_name_or_id: str
     memoryset_override_name_or_id: NotRequired[str | None]
@@ -522,6 +547,7 @@ class RegressionPredictionRequest(TypedDict):
     save_telemetry_synchronously: NotRequired[bool]
     prompt: NotRequired[str | None]
     use_lookup_cache: NotRequired[bool]
+    consistency_level: NotRequired[Literal["Bounded", "Session", "Strong", "Eventual"] | None]
 
 
 class ScorePredictionMemoryLookup(TypedDict):
@@ -636,7 +662,7 @@ class SubConceptMetrics(TypedDict):
     memory_count: int
 
 
-TaskStatus = Literal["INITIALIZED", "DISPATCHED", "PROCESSING", "COMPLETED", "FAILED", "ABORTING", "ABORTED"]
+TaskStatus = Literal["INITIALIZED", "DISPATCHED", "WAITING", "PROCESSING", "COMPLETED", "FAILED", "ABORTING", "ABORTED"]
 
 
 class TaskStatusInfo(TypedDict):
@@ -694,6 +720,7 @@ class DeleteAuthApiKeyByNameOrIdParams(TypedDict):
 
 class GetMemorysetParams(TypedDict):
     type: NotRequired[MemoryType | None]
+    show_hidden: NotRequired[bool | None]
 
 
 class PostMemorysetByNameOrIdCloneParams(TypedDict):
@@ -842,7 +869,7 @@ class GetDatasourceByNameOrIdDownloadParams(TypedDict):
     name_or_id: str
     file_type: NotRequired[Literal["hf_dataset", "json", "csv"]]
     """
-    File type to download: 
+    File type to download:
     * `hf_dataset`: Zipped HuggingFace dataset (default)
     * `json`: Row-oriented JSON array
     * `csv`: CSV file
@@ -1087,6 +1114,7 @@ class ClassificationPredictionRequest(TypedDict):
     save_telemetry_synchronously: NotRequired[bool]
     prompt: NotRequired[str | None]
     use_lookup_cache: NotRequired[bool]
+    consistency_level: NotRequired[Literal["Bounded", "Session", "Strong", "Eventual"] | None]
 
 
 class CloneMemorysetRequest(TypedDict):
@@ -1145,6 +1173,7 @@ class CreateMemorysetRequest(TypedDict):
     index_type: NotRequired[Literal["FLAT", "IVF_FLAT", "IVF_SQ8", "IVF_PQ", "HNSW", "DISKANN"]]
     index_params: NotRequired[dict[str, int | float | str]]
     prompt: NotRequired[str]
+    hidden: NotRequired[bool]
 
 
 class CreateRegressionModelRequest(TypedDict):
@@ -1406,6 +1435,7 @@ class MemorysetAnalysisRequest(TypedDict):
 class MemorysetConceptMetrics(TypedDict):
     concepts: list[ConceptMetrics]
     num_outliers: int
+    updated_at: str
 
 
 class MemorysetMetrics(TypedDict):
@@ -1510,6 +1540,7 @@ class MemorysetMetadata(TypedDict):
     database_uri: str | None
     document_prompt_override: str | None
     query_prompt_override: str | None
+    hidden: bool
 
 
 class PaginatedTask(TypedDict):
@@ -1558,12 +1589,140 @@ class OrcaClient(Client):
             raise ValueError(f"Missing path params: {', '.join(placeholders - path_params.keys())}")
         return path_params, query_params
 
+    @overload
+    def GET(
+        self,
+        path: Literal["/check/alive"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> AliveResponse:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/check/ready"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> ReadyResponse:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/gpu/check/healthy"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> HealthyResponse:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/check/healthy"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> HealthyResponse:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/gpu/config"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> dict[str, str | float | int | bool | None]:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/config"],
+        *,
+        params: None = None,
+        parse_as: Literal["json"] = "json",
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> dict[str, str | float | int | bool | None]:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/gpu/"],
+        *,
+        params: None = None,
+        parse_as: Literal["text"],
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> str:
+        pass
+
     @overload
     def GET(
         self,
         path: Literal["/"],
         *,
         params: None = None,
+        parse_as: Literal["text"],
+        headers: HeaderTypes | None = None,
+        cookies: CookieTypes | None = None,
+        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
+        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
+        extensions: RequestExtensions | None = None,
+    ) -> str:
+        pass
+
+    @overload
+    def GET(
+        self,
+        path: Literal["/auth/root"],
+        *,
+        params: None = None,
         parse_as: Literal["json"] = "json",
         headers: HeaderTypes | None = None,
         cookies: CookieTypes | None = None,
@@ -1571,7 +1730,8 @@ class OrcaClient(Client):
         follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
         timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
         extensions: RequestExtensions | None = None,
-    ) -> Any:
+    ) -> bool:
+        """Return true only when called with a valid root API key; otherwise 401 Unauthenticated."""
         pass
 
     @overload
@@ -2292,22 +2452,6 @@ class OrcaClient(Client):
         """Get the status of a bootstrap classification model task"""
         pass
 
-    @overload
-    def GET(
-        self,
-        path: Literal["/gpu/"],
-        *,
-        params: None = None,
-        parse_as: Literal["json"] = "json",
-        headers: HeaderTypes | None = None,
-        cookies: CookieTypes | None = None,
-        auth: AuthTypes | UseClientDefault = USE_CLIENT_DEFAULT,
-        follow_redirects: bool | UseClientDefault = USE_CLIENT_DEFAULT,
-        timeout: TimeoutTypes | UseClientDefault = USE_CLIENT_DEFAULT,
-        extensions: RequestExtensions | None = None,
-    ) -> Any:
-        pass
-
     def GET(
         self,
         path: str,
@@ -3558,6 +3702,6 @@ logging.getLogger("httpcore").setLevel(logging.ERROR)
 orca_api = OrcaClient(
     event_hooks={"request": [_instrument_request], "response": [_raise_error_for_response]},
     follow_redirects=True,
-    timeout=Timeout(connect=3, read=10, write=2, pool=1),
+    timeout=Timeout(connect=3, read=20, write=10, pool=5),
 )
 """Typed client for the Orca API"""

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/conftest.py

@@ -24,7 +24,7 @@ os.environ["ORCA_SAVE_TELEMETRY_SYNCHRONOUSLY"] = "true"
 
 def skip_in_prod(reason: str):
     """Custom decorator to skip tests when running against production API"""
-    PROD_API_URLs = ["https://api.orcadb.ai", "https://api.dev.orcadb.ai"]
+    PROD_API_URLs = ["https://api.orcadb.ai", "https://api.staging.orcadb.ai"]
     return pytest.mark.skipif(
         os.environ["ORCA_API_URL"] in PROD_API_URLs,
         reason=reason,
@@ -45,7 +45,7 @@ def _create_org_id():
 
 
 @pytest.fixture()
-def base_url_reset():
+def api_url_reset():
     original_base_url = orca_api.base_url
     yield
     orca_api.base_url = original_base_url

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/credentials.py

@@ -35,11 +35,33 @@ class OrcaCredentials:
     """
 
     @staticmethod
-    def get_api_url() -> str:
+    def is_authenticated() -> bool:
         """
-        Get the Orca API base URL that is currently being used
+        Check if you are authenticated to interact with the Orca API
+
+        Returns:
+            True if you are authenticated, False otherwise
         """
-        return str(orca_api.base_url)
+        try:
+            return orca_api.GET("/auth")
+        except ValueError as e:
+            if "Invalid API key" in str(e):
+                return False
+            raise e
+
+    @staticmethod
+    def is_healthy() -> bool:
+        """
+        Check whether the API is healthy
+
+        Returns:
+            True if the API is healthy, False otherwise
+        """
+        try:
+            orca_api.GET("/check/healthy")
+        except Exception:
+            return False
+        return True
 
     @staticmethod
     def list_api_keys() -> list[ApiKeyInfo]:
@@ -58,21 +80,6 @@ class OrcaCredentials:
             for api_key in orca_api.GET("/auth/api_key")
         ]
 
-    @staticmethod
-    def is_authenticated() -> bool:
-        """
-        Check if you are authenticated to interact with the Orca API
-
-        Returns:
-            True if you are authenticated, False otherwise
-        """
-        try:
-            return orca_api.GET("/auth")
-        except ValueError as e:
-            if "Invalid API key" in str(e):
-                return False
-            raise e
-
     @staticmethod
     def create_api_key(name: str, scopes: set[Scope] = {"ADMINISTER"}) -> str:
         """
@@ -104,20 +111,6 @@ class OrcaCredentials:
         """
         orca_api.DELETE("/auth/api_key/{name_or_id}", params={"name_or_id": name})
 
-    @staticmethod
-    def set_headers(headers: dict[str, str]):
-        """
-        Add or override default HTTP headers for all Orca API requests.
-
-        Args:
-            **kwargs: Header names with their string values
-
-        Notes:
-            New keys are merged into the existing headers, this will overwrite headers with the
-            same name, but leave other headers untouched.
-        """
-        orca_api.headers.update(Headers(headers))
-
     @staticmethod
     def set_api_key(api_key: str, check_validity: bool = True):
         """
@@ -133,17 +126,24 @@ class OrcaCredentials:
         Raises:
             ValueError: if the API key is invalid and `check_validity` is True
         """
-        OrcaCredentials.set_headers({"Api-Key": api_key})
+        OrcaCredentials.set_api_headers({"Api-Key": api_key})
         if check_validity:
             orca_api.GET("/auth")
 
     @staticmethod
-    def set_base_url(base_url: str, check_validity: bool = True):
+    def get_api_url() -> str:
+        """
+        Get the base URL of the Orca API that is currently being used
+        """
+        return str(orca_api.base_url)
+
+    @staticmethod
+    def set_api_url(url: str, check_validity: bool = True):
         """
         Set the base URL for the Orca API
 
         Args:
-            base_url: The base URL to set
+            url: The base URL to set
             check_validity: Whether to check if there is an API running at the given base URL
 
         Raises:
@@ -152,27 +152,26 @@ class OrcaCredentials:
         # check if the base url is reachable before setting it
         if check_validity:
             try:
-                httpx.get(base_url, timeout=1)
+                httpx.get(url, timeout=1)
             except ConnectError as e:
-                raise ValueError(f"No API found at {base_url}") from e
+                raise ValueError(f"No API found at {url}") from e
 
-        orca_api.base_url = base_url
+        orca_api.base_url = url
 
         # check if the api passes the health check
         if check_validity:
-            orca_api.GET("/")
+            OrcaCredentials.is_healthy()
 
     @staticmethod
-    def is_healthy() -> bool:
+    def set_api_headers(headers: dict[str, str]):
         """
-        Check whether the API is healthy
+        Add or override default HTTP headers for all Orca API requests.
 
-        Returns:
-            True if the API is healthy, False otherwise
+        Params:
+            headers: Mapping of header names to their string values
+
+        Notes:
+            New keys are merged into the existing headers, this will overwrite headers with the
+            same name, but leave other headers untouched.
         """
-        try:
-            orca_api.GET("/")
-            orca_api.GET("/gpu/")
-        except Exception:
-            return False
-        return True
+        orca_api.headers.update(Headers(headers))

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/credentials_test.py

@@ -38,20 +38,20 @@ def test_set_invalid_api_key(api_key):
     assert not OrcaCredentials.is_authenticated()
 
 
-def test_set_base_url(base_url_reset):
-    OrcaCredentials.set_base_url("http://api.orcadb.ai")
+def test_set_api_url(api_url_reset):
+    OrcaCredentials.set_api_url("http://api.orcadb.ai")
     assert str(orca_api.base_url) == "http://api.orcadb.ai"
 
 
 def test_set_invalid_base_url():
     with pytest.raises(ValueError, match="No API found at http://localhost:1582"):
-        OrcaCredentials.set_base_url("http://localhost:1582")
+        OrcaCredentials.set_api_url("http://localhost:1582")
 
 
 def test_is_healthy():
     assert OrcaCredentials.is_healthy()
 
 
-def test_is_healthy_false(base_url_reset):
-    OrcaCredentials.set_base_url("http://localhost:1582", check_validity=False)
+def test_is_healthy_false(api_url_reset):
+    OrcaCredentials.set_api_url("http://localhost:1582", check_validity=False)
     assert not OrcaCredentials.is_healthy()

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/embedding_model.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from abc import abstractmethod
+from abc import ABC, abstractmethod
 from datetime import datetime
 from typing import TYPE_CHECKING, Literal, Sequence, cast, get_args, overload
 
@@ -23,7 +23,7 @@ if TYPE_CHECKING:
     from .memoryset import LabeledMemoryset
 
 
-class _EmbeddingModel:
+class EmbeddingModelBase(ABC):
     embedding_dim: int
     max_seq_length: int
     uses_context: bool
@@ -39,7 +39,7 @@ class _EmbeddingModel:
 
     @classmethod
     @abstractmethod
-    def all(cls) -> Sequence[_EmbeddingModel]:
+    def all(cls) -> Sequence[EmbeddingModelBase]:
         pass
 
     def _get_instruction_error_message(self) -> str:
@@ -291,7 +291,7 @@ class _ModelDescriptor:
         return self.model
 
 
-class PretrainedEmbeddingModel(_EmbeddingModel):
+class PretrainedEmbeddingModel(EmbeddingModelBase):
     """
     A pretrained embedding model
 
@@ -481,7 +481,7 @@ class PretrainedEmbeddingModel(_EmbeddingModel):
             label_column: Column name of the label
             value_column: Column name of the value
             training_method: Training method to use
-            training_args: Optional override for Hugging Face [`TrainingArguments`](transformers.TrainingArguments).
+            training_args: Optional override for Hugging Face [`TrainingArguments`][transformers.TrainingArguments].
                 If not provided, reasonable training arguments will be used for the specified training method
             if_exists: What to do if a finetuned embedding model with the same name already exists, defaults to
                 `"error"`. Other option is `"open"` to open the existing finetuned embedding model.
@@ -539,7 +539,7 @@ class PretrainedEmbeddingModel(_EmbeddingModel):
         return job if background else job.result()
 
 
-class FinetunedEmbeddingModel(_EmbeddingModel):
+class FinetunedEmbeddingModel(EmbeddingModelBase):
     """
     A finetuned embedding model in the OrcaCloud
 

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/job.py

@@ -7,7 +7,7 @@ from typing import Callable, Generic, TypedDict, TypeVar, cast
 
 from tqdm.auto import tqdm
 
-from .client import TaskStatus, orca_api
+from .client import orca_api
 
 
 class JobConfig(TypedDict):
@@ -26,6 +26,9 @@ class Status(Enum):
     DISPATCHED = "DISPATCHED"
     """The job has been queued and is waiting to be processed"""
 
+    WAITING = "WAITING"
+    """The job is waiting for dependencies to complete"""
+
     PROCESSING = "PROCESSING"
     """The job is being processed"""
 

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/memoryset.py

@@ -21,7 +21,9 @@ from .client import (
     FilterItem,
 )
 from .client import LabeledMemory as LabeledMemoryResponse
-from .client import LabeledMemoryInsert
+from .client import (
+    LabeledMemoryInsert,
+)
 from .client import LabeledMemoryLookup as LabeledMemoryLookupResponse
 from .client import (
     LabeledMemoryUpdate,
@@ -35,7 +37,9 @@ from .client import (
     MemoryType,
 )
 from .client import ScoredMemory as ScoredMemoryResponse
-from .client import ScoredMemoryInsert
+from .client import (
+    ScoredMemoryInsert,
+)
 from .client import ScoredMemoryLookup as ScoredMemoryLookupResponse
 from .client import (
     ScoredMemoryUpdate,
@@ -47,9 +51,9 @@ from .client import (
 )
 from .datasource import Datasource
 from .embedding_model import (
+    EmbeddingModelBase,
     FinetunedEmbeddingModel,
     PretrainedEmbeddingModel,
-    _EmbeddingModel,
 )
 from .job import Job, Status
 
@@ -241,7 +245,7 @@ def _parse_memory_update(update: dict[str, Any], type: MemoryType) -> LabeledMem
             return cast(ScoredMemoryUpdate, payload)
 
 
-class _Memory(ABC):
+class MemoryBase(ABC):
     value: str
     embedding: list[float]
     source_id: str | None
@@ -303,8 +307,6 @@ class _Memory(ABC):
 
         Params:
             value: New value of the memory
-            label: New label of the memory
-            score: New score of the memory
             source_id: New source ID of the memory
             **metadata: New values for metadata properties
 
@@ -345,7 +347,7 @@ class _Memory(ABC):
         }
 
 
-class LabeledMemory(_Memory):
+class LabeledMemory(MemoryBase):
     """
     A row of the [`LabeledMemoryset`][orca_sdk.LabeledMemoryset]
 
@@ -486,7 +488,7 @@ class LabeledMemoryLookup(LabeledMemory):
         )
 
 
-class ScoredMemory(_Memory):
+class ScoredMemory(MemoryBase):
     """
     A row of the [`ScoredMemoryset`][orca_sdk.ScoredMemoryset]
 
@@ -617,11 +619,11 @@ class ScoredMemoryLookup(ScoredMemory):
         )
 
 
-MemoryT = TypeVar("MemoryT", bound=_Memory)
-MemoryLookupT = TypeVar("MemoryLookupT", bound=_Memory)
+MemoryT = TypeVar("MemoryT", bound=MemoryBase)
+MemoryLookupT = TypeVar("MemoryLookupT", bound=MemoryBase)
 
 
-class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
+class MemorysetBase(Generic[MemoryT, MemoryLookupT], ABC):
     """
     A Handle to a collection of memories with labels in the OrcaCloud
 
@@ -644,9 +646,10 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
     created_at: datetime
     updated_at: datetime
     insertion_status: Status
-    embedding_model: _EmbeddingModel
+    embedding_model: EmbeddingModelBase
     index_type: IndexType
     index_params: dict[str, Any]
+    hidden: bool
 
     def __init__(self, metadata: MemorysetMetadata):
         # for internal use only, do not document
@@ -667,9 +670,10 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
         self.index_type = metadata["index_type"]
         self.index_params = metadata["index_params"]
         self.memory_type = metadata["memory_type"]
+        self.hidden = metadata["hidden"]
 
     def __eq__(self, other) -> bool:
-        return isinstance(other, _Memoryset) and self.id == other.id
+        return isinstance(other, MemorysetBase) and self.id == other.id
 
     def __repr__(self) -> str:
         return (
@@ -701,6 +705,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
         index_params: dict[str, Any] = {},
         if_exists: CreateMode = "error",
         background: Literal[True],
+        hidden: bool = False,
     ) -> Job[Self]:
         pass
 
@@ -725,6 +730,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
         index_params: dict[str, Any] = {},
         if_exists: CreateMode = "error",
         background: Literal[False] = False,
+        hidden: bool = False,
     ) -> Self:
         pass
 
@@ -748,6 +754,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
         index_params: dict[str, Any] = {},
         if_exists: CreateMode = "error",
         background: bool = False,
+        hidden: bool = False,
     ) -> Self | Job[Self]:
         """
         Create a new memoryset in the OrcaCloud
@@ -785,6 +792,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             if_exists: What to do if a memoryset with the same name already exists, defaults to
                 `"error"`. Other option is `"open"` to open the existing memoryset.
             background: Whether to run the operation none blocking and return a job handle
+            hidden: Whether the memoryset should be hidden
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -794,7 +802,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
                 `"open"` and the params do not match those of the existing memoryset.
         """
         if embedding_model is None:
-            embedding_model = PretrainedEmbeddingModel.CDE_SMALL
+            embedding_model = PretrainedEmbeddingModel.GTE_BASE
 
         if label_column is None and score_column is None:
             raise ValueError("label_column or score_column must be provided")
@@ -822,6 +830,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             "remove_duplicates": remove_duplicates,
             "index_type": index_type,
             "index_params": index_params,
+            "hidden": hidden,
         }
         if prompt is not None:
             payload["prompt"] = prompt
@@ -862,7 +871,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             name: Name for the new memoryset (must be unique)
             hf_dataset: Hugging Face dataset to create the memoryset from
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -926,7 +935,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
                 argument must be provided to specify the names of the columns.
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -984,7 +993,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             data: List of dictionaries to create the memoryset from
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -1046,7 +1055,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             data: Dictionary of columns to create the memoryset from
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -1109,7 +1118,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             dataframe: Dataframe to create the memoryset from
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -1165,7 +1174,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             pyarrow_table: PyArrow table to create the memoryset from
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -1230,7 +1239,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
                 - dataset directory: Directory containing a saved HuggingFace [`Dataset`][datasets.Dataset]
             background: Whether to run the operation in the background
             kwargs: Additional parameters for creating the memoryset. See
-                [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
+                [`create`][orca_sdk.memoryset.MemorysetBase.create] attributes for details.
 
         Returns:
             Handle to the new memoryset in the OrcaCloud
@@ -1274,14 +1283,20 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             return False
 
     @classmethod
-    def all(cls) -> list[Self]:
+    def all(cls, show_hidden: bool = False) -> list[Self]:
         """
         Get a list of handles to all memorysets in the OrcaCloud
 
+        Params:
+            show_hidden: Whether to include hidden memorysets in results, defaults to `False`
+
         Returns:
             List of handles to all memorysets in the OrcaCloud
         """
-        return [cls(metadata) for metadata in orca_api.GET("/memoryset", params={"type": cls.memory_type})]
+        return [
+            cls(metadata)
+            for metadata in orca_api.GET("/memoryset", params={"type": cls.memory_type, "show_hidden": show_hidden})
+        ]
 
     @classmethod
     def drop(cls, name_or_id: str, if_not_exists: DropMode = "error"):
@@ -1303,7 +1318,14 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             if if_not_exists == "error":
                 raise
 
-    def set(self, *, name: str = UNSET, description: str | None = UNSET, label_names: list[str] = UNSET):
+    def set(
+        self,
+        *,
+        name: str = UNSET,
+        description: str | None = UNSET,
+        label_names: list[str] = UNSET,
+        hidden: bool = UNSET,
+    ):
         """
         Update editable attributes of the memoryset
 
@@ -1322,6 +1344,8 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             payload["description"] = description
         if label_names is not UNSET:
             payload["label_names"] = label_names
+        if hidden is not UNSET:
+            payload["hidden"] = hidden
 
         orca_api.PATCH("/memoryset/{name_or_id}", params={"name_or_id": self.id}, json=payload)
         self.refresh()
@@ -1370,10 +1394,10 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
             embedding_model: Optional new embedding model to use for re-embedding the memory values
                 value is longer than this it will be truncated, will default to the model's max
                 sequence length if not provided
-            document_prompt_override: Optional custom document prompt to use for the cloned memoryset.
-                If not provided, will use the source memoryset's document prompt.
-            query_prompt_override: Optional custom query prompt to use for the cloned memoryset.
-                If not provided, will use the source memoryset's query prompt.
+            max_seq_length_override: Optional custom max sequence length to use for the cloned memoryset.
+                If not provided, will use the source memoryset's max sequence length.
+            prompt: Optional custom prompt to use for the cloned memoryset.
+                If not provided, will use the source memoryset's prompt.
             if_exists: What to do if a memoryset with the same name already exists, defaults to
                 `"error"`. Other option is `"open"` to open the existing memoryset.
 
@@ -1854,7 +1878,6 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
                 to be considered again.
             label_confirmation_cooldown_time: Minimum time (in seconds) since a neighbor's label was confirmed
                 to be considered for suggestions.
-            _current_time: Optional override for the current timestamp (useful for testing).
 
         Returns:
             A list of CascadingEditSuggestion objects, each containing a neighbor and the suggested new label.
@@ -2115,7 +2138,7 @@ class _Memoryset(Generic[MemoryT, MemoryLookupT], ABC):
         return job if background else job.result()
 
 
-class LabeledMemoryset(_Memoryset[LabeledMemory, LabeledMemoryLookup]):
+class LabeledMemoryset(MemorysetBase[LabeledMemory, LabeledMemoryLookup]):
     """
     A Handle to a collection of memories with labels in the OrcaCloud
 
@@ -2145,8 +2168,19 @@ class LabeledMemoryset(_Memoryset[LabeledMemory, LabeledMemoryLookup]):
     def create(cls, name: str, datasource: Datasource, *, label_column: str | None = "label", **kwargs):
         return super().create(name, datasource, label_column=label_column, score_column=None, **kwargs)
 
+    def display_label_analysis(self):
+        """
+        Display an interactive UI to review and act upon the label analysis results
+
+        Note:
+            This method is only available in Jupyter notebooks.
+        """
+        from ._utils.analysis_ui import display_suggested_memory_relabels
+
+        display_suggested_memory_relabels(self)
+
 
-class ScoredMemoryset(_Memoryset[ScoredMemory, ScoredMemoryLookup]):
+class ScoredMemoryset(MemorysetBase[ScoredMemory, ScoredMemoryLookup]):
     """
     A Handle to a collection of memories with scores in the OrcaCloud
 

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/memoryset_test.py

@@ -122,6 +122,26 @@ def test_all_memorysets(readonly_memoryset: LabeledMemoryset):
     assert any(memoryset.name == readonly_memoryset.name for memoryset in memorysets)
 
 
+def test_all_memorysets_hidden(
+    readonly_memoryset: LabeledMemoryset,
+):
+    # Create a hidden memoryset
+    hidden_memoryset = LabeledMemoryset.clone(readonly_memoryset, "test_hidden_memoryset")
+    hidden_memoryset.set(hidden=True)
+
+    # Test that show_hidden=False excludes hidden memorysets
+    visible_memorysets = LabeledMemoryset.all(show_hidden=False)
+    assert len(visible_memorysets) > 0
+    assert readonly_memoryset in visible_memorysets
+    assert hidden_memoryset not in visible_memorysets
+
+    # Test that show_hidden=True includes hidden memorysets
+    all_memorysets = LabeledMemoryset.all(show_hidden=True)
+    assert len(all_memorysets) == len(visible_memorysets) + 1
+    assert readonly_memoryset in all_memorysets
+    assert hidden_memoryset in all_memorysets
+
+
 def test_all_memorysets_unauthenticated(unauthenticated):
     with pytest.raises(ValueError, match="Invalid API key"):
         LabeledMemoryset.all()
@@ -167,6 +187,9 @@ def test_update_memoryset_attributes(writable_memoryset: LabeledMemoryset):
     writable_memoryset.set(label_names=["New label 1", "New label 2"])
     assert writable_memoryset.label_names == ["New label 1", "New label 2"]
 
+    writable_memoryset.set(hidden=True)
+    assert writable_memoryset.hidden is True
+
 
 def test_search(readonly_memoryset: LabeledMemoryset):
     memory_lookups = readonly_memoryset.search(["i love soup", "cats are cute"])

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/orca_sdk/telemetry.py

@@ -152,7 +152,7 @@ class AddMemorySuggestions:
         )
 
 
-class _Prediction(ABC):
+class PredictionBase(ABC):
     prediction_id: str | None
     confidence: float
     anomaly_score: float | None
@@ -460,8 +460,19 @@ class _Prediction(ABC):
         )
         self.refresh()
 
+    def inspect(self) -> None:
+        """
+        Display an interactive UI with the details about this prediction
+
+        Note:
+            This method is only available in Jupyter notebooks.
+        """
+        from ._utils.prediction_result_ui import inspect_prediction_result
+
+        inspect_prediction_result(self)
+
 
-class ClassificationPrediction(_Prediction):
+class ClassificationPrediction(PredictionBase):
     """
     Labeled prediction result from a [`ClassificationModel`][orca_sdk.ClassificationModel]
 
@@ -614,7 +625,7 @@ class ClassificationPrediction(_Prediction):
         )
 
 
-class RegressionPrediction(_Prediction):
+class RegressionPrediction(PredictionBase):
     """
     Score-based prediction result from a [`RegressionModel`][orca_sdk.RegressionModel]
 

{orca_sdk-0.0.101 → orca_sdk-0.0.103}/pyproject.toml

@@ -1,22 +1,30 @@
-[tool.poetry]
+[project]
 name = "orca_sdk"
-version = "0.0.101" # Will be set by CI before building the wheel from the git tag. Do not set manually.
 description = "SDK for interacting with Orca Services"
-authors = ["Orca DB Inc. <dev-rel@orcadb.ai>"]
-license = "Apache-2.0"
+license = {text = "Apache-2.0"}
+authors = [
+    {name = "Orca DB Inc.", email = "dev-rel@orcadb.ai"}
+]
+
+dynamic = ["version", "readme", "dependencies", "requires-python"]
+
+[tool.poetry]
+version = "0.0.103" # Will be set by CI before building the wheel from the git tag. Do not set manually.
 readme = "README.md"
 packages = [{ include = "orca_sdk" }]
 
 [tool.poetry.dependencies]
-python = "^3.11"
-httpx = ">=0.20.0,<0.29.0"
+python = "^3.11,<3.14"
 datasets = "^3.1.0"
+httpx = "^0.28.1"
+python-dotenv = "^1.1.0"
+# TODO: make these optional dependencies
+gradio = "^5.44.1"
+numpy = "^2.1.0"
 pandas = "^2.2.3"
 pyarrow = "^18.0.0"
-torch = "^2.5.1"
-gradio = "5.13.0"
-python-dotenv = "^1.1.0"
 scikit-learn = "^1.6.1"
+torch = "^2.8.0"
 
 [tool.poetry.group.dev]
 optional = true
@@ -28,14 +36,14 @@ pytest = "^8.3.3"
 pytest-asyncio = "^0.25.3"
 pytest-timeout = "^2.3.1"
 pytest-cov = "^6.0.0"
-poethepoet = "^0.31.1"
 pyright = "^1.1.399"
 datamodel-code-generator = "^0.32.0"
 ipykernel = "^6.29.5"
 
-# TODO: Use this once poetry 2.0.0 is released https://github.com/python-poetry/poetry/issues/9448
-# [tool.poetry.requires-plugins]
-# poethepoet = { extras = ["poetry-plugin"], version = "^0.31.0" }
+[tool.poetry.requires-plugins]
+poethepoet = { extras = ["poetry-plugin"], version = "^0.37.0" }
+poetry-plugin-export = ">=1.8"
+
 
 
 [tool.poe]
@@ -46,6 +54,7 @@ test = "pytest"
 typecheck = "pyright"
 codegen = "python -m scripts.codegen"
 lighthouse = { shell = "cd ../lighthouse && poetry run poe dev" }
+vulnerabilities = "trivy fs . --scanners vuln --ignore-unfixed"
 
 [tool.pytest.ini_options]
 log_cli = true
@@ -70,4 +79,3 @@ known_first_party = ["orca_sdk"]
 
 [tool.pyright]
 include = ["./orca_sdk/**", "./scripts/**"]
-