orca-sdk 0.1.0__tar.gz → 0.1.1__tar.gz

orca_sdk-0.1.1/PKG-INFO

@@ -0,0 +1,79 @@
+Metadata-Version: 2.1
+Name: orca_sdk
+Version: 0.1.1
+Summary: SDK for interacting with Orca Services
+License: Apache-2.0
+Author: Orca DB Inc.
+Author-email: dev-rel@orcadb.ai
+Requires-Python: >=3.10,<3.13
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Requires-Dist: attrs (>=22.2.0)
+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: pandas (>=2.2.3,<3.0.0)
+Requires-Dist: pyarrow (>=18.0.0,<19.0.0)
+Requires-Dist: python-dateutil (>=2.8.0,<3.0.0)
+Requires-Dist: torch (>=2.5.1,<3.0.0)
+Description-Content-Type: text/markdown
+
+<!--
+IMPORTANT NOTE:
+- This file will get rendered in the public facing PyPi page here: https://pypi.org/project/orca_sdk/
+- Only content suitable for public consumption should be placed in this file everything else should go into CONTRIBUTING.md
+-->
+
+# OrcaSDK
+
+OrcaSDK is a Python library for building and using retrieval-augmented models with [OrcaCloud](https://orcadb.ai). It enables you to create, deploy, and maintain models that can adapt to changing circumstances without retraining by accessing external data called "memories."
+
+## Documentation
+
+You can find the documentation for all things Orca at [docs.orcadb.ai](https://docs.orcadb.ai). This includes tutorials, how-to guides, and the full interface reference for OrcaSDK.
+
+## Features
+
+- **Labeled Memorysets**: Store and manage labeled examples that your models can use to guide predictions
+- **Classification Models**: Build retrieval-augmented classification models that adapt to new data without retraining
+- **Embedding Models**: Use pre-trained or fine-tuned embedding models to represent your data
+- **Telemetry**: Collect feedback and monitor memory usage to optimize model performance
+- **Datasources**: Easily ingest data from various sources into your memorysets
+
+## Installation
+
+Orca SDK is compatible with Python 3.10 or higher and is available on [PyPI](https://pypi.org/project/orca_sdk/). You can install it with your favorite python package manager:
+
+- Pip: `pip install orca_sdk`
+- Conda: `conda install orca_sdk`
+- Poetry: `poetry add orca_sdk`
+
+## Quick Start
+
+```python
+from dotenv import load_dotenv
+from orca_sdk import OrcaCredentials, LabeledMemoryset, ClassificationModel
+
+# Load your API key from environment variables
+load_dotenv()
+assert OrcaCredentials.is_authenticated()
+
+# Create a labeled memoryset
+memoryset = LabeledMemoryset.from_disk("my_memoryset", "./data.jsonl")
+
+# Create a classification model using the memoryset
+model = ClassificationModel("my_model", memoryset)
+
+# Make predictions
+prediction = model.predict("my input")
+```
+
+For a more detailed walkthrough, check out our [Quick Start Guide](https://docs.orcadb.ai/quickstart/).
+
+## Support
+
+If you have any questions, please reach out to us at support@orcadb.ai.
+

orca_sdk-0.1.1/README.md

@@ -0,0 +1,55 @@
+<!--
+IMPORTANT NOTE:
+- This file will get rendered in the public facing PyPi page here: https://pypi.org/project/orca_sdk/
+- Only content suitable for public consumption should be placed in this file everything else should go into CONTRIBUTING.md
+-->
+
+# OrcaSDK
+
+OrcaSDK is a Python library for building and using retrieval-augmented models with [OrcaCloud](https://orcadb.ai). It enables you to create, deploy, and maintain models that can adapt to changing circumstances without retraining by accessing external data called "memories."
+
+## Documentation
+
+You can find the documentation for all things Orca at [docs.orcadb.ai](https://docs.orcadb.ai). This includes tutorials, how-to guides, and the full interface reference for OrcaSDK.
+
+## Features
+
+- **Labeled Memorysets**: Store and manage labeled examples that your models can use to guide predictions
+- **Classification Models**: Build retrieval-augmented classification models that adapt to new data without retraining
+- **Embedding Models**: Use pre-trained or fine-tuned embedding models to represent your data
+- **Telemetry**: Collect feedback and monitor memory usage to optimize model performance
+- **Datasources**: Easily ingest data from various sources into your memorysets
+
+## Installation
+
+Orca SDK is compatible with Python 3.10 or higher and is available on [PyPI](https://pypi.org/project/orca_sdk/). You can install it with your favorite python package manager:
+
+- Pip: `pip install orca_sdk`
+- Conda: `conda install orca_sdk`
+- Poetry: `poetry add orca_sdk`
+
+## Quick Start
+
+```python
+from dotenv import load_dotenv
+from orca_sdk import OrcaCredentials, LabeledMemoryset, ClassificationModel
+
+# Load your API key from environment variables
+load_dotenv()
+assert OrcaCredentials.is_authenticated()
+
+# Create a labeled memoryset
+memoryset = LabeledMemoryset.from_disk("my_memoryset", "./data.jsonl")
+
+# Create a classification model using the memoryset
+model = ClassificationModel("my_model", memoryset)
+
+# Make predictions
+prediction = model.predict("my input")
+```
+
+For a more detailed walkthrough, check out our [Quick Start Guide](https://docs.orcadb.ai/quickstart/).
+
+## Support
+
+If you have any questions, please reach out to us at support@orcadb.ai.

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/__init__.py

@@ -5,14 +5,19 @@ OrcaSDK is a Python library for building and using retrieval augmented models in
 from ._utils.common import UNSET, CreateMode, DropMode
 from ._utils.task import TaskStatus
 from .classification_model import ClassificationModel
+from .credentials import OrcaCredentials
 from .datasource import Datasource
 from .embedding_model import (
     FinetunedEmbeddingModel,
     PretrainedEmbeddingModel,
     PretrainedEmbeddingModelName,
 )
-from .labeled_memoryset import LabeledMemory, LabeledMemoryLookup, LabeledMemoryset
-from .orca_credentials import OrcaCredentials
+from .memoryset import (
+    FilterItemTuple,
+    LabeledMemory,
+    LabeledMemoryLookup,
+    LabeledMemoryset,
+)
 from .telemetry import LabelPrediction
 
 # only specify things that should show up on the root page of the reference docs because they are in private modules

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/_utils/analysis_ui.py

@@ -7,7 +7,7 @@ from typing import TypedDict, cast
 
 import gradio as gr
 
-from ..labeled_memoryset import LabeledMemory, LabeledMemoryset
+from ..memoryset import LabeledMemory, LabeledMemoryset
 
 # Suppress all httpx logs
 logging.getLogger("httpx").setLevel(logging.CRITICAL)

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/_utils/auth_test.py

@@ -1,6 +1,6 @@
 from uuid import uuid4
 
-from ..orca_credentials import OrcaCredentials
+from ..credentials import OrcaCredentials
 from .auth import _create_api_key, _delete_api_key, _delete_org, _list_api_keys
 
 

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/_utils/prediction_result_ui.py

@@ -5,7 +5,7 @@ from typing import TYPE_CHECKING
 
 import gradio as gr
 
-from ..labeled_memoryset import LabeledMemoryLookup
+from ..memoryset import LabeledMemoryLookup
 
 if TYPE_CHECKING:
     from ..telemetry import LabelPrediction

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/classification_model.py

@@ -34,7 +34,7 @@ from ._generated_api_client.models.prediction_request import PredictionRequest
 from ._utils.common import CreateMode, DropMode
 from ._utils.task import wait_for_task
 from .datasource import Datasource
-from .labeled_memoryset import LabeledMemoryset
+from .memoryset import LabeledMemoryset
 from .telemetry import LabelPrediction, _parse_feedback
 
 

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/classification_model_test.py

@@ -6,7 +6,7 @@ from datasets.arrow_dataset import Dataset
 from .classification_model import ClassificationModel
 from .datasource import Datasource
 from .embedding_model import PretrainedEmbeddingModel
-from .labeled_memoryset import LabeledMemoryset
+from .memoryset import LabeledMemoryset
 
 
 def test_create_model(model: ClassificationModel, memoryset: LabeledMemoryset):

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/conftest.py

@@ -9,10 +9,10 @@ from datasets import ClassLabel, Dataset, Features, Value
 from ._generated_api_client.client import set_headers
 from ._utils.auth import _create_api_key, _delete_org
 from .classification_model import ClassificationModel
+from .credentials import OrcaCredentials
 from .datasource import Datasource
 from .embedding_model import PretrainedEmbeddingModel
-from .labeled_memoryset import LabeledMemoryset
-from .orca_credentials import OrcaCredentials
+from .memoryset import LabeledMemoryset
 
 logging.basicConfig(level=logging.INFO)
 

orca_sdk-0.1.0/orca_sdk/orca_credentials_test.py → orca_sdk-0.1.1/orca_sdk/credentials_test.py

@@ -2,7 +2,7 @@ from uuid import uuid4
 
 import pytest
 
-from .orca_credentials import OrcaCredentials
+from .credentials import OrcaCredentials
 
 
 def test_list_api_keys():

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/embedding_model.py

@@ -28,7 +28,7 @@ from ._utils.task import TaskStatus, wait_for_task
 from .datasource import Datasource
 
 if TYPE_CHECKING:
-    from .labeled_memoryset import LabeledMemoryset
+    from .memoryset import LabeledMemoryset
 
 
 class _EmbeddingModel:
@@ -57,6 +57,18 @@ class _EmbeddingModel:
         pass
 
     def embed(self, value: str | list[str], max_seq_length: int | None = None) -> list[float] | list[list[float]]:
+        """
+        Generate embeddings for a value or list of values
+
+        Params:
+            value: The value or list of values to embed
+            max_seq_length: The maximum sequence length to truncate the input to
+
+        Returns:
+            A matrix of floats representing the embedding for each value if the input is a list of
+                values, or a list of floats representing the embedding for the single value if the
+                input is a single value
+        """
         request = EmbedRequest(values=value if isinstance(value, list) else [value], max_seq_length=max_seq_length)
         if isinstance(self, PretrainedEmbeddingModel):
             embeddings = embed_with_pretrained_model_gpu(self._model_name, body=request)
@@ -201,7 +213,7 @@ class PretrainedEmbeddingModel(_EmbeddingModel, metaclass=_PretrainedEmbeddingMo
 
             return existing
 
-        from .labeled_memoryset import LabeledMemoryset
+        from .memoryset import LabeledMemoryset
 
         train_datasource_id = train_datasource.id if isinstance(train_datasource, Datasource) else None
         train_memoryset_id = train_datasource.id if isinstance(train_datasource, LabeledMemoryset) else None
@@ -265,7 +277,6 @@ class FinetunedEmbeddingModel(_EmbeddingModel):
             f"    name: {self.name},\n"
             f"    embedding_dim: {self.embedding_dim},\n"
             f"    max_seq_length: {self.max_seq_length},\n"
-            f"    status: {self._status}\n"
             f"    base_model: PretrainedEmbeddingModel.{self.base_model_name.value}\n"
             "})"
         )

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/embedding_model_test.py

@@ -9,7 +9,7 @@ from .embedding_model import (
     PretrainedEmbeddingModelName,
     TaskStatus,
 )
-from .labeled_memoryset import LabeledMemoryset
+from .memoryset import LabeledMemoryset
 
 
 def test_open_pretrained_model():

orca_sdk-0.1.0/orca_sdk/labeled_memoryset.py → orca_sdk-0.1.1/orca_sdk/memoryset.py

@@ -442,7 +442,7 @@ class LabeledMemoryset:
         return cls.open(response.id)
 
     @classmethod
-    def from_hf_dataset(cls, name: str, hf_dataset: Dataset, **kwargs) -> LabeledMemoryset:
+    def from_hf_dataset(cls, name: str, hf_dataset: Dataset, **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a Hugging Face [`Dataset`][datasets.Dataset] in the OrcaCloud
 
@@ -455,7 +455,7 @@ class LabeledMemoryset:
         Params:
             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
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
 
@@ -475,7 +475,7 @@ class LabeledMemoryset:
         torch_data: TorchDataLoader | TorchDataset,
         *,
         column_names: list[str] | None = None,
-        **kwargs,
+        **kwargs: Any,
     ) -> LabeledMemoryset:
         """
         Create a new memoryset from a PyTorch [`DataLoader`][torch.utils.data.DataLoader] or
@@ -492,7 +492,7 @@ class LabeledMemoryset:
             torch_data: PyTorch data loader or dataset to create the memoryset from
             column_names: If the provided dataset or data loader returns unnamed tuples, this
                 argument must be provided to specify the names of the columns.
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
 
@@ -505,7 +505,7 @@ class LabeledMemoryset:
         return cls.create(name, datasource, **kwargs)
 
     @classmethod
-    def from_list(cls, name: str, data: list[dict], **kwargs) -> LabeledMemoryset:
+    def from_list(cls, name: str, data: list[dict], **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a list of dictionaries in the OrcaCloud
 
@@ -518,7 +518,7 @@ class LabeledMemoryset:
         Params:
             name: Name for the new memoryset (must be unique)
             data: List of dictionaries to create the memoryset from
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
         Returns:
@@ -534,7 +534,7 @@ class LabeledMemoryset:
         return cls.create(name, datasource, **kwargs)
 
     @classmethod
-    def from_dict(cls, name: str, data: dict, **kwargs) -> LabeledMemoryset:
+    def from_dict(cls, name: str, data: dict, **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a dictionary of columns in the OrcaCloud
 
@@ -547,7 +547,7 @@ class LabeledMemoryset:
         Params:
             name: Name for the new memoryset (must be unique)
             data: Dictionary of columns to create the memoryset from
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
         Returns:
@@ -564,7 +564,7 @@ class LabeledMemoryset:
         return cls.create(name, datasource, **kwargs)
 
     @classmethod
-    def from_pandas(cls, name: str, dataframe: pd.DataFrame, **kwargs) -> LabeledMemoryset:
+    def from_pandas(cls, name: str, dataframe: pd.DataFrame, **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a pandas [`DataFrame`][pandas.DataFrame] in the OrcaCloud
 
@@ -577,7 +577,7 @@ class LabeledMemoryset:
         Params:
             name: Name for the new memoryset (must be unique)
             dataframe: Dataframe to create the memoryset from
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
         Returns:
@@ -587,7 +587,7 @@ class LabeledMemoryset:
         return cls.create(name, datasource, **kwargs)
 
     @classmethod
-    def from_arrow(cls, name: str, pyarrow_table: pa.Table, **kwargs) -> LabeledMemoryset:
+    def from_arrow(cls, name: str, pyarrow_table: pa.Table, **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a PyArrow [`Table`][pyarrow.Table] in the OrcaCloud
 
@@ -600,7 +600,7 @@ class LabeledMemoryset:
         Params:
             name: Name for the new memoryset (must be unique)
             pyarrow_table: PyArrow table to create the memoryset from
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
         Returns:
@@ -612,7 +612,7 @@ class LabeledMemoryset:
         return cls.create(name, datasource, **kwargs)
 
     @classmethod
-    def from_disk(cls, name: str, file_path: str | PathLike, **kwargs) -> LabeledMemoryset:
+    def from_disk(cls, name: str, file_path: str | PathLike, **kwargs: Any) -> LabeledMemoryset:
         """
         Create a new memoryset from a file on disk in the OrcaCloud
 
@@ -632,7 +632,7 @@ class LabeledMemoryset:
                 - .csv: [`CSV`][csv] files
                 - .parquet: [`Parquet`][pyarrow.parquet.ParquetFile] files
                 - dataset directory: Directory containing a saved HuggingFace [`Dataset`][datasets.Dataset]
-            **kwargs: Additional parameters for creating the memoryset. See
+            kwargs: Additional parameters for creating the memoryset. See
                 [`create`][orca_sdk.LabeledMemoryset.create] attributes for details.
 
         Returns:
@@ -932,7 +932,7 @@ class LabeledMemoryset:
         Insert memories into the memoryset
 
         Params:
-            memories: List of memories to insert into the memoryset. This should be a list of
+            items: List of memories to insert into the memoryset. This should be a list of
                 dictionaries with the following keys:
 
                 - `value`: Value of the memory
@@ -1056,14 +1056,14 @@ class LabeledMemoryset:
         Delete memories from the memoryset
 
         Params:
-            memory_ids: unique identifiers of the memories to delete
+            memory_id: unique identifiers of the memories to delete
 
         Examples:
             Delete a single memory:
             >>> memoryset.delete("0195019a-5bc7-7afb-b902-5945ee1fb766")
 
             Delete multiple memories:
-            >>> memoryset.delete(
+            >>> memoryset.delete([
             ...     "0195019a-5bc7-7afb-b902-5945ee1fb766",
             ...     "019501a1-ea08-76b2-9f62-95e4800b4841",
             ... )
@@ -1148,7 +1148,7 @@ class LabeledMemoryset:
         return analysis.result.to_dict()
 
     def display_label_analysis(self):
+        """Display a UI to review and act upon the label analysis results"""
         from ._utils.analysis_ui import display_suggested_memory_relabels
 
-        """Display a UI to review and act upon the label analysis results"""
         display_suggested_memory_relabels(self)

orca_sdk-0.1.0/orca_sdk/labeled_memoryset_test.py → orca_sdk-0.1.1/orca_sdk/memoryset_test.py

@@ -4,7 +4,7 @@ import pytest
 from datasets.arrow_dataset import Dataset
 
 from .embedding_model import PretrainedEmbeddingModel
-from .labeled_memoryset import LabeledMemoryset, TaskStatus
+from .memoryset import LabeledMemoryset, TaskStatus
 
 
 def test_create_memoryset(memoryset: LabeledMemoryset, hf_dataset: Dataset, label_names: list[str]):

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/telemetry.py

@@ -25,7 +25,7 @@ from ._generated_api_client.models import (
 )
 from ._generated_api_client.types import UNSET as CLIENT_UNSET
 from ._utils.prediction_result_ui import inspect_prediction_result
-from .labeled_memoryset import LabeledMemoryLookup, LabeledMemoryset
+from .memoryset import LabeledMemoryLookup, LabeledMemoryset
 
 if TYPE_CHECKING:
     from .classification_model import ClassificationModel

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/orca_sdk/telemetry_test.py

@@ -1,7 +1,7 @@
 import pytest
 
 from .classification_model import ClassificationModel
-from .labeled_memoryset import LabeledMemoryLookup
+from .memoryset import LabeledMemoryLookup
 from .telemetry import FeedbackCategory, LabelPrediction
 
 

{orca_sdk-0.1.0 → orca_sdk-0.1.1}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "orca_sdk"
-version = "0.1.0"
+version = "0.1.1"
 description = "SDK for interacting with Orca Services"
 authors = ["Orca DB Inc. <dev-rel@orcadb.ai>"]
 license = "Apache-2.0"

orca_sdk-0.1.0/PKG-INFO

@@ -1,39 +0,0 @@
-Metadata-Version: 2.1
-Name: orca_sdk
-Version: 0.1.0
-Summary: SDK for interacting with Orca Services
-License: Apache-2.0
-Author: Orca DB Inc.
-Author-email: dev-rel@orcadb.ai
-Requires-Python: >=3.10,<3.13
-Classifier: License :: OSI Approved :: Apache Software License
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Requires-Dist: attrs (>=22.2.0)
-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: pandas (>=2.2.3,<3.0.0)
-Requires-Dist: pyarrow (>=18.0.0,<19.0.0)
-Requires-Dist: python-dateutil (>=2.8.0,<3.0.0)
-Requires-Dist: torch (>=2.5.1,<3.0.0)
-Description-Content-Type: text/markdown
-
-# Orca SDK
-
-Orca SDK is a Python library for interacting with Orca RAC.
-
-## Documentation
-
-TBD
-
-## Installation
-
-TBD
-
-## Support
-
-If you have any questions, please reach out to us at support@orcadb.ai.
-

orca_sdk-0.1.0/README.md

@@ -1,15 +0,0 @@
-# Orca SDK
-
-Orca SDK is a Python library for interacting with Orca RAC.
-
-## Documentation
-
-TBD
-
-## Installation
-
-TBD
-
-## Support
-
-If you have any questions, please reach out to us at support@orcadb.ai.