Add Inference embed and rerank

.github/workflows/testing-integration.yaml

@@ -3,28 +3,6 @@ name: "Integration Tests"
   workflow_call: {}
 
 jobs:
-  plugin-inference:
-    name: Test inference plugin
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        python_version: [3.9, 3.12]
-    steps:
-      - uses: actions/checkout@v4
-      - name: 'Set up Python ${{ matrix.python_version }}'
-        uses: actions/setup-python@v5
-        with:
-          python-version: '${{ matrix.python_version }}'
-      - name: Setup Poetry
-        uses: ./.github/actions/setup-poetry
-        with:
-          include_grpc: 'true'
-      - name: 'Run integration tests'
-        run: poetry run pytest tests/integration/inference -s -vv
-        env:
-          PINECONE_DEBUG_CURL: 'true'
-          PINECONE_API_KEY: '${{ secrets.PINECONE_API_KEY }}'
-
   data-plane-serverless:
     name: Data plane serverless integration tests
     runs-on: ubuntu-latest

codegen/build-oas.sh

@@ -8,12 +8,12 @@ is_early_access=$2 # e.g. true
 # if is_early_access is true, add the "ea" module
 if [ "$is_early_access" = "true" ]; then
 	destination="pinecone/core_ea/openapi"
-	modules=("db_control" "db_data")
+	modules=("db_control" "db_data" "inference")
 	py_module_name="core_ea"
 	template_dir="codegen/python-oas-templates/templates5.2.0"
 else
 	destination="pinecone/core/openapi"
-	modules=("db_control" "db_data")
+	modules=("db_control" "db_data" "inference")
 	py_module_name="core"
 	template_dir="codegen/python-oas-templates/templates5.2.0"
 fi

pinecone/__init__.py

@@ -7,6 +7,7 @@
 
 warnings.filterwarnings("ignore", category=TqdmExperimentalWarning)
 
+from .deprecated_plugins import check_for_deprecated_plugins
 from .deprecation_warnings import *
 from .config import *
 from .exceptions import *
@@ -18,4 +19,9 @@
 
 import logging
 
+# Raise an exception if the user is attempting to use the SDK with deprecated plugins
+# installed in their project.
+check_for_deprecated_plugins()
+
+# Silence annoying log messages from the plugin interface
 logging.getLogger("pinecone_plugin_interface").setLevel(logging.CRITICAL)

pinecone/control/pinecone.py

@@ -30,7 +30,7 @@
 from .langchain_import_warnings import _build_langchain_attribute_error_message
 from pinecone.utils import parse_non_empty_args, docslinks
 
-from pinecone.data import _Index, _AsyncioIndex
+from pinecone.data import _Index, _AsyncioIndex, _Inference
 
 from pinecone_plugin_interface import load_and_install as install_plugins
 
@@ -85,6 +85,8 @@ def __init__(
         self.openapi_config = ConfigBuilder.build_openapi_config(self.config, **kwargs)
         self.pool_threads = pool_threads
 
+        self._inference = None  # Lazy initialization
+
         if index_api:
             self.index_api = index_api
         else:
@@ -102,6 +104,13 @@ def __init__(
 
         self.load_plugins()
 
+    @property
+    def inference(self):
+        """Dynamically create and cache the Inference instance."""
+        if self._inference is None:
+            self._inference = _Inference(config=self.config, openapi_config=self.openapi_config)
+        return self._inference
+
     def load_plugins(self):
         """@private"""
         try:

pinecone/core/openapi/db_control/api/manage_indexes_api.py

@@ -40,8 +40,7 @@
 
 
 class ManageIndexesApi(object):
-    """NOTE: This class is auto generated by OpenAPI Generator
-    Ref: https://openapi-generator.tech
+    """NOTE: This class is auto generated.
 
     Do not edit the class manually.
     """

pinecone/core/openapi/db_data/api/bulk_operations_api.py

@@ -34,8 +34,7 @@
 
 
 class BulkOperationsApi(object):
-    """NOTE: This class is auto generated by OpenAPI Generator
-    Ref: https://openapi-generator.tech
+    """NOTE: This class is auto generated.
 
     Do not edit the class manually.
     """

pinecone/core/openapi/db_data/api/vector_operations_api.py

@@ -45,8 +45,7 @@
 
 
 class VectorOperationsApi(object):
-    """NOTE: This class is auto generated by OpenAPI Generator
-    Ref: https://openapi-generator.tech
+    """NOTE: This class is auto generated.
 
     Do not edit the class manually.
     """

pinecone/core/openapi/inference/__init__.py

@@ -0,0 +1,29 @@
+# flake8: noqa
+
+"""
+Pinecone Inference API
+
+Pinecone is a vector database that makes it easy to search and retrieve billions of high-dimensional vectors.  # noqa: E501
+
+The version of the OpenAPI document: 2025-01
+Contact: support@pinecone.io
+Generated by: https://openapi-generator.tech
+"""
+
+__version__ = "1.0.0"
+
+# import ApiClient
+from pinecone.openapi_support import ApiClient
+
+# import Configuration
+from pinecone.openapi_support import Configuration
+
+# import exceptions
+from pinecone.openapi_support.exceptions import PineconeException
+from pinecone.openapi_support.exceptions import PineconeApiAttributeError
+from pinecone.openapi_support.exceptions import PineconeApiTypeError
+from pinecone.openapi_support.exceptions import PineconeApiValueError
+from pinecone.openapi_support.exceptions import PineconeApiKeyError
+from pinecone.openapi_support.exceptions import PineconeApiException
+
+API_VERSION = "2025-01"

pinecone/core/openapi/inference/api/__init__.py

@@ -0,0 +1,3 @@
+# do not import all apis into this module because that uses a lot of memory and stack frames
+# if you need the ability to import all apis from one package, import them with
+# from pinecone.core.openapi.inference.apis import InferenceApi