From 27f5a621dd6cc8e0d5534641a4b890480ec00f41 Mon Sep 17 00:00:00 2001 From: kirgrim Date: Mon, 14 Apr 2025 16:34:30 +0200 Subject: [PATCH 1/6] feat(py): add vertex ai vector search + samples for big query and firestore --- .../src/genkit/plugins/google_genai/google.py | 124 ++++++++++- .../plugins/google_genai/models/retriever.py | 165 ++++++++++++++ .../google_genai/models/vectorstore.py | 55 +++++ .../LICENSE | 201 ++++++++++++++++++ .../README.md | 20 ++ .../pyproject.toml | 39 ++++ .../src/sample.py | 99 +++++++++ .../LICENSE | 201 ++++++++++++++++++ .../README.md | 20 ++ .../pyproject.toml | 39 ++++ .../src/sample.py | 96 +++++++++ py/uv.lock | 48 +++++ 12 files changed, 1106 insertions(+), 1 deletion(-) create mode 100644 py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py create mode 100644 py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py create mode 100644 py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE create mode 100644 py/samples/google-genai-vertexai-vector-search-bigquery/README.md create mode 100644 py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml create mode 100644 py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py create mode 100644 py/samples/google-genai-vertexai-vector-search-firestore/LICENSE create mode 100644 py/samples/google-genai-vertexai-vector-search-firestore/README.md create mode 100644 py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml create mode 100644 py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py index fd76e802ec..24f9d0daeb 100644 --- a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py +++ b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py @@ -15,11 +15,13 @@ # SPDX-License-Identifier: Apache-2.0 import os +from typing import Any, Type from google import genai from google.auth.credentials import Credentials +from google.cloud import aiplatform_v1, storage from google.genai.client import DebugConfig -from google.genai.types import HttpOptions, HttpOptionsDict +from google.genai.types import HttpOptions, HttpOptionsDict, Operation from genkit.ai import GENKIT_CLIENT_HEADER, GenkitRegistry, Plugin from genkit.plugins.google_genai.models.embedder import ( @@ -34,6 +36,8 @@ VertexAIGeminiVersion, ) from genkit.plugins.google_genai.models.imagen import ImagenModel, ImagenVersion +from genkit.plugins.google_genai.models.retriever import VertexAIVectorStoreRetriever +from genkit.plugins.google_genai.models.vectorstore import IndexConfig GOOGLEAI_PLUGIN_NAME = 'googleai' VERTEXAI_PLUGIN_NAME = 'vertexai' @@ -165,6 +169,124 @@ def initialize(self, ai: GenkitRegistry) -> None: ai.define_model(name=vertexai_name(version), fn=imagen_model.generate, metadata=imagen_model.metadata) +class VertexAIVectorSearch(Plugin): + """VertexAI vector store plugin for Genkit.""" + + name: str = 'vertexAIVectorstore' + + def __init__( + self, + retriever: Type[VertexAIVectorStoreRetriever], + retriever_extra_args: dict[str, Any] | None = None, + credentials: Credentials | None = None, + project: str | None = None, + location: str | None = 'us-central1', + embedder: str | None = None, + embedder_options: dict[str, Any] | None = None, + http_options: HttpOptions | HttpOptionsDict | None = None, + ): + http_options = _inject_attribution_headers(http_options=http_options) + + self.project = project + self.location = location + + self.embedder = embedder + self.embedder_options = embedder_options + + self.retriever_cls = retriever + self.retriever_extra_args = retriever_extra_args or {} + + self._storage_client = storage.Client( + project=self.project, + credentials=credentials, + extra_headers=http_options.headers, + ) + self._index_client = aiplatform_v1.IndexServiceAsyncClient( + credentials=credentials, + ) + self._endpoint_client = aiplatform_v1.IndexEndpointServiceAsyncClient(credentials=credentials) + self._match_service_client = aiplatform_v1.MatchServiceAsyncClient( + credentials=credentials, + ) + + async def create_index( + self, + display_name: str, + description: str | None, + index_config: IndexConfig | None = None, + contents_delta_uri: str | None = None, + ) -> None: + if not index_config: + index_config = IndexConfig() + + index = aiplatform_v1.Index() + index.display_name = display_name + index.description = description + index.metadata = { + 'config': index_config.model_dump(), + 'contentsDeltaUri': contents_delta_uri, + } + + request = aiplatform_v1.CreateIndexRequest( + parent=self.index_location_path, + index=index, + ) + + operation = await self._index_client.create_index(request=request) + + return await operation.result() + + async def deploy_index(self, index_name: str, endpoint_name: str): + deployed_index = aiplatform_v1.DeployedIndex() + deployed_index.id = index_name + deployed_index.index = self.get_index_path(index_name=index_name) + + request = aiplatform_v1.DeployIndexRequest( + index_endpoint=endpoint_name, + deployed_index=deployed_index, + ) + + operation = await self._endpoint_client.deploy_index(request=request) + return operation.result() + + def upload_jsonl_file(self, local_path: str, bucket_name: str, destination_location: str) -> Operation: + bucket = self._storage_client.bucket(bucket_name=bucket_name) + blob = bucket.blob(destination_location) + blob.upload_from_filename(local_path) + + def get_index_path(self, index_name: str) -> str: + return self._index_client.index_path(project=self.project, location=self.location, index=index_name) + + @property + def index_location_path(self) -> str: + return self._index_client.common_location_path(project=self.project, location=self.location) + + def initialize(self, ai: GenkitRegistry) -> None: + """Initialize firestore plugin. + + Register actions with the registry making them available for use in the Genkit framework. + + Args: + ai: The registry to register actions with. + + Returns: + None + """ + retriever = self.retriever_cls( + ai=ai, + name=self.name, + match_service_client=self._match_service_client, + embedder=self.embedder, + embedder_options=self.embedder_options, + **self.retriever_extra_args, + ) + + return ai.define_retriever( + name=vertexai_name(self.name), + fn=retriever.retrieve, + ) + + def _inject_attribution_headers(http_options): """Adds genkit client info to the appropriate http headers.""" if not http_options: diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py new file mode 100644 index 0000000000..1287ad1fef --- /dev/null +++ b/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py @@ -0,0 +1,165 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import json +from abc import ABC, abstractmethod +from typing import Any + +import structlog +from google.cloud import aiplatform_v1, bigquery, firestore +from google.cloud.aiplatform_v1 import FindNeighborsRequest, IndexDatapoint, Neighbor +from pydantic import ValidationError + +from genkit.blocks.document import Document +from genkit.core.typing import Embedding +from genkit.types import ActionRunContext, RetrieverRequest, RetrieverResponse + +logger = structlog.get_logger(__name__) + + +class VertexAIVectorStoreRetriever(ABC): + def __init__( + self, + ai, + name: str, + match_service_client: aiplatform_v1.MatchServiceAsyncClient, + embedder: str, + embedder_options: dict[str, Any] | None = None, + ): + self.ai = ai + self.name = name + self._match_service_client = match_service_client + self.embedder = embedder + self.embedder_options = embedder_options or {} + + async def retrieve(self, request: RetrieverRequest, _: ActionRunContext) -> RetrieverResponse: + document = Document.from_document_data(document_data=request.query) + embeddings = await self.ai.embed( + embedder=self.embedder, + documents=[document], + options=self.embedder_options, + ) + if self.embedder_options: + top_k = self.embedder_options.get('limit') or 3 + else: + top_k = 3 + docs = await self._get_closest_documents( + request=request, + top_k=top_k, + query_embeddings=embeddings.embeddings[0], + ) + + return RetrieverResponse(documents=[d.document for d in docs]) + + async def _get_closest_documents( + self, request: RetrieverRequest, top_k: int, query_embeddings: Embedding + ) -> list[Document]: + metadata = request.query.metadata + if not metadata or 'index_endpoint_path' not in metadata: + raise AttributeError('Request provides no data about index endpoint path') + + index_endpoint_path = metadata['index_endpoint_path'] + deployed_index_id = metadata['deployed_index_id'] + + nn_request = FindNeighborsRequest( + index_endpoint=index_endpoint_path, + deployed_index_id=deployed_index_id, + queries=[ + FindNeighborsRequest.Query( + datapoint=IndexDatapoint(feature_vector=query_embeddings.embedding), + neighbor_count=top_k, + ) + ], + ) + + response = await self._match_service_client.find_neighbors(request=nn_request) + + return await self._retrieve_neighbours_data_from_db(neighbours=response.nearest_neighbors[0].neighbors) + + @abstractmethod + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + pass + + +class BigQueryRetriever(VertexAIVectorStoreRetriever): + def __init__(self, bq_client: bigquery.Client, dataset_id: str, table_id: str, *args, **kwargs): + super().__init__(*args, **kwargs) + self.bq_client = bq_client + self.dataset_id = dataset_id + self.table_id = table_id + + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + ids = [n.datapoint.datapoint_id for n in neighbours if n.datapoint and n.datapoint.datapoint_id] + + if not ids: + return [] + + query = f""" + SELECT * FROM `{self.dataset_id}.{self.table_id}` + WHERE id IN UNNEST(@ids) + """ + + job_config = bigquery.QueryJobConfig(query_parameters=[bigquery.ArrayQueryParameter('ids', 'STRING', ids)]) + + try: + query_job = self.bq_client.query(query, job_config=job_config) + rows = query_job.result() + except Exception as e: + await logger.aerror('Failed to execute BigQuery query: %s', e) + return [] + + documents: list[Document] = [] + + for row in rows: + try: + doc_data = { + 'content': json.loads(row['content']), + } + if row.get('metadata'): + doc_data['metadata'] = json.loads(row['metadata']) + + documents.append(Document(**doc_data)) + except (ValidationError, json.JSONDecodeError, Exception) as error: + doc_id = row.get('id', '') + await logger.awarning(f'Failed to parse document data for document with ID {doc_id}: {error}') + + return documents + + +class FirestoreRetriever(VertexAIVectorStoreRetriever): + def __init__(self, firestore_client: firestore.AsyncClient, collection_name: str, *args, **kwargs): + super().__init__(*args, **kwargs) + self.db = firestore_client + self.collection_name = collection_name + + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + documents: list[Document] = [] + + for neighbor in neighbours: + doc_ref = self.db.collection(self.collection_name).document(document_id=neighbor.datapoint.datapoint_id) + doc_snapshot = await doc_ref.get() + + if doc_snapshot.exists: + doc_data = doc_snapshot.to_dict() or {} + + try: + documents.append(Document(**doc_data)) + except ValidationError as e: + await logger.awarning( + f'Failed to parse document data for ID {neighbor.datapoint.datapoint_id}: {e}' + ) + + return documents diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py new file mode 100644 index 0000000000..b8768a75ae --- /dev/null +++ b/py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py @@ -0,0 +1,55 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import sys + +if sys.version_info < (3, 11): # noqa + from strenum import StrEnum # noqa +else: # noqa + from enum import StrEnum # noqa + +from pydantic import BaseModel, Field + + +# Defines the size of each shard in the index. +class IndexShardSize(StrEnum): + SMALL = 'SHARD_SIZE_SMALL' + MEDIUM = 'SHARD_SIZE_MEDIUM' + LARGE = 'SHARD_SIZE_LARGE' + + +# Specifies the normalization applied to feature vectors. +class FeatureNormType(StrEnum): + NONE = 'NONE' + UNIT_L2_NORMALIZED = 'UNIT_L2_NORM' + + +class DistanceMeasureType(StrEnum): + SQUARED_L2 = 'SQUARED_L2_DISTANCE' + L2 = 'L2_DISTANCE' + COSINE = 'COSINE_DISTANCE' + DOT_PRODUCT = 'DOT_PRODUCT_DISTANCE' + + +class IndexConfig(BaseModel): + dimensions: int = 128 + approximate_neighbors_count: int = Field(default=100, alias='approximateNeighborsCount') + distance_measure_type: DistanceMeasureType | str = Field( + default=DistanceMeasureType.COSINE, alias='distanceMeasureType' + ) + feature_norm_type: FeatureNormType | str = Field(default=FeatureNormType.NONE, alias='featureNormType') + shard_size: IndexShardSize | str = Field(default=IndexShardSize.MEDIUM, alias='shardSize') + algorithm_config: dict | None = Field(default=None, alias='algorithmConfig') diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE b/py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE new file mode 100644 index 0000000000..2205396735 --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright 2025 Google LLC + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/README.md b/py/samples/google-genai-vertexai-vector-search-bigquery/README.md new file mode 100644 index 0000000000..d4634407e6 --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-bigquery/README.md @@ -0,0 +1,20 @@ +# Google GenAI - Vertex AI Vector Search BigQuery + +An example demonstrating the use Vector Search API with BigQuery retriever for Google GenAI - Vertex AI + +## Setup environment + +1. Install [GCP CLI](https://cloud.google.com/sdk/docs/install). +2. Put your GCP project and location in the code to run VertexAI there. +3. Run the sample. + +```bash +uv venv +source .venv/bin/activate +``` + +## Run the sample + +```bash +genkit start -- uv run src/sample.py +``` diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml b/py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml new file mode 100644 index 0000000000..6275c87137 --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml @@ -0,0 +1,39 @@ +[project] +authors = [{ name = "Google" }] +classifiers = [ + "Development Status :: 3 - Alpha", + "Environment :: Console", + "Environment :: Web Environment", + "Intended Audience :: Developers", + "Operating System :: OS Independent", + "License :: OSI Approved :: Apache Software License", + "Programming Language :: Python", + "Programming Language :: Python :: 3 :: Only", + "Programming Language :: Python :: 3.10", + "Programming Language :: Python :: 3.11", + "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", + "Topic :: Scientific/Engineering :: Artificial Intelligence", + "Topic :: Software Development :: Libraries", +] +dependencies = [ + "genkit", + "genkit-plugin-google-genai", + "pydantic>=2.10.5", + "structlog>=25.2.0", + "google-cloud-bigquery", + "strenum>=0.4.15; python_version < '3.11'", +] +description = "An example demonstrating the use Vector Search API with BigQuery retriever for Google GenAI - Vertex AI" +license = { text = "Apache-2.0" } +name = "google-genai-vertexai-vector-search-bigquery" +readme = "README.md" +requires-python = ">=3.10" +version = "0.1.0" + +[build-system] +build-backend = "hatchling.build" +requires = ["hatchling"] + +[tool.hatch.build.targets.wheel] +packages = ["src/sample"] diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py b/py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py new file mode 100644 index 0000000000..cbf6bb6e9f --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py @@ -0,0 +1,99 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import os +import time + +from google.cloud import aiplatform, bigquery +from pydantic import BaseModel + +from genkit.ai import Genkit +from genkit.blocks.document import Document +from genkit.plugins.google_genai import VertexAI +from genkit.plugins.google_genai.google import VertexAIVectorSearch, vertexai_name +from genkit.plugins.google_genai.models.retriever import BigQueryRetriever +from genkit.plugins.vertex_ai import EmbeddingModels + +LOCATION = os.getenv('LOCATION') +PROJECT_ID = os.getenv('PROJECT_ID') +BIGQUERY_DATASET = os.getenv('BIGQUERY_DATASET') +BIGQUERY_TABLE = os.getenv('BIGQUERY_TABLE') +VECTOR_SEARCH_DEPLOYED_INDEX_ID = os.getenv('VECTOR_SEARCH_DEPLOYED_INDEX_ID') +VECTOR_SEARCH_INDEX_ENDPOINT_ID = os.getenv('VECTOR_SEARCH_INDEX_ENDPOINT_ID') +VECTOR_SEARCH_INDEX_ID = os.getenv('VECTOR_SEARCH_INDEX_ID') +VECTOR_SEARCH_PUBLIC_DOMAIN_NAME = os.getenv('VECTOR_SEARCH_PUBLIC_DOMAIN_NAME') + + +bq_client = bigquery.Client(project=PROJECT_ID) +aiplatform.init(project=PROJECT_ID, location=LOCATION) + + +ai = Genkit( + plugins=[ + VertexAI(), + VertexAIVectorSearch( + retriever=BigQueryRetriever, + retriever_extra_args={ + 'bq_client': bq_client, + 'dataset_id': BIGQUERY_DATASET, + 'table_id': BIGQUERY_TABLE, + }, + embedder=EmbeddingModels.TEXT_EMBEDDING_004_ENG, + embedder_options={'taskType': 'RETRIEVAL_DOCUMENT'}, + ), + ] +) + + +class QueryFlowInputSchema(BaseModel): + query: str + k: int + + +class QueryFlowOutputSchema(BaseModel): + result: list[dict] + length: int + time: int + + +@ai.flow(name='queryFlow') +async def query_flow(_input: QueryFlowInputSchema) -> QueryFlowOutputSchema: + start_time = time.time() + query_document = Document.from_text(text=_input.query) + + result: list[Document] = await ai.retrieve( + retriever=vertexai_name(VECTOR_SEARCH_INDEX_ID), + query=query_document, + ) + + end_time = time.time() + + duration = int(end_time - start_time) + + result_data = [] + for doc in result: + result_data.append({ + 'text': doc.content[0].root.text, + 'distance': doc.metadata.get('distance'), + }) + + result_data = sorted(result_data, key=lambda x: x['distance']) + + return QueryFlowOutputSchema( + result=result_data, + length=len(result_data), + time=duration, + ) diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/LICENSE b/py/samples/google-genai-vertexai-vector-search-firestore/LICENSE new file mode 100644 index 0000000000..2205396735 --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-firestore/LICENSE @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright 2025 Google LLC + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/README.md b/py/samples/google-genai-vertexai-vector-search-firestore/README.md new file mode 100644 index 0000000000..670ef11cdb --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-firestore/README.md @@ -0,0 +1,20 @@ +# Google GenAI - Vertex AI Vector Search Firestore + +An example demonstrating the use Vector Search API with Firestore retriever for Google GenAI - Vertex AI + +## Setup environment + +1. Install [GCP CLI](https://cloud.google.com/sdk/docs/install). +2. Put your GCP project and location in the code to run VertexAI there. +3. Run the sample. + +```bash +uv venv +source .venv/bin/activate +``` + +## Run the sample + +```bash +genkit start -- uv run src/sample.py +``` diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml b/py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml new file mode 100644 index 0000000000..020f2557f5 --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml @@ -0,0 +1,39 @@ +[project] +authors = [{ name = "Google" }] +classifiers = [ + "Development Status :: 3 - Alpha", + "Environment :: Console", + "Environment :: Web Environment", + "Intended Audience :: Developers", + "Operating System :: OS Independent", + "License :: OSI Approved :: Apache Software License", + "Programming Language :: Python", + "Programming Language :: Python :: 3 :: Only", + "Programming Language :: Python :: 3.10", + "Programming Language :: Python :: 3.11", + "Programming Language :: Python :: 3.12", + "Programming Language :: Python :: 3.13", + "Topic :: Scientific/Engineering :: Artificial Intelligence", + "Topic :: Software Development :: Libraries", +] +dependencies = [ + "genkit", + "genkit-plugin-google-genai", + "pydantic>=2.10.5", + "structlog>=25.2.0", + "google-cloud-firestore", + "strenum>=0.4.15; python_version < '3.11'", +] +description = "An example demonstrating the use Vector Search API with Firestore retriever for Google GenAI - Vertex AI" +license = { text = "Apache-2.0" } +name = "google-genai-vertexai-vector-search-firestore" +readme = "README.md" +requires-python = ">=3.10" +version = "0.1.0" + +[build-system] +build-backend = "hatchling.build" +requires = ["hatchling"] + +[tool.hatch.build.targets.wheel] +packages = ["src/sample"] diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py b/py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py new file mode 100644 index 0000000000..70634b1ade --- /dev/null +++ b/py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py @@ -0,0 +1,96 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import os +import time + +from google.cloud import aiplatform, firestore +from pydantic import BaseModel + +from genkit.ai import Genkit +from genkit.blocks.document import Document +from genkit.plugins.google_genai import VertexAI +from genkit.plugins.google_genai.google import VertexAIVectorSearch, vertexai_name +from genkit.plugins.google_genai.models.retriever import FirestoreRetriever +from genkit.plugins.vertex_ai import EmbeddingModels + +LOCATION = os.getenv('LOCATION') +PROJECT_ID = os.getenv('PROJECT_ID') +FIRESTORE_COLLECTION = os.getenv('FIRESTORE_COLLECTION') +VECTOR_SEARCH_DEPLOYED_INDEX_ID = os.getenv('VECTOR_SEARCH_DEPLOYED_INDEX_ID') +VECTOR_SEARCH_INDEX_ENDPOINT_ID = os.getenv('VECTOR_SEARCH_INDEX_ENDPOINT_ID') +VECTOR_SEARCH_INDEX_ID = os.getenv('VECTOR_SEARCH_INDEX_ID') +VECTOR_SEARCH_PUBLIC_DOMAIN_NAME = os.getenv('VECTOR_SEARCH_PUBLIC_DOMAIN_NAME') + +firestore_client = firestore.Client(project=PROJECT_ID) +aiplatform.init(project=PROJECT_ID, location=LOCATION) + + +ai = Genkit( + plugins=[ + VertexAI(), + VertexAIVectorSearch( + retriever=FirestoreRetriever, + retriever_extra_args={ + 'firestore_client': firestore_client, + 'collection_name': FIRESTORE_COLLECTION, + }, + embedder=EmbeddingModels.TEXT_EMBEDDING_004_ENG, + embedder_options={'taskType': 'RETRIEVAL_DOCUMENT'}, + ), + ] +) + + +class QueryFlowInputSchema(BaseModel): + query: str + k: int + + +class QueryFlowOutputSchema(BaseModel): + result: list[dict] + length: int + time: int + + +@ai.flow(name='queryFlow') +async def query_flow(_input: QueryFlowInputSchema) -> QueryFlowOutputSchema: + start_time = time.time() + query_document = Document.from_text(text=_input.query) + + result: list[Document] = await ai.retrieve( + retriever=vertexai_name(VECTOR_SEARCH_INDEX_ID), + query=query_document, + ) + + end_time = time.time() + + duration = int(end_time - start_time) + + result_data = [] + for doc in result: + result_data.append({ + 'text': doc.content[0].root.text, + 'distance': doc.metadata.get('distance'), + }) + + result_data = sorted(result_data, key=lambda x: x['distance']) + + return QueryFlowOutputSchema( + result=result_data, + length=len(result_data), + time=duration, + ) diff --git a/py/uv.lock b/py/uv.lock index ef8c60b640..c6e4eebf8b 100644 --- a/py/uv.lock +++ b/py/uv.lock @@ -28,6 +28,8 @@ members = [ "google-genai-image", "google-genai-vertexai-hello", "google-genai-vertexai-image", + "google-genai-vertexai-vector-search-bigquery", + "google-genai-vertexai-vector-search-firestore", "imagen", "menu", "multi-server", @@ -1523,6 +1525,52 @@ requires-dist = [ { name = "pydantic", specifier = ">=2.10.5" }, ] +[[package]] +name = "google-genai-vertexai-vector-search-bigquery" +version = "0.1.0" +source = { editable = "samples/google-genai-vertexai-vector-search-bigquery" } +dependencies = [ + { name = "genkit" }, + { name = "genkit-plugin-google-genai" }, + { name = "google-cloud-bigquery" }, + { name = "pydantic" }, + { name = "strenum", marker = "python_full_version < '3.11'" }, + { name = "structlog" }, +] + +[package.metadata] +requires-dist = [ + { name = "genkit", editable = "packages/genkit" }, + { name = "genkit-plugin-google-genai", editable = "plugins/google-genai" }, + { name = "google-cloud-bigquery" }, + { name = "pydantic", specifier = ">=2.10.5" }, + { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, + { name = "structlog", specifier = ">=25.2.0" }, +] + +[[package]] +name = "google-genai-vertexai-vector-search-firestore" +version = "0.1.0" +source = { editable = "samples/google-genai-vertexai-vector-search-firestore" } +dependencies = [ + { name = "genkit" }, + { name = "genkit-plugin-google-genai" }, + { name = "google-cloud-firestore" }, + { name = "pydantic" }, + { name = "strenum", marker = "python_full_version < '3.11'" }, + { name = "structlog" }, +] + +[package.metadata] +requires-dist = [ + { name = "genkit", editable = "packages/genkit" }, + { name = "genkit-plugin-google-genai", editable = "plugins/google-genai" }, + { name = "google-cloud-firestore" }, + { name = "pydantic", specifier = ">=2.10.5" }, + { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, + { name = "structlog", specifier = ">=25.2.0" }, +] + [[package]] name = "google-resumable-media" version = "2.7.2" From 9fcb2e1004c255acca54a80aec31f536b6329652 Mon Sep 17 00:00:00 2001 From: Abraham Lazaro Martinez Date: Fri, 25 Apr 2025 17:54:25 +0000 Subject: [PATCH 2/6] fix: move to vertexai plugin --- .../src/genkit/plugins/google_genai/google.py | 120 ------- .../plugins/google_genai/models/retriever.py | 165 --------- .../src/genkit/plugins/vertex_ai/__init__.py | 2 + .../plugins/vertex_ai/models/retriever.py | 312 ++++++++++++++++++ .../plugins/vertex_ai}/models/vectorstore.py | 6 +- .../vertex_ai/vector_search/vector_search.py | 234 +++++++++++++ 6 files changed, 552 insertions(+), 287 deletions(-) delete mode 100644 py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py create mode 100644 py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py rename py/plugins/{google-genai/src/genkit/plugins/google_genai => vertex-ai/src/genkit/plugins/vertex_ai}/models/vectorstore.py (88%) create mode 100644 py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/vector_search/vector_search.py diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py index 24f9d0daeb..d50449e71f 100644 --- a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py +++ b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py @@ -36,8 +36,6 @@ VertexAIGeminiVersion, ) from genkit.plugins.google_genai.models.imagen import ImagenModel, ImagenVersion -from genkit.plugins.google_genai.models.retriever import VertexAIVectorStoreRetriever -from genkit.plugins.google_genai.models.vectorstore import IndexConfig GOOGLEAI_PLUGIN_NAME = 'googleai' VERTEXAI_PLUGIN_NAME = 'vertexai' @@ -169,124 +167,6 @@ def initialize(self, ai: GenkitRegistry) -> None: ai.define_model(name=vertexai_name(version), fn=imagen_model.generate, metadata=imagen_model.metadata) -class VertexAIVectorSearch(Plugin): - """VertexAI vector store plugin for Genkit.""" - - name: str = 'vertexAIVectorstore' - - def __init__( - self, - retriever: Type[VertexAIVectorStoreRetriever], - retriever_extra_args: dict[str, Any] | None = None, - credentials: Credentials | None = None, - project: str | None = None, - location: str | None = 'us-central1', - embedder: str | None = None, - embedder_options: dict[str, Any] | None = None, - http_options: HttpOptions | HttpOptionsDict | None = None, - ): - http_options = _inject_attribution_headers(http_options=http_options) - - self.project = project - self.location = location - - self.embedder = embedder - self.embedder_options = embedder_options - - self.retriever_cls = retriever - self.retriever_extra_args = retriever_extra_args or {} - - self._storage_client = storage.Client( - project=self.project, - credentials=credentials, - extra_headers=http_options.headers, - ) - self._index_client = aiplatform_v1.IndexServiceAsyncClient( - credentials=credentials, - ) - self._endpoint_client = aiplatform_v1.IndexEndpointServiceAsyncClient(credentials=credentials) - self._match_service_client = aiplatform_v1.MatchServiceAsyncClient( - credentials=credentials, - ) - - async def create_index( - self, - display_name: str, - description: str | None, - index_config: IndexConfig | None = None, - contents_delta_uri: str | None = None, - ) -> None: - if not index_config: - index_config = IndexConfig() - - index = aiplatform_v1.Index() - index.display_name = display_name - index.description = description - index.metadata = { - 'config': index_config.model_dump(), - 'contentsDeltaUri': contents_delta_uri, - } - - request = aiplatform_v1.CreateIndexRequest( - parent=self.index_location_path, - index=index, - ) - - operation = await self._index_client.create_index(request=request) - - return await operation.result() - - async def deploy_index(self, index_name: str, endpoint_name: str): - deployed_index = aiplatform_v1.DeployedIndex() - deployed_index.id = index_name - deployed_index.index = self.get_index_path(index_name=index_name) - - request = aiplatform_v1.DeployIndexRequest( - index_endpoint=endpoint_name, - deployed_index=deployed_index, - ) - - operation = await self._endpoint_client.deploy_index(request=request) - return operation.result() - - def upload_jsonl_file(self, local_path: str, bucket_name: str, destination_location: str) -> Operation: - bucket = self._storage_client.bucket(bucket_name=bucket_name) - blob = bucket.blob(destination_location) - blob.upload_from_filename(local_path) - - def get_index_path(self, index_name: str) -> str: - return self._index_client.index_path(project=self.project, location=self.location, index=index_name) - - @property - def index_location_path(self) -> str: - return self._index_client.common_location_path(project=self.project, location=self.location) - - def initialize(self, ai: GenkitRegistry) -> None: - """Initialize firestore plugin. - - Register actions with the registry making them available for use in the Genkit framework. - - Args: - ai: The registry to register actions with. - - Returns: - None - """ - retriever = self.retriever_cls( - ai=ai, - name=self.name, - match_service_client=self._match_service_client, - embedder=self.embedder, - embedder_options=self.embedder_options, - **self.retriever_extra_args, - ) - - return ai.define_retriever( - name=vertexai_name(self.name), - fn=retriever.retrieve, - ) - - def _inject_attribution_headers(http_options): """Adds genkit client info to the appropriate http headers.""" if not http_options: diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py deleted file mode 100644 index 1287ad1fef..0000000000 --- a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/retriever.py +++ /dev/null @@ -1,165 +0,0 @@ -# Copyright 2025 Google LLC -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. -# -# SPDX-License-Identifier: Apache-2.0 - -import json -from abc import ABC, abstractmethod -from typing import Any - -import structlog -from google.cloud import aiplatform_v1, bigquery, firestore -from google.cloud.aiplatform_v1 import FindNeighborsRequest, IndexDatapoint, Neighbor -from pydantic import ValidationError - -from genkit.blocks.document import Document -from genkit.core.typing import Embedding -from genkit.types import ActionRunContext, RetrieverRequest, RetrieverResponse - -logger = structlog.get_logger(__name__) - - -class VertexAIVectorStoreRetriever(ABC): - def __init__( - self, - ai, - name: str, - match_service_client: aiplatform_v1.MatchServiceAsyncClient, - embedder: str, - embedder_options: dict[str, Any] | None = None, - ): - self.ai = ai - self.name = name - self._match_service_client = match_service_client - self.embedder = embedder - self.embedder_options = embedder_options or {} - - async def retrieve(self, request: RetrieverRequest, _: ActionRunContext) -> RetrieverResponse: - document = Document.from_document_data(document_data=request.query) - embeddings = await self.ai.embed( - embedder=self.embedder, - documents=[document], - options=self.embedder_options, - ) - if self.embedder_options: - top_k = self.embedder_options.get('limit') or 3 - else: - top_k = 3 - docs = await self._get_closest_documents( - request=request, - top_k=top_k, - query_embeddings=embeddings.embeddings[0], - ) - - return RetrieverResponse(documents=[d.document for d in docs]) - - async def _get_closest_documents( - self, request: RetrieverRequest, top_k: int, query_embeddings: Embedding - ) -> list[Document]: - metadata = request.query.metadata - if not metadata or 'index_endpoint_path' not in metadata: - raise AttributeError('Request provides no data about index endpoint path') - - index_endpoint_path = metadata['index_endpoint_path'] - deployed_index_id = metadata['deployed_index_id'] - - nn_request = FindNeighborsRequest( - index_endpoint=index_endpoint_path, - deployed_index_id=deployed_index_id, - queries=[ - FindNeighborsRequest.Query( - datapoint=IndexDatapoint(feature_vector=query_embeddings.embedding), - neighbor_count=top_k, - ) - ], - ) - - response = await self._match_service_client.find_neighbors(request=nn_request) - - return await self._retrieve_neighbours_data_from_db(neighbours=response.nearest_neighbors[0].neighbors) - - @abstractmethod - async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: - pass - - -class BigQueryRetriever(VertexAIVectorStoreRetriever): - def __init__(self, bq_client: bigquery.Client, dataset_id: str, table_id: str, *args, **kwargs): - super().__init__(*args, **kwargs) - self.bq_client = bq_client - self.dataset_id = dataset_id - self.table_id = table_id - - async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: - ids = [n.datapoint.datapoint_id for n in neighbours if n.datapoint and n.datapoint.datapoint_id] - - if not ids: - return [] - - query = f""" - SELECT * FROM `{self.dataset_id}.{self.table_id}` - WHERE id IN UNNEST(@ids) - """ - - job_config = bigquery.QueryJobConfig(query_parameters=[bigquery.ArrayQueryParameter('ids', 'STRING', ids)]) - - try: - query_job = self.bq_client.query(query, job_config=job_config) - rows = query_job.result() - except Exception as e: - await logger.aerror('Failed to execute BigQuery query: %s', e) - return [] - - documents: list[Document] = [] - - for row in rows: - try: - doc_data = { - 'content': json.loads(row['content']), - } - if row.get('metadata'): - doc_data['metadata'] = json.loads(row['metadata']) - - documents.append(Document(**doc_data)) - except (ValidationError, json.JSONDecodeError, Exception) as error: - doc_id = row.get('id', '') - await logger.awarning(f'Failed to parse document data for document with ID {doc_id}: {error}') - - return documents - - -class FirestoreRetriever(VertexAIVectorStoreRetriever): - def __init__(self, firestore_client: firestore.AsyncClient, collection_name: str, *args, **kwargs): - super().__init__(*args, **kwargs) - self.db = firestore_client - self.collection_name = collection_name - - async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: - documents: list[Document] = [] - - for neighbor in neighbours: - doc_ref = self.db.collection(self.collection_name).document(document_id=neighbor.datapoint.datapoint_id) - doc_snapshot = await doc_ref.get() - - if doc_snapshot.exists: - doc_data = doc_snapshot.to_dict() or {} - - try: - documents.append(Document(**doc_data)) - except ValidationError as e: - await logger.awarning( - f'Failed to parse document data for ID {neighbor.datapoint.datapoint_id}: {e}' - ) - - return documents diff --git a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/__init__.py b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/__init__.py index c0ac5edf03..c635d21132 100644 --- a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/__init__.py +++ b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/__init__.py @@ -26,6 +26,7 @@ from genkit.plugins.vertex_ai.gemini import GeminiVersion from genkit.plugins.vertex_ai.imagen import ImagenOptions, ImagenVersion from genkit.plugins.vertex_ai.plugin_api import VertexAI, vertexai_name +from genkit.plugins.vertex_ai.vector_search.vector_search import VertexAIVectorSearch def package_name() -> str: @@ -46,4 +47,5 @@ def package_name() -> str: GeminiVersion.__name__, ImagenVersion.__name__, ImagenOptions.__name__, + VertexAIVectorSearch.__name__, ] diff --git a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py new file mode 100644 index 0000000000..5f9ffd11bf --- /dev/null +++ b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py @@ -0,0 +1,312 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import json +from abc import ABC, abstractmethod +from typing import Any, BaseModel, Field + +import structlog +from google.cloud import aiplatform_v1, bigquery, firestore +from google.cloud.aiplatform_v1 import FindNeighborsRequest, IndexDatapoint, Neighbor +from pydantic import ValidationError + +from genkit.ai import Genkit +from genkit.blocks.document import Document +from genkit.core.typing import Embedding +from genkit.types import ActionRunContext, RetrieverRequest, RetrieverResponse + +logger = structlog.get_logger(__name__) + + +class DocRetriver(ABC): + """Abstract base class for Vertex AI Vector Search document retrieval. + + This class outlines the core workflow for retrieving relevant documents. + It is not intended to be instantiated directly. Subclasses must implement + the abstract methods to provide concrete retrieval logic depending of the + technology used. + + Attributes: + ai: The Genkit instance. + name: The name of this retriever instance. + match_service_client: The Vertex AI Matching Engine client. + embedder: The name of the embedder to use for generating embeddings. + embedder_options: Options to pass to the embedder. + """ + def __init__( + self, + ai: Genkit, + name: str, + match_service_client: aiplatform_v1.MatchServiceAsyncClient, + embedder: str, + embedder_options: dict[str, Any] | None = None, + ) -> None: + """Initializes the DocRetriever. + + Args: + ai: The Genkit application instance. + name: The name of this retriever instance. + match_service_client: The Vertex AI Matching Engine client. + embedder: The name of the embedder to use for generating embeddings. + embedder_options: Optional dictionary of options to pass to the embedder. + """ + self.ai = ai + self.name = name + self._match_service_client = match_service_client + self.embedder = embedder + self.embedder_options = embedder_options or {} + + async def retrieve(self, request: RetrieverRequest, _: ActionRunContext) -> RetrieverResponse: + """Retrieves documents based on a given query. + + Args: + request: The retrieval request containing the query. + _: The ActionRunContext (unused in this method). + + Returns: + A RetrieverResponse object containing the retrieved documents. + """ + document = Document.from_document_data(document_data=request.query) + + embeddings = await self.ai.embed( + embedder=self.embedder, + documents=[document], + options=self.embedder_options, + ) + + if self.embedder_options: + top_k = self.embedder_options.get('limit') or 3 + else: + top_k = 3 + + docs = await self._get_closest_documents( + request=request, + top_k=top_k, + query_embeddings=embeddings.embeddings[0], + ) + + return RetrieverResponse(documents=[d.document for d in docs]) + + async def _get_closest_documents( + self, request: RetrieverRequest, top_k: int, query_embeddings: Embedding + ) -> list[Document]: + """Retrieves the closest documents from the vector search index based on query embeddings. + + Args: + request: The retrieval request containing the query and metadata. + top_k: The number of nearest neighbors to retrieve. + query_embeddings: The embedding of the query. + + Returns: + A list of Document objects representing the closest documents. + + Raises: + AttributeError: If the request does not contain the necessary + index endpoint path in its metadata. + """ + metadata = request.query.metadata + if not metadata or 'index_endpoint_path' not in metadata: + raise AttributeError('Request provides no data about index endpoint path') + + index_endpoint_path = metadata['index_endpoint_path'] + deployed_index_id = metadata['deployed_index_id'] + + nn_request = FindNeighborsRequest( + index_endpoint=index_endpoint_path, + deployed_index_id=deployed_index_id, + queries=[ + FindNeighborsRequest.Query( + datapoint=IndexDatapoint(feature_vector=query_embeddings.embedding), + neighbor_count=top_k, + ) + ], + ) + + response = await self._match_service_client.find_neighbors(request=nn_request) + + return await self._retrieve_neighbours_data_from_db(neighbours=response.nearest_neighbors[0].neighbors) + + @abstractmethod + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + """Retrieves document data from the database based on neighbor information. + + This method must be implemented by subclasses to define how document + data is fetched from the database using the provided neighbor information. + + Args: + neighbours: A list of Neighbor objects representing the nearest neighbors + found in the vector search index. + + Returns: + A list of Document objects containing the data for the retrieved documents. + """ + raise NotImplementedError + + +class BigQueryRetriever(DocRetriver): + """Retrieves documents from a BigQuery table. + + This class extends DocRetriever to fetch document data from a specified BigQuery + dataset and table. It constructs a query to retrieve documents based on the IDs + obtained from nearest neighbor search results. + + Attributes: + bq_client: The BigQuery client to use for querying. + dataset_id: The ID of the BigQuery dataset. + table_id: The ID of the BigQuery table. + """ + def __init__( + self, bq_client: bigquery.Client, dataset_id: str, table_id: str, *args, **kwargs, + ) -> None: + """Initializes the BigQueryRetriever. + + Args: + bq_client: The BigQuery client to use for querying. + dataset_id: The ID of the BigQuery dataset. + table_id: The ID of the BigQuery table. + *args: Additional positional arguments to pass to the parent class. + **kwargs: Additional keyword arguments to pass to the parent class. + """ + super().__init__(*args, **kwargs) + self.bq_client = bq_client + self.dataset_id = dataset_id + self.table_id = table_id + + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + """Retrieves document data from the BigQuery table for the given neighbors. + + Constructs and executes a BigQuery query to fetch document data based on + the IDs obtained. Handles potential errors during query execution and + document parsing. + + Args: + neighbours: A list of Neighbor objects representing the nearest neighbors. + Each neighbor should contain a datapoint with a datapoint_id. + + Returns: + A list of Document objects containing the retrieved document data. + Returns an empty list if no IDs are found in the neighbors or if the + query fails. + """ + ids = [ + n.datapoint.datapoint_id + for n in neighbours + if n.datapoint and n.datapoint.datapoint_id + ] + + if not ids: + return [] + + query = f""" + SELECT * FROM `{self.dataset_id}.{self.table_id}` + WHERE id IN UNNEST(@ids) + """ + + job_config = bigquery.QueryJobConfig( + query_parameters=[bigquery.ArrayQueryParameter('ids', 'STRING', ids)], + ) + + try: + query_job = self.bq_client.query(query, job_config=job_config) + rows = query_job.result() + except Exception as e: + await logger.aerror('Failed to execute BigQuery query: %s', e) + return [] + + documents: list[Document] = [] + + for row in rows: + try: + doc_data = { + 'content': json.loads(row['content']), + } + if row.get('metadata'): + doc_data['metadata'] = json.loads(row['metadata']) + + documents.append(Document(**doc_data)) + except (ValidationError, json.JSONDecodeError, Exception) as error: + doc_id = row.get('id', '') + await logger.awarning(f'Failed to parse document data for document with ID {doc_id}: {error}') + + return documents + + +class FirestoreRetriever(DocRetriver): + """Retrieves documents from a Firestore collection. + + This class extends DocRetriever to fetch document data from a specified Firestore + collection. It retrieves documents based on IDs obtained from nearest neighbor + search results. + + Attributes: + db: The Firestore client. + collection_name: The name of the Firestore collection. + """ + def __init__( + self, firestore_client: firestore.AsyncClient, collection_name: str, *args, **kwargs, + ) -> None: + """Initializes the FirestoreRetriever. + + Args: + firestore_client: The Firestore client to use for querying. + collection_name: The name of the Firestore collection. + *args: Additional positional arguments to pass to the parent class. + **kwargs: Additional keyword arguments to pass to the parent class. + """ + super().__init__(*args, **kwargs) + self.db = firestore_client + self.collection_name = collection_name + + async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> list[Document]: + """Retrieves document data from the Firestore collection for the given neighbors. + + Fetches document data from Firestore based on the IDs of the nearest neighbors. + Handles potential errors during document retrieval and data parsing. + + Args: + neighbours: A list of Neighbor objects representing the nearest neighbors. + Each neighbor should contain a datapoint with a datapoint_id. + + Returns: + A list of Document objects containing the retrieved document data. + Returns an empty list if no documents are found for the given IDs. + """ + documents: list[Document] = [] + + for neighbor in neighbours: + doc_ref = self.db.collection(self.collection_name).document(document_id=neighbor.datapoint.datapoint_id) + doc_snapshot = await doc_ref.get() + + if doc_snapshot.exists: + doc_data = doc_snapshot.to_dict() or {} + + try: + documents.append(Document(**doc_data)) + except ValidationError as e: + await logger.awarning( + f'Failed to parse document data for ID {neighbor.datapoint.datapoint_id}: {e}' + ) + + return documents + + +class RetrieverOptionsSchema(BaseModel): + """Schema for retriver options. + + Attributes: + limit: Number of documents to retrieve. + """ + limit: int | None = Field(title='Number of documents to retrieve', default=None) diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/vectorstore.py similarity index 88% rename from py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py rename to py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/vectorstore.py index b8768a75ae..6be441f7a3 100644 --- a/py/plugins/google-genai/src/genkit/plugins/google_genai/models/vectorstore.py +++ b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/vectorstore.py @@ -24,20 +24,21 @@ from pydantic import BaseModel, Field -# Defines the size of each shard in the index. class IndexShardSize(StrEnum): + """Defines the size of each shard in the index.""" SMALL = 'SHARD_SIZE_SMALL' MEDIUM = 'SHARD_SIZE_MEDIUM' LARGE = 'SHARD_SIZE_LARGE' -# Specifies the normalization applied to feature vectors. class FeatureNormType(StrEnum): + """Specifies the normalization applied to feature vectors.""" NONE = 'NONE' UNIT_L2_NORMALIZED = 'UNIT_L2_NORM' class DistanceMeasureType(StrEnum): + """Defines the available distance measure methods.""" SQUARED_L2 = 'SQUARED_L2_DISTANCE' L2 = 'L2_DISTANCE' COSINE = 'COSINE_DISTANCE' @@ -45,6 +46,7 @@ class DistanceMeasureType(StrEnum): class IndexConfig(BaseModel): + """Defines the configurations of indexes.""" dimensions: int = 128 approximate_neighbors_count: int = Field(default=100, alias='approximateNeighborsCount') distance_measure_type: DistanceMeasureType | str = Field( diff --git a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/vector_search/vector_search.py b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/vector_search/vector_search.py new file mode 100644 index 0000000000..006d37bb0f --- /dev/null +++ b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/vector_search/vector_search.py @@ -0,0 +1,234 @@ +# Copyright 2025 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# +# SPDX-License-Identifier: Apache-2.0 + +import os +from typing import Any + +import structlog +from google.auth.credentials import Credentials +from google.cloud import aiplatform_v1, storage +from google.genai.types import HttpOptions, HttpOptionsDict, Operation + +from genkit.ai import GENKIT_CLIENT_HEADER, GenkitRegistry, Plugin +from genkit.plugins.vertex_ai import vertexai_name +from genkit.plugins.vertex_ai.models.retriever import ( + DocRetriever, + RetrieverOptionsSchema, +) +from genkit.plugins.vertex_ai.models.vectorstore import IndexConfig + +logger = structlog.get_logger(__name__) + + +class VertexAIVectorSearch(Plugin): + """A plugin for integrating VertexAI Vector Search. + + This class registers VertexAI Vector Stores within a registry, + and allows interaction to retrieve similar documents. + """ + + name: str = 'vertexAIVectorSearch' + + def __init__( + self, + retriever: DocRetriever, + retriever_extra_args: dict[str, Any] | None = None, + credentials: Credentials | None = None, + project: str | None = None, + location: str | None = 'us-central1', + embedder: str | None = None, + embedder_options: dict[str, Any] | None = None, + http_options: HttpOptions | HttpOptionsDict | None = None, + ) -> None: + """Initializes the VertexAIVectorSearch plugin. + + Args: + retriever: The DocRetriever class to use for retrieving documents. + retriever_extra_args: Optional dictionary of extra arguments to pass to the + retriever's constructor. + credentials: Optional Google Cloud credentials to use. If not provided, + the default application credentials will be used. + project: Optional Google Cloud project ID. If not provided, it will be + inferred from the credentials. + location: Optional Google Cloud location (region). Defaults to + 'us-central1'. + embedder: Optional identifier for the embedding model to use. + embedder_options: Optional dictionary of options to pass to the embedding + model. + http_options: Optional HTTP options for API requests. + """ + http_options = _inject_attribution_headers(http_options=http_options) + + self.project = project + self.location = location + + self.embedder = embedder + self.embedder_options = embedder_options + + self.retriever_cls = retriever + self.retriever_extra_args = retriever_extra_args or {} + + self._storage_client = storage.Client( + project=self.project, + credentials=credentials, + extra_headers=http_options.headers, + ) + self._index_client = aiplatform_v1.IndexServiceAsyncClient( + credentials=credentials, + ) + self._endpoint_client = aiplatform_v1.IndexEndpointServiceAsyncClient(credentials=credentials) + self._match_service_client = aiplatform_v1.MatchServiceAsyncClient( + credentials=credentials, + ) + + async def create_index( + self, + display_name: str, + description: str | None, + index_config: IndexConfig | None = None, + contents_delta_uri: str | None = None, + ) -> None: + """Creates a Vertex AI Vector Search index. + + Args: + display_name: The display name for the index. + description: Optional description of the index. + index_config: Optional configuration for the index. If not provided, a + default configuration is used. + contents_delta_uri: Optional URI of the Cloud Storage location for the + contents delta. + """ + if not index_config: + index_config = IndexConfig() + + index = aiplatform_v1.Index() + index.display_name = display_name + index.description = description + index.metadata = { + 'config': index_config.model_dump(), + 'contentsDeltaUri': contents_delta_uri, + } + + request = aiplatform_v1.CreateIndexRequest( + parent=self.index_location_path, + index=index, + ) + + operation = await self._index_client.create_index(request=request) + + logger.debug(await operation.result()) + + async def deploy_index(self, index_name: str, endpoint_name: str) -> None: + """Deploys an index to an endpoint. + + Args: + index_name: The name of the index to deploy. + endpoint_name: The name of the endpoint to deploy the index to. + """ + deployed_index = aiplatform_v1.DeployedIndex() + deployed_index.id = index_name + deployed_index.index = self.get_index_path(index_name=index_name) + + request = aiplatform_v1.DeployIndexRequest( + index_endpoint=endpoint_name, + deployed_index=deployed_index, + ) + + operation = self._endpoint_client.deploy_index(request=request) + + logger.debug(await operation.result()) + + def upload_jsonl_file(self, local_path: str, bucket_name: str, destination_location: str) -> Operation: + """Uploads a JSONL file to Cloud Storage. + + Args: + local_path: The local path to the JSONL file. + bucket_name: The name of the Cloud Storage bucket. + destination_location: The destination path within the bucket. + + Returns: + The upload operation. + """ + bucket = self._storage_client.bucket(bucket_name=bucket_name) + blob = bucket.blob(destination_location) + blob.upload_from_filename(local_path) + + def get_index_path(self, index_name: str) -> str: + """Gets the full resource path of an index. + + Args: + index_name: The name of the index. + + Returns: + The full resource path of the index. + """ + return self._index_client.index_path(project=self.project, location=self.location, index=index_name) + + @property + def index_location_path(self) -> str: + """Gets the resource path of the index location. + + Returns: + The resource path of the index location. + """ + return self._index_client.common_location_path(project=self.project, location=self.location) + + def initialize(self, ai: GenkitRegistry) -> None: + """Initialize plugin with the retriver specified. + + Register actions with the registry making them available for use in the Genkit framework. + + Args: + ai: The registry to register actions with. + """ + retriever = self.retriever_cls( + ai=ai, + name=self.name, + match_service_client=self._match_service_client, + embedder=self.embedder, + embedder_options=self.embedder_options, + **self.retriever_extra_args, + ) + + return ai.define_retriever( + name=vertexai_name(self.name), + config_schema=RetrieverOptionsSchema, + fn=retriever.retrieve, + ) + + +def _inject_attribution_headers(http_options) -> HttpOptions: + """Adds genkit client info to the appropriate http headers.""" + if not http_options: + http_options = HttpOptions() + else: + if isinstance(http_options, dict): + http_options = HttpOptions(**http_options) + + if not http_options.headers: + http_options.headers = {} + + if 'x-goog-api-client' not in http_options.headers: + http_options.headers['x-goog-api-client'] = GENKIT_CLIENT_HEADER + else: + http_options.headers['x-goog-api-client'] += f' {GENKIT_CLIENT_HEADER}' + + if 'user-agent' not in http_options.headers: + http_options.headers['user-agent'] = GENKIT_CLIENT_HEADER + else: + http_options.headers['user-agent'] += f' {GENKIT_CLIENT_HEADER}' + + return http_options From 77a29f755dc369351cfe84935cceda60694000b2 Mon Sep 17 00:00:00 2001 From: Abraham Lazaro Martinez Date: Fri, 25 Apr 2025 17:55:12 +0000 Subject: [PATCH 3/6] finish removing code --- .../google-genai/src/genkit/plugins/google_genai/google.py | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py index d50449e71f..fd76e802ec 100644 --- a/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py +++ b/py/plugins/google-genai/src/genkit/plugins/google_genai/google.py @@ -15,13 +15,11 @@ # SPDX-License-Identifier: Apache-2.0 import os -from typing import Any, Type from google import genai from google.auth.credentials import Credentials -from google.cloud import aiplatform_v1, storage from google.genai.client import DebugConfig -from google.genai.types import HttpOptions, HttpOptionsDict, Operation +from google.genai.types import HttpOptions, HttpOptionsDict from genkit.ai import GENKIT_CLIENT_HEADER, GenkitRegistry, Plugin from genkit.plugins.google_genai.models.embedder import ( From 6639cd8ff0761de8d70443e6c09b7662b4d4d2fa Mon Sep 17 00:00:00 2001 From: Abraham Lazaro Martinez Date: Fri, 25 Apr 2025 18:04:59 +0000 Subject: [PATCH 4/6] fix: typo --- .../src/genkit/plugins/vertex_ai/models/retriever.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py index 5f9ffd11bf..8828749359 100644 --- a/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py +++ b/py/plugins/vertex-ai/src/genkit/plugins/vertex_ai/models/retriever.py @@ -16,12 +16,12 @@ import json from abc import ABC, abstractmethod -from typing import Any, BaseModel, Field +from typing import Any import structlog from google.cloud import aiplatform_v1, bigquery, firestore from google.cloud.aiplatform_v1 import FindNeighborsRequest, IndexDatapoint, Neighbor -from pydantic import ValidationError +from pydantic import BaseModel, Field, ValidationError from genkit.ai import Genkit from genkit.blocks.document import Document @@ -31,7 +31,7 @@ logger = structlog.get_logger(__name__) -class DocRetriver(ABC): +class DocRetriever(ABC): """Abstract base class for Vertex AI Vector Search document retrieval. This class outlines the core workflow for retrieving relevant documents. @@ -156,7 +156,7 @@ async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> raise NotImplementedError -class BigQueryRetriever(DocRetriver): +class BigQueryRetriever(DocRetriever): """Retrieves documents from a BigQuery table. This class extends DocRetriever to fetch document data from a specified BigQuery @@ -244,7 +244,7 @@ async def _retrieve_neighbours_data_from_db(self, neighbours: list[Neighbor]) -> return documents -class FirestoreRetriever(DocRetriver): +class FirestoreRetriever(DocRetriever): """Retrieves documents from a Firestore collection. This class extends DocRetriever to fetch document data from a specified Firestore From dcf7cbdbec2cc5ea06e5ab98a1486ff88bb25fb7 Mon Sep 17 00:00:00 2001 From: Abraham Lazaro Martinez Date: Fri, 25 Apr 2025 18:25:07 +0000 Subject: [PATCH 5/6] fix: samples of vector store --- .../LICENSE | 0 .../README.md | 12 +-- .../pyproject.toml | 6 +- .../src/sample.py | 9 +- .../LICENSE | 0 .../README.md | 12 +-- .../pyproject.toml | 6 +- .../src/sample.py | 9 +- py/uv.lock | 96 +++++++++---------- 9 files changed, 76 insertions(+), 74 deletions(-) rename py/samples/{google-genai-vertexai-vector-search-bigquery => vertex-ai-vector-search-bigquery}/LICENSE (100%) rename py/samples/{google-genai-vertexai-vector-search-bigquery => vertex-ai-vector-search-bigquery}/README.md (56%) rename py/samples/{google-genai-vertexai-vector-search-bigquery => vertex-ai-vector-search-bigquery}/pyproject.toml (88%) rename py/samples/{google-genai-vertexai-vector-search-bigquery => vertex-ai-vector-search-bigquery}/src/sample.py (93%) rename py/samples/{google-genai-vertexai-vector-search-firestore => vertex-ai-vector-search-firestore}/LICENSE (100%) rename py/samples/{google-genai-vertexai-vector-search-firestore => vertex-ai-vector-search-firestore}/README.md (55%) rename py/samples/{google-genai-vertexai-vector-search-firestore => vertex-ai-vector-search-firestore}/pyproject.toml (88%) rename py/samples/{google-genai-vertexai-vector-search-firestore => vertex-ai-vector-search-firestore}/src/sample.py (92%) diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE b/py/samples/vertex-ai-vector-search-bigquery/LICENSE similarity index 100% rename from py/samples/google-genai-vertexai-vector-search-bigquery/LICENSE rename to py/samples/vertex-ai-vector-search-bigquery/LICENSE diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/README.md b/py/samples/vertex-ai-vector-search-bigquery/README.md similarity index 56% rename from py/samples/google-genai-vertexai-vector-search-bigquery/README.md rename to py/samples/vertex-ai-vector-search-bigquery/README.md index d4634407e6..aacaddedf0 100644 --- a/py/samples/google-genai-vertexai-vector-search-bigquery/README.md +++ b/py/samples/vertex-ai-vector-search-bigquery/README.md @@ -1,17 +1,15 @@ -# Google GenAI - Vertex AI Vector Search BigQuery +# Vertex AI - Vector Search BigQuery -An example demonstrating the use Vector Search API with BigQuery retriever for Google GenAI - Vertex AI +An example demonstrating the use Vector Search API with BigQuery retriever for Vertex AI ## Setup environment 1. Install [GCP CLI](https://cloud.google.com/sdk/docs/install). -2. Put your GCP project and location in the code to run VertexAI there. -3. Run the sample. - +2. Run the following code to connect to VertexAI. ```bash -uv venv -source .venv/bin/activate +gcloud auth application-default login` ``` +3. Run the sample. ## Run the sample diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml b/py/samples/vertex-ai-vector-search-bigquery/pyproject.toml similarity index 88% rename from py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml rename to py/samples/vertex-ai-vector-search-bigquery/pyproject.toml index 6275c87137..7eae7e480c 100644 --- a/py/samples/google-genai-vertexai-vector-search-bigquery/pyproject.toml +++ b/py/samples/vertex-ai-vector-search-bigquery/pyproject.toml @@ -18,15 +18,15 @@ classifiers = [ ] dependencies = [ "genkit", - "genkit-plugin-google-genai", + "genkit-plugin-vertex-ai", "pydantic>=2.10.5", "structlog>=25.2.0", "google-cloud-bigquery", "strenum>=0.4.15; python_version < '3.11'", ] -description = "An example demonstrating the use Vector Search API with BigQuery retriever for Google GenAI - Vertex AI" +description = "An example demonstrating the use Vector Search API with BigQuery retriever for Vertex AI" license = { text = "Apache-2.0" } -name = "google-genai-vertexai-vector-search-bigquery" +name = "vertex-ai-vector-search-bigquery" readme = "README.md" requires-python = ">=3.10" version = "0.1.0" diff --git a/py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py b/py/samples/vertex-ai-vector-search-bigquery/src/sample.py similarity index 93% rename from py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py rename to py/samples/vertex-ai-vector-search-bigquery/src/sample.py index cbf6bb6e9f..0a6b5cfdf4 100644 --- a/py/samples/google-genai-vertexai-vector-search-bigquery/src/sample.py +++ b/py/samples/vertex-ai-vector-search-bigquery/src/sample.py @@ -22,9 +22,12 @@ from genkit.ai import Genkit from genkit.blocks.document import Document -from genkit.plugins.google_genai import VertexAI -from genkit.plugins.google_genai.google import VertexAIVectorSearch, vertexai_name -from genkit.plugins.google_genai.models.retriever import BigQueryRetriever +from genkit.plugins.vertex_ai import ( + VertexAI, + VertexAIVectorSearch, + vertexai_name, +) +from genkit.plugins.vertex_ai.models.retriever import BigQueryRetriever from genkit.plugins.vertex_ai import EmbeddingModels LOCATION = os.getenv('LOCATION') diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/LICENSE b/py/samples/vertex-ai-vector-search-firestore/LICENSE similarity index 100% rename from py/samples/google-genai-vertexai-vector-search-firestore/LICENSE rename to py/samples/vertex-ai-vector-search-firestore/LICENSE diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/README.md b/py/samples/vertex-ai-vector-search-firestore/README.md similarity index 55% rename from py/samples/google-genai-vertexai-vector-search-firestore/README.md rename to py/samples/vertex-ai-vector-search-firestore/README.md index 670ef11cdb..5f36adfa17 100644 --- a/py/samples/google-genai-vertexai-vector-search-firestore/README.md +++ b/py/samples/vertex-ai-vector-search-firestore/README.md @@ -1,17 +1,15 @@ -# Google GenAI - Vertex AI Vector Search Firestore +# Vertex AI Vector Search Firestore -An example demonstrating the use Vector Search API with Firestore retriever for Google GenAI - Vertex AI +An example demonstrating the use Vector Search API with Firestore retriever for Vertex AI ## Setup environment 1. Install [GCP CLI](https://cloud.google.com/sdk/docs/install). -2. Put your GCP project and location in the code to run VertexAI there. -3. Run the sample. - +2. Run the following code to connect to VertexAI. ```bash -uv venv -source .venv/bin/activate +gcloud auth application-default login` ``` +3. Run the sample. ## Run the sample diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml b/py/samples/vertex-ai-vector-search-firestore/pyproject.toml similarity index 88% rename from py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml rename to py/samples/vertex-ai-vector-search-firestore/pyproject.toml index 020f2557f5..3413399903 100644 --- a/py/samples/google-genai-vertexai-vector-search-firestore/pyproject.toml +++ b/py/samples/vertex-ai-vector-search-firestore/pyproject.toml @@ -18,15 +18,15 @@ classifiers = [ ] dependencies = [ "genkit", - "genkit-plugin-google-genai", + "genkit-plugin-vertex-ai", "pydantic>=2.10.5", "structlog>=25.2.0", "google-cloud-firestore", "strenum>=0.4.15; python_version < '3.11'", ] -description = "An example demonstrating the use Vector Search API with Firestore retriever for Google GenAI - Vertex AI" +description = "An example demonstrating the use Vector Search API with Firestore retriever for Vertex AI" license = { text = "Apache-2.0" } -name = "google-genai-vertexai-vector-search-firestore" +name = "vertex-ai-vector-search-firestore" readme = "README.md" requires-python = ">=3.10" version = "0.1.0" diff --git a/py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py b/py/samples/vertex-ai-vector-search-firestore/src/sample.py similarity index 92% rename from py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py rename to py/samples/vertex-ai-vector-search-firestore/src/sample.py index 70634b1ade..516deaf693 100644 --- a/py/samples/google-genai-vertexai-vector-search-firestore/src/sample.py +++ b/py/samples/vertex-ai-vector-search-firestore/src/sample.py @@ -22,9 +22,12 @@ from genkit.ai import Genkit from genkit.blocks.document import Document -from genkit.plugins.google_genai import VertexAI -from genkit.plugins.google_genai.google import VertexAIVectorSearch, vertexai_name -from genkit.plugins.google_genai.models.retriever import FirestoreRetriever +from genkit.plugins.vertex_ai import ( + VertexAI, + VertexAIVectorSearch, + vertexai_name, +) +from genkit.plugins.vertex_ai.models.retriever import FirestoreRetriever from genkit.plugins.vertex_ai import EmbeddingModels LOCATION = os.getenv('LOCATION') diff --git a/py/uv.lock b/py/uv.lock index c6e4eebf8b..5657ef7797 100644 --- a/py/uv.lock +++ b/py/uv.lock @@ -28,8 +28,6 @@ members = [ "google-genai-image", "google-genai-vertexai-hello", "google-genai-vertexai-image", - "google-genai-vertexai-vector-search-bigquery", - "google-genai-vertexai-vector-search-firestore", "imagen", "menu", "multi-server", @@ -37,6 +35,8 @@ members = [ "ollama-simple-embed", "short-n-long", "tool-interrupts", + "vertex-ai-vector-search-bigquery", + "vertex-ai-vector-search-firestore", ] [[package]] @@ -1525,52 +1525,6 @@ requires-dist = [ { name = "pydantic", specifier = ">=2.10.5" }, ] -[[package]] -name = "google-genai-vertexai-vector-search-bigquery" -version = "0.1.0" -source = { editable = "samples/google-genai-vertexai-vector-search-bigquery" } -dependencies = [ - { name = "genkit" }, - { name = "genkit-plugin-google-genai" }, - { name = "google-cloud-bigquery" }, - { name = "pydantic" }, - { name = "strenum", marker = "python_full_version < '3.11'" }, - { name = "structlog" }, -] - -[package.metadata] -requires-dist = [ - { name = "genkit", editable = "packages/genkit" }, - { name = "genkit-plugin-google-genai", editable = "plugins/google-genai" }, - { name = "google-cloud-bigquery" }, - { name = "pydantic", specifier = ">=2.10.5" }, - { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, - { name = "structlog", specifier = ">=25.2.0" }, -] - -[[package]] -name = "google-genai-vertexai-vector-search-firestore" -version = "0.1.0" -source = { editable = "samples/google-genai-vertexai-vector-search-firestore" } -dependencies = [ - { name = "genkit" }, - { name = "genkit-plugin-google-genai" }, - { name = "google-cloud-firestore" }, - { name = "pydantic" }, - { name = "strenum", marker = "python_full_version < '3.11'" }, - { name = "structlog" }, -] - -[package.metadata] -requires-dist = [ - { name = "genkit", editable = "packages/genkit" }, - { name = "genkit-plugin-google-genai", editable = "plugins/google-genai" }, - { name = "google-cloud-firestore" }, - { name = "pydantic", specifier = ">=2.10.5" }, - { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, - { name = "structlog", specifier = ">=25.2.0" }, -] - [[package]] name = "google-resumable-media" version = "2.7.2" @@ -4537,6 +4491,52 @@ wheels = [ { url = "https://files.pythonhosted.org/packages/63/9a/0962b05b308494e3202d3f794a6e85abe471fe3cafdbcf95c2e8c713aabd/uvloop-0.21.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:a5c39f217ab3c663dc699c04cbd50c13813e31d917642d459fdcec07555cc553", size = 4660018 }, ] +[[package]] +name = "vertex-ai-vector-search-bigquery" +version = "0.1.0" +source = { editable = "samples/vertex-ai-vector-search-bigquery" } +dependencies = [ + { name = "genkit" }, + { name = "genkit-plugin-vertex-ai" }, + { name = "google-cloud-bigquery" }, + { name = "pydantic" }, + { name = "strenum", marker = "python_full_version < '3.11'" }, + { name = "structlog" }, +] + +[package.metadata] +requires-dist = [ + { name = "genkit", editable = "packages/genkit" }, + { name = "genkit-plugin-vertex-ai", editable = "plugins/vertex-ai" }, + { name = "google-cloud-bigquery" }, + { name = "pydantic", specifier = ">=2.10.5" }, + { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, + { name = "structlog", specifier = ">=25.2.0" }, +] + +[[package]] +name = "vertex-ai-vector-search-firestore" +version = "0.1.0" +source = { editable = "samples/vertex-ai-vector-search-firestore" } +dependencies = [ + { name = "genkit" }, + { name = "genkit-plugin-vertex-ai" }, + { name = "google-cloud-firestore" }, + { name = "pydantic" }, + { name = "strenum", marker = "python_full_version < '3.11'" }, + { name = "structlog" }, +] + +[package.metadata] +requires-dist = [ + { name = "genkit", editable = "packages/genkit" }, + { name = "genkit-plugin-vertex-ai", editable = "plugins/vertex-ai" }, + { name = "google-cloud-firestore" }, + { name = "pydantic", specifier = ">=2.10.5" }, + { name = "strenum", marker = "python_full_version < '3.11'", specifier = ">=0.4.15" }, + { name = "structlog", specifier = ">=25.2.0" }, +] + [[package]] name = "virtualenv" version = "20.30.0" From 168cc547ee56ce91a498ddc6724b2979d0508f79 Mon Sep 17 00:00:00 2001 From: Abraham Lazaro Martinez Date: Fri, 25 Apr 2025 18:32:12 +0000 Subject: [PATCH 6/6] fix: update doc --- .../vertex-ai-vector-search-bigquery/README.md | 13 ++++++++++++- .../vertex-ai-vector-search-bigquery/src/sample.py | 2 +- .../vertex-ai-vector-search-firestore/README.md | 12 +++++++++++- .../vertex-ai-vector-search-firestore/src/sample.py | 2 +- 4 files changed, 25 insertions(+), 4 deletions(-) diff --git a/py/samples/vertex-ai-vector-search-bigquery/README.md b/py/samples/vertex-ai-vector-search-bigquery/README.md index aacaddedf0..47d03ab707 100644 --- a/py/samples/vertex-ai-vector-search-bigquery/README.md +++ b/py/samples/vertex-ai-vector-search-bigquery/README.md @@ -9,7 +9,18 @@ An example demonstrating the use Vector Search API with BigQuery retriever for V ```bash gcloud auth application-default login` ``` -3. Run the sample. +3. Set the following env vars to run the sample +``` +export LOCATION='' +export PROJECT_ID='' +export BIGQUERY_DATASET='' +export BIGQUERY_TABLE='' +export VECTOR_SEARCH_DEPLOYED_INDEX_ID='' +export VECTOR_SEARCH_INDEX_ENDPOINT_ID='' +export VECTOR_SEARCH_INDEX_ID='' +export VECTOR_SEARCH_PUBLIC_DOMAIN_NAME='' +``` +4. Run the sample. ## Run the sample diff --git a/py/samples/vertex-ai-vector-search-bigquery/src/sample.py b/py/samples/vertex-ai-vector-search-bigquery/src/sample.py index 0a6b5cfdf4..71b7713496 100644 --- a/py/samples/vertex-ai-vector-search-bigquery/src/sample.py +++ b/py/samples/vertex-ai-vector-search-bigquery/src/sample.py @@ -23,12 +23,12 @@ from genkit.ai import Genkit from genkit.blocks.document import Document from genkit.plugins.vertex_ai import ( + EmbeddingModels, VertexAI, VertexAIVectorSearch, vertexai_name, ) from genkit.plugins.vertex_ai.models.retriever import BigQueryRetriever -from genkit.plugins.vertex_ai import EmbeddingModels LOCATION = os.getenv('LOCATION') PROJECT_ID = os.getenv('PROJECT_ID') diff --git a/py/samples/vertex-ai-vector-search-firestore/README.md b/py/samples/vertex-ai-vector-search-firestore/README.md index 5f36adfa17..69299cc371 100644 --- a/py/samples/vertex-ai-vector-search-firestore/README.md +++ b/py/samples/vertex-ai-vector-search-firestore/README.md @@ -9,7 +9,17 @@ An example demonstrating the use Vector Search API with Firestore retriever for ```bash gcloud auth application-default login` ``` -3. Run the sample. +3. Set the following env vars to run the sample +``` +export LOCATION='' +export PROJECT_ID='' +export FIRESTORE_COLLECTION='' +export VECTOR_SEARCH_DEPLOYED_INDEX_ID='' +export VECTOR_SEARCH_INDEX_ENDPOINT_ID='' +export VECTOR_SEARCH_INDEX_ID='' +export VECTOR_SEARCH_PUBLIC_DOMAIN_NAME='' +``` +4. Run the sample. ## Run the sample diff --git a/py/samples/vertex-ai-vector-search-firestore/src/sample.py b/py/samples/vertex-ai-vector-search-firestore/src/sample.py index 516deaf693..b34fdba291 100644 --- a/py/samples/vertex-ai-vector-search-firestore/src/sample.py +++ b/py/samples/vertex-ai-vector-search-firestore/src/sample.py @@ -23,12 +23,12 @@ from genkit.ai import Genkit from genkit.blocks.document import Document from genkit.plugins.vertex_ai import ( + EmbeddingModels, VertexAI, VertexAIVectorSearch, vertexai_name, ) from genkit.plugins.vertex_ai.models.retriever import FirestoreRetriever -from genkit.plugins.vertex_ai import EmbeddingModels LOCATION = os.getenv('LOCATION') PROJECT_ID = os.getenv('PROJECT_ID')