gmakstutis/cognee
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge pull request #243 from jinhonglin-ryan/feature/milvus-vectordb

Browse source 
Feature: Integrate Milvus as a Vector Database Provider
This commit is contained in:
Igor Ilic 2024-12-03 13:10:18 +01:00 • committed by GitHub
commit 747a6b9d47
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
9 changed files with 575 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.env.template
View file

@ -14,7 +14,7 @@ GRAPH_DATABASE_URL=
GRAPH_DATABASE_USERNAME=
GRAPH_DATABASE_PASSWORD=
# "qdrant", "pgvector", "weaviate" or "lancedb"
# "qdrant", "pgvector", "weaviate", "milvus" or "lancedb"
VECTOR_DB_PROVIDER="lancedb"
# Not needed if using "lancedb" or "pgvector"
VECTOR_DB_URL=

64
.github/workflows/test_milvus.yml vendored Normal file
View file

@ -0,0 +1,64 @@
name: test | milvus
on:
workflow_dispatch:
pull_request:
branches:
- main
types: [labeled, synchronize]
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
cancel-in-progress: true
env:
RUNTIME__LOG_LEVEL: ERROR
ENV: 'dev'
jobs:
get_docs_changes:
name: docs changes
uses: ./.github/workflows/get_docs_changes.yml
run_milvus:
name: test
needs: get_docs_changes
if: needs.get_docs_changes.outputs.changes_outside_docs == 'true' && ${{ github.event.label.name == 'run-checks' }}
runs-on: ubuntu-latest
strategy:
fail-fast: false
defaults:
run:
shell: bash
steps:
- name: Check out
uses: actions/checkout@master
- name: Setup Python
uses: actions/setup-python@v5
with:
python-version: '3.11.x'
- name: Install Poetry
# https://github.com/snok/install-poetry#running-on-windows
uses: snok/install-poetry@v1.3.2
with:
virtualenvs-create: true
virtualenvs-in-project: true
installer-parallel: true
- name: Install dependencies
run: poetry install -E milvus --no-interaction
- name: Run default basic pipeline
env:
ENV: 'dev'
LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
run: poetry run python ./cognee/tests/test_milvus.py
- name: Clean up disk space
run: |
sudo rm -rf ~/.cache
sudo rm -rf /tmp/*
df -h

12
README.md
View file

@ -53,6 +53,12 @@ pip install 'cognee[qdrant]'
pip install 'cognee[neo4j]'
```
### With pip with Milvus support
```bash
pip install 'cognee[milvus]'
```
### With poetry
```bash
@ -83,6 +89,12 @@ poetry add cognee -E qdrant
poetry add cognee -E neo4j
```
### With poetry with Milvus support
```bash
poetry add cognee -E milvus
```
## 💻 Basic Usage

43
cognee/infrastructure/databases/vector/create_vector_engine.py
View file

@ -1,11 +1,13 @@
from typing import Dict
class VectorConfig(Dict):
vector_db_url: str
vector_db_port: str
vector_db_key: str
vector_db_provider: str
def create_vector_engine(config: VectorConfig, embedding_engine):
if config["vector_db_provider"] == "weaviate":
from .weaviate_db import WeaviateAdapter
@ -16,24 +18,37 @@ def create_vector_engine(config: VectorConfig, embedding_engine):
return WeaviateAdapter(
config["vector_db_url"],
config["vector_db_key"],
embedding_engine = embedding_engine
embedding_engine=embedding_engine
)
elif config["vector_db_provider"] == "qdrant":
if not (config["vector_db_url"] and config["vector_db_key"]):
raise EnvironmentError("Missing requred Qdrant credentials!")
from .qdrant.QDrantAdapter import QDrantAdapter
return QDrantAdapter(
url = config["vector_db_url"],
api_key = config["vector_db_key"],
embedding_engine = embedding_engine
url=config["vector_db_url"],
api_key=config["vector_db_key"],
embedding_engine=embedding_engine
)
elif config['vector_db_provider'] == 'milvus':
from .milvus.MilvusAdapter import MilvusAdapter
if not config["vector_db_url"]:
raise EnvironmentError("Missing required Milvus credentials!")
return MilvusAdapter(
url=config["vector_db_url"],
api_key=config['vector_db_key'],
embedding_engine=embedding_engine
)
elif config["vector_db_provider"] == "pgvector":
from cognee.infrastructure.databases.relational import get_relational_config
# Get configuration for postgres database
relational_config = get_relational_config()
db_username = relational_config.db_username
@ -52,8 +67,8 @@ def create_vector_engine(config: VectorConfig, embedding_engine):
from .pgvector.PGVectorAdapter import PGVectorAdapter
return PGVectorAdapter(
connection_string,
config["vector_db_key"],
connection_string,
config["vector_db_key"],
embedding_engine,
)
@ -64,16 +79,16 @@ def create_vector_engine(config: VectorConfig, embedding_engine):
from ..hybrid.falkordb.FalkorDBAdapter import FalkorDBAdapter
return FalkorDBAdapter(
database_url = config["vector_db_url"],
database_port = config["vector_db_port"],
embedding_engine = embedding_engine,
database_url=config["vector_db_url"],
database_port=config["vector_db_port"],
embedding_engine=embedding_engine,
)
else:
from .lancedb.LanceDBAdapter import LanceDBAdapter
return LanceDBAdapter(
url = config["vector_db_url"],
api_key = config["vector_db_key"],
embedding_engine = embedding_engine,
url=config["vector_db_url"],
api_key=config["vector_db_key"],
embedding_engine=embedding_engine,
)

252
cognee/infrastructure/databases/vector/milvus/MilvusAdapter.py Normal file
View file

@ -0,0 +1,252 @@
from __future__ import annotations
import asyncio
import logging
from typing import List, Optional
from uuid import UUID
from cognee.infrastructure.engine import DataPoint
from ..vector_db_interface import VectorDBInterface
from ..models.ScoredResult import ScoredResult
from ..embeddings.EmbeddingEngine import EmbeddingEngine
logger = logging.getLogger("MilvusAdapter")
class IndexSchema(DataPoint):
text: str
_metadata: dict = {
"index_fields": ["text"]
}
class MilvusAdapter(VectorDBInterface):
name = "Milvus"
url: str
api_key: Optional[str]
embedding_engine: EmbeddingEngine = None
def __init__(self, url: str, api_key: Optional[str], embedding_engine: EmbeddingEngine):
self.url = url
self.api_key = api_key
self.embedding_engine = embedding_engine
def get_milvus_client(self) -> "MilvusClient":
from pymilvus import MilvusClient
if self.api_key:
client = MilvusClient(uri=self.url, token=self.api_key)
else:
client = MilvusClient(uri=self.url)
return client
async def embed_data(self, data: List[str]) -> list[list[float]]:
return await self.embedding_engine.embed_text(data)
async def has_collection(self, collection_name: str) -> bool:
future = asyncio.Future()
client = self.get_milvus_client()
future.set_result(client.has_collection(collection_name=collection_name))
return await future
async def create_collection(
self,
collection_name: str,
payload_schema=None,
):
from pymilvus import DataType, MilvusException
client = self.get_milvus_client()
if client.has_collection(collection_name=collection_name):
logger.info(f"Collection '{collection_name}' already exists.")
return True
try:
dimension = self.embedding_engine.get_vector_size()
assert dimension > 0, "Embedding dimension must be greater than 0."
schema = client.create_schema(
auto_id=False,
enable_dynamic_field=False,
)
schema.add_field(
field_name="id",
datatype=DataType.VARCHAR,
is_primary=True,
max_length=36
)
schema.add_field(
field_name="vector",
datatype=DataType.FLOAT_VECTOR,
dim=dimension
)
schema.add_field(
field_name="text",
datatype=DataType.VARCHAR,
max_length=60535
)
index_params = client.prepare_index_params()
index_params.add_index(
field_name="vector",
metric_type="COSINE"
)
client.create_collection(
collection_name=collection_name,
schema=schema,
index_params=index_params
)
client.load_collection(collection_name)
logger.info(f"Collection '{collection_name}' created successfully.")
return True
except MilvusException as e:
logger.error(f"Error creating collection '{collection_name}': {str(e)}")
raise e
async def create_data_points(
self,
collection_name: str,
data_points: List[DataPoint]
):
from pymilvus import MilvusException
client = self.get_milvus_client()
data_vectors = await self.embed_data(
[data_point.get_embeddable_data() for data_point in data_points]
)
insert_data = [
{
"id": str(data_point.id),
"vector": data_vectors[index],
"text": data_point.text,
}
for index, data_point in enumerate(data_points)
]
try:
result = client.insert(
collection_name=collection_name,
data=insert_data
)
logger.info(
f"Inserted {result.get('insert_count', 0)} data points into collection '{collection_name}'."
)
return result
except MilvusException as e:
logger.error(f"Error inserting data points into collection '{collection_name}': {str(e)}")
raise e
async def create_vector_index(self, index_name: str, index_property_name: str):
await self.create_collection(f"{index_name}_{index_property_name}")
async def index_data_points(self, index_name: str, index_property_name: str, data_points: List[DataPoint]):
formatted_data_points = [
IndexSchema(
id=data_point.id,
text=getattr(data_point, data_point._metadata["index_fields"][0]),
)
for data_point in data_points
]
collection_name = f"{index_name}_{index_property_name}"
await self.create_data_points(collection_name, formatted_data_points)
async def retrieve(self, collection_name: str, data_point_ids: list[str]):
from pymilvus import MilvusException
client = self.get_milvus_client()
try:
filter_expression = f"""id in [{", ".join(f'"{id}"' for id in data_point_ids)}]"""
results = client.query(
collection_name=collection_name,
expr=filter_expression,
output_fields=["*"],
)
return results
except MilvusException as e:
logger.error(f"Error retrieving data points from collection '{collection_name}': {str(e)}")
raise e
async def search(
self,
collection_name: str,
query_text: Optional[str] = None,
query_vector: Optional[List[float]] = None,
limit: int = 5,
with_vector: bool = False,
):
from pymilvus import MilvusException
client = self.get_milvus_client()
if query_text is None and query_vector is None:
raise ValueError("One of query_text or query_vector must be provided!")
try:
query_vector = query_vector or (await self.embed_data([query_text]))[0]
output_fields = ["id", "text"]
if with_vector:
output_fields.append("vector")
results = client.search(
collection_name=collection_name,
data=[query_vector],
anns_field="vector",
limit=limit,
output_fields=output_fields,
search_params={
"metric_type": "COSINE",
},
)
return [
ScoredResult(
id=UUID(result["id"]),
score=result["distance"],
payload=result.get("entity", {}),
)
for result in results[0]
]
except MilvusException as e:
logger.error(f"Error during search in collection '{collection_name}': {str(e)}")
raise e
async def batch_search(self, collection_name: str, query_texts: List[str], limit: int, with_vectors: bool = False):
query_vectors = await self.embed_data(query_texts)
return await asyncio.gather(
*[self.search(collection_name=collection_name,
query_vector=query_vector,
limit=limit,
with_vector=with_vectors,
) for query_vector in query_vectors]
)
async def delete_data_points(self, collection_name: str, data_point_ids: list[str]):
from pymilvus import MilvusException
client = self.get_milvus_client()
try:
filter_expression = f"""id in [{", ".join(f'"{id}"' for id in data_point_ids)}]"""
delete_result = client.delete(
collection_name=collection_name,
filter=filter_expression
)
logger.info(f"Deleted data points with IDs {data_point_ids} from collection '{collection_name}'.")
return delete_result
except MilvusException as e:
logger.error(f"Error deleting data points from collection '{collection_name}': {str(e)}")
raise e
async def prune(self):
client = self.get_milvus_client()
if client:
collections = client.list_collections()
for collection_name in collections:
client.drop_collection(collection_name=collection_name)
client.close()

1
cognee/infrastructure/databases/vector/milvus/__init__.py Normal file
View file

@ -0,0 +1 @@
from .MilvusAdapter import MilvusAdapter

84
cognee/tests/test_milvus.py Normal file
View file

@ -0,0 +1,84 @@
import os
import logging
import pathlib
import cognee
from cognee.api.v1.search import SearchType
logging.basicConfig(level=logging.DEBUG)
async def main():
cognee.config.set_vector_db_provider("milvus")
data_directory_path = str(
pathlib.Path(os.path.join(pathlib.Path(__file__).parent, ".data_storage/test_milvus")).resolve())
cognee.config.data_root_directory(data_directory_path)
cognee_directory_path = str(
pathlib.Path(os.path.join(pathlib.Path(__file__).parent, ".cognee_system/test_milvus")).resolve())
cognee.config.system_root_directory(cognee_directory_path)
cognee.config.set_vector_db_config(
{
"vector_db_url": os.path.join(cognee_directory_path, "databases/milvus.db"),
"vector_db_key": "",
"vector_db_provider": "milvus"
}
)
await cognee.prune.prune_data()
await cognee.prune.prune_system(metadata=True)
dataset_name = "cs_explanations"
explanation_file_path = os.path.join(pathlib.Path(__file__).parent, "test_data/Natural_language_processing.txt")
await cognee.add([explanation_file_path], dataset_name)
text = """A quantum computer is a computer that takes advantage of quantum mechanical phenomena.
At small scales, physical matter exhibits properties of both particles and waves, and quantum computing leverages this behavior, specifically quantum superposition and entanglement, using specialized hardware that supports the preparation and manipulation of quantum states.
Classical physics cannot explain the operation of these quantum devices, and a scalable quantum computer could perform some calculations exponentially faster (with respect to input size scaling) than any modern "classical" computer. In particular, a large-scale quantum computer could break widely used encryption schemes and aid physicists in performing physical simulations; however, the current state of the technology is largely experimental and impractical, with several obstacles to useful applications. Moreover, scalable quantum computers do not hold promise for many practical tasks, and for many important tasks quantum speedups are proven impossible.
The basic unit of information in quantum computing is the qubit, similar to the bit in traditional digital electronics. Unlike a classical bit, a qubit can exist in a superposition of its two "basis" states. When measuring a qubit, the result is a probabilistic output of a classical bit, therefore making quantum computers nondeterministic in general. If a quantum computer manipulates the qubit in a particular way, wave interference effects can amplify the desired measurement results. The design of quantum algorithms involves creating procedures that allow a quantum computer to perform calculations efficiently and quickly.
Physically engineering high-quality qubits has proven challenging. If a physical qubit is not sufficiently isolated from its environment, it suffers from quantum decoherence, introducing noise into calculations. Paradoxically, perfectly isolating qubits is also undesirable because quantum computations typically need to initialize qubits, perform controlled qubit interactions, and measure the resulting quantum states. Each of those operations introduces errors and suffers from noise, and such inaccuracies accumulate.
In principle, a non-quantum (classical) computer can solve the same computational problems as a quantum computer, given enough time. Quantum advantage comes in the form of time complexity rather than computability, and quantum complexity theory shows that some quantum algorithms for carefully selected tasks require exponentially fewer computational steps than the best known non-quantum algorithms. Such tasks can in theory be solved on a large-scale quantum computer whereas classical computers would not finish computations in any reasonable amount of time. However, quantum speedup is not universal or even typical across computational tasks, since basic tasks such as sorting are proven to not allow any asymptotic quantum speedup. Claims of quantum supremacy have drawn significant attention to the discipline, but are demonstrated on contrived tasks, while near-term practical use cases remain limited.
"""
await cognee.add([text], dataset_name)
await cognee.cognify([dataset_name])
from cognee.infrastructure.databases.vector import get_vector_engine
vector_engine = get_vector_engine()
random_node = (await vector_engine.search("entity_name", "Quantum computer"))[0]
random_node_name = random_node.payload["text"]
search_results = await cognee.search(SearchType.INSIGHTS, query_text=random_node_name)
assert len(search_results) != 0, "The search results list is empty."
print("\n\nExtracted INSIGHTS are:\n")
for result in search_results:
print(f"{result}\n")
search_results = await cognee.search(SearchType.CHUNKS, query_text=random_node_name)
assert len(search_results) != 0, "The search results list is empty."
print("\n\nExtracted CHUNKS are:\n")
for result in search_results:
print(f"{result}\n")
search_results = await cognee.search(SearchType.SUMMARIES, query_text=random_node_name)
assert len(search_results) != 0, "The search results list is empty."
print("\nExtracted SUMMARIES are:\n")
for result in search_results:
print(f"{result}\n")
history = await cognee.get_search_history()
assert len(history) == 6, "Search history is not correct."
await cognee.prune.prune_data()
assert not os.path.isdir(data_directory_path), "Local data files are not deleted"
await cognee.prune.prune_system(metadata=True)
milvus_client = get_vector_engine().get_milvus_client()
collections = milvus_client.list_collections()
assert len(collections) == 0, "Milvus vector database is not empty"
if __name__ == "__main__":
import asyncio
asyncio.run(main())

131
poetry.lock generated
View file

@ -3602,6 +3602,22 @@ files = [
{file = "mergedeep-1.3.4.tar.gz", hash = "sha256:0096d52e9dad9939c3d975a774666af186eda617e6ca84df4c94dec30004f2a8"},
]
[[package]]
name = "milvus-lite"
version = "2.4.10"
description = "A lightweight version of Milvus wrapped with Python."
optional = true
python-versions = ">=3.7"
files = [
{file = "milvus_lite-2.4.10-py3-none-macosx_10_9_x86_64.whl", hash = "sha256:fc4246d3ed7d1910847afce0c9ba18212e93a6e9b8406048436940578dfad5cb"},
{file = "milvus_lite-2.4.10-py3-none-macosx_11_0_arm64.whl", hash = "sha256:74a8e07c5e3b057df17fbb46913388e84df1dc403a200f4e423799a58184c800"},
{file = "milvus_lite-2.4.10-py3-none-manylinux2014_aarch64.whl", hash = "sha256:240c7386b747bad696ecb5bd1f58d491e86b9d4b92dccee3315ed7256256eddc"},
{file = "milvus_lite-2.4.10-py3-none-manylinux2014_x86_64.whl", hash = "sha256:211d2e334a043f9282bdd9755f76b9b2d93b23bffa7af240919ffce6a8dfe325"},
]
[package.dependencies]
tqdm = "*"
[[package]]
name = "mistune"
version = "3.0.2"
@ -5360,6 +5376,31 @@ pyyaml = "*"
[package.extras]
extra = ["pygments (>=2.12)"]
[[package]]
name = "pymilvus"
version = "2.5.0"
description = "Python Sdk for Milvus"
optional = true
python-versions = ">=3.8"
files = [
{file = "pymilvus-2.5.0-py3-none-any.whl", hash = "sha256:a0e8653d8fe78019abfda79b3404ef7423f312501e8cbd7dc728051ce8732652"},
{file = "pymilvus-2.5.0.tar.gz", hash = "sha256:4da14a3bd957a4921166f9355fd1f1ac5c5e4e80b46f12f64d9c9a6dcb8cb395"},
]
[package.dependencies]
grpcio = ">=1.49.1,<=1.67.1"
milvus-lite = {version = ">=2.4.0", markers = "sys_platform != \"win32\""}
pandas = ">=1.2.4"
protobuf = ">=3.20.0"
python-dotenv = ">=1.0.1,<2.0.0"
setuptools = ">69"
ujson = ">=2.0.0"
[package.extras]
bulk-writer = ["azure-storage-blob", "minio (>=7.0.0)", "pyarrow (>=12.0.0)", "requests"]
dev = ["black", "grpcio (==1.62.2)", "grpcio-testing (==1.62.2)", "grpcio-tools (==1.62.2)", "pytest (>=5.3.4)", "pytest-cov (>=2.8.1)", "pytest-timeout (>=1.3.4)", "ruff (>0.4.0)"]
model = ["milvus-model (>=0.1.0)"]
[[package]]
name = "pyparsing"
version = "3.2.0"
@ -7075,6 +7116,93 @@ files = [
{file = "tzdata-2024.2.tar.gz", hash = "sha256:7d85cc416e9382e69095b7bdf4afd9e3880418a2413feec7069d533d6b4e31cc"},
]
[[package]]
name = "ujson"
version = "5.10.0"
description = "Ultra fast JSON encoder and decoder for Python"
optional = true
python-versions = ">=3.8"
files = [
{file = "ujson-5.10.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:2601aa9ecdbee1118a1c2065323bda35e2c5a2cf0797ef4522d485f9d3ef65bd"},
{file = "ujson-5.10.0-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:348898dd702fc1c4f1051bc3aacbf894caa0927fe2c53e68679c073375f732cf"},
{file = "ujson-5.10.0-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:22cffecf73391e8abd65ef5f4e4dd523162a3399d5e84faa6aebbf9583df86d6"},
{file = "ujson-5.10.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:26b0e2d2366543c1bb4fbd457446f00b0187a2bddf93148ac2da07a53fe51569"},
{file = "ujson-5.10.0-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:caf270c6dba1be7a41125cd1e4fc7ba384bf564650beef0df2dd21a00b7f5770"},
{file = "ujson-5.10.0-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:a245d59f2ffe750446292b0094244df163c3dc96b3ce152a2c837a44e7cda9d1"},
{file = "ujson-5.10.0-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:94a87f6e151c5f483d7d54ceef83b45d3a9cca7a9cb453dbdbb3f5a6f64033f5"},
{file = "ujson-5.10.0-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:29b443c4c0a113bcbb792c88bea67b675c7ca3ca80c3474784e08bba01c18d51"},
{file = "ujson-5.10.0-cp310-cp310-win32.whl", hash = "sha256:c18610b9ccd2874950faf474692deee4223a994251bc0a083c114671b64e6518"},
{file = "ujson-5.10.0-cp310-cp310-win_amd64.whl", hash = "sha256:924f7318c31874d6bb44d9ee1900167ca32aa9b69389b98ecbde34c1698a250f"},
{file = "ujson-5.10.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:a5b366812c90e69d0f379a53648be10a5db38f9d4ad212b60af00bd4048d0f00"},
{file = "ujson-5.10.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:502bf475781e8167f0f9d0e41cd32879d120a524b22358e7f205294224c71126"},
{file = "ujson-5.10.0-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:5b91b5d0d9d283e085e821651184a647699430705b15bf274c7896f23fe9c9d8"},
{file = "ujson-5.10.0-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:129e39af3a6d85b9c26d5577169c21d53821d8cf68e079060602e861c6e5da1b"},
{file = "ujson-5.10.0-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f77b74475c462cb8b88680471193064d3e715c7c6074b1c8c412cb526466efe9"},
{file = "ujson-5.10.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:7ec0ca8c415e81aa4123501fee7f761abf4b7f386aad348501a26940beb1860f"},
{file = "ujson-5.10.0-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:ab13a2a9e0b2865a6c6db9271f4b46af1c7476bfd51af1f64585e919b7c07fd4"},
{file = "ujson-5.10.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:57aaf98b92d72fc70886b5a0e1a1ca52c2320377360341715dd3933a18e827b1"},
{file = "ujson-5.10.0-cp311-cp311-win32.whl", hash = "sha256:2987713a490ceb27edff77fb184ed09acdc565db700ee852823c3dc3cffe455f"},
{file = "ujson-5.10.0-cp311-cp311-win_amd64.whl", hash = "sha256:f00ea7e00447918ee0eff2422c4add4c5752b1b60e88fcb3c067d4a21049a720"},
{file = "ujson-5.10.0-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:98ba15d8cbc481ce55695beee9f063189dce91a4b08bc1d03e7f0152cd4bbdd5"},
{file = "ujson-5.10.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:a9d2edbf1556e4f56e50fab7d8ff993dbad7f54bac68eacdd27a8f55f433578e"},
{file = "ujson-5.10.0-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6627029ae4f52d0e1a2451768c2c37c0c814ffc04f796eb36244cf16b8e57043"},
{file = "ujson-5.10.0-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f8ccb77b3e40b151e20519c6ae6d89bfe3f4c14e8e210d910287f778368bb3d1"},
{file = "ujson-5.10.0-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f3caf9cd64abfeb11a3b661329085c5e167abbe15256b3b68cb5d914ba7396f3"},
{file = "ujson-5.10.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:6e32abdce572e3a8c3d02c886c704a38a1b015a1fb858004e03d20ca7cecbb21"},
{file = "ujson-5.10.0-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:a65b6af4d903103ee7b6f4f5b85f1bfd0c90ba4eeac6421aae436c9988aa64a2"},
{file = "ujson-5.10.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:604a046d966457b6cdcacc5aa2ec5314f0e8c42bae52842c1e6fa02ea4bda42e"},
{file = "ujson-5.10.0-cp312-cp312-win32.whl", hash = "sha256:6dea1c8b4fc921bf78a8ff00bbd2bfe166345f5536c510671bccececb187c80e"},
{file = "ujson-5.10.0-cp312-cp312-win_amd64.whl", hash = "sha256:38665e7d8290188b1e0d57d584eb8110951a9591363316dd41cf8686ab1d0abc"},
{file = "ujson-5.10.0-cp313-cp313-macosx_10_9_x86_64.whl", hash = "sha256:618efd84dc1acbd6bff8eaa736bb6c074bfa8b8a98f55b61c38d4ca2c1f7f287"},
{file = "ujson-5.10.0-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:38d5d36b4aedfe81dfe251f76c0467399d575d1395a1755de391e58985ab1c2e"},
{file = "ujson-5.10.0-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:67079b1f9fb29ed9a2914acf4ef6c02844b3153913eb735d4bf287ee1db6e557"},
{file = "ujson-5.10.0-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d7d0e0ceeb8fe2468c70ec0c37b439dd554e2aa539a8a56365fd761edb418988"},
{file = "ujson-5.10.0-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:59e02cd37bc7c44d587a0ba45347cc815fb7a5fe48de16bf05caa5f7d0d2e816"},
{file = "ujson-5.10.0-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:2a890b706b64e0065f02577bf6d8ca3b66c11a5e81fb75d757233a38c07a1f20"},
{file = "ujson-5.10.0-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:621e34b4632c740ecb491efc7f1fcb4f74b48ddb55e65221995e74e2d00bbff0"},
{file = "ujson-5.10.0-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:b9500e61fce0cfc86168b248104e954fead61f9be213087153d272e817ec7b4f"},
{file = "ujson-5.10.0-cp313-cp313-win32.whl", hash = "sha256:4c4fc16f11ac1612f05b6f5781b384716719547e142cfd67b65d035bd85af165"},
{file = "ujson-5.10.0-cp313-cp313-win_amd64.whl", hash = "sha256:4573fd1695932d4f619928fd09d5d03d917274381649ade4328091ceca175539"},
{file = "ujson-5.10.0-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:a984a3131da7f07563057db1c3020b1350a3e27a8ec46ccbfbf21e5928a43050"},
{file = "ujson-5.10.0-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:73814cd1b9db6fc3270e9d8fe3b19f9f89e78ee9d71e8bd6c9a626aeaeaf16bd"},
{file = "ujson-5.10.0-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:61e1591ed9376e5eddda202ec229eddc56c612b61ac6ad07f96b91460bb6c2fb"},
{file = "ujson-5.10.0-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d2c75269f8205b2690db4572a4a36fe47cd1338e4368bc73a7a0e48789e2e35a"},
{file = "ujson-5.10.0-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:7223f41e5bf1f919cd8d073e35b229295aa8e0f7b5de07ed1c8fddac63a6bc5d"},
{file = "ujson-5.10.0-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:d4dc2fd6b3067c0782e7002ac3b38cf48608ee6366ff176bbd02cf969c9c20fe"},
{file = "ujson-5.10.0-cp38-cp38-musllinux_1_2_i686.whl", hash = "sha256:232cc85f8ee3c454c115455195a205074a56ff42608fd6b942aa4c378ac14dd7"},
{file = "ujson-5.10.0-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:cc6139531f13148055d691e442e4bc6601f6dba1e6d521b1585d4788ab0bfad4"},
{file = "ujson-5.10.0-cp38-cp38-win32.whl", hash = "sha256:e7ce306a42b6b93ca47ac4a3b96683ca554f6d35dd8adc5acfcd55096c8dfcb8"},
{file = "ujson-5.10.0-cp38-cp38-win_amd64.whl", hash = "sha256:e82d4bb2138ab05e18f089a83b6564fee28048771eb63cdecf4b9b549de8a2cc"},
{file = "ujson-5.10.0-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:dfef2814c6b3291c3c5f10065f745a1307d86019dbd7ea50e83504950136ed5b"},
{file = "ujson-5.10.0-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:4734ee0745d5928d0ba3a213647f1c4a74a2a28edc6d27b2d6d5bd9fa4319e27"},
{file = "ujson-5.10.0-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d47ebb01bd865fdea43da56254a3930a413f0c5590372a1241514abae8aa7c76"},
{file = "ujson-5.10.0-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:dee5e97c2496874acbf1d3e37b521dd1f307349ed955e62d1d2f05382bc36dd5"},
{file = "ujson-5.10.0-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:7490655a2272a2d0b072ef16b0b58ee462f4973a8f6bbe64917ce5e0a256f9c0"},
{file = "ujson-5.10.0-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:ba17799fcddaddf5c1f75a4ba3fd6441f6a4f1e9173f8a786b42450851bd74f1"},
{file = "ujson-5.10.0-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:2aff2985cef314f21d0fecc56027505804bc78802c0121343874741650a4d3d1"},
{file = "ujson-5.10.0-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:ad88ac75c432674d05b61184178635d44901eb749786c8eb08c102330e6e8996"},
{file = "ujson-5.10.0-cp39-cp39-win32.whl", hash = "sha256:2544912a71da4ff8c4f7ab5606f947d7299971bdd25a45e008e467ca638d13c9"},
{file = "ujson-5.10.0-cp39-cp39-win_amd64.whl", hash = "sha256:3ff201d62b1b177a46f113bb43ad300b424b7847f9c5d38b1b4ad8f75d4a282a"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-macosx_10_9_x86_64.whl", hash = "sha256:5b6fee72fa77dc172a28f21693f64d93166534c263adb3f96c413ccc85ef6e64"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-macosx_11_0_arm64.whl", hash = "sha256:61d0af13a9af01d9f26d2331ce49bb5ac1fb9c814964018ac8df605b5422dcb3"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:ecb24f0bdd899d368b715c9e6664166cf694d1e57be73f17759573a6986dd95a"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:fbd8fd427f57a03cff3ad6574b5e299131585d9727c8c366da4624a9069ed746"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:beeaf1c48e32f07d8820c705ff8e645f8afa690cca1544adba4ebfa067efdc88"},
{file = "ujson-5.10.0-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:baed37ea46d756aca2955e99525cc02d9181de67f25515c468856c38d52b5f3b"},
{file = "ujson-5.10.0-pp38-pypy38_pp73-macosx_10_9_x86_64.whl", hash = "sha256:7663960f08cd5a2bb152f5ee3992e1af7690a64c0e26d31ba7b3ff5b2ee66337"},
{file = "ujson-5.10.0-pp38-pypy38_pp73-macosx_11_0_arm64.whl", hash = "sha256:d8640fb4072d36b08e95a3a380ba65779d356b2fee8696afeb7794cf0902d0a1"},
{file = "ujson-5.10.0-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:78778a3aa7aafb11e7ddca4e29f46bc5139131037ad628cc10936764282d6753"},
{file = "ujson-5.10.0-pp38-pypy38_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:b0111b27f2d5c820e7f2dbad7d48e3338c824e7ac4d2a12da3dc6061cc39c8e6"},
{file = "ujson-5.10.0-pp38-pypy38_pp73-win_amd64.whl", hash = "sha256:c66962ca7565605b355a9ed478292da628b8f18c0f2793021ca4425abf8b01e5"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-macosx_10_9_x86_64.whl", hash = "sha256:ba43cc34cce49cf2d4bc76401a754a81202d8aa926d0e2b79f0ee258cb15d3a4"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-macosx_11_0_arm64.whl", hash = "sha256:ac56eb983edce27e7f51d05bc8dd820586c6e6be1c5216a6809b0c668bb312b8"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f44bd4b23a0e723bf8b10628288c2c7c335161d6840013d4d5de20e48551773b"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:7c10f4654e5326ec14a46bcdeb2b685d4ada6911050aa8baaf3501e57024b804"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:0de4971a89a762398006e844ae394bd46991f7c385d7a6a3b93ba229e6dac17e"},
{file = "ujson-5.10.0-pp39-pypy39_pp73-win_amd64.whl", hash = "sha256:e1402f0564a97d2a52310ae10a64d25bcef94f8dd643fcf5d310219d915484f7"},
{file = "ujson-5.10.0.tar.gz", hash = "sha256:b3cd8f3c5d8c7738257f1018880444f7b7d9b66232c64649f562d7ba86ad4bc1"},
]
[[package]]
name = "uri-template"
version = "1.3.0"
@ -7635,6 +7763,7 @@ groq = ["groq"]
langchain = ["langchain_text_splitters", "langsmith"]
langfuse = ["langfuse"]
llama-index = ["llama-index-core"]
milvus = ["pymilvus"]
neo4j = ["neo4j"]
notebook = []
postgres = ["asyncpg", "pgvector", "psycopg2"]
@ -7645,4 +7774,4 @@ weaviate = ["weaviate-client"]
[metadata]
lock-version = "2.0"
python-versions = ">=3.9.0,<3.12"
content-hash = "6b57d44b0924bcf64397b3807c2a6ba369166e1d2102b5312c8f8ae2d5323376"
content-hash = "d6b10b74a910202f224ff34fa06ad3d2767796a6492a96724de0d608ac0356c5"

3
pyproject.toml
View file

@ -70,6 +70,7 @@ asyncpg = {version = "0.30.0", optional = true}
pgvector = {version = "^0.3.5", optional = true}
psycopg2 = {version = "^2.9.10", optional = true}
llama-index-core = {version = "^0.11.22", optional = true}
pymilvus = {version = "^2.5.0", optional = true}
[tool.poetry.extras]
filesystem = ["s3fs", "botocore"]
@ -84,7 +85,7 @@ posthog = ["posthog"]
falkordb = ["falkordb"]
groq = ["groq"]
langfuse = ["langfuse"]
milvus = ["pymilvus"]
[tool.poetry.group.dev.dependencies]
pytest = "^7.4.0"