import importlib
import os
import tempfile
from pathlib import Path
from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
from langchain_core.documents import Document
from langchain_core.embeddings import Embeddings
from langchain_core.vectorstores import VectorStore
from pydantic import ConfigDict
[docs]
class NeuralDBVectorStore(VectorStore):
"""Vectorstore that uses ThirdAI's NeuralDB.
To use, you should have the ``thirdai[neural_db]`` python package installed.
Example:
.. code-block:: python
from langchain_community.vectorstores import NeuralDBVectorStore
from thirdai import neural_db as ndb
db = ndb.NeuralDB()
vectorstore = NeuralDBVectorStore(db=db)
"""
[docs]
def __init__(self, db: Any) -> None:
self.db = db
db: Any = None #: :meta private:
"""NeuralDB instance"""
model_config = ConfigDict(
extra="forbid",
)
@staticmethod
def _verify_thirdai_library(thirdai_key: Optional[str] = None): # type: ignore[no-untyped-def]
try:
from thirdai import licensing
importlib.util.find_spec("thirdai.neural_db")
licensing.activate(thirdai_key or os.getenv("THIRDAI_KEY"))
except ImportError:
raise ImportError(
"Could not import thirdai python package and neuraldb dependencies. "
"Please install it with `pip install thirdai[neural_db]`."
)
[docs]
@classmethod
def from_scratch( # type: ignore[no-untyped-def, no-untyped-def]
cls,
thirdai_key: Optional[str] = None,
**model_kwargs,
):
"""
Create a NeuralDBVectorStore from scratch.
To use, set the ``THIRDAI_KEY`` environment variable with your ThirdAI
API key, or pass ``thirdai_key`` as a named parameter.
Example:
.. code-block:: python
from langchain_community.vectorstores import NeuralDBVectorStore
vectorstore = NeuralDBVectorStore.from_scratch(
thirdai_key="your-thirdai-key",
)
vectorstore.insert([
"/path/to/doc.pdf",
"/path/to/doc.docx",
"/path/to/doc.csv",
])
documents = vectorstore.similarity_search("AI-driven music therapy")
"""
NeuralDBVectorStore._verify_thirdai_library(thirdai_key)
from thirdai import neural_db as ndb
return cls(db=ndb.NeuralDB(**model_kwargs)) # type: ignore[call-arg]
[docs]
@classmethod
def from_checkpoint( # type: ignore[no-untyped-def]
cls,
checkpoint: Union[str, Path],
thirdai_key: Optional[str] = None,
):
"""
Create a NeuralDBVectorStore with a base model from a saved checkpoint
To use, set the ``THIRDAI_KEY`` environment variable with your ThirdAI
API key, or pass ``thirdai_key`` as a named parameter.
Example:
.. code-block:: python
from langchain_community.vectorstores import NeuralDBVectorStore
vectorstore = NeuralDBVectorStore.from_checkpoint(
checkpoint="/path/to/checkpoint.ndb",
thirdai_key="your-thirdai-key",
)
vectorstore.insert([
"/path/to/doc.pdf",
"/path/to/doc.docx",
"/path/to/doc.csv",
])
documents = vectorstore.similarity_search("AI-driven music therapy")
"""
NeuralDBVectorStore._verify_thirdai_library(thirdai_key)
from thirdai import neural_db as ndb
return cls(db=ndb.NeuralDB.from_checkpoint(checkpoint)) # type: ignore[call-arg]
[docs]
@classmethod
def from_texts(
cls,
texts: List[str],
embedding: Embeddings,
metadatas: Optional[List[dict]] = None,
**kwargs: Any,
) -> "NeuralDBVectorStore":
"""Return VectorStore initialized from texts and embeddings."""
model_kwargs = {}
if "thirdai_key" in kwargs:
model_kwargs["thirdai_key"] = kwargs["thirdai_key"]
del kwargs["thirdai_key"]
vectorstore = cls.from_scratch(**model_kwargs)
vectorstore.add_texts(texts, metadatas, **kwargs)
return vectorstore
[docs]
def add_texts(
self,
texts: Iterable[str],
metadatas: Optional[List[dict]] = None,
**kwargs: Any,
) -> List[str]:
"""Run more texts through the embeddings and add to the vectorstore.
Args:
texts: Iterable of strings to add to the vectorstore.
metadatas: Optional list of metadatas associated with the texts.
kwargs: vectorstore specific parameters
Returns:
List of ids from adding the texts into the vectorstore.
"""
import pandas as pd
from thirdai import neural_db as ndb
df = pd.DataFrame({"texts": texts})
if metadatas:
df = pd.concat([df, pd.DataFrame.from_records(metadatas)], axis=1)
temp = tempfile.NamedTemporaryFile("w", delete=False, delete_on_close=False) # type: ignore[call-overload]
df.to_csv(temp)
source_id = self.insert([ndb.CSV(temp.name)], **kwargs)[0]
offset = self.db._savable_state.documents.get_source_by_id(source_id)[1]
return [str(offset + i) for i in range(len(texts))] # type: ignore[arg-type]
[docs]
def insert( # type: ignore[no-untyped-def, no-untyped-def]
self,
sources: List[Any],
train: bool = True,
fast_mode: bool = True,
**kwargs,
):
"""Inserts files / document sources into the vectorstore.
Args:
train: When True this means that the underlying model in the
NeuralDB will undergo unsupervised pretraining on the inserted files.
Defaults to True.
fast_mode: Much faster insertion with a slight drop in performance.
Defaults to True.
"""
sources = self._preprocess_sources(sources)
self.db.insert(
sources=sources,
train=train,
fast_approximation=fast_mode,
**kwargs,
)
def _preprocess_sources(self, sources): # type: ignore[no-untyped-def]
"""Checks if the provided sources are string paths. If they are, convert
to NeuralDB document objects.
Args:
sources: list of either string paths to PDF, DOCX or CSV files, or
NeuralDB document objects.
"""
from thirdai import neural_db as ndb
if not sources:
return sources
preprocessed_sources = []
for doc in sources:
if not isinstance(doc, str):
preprocessed_sources.append(doc)
else:
if doc.lower().endswith(".pdf"):
preprocessed_sources.append(ndb.PDF(doc))
elif doc.lower().endswith(".docx"):
preprocessed_sources.append(ndb.DOCX(doc))
elif doc.lower().endswith(".csv"):
preprocessed_sources.append(ndb.CSV(doc))
else:
raise RuntimeError(
f"Could not automatically load {doc}. Only files "
"with .pdf, .docx, or .csv extensions can be loaded "
"automatically. For other formats, please use the "
"appropriate document object from the ThirdAI library."
)
return preprocessed_sources
[docs]
def upvote(self, query: str, document_id: Union[int, str]): # type: ignore[no-untyped-def]
"""The vectorstore upweights the score of a document for a specific query.
This is useful for fine-tuning the vectorstore to user behavior.
Args:
query: text to associate with `document_id`
document_id: id of the document to associate query with.
"""
self.db.text_to_result(query, int(document_id))
[docs]
def upvote_batch(self, query_id_pairs: List[Tuple[str, int]]): # type: ignore[no-untyped-def]
"""Given a batch of (query, document id) pairs, the vectorstore upweights
the scores of the document for the corresponding queries.
This is useful for fine-tuning the vectorstore to user behavior.
Args:
query_id_pairs: list of (query, document id) pairs. For each pair in
this list, the model will upweight the document id for the query.
"""
self.db.text_to_result_batch(
[(query, int(doc_id)) for query, doc_id in query_id_pairs]
)
[docs]
def associate(self, source: str, target: str): # type: ignore[no-untyped-def]
"""The vectorstore associates a source phrase with a target phrase.
When the vectorstore sees the source phrase, it will also consider results
that are relevant to the target phrase.
Args:
source: text to associate to `target`.
target: text to associate `source` to.
"""
self.db.associate(source, target)
[docs]
def associate_batch(self, text_pairs: List[Tuple[str, str]]): # type: ignore[no-untyped-def]
"""Given a batch of (source, target) pairs, the vectorstore associates
each source phrase with the corresponding target phrase.
Args:
text_pairs: list of (source, target) text pairs. For each pair in
this list, the source will be associated with the target.
"""
self.db.associate_batch(text_pairs)
[docs]
def similarity_search(
self, query: str, k: int = 10, **kwargs: Any
) -> List[Document]:
"""Retrieve {k} contexts with for a given query
Args:
query: Query to submit to the model
k: The max number of context results to retrieve. Defaults to 10.
"""
try:
references = self.db.search(query=query, top_k=k, **kwargs)
return [
Document(
page_content=ref.text,
metadata={
"id": ref.id,
"upvote_ids": ref.upvote_ids,
"source": ref.source,
"metadata": ref.metadata,
"score": ref.score,
"context": ref.context(1),
},
)
for ref in references
]
except Exception as e:
raise ValueError(f"Error while retrieving documents: {e}") from e
[docs]
def save(self, path: str): # type: ignore[no-untyped-def]
"""Saves a NeuralDB instance to disk. Can be loaded into memory by
calling NeuralDB.from_checkpoint(path)
Args:
path: path on disk to save the NeuralDB instance to.
"""
self.db.save(path)
[docs]
class NeuralDBClientVectorStore(VectorStore):
"""Vectorstore that uses ThirdAI's NeuralDB Enterprise Python Client for NeuralDBs.
To use, you should have the ``thirdai[neural_db]`` python package installed.
Example:
.. code-block:: python
from langchain_community.vectorstores import NeuralDBClientVectorStore
from thirdai.neural_db import ModelBazaar, NeuralDBClient
bazaar = ModelBazaar(base_url="http://{NEURAL_DB_ENTERPRISE_IP}/api/")
bazaar.log_in(email="user@thirdai.com", password="1234")
ndb_client = NeuralDBClient(
deployment_identifier="user/model-0:user/deployment-0",
base_url="http://{NEURAL_DB_ENTERPRISE_IP}/api/",
bazaar=bazaar
)
vectorstore = NeuralDBClientVectorStore(db=ndb_client)
retriever = vectorstore.as_retriever(search_kwargs={'k':5})
"""
[docs]
def __init__(self, db: Any) -> None:
self.db = db
db: Any = None #: :meta private:
"""NeuralDB Client instance"""
model_config = ConfigDict(
extra="forbid",
)
[docs]
def similarity_search(
self, query: str, k: int = 10, **kwargs: Any
) -> List[Document]:
"""Retrieve {k} contexts with for a given query
Args:
query: Query to submit to the model
k: The max number of context results to retrieve. Defaults to 10.
"""
try:
references = self.db.search(query=query, top_k=k, **kwargs)["references"]
return [
Document(
page_content=ref["text"],
metadata={
"id": ref["id"],
"source": ref["source"],
"metadata": ref["metadata"],
"score": ref["source"],
"context": ref["context"],
},
)
for ref in references
]
except Exception as e:
raise ValueError(f"Error while retrieving documents: {e}") from e
[docs]
def insert(self, documents: List[Dict[str, Any]]): # type: ignore[no-untyped-def, no-untyped-def]
"""
Inserts documents into the VectorStore and return the corresponding Sources.
Args:
documents (List[Dict[str, Any]]): A list of dictionaries that
represent documents to be inserted to the VectorStores.
The document dictionaries must be in the following format:
{"document_type": "DOCUMENT_TYPE", **kwargs} where "DOCUMENT_TYPE"
is one of the following:
"PDF", "CSV", "DOCX", "URL", "SentenceLevelPDF", "SentenceLevelDOCX",
"Unstructured", "InMemoryText".
The kwargs for each document type are shown below:
class PDF(Document):
document_type: Literal["PDF"]
path: str
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
version: str = "v1"
chunk_size: int = 100
stride: int = 40
emphasize_first_words: int = 0
ignore_header_footer: bool = True
ignore_nonstandard_orientation: bool = True
class CSV(Document):
document_type: Literal["CSV"]
path: str
id_column: Optional[str] = None
strong_columns: Optional[List[str]] = None
weak_columns: Optional[List[str]] = None
reference_columns: Optional[List[str]] = None
save_extra_info: bool = True
metadata: Optional[dict[str, Any]] = None
has_offset: bool = False
on_disk: bool = False
class DOCX(Document):
document_type: Literal["DOCX"]
path: str
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
class URL(Document):
document_type: Literal["URL"]
url: str
save_extra_info: bool = True
title_is_strong: bool = False
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
class SentenceLevelPDF(Document):
document_type: Literal["SentenceLevelPDF"]
path: str
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
class SentenceLevelDOCX(Document):
document_type: Literal["SentenceLevelDOCX"]
path: str
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
class Unstructured(Document):
document_type: Literal["Unstructured"]
path: str
save_extra_info: bool = True
metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
class InMemoryText(Document):
document_type: Literal["InMemoryText"]
name: str
texts: list[str]
metadatas: Optional[list[dict[str, Any]]] = None
global_metadata: Optional[dict[str, Any]] = None
on_disk: bool = False
For Document types with the arg "path", ensure that
the path exists on your local machine.
"""
return self.db.insert(documents)
[docs]
def remove_documents(self, source_ids: List[str]): # type: ignore[no-untyped-def]
"""
Deletes documents from the VectorStore using source ids.
Args:
files (List[str]): A list of source ids to delete from the VectorStore.
"""
self.db.delete(source_ids)
