Spaces:
Running
Running
File size: 8,654 Bytes
e70c280 5583f46 e70c280 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 | """
File: /embeddings.py
Created Date: Sunday October 19th 2025
Author: Christian Nonis <alch.infoemail@gmail.com>
-----
Last Modified: Thursday January 29th 2026 8:43:59 pm
Modified By: Christian Nonis <alch.infoemail@gmail.com>
-----
"""
import uuid
try:
from tenacity import (
retry,
stop_after_attempt,
wait_exponential,
retry_if_exception_type,
RetryError,
)
except ModuleNotFoundError:
def retry(*args, **kwargs):
def _decorator(func):
return func
return _decorator
def stop_after_attempt(*args, **kwargs):
return None
def wait_exponential(*args, **kwargs):
return None
def retry_if_exception_type(*args, **kwargs):
return None
class RetryError(Exception):
pass
from src.adapters.interfaces.embeddings import EmbeddingsClient, VectorStoreClient
from src.constants.embeddings import Vector
try:
from src.lib.embeddings.client import EmbeddingError
except ModuleNotFoundError:
class EmbeddingError(Exception):
pass
class EmbeddingFailureStrategy:
def on_single_failure(self, error: EmbeddingError) -> Vector:
raise NotImplementedError
def on_batch_failure(self, error: EmbeddingError, count: int) -> list[Vector]:
raise NotImplementedError
class ReturnEmptyVectorStrategy(EmbeddingFailureStrategy):
def on_single_failure(self, error: EmbeddingError) -> Vector:
print(f"Embedding failed in adapter, returning empty vector: {error}")
return Vector(id=str(uuid.uuid4()), embeddings=[], metadata={})
def on_batch_failure(self, error: EmbeddingError, count: int) -> list[Vector]:
print(f"Embedding failed in adapter, returning empty vectors: {error}")
return [Vector(id=str(uuid.uuid4()), embeddings=[], metadata={}) for _ in range(count)]
class RaiseEmbeddingFailureStrategy(EmbeddingFailureStrategy):
def on_single_failure(self, error: EmbeddingError) -> Vector:
raise error
def on_batch_failure(self, error: EmbeddingError, count: int) -> list[Vector]:
raise error
class EmbeddingsAdapter:
def __init__(self, failure_strategy: EmbeddingFailureStrategy | None = None):
self.embeddings = None
self.failure_strategy = failure_strategy or ReturnEmptyVectorStrategy()
def add_client(self, client: EmbeddingsClient) -> None:
"""
Add a embeddings client to the adapter.
"""
self.embeddings = client
def set_failure_strategy(self, strategy: EmbeddingFailureStrategy) -> None:
self.failure_strategy = strategy
def _to_embedding_error(self, error: Exception) -> EmbeddingError:
if isinstance(error, EmbeddingError):
return error
if isinstance(error, RetryError):
last_attempt = getattr(error, "last_attempt", None)
if last_attempt and hasattr(last_attempt, "exception"):
last_exception = last_attempt.exception()
if isinstance(last_exception, EmbeddingError):
return last_exception
if isinstance(last_exception, Exception):
return EmbeddingError(str(last_exception))
return EmbeddingError(str(error))
@retry(
stop=stop_after_attempt(5),
wait=wait_exponential(multiplier=1, min=2, max=30),
retry=retry_if_exception_type(EmbeddingError),
)
def _embed_text_with_retry(self, text: str) -> list[float]:
return self.embeddings.embed_text(text)
@retry(
stop=stop_after_attempt(5),
wait=wait_exponential(multiplier=1, min=2, max=30),
retry=retry_if_exception_type(EmbeddingError),
)
def _embed_texts_with_retry(self, texts: list[str]) -> list[list[float]]:
return self.embeddings.embed_texts(texts)
def embed_text(self, text: str) -> Vector:
"""
Embed a text and return a vector.
"""
try:
embeddings = self._embed_text_with_retry(text)
return Vector(id=str(uuid.uuid4()), embeddings=embeddings, metadata={})
except (EmbeddingError, RetryError) as e:
return self.failure_strategy.on_single_failure(self._to_embedding_error(e))
@retry(
stop=stop_after_attempt(5),
wait=wait_exponential(multiplier=1, min=2, max=30),
retry=retry_if_exception_type(EmbeddingError),
)
def _embed_query_with_retry(self, text: str) -> list[float]:
return self.embeddings.embed_query(text)
def embed_query(self, text: str) -> Vector:
"""Apply provider-specific query steering before the shared vector transform."""
try:
return Vector(id=str(uuid.uuid4()), embeddings=self._embed_query_with_retry(text), metadata={})
except (EmbeddingError, RetryError) as e:
return self.failure_strategy.on_single_failure(self._to_embedding_error(e))
def embed_texts(self, texts: list[str]) -> list[Vector]:
"""
Embed a list of texts and return a list of vectors.
"""
try:
embeddings_list = self._embed_texts_with_retry(texts)
return [
Vector(id=str(uuid.uuid4()), embeddings=embeddings, metadata={})
for embeddings in embeddings_list
]
except (EmbeddingError, RetryError) as e:
return self.failure_strategy.on_batch_failure(
self._to_embedding_error(e), len(texts)
)
class VectorStoreAdapter:
def __init__(self):
self.vector_store = None
def add_client(self, client: VectorStoreClient) -> None:
"""
Add a vector store client to the adapter.
"""
self.vector_store = client
def add_vectors(
self, vectors: list[Vector], store: str, brain_id: str = "default"
) -> list[str]:
"""
Add vectors to the vector store.
"""
return self.vector_store.add_vectors(vectors, store, brain_id)
def search_vectors(
self,
data_vector: list[float],
brain_id: str = "default",
store: str = "default",
k: int = 10,
) -> list[Vector]:
"""
Search vectors in the vector store and return the top k vectors.
"""
from src.utils.vector_search import stable_top_k_vectors
vectors = self.vector_store.search_vectors(data_vector, brain_id, store, k)
return stable_top_k_vectors(list(vectors), k)
def get_by_ids(
self, ids: list[str], store: str, brain_id: str = "default"
) -> list[Vector]:
"""
Get vectors by their IDs.
"""
return self.vector_store.get_by_ids(ids, store, brain_id)
def search_similar_by_ids(
self,
vector_ids: list[str],
brain_id: str,
store: str,
min_similarity: float,
limit: int = 10,
) -> dict[str, list[Vector]]:
"""
Finds vectors similar to the provided vector IDs within a specified store and brain.
Parameters:
vector_ids (list[str]): IDs of the source vectors to find similarities for.
brain_id (str): Identifier of the brain/namespace to search within.
store (str): Name of the vector store to query.
min_similarity (float): Minimum similarity threshold for returned vectors (inclusive).
limit (int): Maximum number of similar vectors to return per source ID.
Returns:
dict[str, list[Vector]]: Mapping from each source vector ID to a list of similar Vectors.
Each list contains at most `limit` items, includes only vectors with similarity >= `min_similarity`,
and is ordered by descending similarity.
"""
return self.vector_store.search_similar_by_ids(
vector_ids, brain_id, store, min_similarity, limit
)
def remove_vectors(
self, ids: list[str], store: str, brain_id: str = "default"
) -> None:
"""
Remove vectors from the vector store.
"""
return self.vector_store.remove_vectors(ids, store, brain_id)
def list_vectors(
self,
store: str,
brain_id: str = "default",
limit: int = 10,
skip: int = 0,
include_embeddings: bool = False,
) -> tuple[list[Vector], int]:
"""
List vectors in a store with pagination.
"""
return self.vector_store.list_vectors(
store, brain_id, limit, skip, include_embeddings
)
_embeddings_adapter = EmbeddingsAdapter()
_vector_store_adapter = VectorStoreAdapter()
|