from __future__ import annotations
import logging
import os
import warnings
from typing import (
Any,
Dict,
Iterable,
List,
Literal,
Mapping,
Optional,
Sequence,
Set,
Tuple,
Union,
cast,
)
import openai
import tiktoken
from langchain_core.embeddings import Embeddings
from langchain_core.pydantic_v1 import (
BaseModel,
Extra,
Field,
SecretStr,
root_validator,
)
from langchain_core.utils import (
convert_to_secret_str,
get_from_dict_or_env,
get_pydantic_field_names,
)
logger = logging.getLogger(__name__)
def _process_batched_chunked_embeddings(
num_texts: int,
tokens: List[Union[List[int], str]],
batched_embeddings: List[List[float]],
indices: List[int],
skip_empty: bool,
) -> List[Optional[List[float]]]:
# for each text, this is the list of embeddings (list of list of floats)
# corresponding to the chunks of the text
results: List[List[List[float]]] = [[] for _ in range(num_texts)]
# for each text, this is the token length of each chunk
# for transformers tokenization, this is the string length
# for tiktoken, this is the number of tokens
num_tokens_in_batch: List[List[int]] = [[] for _ in range(num_texts)]
for i in range(len(indices)):
if skip_empty and len(batched_embeddings[i]) == 1:
continue
results[indices[i]].append(batched_embeddings[i])
num_tokens_in_batch[indices[i]].append(len(tokens[i]))
# for each text, this is the final embedding
embeddings: List[Optional[List[float]]] = []
for i in range(num_texts):
# an embedding for each chunk
_result: List[List[float]] = results[i]
if len(_result) == 0:
# this will be populated with the embedding of an empty string
# in the sync or async code calling this
embeddings.append(None)
continue
elif len(_result) == 1:
# if only one embedding was produced, use it
embeddings.append(_result[0])
continue
else:
# else we need to weighted average
# should be same as
# average = np.average(_result, axis=0, weights=num_tokens_in_batch[i])
total_weight = sum(num_tokens_in_batch[i])
average = [
sum(
val * weight
for val, weight in zip(embedding, num_tokens_in_batch[i])
)
/ total_weight
for embedding in zip(*_result)
]
# should be same as
# embeddings.append((average / np.linalg.norm(average)).tolist())
magnitude = sum(val**2 for val in average) ** 0.5
embeddings.append([val / magnitude for val in average])
return embeddings
[docs]
class OpenAIEmbeddings(BaseModel, Embeddings):
"""OpenAI embedding models.
To use, you should have the
environment variable ``OPENAI_API_KEY`` set with your API key or pass it
as a named parameter to the constructor.
In order to use the library with Microsoft Azure endpoints, use
AzureOpenAIEmbeddings.
Example:
.. code-block:: python
from langchain_openai import OpenAIEmbeddings
model = OpenAIEmbeddings(model="text-embedding-3-large")
"""
client: Any = Field(default=None, exclude=True) #: :meta private:
async_client: Any = Field(default=None, exclude=True) #: :meta private:
model: str = "text-embedding-ada-002"
dimensions: Optional[int] = None
"""The number of dimensions the resulting output embeddings should have.
Only supported in `text-embedding-3` and later models.
"""
# to support Azure OpenAI Service custom deployment names
deployment: Optional[str] = model
# TODO: Move to AzureOpenAIEmbeddings.
openai_api_version: Optional[str] = Field(default=None, alias="api_version")
"""Automatically inferred from env var `OPENAI_API_VERSION` if not provided."""
# to support Azure OpenAI Service custom endpoints
openai_api_base: Optional[str] = Field(default=None, alias="base_url")
"""Base URL path for API requests, leave blank if not using a proxy or service
emulator."""
# to support Azure OpenAI Service custom endpoints
openai_api_type: Optional[str] = None
# to support explicit proxy for OpenAI
openai_proxy: Optional[str] = None
embedding_ctx_length: int = 8191
"""The maximum number of tokens to embed at once."""
openai_api_key: Optional[SecretStr] = Field(default=None, alias="api_key")
"""Automatically inferred from env var `OPENAI_API_KEY` if not provided."""
openai_organization: Optional[str] = Field(default=None, alias="organization")
"""Automatically inferred from env var `OPENAI_ORG_ID` if not provided."""
allowed_special: Union[Literal["all"], Set[str], None] = None
disallowed_special: Union[Literal["all"], Set[str], Sequence[str], None] = None
chunk_size: int = 1000
"""Maximum number of texts to embed in each batch"""
max_retries: int = 2
"""Maximum number of retries to make when generating."""
request_timeout: Optional[Union[float, Tuple[float, float], Any]] = Field(
default=None, alias="timeout"
)
"""Timeout for requests to OpenAI completion API. Can be float, httpx.Timeout or
None."""
headers: Any = None
tiktoken_enabled: bool = True
"""Set this to False for non-OpenAI implementations of the embeddings API, e.g.
the `--extensions openai` extension for `text-generation-webui`"""
tiktoken_model_name: Optional[str] = None
"""The model name to pass to tiktoken when using this class.
Tiktoken is used to count the number of tokens in documents to constrain
them to be under a certain limit. By default, when set to None, this will
be the same as the embedding model name. However, there are some cases
where you may want to use this Embedding class with a model name not
supported by tiktoken. This can include when using Azure embeddings or
when using one of the many model providers that expose an OpenAI-like
API but with different models. In those cases, in order to avoid erroring
when tiktoken is called, you can specify a model name to use here."""
show_progress_bar: bool = False
"""Whether to show a progress bar when embedding."""
model_kwargs: Dict[str, Any] = Field(default_factory=dict)
"""Holds any model parameters valid for `create` call not explicitly specified."""
skip_empty: bool = False
"""Whether to skip empty strings when embedding or raise an error.
Defaults to not skipping."""
default_headers: Union[Mapping[str, str], None] = None
default_query: Union[Mapping[str, object], None] = None
# Configure a custom httpx client. See the
# [httpx documentation](https://www.python-httpx.org/api/#client) for more details.
retry_min_seconds: int = 4
"""Min number of seconds to wait between retries"""
retry_max_seconds: int = 20
"""Max number of seconds to wait between retries"""
http_client: Union[Any, None] = None
"""Optional httpx.Client. Only used for sync invocations. Must specify
http_async_client as well if you'd like a custom client for async invocations.
"""
http_async_client: Union[Any, None] = None
"""Optional httpx.AsyncClient. Only used for async invocations. Must specify
http_client as well if you'd like a custom client for sync invocations."""
check_embedding_ctx_length: bool = True
"""Whether to check the token length of inputs and automatically split inputs
longer than embedding_ctx_length."""
class Config:
"""Configuration for this pydantic object."""
extra = Extra.forbid
allow_population_by_field_name = True
@root_validator(pre=True)
def build_extra(cls, values: Dict[str, Any]) -> Dict[str, Any]:
"""Build extra kwargs from additional params that were passed in."""
all_required_field_names = get_pydantic_field_names(cls)
extra = values.get("model_kwargs", {})
for field_name in list(values):
if field_name in extra:
raise ValueError(f"Found {field_name} supplied twice.")
if field_name not in all_required_field_names:
warnings.warn(
f"""WARNING! {field_name} is not default parameter.
{field_name} was transferred to model_kwargs.
Please confirm that {field_name} is what you intended."""
)
extra[field_name] = values.pop(field_name)
invalid_model_kwargs = all_required_field_names.intersection(extra.keys())
if invalid_model_kwargs:
raise ValueError(
f"Parameters {invalid_model_kwargs} should be specified explicitly. "
f"Instead they were passed in as part of `model_kwargs` parameter."
)
values["model_kwargs"] = extra
return values
@root_validator()
def validate_environment(cls, values: Dict) -> Dict:
"""Validate that api key and python package exists in environment."""
openai_api_key = get_from_dict_or_env(
values, "openai_api_key", "OPENAI_API_KEY"
)
values["openai_api_key"] = (
convert_to_secret_str(openai_api_key) if openai_api_key else None
)
values["openai_api_base"] = values["openai_api_base"] or os.getenv(
"OPENAI_API_BASE"
)
values["openai_api_type"] = get_from_dict_or_env(
values, "openai_api_type", "OPENAI_API_TYPE", default=""
)
values["openai_proxy"] = get_from_dict_or_env(
values, "openai_proxy", "OPENAI_PROXY", default=""
)
if values["openai_api_type"] in ("azure", "azure_ad", "azuread"):
default_api_version = "2023-05-15"
# Azure OpenAI embedding models allow a maximum of 16 texts
# at a time in each batch
# See: https://learn.microsoft.com/en-us/azure/ai-services/openai/reference#embeddings
values["chunk_size"] = min(values["chunk_size"], 16)
else:
default_api_version = ""
values["openai_api_version"] = get_from_dict_or_env(
values,
"openai_api_version",
"OPENAI_API_VERSION",
default=default_api_version,
)
# Check OPENAI_ORGANIZATION for backwards compatibility.
values["openai_organization"] = (
values["openai_organization"]
or os.getenv("OPENAI_ORG_ID")
or os.getenv("OPENAI_ORGANIZATION")
)
if values["openai_api_type"] in ("azure", "azure_ad", "azuread"):
raise ValueError(
"If you are using Azure, "
"please use the `AzureOpenAIEmbeddings` class."
)
client_params = {
"api_key": (
values["openai_api_key"].get_secret_value()
if values["openai_api_key"]
else None
),
"organization": values["openai_organization"],
"base_url": values["openai_api_base"],
"timeout": values["request_timeout"],
"max_retries": values["max_retries"],
"default_headers": values["default_headers"],
"default_query": values["default_query"],
}
if values["openai_proxy"] and (
values["http_client"] or values["http_async_client"]
):
openai_proxy = values["openai_proxy"]
http_client = values["http_client"]
http_async_client = values["http_async_client"]
raise ValueError(
"Cannot specify 'openai_proxy' if one of "
"'http_client'/'http_async_client' is already specified. Received:\n"
f"{openai_proxy=}\n{http_client=}\n{http_async_client=}"
)
if not values.get("client"):
if values["openai_proxy"] and not values["http_client"]:
try:
import httpx
except ImportError as e:
raise ImportError(
"Could not import httpx python package. "
"Please install it with `pip install httpx`."
) from e
values["http_client"] = httpx.Client(proxy=values["openai_proxy"])
sync_specific = {"http_client": values["http_client"]}
values["client"] = openai.OpenAI(
**client_params, **sync_specific
).embeddings
if not values.get("async_client"):
if values["openai_proxy"] and not values["http_async_client"]:
try:
import httpx
except ImportError as e:
raise ImportError(
"Could not import httpx python package. "
"Please install it with `pip install httpx`."
) from e
values["http_async_client"] = httpx.AsyncClient(
proxy=values["openai_proxy"]
)
async_specific = {"http_client": values["http_async_client"]}
values["async_client"] = openai.AsyncOpenAI(
**client_params, **async_specific
).embeddings
return values
@property
def _invocation_params(self) -> Dict[str, Any]:
params: Dict = {"model": self.model, **self.model_kwargs}
if self.dimensions is not None:
params["dimensions"] = self.dimensions
return params
def _tokenize(
self, texts: List[str], chunk_size: int
) -> Tuple[Iterable[int], List[Union[List[int], str]], List[int]]:
"""
Take the input `texts` and `chunk_size` and return 3 iterables as a tuple:
We have `batches`, where batches are sets of individual texts
we want responses from the openai api. The length of a single batch is
`chunk_size` texts.
Each individual text is also split into multiple texts based on the
`embedding_ctx_length` parameter (based on number of tokens).
This function returns a 3-tuple of the following:
_iter: An iterable of the starting index in `tokens` for each *batch*
tokens: A list of tokenized texts, where each text has already been split
into sub-texts based on the `embedding_ctx_length` parameter. In the
case of tiktoken, this is a list of token arrays. In the case of
HuggingFace transformers, this is a list of strings.
indices: An iterable of the same length as `tokens` that maps each token-array
to the index of the original text in `texts`.
"""
tokens: List[Union[List[int], str]] = []
indices: List[int] = []
model_name = self.tiktoken_model_name or self.model
# If tiktoken flag set to False
if not self.tiktoken_enabled:
try:
from transformers import AutoTokenizer
except ImportError:
raise ValueError(
"Could not import transformers python package. "
"This is needed for OpenAIEmbeddings to work without "
"`tiktoken`. Please install it with `pip install transformers`. "
)
tokenizer = AutoTokenizer.from_pretrained(
pretrained_model_name_or_path=model_name
)
for i, text in enumerate(texts):
# Tokenize the text using HuggingFace transformers
tokenized: List[int] = tokenizer.encode(text, add_special_tokens=False)
# Split tokens into chunks respecting the embedding_ctx_length
for j in range(0, len(tokenized), self.embedding_ctx_length):
token_chunk: List[int] = tokenized[
j : j + self.embedding_ctx_length
]
# Convert token IDs back to a string
chunk_text: str = tokenizer.decode(token_chunk)
tokens.append(chunk_text)
indices.append(i)
else:
try:
encoding = tiktoken.encoding_for_model(model_name)
except KeyError:
encoding = tiktoken.get_encoding("cl100k_base")
encoder_kwargs: Dict[str, Any] = {
k: v
for k, v in {
"allowed_special": self.allowed_special,
"disallowed_special": self.disallowed_special,
}.items()
if v is not None
}
for i, text in enumerate(texts):
if self.model.endswith("001"):
# See: https://github.com/openai/openai-python/
# issues/418#issuecomment-1525939500
# replace newlines, which can negatively affect performance.
text = text.replace("\n", " ")
if encoder_kwargs:
token = encoding.encode(text, **encoder_kwargs)
else:
token = encoding.encode_ordinary(text)
# Split tokens into chunks respecting the embedding_ctx_length
for j in range(0, len(token), self.embedding_ctx_length):
tokens.append(token[j : j + self.embedding_ctx_length])
indices.append(i)
if self.show_progress_bar:
try:
from tqdm.auto import tqdm
_iter: Iterable = tqdm(range(0, len(tokens), chunk_size))
except ImportError:
_iter = range(0, len(tokens), chunk_size)
else:
_iter = range(0, len(tokens), chunk_size)
return _iter, tokens, indices
# please refer to
# https://github.com/openai/openai-cookbook/blob/main/examples/Embedding_long_inputs.ipynb
def _get_len_safe_embeddings(
self, texts: List[str], *, engine: str, chunk_size: Optional[int] = None
) -> List[List[float]]:
"""
Generate length-safe embeddings for a list of texts.
This method handles tokenization and embedding generation, respecting the
set embedding context length and chunk size. It supports both tiktoken
and HuggingFace tokenizer based on the tiktoken_enabled flag.
Args:
texts (List[str]): A list of texts to embed.
engine (str): The engine or model to use for embeddings.
chunk_size (Optional[int]): The size of chunks for processing embeddings.
Returns:
List[List[float]]: A list of embeddings for each input text.
"""
_chunk_size = chunk_size or self.chunk_size
_iter, tokens, indices = self._tokenize(texts, _chunk_size)
batched_embeddings: List[List[float]] = []
for i in _iter:
response = self.client.create(
input=tokens[i : i + _chunk_size], **self._invocation_params
)
if not isinstance(response, dict):
response = response.model_dump()
batched_embeddings.extend(r["embedding"] for r in response["data"])
embeddings = _process_batched_chunked_embeddings(
len(texts), tokens, batched_embeddings, indices, self.skip_empty
)
_cached_empty_embedding: Optional[List[float]] = None
def empty_embedding() -> List[float]:
nonlocal _cached_empty_embedding
if _cached_empty_embedding is None:
average_embedded = self.client.create(
input="", **self._invocation_params
)
if not isinstance(average_embedded, dict):
average_embedded = average_embedded.model_dump()
_cached_empty_embedding = average_embedded["data"][0]["embedding"]
return _cached_empty_embedding
return [e if e is not None else empty_embedding() for e in embeddings]
# please refer to
# https://github.com/openai/openai-cookbook/blob/main/examples/Embedding_long_inputs.ipynb
async def _aget_len_safe_embeddings(
self, texts: List[str], *, engine: str, chunk_size: Optional[int] = None
) -> List[List[float]]:
"""
Asynchronously generate length-safe embeddings for a list of texts.
This method handles tokenization and asynchronous embedding generation,
respecting the set embedding context length and chunk size. It supports both
`tiktoken` and HuggingFace `tokenizer` based on the tiktoken_enabled flag.
Args:
texts (List[str]): A list of texts to embed.
engine (str): The engine or model to use for embeddings.
chunk_size (Optional[int]): The size of chunks for processing embeddings.
Returns:
List[List[float]]: A list of embeddings for each input text.
"""
_chunk_size = chunk_size or self.chunk_size
_iter, tokens, indices = self._tokenize(texts, _chunk_size)
batched_embeddings: List[List[float]] = []
_chunk_size = chunk_size or self.chunk_size
for i in range(0, len(tokens), _chunk_size):
response = await self.async_client.create(
input=tokens[i : i + _chunk_size], **self._invocation_params
)
if not isinstance(response, dict):
response = response.model_dump()
batched_embeddings.extend(r["embedding"] for r in response["data"])
embeddings = _process_batched_chunked_embeddings(
len(texts), tokens, batched_embeddings, indices, self.skip_empty
)
_cached_empty_embedding: Optional[List[float]] = None
async def empty_embedding() -> List[float]:
nonlocal _cached_empty_embedding
if _cached_empty_embedding is None:
average_embedded = await self.async_client.create(
input="", **self._invocation_params
)
if not isinstance(average_embedded, dict):
average_embedded = average_embedded.model_dump()
_cached_empty_embedding = average_embedded["data"][0]["embedding"]
return _cached_empty_embedding
return [e if e is not None else await empty_embedding() for e in embeddings]
[docs]
def embed_documents(
self, texts: List[str], chunk_size: Optional[int] = 0
) -> List[List[float]]:
"""Call out to OpenAI's embedding endpoint for embedding search docs.
Args:
texts: The list of texts to embed.
chunk_size: The chunk size of embeddings. If None, will use the chunk size
specified by the class.
Returns:
List of embeddings, one for each text.
"""
if not self.check_embedding_ctx_length:
embeddings: List[List[float]] = []
for text in texts:
response = self.client.create(input=text, **self._invocation_params)
if not isinstance(response, dict):
response = response.dict()
embeddings.extend(r["embedding"] for r in response["data"])
return embeddings
# NOTE: to keep things simple, we assume the list may contain texts longer
# than the maximum context and use length-safe embedding function.
engine = cast(str, self.deployment)
return self._get_len_safe_embeddings(texts, engine=engine)
[docs]
async def aembed_documents(
self, texts: List[str], chunk_size: Optional[int] = 0
) -> List[List[float]]:
"""Call out to OpenAI's embedding endpoint async for embedding search docs.
Args:
texts: The list of texts to embed.
chunk_size: The chunk size of embeddings. If None, will use the chunk size
specified by the class.
Returns:
List of embeddings, one for each text.
"""
if not self.check_embedding_ctx_length:
embeddings: List[List[float]] = []
for text in texts:
response = await self.async_client.create(
input=text, **self._invocation_params
)
if not isinstance(response, dict):
response = response.dict()
embeddings.extend(r["embedding"] for r in response["data"])
return embeddings
# NOTE: to keep things simple, we assume the list may contain texts longer
# than the maximum context and use length-safe embedding function.
engine = cast(str, self.deployment)
return await self._aget_len_safe_embeddings(texts, engine=engine)
[docs]
def embed_query(self, text: str) -> List[float]:
"""Call out to OpenAI's embedding endpoint for embedding query text.
Args:
text: The text to embed.
Returns:
Embedding for the text.
"""
return self.embed_documents([text])[0]
[docs]
async def aembed_query(self, text: str) -> List[float]:
"""Call out to OpenAI's embedding endpoint async for embedding query text.
Args:
text: The text to embed.
Returns:
Embedding for the text.
"""
embeddings = await self.aembed_documents([text])
return embeddings[0]
