Spaces:
Build error
Build error
File size: 4,256 Bytes
1c744c7 |
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 |
"""Wrapper around OpenAI embedding models."""
from typing import Any, Dict, List, Optional
from pydantic import BaseModel, Extra, root_validator
from langchain.embeddings.base import Embeddings
from langchain.utils import get_from_dict_or_env
from tenacity import (
retry,
retry_if_exception_type,
stop_after_attempt,
wait_exponential,
)
from openai.error import Timeout, APIError, APIConnectionError, RateLimitError
class OpenAIEmbeddings(BaseModel, Embeddings):
"""Wrapper around OpenAI embedding models.
To use, you should have the ``openai`` python package installed, and the
environment variable ``OPENAI_API_KEY`` set with your API key or pass it
as a named parameter to the constructor.
Example:
.. code-block:: python
from langchain.embeddings import OpenAIEmbeddings
openai = OpenAIEmbeddings(openai_api_key="my-api-key")
"""
client: Any #: :meta private:
document_model_name: str = "text-embedding-ada-002"
query_model_name: str = "text-embedding-ada-002"
openai_api_key: Optional[str] = None
class Config:
"""Configuration for this pydantic object."""
extra = Extra.forbid
# TODO: deprecate this
@root_validator(pre=True, allow_reuse=True)
def get_model_names(cls, values: Dict) -> Dict:
"""Get model names from just old model name."""
if "model_name" in values:
if "document_model_name" in values:
raise ValueError(
"Both `model_name` and `document_model_name` were provided, "
"but only one should be."
)
if "query_model_name" in values:
raise ValueError(
"Both `model_name` and `query_model_name` were provided, "
"but only one should be."
)
model_name = values.pop("model_name")
values["document_model_name"] = f"text-search-{model_name}-doc-001"
values["query_model_name"] = f"text-search-{model_name}-query-001"
return values
@root_validator(allow_reuse=True)
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"
)
try:
import openai
openai.api_key = openai_api_key
values["client"] = openai.Embedding
except ImportError:
raise ValueError(
"Could not import openai python package. "
"Please it install it with `pip install openai`."
)
return values
@retry(
reraise=True,
stop=stop_after_attempt(100),
wait=wait_exponential(multiplier=1, min=10, max=60),
retry=(
retry_if_exception_type(Timeout)
| retry_if_exception_type(APIError)
| retry_if_exception_type(APIConnectionError)
| retry_if_exception_type(RateLimitError)
),
)
def _embedding_func(self, text: str, *, engine: str) -> List[float]:
"""Call out to OpenAI's embedding endpoint with exponential backoff."""
# replace newlines, which can negatively affect performance.
text = text.replace("\n", " ")
return self.client.create(input=[text], engine=engine)["data"][0]["embedding"]
def embed_documents(self, texts: List[str]) -> List[List[float]]:
"""Call out to OpenAI's embedding endpoint for embedding search docs.
Args:
texts: The list of texts to embed.
Returns:
List of embeddings, one for each text.
"""
responses = [
self._embedding_func(text, engine=self.document_model_name)
for text in texts
]
return responses
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:
Embeddings for the text.
"""
embedding = self._embedding_func(text, engine=self.query_model_name)
return embedding |