Source code for langchain_community.embeddings.huggingface_hub
import json
import os
from typing import Any, Dict, List, Optional
from langchain_core.embeddings import Embeddings
from langchain_core.pydantic_v1 import BaseModel, Extra, root_validator
DEFAULT_MODEL = "sentence-transformers/all-mpnet-base-v2"
VALID_TASKS = ("feature-extraction",)
[docs]class HuggingFaceHubEmbeddings(BaseModel, Embeddings):
"""HuggingFaceHub嵌入模型。
要使用,您应该已安装``huggingface_hub`` python包,并且
将环境变量``HUGGINGFACEHUB_API_TOKEN``设置为您的API令牌,或者传递
它作为构造函数的一个命名参数。
示例:
.. code-block:: python
from langchain_community.embeddings import HuggingFaceHubEmbeddings
model = "sentence-transformers/all-mpnet-base-v2"
hf = HuggingFaceHubEmbeddings(
model=model,
task="feature-extraction",
huggingfacehub_api_token="my-api-key",
)
"""
client: Any #: :meta private:
async_client: Any #: :meta private:
model: Optional[str] = None
"""要使用的模型名称。"""
repo_id: Optional[str] = None
"""Huggingfacehub存储库ID,用于向后兼容。"""
task: Optional[str] = "feature-extraction"
"""调用模型的任务。"""
model_kwargs: Optional[dict] = None
"""传递给模型的关键字参数。"""
huggingfacehub_api_token: Optional[str] = None
class Config:
"""此pydantic对象的配置。"""
extra = Extra.forbid
@root_validator()
def validate_environment(cls, values: Dict) -> Dict:
"""验证环境中是否存在API密钥和Python包。"""
huggingfacehub_api_token = values["huggingfacehub_api_token"] or os.getenv(
"HUGGINGFACEHUB_API_TOKEN"
)
try:
from huggingface_hub import AsyncInferenceClient, InferenceClient
if values["model"]:
values["repo_id"] = values["model"]
elif values["repo_id"]:
values["model"] = values["repo_id"]
else:
values["model"] = DEFAULT_MODEL
values["repo_id"] = DEFAULT_MODEL
client = InferenceClient(
model=values["model"],
token=huggingfacehub_api_token,
)
async_client = AsyncInferenceClient(
model=values["model"],
token=huggingfacehub_api_token,
)
if values["task"] not in VALID_TASKS:
raise ValueError(
f"Got invalid task {values['task']}, "
f"currently only {VALID_TASKS} are supported"
)
values["client"] = client
values["async_client"] = async_client
except ImportError:
raise ImportError(
"Could not import huggingface_hub python package. "
"Please install it with `pip install huggingface_hub`."
)
return values
[docs] def embed_documents(self, texts: List[str]) -> List[List[float]]:
"""调用HuggingFaceHub的嵌入端点来进行嵌入搜索文档。
参数:
texts:要嵌入的文本列表。
返回:
每个文本的嵌入列表。
"""
# replace newlines, which can negatively affect performance.
texts = [text.replace("\n", " ") for text in texts]
_model_kwargs = self.model_kwargs or {}
responses = self.client.post(
json={"inputs": texts, "parameters": _model_kwargs}, task=self.task
)
return json.loads(responses.decode())
[docs] async def aembed_documents(self, texts: List[str]) -> List[List[float]]:
"""异步调用HuggingFaceHub的嵌入端点以进行嵌入搜索文档。
参数:
texts:要嵌入的文本列表。
返回:
嵌入列表,每个文本对应一个嵌入。
"""
# replace newlines, which can negatively affect performance.
texts = [text.replace("\n", " ") for text in texts]
_model_kwargs = self.model_kwargs or {}
responses = await self.async_client.post(
json={"inputs": texts, "parameters": _model_kwargs}, task=self.task
)
return json.loads(responses.decode())
[docs] def embed_query(self, text: str) -> List[float]:
"""调用HuggingFaceHub的嵌入端点来嵌入查询文本。
参数:
text:要嵌入的文本。
返回:
文本的嵌入。
"""
response = self.embed_documents([text])[0]
return response
[docs] async def aembed_query(self, text: str) -> List[float]:
"""异步调用HuggingFaceHub的嵌入端点,用于嵌入查询文本。
参数:
text: 要嵌入的文本。
返回:
文本的嵌入。
"""
response = (await self.aembed_documents([text]))[0]
return response