Spaces:
Runtime error
Runtime error
File size: 5,306 Bytes
ed4d993 |
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 |
from typing import Any, Dict, List, Optional
from langchain_core.embeddings import Embeddings
from langchain_core.pydantic_v1 import BaseModel, Extra, root_validator
from langchain_core.utils import get_from_dict_or_env
from packaging.version import parse
__all__ = ["GradientEmbeddings"]
class GradientEmbeddings(BaseModel, Embeddings):
"""Gradient.ai Embedding models.
GradientLLM is a class to interact with Embedding Models on gradient.ai
To use, set the environment variable ``GRADIENT_ACCESS_TOKEN`` with your
API token and ``GRADIENT_WORKSPACE_ID`` for your gradient workspace,
or alternatively provide them as keywords to the constructor of this class.
Example:
.. code-block:: python
from langchain_community.embeddings import GradientEmbeddings
GradientEmbeddings(
model="bge-large",
gradient_workspace_id="12345614fc0_workspace",
gradient_access_token="gradientai-access_token",
)
"""
model: str
"Underlying gradient.ai model id."
gradient_workspace_id: Optional[str] = None
"Underlying gradient.ai workspace_id."
gradient_access_token: Optional[str] = None
"""gradient.ai API Token, which can be generated by going to
https://auth.gradient.ai/select-workspace
and selecting "Access tokens" under the profile drop-down.
"""
gradient_api_url: str = "https://api.gradient.ai/api"
"""Endpoint URL to use."""
query_prompt_for_retrieval: Optional[str] = None
"""Query pre-prompt"""
client: Any = None #: :meta private:
"""Gradient client."""
# LLM call kwargs
class Config:
"""Configuration for this pydantic object."""
extra = Extra.forbid
@root_validator(allow_reuse=True)
def validate_environment(cls, values: Dict) -> Dict:
"""Validate that api key and python package exists in environment."""
values["gradient_access_token"] = get_from_dict_or_env(
values, "gradient_access_token", "GRADIENT_ACCESS_TOKEN"
)
values["gradient_workspace_id"] = get_from_dict_or_env(
values, "gradient_workspace_id", "GRADIENT_WORKSPACE_ID"
)
values["gradient_api_url"] = get_from_dict_or_env(
values, "gradient_api_url", "GRADIENT_API_URL"
)
try:
import gradientai
except ImportError:
raise ImportError(
'GradientEmbeddings requires `pip install -U "gradientai>=1.4.0"`.'
)
if parse(gradientai.__version__) < parse("1.4.0"):
raise ImportError(
'GradientEmbeddings requires `pip install -U "gradientai>=1.4.0"`.'
)
gradient = gradientai.Gradient(
access_token=values["gradient_access_token"],
workspace_id=values["gradient_workspace_id"],
host=values["gradient_api_url"],
)
values["client"] = gradient.get_embeddings_model(slug=values["model"])
return values
def embed_documents(self, texts: List[str]) -> List[List[float]]:
"""Call out to Gradient's embedding endpoint.
Args:
texts: The list of texts to embed.
Returns:
List of embeddings, one for each text.
"""
inputs = [{"input": text} for text in texts]
result = self.client.embed(inputs=inputs).embeddings
return [e.embedding for e in result]
async def aembed_documents(self, texts: List[str]) -> List[List[float]]:
"""Async call out to Gradient's embedding endpoint.
Args:
texts: The list of texts to embed.
Returns:
List of embeddings, one for each text.
"""
inputs = [{"input": text} for text in texts]
result = (await self.client.aembed(inputs=inputs)).embeddings
return [e.embedding for e in result]
def embed_query(self, text: str) -> List[float]:
"""Call out to Gradient's embedding endpoint.
Args:
text: The text to embed.
Returns:
Embeddings for the text.
"""
query = (
f"{self.query_prompt_for_retrieval} {text}"
if self.query_prompt_for_retrieval
else text
)
return self.embed_documents([query])[0]
async def aembed_query(self, text: str) -> List[float]:
"""Async call out to Gradient's embedding endpoint.
Args:
text: The text to embed.
Returns:
Embeddings for the text.
"""
query = (
f"{self.query_prompt_for_retrieval} {text}"
if self.query_prompt_for_retrieval
else text
)
embeddings = await self.aembed_documents([query])
return embeddings[0]
class TinyAsyncGradientEmbeddingClient: #: :meta private:
"""Deprecated, TinyAsyncGradientEmbeddingClient was removed.
This class is just for backwards compatibility with older versions
of langchain_community.
It might be entirely removed in the future.
"""
def __init__(self, *args, **kwargs) -> None: # type: ignore[no-untyped-def]
raise ValueError("Deprecated,TinyAsyncGradientEmbeddingClient was removed.")
|