github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-09-28 06:48:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

huggingface[major]: upgrade pydantic (#26048)

Browse Source
This commit is contained in:
ccurme
2024-09-04 15:08:43 -04:00
committed by GitHub
parent 63ba12d8e0
commit ed771f2d2b
10 changed files with 90 additions and 221 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
libs/partners/huggingface/langchain_huggingface/chat_models/huggingface.py
View File

@@ -29,10 +29,11 @@ from langchain_core.messages import (
ToolMessage,
)
from langchain_core.outputs import ChatGeneration, ChatResult, LLMResult
from langchain_core.pydantic_v1 import root_validator
from langchain_core.runnables import Runnable
from langchain_core.tools import BaseTool
from langchain_core.utils.function_calling import convert_to_openai_tool
from pydantic import model_validator
from typing_extensions import Self
from langchain_huggingface.llms.huggingface_endpoint import HuggingFaceEndpoint
from langchain_huggingface.llms.huggingface_pipeline import HuggingFacePipeline
@@ -265,7 +266,7 @@ class ChatHuggingFace(BaseChatModel):
Tool calling:
.. code-block:: python
from langchain_core.pydantic_v1 import BaseModel, Field
from pydantic import BaseModel, Field
class GetWeather(BaseModel):
'''Get the current weather in a given location'''
@@ -325,20 +326,20 @@ class ChatHuggingFace(BaseChatModel):
else self.tokenizer
)
@root_validator(pre=False, skip_on_failure=True)
def validate_llm(cls, values: dict) -> dict:
@model_validator(mode="after")
def validate_llm(self) -> Self:
if (
not _is_huggingface_hub(values["llm"])
and not _is_huggingface_textgen_inference(values["llm"])
and not _is_huggingface_endpoint(values["llm"])
and not _is_huggingface_pipeline(values["llm"])
not _is_huggingface_hub(self.llm)
and not _is_huggingface_textgen_inference(self.llm)
and not _is_huggingface_endpoint(self.llm)
and not _is_huggingface_pipeline(self.llm)
):
raise TypeError(
"Expected llm to be one of HuggingFaceTextGenInference, "
"HuggingFaceEndpoint, HuggingFaceHub, HuggingFacePipeline "
f"received {type(values['llm'])}"
f"received {type(self.llm)}"
)
return values
return self
def _create_chat_result(self, response: TGI_RESPONSE) -> ChatResult:
generations = []

10
libs/partners/huggingface/langchain_huggingface/embeddings/huggingface.py
View File

@@ -1,7 +1,7 @@
from typing import Any, Dict, List, Optional # type: ignore[import-not-found]
from langchain_core.embeddings import Embeddings
from langchain_core.pydantic_v1 import BaseModel, Field
from pydantic import BaseModel, ConfigDict, Field
DEFAULT_MODEL_NAME = "sentence-transformers/all-mpnet-base-v2"
@@ -62,10 +62,10 @@ class HuggingFaceEmbeddings(BaseModel, Embeddings):
self.model_name, cache_folder=self.cache_folder, **self.model_kwargs
)
class Config:
"""Configuration for this pydantic object."""
extra = "forbid"
model_config = ConfigDict(
extra="forbid",
protected_namespaces=(),
)
def embed_documents(self, texts: List[str]) -> List[List[float]]:
"""Compute doc embeddings using a HuggingFace transformer model.

58
libs/partners/huggingface/langchain_huggingface/embeddings/huggingface_endpoint.py
View File

@@ -1,9 +1,11 @@
import json
from typing import Any, Dict, List, Optional
import os
from typing import Any, List, Optional
from langchain_core.embeddings import Embeddings
from langchain_core.pydantic_v1 import BaseModel, root_validator
from langchain_core.utils import get_from_dict_or_env
from langchain_core.utils import from_env
from pydantic import BaseModel, ConfigDict, Field, model_validator
from typing_extensions import Self
DEFAULT_MODEL = "sentence-transformers/all-mpnet-base-v2"
VALID_TASKS = ("feature-extraction",)
@@ -39,22 +41,20 @@ class HuggingFaceEndpointEmbeddings(BaseModel, Embeddings):
model_kwargs: Optional[dict] = None
"""Keyword arguments to pass to the model."""
huggingfacehub_api_token: Optional[str] = None
huggingfacehub_api_token: Optional[str] = Field(
default_factory=from_env("HUGGINGFACEHUB_API_TOKEN", default=None)
)
class Config:
"""Configuration for this pydantic object."""
model_config = ConfigDict(
extra="forbid",
protected_namespaces=(),
)
extra = "forbid"
@root_validator(pre=False, skip_on_failure=True)
def validate_environment(cls, values: Dict) -> Dict:
@model_validator(mode="after")
def validate_environment(self) -> Self:
"""Validate that api key and python package exists in environment."""
values["huggingfacehub_api_token"] = get_from_dict_or_env(
values, "huggingfacehub_api_token", "HUGGINGFACEHUB_API_TOKEN", None
)
huggingfacehub_api_token = get_from_dict_or_env(
values, "huggingfacehub_api_token", "HF_TOKEN", None
huggingfacehub_api_token = self.huggingfacehub_api_token or os.getenv(
"HF_TOKEN"
)
try:
@@ -63,38 +63,38 @@ class HuggingFaceEndpointEmbeddings(BaseModel, Embeddings):
InferenceClient,
)
if values["model"]:
values["repo_id"] = values["model"]
elif values["repo_id"]:
values["model"] = values["repo_id"]
if self.model:
self.repo_id = self.model
elif self.repo_id:
self.model = self.repo_id
else:
values["model"] = DEFAULT_MODEL
values["repo_id"] = DEFAULT_MODEL
self.model = DEFAULT_MODEL
self.repo_id = DEFAULT_MODEL
client = InferenceClient(
model=values["model"],
model=self.model,
token=huggingfacehub_api_token,
)
async_client = AsyncInferenceClient(
model=values["model"],
model=self.model,
token=huggingfacehub_api_token,
)
if values["task"] not in VALID_TASKS:
if self.task not in VALID_TASKS:
raise ValueError(
f"Got invalid task {values['task']}, "
f"Got invalid task {self.task}, "
f"currently only {VALID_TASKS} are supported"
)
values["client"] = client
values["async_client"] = async_client
self.client = client
self.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
return self
def embed_documents(self, texts: List[str]) -> List[List[float]]:
"""Call out to HuggingFaceHub's embedding endpoint for embedding search docs.

51
libs/partners/huggingface/langchain_huggingface/llms/huggingface_endpoint.py
View File

@@ -9,8 +9,9 @@ from langchain_core.callbacks import (
)
from langchain_core.language_models.llms import LLM
from langchain_core.outputs import GenerationChunk
from langchain_core.pydantic_v1 import Field, root_validator
from langchain_core.utils import get_from_dict_or_env, get_pydantic_field_names
from langchain_core.utils import from_env, get_pydantic_field_names
from pydantic import ConfigDict, Field, model_validator
from typing_extensions import Self
logger = logging.getLogger(__name__)
@@ -71,7 +72,9 @@ class HuggingFaceEndpoint(LLM):
should be pass as env variable in `HF_INFERENCE_ENDPOINT`"""
repo_id: Optional[str] = None
"""Repo to use. If endpoint_url is not specified then this needs to given"""
huggingfacehub_api_token: Optional[str] = None
huggingfacehub_api_token: Optional[str] = Field(
default_factory=from_env("HUGGINGFACEHUB_API_TOKEN", default=None)
)
max_new_tokens: int = 512
"""Maximum number of generated tokens"""
top_k: Optional[int] = None
@@ -118,13 +121,13 @@ class HuggingFaceEndpoint(LLM):
"""Task to call the model with.
Should be a task that returns `generated_text` or `summary_text`."""
class Config:
"""Configuration for this pydantic object."""
model_config = ConfigDict(
extra="forbid",
)
extra = "forbid"
@root_validator(pre=True)
def build_extra(cls, values: Dict[str, Any]) -> Dict[str, Any]:
@model_validator(mode="before")
@classmethod
def build_extra(cls, values: Dict[str, Any]) -> 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", {})
@@ -182,8 +185,8 @@ class HuggingFaceEndpoint(LLM):
)
return values
@root_validator(pre=False, skip_on_failure=True)
def validate_environment(cls, values: Dict) -> Dict:
@model_validator(mode="after")
def validate_environment(self) -> Self:
"""Validate that package is installed and that the API token is valid."""
try:
from huggingface_hub import login # type: ignore[import]
@@ -194,12 +197,8 @@ class HuggingFaceEndpoint(LLM):
"Please install it with `pip install huggingface_hub`."
)
values["huggingfacehub_api_token"] = get_from_dict_or_env(
values, "huggingfacehub_api_token", "HUGGINGFACEHUB_API_TOKEN", None
)
huggingfacehub_api_token = get_from_dict_or_env(
values, "huggingfacehub_api_token", "HF_TOKEN", None
huggingfacehub_api_token = self.huggingfacehub_api_token or os.getenv(
"HF_TOKEN"
)
if huggingfacehub_api_token is not None:
@@ -213,20 +212,20 @@ class HuggingFaceEndpoint(LLM):
from huggingface_hub import AsyncInferenceClient, InferenceClient
values["client"] = InferenceClient(
model=values["model"],
timeout=values["timeout"],
self.client = InferenceClient(
model=self.model,
timeout=self.timeout,
token=huggingfacehub_api_token,
**values["server_kwargs"],
**self.server_kwargs,
)
values["async_client"] = AsyncInferenceClient(
model=values["model"],
timeout=values["timeout"],
self.async_client = AsyncInferenceClient(
model=self.model,
timeout=self.timeout,
token=huggingfacehub_api_token,
**values["server_kwargs"],
**self.server_kwargs,
)
return values
return self
@property
def _default_params(self) -> Dict[str, Any]:

8
libs/partners/huggingface/langchain_huggingface/llms/huggingface_pipeline.py
View File

@@ -7,6 +7,7 @@ from typing import Any, Iterator, List, Mapping, Optional
from langchain_core.callbacks import CallbackManagerForLLMRun
from langchain_core.language_models.llms import BaseLLM
from langchain_core.outputs import Generation, GenerationChunk, LLMResult
from pydantic import ConfigDict
DEFAULT_MODEL_ID = "gpt2"
DEFAULT_TASK = "text-generation"
@@ -63,10 +64,9 @@ class HuggingFacePipeline(BaseLLM):
batch_size: int = DEFAULT_BATCH_SIZE
"""Batch size to use when passing multiple documents to generate."""
class Config:
"""Configuration for this pydantic object."""
extra = "forbid"
model_config = ConfigDict(
extra="forbid",
)
@classmethod
def from_model_id(