Langchain x Huggingface: Critical Bug - HuggingFaceEndpoint connects to random mistral-nemo endpoint

[SMAntony]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

from langchain_huggingface import HuggingFaceEmbeddings

endpoint_url = "https://www.google.com/" #or any random endpoint

llm = HuggingFaceEndpoint(
    name="chat_agent",
    endpoint_url=endpoint_url,
    task="text-generation",
    # callbacks=common_callbacks + self.callbacks,
    # streaming=streaming,
    # stop_sequences=self.stop_sequences,
    seed=None,
    # **self.generation_config,
)

llm.invoke("hi") # goes to a api-inference huggingface Mistral-Nemo-Instruct endpoint
# Not sure if it is designed to be accessed.
# \.venv\Lib\site-packages\huggingface_hub\utils\_deprecation.py:131: FutureWarning: 'post' (from 'huggingface_hub.inference._client') is deprecated and will be removed from version '0.31.0'. Making direct POST requests to the inference server is not supported anymore. Please use task methods instead (e.g. `InferenceClient.chat_completion`). If your use case is not supported, please open an issue in https://github.com/huggingface/huggingface_hub.
#  warnings.warn(warning_message, FutureWarning)
# '\ndef create():\n  while True:\n    name = input("Enter your name: ")\n    if name.isalpha():\n      print(f"Hello, {name}!")\n      break\n    else:\n      print("Invalid name. Please enter a name with only letters.")\n\ncreate()'

Error Message and Stack Trace (if applicable)

No response

Description

• I am trying to connect to the endpoint url and infer from it
• But it is connecting to some different huggingface inference api when calling llm.invoke()
• This shouldn't be the expected behavior

System Info

System Information
------------------
> OS:  Windows
> OS Version:  10.0.26100
> Python Version:  3.11.9 (tags/v3.11.9:de54cf5, Apr  2 2024, 10:12:12) [MSC v.1938 64 bit (AMD64)]

Package Information
-------------------
> langchain_core: 0.3.33
> langchain: 0.3.17
> langchain_community: 0.3.16
> langsmith: 0.3.4
> langchain_chroma: 0.1.4
> langchain_experimental: 0.3.4
> langchain_huggingface: 0.1.2
> langchain_openai: 0.2.0
> langchain_qdrant: 0.1.4
> langchain_text_splitters: 0.3.5
> langgraph_sdk: 0.1.51

Optional packages not installed
-------------------------------
> langserve

Other Dependencies
------------------
> aiohttp: 3.11.11
> async-timeout: 5.0.1
> chromadb: 0.5.20
> dataclasses-json: 0.6.7
> fastapi: 0.111.0
> fastembed: Installed. No version info available.
> httpx: 0.28.1
> httpx-sse: 0.4.0
> huggingface-hub: 0.28.1
> jsonpatch: 1.33
> langsmith-pyo3: Installed. No version info available.
> numpy: 1.26.4
> openai: 1.61.0
> orjson: 3.10.15
> packaging: 24.2
> pydantic: 2.10.6
> pydantic-settings: 2.7.1
> pytest: Installed. No version info available.
> PyYAML: 6.0.2
> qdrant-client: 1.13.2
> requests: 2.32.3
> requests-toolbelt: 1.0.0
> rich: 13.9.4
> sentence-transformers: 3.4.1
> SQLAlchemy: 2.0.37
> tenacity: 9.0.0
> tiktoken: 0.8.0
> tokenizers: 0.21.0
> transformers: 4.48.2
> typing-extensions: 4.12.2
> zstandard: 0.23.0

[SMAntony]

I think this is a issue with huggingface_hub. After downgrading to hf_hub version 0.27.0 (from latest), this issue is fixed.