HuggingfacePipeline - ValueError: alternative_import must be a fully qualified module path

[ErikTromp]

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_community.llms.huggingface_pipeline import HuggingFacePipeline
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

model_id = 'meta-llama/Meta-Llama-3-8B-Instruct'
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=10)
hf = HuggingFacePipeline(pipeline=pipe)```

### Error Message and Stack Trace (if applicable)

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/e/anaconda3/envs/rag/lib/python3.11/site-packages/langchain_core/_api/deprecation.py", line 182, in warn_if_direct_instance
    emit_warning()
  File "/home/e/anaconda3/envs/rag/lib/python3.11/site-packages/langchain_core/_api/deprecation.py", line 119, in emit_warning
    warn_deprecated(
  File "/home/e/anaconda3/envs/rag/lib/python3.11/site-packages/langchain_core/_api/deprecation.py", line 345, in warn_deprecated
    raise ValueError("alternative_import must be a fully qualified module path")
ValueError: alternative_import must be a fully qualified module path

### Description

Documentation shows how to use HuggingFacePipeline but using that code leads to an error. HF Pipeline can no longer be used.

### System Info

Windows
Langchain 0.2

[Aillian]

temp fix
pip install langchain-community==0.2.1 langchain-core==0.2.1

[ErikTromp]

I was already on 0.2.1 so that isn't the solution, the docs are just wrong. What did work was:

• Install langchain-huggingface==0.0.2
• Change the imports:
  from langchain_huggingface.embeddings import HuggingFaceEmbeddings
  from langchain_huggingface.llms import HuggingFacePipeline

[GMartin-dev]

I'm still having this issue langchain-huggingface pulls combination of dependencies that conflicts with pydantic version for other langchain modules, for me downgrading :
pip install langchain-community==0.2.1 langchain-core==0.2.1
Did the trick until the pydantic conflict is solved.

[ErikTromp]

For me it works with both 0.2.1 and 0.2.2 given the change of imports. Not using pydantic so can't say anything there.