[BUG]: HuggingFacePipeline not initializes correctly when passing transformers pipeline

Checked other resources

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

Example Code

from transformers import pipeline
from langchain_huggingface import HuggingFacePipeline

model_id = "microsoft/Phi-3.5-mini-instruct"
pipe = pipeline("text-generation", model=model_id)

llm = HuggingFacePipeline(pipeline=pipe)
llm.model_id # Output: 'gpt2'

While initializing the HuggingFacePipeline through model_id gives,

from langchain_huggingface import ChatHuggingFace, HuggingFacePipeline

llm = HuggingFacePipeline.from_model_id(
    model_id="microsoft/Phi-3.5-mini-instruct",
    task="text-generation",
)
llm.model_id # Output is : 'microsoft/Phi-3.5-mini-instruct'

Error Message and Stack Trace (if applicable)

No response

Description

I am using langchain, when using HuggingFacePipeline first I initialized it with custom transformers pipeline doing that results in incorrect model name as 'gpt-2'. While using the same module and initialization through .model_id() recognizes model_id correctly as phi3. Due to the model_id mismatch the other modules are getting impacted like using ChatHuggingFace because ChatHuggingFace uses model_id to initialize the tokenizer.

System Info

"pip freeze | grep langchain"

langchain==0.2.15
langchain-community==0.0.27
langchain-core==0.2.37
langchain-huggingface==0.0.3
langchain-text-splitters==0.2.2

platform: Linux Python: 3.10.6

langchain-ai / langchain