[BUG] error with embedder config using huggingface provider

[prusswan]

Description

Despite the fix in #4005, I am getting error when running crewai run using huggingface provider for embedder. Crewai version is 1.7.1/1.7.2

It is unclear what the pydantic validation error is referring to: "Error: could not convert string to float:"

Steps to Reproduce

Using the config from https://docs.crewai.com/en/concepts/memory#hugging-face-embeddings

        # ... snipped 
        return Crew(
            agents=self.agents,  # Automatically collected by the @agent decorator
            tasks=self.tasks,    # Automatically collected by the @task decorator.
            process=Process.sequential,
            verbose=True,
            memory=True,
            embedder={
                "provider":"huggingface",
                "config": {
                    #"api_key": "your-hf-token",  # Optional for public models
                    "model": "sentence-transformers/all-MiniLM-L6-v2"
                }
            }
        )

Expected behavior

crewai run should work without error

Screenshots/Code snippets

N/A

Operating System

Windows 11

Python Version

3.12

crewAI Version

1.7.1

crewAI Tools Version

1.7.1

Virtual Environment

Venv

Evidence

> crewai run
Running the Crew
Traceback (most recent call last):
  File "D:\data\crewai_ia\market-researcher\src\seo_crew\main.py", line 28, in run
    seo_crew_instance = SeoCrew().crew()
                        ^^^^^^^^^^^^^^^^
  File "D:\data\crewai_ia\market-researcher\.venv\Lib\site-packages\crewai\project\utils.py", line 70, in wrapper
    result = meth(*args, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^
  File "D:\data\crewai_ia\market-researcher\.venv\Lib\site-packages\crewai\project\annotations.py", line 237, in wrapper
    crew_instance: Crew = _call_method(meth, self, *args, **kwargs)
                          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\data\crewai_ia\market-researcher\.venv\Lib\site-packages\crewai\project\annotations.py", line 163, in _call_method
    result = method(*args, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\data\crewai_ia\market-researcher\src\seo_crew\crew.py", line 91, in crew
    return Crew(
           ^^^^^
  File "D:\data\crewai_ia\market-researcher\.venv\Lib\site-packages\pydantic\main.py", line 253, in __init__
    validated_self = self.__pydantic_validator__.validate_python(data, self_instance=self)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
pydantic_core._pydantic_core.ValidationError: 1 validation error for Crew
  Value error, Failed to initialize embedder. Please check your configuration or connection.
Provider: huggingface
Error: could not convert string to float: 'error' [type=value_error, input_value={'agents': [Agent(role=Ke...ers/all-MiniLM-L6-v2'}}}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.11/v/value_error

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "D:\data\crewai_ia\market-researcher\.venv\Scripts\run_crew.exe\__main__.py", line 10, in <module>
  File "D:\data\crewai_ia\market-researcher\src\seo_crew\main.py", line 32, in run
    raise Exception(f"An error occurred while running the crew: {e}")
Exception: An error occurred while running the crew: 1 validation error for Crew
  Value error, Failed to initialize embedder. Please check your configuration or connection.
Provider: huggingface
Error: could not convert string to float: 'error' [type=value_error, input_value={'agents': [Agent(role=Ke...ers/all-MiniLM-L6-v2'}}}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.11/v/value_error
An error occurred while running the crew: Command '['uv', 'run', 'run_crew']' returned non-zero exit status 1.

Possible Solution

It's not a solution, but the error does not appear when memory=False. I was also able to use a locally hosted embedding model using openai provider.

Additional context

none