`content_image_url` throws error in local vision model

Hi- thanks for creating this package! I'm testing out some local vision workflows with ollama and llama3.2-vision. I seem to get an error when using the content_image_url function to fetch an image:

from chatlas import ChatOllama, content_image_url, content_image_file

# Initialize chat with ollama llama3.2-vision
chat = ChatOllama(model="llama3.2-vision")

# Using content_image_url
chat.chat(
    "What do you see in this image?",
    content_image_url("https://www.python.org/static/img/python-logo.png"),
    echo="all"
)

Throws this error:

                                                           👤 User turn:

What do you see in this image?

<< 👤 other content >>

🌆 python-logo.png


                                                        🤖 Assistant turn:
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "\.venv\lib\site-packages\chatlas\_chat.py", line 356, in chat
    for _ in response:
  File "\.venv\lib\site-packages\chatlas\_chat.py", line 1169, in __next__
    chunk = next(self._generator)
  File "\.venv\lib\site-packages\chatlas\_chat.py", line 833, in _chat_impl
    for chunk in self._submit_turns(
  File "\.venv\lib\site-packages\chatlas\_chat.py", line 884, in _submit_turns
    response = self.provider.chat_perform(
  File "\.venv\lib\site-packages\chatlas\_openai.py", line 247, in chat_perform
    return self._client.chat.completions.create(**kwargs)  # type: ignore
  File "\.venv\lib\site-packages\openai\_utils\_utils.py", line 275, in wrapper
    return func(*args, **kwargs)
  File "\.venv\lib\site-packages\openai\resources\chat\completions.py", line 829, in create
    return self._post(
  File "\.venv\lib\site-packages\openai\_base_client.py", line 1280, in post
    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
  File "\.venv\lib\site-packages\openai\_base_client.py", line 957, in request
    return self._request(
  File "\.venv\lib\site-packages\openai\_base_client.py", line 1061, in _request
    raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': 'invalid image input', 'type': 'invalid_request_error', 'param': None, 'code': None}}

The content_image_url("https://www.python.org/static/img/python-logo.png") seems to fetch the image and store it as a ContentImageRemote object.

Also, this works fine with content_image_file function:

chat.chat(
    "What do you see in this image?",
    content_image_file("python-logo.png"),
    echo="all"
)

                                👤 User turn:                                                                
                                                                                                             
What do you see in this image?                                                                               
                                                                                                             
<< 👤 other content >>                                                                                       
                                                                                                             
🌆 Dw2mJn+YhmuMAAAAAElFTkSuQmCC                                                                              
                                                                                                             
                                                                                                             
                              🤖 Assistant turn:                                                             
                                                                                                             
The image features the Python logo, which is a stylized representation of the                                
word "python" and serves as an emblem for the official implementation of the                                 
Python programming language.                                                                                 
                                                                                                             
Key Features:                                                                                                
                                                                                                             
 • Color scheme:                                                                                             
    • Primary color: Blue                                                                                    
    • Secondary color: Yellow                                                                                
 • Positioning: Left side of the screen                                                                      
                                                                                                             
Purpose: The logo is used to represent the official Python programming                                       
language, providing a clear visual identifier for users.    

I've tested this with version 0.2.0 and version 0.2.1.dev1+g6e81eb3 of chatlas.

Start with content_image_url and the ContentImageRemote path, then trace the request through chatlas/_chat.py and chatlas/_openai.py. Compare the remote-image request with the working content_image_file example and reproduce the Ollama llama3.2-vision call; done means the URL image is accepted without the invalid image input error.