Error when testing code example for megaparsevision with Claude 3.5 Sonnet

[OSS-GR]

I was testing the megaparsevision code and I used Claude 3.5 sonnet and it spit out this error

You may be able to resolve this warning by setting `model_config['protected_namespaces'] = ()`.
  warnings.warn(

Traceback (most recent call last):
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/megaparse/core/megaparse.py", line 91, in load
    parsed_document: str = loop.run_until_complete(
                           ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/.pyenv/versions/3.12.0/lib/python3.12/asyncio/base_events.py", line 664, in run_until_complete
    return future.result()
           ^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/megaparse/core/parser/megaparse_vision.py", line 141, in convert
    self.parsed_chunks = await asyncio.gather(*tasks)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/megaparse/core/parser/megaparse_vision.py", line 114, in send_to_mlm
    response = await self.model.ainvoke([message])
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 307, in ainvoke
    llm_result = await self.agenerate_prompt(
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 796, in agenerate_prompt
    return await self.agenerate(
           ^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 756, in agenerate
    raise exceptions[0]
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 924, in _agenerate_with_cache
    result = await self._agenerate(
             ^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/langchain_anthropic/chat_models.py", line 812, in _agenerate
    data = await self._async_client.messages.create(**payload)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/anthropic/resources/messages.py", line 1817, in create
    return await self._post(
           ^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/anthropic/_base_client.py", line 1842, in post
    return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/anthropic/_base_client.py", line 1536, in request
    return await self._request(
           ^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/anthropic/_base_client.py", line 1637, in _request
    raise self._make_status_error_from_response(err.response) from None
anthropic.BadRequestError: Error code: 400 - {'type': 'error', 'error': {'type': 'invalid_request_error', 'message': 'messages.0.content.3.image.source.base64.data: The image specified was not valid'}}

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/Users/orestis/Repos/test_megaparser/basic_test.py", line 15, in <module>
    response = megaparse.load("./monthly-budget-breakdown-revised.pdf")
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/orestis/Repos/test_megaparser/megaparser/lib/python3.12/site-packages/megaparse/core/megaparse.py", line 101, in load
    raise ValueError(f"Error while parsing {file_path}: {e}")
ValueError: Error while parsing monthly-budget-breakdown-revised.pdf: Error code: 400 - {'type': 'error', 'error': {'type': 'invalid_request_error', 'message': 'messages.0.content.3.image.source.base64.data: The image specified was not valid'}}

Here is the code itself:

from langchain_anthropic import ChatAnthropic
from megaparse.core.parser.unstructured_parser import UnstructuredParser
from megaparse.core.parser.megaparse_vision import MegaParseVision
import os

# Initialize the language model
model = ChatAnthropic(model="claude-3-5-sonnet-20241022", api_key=os.getenv("ANTHROPIC_API_KEY"))

# Set up the parser
parser = MegaParseVision(model=model)
megaparse = MegaParse(parser)

# Load and process the document
response = megaparse.load("./monthly-budget-breakdown-revised.pdf")
print(response)

# Save the processed content to a markdown file
megaparse.save("./test.md")```


It worked with the UnstructuredParser so there is no issue with the API key as far as I can tell.

[eridgd]

Anthropic requires images to be in JPEG. Make this change in:

MegaParse/libs/megaparse/src/megaparse/parser/megaparse_vision.py

Lines 71 to 89 in 32101e9

	def process_file(self, file_path: str, image_format: str = "PNG") -> List[str]:
	"""
	Process a PDF file and convert its pages to base64 encoded images.
	:param file_path: Path to the PDF file
	:param image_format: Format to save the images (default: PNG)
	:return: List of base64 encoded images
	"""
	try:
	images = convert_from_path(file_path)
	images_base64 = []
	for image in images:
	buffered = BytesIO()
	image.save(buffered, format=image_format)
	image_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")
	images_base64.append(image_base64)
	return images_base64
	except Exception as e:
	raise ValueError(f"Error processing PDF file: {str(e)}")

def process_file(self, file_path: str, image_format: str = "JPEG") -> List[str]:
    """
    Process a PDF file and convert its pages to base64 encoded images.

    :param file_path: Path to the PDF file
    :param image_format: Format to save the images (default: JPEG)
    :return: List of base64 encoded images
    """
    try:
        images = convert_from_path(file_path)
        images_base64 = []
        for image in images:
            buffered = BytesIO()
            image.save(buffered, format=image_format)
            image_base64 = base64.b64encode(buffered.getvalue()).decode("utf-8")
            images_base64.append(image_base64)
        return images_base64
    except Exception as e:
        raise ValueError(f"Error processing PDF file: {str(e)}")