401 Authorization Error with BentoML Cloud Extractor

[mpyo90]

Great project, thanks for your work.

When attempting to convert a PDF to EPUB using the Oomol Studio workflow, the process fails with a 401 Authorization Required error from the BentoML cloud service during the OCR analysis step.

Error Details

The error occurs in /root/.oomol-studio/registry-package-store/pdf-craft-0.2.15/shared/cloud_extractor.py at line 81:
AssertionError: Task data is None
The task receives a 401 Unauthorized response from the BentoML service:
html

401 Authorization Required

It seems there might be an issue with your access token or it may not have been set in the header.

Full Error Log

task id: 41bfa04eb2a4466eaca34683a2c66833
task failed: 41bfa04eb2a4466eaca34683a2c66833 {'data': {'task_id': '41bfa04eb2a4466eaca34683a2c66833', 'status': 'failed', 'created_at': '', 'result': '\n\n\n \n <title>401 Unauthorized</title>
...
Traceback (most recent call last):
File "/root/.local/share/uv/python/cpython-3.11.13-linux-aarch64-gnu/lib/python3.11/site-packages/pdf_craft/analysers/analyser.py", line 56, in analyse
generate_ocr_pages(
File "/root/.local/share/uv/python/cpython-3.11.13-linux-aarch64-gnu/lib/python3.11/site-packages/pdf_craft/analysers/ocr/generation.py", line 38, in generate_ocr_pages
for page_index, page_xml in extract_ocr_page_xmls(
File "/root/.oomol-studio/registry-package-store/pdf-craft-0.2.15/shared/cloud_extractor.py", line 81, in extract
assert task_data is not None, "Task data is None"
AssertionError: Task data is None

Any idea how to solve the problem ?