Skip to content

Added information and error message for Python 3.14 users facing mark… #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
havelentzas wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Added information and error message for Python 3.14 users facing mark… #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ The 7 stages that genuinely need page rasters — the math chain (`01e`, `01e2`,

## Requirements

- Python 3.10+
- Python 3.10+ (Python 3.14 users: PDF input is broken; see note below)
- An Anthropic API key ([console.anthropic.com](https://console.anthropic.com/))
- `marker-pdf` (installed automatically via `pip install -e .`; pulls in PyTorch and a few GB of ML models on first use — see "PDF handling and cost" above)
- `qpdf` on `PATH` for PDF preprocessing (optional; python fallback exists)
Expand All @@ -53,6 +53,8 @@ pip install -e .

Initial install takes 5–10 minutes because of `marker-pdf` and its PyTorch dependency. The first PDF conversion downloads model weights (~3–5 GB) the first time.

> **Python 3.14: PDF input is broken; `.md` / `.tex` input works fine.** `marker-pdf` depends on `surya-ocr` 0.12.x, which was built against `transformers` 4.43.x. On Python 3.14, `tokenizers` 0.20.x (required to downgrade transformers) cannot be compiled from source because `pyo3` 0.22.5 only supports up to Python 3.13. The `transformers` version that installs in a Python 3.14 venv (4.57+) has breaking changes for surya's composition config classes, causing `KeyError: 'encoder'` and subsequent errors during PDF conversion. Workaround: convert your PDF to markdown first (e.g. with `pymupdf4llm`) and pass the `.md` file — the pipeline accepts `.md` / `.markdown` / `.txt` / `.tex` directly and skips marker entirely. To use PDF input, run on Python 3.12 or 3.13.

## API key setup

`presubmit` calls the Anthropic API directly via the official Python SDK. It does **not** authenticate via the `claude` CLI's OAuth subscription or via your claude.ai login — those are different auth surfaces. You need a personal API key on your Anthropic account.
Expand Down
1 change: 1 addition & 0 deletions pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@ classifiers = [
"Programming Language :: Python :: 3.10",
"Programming Language :: Python :: 3.11",
"Programming Language :: Python :: 3.12",
"Programming Language :: Python :: 3.13",
"Topic :: Scientific/Engineering",
"Topic :: Text Processing",
]
Expand Down
9 changes: 8 additions & 1 deletion src/presubmit/pipeline.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -188,10 +188,17 @@ def _marker_convert(pdf_path: Path, md_path: Path) -> None:
".md / .markdown / .tex file as input to bypass marker."
) from e
except Exception as e:
py314 = sys.version_info >= (3, 14)
hint = (
" PDF input is broken on Python 3.14 — pass a .md file instead, "
"or downgrade to Python 3.12/3.13. See README."
) if py314 else (
" Pass a pre-converted .md file to bypass marker."
)
raise PipelineError(
f"marker-pdf failed converting {pdf_path}. "
f"Latest error: {e}. "
f"Earlier (<=0.x API) error: {last_err}."
f"Earlier (<=0.x API) error: {last_err}.{hint}"
) from e


Expand Down