1. Skip to content

1. Choose your path

Use this guide to pick the fastest entrypoint for your goal. ModSSC supports several workflows: package-level Python APIs, CLI bricks, end-to-end benchmark configs, example scripts, and interactive notebooks.

1.1 Fastest routes by goal

Goal Start here Then continue with
First successful install and run Installation and Quickstart Inductive tutorial or Transductive tutorial
Learn the concepts before running code Concepts Example scripts
Use the Python API directly Quickstart Example scripts, Notebook tour, API reference
Run reproducible benchmarks from YAML configs Benchmarks Configuration reference, Bench config cookbook, Reproducibility
Set up modality-specific dependencies Optional extras and platform support Manage datasets, Run preprocessing plans, Build graphs and views
Work interactively in notebooks Notebook tour The matching how-to page for the same brick
Extend ModSSC with new datasets or methods Architecture Contributing and development, Catalogs and registries

If you cloned the repository and want the broadest surface area available locally:

python -m pip install -e "."
python -m modssc --help
python -m bench.main --config bench/configs/experiments/toy_inductive.yaml

If you prefer a lighter Python-first path:

python -m pip install modssc
python examples/00_inductive_toy_pseudo_label.py

Use a source install when you need the benchmark runner, authored benchmark configs, notebooks, or example scripts from the repository. Use the PyPI install when you only need the packaged library and CLI.

1.3 Which learning format should you choose

  • Use the tutorials when you want a full end-to-end run with one opinionated config.
  • Use the how-to guides when you want to understand one brick in isolation.
  • Use the example scripts when you want copy-pasteable Python files.
  • Use the notebooks when you want an interactive, exploratory workflow.
  • Use the reference when you already know what you want and only need precise command or schema details.
Sources
  1. README.md
  2. mkdocs.yml
  3. examples/
  4. notebooks/
  5. bench/configs/experiments/minimal/README.md
  6. bench/configs/experiments/toy_inductive.yaml
  7. bench/configs/experiments/toy_transductive.yaml