Colab-ready notebooks for the tutorial Under-Resourced Studies of Under-Resourced Languages: Practical, Reproducible LLM-as-Annotator Pipelines Across Scripts and Domains.
Run order:
00_setup_and_data.ipynb01_prompting_zero_few_shot.ipynb02_structured_outputs_and_validation.ipynb03_evaluation_and_error_analysis.ipynb04_sampling_and_bootstrapping.ipynbThe notebooks default to USE_API = False; they therefore run without API keys using deterministic fallback predictions. Replace the toy data with validated project data before using the pipeline for research.