AutomatedScientist/jupyter-interleaved-tool-call

Aurora-M manual pipeline This repo is a small, deterministic data pipeline for the Aurora-M notebook corpus. It reads JSONL records, converts markdown fenced code blocks into explicit tool-call markers, and then scores how parseable those tool-call blocks are. Core constraints (final behavior): No Docker, no uv, and no execution of extracted code. Scoring uses ast.parse only. Work is streaming/row-based and resilient: malformed rows/snippets should not kill the full run.… See the full description on the dataset page: https://huggingface.co/datasets/AutomatedScientist/jupyter-interleaved-tool-c

<iframe src="https://silytics.si/embed/item/4b33246b-fb56-4258-9a0a-f4fcfddc49dc" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.