Reproduce “Automated Code generation for Information Technology Tasks in YAML through Large Language Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Saurabh Pujar, Luca Buratti, Xiaojie Guo, Nicolas Dupuis, Burn Lewis, Sahil Suneja, Atin Sood, Ganesh Nalawade and 3 more
Ansible Wisdom generates Ansible YAML tasks from natural-language task descriptions by treating the task’s name field as a prompt and completing the remaining instructions. It uses a decoder-based language model adapted with YAML and Ansible data, and an Ansible-specific prompt format. On a held-out test set, it scores BLEU 66.67, but evaluation measures reference similarity and schema, not execution.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.