Seed data collected from technical Q&A platforms, programming communities, and command-line instruction resources, used as source material for the SETA-Synth pipeline.
This repository contains the seed-side inputs for synthesizing terminal-agent tasks and environments.
The dataset is organised by source, then by seed ID or source-specific item ID:
{source}/
└── {item_id}/
├── main.json # seed record for Q&A and… See the full description on the dataset page:
https://huggingface.co/datasets/AnonymousSubmissionUnderDouble-BlindRevi/Anonymous_Submission_Seed.
