Datasets:

camel-ai
/

seta-env-seed2synth-seed

license: other
task_categories:
  - question-answering
  - text-generation
language:
  - en
tags:
  - seed-data
  - stack-overflow
  - unix-linux
  - terminal
  - shell
  - synthetic-data
pretty_name: SETA Env Seed-to-Synth Seed Data
size_categories:
  - 10K<n<100K

SETA Env Seed-to-Synth Seed Data

Seed data collected from technical Q&A platforms and programming communities, used as source material for the SETA (Synthetic Environment Terminal Agent) data synthesis pipeline.

Dataset Structure

The dataset is organised by source, then by seed ID:

{source}/
└── {seed_id}/
    ├── main.json        # primary Q&A pair with metadata
    ├── related_1.json   # related question/post #1
    ├── related_2.json   # related question/post #2
    └── related_3.json   # related question/post #3

Sources

Source	Description
`stack_overflow`	Programming Q&A from Stack Overflow (CC BY-SA)
`unix_linux_se`	Unix & Linux Stack Exchange Q&A (CC BY-SA)

Item Format

Each main.json contains:

{
  "question_id": 27289,
  "accepted_answer_id": 27291,
  "title": "How can I run a cron command with existing environmental variables?",
  "link": "https://unix.stackexchange.com/questions/27289",
  "score": 161,
  "content_license": "CC BY-SA 3.0",
  "tags": ["shell", "environment-variables", "cron"],
  "question_text_markdown": "...",
  "question_text": "...",
  "answer_text_markdown": "...",
  "answer_text": "...",
  "category": "shell-scripting"
}

Usage

Download a specific source

Warning: Sources can be large. Always download only what you need.

Use the provided download_from_hf.py script:

# List available sources on the remote
python download_from_hf.py --list

# Download a single source
python download_from_hf.py unix_linux_se

# Download multiple sources
python download_from_hf.py unix_linux_se,stack_overflow

# Download a specific seed item only
python download_from_hf.py unix_linux_se/27289

# Download to a custom directory
python download_from_hf.py unix_linux_se --seed-data-dir /path/to/output

# Dry run (preview without downloading)
python download_from_hf.py unix_linux_se --dry-run

Or with the huggingface_hub Python library directly:

from huggingface_hub import snapshot_download

# Download a single source
snapshot_download(
    repo_id="camel-ai/seta-env-seed2synth-seed",
    repo_type="dataset",
    allow_patterns="unix_linux_se/**",
    local_dir="./seed_data",
)

# Download a single seed item
snapshot_download(
    repo_id="camel-ai/seta-env-seed2synth-seed",
    repo_type="dataset",
    allow_patterns="unix_linux_se/27289/**",
    local_dir="./seed_data",
)

Upload a specific source

Warning: Do not upload the entire dataset at once — upload only the sources or items you have collected or updated.

Use upload_to_hf.py:

# List available local sources
python upload_to_hf.py --list

# Upload a single source
python upload_to_hf.py unix_linux_se

# Upload multiple sources
python upload_to_hf.py unix_linux_se,stack_overflow

# Upload a specific seed item only
python upload_to_hf.py unix_linux_se/27289

# Dry run (preview without uploading)
python upload_to_hf.py unix_linux_se --dry-run

Uploads use upload_large_folder which is resumable — if interrupted, re-running the same command will skip already-uploaded files.

Note: Uploading merges with existing remote content. Files already on the remote that are not in your local folder are left untouched.

License

Content is sourced from Stack Overflow and Unix & Linux Stack Exchange under CC BY-SA.