Datasets:
Hugging Face quickstart
The expanded repo/ folder becomes the root of PureOne/AELYRION-NOULITH-III after upload.
Download a deposited snapshot
python -m pip install "huggingface_hub>=0.35,<3"
from huggingface_hub import snapshot_download
snapshot_download(
repo_id="PureOne/AELYRION-NOULITH-III",
repo_type="dataset",
local_dir="AELYRION-NOULITH-III",
)
This remote example requires the prepared repository to have been uploaded. For a pinned
reproduction, pass revision set to the actual deposited commit hash.
Run the exact example
From the downloaded repository root:
python -m pip install -e .
python hub_tools/quickstart_exact.py
Defaults: six channels, 41 scalar components, twin rank 60 of 64, four synchronizing twin constraints, and one edge-scale class. A generic visible witness is verified in the original sparse algebra. All certificate coefficients are exact rationals.
--m selects another channel count. Above eight, the quickstart omits the exponential
independent rank calculation and reports that omission.
Run tests and exhaustive finite checks
python -m pip install -r requirements.txt
python -m pytest -q
python scripts/exhaustive_v3.py
The supplied suite has 1,028 cases. The graph script performs two distinct exhaustive six-channel enumerations and finds 810 minimum-size synchronizing families. These remain finite computational checks, separate from the general written proofs.
Recreate the notebooks and PDF
python -m pip install -r requirements-notebook.txt
python scripts/reproduce.py --notebook --skip-pdf
Both original notebooks are already executed. The command regenerates data, validation
records, and notebook outputs, intentionally changing the delivered bytes.
For the PDF, install TeX Live or MiKTeX with pdflatex and bibtex, then run:
python scripts/build_manuscript.py
The delivered PDF has 67 pages and can be read without rebuilding.
Load an individual table
python -m pip install "datasets>=4,<6"
from datasets import load_dataset
table = load_dataset(
"PureOne/AELYRION-NOULITH-III",
"v3_exact_rank_strata",
split="train",
)
See DATA_DICTIONARY.md for row semantics. All configurations have a single logical split; there is no held-out ML benchmark.
Integrity
python hub_tools/verify_release.py
After intentional edits:
python hub_tools/prepare_viewer_data.py
python hub_tools/verify_release.py --write
The manifest verifies prepared-file integrity, not authorship or mathematical truth. The original ZIP has its own unchanged internal manifest.
API
Channel labels are zero-based; coefficient keys are ascending pairs. Use exact SymPy
rationals. certify_scalar enforces m>=6 and abs(q_ij+1)<1/20 and rejects floating-point
coefficients. API_REFERENCE.md lists the existing entry points.
Check the Dataset configuration locally
After dependency installation, this uses local files and offline Hub settings:
python -m pip install -r requirements-hub-validation.txt
python hub_tools/validate_dataset_configs.py