qoro / pushquantum-2026-leaderboard
Public| Problem | Instances | Binary variables |
|---|---|---|
| Portfolio optimisation (QOBLIB 06) | po_a003_t02_orig to po_a400_t15_s02, each at QOBLIB's budget for its asset count (4 for 10 assets, 20 for 50, 50 for 200, 100 for 400) and with any λ on QOBLIB's grid: 0, 1e-6, 1e-5, 5e-5, 1e-4, 5e-4, 1e-3 or 1e-2 | 58 to about 36 000 |
| Network design (QOBLIB 08) | network05 to network24 | 20 to 552 |
qoblib_tools.py, recomputes QOBLIB's official objective from your
solution file, and a solution that breaks a rule doesn't count. Lower is better for both problems.
Every result is shown with its
.sol files with notebook 01 or 02 of the kit. Keep the instance, budget and
lambda lines of a portfolio file: they say which instance it solves.pushquantum-2026-qoro of your own aqora store, under its file name.
Beside it, put a .json file of the same name with the qubits of your largest quantum program and a
few words on your method:
import json
from pathlib import Path
import aqora
store = aqora.Store()
name = "network12" # networkNN, or po_<instance>_b<budget>_l<lambda>
store.put(f"pushquantum-2026-qoro/{name}.sol", Path(f"{name}.sol").read_bytes())
store.put(
f"pushquantum-2026-qoro/{name}.json",
json.dumps({"qubits": 30, "method": "partitioned QAOA, p = 2, + local search"}).encode(),
content_type="application/json",
)
.sol file in the folder is
scored, up to 50 of them.pushquantum-2026-qoro of your aqora
store. It only asks to read your storage, and it keeps a copy of exactly what it scored.qoblib_tools.py,
which sits unchanged next to this notebook. It is the same file the problem notebooks download, so
you can check a solution yourself before you submit it, as the kit's README shows under Scoring.