Add qqa (GPU-parallel Quasi-Quantum Annealing) to Code Repositories

[Yuma-Ichikawa]

This PR adds qqa to the Other Resources → Code Repositories section. It is a maintained, pip-installable PyTorch toolkit for combinatorial optimisation that complements the existing entries (cvxpy, MIPLearn, BOTorch, EvoTorch, TorchOpt, ...) by covering the unsupervised-learning solver for QUBO / Ising / spin-glass problems niche, which is currently empty in the list.

Highlights:

• GPU-parallel Quasi-Quantum Annealing (QQA / PQQA) solver — Ichikawa & Arai, ICLR 2025.
• PI-GNN / CRA-PI-GNN / CPRA graph-neural-network backends.
• Simulated Annealing baseline (Glauber-style fast path for QUBO + generic single-spin fallback).
• 17 problem classes (MIS, MaxCut, Coloring, Edwards–Anderson, Sherrington–Kirkpatrick, binary perceptron, Hopfield memory, ...).
• CLI + interactive Streamlit dashboard + MkDocs Material docs.
• BSD-3-Clause-Clear, on PyPI (pip install qqa), CI / lint / coverage all green.
• Software DOI: 10.5281/zenodo.19648231.

Happy to adjust description or formatting per your style preferences — let me know!