In TSP, use spin_glass_solver(G) (for non-Monte-Carlo mode)

The Monte Carlo mode of of the TSP solvers works so well, one wouldn't use the MAXCUT-based mode unless one had a very specific reason for the problem under consideration. Then, one likely wants high-quality MAXCUT solutions as intermediaries, and it makes more sense to use spin_glass_solver(G) (by now).

Full Changelog: v9.2.4...v9.2.5

sha1sum results:
05a100a8356172ada9616262f1c1bf46a6aedf43 pyqrackising-9.2.5-py3-none-macosx_14_0_arm64.whl
e181c252a1f3dd0d6b8abea93c47604a51887f4c pyqrackising-9.2.5-py3-none-macosx_15_0_arm64.whl
051a106d91cd486e62962a5e8e20aa92121c5121 pyqrackising-9.2.5-py3-none-manylinux_2_35_x86_64.whl
06d897993f4d9c1264cacf90047481b499d6b451 pyqrackising-9.2.5-py3-none-manylinux_2_39_x86_64.whl
fced6cf86f3c773ada882e914adb6d390e20f274 pyqrackising-9.2.5-py3-none-win_amd64.whl
e7c286169b14196953170ba1e307a2056a0c8d0f pyqrackising-9.2.5.tar.gz