[Feature] Add reduction parameter to On-Policy losses. #617

Job	Run time
unittests-atari-dqn	0s
unittests-d4rl	0s
unittests-envpool	0s
unittests-gendgrl	0s
unittests-minari	0s
unittests-openx	0s
unittests-roboset	0s
unittests-smacv2	0s
unittests-vd4rl	0s
unittests-pettingzoo / linux-job	9m 23s
unittests-brax (3.9, 12.1) / linux-job	9m 10s
unittests-gym (3.9, 12.1) / linux-job	17m 7s
unittests-jumanji (3.9, 12.1) / linux-job	8m 53s
unittests-robohive (3.9, 12.1) / linux-job	10m 3s
unittests-sklearn (3.9, 12.1) / linux-job	5m 43s
unittests-vmas (3.9, 12.1) / linux-job	13m 23s
	1h 13m 42s

Provide feedback