coach/benchmarks/acer

ACER

Each experiment uses 3 seeds. The parameters used for ACER are the same parameters as described in the original paper, except for the optimizer (changed to ADAM) and learning rate (1e-4) used.

Breakout ACER - 16 workers

coach -p Atari_ACER -lvl breakout -n 16

Space Invaders ACER - 16 workers

coach -p Atari_ACER -lvl space_invaders -n 16

Pong ACER - 16 workers

coach -p Atari_ACER -lvl pong -n 16