PPO playing Acrobot-v1 from https://github.com/sgoodfriend/rl-algo-impls/tree/5598ebc4b03054f16eebe76792486ba7bcacfc5c
341188c
from typing import Callable | |
Schedule = Callable[[float], float] | |
def linear_schedule( | |
start_val: float, end_val: float, end_fraction: float = 1.0 | |
) -> Schedule: | |
def func(progress_fraction: float) -> float: | |
if progress_fraction >= end_fraction: | |
return end_val | |
else: | |
return start_val + (end_val - start_val) * progress_fraction / end_fraction | |
return func | |
def constant_schedule(val: float) -> Schedule: | |
return lambda f: val | |