Compared to "CartPole-v0" environment, one of the most common gym environment for reinforcement learning, "PCSE-v0" leads to monotonically increasing memory usage that ends up killing the kernel. Both env.reset() and env.step() turns out to require cpu memory.
Compared to "CartPole-v0" environment, one of the most common gym environment for reinforcement learning, "PCSE-v0" leads to monotonically increasing memory usage that ends up killing the kernel. Both env.reset() and env.step() turns out to require cpu memory.