fix(dqn): increase vanilla_dqn_boltzmann_cartpole max frames

kengz / SLM-Lab

Modular Deep Reinforcement Learning framework in PyTorch. Companion library of the book "Foundations of Deep Reinforcement Learning".

https://slm-lab.gitbook.io/slm-lab/

MIT License

1.23k stars 263 forks source link

fix(dqn): increase vanilla_dqn_boltzmann_cartpole max frames #503

Closed kengz closed 2 years ago

kengz commented 2 years ago

fix error when max_frame is lower than log_frequency

Fix error in benchmark spec vanilla_dqn_boltzmann_cartpole when max_frame is lower than log_frequency, causing metrics calculation to not have enough data points.