From d7c537707a95d84556607ae965d34b0dcc7dfea3 Mon Sep 17 00:00:00 2001 From: Shangtong Zhang Date: Sun, 20 May 2018 21:56:14 -0600 Subject: [PATCH] Fix a minor bug of option critic --- deep_rl/agent/OptionCritic_agent.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/deep_rl/agent/OptionCritic_agent.py b/deep_rl/agent/OptionCritic_agent.py index a0602e7..32992bd 100644 --- a/deep_rl/agent/OptionCritic_agent.py +++ b/deep_rl/agent/OptionCritic_agent.py @@ -52,7 +52,7 @@ class OptionCriticAgent(BaseAgent): options_next = np.copy(options) dice = np.random.rand(len(options_next)) for j in range(len(dice)): - if dice[j] < np_betas_next[j]: + if dice[j] < np_betas_next[j] or terminals[j]: options_next[j] = self.policy.sample(np_q_options_next[j]) for i, terminal in enumerate(terminals): if terminals[i]: