From 7b615016cfdc45f26d3571430fe14b2f37a932dc Mon Sep 17 00:00:00 2001 From: Arthur Juliani Date: Fri, 14 Dec 2018 11:28:28 -0800 Subject: [PATCH] Move done and reward to buffer from demonstration --- ml-agents/mlagents/trainers/demo_loader.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/ml-agents/mlagents/trainers/demo_loader.py b/ml-agents/mlagents/trainers/demo_loader.py index 5819e7be96..7ac0f1eba6 100644 --- a/ml-agents/mlagents/trainers/demo_loader.py +++ b/ml-agents/mlagents/trainers/demo_loader.py @@ -18,6 +18,8 @@ def make_demo_buffer(brain_infos, brain_params, sequence_length): current_brain_info = brain_infos[idx] next_brain_info = brain_infos[idx + 1] demo_buffer[0].last_brain_info = current_brain_info + demo_buffer[0]['done'].append(next_brain_info.local_done[0]) + demo_buffer[0]['rewards'].append(next_brain_info.rewards[0]) for i in range(brain_params.number_visual_observations): demo_buffer[0]['visual_obs%d' % i] \ .append(current_brain_info.visual_observations[i][0])