From b10a58c4ae49d7b08c924bb0acb0af9f64e17b72 Mon Sep 17 00:00:00 2001 From: Andrew Cohen Date: Tue, 16 Mar 2021 16:27:05 -0400 Subject: [PATCH 1/3] fix behavior name in FoodCollector yaml --- config/ppo/FoodCollector.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/config/ppo/FoodCollector.yaml b/config/ppo/FoodCollector.yaml index 6cddcf1b25..3cd735e9f8 100644 --- a/config/ppo/FoodCollector.yaml +++ b/config/ppo/FoodCollector.yaml @@ -1,5 +1,5 @@ behaviors: - FoodCollector: + GridFoodCollector: trainer_type: ppo hyperparameters: batch_size: 1024 From 018a47594ed5a8a647da0ae2ef66de7ee1d5a131 Mon Sep 17 00:00:00 2001 From: Andrew Cohen Date: Tue, 16 Mar 2021 17:31:33 -0400 Subject: [PATCH 2/3] add grid food configs --- config/ppo/FoodCollector.yaml | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/config/ppo/FoodCollector.yaml b/config/ppo/FoodCollector.yaml index 3cd735e9f8..176e4ca01c 100644 --- a/config/ppo/FoodCollector.yaml +++ b/config/ppo/FoodCollector.yaml @@ -1,4 +1,29 @@ behaviors: + FoodCollector: + trainer_type: ppo + hyperparameters: + batch_size: 1024 + buffer_size: 10240 + learning_rate: 0.0003 + beta: 0.005 + epsilon: 0.2 + lambd: 0.95 + num_epoch: 3 + learning_rate_schedule: linear + network_settings: + normalize: false + hidden_units: 256 + num_layers: 1 + vis_encode_type: simple + reward_signals: + extrinsic: + gamma: 0.99 + strength: 1.0 + keep_checkpoints: 5 + max_steps: 2000000 + time_horizon: 64 + summary_freq: 10000 + threaded: true GridFoodCollector: trainer_type: ppo hyperparameters: From fdd7a6f15ee89452e6f25a26b599f39239cf230a Mon Sep 17 00:00:00 2001 From: Andrew Cohen Date: Tue, 16 Mar 2021 17:37:46 -0400 Subject: [PATCH 3/3] Revert "add grid food configs" This reverts commit 018a47594ed5a8a647da0ae2ef66de7ee1d5a131. --- config/ppo/FoodCollector.yaml | 25 ------------------------- 1 file changed, 25 deletions(-) diff --git a/config/ppo/FoodCollector.yaml b/config/ppo/FoodCollector.yaml index 176e4ca01c..3cd735e9f8 100644 --- a/config/ppo/FoodCollector.yaml +++ b/config/ppo/FoodCollector.yaml @@ -1,29 +1,4 @@ behaviors: - FoodCollector: - trainer_type: ppo - hyperparameters: - batch_size: 1024 - buffer_size: 10240 - learning_rate: 0.0003 - beta: 0.005 - epsilon: 0.2 - lambd: 0.95 - num_epoch: 3 - learning_rate_schedule: linear - network_settings: - normalize: false - hidden_units: 256 - num_layers: 1 - vis_encode_type: simple - reward_signals: - extrinsic: - gamma: 0.99 - strength: 1.0 - keep_checkpoints: 5 - max_steps: 2000000 - time_horizon: 64 - summary_freq: 10000 - threaded: true GridFoodCollector: trainer_type: ppo hyperparameters: