yaml ready to go

bdaiinstitute · Mar 30, 2024 · 8fb0e06 · 8fb0e06
1 parent 5c5307e
commit 8fb0e06
Showing 1 changed file with 54 additions and 54 deletions.
diff --git a/scripts/configs/active_sampler_learning.yaml b/scripts/configs/active_sampler_learning.yaml
@@ -1,42 +1,42 @@
 # Final active sampler learning experiments.
 ---
 APPROACHES:
-  # planning_progress_explore:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "planning_progress"
-  # task_repeat_explore:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "task_repeat"
-  # competence_gradient_explore_ucb:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "competence_gradient"
-  #     active_sampler_explore_use_ucb_bonus: True
-  # success_rate_explore_ucb:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "success_rate"
-  #     active_sampler_explore_use_ucb_bonus: True
-  # skill_diversity_explore_ucb:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "skill_diversity"
-  # random_score_explore:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "active_sampler"
-  #     active_sampler_explore_task_strategy: "random"
-  # random_nsrts_explore:
-  #   NAME: "active_sampler_learning"
-  #   FLAGS:
-  #     explorer: "random_nsrts"
+  planning_progress_explore:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "planning_progress"
+  task_repeat_explore:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "task_repeat"
+  competence_gradient_explore_ucb:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "competence_gradient"
+      active_sampler_explore_use_ucb_bonus: True
+  success_rate_explore_ucb:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "success_rate"
+      active_sampler_explore_use_ucb_bonus: True
+  skill_diversity_explore_ucb:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "skill_diversity"
+  random_score_explore:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "active_sampler"
+      active_sampler_explore_task_strategy: "random"
+  random_nsrts_explore:
+    NAME: "active_sampler_learning"
+    FLAGS:
+      explorer: "random_nsrts"
   maple_q:
     NAME: "maple_q"
     FLAGS:
@@ -51,23 +51,23 @@ ENVS:
   #     active_sampler_learning_feature_selection: all
   #     active_sampler_learning_explore_pursue_goal_interval: 1
   #     max_num_steps_interaction_request: 150
-  # ball_and_cup_sticky_table:
-  #   NAME: "ball_and_cup_sticky_table"
-  #   FLAGS:
-  #     sticky_table_place_smooth_fall_prob: 1.00
-  #     sticky_table_place_sticky_fall_prob: 0.00
-  #     sticky_table_pick_success_prob: 1.0
-  #     sticky_table_num_sticky_tables: 1
-  #     sticky_table_num_tables: 5
-  #     sticky_table_place_ball_fall_prob: 1.00
-  #     active_sampler_learning_explore_length_base: 25
-  #     active_sampler_learning_exploration_epsilon: 0.5
-  #     skill_competence_model_optimistic_recency_size: 2
-  #     skill_competence_model_optimistic_window_size: 2
-  #     horizon: 8
-  #     active_sampler_learning_explore_length_base: 100000  # effectively disable
-  #     active_sampler_learning_feature_selection: oracle
-  #     max_num_steps_interaction_request: 100
+  ball_and_cup_sticky_table:
+    NAME: "ball_and_cup_sticky_table"
+    FLAGS:
+      sticky_table_place_smooth_fall_prob: 1.00
+      sticky_table_place_sticky_fall_prob: 0.00
+      sticky_table_pick_success_prob: 1.0
+      sticky_table_num_sticky_tables: 1
+      sticky_table_num_tables: 5
+      sticky_table_place_ball_fall_prob: 1.00
+      active_sampler_learning_explore_length_base: 25
+      active_sampler_learning_exploration_epsilon: 0.5
+      skill_competence_model_optimistic_recency_size: 2
+      skill_competence_model_optimistic_window_size: 2
+      horizon: 8
+      active_sampler_learning_explore_length_base: 100000  # effectively disable
+      active_sampler_learning_feature_selection: all
+      max_num_steps_interaction_request: 100
   spot_sweeping_sim:
     NAME: "spot_main_sweep_env"
     FLAGS:
@@ -76,7 +76,7 @@ ENVS:
       skill_competence_model_optimistic_window_size: 2
       horizon: 10
       active_sampler_learning_explore_length_base: 100000  # effectively disable
-      active_sampler_learning_feature_selection: oracle
+      active_sampler_learning_feature_selection: all
       spot_run_dry: True
       approach_wrapper: spot_wrapper
       perceiver: spot_perceiver