Releases: Learning-and-Intelligent-Systems/predicators
Releases · Learning-and-Intelligent-Systems/predicators
Active Predicate Learning
Code for "Embodied Active Learning of Relational State Abstractions for Bilevel Planning" by Amber Li and Tom Silver.
Ignore Effects (August 2022)
Learning Operators with Ignore Effects for Bilevel Planning in Continuous Domains
Requirements
- Python 3.8+
- Code tested on Ubuntu 18.04 and 20.04 and Mac OS Big Sur
Installation
- Virtual environments recommended
pip install requirements.txtexport PYTHONHASHSEED=0export PYTHONPATH="${PYTHONPATH}:</path/to/this/folder/>
Running
Example:
python src/main.py --env screws --approach nsrt_learning --strips_learner backchaining --num_train_tasks 50
Other Environment Names
- repeated_nextto_single_option
- repeated_nextto_painting
- painting
- satellites
- satellites_simple
Skill Learning (June 2022)
Learning Neuro-Symbolic Skills for Bilevel Planning
Paper: http://arxiv.org/abs/2206.10680
Requirements
- Python 3.8+
- Code tested on Ubuntu 18.04 and MacOS Big Sur
Installation
- Virtual environments recommended
pip install requirements.txtexport PYTHONHASHSEED=0export PYTHONPATH="${PYTHONPATH}:</path/to/this/folder/>
Running
Example:
python src/main.py --env cover_multistep_options --num_train_tasks 250 \
--seed 0 --approach nsrt_learning --option_learner direct_bc \
--min_perc_data_for_nsrt 1 --segmenter contacts --timeout 300
Other environment names:
doors(add flags:--included_options MoveToDoor,MoveThroughDoor)stick_buttoncoffee
Citation
@misc{silver2022learning,
title={Learning Neuro-Symbolic Skills for Bilevel Planning},
author={Tom Silver and Ashay Athalye and Joshua B. Tenenbaum and Tomas Lozano-Perez and Leslie Pack Kaelbling},
year={2022},
eprint={2206.10680},
}
march-2022-experiments
Experiments for the paper:
Predicate Invention for Bilevel Planning
Tom Silver*, Rohan Chitnis*, Nishanth Kumar, Willie McClinton, Tomás Lozano-Pérez, Leslie Pack Kaelbling, Joshua Tenenbaum
AAAI 2023
https://arxiv.org/abs/2203.09634
experiments for RLDM 2022 submission
AGGREGATED DATA OVER SEEDS:
NUM_TRAIN_TASKS NUM_SOLVED AVG_NUM_PREDS AVG_TEST_TIME AVG_NODES_CREATED LEARNING_TIME NUM_SEEDS
EXPERIMENT_ID
blocks_branchfac_200demo 200.00 (0.00) 50.00 (0.00) 6.00 (0.00) 0.29 (0.09) 2886.93 (1494.24) 485.26 (11.95) 10
blocks_downrefUNUSED_200demo 200.00 (0.00) 41.50 (13.45) 7.70 (2.65) 0.35 (0.27) 1915.20 (738.42) 5542.41 (6569.45) 10
blocks_downrefeval_200demo 200.00 (0.00) 50.00 (0.00) 2.00 (0.00) 0.29 (0.09) 2886.93 (1494.24) 14.76 (4.62) 10
blocks_downrefscore_200demo 200.00 (0.00) 42.10 (12.76) 2.00 (0.00) 0.33 (0.15) 1731.77 (1076.78) 90.82 (126.12) 10
blocks_energy_200demo 200.00 (0.00) 41.90 (5.86) 9.10 (1.22) 0.83 (0.40) 5764.24 (2870.76) 3567.16 (571.81) 10
blocks_main_200demo 200.00 (0.00) 49.80 (0.40) 6.00 (0.00) 0.27 (0.09) 2729.36 (1561.74) 4005.58 (266.70) 10
blocks_mainhadd_200demo 200.00 (0.00) 50.00 (0.00) 6.30 (0.90) 0.07 (0.01) 119.63 (33.65) 2395.41 (491.45) 10
blocks_noinventallexclude_200demo 200.00 (0.00) 1.60 (1.20) 2.00 (0.00) 3.63 (3.17) 409.67 (100.78) 679.97 (20.79) 10
blocks_noinventnoexclude_200demo 200.00 (0.00) 50.00 (0.00) 5.00 (0.00) 0.23 (0.07) 2886.93 (1494.24) 319.36 (11.72) 10
blocks_noinventnoexcludehadd_200demo 200.00 (0.00) 50.00 (0.00) 5.00 (0.00) 0.20 (0.10) 4206.11 (2952.31) 318.04 (10.29) 10
blocks_prederror_200demo 200.00 (0.00) 8.70 (2.24) 3.00 (0.00) 0.27 (0.36) 30.68 (8.69) 554.58 (53.65) 10
blocks_random 50.00 (0.00) 0.50 (0.67) 0.00 (0.00) 0.02 (0.01) 0.00 (0.00) 0.00 (0.00) 10
cover_regrasp_branchfac_200demo 200.00 (0.00) 3.60 (1.20) 3.00 (0.00) 0.00 (0.00) 5.00 (0.00) 101.21 (1.31) 10
cover_regrasp_downrefUNUSED_200demo 200.00 (0.00) 27.50 (2.11) 5.20 (0.75) 0.00 (0.00) 7.79 (0.60) 508.14 (150.75) 10
cover_regrasp_downrefeval_200demo 200.00 (0.00) 27.90 (3.36) 1.00 (0.00) 0.01 (0.00) 7.48 (0.34) 5.17 (13.02) 10
cover_regrasp_downrefscore_200demo 200.00 (0.00) 50.00 (0.00) 1.00 (0.00) 0.02 (0.02) 11.03 (1.81) 1.05 (1.15) 10
cover_regrasp_energy_200demo 200.00 (0.00) 50.00 (0.00) 5.00 (0.00) 0.11 (0.10) 12.75 (0.67) 398.96 (9.50) 10
cover_regrasp_main_200demo 200.00 (0.00) 49.40 (1.80) 7.00 (1.84) 0.09 (0.20) 9.83 (0.57) 883.64 (368.51) 10
cover_regrasp_noinventallexclude_200demo 200.00 (0.00) 3.60 (1.20) 1.00 (0.00) 0.00 (0.00) 5.00 (0.00) 232.57 (18.94) 10
cover_regrasp_noinventnoexclude_200demo 200.00 (0.00) 50.00 (0.00) 6.00 (0.00) 0.01 (0.00) 12.23 (0.51) 222.01 (148.46) 10
cover_regrasp_prederror_200demo 200.00 (0.00) 50.00 (0.00) 5.00 (0.00) 0.01 (0.00) 10.10 (0.38) 229.49 (3.06) 10
cover_regrasp_random 50.00 (0.00) 2.00 (1.90) 0.00 (0.00) 0.01 (0.00) 0.00 (0.00) 0.00 (0.00) 10
painting_branchfac_200demo 200.00 (0.00) 4.30 (4.50) 19.11 (0.31) 4.21 (2.08) 570.90 (566.96) 3683.97 (149.09) 10
painting_downrefUNUSED_200demo 200.00 (0.00) 49.50 (0.67) 21.50 (1.86) 0.21 (0.04) 465.33 (106.51) 60198.09 (4420.82) 10
painting_downrefeval_200demo 200.00 (0.00) 47.30 (5.85) 4.00 (0.00) 0.21 (0.03) 448.13 (117.30) 32.38 (58.14) 10
painting_downrefscore_200demo 200.00 (0.00) 49.80 (0.60) 4.00 (0.00) 0.21 (0.03) 467.90 (102.40) 12.78 (6.32) 10
painting_energy_200demo 200.00 (0.00) 49.40 (0.80) 18.20 (2.48) 0.76 (0.48) 2738.32 (1307.99) 12760.43 (3403.07) 10
painting_main_200demo 200.00 (0.00) 47.60 (6.56) 22.10 (1.45) 0.22 (0.03) 461.40 (104.45) 72425.74 (13740.91) 10
painting_noinventallexclude_200demo 200.00 (0.00) 0.00 (0.00) nan (nan) nan (nan) nan (nan) 577.33 (19.46) 10
painting_noinventnoexclude_200demo 200.00 (0.00) 50.00 (0.00) 13.00 (0.00) 0.45 (0.07) 2474.79 (677.15) 667.79 (35.78) 10
painting_prederror_200demo 200.00 (0.00) 0.00 (0.00) nan (nan) nan (nan) nan (nan) 1057.66 (80.21) 10
painting_random 50.00 (0.00) 0.00 (0.00) 0.00 (0.00) nan (nan) 0.00 (0.00) 0.00 (0.00) 10
tools_branchfac_200demo 200.00 (0.00) 36.80 (3.97) 33.70 (2.45) 0.11 (0.06) 112.41 (68.15) 5791.23 (656.16) 10
tools_downrefUNUSED_200demo 200.00 (0.00) 26.70 (5.53) 28.90 (5.68) 0.08 (0.02) 152.63 (55.82) 18061.51 (3320.13) 10
tools_downrefeval_200demo 200.00 (0.00) 26.10 (6.59) 6.00 (0.00) 0.08 (0.04) 177.24 (101.63) 5.40 (3.64) 10
tools_downrefscore_200demo 200.00 (0.00) 48.60 (2.15) 6.00 (0.00) 0.51 (0.24) 1504.71 (1393.35) 38.35 (21.15) 10
tools_energy_200demo 200.00 (0.00) 30.20 (2.40) 30.30 (2.24) 0.25 (0.07) 688.29 (219.66) 9866.25 (1228.92) 10
tools_main_200demo 200.00 (0.00) 48.80 (1.54) 27.40 (4.39) 0.46 (0.14) 1938.66 (1594.13) 17151.09 (1500.10) 10
tools_noinventallexclude_200demo 200.00 (0.00) 0.00 (0.00) nan (nan) nan (nan) nan (nan) 720.90 (150.29) 10
tools_noinventnoexclude_200demo 200.00 (0.00) 50.00 (0.00) 15.00 (0.00) 0.55 (0.13) 3857.64 (977.09) 235.10 (8.21) 10
tools_prederror_200demo 200.00 (0.00) 12.40 (1.62) 12.00 (0.00) 0.98 (0.23) 5538.06 (1442.38) 823.12 (14.36) 10
tools_random 50.00 (0.00) 0.00 (0.00) 0.00 (0.00) nan (nan) 0.00 (0.00) 0.00 (0.00) 10