Commit Graph

  • 50c9b3f41d Update README.md main Arec Jamgochian 2024-04-01 16:55:43 -07:00
  • 6e8986b3d4 Merge pull request #6 from sisl/dev-idm-videos icra23 Etienne Bührle 2022-10-04 21:03:25 +02:00
  • bc33b786aa helper script for copying videos dev-idm-videos ebuehrle 2022-10-04 20:52:51 +02:00
  • dd201738cb when using IDM expert, update interaction graph in order to correctly display ego target agent ebuehrle 2022-10-04 20:51:19 +02:00
  • 5fb358d725 do not enable IDM for expert data ebuehrle 2022-09-22 11:15:49 +02:00
  • 740e0ea9f4 Merge branch 'save-videos' into dev-idm-videos ebuehrle 2022-09-22 10:55:38 +02:00
  • 88213e7d76 remove --skip_running flag in evaluate_models.sh ebuehrle 2022-09-22 10:43:28 +02:00
  • 388c80007e Revert "committing learned policies and summary files" ebuehrle 2022-09-22 10:42:23 +02:00
  • 3aaf252dbe committing learned policies and summary files idm-override Arec Jamgochian 2022-08-29 01:30:11 -07:00
  • 779a0ea89f adding use idm to test scripts and running test suite Arec Jamgochian 2022-08-24 02:44:21 -07:00
  • 3a09a6eb7d adding idm override option flag, set to true. current running experiments for gail and shail experiment A to see how different times are. Since were on cpus on the cluster, guessing it will be 10x Arec Jamgochian 2022-08-07 16:33:09 -07:00
  • 9c9ee8f21b clearing stale __init__ iros22 Arec 2022-04-11 21:46:40 -07:00
  • f814516072 Update README.md Arec Jamgochian 2022-04-11 10:49:29 -07:00
  • 3fa370eb8a Add flag to skip seeds save-videos ebuehrle 2022-03-05 07:12:34 +01:00
  • a576f0fb18 Close figures ebuehrle 2022-03-04 16:14:35 +01:00
  • 575e299fc8 Generate videos of expert data ebuehrle 2022-03-04 16:07:15 +01:00
  • 1e70303c57 Optionally save videos of policy evaluations ebuehrle 2022-03-04 15:58:18 +01:00
  • a9feec4f38 Merge branch 'main' of https://github.com/sisl/InteractionImitation into main Arec Jamgochian 2022-03-03 20:20:48 -08:00
  • 3d3b3d510a removing demo generation script and including it in readme. updating expert data generator to loop over tracks and match readme Arec Jamgochian 2022-03-03 20:20:36 -08:00
  • 3ac9465997 Update README.md Arec Jamgochian 2022-03-04 05:19:26 +01:00
  • 597b9af5d4 Updating readme Arec Jamgochian 2022-03-04 05:13:25 +01:00
  • 37f44605d2 purging unused files Arec Jamgochian 2022-03-03 19:53:45 -08:00
  • 4076b0361a removed scratch except for expert data generator, which i moved to the top Arec Jamgochian 2022-03-03 19:37:29 -08:00
  • fc04f8e9ee adding final commands for tuning, training final models, testing them Arec Jamgochian 2022-03-03 12:12:12 -08:00
  • f5f1c24f45 Remove unused method idm_upgrade Johannes Fischer 2022-03-01 10:00:40 +01:00
  • f93e130498 Merge branch 'main' into idm_upgrade Johannes Fischer 2022-03-01 09:58:07 +01:00
  • 2965dc9982 Update scratch for plotting Johannes Fischer 2022-03-01 09:57:36 +01:00
  • fa0e20998d Update IDM Johannes Fischer 2022-03-01 09:56:14 +01:00
  • e0d205bbd7 updating experiment B vecenv overfit-a Arec Jamgochian 2022-02-28 21:44:10 -08:00
  • 910615cc00 Add best config from previous runs ebuehrle 2022-03-01 06:34:51 +01:00
  • 57a42f70ec purging etienne experiment files, adding hail metric files Arec Jamgochian 2022-02-28 19:03:57 -08:00
  • 5f6ad37c37 updating bc experiment, adding automatic config copying, fixing idm/expert running from file issues in eval_experiments Arec Jamgochian 2022-02-28 18:56:28 -08:00
  • 7a134d98c9 120 episodes per rolloout ebuehrle 2022-02-28 22:41:39 +01:00
  • 261e35eae8 More policy iterations per epoch ebuehrle 2022-02-28 22:39:28 +01:00
  • e602aa0641 Merge branch 'main' of https://github.com/sisl/InteractionImitation into main Arec Jamgochian 2022-02-28 03:35:03 -08:00
  • 5ada1cc543 updating weight decay Arec Jamgochian 2022-02-28 03:35:00 -08:00
  • a9314c4657 Copy over config files ebuehrle 2022-02-28 12:23:28 +01:00
  • a37995694d Load policy config in evaluation ebuehrle 2022-02-28 12:23:02 +01:00
  • 62c28d0cfa adding big tune run options Arec Jamgochian 2022-02-28 03:04:59 -08:00
  • e28459a168 Merge branch 'main' into idm_upgrade Johannes Fischer 2022-02-28 11:47:11 +01:00
  • 6068c87402 Only take targets driving roughly in the same direction as IDM target idm_upgrade_proj Johannes Fischer 2022-02-28 11:37:08 +01:00
  • 6f181a7351 Merge branch 'main' of https://github.com/sisl/InteractionImitation into main Arec Jamgochian 2022-02-28 01:57:28 -08:00
  • 81e38f55ab adding shail-etienne policies Arec Jamgochian 2022-02-28 01:57:25 -08:00
  • 51eb810f5c COnfigure HAIL in eval_experiments ebuehrle 2022-02-28 10:52:26 +01:00
  • b94344214b Merge branch 'main' into idm_upgrade Johannes Fischer 2022-02-28 10:50:43 +01:00
  • 6d867466c6 Change IDM default params Johannes Fischer 2022-02-28 10:49:58 +01:00
  • 8cc71f07de Configure HAIL in eval script ebuehrle 2022-02-28 10:45:19 +01:00
  • 57076793c1 Add configs for SHAIL ebuehrle 2022-02-28 10:43:12 +01:00
  • da6d1d812c Add some configs ebuehrle 2022-02-28 10:33:10 +01:00
  • 88cae70477 fixing big in loading precalculated metrics for averaging Arec Jamgochian 2022-02-28 00:59:30 -08:00
  • 7814c7c30d ignoring pngs Arec Jamgochian 2022-02-28 00:37:30 -08:00
  • 5533faab40 adding metric pkls Arec Jamgochian 2022-02-28 00:36:52 -08:00
  • 1ee3805aba adding trained test policies, as well as fixing evaluate script and commands Arec Jamgochian 2022-02-27 23:51:12 -08:00
  • e37911eeff removing gail-trpo since performance is about the same as gail, adding experiment evaluation script, updating metric averaging to work Arec Jamgochian 2022-02-27 23:03:28 -08:00
  • e110570092 adding final bc configs, updating experiment to validate in longer intervals Arec Jamgochian 2022-02-27 20:40:24 -08:00
  • 5f77398fd0 commiting json files and intermediate settings used to run baseline methods around 02-26, 16ocklock. Next will set off big tuning runs for gail, gailtrpo, and bc (with committed values) Arec Jamgochian 2022-02-27 16:27:01 -08:00
  • 2cacdf691f Merge branch 'main' of https://github.com/sisl/InteractionImitation into main Arec Jamgochian 2022-02-27 14:30:32 -08:00
  • e52b658aff adding all bestconfigs from big tune run (not sure what happened in last commit) Arec Jamgochian 2022-02-27 14:29:49 -08:00
  • 578459e37b best_configs/hail_expA.json Arec Jamgochian 2022-02-27 14:27:29 -08:00
  • 25e4d5679f Make envs independent in BC experiment ebuehrle 2022-02-27 22:53:12 +01:00
  • 7f1557144c committing etiennes best configs, leading to 73% success rate in shail expB Arec Jamgochian 2022-02-27 13:43:59 -08:00
  • 7039468e46 Add tuning script for BC. Todo: if enough GPUs, enable resources in run function ebuehrle 2022-02-27 15:03:52 +01:00
  • cc3848e8e0 Remove unused parameters ebuehrle 2022-02-27 15:02:32 +01:00
  • 4443492d3e Add tuning script for GAIL (TRPO) ebuehrle 2022-02-27 14:28:25 +01:00
  • 4b4fa7e09a Add tuning script for GAIL (PPO) ebuehrle 2022-02-27 14:14:08 +01:00
  • 569e0756ca getting loading of configs, overwriting with seeds, running with tune, and moving back to directory working. adding check for either training or testing, and allowing specification of number of test cpus to split seeds over Arec Jamgochian 2022-02-27 02:25:16 -08:00
  • 21cbd1c956 getting best shail/hail configs outsave up and running, adding shell for test runs Arec Jamgochian 2022-02-27 01:22:39 -08:00
  • acb4fdc518 renaming shail policy file, removing ogail file since its unnecessary Arec Jamgochian 2022-02-27 00:00:42 -08:00
  • 3280041efa committing what is hopefully final run of sgail for both experiments, A and B Arec Jamgochian 2022-02-26 23:48:58 -08:00
  • 3b9051505e naming convention doesnt like env key, leakyrelu is no bueno, running quick experiment to see which option sets can work with our time left. the options that look 4s ahead are definitely out of the questions, possibly the 2s ones aswell Arec Jamgochian 2022-02-26 16:02:29 -08:00
  • 71e3c5f816 Parametrize discriminator architecture ebuehrle 2022-02-26 14:46:17 +01:00
  • 35e6fb299c Use collision rate as main metric ebuehrle 2022-02-26 14:28:58 +01:00
  • 1e24612347 Merge branch 'main' of https://github.com/sisl/InteractionImitation ebuehrle 2022-02-26 14:06:22 +01:00
  • 6876cf9625 Track collision rate ebuehrle 2022-02-26 13:50:53 +01:00
  • dc9cbf329b Enable safe options on evaluation env ebuehrle 2022-02-26 12:24:25 +01:00
  • f9d3cceed5 Save checkpoints and config to folder, move params to config ebuehrle 2022-02-26 12:23:42 +01:00
  • c28c6c05b7 added ogail script and splitting up feasability on next line to avoid calculating it unnecessarily (though it might be fine as is) Arec Jamgochian 2022-02-25 18:39:33 -08:00
  • fa98601fa6 fixing issues with lazylinear sequential, setting off a big run Arec Jamgochian 2022-02-25 17:48:56 -08:00
  • 2dfd7e3c2b add policy saving Arec Jamgochian 2022-02-25 17:00:46 -08:00
  • 59083ebce3 making option list indexable to visualize in tensorboard. noticing run from last night had much better performance under more long options, unclear if due to choice or environment factors, like episodes lasting longer. making stoponcollision and seed parameters. fixing ability to see reward metrics under ray/tune/, problem was they were being returned as tensors, not floats Arec Jamgochian 2022-02-25 16:36:14 -08:00
  • d1f23e6d25 readding expertdata to gitignore, just manually adding expert joint state and action files Arec Jamgochian 2022-02-25 01:27:02 -08:00
  • 46564231ef changing gitignore to upload expert data for evaluation, adding multiple roundabout trackfile environments in vecenv Arec Jamgochian 2022-02-25 01:21:16 -08:00
  • bf9ce84fe4 changing when directory name gets saved Arec Jamgochian 2022-02-25 00:42:44 -08:00
  • 99aa50a08b added support for different activations, number of hidden layers, options, running a big run over this Arec Jamgochian 2022-02-25 00:00:04 -08:00
  • 7feea74eb8 adding stablebaselines, adding safe loading for nonCuda cluster Arec Jamgochian 2022-02-24 23:18:02 -08:00
  • 5a5d8a7aff WIP: Support different plan lengths ebuehrle 2022-02-25 01:32:43 +01:00
  • 02c1813b00 Incrementing agent expert data, smaller policy network ebuehrle 2022-02-25 01:23:15 +01:00
  • 0d193d4af3 Scratch for horner scheme Johannes Fischer 2022-02-24 14:16:56 +01:00
  • 8e12996dfe fix typo Johannes Fischer 2022-02-24 14:18:26 +01:00
  • 495b87e70e Scratch for horner scheme horner_scheme Johannes Fischer 2022-02-24 14:16:56 +01:00
  • febceed651 Add expert data ebuehrle 2022-02-23 18:17:30 +01:00
  • a3280893af Update IDM script Johannes Fischer 2022-02-23 18:16:01 +01:00
  • 336cf02278 Parameterize hidden layer size of policy, add some candidates to grid search ebuehrle 2022-02-23 18:14:23 +01:00
  • 7f64ec7bb0 Move hyperparameters to config object ebuehrle 2022-02-23 17:50:21 +01:00
  • 406c4ad9ee Fix tune by moving file ebuehrle 2022-02-23 17:32:11 +01:00
  • 9c3cb4fb55 Add IDM script Johannes Fischer 2022-02-23 17:21:02 +01:00
  • f037c119cc Set up for ray tune ebuehrle 2022-02-23 16:55:13 +01:00
  • 91f88983b0 Add learning rate schedule to SHAIL-PPO ebuehrle 2022-02-23 14:22:02 +01:00
  • a1db6aa553 Make IDM use vehicle on ego path a reference Johannes Fischer 2022-02-22 22:00:46 +01:00
  • 68b066ec53 adding success rate, total distance, and survive time Arec 2022-02-22 00:48:35 -08:00