Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
* Add citing * Revert "Add citing" * Add citing * Add third-party references * Refactor metrics * behavior metric * Rectify the black version and reformatted * refactored * Fixed up: compatible with stacked observation space * refactored framestack wrapper * collect config generation * add radar plot tools * test passed for evaluation * test passed for evaluation * update maps * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * unify yaml configuration * remove comments * support multiple algorithm evaluation * change package version ray==0.8.7, tensorflow==1.15 * fixed some bugs, make evaluate runnable * update readme * Decouple logging keep headless envision (#433) (#451) * decoupled logging in client.py and have a headless option for client * modified envision creation to add headless * remove unneeded imports * address comments * Missing Fixes from 0_4_12 -> develop (#481) * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures * Fix pybullet log polution (again) (#475) * Add entries to dockerignore, gitignore, requirements. (#476) * Quick fix 4lane t collesion (#456) (#490) * removed TargetPose agents in 4lane_t * Change naming of smarts pybullet_action * Fix Missing Features in Develop Branch (#484) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Included ultra directories in header test * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: Jenish Patel <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: JenishPatel99 <[email protected]> * Clarify steps and fix missing step * Apply suggestions from code review Apply Adai's suggestions. Co-authored-by: adai <[email protected]> * Long-duration determinism test (#479) * Added long-duration determinism test with run time ~5 hours. * fixing the DISPLAY (#486) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> * fixing the DISPLAY * setup.py commented Co-authored-by: Tucker <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> * Clean up args and help strings in evaluate.py and train.py (#487) * Ignore generated tasks in ULTRA * Remove task 0 * Clean up arg help-strings in train.py * Rename task to scenario_info in train * Format ultra/train.py * Clean up arg help-strings in evaluate.py * Add task 0 back * Remove comments from train.py * Ultra add design documentation (#497) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> * Add design documentation in ULTRA * Move design flowchart to docs folder * Remove cached ultra-flowchart Co-authored-by: Tucker <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: AlexLewandowski <[email protected]> * Updated train.py and evaluate.py to include new flags (#507) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> * Updated train.py to include new flags * formatted * formatting again * addressed some formatting issues * Resolve utils/episode.py * replaced json with ijson * Format train.py and evaluate.py * Add policy flag to evaluate.py Co-authored-by: Tucker <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: AlexLewandowski <[email protected]> * Updated previous ultra tests and integrated them into workflow (#511) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> * uploading the tests for ci * Updated train.py to include new flags * formatted * WIP: first commit * restored config file * test push to examine new ultra ci * added missing dependencies for ultra ci * fixed reward apdater test * setup.py added * Editted gitignore file to add couple of .glb files * added few other files * run train_test.py in ultra ci * renamed test files * disable train_cli due to ray error * ultra ci strategy changed * disable train test * formatted train.py * discard previous commit * smarts ci added back * Added headers * disable episode counter test * Format ultra/train.py * formatted * run tests again Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: Tucker <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: AlexLewandowski <[email protected]> * Add MAD mode (#493) (#495) * added mode to only remove agent vehicles when sumo teardown * added docstring about parameter descriptions in sumo_traffic_simulation * ULTRA Initial Documentation (#449) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Simplified README * Instructions for training and evaluating a baseline agent * Instructions for creating a custom agent * Instructions for ULTRA setup * Other miscellaneous documentation * Added ULTRA paper link to README * Update ultra/README.md to include citation Co-authored-by: Alex Lewandowski <[email protected]> * Update ultra/README.md to fix missing link and other miscellaneous changes * Changed spec flag in ultra/evaluate.py to experiment-dir * Added an agent.md link to ultra/README.md * Rewrite ULTRA features in ultra/README.md Co-authored-by: Tucker <[email protected]> * Update agent's custom options in ultra/docs/agent.md Co-authored-by: Tucker <[email protected]> * Change wording in ultra/docs/custom_agent.md to be order-independent Co-authored-by: Tucker <[email protected]> * Reorder AgentSpec params to match code * Remove the 'the' in the AgentSpec params descriptions * Elaborate on the adapters of the AgentSpec params * Remove the initial AgentSpec params list * Add initial ULTRA demo to ultra/README.md * Change initial instructions for training/evaluating agents * Change custom agent creation instructions * Correct the agent locator Co-authored-by: Tucker <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: Alex Lewandowski <[email protected]> Co-authored-by: Tucker <[email protected]> * Update ULTRA/README.md with flowchart * fix format in train.py (#523) * Continuous deployment (#499) * Automated release of PyPI package and docker image upon new code release. * fix collision between agent and bubble agent vehicle (#485) * Bubble manager will choose BoxChassis for hijacking vehicles with non-dynamic actions * added warnings when vehicles with AckermanChassis call set_pose after step * Fixed vehicle generation to use boxchassis if action is not in smarts.dynamic_actions * Add initial speed uturn (#516) * Add initial speed for uturn task * Fix formatting issue * Apply Liam comments * Fix initial speed bug * Add Tucker's comment on using dataclass * Fix format on ULTRA * Fix typo in types * New refactor uturn cutin (#517) * Refactor uturn and cutin * Apply Tucker's last comment on removing the attribute * Give improved warning about missing goal position (#512) * Run CI on pull request from forked repo (#524) * Bump version 0.4.12->0.4.13 * Remove ijson dependency in ULTRA (#477) * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * ULTRA now uses a stable version of SMARTS, slight tweak in the installation process. README needs to updated * Added a few comments to setup.py * generated header for setup.py * Trying to commit setup.py again * formatted using black * Updated install.sh to fix path issue * Removed extra lines and added comments * Make removing ijson only change Co-authored-by: Tucker <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: AlexLewandowski <[email protected]> * Change distribution name for ULTRA (#532) * Change distribution name for ULTRA * Add MIT license to ULTRA * Add long description to ULTRA setup * Remove whitespace in citation * Update tests cases for ultra-ci and integrate them into workflow (#520) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format * Fix the teleport issue for BM (#442) * Fix the teleport issue for BM * Put traffic.destroy after agent_manger.destroy * Ultra Merge (#430) * migrating to smarts * initial changes to SAC and ultra registry * added info_adapter and removed src * adding environment score * moved env_score to info * removed unnecessary variable * applied change to ppo * updating dqn * changing the action space to discrete * updating ddpg * updating policy params * adding old README * adding docker for ultra * removing worker.py and run_agent.py * updating docker file * Update Dockerfile * Update run_container.sh * Update Dockerfile * Fix registry import (#385) * ULTRA Fix `test-formatting` (#386) * Fix registry import * `make format` * Update Dockerfile * Removed proxy details. (#389) * Fixed ULTRA README.md values and directories that were old. (#398) * renaming UltraAgentSpec to BaselineAgentSpec for clarity * overriding step module for ultra * cleanup highway changes * cleanup highway changes * fixing the format * removing tests for updates in future * applied comments to cleanup * removing docker folder * removing proxy from Makefile and cleanup Dockerfile * README docker step update * README update remove badges * update docker steo * adding headers * fixing the format Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> * Release v0.4.12 (#450) * Fix history vehicles key error + not recreate SMARTS instance (#230) (#439) * Not recreate SMARTS instance during imitation learning * Fix bugs in imitation learning to run interaction dataset * Fixes to improve mini-city (#410) * Fixes to make the mini-city work nicely * Address comments * Remove all vehicle in SUMO when reset (#435) * Fix OpEn for large vehicle (#446) * Option for resetting a scenario without restart or reload SUMO (#441) * Remove all vehicle in SUMO when reset * Reset without restart or reload SUMO * Upgrade to ray 1 0 1 post1 With Fix (#420) * Upgrade ray version * update requirements * Fix rllib_agent * Add jupyter to requirements * Remove jupyter in setup * Remove aggressive mode args * Fix analysis crash * Fix mode param * Fix pandas collumn error * Add mode to get_best_logdir * Bump version 0.4.11->0.4.12 * Fix rllib tests * Fix minicity vehicle type generation * Address unneeded * and self in callback signatures Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> * uploading the tests for ci * Updated train.py to include new flags * formatted * WIP: first commit * restored config file * test push to examine new ultra ci * added missing dependencies for ultra ci * fixed reward apdater test * setup.py added * Editted gitignore file to add couple of .glb files * added few other files * run train_test.py in ultra ci * renamed test files * disable train_cli due to ray error * ultra ci strategy changed * disable train test * formatted train.py * discard previous commit * smarts ci added back * Added headers * disable episode counter test * Format ultra/train.py * formatted * run tests again * formatting * changed ci-ultra-test file * execute ultra tests * try again * changed workflow files * execute ultra-ci * test ultra-ci * fix ultra-ci * fix ultra-ci * change dir to ultra in ultra-ci * revert to previous ultra-ci * revert to previous ultra-ci * dill dependency added to requirements.txt * simplified the jobs in ultra-ci * added pull request trigger Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: Tucker <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: AlexLewandowski <[email protected]> * ULTRA fix scenario analysis runtime errors (#508) * Fix scenario analysis bugs * Change argument names and format * Change ULTRA acronym in README (#550) * Changed ultra title in REAMDE * Update README and setup ULTRA acronym * Fix formatting * Change version number Co-authored-by: AlexLewandowski <[email protected]> * Fix transient performace for lane change in laner (#552) * Fix transient performace for lane change in laner * Fix the heading and gains during lane transition * Update evaluate.py to remove policy flag and add log directory flag (#544) * updated evalaute, train and episode files * formatted * WIP : fixing evaluate test * updated evaluation tests * generating model workaround * added scenarios to allow training * Added custom test to improve evaluating time * formatted * removed unnecessary test * moved cli commands to ultra-ci from evaluate * test ultra-ci * expanded model path variablity * final clean up * formatted * removed ray key-word args * restored original eval episodes * removed cpu arg * Update ultra/scenarios/README.md (#504) * Reset initial vehicle speed in UTurn (#560) * Reset initial vehicle speed in UTurn * Remove front force correction for low speed * Fix place of correction * Further reduce front lateral coefficient * SMARTS_reconnect_envision_when_disconnected (#341) (#545) * SMARTS will forever try to reconnect envision, envision will check if there are running simulation every 3 seconds * Parallel process to send state to Envision server (#567) * Replaced multithreading with multiprocessing and queue. * Modified test to support multiprocessing Envision client. * Add docs report format (#528) * Add docs report format * Update readme bug report reference * Add speed mode to update speed on sumo (#564) * Change ULTRA folder structure and add inits (#556) * Change folder structure and add inits * Change paths in ultra/tests * Change workflow to cd to ultra first * Remove ultra prefix from paths in test_evaluate * Move config to root of ultra * Change path of config in ultra_env * updated evaluate test and corrected installation process in ultra-ci * removed requirements.txt * move to ultra directory after install dependencies * added more episodes to enable evaluation * enable all tests * changed smarts dependency to v0.4.12 * Added new ci job * fixed typo in ultra-ci * ultra ci: put jobs in sequence * split jobs into seperate CI * fix syntax * revert back to previous test settings * delete package test for now * retry ultra-ci * try disabling test_ultra_package test * list all python packages * try adding path * change smarts dep to v0.4.6 * now try changing smarts dep to v0.4.12 * change smarts dep to v0.4.11 * syntax fix * run ci to show failure using v0.4.12 * test whl construction and installation * Fixed ultra_test_package issue * path issue resolve * Enable all tests (wrap up) * formatted * version bump v0.1.0 -> v0.1.1 * conformed to pytest assertions * regex warning suppressed * logic error fixed Co-authored-by: Jenish Patel <[email protected]> * benchmark: fixed cruising adapter not used * benchmark: temporary fixed global var not update * analysis test updated (#577) * Fix headers * Make format * Revert setup requirements changes * Split shadowing and hijacking for consistent vehicles in bubble (#553) * Split shadowing and hijacking for consistant vehicles in bubble * Fix bubble manager test * Fix bubble bugs * Fix bubble test * Fix Docker Cache Problem (#576) * Undo unnecessary changes * Create Uturn cutin example scenarios (#571) * Added cut-in scenario and OpEn works fine on cut-in * Added uturn scenario and OpEn meets error on uturn scenario * New add doc uturn cutin (#518) * Add document fot uturn and cutin * Create rst file using the pdf document * Fix typo in docs * Update issue templates (#594) * Added left control panel on envision, squashed all commits * Added back consoles and fix checkbox toggle * added pause button on envision (#575) (#585) * added pause button on envision * updated built files * Fix Python SCL Imports * Add missing import * Fix header * Apply suggestions from code review Co-authored-by: adai <[email protected]> * Minimal docker image. (#587) * Add max-episode-steps flag and use in tests (#584) * Add max-episode-steps flag and use in tests * Fix formatting * Fix formatting for ultra_env * Cleaned up analysis test * Change max_steps to max_episode_steps * changed kwargs name Co-authored-by: Jenish Patel <[email protected]> * Update black and add isort * Fix isort targets in makefile * Finalize isort folder targets * Add skip to sumo imports to ensure order * Finish isort config * Run make format * Upgrade black in formatting ci * Update open gains with curvature (#598) * Update open gains with curvature * Ensure new solver works with Kyber * Fix cutin test * Fix issues with black * Remove old solver * Bump version to 0.4.14 * Apply review suggestions * Remove introduced changes from master Co-authored-by: Ming Zhou <[email protected]> Co-authored-by: Jeffer Jingfei Peng <[email protected]> Co-authored-by: Liam Chen <[email protected]> Co-authored-by: iman512003 <[email protected]> Co-authored-by: adai <[email protected]> Co-authored-by: Kimia Hassanzadeh <[email protected]> Co-authored-by: christianjans <[email protected]> Co-authored-by: Jenish Patel <[email protected]> Co-authored-by: JenishPatel99 <[email protected]> Co-authored-by: Alex Lewandowski <[email protected]> Co-authored-by: jiayu <[email protected]> Co-authored-by: Jeffer Peng <[email protected]>
- Loading branch information
13 people
authored
Mar 1, 2021
1 parent
feee8fd
commit c847304