-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[RLlib] Refine MultiAgentEpisode
and add test cases.
#40799
[RLlib] Refine MultiAgentEpisode
and add test cases.
#40799
Commits on Oct 11, 2023
-
Initialized MultiAgentEpisode.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for afc0e17 - Browse repository at this point
Copy the full SHA afc0e17View commit details
Commits on Oct 12, 2023
-
Added timestep mapping and necessary methods to MultiAgentEpisode.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for c6c8275 - Browse repository at this point
Copy the full SHA c6c8275View commit details
Commits on Oct 16, 2023
-
Added get_<data> - methods to 'MultiAgentEpisode' and refined initial…
…ization, timestep mapping and class data. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 7b423ce - Browse repository at this point
Copy the full SHA 7b423ceView commit details -
Added 'create_successor', Äget_state()' and 'from_state()'. Agent mod…
…ule states will only be stored in the 'SingleAgentEpisode's. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for d7a50fa - Browse repository at this point
Copy the full SHA d7a50faView commit details
Commits on Oct 17, 2023
-
Refactored 'self.t' and 'self.t_started'. Added 'to_sample_batch', 'g…
…et_return' and '__len__'. Moved episode files into 'rllib/env'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 02ea5b8 - Browse repository at this point
Copy the full SHA 02ea5b8View commit details
Commits on Oct 18, 2023
-
Added list conversion to 'from_sample_batch' in case episode is not d…
…one. Furthermore moved 'SingleAgentEpisode' and 'MultiAgentEpisode' towards 'rllib/env'. I also added unit testing for 'SingleAgentEpisode'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for ac189c5 - Browse repository at this point
Copy the full SHA ac189c5View commit details -
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for c84525a - Browse repository at this point
Copy the full SHA c84525aView commit details
Commits on Oct 25, 2023
-
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for f3d88ec - Browse repository at this point
Copy the full SHA f3d88ecView commit details -
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 9112f38 - Browse repository at this point
Copy the full SHA 9112f38View commit details -
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 7f860a5 - Browse repository at this point
Copy the full SHA 7f860a5View commit details -
Configuration menu - View commit details
-
Copy full SHA for a8b3b1d - Browse repository at this point
Copy the full SHA a8b3b1dView commit details
Commits on Oct 26, 2023
-
Added documentation to 'SingleAgentEpisode' and merged existing episo…
…de into branch episode. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b4ea186 - Browse repository at this point
Copy the full SHA b4ea186View commit details -
Fixed some bugs found during testing and finished testing for 'Single…
…AgentEpisode'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for fd71c17 - Browse repository at this point
Copy the full SHA fd71c17View commit details -
Implemented review from @sven1977 and changed docstrings a bit.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 03cce90 - Browse repository at this point
Copy the full SHA 03cce90View commit details
Commits on Oct 27, 2023
-
Changed import for 'SingleAgentEpisode' to fix some errors in CI tests.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 47e88e6 - Browse repository at this point
Copy the full SHA 47e88e6View commit details -
Refactored 'get_observations|rewards|actions' into helper funciton.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 1a62530 - Browse repository at this point
Copy the full SHA 1a62530View commit details -
Refactored 'get_observations|rewards|actions' into helper funciton.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 9056dfe - Browse repository at this point
Copy the full SHA 9056dfeView commit details -
Added infos to 'MUltiAgentEpiosde'.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 0d8e324 - Browse repository at this point
Copy the full SHA 0d8e324View commit details -
Fixed imports and the resulting bug in 'single_agent_gym_env_runner.py'.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 879b080 - Browse repository at this point
Copy the full SHA 879b080View commit details -
Added 'extra_model_outputs' to 'MultiAgentEpisode'.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b53e122 - Browse repository at this point
Copy the full SHA b53e122View commit details -
Merge branch 'multi-agent-episode' into infos-and-extra-model-outs-fo…
…r-mae Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 77e6296 - Browse repository at this point
Copy the full SHA 77e6296View commit details -
Implemented @sven1977 's review.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 96e9d21 - Browse repository at this point
Copy the full SHA 96e9d21View commit details -
Implemented @sven1977 's review.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 277282e - Browse repository at this point
Copy the full SHA 277282eView commit details -
Implemented @sven1977 's review.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for ccce22c - Browse repository at this point
Copy the full SHA ccce22cView commit details -
Merged branch 'multi-agent-epsiode' into branch.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 72bc8f5 - Browse repository at this point
Copy the full SHA 72bc8f5View commit details
Commits on Oct 28, 2023
-
Started testing for MultiAgentEpisode.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 9300309 - Browse repository at this point
Copy the full SHA 9300309View commit details
Commits on Oct 30, 2023
-
Added buffers to the 'MultiAgentEpisode' and a corresponding logic to…
… 'add_timestep()'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for c786e06 - Browse repository at this point
Copy the full SHA c786e06View commit details
Commits on Oct 31, 2023
-
Refined 'add_timestep()' in 'MultiAgentEpisode' to handle also states…
… and agents that terminate before stepping first time. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 68d5478 - Browse repository at this point
Copy the full SHA 68d5478View commit details -
Added first test with multi-agent test environment and refined a coup…
…le of functionalitites in 'MultiAgentEPisode'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for d18d4a7 - Browse repository at this point
Copy the full SHA d18d4a7View commit details -
Added test scenarios and modified testing environment.
Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for de6bab0 - Browse repository at this point
Copy the full SHA de6bab0View commit details -
Added a test for 'add_initial_observation' and fixed some minor bug i…
…n the generation of the global timestep mapping. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 09f9e3a - Browse repository at this point
Copy the full SHA 09f9e3aView commit details
Commits on Nov 1, 2023
-
Added test for 'add_timestep', made intensive testing, modified 'add_…
…timestep' to deal directly with 'done' agents and also to skip partial testing for is_terminated/is_truncated. Furthermore, adjusted all functions to deal with 'terminated' and Ätruncated' dictionaries that contain only ready agents. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b57bef7 - Browse repository at this point
Copy the full SHA b57bef7View commit details
Commits on Nov 3, 2023
-
Added more testing for 'add_timestep()' and '__init__()'. Fixed some …
…bugs in 'add_timestep()' and '__init__()' methods in 'MultiAgentEpisode' to handle agents that never stepped and to handle the '__all__' in 'terminated' and 'truncated'. Changed '_agent_ids' to 'set' instead of 'list'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 4e07c41 - Browse repository at this point
Copy the full SHA 4e07c41View commit details -
Added global reward list and global reward timestep mapping to keep t…
…rack of 'orphane' rewards in multi-agent environemnts that occur in between an agent's action and the nex tobservation or before the initial observation. Furthermore, changed data extraction for single agents to take care of siutations in which collected data contains 'orphane' actions, i.e. actions that miss their subsequent observation. Orphane actions, states, and extra model outputs are put into the buffer. Ensured that tests are making use of these functionalities. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for fda7ef8 - Browse repository at this point
Copy the full SHA fda7ef8View commit details
Commits on Nov 8, 2023
-
Added test cases for 'create_successor' in the 'MultiAgentEpisode'. I…
…ntense testing led to further changes in the 'MultiAgentEpisode', specifically as we do need for the successor results from 'get_observations()', 'get_infos', etc. in type 'List[MultiAgentDict]' and not 'MultiAgentDict'. Furthermore, 'terminateds' and 'truncateds' hat to be provided with a getter. Needs more testing. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b8b530c - Browse repository at this point
Copy the full SHA b8b530cView commit details
Commits on Nov 9, 2023
-
Added functionalities to transfer buffers to successors of 'MultiAgen…
…tEpisode's. Also added corresponding tests. Rnamed 'global_rewards' to 'partial_rewards'. This is an intermediate commit to have a safe state to return to. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 8d6ac06 - Browse repository at this point
Copy the full SHA 8d6ac06View commit details -
Added more tests to the 'test_create_successor'. Found some incpnsist…
…encies in the use of the ' partial_rewards'. Has to be fixed before stepping forward. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 2d375ee - Browse repository at this point
Copy the full SHA 2d375eeView commit details
Commits on Nov 13, 2023
-
Removed states and related methods from 'MultiAgentEpisode' as they w…
…ere removed from 'SingleAgentEpisode by @sven1977. Added test for getters. Needed to change '_getattr_by_index' as using buffered actions is non-trivial. Had to add a 'global_actions_t' timestep_mapping for the actions as they could be buffered and the original timestep would get lost. Testing is not finished. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 9aabdac - Browse repository at this point
Copy the full SHA 9aabdacView commit details
Commits on Nov 14, 2023
-
Finished 'get_actions()' and 'get_extra_model_outputs()' for the case…
… of using buffered actions. Thereby refactored '_getattr_by_index()' extensively to be used more generically. Furthermore, wrote corresponding tests. Fixed a bug in the 'create_successor()' method. All tests run. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 53121d8 - Browse repository at this point
Copy the full SHA 53121d8View commit details
Commits on Nov 15, 2023
-
Finished testing for 'get_rewards()' without partial and buffered rew…
…ards, but with receiving either 'MultiAgentDict' or 'List[MutliAgentDict]'. Made minor changes to '_IndexMapping' and '_getattr_by_index()'. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 225eb68 - Browse repository at this point
Copy the full SHA 225eb68View commit details -
Added methods to '_IndexMapping' for more complex index searching. Ad…
…ded buffered rewards to the 'get_rewards' method. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for f0fb853 - Browse repository at this point
Copy the full SHA f0fb853View commit details
Commits on Nov 17, 2023
-
Added functionality for buffered_rewards and partial rewards. 'get_re…
…wards' is complete now and needs to be tested more. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 7d408aa - Browse repository at this point
Copy the full SHA 7d408aaView commit details
Commits on Nov 18, 2023
-
Added the file for testing the 'MultiAgentEpisode'. Worked on the fil…
…e for some days now. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 6bdc9e6 - Browse repository at this point
Copy the full SHA 6bdc9e6View commit details
Commits on Nov 20, 2023
-
Finished testing for getters and fixed some minor bugs I ran over whi…
…le testing. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for b42decf - Browse repository at this point
Copy the full SHA b42decfView commit details
Commits on Nov 22, 2023
-
Fixed several minor bugs and modified 'create_successor()' and 'conca…
…t_episode()' to use the corresponding 'SingleAgentEpisode''s methods, to contain as initial observation in the successor always the last observation of an agent. Adjusted tests accordingly and added multiple new ones. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for c0802ad - Browse repository at this point
Copy the full SHA c0802adView commit details
Commits on Nov 23, 2023
-
Finished testing for 'concat_episode()'. Also added a '_copy_buffer' …
…method to 'MultiAgentEpisode' for immutable copying of buffers between episodes. Refined the test file. Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 040ad60 - Browse repository at this point
Copy the full SHA 040ad60View commit details
Commits on Nov 24, 2023
-
Added tests for 'to_sample_batch', '__len__', and 'get_returns'. Modi…
…fied these functions to account for empty episodes and agents that are done.' Signed-off-by: Simon Zehnder <simon.zehnder@gmail.com>
Configuration menu - View commit details
-
Copy full SHA for 7d570e1 - Browse repository at this point
Copy the full SHA 7d570e1View commit details
Commits on Nov 27, 2023
-
Configuration menu - View commit details
-
Copy full SHA for d23f959 - Browse repository at this point
Copy the full SHA d23f959View commit details
Commits on Nov 28, 2023
-
Configuration menu - View commit details
-
Copy full SHA for a8937f3 - Browse repository at this point
Copy the full SHA a8937f3View commit details -
Configuration menu - View commit details
-
Copy full SHA for 721ff00 - Browse repository at this point
Copy the full SHA 721ff00View commit details -
Merge branch 'master' of https://github.com/ray-project/ray into test…
…ing-multi-agent-episode
Configuration menu - View commit details
-
Copy full SHA for 3a64881 - Browse repository at this point
Copy the full SHA 3a64881View commit details -
Configuration menu - View commit details
-
Copy full SHA for 18ea675 - Browse repository at this point
Copy the full SHA 18ea675View commit details