Marek Wolan
3af22ac2ed
Merge remote-tracking branch 'origin/dev' into feature/2623-action-masking
2024-07-11 15:50:42 +01:00
Archer.Bowen
239f5b86c0
#2716 Agent logging now sits outside of the simulation output log directory, updated dev-mode CLI to include agent logging and added additional tests.
2024-07-10 13:36:37 +01:00
Marek Wolan
7201b7b8e0
2623 Add e2e tests for action masking
2024-07-10 11:01:42 +01:00
Archer.Bowen
48645d2e72
#2716 Initial Implementation + Initial Tests and updated changelog and sphinx documentation.
2024-07-09 16:46:31 +01:00
Marek Wolan
faf268a9b9
2623 move action mask generation to game and fix MARL masking
2024-07-09 15:59:50 +01:00
Marek Wolan
5367f9ad53
2623 Ray single agent action masking
2024-07-09 15:27:03 +01:00
Marek Wolan
470fa28ee1
2623 Implement basic action masking logic
2024-07-09 13:13:13 +01:00
Marek Wolan
28dabad66b
#2682 Backport changes to core PrimAITE
2024-06-25 11:04:52 +01:00
Marek Wolan
fad8d936e6
Merge remote-tracking branch 'origin/dev' into feature/2646_Update-pre-commit-to-check-for-valid-copyright
2024-06-13 12:52:09 +01:00
Marek Wolan
c538aee022
Rename benchmark variables and add h1 to notebooks
2024-06-10 12:58:07 +01:00
Chris McCarthy
81bcf99855
#2646 - Added a custom pre-commit hook that ensure the copyright clause is added to .py and .rst files.
2024-06-05 09:11:37 +01:00
Czar Echavez
472040aa70
Merge remote-tracking branch 'origin/dev' into feature/2628-update-benchmarking-script-branched
2024-06-01 00:20:21 +01:00
Marek Wolan
53f01ca1e1
Merged PR 395: Fix reward logging and get ray to stop complaining
...
## Summary
Change the agent history log to also spit out rewards correctly.
Make ray adhere to new ray API
## Test process
Unit test pass
All notebooks run without crash
## Checklist
- [ ] PR is linked to a **work item**
- [ ] **acceptance criteria** of linked ticket are met
- [x] performed **self-review** of the code
- [ ] written **tests** for any new functionality added with this PR
- [ ] updated the **documentation** if this PR changes or adds functionality
- [ ] written/updated **design docs** if this PR implements new functionality
- [ ] updated the **change log**
- [x] ran **pre-commit** checks for code style
- [x] attended to any **TO-DOs** left in the code
2024-05-31 14:38:05 +00:00
Marek Wolan
e48b71ea1a
get ray to stop crashing
2024-05-31 15:25:08 +01:00
Marek Wolan
c5f131ece5
fix reward logging
2024-05-31 15:00:18 +01:00
Czar Echavez
add8cc0ab1
#2628 : temp commit
2024-05-31 13:47:02 +01:00
Marek Wolan
f8336d07bd
#2626 fix too many open files bug
2024-05-31 13:28:56 +01:00
Marek Wolan
c7b2ed7f86
Merge branch 'dev' into feature/2588-minimal-deps
2024-05-31 12:11:16 +01:00
Marek Wolan
a161a3a5ec
#2639 Fix reset signature
2024-05-30 15:34:46 +01:00
Marek Wolan
0c19546322
#2588 optional RL deps
2024-05-29 14:54:45 +01:00
Charlie Crane
f66d9d5817
Merge branch 'dev' into feature/2457-Set_link_bandwidth_via_config
2024-05-14 14:44:20 +01:00
Charlie Crane
71684dc5f0
#2457 - Initial commit removing the hardcoding of bandwidth values within links. Now pull it from configuration files, defaulting to 100 if not implemented
2024-05-13 14:42:15 +01:00
Czar Echavez
ec6d2bf640
#2533 : fix missing implementation of output_dir override in io.py + moving the dev mode override in SIM_OUTPUT to prevent the override from being overridden
2024-05-01 16:53:37 +01:00
Czar Echavez
a2fb04e6f6
#2533 : set default dev output path + clean up
2024-05-01 14:02:40 +01:00
Czar Echavez
729f9c5064
#2533 : fix primaite config recreated when running setup
2024-04-30 19:15:54 +01:00
Czar Echavez
73f64cd89c
Merge remote-tracking branch 'origin/dev' into feature/2533-improve-dev-tools-in-primaite-cli
2024-04-30 18:27:05 +01:00
Charlie Crane
9b3699389a
#2523 - Adding in some additional logging for other agent classes. This currently prints total_reward instead of average reward
2024-04-30 15:36:59 +01:00
Charlie Crane
b8c46a92e9
#2523 - Minor typos/corrections to make things run locally. Added one logging line to record Blue agent reward per step in environment.py
2024-04-30 14:16:29 +01:00
Czar Echavez
c94fa653f7
Merge remote-tracking branch 'origin/dev' into feature/2533-improve-dev-tools-in-primaite-cli
2024-04-29 16:24:19 +01:00
Czar Echavez
a7492ba39f
#2447 : cli commands for dev mode
2024-04-29 13:45:10 +01:00
Marek Wolan
be5cf0b877
Merge remote-tracking branch 'origin/dev' into feature/2476-training-schedules-mockup
2024-04-25 15:17:08 +01:00
Marek Wolan
66f31e8ed1
#2476 Add test for episode scheduler
2024-04-25 15:09:46 +01:00
Marek Wolan
42ce264e73
#2476 Fix string formatting
2024-04-25 13:54:05 +01:00
Marek Wolan
e3f9162a77
Merge remote-tracking branch 'origin/dev' into feature/2476-training-schedules-mockup
2024-04-25 13:50:11 +01:00
Czar Echavez
96b6fb81da
#2447 : added mode to primaite cli + printing session to current working directory if in dev mode
2024-04-25 13:32:57 +01:00
Marek Wolan
a92898d001
#2476 Finalise explanation notebook for episode schedule
2024-04-25 13:25:26 +01:00
Marek Wolan
28c8b7c9d9
#2476 Get episode schedule working
2024-04-23 11:51:50 +01:00
Czar Echavez
34773ed225
#2470 : implement PR suggestions
2024-04-18 16:38:42 +01:00
Czar Echavez
94ca13c0f9
#2470 : add log level via config + test
2024-04-18 15:14:40 +01:00
Marek Wolan
8d0d323e0b
#2374 Remove primaite session
2024-04-16 11:26:17 +01:00
Marek Wolan
0828f70b4c
#2459 back-sync b8 changes into core
2024-04-15 11:50:08 +01:00
Marek Wolan
82143a2a2e
#2446 Fix io config parsing order
2024-04-02 00:31:06 +01:00
Marek Wolan
b4d310eda2
Align step counts in logging
2024-03-15 16:17:38 +00:00
Marek Wolan
04c86e30c9
Fix some stuff
2024-03-15 11:15:02 +00:00
Marek Wolan
d33c80d0d6
Minor fixes
2024-03-14 14:33:04 +00:00
Marek Wolan
c3f1cfb33d
Add shared reward
2024-03-11 22:53:39 +00:00
Marek Wolan
7599655879
Add agent action history
2024-03-11 20:10:08 +00:00
Marek Wolan
0447a05084
Add call validation
2024-03-08 15:57:43 +00:00
Marek Wolan
8589ce449a
Merged PR 295: Agent action logging
...
## Summary
Added a new optional capability to create a JSON log each episode with a list of action each agent took each step (including scripted, RL, and red agents).
Also I had to slightly refactor the IO system to not rely on PrimaiteSession, as it's gonna be deprecated soon. Therefore the IO module is now linked to the gym environment. Each time you init a gym environment, it creates a session directory.
## Test process
Tried the SB3, Ray SARL and Ray MARL notebooks to see that the outputs get generated.
## Checklist
- [x] PR is linked to a **work item**
- [x] **acceptance criteria** of linked ticket are met
- [x] performed **self-review** of the code
- [ ] written **tests** for any new functionality added with this PR
- [x] updated the **documentation** if this PR changes or adds functionality
- [ ] written/updated **design docs** if this PR implements new functionality
- [x] updated the **change log**
- [x] ran **pre-commit** checks for code style
- [x] attended to any **TO-DOs** left in the code
Move IO to environments from session and add agent logging
Related work items: #2278
2024-03-05 16:39:44 +00:00
Marek Wolan
a7bfc56b98
Apply documentation changes based on PR review.
2024-03-05 11:21:49 +00:00