GitHub
0ff8f9af
Create ML-Agents Package ( #3267 )
Convert the UnitySDK to a Packman Package.
- Separate Examples into a sample project.
- Move core UnitySDK Code into com.unity.ml-agents.
- Create asmdefs for the ml-agents package.
- Add package validation tests for win/linux/max.
- Update protobuf generation scripts.
- Add Barracuda as a package dependency for ML-Agents. (users no longer have to install it themselves).
5 年前
Chris Elion
4da6f17b
Take tagmanager asset from previous commit of master. ( #3298 )
5 年前
GitHub
ceebacd6
Convert pyramids to Raycast sensor ( #3299 )
5 年前
GitHub
01b29cec
Ray perception sensor tests ( #3297 )
* tests WIP
* add test for misses, remove obsolete test
* add filtering test
* cleanup
* 2d default, remove constants test
* use built in tags
* copy old physics settings
* manually sync physics transforms
5 年前
GitHub
be0e5635
remove deprecated RayPerceptionMB, legacy codepath: ( #3304 )
5 年前
GitHub
2990116f
Fix Walker bug with dirToTarget ( #3309 )
5 年前
Ervin Teng
db249ceb
Merge branch 'master' into develop-splitpolicyoptimizer
5 年前
GitHub
6ddcef36
Upgrade to barracuda 0.4.0-preview. ( #3310 )
5 年前
GitHub
fed3efdc
Done After Set Reward ( #3311 )
5 年前
GitHub
2db09cef
Model override from commandline ( #3265 )
* WIP model override from commandline
* Agent lazy init, multiple overrides
* MLAgentsExamples namespace
* add model override to 3dball
5 年前
GitHub
5c192cb5
BehaviorParameters - accessors for undecorated behavior name ( #3316 )
* BehaviorParameters - accessors for undecorated behavior name
* one fully-qualified is enough
5 年前
GitHub
620fa24a
Track reward for inference ( #3320 )
* WIP
* add reward stats
* const, dont write timers on mobile
5 年前
GitHub
93fa865e
Update Barracuda to 0.5.0-preview. ( #3329 )
5 年前
GitHub
3afec7c1
Avoid scores and Unity logo overlap ( #3324 )
5 年前
Christopher Goy
a6b472a8
Upgrade project to 2019.3.0f6.
5 年前
GitHub
7ad624da
testing: exit playback specified number of episodes ( #3332 )
5 年前
GitHub
3bd57902
fix MissingReferenceException in GridWorld ( #3346 )
5 年前
GitHub
4d252428
move Startup ( #3357 )
5 年前
GitHub
ae97ab3a
Soccer refactor ( #3331 )
5 年前
Ervin Teng
d4ee7346
Merge commit 'f9c05a61d574305497789b5997f1ae3ea1b1ad3b' into develop-splitpolicyoptimizer
5 年前
GitHub
ae9d12b5
Remove monitor ( #3372 )
* Removing the monitor
* Editing the migrating docs
* Readding the Monitor in the examples folder
* Editing the Migrating doc
5 年前
GitHub
386ba66c
Develop observation collector ( #3352 )
* Add the VectorSensor to the CollectObservation call
* Example of API change for BalanceBall
* Modified the Examples
* Changes to the migrating doc
* Editing the docs
* Update docs/Learning-Environment-Design-Agents.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Migrating.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Migrating.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Getting-Started-with-Balance-Ball.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* addressing comments
* Removed the MLAgents.Sensor namespace
* Removing the MLAgents.Sensor namespace from the tests
* Editing the migrating docs
Co-authored-by: Chris Elion <celion@gmail.com>
5 年前
GitHub
84161e7a
csharp cleanup ( #3392 )
* clean up examples and tests
* more cleanup
* m_UseChildSensors
5 年前
GitHub
d20bda06
csharp cleanup ( #3392 ) ( #3395 )
* clean up examples and tests
* more cleanup
* m_UseChildSensors
5 年前
GitHub
92a8aed2
Pass action masker as input to CollectObservations ( #3389 )
* Sentencing Action masking the same as observations
I am rather unsure about the doubling of the CollectObservation methods (and the copy pasta that comes along)
Need to edit the documentation and the migrating doc once we agree we want to do this
* Addressing the comments
* Improvements to the documentation
* Editing the documentation
5 年前
GitHub
c5c5f9b0
add DecisionRequester to FoodCollector prefabs ( #3414 )
5 年前
GitHub
981c16ec
set Agent.maxSteps for FoodCollector prefabs ( #3416 )
5 年前
GitHub
352dca12
Adding new food collector model ( #3423 )
5 年前
GitHub
c982fa3d
adding new models ( #3413 )
5 年前
GitHub
44f88933
Make a serialization upgrade path for maxStep. ( #3424 )
- Test comment out as we need a package that is only verified for 2019.2.
5 年前
Andrew Cohen
23f74f21
soccer fives
5 年前
GitHub
6f5bb92a
Make a serialization upgrade path for maxStep. ( #3424 ) ( #3434 )
- Test comment out as we need a package that is only verified for 2019.2.
5 年前
Andrew Cohen
7d90b042
soccerfives curricula
5 年前
Andrew Cohen
db770e7c
SoccerFives brain
5 年前
Anupam Bhatnagar
d8c79f48
resolving merge conflicts
5 年前
GitHub
9a4b151c
Merge pull request #3441 from Unity-Technologies/master-into-release-0.14.0
Master into release 0.14.0 copy
5 年前
GitHub
18e1d01b
Update section name in changelog.md ( #3499 )
5 年前
Ervin Teng
5ef902bf
Merge branch 'master' into develop-splitpolicyoptimizer
5 年前
GitHub
443686c5
undo material changes ( #3503 )
5 年前
GitHub
d072e091
[upkeep] Add a dev project to take advantage of package that only work with 2019.x or newer. ( #3452 )
5 年前
GitHub
1ebed95a
Remove internal registry. ( #3465 )
5 年前
Andrew Cohen
e4d776c3
Merge branch 'master' into soccer-fives
5 年前
Andrew Cohen
a3ec270f
SoccerFives brain
5 年前
GitHub
c55cb4df
Replace Agent.GetStepCount with Agent.StepCount` ( #3476 )
5 年前
GitHub
472f9f0e
Merge branch 'master' into develop-badEnvReturnCode
5 年前
Andrew Cohen
b7d77740
Merge branch 'master' into soccer-fives
5 年前
Ervin Teng
1859f252
Merge commit 'fbcdd83c087135f870e785cc72e5ff9a7e898e3a' into develop-splitpolicyoptimizer
5 年前
Andrew Cohen
926bc888
added positions to behaviorparameters
5 年前
Andrew Cohen
39a76867
added more backward raycasts to twos and fives
5 年前
Andrew Cohen
4edb7f41
updated config/soccer brains
5 年前
Ervin Teng
1156b9b3
Merge branch 'develop-splitpolicyoptimizer' into develop-removeactionholder
5 年前
GitHub
8530b724
Fix package validation errors for sensor-based classes ( #3501 )
* Fix package validation errors in ISensor and CameraSensor.
* Fix docs in CameraSensor and add a helper method.
* Fix package validation errors in CameraSensorComponent.
* Fix package validation errors in SensorComponent.
* Additonal fixes to ISensor.
* Fixes package validation errors for the remaining sensor-based classses.
* Fixed docs based on PR comments. Only doc-changes.
* Fold back docs that were deleted due to merge conflict. (Was easier to delete and re-add.)
* More missing docs.
* Fixed incorrect location for doc strings.
* Adding periods at the end of param/return tags.
* Added a few periods.
5 年前
Anupam Bhatnagar
e04fcd71
Merge branch 'master' into master-into-release-0.14.1
5 年前
Andrew Cohen
30725c27
2v1 soccer config and env
5 年前
Andrew Cohen
de73baa9
Merge branch 'master' into soccer-fives
5 年前
GitHub
f25bf7d3
Reintroduce MLAgents.Sensors namespace ( #3509 )
* Reintroduced the namespace MLAgents.Sensors
* Documentation changes
* updated the changelog
5 年前
GitHub
8ce9dcfd
add DoneReason enum to Agent ( #3517 )
5 年前
GitHub
d9898be5
Remove test scene from build settings and codebase. ( #3524 )
5 年前
GitHub
9a371b17
[Renaming] SetActionMask -> SetDiscreteActionMask + added the virtual method CollectDiscreteActionMasks ( #3525 )
* Code edits
* Modified the markdowns
* Update com.unity.ml-agents/CHANGELOG.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Learning-Environment-Design-Agents.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Learning-Environment-Design-Agents.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Renaming files and methods
* Addressing comments
* Update docs/Learning-Environment-Design-Agents.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
Co-authored-by: Chris Elion <celion@gmail.com>
5 年前
GitHub
e1a0f41b
Added the MLAgents.Demonstrations namespace ( #3532 )
* Added the MLAgents.Demonstrations namespace
* Added the MLAgents.Editor namespace
* Overrided the .demo.meta files due to the change in namespace
5 年前
GitHub
b9bd4df2
Modified some namespaces ( #3533 )
* Added the MLAgents.Demonstrations namespace
* Added the MLAgents.Editor namespace
* Overrided the .demo.meta files due to the change in namespace
* More namespace changes
* Added the sidechannels namespace
* Modified changelog and migrating docs
5 年前
GitHub
e7ec5007
[change] Make Agent non-abstract, update Basic scene. ( #3528 )
5 年前
GitHub
4e747130
Renaming AgentInfo.actionMasks to AgentInfo.DiscreteActionMasks ( #3539 )
5 年前
GitHub
47755a62
Made BehaviorParameters internal ( #3546 )
* Made the BrainParameters internal
* Editing the docs
* [skip-ci] A lot more controversial
* [skip ci] Added formerly serialized as
* Use cached BehaviorParameters
5 年前
Andrew Cohen
573b1f6d
Merge branch 'master' into soccer-fives
5 年前
GitHub
251aa7b3
Removed the IFloatProperties interface ( #3570 )
* [skip ci] Remove IFloatProperties interface
* [skip ci] Update the examples
* Updating the documentation
5 年前
GitHub
e3af96ca
Merge branch 'master' into develop-demo-load-seek
5 年前
GitHub
c315979f
Add command-line flag for example env scene selection ( #3473 )
The Startup scene used for us to retrain and bundle example scenes
together already accepts the SCENE_NAME environment variable to choose
a scene. This change adds support for command-line selection of the
scene name so that `--env-args` can be used to choose the scene.
If both the environment variable and command-line flag are provided, the
scene will use the environment variable.
5 年前
GitHub
5b7975ad
[bugfix] Fix MLA-793 Make Unity lifecycle methods protected. Added tests for changes ( #3590 )
5 年前
Chris Elion
841b0937
SideChannel helper messages
5 年前
Chris Elion
3196300e
undo accidental changes and formatting
5 年前
GitHub
411bb64a
Renaming Agent's methods ( #3557 )
* [skip ci] Renamed methods in the Agent class
WARNING, the user when implementing obsolete methods will see the message :Member `old method` overrides obsolete member `old method`. Add the Obsolete attribute to `old method`. It will not suggest the new method to override.
* [skip ci] Updated the example environment
* [skip ci] Updated migrating and changelog
* [skip ci] Editing the docs
* [skip ci] Missing docs
* :+1
* Update docs/Getting-Started-with-Balance-Ball.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Learning-Environment-Create-New.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update docs/Learning-Environment-Create-New.md
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* [skip ci] documentation changes
* [skip ci] Update docs/Getting-Started-with-Balance-Ball.md
* [skip ci] Update docs/Getting-Started-with-Balance-Ball.md
* [skip ci] Update docs/Gett...
5 年前
GitHub
345329f0
[bug-fix] Re-record GridWorld Demo ( #3585 )
5 年前
Chris Elion
fcfd0f7f
add and use Agent.ReloadPolicy()
5 年前
Chris Elion
fa5e7e6d
Merge remote-tracking branch 'origin/master' into develop-BehaviorParams-public
5 年前
GitHub
ab99343a
Merge pull request #3595 from Unity-Technologies/develop-BehaviorParams-public
Make BehaviorParameters public again, update editor
5 年前
Andrew Cohen
b1cfa74d
Merge branch 'master' into develop-test-imitation
5 年前
GitHub
000ceedf
[bug-fix] Update Pyramids demo ( #3613 )
5 年前
GitHub
68bbee63
Update examples project to 2018.4.18f1 ( #3618 )
From 2018.4.14f1. An internal package dependency was updated as
a side effect.
5 年前
GitHub
2ba1e9d4
Remove dead components from the examples scenes ( #3619 )
5 年前
Jonathan Harper
066df247
Remove space from Product Name for examples
In #2588 it was suggested that the space in the Product Name for
our example environments causes confusion when using a default build
because of the need to escape the space in the build filename.
This change removes the space from the Product Name in the project's
player settings.
5 年前
Jonathan Harper
391cb49d
Update examples project to 2018.4.18f1 ( #3618 )
From 2018.4.14f1. An internal package dependency was updated as
a side effect.
5 年前
GitHub
e83bc02b
Remove space from Product Name for examples ( #3612 )
In #2588 it was suggested that the space in the Product Name for
our example environments causes confusion when using a default build
because of the need to escape the space in the build filename.
This change removes the space from the Product Name in the project's
player settings.
5 年前
GitHub
1d9cce7b
Remove dead components from the examples scenes ( #3619 ) ( #3624 )
5 年前
Jonathan Harper
c36f3534
Remove space from Product Name for examples
In #2588 it was suggested that the space in the Product Name for
our example environments causes confusion when using a default build
because of the need to escape the space in the build filename.
This change removes the space from the Product Name in the project's
player settings.
5 年前
GitHub
8ed45da0
Updating the NN models ( #3632 )
* Updating the NN models
* Update gridworld
* [skip ci] Update BallHard
* Update hallway
5 年前
GitHub
ec278616
Hotfixes for Release 0.15.1 ( #3698 )
* [bug-fix] Increase height of wall in CrawlerStatic (#3650 )
* [bug-fix] Improve performance for PPO with continuous actions (#3662 )
* Corrected a typo in a name of a function (#3670 )
OnEpsiodeBegin was corrected to OnEpisodeBegin in Migrating.md document
* Add Academy.AutomaticSteppingEnabled to migration (#3666 )
* Fix editor port in Dockerfile (#3674 )
* Hotfix memory leak on Python (#3664 )
* Hotfix memory leak on Python
* Fixing
* Fixing a bug in the heuristic policy. A decision should not be requested when the agent is done
* [bug-fix] Make Python able to deal with 0-step episodes (#3671 )
* adding some comments
Co-authored-by: Ervin T <ervin@unity3d.com>
* Remove vis_encode_type from list of required (#3677 )
* Update changelog (#3678 )
* Shorten timeout duration for environment close (#3679 )
The timeout duration for closing an environment was set to the
same duration as the timeout when waiting ...
5 年前
Andrew Cohen
53bea15c
Merge branch 'master' into soccer-fives
5 年前
Andrew Cohen
ac261e36
Merge branch 'master' into self-play-mutex
5 年前
GitHub
de0cf96f
[bug-fix] Increase height of wall in CrawlerStatic ( #3650 )
5 年前
GitHub
4af8d5cd
merge release 0.15.0 into master ( #3652 )
* Bumping version on the release (#3615 )
* Update examples project to 2018.4.18f1 (#3618 )
From 2018.4.14f1. An internal package dependency was updated as
a side effect.
* Remove dead components from the examples scenes (#3619 ) (#3624 )
* Improve warnings and exception if using unsupported combo
* add meta file
* fix unit test
* enforce onnx conversion (expect tf2 CI to fail) (#3600 )
* Update error message
* Updated the release branch docs (#3621 )
* Updated the release branch docs
* Edited the README
* make sure top-level timer is closed before writing
* Remove space from Product Name for examples
In #2588 it was suggested that the space in the Product Name for
our example environments causes confusion when using a default build
because of the need to escape the space in the build filename.
This change removes the space from the Product Name in the project's
player settings.
* [bug-fix] Incr...
5 年前
Andrew Cohen
072b4135
soccer 2v1 on the cloud
5 年前
Andrew Cohen
033b58ba
bonus to single agent
5 年前
Andrew Cohen
fbae3d48
added soccer brains
5 年前
Andrew Cohen
fb993986
Merge branch 'master' into self-play-mutex
5 年前
Andrew Cohen
b42c9482
Merge branch 'self-play-mutex' into soccer-2v1
5 年前
GitHub
11c518a3
Stats SideChannel (for custom TensorBoard metrics) ( #3660 )
5 年前
Andrew Cohen
03cd3cd6
striker vs goalie env
5 年前
Andrew Cohen
941b8ae7
Strikers vs goalie added
5 年前
GitHub
6612b496
Deprecating Academy.Instance.FloatProperties ( #3696 )
* Deprecating Academy.Instance.FloatProperties
* Made the registered side channels a static property and created the sideChannelUtils class to handle side channel stuff
* Clearing the sending message queue in the Academy when the communicaor is not on
* addressing comments
5 年前
Andrew Cohen
e892cb9a
Added Generic soccer agent for draws
5 年前
Andrew Cohen
a7a372b9
Merge branch 'master' into self-play-mutex
5 年前
GitHub
f623b722
Anchor GridWorld UI to top-left ( #3706 )
5 年前
Andrew Cohen
c7a34413
Merge branch 'self-play-mutex' into soccer-2v1
5 年前
Andrew Cohen
61d38b15
rerunning all self-play
5 年前
Andrew Cohen
d9f1a2f5
more experiments for self-play
5 年前
GitHub
914c896c
reuse built player in CI tests ( #3709 )
* reuse built player in training tests
* cleanup backcompat path
* verbose
* move exe directories instead
* don't build in verbose mode
5 年前
Yuan Gao
c9403183
Initial trial on 3dballhard, bouncer, basics
5 年前
Andrew Cohen
4f27c206
made striker wall prettier
5 年前
Andrew Cohen
d7b8cf16
CubeWars
5 年前
Andrew Cohen
59b88be6
Merge branch 'master' into self-play-mutex
5 年前
GitHub
9093c796
Make sure Academy is initialized in ProjectSettingsOverrides.Awake()
5 年前
Ervin Teng
06fa3d39
Merge branch 'master' into develop-sac-apex
5 年前
Anupam Bhatnagar
50e52d9c
Merge branch 'master' into distributed-training
5 年前
Andrew Cohen
72706301
soccer curriculum
5 年前
Andrew Cohen
6f8c4e5e
reward for shooting big agent
5 年前
bhh
823fa3a5
initial snake proto
5 年前
GitHub
eeb0c74d
Add CompletedEpisodes counter ( #3724 )
5 年前
Andrew Cohen
ccd1a991
striker and goalie brains
5 年前
Andrew Cohen
c13259fd
curriculum for small soldiers
5 年前
bhh
1f4a7c58
misc
5 年前
Yuan Gao
24a681bf
Updated the prefebs to enable inference
5 年前
Andrew Cohen
c5f9abce
new blue goalie wall
5 年前
GitHub
92870d43
Fix missing action axis in Tennis ( #3732 )
* Fix missing action axis in Tennis
* Update Project/Assets/ML-Agents/Examples/Tennis/Scripts/TennisAgent.cs
Fix typo
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
Co-authored-by: Chris Elion <chris.elion@unity3d.com>
5 年前
bhh
e18ea8ab
updating snake model
5 年前
Yuan Gao
51787b5f
Updated the editor build setting
5 年前
Andrew Cohen
9fed4985
tennis curriculum
5 年前
Andrew Cohen
5b3e899c
more shoot reward
5 年前
GitHub
89237f96
Reset StackingSensor when the Agent resets ( #3727 )
* sensors.Reset() WIP
* fix test implementations
* call reset from Agent
5 年前
bhh
5c24f925
more model tweaks
5 年前
Yuan Gao
524079ab
Revert "Updated the editor build setting"
This reverts commit d5c7c9c30e81c4b352b25ab321862d06570377b7.
5 年前
Andrew Cohen
34dcc1c1
modified striker wall
5 年前
Andrew Cohen
804f8345
fixed angle obs
5 年前
Andrew Cohen
1b9c643b
Merge branch 'master' into self-play-mutex
5 年前
bhh
35736d30
added scripts
5 年前
Andrew Cohen
b8ffdc4b
cloud experiments
5 年前
Andrew Cohen
c47646fa
stacked vector obs
5 年前
bhh
a4461583
updated script
5 年前
Andrew Cohen
7a7eb324
Merge branch 'master' into internal-policy-ghost
5 年前
bhh
e8035a23
training works now
5 年前
Andrew Cohen
adf93fd8
Goalie Speed
5 年前
bhh
d07a48a3
added more platforms
5 年前
Andrew Cohen
e4f7f2a6
removed curriculum tennis
5 年前
bhh
4e8a65ef
fixed char forward since it's a rotated capsule
5 年前
Andrew Cohen
2e7f8f41
Merge branch 'develop-cubewars' into asymm-envs
5 年前
Andrew Cohen
930d6fa3
Merge branch 'self-play-mutex' into soccer-2v1
5 年前
bhh
0a46fc45
added new models
5 年前
bhh
59933799
Worm dynamic scene ready
5 年前
Andrew Cohen
b1f01bb5
fixed side channels -> float prop
5 年前
bhh
5262c14f
update model
5 年前
bhh
dc9fcd46
loosened joints retrained looking good
5 年前
Andrew Cohen
23931551
added existential penaalty to generic soccer agent
5 年前
bhh
a9c5d24b
cleaned up code and removed old prefabs
5 年前
Andrew Cohen
b6a621e9
cooldowns/shockwave feature of big agent
5 年前
Andrew Cohen
e4937807
new soccer brains
5 年前
bhh
1ecc8924
final training done. ready to go.
5 年前
Andrew Cohen
58e99f61
fixed large agent falling
5 年前
Andrew Cohen
8e271ee8
new tennis brain
5 年前
bhh
b32dab7a
added dynamic 3.5M step model
5 年前
GitHub
aae58330
Merge branch 'master' into develop-add-inference-examples
5 年前
Andrew Cohen
5cd88455
increased reward for shooting big agent
5 年前
bhh
5f2c7a4c
updated static model, added 28M step dyamic model for testing. redoing dynamic model now
5 年前
Yuan Gao
06e16a61
Revert the EditorBuildSetting file
5 年前
Andrew Cohen
b0c506a6
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
5a5e13fa
soccertwos config
5 年前
GitHub
6ef56c83
Merge pull request #3749 from Unity-Technologies/develop-add-inference-examples
Add ModelOverrider to all of the Agent prefabs to enable Barracuda Inference with specified .nn model file
5 年前
bhh
58354f1b
updated dynamic 3.5M model
5 年前
Andrew Cohen
d8c7b60c
SideChannels replace FloatProp
5 年前
Andrew Cohen
9bec6cae
strikers vs goalie brains
5 年前
Andrew Cohen
5e864e6f
Merge branch 'soccer-2v1' into asymm-envs
5 年前
bhh
c3e74451
Remove 28M model
5 年前
Andrew Cohen
c07e0fce
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
933d7b32
Merge branch 'self-play-mutex' into soccer-2v1
5 年前
Ervin Teng
5e980ec1
Merge branch 'master' into develop-sac-apex
5 年前
Andrew Cohen
44e6fa7b
soccer 1e8 timesteps/Tennis existential penalty
5 年前
Andrew Cohen
514cd342
removed kickpower so goalie uses face
5 年前
Andrew Cohen
7a932b55
new cubewar brains
5 年前
Andrew Cohen
900ae050
new SoccerTwos brain
5 年前
Andrew Cohen
2988727e
cubewars fixes
5 年前
GitHub
dd6aa7e2
Agent.Heuristic takes an float[] ( #3765 )
5 年前
bhh
4eef4185
added retrained worm static model
5 年前
Andrew Cohen
f41695b9
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
9227f74d
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
6f1f89f6
new soccertwos brain
5 年前
GitHub
9f43ef72
[MLA-488] Fix observations in demonstration drawer ( #3771 )
* WIP observation shapes
* WIP observation shapes
* fix Observation shape serialization
* rename field, redo screenshot
* docstring
* changelog
5 年前
Hunter-Unity
2751b3a4
updated crawlerAgent code to match worm env
5 年前
Andrew Cohen
7c84b6d8
new CubeWar brains
5 年前
Hunter-Unity
1367670d
added new model with worm settings
5 年前
Andrew Cohen
02d26c3f
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Hunter-Unity
6d1b23af
put rewards back in fu for testing
5 年前
Hunter-Unity
dd3b1fbf
comparing rewards in fu vs in decision
5 年前
GitHub
41e4efca
Update art for asymmetric self-play envs ( #3717 )
5 年前
Andrew Cohen
19e579af
new tennis/soccer brains
5 年前
Hunter-Unity
75c7fb85
added more nn files from testing
5 年前
Andrew Cohen
384f6439
reduced laser cd/increased heal
5 年前
Andrew Cohen
931ae7ac
new Goalie/Striker brain
5 年前
Hunter-Unity
9bb867a5
change crawler body mass to 12
5 年前
GitHub
24b387d3
Merge branch 'master' into develop-sac-apex
5 年前
Andrew Cohen
4bc36520
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Hunter-Unity
85ee0ed3
cp
5 年前
Hunter-Unity
76b0686c
added new scene for posing
5 年前
Andrew Cohen
80469267
Merge branch 'internal-policy-ghost' into soccer-2v1
5 年前
Hunter-Unity
ff9a6a91
updated size on grid material
5 年前
Andrew Cohen
9ae19e9d
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Hunter-Unity
77f231b5
cp
5 年前
Andrew Cohen
4468280a
Merge branch 'soccer-2v1' into asymm-envs
5 年前
GitHub
9695b89a
StrikerVsGoalie and SoccerTwos env improvements ( #3699 )
5 年前
Hunter-Unity
ee22bd50
added corgi/probuilder package
5 年前
Ervin Teng
0bc4cc1e
Merge branch 'master' of github.com:Unity-Technologies/ml-agents into develop-sac-apex
5 年前
Andrew Cohen
66c15346
added striker/goalie brain to prefab
5 年前
Hunter-Unity
196c70c7
added posed crawler
5 年前
Arthur Juliani
c577ce26
Merge remote-tracking branch 'origin/master' into develop-add-fire
5 年前
Hunter-Unity
4a9df3a8
posed scene 1
5 年前
GitHub
bc45453b
AcademyStepper to DontDestroyOnLoad ( #3789 )
* AcademyStepper to DontDestroyOnLoad
* Adding a try-catch around DontDestroyOnLoad because it cannot be used in Editor Tests
5 年前
Hunter-Unity
f52150ef
updated light angle
5 年前
Andrew Cohen
9bec75ee
Merge branch 'master' into soccer-2v1
5 年前
Hunter-Unity
fd2f8d61
final pose
5 年前
GitHub
884e1455
Rename monster cube and change fire to red ( #3718 )
5 年前
Andrew Cohen
bdf1c8b9
new soccer brain
5 年前
HH
27ded81d
moved posed scene to SharedAssets
4 年前
Andrew Cohen
bc5c59ec
update soccer-2v1 reward function from test branch
5 年前
Arthur Juliani
3769d943
Merge remote-tracking branch 'origin/master' into develop-add-fire
5 年前
HH
a5e1878a
adjust characters
4 年前
Andrew Cohen
8431ecb5
tennis reward fix
5 年前
Andrew Cohen
5d659946
update tennis reward function
5 年前
HH
e3a08bfb
tweak rot
4 年前
Andrew Cohen
9a25be8c
cumulative final reward soccer
5 年前
Andrew Cohen
1ac4dfb3
update Tennis max step
5 年前
HH
121c75fd
add meta files
4 年前
Ervin Teng
9cd2c034
Merge branch 'master' of github.com:Unity-Technologies/ml-agents into develop-sac-apex
5 年前
Andrew Cohen
f5087cee
modify goalie
5 年前
HH
048bf088
Release 7 scene added
4 年前
Andrew Cohen
95f3ffab
fix max step tennis
5 年前
HH
e54abece
add 2020 year end blog images scenes
4 年前
Andrew Cohen
185d4b35
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
47548ee4
tennis curriculum
5 年前
Andrew Cohen
e2be590d
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
1807f698
increase drag
5 年前
Andrew Cohen
c8cae0dd
reduced bounciness/added downward force tennis
5 年前
Arthur Juliani
3c82bf59
Training runs, but doesn’t actually work
5 年前
Andrew Cohen
3bd33889
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
9f36cd36
added floorhit obs tennis
5 年前
Andrew Cohen
f5c551f2
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
fdcf3f43
Revert "added floorhit obs tennis"
This reverts commit 1d3e1cffda2e49da9cf1944eca4f250d9e551c39.
5 年前
Andrew Cohen
e1660b49
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
8ad1323e
increased scale tennis env
5 年前
Andrew Cohen
29c23484
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
e5c62cb8
update striker vs goalie brain/retrain
5 年前
Andrew Cohen
e2d68e4f
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
1d020fa7
Merge branch 'soccer-2v1' into asymm-envs
5 年前
Andrew Cohen
4d5b1b33
fixed agent action constraint tennis
5 年前
GitHub
0a7e53be
Code Style - apply PascalCase ( #3828 )
* cleanup a few classes
* cleanup raycast code
* more capitalization
* more renames
* changelog and migration
* fix MaxStep in docs
* doc string
5 年前
Andrew Cohen
c4946d31
increase curr to .2
5 年前
GitHub
f8909ab1
Add New 3 Joint Ragdoll Worm Environment ( #3798 )
Co-authored-by: Arthur Juliani <awjuliani@gmail.com>
5 年前
Andrew Cohen
13fd97de
small per timestep reward
5 年前
Andrew Cohen
0c7b4ac4
Merge branch 'soccer-2v1' into asymm-envs
5 年前
GitHub
4399e29f
UI for Ray stacks, rename WriteAdapter to ObservationWriter ( #3834 )
* UI for Ray stacks, rename WriteAdapter to ObservationWriter
* move test
* changelog and migration
5 年前
Andrew Cohen
45d35fa4
downward force tennis agent
5 年前
GitHub
ea0c6fa0
[WIP] Side Channel Design Changes ( #3807 )
* Make EnvironmentParameters a first-class citizen in the API
Missing: Python conterparts and testing.
* Minor comment fix to Engine Parameters
* A second minor fix.
* Make EngineConfigChannel Internal and add a singleton/sealed accessor
* Make StatsSideChannel Internal and add a singleton/sealed accessor
* Changes to SideChannelUtils
- Disallow two sidechannels of the same type to be added
- Remove GetSideChannels that return a list as that is now unnecessary
- Make most methods except (register/unregister) internal to limit users impacting the “system-level” side channels
- Add an improved comment to SideChannel.cs
* Added Dispose methods to system-level sidechannel wrappers
- Specifically to StatsRecorder, EnvironmentParameters and EngineParameters.
- Updated Academy.Dispose to take advantage of these.
- Updated Editor tests to cover all three “system-level” side channels.
Kudos to Unit Tests (TestAcade...
5 年前
Andrew Cohen
f60df1c9
lower agent start
5 年前
Andrew Cohen
29627181
more downward force/constrain y
5 年前
GitHub
5f89f60e
add a description to ProjectSettingsOverrides ( #3838 )
5 年前
Andrew Cohen
e5b883db
added bounce obs to agent/more downward force on ball
5 年前
Andrew Cohen
028a8d59
larger network/6 stacked obs
5 年前
Andrew Cohen
ca6cdff3
fixed broken prefab...
5 年前
GitHub
1e0b022f
[MLA-850] rename namespaces to Unity.MLAgents ( #3843 )
* rename in protos
* rename in C#
* doc changes, migration, changelog
* PR numbers
* fix standalone test path
5 年前
Andrew Cohen
1c4ba1a5
add timestep bonus to loss
5 年前
Andrew Cohen
428b1dfa
Hit bonus for whole exp
5 年前
Andrew Cohen
a1143427
increased entro bonus tennis
5 年前
Andrew Cohen
443935f4
remove time bonus loss
5 年前
GitHub
3d15d51f
apply rider suggestions in Examples ( #3846 )
5 年前
GitHub
c7722f73
[barracuda] Update Barracuda to 0.7.0-preview. ( #3873 )
5 年前
Andrew Cohen
5d22b819
added timepenalty to obs
5 年前
GitHub
4092d937
[Bug fix] Hard reset when team changes ( #3870 )
5 年前
Arthur Juliani
212e2d1d
Merge remote-tracking branch 'origin/master' into develop-add-fire
5 年前
Andrew Cohen
a6e6e63e
timestep penalty on loss only
5 年前
GitHub
f86fc81d
[refactor] Move configuration files to single YAML file ( #3791 )
5 年前
GitHub
731fb88b
[barracuda] Update Barracuda to 0.7.0-preview ( #3875 )
5 年前
Andrew Cohen
b80c48bf
new cube wars brains
5 年前
Andrew Cohen
fb7aa862
remove timepenalty obs
5 年前
Andrew Cohen
251dcc76
remove timepenalty from tennis
5 年前
Andrew Cohen
b0243014
broken prefab..
5 年前
Andrew Cohen
1b10ef6d
clip speed
5 年前
Andrew Cohen
4769cb1e
proximity bonus
5 年前
GitHub
48f33cce
Commit ExpertGrid meta file ( #3883 )
5 年前
Andrew Cohen
39e0bbe9
remove debug log
5 年前
Andrew Cohen
b7bd4c2c
reduce winning reward
5 年前
Andrew Cohen
1c2e1d79
increase beta
5 年前
Andrew Cohen
e0723381
new tennis brain
5 年前
Andrew Cohen
7da1869a
new tennis brain
5 年前
Andrew Cohen
a56643bb
tennis brain to prefab
5 年前
Andrew Cohen
376af981
lower agent height
5 年前
GitHub
f5ef4835
new trained models (except Tennis) ( #3894 )
5 年前
Andrew Cohen
b80d6228
randomize ball/agent spawn
5 年前
GitHub
d8b93f8f
[Bug fix] Hard reset when team changes ( #3870 ) ( #3899 )
5 年前
Andrew Cohen
2c42f577
Merge branch 'master' into asymm-envs
5 年前
Andrew Cohen
843ddfc3
updated heuristics of cubewars
5 年前
Andrew Cohen
d5428487
addforce and static walls
5 年前
Andrew Cohen
cde22a14
fix clipping
5 年前
Andrew Cohen
d77f2566
energy usage penalty to prevent superstition on serve
5 年前
Andrew Cohen
0d943676
inc energy pen
5 年前
Andrew Cohen
1f2453ef
slower tennis
5 年前
Andrew Cohen
a8f2f613
no energy penalty
5 年前
Chris Elion
68b68396
Merge remote-tracking branch 'origin/master' into release_1_to_master
5 年前
Andrew Cohen
8ef0b3a8
opponent observations
5 年前
Andrew Cohen
862a3c02
faster now that it works?
5 年前
Andrew Cohen
1819fbad
0f max height
5 年前
GitHub
d8956399
Merge pull request #3914 from Unity-Technologies/release_1_to_master
Merge Release 1 to master
5 年前
Andrew Cohen
acb85908
hitbonus
5 年前
Andrew Cohen
4907a9cd
slower tennis
5 年前
vincentpierre
c34dd5b6
Merge branch 'master' into develop-gym-wrapper
5 年前
Andrew Cohen
69acdeec
fixed reset tennis
5 年前
Andrew Cohen
98c878ba
fixed rotate action
5 年前
Andrew Cohen
6c42c221
remove debug print
5 年前
Andrew Cohen
a2f8319a
Merge branch 'master' into asymm-envs
5 年前
Andrew Cohen
db1a77e7
rotate over tiemsteps
5 年前
Andrew Cohen
0deacd92
much slower agent
5 年前
Arthur Juliani
89ad3020
Merge remote-tracking branch 'origin/master' into develop-add-fire
# Conflicts:
# ml-agents/mlagents/trainers/policy/tf_policy.py
5 年前
Hunter-Unity
e891d9b5
about to implement orientation cube
5 年前
Andrew Cohen
f74ac6ae
remove rotation hindrance
5 年前
Hunter-Unity
b60cd47c
oCube spawining works. ready to train
5 年前
GitHub
a54aef02
[MLA-1223] Backport Heuristic fixes ( #4176 )
* [bugfix] Make FoodCollector heuristic playable (#4147 )
* Modified the documentation of the Heuristic method (default action = previous action) (#4174 )
Co-authored-by: Ervin T <ervin@unity3d.com>
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
4 年前
Andrew Cohen
8376b862
lower tennis height
5 年前
Hunter-Unity
8a74e6f0
working. about to try com
5 年前
Hunter-Unity
b7ce8a65
ready for training
5 年前
Andrew Cohen
c702bb27
faster ball
5 年前
Hunter-Unity
0bf25b6d
add random rot on episode start
5 年前
Andrew Cohen
b13efd5d
slow down ball/nerf jump
5 年前
Hunter-Unity
f6c8d9fc
feet now alternate but runs backwards
5 年前
GitHub
d09d8f13
better error message, optionally exit if no file ( #3934 )
5 年前
Hunter-Unity
d755b77a
still running with right leg in front
5 年前
Andrew Cohen
d7c2c163
please no more
5 年前
Hunter-Unity
9d46b450
increased joint strength to 40k
5 年前
Hunter-Unity
56a22426
removed texture example
5 年前
Hunter-Unity
3edca8d0
reduced maxAngVel, enabled enhanced determinism, cont spec
5 年前
Arthur Juliani
2b3a6347
Merge remote-tracking branch 'origin/master' into develop-add-fire
5 年前
Hunter-Unity
546c610f
rebuilt walker ragdoll to scale 1
5 年前
Andrew Cohen
765dfb4d
faster x axis/less rotation
5 年前
Hunter-Unity
182aaee3
rebuilt ragdoll ready
5 年前
Andrew Cohen
9beb2a15
serve spawns forward
5 年前
Hunter-Unity
bfb4eddc
update walker pair prefab
5 年前
Andrew Cohen
0b45d365
ball spawns forward
5 年前
Hunter-Unity
b1e5a6f8
fixed bp heirarchy
5 年前
Andrew Cohen
138dc0e3
ball toss
5 年前
Hunter-Unity
6f91aeae
added trained model, renamed scene, usecollisioncallbacks
5 年前
Hunter-Unity
78256e86
updated dynamic platforms
5 年前
Andrew Cohen
95424b93
slower
5 年前
Hunter-Unity
c4943a05
added dynamic walker tf file. max speed 5
5 年前
Hunter-Unity
5f149660
DynamicWalker working. has working nn file
5 年前
Andrew Cohen
c5ce18c7
remove x/y vel, smaller network
5 年前
Hunter-Unity
e037175e
collect local rotations
5 年前
Andrew Cohen
97119431
fix prefab
5 年前
Hunter-Unity
0b02b434
added new dynamic nn file
5 年前
Andrew Cohen
fd7ee405
normalize by hand
5 年前
Hunter-Unity
9e20feef
hip facing reward
5 年前
Andrew Cohen
7e3c3b45
remove hit bonus
5 年前
Andrew Cohen
84f231ce
time penalty
5 年前
Hunter-Unity
d29cc962
fix hip rotation
5 年前
Andrew Cohen
0ae5e040
rename some vars
5 年前
Hunter-Unity
7b35627d
about to clean up code
5 年前
Andrew Cohen
cc79fa0e
no opp obs
5 年前
vincentpierre
6ddfe74f
Merge branch 'master' into develop-gym-wrapper
5 年前
Hunter-Unity
81e976ac
added dirIndicator and orentCubeGizmo
5 年前
Andrew Cohen
f5aa9fd6
no time penalty no opp
5 年前
Hunter-Unity
89e95781
clean up
5 年前
Hunter-Unity
6e0dbc22
cleanup
5 年前
Andrew Cohen
f21304a9
ball
5 年前
Hunter-Unity
12cbca77
updated WalkerStatic scene with new ragdoll
5 年前
Hunter-Unity
10f12c25
cleanup
5 年前
GitHub
4cb7bfe7
add debugging option to ModelOverrider, handle scene resets ( #3997 )
5 年前
Hunter-Unity
da6d25c9
updated walker dynamic demo file. cleanup
5 年前
Hunter-Unity
1316aa98
iterate through list not dict to collect observations
5 年前
Andrew Cohen
13c2a209
added opp, decay eps removed
5 年前
Hunter-Unity
9add4a97
increase gravity to 1.5
5 年前
Hunter-Unity
99eadde6
try 100M steps on walkerdynamic
5 年前
GitHub
f8df9fd2
add ModelOverride to SoccerTwos and Worm scenes ( #4002 )
5 年前
Hunter-Unity
c9821f85
100M steps
5 年前
Hunter-Unity
07266f46
add dir vector obsv
5 年前
Andrew Cohen
6f637d7f
soccer diverged
5 年前
GitHub
ef530217
use directory for model overrides instead ( #4003 )
* use directory for model overrides instead
* fix maxEpisodes setting
5 年前
Andrew Cohen
59a60c1e
Merge branch 'master' into asymm-envs
5 年前
Hunter-Unity
7b9dfa04
testing new nn models
5 年前
Andrew Cohen
efdf8cdf
fixes due to change in master
5 年前
Hunter-Unity
85958dad
try 8x mem for cloud
5 年前
Hunter-Unity
b06dd988
8x batch size for cloud test
5 年前
GitHub
8ba04736
Comment about Goalies and Striker removed from PushBlock Agent ( #4011 )
Probably a Copy&Paste mistake.
5 年前
Hunter-Unity
6b92b01a
epoch 10
5 年前
Chris Elion
7f8258a8
update yamato tests from master
4 年前
Andrew Cohen
a0dc8789
test new sampling method
5 年前
Hunter-Unity
a7693501
cp
5 年前
Hunter-Unity
f17b1075
increase timescale for cloudtraining
5 年前
HH
ad90e6b9
about to implement orientation cube
5 年前
Hunter-Unity
769dbec5
cp
5 年前
HH
9ad4c76c
oCube spawining works. ready to train
5 年前
HH
8839b08a
working. about to try com
5 年前
Andrew Cohen
fe0a077e
passing sampler configs to c#
5 年前
Hunter-Unity
a3f7b980
cp
5 年前
HH
f2ff4fd8
ready for training
5 年前
Chris Elion
2f7174f0
migrate from old branch
5 年前
Andrew Cohen
4464ca46
ignoring commit checks
4 年前
HH
ab4e288a
add random rot on episode start
5 年前
Chris Elion
1e96ad52
update crawler
5 年前
Hunter-Unity
8bcce546
cleanup & put direction indicator in separate script
5 年前
GitHub
9c3b179a
[MLA-1981] Don't allow connections to newer python trainers ( #5370 )
4 年前
HH
cd9a943b
feet now alternate but runs backwards
5 年前
Chris Elion
20b5a157
update scenes and get them training
5 年前
GitHub
ea26ba4f
update to 1.0.8 ( #5395 )
* update to 1.0.8
* remove com.unity.purchasing dep from project
* remove 2018.4 from tests, make sure to user 2020.3
* copy yamato util from main
* test 2020.3 (not .2)
* use downloader in standalone build test
* remove trunk tests, fix version parsing in unit test
* add 2018.4 back to package tests
4 年前
HH
e88f3a32
still running with right leg in front
5 年前
Chris Elion
dc4a001f
copy sensor from robot branch, fix space
5 年前
Chris Elion
3826a38f
remove com.unity.purchasing dep from project
4 年前
HH
2fa61a51
increased joint strength to 40k
5 年前
HH
6b66210a
removed texture example
5 年前
HH
f7e650a6
reduced maxAngVel, enabled enhanced determinism, cont spec
5 年前
HH
ae3e3a2c
rebuilt walker ragdoll to scale 1
5 年前
HH
6b18a012
rebuilt ragdoll ready
5 年前
HH
a1cce613
update walker pair prefab
5 年前
HH
eca1dac5
fixed bp heirarchy
5 年前
Andrew Cohen
e386b829
some cleanups/ interval error checking
4 年前
HH
f879b80a
added trained model, renamed scene, usecollisioncallbacks
5 年前
HH
6df39dfd
updated dynamic platforms
5 年前
HH
29bfb845
added dynamic walker tf file. max speed 5
5 年前
HH
c454a961
DynamicWalker working. has working nn file
5 年前
GitHub
101a8e00
Add Dynamic Walker. Improved Ragdoll Stability/Performance ( #4037 )
* about to implement orientation cube
* oCube spawining works. ready to train
* working. about to try com
* ready for training
* add random rot on episode start
* feet now alternate but runs backwards
* still running with right leg in front
* increased joint strength to 40k
* removed texture example
* reduced maxAngVel, enabled enhanced determinism, cont spec
* rebuilt walker ragdoll to scale 1
* rebuilt ragdoll ready
* update walker pair prefab
* fixed bp heirarchy
* added trained model, renamed scene, usecollisioncallbacks
* updated dynamic platforms
* added dynamic walker tf file. max speed 5
* DynamicWalker working. has working nn file
* collect local rotations
* added new dynamic nn file
* hip facing reward
* Create WalkerDynamic.yaml
* fix hip rotation
* about to clean up code
* added dirIndicator and orentCubeGizmo
* clean up
* clea...
4 年前
Andrew Cohen
22786526
Merge branch 'master' into asymm-envs
5 年前
HH
5de12753
collect local rotations
5 年前
GitHub
8d29368d
Cleanup - remove walker demo file ( #4068 )
* cleanup - remove walker demo file
* move properties in WalkerAgent script. no logic changes
4 年前
HH
8bee075b
added new dynamic nn file
5 年前
Andrew Cohen
9005ad3e
normalize
5 年前
HH
5bf43487
hip facing reward
5 年前
Andrew Cohen
d7e8d25c
reset params broken
5 年前
Andrew Cohen
8e69cd88
fix tennis prefab
5 年前
HH
03da6557
fix hip rotation
5 年前
Andrew Cohen
43d5ef17
fixed opponent setting
5 年前
HH
52795795
about to clean up code
5 年前
HH
0b09b3ab
added dirIndicator and orentCubeGizmo
5 年前
Andrew Cohen
4ba0d98c
cubewar and tennis stability test
5 年前
HH
602638ba
clean up
5 年前
HH
e585f986
cleanup
5 年前
Andrew Cohen
c0f7052b
Merge branch 'master' into develop-sampler-refactor
4 年前
HH
6701bb3b
updated WalkerStatic scene with new ragdoll
5 年前
Andrew Cohen
34ecc7e6
Merge branch 'master' into asymm-envs
5 年前
HH
4bb3f544
cleanup
5 年前
Andrew Cohen
33458d24
running cubewar
5 年前
HH
2d2844bd
updated walker dynamic demo file. cleanup
5 年前
Andrew Cohen
08b9aabc
half damage small agent
5 年前
HH
80a8b040
iterate through list not dict to collect observations
5 年前
Andrew Cohen
99b18d2e
revert tennis timepenalty
5 年前
HH
5b97c300
increase gravity to 1.5
5 年前
GitHub
05d80169
Re-train models for example scenes ( #4106 )
4 年前
Andrew Cohen
34f3ac64
updated cube war
5 年前
HH
d7e8c5e4
try 100M steps on walkerdynamic
5 年前
Andrew Cohen
723403d1
less hit reward cubewars
5 年前
HH
8f463c55
100M steps
5 年前
HH
005f377a
add dir vector obsv
5 年前
GitHub
fefbc038
Merge pull request #4109 from Unity-Technologies/release_3_merge_master
Release 3 merge to master
4 年前
HH
08dd453b
testing new nn models
5 年前
GitHub
9514cbbb
Update Crawler Ragdoll & Add DirectionIndicator ( #4097 )
* about to implement orientation cube
* oCube spawining works. ready to train
* working. about to try com
* ready for training
* add random rot on episode start
* feet now alternate but runs backwards
* still running with right leg in front
* increased joint strength to 40k
* removed texture example
* reduced maxAngVel, enabled enhanced determinism, cont spec
* rebuilt walker ragdoll to scale 1
* rebuilt ragdoll ready
* update walker pair prefab
* fixed bp heirarchy
* added trained model, renamed scene, usecollisioncallbacks
* updated dynamic platforms
* added dynamic walker tf file. max speed 5
* DynamicWalker working. has working nn file
* collect local rotations
* added new dynamic nn file
* hip facing reward
* Create WalkerDynamic.yaml
* fix hip rotation
* about to clean up code
* added dirIndicator and orentCubeGizmo
* clean up
* cleanup
* up...
4 年前
Andrew Cohen
03eef40b
constrain x tennis
5 年前
Andrew Cohen
21f871db
Merge branch 'develop-constant-decay' into asymm-envs
5 年前
HH
ef3be52c
try 8x mem for cloud
5 年前
HH
25d7ba5e
8x batch size for cloud test
5 年前
Andrew Cohen
0c17dc1b
cannot hit scenery tennis
5 年前
HH
2cce3bbe
epoch 10
5 年前
GitHub
59fb7aa5
ML-Agents Extensions package ( #4105 )
* extensions package first pass
* add extensions test to yamato
* cleanup docs folder
* triggers and pack
* yamato - different jobs for each package
* disable coverage checks for extensions
* trigger
* Unity Companion License
4 年前
Andrew Cohen
7475ad11
tunneling is a loss
5 年前
Andrew Cohen
31a5b2ee
4096 batch
5 年前
HH
c6264bd1
cp
5 年前
Arthur Juliani
9724c9ac
Merge master
4 年前
HH
f7dd600f
increase timescale for cloudtraining
5 年前
Andrew Cohen
346a90ba
move agent back
5 年前
HH
65b80abb
cp
5 年前
Andrew Cohen
1625cca0
faster tennis ball
5 年前
yanchaosun
c2d6f5c0
basic implementation
4 年前
Andrew Cohen
3ba64554
close
5 年前
HH
ba835a22
cp
5 年前
Andrew Cohen
c8ecf7ea
remove opp obs
5 年前
GitHub
8f33b60c
Add 'old' crawler env ( #4171 )
* reverted old crawler
* remove crawler prefab
* update Crawler prefabs
* put together prefabs
* old and new crawler envs
4 年前
Andrew Cohen
b37c4493
floor material
4 年前
Andrew Cohen
e76bd1ac
remove tp
5 年前
HH
3e204ae5
cleanup & put direction indicator in separate script
5 年前
Andrew Cohen
30161f21
reduce max step
5 年前
Chris Elion
ef1c3b91
custom one-hot visual sensor
4 年前
Andrew Cohen
3d228cb5
slower x
5 年前
Chris Elion
ed8a249c
fix commit
4 年前
Andrew Cohen
a5383d7e
move agent service forward
5 年前
Chris Elion
1a485482
comments
4 年前
Andrew Cohen
72b0e0ab
bouncier ball
5 年前
HH
81df35a9
added final nn files
5 年前
Chris Elion
2677d314
updated scene
4 年前
Andrew Cohen
6c5741a1
move agents back
5 年前
HH
8c4300e3
more cleanup
5 年前
Jonathan Harper
5906cd61
WIP dump - working prototype
Updates to CameraSensor / CameraSensorComponent:
* new settings object to set up layer masks, depth, etc
* support for arbitrary # channels returned by the sensor
* channels include depth, layer mask, RGB, or grayscale
Updates to VisualFoodCollector:
* food doesn't roll (for now), simplifies learning without stacking
* added layers
Updates to mlagents_envs:
* decode multiple pngs packed into a single observation
4 年前
Andrew Cohen
1f305f23
no latest model
5 年前
HH
be27b1e5
new walker image for docs
5 年前
Andrew Cohen
e7e25c16
correct max step...
5 年前
Jonathan Harper
7656f419
More experimentation
4 年前
Andrew Cohen
6852707a
better
5 年前
GitHub
a28e2767
Update add-fire to latest master, including Policy refactor ( #4263 )
* Update Dockerfile
* Separate send environment data from reset (#4128 )
* Fixed a typo on ML-Agents-Overview.md (#4130 )
Fixed redundant "to" word from the sentence since it is probably a typo in document.
* Updated the badge’s link to point to the newest doc version
* Replaced all of the doc to release_3_doc
* Fix 3DBall and 3DBallHard SAC regressions (#4132 )
* Move memory validation to settings
* Update docs
* Add settings test
* Update to release_3 in installation.md (#4144 )
* rename to SideChannelManager +backcompat (#4137 )
* Remove comment about logo with --help (#4148 )
* [bugfix] Make FoodCollector heuristic playable (#4147 )
* Make FoodCollector heuristic playable
* Update changelog
* script to check for old release links and references (#4153 )
* Remove package validation suite from Project (#4146 )
* RayPerceptionSensor: handle empty and invalid tags (#4155...
4 年前
GitHub
bece7888
[bugfix] Make FoodCollector heuristic playable ( #4147 )
* Make FoodCollector heuristic playable
* Update changelog
4 年前
Andrew Cohen
70b36614
increase max step
5 年前
HH
f1e2324c
cleanup
5 年前
Andrew Cohen
01af358f
tp
5 年前
GitHub
4642488e
Remove package validation suite from Project ( #4146 )
4 年前
HH
b793d93a
update crawler code
5 年前
HH
7f9b9567
rebuilt crawler ragdoll to scale of 1
5 年前
HH
375bdd52
Update CrawlerDynamic and CrawlerStatic scenes
5 年前
HH
e05e152a
adjusted crawler joint and mass
5 年前
HH
15cb62b9
add trained crawler nn file
5 年前
Andrew Cohen
acd219ce
faster ball real max step
5 年前
HH
3230d564
try invtranspoint
5 年前
GitHub
7f0bb6bd
Yamato inference tests ( #4066 )
* better errors for missing constants
* run inference in yamato after training
* add extension
* debug subprocess args
* fix exe path
* search for executable
* fix dumb bug
* -batchmode
* fail if inference fails
* install tf2onnx on yamato
* allow onnx for overrides (expect to fail now)
* enable logs
* fix commandline arg
* catch exception from SetModel and exit
* cleanup error message
* model artifacts, logs as artifacts, fix pip
* don't run onnx
* cleanup and comment
* update extension handling
4 年前
vincentpierre
599d7e9f
Merging master
4 年前
HH
86ac4c3e
cp
5 年前
Ervin Teng
28596129
Send sensor types to Python
4 年前
HH
438b02db
walk use tp
5 年前
GitHub
323f2417
FoodCollectorAgent - don't convert bool to int ( #4169 )
4 年前
HH
7afa1761
Merge branch 'master' into hh/develop/ragdoll-updates
4 年前
HH
d6d3e16b
clean up code
4 年前
Ervin Teng
2b622c78
Example 3DBall with TaskSensor
4 年前
HH
59135ac4
update prefabs
4 年前
HH
e1c94e22
crawler shorter legs
4 年前
GitHub
6ee553d8
Modified the documentation of the Heuristic method (default action = previous action) ( #4174 )
* Modifying the documentation to explain that Heuristic method default action will be the previous action decided by the heuristic. Changing this behavior would be a breking change.
* Rephrase the working of the documentation of the default action of the Heuristic method
* Forgot an import
4 年前
Ervin Teng
2fc4fe16
Add AgentParametersChannel
4 年前
GitHub
0e0daf47
[add-fire] Merge post-0.19.0 master into add-fire ( #4328 )
4 年前
HH
41c9334a
add new crawler demo files
4 年前
GitHub
9d2e4268
Revert "[add-fire] Merge post-0.19.0 master into add-fire ( #4328 )" ( #4330 )
This reverts commit 9913e71b6f35f1e11027a4a571a65533caf285ac.
4 年前
GitHub
8b913a96
Add TargetController/OrientationCubeController Components & Bugfix ( #4157 )
* added Target and OCube controllers. updated crawler envs
* update walker prefab
* add refs to prefab
* Update Crawler.prefab
* update platform, ragdoll, ocube prefabs
* reformat file
* reformat files
* fix behavior name
* add final retrained crawler and walker nn files
* collect hip ocube rot in world space
* update crawler observations and update prefabs
* change to 20M steps
* update crwl prefab to 142 observ
* update obsvs to 241. add expvel reward
* change walkspeed to 3
* add new crawler and walker nn files
* adjust rewards
* enable other pairs
* add RewardManager
* cleanup about to do final training
* cleanup add nn files for increased facing rew reduced height rew
* try no facing rew
* add vel only policy, try dy target
* inc torq on cube
* added dynamic cube nn. gonna try 40M steps
* add 40M step test, more cleanup
* ch...
4 年前
HH
8317e113
Dr Crocodile - added models and level 1
4 年前
HH
82b5af24
cleanup and update docs
4 年前
HH
b812af99
add more pickup items
4 年前
GitHub
6526ac1d
Add old walker ( #4212 )
* old walker
* update walker to master
* revert walker to release 3
* fix prefab
* fix obs size
* rename behavior
* camera follow
4 年前
HH
bcc2f085
added lights
4 年前
HH
f53bc0b3
cleanup
4 年前
HH
d091183a
update rooms
4 年前
HH
1fe76781
remove cube left in scene
4 年前
HH
0854bae7
cleanup
4 年前
GitHub
32dd1e86
update walker ( #4217 )
4 年前
HH
8ff77161
remove tab
4 年前
HH
8fcca582
cp
4 年前
GitHub
559549e4
Add dynamics change to crawler ( #4218 )
* update walker
* add dynamics change crawler
4 年前
HH
62abef9f
more cleanup
4 年前
HH
582d0cd7
remove lightmaps
4 年前
HH
78455cde
add a2 env
4 年前
HH
44594919
update leg joint settings
4 年前
GitHub
09bfeb0e
enforce warnings-as-errors, fix warning ( #4191 )
4 年前
HH
98751025
add trees
4 年前
HH
96d7b2a5
add new nn files
4 年前
HH
57e8a97b
move cube update logic outside if statement
4 年前
GitHub
8eefdcd3
Refactor of Curriculum and parameter sampling ( #4160 )
* Introduced the Constant Parameter Sampler that will be useful later as samplers and floats can be used interchangeably
* Refactored the settings.py to refect the new format of the config.yaml
* First working version
* Added the unit tests
* Update to Upgrade for Updates
* fixing the tests
* Upgraded the config files
* Fixes
* Additional error catching
* addressing some comments
* Making the code nicer with cattr
* Added and registered an unstructure hook for PrameterRandomization
* Updating C# Walljump
* Adding comments
* Add test for settings export (#4164 )
* Add test for settings export
* Update ml-agents/mlagents/trainers/tests/test_settings.py
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
* Including environment parameters for the test for settings export
* First documentation up...
4 年前
HH
27289f39
init
4 年前
HH
6486c762
removed demo file
4 年前
HH
decf9a0a
Add reward manager and hurryUpReward
4 年前
HH
177336bf
normalized max raycast dist to 1
4 年前
HH
51969c6f
fix hurry reward/ add awful first training
4 年前
HH
1ea8ad6f
Turn off head height and hurry rew
4 年前
HH
0fdac847
Merge branch 'master' into hh/develop/crawler-ragdoll-updates
4 年前
HH
ce8e55b4
changed max speed to 15. added small hh rew
4 年前
HH
9a531f38
add new nn files and update prefabs
4 年前
GitHub
30a5274c
better logging for NaN rewards ( #4205 )
4 年前
HH
35235345
add NaN check for reward manager. start vel penalty
4 年前
HH
2d5516c9
try obsv rel to body
4 年前
HH
3a2c2605
add bpVel pen
4 年前
HH
29d7eb2b
try loc pos
4 年前
HH
9c787458
add new BPVelPen nn file
4 年前
HH
9d8c4aa2
walker no oc
4 年前
HH
5f0d0658
remove outdated nn file
4 年前
HH
c464a9f1
try inv tr dir
4 年前
HH
f83c2f8a
add randomize speed bool
4 年前
HH
76e8a48f
fix missing prefab
4 年前
HH
3c2ff715
try rewad product
4 年前
Arthur Juliani
6bee0fd1
Merge master
4 年前
HH
b0ef077a
change coeff to 1
4 年前
HH
58102fdc
try avg vel of all bp for reward
4 年前
HH
3b07fb7d
move outside loop
4 年前
HH
13279c74
try linear inverselerp for vel
4 年前
GitHub
20f1386a
Don't drop multiple stats from the same step ( #4236 )
4 年前
HH
1e10aea6
add avg rew matchspeed15 nn file. looks much better
4 年前
HH
4b7c2259
save scene
4 年前
HH
5a2fff7d
no hand penalty, random walk speed
4 年前
Andrew Cohen
e10c9abd
update crawler
4 年前
HH
375fcfff
fix inverse lerp
4 年前
HH
9e6edb6c
try new reward falloff
4 年前
HH
ada8cccc
added new nn file. don't allow hand contact
4 年前
GitHub
82602a69
Longer demos for ragdoll envs ( #4247 )
4 年前
HH
74478199
update obsv
4 年前
HH
8bbfb46c
remove hh rew. add trained no-hh model
4 年前
Andrew Cohen
d8c123a0
Merge branch 'master' into sensitivity
4 年前
HH
16b33ca8
add new nn file
4 年前
HH
e2217a9a
new curve
4 年前
HH
9aed2094
add new models. try no reset
4 年前
Ruo-Ping Dong
262f38ea
add basketball example
4 年前
HH
b99fcfdd
add hh rew
4 年前
Ruo-Ping Dong
9fb30ad9
add basketball example
4 年前
HH
652abe24
clamp hh
4 年前
HH
cc0c893d
zero rewards if ground contact
4 年前
HH
675f1ae4
switch to approved with moving target
4 年前
HH
611623e7
try new dot
4 年前
HH
1e3221f8
add shifted dot and reg dot nn file
4 年前
GitHub
42d61c09
Crawler fix ( #4270 )
* fixed crawler issue
* fixed crawler again
4 年前
HH
00cb4c89
add WalkerStaticVariableSpeedScene and PPO config
4 年前
HH
281a4e60
add a NaN debug for action values
4 年前
HH
7c63197e
start dynamic cleanup and more debug for NaNs
4 年前
HH
40b5cb96
more cleanup
4 年前
Arthur Juliani
1a123641
Merge remote-tracking branch 'origin/master' into r5-master
4 年前
HH
bdc9c1a4
add WalkerDynamicVarialbeSpeed scene and update prefabs
4 年前
GitHub
3a0d6e54
[MLA-1135] Physics sensors - optional reference body ( #4276 )
4 年前
HH
950f9a8f
add trained static walker nn file
4 年前
HH
f8bd79cb
About to do cleanup
4 年前
GitHub
80e54ea0
[examples] Behavior Name correction ( #4294 )
4 年前
HH
977287dd
add all scenes
4 年前
HH
0840139d
add new no hh nn models and update prefabs
4 年前
HH
d3b73212
add hh rew
4 年前
HH
292e8743
try 15k strength. reset jdcontroller to master
4 年前
Andrew Cohen
78943972
add l2 penalty 3dball
4 年前
HH
4a1a3210
remove h rew 10k strength
4 年前
Andrew Cohen
7b843666
l2 3dball
4 年前
GitHub
323d911b
Add internal API for Actuators along with their tests. ( #4297 )
4 年前
HH
af8b2e3b
trying to figure out shuffle foot regression. added 10k no hh model
4 年前
Andrew Cohen
c6416bef
3dball hard rew
4 年前
Ervin Teng
dc937d5c
Merge branch 'master' into develop-add-fire-mm
4 年前
HH
7083559c
about to train 20k strength, no hh, no rolling targ 30M
4 年前
Andrew Cohen
1c4d9554
fix action dims
4 年前
yanchaosun
42c9ba43
reuse encoder and linear
4 年前
GitHub
1e76f8d0
Merge pull request #4331 from Unity-Technologies/develop-add-fire-mm2
[add-fire] Merge post-0.19.0 master into add-fire (ver. 2)
4 年前
Ervin Teng
4ebccf97
Merge branch 'develop-add-fire' into develop-add-fire-sac-lst
4 年前
Andrew Cohen
598826fe
Merge branch 'develop-add-fire' into develop-add-fire-bc
4 年前
Ruo-Ping Dong
d3eb6c46
Merge branch 'develop-add-fire' into develop-add-fire-checkpoint
4 年前
HH
9c81a98f
fixed shuffle step regr with 20k no hh
4 年前
GitHub
957010bf
Fix worm environment. ( #4337 )
* Fix worm environment.
* Remove dead code from WormAgent.cs
* remove target logic. change obsv loop to list instead of dict
Co-authored-by: HH <brandonh@unity3d.com>
4 年前
HH
ee810790
update prefabs with new models. walkerstatic failed to train
4 年前
yanchaosun
b40bd941
new 3dball rewards
4 年前
HH
e6a52bd4
saved scene
4 年前
HH
0560848f
implemented distToTarget Instead of targetPos
4 年前
GitHub
7852bc20
Update NN files for the release. ( #4353 )
4 年前
HH
f06849e0
add dist observ nn files
4 年前
GitHub
10743ac0
Remove XR settings asset. ( #4355 )
4 年前
HH
7979a7a3
more cleanup
4 年前
HH
8eaddb61
Merge branch 'master' into hh/develop/loco-walker-variable-speed
4 年前
HH
51cbb887
reduce maxSpeed to 10, update prefabs
4 年前
HH
c34ea517
max dist 50 avg core vel
4 年前
Andrew Cohen
ad723e0e
fixed crawler issue
4 年前
HH
f85b79da
cleanup
4 年前
Andrew Cohen
18704e34
fixed crawler again
4 年前
HH
c038362c
use all bp for avg vel
4 年前
Andrew Cohen
b2efa00b
update crawler rewards and gravity
4 年前
GitHub
7795ce5f
increase grav multiplier ( #4283 )
4 年前
GitHub
b51347ac
New Variable Speed Walker Environments ( #4301 )
* init
* Add reward manager and hurryUpReward
* fix hurry reward/ add awful first training
* Turn off head height and hurry rew
* changed max speed to 15. added small hh rew
* add NaN check for reward manager. start vel penalty
* add bpVel pen
* add new BPVelPen nn file
* remove outdated nn file
* add randomize speed bool
* try rewad product
* change coeff to 1
* try avg vel of all bp for reward
* move outside loop
* try linear inverselerp for vel
* add avg rew matchspeed15 nn file. looks much better
* save scene
* no hand penalty, random walk speed
* fix inverse lerp
* try new reward falloff
* cleanup
* added new nn file. don't allow hand contact
* update obsv
* remove hh rew. add trained no-hh model
* add new nn file
* new curve
* add new models. try no reset
* add hh rew
* clamp hh
* zero rewards if ground contact
* switch to approved with movi...
4 年前
HH
f5d3ef52
reimplement cube relTargetPos
4 年前
yanchaosun
d694bef9
merge
4 年前
HH
691c06e8
init
4 年前
HH
1bbd76fe
update prefabs
4 年前
Ervin Teng
d65a9326
Merge branch 'master' into develop-add-fire-mm3
4 年前
Ruo-Ping Dong
d57aa9ab
Merge branch 'develop-add-fire-mm3' into develop-add-fire-checkpoint
4 年前
HH
ef62939e
updating prefabs
4 年前
HH
d384a902
add relPos clamped to 100m models
4 年前
HH
6c67bf4e
cleanup
4 年前
GitHub
bd6bcd2f
Merge master and add Saver class for save/load checkpoints
4 年前
HH
756099ec
more prefab cleanup
4 年前
vincentpierre
2d7d605d
reverting Project settings
4 年前
GitHub
bb9417f7
Update example environments to use the Actuator API ( #4363 )
4 年前
HH
1f6533c7
more cleanup
4 年前
Ervin Teng
42e25b25
Merge branch 'develop-add-fire' into develop-add-fire-memoryclass
4 年前
GitHub
8bbcfc28
fix deleted observations ( #4366 )
4 年前
HH
0779523e
remove unused prefabs
4 年前
Christopher Goy
5a233353
Merge remote-tracking branch 'origin/master' into release_6-to-master
4 年前
HH
ad012869
remove unused code
4 年前
GitHub
e7916b08
add pre-commit hook for dotnet-format ( #4362 )
4 年前
HH
b2b2e28a
replace demo files
4 年前
Christopher Goy
061a6c43
Merge remote-tracking branch 'origin/master' into release_6-to-master
4 年前
HH
e98c0280
remove demorecorder
4 年前
GitHub
99ccb7b9
Merge pull request #4369 from Unity-Technologies/release_6-to-master
Merge release 6 back into master
4 年前
HH
af4792a6
reset ppo learn.py to master
4 年前
GitHub
9cc4e0b2
update demonstration meta files ( #4367 )
4 年前
HH
2080c287
Merge branch 'master' into hh/develop/loco-crawler-variable-speed
4 年前
HH
d4bd7fe6
Merge branch 'master' into hh/develop/loco-walker-variable-speed
4 年前
Ervin Teng
d52443a5
Merge branch 'master' into develop-shortenstrikervsgoalie
4 年前
HH
1dd1dd81
spawn a target
4 年前
GitHub
31919e08
[MLA-1267] Account for actuators in training and inference. ( #4371 )
4 年前
HH
011e55df
add brains
4 年前
HH
7e7743d1
update static prefabs
4 年前
HH
ab60431e
cleanup from PR review
4 年前
yanchaosun
49d6b70c
crawler: max episode length=1000; new config: 1 forward layer
4 年前
HH
cf00e6e9
enable enhanced determinism
4 年前
HH
510c146e
more cleanup
4 年前
HH
fadb21fa
add code comments
4 年前
HH
24cf37c2
reset manifest
4 年前
HH
83498690
observe velocity delta
4 年前
HH
e3b1c5cf
add nn files. update to 15M steps
4 年前
HH
f8059bb5
add additional velocity observations
4 年前
HH
5ee69715
update prefabs
4 年前
HH
60916d58
add hh
4 年前
HH
5e2d40f1
increase max speed to 15
4 年前
HH
7dfce2b2
add trained models. remove hh rew
4 年前
HH
f2186074
add new local model for 15 speed
4 年前
HH
07115ce4
remove multiple walk dir methods because its confusing
4 年前
HH
6526c72c
update prefabs
4 年前
HH
3006f5cd
update walker static vs prefabs
4 年前
yanchaosun
8f8ec11b
crawler: add time reward
4 年前
HH
f0a12c70
update configs/prefabs
4 年前
Scott Jordan
52ec9230
Merge branch 'develop-taggedobservations' into active-variablespeed
4 年前
HH
702567d7
add new trained models and romove old ones
4 年前
HH
a9d9ea4c
Merge branch 'master' into hh/develop/loco-crawler-variable-speed
4 年前
Scott Jordan
3d98516d
incorporated task parameter channel branch
added the ability to set task parameters from python
4 年前
HH
64dde24e
add new demo files
4 年前
HH
19097dc5
cleanup
4 年前
Scott Jordan
d695c044
initial addition of active learning (incomplete)
4 年前
HH
f7716bb0
reset script to master
4 年前
yanchaosun
82a4914e
1k time reward 0.001
4 年前
HH
4b6f735d
added final nn models
4 年前
HH
ebc45428
custom setter for TargetWalkingSpeed
4 年前
Anupam Bhatnagar
f4f1a8d9
merge master into trainer-plugin branch
4 年前
HH
d6f1cb2d
add new demos and do more cleanup.
4 年前
Scott Jordan
56745026
Initial commit of running active learning code
Active learning code is running on walker variable speed. Needs to be tested to see if it is working.
4 年前
HH
ced04826
add meta files
4 年前
Scott Jordan
78f8a9a2
Updated task manager
active learning is no optional and defaults to uniform sampling of tasks.
Renamed ActiveLearningTaskManager to just TaskManager
4 年前
HH
2bf6ae0e
cleanup per PR suggestions
4 年前
yanchaosun
db30f918
push block
4 年前
yanchaosun
4be4f1d1
new reacher env
4 年前
GitHub
6db9520c
enable enhanceddeterminism in physics settings ( #4423 )
4 年前
GitHub
a117c932
Grid Sensor ( #4399 )
Co-authored-by: Chris Elion <chris.elion@unity3d.com>
4 年前
yanchaosun
e39986ed
block larger feature size; reacher fix and new reward
4 年前
yanchaosun
7dac3284
push block more steps
4 年前
GitHub
582859b6
New Crawler Variable Speed Scenes ( #4382 )
* init
* updating prefabs
* spawn a target
* add brains
* update static prefabs
* enable enhanced determinism
* reset manifest
* add nn files. update to 15M steps
* update prefabs
* increase max speed to 15
* add new local model for 15 speed
* update prefabs
* add configs
* update configs/prefabs
* cleanup
* added final nn models
* add new demos and do more cleanup.
* add meta files
* add RigidbodySensor
* update prefab. about to retrain
* remove body pen
* add fixed crawler & retrained nn file, new demos
* train 10M steps
* Update Crawler Docs
* more prefab cleanup
* add meta files
* Update Project/Assets/ML-Agents/Examples/Crawler/Scripts/CrawlerAgent.cs
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
* remove unused prefab
* update comment
* add summary tags
* cleanup and add more comments
* remove unused prefab
* Update P...
4 年前
GitHub
cc10cd82
Worm Ragdoll & Env Updates ( #4413 )
* add worm updates
* add rewman
* cp
* normalize rewards
* only cookie
* try 20M. Add3.5Mnn file
* reduce strength to 3000spring
* facing reward troubleshooting
* Update WormAgent.cs
* troubleshoot nan
* try product of rewards
* train 5M steps
* try end episode on target touch
* fix joint obsv
* use 7M steps
* added nn file for observation joint fix. looks great
* don't end episode
* remove old code
* refactor to patterns used in walker & crawler
* add auto-setup code
* reformat
* use head vel
* remove unneeded observ. update prefabs
* update static scenes
* keeps rolling. added debug. try 5 m/s
* gate the facing reward based on angle tolerance
* added 10ms_angle30rew_nn files
* use fromto rot
* use 7M steps
* add new trained files. cleanup code and prefabs
* use avgvel. add code comments
* remove unused method
* add more comments
* Update Learning-E...
4 年前
yanchaosun
27dffa4d
new reacher reward
4 年前
yanchaosun
883361ee
reacher new reward: action penalty and constant not-reaching-goal penalty
4 年前
yanchaosun
7bc457f8
change reward function: stress less on action
4 年前
yanchaosun
1d475ec4
prefab
4 年前
yanchaosun
8e2945cf
stack observations
4 年前
GitHub
fc46ff58
Add Smoothing to CameraFollow script ( #4422 )
* update cam script & add comments
* use fixedDeltaTime in SmoothDamp
4 年前
yanchaosun
85549b2b
reacher: stack observation. with the original reward function
4 年前
GitHub
fcbc47b5
load onnx files for testing ( #4208 )
* load onnx files for testing
* changelog
* update interface
4 年前
yanchaosun
92c3facf
distance based penalty
4 年前
HH
17162d1d
rebuild hallway walls using primatives
4 年前
Ervin Teng
60eacc0d
Merge branch 'master' into develop-adjust-cpu-settings
4 年前
yanchaosun
68544337
lift pink balls to another level
4 年前
yanchaosun
ecef019c
fix
4 年前
GitHub
8334562a
rebuild hallway walls using primatives ( #4445 )
4 年前
yanchaosun
1a9aaaf6
model weights and large transfer learning weight
4 年前
yanchaosun
96ffdb03
new reachers
4 年前
yanchaosun
8f67cd40
smaller learning rate
4 年前
yanchaosun
abbbd0ea
dummy 2 bug fixed
4 年前
Andrew Cohen
3997b14b
Merge branch 'master' into develop-hybrid-actions
4 年前
GitHub
056e9bcc
[release7] Fix missing script warnings when starting the scene ( #4486 )
* fix missing scrips in GridFoodCollector
* remove comments, enable timers
4 年前
GitHub
fb177485
Merge pull request #4500 from Unity-Technologies/r7-merge-to-master
[release7] merge release branch back to master
4 年前
Ervin Teng
f83f10a2
Merge branch 'master' into develop-policyonly
4 年前
Ervin Teng
f9ff3efe
Merge branch 'develop-policyonly' into develop-sac-targetq
4 年前
Andrew Cohen
7c0aa77b
Merge branch 'develop-actions-out' into develop-hybrid-actions
4 年前
GitHub
f5bb26d2
WallJump - handle Agent starting before ModelOverrider ( #4502 )
* handle Agent starting before ModelOverrider
* allow multiple override types
4 年前
HH
9721667b
adjusted debug gizmos
4 年前
HH
2b8f8631
Update SoccerTwos.unity
4 年前
HH
0d42b277
train combo. added nn files.
4 年前
HH
583dd767
not training well. not sure why
4 年前
HH
d02c90f6
added more variants
4 年前
HH
ac3b8705
creature crawler working
4 年前
HH
4d720feb
added trained 6 leg creature
4 年前
HH
1912e47a
Dynamic Sensor Benchmarks In
4 年前
GitHub
9e1a28c2
Add vector flag of agent's frozen state to VisualFoodCollector ( #4511 )
VisualFoodCollector is now an example environment of using a mix of visual and vector observation and is able to train with default config file.
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
4 年前
HH
4163ce43
Update CrawlerGRAVEYARD.unity
4 年前
HH
c0bcb416
adjust default settings
4 年前
GitHub
b33e310f
Add Visual3DBall scene ( #4513 )
* Add Visual3DBall scene which use visual observations with stacking
4 年前
HH
72278e8a
change scene name
4 年前
HH
09b94c96
cp
4 年前
HH
ca8517c5
working
4 年前
HH
533b479a
Update PushBlockActuatedSensor.unity
4 年前
GitHub
1f9f1798
[Test] Make tests to compare compressed and non-compressed obs ( #4547 )
* Move folders around
* [Make tests to compare compressed and non-compressed obs
> Make `Utilities.TextureToTensorProxy` public (debatable)
> GridSensor can now be compressed or uncompressed
> Added 2 scenes and an integration test to make sure compressed and uncompressed obs are the same
* fix typo
* renaming file so pytest will not try to run it
* rename yamato config file
* Need standalone build for 2019.4
* Running only on 2018.4 since it seems to be the only version thhat can build
* Typo in the name of the testing script
* Renaming the scene
So this is funny Python bug but
`a.strip(".unity")` will remove the last letter of `a` (before the .unity) if the last letter is "t"
not kidding, try :
```
a = "fewgfwegwrgvrt.unity"
a.strip(".unity")
```
* using splitext rather than strip to find executable name`
* Rename and move TextureToTensor
* Addressing comments
* re...
4 年前
GitHub
30c8caeb
Moving back the template to the Examples folder rather than shared assets ( #4559 )
4 年前
GitHub
69336624
remove demonstration recorder from crawler agent ( #4556 )
4 年前
GitHub
c02ecf14
Add useVectorFrozenFlag option in FoodCollector ( #4552 )
4 年前
Ruo-Ping Dong
336a2af3
remove demonstration recorder from crawler agent ( #4556 )
4 年前
Ruo-Ping Dong
7d42c559
Merge branch 'master' into r8_merge_to_master
4 年前
GitHub
dc40c926
Merge release 8 back to master ( #4570 )
4 年前
Andrew Cohen
e5f14400
Merge branch 'master' into develop-hybrid-actions-singleton
4 年前
Andrew Cohen
2f870407
bullet hell game
4 年前
Ervin Teng
73844bf3
Add battle food arena scene
4 年前
Ervin Teng
56196761
hyperparameteers and tweaks
4 年前
Chris Elion
674540de
try training tests on windows (breaks other tests)
4 年前
Ervin Teng
aeb0674e
Lots of small fixes, better obs
4 年前
Ervin Teng
2e03fc8b
Fix agent obs
4 年前
Ervin Teng
570a6291
Add walls and make spawn areas
4 年前
Ervin Teng
6f510af8
Add laser that stops at walls
4 年前
GitHub
a4ba90ec
[bug-fix] Fix ONNX export/Barracuda import for continuous actions ( #4608 )
* Use torch.ones rather than expand
* use tf.cat instead of tf.expand for Barracuda
* fix error msg handling
* handle loading exception
Co-authored-by: Chris Elion <chris.elion@unity3d.com>
4 年前
Ervin Teng
798a5613
Cooldown timer for laser
4 年前
Ervin Teng
e8e83b95
Add observation to agent
4 年前
Ervin Teng
17f4520e
Add Deathmatch mode
4 年前
Ervin Teng
e2caa20f
Simple scene
4 年前
vincentpierre
a3a9a56b
Merge branch 'exp-multi-head-attention' into exp-bullet-hell
4 年前
Ruo-Ping Dong
9e08be87
Merge branch 'master' into release_9_branch_merge
4 年前
HH
307ea28d
Added FPS scene with new camera, character, and gun. no agent yet
4 年前
Ervin Teng
8a8cd93b
Some more tweaks
4 年前
GitHub
90a9d214
Match3 example ( #4515 )
4 年前
HH
6d608fc1
cp
4 年前
Ervin Teng
83837970
Bring back cooldown
4 年前
GitHub
1b933a0a
Add Graphics to the Match3 env ( #4627 )
* first pass - matching and gravity work
* valid moves WIP
* move enumeration, simple checking
* move detection
* fix up/down directions
* actuator and heuristic
* sensor
* reset if no moves
* add masking, retrain
* cleanup
* config update
* optionally force heuristic mode
* compressed observations, reset board if no moves
* TFModel meta files
* WIP abstract
* random move heuristic
* rename Move fields, migrate IsValid logic
* comments
* reduce dependency on Match3Agent
* abstractBoard as monobehavior
* move actuator to extensions
* move sensor to extensions
* update components in scene
* undo accidental checkins
* drawing cleanup, custom maxMoves logic
* meta file
* prefab, use 8 agents
* different behaviors
* timers, early out
* unit tests
* docs
* Apply suggestions from code review
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>...
4 年前
HH
50365b74
add strafe
4 年前
Ervin Teng
a572f45a
Actually remove cap
4 年前
vincentpierre
b09dfd0a
Still not working
4 年前
Ervin Teng
89489ae0
Invert divide by 3 in log prob
4 年前
HH
f7b4daac
input fix
4 年前
Ervin Teng
4d29947a
Remove central wall
4 年前
vincentpierre
c264b4da
additional changes
4 年前
GitHub
88d3ec3e
Merge master into hybrid actions staging branch ( #4704 )
4 年前
HH
18ab67a0
add agent script
4 年前
vincentpierre
daa75eb5
block agent rotation
4 年前
HH
281e0be1
added sensors & controls UI
4 年前
Ervin Teng
b1559550
Change to 1v1
4 年前
vincentpierre
0b3e24dd
block agent rotation
4 年前
Ervin Teng
66f3d2ad
Merge branch 'develop-contentropy' into develop-torch-clip
4 年前
HH
bf7ef06d
add more structure types
4 年前
vincentpierre
9b3d4ade
Fixing some issues
4 年前
HH
1065a4e5
update projectile
4 年前
vincentpierre
864d8d6c
fixing an important issue in the sensor
4 年前
Ervin Teng
e05e0aef
Collaborative walljump
4 年前
Chris Elion
8cf87ed6
match3 settings
4 年前
HH
583dc7ff
remove print statement
4 年前
vincentpierre
87a79b1d
update
4 年前
Ervin Teng
e1378efc
Merge commit '6d729a0a2b2ba1fc946720cdb7871c9be3e38d45' into develop-fix-nan
4 年前
Ervin Teng
5ee6efb0
Make collab env harder
4 年前
HH
c4f623f5
change input to raw, add shield controller, gun controller
4 年前
HH
e9fe3a29
add hop-strafe, optional player input, misc improvements
4 年前
HH
c1af1229
update strafe code. getkeydown not working in agent script
4 年前
HH
3f05d962
add health controller
4 年前
vincentpierre
e14e1c4d
Improvements and new tests
4 年前
HH
4f789c09
remove timescale adjust
4 年前
Andrew Cohen
d62f6b0a
modify bullet/attn
4 年前
Andrew Cohen
d6528cac
added noise to spawns soccer
4 年前
HH
2387e68a
lots of stuff. locomotion feels good
4 年前
HH
8b49dc9d
updated input and agent script
4 年前
Ervin Teng
ce7d34a3
Revert "Invert divide by 3 in log prob"
This reverts commit a708af66e740f19df5082b4b4e152a566c703385.
4 年前
HH
63381e18
health bar working now. move input to FU
4 年前
Ervin Teng
3b15cc32
Multiprocessing but Stats are quite broken
4 年前
HH
48db7a48
save scene
4 年前
HH
e344129b
fix gun timing
4 年前
HH
669d4fc4
fixed strafe input issue
4 年前
HH
adfdeebf
about to update editor ver
4 年前
HH
3775117b
updated editor
4 年前
HH
fe846021
add new input system
4 年前
HH
f6b37bea
updated input, added dash
4 年前
Ervin Teng
9d20139b
Merge branch 'develop-stack-walljump' into develop-centralizedcritic
4 年前
brccabral
bbdbe0ec
set multiple obstacles from area
added a parameter to GridArea to control how many obstacles will be created in the scene
4 年前
HH
fe569079
updated particles & input
4 年前
Ervin Teng
7a0ebfbd
Pretty broken
4 年前
brccabral
349e98a6
set number of obstacles to 1 in scene
4 年前
HH
87b99a85
fixed input not triggering properly + misc jank
4 年前
HH
5ad04693
add deathcube
4 年前
HH
7927909e
added death explosion
4 年前
HH
c5a6783a
fix cam impulse
4 年前
Ervin Teng
ab01834c
Set group ids for some envs
4 年前
HH
40d7e3b2
add large blocks
4 年前
Andrew Cohen
ef8f70e8
Add WalljumpPushblock env
4 年前
Arthur Juliani
15052e1f
Add sensor type
4 年前
Arthur Juliani
6b32ff21
Allow gridworld agent to have different goals
4 年前
Ervin Teng
56fe741d
Remove erroneous file
4 年前
HH
34959fa9
add heuristic ai behavior
4 年前
vincentpierre
90da7426
Fix masking for torch
4 年前
Arthur Juliani
1737be14
Update protobufs
4 年前
Arthur Juliani
b3638ad6
Update prefabs
4 年前
HH
2cd5cc5d
fix dash input bug. working
4 年前
Arthur Juliani
6933b9c2
Add GoalSensor and component
4 年前
Arthur Juliani
72da0652
Fix white space
4 年前
HH
31bdedb6
added heuristic ai spawn
4 年前
GitHub
76faf383
Merge pull request #4733 from Unity-Technologies/gc-food-goals
Adds goal signal to GridWorld environment
4 年前
Arthur Juliani
a35ea7a9
Rename parameter -> goal
4 年前
HH
93bf2f07
added music
4 年前
GitHub
2c744129
Merge pull request #4714 from Unity-Technologies/develop-fix-mask
Fix masking for torch
4 年前
Arthur Juliani
b63ecb03
Fix whitespace
4 年前
HH
9e0a78f7
update UI and input actions
4 年前
GitHub
990f801a
Develop hybrid action staging ( #4702 )
Co-authored-by: Ervin T <ervin@unity3d.com>
Co-authored-by: Vincent-Pierre BERGES <vincentpierre@unity3d.com>
Co-authored-by: Ruo-Ping Dong <ruoping.dong@unity3d.com>
Co-authored-by: Chris Elion <chris.elion@unity3d.com>
4 年前
vincentpierre
89605d02
Replaced .nn models with .onnx models
Missing so far :
- Visual3DBall
- CrawlerDynamicVariableSpeed
- CrawlerStaticVariableSpeed
- GridFoodCollector
- VisualFoodCollector
- GridWorld
- Match3
4 年前
HH
bdc15070
update controls UI
4 年前
vincentpierre
c18f687d
Modified all the models (except visual food collector)
4 年前
vincentpierre
7a5cc9ec
Merge master into develop-rm-tf
4 年前
HH
a26b341f
update restart logic
4 年前
HH
d8384d5a
start work on rocket launcher
4 年前
vincentpierre
c1587bce
Solving merge conflicts
4 年前
HH
c24323e0
add gun sounds
4 年前
HH
75c2bb83
working
4 年前
Andrew Cohen
4ebc6c44
ml-agents-envs pass
4 年前
GitHub
ded1f79b
Merge pull request #4732 from Unity-Technologies/goal-sensors
Adds SensorTypes and GoalSensors
4 年前
HH
00448e8c
reduce restart time
4 年前
Andrew Cohen
b5d1c071