b6ce30bf
Add curriculum support to PPO
7 年前
9d76583e
Fixed typo ( #105 )
7 年前
c16e0ac3
modified the socket to receive states and images of any size
7 年前
9a7445f7
Update best-practices-ppo.md
7 年前
447da485
fix on the CoreBrains so that if one Corebrain gets eraised, it will be reinstanciated
7 年前
332061aa
Rename
7 年前
d421a300
updated the tests of unityagents
7 年前
0d7f6726
Add ppo to table of contents
7 年前
250eb8e1
better checking of the format of the curriculum file
7 年前
3bd12269
Add best practices for PPO
7 年前
e8429059
bug fix for python3
7 年前
d14d88e2
Clarify text
7 年前
360984c4
curriculum.json params must have 4 entries
7 年前
3e0ae561
Clarify wording
7 年前
ac910514
initial commit of the curriculum with broadcast. Improved the Unity python handshake
7 年前
71086699
Update Making-a-new-Unity-Environment.md
7 年前
5cae720d
modified Environment to send a specific error when no external brains are in the environment
7 年前
cca45d1f
BrainEditor: Replaced `new SerializedObject()` with property. ( #88 )
* Fixed `Academy` children `Brain` gathering.
* Brain-Editor: Replaced `new SerializedObject()` with `serializedObject`.
* BrainEditor: Simplified `if` condition for `BrainParameters` checks.
7 年前
94025cbf
Update Making-a-new-Unity-Environment.md
7 年前
4d3d63e5
Add Mean reward error
7 年前
2d5d735f
new Monitor
7 年前
4b7e0d4b
Make clear meaning of <env_name>
7 年前
9933b56e
modification to get started on the recurrent NN rewrite
7 年前
6054fcbb
Merge pull request #55 from mplantady/master
Fix brain type not saved in inspector
7 年前
b237e6c2
fix the basic decision to return 1 instead of null
7 年前
a16f7b0d
Typo on discreet action in the Agent monitor
7 年前
9ff529f7
Fixed `Academy` children `Brain` gathering. ( #63 )
7 年前
5390bb09
need to convert the state to int if the state is discrete
7 年前
0a0fbbeb
Fix brain type not saved in inspector
7 年前
81d896f1
Fixed typo in `BrainEditor`. ( #64 )
Replaced `myst` with `must`.
7 年前
431fc43c
Merge branch 'master' of https://github.com/Unity-Technologies/ml-agents into dev-broadcast
7 年前
c190eb22
Randomize ppo training batch
7 年前
3a154c62
Update Unity-Agents---Python-API.md
7 年前
adac2683
Fix for multi-agent with observations
7 年前
0c9b296e
Minor changes ( #43 )
I walked through the whole example and made some minor revision.
7 年前
cd3bfb87
Added worker-id flag and pass through to enviroment in order to more easily manage multiple running simulations. ( #40 )
7 年前
20b8e021
Update Getting-Started-with-Balance-Ball.md
7 年前
12753429
Update Getting-Started-with-Balance-Ball.md
7 年前
77f93ed8
Link to installation documentation
7 年前
2acfc2ac
Add 64-bit Python requirement
7 年前
bbcc2af4
Explain use of epsilon in model embedding
7 年前
7d229e5c
Update Getting-Started-with-Balance-Ball.md
7 年前
3f85bb56
Merge branch 'master' into dev-broadcast
7 年前
46853a5a
Update installation.md
7 年前
aee5d336
Fix discrete state ( #33 )
* made BrainParameters a class to set default values
Modified the error message if the state is discrete
* Add discrete state support to PPO and provide discrete state example environment
* Add flexibility to continuous control as well
* Finish PPO flexible model generation implementation
* Fix formatting
* Support color observations
* Add best practices document
* bug fix for non square observations
* Update Readme.md
* Remove scipy dependency
* Add installation doc
7 年前
6cab43f9
Update Training-on-Amazon-Web-Service.md
7 年前
8071e029
Update Training-on-Amazon-Web-Service.md
7 年前
0df8326e
minor fixes
7 年前
4d3716fe
default discrete action is now 0
bug fix for discrete broadcast action (the action size should be one in Agents.cs)
modified Tennis so that the default action is no action
modified the TemplateDecsion.cs to ensure non null values are sent from Decide() and MakeMemory()
7 年前
65df8ae9
fix on the instanciation of coreBrains,
fix on the conversion of actions to arrays in the BrainInfo received from step
7 年前
Arthur Juliani
vincentpierre
GitHub
mplantady
Maxime Plantady