* Changing learn.py log messages.
- learn.py refers to the mlagents-learn script now.
- If a non-existant trainer config is passed, the log message
correctly points that out now.
* Changing the curriculum arg from file to dir.
* Fixing learn.py, trainer_controller.py, and Docker
- learn.py has been moved under trainers.
- this was a two line change
- learn.py will no longer be run as a main method
- docopt arguments are strings by default. learn.py now uses
this assumption to correctly parse arguments.
- trainer_controller.py now considers the Docker volume when
accepting a trainer config file path.
- the Docker container now uses mlagents-learn.
* Removing extraneous unity-volume ref.
* Initial Commit
Ported most functionalities, still need to :
- Documentation
- Add Comments
- Custom drawer for BrainParameters
- Fix the UnitTests
- Review Functionalities
* Added Custom Drawer for the Brain Parameters
* Improvements to the HubDrawer
* Modified the Brain Editors
* Minor bug fixes and UI changes
* Modified the Help Boxes of the Drawers
* Modified Brain class, renamed Initialize and made DecideAction virtual
* Fix the UnityTests
* Simpler Brain creation menu
* Renamed Internal Brain to Learning Brain
* modified the parameters to remove reference to External or Internal in the Protobuf objects
* Updated the protobuf generated files
* Fix the Pytests
* Removed the graph scope from the Learning Brain
* cleaner logic than try catch
* Removed the isExternal field of the brain and put the isTraining logic into LearningBrain and Training Hub
* Modified how the Brain finds the A...
* Initial Commit
* attempt at refactor
* Put all static methods into the CoreInternalBrain
* improvements
* more testing
* modifications
* renamed epsilon
* misc
* Now supports discrete actions
* added discrete support and RNN and visual. Left to do is refactor and save variables into models
* code cleaning
* made a tensor generator and applier
* fix on the models.py file
* Moved the Checks to a different Class
* Added some unit tests
* BugFix
* Need to generate the output tensors as well as inputs before executing the graph
* Made NodeNames static and created a new namespace
* Added comments to the TensorAppliers
* Started adding comments on the TensorGenerators code
* Added comments for the Tensor Generator
* Moving the helper classes into a separate folder
* Added initial comments to the TensorChecks
* Renamed NodeNames -> TensorNames
* Removing warnings in tests
* Now using Aut...
* Check that worker port is available in RpcCommunicator
Previously the RpcCommunicator did not check the port or create the
RPC server until `initialize()` was called. Since "initialize"
requires the environment to be available, this means we might create
a new environment which connects to an existing RPC server running
in another process. This causes both training runs to fail.
As a remedy to this issue, this commit moves the server creation into
the RpcCommunicator constructor and adds an explicit socket binding
check to the requested port.
* Fixes suggested by Codacy
* Update rpc_communicator.py
* Addressing feedback: formatting & consistency
We check for the single brain case in UnityEnvironment by checking
for applicable non-dict types in the step arguments. However for ints
and floats we just use `np.int_` and `np.float_` for the check, which
are the defaults for your system.
This means if you are using an application (like baselines in #1448)
which uses the wrong int/float size an error will be thrown. This
change explicitly allows both 32 and 64-bit numbers.
* Enable buffer padding to be set other than 0
Allows buffer padding in AgentBufferField to be set to a custom value. In particular, 0-padding for `action_masks` causes a divide-by-zero error, and should be padded with 1’s instead.
This is done as a parameter passed to the `append` method, so that the pad value can be set right after the instantiation of an AgentBufferField.
* Documentation tweaks and updates (#1479)
* Add blurb about using the --load flag in the intro guide, and typo fix.
* Add section in tutorial to create multiple area learning environment.
* Add mention of Done() method in agent design
* fixed the windows ctrl-c bug
* fixed typo
* removed some uncessary printing
* nothing
* make the import of the win api conditional
* removved the duplicate code
* added the ability to use python debugger on ml-agents
* added newline at the end, changed the import to be complete path
* changed the info.log into policy.export_model, changed the sys.platform to use startswith
* fixed a bug
* remove the printing of the path
* tweaked the info message to notify the user about the expected error message
* removed some logging according to comments
* removed the sys import
* Revert "Documentation tweaks and updates (#1479)"
This reverts commit 84ef07a4525fa8a89f4...
* Remove env creation logic from TrainerController
Currently TrainerController includes logic related to creating the
UnityEnvironment, which causes poor separation of concerns between
the learn.py application script, TrainerController and UnityEnvironment:
* TrainerController must know about the proper way to instantiate the
UnityEnvironment, which may differ from application to application.
This also makes mocking or subclassing UnityEnvironment more
difficult.
* Many arguments are passed by learn.py to TrainerController and passed
along to UnityEnvironment.
This change moves environment construction logic into learn.py, as part
of the greater refactor to separate trainer logic from actor / environment.
* Switched default Mac GFX API to Metal
* Added Barracuda pre-0.1.5
* Added basic integration with Barracuda Inference Engine
* Use predefined outputs the same way as for TF engine
* Fixed discrete action + LSTM support
* Switch Unity Mac Editor to Metal GFX API
* Fixed null model handling
* All examples converted to support Barracuda
* Added model conversion from Tensorflow to Barracuda
copied the barracuda.py file to ml-agents/mlagents/trainers
copied the tensorflow_to_barracuda.py file to ml-agents/mlagents/trainers
modified the tensorflow_to_barracuda.py file so it could be called from mlagents
modified ml-agents/mlagents/trainers/policy.py to convert the tf models to barracuda compatible .bytes file
* Added missing iOS BLAS plugin
* Added forgotten prefab changes
* Removed GLCore GFX backend for Mac, because it doesn't support Compute shaders
* Exposed GPU support for LearningBrain inference
...
* Ticked API :
- Ticked API for pypi for mlagents
- Ticked API for pypi for unity-gym
- Ticked Communication number for API
- Ticked Model Loader number for API
* Ticked the API for the pytest
Removing this function breaks some tests, and the only way around
this at this time is a bigger refactor or hacky fixes to tests.
For now, I'd suggest we just revert this small part of a change
and keep a refactor in mind for the future.
* Move 'take_action' into Policy class
This refactor is part of Actor-Trainer separation. Since policies
will be distributed across actors in separate processes which share
a single trainer, taking an action should be the responsibility of
the policy.
This change makes a few smaller changes:
* Combines `take_action` logic between trainers, making it more
generic
* Adds an `ActionInfo` data class to be more explicit about the
data returned by the policy, only used by TrainerController and
policy for now.
* Moves trainer stats logic out of `take_action` and into
`add_experiences`
* Renames 'take_action' to 'get_action'
This commit adds support for running Unity environments in parallel.
An abstract base class was created for UnityEnvironment which a new
SubprocessUnityEnvironment inherits from.
SubprocessUnityEnvironment communicates through a pipe in order to
send commands which will be run in parallel to its workers.
A few significant changes needed to be made as a side-effect:
* UnityEnvironments are created via a factory method (a closure)
rather than being directly created by the main process.
* In mlagents-learn "worker-id" has been replaced by "base-port"
and "num-envs", and worker_ids are automatically assigned across runs.
* BrainInfo objects now convert all fields to numpy arrays or lists to
avoid serialization issues.
On Windows the interrupt for subprocesses works in a different
way from OSX/Linux. The result is that child subprocesses and
their pipes may close while the parent process is still running
during a keyboard (ctrl+C) interrupt.
To handle this, this change adds handling for EOFError and
BrokenPipeError exceptions when interacting with subprocess
environments. Additional management is also added to be sure
when using parallel runs using the "num-runs" option that
the threads for each run are joined and KeyboardInterrupts are
handled.
These changes made the "_win_handler" we used to specially
manage interrupts on Windows unnecessary, so they have been
removed.
A change was made to the way the "train_mode" flag was used by
environments when SubprocessUnityEnvironment was added which was
intended to be part of a separate change set. This broke the CLI
'--slow' flag. This change undoes those changes, so that the slow
/ fast simulation option works correctly.
As a minor additional change, the remaining tests from top level
'tests' folders have been moved into the new test folders.
When using parallel SubprocessUnityEnvironment instances along
with Academy Done(), a new step might be taken when reset should
have been called because some environments may have been done while
others were not (making "global done" less useful).
This change manages the reset on `global_done` at the level of the
environment worker, and removes the global reset from
TrainerController.
* WIP precommit on top level
* update CI
* circleci fixes
* intentionally fail black
* use --show-diff-on-failure in CI
* fix command order
* rebreak a file
* apply black
* WIP enable mypy
* run mypy on each package
* fix trainer_metrics mypy errors
* more mypy errors
* more mypy
* Fix some partially typed functions
* types for take_action_outputs
* fix formatting
* cleanup
* generate stubs for proto objects
* fix ml-agents-env mypy errors
* disallow-incomplete-defs for gym-unity
* Add CI notes to CONTRIBUTING.md
* Create new class (RewardSignal) that represents a reward signal.
* Add value heads for each reward signal in the PPO model.
* Make summaries agnostic to the type of reward signals, and log weighted rewards per reward signal.
* Move extrinsic and curiosity rewards into this new structure.
* Allow defining multiple reward signals in YAML file. Add documentation for this new structure.
At each step, an unused `last_reward` variable in the TF graph is
updated in our PPO trainer. There are also related unused methods
in various places in the codebase. This change removes them.
Previously in v0.8 we added parallel environments via the
SubprocessUnityEnvironment, which exposed the same abstraction as
UnityEnvironment while actually wrapping many parallel environments
via subprocesses.
Wrapping many environments with the same interface as a single
environment had some downsides, however:
* Ordering needed to be preserved for agents across different envs,
complicating the SubprocessEnvironment logic
* Asynchronous environments with steps taken out of sync with the
trainer aren't viable with the Environment abstraction
This PR introduces a new EnvManager abstraction which exposes a
reduced subset of the UnityEnvironment abstraction and a
SubprocessEnvManager implementation which replaces the
SubprocessUnityEnvironment.
TrainerController depended on an external_brains dictionary with
brain params in its constructor but only used it in a single function
call. The same function call (start_learning) takes the environment
as an argument, which is the source of the external_brains.
This change removes the dependency of TrainerController on external
brains and removes the two class members related to external_brains
and retrieves the brains directly from the environment.
* Timer proof-of-concept
* micro optimizations
* add some timers
* cleanup, add asserts
* Cleanup (no start/end methods) and handle exceptions
* unit test and decorator
* move output code, add a decorator
* cleanup
* module docstring
* actually write the timings when done with training
* use __qualname__ instead
* add a few more timers
* fix mock import
* fix unit test
* don't need fwd reference
* cleanup root
* always write timers, add comments
* undo accidental change
Based on the new reward signals architecture, add BC pretrainer and GAIL for PPO. Main changes:
- A new GAILRewardSignal and GAILModel for GAIL/VAIL
- A BCModule component (not a reward signal) to do pretraining during RL
- Documentation for both of these
- Change to Demo Loader that lets you load multiple demo files in a folder
- Example Demo files for all of our tested sample environments (for future regression testing)
Fixes shuffling issue with newer versions of numpy (#1798).
* make get_value_estimates output a dict of floats
* Use np.append instead of convert to list, unconvert
* Add type hints and test for get_value_estimates
* Don't 0 value bootstrap for GAIL and Curiosity
* Add gradient penalties to GAN to help with stability
* Add gail_config.yaml with GAIL examples
* Cleaned up trainer_config.yaml and unnecessary gammas
* Documentation updates
* Code cleanup
* Removes unused SubprocessEnvManager import in trainer_controller
* Removes unused `steps` argument to `TrainerController._save_model`
* Consolidates unnecessary branching for curricula in
`TrainerController.advance`
* Moves `reward_buffer` into `TFPolicy` from `PPOPolicy` and adds
`BCTrainer` support so that we don't have a broken interface /
undefined behavior when BCTrainer is used with curricula.
* Add Sampler and SamplerManager
* Enable resampling of reset parameters during training
* Documentation for Sampler and example YAML configuration file
* Removed obsolete 'TestDstWrongShape' test as it does not reflect how Barracuda tensors work
* Added proper test cleanup, to avoid warning messages from finalizer thread.
* Hotfix for recurrent + continous action nets in ML Agents
* Included explicit version # for ZN
* added explicit version for KR docs
* minor fix in installation doc
* Consistency with numbers for reset parameters
* Removed extra verbiage. minor consistency
* minor consistency
* Cleaned up IL language
* moved parameter sampling above in list
* Cleaned up language in Env Parameter sampling
* Cleaned up migrating content
* updated consistency of Reset Parameter Sampling
* Rename Training-Generalization-Learning.md to Training-Generalization-Reinforcement-Learning-Agents.md
* Updated doc link for generalization
* Rename Training-Generalization-Reinforcement-Learning-Agents.md to Training-Generalized-Reinforcement-Learning-Agents.md
* Re-wrote the intro paragraph for generalization
* add titles, cleaned up language for reset params
* Update Training-Generalized-Reinforcement-Learning-Agents.md
* cleanup of generalization doc
* More cleanu...
This fixes an issue where stopping the game when training in the Editor won't end training, due to the new asynchronous SubprocessEnvManager changes. Another minor change was made to move the `env_manager.close()` in TrainerController to the end of `start_learning` so that we are more likely to save the model if something goes wrong during the environment shutdown (this occurs sometimes on Windows machines).
This change moves trainer initialization outside of TrainerController,
reducing some of the constructor arguments of TrainerController and
setting up the ability for trainers to be initialized in the case where
a TrainerController isn't needed.
- Move common functions to trainer.py, model.pyfromppo/trainer.py, ppo/policy.pyandppo/model.py'
- Introduce RLTrainer class and move most of add_experiences and some common reward
signal code there. PPO and SAC will inherit from this, not so much BC Trainer.
- Add methods to Buffer to enable sampling, truncating, and save/loading.
- Add scoping to create encoders in model.py
- Fix issue with BC Trainer `increment_steps`.
- Fix issue with Demonstration Recorder and visual observations (memory leak fix was deleting vis obs too early).
- Make Samplers sample from the same random seed every time, so generalization runs are repeatable.
- Fix crash when using GAIL, Curiosity, and visual observations together.
* Adds evaluate_batch to reward signals. Evaluates on minibatch rather than on BrainInfo.
* Changes the way reward signal results are reported in rl_trainer so that we get the pure, unprocessed environment reward separate from the reward signals.
* Moves end_episode to rl_trainer
* Fixed bug with BCModule with RNN
* Add Soft Actor-Critic model, trainer, and policy and sac_trainer_config.yaml
* Add documentation for SAC and tweak PPO documentation to reference the new pages.
* Add tests for SAC, change simple_rl test to run both PPO and SAC.
* Initial Commit
* Remove the Academy Done flag from the protobuf definitions
* remove global_done in the environment
* Removed irrelevant unitTests
* Remove the max_step from the Academy inspector
* Removed global_done from the python scripts
* Modified and removed some tests
* This actually does not break either curriculum nor generalization training
* Replace global_done with reserved.
Addressing Chris Elion's comment regarding the deprecation of the global_done field. We will use a reserved field to make sure the global done does not get replaced in the future causing errors.
* Removed unused fake brain
* Tested that the first call to step was the same as a reset call
* black formating
* Added documentation changes
* Editing the migrating doc
* Addressing comments on the Migrating doc
* Addressing comments :
- Removing dead code
- Resolving forgotten merged conflicts
- Editing documentations...
* Normalize observations when adding experiences
This change moves normalization of vector observations into the trainer's
"add_experiences" interface.
Prior to this change, normalization occurred at inference time. This
was somewhat confusing since usually executing a forward pass shouldn't
have side-effects which would change the training step. Also, in a
asynchronous or distributed setting where we copy the neural network
weights from a trainer to a remote actor / inference worker we'd end up
with training issues because of the weights being different on the trainer
than the workers.
We have been ignoring unused imports and star imports via flake8. These are
both bad practice and grow over time without automated checking. This
commit attempts to fix all existing import errors and add back the corresponding
flake8 checks.
Our multi-GPU training had a regression such that freezing the
graph was broken. This change fixes that issue by making a few
changes:
* Removes the top level "tower" variable scope added by multi-GPU
so that the output nodes have correct names
* Removes the use of "freeze_graph" and replaces it with our own similar
functionality.
* Adds the "auto reuse" to network layers which require them
* Feature Deprecation : Online Behavioral Cloning
In this PR :
- Delete the online_bc_trainer
- Delete the tests for online bc
- delete the configuration file for online bc training
* Deleting the BCTeacherHelper.cs Script
TODO :
- Remove usages in the scene
- Documentation Edits
*DO NOT MERGE*
* IMPORTANT : REMOVED ALL IL SCENES
- Removed all the IL scenes from the Examples folder
* Removed all mentions of online BC training in the Documentation
* Made a note in the Migrating.md doc about the removal of the Online BC feature.
* Feature Deprecation : Online Behavioral Cloning
In this PR :
- Delete the online_bc_trainer
- Delete the tests for online bc
- delete the configuration file for online bc training
* Deleting the BCTeacherHelper.cs Script
TODO :
- Remove usages in the scene
- Documentation Edits
*DO NOT MERGE*
* IMPORTANT : REMOVED ALL IL SCENES
- Removed all the IL scenes from the Examples folder
* Removed all mentions of online BC training in the Documentation
* Made a note in the Migrating.md doc about the removal of the Online BC feature.
* Modified the Academy UI to remove the control checkbox and replaced it with a train in the editor checkbox
* Removed the Broadcast functionality from the non-Learning brains
* Bug fix
* Note that the scenes are broken since the BroadcastHub has changed
* Modified the LL-API for Python to remove the broadcasting functiuonality.
* All unit tests are running
* Modifie...
* Feature Deprecation : Online Behavioral Cloning
In this PR :
- Delete the online_bc_trainer
- Delete the tests for online bc
- delete the configuration file for online bc training
* Deleting the BCTeacherHelper.cs Script
TODO :
- Remove usages in the scene
- Documentation Edits
*DO NOT MERGE*
* IMPORTANT : REMOVED ALL IL SCENES
- Removed all the IL scenes from the Examples folder
* Removed all mentions of online BC training in the Documentation
* Made a note in the Migrating.md doc about the removal of the Online BC feature.
* Modified the Academy UI to remove the control checkbox and replaced it with a train in the editor checkbox
* Removed the Broadcast functionality from the non-Learning brains
* Bug fix
* Note that the scenes are broken since the BroadcastHub has changed
* Modified the LL-API for Python to remove the broadcasting functiuonality.
* All unit tests are running
* Modified the scen...
* Add test for curiosity + SAC
* Use actions for all curiosity (need to test on PPO)
* Fix issue with reward signals updating multiple times
* Put curiosity actions in the right placeholder
* Test PPO curiosity update
* ISensor and SensorBase
* camera and rendertex first pass
* use isensors for visual obs
* Update gridworld with CameraSensors
* compressed obs for reals
* Remove AgentInfo.visualObservations
* better separation of train and inference sensor calls
* compressed obs proto - need CI to generate code
* int32
* get proto name right
* run protoc locally for new fiels
* apply generated proto patch (pyi files were weird)
* don't repeat bytes
* hook up compressedobs
* dont send BrainParameters until there's an AgentInfo
* python BrainParameters now needs an AgentInfo to create
* remove last (I hope) dependency on camerares
* remove CameraResolutions and AgentInfo.visual_observations
* update mypy-protobuf version
* cleanup todos
* python cleanup
* more unit test fixes
* more unit test fix
* camera sensors for VisualFood collector, record demo
* SensorCompon...
* Initial commit removing memories from C# and deprecating memory fields in proto
* initial changes to Python
* Adding functionalities
* Fixes
* adding the memories to the dictionary
* Fixing bugs
* tweeks
* Resolving bugs
* Recreating the proto
* Addressing comments
* Passing by reference does not work. Do not merge
* Fixing huge bug in Inference
* Applying patches
* fixing tests
* Addressing comments
* Renaming variable to reflect type
* test
* Modifying the .proto files
* attempt 1 at refactoring Python
* works for ppo hallway
* changing the documentation
* now works with both sac and ppo both training and inference
* Ned to fix the tests
* TODOs :
- Fix the demonstration recorder
- Fix the demonstration loader
- verify the intrinsic reward signals work
- Fix the tests on Python
- Fix the C# tests
* Regenerating the protos
* fix proto typo
* protos and modifying the C# demo recorder
* modified the demo loader
* Demos are loading
* IMPORTANT : THESE ARE THE FILES USED FOR CONVERSION FROM OLD TO NEW FORMAT
* Modified all the demo files
* Fixing all the tests
* fixing ci
* addressing comments
* removing reference to memories in the ll-api
* allow --version argument in mlagents-learn
* Develop version print add strings (#2945)
* add __version__ to libs
* more version info
* use actual version
This is the first in a series of PRs that intend to move the agent processing logic (add_experiences and process_experiences) out of the trainer and into a separate class. The plan is to do so in steps:
- Split the processing buffers (keeping track of agent trajectories and assembling trajectories) and update buffer (complete trajectories to be used for training) within the Trainer (this PR)
- Move the processing buffer and add/process experiences into a separate, outside class
- Change the data type of the update buffer to be a Trajectory
- Place and read Trajectories from queues, add subscription mechanism for both AgentProcessor and Trainers
* [WIP] Side Channel initial layout
* Working prototype for raw bytes
* fixing format mistake
* Added some errors and some unit tests in C#
* Added the side channel for the Engine Configuration. (#2958)
* Added the side channel for the Engine Configuration.
Note that this change does not require modifying a lot of files :
- Adding a sender in Python
- Adding a receiver in C#
- subscribe the receiver to the communicator (here is a one liner in the Academy)
- Add the side channel to the Python UnityEnvironment (not represented here)
Adding the side channel to the environment would look like such :
```python
from mlagents.envs.environment import UnityEnvironment
from mlagents.envs.side_channel.raw_bytes_channel import RawBytesChannel
from mlagents.envs.side_channel.engine_configuration_channel import EngineConfigurationChannel
channel0 = RawBytesChannel()
channel1 = EngineConfigurationChanne...
* initial commit for LL-API
* fixing ml-agents-envs tests
* Implementing action masks
* training is fixed for 3DBall
* Tests all fixed, gym is broken and missing documentation changes
* adding case where no vector obs
* Fixed Gym
* fixing tests of float64
* fixing float64
* reverting some of brain.py
* removing old proto apis
* comment type fixes
* added properties to AgentGroupSpec and edited the notebooks.
* clearing the notebook outputs
* Update gym-unity/gym_unity/tests/test_gym.py
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update gym-unity/gym_unity/tests/test_gym.py
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update ml-agents-envs/mlagents/envs/base_env.py
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* Update ml-agents-envs/mlagents/envs/base_env.py
Co-Authored-By: Chris Elion <chris.elion@unity3d.com>
* addressing first comments
* NaN checks for r...
Our tests were using pytest fixtures by actually calling the fixture
methods, but in newer 5.x versions of pytest this causes test failures.
The recommended method for using fixtures is dependency injection.
This change updates the relevant test fixtures to either not use
`pytest.fixture` or to use dependency injection to pass the fixture.
The version range requirements in `test_requirements.txt` were also
updated accordingly.
* added team id and identifier concat to behavior parameters
* splitting brain params into brain name and identifiers
* set team id in prefab
* recieves brain_name and identifier on python side
* added team id and identifier concat to behavior parameters
* splitting brain params into brain name and identifiers
* set team id in prefab
* recieves brain_name and identifier on python side
* rebased with develop
* Correctly calls concatBehaviorIdentifiers
* added team id and identifier concat to behavior parameters
* splitting brain params into brain name and identifiers
* set team id in prefab
* recieves brain_name and identifier on python side
* rebased with develop
* Correctly calls concatBehaviorIdentifiers
* trainer_controller expects name_behavior_ids
* add_policy and create_policy separated
* adjusting tests to expect trainer.add_policy to be called
* fixing tests
* fixed naming ...
Previously the Curriculum and MetaCurriculum classes required file / folder
paths for initialization. These methods loaded the configuration for the
curricula from the filesystem. Requiring files for configuring curricula
makes testing and updating our config format more difficult.
This change moves the file loading into static methods, so that Curricula /
MetaCurricula can be initialized from dictionaries only.
* pass shape to WriteAdapter
* handle floats on python side
* cleanup
* whitespace
* rename GetFloatObservationShape, support uncompressed in RenderTexture sensor
* numpy float32
* remove unused using
* Float sensor and unit test
* replace asserts with exceptions, docstrings
The "num-runs" command-line option provides the ability to run multiple
identically-configured training runs in separate processes by running
mlagents-learn only once. This is a rarely used ML-Agents feature,
but it adds complexity to other parts of the system by adding the need
to support multiprocessing and managing of ports for the parallel training
runs. It also doesn't provide truly reproducible experiments, since there
is no guarantee of resource isolation between the trials.
This commit removes the --num-runs option, with the idea that users will
manage parallel or sequential runs of the same experiment themselves in the
future.
This change adds a new 'mlagents-run-experiment' endpoint which
accepts a single YAML/JSON file providing all of the information that
mlagents-learn accepts via command-line arguments and file inputs.
As part of this change the curriculum configuration is simplified to
accept only a single file for all the curricula in an environment
rather than a file for each behavior.
This PR makes it so that the env_manager only sends one current BrainInfo and the previous actions (if any) to the AgentManager. The list of agents was added to the ActionInfo and used appropriately.
This PR moves the AgentManagers from the TrainerController into the env_manager. This way, the TrainerController only needs to create the components (Trainers, AgentManagers) and call advance() on the EnvManager and the Trainers.
* Made the Agent reset immediately
* fixing the C# tests
* Fixing the tests still
* Trying with incremental episode ids
* deleting buffer rather than using an empty list
* Addressing the comments
* Forgot to edit the comment on AgentInfo
* Updating the migrating doc
* Fixed an obvious bug
* cleaning after an agent is done in agent processor
* Fixing the pytest errors
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).
In the previous PR, steps were processed when the env manager was reset. This was an issue for the very first reset, where we don't actually know which agent groups (and AgentManagers) we needed to send the steps to. These steps were being thrown away.
This PR moves the processing of steps to advance(), so that the initial reset steps are simply processed when the next advance(). This also removes the need for an additional block of code in TrainerController to handle the initial reset.
Tensorflow doesn't prescribe any particular file suffix for checkpoint files, but they
are commonly referred to as "ckpt" as a shorthand for "checkpoint". However ours
is somewhat confusingly "cptk". This change simply changes our checkpoint suffix
to "ckpt".
* Bumping versions on master
* Bumping package version
* Made the package version 0.15.0-preview
* Reverting the API version that was bumped by mistake
* [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 ...
* 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>
The "docker target" feature and associated command-line flag
--docker-target-name were created for use with the now-deprecated
Docker setup. This feature redirects the paths used by learn.py
for the environment and config files to be based from a directory
other than the current working directory. Additionally it wrapped
the environment execution with xvfb-run.
This commit removes the "docker target" feature because:
* Renaming the paths doesn't fix any problem. Absolute paths can
already be passed for configs and environment executables.
* Use of xserver, Xvfb, or xvfb-run are independent of mlagents-learn
and can be used outside of the mlagents-learn call. Further, xvfb-run
is not the only solution for software rendering.
This commit surfaces exceptions from environment worker subprocesses,
and changes the SubprocessEnvManager to raise those exceptions when
caught. Additionally TrainerController was changed to treat environment
exceptions differently than KeyboardInterrupts. We now raise the
environment exceptions after exporting the model, so that ML-Agents will
correctly exit with a non-zero return code.
* [skip ci] WIP : Modify the base_env.py file
* [skip ci] typo
* [skip ci] renamed some methods
* [skip ci] Incorporated changes from our meeting
* [skip ci] everything is broken
* [skip ci] everything is broken
* [skip ci] formatting
* Fixing the gym tests
* Fixing bug, C# has an error that needs fixing
* Fixing the test
* relaxing the threshold of 0.99 to 0.9
* fixing the C# side
* formating
* Fixed the llapi integratio test
* [Increasing steps for testing]
* Fixing the python tests
* Need __contains__ after all
* changing the max_steps in the tests
* addressing comments
* Making env_manager logic clearer as proposed in the comments
* Remove duplicated logic and added back in episode length (#3728)
* removing mentions of multi-agent in gym and changed the docstring in base_env.py
* Edited the Documentation for the changes to the LLAPI (#3733)
* Edite...