您最多选择25个主题
主题必须以中文或者字母或数字开头，可以包含连字符 (-)，并且长度不得超过35个字符
615 行
22 KiB
615 行
22 KiB
using UnityEngine;
|
|
using System.Collections.Generic;
|
|
using UnityEngine.Serialization;
|
|
#if UNITY_EDITOR
|
|
using UnityEditor;
|
|
#endif
|
|
using MLAgents.InferenceBrain;
|
|
using Barracuda;
|
|
|
|
/**
|
|
* Welcome to Unity Machine Learning Agents (ML-Agents).
|
|
*
|
|
* The ML-Agents toolkit contains four entities: Academy, Agent, Communicator and
|
|
* Python API. The academy and connected agents live within
|
|
* a learning environment (herein called Environment), while the communicator
|
|
* manages the communication between the learning environment and the Python
|
|
* API. For more information on each of these entities, in addition to how to
|
|
* set-up a learning environment and train the behavior of characters in a
|
|
* Unity scene, please browse our documentation pages on GitHub:
|
|
* https://github.com/Unity-Technologies/ml-agents/blob/master/docs/
|
|
*/
|
|
|
|
namespace MLAgents
|
|
{
|
|
/// <summary>
|
|
/// Wraps the environment-level parameters that are provided within the
|
|
/// Editor. These parameters can be provided for training and inference
|
|
/// modes separately and represent screen resolution, rendering quality and
|
|
/// frame rate.
|
|
/// </summary>
|
|
[System.Serializable]
|
|
public class EnvironmentConfiguration
|
|
{
|
|
[Tooltip("Width of the environment window in pixels.")]
|
|
public int width;
|
|
|
|
[Tooltip("Height of the environment window in pixels.")]
|
|
public int height;
|
|
|
|
[Tooltip("Rendering quality of environment. (Higher is better quality.)")]
|
|
[Range(0, 5)]
|
|
public int qualityLevel;
|
|
|
|
[Tooltip("Speed at which environment is run. (Higher is faster.)")]
|
|
[Range(1f, 100f)]
|
|
public float timeScale;
|
|
|
|
[Tooltip("Frames per second (FPS) engine attempts to maintain.")]
|
|
public int targetFrameRate;
|
|
|
|
/// Initializes a new instance of the
|
|
/// <see cref="EnvironmentConfiguration"/> class.
|
|
/// <param name="width">Width of environment window (pixels).</param>
|
|
/// <param name="height">Height of environment window (pixels).</param>
|
|
/// <param name="qualityLevel">
|
|
/// Rendering quality of environment. Ranges from 0 to 5, with higher.
|
|
/// </param>
|
|
/// <param name="timeScale">
|
|
/// Speed at which environment is run. Ranges from 1 to 100, with higher
|
|
/// values representing faster speed.
|
|
/// </param>
|
|
/// <param name="targetFrameRate">
|
|
/// Target frame rate (per second) that the engine tries to maintain.
|
|
/// </param>
|
|
public EnvironmentConfiguration(
|
|
int width, int height, int qualityLevel,
|
|
float timeScale, int targetFrameRate)
|
|
{
|
|
this.width = width;
|
|
this.height = height;
|
|
this.qualityLevel = qualityLevel;
|
|
this.timeScale = timeScale;
|
|
this.targetFrameRate = targetFrameRate;
|
|
}
|
|
}
|
|
|
|
/// <summary>
|
|
/// An Academy is where Agent objects go to train their behaviors.
|
|
/// Currently, this class is expected to be extended to
|
|
/// implement the desired academy behavior.
|
|
/// </summary>
|
|
/// <remarks>
|
|
/// When an academy is run, it can either be in inference or training mode.
|
|
/// The mode is determined by the presence or absence of a Communicator. In
|
|
/// the presence of a communicator, the academy is run in training mode where
|
|
/// the states and observations of each agent are sent through the
|
|
/// communicator. In the absence of a communicator, the academy is run in
|
|
/// inference mode where the agent behavior is determined by the Policy
|
|
/// attached to it.
|
|
/// </remarks>
|
|
[HelpURL("https://github.com/Unity-Technologies/ml-agents/blob/master/" +
|
|
"docs/Learning-Environment-Design-Academy.md")]
|
|
public abstract class Academy : MonoBehaviour
|
|
{
|
|
private const string k_ApiVersion = "API-10";
|
|
|
|
/// Temporary storage for global gravity value
|
|
/// Used to restore oringal value when deriving Academy modifies it
|
|
private Vector3 m_OriginalGravity;
|
|
|
|
/// Temporary storage for global fixedDeltaTime value
|
|
/// Used to restore original value when deriving Academy modifies it
|
|
private float m_OriginalFixedDeltaTime;
|
|
|
|
/// Temporary storage for global maximumDeltaTime value
|
|
/// Used to restore original value when deriving Academy modifies it
|
|
private float m_OriginalMaximumDeltaTime;
|
|
|
|
// Fields provided in the Inspector
|
|
|
|
[FormerlySerializedAs("maxSteps")]
|
|
[SerializeField]
|
|
[Tooltip("The engine-level settings which correspond to rendering " +
|
|
"quality and engine speed during Training.")]
|
|
EnvironmentConfiguration m_TrainingConfiguration =
|
|
new EnvironmentConfiguration(80, 80, 1, 100.0f, -1);
|
|
|
|
[FormerlySerializedAs("inferenceConfiguration")]
|
|
[SerializeField]
|
|
[Tooltip("The engine-level settings which correspond to rendering " +
|
|
"quality and engine speed during Inference.")]
|
|
EnvironmentConfiguration m_InferenceConfiguration =
|
|
new EnvironmentConfiguration(1280, 720, 5, 1.0f, 60);
|
|
|
|
/// <summary>
|
|
/// Contains a mapping from parameter names to float values. They are
|
|
/// used in <see cref="AcademyReset"/> and <see cref="AcademyStep"/>
|
|
/// to modify elements in the environment at reset time.
|
|
/// </summary>
|
|
/// <remarks>
|
|
/// Default reset parameters are specified in the academy Editor, and can
|
|
/// be modified when training by passing a config
|
|
/// dictionary at reset.
|
|
/// </remarks>
|
|
[SerializeField]
|
|
[Tooltip("List of custom parameters that can be changed in the " +
|
|
"environment when it resets.")]
|
|
public ResetParameters resetParameters;
|
|
public CommunicatorObjects.CustomResetParametersProto customResetParameters;
|
|
|
|
// Fields not provided in the Inspector.
|
|
|
|
/// <summary>
|
|
/// Returns whether or not the communicator is on.
|
|
/// </summary>
|
|
/// <returns>
|
|
/// <c>true</c>, if communicator is on, <c>false</c> otherwise.
|
|
/// </returns>
|
|
public bool IsCommunicatorOn
|
|
{
|
|
get { return Communicator != null; }
|
|
}
|
|
|
|
/// If true, the Academy will use inference settings. This field is
|
|
/// initialized in <see cref="Awake"/> depending on the presence
|
|
/// or absence of a communicator. Furthermore, it can be modified during
|
|
/// training via <see cref="SetIsInference"/>.
|
|
bool m_IsInference = true;
|
|
|
|
/// The number of episodes completed by the environment. Incremented
|
|
/// each time the environment is reset.
|
|
int m_EpisodeCount;
|
|
|
|
/// The number of steps completed within the current episode. Incremented
|
|
/// each time a step is taken in the environment. Is reset to 0 during
|
|
/// <see cref="AcademyReset"/>.
|
|
int m_StepCount;
|
|
|
|
/// The number of total number of steps completed during the whole simulation. Incremented
|
|
/// each time a step is taken in the environment.
|
|
int m_TotalStepCount;
|
|
|
|
/// Flag that indicates whether the inference/training mode of the
|
|
/// environment was switched by the training process. This impacts the
|
|
/// engine settings at the next environment step.
|
|
bool m_ModeSwitched;
|
|
|
|
/// Pointer to the communicator currently in use by the Academy.
|
|
public ICommunicator Communicator;
|
|
|
|
private bool m_Initialized;
|
|
private List<ModelRunner> m_ModelRunners = new List<ModelRunner>();
|
|
|
|
// Flag used to keep track of the first time the Academy is reset.
|
|
bool m_FirstAcademyReset;
|
|
|
|
// The Academy uses a series of events to communicate with agents
|
|
// to facilitate synchronization. More specifically, it ensure
|
|
// that all the agents performs their steps in a consistent order (i.e. no
|
|
// agent can act based on a decision before another agent has had a chance
|
|
// to request a decision).
|
|
|
|
// Signals to all the Agents at each environment step so they can use
|
|
// their Policy to decide on their next action.
|
|
public event System.Action DecideAction;
|
|
|
|
// Signals to all the listeners that the academy is being destroyed
|
|
public event System.Action DestroyAction;
|
|
|
|
// Signals to all the agents at each environment step along with the
|
|
// Academy's maxStepReached, done and stepCount values. The agents rely
|
|
// on this event to update their own values of max step reached and done
|
|
// in addition to aligning on the step count of the global episode.
|
|
public event System.Action<int> AgentSetStatus;
|
|
|
|
// Signals to all the agents at each environment step so they can reset
|
|
// if their flag has been set to done (assuming the agent has requested a
|
|
// decision).
|
|
public event System.Action AgentResetIfDone;
|
|
|
|
// Signals to all the agents at each environment step so they can send
|
|
// their state to their Policy if they have requested a decision.
|
|
public event System.Action AgentSendState;
|
|
|
|
// Signals to all the agents at each environment step so they can act if
|
|
// they have requested a decision.
|
|
public event System.Action AgentAct;
|
|
|
|
// Signals to all the agents each time the Academy force resets.
|
|
public event System.Action AgentForceReset;
|
|
|
|
/// <summary>
|
|
/// MonoBehavior function called at the very beginning of environment
|
|
/// creation. Academy uses this time to initialize internal data
|
|
/// structures, initialize the environment and check for the existence
|
|
/// of a communicator.
|
|
/// </summary>
|
|
void Awake()
|
|
{
|
|
LazyInitialization();
|
|
}
|
|
|
|
public void LazyInitialization()
|
|
{
|
|
if (!m_Initialized)
|
|
{
|
|
InitializeEnvironment();
|
|
m_Initialized = true;
|
|
}
|
|
}
|
|
|
|
// Used to read Python-provided environment parameters
|
|
private static int ReadArgs()
|
|
{
|
|
var args = System.Environment.GetCommandLineArgs();
|
|
var inputPort = "";
|
|
for (var i = 0; i < args.Length; i++)
|
|
{
|
|
if (args[i] == "--port")
|
|
{
|
|
inputPort = args[i + 1];
|
|
}
|
|
}
|
|
|
|
return int.Parse(inputPort);
|
|
}
|
|
|
|
/// <summary>
|
|
/// Initializes the environment, configures it and initialized the Academy.
|
|
/// </summary>
|
|
private void InitializeEnvironment()
|
|
{
|
|
m_OriginalGravity = Physics.gravity;
|
|
m_OriginalFixedDeltaTime = Time.fixedDeltaTime;
|
|
m_OriginalMaximumDeltaTime = Time.maximumDeltaTime;
|
|
|
|
InitializeAcademy();
|
|
|
|
// Try to launch the communicator by using the arguments passed at launch
|
|
try
|
|
{
|
|
Communicator = new RpcCommunicator(
|
|
new CommunicatorInitParameters
|
|
{
|
|
port = ReadArgs()
|
|
});
|
|
}
|
|
catch
|
|
{
|
|
#if UNITY_EDITOR
|
|
Communicator = new RpcCommunicator(
|
|
new CommunicatorInitParameters
|
|
{
|
|
port = 5004
|
|
});
|
|
#endif
|
|
}
|
|
|
|
if (Communicator != null)
|
|
{
|
|
// We try to exchange the first message with Python. If this fails, it means
|
|
// no Python Process is ready to train the environment. In this case, the
|
|
//environment must use Inference.
|
|
try
|
|
{
|
|
var unityRLInitParameters = Communicator.Initialize(
|
|
new CommunicatorInitParameters
|
|
{
|
|
version = k_ApiVersion,
|
|
name = gameObject.name,
|
|
environmentResetParameters = new EnvironmentResetParameters
|
|
{
|
|
resetParameters = resetParameters,
|
|
customResetParameters = customResetParameters
|
|
}
|
|
});
|
|
Random.InitState(unityRLInitParameters.seed);
|
|
}
|
|
catch
|
|
{
|
|
Communicator = null;
|
|
}
|
|
|
|
if (Communicator != null)
|
|
{
|
|
Communicator.QuitCommandReceived += OnQuitCommandReceived;
|
|
Communicator.ResetCommandReceived += OnResetCommand;
|
|
Communicator.RLInputReceived += OnRLInputReceived;
|
|
}
|
|
}
|
|
|
|
// If a communicator is enabled/provided, then we assume we are in
|
|
// training mode. In the absence of a communicator, we assume we are
|
|
// in inference mode.
|
|
|
|
SetIsInference(!IsCommunicatorOn);
|
|
|
|
DecideAction += () => { };
|
|
DestroyAction += () => { };
|
|
AgentSetStatus += i => { };
|
|
AgentResetIfDone += () => { };
|
|
AgentSendState += () => { };
|
|
AgentAct += () => { };
|
|
AgentForceReset += () => { };
|
|
|
|
ConfigureEnvironment();
|
|
}
|
|
|
|
static void OnQuitCommandReceived()
|
|
{
|
|
#if UNITY_EDITOR
|
|
EditorApplication.isPlaying = false;
|
|
#endif
|
|
Application.Quit();
|
|
}
|
|
|
|
private void OnResetCommand(EnvironmentResetParameters newResetParameters)
|
|
{
|
|
UpdateResetParameters(newResetParameters);
|
|
ForcedFullReset();
|
|
}
|
|
|
|
void OnRLInputReceived(UnityRLInputParameters inputParams)
|
|
{
|
|
m_IsInference = !inputParams.isTraining;
|
|
}
|
|
|
|
private void UpdateResetParameters(EnvironmentResetParameters newResetParameters)
|
|
{
|
|
if (newResetParameters.resetParameters != null)
|
|
{
|
|
foreach (var kv in newResetParameters.resetParameters)
|
|
{
|
|
resetParameters[kv.Key] = kv.Value;
|
|
}
|
|
}
|
|
customResetParameters = newResetParameters.customResetParameters;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Configures the environment settings depending on the training/inference
|
|
/// mode and the corresponding parameters passed in the Editor.
|
|
/// </summary>
|
|
void ConfigureEnvironment()
|
|
{
|
|
if (m_IsInference)
|
|
{
|
|
ConfigureEnvironmentHelper(m_InferenceConfiguration);
|
|
Monitor.SetActive(true);
|
|
}
|
|
else
|
|
{
|
|
ConfigureEnvironmentHelper(m_TrainingConfiguration);
|
|
Monitor.SetActive(false);
|
|
}
|
|
}
|
|
|
|
/// <summary>
|
|
/// Helper method for initializing the environment based on the provided
|
|
/// configuration.
|
|
/// </summary>
|
|
/// <param name="config">
|
|
/// Environment configuration (specified in the Editor).
|
|
/// </param>
|
|
static void ConfigureEnvironmentHelper(EnvironmentConfiguration config)
|
|
{
|
|
Screen.SetResolution(config.width, config.height, false);
|
|
QualitySettings.SetQualityLevel(config.qualityLevel, true);
|
|
Time.timeScale = config.timeScale;
|
|
Time.captureFramerate = 60;
|
|
Application.targetFrameRate = config.targetFrameRate;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Initializes the academy and environment. Called during the waking-up
|
|
/// phase of the environment before any of the scene objects/agents have
|
|
/// been initialized.
|
|
/// </summary>
|
|
public virtual void InitializeAcademy()
|
|
{
|
|
}
|
|
|
|
/// <summary>
|
|
/// Specifies the academy behavior at every step of the environment.
|
|
/// </summary>
|
|
public virtual void AcademyStep()
|
|
{
|
|
}
|
|
|
|
/// <summary>
|
|
/// Specifies the academy behavior when being reset (i.e. at the completion
|
|
/// of a global episode).
|
|
/// </summary>
|
|
public virtual void AcademyReset()
|
|
{
|
|
}
|
|
|
|
/// <summary>
|
|
/// Returns the <see cref="m_IsInference"/> flag.
|
|
/// </summary>
|
|
/// <returns>
|
|
/// <c>true</c>, if current mode is inference, <c>false</c> if training.
|
|
/// </returns>
|
|
public bool GetIsInference()
|
|
{
|
|
return m_IsInference;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Sets the <see cref="m_IsInference"/> flag to the provided value. If
|
|
/// the new flag differs from the current flag value, this signals that
|
|
/// the environment configuration needs to be updated.
|
|
/// </summary>
|
|
/// <param name="isInference">
|
|
/// Environment mode, if true then inference, otherwise training.
|
|
/// </param>
|
|
public void SetIsInference(bool isInference)
|
|
{
|
|
if (m_IsInference != isInference)
|
|
{
|
|
m_IsInference = isInference;
|
|
|
|
// This signals to the academy that at the next environment step
|
|
// the engine configurations need updating to the respective mode
|
|
// (i.e. training vs inference) configuration.
|
|
m_ModeSwitched = true;
|
|
}
|
|
}
|
|
|
|
/// <summary>
|
|
/// Returns the current episode counter.
|
|
/// </summary>
|
|
/// <returns>
|
|
/// Current episode number.
|
|
/// </returns>
|
|
public int GetEpisodeCount()
|
|
{
|
|
return m_EpisodeCount;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Returns the current step counter (within the current episode).
|
|
/// </summary>
|
|
/// <returns>
|
|
/// Current step count.
|
|
/// </returns>
|
|
public int GetStepCount()
|
|
{
|
|
return m_StepCount;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Returns the total step counter.
|
|
/// </summary>
|
|
/// <returns>
|
|
/// Total step count.
|
|
/// </returns>
|
|
public int GetTotalStepCount()
|
|
{
|
|
return m_TotalStepCount;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Forces the full reset. The done flags are not affected. Is either
|
|
/// called the first reset at inference and every external reset
|
|
/// at training.
|
|
/// </summary>
|
|
void ForcedFullReset()
|
|
{
|
|
EnvironmentReset();
|
|
AgentForceReset?.Invoke();
|
|
m_FirstAcademyReset = true;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Performs a single environment update to the Academy, and Agent
|
|
/// objects within the environment.
|
|
/// </summary>
|
|
void EnvironmentStep()
|
|
{
|
|
if (m_ModeSwitched)
|
|
{
|
|
ConfigureEnvironment();
|
|
m_ModeSwitched = false;
|
|
}
|
|
if (!m_FirstAcademyReset)
|
|
{
|
|
ForcedFullReset();
|
|
}
|
|
|
|
AgentSetStatus?.Invoke(m_StepCount);
|
|
|
|
using (TimerStack.Instance.Scoped("AgentResetIfDone"))
|
|
{
|
|
AgentResetIfDone?.Invoke();
|
|
}
|
|
|
|
using (TimerStack.Instance.Scoped("AgentSendState"))
|
|
{
|
|
AgentSendState?.Invoke();
|
|
}
|
|
|
|
using (TimerStack.Instance.Scoped("DecideAction"))
|
|
{
|
|
DecideAction?.Invoke();
|
|
}
|
|
|
|
using (TimerStack.Instance.Scoped("AcademyStep"))
|
|
{
|
|
AcademyStep();
|
|
}
|
|
|
|
using (TimerStack.Instance.Scoped("AgentAct"))
|
|
{
|
|
AgentAct?.Invoke();
|
|
}
|
|
|
|
m_StepCount += 1;
|
|
m_TotalStepCount += 1;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Resets the environment, including the Academy.
|
|
/// </summary>
|
|
void EnvironmentReset()
|
|
{
|
|
m_StepCount = 0;
|
|
m_EpisodeCount++;
|
|
AcademyReset();
|
|
}
|
|
|
|
/// <summary>
|
|
/// MonoBehaviour function that dictates each environment step.
|
|
/// </summary>
|
|
void FixedUpdate()
|
|
{
|
|
EnvironmentStep();
|
|
}
|
|
|
|
/// <summary>
|
|
/// Creates or retrieves an existing ModelRunner that uses the same
|
|
/// NNModel and the InferenceDevice as provided.
|
|
/// </summary>
|
|
/// <param name="model"> The NNModel the ModelRunner must use </param>
|
|
/// <param name="brainParameters"> The brainParameters used to create
|
|
/// the ModelRunner </param>
|
|
/// <param name="inferenceDevice"> The inference device (CPU or GPU)
|
|
/// the ModelRunner will use </param>
|
|
/// <returns> The ModelRunner compatible with the input settings</returns>
|
|
public ModelRunner GetOrCreateModelRunner(
|
|
NNModel model, BrainParameters brainParameters, InferenceDevice inferenceDevice)
|
|
{
|
|
var modelRunner = m_ModelRunners.Find(x => x.HasModel(model, inferenceDevice));
|
|
if (modelRunner == null)
|
|
{
|
|
modelRunner = new ModelRunner(
|
|
model, brainParameters, inferenceDevice);
|
|
m_ModelRunners.Add(modelRunner);
|
|
}
|
|
return modelRunner;
|
|
}
|
|
|
|
/// <summary>
|
|
/// Cleanup function
|
|
/// </summary>
|
|
protected virtual void OnDestroy()
|
|
{
|
|
Physics.gravity = m_OriginalGravity;
|
|
Time.fixedDeltaTime = m_OriginalFixedDeltaTime;
|
|
Time.maximumDeltaTime = m_OriginalMaximumDeltaTime;
|
|
|
|
// Signal to listeners that the academy is being destroyed now
|
|
DestroyAction?.Invoke();
|
|
|
|
foreach (var mr in m_ModelRunners)
|
|
{
|
|
mr.Dispose();
|
|
}
|
|
|
|
// TODO - Pass worker ID or some other identifier,
|
|
// so that multiple envs won't overwrite each others stats.
|
|
TimerStack.Instance.SaveJsonTimers();
|
|
}
|
|
}
|
|
}
|