Unity 机器学习代理工具包 (ML-Agents) 是一个开源项目,它使游戏和模拟能够作为训练智能代理的环境。
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
 
 
 
 
 

532 行
23 KiB

import atexit
import glob
import logging
import numpy as np
import os
import subprocess
from typing import Dict, List, Optional, Any
from mlagents_envs.side_channel.side_channel import SideChannel
from mlagents_envs.base_env import (
BaseEnv,
BatchedStepResult,
AgentGroupSpec,
AgentGroup,
AgentId,
)
from mlagents_envs.timers import timed, hierarchical_timer
from mlagents_envs.exception import (
UnityEnvironmentException,
UnityCommunicationException,
UnityActionException,
UnityTimeOutException,
)
from mlagents_envs.communicator_objects.command_pb2 import STEP, RESET
from mlagents_envs.rpc_utils import (
agent_group_spec_from_proto,
batched_step_result_from_proto,
)
from mlagents_envs.communicator_objects.unity_rl_input_pb2 import UnityRLInputProto
from mlagents_envs.communicator_objects.unity_rl_output_pb2 import UnityRLOutputProto
from mlagents_envs.communicator_objects.agent_action_pb2 import AgentActionProto
from mlagents_envs.communicator_objects.unity_output_pb2 import UnityOutputProto
from mlagents_envs.communicator_objects.unity_rl_initialization_input_pb2 import (
UnityRLInitializationInputProto,
)
from mlagents_envs.communicator_objects.unity_input_pb2 import UnityInputProto
from .rpc_communicator import RpcCommunicator
from sys import platform
import signal
import struct
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger("mlagents_envs")
class UnityEnvironment(BaseEnv):
SCALAR_ACTION_TYPES = (int, np.int32, np.int64, float, np.float32, np.float64)
SINGLE_BRAIN_ACTION_TYPES = SCALAR_ACTION_TYPES + (list, np.ndarray)
API_VERSION = "API-12"
def __init__(
self,
file_name: Optional[str] = None,
worker_id: int = 0,
base_port: int = 5005,
seed: int = 0,
docker_training: bool = False,
no_graphics: bool = False,
timeout_wait: int = 60,
args: Optional[List[str]] = None,
side_channels: Optional[List[SideChannel]] = None,
):
"""
Starts a new unity environment and establishes a connection with the environment.
Notice: Currently communication between Unity and Python takes place over an open socket without authentication.
Ensure that the network where training takes place is secure.
:string file_name: Name of Unity environment binary.
:int base_port: Baseline port number to connect to Unity environment over. worker_id increments over this.
:int worker_id: Number to add to communication port (5005) [0]. Used for asynchronous agent scenarios.
:bool docker_training: Informs this class whether the process is being run within a container.
:bool no_graphics: Whether to run the Unity simulator in no-graphics mode
:int timeout_wait: Time (in seconds) to wait for connection from environment.
:bool train_mode: Whether to run in training mode, speeding up the simulation, by default.
:list args: Addition Unity command line arguments
:list side_channels: Additional side channel for no-rl communication with Unity
"""
args = args or []
atexit.register(self._close)
self.port = base_port + worker_id
self._buffer_size = 12000
self._version_ = UnityEnvironment.API_VERSION
# If true, this means the environment was successfully loaded
self._loaded = False
# The process that is started. If None, no process was started
self.proc1 = None
self.timeout_wait: int = timeout_wait
self.communicator = self.get_communicator(worker_id, base_port, timeout_wait)
self.worker_id = worker_id
self.side_channels: Dict[int, SideChannel] = {}
if side_channels is not None:
for _sc in side_channels:
if _sc.channel_type in self.side_channels:
raise UnityEnvironmentException(
"There cannot be two side channels with the same channel type {0}.".format(
_sc.channel_type
)
)
self.side_channels[_sc.channel_type] = _sc
# If the environment name is None, a new environment will not be launched
# and the communicator will directly try to connect to an existing unity environment.
# If the worker-id is not 0 and the environment name is None, an error is thrown
if file_name is None and worker_id != 0:
raise UnityEnvironmentException(
"If the environment name is None, "
"the worker-id must be 0 in order to connect with the Editor."
)
if file_name is not None:
self.executable_launcher(file_name, docker_training, no_graphics, args)
else:
logger.info(
f"Listening on port {self.port}. "
f"Start training by pressing the Play button in the Unity Editor."
)
self._loaded = True
rl_init_parameters_in = UnityRLInitializationInputProto(seed=seed)
try:
aca_output = self.send_academy_parameters(rl_init_parameters_in)
aca_params = aca_output.rl_initialization_output
except UnityTimeOutException:
self._close()
raise
# TODO : think of a better way to expose the academyParameters
self._unity_version = aca_params.version
if self._unity_version != self._version_:
self._close()
raise UnityEnvironmentException(
f"The API number is not compatible between Unity and python. "
f"Python API: {self._version_}, Unity API: {self._unity_version}.\n"
f"Please go to https://github.com/Unity-Technologies/ml-agents/releases/tag/latest_release"
f"to download the latest version of ML-Agents."
)
self._env_state: Dict[str, BatchedStepResult] = {}
self._env_specs: Dict[str, AgentGroupSpec] = {}
self._env_actions: Dict[str, np.ndarray] = {}
self._is_first_message = True
self._update_group_specs(aca_output)
@staticmethod
def get_communicator(worker_id, base_port, timeout_wait):
return RpcCommunicator(worker_id, base_port, timeout_wait)
def executable_launcher(self, file_name, docker_training, no_graphics, args):
cwd = os.getcwd()
file_name = (
file_name.strip()
.replace(".app", "")
.replace(".exe", "")
.replace(".x86_64", "")
.replace(".x86", "")
)
true_filename = os.path.basename(os.path.normpath(file_name))
logger.debug("The true file name is {}".format(true_filename))
launch_string = None
if platform == "linux" or platform == "linux2":
candidates = glob.glob(os.path.join(cwd, file_name) + ".x86_64")
if len(candidates) == 0:
candidates = glob.glob(os.path.join(cwd, file_name) + ".x86")
if len(candidates) == 0:
candidates = glob.glob(file_name + ".x86_64")
if len(candidates) == 0:
candidates = glob.glob(file_name + ".x86")
if len(candidates) > 0:
launch_string = candidates[0]
elif platform == "darwin":
candidates = glob.glob(
os.path.join(
cwd, file_name + ".app", "Contents", "MacOS", true_filename
)
)
if len(candidates) == 0:
candidates = glob.glob(
os.path.join(file_name + ".app", "Contents", "MacOS", true_filename)
)
if len(candidates) == 0:
candidates = glob.glob(
os.path.join(cwd, file_name + ".app", "Contents", "MacOS", "*")
)
if len(candidates) == 0:
candidates = glob.glob(
os.path.join(file_name + ".app", "Contents", "MacOS", "*")
)
if len(candidates) > 0:
launch_string = candidates[0]
elif platform == "win32":
candidates = glob.glob(os.path.join(cwd, file_name + ".exe"))
if len(candidates) == 0:
candidates = glob.glob(file_name + ".exe")
if len(candidates) > 0:
launch_string = candidates[0]
if launch_string is None:
self._close()
raise UnityEnvironmentException(
"Couldn't launch the {0} environment. "
"Provided filename does not match any environments.".format(
true_filename
)
)
else:
logger.debug("This is the launch string {}".format(launch_string))
# Launch Unity environment
if not docker_training:
subprocess_args = [launch_string]
if no_graphics:
subprocess_args += ["-nographics", "-batchmode"]
subprocess_args += ["--port", str(self.port)]
subprocess_args += args
try:
self.proc1 = subprocess.Popen(
subprocess_args,
# start_new_session=True means that signals to the parent python process
# (e.g. SIGINT from keyboard interrupt) will not be sent to the new process on POSIX platforms.
# This is generally good since we want the environment to have a chance to shutdown,
# but may be undesirable in come cases; if so, we'll add a command-line toggle.
# Note that on Windows, the CTRL_C signal will still be sent.
start_new_session=True,
)
except PermissionError as perm:
# This is likely due to missing read or execute permissions on file.
raise UnityEnvironmentException(
f"Error when trying to launch environment - make sure "
f"permissions are set correctly. For example "
f'"chmod -R 755 {launch_string}"'
) from perm
else:
# Comments for future maintenance:
# xvfb-run is a wrapper around Xvfb, a virtual xserver where all
# rendering is done to virtual memory. It automatically creates a
# new virtual server automatically picking a server number `auto-servernum`.
# The server is passed the arguments using `server-args`, we are telling
# Xvfb to create Screen number 0 with width 640, height 480 and depth 24 bits.
# Note that 640 X 480 are the default width and height. The main reason for
# us to add this is because we'd like to change the depth from the default
# of 8 bits to 24.
# Unfortunately, this means that we will need to pass the arguments through
# a shell which is why we set `shell=True`. Now, this adds its own
# complications. E.g SIGINT can bounce off the shell and not get propagated
# to the child processes. This is why we add `exec`, so that the shell gets
# launched, the arguments are passed to `xvfb-run`. `exec` replaces the shell
# we created with `xvfb`.
#
docker_ls = (
"exec xvfb-run --auto-servernum"
" --server-args='-screen 0 640x480x24'"
" {0} --port {1}"
).format(launch_string, str(self.port))
self.proc1 = subprocess.Popen(
docker_ls,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
shell=True,
)
def _update_group_specs(self, output: UnityOutputProto) -> None:
init_output = output.rl_initialization_output
for brain_param in init_output.brain_parameters:
# Each BrainParameter in the rl_initialization_output should have at least one AgentInfo
# Get that agent, because we need some of its observations.
agent_infos = output.rl_output.agentInfos[brain_param.brain_name]
if agent_infos.value:
agent = agent_infos.value[0]
new_spec = agent_group_spec_from_proto(brain_param, agent)
self._env_specs[brain_param.brain_name] = new_spec
logger.info(f"Connected new brain:\n{brain_param.brain_name}")
def _update_state(self, output: UnityRLOutputProto) -> None:
"""
Collects experience information from all external brains in environment at current step.
"""
for brain_name in self._env_specs.keys():
if brain_name in output.agentInfos:
agent_info_list = output.agentInfos[brain_name].value
self._env_state[brain_name] = batched_step_result_from_proto(
agent_info_list, self._env_specs[brain_name]
)
else:
self._env_state[brain_name] = BatchedStepResult.empty(
self._env_specs[brain_name]
)
self._parse_side_channel_message(self.side_channels, output.side_channel)
def reset(self) -> None:
if self._loaded:
outputs = self.communicator.exchange(self._generate_reset_input())
if outputs is None:
raise UnityCommunicationException("Communicator has stopped.")
self._update_group_specs(outputs)
rl_output = outputs.rl_output
self._update_state(rl_output)
self._is_first_message = False
self._env_actions.clear()
else:
raise UnityEnvironmentException("No Unity environment is loaded.")
@timed
def step(self) -> None:
if self._is_first_message:
return self.reset()
if not self._loaded:
raise UnityEnvironmentException("No Unity environment is loaded.")
# fill the blanks for missing actions
for group_name in self._env_specs:
if group_name not in self._env_actions:
n_agents = 0
if group_name in self._env_state:
n_agents = self._env_state[group_name].n_agents()
self._env_actions[group_name] = self._env_specs[
group_name
].create_empty_action(n_agents)
step_input = self._generate_step_input(self._env_actions)
with hierarchical_timer("communicator.exchange"):
outputs = self.communicator.exchange(step_input)
if outputs is None:
raise UnityCommunicationException("Communicator has stopped.")
self._update_group_specs(outputs)
rl_output = outputs.rl_output
self._update_state(rl_output)
self._env_actions.clear()
def get_agent_groups(self) -> List[AgentGroup]:
return list(self._env_specs.keys())
def _assert_group_exists(self, agent_group: str) -> None:
if agent_group not in self._env_specs:
raise UnityActionException(
"The group {0} does not correspond to an existing agent group "
"in the environment".format(agent_group)
)
def set_actions(self, agent_group: AgentGroup, action: np.ndarray) -> None:
self._assert_group_exists(agent_group)
if agent_group not in self._env_state:
return
spec = self._env_specs[agent_group]
expected_type = np.float32 if spec.is_action_continuous() else np.int32
expected_shape = (self._env_state[agent_group].n_agents(), spec.action_size)
if action.shape != expected_shape:
raise UnityActionException(
"The group {0} needs an input of dimension {1} but received input of dimension {2}".format(
agent_group, expected_shape, action.shape
)
)
if action.dtype != expected_type:
action = action.astype(expected_type)
self._env_actions[agent_group] = action
def set_action_for_agent(
self, agent_group: AgentGroup, agent_id: AgentId, action: np.ndarray
) -> None:
self._assert_group_exists(agent_group)
if agent_group not in self._env_state:
return
spec = self._env_specs[agent_group]
expected_shape = (spec.action_size,)
if action.shape != expected_shape:
raise UnityActionException(
"The Agent {0} in group {1} needs an input of dimension {2} but received input of dimension {3}".format(
agent_id, agent_group, expected_shape, action.shape
)
)
expected_type = np.float32 if spec.is_action_continuous() else np.int32
if action.dtype != expected_type:
action = action.astype(expected_type)
if agent_group not in self._env_actions:
self._env_actions[agent_group] = spec.create_empty_action(
self._env_state[agent_group].n_agents()
)
try:
index = np.where(self._env_state[agent_group].agent_id == agent_id)[0][0]
except IndexError as ie:
raise IndexError(
"agent_id {} is did not request a decision at the previous step".format(
agent_id
)
) from ie
self._env_actions[agent_group][index] = action
def get_step_result(self, agent_group: AgentGroup) -> BatchedStepResult:
self._assert_group_exists(agent_group)
return self._env_state[agent_group]
def get_agent_group_spec(self, agent_group: AgentGroup) -> AgentGroupSpec:
self._assert_group_exists(agent_group)
return self._env_specs[agent_group]
def close(self):
"""
Sends a shutdown signal to the unity environment, and closes the socket connection.
"""
if self._loaded:
self._close()
else:
raise UnityEnvironmentException("No Unity environment is loaded.")
def _close(self):
self._loaded = False
self.communicator.close()
if self.proc1 is not None:
# Wait a bit for the process to shutdown, but kill it if it takes too long
try:
self.proc1.wait(timeout=self.timeout_wait)
signal_name = self.returncode_to_signal_name(self.proc1.returncode)
signal_name = f" ({signal_name})" if signal_name else ""
return_info = f"Environment shut down with return code {self.proc1.returncode}{signal_name}."
logger.info(return_info)
except subprocess.TimeoutExpired:
logger.info("Environment timed out shutting down. Killing...")
self.proc1.kill()
# Set to None so we don't try to close multiple times.
self.proc1 = None
@classmethod
def _flatten(cls, arr: Any) -> List[float]:
"""
Converts arrays to list.
:param arr: numpy vector.
:return: flattened list.
"""
if isinstance(arr, cls.SCALAR_ACTION_TYPES):
arr = [float(arr)]
if isinstance(arr, np.ndarray):
arr = arr.tolist()
if len(arr) == 0:
return arr
if isinstance(arr[0], np.ndarray):
# pylint: disable=no-member
arr = [item for sublist in arr for item in sublist.tolist()]
if isinstance(arr[0], list):
# pylint: disable=not-an-iterable
arr = [item for sublist in arr for item in sublist]
arr = [float(x) for x in arr]
return arr
@staticmethod
def _parse_side_channel_message(
side_channels: Dict[int, SideChannel], data: bytes
) -> None:
offset = 0
while offset < len(data):
try:
channel_type, message_len = struct.unpack_from("<ii", data, offset)
offset = offset + 8
message_data = data[offset : offset + message_len]
offset = offset + message_len
except Exception:
raise UnityEnvironmentException(
"There was a problem reading a message in a SideChannel. "
"Please make sure the version of MLAgents in Unity is "
"compatible with the Python version."
)
if len(message_data) != message_len:
raise UnityEnvironmentException(
"The message received by the side channel {0} was "
"unexpectedly short. Make sure your Unity Environment "
"sending side channel data properly.".format(channel_type)
)
if channel_type in side_channels:
side_channels[channel_type].on_message_received(message_data)
else:
logger.warning(
"Unknown side channel data received. Channel type "
": {0}.".format(channel_type)
)
@staticmethod
def _generate_side_channel_data(side_channels: Dict[int, SideChannel]) -> bytearray:
result = bytearray()
for channel_type, channel in side_channels.items():
for message in channel.message_queue:
result += struct.pack("<ii", channel_type, len(message))
result += message
channel.message_queue = []
return result
@timed
def _generate_step_input(
self, vector_action: Dict[str, np.ndarray]
) -> UnityInputProto:
rl_in = UnityRLInputProto()
for b in vector_action:
n_agents = self._env_state[b].n_agents()
if n_agents == 0:
continue
for i in range(n_agents):
action = AgentActionProto(vector_actions=vector_action[b][i])
rl_in.agent_actions[b].value.extend([action])
rl_in.command = STEP
rl_in.side_channel = bytes(self._generate_side_channel_data(self.side_channels))
return self.wrap_unity_input(rl_in)
def _generate_reset_input(self) -> UnityInputProto:
rl_in = UnityRLInputProto()
rl_in.command = RESET
rl_in.side_channel = bytes(self._generate_side_channel_data(self.side_channels))
return self.wrap_unity_input(rl_in)
def send_academy_parameters(
self, init_parameters: UnityRLInitializationInputProto
) -> UnityOutputProto:
inputs = UnityInputProto()
inputs.rl_initialization_input.CopyFrom(init_parameters)
return self.communicator.initialize(inputs)
@staticmethod
def wrap_unity_input(rl_input: UnityRLInputProto) -> UnityInputProto:
result = UnityInputProto()
result.rl_input.CopyFrom(rl_input)
return result
@staticmethod
def returncode_to_signal_name(returncode: int) -> Optional[str]:
"""
Try to convert return codes into their corresponding signal name.
E.g. returncode_to_signal_name(-2) -> "SIGINT"
"""
try:
# A negative value -N indicates that the child was terminated by signal N (POSIX only).
s = signal.Signals(-returncode) # pylint: disable=no-member
return s.name
except Exception:
# Should generally be a ValueError, but catch everything just in case.
return None