gRPCImitationProtocol

Full path: schola.core.protocols.protobuf.offlinegRPC.gRPCImitationProtocol

schola.core.protocols.protobuf.offlinegRPC.gRPCImitationProtocol

gRPCImitationProtocol

gRPCImitationProtocol(url, port=None, protocol_start_timeout=45)

Bases: BaseImitationProtocol, SocketProtocolMixin

Methods

Item	Description
init	—
close	Close the Unreal Connection.
get_data	Get demonstration data from the environment.
get_definition	Get the environment definition for imitation learning.
on_close()	Close the Unreal Connection.
on_start()	Bind the tcp_socket
send_startup_msg	Send the startup message for imitation learning data collection.
start	Open the Connection to Unreal Engine.

Attributes

Item	Description
address	Returns the address of the connection
channel_connected	Returns whether the connection is active or not
has_socket	Returns whether the connection is active or not
mixin_properties	Get mixin-specific properties.
properties	Get protocol-specific properties.

Parameters

url (str)

port (int)

protocol_start_timeout (int)

init

__init__(url, port=None, protocol_start_timeout=45)

Parameters

url (str)

port (int)

protocol_start_timeout (int)

---

channel_connected

channel_connected: bool

Returns whether the connection is active or not

---

close

close()

Close the Unreal Connection. Method must be safe to call multiple times.

Returns

None

---

get_data

get_data()

Get demonstration data from the environment.

Returns

A tuple containing:

• Observations for each timestep
• Rewards for each timestep
• Termination flags
• Truncation flags
• Info dicts
• Initial agent observations
• Initial agent info dicts
• Demonstration actions

Return type: Tuple[List[Dict[str,Any]], List[float], List[Dict[str,bool]], List[Dict[str,bool]], List[Dict[str,str]], Dict[int,Dict[str, Any]], Dict[int,Dict[str, str]], Dict[int, Dict[str,Any]]]

---

get_definition

get_definition()

Get the environment definition for imitation learning.

Returns

A tuple containing:

• List of agent IDs per environment
• Agent types indexed by environment and agent
• Observation spaces indexed by environment and agent
• Action spaces indexed by environment and agent

Return type: Tuple[List[List[str]], Dict[int, Dict[str, str]], Dict[int,Dict[str,gym.Space]], Dict[int,Dict[str,gym.Space]]]

---

properties

properties: Dict[str, Any]

Get protocol-specific properties.

---

send_startup_msg

send_startup_msg(seeds=None, options=None)

Send the startup message for imitation learning data collection.

Parameters

seeds (List, optional) : List of random seeds for each environment.

options (List, optional) : List of startup options for each environment.

---

start

start()

Open the Connection to Unreal Engine.

Returns

None

---