relayrl_framework 0.5.0-alpha

A distributed, system-oriented multi-agent reinforcement learning framework.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42

{
    "inference_server_config": {
        "config_update_polling_seconds": 10.0,
        "transport_config": {
            "addresses": {
                "_comment": "gRPC uses only this address (prefix is unused).",
                "model_server": {
                    "prefix": "tcp://",
                    "host": "127.0.0.1",
                    "port": "50051"
                },
                "trajectory_server": {
                    "prefix": "tcp://",
                    "host": "127.0.0.1",
                    "port": "7776"
                },
                "agent_listener": {
                    "prefix": "tcp://",
                    "host": "127.0.0.1",
                    "port": "7777"
                },
                "scaling_server": {
                    "prefix": "tcp://",
                    "host": "127.0.0.1",
                    "port": "7778"
                },
                "_comment2": "Only used when client-side inference is disabled or when client-side inference is used as fallback.",
                "inference_server": {
                    "prefix": "tcp://",
                    "host": "127.0.0.1",
                    "port": "7779"
                }
            },
            "local_model_module": {
                "directory": "model_module",
                "model_name": "inference_model",
                "format": "pt"
            },
            "max_traj_length": 100000000
        }
    }
}