File size: 3,813 Bytes
2952f21 |
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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
using System;
using Godot;
using Dictionary = Godot.Collections.Dictionary;
using Array = Godot.Collections.Array;
[GlobalClass]
public abstract partial class AIControllerSharp3D : Node3D {
public enum ControlModes { INHERIT_FROM_SYNC, HUMAN, TRAINING, ONNX_INFERENCE, RECORD_EXPERT_DEMOS }
[Export] public ControlModes control_mode = ControlModes.INHERIT_FROM_SYNC;
[Export] public string onnx_model_path = "";
[Export] public int reset_after = 1000;
[ExportGroup("Record expert demos mode options")]
// Path where the demos will be saved. The file can later be used for imitation learning.
[Export] public String expert_demo_save_path;
// The action that erases the last recorded episode from the currently recorded data.
[Export] public InputEvent remove_last_episode_key;
// Action will be repeated for n frames. Will introduce control lag if larger than 1.
// Can be used to ensure that action_repeat on inference && training matches
// the recorded demonstrations.
[Export] public int action_repeat = 1;
[ExportGroup("Debug")]
[Export] protected bool DebugOn;
// ONNXModel GDScript object
public Resource onnx_model;
public string heuristic = "human";
public bool done;
public float reward;
public int n_steps;
public bool needs_reset;
public Node2D _player;
public override void _Ready() {
base._Ready();
AddToGroup("AGENT");
}
public virtual void init(Node2D player) {
if (DebugOn) GD.Print("Initializing AIController...");
_player = player;
}
public virtual float get_reward() {
return reward;
}
//-- Methods that need implementing using the "extend script" option in Godot --#
public abstract Dictionary get_obs();
public abstract Dictionary get_action_space();
public abstract void set_action(Dictionary action);
//-----------------------------------------------------------------------------#
//-- Methods that sometimes need implementing using the "extend script" option in Godot --#
// Only needed if you are recording expert demos with this AIController
public virtual Array get_action() {
throw new NotImplementedException();
}
// -----------------------------------------------------------------------------#
public override void _PhysicsProcess(double delta) {
base._PhysicsProcess(delta);
n_steps += 1;
if(n_steps > reset_after) {
if (DebugOn) GD.Print("Timeout reached. Setting 'needs_reset' to true.");
needs_reset = true;
}
}
public virtual Dictionary get_obs_space() {
// may need overriding if the obs space is complex
Dictionary obs = get_obs();
Array size = new Array
{
((Array)obs["obs"]).Count
};
return new Dictionary()
{
{
"obs", new Dictionary()
{
{ "size", size },
{ "space", "box" }
}
}
};
}
public void reset() {
if (DebugOn) GD.Print("Resetting AIController...");
n_steps = 0;
needs_reset = false;
}
public void reset_if_done() {
if (DebugOn) GD.Print("Resetting if done...");
if(done) {
reset();
}
}
public void set_heuristic(String h) {
if (DebugOn) GD.Print("Setting heuristic...");
// sets the heuristic from "human" || "model" nothing to change here
heuristic = h;
}
public bool get_done() {
return done;
}
public void set_done_false() {
done = false;
}
public void zero_reward() {
reward = 0.0f;
}
}
|