results.json
4.5 KB · 197 lines · json Raw
1 {
2 "environment": {
3 "base_environment": "LunarLander-v3",
4 "custom_objective": "Complete a full directed rotation, recover upright, enter the landing zone and land safely.",
5 "reward_version": "v4-soft-touchdown",
6 "observation_space": {
7 "base_dimensions": 8,
8 "added_dimensions": [
9 "signed_rotation_progress",
10 "flip_completed",
11 "recovery_completed"
12 ],
13 "wrapped_dimensions": 11
14 },
15 "reward_config": {
16 "required_rotations": 1.0,
17 "rotation_direction": 1,
18 "upright_tolerance_radians": 0.3,
19 "recovery_angular_velocity_tolerance": 0.5,
20 "pre_flip_original_reward_weight": 0.15,
21 "post_flip_original_reward_weight": 3.0,
22 "rotation_progress_bonus": 300.0,
23 "flip_completion_bonus": 500.0,
24 "recovery_bonus": 250.0,
25 "flip_landing_bonus": 2000.0,
26 "landing_without_flip_penalty": 300.0,
27 "no_flip_terminal_penalty": 300.0,
28 "failed_landing_penalty": 800.0,
29 "outside_zone_landing_penalty": 500.0,
30 "in_zone_crash_penalty": 1200.0,
31 "post_flip_shaping_weight": 2.0,
32 "post_flip_shaping_gamma": 0.999,
33 "post_flip_shaping_clip": 30.0,
34 "post_flip_center_weight": 80.0,
35 "post_flip_horizontal_speed_weight": 50.0,
36 "post_flip_vertical_speed_weight": 100.0,
37 "post_flip_angle_weight": 70.0,
38 "post_flip_angular_speed_weight": 40.0,
39 "post_flip_leg_contact_weight": 30.0,
40 "landing_zone_half_width": 0.2,
41 "post_flip_zone_excess_weight": 120.0,
42 "post_flip_target_vx_gain": 0.5,
43 "post_flip_max_target_vx": 0.35,
44 "post_flip_horizontal_deadband": 0.08,
45 "post_flip_target_vy_high": -0.45,
46 "post_flip_target_vy_near_ground": -0.12,
47 "near_ground_height": 0.6,
48 "safe_touchdown_vertical_speed": 0.18,
49 "near_ground_overspeed_weight": 120.0
50 },
51 "change_notes": [
52 "Continued from the selected Phase C checkpoint.",
53 "Added an altitude-dependent vertical-speed target.",
54 "Added a near-ground quadratic descent-overspeed penalty.",
55 "Added a dedicated in-zone crash penalty.",
56 "Reduced risky late horizontal corrections using a deadband."
57 ]
58 },
59 "evaluation": {
60 "n_episodes": 100,
61 "mean_shaped_reward": -325.2546577557927,
62 "std_shaped_reward": 435.5396409711088,
63 "shaped_course_score": -760.7942987269015,
64 "min_shaped_reward": -3342.837273847694,
65 "max_shaped_reward": -88.07963784059791,
66 "mean_original_reward": -706.4599452970039,
67 "std_original_reward": 73.7189566670546,
68 "original_course_score": -780.1789019640585,
69 "min_original_reward": -818.5765463852929,
70 "max_original_reward": -463.0414724621158,
71 "original_reward_200_rate": 0.0,
72 "flip_completion_rate": 0.36,
73 "recovery_completion_rate": 0.12,
74 "recovery_given_flip_rate": 0.3333333333333333,
75 "safe_landing_rate": 0.0,
76 "flip_landing_rate": 0.0,
77 "landing_given_flip_rate": 0.0,
78 "stable_landing_rate": 0.0,
79 "inside_landing_zone_rate": 0.2,
80 "outside_zone_landing_rate": 0.0,
81 "in_zone_crash_rate": 0.08
82 },
83 "evaluation_seeds": [
84 20000,
85 20001,
86 20002,
87 20003,
88 20004,
89 20005,
90 20006,
91 20007,
92 20008,
93 20009,
94 20010,
95 20011,
96 20012,
97 20013,
98 20014,
99 20015,
100 20016,
101 20017,
102 20018,
103 20019,
104 20020,
105 20021,
106 20022,
107 20023,
108 20024,
109 20025,
110 20026,
111 20027,
112 20028,
113 20029,
114 20030,
115 20031,
116 20032,
117 20033,
118 20034,
119 20035,
120 20036,
121 20037,
122 20038,
123 20039,
124 20040,
125 20041,
126 20042,
127 20043,
128 20044,
129 20045,
130 20046,
131 20047,
132 20048,
133 20049,
134 20050,
135 20051,
136 20052,
137 20053,
138 20054,
139 20055,
140 20056,
141 20057,
142 20058,
143 20059,
144 20060,
145 20061,
146 20062,
147 20063,
148 20064,
149 20065,
150 20066,
151 20067,
152 20068,
153 20069,
154 20070,
155 20071,
156 20072,
157 20073,
158 20074,
159 20075,
160 20076,
161 20077,
162 20078,
163 20079,
164 20080,
165 20081,
166 20082,
167 20083,
168 20084,
169 20085,
170 20086,
171 20087,
172 20088,
173 20089,
174 20090,
175 20091,
176 20092,
177 20093,
178 20094,
179 20095,
180 20096,
181 20097,
182 20098,
183 20099
184 ],
185 "replay": {
186 "seed": 20016,
187 "shaped_reward": -88.07963784059791,
188 "original_reward": -679.112379167114,
189 "steps": 93,
190 "flip_completed": true,
191 "landed_safely": false,
192 "rotations_completed": 1.0265225981958122,
193 "recovery_completed": true,
194 "custom_outcome": "flip_but_failed_landing",
195 "flip_and_landed": false
196 }
197 }