results.json
4.7 KB · 203 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": "v5-soft-touchdown-refinement",
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": 0.3,
20 "pre_flip_original_reward_weight": 0.15,
21 "post_flip_original_reward_weight": 3.0,
22 "pre_flip_center_weight": 25.0,
23 "pre_flip_horizontal_speed_weight": 15.0,
24 "pre_flip_shaping_weight": 1.0,
25 "pre_flip_shaping_clip": 10.0,
26 "flip_corridor_half_width": 0.35,
27 "flip_entry_penalty_weight": 600.0,
28 "rotation_progress_bonus": 300.0,
29 "flip_completion_bonus": 500.0,
30 "recovery_bonus": 250.0,
31 "flip_landing_bonus": 2000.0,
32 "landing_without_flip_penalty": 600.0,
33 "no_flip_terminal_penalty": 500.0,
34 "failed_landing_penalty": 900.0,
35 "outside_zone_landing_penalty": 1200.0,
36 "in_zone_crash_penalty": 1500.0,
37 "post_flip_shaping_weight": 2.0,
38 "post_flip_shaping_gamma": 0.999,
39 "post_flip_shaping_clip": 30.0,
40 "post_flip_center_weight": 100.0,
41 "post_flip_horizontal_speed_weight": 50.0,
42 "post_flip_vertical_speed_weight": 100.0,
43 "post_flip_angle_weight": 70.0,
44 "post_flip_angular_speed_weight": 40.0,
45 "post_flip_leg_contact_weight": 30.0,
46 "landing_zone_half_width": 0.2,
47 "post_flip_zone_excess_weight": 180.0,
48 "post_flip_target_vx_gain": 0.6,
49 "post_flip_max_target_vx": 0.4,
50 "post_flip_horizontal_deadband": 0.08,
51 "post_flip_target_vy_high": -0.45,
52 "post_flip_target_vy_near_ground": -0.12,
53 "near_ground_height": 0.6,
54 "safe_touchdown_vertical_speed": 0.14,
55 "near_ground_overspeed_weight": 120.0
56 },
57 "change_notes": [
58 "Continued from the selected Phase D checkpoint.",
59 "Added an altitude-dependent vertical-speed target.",
60 "Added a near-ground quadratic descent-overspeed penalty.",
61 "Added a dedicated in-zone crash penalty.",
62 "Reduced risky late horizontal corrections using a deadband."
63 ]
64 },
65 "evaluation": {
66 "n_episodes": 100,
67 "mean_shaped_reward": -413.80892156538437,
68 "std_shaped_reward": 213.20887834470014,
69 "shaped_course_score": -627.0177999100845,
70 "min_shaped_reward": -1183.712121425813,
71 "max_shaped_reward": -294.95529900279223,
72 "mean_original_reward": -671.0025925747099,
73 "std_original_reward": 62.16580799530145,
74 "original_course_score": -733.1684005700113,
75 "min_original_reward": -751.9068113021244,
76 "max_original_reward": -397.9538147700128,
77 "original_reward_200_rate": 0.0,
78 "flip_completion_rate": 0.15,
79 "recovery_completion_rate": 0.09,
80 "recovery_given_flip_rate": 0.6,
81 "safe_landing_rate": 0.0,
82 "flip_landing_rate": 0.0,
83 "landing_given_flip_rate": 0.0,
84 "stable_landing_rate": 0.0,
85 "inside_landing_zone_rate": 0.44,
86 "outside_zone_landing_rate": 0.0,
87 "in_zone_crash_rate": 0.13
88 },
89 "evaluation_seeds": [
90 20000,
91 20001,
92 20002,
93 20003,
94 20004,
95 20005,
96 20006,
97 20007,
98 20008,
99 20009,
100 20010,
101 20011,
102 20012,
103 20013,
104 20014,
105 20015,
106 20016,
107 20017,
108 20018,
109 20019,
110 20020,
111 20021,
112 20022,
113 20023,
114 20024,
115 20025,
116 20026,
117 20027,
118 20028,
119 20029,
120 20030,
121 20031,
122 20032,
123 20033,
124 20034,
125 20035,
126 20036,
127 20037,
128 20038,
129 20039,
130 20040,
131 20041,
132 20042,
133 20043,
134 20044,
135 20045,
136 20046,
137 20047,
138 20048,
139 20049,
140 20050,
141 20051,
142 20052,
143 20053,
144 20054,
145 20055,
146 20056,
147 20057,
148 20058,
149 20059,
150 20060,
151 20061,
152 20062,
153 20063,
154 20064,
155 20065,
156 20066,
157 20067,
158 20068,
159 20069,
160 20070,
161 20071,
162 20072,
163 20073,
164 20074,
165 20075,
166 20076,
167 20077,
168 20078,
169 20079,
170 20080,
171 20081,
172 20082,
173 20083,
174 20084,
175 20085,
176 20086,
177 20087,
178 20088,
179 20089,
180 20090,
181 20091,
182 20092,
183 20093,
184 20094,
185 20095,
186 20096,
187 20097,
188 20098,
189 20099
190 ],
191 "replay": {
192 "seed": 20000,
193 "shaped_reward": -294.95529900279223,
194 "original_reward": -669.6128317140386,
195 "steps": 133,
196 "flip_completed": false,
197 "landed_safely": false,
198 "rotations_completed": 0.9970662914171838,
199 "recovery_completed": false,
200 "custom_outcome": "episode_ended_without_flip",
201 "flip_and_landed": false
202 }
203 }