Linslab commited on
Commit
c4568d7
·
verified ·
1 Parent(s): 42660db

Add level3_from_level2 PI0.5 multi-task LoRA transfer (level_2 -> level_3, 26 tasks, step 50000, 26/26 tasks evaluated, mean 0.002 over 50 ep)

Browse files
Files changed (30) hide show
  1. multi_task/level3_from_level2/eval_reports/ClutterGrasp-v1.json +171 -0
  2. multi_task/level3_from_level2/eval_reports/DualArmAssembleBracket-v1.json +171 -0
  3. multi_task/level3_from_level2/eval_reports/DualArmFoldCarton-v1.json +171 -0
  4. multi_task/level3_from_level2/eval_reports/DualArmLayerClayInDish-v1.json +171 -0
  5. multi_task/level3_from_level2/eval_reports/DualArmMakeLine-v1.json +171 -0
  6. multi_task/level3_from_level2/eval_reports/DualArmOpenFridge-v1.json +171 -0
  7. multi_task/level3_from_level2/eval_reports/DualArmOrganizeCabinet-v1.json +171 -0
  8. multi_task/level3_from_level2/eval_reports/DualArmPingpongBalance-v1.json +171 -0
  9. multi_task/level3_from_level2/eval_reports/DualArmPlugCharger-v1.json +171 -0
  10. multi_task/level3_from_level2/eval_reports/DualArmSetupTable-v1.json +171 -0
  11. multi_task/level3_from_level2/eval_reports/DualArmShapeDoll-v1.json +171 -0
  12. multi_task/level3_from_level2/eval_reports/DualArmThreadCloth-v1.json +171 -0
  13. multi_task/level3_from_level2/eval_reports/DualArmThreadRopeThroughTube-v1.json +171 -0
  14. multi_task/level3_from_level2/eval_reports/DualArmThreading-v1.json +171 -0
  15. multi_task/level3_from_level2/eval_reports/DualArmWeaveCloth-v1.json +171 -0
  16. multi_task/level3_from_level2/eval_reports/GroceryPrepare-v1.json +171 -0
  17. multi_task/level3_from_level2/eval_reports/HammerNail-v1.json +171 -0
  18. multi_task/level3_from_level2/eval_reports/HitBox-v1.json +171 -0
  19. multi_task/level3_from_level2/eval_reports/LiquidMeasuring-v1.json +171 -0
  20. multi_task/level3_from_level2/eval_reports/MazeBlockPush-v1.json +171 -0
  21. multi_task/level3_from_level2/eval_reports/MobilePickPlace-v1.json +171 -0
  22. multi_task/level3_from_level2/eval_reports/PressFitSoftPeg-v1.json +171 -0
  23. multi_task/level3_from_level2/eval_reports/PushTargetPose-v1.json +171 -0
  24. multi_task/level3_from_level2/eval_reports/SeparateTwoRopesEnv-v1.json +171 -0
  25. multi_task/level3_from_level2/eval_reports/Stack6Blocks-v1.json +171 -0
  26. multi_task/level3_from_level2/eval_reports/TweezersPickPlace-v1.json +171 -0
  27. multi_task/level3_from_level2/eval_result.json +200 -0
  28. multi_task/level3_from_level2/lora_step_0050000.npz +3 -0
  29. multi_task/level3_from_level2/metadata.json +108 -0
  30. multi_task/level3_from_level2/norm_stats.json +424 -0
multi_task/level3_from_level2/eval_reports/ClutterGrasp-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "ClutterGrasp-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ -102.701,
64
+ 333.777,
65
+ 1590.036,
66
+ 2494.553,
67
+ -69.682,
68
+ NaN,
69
+ 248.068,
70
+ -82.77,
71
+ -120.725,
72
+ 520.835,
73
+ -105.34,
74
+ 4918.164,
75
+ 5293.217,
76
+ 295.432,
77
+ -93.657,
78
+ 7494.0,
79
+ -131.41,
80
+ -105.674,
81
+ -90.338,
82
+ -144.956,
83
+ -168.156,
84
+ -73.539,
85
+ -98.629,
86
+ 7070.703,
87
+ -84.774,
88
+ -116.495,
89
+ -97.693,
90
+ 2147.689,
91
+ 6396.06,
92
+ 1339.257,
93
+ -97.911,
94
+ 1106.928,
95
+ 165.026,
96
+ 5214.071,
97
+ -124.146,
98
+ -81.091,
99
+ -77.646,
100
+ -78.477,
101
+ -92.543,
102
+ -119.411,
103
+ 5476.19,
104
+ 700.063,
105
+ -136.87,
106
+ -140.554,
107
+ 7151.135,
108
+ -112.503,
109
+ 1971.415,
110
+ -80.835,
111
+ -122.663,
112
+ -114.757
113
+ ],
114
+ "episode_steps": [
115
+ 2000,
116
+ 2000,
117
+ 2000,
118
+ 2000,
119
+ 2000,
120
+ 2000,
121
+ 2000,
122
+ 2000,
123
+ 2000,
124
+ 2000,
125
+ 2000,
126
+ 2000,
127
+ 2000,
128
+ 2000,
129
+ 2000,
130
+ 2000,
131
+ 2000,
132
+ 2000,
133
+ 2000,
134
+ 2000,
135
+ 2000,
136
+ 2000,
137
+ 2000,
138
+ 2000,
139
+ 2000,
140
+ 2000,
141
+ 2000,
142
+ 2000,
143
+ 2000,
144
+ 2000,
145
+ 2000,
146
+ 2000,
147
+ 2000,
148
+ 2000,
149
+ 2000,
150
+ 2000,
151
+ 2000,
152
+ 2000,
153
+ 2000,
154
+ 2000,
155
+ 2000,
156
+ 2000,
157
+ 2000,
158
+ 2000,
159
+ 2000,
160
+ 2000,
161
+ 2000,
162
+ 2000,
163
+ 2000,
164
+ 2000
165
+ ],
166
+ "return_mean": NaN,
167
+ "embodiment": "rigid_body/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.04,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmAssembleBracket-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmAssembleBracket-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 12.1,
64
+ 17.963,
65
+ 0.606,
66
+ 0.44,
67
+ 1.558,
68
+ 16.546,
69
+ 15.035,
70
+ 1.066,
71
+ 0.842,
72
+ 20.056,
73
+ 14.96,
74
+ 13.393,
75
+ 0.687,
76
+ 8.413,
77
+ 27.512,
78
+ 15.673,
79
+ 10.289,
80
+ 1.035,
81
+ 0.794,
82
+ 28.982,
83
+ 22.373,
84
+ 1.521,
85
+ 0.948,
86
+ 14.716,
87
+ 1.583,
88
+ 14.746,
89
+ 15.216,
90
+ 1.28,
91
+ 0.899,
92
+ 0.737,
93
+ 1.091,
94
+ 17.364,
95
+ 20.648,
96
+ 18.632,
97
+ 15.724,
98
+ 12.023,
99
+ 15.851,
100
+ 18.742,
101
+ 11.383,
102
+ 2.491,
103
+ 20.414,
104
+ 19.173,
105
+ 0.683,
106
+ 1.035,
107
+ 1.065,
108
+ 16.786,
109
+ 13.168,
110
+ 0.741,
111
+ 1.127,
112
+ 24.244
113
+ ],
114
+ "episode_steps": [
115
+ 1800,
116
+ 1800,
117
+ 1800,
118
+ 1800,
119
+ 1800,
120
+ 1800,
121
+ 1800,
122
+ 1800,
123
+ 1800,
124
+ 1800,
125
+ 1800,
126
+ 1800,
127
+ 1800,
128
+ 1800,
129
+ 1800,
130
+ 1800,
131
+ 1800,
132
+ 1800,
133
+ 1800,
134
+ 1800,
135
+ 1800,
136
+ 1800,
137
+ 1800,
138
+ 1800,
139
+ 1800,
140
+ 1800,
141
+ 1800,
142
+ 1800,
143
+ 1800,
144
+ 1800,
145
+ 1800,
146
+ 1800,
147
+ 1800,
148
+ 1800,
149
+ 1800,
150
+ 1800,
151
+ 1800,
152
+ 1800,
153
+ 1800,
154
+ 1800,
155
+ 1800,
156
+ 1800,
157
+ 1800,
158
+ 1800,
159
+ 1800,
160
+ 1800,
161
+ 1800,
162
+ 1800,
163
+ 1800,
164
+ 1800
165
+ ],
166
+ "return_mean": 10.287080000000001,
167
+ "embodiment": "rigid_body/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmFoldCarton-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmFoldCarton-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 9461.444,
64
+ 9058.281,
65
+ 5792.33,
66
+ 5326.19,
67
+ 5559.97,
68
+ 5890.31,
69
+ 5092.618,
70
+ 5762.759,
71
+ 5921.023,
72
+ 6236.28,
73
+ 5664.5,
74
+ 8555.656,
75
+ 5474.803,
76
+ 9609.662,
77
+ 5906.456,
78
+ 5352.047,
79
+ 6692.73,
80
+ 5343.926,
81
+ 6228.451,
82
+ 5609.719,
83
+ 5448.731,
84
+ 5898.349,
85
+ 5618.078,
86
+ 5446.383,
87
+ 5194.633,
88
+ 4988.619,
89
+ 5251.897,
90
+ 5763.418,
91
+ 5370.427,
92
+ 5478.599,
93
+ 5440.11,
94
+ 5663.422,
95
+ 5461.8,
96
+ 6627.439,
97
+ 5641.605,
98
+ 5739.761,
99
+ 5767.1,
100
+ 5648.267,
101
+ 5851.328,
102
+ 5467.301,
103
+ 5021.594,
104
+ 5599.05,
105
+ 5467.912,
106
+ 5871.398,
107
+ 5384.613,
108
+ 6541.564,
109
+ 5794.427,
110
+ 5806.257,
111
+ 5512.969,
112
+ 5514.84
113
+ ],
114
+ "episode_steps": [
115
+ 2400,
116
+ 2400,
117
+ 2400,
118
+ 2400,
119
+ 2400,
120
+ 2400,
121
+ 2400,
122
+ 2400,
123
+ 2400,
124
+ 2400,
125
+ 2400,
126
+ 2400,
127
+ 2400,
128
+ 2400,
129
+ 2400,
130
+ 2400,
131
+ 2400,
132
+ 2400,
133
+ 2400,
134
+ 2400,
135
+ 2400,
136
+ 2400,
137
+ 2400,
138
+ 2400,
139
+ 2400,
140
+ 2400,
141
+ 2400,
142
+ 2400,
143
+ 2400,
144
+ 2400,
145
+ 2400,
146
+ 2400,
147
+ 2400,
148
+ 2400,
149
+ 2400,
150
+ 2400,
151
+ 2400,
152
+ 2400,
153
+ 2400,
154
+ 2400,
155
+ 2400,
156
+ 2400,
157
+ 2400,
158
+ 2400,
159
+ 2400,
160
+ 2400,
161
+ 2400,
162
+ 2400,
163
+ 2400,
164
+ 2400
165
+ ],
166
+ "return_mean": 5936.42092,
167
+ "embodiment": "rigid_body/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmLayerClayInDish-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmLayerClayInDish-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 424.953,
64
+ 420.0,
65
+ 420.0,
66
+ 420.0,
67
+ 454.402,
68
+ 420.0,
69
+ 420.0,
70
+ 420.0,
71
+ 681.425,
72
+ 551.24,
73
+ 420.0,
74
+ 420.0,
75
+ 432.321,
76
+ 420.0,
77
+ 420.0,
78
+ 420.0,
79
+ 420.0,
80
+ 421.865,
81
+ 455.118,
82
+ 442.592,
83
+ 420.0,
84
+ 420.0,
85
+ 420.0,
86
+ 494.7,
87
+ 420.0,
88
+ 644.408,
89
+ 420.0,
90
+ 466.82,
91
+ 420.0,
92
+ 672.455,
93
+ 517.247,
94
+ 420.0,
95
+ 420.0,
96
+ 529.344,
97
+ 420.0,
98
+ 420.0,
99
+ 420.0,
100
+ 420.0,
101
+ 420.0,
102
+ 420.0,
103
+ 420.0,
104
+ 420.0,
105
+ 455.357,
106
+ 420.0,
107
+ 420.0,
108
+ 420.0,
109
+ 420.0,
110
+ 420.0,
111
+ 420.0,
112
+ 420.0
113
+ ],
114
+ "episode_steps": [
115
+ 600,
116
+ 600,
117
+ 600,
118
+ 600,
119
+ 600,
120
+ 600,
121
+ 600,
122
+ 600,
123
+ 600,
124
+ 600,
125
+ 600,
126
+ 600,
127
+ 600,
128
+ 600,
129
+ 600,
130
+ 600,
131
+ 600,
132
+ 600,
133
+ 600,
134
+ 600,
135
+ 600,
136
+ 600,
137
+ 600,
138
+ 600,
139
+ 600,
140
+ 600,
141
+ 600,
142
+ 600,
143
+ 600,
144
+ 600,
145
+ 600,
146
+ 600,
147
+ 600,
148
+ 600,
149
+ 600,
150
+ 600,
151
+ 600,
152
+ 600,
153
+ 600,
154
+ 600,
155
+ 600,
156
+ 600,
157
+ 600,
158
+ 600,
159
+ 600,
160
+ 600,
161
+ 600,
162
+ 600,
163
+ 600,
164
+ 600
165
+ ],
166
+ "return_mean": 446.88494,
167
+ "embodiment": "deformable/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmMakeLine-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmMakeLine-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 13191.022,
64
+ 11743.685,
65
+ 11049.028,
66
+ 13010.448,
67
+ 11922.528,
68
+ 12143.233,
69
+ 12370.073,
70
+ 12348.794,
71
+ 12567.115,
72
+ 11604.863,
73
+ 12566.997,
74
+ 12340.935,
75
+ 10847.524,
76
+ 12031.923,
77
+ 11799.289,
78
+ 12132.53,
79
+ 12495.139,
80
+ 11606.909,
81
+ 12145.66,
82
+ 12339.059,
83
+ 13354.16,
84
+ 14572.492,
85
+ 13996.093,
86
+ 13406.842,
87
+ 14009.339,
88
+ 12897.643,
89
+ 11455.245,
90
+ 12059.776,
91
+ 12181.661,
92
+ 12052.948,
93
+ 12397.184,
94
+ 12697.518,
95
+ 10844.666,
96
+ 15056.8,
97
+ 11453.126,
98
+ 14195.823,
99
+ 12124.038,
100
+ 11944.403,
101
+ 12891.294,
102
+ 12449.421,
103
+ 13059.952,
104
+ 12673.181,
105
+ 12578.186,
106
+ 11249.68,
107
+ 12782.983,
108
+ 12667.521,
109
+ 12616.911,
110
+ 12093.43,
111
+ 12584.944,
112
+ 11969.772
113
+ ],
114
+ "episode_steps": [
115
+ 2000,
116
+ 2000,
117
+ 2000,
118
+ 2000,
119
+ 2000,
120
+ 2000,
121
+ 2000,
122
+ 2000,
123
+ 2000,
124
+ 2000,
125
+ 2000,
126
+ 2000,
127
+ 2000,
128
+ 2000,
129
+ 2000,
130
+ 2000,
131
+ 2000,
132
+ 2000,
133
+ 2000,
134
+ 2000,
135
+ 2000,
136
+ 2000,
137
+ 2000,
138
+ 2000,
139
+ 2000,
140
+ 2000,
141
+ 2000,
142
+ 2000,
143
+ 2000,
144
+ 2000,
145
+ 2000,
146
+ 2000,
147
+ 2000,
148
+ 2000,
149
+ 2000,
150
+ 2000,
151
+ 2000,
152
+ 2000,
153
+ 2000,
154
+ 2000,
155
+ 2000,
156
+ 2000,
157
+ 2000,
158
+ 2000,
159
+ 2000,
160
+ 2000,
161
+ 2000,
162
+ 2000,
163
+ 2000,
164
+ 2000
165
+ ],
166
+ "return_mean": 12451.475719999999,
167
+ "embodiment": "rigid_body/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmOpenFridge-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmOpenFridge-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 1536.344,
64
+ 1335.538,
65
+ 1334.558,
66
+ 1332.136,
67
+ 1548.509,
68
+ 1339.22,
69
+ 1337.652,
70
+ 1511.712,
71
+ 1518.044,
72
+ 1453.153,
73
+ 1616.541,
74
+ 1338.257,
75
+ 1420.738,
76
+ 1533.831,
77
+ 1517.542,
78
+ 1519.106,
79
+ 1484.2,
80
+ 1337.552,
81
+ 1332.679,
82
+ 1448.39,
83
+ 1446.068,
84
+ 1340.085,
85
+ 1551.231,
86
+ 1335.821,
87
+ 1452.329,
88
+ 1393.654,
89
+ 1589.376,
90
+ 1334.93,
91
+ 1415.734,
92
+ 1338.78,
93
+ 1336.225,
94
+ 1403.108,
95
+ 1454.055,
96
+ 1330.608,
97
+ 1395.782,
98
+ 1333.432,
99
+ 1548.092,
100
+ 1484.238,
101
+ 1339.681,
102
+ 1503.042,
103
+ 1338.605,
104
+ 1417.404,
105
+ 1489.703,
106
+ 1429.259,
107
+ 1507.258,
108
+ 1334.046,
109
+ 1336.047,
110
+ 1341.894,
111
+ 1606.215,
112
+ 1335.349
113
+ ],
114
+ "episode_steps": [
115
+ 650,
116
+ 650,
117
+ 650,
118
+ 650,
119
+ 650,
120
+ 650,
121
+ 650,
122
+ 650,
123
+ 650,
124
+ 650,
125
+ 650,
126
+ 650,
127
+ 650,
128
+ 650,
129
+ 650,
130
+ 650,
131
+ 650,
132
+ 650,
133
+ 650,
134
+ 650,
135
+ 650,
136
+ 650,
137
+ 650,
138
+ 650,
139
+ 650,
140
+ 650,
141
+ 650,
142
+ 650,
143
+ 650,
144
+ 650,
145
+ 650,
146
+ 650,
147
+ 650,
148
+ 650,
149
+ 650,
150
+ 650,
151
+ 650,
152
+ 650,
153
+ 650,
154
+ 650,
155
+ 650,
156
+ 650,
157
+ 650,
158
+ 650,
159
+ 650,
160
+ 650,
161
+ 650,
162
+ 650,
163
+ 650,
164
+ 650
165
+ ],
166
+ "return_mean": 1425.15506,
167
+ "embodiment": "rigid_body/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmOrganizeCabinet-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmOrganizeCabinet-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 3256.927,
64
+ 223.672,
65
+ 833.013,
66
+ 2484.619,
67
+ 419.054,
68
+ 715.382,
69
+ 2454.291,
70
+ 290.792,
71
+ 3092.962,
72
+ 2727.653,
73
+ 347.336,
74
+ 989.184,
75
+ 866.216,
76
+ 1685.324,
77
+ 364.095,
78
+ 1439.248,
79
+ 1704.713,
80
+ 828.752,
81
+ 840.205,
82
+ 1619.286,
83
+ 267.61,
84
+ 1757.994,
85
+ 1122.736,
86
+ 869.531,
87
+ 921.757,
88
+ 255.571,
89
+ 263.016,
90
+ 418.048,
91
+ 1934.781,
92
+ 307.225,
93
+ 799.332,
94
+ 1708.443,
95
+ 806.577,
96
+ 1632.967,
97
+ 185.357,
98
+ 1765.039,
99
+ 2501.926,
100
+ 2381.271,
101
+ 1799.048,
102
+ 155.648,
103
+ 2527.278,
104
+ 339.008,
105
+ 2286.65,
106
+ 3146.135,
107
+ 3301.98,
108
+ 1755.532,
109
+ 1034.679,
110
+ 2466.927,
111
+ 243.473,
112
+ 2311.044
113
+ ],
114
+ "episode_steps": [
115
+ 5000,
116
+ 5000,
117
+ 5000,
118
+ 5000,
119
+ 5000,
120
+ 5000,
121
+ 5000,
122
+ 5000,
123
+ 5000,
124
+ 5000,
125
+ 5000,
126
+ 5000,
127
+ 5000,
128
+ 5000,
129
+ 5000,
130
+ 5000,
131
+ 5000,
132
+ 5000,
133
+ 5000,
134
+ 5000,
135
+ 5000,
136
+ 5000,
137
+ 5000,
138
+ 5000,
139
+ 5000,
140
+ 5000,
141
+ 5000,
142
+ 5000,
143
+ 5000,
144
+ 5000,
145
+ 5000,
146
+ 5000,
147
+ 5000,
148
+ 5000,
149
+ 5000,
150
+ 5000,
151
+ 5000,
152
+ 5000,
153
+ 5000,
154
+ 5000,
155
+ 5000,
156
+ 5000,
157
+ 5000,
158
+ 5000,
159
+ 5000,
160
+ 5000,
161
+ 5000,
162
+ 5000,
163
+ 5000,
164
+ 5000
165
+ ],
166
+ "return_mean": 1368.98614,
167
+ "embodiment": "rigid_body/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmPingpongBalance-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmPingpongBalance-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 82.976,
64
+ 99.337,
65
+ 71.536,
66
+ 102.271,
67
+ 107.443,
68
+ 76.463,
69
+ 101.088,
70
+ 97.119,
71
+ 106.704,
72
+ 100.091,
73
+ 101.735,
74
+ 106.496,
75
+ 93.63,
76
+ 76.969,
77
+ 104.628,
78
+ 71.109,
79
+ 99.296,
80
+ 99.142,
81
+ 69.993,
82
+ 97.671,
83
+ 107.598,
84
+ 102.114,
85
+ 102.255,
86
+ 99.467,
87
+ 101.064,
88
+ 87.247,
89
+ 105.786,
90
+ 74.789,
91
+ 105.239,
92
+ 102.633,
93
+ 91.986,
94
+ 108.736,
95
+ 97.85,
96
+ 105.126,
97
+ 107.374,
98
+ 102.86,
99
+ 98.093,
100
+ 100.609,
101
+ 94.967,
102
+ 96.277,
103
+ 72.8,
104
+ 110.642,
105
+ 105.393,
106
+ 95.273,
107
+ 97.14,
108
+ 109.149,
109
+ 112.178,
110
+ 99.891,
111
+ 100.148,
112
+ 95.915
113
+ ],
114
+ "episode_steps": [
115
+ 400,
116
+ 400,
117
+ 400,
118
+ 400,
119
+ 400,
120
+ 400,
121
+ 400,
122
+ 400,
123
+ 400,
124
+ 400,
125
+ 400,
126
+ 400,
127
+ 400,
128
+ 400,
129
+ 400,
130
+ 400,
131
+ 400,
132
+ 400,
133
+ 400,
134
+ 400,
135
+ 400,
136
+ 400,
137
+ 400,
138
+ 400,
139
+ 400,
140
+ 400,
141
+ 400,
142
+ 400,
143
+ 400,
144
+ 400,
145
+ 400,
146
+ 400,
147
+ 400,
148
+ 400,
149
+ 400,
150
+ 400,
151
+ 400,
152
+ 400,
153
+ 400,
154
+ 400,
155
+ 400,
156
+ 400,
157
+ 400,
158
+ 400,
159
+ 400,
160
+ 400,
161
+ 400,
162
+ 400,
163
+ 400,
164
+ 400
165
+ ],
166
+ "return_mean": 97.12592000000001,
167
+ "embodiment": "rigid_body/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmPlugCharger-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmPlugCharger-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 83.003,
64
+ 113.664,
65
+ 55.63,
66
+ 120.353,
67
+ 107.896,
68
+ 127.458,
69
+ 206.061,
70
+ 31.755,
71
+ 298.266,
72
+ 63.694,
73
+ 99.255,
74
+ 100.968,
75
+ 41.272,
76
+ 98.614,
77
+ 91.06,
78
+ 72.771,
79
+ 40.925,
80
+ 294.084,
81
+ 87.839,
82
+ 165.791,
83
+ 108.989,
84
+ 110.614,
85
+ 48.849,
86
+ 79.727,
87
+ 155.117,
88
+ 142.183,
89
+ 149.544,
90
+ 78.719,
91
+ 139.066,
92
+ 145.026,
93
+ 80.233,
94
+ 99.193,
95
+ 48.489,
96
+ 101.0,
97
+ 83.747,
98
+ 97.149,
99
+ 97.898,
100
+ 44.064,
101
+ 91.977,
102
+ 84.345,
103
+ 72.723,
104
+ 109.392,
105
+ 186.399,
106
+ 93.346,
107
+ 154.75,
108
+ 122.018,
109
+ 39.76,
110
+ 222.241,
111
+ 129.044,
112
+ 153.299
113
+ ],
114
+ "episode_steps": [
115
+ 2000,
116
+ 2000,
117
+ 2000,
118
+ 2000,
119
+ 2000,
120
+ 2000,
121
+ 2000,
122
+ 2000,
123
+ 2000,
124
+ 2000,
125
+ 2000,
126
+ 2000,
127
+ 2000,
128
+ 2000,
129
+ 2000,
130
+ 2000,
131
+ 2000,
132
+ 2000,
133
+ 2000,
134
+ 2000,
135
+ 2000,
136
+ 2000,
137
+ 2000,
138
+ 2000,
139
+ 2000,
140
+ 2000,
141
+ 2000,
142
+ 2000,
143
+ 2000,
144
+ 2000,
145
+ 2000,
146
+ 2000,
147
+ 2000,
148
+ 2000,
149
+ 2000,
150
+ 2000,
151
+ 2000,
152
+ 2000,
153
+ 2000,
154
+ 2000,
155
+ 2000,
156
+ 2000,
157
+ 2000,
158
+ 2000,
159
+ 2000,
160
+ 2000,
161
+ 2000,
162
+ 2000,
163
+ 2000,
164
+ 2000
165
+ ],
166
+ "return_mean": 111.3852,
167
+ "embodiment": "rigid_body/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmSetupTable-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmSetupTable-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 42.122,
64
+ 43.423,
65
+ 58.75,
66
+ 52.211,
67
+ 46.287,
68
+ 70.099,
69
+ 47.186,
70
+ 74.141,
71
+ 46.461,
72
+ 48.504,
73
+ 69.108,
74
+ 106.745,
75
+ 55.116,
76
+ 55.099,
77
+ 43.999,
78
+ 50.106,
79
+ 46.147,
80
+ 51.766,
81
+ 55.359,
82
+ 61.115,
83
+ 82.771,
84
+ 69.829,
85
+ 45.683,
86
+ 43.423,
87
+ 46.158,
88
+ 49.627,
89
+ 52.938,
90
+ 49.727,
91
+ 50.348,
92
+ 49.703,
93
+ 70.605,
94
+ 55.257,
95
+ 63.891,
96
+ 51.285,
97
+ 48.379,
98
+ 49.393,
99
+ 49.942,
100
+ 50.432,
101
+ 40.824,
102
+ 52.531,
103
+ 52.958,
104
+ 87.738,
105
+ 46.338,
106
+ 53.955,
107
+ 68.532,
108
+ 59.338,
109
+ 83.97,
110
+ 49.671,
111
+ 69.544,
112
+ 75.088
113
+ ],
114
+ "episode_steps": [
115
+ 750,
116
+ 750,
117
+ 750,
118
+ 750,
119
+ 750,
120
+ 750,
121
+ 750,
122
+ 750,
123
+ 750,
124
+ 750,
125
+ 750,
126
+ 750,
127
+ 750,
128
+ 750,
129
+ 750,
130
+ 750,
131
+ 750,
132
+ 750,
133
+ 750,
134
+ 750,
135
+ 750,
136
+ 750,
137
+ 750,
138
+ 750,
139
+ 750,
140
+ 750,
141
+ 750,
142
+ 750,
143
+ 750,
144
+ 750,
145
+ 750,
146
+ 750,
147
+ 750,
148
+ 750,
149
+ 750,
150
+ 750,
151
+ 750,
152
+ 750,
153
+ 750,
154
+ 750,
155
+ 750,
156
+ 750,
157
+ 750,
158
+ 750,
159
+ 750,
160
+ 750,
161
+ 750,
162
+ 750,
163
+ 750,
164
+ 750
165
+ ],
166
+ "return_mean": 56.87244,
167
+ "embodiment": "rigid_body/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmShapeDoll-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmShapeDoll-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 384.159,
64
+ 433.938,
65
+ 334.578,
66
+ 373.636,
67
+ 380.305,
68
+ 358.317,
69
+ 371.957,
70
+ 403.688,
71
+ 365.836,
72
+ 430.256,
73
+ 346.928,
74
+ 380.191,
75
+ 360.743,
76
+ 424.211,
77
+ 370.308,
78
+ 352.393,
79
+ 427.248,
80
+ 383.716,
81
+ 411.867,
82
+ 356.856,
83
+ 351.47,
84
+ 410.022,
85
+ 383.761,
86
+ 395.148,
87
+ 370.091,
88
+ 408.118,
89
+ 349.411,
90
+ 352.6,
91
+ 378.045,
92
+ 385.033,
93
+ 374.903,
94
+ 405.351,
95
+ 362.951,
96
+ 427.891,
97
+ 391.601,
98
+ 361.764,
99
+ 414.39,
100
+ 390.9,
101
+ 423.222,
102
+ 384.175,
103
+ 415.772,
104
+ 406.329,
105
+ 380.971,
106
+ 421.444,
107
+ 388.923,
108
+ 368.211,
109
+ 407.002,
110
+ 377.199,
111
+ 384.143,
112
+ 384.551
113
+ ],
114
+ "episode_steps": [
115
+ 1600,
116
+ 1600,
117
+ 1600,
118
+ 1600,
119
+ 1600,
120
+ 1600,
121
+ 1600,
122
+ 1600,
123
+ 1600,
124
+ 1600,
125
+ 1600,
126
+ 1600,
127
+ 1600,
128
+ 1600,
129
+ 1600,
130
+ 1600,
131
+ 1600,
132
+ 1600,
133
+ 1600,
134
+ 1600,
135
+ 1600,
136
+ 1600,
137
+ 1600,
138
+ 1600,
139
+ 1600,
140
+ 1600,
141
+ 1600,
142
+ 1600,
143
+ 1600,
144
+ 1600,
145
+ 1600,
146
+ 1600,
147
+ 1600,
148
+ 1600,
149
+ 1600,
150
+ 1600,
151
+ 1600,
152
+ 1600,
153
+ 1600,
154
+ 1600,
155
+ 1600,
156
+ 1600,
157
+ 1600,
158
+ 1600,
159
+ 1600,
160
+ 1600,
161
+ 1600,
162
+ 1600,
163
+ 1600,
164
+ 1600
165
+ ],
166
+ "return_mean": 386.13046,
167
+ "embodiment": "deformable/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmThreadCloth-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmThreadCloth-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.02,
7
+ "successes": 1,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 1.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 240.783,
64
+ 134.636,
65
+ 137.192,
66
+ 66.814,
67
+ 206.853,
68
+ 207.264,
69
+ 175.011,
70
+ 163.042,
71
+ 248.278,
72
+ 235.826,
73
+ 156.332,
74
+ 71.721,
75
+ 202.293,
76
+ 64.934,
77
+ 217.516,
78
+ 169.357,
79
+ 193.805,
80
+ 181.32,
81
+ 196.564,
82
+ 223.19,
83
+ 91.585,
84
+ 208.391,
85
+ 183.356,
86
+ 188.547,
87
+ 109.988,
88
+ 161.457,
89
+ 224.983,
90
+ 201.14,
91
+ 97.8,
92
+ 158.869,
93
+ 230.942,
94
+ 230.903,
95
+ 183.805,
96
+ 69.319,
97
+ 127.004,
98
+ 257.423,
99
+ 200.679,
100
+ 192.642,
101
+ 218.37,
102
+ 169.174,
103
+ 227.785,
104
+ 144.225,
105
+ 233.242,
106
+ 144.463,
107
+ 108.354,
108
+ 128.166,
109
+ 151.789,
110
+ 142.696,
111
+ 120.334,
112
+ 158.83
113
+ ],
114
+ "episode_steps": [
115
+ 400,
116
+ 272,
117
+ 400,
118
+ 400,
119
+ 400,
120
+ 400,
121
+ 400,
122
+ 400,
123
+ 400,
124
+ 400,
125
+ 400,
126
+ 400,
127
+ 400,
128
+ 400,
129
+ 400,
130
+ 400,
131
+ 400,
132
+ 400,
133
+ 400,
134
+ 400,
135
+ 400,
136
+ 400,
137
+ 400,
138
+ 400,
139
+ 400,
140
+ 400,
141
+ 400,
142
+ 400,
143
+ 400,
144
+ 400,
145
+ 400,
146
+ 400,
147
+ 400,
148
+ 400,
149
+ 400,
150
+ 400,
151
+ 400,
152
+ 400,
153
+ 400,
154
+ 400,
155
+ 400,
156
+ 400,
157
+ 400,
158
+ 400,
159
+ 400,
160
+ 400,
161
+ 400,
162
+ 400,
163
+ 400,
164
+ 400
165
+ ],
166
+ "return_mean": 171.17984,
167
+ "embodiment": "deformable/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.04,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmThreadRopeThroughTube-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmThreadRopeThroughTube-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 0.0,
64
+ 19.5,
65
+ 0.0,
66
+ 0.0,
67
+ 0.0,
68
+ 0.0,
69
+ 0.0,
70
+ 0.0,
71
+ 8.0,
72
+ 0.0,
73
+ 0.0,
74
+ 0.0,
75
+ 0.0,
76
+ 0.0,
77
+ 0.0,
78
+ 0.0,
79
+ 0.0,
80
+ 1.5,
81
+ 0.0,
82
+ 0.0,
83
+ 0.0,
84
+ 0.0,
85
+ 0.0,
86
+ 0.0,
87
+ 0.0,
88
+ 0.0,
89
+ 24.629,
90
+ 0.0,
91
+ 0.0,
92
+ 0.0,
93
+ 1.5,
94
+ 1.5,
95
+ 0.0,
96
+ 0.0,
97
+ 69.386,
98
+ 0.0,
99
+ 0.0,
100
+ 0.0,
101
+ 0.0,
102
+ 0.0,
103
+ 0.0,
104
+ 0.0,
105
+ 0.0,
106
+ 0.0,
107
+ 0.0,
108
+ 0.0,
109
+ 0.0,
110
+ 0.0,
111
+ 0.0,
112
+ 2.5
113
+ ],
114
+ "episode_steps": [
115
+ 470,
116
+ 470,
117
+ 470,
118
+ 470,
119
+ 470,
120
+ 470,
121
+ 470,
122
+ 470,
123
+ 470,
124
+ 470,
125
+ 470,
126
+ 470,
127
+ 470,
128
+ 470,
129
+ 470,
130
+ 470,
131
+ 470,
132
+ 470,
133
+ 470,
134
+ 470,
135
+ 470,
136
+ 470,
137
+ 470,
138
+ 470,
139
+ 470,
140
+ 470,
141
+ 470,
142
+ 470,
143
+ 470,
144
+ 470,
145
+ 470,
146
+ 470,
147
+ 470,
148
+ 470,
149
+ 470,
150
+ 470,
151
+ 470,
152
+ 470,
153
+ 470,
154
+ 470,
155
+ 470,
156
+ 470,
157
+ 470,
158
+ 470,
159
+ 470,
160
+ 470,
161
+ 470,
162
+ 470,
163
+ 470,
164
+ 470
165
+ ],
166
+ "return_mean": 2.5702999999999996,
167
+ "embodiment": "deformable/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmThreading-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmThreading-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.04,
7
+ "successes": 2,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 1.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 1.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 224.452,
64
+ NaN,
65
+ 220.073,
66
+ 221.234,
67
+ 239.758,
68
+ 221.855,
69
+ 305.72,
70
+ 227.11,
71
+ 285.19,
72
+ 220.312,
73
+ 193.786,
74
+ 221.768,
75
+ 255.641,
76
+ 222.892,
77
+ 224.132,
78
+ 331.131,
79
+ 204.903,
80
+ 251.134,
81
+ 238.083,
82
+ 222.998,
83
+ 224.033,
84
+ 191.598,
85
+ 224.661,
86
+ 218.593,
87
+ 203.051,
88
+ 194.601,
89
+ 223.667,
90
+ 297.535,
91
+ 226.145,
92
+ 218.056,
93
+ 301.055,
94
+ 279.49,
95
+ 222.514,
96
+ 225.148,
97
+ 214.588,
98
+ 216.388,
99
+ 222.715,
100
+ 212.239,
101
+ 227.879,
102
+ 214.903,
103
+ 211.926,
104
+ 242.696,
105
+ 225.212,
106
+ 227.501,
107
+ 336.279,
108
+ 221.728,
109
+ 218.855,
110
+ 224.248,
111
+ 220.277,
112
+ 233.999
113
+ ],
114
+ "episode_steps": [
115
+ 650,
116
+ 650,
117
+ 650,
118
+ 650,
119
+ 650,
120
+ 650,
121
+ 650,
122
+ 650,
123
+ 650,
124
+ 650,
125
+ 650,
126
+ 650,
127
+ 516,
128
+ 650,
129
+ 650,
130
+ 650,
131
+ 650,
132
+ 513,
133
+ 650,
134
+ 650,
135
+ 650,
136
+ 650,
137
+ 650,
138
+ 650,
139
+ 650,
140
+ 650,
141
+ 650,
142
+ 650,
143
+ 650,
144
+ 650,
145
+ 650,
146
+ 650,
147
+ 650,
148
+ 650,
149
+ 650,
150
+ 650,
151
+ 650,
152
+ 650,
153
+ 650,
154
+ 650,
155
+ 650,
156
+ 650,
157
+ 650,
158
+ 650,
159
+ 650,
160
+ 650,
161
+ 650,
162
+ 650,
163
+ 650,
164
+ 650
165
+ ],
166
+ "return_mean": NaN,
167
+ "embodiment": "deformable/dual_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/DualArmWeaveCloth-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "DualArmWeaveCloth-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 0.0,
64
+ 31.289,
65
+ 38.044,
66
+ 106.222,
67
+ 59.644,
68
+ 0.0,
69
+ 22.311,
70
+ 36.0,
71
+ 46.933,
72
+ 57.689,
73
+ 71.556,
74
+ 22.222,
75
+ 75.022,
76
+ 49.422,
77
+ 47.556,
78
+ 44.8,
79
+ 80.0,
80
+ 54.311,
81
+ 63.467,
82
+ 72.178,
83
+ 35.111,
84
+ 80.711,
85
+ 33.422,
86
+ 51.378,
87
+ 4.711,
88
+ 23.644,
89
+ 48.356,
90
+ 42.311,
91
+ 0.0,
92
+ 34.667,
93
+ 64.178,
94
+ 37.422,
95
+ 71.911,
96
+ 0.0,
97
+ 66.667,
98
+ 0.0,
99
+ 37.067,
100
+ 10.933,
101
+ 0.0,
102
+ 33.867,
103
+ 34.222,
104
+ 45.689,
105
+ 54.044,
106
+ 69.778,
107
+ 0.0,
108
+ 176.444,
109
+ 20.356,
110
+ 41.778,
111
+ 0.0,
112
+ 12.889
113
+ ],
114
+ "episode_steps": [
115
+ 500,
116
+ 500,
117
+ 500,
118
+ 500,
119
+ 500,
120
+ 500,
121
+ 500,
122
+ 500,
123
+ 500,
124
+ 500,
125
+ 500,
126
+ 500,
127
+ 500,
128
+ 500,
129
+ 500,
130
+ 500,
131
+ 500,
132
+ 500,
133
+ 500,
134
+ 500,
135
+ 500,
136
+ 500,
137
+ 500,
138
+ 500,
139
+ 500,
140
+ 500,
141
+ 500,
142
+ 500,
143
+ 500,
144
+ 500,
145
+ 500,
146
+ 500,
147
+ 500,
148
+ 500,
149
+ 500,
150
+ 500,
151
+ 500,
152
+ 500,
153
+ 500,
154
+ 500,
155
+ 500,
156
+ 500,
157
+ 500,
158
+ 500,
159
+ 500,
160
+ 500,
161
+ 500,
162
+ 500,
163
+ 500,
164
+ 500
165
+ ],
166
+ "return_mean": 42.20444,
167
+ "embodiment": "deformable/dual_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/GroceryPrepare-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "GroceryPrepare-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 2333.583,
64
+ 1852.986,
65
+ 1891.725,
66
+ 1844.542,
67
+ 2037.935,
68
+ 1951.72,
69
+ 1959.056,
70
+ 2163.779,
71
+ 2003.453,
72
+ 2325.772,
73
+ 2313.828,
74
+ 1869.459,
75
+ 2085.251,
76
+ 2195.032,
77
+ 2242.702,
78
+ 2216.855,
79
+ 2212.18,
80
+ 1919.391,
81
+ 1879.152,
82
+ 2329.389,
83
+ 2199.778,
84
+ 2104.275,
85
+ 2065.767,
86
+ 1977.567,
87
+ 1991.075,
88
+ 2074.859,
89
+ 2167.221,
90
+ 1776.275,
91
+ 2236.226,
92
+ 1964.537,
93
+ 1970.113,
94
+ 2185.301,
95
+ 2017.235,
96
+ 1821.834,
97
+ 2279.322,
98
+ 1879.427,
99
+ 2293.881,
100
+ 2240.039,
101
+ 1869.151,
102
+ 2187.609,
103
+ 2111.043,
104
+ 2140.395,
105
+ 2171.414,
106
+ 1993.378,
107
+ 2051.239,
108
+ 1861.75,
109
+ 1914.148,
110
+ 1929.071,
111
+ 2302.669,
112
+ 1851.673
113
+ ],
114
+ "episode_steps": [
115
+ 1600,
116
+ 1600,
117
+ 1600,
118
+ 1600,
119
+ 1600,
120
+ 1600,
121
+ 1600,
122
+ 1600,
123
+ 1600,
124
+ 1600,
125
+ 1600,
126
+ 1600,
127
+ 1600,
128
+ 1600,
129
+ 1600,
130
+ 1600,
131
+ 1600,
132
+ 1600,
133
+ 1600,
134
+ 1600,
135
+ 1600,
136
+ 1600,
137
+ 1600,
138
+ 1600,
139
+ 1600,
140
+ 1600,
141
+ 1600,
142
+ 1600,
143
+ 1600,
144
+ 1600,
145
+ 1600,
146
+ 1600,
147
+ 1600,
148
+ 1600,
149
+ 1600,
150
+ 1600,
151
+ 1600,
152
+ 1600,
153
+ 1600,
154
+ 1600,
155
+ 1600,
156
+ 1600,
157
+ 1600,
158
+ 1600,
159
+ 1600,
160
+ 1600,
161
+ 1600,
162
+ 1600,
163
+ 1600,
164
+ 1600
165
+ ],
166
+ "return_mean": 2065.12124,
167
+ "embodiment": "rigid_body/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/HammerNail-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "HammerNail-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ -77.888,
64
+ -82.361,
65
+ -61.309,
66
+ -61.592,
67
+ -68.451,
68
+ -64.835,
69
+ -54.843,
70
+ -70.639,
71
+ -64.65,
72
+ -69.158,
73
+ -66.401,
74
+ -80.739,
75
+ -71.94,
76
+ -72.454,
77
+ -61.675,
78
+ -59.952,
79
+ -67.811,
80
+ -44.252,
81
+ -68.104,
82
+ -62.45,
83
+ -72.672,
84
+ -67.842,
85
+ -66.839,
86
+ -73.89,
87
+ -49.532,
88
+ -77.133,
89
+ -72.273,
90
+ -44.185,
91
+ -50.956,
92
+ -47.865,
93
+ -62.078,
94
+ -63.868,
95
+ -71.569,
96
+ -68.141,
97
+ -62.477,
98
+ -72.236,
99
+ -59.465,
100
+ -80.268,
101
+ -70.796,
102
+ -72.587,
103
+ -63.817,
104
+ -70.545,
105
+ -63.748,
106
+ -58.038,
107
+ -77.811,
108
+ -68.216,
109
+ -74.002,
110
+ -67.503,
111
+ -63.183,
112
+ -67.854
113
+ ],
114
+ "episode_steps": [
115
+ 700,
116
+ 700,
117
+ 700,
118
+ 700,
119
+ 700,
120
+ 700,
121
+ 700,
122
+ 700,
123
+ 700,
124
+ 700,
125
+ 700,
126
+ 700,
127
+ 700,
128
+ 700,
129
+ 700,
130
+ 700,
131
+ 700,
132
+ 700,
133
+ 700,
134
+ 700,
135
+ 700,
136
+ 700,
137
+ 700,
138
+ 700,
139
+ 700,
140
+ 700,
141
+ 700,
142
+ 700,
143
+ 700,
144
+ 700,
145
+ 700,
146
+ 700,
147
+ 700,
148
+ 700,
149
+ 700,
150
+ 700,
151
+ 700,
152
+ 700,
153
+ 700,
154
+ 700,
155
+ 700,
156
+ 700,
157
+ 700,
158
+ 700,
159
+ 700,
160
+ 700,
161
+ 700,
162
+ 700,
163
+ 700,
164
+ 700
165
+ ],
166
+ "return_mean": -66.25786,
167
+ "embodiment": "rigid_body/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/HitBox-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "HitBox-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 517.546,
64
+ 292.072,
65
+ 267.622,
66
+ 279.031,
67
+ 528.518,
68
+ 346.608,
69
+ 328.034,
70
+ 460.764,
71
+ 297.115,
72
+ 273.464,
73
+ 459.69,
74
+ 472.424,
75
+ 477.458,
76
+ 275.312,
77
+ 490.36,
78
+ 481.344,
79
+ 331.032,
80
+ 435.694,
81
+ 403.94,
82
+ 416.515,
83
+ 436.297,
84
+ 259.179,
85
+ 410.365,
86
+ 313.576,
87
+ 280.397,
88
+ 287.959,
89
+ 325.218,
90
+ 372.662,
91
+ 283.515,
92
+ 486.289,
93
+ 419.705,
94
+ 322.409,
95
+ 279.083,
96
+ 672.983,
97
+ 359.033,
98
+ 304.494,
99
+ 361.278,
100
+ 303.317,
101
+ 582.315,
102
+ 301.622,
103
+ 429.631,
104
+ 432.428,
105
+ 319.346,
106
+ 351.088,
107
+ 326.027,
108
+ 592.331,
109
+ 424.82,
110
+ 263.516,
111
+ 280.084,
112
+ 275.059
113
+ ],
114
+ "episode_steps": [
115
+ 520,
116
+ 520,
117
+ 520,
118
+ 520,
119
+ 520,
120
+ 520,
121
+ 520,
122
+ 520,
123
+ 520,
124
+ 520,
125
+ 520,
126
+ 520,
127
+ 520,
128
+ 520,
129
+ 520,
130
+ 520,
131
+ 520,
132
+ 520,
133
+ 520,
134
+ 520,
135
+ 520,
136
+ 520,
137
+ 520,
138
+ 520,
139
+ 520,
140
+ 520,
141
+ 520,
142
+ 520,
143
+ 520,
144
+ 520,
145
+ 520,
146
+ 520,
147
+ 520,
148
+ 520,
149
+ 520,
150
+ 520,
151
+ 520,
152
+ 520,
153
+ 520,
154
+ 520,
155
+ 520,
156
+ 520,
157
+ 520,
158
+ 520,
159
+ 520,
160
+ 520,
161
+ 520,
162
+ 520,
163
+ 520,
164
+ 520
165
+ ],
166
+ "return_mean": 377.81138,
167
+ "embodiment": "deformable/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/LiquidMeasuring-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "LiquidMeasuring-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 64.917,
64
+ 60.0,
65
+ 119.692,
66
+ 111.65,
67
+ 21.477,
68
+ 131.708,
69
+ 95.621,
70
+ 45.111,
71
+ 117.843,
72
+ 48.182,
73
+ 74.817,
74
+ 21.416,
75
+ 112.51,
76
+ 70.618,
77
+ 49.755,
78
+ 159.297,
79
+ 43.236,
80
+ 76.423,
81
+ 27.257,
82
+ 138.411,
83
+ 39.157,
84
+ 38.915,
85
+ 37.604,
86
+ 71.201,
87
+ 37.721,
88
+ 58.845,
89
+ 41.729,
90
+ 81.174,
91
+ 112.685,
92
+ 77.685,
93
+ 110.366,
94
+ 75.35,
95
+ 68.57,
96
+ 174.532,
97
+ 177.329,
98
+ 183.142,
99
+ 66.02,
100
+ 52.053,
101
+ 125.355,
102
+ 33.026,
103
+ 218.061,
104
+ 29.954,
105
+ 196.712,
106
+ 88.339,
107
+ 50.314,
108
+ 97.508,
109
+ 79.685,
110
+ 34.804,
111
+ 149.823,
112
+ 78.673
113
+ ],
114
+ "episode_steps": [
115
+ 1000,
116
+ 1000,
117
+ 1000,
118
+ 1000,
119
+ 1000,
120
+ 1000,
121
+ 1000,
122
+ 1000,
123
+ 1000,
124
+ 1000,
125
+ 1000,
126
+ 1000,
127
+ 1000,
128
+ 1000,
129
+ 1000,
130
+ 1000,
131
+ 1000,
132
+ 1000,
133
+ 1000,
134
+ 1000,
135
+ 1000,
136
+ 1000,
137
+ 1000,
138
+ 1000,
139
+ 1000,
140
+ 1000,
141
+ 1000,
142
+ 1000,
143
+ 1000,
144
+ 1000,
145
+ 1000,
146
+ 1000,
147
+ 1000,
148
+ 1000,
149
+ 1000,
150
+ 1000,
151
+ 1000,
152
+ 1000,
153
+ 1000,
154
+ 1000,
155
+ 1000,
156
+ 1000,
157
+ 1000,
158
+ 1000,
159
+ 1000,
160
+ 1000,
161
+ 1000,
162
+ 1000,
163
+ 1000,
164
+ 1000
165
+ ],
166
+ "return_mean": 85.52546000000001,
167
+ "embodiment": "deformable/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/MazeBlockPush-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "MazeBlockPush-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ -674.812,
64
+ -754.625,
65
+ -340.238,
66
+ -623.604,
67
+ -601.92,
68
+ -655.05,
69
+ -510.015,
70
+ -570.37,
71
+ -534.644,
72
+ -481.014,
73
+ -729.282,
74
+ -403.866,
75
+ -615.851,
76
+ -830.092,
77
+ -711.088,
78
+ -561.512,
79
+ -561.836,
80
+ -553.104,
81
+ -549.8,
82
+ -566.367,
83
+ -724.556,
84
+ -673.923,
85
+ -743.228,
86
+ -606.156,
87
+ -630.052,
88
+ -554.764,
89
+ -573.315,
90
+ -435.388,
91
+ -1069.366,
92
+ -512.538,
93
+ -716.473,
94
+ -623.247,
95
+ -564.08,
96
+ -583.408,
97
+ -426.452,
98
+ -597.259,
99
+ -636.777,
100
+ -587.332,
101
+ -588.443,
102
+ -467.671,
103
+ -733.592,
104
+ -581.544,
105
+ -628.976,
106
+ -654.542,
107
+ -541.603,
108
+ -774.379,
109
+ -546.539,
110
+ -721.699,
111
+ -431.9,
112
+ -562.506
113
+ ],
114
+ "episode_steps": [
115
+ 1000,
116
+ 1000,
117
+ 1000,
118
+ 1000,
119
+ 1000,
120
+ 1000,
121
+ 1000,
122
+ 1000,
123
+ 1000,
124
+ 1000,
125
+ 1000,
126
+ 1000,
127
+ 1000,
128
+ 1000,
129
+ 1000,
130
+ 1000,
131
+ 1000,
132
+ 1000,
133
+ 1000,
134
+ 1000,
135
+ 1000,
136
+ 1000,
137
+ 1000,
138
+ 1000,
139
+ 1000,
140
+ 1000,
141
+ 1000,
142
+ 1000,
143
+ 1000,
144
+ 1000,
145
+ 1000,
146
+ 1000,
147
+ 1000,
148
+ 1000,
149
+ 1000,
150
+ 1000,
151
+ 1000,
152
+ 1000,
153
+ 1000,
154
+ 1000,
155
+ 1000,
156
+ 1000,
157
+ 1000,
158
+ 1000,
159
+ 1000,
160
+ 1000,
161
+ 1000,
162
+ 1000,
163
+ 1000,
164
+ 1000
165
+ ],
166
+ "return_mean": -606.4159599999999,
167
+ "embodiment": "rigid_body/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/MobilePickPlace-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "MobilePickPlace-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 22.449,
64
+ 13.624,
65
+ 16.586,
66
+ 11.531,
67
+ 26.522,
68
+ 5.868,
69
+ 16.591,
70
+ 15.625,
71
+ 22.334,
72
+ 16.042,
73
+ 27.867,
74
+ 5.355,
75
+ 13.031,
76
+ 20.607,
77
+ 20.805,
78
+ 26.538,
79
+ 16.655,
80
+ 18.915,
81
+ 1.214,
82
+ 36.813,
83
+ 16.887,
84
+ 22.073,
85
+ 20.449,
86
+ 12.472,
87
+ 19.175,
88
+ 10.241,
89
+ 21.632,
90
+ 16.037,
91
+ 11.279,
92
+ 24.623,
93
+ 11.539,
94
+ 19.673,
95
+ 14.027,
96
+ 8.297,
97
+ 12.15,
98
+ 10.096,
99
+ 21.406,
100
+ 18.786,
101
+ 5.653,
102
+ 15.819,
103
+ 16.926,
104
+ 10.634,
105
+ 17.541,
106
+ 12.185,
107
+ 258.141,
108
+ 252.323,
109
+ 5.07,
110
+ 2.368,
111
+ 26.214,
112
+ 24.775
113
+ ],
114
+ "episode_steps": [
115
+ 800,
116
+ 800,
117
+ 800,
118
+ 800,
119
+ 800,
120
+ 800,
121
+ 800,
122
+ 800,
123
+ 800,
124
+ 800,
125
+ 800,
126
+ 800,
127
+ 800,
128
+ 800,
129
+ 800,
130
+ 800,
131
+ 800,
132
+ 800,
133
+ 800,
134
+ 800,
135
+ 800,
136
+ 800,
137
+ 800,
138
+ 800,
139
+ 800,
140
+ 800,
141
+ 800,
142
+ 800,
143
+ 800,
144
+ 800,
145
+ 800,
146
+ 800,
147
+ 800,
148
+ 800,
149
+ 800,
150
+ 800,
151
+ 800,
152
+ 800,
153
+ 800,
154
+ 800,
155
+ 800,
156
+ 800,
157
+ 800,
158
+ 800,
159
+ 800,
160
+ 800,
161
+ 800,
162
+ 800,
163
+ 800,
164
+ 800
165
+ ],
166
+ "return_mean": 25.86926,
167
+ "embodiment": "rigid_body/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/PressFitSoftPeg-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "PressFitSoftPeg-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 494.936,
64
+ 392.072,
65
+ 2608.264,
66
+ 192.603,
67
+ 92.928,
68
+ 456.67,
69
+ 682.861,
70
+ 530.182,
71
+ 588.837,
72
+ 1320.581,
73
+ 588.54,
74
+ 284.956,
75
+ 623.466,
76
+ 158.403,
77
+ 437.147,
78
+ 428.617,
79
+ 474.821,
80
+ 2679.206,
81
+ 1320.797,
82
+ 354.332,
83
+ 432.508,
84
+ 258.354,
85
+ 1256.674,
86
+ 752.303,
87
+ 482.762,
88
+ 418.207,
89
+ 663.644,
90
+ 527.875,
91
+ 610.094,
92
+ 417.471,
93
+ 315.388,
94
+ 221.329,
95
+ 541.132,
96
+ 336.906,
97
+ 265.57,
98
+ 387.733,
99
+ 285.426,
100
+ 114.741,
101
+ 1249.881,
102
+ 236.007,
103
+ 669.307,
104
+ 379.173,
105
+ 677.293,
106
+ 414.44,
107
+ 354.548,
108
+ 142.985,
109
+ 269.169,
110
+ 792.432,
111
+ 440.72,
112
+ 2284.677
113
+ ],
114
+ "episode_steps": [
115
+ 1600,
116
+ 1600,
117
+ 1600,
118
+ 1600,
119
+ 1600,
120
+ 1600,
121
+ 1600,
122
+ 1600,
123
+ 1600,
124
+ 1600,
125
+ 1600,
126
+ 1600,
127
+ 1600,
128
+ 1600,
129
+ 1600,
130
+ 1600,
131
+ 1600,
132
+ 1600,
133
+ 1600,
134
+ 1600,
135
+ 1600,
136
+ 1600,
137
+ 1600,
138
+ 1600,
139
+ 1600,
140
+ 1600,
141
+ 1600,
142
+ 1600,
143
+ 1600,
144
+ 1600,
145
+ 1600,
146
+ 1600,
147
+ 1600,
148
+ 1600,
149
+ 1600,
150
+ 1600,
151
+ 1600,
152
+ 1600,
153
+ 1600,
154
+ 1600,
155
+ 1600,
156
+ 1600,
157
+ 1600,
158
+ 1600,
159
+ 1600,
160
+ 1600,
161
+ 1600,
162
+ 1600,
163
+ 1600,
164
+ 1600
165
+ ],
166
+ "return_mean": 618.17936,
167
+ "embodiment": "deformable/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/PushTargetPose-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "PushTargetPose-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ -548.65,
64
+ -515.571,
65
+ -494.366,
66
+ -613.657,
67
+ -590.325,
68
+ -522.859,
69
+ -480.187,
70
+ -712.254,
71
+ -591.961,
72
+ -717.6,
73
+ -570.997,
74
+ -377.463,
75
+ -314.543,
76
+ -517.084,
77
+ -586.867,
78
+ -553.585,
79
+ -607.301,
80
+ -416.768,
81
+ -587.945,
82
+ -714.624,
83
+ -593.253,
84
+ -631.924,
85
+ -459.067,
86
+ -534.675,
87
+ -586.531,
88
+ -643.01,
89
+ -631.638,
90
+ -591.719,
91
+ -589.748,
92
+ -699.948,
93
+ -542.168,
94
+ -555.172,
95
+ -332.813,
96
+ -627.045,
97
+ -604.547,
98
+ -585.584,
99
+ -552.171,
100
+ -653.58,
101
+ -541.846,
102
+ -712.591,
103
+ -557.171,
104
+ -426.772,
105
+ -346.4,
106
+ -611.711,
107
+ -598.177,
108
+ -582.146,
109
+ -562.074,
110
+ -688.592,
111
+ -588.225,
112
+ -567.678
113
+ ],
114
+ "episode_steps": [
115
+ 1200,
116
+ 1200,
117
+ 1200,
118
+ 1200,
119
+ 1200,
120
+ 1200,
121
+ 1200,
122
+ 1200,
123
+ 1200,
124
+ 1200,
125
+ 1200,
126
+ 1200,
127
+ 1200,
128
+ 1200,
129
+ 1200,
130
+ 1200,
131
+ 1200,
132
+ 1200,
133
+ 1200,
134
+ 1200,
135
+ 1200,
136
+ 1200,
137
+ 1200,
138
+ 1200,
139
+ 1200,
140
+ 1200,
141
+ 1200,
142
+ 1200,
143
+ 1200,
144
+ 1200,
145
+ 1200,
146
+ 1200,
147
+ 1200,
148
+ 1200,
149
+ 1200,
150
+ 1200,
151
+ 1200,
152
+ 1200,
153
+ 1200,
154
+ 1200,
155
+ 1200,
156
+ 1200,
157
+ 1200,
158
+ 1200,
159
+ 1200,
160
+ 1200,
161
+ 1200,
162
+ 1200,
163
+ 1200,
164
+ 1200
165
+ ],
166
+ "return_mean": -564.65166,
167
+ "embodiment": "rigid_body/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/SeparateTwoRopesEnv-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "SeparateTwoRopesEnv-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 1833.105,
64
+ 1146.046,
65
+ 1858.325,
66
+ 2371.713,
67
+ 3544.06,
68
+ 2221.341,
69
+ 2785.422,
70
+ 2357.311,
71
+ 1218.878,
72
+ 2682.828,
73
+ 2011.517,
74
+ 1897.322,
75
+ 1379.783,
76
+ 1542.494,
77
+ 1509.357,
78
+ 1187.285,
79
+ 1200.375,
80
+ 1618.844,
81
+ 1017.316,
82
+ 5254.226,
83
+ 1793.191,
84
+ 1699.475,
85
+ 2079.897,
86
+ 1364.667,
87
+ 1391.95,
88
+ 1730.692,
89
+ 1676.34,
90
+ 1731.407,
91
+ 1400.381,
92
+ 1655.092,
93
+ 1225.332,
94
+ 1606.973,
95
+ 1227.617,
96
+ 1366.804,
97
+ 1774.207,
98
+ 1487.975,
99
+ 1762.081,
100
+ 1436.706,
101
+ 1025.958,
102
+ 1107.112,
103
+ 1762.932,
104
+ 1509.134,
105
+ 1298.699,
106
+ 1812.694,
107
+ 1495.38,
108
+ 5636.531,
109
+ 1613.214,
110
+ 2150.108,
111
+ 1960.245,
112
+ 2515.816
113
+ ],
114
+ "episode_steps": [
115
+ 2000,
116
+ 2000,
117
+ 2000,
118
+ 2000,
119
+ 2000,
120
+ 2000,
121
+ 2000,
122
+ 2000,
123
+ 2000,
124
+ 2000,
125
+ 2000,
126
+ 2000,
127
+ 2000,
128
+ 2000,
129
+ 2000,
130
+ 2000,
131
+ 2000,
132
+ 2000,
133
+ 2000,
134
+ 2000,
135
+ 2000,
136
+ 2000,
137
+ 2000,
138
+ 2000,
139
+ 2000,
140
+ 2000,
141
+ 2000,
142
+ 2000,
143
+ 2000,
144
+ 2000,
145
+ 2000,
146
+ 2000,
147
+ 2000,
148
+ 2000,
149
+ 2000,
150
+ 2000,
151
+ 2000,
152
+ 2000,
153
+ 2000,
154
+ 2000,
155
+ 2000,
156
+ 2000,
157
+ 2000,
158
+ 2000,
159
+ 2000,
160
+ 2000,
161
+ 2000,
162
+ 2000,
163
+ 2000,
164
+ 2000
165
+ ],
166
+ "return_mean": 1858.72316,
167
+ "embodiment": "deformable/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.06,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/Stack6Blocks-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "Stack6Blocks-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ 1372.169,
64
+ 1347.511,
65
+ 1566.797,
66
+ 1408.813,
67
+ 1429.012,
68
+ 1430.028,
69
+ 1382.404,
70
+ 1458.603,
71
+ 1498.088,
72
+ 1651.029,
73
+ 1443.715,
74
+ 1430.241,
75
+ 1434.018,
76
+ 1397.502,
77
+ 1394.459,
78
+ 1438.497,
79
+ 1478.297,
80
+ 1353.731,
81
+ 1495.735,
82
+ 1408.506,
83
+ 1460.777,
84
+ 1393.847,
85
+ 1415.539,
86
+ 1566.165,
87
+ 1444.115,
88
+ 1438.587,
89
+ 1346.133,
90
+ 1451.801,
91
+ 1443.595,
92
+ 1419.771,
93
+ 1442.618,
94
+ 1423.925,
95
+ 1484.354,
96
+ 1391.094,
97
+ 1369.711,
98
+ 1443.035,
99
+ 1403.89,
100
+ 1429.418,
101
+ 1472.197,
102
+ 1360.091,
103
+ 1440.251,
104
+ 1461.023,
105
+ 1568.81,
106
+ 1447.994,
107
+ 1414.571,
108
+ 1378.775,
109
+ 1369.306,
110
+ 1388.668,
111
+ 1398.779,
112
+ 1400.048
113
+ ],
114
+ "episode_steps": [
115
+ 2400,
116
+ 2400,
117
+ 2400,
118
+ 2400,
119
+ 2400,
120
+ 2400,
121
+ 2400,
122
+ 2400,
123
+ 2400,
124
+ 2400,
125
+ 2400,
126
+ 2400,
127
+ 2400,
128
+ 2400,
129
+ 2400,
130
+ 2400,
131
+ 2400,
132
+ 2400,
133
+ 2400,
134
+ 2400,
135
+ 2400,
136
+ 2400,
137
+ 2400,
138
+ 2400,
139
+ 2400,
140
+ 2400,
141
+ 2400,
142
+ 2400,
143
+ 2400,
144
+ 2400,
145
+ 2400,
146
+ 2400,
147
+ 2400,
148
+ 2400,
149
+ 2400,
150
+ 2400,
151
+ 2400,
152
+ 2400,
153
+ 2400,
154
+ 2400,
155
+ 2400,
156
+ 2400,
157
+ 2400,
158
+ 2400,
159
+ 2400,
160
+ 2400,
161
+ 2400,
162
+ 2400,
163
+ 2400,
164
+ 2400
165
+ ],
166
+ "return_mean": 1433.76086,
167
+ "embodiment": "rigid_body/single_arm_gripper",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_reports/TweezersPickPlace-v1.json ADDED
@@ -0,0 +1,171 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "task": "TweezersPickPlace-v1",
3
+ "suite": "lv2to3",
4
+ "step": 50000,
5
+ "episodes": 50,
6
+ "success_rate": 0.0,
7
+ "successes": 0,
8
+ "episode_seed_base": 1000,
9
+ "episode_seeds": "1000-1049",
10
+ "episode_success": [
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0,
55
+ 0.0,
56
+ 0.0,
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0
61
+ ],
62
+ "episode_return": [
63
+ -135.049,
64
+ -132.403,
65
+ -132.067,
66
+ -132.816,
67
+ -135.982,
68
+ -132.514,
69
+ -129.82,
70
+ -131.284,
71
+ -133.142,
72
+ -130.059,
73
+ -124.138,
74
+ -130.249,
75
+ -132.904,
76
+ -130.407,
77
+ -129.973,
78
+ -130.492,
79
+ -134.756,
80
+ -130.587,
81
+ -134.924,
82
+ -132.403,
83
+ -133.006,
84
+ -131.886,
85
+ -133.87,
86
+ -129.198,
87
+ -134.184,
88
+ -135.096,
89
+ -134.114,
90
+ -131.486,
91
+ -132.413,
92
+ -128.231,
93
+ -133.353,
94
+ -131.169,
95
+ -132.341,
96
+ -133.694,
97
+ -125.856,
98
+ -134.897,
99
+ -130.609,
100
+ -132.28,
101
+ -132.297,
102
+ -129.444,
103
+ -129.74,
104
+ -130.284,
105
+ -132.187,
106
+ -131.052,
107
+ -134.388,
108
+ -132.708,
109
+ -132.303,
110
+ -135.031,
111
+ -133.988,
112
+ -134.269
113
+ ],
114
+ "episode_steps": [
115
+ 600,
116
+ 600,
117
+ 600,
118
+ 600,
119
+ 600,
120
+ 600,
121
+ 600,
122
+ 600,
123
+ 600,
124
+ 600,
125
+ 600,
126
+ 600,
127
+ 600,
128
+ 600,
129
+ 600,
130
+ 600,
131
+ 600,
132
+ 600,
133
+ 600,
134
+ 600,
135
+ 600,
136
+ 600,
137
+ 600,
138
+ 600,
139
+ 600,
140
+ 600,
141
+ 600,
142
+ 600,
143
+ 600,
144
+ 600,
145
+ 600,
146
+ 600,
147
+ 600,
148
+ 600,
149
+ 600,
150
+ 600,
151
+ 600,
152
+ 600,
153
+ 600,
154
+ 600,
155
+ 600,
156
+ 600,
157
+ 600,
158
+ 600,
159
+ 600,
160
+ 600,
161
+ 600,
162
+ 600,
163
+ 600,
164
+ 600
165
+ ],
166
+ "return_mean": -132.02686,
167
+ "embodiment": "rigid_body/single_arm_dex",
168
+ "level": "level_3",
169
+ "baseline_success_rate": 0.0,
170
+ "baseline": "multi_task/level_3/lora_step_0080000.npz (same tasks, trained from pi05_base)"
171
+ }
multi_task/level3_from_level2/eval_result.json ADDED
@@ -0,0 +1,200 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "suite": "level3_from_level2",
3
+ "source_level": "level_2",
4
+ "target_level": "level_3",
5
+ "step": 50000,
6
+ "num_tasks": 26,
7
+ "tasks_evaluated": 26,
8
+ "evaluation_complete": true,
9
+ "pending_tasks": [],
10
+ "episodes_per_task": 50,
11
+ "episode_seeds": "1000-1049",
12
+ "mean_success_rate": 0.0023076923076923075,
13
+ "baseline_mean_success_rate": 0.005384615384615385,
14
+ "per_task": {
15
+ "GroceryPrepare-v1": {
16
+ "success_rate": 0.0,
17
+ "successes": 0,
18
+ "return_mean": 2065.12124,
19
+ "baseline_success_rate": 0.0,
20
+ "embodiment": "rigid_body/single_arm_gripper"
21
+ },
22
+ "Stack6Blocks-v1": {
23
+ "success_rate": 0.0,
24
+ "successes": 0,
25
+ "return_mean": 1433.76086,
26
+ "baseline_success_rate": 0.0,
27
+ "embodiment": "rigid_body/single_arm_gripper"
28
+ },
29
+ "PushTargetPose-v1": {
30
+ "success_rate": 0.0,
31
+ "successes": 0,
32
+ "return_mean": -564.65166,
33
+ "baseline_success_rate": 0.0,
34
+ "embodiment": "rigid_body/single_arm_gripper"
35
+ },
36
+ "MazeBlockPush-v1": {
37
+ "success_rate": 0.0,
38
+ "successes": 0,
39
+ "return_mean": -606.4159599999999,
40
+ "baseline_success_rate": 0.0,
41
+ "embodiment": "rigid_body/single_arm_gripper"
42
+ },
43
+ "TweezersPickPlace-v1": {
44
+ "success_rate": 0.0,
45
+ "successes": 0,
46
+ "return_mean": -132.02686,
47
+ "baseline_success_rate": 0.0,
48
+ "embodiment": "rigid_body/single_arm_dex"
49
+ },
50
+ "ClutterGrasp-v1": {
51
+ "success_rate": 0.0,
52
+ "successes": 0,
53
+ "return_mean": NaN,
54
+ "baseline_success_rate": 0.04,
55
+ "embodiment": "rigid_body/single_arm_dex"
56
+ },
57
+ "MobilePickPlace-v1": {
58
+ "success_rate": 0.0,
59
+ "successes": 0,
60
+ "return_mean": 25.86926,
61
+ "baseline_success_rate": 0.0,
62
+ "embodiment": "rigid_body/single_arm_dex"
63
+ },
64
+ "HammerNail-v1": {
65
+ "success_rate": 0.0,
66
+ "successes": 0,
67
+ "return_mean": -66.25786,
68
+ "baseline_success_rate": 0.0,
69
+ "embodiment": "rigid_body/single_arm_dex"
70
+ },
71
+ "DualArmFoldCarton-v1": {
72
+ "success_rate": 0.0,
73
+ "successes": 0,
74
+ "return_mean": 5936.42092,
75
+ "baseline_success_rate": 0.0,
76
+ "embodiment": "rigid_body/dual_arm_gripper"
77
+ },
78
+ "DualArmOpenFridge-v1": {
79
+ "success_rate": 0.0,
80
+ "successes": 0,
81
+ "return_mean": 1425.15506,
82
+ "baseline_success_rate": 0.0,
83
+ "embodiment": "rigid_body/dual_arm_gripper"
84
+ },
85
+ "DualArmAssembleBracket-v1": {
86
+ "success_rate": 0.0,
87
+ "successes": 0,
88
+ "return_mean": 10.287080000000001,
89
+ "baseline_success_rate": 0.0,
90
+ "embodiment": "rigid_body/dual_arm_gripper"
91
+ },
92
+ "DualArmOrganizeCabinet-v1": {
93
+ "success_rate": 0.0,
94
+ "successes": 0,
95
+ "return_mean": 1368.98614,
96
+ "baseline_success_rate": 0.0,
97
+ "embodiment": "rigid_body/dual_arm_gripper"
98
+ },
99
+ "DualArmPingpongBalance-v1": {
100
+ "success_rate": 0.0,
101
+ "successes": 0,
102
+ "return_mean": 97.12592000000001,
103
+ "baseline_success_rate": 0.0,
104
+ "embodiment": "rigid_body/dual_arm_dex"
105
+ },
106
+ "DualArmSetupTable-v1": {
107
+ "success_rate": 0.0,
108
+ "successes": 0,
109
+ "return_mean": 56.87244,
110
+ "baseline_success_rate": 0.0,
111
+ "embodiment": "rigid_body/dual_arm_dex"
112
+ },
113
+ "DualArmMakeLine-v1": {
114
+ "success_rate": 0.0,
115
+ "successes": 0,
116
+ "return_mean": 12451.475719999999,
117
+ "baseline_success_rate": 0.0,
118
+ "embodiment": "rigid_body/dual_arm_dex"
119
+ },
120
+ "DualArmPlugCharger-v1": {
121
+ "success_rate": 0.0,
122
+ "successes": 0,
123
+ "return_mean": 111.3852,
124
+ "baseline_success_rate": 0.0,
125
+ "embodiment": "rigid_body/dual_arm_dex"
126
+ },
127
+ "DualArmLayerClayInDish-v1": {
128
+ "success_rate": 0.0,
129
+ "successes": 0,
130
+ "return_mean": 446.88494,
131
+ "baseline_success_rate": 0.0,
132
+ "embodiment": "deformable/dual_arm_dex"
133
+ },
134
+ "DualArmThreadCloth-v1": {
135
+ "success_rate": 0.02,
136
+ "successes": 1,
137
+ "return_mean": 171.17984,
138
+ "baseline_success_rate": 0.04,
139
+ "embodiment": "deformable/dual_arm_dex"
140
+ },
141
+ "DualArmThreading-v1": {
142
+ "success_rate": 0.04,
143
+ "successes": 2,
144
+ "return_mean": NaN,
145
+ "baseline_success_rate": 0.0,
146
+ "embodiment": "deformable/dual_arm_dex"
147
+ },
148
+ "DualArmWeaveCloth-v1": {
149
+ "success_rate": 0.0,
150
+ "successes": 0,
151
+ "return_mean": 42.20444,
152
+ "baseline_success_rate": 0.0,
153
+ "embodiment": "deformable/dual_arm_gripper"
154
+ },
155
+ "DualArmShapeDoll-v1": {
156
+ "success_rate": 0.0,
157
+ "successes": 0,
158
+ "return_mean": 386.13046,
159
+ "baseline_success_rate": 0.0,
160
+ "embodiment": "deformable/dual_arm_gripper"
161
+ },
162
+ "DualArmThreadRopeThroughTube-v1": {
163
+ "success_rate": 0.0,
164
+ "successes": 0,
165
+ "return_mean": 2.5702999999999996,
166
+ "baseline_success_rate": 0.0,
167
+ "embodiment": "deformable/dual_arm_gripper"
168
+ },
169
+ "HitBox-v1": {
170
+ "success_rate": 0.0,
171
+ "successes": 0,
172
+ "return_mean": 377.81138,
173
+ "baseline_success_rate": 0.0,
174
+ "embodiment": "deformable/single_arm_dex"
175
+ },
176
+ "PressFitSoftPeg-v1": {
177
+ "success_rate": 0.0,
178
+ "successes": 0,
179
+ "return_mean": 618.17936,
180
+ "baseline_success_rate": 0.0,
181
+ "embodiment": "deformable/single_arm_dex"
182
+ },
183
+ "LiquidMeasuring-v1": {
184
+ "success_rate": 0.0,
185
+ "successes": 0,
186
+ "return_mean": 85.52546000000001,
187
+ "baseline_success_rate": 0.0,
188
+ "embodiment": "deformable/single_arm_gripper"
189
+ },
190
+ "SeparateTwoRopesEnv-v1": {
191
+ "success_rate": 0.0,
192
+ "successes": 0,
193
+ "return_mean": 1858.72316,
194
+ "baseline_success_rate": 0.06,
195
+ "embodiment": "deformable/single_arm_gripper"
196
+ }
197
+ },
198
+ "note": "Only the final checkpoint (step 50000) was evaluated, 50 episodes per task; baseline = level_3 policy trained 80000 steps from pi05_base on the same tasks (multi_task/level_3).",
199
+ "campaign": "pi05_transfer_20261101"
200
+ }
multi_task/level3_from_level2/lora_step_0050000.npz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e70b239f67df087c387c75c92c3a6251cf8cea7b2f5ca684b3d55587f8604045
3
+ size 1867992394
multi_task/level3_from_level2/metadata.json ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "suite": "level3_from_level2",
3
+ "method": "PI0.5 LoRA + trainable weights, multi-task; level-to-level transfer",
4
+ "source_level": "level_2",
5
+ "target_level": "level_3",
6
+ "init_checkpoint": "multi_task/level_2/lora_step_0080000.npz",
7
+ "init_note": "All weights (base + LoRA + trainable towers) initialised from the level_2 multi-task policy at step 80000 (its full orbax params); the optimizer, warmup and cosine schedule were restarted. action_dim differs between source and target (44 -> 50): action_in_proj/kernel, action_out_proj/kernel and action_out_proj/bias were copied on the overlapping slice and the new slots initialised like a fresh Flax Linear layer; every other tensor was copied as is.",
8
+ "base_checkpoint": "gs://openpi-assets/checkpoints/pi05_base/params (via the source level policy)",
9
+ "checkpoint": "lora_step_0050000.npz",
10
+ "checkpoint_sha256": "e70b239f67df087c387c75c92c3a6251cf8cea7b2f5ca684b3d55587f8604045",
11
+ "checkpoint_bytes": 1867992394,
12
+ "tensor_count": 51,
13
+ "trainable_params": 466993954,
14
+ "tasks": [
15
+ "GroceryPrepare-v1",
16
+ "Stack6Blocks-v1",
17
+ "PushTargetPose-v1",
18
+ "MazeBlockPush-v1",
19
+ "TweezersPickPlace-v1",
20
+ "ClutterGrasp-v1",
21
+ "MobilePickPlace-v1",
22
+ "HammerNail-v1",
23
+ "DualArmFoldCarton-v1",
24
+ "DualArmOpenFridge-v1",
25
+ "DualArmAssembleBracket-v1",
26
+ "DualArmOrganizeCabinet-v1",
27
+ "DualArmPingpongBalance-v1",
28
+ "DualArmSetupTable-v1",
29
+ "DualArmMakeLine-v1",
30
+ "DualArmPlugCharger-v1",
31
+ "DualArmLayerClayInDish-v1",
32
+ "DualArmThreadCloth-v1",
33
+ "DualArmThreading-v1",
34
+ "DualArmWeaveCloth-v1",
35
+ "DualArmShapeDoll-v1",
36
+ "DualArmThreadRopeThroughTube-v1",
37
+ "HitBox-v1",
38
+ "PressFitSoftPeg-v1",
39
+ "LiquidMeasuring-v1",
40
+ "SeparateTwoRopesEnv-v1"
41
+ ],
42
+ "num_tasks": 26,
43
+ "prompts": {
44
+ "GroceryPrepare-v1": "perform the ManiLadder task: grocery prepare",
45
+ "Stack6Blocks-v1": "perform the ManiLadder task: stack6 blocks",
46
+ "PushTargetPose-v1": "perform the ManiLadder task: push target pose",
47
+ "MazeBlockPush-v1": "perform the ManiLadder task: maze block push",
48
+ "TweezersPickPlace-v1": "perform the ManiLadder task: tweezers pick place",
49
+ "ClutterGrasp-v1": "perform the ManiLadder task: clutter grasp",
50
+ "MobilePickPlace-v1": "perform the ManiLadder task: mobile pick place",
51
+ "HammerNail-v1": "perform the ManiLadder task: hammer nail",
52
+ "DualArmFoldCarton-v1": "perform the ManiLadder task: dual arm fold carton",
53
+ "DualArmOpenFridge-v1": "perform the ManiLadder task: dual arm open fridge",
54
+ "DualArmAssembleBracket-v1": "perform the ManiLadder task: dual arm assemble bracket",
55
+ "DualArmOrganizeCabinet-v1": "perform the ManiLadder task: dual arm organize cabinet",
56
+ "DualArmPingpongBalance-v1": "perform the ManiLadder task: dual arm pingpong balance",
57
+ "DualArmSetupTable-v1": "perform the ManiLadder task: dual arm setup table",
58
+ "DualArmMakeLine-v1": "perform the ManiLadder task: dual arm make line",
59
+ "DualArmPlugCharger-v1": "perform the ManiLadder task: dual arm plug charger",
60
+ "DualArmLayerClayInDish-v1": "perform the ManiLadder task: dual arm layer clay in dish",
61
+ "DualArmThreadCloth-v1": "perform the ManiLadder task: dual arm thread cloth",
62
+ "DualArmThreading-v1": "perform the ManiLadder task: dual arm threading",
63
+ "DualArmWeaveCloth-v1": "perform the ManiLadder task: dual arm weave cloth",
64
+ "DualArmShapeDoll-v1": "perform the ManiLadder task: dual arm shape doll",
65
+ "DualArmThreadRopeThroughTube-v1": "perform the ManiLadder task: dual arm thread rope through tube",
66
+ "HitBox-v1": "perform the ManiLadder task: hit box",
67
+ "PressFitSoftPeg-v1": "perform the ManiLadder task: press fit soft peg",
68
+ "LiquidMeasuring-v1": "perform the ManiLadder task: liquid measuring",
69
+ "SeparateTwoRopesEnv-v1": "perform the ManiLadder task: separate two ropes env"
70
+ },
71
+ "num_demos_per_task": 20,
72
+ "action_dim": 50,
73
+ "state_dim": 50,
74
+ "state_input": "zero vector",
75
+ "action_horizon": 50,
76
+ "max_token_len": 200,
77
+ "image_size": 224,
78
+ "exec_steps": 16,
79
+ "sample_steps": 10,
80
+ "control_mode": "pd_ee_delta_pose",
81
+ "train_steps": 50000,
82
+ "batch_size": 64,
83
+ "lr": 5e-05,
84
+ "warmup_steps": 1000,
85
+ "lr_schedule": "cosine decay 5e-5 -> 5e-6 over all 50000 steps",
86
+ "data_parallel_gpus": 8,
87
+ "gpu": "NVIDIA A100 80GB",
88
+ "fsdp_devices": 1,
89
+ "online_eval": false,
90
+ "norm_stats_file": "norm_stats.json",
91
+ "norm_stats_sha256": "a56198ac52c8710376ba23fb17460d6ffee6e7a2c4d90b2e2f98b3d319e3e4d4",
92
+ "norm_stats_note": "identical to multi_task/level_3/norm_stats.json: quantile stats of the target level, computed per dimension while ignoring padded (all-zero) action/state dimensions",
93
+ "training_data": "same dataset as multi_task/level_3: Linslab/ManiLadder_Dataset raw/<task>, 20 demos per task, revision 1190737391cc",
94
+ "training_data_overrides": {
95
+ "MazeBlockPush-v1": "local demos collected 2026-09-09 after the layout-pool change (HF copy of 2026-08-28 predates the env)"
96
+ },
97
+ "excluded_tasks": {
98
+ "BagHanging-v1": "level 3, excluded on request",
99
+ "BagDropIn-v1": "level 3, excluded on request"
100
+ },
101
+ "eval_episodes": 50,
102
+ "eval_seed_base": 1000,
103
+ "evaluated_steps": [
104
+ 50000
105
+ ],
106
+ "campaign": "pi05_transfer_20261101",
107
+ "trained": "2026-11-01 04:20 -> 18:17 (lyg2056)"
108
+ }
multi_task/level3_from_level2/norm_stats.json ADDED
@@ -0,0 +1,424 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "norm_stats": {
3
+ "state": {
4
+ "mean": [
5
+ 0.0,
6
+ 0.0,
7
+ 0.0,
8
+ 0.0,
9
+ 0.0,
10
+ 0.0,
11
+ 0.0,
12
+ 0.0,
13
+ 0.0,
14
+ 0.0,
15
+ 0.0,
16
+ 0.0,
17
+ 0.0,
18
+ 0.0,
19
+ 0.0,
20
+ 0.0,
21
+ 0.0,
22
+ 0.0,
23
+ 0.0,
24
+ 0.0,
25
+ 0.0,
26
+ 0.0,
27
+ 0.0,
28
+ 0.0,
29
+ 0.0,
30
+ 0.0,
31
+ 0.0,
32
+ 0.0,
33
+ 0.0,
34
+ 0.0,
35
+ 0.0,
36
+ 0.0,
37
+ 0.0,
38
+ 0.0,
39
+ 0.0,
40
+ 0.0,
41
+ 0.0,
42
+ 0.0,
43
+ 0.0,
44
+ 0.0,
45
+ 0.0,
46
+ 0.0,
47
+ 0.0,
48
+ 0.0,
49
+ 0.0,
50
+ 0.0,
51
+ 0.0,
52
+ 0.0,
53
+ 0.0,
54
+ 0.0
55
+ ],
56
+ "std": [
57
+ 0.0,
58
+ 0.0,
59
+ 0.0,
60
+ 0.0,
61
+ 0.0,
62
+ 0.0,
63
+ 0.0,
64
+ 0.0,
65
+ 0.0,
66
+ 0.0,
67
+ 0.0,
68
+ 0.0,
69
+ 0.0,
70
+ 0.0,
71
+ 0.0,
72
+ 0.0,
73
+ 0.0,
74
+ 0.0,
75
+ 0.0,
76
+ 0.0,
77
+ 0.0,
78
+ 0.0,
79
+ 0.0,
80
+ 0.0,
81
+ 0.0,
82
+ 0.0,
83
+ 0.0,
84
+ 0.0,
85
+ 0.0,
86
+ 0.0,
87
+ 0.0,
88
+ 0.0,
89
+ 0.0,
90
+ 0.0,
91
+ 0.0,
92
+ 0.0,
93
+ 0.0,
94
+ 0.0,
95
+ 0.0,
96
+ 0.0,
97
+ 0.0,
98
+ 0.0,
99
+ 0.0,
100
+ 0.0,
101
+ 0.0,
102
+ 0.0,
103
+ 0.0,
104
+ 0.0,
105
+ 0.0,
106
+ 0.0
107
+ ],
108
+ "q01": [
109
+ 0.0,
110
+ 0.0,
111
+ 0.0,
112
+ 0.0,
113
+ 0.0,
114
+ 0.0,
115
+ 0.0,
116
+ 0.0,
117
+ 0.0,
118
+ 0.0,
119
+ 0.0,
120
+ 0.0,
121
+ 0.0,
122
+ 0.0,
123
+ 0.0,
124
+ 0.0,
125
+ 0.0,
126
+ 0.0,
127
+ 0.0,
128
+ 0.0,
129
+ 0.0,
130
+ 0.0,
131
+ 0.0,
132
+ 0.0,
133
+ 0.0,
134
+ 0.0,
135
+ 0.0,
136
+ 0.0,
137
+ 0.0,
138
+ 0.0,
139
+ 0.0,
140
+ 0.0,
141
+ 0.0,
142
+ 0.0,
143
+ 0.0,
144
+ 0.0,
145
+ 0.0,
146
+ 0.0,
147
+ 0.0,
148
+ 0.0,
149
+ 0.0,
150
+ 0.0,
151
+ 0.0,
152
+ 0.0,
153
+ 0.0,
154
+ 0.0,
155
+ 0.0,
156
+ 0.0,
157
+ 0.0,
158
+ 0.0
159
+ ],
160
+ "q99": [
161
+ 0.0,
162
+ 0.0,
163
+ 0.0,
164
+ 0.0,
165
+ 0.0,
166
+ 0.0,
167
+ 0.0,
168
+ 0.0,
169
+ 0.0,
170
+ 0.0,
171
+ 0.0,
172
+ 0.0,
173
+ 0.0,
174
+ 0.0,
175
+ 0.0,
176
+ 0.0,
177
+ 0.0,
178
+ 0.0,
179
+ 0.0,
180
+ 0.0,
181
+ 0.0,
182
+ 0.0,
183
+ 0.0,
184
+ 0.0,
185
+ 0.0,
186
+ 0.0,
187
+ 0.0,
188
+ 0.0,
189
+ 0.0,
190
+ 0.0,
191
+ 0.0,
192
+ 0.0,
193
+ 0.0,
194
+ 0.0,
195
+ 0.0,
196
+ 0.0,
197
+ 0.0,
198
+ 0.0,
199
+ 0.0,
200
+ 0.0,
201
+ 0.0,
202
+ 0.0,
203
+ 0.0,
204
+ 0.0,
205
+ 0.0,
206
+ 0.0,
207
+ 0.0,
208
+ 0.0,
209
+ 0.0,
210
+ 0.0
211
+ ]
212
+ },
213
+ "actions": {
214
+ "mean": [
215
+ -0.033782027661800385,
216
+ -0.022587396204471588,
217
+ -0.29355403780937195,
218
+ 0.000647504348307848,
219
+ 0.002954313298687339,
220
+ 0.0004305039474274963,
221
+ -0.07470420747995377,
222
+ 0.019150469452142715,
223
+ 0.014058200642466545,
224
+ 0.003928771708160639,
225
+ -0.012139490805566311,
226
+ -0.010567662306129932,
227
+ -0.06447278708219528,
228
+ 0.0631028339266777,
229
+ 0.0007679514237679541,
230
+ 0.006326192058622837,
231
+ 0.004275909159332514,
232
+ 0.0032503237016499043,
233
+ 0.002842068439349532,
234
+ 0.02700621262192726,
235
+ 0.032175980508327484,
236
+ 0.022992270067334175,
237
+ -0.0005415640771389008,
238
+ 0.004833668936043978,
239
+ 0.0014543588040396571,
240
+ -0.00624785665422678,
241
+ -0.021385399624705315,
242
+ -0.0000353561517840717,
243
+ 0.0014680171152576804,
244
+ 0.008543578907847404,
245
+ 0.007964592427015305,
246
+ 0.0029826026875525713,
247
+ -0.000026650364816305228,
248
+ 0.005793577991425991,
249
+ 0.0056790742091834545,
250
+ 0.0045994482934474945,
251
+ 0.00008398477075388655,
252
+ 0.00603652186691761,
253
+ 0.006156971678137779,
254
+ 0.0054742698557674885,
255
+ 0.004678060300648212,
256
+ 0.0014415249461308122,
257
+ 0.025179289281368256,
258
+ 0.01766735129058361,
259
+ 0.0006659116479568183,
260
+ 0.00037842997699044645,
261
+ -0.0013420659815892577,
262
+ -0.00010432677663629875,
263
+ 0.002894642995670438,
264
+ 0.0011669923551380634
265
+ ],
266
+ "std": [
267
+ 0.28698259592056274,
268
+ 0.19696193933486938,
269
+ 0.9252079129219055,
270
+ 0.032511353492736816,
271
+ 0.04559716209769249,
272
+ 0.04620357230305672,
273
+ 0.6854458451271057,
274
+ 0.07256162166595459,
275
+ 0.06765808165073395,
276
+ 0.27136313915252686,
277
+ 0.0849372148513794,
278
+ 0.17261932790279388,
279
+ 0.4691751003265381,
280
+ 0.5133065581321716,
281
+ 0.008760109543800354,
282
+ 0.06635921448469162,
283
+ 0.057061173021793365,
284
+ 0.05286990851163864,
285
+ 0.0564114972949028,
286
+ 0.16191180050373077,
287
+ 0.10301927477121353,
288
+ 0.08579910546541214,
289
+ 0.027703961357474327,
290
+ 0.05496024712920189,
291
+ 0.040191732347011566,
292
+ 0.04620763659477234,
293
+ 0.15451842546463013,
294
+ 0.0016511352732777596,
295
+ 0.015850326046347618,
296
+ 0.04649820178747177,
297
+ 0.0505736842751503,
298
+ 0.047922283411026,
299
+ 0.001638531917706132,
300
+ 0.0400613397359848,
301
+ 0.03152022138237953,
302
+ 0.03210466727614403,
303
+ 0.019149769097566605,
304
+ 0.044146161526441574,
305
+ 0.03192608803510666,
306
+ 0.03284692019224167,
307
+ 0.04730696231126785,
308
+ 0.02281091921031475,
309
+ 0.104012131690979,
310
+ 0.07011067122220993,
311
+ 0.01952323690056801,
312
+ 0.01839304529130459,
313
+ 0.037878140807151794,
314
+ 0.003996575251221657,
315
+ 0.04017985239624977,
316
+ 0.024547340348362923
317
+ ],
318
+ "q01": [
319
+ -1.293516993522644,
320
+ -0.9122940247058868,
321
+ -3.1471363578796385,
322
+ -0.12001893310546874,
323
+ -0.06008779774308204,
324
+ -0.1309176956653595,
325
+ -1.0,
326
+ -0.12031810512542718,
327
+ -0.20866160240173337,
328
+ -1.0018507961273193,
329
+ -0.5926298097372055,
330
+ -1.1243762382268905,
331
+ -3.144608434963226,
332
+ -1.0018370251655577,
333
+ -0.009388944998383521,
334
+ -0.3939734101295471,
335
+ -0.356801942884922,
336
+ -0.31038713455200195,
337
+ -0.2772219181060791,
338
+ -0.11667379987239834,
339
+ -0.14657057349681846,
340
+ -0.14596793406009678,
341
+ -0.2190624426305294,
342
+ -0.08032799344062802,
343
+ -0.09003404917716984,
344
+ -0.280207644701004,
345
+ -0.8257047739863396,
346
+ -0.02669478952884674,
347
+ -0.031201852294802664,
348
+ -0.06293542943894863,
349
+ -0.13904655757546427,
350
+ -0.3136264556646347,
351
+ -0.007869197897613048,
352
+ -0.04900383576750755,
353
+ -0.05303184688091278,
354
+ -0.04026162251830101,
355
+ -0.19190412923693656,
356
+ -0.19206023216247559,
357
+ -9.76724004431162e-6,
358
+ -0.025828974321484566,
359
+ -0.06144549695253371,
360
+ -0.024112636876106264,
361
+ -0.06349016291499138,
362
+ -0.012862106871604925,
363
+ -0.24201032519340515,
364
+ -0.27323365211486816,
365
+ -0.4351250231266022,
366
+ -0.07336171716451645,
367
+ -0.3342823386192322,
368
+ -0.19350360333919525
369
+ ],
370
+ "q99": [
371
+ 0.9992663357734681,
372
+ 0.498039038705826,
373
+ 0.17883401031494195,
374
+ 0.11959322891235358,
375
+ 0.17566223419904714,
376
+ 0.1708689628124237,
377
+ 0.9996,
378
+ 0.2583850477695464,
379
+ 0.24012254574298852,
380
+ 0.9977392828941345,
381
+ 0.04414243321418765,
382
+ 0.31903822612762456,
383
+ 0.2490699649333954,
384
+ 0.998206401729584,
385
+ 0.049844462689012284,
386
+ 0.40833163261413574,
387
+ 0.3446492552757263,
388
+ 0.32548064321279524,
389
+ 0.2565118074417114,
390
+ 0.999718154013157,
391
+ 0.4331694315671921,
392
+ 0.3009268979549409,
393
+ 0.354452908039093,
394
+ 0.3597899972796441,
395
+ 0.21706091115474702,
396
+ 0.00966519737243654,
397
+ 0.00725956751108181,
398
+ 0.015374016743153368,
399
+ 0.08456174169778824,
400
+ 0.4509677290916443,
401
+ 0.1999645719647407,
402
+ 0.3613354504108993,
403
+ 0.003934723903238773,
404
+ 0.6071873307228088,
405
+ 0.41866064071655273,
406
+ 0.47751927375793457,
407
+ 0.2165757715702057,
408
+ 0.6177723407745361,
409
+ 0.4170748886466027,
410
+ 0.47528305649757385,
411
+ 0.32349352074861526,
412
+ 0.11218974010944366,
413
+ 0.604661538207531,
414
+ 0.32043952921628954,
415
+ 0.5277910232543945,
416
+ 0.527792751789093,
417
+ 0.811981201171875,
418
+ 0.10148198902606964,
419
+ 0.6083123683929443,
420
+ 0.5887998342514038
421
+ ]
422
+ }
423
+ }
424
+ }