Robotics
LeRobot
Safetensors
act
imitation-learning
reinforcement-learning
grpo
mujoco
bimanual
roboeval
hybird-control
Sa74ll commited on
Commit
ce5f4e1
·
verified ·
1 Parent(s): 18ef0b5

Variant T: official run results and card numbers

Browse files
Files changed (2) hide show
  1. README.md +4 -4
  2. official_run/results.json +82 -82
README.md CHANGED
@@ -20,7 +20,7 @@ datasets:
20
  # LocalModels / ACT-bimanual
21
 
22
  Entry **ACT-bimanual** by team **LocalModels** (Saleh Alsharideh, Independent) to the CoRL 2026 RoboEval Diagnostic
23
- Bimanual Challenge. **Official score 0.9866**, 399 of 400 episodes successful.
24
 
25
  Seven tasks run hand-designed RGB-D controllers (no learned weights); LiftPot runs an Action Chunking Transformer (ACT)
26
  trained on the demonstrations and fine-tuned in simulation with GRPO. Code, report and explainer videos:
@@ -50,14 +50,14 @@ python -m roboeval.challenge.evaluate --submission submission --workers W --gpus
50
  | Task | Policy | Successes | Score |
51
  |---|---|---|---|
52
  | CubeHandover | controller | 50 | 1.000 |
53
- | LiftPot | ACT | 50 | 0.997 |
54
  | LiftTray | controller | 50 | 1.000 |
55
  | PackBox | controller | 49 | 0.954 |
56
- | PickSingleBookFromTable | controller | 50 | 0.988 |
57
  | RotateValve | controller | 50 | 1.000 |
58
  | StackSingleBookShelf | controller | 50 | 1.000 |
59
  | StackTwoBlocks | controller | 50 | 0.953 |
60
- | **Suite** | | **399 / 400** | **0.9866** |
61
 
62
  Training data and every use of simulator state: the report, Sec. 3. The files here equal the submitted ones except
63
  three edits that do not change behaviour (no contact e-mail in `metadata.yaml`, local paths shortened in the LiftPot
 
20
  # LocalModels / ACT-bimanual
21
 
22
  Entry **ACT-bimanual** by team **LocalModels** (Saleh Alsharideh, Independent) to the CoRL 2026 RoboEval Diagnostic
23
+ Bimanual Challenge. **Official score 0.9873**, 399 of 400 episodes successful.
24
 
25
  Seven tasks run hand-designed RGB-D controllers (no learned weights); LiftPot runs an Action Chunking Transformer (ACT)
26
  trained on the demonstrations and fine-tuned in simulation with GRPO. Code, report and explainer videos:
 
50
  | Task | Policy | Successes | Score |
51
  |---|---|---|---|
52
  | CubeHandover | controller | 50 | 1.000 |
53
+ | LiftPot | ACT | 50 | 1.000 |
54
  | LiftTray | controller | 50 | 1.000 |
55
  | PackBox | controller | 49 | 0.954 |
56
+ | PickSingleBookFromTable | controller | 50 | 0.991 |
57
  | RotateValve | controller | 50 | 1.000 |
58
  | StackSingleBookShelf | controller | 50 | 1.000 |
59
  | StackTwoBlocks | controller | 50 | 0.953 |
60
+ | **Suite** | | **399 / 400** | **0.9873** |
61
 
62
  Training data and every use of simulator state: the report, Sec. 3. The files here equal the submitted ones except
63
  three edits that do not change behaviour (no contact e-mail in `metadata.yaml`, local paths shortened in the LiftPot
official_run/results.json CHANGED
@@ -122,44 +122,44 @@
122
  "std": 0.0
123
  },
124
  "tp": {
125
- "mean": 0.99,
126
- "std": 0.06999999999999999
127
  },
128
  "bavd": {
129
- "mean": 0.29281951315243276,
130
- "std": 0.09462814601254968
131
  },
132
  "bgvd": {
133
- "mean": 0.010165391094207603,
134
- "std": 0.0046259623922256694
135
  },
136
  "cpl": {
137
- "mean": 0.6704045530620671,
138
- "std": 0.07377597417639128
139
  },
140
  "ct": {
141
- "mean": 2.967,
142
- "std": 0.1891057904983346
143
  },
144
  "ecc": {
145
  "mean": 0.0,
146
  "std": 0.0
147
  },
148
  "jpl": {
149
- "mean": 4.098847958860526,
150
- "std": 0.5163070688299158
151
  },
152
  "mcj": {
153
- "mean": 9.31802731270813,
154
- "std": 1.8950421588022368
155
  },
156
  "mjj": {
157
- "mean": 54.52174808567004,
158
- "std": 10.309689388763909
159
  },
160
  "opl": {
161
- "mean": 2.3032091026371906,
162
- "std": 0.39533000540578506
163
  },
164
  "scc": {
165
  "mean": 0.0,
@@ -170,8 +170,8 @@
170
  "std": 0.0
171
  },
172
  "tl": {
173
- "mean": 59.34,
174
- "std": 3.7821158099666907
175
  }
176
  },
177
  "terminations": {
@@ -546,52 +546,52 @@
546
  "std": 0.0
547
  },
548
  "bavd": {
549
- "mean": 0.9597556843993951,
550
- "std": 0.25022324245165284
551
  },
552
  "bgvd": {
553
- "mean": 0.21264552731291023,
554
- "std": 0.03902639572426828
555
  },
556
  "cpl": {
557
- "mean": 1.1374091543412113,
558
- "std": 0.44072051299774695
559
  },
560
  "ct": {
561
- "mean": 6.22,
562
- "std": 2.3655866080107915
563
  },
564
  "ecc": {
565
  "mean": 0.0,
566
  "std": 0.0
567
  },
568
  "jpl": {
569
- "mean": 7.1145422603480055,
570
- "std": 2.5897285047632086
571
  },
572
  "mcj": {
573
- "mean": 3.726751704228263,
574
- "std": 0.6317925151914751
575
  },
576
  "mjj": {
577
- "mean": 24.362044697427763,
578
- "std": 8.632357968377242
579
  },
580
  "opl": {
581
- "mean": 2.344576096088616,
582
- "std": 0.8547016352563385
583
  },
584
  "scc": {
585
  "mean": 0.0,
586
  "std": 0.0
587
  },
588
  "sc": {
589
- "mean": 0.02,
590
- "std": 0.13999999999999999
591
  },
592
  "tl": {
593
- "mean": 124.4,
594
- "std": 47.31173216021582
595
  }
596
  },
597
  "terminations": {
@@ -678,19 +678,19 @@
678
  "per_family": {
679
  "LiftPot": {
680
  "success": 1.0,
681
- "tp": 0.99,
682
- "bavd": 0.29281951315243276,
683
- "bgvd": 0.010165391094207603,
684
- "cpl": 0.6704045530620671,
685
- "ct": 2.967,
686
  "ecc": 0.0,
687
- "jpl": 4.098847958860526,
688
- "mcj": 9.31802731270813,
689
- "mjj": 54.52174808567004,
690
- "opl": 2.3032091026371906,
691
  "scc": 0.0,
692
  "sc": 0.0,
693
- "tl": 59.34
694
  },
695
  "LiftTray": {
696
  "success": 1.0,
@@ -775,18 +775,18 @@
775
  "PickSingleBookFromTable": {
776
  "success": 1.0,
777
  "tp": 1.0,
778
- "bavd": 0.9597556843993951,
779
- "bgvd": 0.21264552731291023,
780
- "cpl": 1.1374091543412113,
781
- "ct": 6.22,
782
  "ecc": 0.0,
783
- "jpl": 7.1145422603480055,
784
- "mcj": 3.726751704228263,
785
- "mjj": 24.362044697427763,
786
- "opl": 2.344576096088616,
787
  "scc": 0.0,
788
- "sc": 0.02,
789
- "tl": 124.4
790
  },
791
  "RotateValve": {
792
  "success": 1.0,
@@ -807,36 +807,36 @@
807
  },
808
  "suite": {
809
  "success": 0.9975,
810
- "tp": 0.99725,
811
- "bavd": 0.39505469212048305,
812
- "bgvd": 0.056357355682837666,
813
- "cpl": 1.186157018476982,
814
- "ct": 5.11475,
815
  "ecc": 0.615,
816
- "jpl": 5.6227863565119485,
817
- "mcj": 6.104427140596522,
818
- "mjj": 31.951006449470228,
819
- "opl": 1.9129712498958864,
820
  "scc": 0.0025,
821
- "sc": 0.0025,
822
- "tl": 102.295
823
  }
824
  },
825
  "score": {
826
  "overall": {
827
- "score": 0.9865902931563009,
828
- "outcome": 0.997425,
829
- "safety": 0.9871775025799794,
830
- "efficiency": 0.9658262433878773,
831
  "coordination": 0.9978352329278946
832
  },
833
  "tasks": {
834
  "LiftPot": {
835
- "outcome": 0.9969999999999999,
836
  "safety": 1.0,
837
  "efficiency": 1.0,
838
  "coordination": 1.0,
839
- "score": 0.9969999999999999
840
  },
841
  "LiftTray": {
842
  "outcome": 1.0,
@@ -875,10 +875,10 @@
875
  },
876
  "PickSingleBookFromTable": {
877
  "outcome": 1.0,
878
- "safety": 0.996078431372549,
879
- "efficiency": 0.947767624751622,
880
  "coordination": 1.0,
881
- "score": 0.9883770543620891
882
  },
883
  "RotateValve": {
884
  "outcome": 1.0,
@@ -890,7 +890,7 @@
890
  }
891
  },
892
  "provenance": {
893
- "created_utc": "2026-11-02T11:29:21+00:00",
894
  "roboeval_version": "4.1.0",
895
  "roboeval_metrics_version": "0.1.0",
896
  "git_commit": "c8b79598c9fb9021652128020517d1847abe4c95",
@@ -915,6 +915,6 @@
915
  "MKL_CBWR": "COMPATIBLE"
916
  }
917
  },
918
- "command": "/home/user/roboeval_orch/RoboEval/roboeval/challenge/evaluate.py --submission submission_pick22r --workers 6 --gpus 0"
919
  }
920
  }
 
122
  "std": 0.0
123
  },
124
  "tp": {
125
+ "mean": 1.0,
126
+ "std": 0.0
127
  },
128
  "bavd": {
129
+ "mean": 0.2945146608620109,
130
+ "std": 0.09876589515583759
131
  },
132
  "bgvd": {
133
+ "mean": 0.010150512516947072,
134
+ "std": 0.004701519102875442
135
  },
136
  "cpl": {
137
+ "mean": 0.6684489962231973,
138
+ "std": 0.07009133802024205
139
  },
140
  "ct": {
141
+ "mean": 2.958,
142
+ "std": 0.19192706948213428
143
  },
144
  "ecc": {
145
  "mean": 0.0,
146
  "std": 0.0
147
  },
148
  "jpl": {
149
+ "mean": 4.07982816465198,
150
+ "std": 0.4917501124475921
151
  },
152
  "mcj": {
153
+ "mean": 9.406648073650887,
154
+ "std": 1.9370018609342161
155
  },
156
  "mjj": {
157
+ "mean": 54.206255822281726,
158
+ "std": 11.447600563867148
159
  },
160
  "opl": {
161
+ "mean": 2.2894874527833933,
162
+ "std": 0.37732823525190384
163
  },
164
  "scc": {
165
  "mean": 0.0,
 
170
  "std": 0.0
171
  },
172
  "tl": {
173
+ "mean": 59.16,
174
+ "std": 3.838541389642686
175
  }
176
  },
177
  "terminations": {
 
546
  "std": 0.0
547
  },
548
  "bavd": {
549
+ "mean": 1.002609334236493,
550
+ "std": 0.2746527217467642
551
  },
552
  "bgvd": {
553
+ "mean": 0.2099936315236155,
554
+ "std": 0.03642409101543928
555
  },
556
  "cpl": {
557
+ "mean": 1.1383702368725983,
558
+ "std": 0.4376794275831977
559
  },
560
  "ct": {
561
+ "mean": 6.023,
562
+ "std": 2.357948472719453
563
  },
564
  "ecc": {
565
  "mean": 0.0,
566
  "std": 0.0
567
  },
568
  "jpl": {
569
+ "mean": 7.113438047014352,
570
+ "std": 2.577057582722626
571
  },
572
  "mcj": {
573
+ "mean": 3.91998318597079,
574
+ "std": 0.6575488557751517
575
  },
576
  "mjj": {
577
+ "mean": 25.840862916151305,
578
+ "std": 9.573895820755293
579
  },
580
  "opl": {
581
+ "mean": 2.3479093411990015,
582
+ "std": 0.8556182028029746
583
  },
584
  "scc": {
585
  "mean": 0.0,
586
  "std": 0.0
587
  },
588
  "sc": {
589
+ "mean": 0.0,
590
+ "std": 0.0
591
  },
592
  "tl": {
593
+ "mean": 120.46,
594
+ "std": 47.15896945438905
595
  }
596
  },
597
  "terminations": {
 
678
  "per_family": {
679
  "LiftPot": {
680
  "success": 1.0,
681
+ "tp": 1.0,
682
+ "bavd": 0.2945146608620109,
683
+ "bgvd": 0.010150512516947072,
684
+ "cpl": 0.6684489962231973,
685
+ "ct": 2.958,
686
  "ecc": 0.0,
687
+ "jpl": 4.07982816465198,
688
+ "mcj": 9.406648073650887,
689
+ "mjj": 54.206255822281726,
690
+ "opl": 2.2894874527833933,
691
  "scc": 0.0,
692
  "sc": 0.0,
693
+ "tl": 59.16
694
  },
695
  "LiftTray": {
696
  "success": 1.0,
 
775
  "PickSingleBookFromTable": {
776
  "success": 1.0,
777
  "tp": 1.0,
778
+ "bavd": 1.002609334236493,
779
+ "bgvd": 0.2099936315236155,
780
+ "cpl": 1.1383702368725983,
781
+ "ct": 6.023,
782
  "ecc": 0.0,
783
+ "jpl": 7.113438047014352,
784
+ "mcj": 3.91998318597079,
785
+ "mjj": 25.840862916151305,
786
+ "opl": 2.3479093411990015,
787
  "scc": 0.0,
788
+ "sc": 0.0,
789
+ "tl": 120.46
790
  },
791
  "RotateValve": {
792
  "success": 1.0,
 
807
  },
808
  "suite": {
809
  "success": 0.9975,
810
+ "tp": 0.9984999999999999,
811
+ "bavd": 0.40062329181381756,
812
+ "bgvd": 0.056024008887018265,
813
+ "cpl": 1.1860327091885465,
814
+ "ct": 5.089,
815
  "ecc": 0.615,
816
+ "jpl": 5.620270855569174,
817
+ "mcj": 6.1396586709321825,
818
+ "mjj": 32.096422193887136,
819
+ "opl": 1.91167269930296,
820
  "scc": 0.0025,
821
+ "sc": 0.0,
822
+ "tl": 101.78
823
  }
824
  },
825
  "score": {
826
  "overall": {
827
+ "score": 0.9872581084705565,
828
+ "outcome": 0.9978,
829
+ "safety": 0.9876676986584108,
830
+ "efficiency": 0.9665550258415083,
831
  "coordination": 0.9978352329278946
832
  },
833
  "tasks": {
834
  "LiftPot": {
835
+ "outcome": 1.0,
836
  "safety": 1.0,
837
  "efficiency": 1.0,
838
  "coordination": 1.0,
839
+ "score": 1.0
840
  },
841
  "LiftTray": {
842
  "outcome": 1.0,
 
875
  },
876
  "PickSingleBookFromTable": {
877
  "outcome": 1.0,
878
+ "safety": 1.0,
879
+ "efficiency": 0.9535978843806703,
880
  "coordination": 1.0,
881
+ "score": 0.9907195768761341
882
  },
883
  "RotateValve": {
884
  "outcome": 1.0,
 
890
  }
891
  },
892
  "provenance": {
893
+ "created_utc": "2026-11-02T11:44:40+00:00",
894
  "roboeval_version": "4.1.0",
895
  "roboeval_metrics_version": "0.1.0",
896
  "git_commit": "c8b79598c9fb9021652128020517d1847abe4c95",
 
915
  "MKL_CBWR": "COMPATIBLE"
916
  }
917
  },
918
+ "command": "/home/user/roboeval_orch/RoboEval/roboeval/challenge/evaluate.py --submission submission_pick22t --workers 6 --gpus 0"
919
  }
920
  }