Hugging Face's logo

RobotControlStack
/
duobench-act-sim-carry_pot

Robotics
LeRobot
Safetensors
duobench
bimanual
manipulation
simulation
act
Model card Files
xet
 Community
Juelg commited on
Commit
425dba7
·
verified ·
1 Parent(s): 266765b

Add model

Browse files
Files changed (8) hide show
  1. README.md +58 -0
  2. config.json +79 -0
  3. model.safetensors +3 -0
  4. policy_postprocessor.json +32 -0
  5. policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
  6. policy_preprocessor.json +72 -0
  7. policy_preprocessor_step_3_normalizer_processor.safetensors +3 -0
  8. train_config.json +211 -0
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: lerobot
4
+ pipeline_tag: robotics
5
+ datasets:
6
+ - RobotControlStack/duobench
7
+ tags:
8
+ - duobench
9
+ - bimanual
10
+ - manipulation
11
+ - simulation
12
+ - act
13
+ ---
14
+
15
+ # DuoBench ACT (sim)
16
+
17
+ ACT policy for the bimanual FR3 Duo setup of [DuoBench](https://arxiv.org/abs/2606.11901), trained on simulation data only for the DuoBench `carry_pot` sim task (single-task policy).
18
+ Trained from scratch with [LeRobot](https://github.com/huggingface/lerobot) on the [DuoBench dataset](https://huggingface.co/datasets/RobotControlStack/duobench).
19
+
20
+ Project page: https://duobench.github.io/
21
+ DuoBench Code: https://github.com/RobotControlStack/duobench
22
+
23
+ ## Training
24
+ - Data: 50 demonstrations per task from the `<task>/sim` folders of [RobotControlStack/duobench](https://huggingface.co/datasets/RobotControlStack/duobench)
25
+ - Inputs: `head`, `left_wrist` and `right_wrist` cameras (224x224) and the 16-D state (2 x 7 joints + 2 grippers); output: 16-D joint and gripper action
26
+ - 150,000 steps, batch size 64, learning rate 1e-5, action chunk 100, seed 1000
27
+ - LeRobot commit [`c8ce413d`](https://github.com/huggingface/lerobot/tree/c8ce413d738da15a2eed2d0832315779ea28cbf9); the full configuration is in `train_config.json`
28
+
29
+ ## Results
30
+ Success rate (%) in DuoBench simulation over 300 rollouts per task (seeds 0-299) with a Wilson 95% interval.
31
+
32
+ | Task | Success | 95% CI |
33
+ |---|---|---|
34
+ | carry_pot | 54.0 | [48.3, 59.6] |
35
+
36
+ ## Usage
37
+ ```python
38
+ # LeRobot at commit c8ce413d
39
+ from lerobot.policies.act.modeling_act import ACTPolicy
40
+
41
+ policy = ACTPolicy.from_pretrained("RobotControlStack/duobench-act-sim-carry_pot")
42
+ ```
43
+ See the [DuoBench code](https://github.com/RobotControlStack/duobench) for the simulation environments and evaluation.
44
+
45
+ ## Limitations
46
+ - Trained on simulation data only; not meant for the real robot.
47
+ - Loading needs a LeRobot version compatible with the commit above.
48
+
49
+ ## Citation
50
+ If you find this model or DuoBench useful for your work, please consider citing it:
51
+ ```
52
+ @misc{duobench,
53
+ title={{DuoBench}: A Reproducible Benchmark for Bimanual Manipulation in Simulation and the Real World},
54
+ author={Tobias J{\"u}lg and Seongjin Bien and Simon Hilber and Yannik Blei and Pierre Krack and Maximilian Li and Sven Parusel and Rudolf Lioutikov and Florian Walter and Wolfram Burgard},
55
+ year={2026},
56
+ url={https://arxiv.org/abs/2606.11901}
57
+ }
58
+ ```
config.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "act",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.images.head": {
6
+ "type": "VISUAL",
7
+ "shape": [
8
+ 3,
9
+ 224,
10
+ 224
11
+ ]
12
+ },
13
+ "observation.images.left_wrist": {
14
+ "type": "VISUAL",
15
+ "shape": [
16
+ 3,
17
+ 224,
18
+ 224
19
+ ]
20
+ },
21
+ "observation.images.right_wrist": {
22
+ "type": "VISUAL",
23
+ "shape": [
24
+ 3,
25
+ 224,
26
+ 224
27
+ ]
28
+ },
29
+ "observation.state": {
30
+ "type": "STATE",
31
+ "shape": [
32
+ 16
33
+ ]
34
+ }
35
+ },
36
+ "output_features": {
37
+ "action": {
38
+ "type": "ACTION",
39
+ "shape": [
40
+ 16
41
+ ]
42
+ }
43
+ },
44
+ "device": "cuda",
45
+ "use_amp": false,
46
+ "use_peft": false,
47
+ "push_to_hub": false,
48
+ "repo_id": null,
49
+ "private": null,
50
+ "tags": null,
51
+ "license": null,
52
+ "pretrained_path": null,
53
+ "chunk_size": 100,
54
+ "n_action_steps": 100,
55
+ "normalization_mapping": {
56
+ "VISUAL": "MEAN_STD",
57
+ "STATE": "MEAN_STD",
58
+ "ACTION": "MEAN_STD"
59
+ },
60
+ "vision_backbone": "resnet18",
61
+ "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
62
+ "replace_final_stride_with_dilation": false,
63
+ "pre_norm": false,
64
+ "dim_model": 512,
65
+ "n_heads": 8,
66
+ "dim_feedforward": 3200,
67
+ "feedforward_activation": "relu",
68
+ "n_encoder_layers": 4,
69
+ "n_decoder_layers": 1,
70
+ "use_vae": true,
71
+ "latent_dim": 32,
72
+ "n_vae_encoder_layers": 4,
73
+ "temporal_ensemble_coeff": null,
74
+ "dropout": 0.1,
75
+ "kl_weight": 10.0,
76
+ "optimizer_lr": 1e-05,
77
+ "optimizer_weight_decay": 0.0001,
78
+ "optimizer_lr_backbone": 1e-05
79
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:681afac6a89b139add4f2f65346eca7571319818d6c7b823db459da45ea57c72
3
+ size 206781696
policy_postprocessor.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_postprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "unnormalizer_processor",
6
+ "config": {
7
+ "eps": 1e-08,
8
+ "features": {
9
+ "action": {
10
+ "type": "ACTION",
11
+ "shape": [
12
+ 16
13
+ ]
14
+ }
15
+ },
16
+ "norm_map": {
17
+ "VISUAL": "MEAN_STD",
18
+ "STATE": "MEAN_STD",
19
+ "ACTION": "MEAN_STD"
20
+ }
21
+ },
22
+ "state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
23
+ },
24
+ {
25
+ "registry_name": "device_processor",
26
+ "config": {
27
+ "device": "cpu",
28
+ "float_dtype": null
29
+ }
30
+ }
31
+ ]
32
+ }
policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3361cae90d4f5f0b5c5ce42065c58b770aec45dde94ebddbd0e67ae4751a0103
3
+ size 9456
policy_preprocessor.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_preprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "rename_observations_processor",
6
+ "config": {
7
+ "rename_map": {}
8
+ }
9
+ },
10
+ {
11
+ "registry_name": "to_batch_processor",
12
+ "config": {}
13
+ },
14
+ {
15
+ "registry_name": "device_processor",
16
+ "config": {
17
+ "device": "cuda",
18
+ "float_dtype": null
19
+ }
20
+ },
21
+ {
22
+ "registry_name": "normalizer_processor",
23
+ "config": {
24
+ "eps": 1e-08,
25
+ "features": {
26
+ "observation.images.head": {
27
+ "type": "VISUAL",
28
+ "shape": [
29
+ 3,
30
+ 224,
31
+ 224
32
+ ]
33
+ },
34
+ "observation.images.left_wrist": {
35
+ "type": "VISUAL",
36
+ "shape": [
37
+ 3,
38
+ 224,
39
+ 224
40
+ ]
41
+ },
42
+ "observation.images.right_wrist": {
43
+ "type": "VISUAL",
44
+ "shape": [
45
+ 3,
46
+ 224,
47
+ 224
48
+ ]
49
+ },
50
+ "observation.state": {
51
+ "type": "STATE",
52
+ "shape": [
53
+ 16
54
+ ]
55
+ },
56
+ "action": {
57
+ "type": "ACTION",
58
+ "shape": [
59
+ 16
60
+ ]
61
+ }
62
+ },
63
+ "norm_map": {
64
+ "VISUAL": "MEAN_STD",
65
+ "STATE": "MEAN_STD",
66
+ "ACTION": "MEAN_STD"
67
+ }
68
+ },
69
+ "state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
70
+ }
71
+ ]
72
+ }
policy_preprocessor_step_3_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3361cae90d4f5f0b5c5ce42065c58b770aec45dde94ebddbd0e67ae4751a0103
3
+ size 9456
train_config.json ADDED
@@ -0,0 +1,211 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "duobench/carry_pot/sim",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": false,
8
+ "max_num_transforms": 3,
9
+ "random_order": false,
10
+ "tfs": {
11
+ "brightness": {
12
+ "weight": 1.0,
13
+ "type": "ColorJitter",
14
+ "kwargs": {
15
+ "brightness": [
16
+ 0.8,
17
+ 1.2
18
+ ]
19
+ }
20
+ },
21
+ "contrast": {
22
+ "weight": 1.0,
23
+ "type": "ColorJitter",
24
+ "kwargs": {
25
+ "contrast": [
26
+ 0.8,
27
+ 1.2
28
+ ]
29
+ }
30
+ },
31
+ "saturation": {
32
+ "weight": 1.0,
33
+ "type": "ColorJitter",
34
+ "kwargs": {
35
+ "saturation": [
36
+ 0.5,
37
+ 1.5
38
+ ]
39
+ }
40
+ },
41
+ "hue": {
42
+ "weight": 1.0,
43
+ "type": "ColorJitter",
44
+ "kwargs": {
45
+ "hue": [
46
+ -0.05,
47
+ 0.05
48
+ ]
49
+ }
50
+ },
51
+ "sharpness": {
52
+ "weight": 1.0,
53
+ "type": "SharpnessJitter",
54
+ "kwargs": {
55
+ "sharpness": [
56
+ 0.5,
57
+ 1.5
58
+ ]
59
+ }
60
+ },
61
+ "affine": {
62
+ "weight": 1.0,
63
+ "type": "RandomAffine",
64
+ "kwargs": {
65
+ "degrees": [
66
+ -5.0,
67
+ 5.0
68
+ ],
69
+ "translate": [
70
+ 0.05,
71
+ 0.05
72
+ ]
73
+ }
74
+ }
75
+ }
76
+ },
77
+ "revision": null,
78
+ "use_imagenet_stats": true,
79
+ "video_backend": "torchcodec",
80
+ "return_uint8": false,
81
+ "streaming": false
82
+ },
83
+ "env": null,
84
+ "policy": {
85
+ "type": "act",
86
+ "n_obs_steps": 1,
87
+ "input_features": {
88
+ "observation.images.head": {
89
+ "type": "VISUAL",
90
+ "shape": [
91
+ 3,
92
+ 224,
93
+ 224
94
+ ]
95
+ },
96
+ "observation.images.left_wrist": {
97
+ "type": "VISUAL",
98
+ "shape": [
99
+ 3,
100
+ 224,
101
+ 224
102
+ ]
103
+ },
104
+ "observation.images.right_wrist": {
105
+ "type": "VISUAL",
106
+ "shape": [
107
+ 3,
108
+ 224,
109
+ 224
110
+ ]
111
+ },
112
+ "observation.state": {
113
+ "type": "STATE",
114
+ "shape": [
115
+ 16
116
+ ]
117
+ }
118
+ },
119
+ "output_features": {
120
+ "action": {
121
+ "type": "ACTION",
122
+ "shape": [
123
+ 16
124
+ ]
125
+ }
126
+ },
127
+ "device": "cuda",
128
+ "use_amp": false,
129
+ "use_peft": false,
130
+ "push_to_hub": false,
131
+ "repo_id": null,
132
+ "private": null,
133
+ "tags": null,
134
+ "license": null,
135
+ "pretrained_path": null,
136
+ "chunk_size": 100,
137
+ "n_action_steps": 100,
138
+ "normalization_mapping": {
139
+ "VISUAL": "MEAN_STD",
140
+ "STATE": "MEAN_STD",
141
+ "ACTION": "MEAN_STD"
142
+ },
143
+ "vision_backbone": "resnet18",
144
+ "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
145
+ "replace_final_stride_with_dilation": false,
146
+ "pre_norm": false,
147
+ "dim_model": 512,
148
+ "n_heads": 8,
149
+ "dim_feedforward": 3200,
150
+ "feedforward_activation": "relu",
151
+ "n_encoder_layers": 4,
152
+ "n_decoder_layers": 1,
153
+ "use_vae": true,
154
+ "latent_dim": 32,
155
+ "n_vae_encoder_layers": 4,
156
+ "temporal_ensemble_coeff": null,
157
+ "dropout": 0.1,
158
+ "kl_weight": 10.0,
159
+ "optimizer_lr": 1e-05,
160
+ "optimizer_weight_decay": 0.0001,
161
+ "optimizer_lr_backbone": 1e-05
162
+ },
163
+ "reward_model": null,
164
+ "output_dir": null,
165
+ "job_name": "act_carry_pot",
166
+ "resume": false,
167
+ "seed": 1000,
168
+ "cudnn_deterministic": false,
169
+ "num_workers": 16,
170
+ "batch_size": 64,
171
+ "prefetch_factor": 4,
172
+ "persistent_workers": true,
173
+ "steps": 150000,
174
+ "eval_freq": 20000,
175
+ "log_freq": 200,
176
+ "tolerance_s": 0.0001,
177
+ "save_checkpoint": true,
178
+ "save_freq": 5000,
179
+ "use_policy_training_preset": true,
180
+ "optimizer": {
181
+ "type": "adamw",
182
+ "lr": 1e-05,
183
+ "weight_decay": 0.0001,
184
+ "grad_clip_norm": 10.0,
185
+ "betas": [
186
+ 0.9,
187
+ 0.999
188
+ ],
189
+ "eps": 1e-08
190
+ },
191
+ "scheduler": null,
192
+ "eval": {
193
+ "n_episodes": 50,
194
+ "batch_size": 50,
195
+ "use_async_envs": true
196
+ },
197
+ "wandb": {
198
+ "enable": false,
199
+ "disable_artifact": false,
200
+ "project": null,
201
+ "entity": null,
202
+ "notes": null,
203
+ "run_id": null,
204
+ "mode": null,
205
+ "add_tags": true
206
+ },
207
+ "peft": null,
208
+ "sample_weighting": null,
209
+ "rename_map": {},
210
+ "checkpoint_path": null
211
+ }