Hugging Face's logo

openEuler
/
pi05

Robotics
LeRobot
Safetensors
pi05
pytorch_model_hub_mixin
IB-Robot
LeRobot
so-101
banana-pick
vision-language-action
Model card Files
xet
 Community
7
BreezeWu commited on
Commit
d0619a7
·
verified ·
1 Parent(s): 9f91d2f

Upload 8 files

Browse files
Files changed (8) hide show
  1. config.json +105 -0
  2. inference_manifest.json +134 -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 +85 -0
  7. policy_preprocessor_step_2_normalizer_processor.safetensors +3 -0
  8. train_config.json +244 -0
config.json ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "pi05",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 6
9
+ ]
10
+ },
11
+ "observation.current": {
12
+ "type": "STATE",
13
+ "shape": [
14
+ 6
15
+ ]
16
+ },
17
+ "observation.images.top": {
18
+ "type": "VISUAL",
19
+ "shape": [
20
+ 3,
21
+ 480,
22
+ 640
23
+ ]
24
+ },
25
+ "observation.images.wrist": {
26
+ "type": "VISUAL",
27
+ "shape": [
28
+ 3,
29
+ 480,
30
+ 640
31
+ ]
32
+ }
33
+ },
34
+ "output_features": {
35
+ "action": {
36
+ "type": "ACTION",
37
+ "shape": [
38
+ 6
39
+ ]
40
+ }
41
+ },
42
+ "device": "cuda",
43
+ "use_amp": false,
44
+ "use_peft": false,
45
+ "push_to_hub": true,
46
+ "repo_id": "1arm_2cam_banana_pick_v1_20260407",
47
+ "private": null,
48
+ "tags": null,
49
+ "license": null,
50
+ "pretrained_path": "lerobot/pi05_base",
51
+ "paligemma_variant": "gemma_2b",
52
+ "action_expert_variant": "gemma_300m",
53
+ "dtype": "bfloat16",
54
+ "chunk_size": 50,
55
+ "n_action_steps": 50,
56
+ "max_state_dim": 32,
57
+ "max_action_dim": 32,
58
+ "num_inference_steps": 10,
59
+ "time_sampling_beta_alpha": 1.5,
60
+ "time_sampling_beta_beta": 1.0,
61
+ "time_sampling_scale": 0.999,
62
+ "time_sampling_offset": 0.001,
63
+ "min_period": 0.004,
64
+ "max_period": 4.0,
65
+ "use_relative_actions": false,
66
+ "relative_exclude_joints": [
67
+ "gripper"
68
+ ],
69
+ "action_feature_names": [
70
+ "shoulder_pan.pos",
71
+ "shoulder_lift.pos",
72
+ "elbow_flex.pos",
73
+ "wrist_flex.pos",
74
+ "wrist_roll.pos",
75
+ "gripper.pos"
76
+ ],
77
+ "rtc_config": null,
78
+ "image_resolution": [
79
+ 224,
80
+ 224
81
+ ],
82
+ "empty_cameras": 0,
83
+ "tokenizer_max_length": 200,
84
+ "normalization_mapping": {
85
+ "VISUAL": "IDENTITY",
86
+ "STATE": "QUANTILES",
87
+ "ACTION": "QUANTILES"
88
+ },
89
+ "gradient_checkpointing": true,
90
+ "compile_model": false,
91
+ "compile_mode": "max-autotune",
92
+ "freeze_vision_encoder": false,
93
+ "train_expert_only": false,
94
+ "optimizer_lr": 2.5e-05,
95
+ "optimizer_betas": [
96
+ 0.9,
97
+ 0.95
98
+ ],
99
+ "optimizer_eps": 1e-08,
100
+ "optimizer_weight_decay": 0.01,
101
+ "optimizer_grad_clip_norm": 1.0,
102
+ "scheduler_warmup_steps": 1000,
103
+ "scheduler_decay_steps": 30000,
104
+ "scheduler_decay_lr": 2.5e-06
105
+ }
inference_manifest.json ADDED
@@ -0,0 +1,134 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bundle": {
3
+ "digest": {
4
+ "algorithm": "sha256",
5
+ "scope": "structure",
6
+ "value": "aef5f2953c5194c3db86b22c9e12956850542284ff8506129d6fb9a5898786cb"
7
+ },
8
+ "files": [
9
+ {
10
+ "path": "bert-base-uncased/added_tokens.json"
11
+ },
12
+ {
13
+ "path": "bert-base-uncased/special_tokens_map.json"
14
+ },
15
+ {
16
+ "path": "bert-base-uncased/tokenizer.json"
17
+ },
18
+ {
19
+ "path": "bert-base-uncased/tokenizer.model"
20
+ },
21
+ {
22
+ "path": "bert-base-uncased/tokenizer_config.json"
23
+ },
24
+ {
25
+ "path": "config.json"
26
+ },
27
+ {
28
+ "path": "model.safetensors"
29
+ },
30
+ {
31
+ "path": "policy_postprocessor.json"
32
+ },
33
+ {
34
+ "path": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
35
+ },
36
+ {
37
+ "path": "policy_preprocessor.json"
38
+ },
39
+ {
40
+ "path": "policy_preprocessor_step_2_normalizer_processor.safetensors"
41
+ }
42
+ ],
43
+ "name": "pi05",
44
+ "revision": 1,
45
+ "uuid": "9a7ed292-8e4f-4771-864b-20e3de990ff7"
46
+ },
47
+ "deployments": {
48
+ "torch-cpu": {
49
+ "execution_contract": {
50
+ "cancellation_granularity": "request_boundary",
51
+ "execution_structure": "direct",
52
+ "state_scope": "request"
53
+ },
54
+ "revision": 1,
55
+ "runtime_profile": {
56
+ "backend": "torch",
57
+ "profile": {
58
+ "device": "cpu"
59
+ },
60
+ "target": {
61
+ "runtime": "torch"
62
+ }
63
+ },
64
+ "uuid": "a3097e02-bdb9-4273-80d2-ec7a1ee48dc0"
65
+ },
66
+ "torch-cuda": {
67
+ "execution_contract": {
68
+ "cancellation_granularity": "request_boundary",
69
+ "execution_structure": "direct",
70
+ "state_scope": "request"
71
+ },
72
+ "revision": 1,
73
+ "runtime_profile": {
74
+ "backend": "torch",
75
+ "profile": {
76
+ "device": "cuda"
77
+ },
78
+ "target": {
79
+ "runtime": "torch"
80
+ }
81
+ },
82
+ "uuid": "4aa27107-60a4-40d3-989b-2e1557f48e04"
83
+ }
84
+ },
85
+ "model": {
86
+ "inputs": [
87
+ {
88
+ "dtype": "float32",
89
+ "semantic": "observation.state",
90
+ "shape": [
91
+ 6
92
+ ]
93
+ },
94
+ {
95
+ "dtype": "float32",
96
+ "semantic": "observation.current",
97
+ "shape": [
98
+ 6
99
+ ]
100
+ },
101
+ {
102
+ "dtype": "float32",
103
+ "semantic": "observation.images.top",
104
+ "shape": [
105
+ 3,
106
+ 480,
107
+ 640
108
+ ]
109
+ },
110
+ {
111
+ "dtype": "float32",
112
+ "semantic": "observation.images.wrist",
113
+ "shape": [
114
+ 3,
115
+ 480,
116
+ 640
117
+ ]
118
+ }
119
+ ],
120
+ "interface": "policy",
121
+ "model_type": "pi05",
122
+ "operation": "predict",
123
+ "outputs": [
124
+ {
125
+ "dtype": "float32",
126
+ "semantic": "action",
127
+ "shape": [
128
+ 6
129
+ ]
130
+ }
131
+ ]
132
+ },
133
+ "schema_version": 3
134
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc3a2749c4de0b99c6b5a5c9cedc2c07464df146ca000dc4f09fa1dd7a81b1b8
3
+ size 9354050752
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
+ 6
13
+ ]
14
+ }
15
+ },
16
+ "norm_map": {
17
+ "VISUAL": "IDENTITY",
18
+ "STATE": "QUANTILES",
19
+ "ACTION": "QUANTILES"
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:b5f2adea2aea10fb1afb86c552d4c6f6f29421f68df10a99e0fc200239356631
3
+ size 8588
policy_preprocessor.json ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "normalizer_processor",
16
+ "config": {
17
+ "eps": 1e-08,
18
+ "features": {
19
+ "observation.state": {
20
+ "type": "STATE",
21
+ "shape": [
22
+ 6
23
+ ]
24
+ },
25
+ "observation.current": {
26
+ "type": "STATE",
27
+ "shape": [
28
+ 6
29
+ ]
30
+ },
31
+ "observation.images.top": {
32
+ "type": "VISUAL",
33
+ "shape": [
34
+ 3,
35
+ 480,
36
+ 640
37
+ ]
38
+ },
39
+ "observation.images.wrist": {
40
+ "type": "VISUAL",
41
+ "shape": [
42
+ 3,
43
+ 480,
44
+ 640
45
+ ]
46
+ },
47
+ "action": {
48
+ "type": "ACTION",
49
+ "shape": [
50
+ 6
51
+ ]
52
+ }
53
+ },
54
+ "norm_map": {
55
+ "VISUAL": "IDENTITY",
56
+ "STATE": "QUANTILES",
57
+ "ACTION": "QUANTILES"
58
+ }
59
+ },
60
+ "state_file": "policy_preprocessor_step_2_normalizer_processor.safetensors"
61
+ },
62
+ {
63
+ "registry_name": "pi05_prepare_state_tokenizer_processor_step",
64
+ "config": {}
65
+ },
66
+ {
67
+ "registry_name": "tokenizer_processor",
68
+ "config": {
69
+ "max_length": 200,
70
+ "task_key": "task",
71
+ "padding_side": "right",
72
+ "padding": "max_length",
73
+ "truncation": true,
74
+ "tokenizer_name": "bert-base-uncased"
75
+ }
76
+ },
77
+ {
78
+ "registry_name": "device_processor",
79
+ "config": {
80
+ "device": "cuda",
81
+ "float_dtype": null
82
+ }
83
+ }
84
+ ]
85
+ }
policy_preprocessor_step_2_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5f2adea2aea10fb1afb86c552d4c6f6f29421f68df10a99e0fc200239356631
3
+ size 8588
train_config.json ADDED
@@ -0,0 +1,244 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "/home/lwh/code/lerobot/dataset/1arm_2cam_banana_pick_v1_20260407",
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
+ "streaming": false
81
+ },
82
+ "env": null,
83
+ "policy": {
84
+ "type": "pi05",
85
+ "n_obs_steps": 1,
86
+ "input_features": {
87
+ "observation.state": {
88
+ "type": "STATE",
89
+ "shape": [
90
+ 6
91
+ ]
92
+ },
93
+ "observation.current": {
94
+ "type": "STATE",
95
+ "shape": [
96
+ 6
97
+ ]
98
+ },
99
+ "observation.images.top": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ },
107
+ "observation.images.wrist": {
108
+ "type": "VISUAL",
109
+ "shape": [
110
+ 3,
111
+ 480,
112
+ 640
113
+ ]
114
+ }
115
+ },
116
+ "output_features": {
117
+ "action": {
118
+ "type": "ACTION",
119
+ "shape": [
120
+ 6
121
+ ]
122
+ }
123
+ },
124
+ "device": "cuda",
125
+ "use_amp": false,
126
+ "use_peft": false,
127
+ "push_to_hub": true,
128
+ "repo_id": "1arm_2cam_banana_pick_v1_20260407",
129
+ "private": null,
130
+ "tags": null,
131
+ "license": null,
132
+ "pretrained_path": "lerobot/pi05_base",
133
+ "paligemma_variant": "gemma_2b",
134
+ "action_expert_variant": "gemma_300m",
135
+ "dtype": "bfloat16",
136
+ "chunk_size": 50,
137
+ "n_action_steps": 50,
138
+ "max_state_dim": 32,
139
+ "max_action_dim": 32,
140
+ "num_inference_steps": 10,
141
+ "time_sampling_beta_alpha": 1.5,
142
+ "time_sampling_beta_beta": 1.0,
143
+ "time_sampling_scale": 0.999,
144
+ "time_sampling_offset": 0.001,
145
+ "min_period": 0.004,
146
+ "max_period": 4.0,
147
+ "use_relative_actions": false,
148
+ "relative_exclude_joints": [
149
+ "gripper"
150
+ ],
151
+ "action_feature_names": [
152
+ "shoulder_pan.pos",
153
+ "shoulder_lift.pos",
154
+ "elbow_flex.pos",
155
+ "wrist_flex.pos",
156
+ "wrist_roll.pos",
157
+ "gripper.pos"
158
+ ],
159
+ "rtc_config": null,
160
+ "image_resolution": [
161
+ 224,
162
+ 224
163
+ ],
164
+ "empty_cameras": 0,
165
+ "tokenizer_max_length": 200,
166
+ "normalization_mapping": {
167
+ "VISUAL": "IDENTITY",
168
+ "STATE": "QUANTILES",
169
+ "ACTION": "QUANTILES"
170
+ },
171
+ "gradient_checkpointing": true,
172
+ "compile_model": false,
173
+ "compile_mode": "max-autotune",
174
+ "freeze_vision_encoder": false,
175
+ "train_expert_only": false,
176
+ "optimizer_lr": 2.5e-05,
177
+ "optimizer_betas": [
178
+ 0.9,
179
+ 0.95
180
+ ],
181
+ "optimizer_eps": 1e-08,
182
+ "optimizer_weight_decay": 0.01,
183
+ "optimizer_grad_clip_norm": 1.0,
184
+ "scheduler_warmup_steps": 1000,
185
+ "scheduler_decay_steps": 30000,
186
+ "scheduler_decay_lr": 2.5e-06
187
+ },
188
+ "output_dir": "outputs/1arm_2cam_banana_pick_v1_20260407_pi05",
189
+ "job_name": "pi05_training",
190
+ "resume": false,
191
+ "seed": 1000,
192
+ "cudnn_deterministic": false,
193
+ "num_workers": 4,
194
+ "batch_size": 8,
195
+ "steps": 100000,
196
+ "eval_freq": 20000,
197
+ "log_freq": 200,
198
+ "tolerance_s": 0.0001,
199
+ "save_checkpoint": true,
200
+ "save_freq": 5000,
201
+ "use_policy_training_preset": true,
202
+ "optimizer": {
203
+ "type": "adamw",
204
+ "lr": 2.5e-05,
205
+ "weight_decay": 0.01,
206
+ "grad_clip_norm": 1.0,
207
+ "betas": [
208
+ 0.9,
209
+ 0.95
210
+ ],
211
+ "eps": 1e-08
212
+ },
213
+ "scheduler": {
214
+ "type": "cosine_decay_with_warmup",
215
+ "num_warmup_steps": 1000,
216
+ "num_decay_steps": 30000,
217
+ "peak_lr": 2.5e-05,
218
+ "decay_lr": 2.5e-06
219
+ },
220
+ "eval": {
221
+ "n_episodes": 50,
222
+ "batch_size": 22,
223
+ "use_async_envs": true
224
+ },
225
+ "wandb": {
226
+ "enable": false,
227
+ "disable_artifact": false,
228
+ "project": "lerobot",
229
+ "entity": null,
230
+ "notes": null,
231
+ "run_id": null,
232
+ "mode": null,
233
+ "add_tags": true
234
+ },
235
+ "tensorboard_dir": null,
236
+ "peft": null,
237
+ "use_rabc": false,
238
+ "rabc_progress_path": null,
239
+ "rabc_kappa": 0.01,
240
+ "rabc_epsilon": 1e-06,
241
+ "rabc_head_mode": "sparse",
242
+ "rename_map": {},
243
+ "checkpoint_path": null
244
+ }