myx160 commited on
Commit
e4c80ab
·
verified ·
1 Parent(s): 768fe6c

Upload latest LeRobot model: pi05_g1d_stack_block_ActionExpert checkpoint 010000

Browse files
config.json ADDED
@@ -0,0 +1,125 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "pi05",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 16
9
+ ]
10
+ },
11
+ "observation.images.cam_left_high": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 480,
16
+ 640
17
+ ]
18
+ },
19
+ "observation.images.cam_right_high": {
20
+ "type": "VISUAL",
21
+ "shape": [
22
+ 3,
23
+ 480,
24
+ 640
25
+ ]
26
+ },
27
+ "observation.images.cam_left_wrist": {
28
+ "type": "VISUAL",
29
+ "shape": [
30
+ 3,
31
+ 480,
32
+ 640
33
+ ]
34
+ },
35
+ "observation.images.cam_right_wrist": {
36
+ "type": "VISUAL",
37
+ "shape": [
38
+ 3,
39
+ 480,
40
+ 640
41
+ ]
42
+ }
43
+ },
44
+ "output_features": {
45
+ "action": {
46
+ "type": "ACTION",
47
+ "shape": [
48
+ 16
49
+ ]
50
+ }
51
+ },
52
+ "device": "cuda",
53
+ "use_amp": false,
54
+ "use_peft": false,
55
+ "push_to_hub": false,
56
+ "repo_id": null,
57
+ "private": null,
58
+ "tags": null,
59
+ "license": null,
60
+ "pretrained_path": null,
61
+ "paligemma_variant": "gemma_2b",
62
+ "action_expert_variant": "gemma_300m",
63
+ "dtype": "bfloat16",
64
+ "chunk_size": 10,
65
+ "n_action_steps": 10,
66
+ "max_state_dim": 32,
67
+ "max_action_dim": 32,
68
+ "num_inference_steps": 10,
69
+ "time_sampling_beta_alpha": 1.5,
70
+ "time_sampling_beta_beta": 1.0,
71
+ "time_sampling_scale": 0.999,
72
+ "time_sampling_offset": 0.001,
73
+ "min_period": 0.004,
74
+ "max_period": 4.0,
75
+ "use_relative_actions": true,
76
+ "relative_exclude_joints": [
77
+ "gripper"
78
+ ],
79
+ "action_feature_names": [
80
+ "kLeftShoulderPitch",
81
+ "kLeftShoulderRoll",
82
+ "kLeftShoulderYaw",
83
+ "kLeftElbow",
84
+ "kLeftWristRoll",
85
+ "kLeftWristPitch",
86
+ "kLeftWristYaw",
87
+ "kRightShoulderPitch",
88
+ "kRightShoulderRoll",
89
+ "kRightShoulderYaw",
90
+ "kRightElbow",
91
+ "kRightWristRoll",
92
+ "kRightWristPitch",
93
+ "kRightWristYaw",
94
+ "kLeftGripper",
95
+ "kRightGripper"
96
+ ],
97
+ "rtc_config": null,
98
+ "image_resolution": [
99
+ 224,
100
+ 224
101
+ ],
102
+ "empty_cameras": 0,
103
+ "tokenizer_max_length": 200,
104
+ "normalization_mapping": {
105
+ "VISUAL": "IDENTITY",
106
+ "STATE": "QUANTILES",
107
+ "ACTION": "QUANTILES"
108
+ },
109
+ "gradient_checkpointing": false,
110
+ "compile_model": false,
111
+ "compile_mode": "max-autotune",
112
+ "freeze_vision_encoder": false,
113
+ "train_expert_only": true,
114
+ "optimizer_lr": 2.5e-05,
115
+ "optimizer_betas": [
116
+ 0.9,
117
+ 0.95
118
+ ],
119
+ "optimizer_eps": 1e-08,
120
+ "optimizer_weight_decay": 0.01,
121
+ "optimizer_grad_clip_norm": 1.0,
122
+ "scheduler_warmup_steps": 1000,
123
+ "scheduler_decay_steps": 30000,
124
+ "scheduler_decay_lr": 2.5e-06
125
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abd64b7339172e50389d86c074664fc769d0667cecca98d7cf64fee8622f7355
3
+ size 9354050752
policy_postprocessor.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "IDENTITY",
18
+ "STATE": "QUANTILES",
19
+ "ACTION": "QUANTILES"
20
+ }
21
+ },
22
+ "state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
23
+ },
24
+ {
25
+ "registry_name": "absolute_actions_processor",
26
+ "config": {
27
+ "enabled": true
28
+ }
29
+ },
30
+ {
31
+ "registry_name": "device_processor",
32
+ "config": {
33
+ "device": "cpu",
34
+ "float_dtype": null
35
+ }
36
+ }
37
+ ]
38
+ }
policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb77cf840b8a740c6a00413a6c8fdfa2161a1d8a005933103331d5c083ae3837
3
+ size 6652
policy_preprocessor.json ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "relative_actions_processor",
16
+ "config": {
17
+ "enabled": true,
18
+ "exclude_joints": [
19
+ "gripper"
20
+ ],
21
+ "action_names": [
22
+ "kLeftShoulderPitch",
23
+ "kLeftShoulderRoll",
24
+ "kLeftShoulderYaw",
25
+ "kLeftElbow",
26
+ "kLeftWristRoll",
27
+ "kLeftWristPitch",
28
+ "kLeftWristYaw",
29
+ "kRightShoulderPitch",
30
+ "kRightShoulderRoll",
31
+ "kRightShoulderYaw",
32
+ "kRightElbow",
33
+ "kRightWristRoll",
34
+ "kRightWristPitch",
35
+ "kRightWristYaw",
36
+ "kLeftGripper",
37
+ "kRightGripper"
38
+ ]
39
+ }
40
+ },
41
+ {
42
+ "registry_name": "normalizer_processor",
43
+ "config": {
44
+ "eps": 1e-08,
45
+ "features": {
46
+ "observation.state": {
47
+ "type": "STATE",
48
+ "shape": [
49
+ 16
50
+ ]
51
+ },
52
+ "observation.images.cam_left_high": {
53
+ "type": "VISUAL",
54
+ "shape": [
55
+ 3,
56
+ 480,
57
+ 640
58
+ ]
59
+ },
60
+ "observation.images.cam_right_high": {
61
+ "type": "VISUAL",
62
+ "shape": [
63
+ 3,
64
+ 480,
65
+ 640
66
+ ]
67
+ },
68
+ "observation.images.cam_left_wrist": {
69
+ "type": "VISUAL",
70
+ "shape": [
71
+ 3,
72
+ 480,
73
+ 640
74
+ ]
75
+ },
76
+ "observation.images.cam_right_wrist": {
77
+ "type": "VISUAL",
78
+ "shape": [
79
+ 3,
80
+ 480,
81
+ 640
82
+ ]
83
+ },
84
+ "action": {
85
+ "type": "ACTION",
86
+ "shape": [
87
+ 16
88
+ ]
89
+ }
90
+ },
91
+ "norm_map": {
92
+ "VISUAL": "IDENTITY",
93
+ "STATE": "QUANTILES",
94
+ "ACTION": "QUANTILES"
95
+ }
96
+ },
97
+ "state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
98
+ },
99
+ {
100
+ "registry_name": "pi05_prepare_state_tokenizer_processor_step",
101
+ "config": {}
102
+ },
103
+ {
104
+ "registry_name": "tokenizer_processor",
105
+ "config": {
106
+ "max_length": 200,
107
+ "task_key": "task",
108
+ "padding_side": "right",
109
+ "padding": "max_length",
110
+ "truncation": true,
111
+ "tokenizer_name": "google/paligemma-3b-pt-224"
112
+ }
113
+ },
114
+ {
115
+ "registry_name": "device_processor",
116
+ "config": {
117
+ "device": "cuda",
118
+ "float_dtype": null
119
+ }
120
+ }
121
+ ]
122
+ }
policy_preprocessor_step_3_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:686d8cd6d2af4a5b249f385956233d848d3abc7dbf0deb6ef42a23fa281b1cf9
3
+ size 6668
train_config.json ADDED
@@ -0,0 +1,263 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "local/G1_Dex1_Stack_Block_ActionExpert",
4
+ "root": "/mnt/cpfs/myx/unitree_data/lerobot_ppu_outputs/G1_Dex1_Stack_Block_regrouped",
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": "pyav",
80
+ "return_uint8": false,
81
+ "streaming": false
82
+ },
83
+ "env": null,
84
+ "policy": {
85
+ "type": "pi05",
86
+ "n_obs_steps": 1,
87
+ "input_features": {
88
+ "observation.state": {
89
+ "type": "STATE",
90
+ "shape": [
91
+ 16
92
+ ]
93
+ },
94
+ "observation.images.cam_left_high": {
95
+ "type": "VISUAL",
96
+ "shape": [
97
+ 3,
98
+ 480,
99
+ 640
100
+ ]
101
+ },
102
+ "observation.images.cam_right_high": {
103
+ "type": "VISUAL",
104
+ "shape": [
105
+ 3,
106
+ 480,
107
+ 640
108
+ ]
109
+ },
110
+ "observation.images.cam_left_wrist": {
111
+ "type": "VISUAL",
112
+ "shape": [
113
+ 3,
114
+ 480,
115
+ 640
116
+ ]
117
+ },
118
+ "observation.images.cam_right_wrist": {
119
+ "type": "VISUAL",
120
+ "shape": [
121
+ 3,
122
+ 480,
123
+ 640
124
+ ]
125
+ }
126
+ },
127
+ "output_features": {
128
+ "action": {
129
+ "type": "ACTION",
130
+ "shape": [
131
+ 16
132
+ ]
133
+ }
134
+ },
135
+ "device": "cuda",
136
+ "use_amp": false,
137
+ "use_peft": false,
138
+ "push_to_hub": false,
139
+ "repo_id": null,
140
+ "private": null,
141
+ "tags": null,
142
+ "license": null,
143
+ "pretrained_path": null,
144
+ "paligemma_variant": "gemma_2b",
145
+ "action_expert_variant": "gemma_300m",
146
+ "dtype": "bfloat16",
147
+ "chunk_size": 10,
148
+ "n_action_steps": 10,
149
+ "max_state_dim": 32,
150
+ "max_action_dim": 32,
151
+ "num_inference_steps": 10,
152
+ "time_sampling_beta_alpha": 1.5,
153
+ "time_sampling_beta_beta": 1.0,
154
+ "time_sampling_scale": 0.999,
155
+ "time_sampling_offset": 0.001,
156
+ "min_period": 0.004,
157
+ "max_period": 4.0,
158
+ "use_relative_actions": true,
159
+ "relative_exclude_joints": [
160
+ "gripper"
161
+ ],
162
+ "action_feature_names": [
163
+ "kLeftShoulderPitch",
164
+ "kLeftShoulderRoll",
165
+ "kLeftShoulderYaw",
166
+ "kLeftElbow",
167
+ "kLeftWristRoll",
168
+ "kLeftWristPitch",
169
+ "kLeftWristYaw",
170
+ "kRightShoulderPitch",
171
+ "kRightShoulderRoll",
172
+ "kRightShoulderYaw",
173
+ "kRightElbow",
174
+ "kRightWristRoll",
175
+ "kRightWristPitch",
176
+ "kRightWristYaw",
177
+ "kLeftGripper",
178
+ "kRightGripper"
179
+ ],
180
+ "rtc_config": null,
181
+ "image_resolution": [
182
+ 224,
183
+ 224
184
+ ],
185
+ "empty_cameras": 0,
186
+ "tokenizer_max_length": 200,
187
+ "normalization_mapping": {
188
+ "VISUAL": "IDENTITY",
189
+ "STATE": "QUANTILES",
190
+ "ACTION": "QUANTILES"
191
+ },
192
+ "gradient_checkpointing": false,
193
+ "compile_model": false,
194
+ "compile_mode": "max-autotune",
195
+ "freeze_vision_encoder": false,
196
+ "train_expert_only": true,
197
+ "optimizer_lr": 2.5e-05,
198
+ "optimizer_betas": [
199
+ 0.9,
200
+ 0.95
201
+ ],
202
+ "optimizer_eps": 1e-08,
203
+ "optimizer_weight_decay": 0.01,
204
+ "optimizer_grad_clip_norm": 1.0,
205
+ "scheduler_warmup_steps": 1000,
206
+ "scheduler_decay_steps": 30000,
207
+ "scheduler_decay_lr": 2.5e-06
208
+ },
209
+ "reward_model": null,
210
+ "output_dir": "/mnt/cpfs/myx/unitree_data/lerobot_ppu_outputs/train_runs/pi05_g1d_stack_block_ActionExpert",
211
+ "job_name": "pi05_g1d_stack_block",
212
+ "resume": false,
213
+ "seed": 1000,
214
+ "cudnn_deterministic": false,
215
+ "num_workers": 8,
216
+ "batch_size": 16,
217
+ "prefetch_factor": 4,
218
+ "persistent_workers": true,
219
+ "steps": 10000,
220
+ "eval_freq": 0,
221
+ "log_freq": 100,
222
+ "tolerance_s": 0.0001,
223
+ "save_checkpoint": true,
224
+ "save_freq": 5000,
225
+ "use_policy_training_preset": true,
226
+ "optimizer": {
227
+ "type": "adamw",
228
+ "lr": 2.5e-05,
229
+ "weight_decay": 0.01,
230
+ "grad_clip_norm": 1.0,
231
+ "betas": [
232
+ 0.9,
233
+ 0.95
234
+ ],
235
+ "eps": 1e-08
236
+ },
237
+ "scheduler": {
238
+ "type": "cosine_decay_with_warmup",
239
+ "num_warmup_steps": 1000,
240
+ "num_decay_steps": 30000,
241
+ "peak_lr": 2.5e-05,
242
+ "decay_lr": 2.5e-06
243
+ },
244
+ "eval": {
245
+ "n_episodes": 50,
246
+ "batch_size": 50,
247
+ "use_async_envs": true
248
+ },
249
+ "wandb": {
250
+ "enable": true,
251
+ "disable_artifact": false,
252
+ "project": "g1d_lerobot",
253
+ "entity": null,
254
+ "notes": null,
255
+ "run_id": "dgmypkwq",
256
+ "mode": "offline",
257
+ "add_tags": true
258
+ },
259
+ "peft": null,
260
+ "sample_weighting": null,
261
+ "rename_map": {},
262
+ "checkpoint_path": null
263
+ }