Hugging Face's logo

drashutoshspace
/
moonbot_pi0_three_blocks_stack

Robotics
LeRobot
Safetensors
pi0
moonbot
Model card Files
xet
 Community
drashutoshspace commited on
Commit
c87eb8f
·
verified ·
1 Parent(s): a3db308

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. checkpoint-010000/config.json +110 -0
  2. checkpoint-010000/model.safetensors +3 -0
  3. checkpoint-010000/policy_postprocessor.json +38 -0
  4. checkpoint-010000/policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
  5. checkpoint-010000/policy_preprocessor.json +97 -0
  6. checkpoint-010000/policy_preprocessor_step_6_normalizer_processor.safetensors +3 -0
  7. checkpoint-010000/train_config.json +265 -0
checkpoint-010000/config.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "pi0",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.images.front": {
6
+ "type": "VISUAL",
7
+ "shape": [
8
+ 3,
9
+ 480,
10
+ 640
11
+ ]
12
+ },
13
+ "observation.images.eef": {
14
+ "type": "VISUAL",
15
+ "shape": [
16
+ 3,
17
+ 480,
18
+ 640
19
+ ]
20
+ },
21
+ "observation.images.right": {
22
+ "type": "VISUAL",
23
+ "shape": [
24
+ 3,
25
+ 480,
26
+ 640
27
+ ]
28
+ },
29
+ "observation.state": {
30
+ "type": "STATE",
31
+ "shape": [
32
+ 21
33
+ ]
34
+ }
35
+ },
36
+ "output_features": {
37
+ "action": {
38
+ "type": "ACTION",
39
+ "shape": [
40
+ 8
41
+ ]
42
+ }
43
+ },
44
+ "device": "cuda",
45
+ "use_amp": false,
46
+ "use_peft": false,
47
+ "push_to_hub": false,
48
+ "repo_id": "lerobot/moonbot",
49
+ "private": null,
50
+ "tags": null,
51
+ "license": null,
52
+ "pretrained_path": "lerobot/pi0_base",
53
+ "pretrained_revision": null,
54
+ "paligemma_variant": "gemma_2b",
55
+ "action_expert_variant": "gemma_300m",
56
+ "dtype": "bfloat16",
57
+ "chunk_size": 50,
58
+ "n_action_steps": 50,
59
+ "max_state_dim": 32,
60
+ "max_action_dim": 32,
61
+ "num_inference_steps": 10,
62
+ "time_sampling_beta_alpha": 1.5,
63
+ "time_sampling_beta_beta": 1.0,
64
+ "time_sampling_scale": 0.999,
65
+ "time_sampling_offset": 0.001,
66
+ "min_period": 0.004,
67
+ "max_period": 4.0,
68
+ "use_relative_actions": false,
69
+ "relative_exclude_joints": [
70
+ "gripper"
71
+ ],
72
+ "action_feature_names": [
73
+ "position.leg302_joint1",
74
+ "position.leg302_joint2",
75
+ "position.leg302_joint3",
76
+ "position.leg302_joint4",
77
+ "position.leg302_joint5",
78
+ "position.leg302_joint6",
79
+ "position.leg302_joint7",
80
+ "position.leg302_grip2"
81
+ ],
82
+ "rtc_config": null,
83
+ "image_resolution": [
84
+ 224,
85
+ 224
86
+ ],
87
+ "empty_cameras": 0,
88
+ "normalization_mapping": {
89
+ "VISUAL": "IDENTITY",
90
+ "STATE": "MEAN_STD",
91
+ "ACTION": "MEAN_STD"
92
+ },
93
+ "gradient_checkpointing": true,
94
+ "compile_model": false,
95
+ "compile_mode": "max-autotune",
96
+ "freeze_vision_encoder": false,
97
+ "train_expert_only": false,
98
+ "optimizer_lr": 2.5e-05,
99
+ "optimizer_betas": [
100
+ 0.9,
101
+ 0.95
102
+ ],
103
+ "optimizer_eps": 1e-08,
104
+ "optimizer_weight_decay": 0.01,
105
+ "optimizer_grad_clip_norm": 1.0,
106
+ "scheduler_warmup_steps": 1000,
107
+ "scheduler_decay_steps": 30000,
108
+ "scheduler_decay_lr": 2.5e-06,
109
+ "tokenizer_max_length": 48
110
+ }
checkpoint-010000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23847cb941525e7ac4c7d2756a40a030133c10b2d99b02b55b3235a0906798e8
3
+ size 8892502944
checkpoint-010000/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
+ 8
13
+ ]
14
+ }
15
+ },
16
+ "norm_map": {
17
+ "VISUAL": "IDENTITY",
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": "absolute_actions_processor",
26
+ "config": {
27
+ "enabled": false
28
+ }
29
+ },
30
+ {
31
+ "registry_name": "device_processor",
32
+ "config": {
33
+ "device": "cpu",
34
+ "float_dtype": null
35
+ }
36
+ }
37
+ ]
38
+ }
checkpoint-010000/policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:054698bd33960e5455ea9303973141f654e13573fea4bd44bf4eab0ab9c6b137
3
+ size 11388
checkpoint-010000/policy_preprocessor.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "pi0_new_line_processor",
16
+ "config": {}
17
+ },
18
+ {
19
+ "registry_name": "tokenizer_processor",
20
+ "config": {
21
+ "max_length": 48,
22
+ "task_key": "task",
23
+ "padding_side": "right",
24
+ "padding": "max_length",
25
+ "truncation": true,
26
+ "tokenizer_name": "google/paligemma-3b-pt-224"
27
+ }
28
+ },
29
+ {
30
+ "registry_name": "device_processor",
31
+ "config": {
32
+ "device": "cuda",
33
+ "float_dtype": null
34
+ }
35
+ },
36
+ {
37
+ "registry_name": "relative_actions_processor",
38
+ "config": {
39
+ "enabled": false,
40
+ "exclude_joints": [
41
+ "gripper"
42
+ ],
43
+ "action_names": null
44
+ }
45
+ },
46
+ {
47
+ "registry_name": "normalizer_processor",
48
+ "config": {
49
+ "eps": 1e-08,
50
+ "features": {
51
+ "observation.images.front": {
52
+ "type": "VISUAL",
53
+ "shape": [
54
+ 3,
55
+ 480,
56
+ 640
57
+ ]
58
+ },
59
+ "observation.images.eef": {
60
+ "type": "VISUAL",
61
+ "shape": [
62
+ 3,
63
+ 480,
64
+ 640
65
+ ]
66
+ },
67
+ "observation.images.right": {
68
+ "type": "VISUAL",
69
+ "shape": [
70
+ 3,
71
+ 480,
72
+ 640
73
+ ]
74
+ },
75
+ "observation.state": {
76
+ "type": "STATE",
77
+ "shape": [
78
+ 21
79
+ ]
80
+ },
81
+ "action": {
82
+ "type": "ACTION",
83
+ "shape": [
84
+ 8
85
+ ]
86
+ }
87
+ },
88
+ "norm_map": {
89
+ "VISUAL": "IDENTITY",
90
+ "STATE": "MEAN_STD",
91
+ "ACTION": "MEAN_STD"
92
+ }
93
+ },
94
+ "state_file": "policy_preprocessor_step_6_normalizer_processor.safetensors"
95
+ }
96
+ ]
97
+ }
checkpoint-010000/policy_preprocessor_step_6_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8867128cc9f256c500fccd095fb037acd89c840bab9b1b44e7070a95dfb196b9
3
+ size 11404
checkpoint-010000/train_config.json ADDED
@@ -0,0 +1,265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "gdiazsrl/three_blocks_stack_sep22",
4
+ "repo_type": "dataset",
5
+ "root": "/home/amishra1-iit.local/work/vlm/datasets/three_blocks_stack_sep22",
6
+ "episodes": null,
7
+ "image_transforms": {
8
+ "enable": false,
9
+ "max_num_transforms": 3,
10
+ "random_order": false,
11
+ "tfs": {
12
+ "brightness": {
13
+ "weight": 1.0,
14
+ "type": "ColorJitter",
15
+ "kwargs": {
16
+ "brightness": [
17
+ 0.8,
18
+ 1.2
19
+ ]
20
+ }
21
+ },
22
+ "contrast": {
23
+ "weight": 1.0,
24
+ "type": "ColorJitter",
25
+ "kwargs": {
26
+ "contrast": [
27
+ 0.8,
28
+ 1.2
29
+ ]
30
+ }
31
+ },
32
+ "saturation": {
33
+ "weight": 1.0,
34
+ "type": "ColorJitter",
35
+ "kwargs": {
36
+ "saturation": [
37
+ 0.5,
38
+ 1.5
39
+ ]
40
+ }
41
+ },
42
+ "hue": {
43
+ "weight": 1.0,
44
+ "type": "ColorJitter",
45
+ "kwargs": {
46
+ "hue": [
47
+ -0.05,
48
+ 0.05
49
+ ]
50
+ }
51
+ },
52
+ "sharpness": {
53
+ "weight": 1.0,
54
+ "type": "SharpnessJitter",
55
+ "kwargs": {
56
+ "sharpness": [
57
+ 0.5,
58
+ 1.5
59
+ ]
60
+ }
61
+ },
62
+ "affine": {
63
+ "weight": 1.0,
64
+ "type": "RandomAffine",
65
+ "kwargs": {
66
+ "degrees": [
67
+ -5.0,
68
+ 5.0
69
+ ],
70
+ "translate": [
71
+ 0.05,
72
+ 0.05
73
+ ]
74
+ }
75
+ }
76
+ }
77
+ },
78
+ "revision": null,
79
+ "use_imagenet_stats": true,
80
+ "video_backend": "pyav",
81
+ "return_uint8": false,
82
+ "depth_output_unit": "mm",
83
+ "streaming": false,
84
+ "eval_split": 0.0
85
+ },
86
+ "env": null,
87
+ "policy": {
88
+ "type": "pi0",
89
+ "n_obs_steps": 1,
90
+ "input_features": {
91
+ "observation.images.front": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.eef": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
105
+ ]
106
+ },
107
+ "observation.images.right": {
108
+ "type": "VISUAL",
109
+ "shape": [
110
+ 3,
111
+ 480,
112
+ 640
113
+ ]
114
+ },
115
+ "observation.state": {
116
+ "type": "STATE",
117
+ "shape": [
118
+ 21
119
+ ]
120
+ }
121
+ },
122
+ "output_features": {
123
+ "action": {
124
+ "type": "ACTION",
125
+ "shape": [
126
+ 8
127
+ ]
128
+ }
129
+ },
130
+ "device": "cuda",
131
+ "use_amp": false,
132
+ "use_peft": false,
133
+ "push_to_hub": false,
134
+ "repo_id": "lerobot/moonbot",
135
+ "private": null,
136
+ "tags": null,
137
+ "license": null,
138
+ "pretrained_path": "lerobot/pi0_base",
139
+ "pretrained_revision": null,
140
+ "paligemma_variant": "gemma_2b",
141
+ "action_expert_variant": "gemma_300m",
142
+ "dtype": "bfloat16",
143
+ "chunk_size": 50,
144
+ "n_action_steps": 50,
145
+ "max_state_dim": 32,
146
+ "max_action_dim": 32,
147
+ "num_inference_steps": 10,
148
+ "time_sampling_beta_alpha": 1.5,
149
+ "time_sampling_beta_beta": 1.0,
150
+ "time_sampling_scale": 0.999,
151
+ "time_sampling_offset": 0.001,
152
+ "min_period": 0.004,
153
+ "max_period": 4.0,
154
+ "use_relative_actions": false,
155
+ "relative_exclude_joints": [
156
+ "gripper"
157
+ ],
158
+ "action_feature_names": [
159
+ "position.leg302_joint1",
160
+ "position.leg302_joint2",
161
+ "position.leg302_joint3",
162
+ "position.leg302_joint4",
163
+ "position.leg302_joint5",
164
+ "position.leg302_joint6",
165
+ "position.leg302_joint7",
166
+ "position.leg302_grip2"
167
+ ],
168
+ "rtc_config": null,
169
+ "image_resolution": [
170
+ 224,
171
+ 224
172
+ ],
173
+ "empty_cameras": 0,
174
+ "normalization_mapping": {
175
+ "VISUAL": "IDENTITY",
176
+ "STATE": "MEAN_STD",
177
+ "ACTION": "MEAN_STD"
178
+ },
179
+ "gradient_checkpointing": true,
180
+ "compile_model": false,
181
+ "compile_mode": "max-autotune",
182
+ "freeze_vision_encoder": false,
183
+ "train_expert_only": false,
184
+ "optimizer_lr": 2.5e-05,
185
+ "optimizer_betas": [
186
+ 0.9,
187
+ 0.95
188
+ ],
189
+ "optimizer_eps": 1e-08,
190
+ "optimizer_weight_decay": 0.01,
191
+ "optimizer_grad_clip_norm": 1.0,
192
+ "scheduler_warmup_steps": 1000,
193
+ "scheduler_decay_steps": 30000,
194
+ "scheduler_decay_lr": 2.5e-06,
195
+ "tokenizer_max_length": 48
196
+ },
197
+ "reward_model": null,
198
+ "output_dir": "/home/amishra1-iit.local/work/vlm/train/three_blocks_stack_v1_pi0",
199
+ "job_name": "pi0_three_blocks_stack",
200
+ "resume": false,
201
+ "seed": 1000,
202
+ "cudnn_deterministic": false,
203
+ "num_workers": 2,
204
+ "batch_size": 16,
205
+ "prefetch_factor": 4,
206
+ "persistent_workers": true,
207
+ "dataloader_multiprocessing_context": "spawn",
208
+ "steps": 20000,
209
+ "env_eval_freq": 20000,
210
+ "log_freq": 200,
211
+ "eval_steps": 0,
212
+ "max_eval_samples": 0,
213
+ "tolerance_s": 0.0001,
214
+ "save_checkpoint": true,
215
+ "save_freq": 5000,
216
+ "use_policy_training_preset": true,
217
+ "optimizer": {
218
+ "type": "adamw",
219
+ "lr": 2.5e-05,
220
+ "weight_decay": 0.01,
221
+ "grad_clip_norm": 1.0,
222
+ "betas": [
223
+ 0.9,
224
+ 0.95
225
+ ],
226
+ "eps": 1e-08
227
+ },
228
+ "scheduler": {
229
+ "type": "cosine_decay_with_warmup",
230
+ "num_warmup_steps": 1000,
231
+ "num_decay_steps": 30000,
232
+ "peak_lr": 2.5e-05,
233
+ "decay_lr": 2.5e-06
234
+ },
235
+ "eval": {
236
+ "n_episodes": 50,
237
+ "batch_size": 33,
238
+ "use_async_envs": true,
239
+ "recording": false,
240
+ "recording_repo_id": null,
241
+ "recording_private": false
242
+ },
243
+ "wandb": {
244
+ "enable": true,
245
+ "disable_artifact": false,
246
+ "project": "lerobot",
247
+ "entity": null,
248
+ "notes": null,
249
+ "run_id": "bkbszs91",
250
+ "mode": null,
251
+ "add_tags": true
252
+ },
253
+ "peft": null,
254
+ "job": {
255
+ "target": null,
256
+ "image": "huggingface/lerobot-gpu:latest",
257
+ "timeout": "2d",
258
+ "detach": false,
259
+ "tags": []
260
+ },
261
+ "save_checkpoint_to_hub": false,
262
+ "sample_weighting": null,
263
+ "rename_map": {},
264
+ "checkpoint_path": null
265
+ }