Robotics
LeRobot
Safetensors
smolvla
tinkhireeva commited on
Commit
329ee82
·
verified ·
1 Parent(s): 6af59e5

Upload folder using huggingface_hub

Browse files
config.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "smolvla",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 6
9
+ ]
10
+ },
11
+ "observation.images.bottom": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 480,
16
+ 640
17
+ ]
18
+ },
19
+ "observation.images.up": {
20
+ "type": "VISUAL",
21
+ "shape": [
22
+ 3,
23
+ 480,
24
+ 640
25
+ ]
26
+ },
27
+ "observation.images.wrist": {
28
+ "type": "VISUAL",
29
+ "shape": [
30
+ 3,
31
+ 480,
32
+ 640
33
+ ]
34
+ }
35
+ },
36
+ "output_features": {
37
+ "action": {
38
+ "type": "ACTION",
39
+ "shape": [
40
+ 6
41
+ ]
42
+ }
43
+ },
44
+ "device": "cuda",
45
+ "use_amp": false,
46
+ "push_to_hub": true,
47
+ "repo_id": "tinkhireeva/so101_sort_yellow_objects_smolvla",
48
+ "private": null,
49
+ "tags": null,
50
+ "license": null,
51
+ "pretrained_path": "lerobot/smolvla_base",
52
+ "chunk_size": 50,
53
+ "n_action_steps": 50,
54
+ "normalization_mapping": {
55
+ "VISUAL": "IDENTITY",
56
+ "STATE": "MEAN_STD",
57
+ "ACTION": "MEAN_STD"
58
+ },
59
+ "max_state_dim": 32,
60
+ "max_action_dim": 32,
61
+ "resize_imgs_with_padding": [
62
+ 512,
63
+ 512
64
+ ],
65
+ "empty_cameras": 0,
66
+ "adapt_to_pi_aloha": false,
67
+ "use_delta_joint_actions_aloha": false,
68
+ "tokenizer_max_length": 48,
69
+ "num_steps": 10,
70
+ "use_cache": true,
71
+ "freeze_vision_encoder": true,
72
+ "train_expert_only": true,
73
+ "train_state_proj": true,
74
+ "optimizer_lr": 0.0001,
75
+ "optimizer_betas": [
76
+ 0.9,
77
+ 0.95
78
+ ],
79
+ "optimizer_eps": 1e-08,
80
+ "optimizer_weight_decay": 1e-10,
81
+ "optimizer_grad_clip_norm": 10.0,
82
+ "scheduler_warmup_steps": 1000,
83
+ "scheduler_decay_steps": 30000,
84
+ "scheduler_decay_lr": 2.5e-06,
85
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
86
+ "load_vlm_weights": true,
87
+ "add_image_special_tokens": false,
88
+ "attention_mode": "cross_attn",
89
+ "prefix_length": 0,
90
+ "pad_language_to": "max_length",
91
+ "num_expert_layers": 0,
92
+ "num_vlm_layers": 16,
93
+ "self_attn_every_n_layers": 2,
94
+ "expert_width_multiplier": 0.75,
95
+ "min_period": 0.004,
96
+ "max_period": 4.0
97
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40a6397af57b60e55a9204ddb581a4320974a50afa9795ab8082b4a9fd8ee4a8
3
+ size 906712520
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": "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:a7a14f7415c2f3f2f19379eb0535f1c1b4c31c409c42e989f3f178e38d8765c4
3
+ size 8528
policy_preprocessor.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "smolvla_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": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct"
27
+ }
28
+ },
29
+ {
30
+ "registry_name": "device_processor",
31
+ "config": {
32
+ "device": "cuda",
33
+ "float_dtype": null
34
+ }
35
+ },
36
+ {
37
+ "registry_name": "normalizer_processor",
38
+ "config": {
39
+ "eps": 1e-08,
40
+ "features": {
41
+ "observation.state": {
42
+ "type": "STATE",
43
+ "shape": [
44
+ 6
45
+ ]
46
+ },
47
+ "observation.images.bottom": {
48
+ "type": "VISUAL",
49
+ "shape": [
50
+ 3,
51
+ 480,
52
+ 640
53
+ ]
54
+ },
55
+ "observation.images.up": {
56
+ "type": "VISUAL",
57
+ "shape": [
58
+ 3,
59
+ 480,
60
+ 640
61
+ ]
62
+ },
63
+ "observation.images.wrist": {
64
+ "type": "VISUAL",
65
+ "shape": [
66
+ 3,
67
+ 480,
68
+ 640
69
+ ]
70
+ },
71
+ "action": {
72
+ "type": "ACTION",
73
+ "shape": [
74
+ 6
75
+ ]
76
+ }
77
+ },
78
+ "norm_map": {
79
+ "VISUAL": "IDENTITY",
80
+ "STATE": "MEAN_STD",
81
+ "ACTION": "MEAN_STD"
82
+ }
83
+ },
84
+ "state_file": "policy_preprocessor_step_5_normalizer_processor.safetensors"
85
+ }
86
+ ]
87
+ }
policy_preprocessor_step_5_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a7a14f7415c2f3f2f19379eb0535f1c1b4c31c409c42e989f3f178e38d8765c4
3
+ size 8528
train_config.json ADDED
@@ -0,0 +1,210 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "tinkhireeva/so101_sort_yellow_objects",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec",
66
+ "streaming": false
67
+ },
68
+ "env": null,
69
+ "policy": {
70
+ "type": "smolvla",
71
+ "n_obs_steps": 1,
72
+ "input_features": {
73
+ "observation.state": {
74
+ "type": "STATE",
75
+ "shape": [
76
+ 6
77
+ ]
78
+ },
79
+ "observation.images.bottom": {
80
+ "type": "VISUAL",
81
+ "shape": [
82
+ 3,
83
+ 480,
84
+ 640
85
+ ]
86
+ },
87
+ "observation.images.up": {
88
+ "type": "VISUAL",
89
+ "shape": [
90
+ 3,
91
+ 480,
92
+ 640
93
+ ]
94
+ },
95
+ "observation.images.wrist": {
96
+ "type": "VISUAL",
97
+ "shape": [
98
+ 3,
99
+ 480,
100
+ 640
101
+ ]
102
+ }
103
+ },
104
+ "output_features": {
105
+ "action": {
106
+ "type": "ACTION",
107
+ "shape": [
108
+ 6
109
+ ]
110
+ }
111
+ },
112
+ "device": "cuda",
113
+ "use_amp": false,
114
+ "push_to_hub": true,
115
+ "repo_id": "tinkhireeva/so101_sort_yellow_objects_smolvla",
116
+ "private": null,
117
+ "tags": null,
118
+ "license": null,
119
+ "pretrained_path": "lerobot/smolvla_base",
120
+ "chunk_size": 50,
121
+ "n_action_steps": 50,
122
+ "normalization_mapping": {
123
+ "VISUAL": "IDENTITY",
124
+ "STATE": "MEAN_STD",
125
+ "ACTION": "MEAN_STD"
126
+ },
127
+ "max_state_dim": 32,
128
+ "max_action_dim": 32,
129
+ "resize_imgs_with_padding": [
130
+ 512,
131
+ 512
132
+ ],
133
+ "empty_cameras": 0,
134
+ "adapt_to_pi_aloha": false,
135
+ "use_delta_joint_actions_aloha": false,
136
+ "tokenizer_max_length": 48,
137
+ "num_steps": 10,
138
+ "use_cache": true,
139
+ "freeze_vision_encoder": true,
140
+ "train_expert_only": true,
141
+ "train_state_proj": true,
142
+ "optimizer_lr": 0.0001,
143
+ "optimizer_betas": [
144
+ 0.9,
145
+ 0.95
146
+ ],
147
+ "optimizer_eps": 1e-08,
148
+ "optimizer_weight_decay": 1e-10,
149
+ "optimizer_grad_clip_norm": 10.0,
150
+ "scheduler_warmup_steps": 1000,
151
+ "scheduler_decay_steps": 30000,
152
+ "scheduler_decay_lr": 2.5e-06,
153
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
154
+ "load_vlm_weights": true,
155
+ "add_image_special_tokens": false,
156
+ "attention_mode": "cross_attn",
157
+ "prefix_length": 0,
158
+ "pad_language_to": "max_length",
159
+ "num_expert_layers": 0,
160
+ "num_vlm_layers": 16,
161
+ "self_attn_every_n_layers": 2,
162
+ "expert_width_multiplier": 0.75,
163
+ "min_period": 0.004,
164
+ "max_period": 4.0
165
+ },
166
+ "output_dir": "outputs/train/so101_sort_yellow_objects_smolvla",
167
+ "job_name": "so101_sort_yellow_objects_smolvla",
168
+ "resume": false,
169
+ "seed": 1000,
170
+ "num_workers": 4,
171
+ "batch_size": 16,
172
+ "steps": 300000,
173
+ "eval_freq": 20000,
174
+ "log_freq": 200,
175
+ "save_checkpoint": true,
176
+ "save_freq": 20000,
177
+ "use_policy_training_preset": true,
178
+ "optimizer": {
179
+ "type": "adamw",
180
+ "lr": 0.0001,
181
+ "weight_decay": 1e-10,
182
+ "grad_clip_norm": 10.0,
183
+ "betas": [
184
+ 0.9,
185
+ 0.95
186
+ ],
187
+ "eps": 1e-08
188
+ },
189
+ "scheduler": {
190
+ "type": "cosine_decay_with_warmup",
191
+ "num_warmup_steps": 1000,
192
+ "num_decay_steps": 30000,
193
+ "peak_lr": 0.0001,
194
+ "decay_lr": 2.5e-06
195
+ },
196
+ "eval": {
197
+ "n_episodes": 50,
198
+ "batch_size": 50,
199
+ "use_async_envs": false
200
+ },
201
+ "wandb": {
202
+ "enable": false,
203
+ "disable_artifact": false,
204
+ "project": "lerobot",
205
+ "entity": null,
206
+ "notes": null,
207
+ "run_id": null,
208
+ "mode": null
209
+ }
210
+ }