Hugging Face's logo

robot-learning-group47
/
eval1_final_final

Safetensors
Model card Files
xet
 Community
nkim7 commited on
Commit
afb967d
·
verified ·
1 Parent(s): 6c50c9c

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. config.json +117 -0
  2. model.safetensors +3 -0
  3. policy_postprocessor.json +32 -0
  4. policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
  5. policy_preprocessor.json +105 -0
  6. policy_preprocessor_step_5_normalizer_processor.safetensors +3 -0
  7. train_config.json +255 -0
config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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.camera1": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 256,
16
+ 256
17
+ ]
18
+ },
19
+ "observation.images.camera2": {
20
+ "type": "VISUAL",
21
+ "shape": [
22
+ 3,
23
+ 256,
24
+ 256
25
+ ]
26
+ },
27
+ "observation.images.camera3": {
28
+ "type": "VISUAL",
29
+ "shape": [
30
+ 3,
31
+ 256,
32
+ 256
33
+ ]
34
+ },
35
+ "observation.images.empty_camera_0": {
36
+ "type": "VISUAL",
37
+ "shape": [
38
+ 3,
39
+ 480,
40
+ 640
41
+ ]
42
+ },
43
+ "observation.images.empty_camera_1": {
44
+ "type": "VISUAL",
45
+ "shape": [
46
+ 3,
47
+ 480,
48
+ 640
49
+ ]
50
+ }
51
+ },
52
+ "output_features": {
53
+ "action": {
54
+ "type": "ACTION",
55
+ "shape": [
56
+ 6
57
+ ]
58
+ }
59
+ },
60
+ "device": "cuda",
61
+ "use_amp": true,
62
+ "use_peft": false,
63
+ "push_to_hub": false,
64
+ "repo_id": null,
65
+ "private": null,
66
+ "tags": null,
67
+ "license": null,
68
+ "pretrained_path": "lerobot/smolvla_base",
69
+ "chunk_size": 50,
70
+ "n_action_steps": 50,
71
+ "normalization_mapping": {
72
+ "VISUAL": "IDENTITY",
73
+ "STATE": "MEAN_STD",
74
+ "ACTION": "MEAN_STD"
75
+ },
76
+ "max_state_dim": 32,
77
+ "max_action_dim": 32,
78
+ "resize_imgs_with_padding": [
79
+ 256,
80
+ 256
81
+ ],
82
+ "empty_cameras": 2,
83
+ "adapt_to_pi_aloha": false,
84
+ "use_delta_joint_actions_aloha": false,
85
+ "tokenizer_max_length": 48,
86
+ "num_steps": 10,
87
+ "use_cache": true,
88
+ "freeze_vision_encoder": true,
89
+ "train_expert_only": true,
90
+ "train_state_proj": true,
91
+ "optimizer_lr": 0.0001,
92
+ "optimizer_betas": [
93
+ 0.9,
94
+ 0.95
95
+ ],
96
+ "optimizer_eps": 1e-08,
97
+ "optimizer_weight_decay": 1e-10,
98
+ "optimizer_grad_clip_norm": 10.0,
99
+ "scheduler_warmup_steps": 1000,
100
+ "scheduler_decay_steps": 30000,
101
+ "scheduler_decay_lr": 2.5e-06,
102
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
103
+ "load_vlm_weights": true,
104
+ "add_image_special_tokens": false,
105
+ "attention_mode": "cross_attn",
106
+ "prefix_length": 0,
107
+ "pad_language_to": "max_length",
108
+ "num_expert_layers": 0,
109
+ "num_vlm_layers": 16,
110
+ "self_attn_every_n_layers": 2,
111
+ "expert_width_multiplier": 0.75,
112
+ "min_period": 0.004,
113
+ "max_period": 4.0,
114
+ "rtc_config": null,
115
+ "compile_model": false,
116
+ "compile_mode": "max-autotune"
117
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3357bafd45b6c8d1edca5e71399d4c81ec45ecb6b864f7f8498d34dc16151bf
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:e1d60a70c6f111c64e5ecb021f74f5dd29b1817ab0646c0f6c4265daf7b67f90
3
+ size 6560
policy_preprocessor.json ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_preprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "rename_observations_processor",
6
+ "config": {
7
+ "rename_map": {
8
+ "observation.images.front": "observation.images.camera1"
9
+ }
10
+ }
11
+ },
12
+ {
13
+ "registry_name": "to_batch_processor",
14
+ "config": {}
15
+ },
16
+ {
17
+ "registry_name": "smolvla_new_line_processor",
18
+ "config": {}
19
+ },
20
+ {
21
+ "registry_name": "tokenizer_processor",
22
+ "config": {
23
+ "max_length": 48,
24
+ "task_key": "task",
25
+ "padding_side": "right",
26
+ "padding": "max_length",
27
+ "truncation": true,
28
+ "tokenizer_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct"
29
+ }
30
+ },
31
+ {
32
+ "registry_name": "device_processor",
33
+ "config": {
34
+ "device": "cuda",
35
+ "float_dtype": null
36
+ }
37
+ },
38
+ {
39
+ "registry_name": "normalizer_processor",
40
+ "config": {
41
+ "eps": 1e-08,
42
+ "features": {
43
+ "observation.state": {
44
+ "type": "STATE",
45
+ "shape": [
46
+ 6
47
+ ]
48
+ },
49
+ "observation.images.camera1": {
50
+ "type": "VISUAL",
51
+ "shape": [
52
+ 3,
53
+ 256,
54
+ 256
55
+ ]
56
+ },
57
+ "observation.images.camera2": {
58
+ "type": "VISUAL",
59
+ "shape": [
60
+ 3,
61
+ 256,
62
+ 256
63
+ ]
64
+ },
65
+ "observation.images.camera3": {
66
+ "type": "VISUAL",
67
+ "shape": [
68
+ 3,
69
+ 256,
70
+ 256
71
+ ]
72
+ },
73
+ "observation.images.empty_camera_0": {
74
+ "type": "VISUAL",
75
+ "shape": [
76
+ 3,
77
+ 480,
78
+ 640
79
+ ]
80
+ },
81
+ "observation.images.empty_camera_1": {
82
+ "type": "VISUAL",
83
+ "shape": [
84
+ 3,
85
+ 480,
86
+ 640
87
+ ]
88
+ },
89
+ "action": {
90
+ "type": "ACTION",
91
+ "shape": [
92
+ 6
93
+ ]
94
+ }
95
+ },
96
+ "norm_map": {
97
+ "VISUAL": "IDENTITY",
98
+ "STATE": "MEAN_STD",
99
+ "ACTION": "MEAN_STD"
100
+ }
101
+ },
102
+ "state_file": "policy_preprocessor_step_5_normalizer_processor.safetensors"
103
+ }
104
+ ]
105
+ }
policy_preprocessor_step_5_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1d60a70c6f111c64e5ecb021f74f5dd29b1817ab0646c0f6c4265daf7b67f90
3
+ size 6560
train_config.json ADDED
@@ -0,0 +1,255 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "robot-learning-group47/eval1_60_fixed_banana_fixed_bowl",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
8
+ "max_num_transforms": 2,
9
+ "random_order": false,
10
+ "tfs": {
11
+ "brightness": {
12
+ "weight": 1.0,
13
+ "type": "ColorJitter",
14
+ "kwargs": {
15
+ "brightness": [
16
+ 0.85,
17
+ 1.15
18
+ ]
19
+ }
20
+ },
21
+ "contrast": {
22
+ "weight": 1.0,
23
+ "type": "ColorJitter",
24
+ "kwargs": {
25
+ "contrast": [
26
+ 0.9,
27
+ 1.1
28
+ ]
29
+ }
30
+ },
31
+ "saturation": {
32
+ "weight": 0.0,
33
+ "type": "ColorJitter",
34
+ "kwargs": {
35
+ "saturation": [
36
+ 1.0,
37
+ 1.0
38
+ ]
39
+ }
40
+ },
41
+ "hue": {
42
+ "weight": 0.0,
43
+ "type": "ColorJitter",
44
+ "kwargs": {
45
+ "hue": [
46
+ 0.0,
47
+ 0.0
48
+ ]
49
+ }
50
+ },
51
+ "sharpness": {
52
+ "weight": 0.0,
53
+ "type": "RandomAdjustSharpness",
54
+ "kwargs": {
55
+ "sharpness": [
56
+ 1.0,
57
+ 1.0
58
+ ]
59
+ }
60
+ },
61
+ "affine": {
62
+ "weight": 0.0,
63
+ "type": "RandomAffine",
64
+ "kwargs": {
65
+ "degrees": [
66
+ 0,
67
+ 0
68
+ ],
69
+ "translate": [
70
+ 0.0,
71
+ 0.0
72
+ ]
73
+ }
74
+ }
75
+ }
76
+ },
77
+ "revision": "main",
78
+ "use_imagenet_stats": true,
79
+ "video_backend": "pyav",
80
+ "streaming": false
81
+ },
82
+ "env": null,
83
+ "policy": {
84
+ "type": "smolvla",
85
+ "n_obs_steps": 1,
86
+ "input_features": {
87
+ "observation.state": {
88
+ "type": "STATE",
89
+ "shape": [
90
+ 6
91
+ ]
92
+ },
93
+ "observation.images.camera1": {
94
+ "type": "VISUAL",
95
+ "shape": [
96
+ 3,
97
+ 256,
98
+ 256
99
+ ]
100
+ },
101
+ "observation.images.camera2": {
102
+ "type": "VISUAL",
103
+ "shape": [
104
+ 3,
105
+ 256,
106
+ 256
107
+ ]
108
+ },
109
+ "observation.images.camera3": {
110
+ "type": "VISUAL",
111
+ "shape": [
112
+ 3,
113
+ 256,
114
+ 256
115
+ ]
116
+ },
117
+ "observation.images.empty_camera_0": {
118
+ "type": "VISUAL",
119
+ "shape": [
120
+ 3,
121
+ 480,
122
+ 640
123
+ ]
124
+ },
125
+ "observation.images.empty_camera_1": {
126
+ "type": "VISUAL",
127
+ "shape": [
128
+ 3,
129
+ 480,
130
+ 640
131
+ ]
132
+ }
133
+ },
134
+ "output_features": {
135
+ "action": {
136
+ "type": "ACTION",
137
+ "shape": [
138
+ 6
139
+ ]
140
+ }
141
+ },
142
+ "device": "cuda",
143
+ "use_amp": true,
144
+ "use_peft": false,
145
+ "push_to_hub": false,
146
+ "repo_id": null,
147
+ "private": null,
148
+ "tags": null,
149
+ "license": null,
150
+ "pretrained_path": "lerobot/smolvla_base",
151
+ "chunk_size": 50,
152
+ "n_action_steps": 50,
153
+ "normalization_mapping": {
154
+ "VISUAL": "IDENTITY",
155
+ "STATE": "MEAN_STD",
156
+ "ACTION": "MEAN_STD"
157
+ },
158
+ "max_state_dim": 32,
159
+ "max_action_dim": 32,
160
+ "resize_imgs_with_padding": [
161
+ 256,
162
+ 256
163
+ ],
164
+ "empty_cameras": 2,
165
+ "adapt_to_pi_aloha": false,
166
+ "use_delta_joint_actions_aloha": false,
167
+ "tokenizer_max_length": 48,
168
+ "num_steps": 10,
169
+ "use_cache": true,
170
+ "freeze_vision_encoder": true,
171
+ "train_expert_only": true,
172
+ "train_state_proj": true,
173
+ "optimizer_lr": 0.0001,
174
+ "optimizer_betas": [
175
+ 0.9,
176
+ 0.95
177
+ ],
178
+ "optimizer_eps": 1e-08,
179
+ "optimizer_weight_decay": 1e-10,
180
+ "optimizer_grad_clip_norm": 10.0,
181
+ "scheduler_warmup_steps": 1000,
182
+ "scheduler_decay_steps": 30000,
183
+ "scheduler_decay_lr": 2.5e-06,
184
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
185
+ "load_vlm_weights": true,
186
+ "add_image_special_tokens": false,
187
+ "attention_mode": "cross_attn",
188
+ "prefix_length": 0,
189
+ "pad_language_to": "max_length",
190
+ "num_expert_layers": 0,
191
+ "num_vlm_layers": 16,
192
+ "self_attn_every_n_layers": 2,
193
+ "expert_width_multiplier": 0.75,
194
+ "min_period": 0.004,
195
+ "max_period": 4.0,
196
+ "rtc_config": null,
197
+ "compile_model": false,
198
+ "compile_mode": "max-autotune"
199
+ },
200
+ "output_dir": "/home/shadeform/outputs/eval1_fixed_bowl_smolvla_v1",
201
+ "job_name": "eval1_fixed_bowl_smolvla_v1",
202
+ "resume": false,
203
+ "seed": 1000,
204
+ "num_workers": 8,
205
+ "batch_size": 64,
206
+ "steps": 10000,
207
+ "eval_freq": 10000,
208
+ "log_freq": 100,
209
+ "tolerance_s": 0.0001,
210
+ "save_checkpoint": true,
211
+ "save_freq": 2000,
212
+ "use_policy_training_preset": false,
213
+ "optimizer": {
214
+ "type": "adamw",
215
+ "lr": 0.0001,
216
+ "weight_decay": 1e-10,
217
+ "grad_clip_norm": 10.0,
218
+ "betas": [
219
+ 0.9,
220
+ 0.95
221
+ ],
222
+ "eps": 1e-08
223
+ },
224
+ "scheduler": {
225
+ "type": "cosine_decay_with_warmup",
226
+ "num_warmup_steps": 1000,
227
+ "num_decay_steps": 10000,
228
+ "peak_lr": 0.0001,
229
+ "decay_lr": 2.5e-06
230
+ },
231
+ "eval": {
232
+ "n_episodes": 50,
233
+ "batch_size": 50,
234
+ "use_async_envs": false
235
+ },
236
+ "wandb": {
237
+ "enable": true,
238
+ "disable_artifact": false,
239
+ "project": "smolvla-eval2",
240
+ "entity": null,
241
+ "notes": null,
242
+ "run_id": "v7yw6j14",
243
+ "mode": null
244
+ },
245
+ "peft": null,
246
+ "use_rabc": false,
247
+ "rabc_progress_path": null,
248
+ "rabc_kappa": 0.01,
249
+ "rabc_epsilon": 1e-06,
250
+ "rabc_head_mode": "sparse",
251
+ "rename_map": {
252
+ "observation.images.front": "observation.images.camera1"
253
+ },
254
+ "checkpoint_path": null
255
+ }