Hugging Face's logo

jasontchan
/
no-emg-task2

Safetensors
Model card Files
xet
Community
jasontchan commited on
Commit
641a45e
·
verified ·
1 Parent(s): 7a16e79

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +103 -0
  2. model.safetensors +3 -0
  3. train_config.json +216 -0
config.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "smolvla",
3
+ "n_obs_steps": 1,
4
+ "normalization_mapping": {
5
+ "VISUAL": "IDENTITY",
6
+ "STATE": "MEAN_STD",
7
+ "ACTION": "MEAN_STD"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.front": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 1080,
21
+ 1920
22
+ ]
23
+ },
24
+ "observation.images.top": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 720,
29
+ 1280
30
+ ]
31
+ },
32
+ "observation.images.gripper": {
33
+ "type": "VISUAL",
34
+ "shape": [
35
+ 3,
36
+ 1080,
37
+ 1920
38
+ ]
39
+ }
40
+ },
41
+ "output_features": {
42
+ "action": {
43
+ "type": "ACTION",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ }
48
+ },
49
+ "device": "cuda",
50
+ "use_amp": false,
51
+ "emg_mask_prob": 0.0,
52
+ "emg_spectrogram": false,
53
+ "n_fft": 64,
54
+ "emg_window_size": 0,
55
+ "emg_n_bins": 0,
56
+ "emg_gate": 0.0,
57
+ "emg_shift": 0,
58
+ "use_camera_keys": {
59
+ "observation.images.top": true,
60
+ "observation.images.gripper": true,
61
+ "observation.images.front": true
62
+ },
63
+ "chunk_size": 50,
64
+ "n_action_steps": 50,
65
+ "max_state_dim": 32,
66
+ "max_action_dim": 32,
67
+ "resize_imgs_with_padding": [
68
+ 512,
69
+ 512
70
+ ],
71
+ "empty_cameras": 0,
72
+ "adapt_to_pi_aloha": false,
73
+ "use_delta_joint_actions_aloha": false,
74
+ "tokenizer_max_length": 48,
75
+ "num_steps": 10,
76
+ "use_cache": true,
77
+ "freeze_vision_encoder": true,
78
+ "train_expert_only": true,
79
+ "train_state_proj": true,
80
+ "optimizer_lr": 0.0001,
81
+ "optimizer_betas": [
82
+ 0.9,
83
+ 0.95
84
+ ],
85
+ "optimizer_eps": 1e-08,
86
+ "optimizer_weight_decay": 1e-10,
87
+ "optimizer_grad_clip_norm": 10.0,
88
+ "scheduler_warmup_steps": 1000,
89
+ "scheduler_decay_steps": 30000,
90
+ "scheduler_decay_lr": 2.5e-06,
91
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
92
+ "load_vlm_weights": true,
93
+ "add_image_special_tokens": false,
94
+ "attention_mode": "cross_attn",
95
+ "prefix_length": 0,
96
+ "pad_language_to": "max_length",
97
+ "num_expert_layers": 0,
98
+ "num_vlm_layers": 16,
99
+ "self_attn_every_n_layers": 2,
100
+ "expert_width_multiplier": 0.75,
101
+ "min_period": 0.004,
102
+ "max_period": 4.0
103
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67021ead6a50d6c652d12b746d0250cf6c75754407386de7932b454306e2650b
3
+ size 907235268
train_config.json ADDED
@@ -0,0 +1,216 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "jasontchan/emg-gestures-task2",
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
+ }
62
+ },
63
+ "revision": null,
64
+ "use_imagenet_stats": true,
65
+ "video_backend": "torchcodec"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "smolvla",
70
+ "n_obs_steps": 1,
71
+ "normalization_mapping": {
72
+ "VISUAL": "IDENTITY",
73
+ "STATE": "MEAN_STD",
74
+ "ACTION": "MEAN_STD"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.front": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 1080,
88
+ 1920
89
+ ]
90
+ },
91
+ "observation.images.top": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 720,
96
+ 1280
97
+ ]
98
+ },
99
+ "observation.images.gripper": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 1080,
104
+ 1920
105
+ ]
106
+ }
107
+ },
108
+ "output_features": {
109
+ "action": {
110
+ "type": "ACTION",
111
+ "shape": [
112
+ 6
113
+ ]
114
+ }
115
+ },
116
+ "device": "cuda",
117
+ "use_amp": false,
118
+ "emg_mask_prob": 0.0,
119
+ "emg_spectrogram": false,
120
+ "n_fft": 64,
121
+ "emg_window_size": 0,
122
+ "emg_n_bins": 0,
123
+ "emg_gate": 0.0,
124
+ "emg_shift": 0,
125
+ "use_camera_keys": {
126
+ "observation.images.top": true,
127
+ "observation.images.gripper": true,
128
+ "observation.images.front": true
129
+ },
130
+ "chunk_size": 50,
131
+ "n_action_steps": 50,
132
+ "max_state_dim": 32,
133
+ "max_action_dim": 32,
134
+ "resize_imgs_with_padding": [
135
+ 512,
136
+ 512
137
+ ],
138
+ "empty_cameras": 0,
139
+ "adapt_to_pi_aloha": false,
140
+ "use_delta_joint_actions_aloha": false,
141
+ "tokenizer_max_length": 48,
142
+ "num_steps": 10,
143
+ "use_cache": true,
144
+ "freeze_vision_encoder": true,
145
+ "train_expert_only": true,
146
+ "train_state_proj": true,
147
+ "optimizer_lr": 0.0001,
148
+ "optimizer_betas": [
149
+ 0.9,
150
+ 0.95
151
+ ],
152
+ "optimizer_eps": 1e-08,
153
+ "optimizer_weight_decay": 1e-10,
154
+ "optimizer_grad_clip_norm": 10.0,
155
+ "scheduler_warmup_steps": 1000,
156
+ "scheduler_decay_steps": 30000,
157
+ "scheduler_decay_lr": 2.5e-06,
158
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
159
+ "load_vlm_weights": true,
160
+ "add_image_special_tokens": false,
161
+ "attention_mode": "cross_attn",
162
+ "prefix_length": 0,
163
+ "pad_language_to": "max_length",
164
+ "num_expert_layers": 0,
165
+ "num_vlm_layers": 16,
166
+ "self_attn_every_n_layers": 2,
167
+ "expert_width_multiplier": 0.75,
168
+ "min_period": 0.004,
169
+ "max_period": 4.0
170
+ },
171
+ "output_dir": "outputs/train/no-emg-task2",
172
+ "job_name": "no-emg-task2",
173
+ "resume": false,
174
+ "seed": 1000,
175
+ "num_workers": 4,
176
+ "batch_size": 8,
177
+ "steps": 100000,
178
+ "eval_freq": 20000,
179
+ "log_freq": 200,
180
+ "save_checkpoint": true,
181
+ "save_freq": 20000,
182
+ "use_policy_training_preset": true,
183
+ "optimizer": {
184
+ "type": "adamw",
185
+ "lr": 0.0001,
186
+ "weight_decay": 1e-10,
187
+ "grad_clip_norm": 10.0,
188
+ "betas": [
189
+ 0.9,
190
+ 0.95
191
+ ],
192
+ "eps": 1e-08
193
+ },
194
+ "scheduler": {
195
+ "type": "cosine_decay_with_warmup",
196
+ "num_warmup_steps": 1000,
197
+ "num_decay_steps": 30000,
198
+ "peak_lr": 0.0001,
199
+ "decay_lr": 2.5e-06
200
+ },
201
+ "eval": {
202
+ "n_episodes": 50,
203
+ "batch_size": 50,
204
+ "use_async_envs": false
205
+ },
206
+ "wandb": {
207
+ "enable": true,
208
+ "disable_artifact": false,
209
+ "project": "lerobot",
210
+ "entity": null,
211
+ "notes": null,
212
+ "run_id": "74wrafh8",
213
+ "mode": null
214
+ },
215
+ "use_emg": false
216
+ }