Hugging Face's logo

astro189
/
coffee_pouring

Safetensors
Model card Files
xet
 Community
astro189 commited on
Commit
5e0e709
·
1 Parent(s): 1465678

Add 030000 checkpoint weights

Browse files
Files changed (3) hide show
  1. config.json +83 -0
  2. model.safetensors +3 -0
  3. train_config.json +247 -0
config.json ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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.gripper": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ },
24
+ "observation.images.top": {
25
+ "type": "VISUAL",
26
+ "shape": [
27
+ 3,
28
+ 480,
29
+ 640
30
+ ]
31
+ }
32
+ },
33
+ "output_features": {
34
+ "action": {
35
+ "type": "ACTION",
36
+ "shape": [
37
+ 6
38
+ ]
39
+ }
40
+ },
41
+ "device": "cuda",
42
+ "use_amp": false,
43
+ "chunk_size": 50,
44
+ "n_action_steps": 50,
45
+ "max_state_dim": 32,
46
+ "max_action_dim": 32,
47
+ "resize_imgs_with_padding": [
48
+ 512,
49
+ 512
50
+ ],
51
+ "empty_cameras": 0,
52
+ "adapt_to_pi_aloha": false,
53
+ "use_delta_joint_actions_aloha": false,
54
+ "tokenizer_max_length": 48,
55
+ "num_steps": 10,
56
+ "use_cache": true,
57
+ "freeze_vision_encoder": true,
58
+ "train_expert_only": true,
59
+ "train_state_proj": true,
60
+ "optimizer_lr": 0.0001,
61
+ "optimizer_betas": [
62
+ 0.9,
63
+ 0.95
64
+ ],
65
+ "optimizer_eps": 1e-08,
66
+ "optimizer_weight_decay": 1e-10,
67
+ "optimizer_grad_clip_norm": 10.0,
68
+ "scheduler_warmup_steps": 1000,
69
+ "scheduler_decay_steps": 30000,
70
+ "scheduler_decay_lr": 2.5e-06,
71
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
72
+ "load_vlm_weights": true,
73
+ "add_image_special_tokens": false,
74
+ "attention_mode": "cross_attn",
75
+ "prefix_length": 0,
76
+ "pad_language_to": "max_length",
77
+ "num_expert_layers": 0,
78
+ "num_vlm_layers": 16,
79
+ "self_attn_every_n_layers": 2,
80
+ "expert_width_multiplier": 0.75,
81
+ "min_period": 0.004,
82
+ "max_period": 4.0
83
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de17ff9d659108dbe23265fb7e6a7168f911ad120249dd20215a5516da13aa26
3
+ size 906713296
train_config.json ADDED
@@ -0,0 +1,247 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "astro189/record_scene_1",
4
+ "root": null,
5
+ "episodes": [
6
+ 0,
7
+ 1,
8
+ 2,
9
+ 3,
10
+ 4,
11
+ 5,
12
+ 6,
13
+ 7,
14
+ 8,
15
+ 9,
16
+ 10,
17
+ 11,
18
+ 12,
19
+ 13,
20
+ 14,
21
+ 15,
22
+ 16,
23
+ 17,
24
+ 18,
25
+ 19,
26
+ 20,
27
+ 21,
28
+ 22,
29
+ 23,
30
+ 24,
31
+ 25,
32
+ 26,
33
+ 27,
34
+ 28,
35
+ 29,
36
+ 30,
37
+ 31,
38
+ 32,
39
+ 33,
40
+ 34,
41
+ 35,
42
+ 36,
43
+ 37,
44
+ 38,
45
+ 39,
46
+ 40,
47
+ 41,
48
+ 42,
49
+ 43,
50
+ 44,
51
+ 45,
52
+ 46,
53
+ 47,
54
+ 48,
55
+ 49,
56
+ 50
57
+ ],
58
+ "image_transforms": {
59
+ "enable": false,
60
+ "max_num_transforms": 3,
61
+ "random_order": false,
62
+ "tfs": {
63
+ "brightness": {
64
+ "weight": 1.0,
65
+ "type": "ColorJitter",
66
+ "kwargs": {
67
+ "brightness": [
68
+ 0.8,
69
+ 1.2
70
+ ]
71
+ }
72
+ },
73
+ "contrast": {
74
+ "weight": 1.0,
75
+ "type": "ColorJitter",
76
+ "kwargs": {
77
+ "contrast": [
78
+ 0.8,
79
+ 1.2
80
+ ]
81
+ }
82
+ },
83
+ "saturation": {
84
+ "weight": 1.0,
85
+ "type": "ColorJitter",
86
+ "kwargs": {
87
+ "saturation": [
88
+ 0.5,
89
+ 1.5
90
+ ]
91
+ }
92
+ },
93
+ "hue": {
94
+ "weight": 1.0,
95
+ "type": "ColorJitter",
96
+ "kwargs": {
97
+ "hue": [
98
+ -0.05,
99
+ 0.05
100
+ ]
101
+ }
102
+ },
103
+ "sharpness": {
104
+ "weight": 1.0,
105
+ "type": "SharpnessJitter",
106
+ "kwargs": {
107
+ "sharpness": [
108
+ 0.5,
109
+ 1.5
110
+ ]
111
+ }
112
+ }
113
+ }
114
+ },
115
+ "revision": null,
116
+ "use_imagenet_stats": true,
117
+ "video_backend": "torchcodec"
118
+ },
119
+ "env": null,
120
+ "policy": {
121
+ "type": "smolvla",
122
+ "n_obs_steps": 1,
123
+ "normalization_mapping": {
124
+ "VISUAL": "IDENTITY",
125
+ "STATE": "MEAN_STD",
126
+ "ACTION": "MEAN_STD"
127
+ },
128
+ "input_features": {
129
+ "observation.state": {
130
+ "type": "STATE",
131
+ "shape": [
132
+ 6
133
+ ]
134
+ },
135
+ "observation.images.gripper": {
136
+ "type": "VISUAL",
137
+ "shape": [
138
+ 3,
139
+ 480,
140
+ 640
141
+ ]
142
+ },
143
+ "observation.images.top": {
144
+ "type": "VISUAL",
145
+ "shape": [
146
+ 3,
147
+ 480,
148
+ 640
149
+ ]
150
+ }
151
+ },
152
+ "output_features": {
153
+ "action": {
154
+ "type": "ACTION",
155
+ "shape": [
156
+ 6
157
+ ]
158
+ }
159
+ },
160
+ "device": "cuda",
161
+ "use_amp": false,
162
+ "chunk_size": 50,
163
+ "n_action_steps": 50,
164
+ "max_state_dim": 32,
165
+ "max_action_dim": 32,
166
+ "resize_imgs_with_padding": [
167
+ 512,
168
+ 512
169
+ ],
170
+ "empty_cameras": 0,
171
+ "adapt_to_pi_aloha": false,
172
+ "use_delta_joint_actions_aloha": false,
173
+ "tokenizer_max_length": 48,
174
+ "num_steps": 10,
175
+ "use_cache": true,
176
+ "freeze_vision_encoder": true,
177
+ "train_expert_only": true,
178
+ "train_state_proj": true,
179
+ "optimizer_lr": 0.0001,
180
+ "optimizer_betas": [
181
+ 0.9,
182
+ 0.95
183
+ ],
184
+ "optimizer_eps": 1e-08,
185
+ "optimizer_weight_decay": 1e-10,
186
+ "optimizer_grad_clip_norm": 10.0,
187
+ "scheduler_warmup_steps": 1000,
188
+ "scheduler_decay_steps": 30000,
189
+ "scheduler_decay_lr": 2.5e-06,
190
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
191
+ "load_vlm_weights": true,
192
+ "add_image_special_tokens": false,
193
+ "attention_mode": "cross_attn",
194
+ "prefix_length": 0,
195
+ "pad_language_to": "max_length",
196
+ "num_expert_layers": 0,
197
+ "num_vlm_layers": 16,
198
+ "self_attn_every_n_layers": 2,
199
+ "expert_width_multiplier": 0.75,
200
+ "min_period": 0.004,
201
+ "max_period": 4.0
202
+ },
203
+ "output_dir": "outputs/train/2025-06-15/11-37-30_smolvla",
204
+ "job_name": "smolvla",
205
+ "resume": false,
206
+ "seed": 1000,
207
+ "num_workers": 4,
208
+ "batch_size": 8,
209
+ "steps": 40000,
210
+ "eval_freq": 5000,
211
+ "log_freq": 200,
212
+ "save_checkpoint": true,
213
+ "save_freq": 10000,
214
+ "use_policy_training_preset": true,
215
+ "optimizer": {
216
+ "type": "adamw",
217
+ "lr": 0.0001,
218
+ "weight_decay": 1e-10,
219
+ "grad_clip_norm": 10.0,
220
+ "betas": [
221
+ 0.9,
222
+ 0.95
223
+ ],
224
+ "eps": 1e-08
225
+ },
226
+ "scheduler": {
227
+ "type": "cosine_decay_with_warmup",
228
+ "num_warmup_steps": 1000,
229
+ "num_decay_steps": 30000,
230
+ "peak_lr": 0.0001,
231
+ "decay_lr": 2.5e-06
232
+ },
233
+ "eval": {
234
+ "n_episodes": 50,
235
+ "batch_size": 50,
236
+ "use_async_envs": false
237
+ },
238
+ "wandb": {
239
+ "enable": true,
240
+ "disable_artifact": true,
241
+ "project": "coffee-pouring",
242
+ "entity": "astro189-indian-institute-of-information-technology-bhag",
243
+ "notes": "pouring coffee",
244
+ "run_id": "4bl10as2",
245
+ "mode": null
246
+ }
247
+ }