Hugging Face's logo

hcltech-robotics
/
diffusion_glasses_exp_5

Safetensors
Model card Files
xet
Community
tfaith commited on
Commit
1384d3c
·
verified ·
1 Parent(s): 37b476d

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +91 -0
  2. model.safetensors +3 -0
  3. train_config.json +200 -0
config.json ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "n_obs_steps": 2,
3
+ "normalization_mapping": {
4
+ "VISUAL": "MEAN_STD",
5
+ "STATE": "MIN_MAX",
6
+ "ACTION": "MIN_MAX"
7
+ },
8
+ "input_features": {
9
+ "observation.state": {
10
+ "type": "STATE",
11
+ "shape": [
12
+ 6
13
+ ]
14
+ },
15
+ "observation.images.top": {
16
+ "type": "VISUAL",
17
+ "shape": [
18
+ 3,
19
+ 480,
20
+ 640
21
+ ]
22
+ },
23
+ "observation.images.side": {
24
+ "type": "VISUAL",
25
+ "shape": [
26
+ 3,
27
+ 480,
28
+ 640
29
+ ]
30
+ },
31
+ "observation.images.arm": {
32
+ "type": "VISUAL",
33
+ "shape": [
34
+ 3,
35
+ 480,
36
+ 640
37
+ ]
38
+ }
39
+ },
40
+ "output_features": {
41
+ "action": {
42
+ "type": "ACTION",
43
+ "shape": [
44
+ 6
45
+ ]
46
+ }
47
+ },
48
+ "device": "cuda",
49
+ "use_amp": false,
50
+ "horizon": 16,
51
+ "n_action_steps": 8,
52
+ "drop_n_last_frames": 7,
53
+ "vision_backbone": "resnet18",
54
+ "crop_shape": [
55
+ 84,
56
+ 84
57
+ ],
58
+ "crop_is_random": true,
59
+ "pretrained_backbone_weights": null,
60
+ "use_group_norm": true,
61
+ "spatial_softmax_num_keypoints": 32,
62
+ "use_separate_rgb_encoder_per_camera": false,
63
+ "down_dims": [
64
+ 512,
65
+ 1024,
66
+ 2048
67
+ ],
68
+ "kernel_size": 5,
69
+ "n_groups": 8,
70
+ "diffusion_step_embed_dim": 128,
71
+ "use_film_scale_modulation": true,
72
+ "noise_scheduler_type": "DDPM",
73
+ "num_train_timesteps": 100,
74
+ "beta_schedule": "squaredcos_cap_v2",
75
+ "beta_start": 0.0001,
76
+ "beta_end": 0.02,
77
+ "prediction_type": "epsilon",
78
+ "clip_sample": true,
79
+ "clip_sample_range": 1.0,
80
+ "num_inference_steps": null,
81
+ "do_mask_loss_for_padding": false,
82
+ "optimizer_lr": 0.0001,
83
+ "optimizer_betas": [
84
+ 0.95,
85
+ 0.999
86
+ ],
87
+ "optimizer_eps": 1e-08,
88
+ "optimizer_weight_decay": 1e-06,
89
+ "scheduler_name": "cosine",
90
+ "scheduler_warmup_steps": 500
91
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3bd7a1793420db06d28bc2cd96b409765d2ebe909d1a1e6602fa0aaf4467e2de
3
+ size 1081198064
train_config.json ADDED
@@ -0,0 +1,200 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "tfoldi/glasses_exp_5",
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": "pyav"
66
+ },
67
+ "env": null,
68
+ "policy": {
69
+ "type": "diffusion",
70
+ "n_obs_steps": 2,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "input_features": {
77
+ "observation.state": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 6
81
+ ]
82
+ },
83
+ "observation.images.top": {
84
+ "type": "VISUAL",
85
+ "shape": [
86
+ 3,
87
+ 480,
88
+ 640
89
+ ]
90
+ },
91
+ "observation.images.side": {
92
+ "type": "VISUAL",
93
+ "shape": [
94
+ 3,
95
+ 480,
96
+ 640
97
+ ]
98
+ },
99
+ "observation.images.arm": {
100
+ "type": "VISUAL",
101
+ "shape": [
102
+ 3,
103
+ 480,
104
+ 640
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
+ "horizon": 16,
119
+ "n_action_steps": 8,
120
+ "drop_n_last_frames": 7,
121
+ "vision_backbone": "resnet18",
122
+ "crop_shape": [
123
+ 84,
124
+ 84
125
+ ],
126
+ "crop_is_random": true,
127
+ "pretrained_backbone_weights": null,
128
+ "use_group_norm": true,
129
+ "spatial_softmax_num_keypoints": 32,
130
+ "use_separate_rgb_encoder_per_camera": false,
131
+ "down_dims": [
132
+ 512,
133
+ 1024,
134
+ 2048
135
+ ],
136
+ "kernel_size": 5,
137
+ "n_groups": 8,
138
+ "diffusion_step_embed_dim": 128,
139
+ "use_film_scale_modulation": true,
140
+ "noise_scheduler_type": "DDPM",
141
+ "num_train_timesteps": 100,
142
+ "beta_schedule": "squaredcos_cap_v2",
143
+ "beta_start": 0.0001,
144
+ "beta_end": 0.02,
145
+ "prediction_type": "epsilon",
146
+ "clip_sample": true,
147
+ "clip_sample_range": 1.0,
148
+ "num_inference_steps": null,
149
+ "do_mask_loss_for_padding": false,
150
+ "optimizer_lr": 0.0001,
151
+ "optimizer_betas": [
152
+ 0.95,
153
+ 0.999
154
+ ],
155
+ "optimizer_eps": 1e-08,
156
+ "optimizer_weight_decay": 1e-06,
157
+ "scheduler_name": "cosine",
158
+ "scheduler_warmup_steps": 500
159
+ },
160
+ "output_dir": "outputs/train/diffusion_glasses_exp_5",
161
+ "job_name": "diffusion_glasses_exp_5",
162
+ "resume": false,
163
+ "seed": 1000,
164
+ "num_workers": 4,
165
+ "batch_size": 8,
166
+ "steps": 100000,
167
+ "eval_freq": 20000,
168
+ "log_freq": 200,
169
+ "save_checkpoint": true,
170
+ "save_freq": 20000,
171
+ "use_policy_training_preset": true,
172
+ "optimizer": {
173
+ "type": "adam",
174
+ "lr": 0.0001,
175
+ "weight_decay": 1e-06,
176
+ "grad_clip_norm": 10.0,
177
+ "betas": [
178
+ 0.95,
179
+ 0.999
180
+ ],
181
+ "eps": 1e-08
182
+ },
183
+ "scheduler": {
184
+ "type": "diffuser",
185
+ "num_warmup_steps": 500,
186
+ "name": "cosine"
187
+ },
188
+ "eval": {
189
+ "n_episodes": 50,
190
+ "batch_size": 50,
191
+ "use_async_envs": false
192
+ },
193
+ "wandb": {
194
+ "enable": true,
195
+ "disable_artifact": false,
196
+ "project": "lerobot",
197
+ "entity": null,
198
+ "notes": null
199
+ }
200
+ }