Hugging Face's logo

Benxiaogu
/
dpmoe_bbox_16w

Safetensors
Model card Files
xet
Community
Benxiaogu commited on
Commit
98a05aa
·
verified ·
1 Parent(s): e7764be

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +83 -0
  2. model.safetensors +3 -0
  3. train_config.json +193 -0
config.json ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "diffusion",
3
+ "n_obs_steps": 2,
4
+ "normalization_mapping": {
5
+ "VISUAL": "MEAN_STD",
6
+ "STATE": "MIN_MAX",
7
+ "ACTION": "MIN_MAX"
8
+ },
9
+ "input_features": {
10
+ "observation.state": {
11
+ "type": "STATE",
12
+ "shape": [
13
+ 6
14
+ ]
15
+ },
16
+ "observation.images.top": {
17
+ "type": "VISUAL",
18
+ "shape": [
19
+ 3,
20
+ 480,
21
+ 640
22
+ ]
23
+ }
24
+ },
25
+ "output_features": {
26
+ "action": {
27
+ "type": "ACTION",
28
+ "shape": [
29
+ 6
30
+ ]
31
+ }
32
+ },
33
+ "device": "cuda",
34
+ "use_amp": false,
35
+ "push_to_hub": true,
36
+ "repo_id": "dpmoe_test",
37
+ "private": null,
38
+ "tags": null,
39
+ "license": null,
40
+ "horizon": 16,
41
+ "n_action_steps": 8,
42
+ "drop_n_last_frames": 7,
43
+ "vision_backbone": "resnet18",
44
+ "crop_shape": null,
45
+ "crop_is_random": true,
46
+ "pretrained_backbone_weights": null,
47
+ "use_group_norm": true,
48
+ "spatial_softmax_num_keypoints": 32,
49
+ "use_separate_rgb_encoder_per_camera": false,
50
+ "mask_bbox": true,
51
+ "mask_vision_backbone": "resnet18",
52
+ "use_groundingdino": false,
53
+ "mask_model": "/home/dawn/workspace/GroundingDINO/weights/groundingdino_swinb_cogcoor.pth",
54
+ "mask_prompt": "Pick up the bule cube and place it into box.",
55
+ "down_dims": [
56
+ 512,
57
+ 1024,
58
+ 2048
59
+ ],
60
+ "kernel_size": 5,
61
+ "n_groups": 8,
62
+ "diffusion_step_embed_dim": 128,
63
+ "use_film_scale_modulation": true,
64
+ "noise_scheduler_type": "DDPM",
65
+ "num_train_timesteps": 100,
66
+ "beta_schedule": "squaredcos_cap_v2",
67
+ "beta_start": 0.0001,
68
+ "beta_end": 0.02,
69
+ "prediction_type": "epsilon",
70
+ "clip_sample": true,
71
+ "clip_sample_range": 1.0,
72
+ "num_inference_steps": null,
73
+ "do_mask_loss_for_padding": false,
74
+ "optimizer_lr": 0.0001,
75
+ "optimizer_betas": [
76
+ 0.95,
77
+ 0.999
78
+ ],
79
+ "optimizer_eps": 1e-08,
80
+ "optimizer_weight_decay": 1e-06,
81
+ "scheduler_name": "cosine",
82
+ "scheduler_warmup_steps": 500
83
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a73d8676948a97e0eb2c67003e08208bc22715e99bf93fc2dca28abe4fdd7b8
3
+ size 1111318144
train_config.json ADDED
@@ -0,0 +1,193 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "Benxiaogu/blue_cube_bbox",
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": "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
+ },
92
+ "output_features": {
93
+ "action": {
94
+ "type": "ACTION",
95
+ "shape": [
96
+ 6
97
+ ]
98
+ }
99
+ },
100
+ "device": "cuda",
101
+ "use_amp": false,
102
+ "push_to_hub": true,
103
+ "repo_id": "dpmoe_test",
104
+ "private": null,
105
+ "tags": null,
106
+ "license": null,
107
+ "horizon": 16,
108
+ "n_action_steps": 8,
109
+ "drop_n_last_frames": 7,
110
+ "vision_backbone": "resnet18",
111
+ "crop_shape": null,
112
+ "crop_is_random": true,
113
+ "pretrained_backbone_weights": null,
114
+ "use_group_norm": true,
115
+ "spatial_softmax_num_keypoints": 32,
116
+ "use_separate_rgb_encoder_per_camera": false,
117
+ "mask_bbox": true,
118
+ "mask_vision_backbone": "resnet18",
119
+ "use_groundingdino": false,
120
+ "mask_model": "/home/dawn/workspace/GroundingDINO/weights/groundingdino_swinb_cogcoor.pth",
121
+ "mask_prompt": "Pick up the bule cube and place it into box.",
122
+ "down_dims": [
123
+ 512,
124
+ 1024,
125
+ 2048
126
+ ],
127
+ "kernel_size": 5,
128
+ "n_groups": 8,
129
+ "diffusion_step_embed_dim": 128,
130
+ "use_film_scale_modulation": true,
131
+ "noise_scheduler_type": "DDPM",
132
+ "num_train_timesteps": 100,
133
+ "beta_schedule": "squaredcos_cap_v2",
134
+ "beta_start": 0.0001,
135
+ "beta_end": 0.02,
136
+ "prediction_type": "epsilon",
137
+ "clip_sample": true,
138
+ "clip_sample_range": 1.0,
139
+ "num_inference_steps": null,
140
+ "do_mask_loss_for_padding": false,
141
+ "optimizer_lr": 0.0001,
142
+ "optimizer_betas": [
143
+ 0.95,
144
+ 0.999
145
+ ],
146
+ "optimizer_eps": 1e-08,
147
+ "optimizer_weight_decay": 1e-06,
148
+ "scheduler_name": "cosine",
149
+ "scheduler_warmup_steps": 500
150
+ },
151
+ "output_dir": "outputs/train/dpmoe",
152
+ "job_name": "dpmoe",
153
+ "resume": false,
154
+ "seed": 1000,
155
+ "num_workers": 4,
156
+ "batch_size": 16,
157
+ "steps": 400000,
158
+ "eval_freq": 20000,
159
+ "log_freq": 200,
160
+ "save_checkpoint": true,
161
+ "save_freq": 80000,
162
+ "use_policy_training_preset": true,
163
+ "optimizer": {
164
+ "type": "adam",
165
+ "lr": 0.0001,
166
+ "weight_decay": 1e-06,
167
+ "grad_clip_norm": 10.0,
168
+ "betas": [
169
+ 0.95,
170
+ 0.999
171
+ ],
172
+ "eps": 1e-08
173
+ },
174
+ "scheduler": {
175
+ "type": "diffuser",
176
+ "num_warmup_steps": 500,
177
+ "name": "cosine"
178
+ },
179
+ "eval": {
180
+ "n_episodes": 50,
181
+ "batch_size": 50,
182
+ "use_async_envs": false
183
+ },
184
+ "wandb": {
185
+ "enable": true,
186
+ "disable_artifact": false,
187
+ "project": "lerobot",
188
+ "entity": null,
189
+ "notes": null,
190
+ "run_id": "xng4stig",
191
+ "mode": null
192
+ }
193
+ }