Hugging Face's logo

takuzennn
/
pi0_transfer_cube

Safetensors
Model card Files Community
1
takuzennn commited on
Commit
52740c9
·
verified ·
1 Parent(s): 6befd29

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +62 -0
  2. model.safetensors +3 -0
  3. train_config.json +173 -0
config.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "pi0",
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.images.top": {
11
+ "type": "VISUAL",
12
+ "shape": [
13
+ 3,
14
+ 480,
15
+ 640
16
+ ]
17
+ },
18
+ "observation.state": {
19
+ "type": "STATE",
20
+ "shape": [
21
+ 14
22
+ ]
23
+ }
24
+ },
25
+ "output_features": {
26
+ "action": {
27
+ "type": "ACTION",
28
+ "shape": [
29
+ 14
30
+ ]
31
+ }
32
+ },
33
+ "chunk_size": 50,
34
+ "n_action_steps": 50,
35
+ "max_state_dim": 32,
36
+ "max_action_dim": 32,
37
+ "resize_imgs_with_padding": [
38
+ 224,
39
+ 224
40
+ ],
41
+ "empty_cameras": 0,
42
+ "adapt_to_pi_aloha": false,
43
+ "use_delta_joint_actions_aloha": false,
44
+ "tokenizer_max_length": 48,
45
+ "proj_width": 1024,
46
+ "num_steps": 10,
47
+ "use_cache": true,
48
+ "attention_implementation": "eager",
49
+ "freeze_vision_encoder": true,
50
+ "train_expert_only": false,
51
+ "train_state_proj": true,
52
+ "optimizer_lr": 2.5e-05,
53
+ "optimizer_betas": [
54
+ 0.9,
55
+ 0.95
56
+ ],
57
+ "optimizer_eps": 1e-08,
58
+ "optimizer_weight_decay": 1e-10,
59
+ "scheduler_warmup_steps": 1000,
60
+ "scheduler_decay_steps": 30000,
61
+ "scheduler_decay_lr": 2.5e-06
62
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d35e11fa1bf9152a1ca6cbb8e881b008118b6367d76c5a17a1ad2a2281b37354
3
+ size 7536022744
train_config.json ADDED
@@ -0,0 +1,173 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "lerobot/aloha_sim_transfer_cube_human",
4
+ "episodes": null,
5
+ "image_transforms": {
6
+ "enable": false,
7
+ "max_num_transforms": 3,
8
+ "random_order": false,
9
+ "tfs": {
10
+ "brightness": {
11
+ "weight": 1.0,
12
+ "type": "ColorJitter",
13
+ "kwargs": {
14
+ "brightness": [
15
+ 0.8,
16
+ 1.2
17
+ ]
18
+ }
19
+ },
20
+ "contrast": {
21
+ "weight": 1.0,
22
+ "type": "ColorJitter",
23
+ "kwargs": {
24
+ "contrast": [
25
+ 0.8,
26
+ 1.2
27
+ ]
28
+ }
29
+ },
30
+ "saturation": {
31
+ "weight": 1.0,
32
+ "type": "ColorJitter",
33
+ "kwargs": {
34
+ "saturation": [
35
+ 0.5,
36
+ 1.5
37
+ ]
38
+ }
39
+ },
40
+ "hue": {
41
+ "weight": 1.0,
42
+ "type": "ColorJitter",
43
+ "kwargs": {
44
+ "hue": [
45
+ -0.05,
46
+ 0.05
47
+ ]
48
+ }
49
+ },
50
+ "sharpness": {
51
+ "weight": 1.0,
52
+ "type": "SharpnessJitter",
53
+ "kwargs": {
54
+ "sharpness": [
55
+ 0.5,
56
+ 1.5
57
+ ]
58
+ }
59
+ }
60
+ }
61
+ },
62
+ "local_files_only": false,
63
+ "use_imagenet_stats": true,
64
+ "video_backend": "pyav"
65
+ },
66
+ "env": null,
67
+ "policy": {
68
+ "type": "pi0",
69
+ "n_obs_steps": 1,
70
+ "normalization_mapping": {
71
+ "VISUAL": "IDENTITY",
72
+ "STATE": "MEAN_STD",
73
+ "ACTION": "MEAN_STD"
74
+ },
75
+ "input_features": {
76
+ "observation.images.top": {
77
+ "type": "VISUAL",
78
+ "shape": [
79
+ 3,
80
+ 480,
81
+ 640
82
+ ]
83
+ },
84
+ "observation.state": {
85
+ "type": "STATE",
86
+ "shape": [
87
+ 14
88
+ ]
89
+ }
90
+ },
91
+ "output_features": {
92
+ "action": {
93
+ "type": "ACTION",
94
+ "shape": [
95
+ 14
96
+ ]
97
+ }
98
+ },
99
+ "chunk_size": 50,
100
+ "n_action_steps": 50,
101
+ "max_state_dim": 32,
102
+ "max_action_dim": 32,
103
+ "resize_imgs_with_padding": [
104
+ 224,
105
+ 224
106
+ ],
107
+ "empty_cameras": 0,
108
+ "adapt_to_pi_aloha": false,
109
+ "use_delta_joint_actions_aloha": false,
110
+ "tokenizer_max_length": 48,
111
+ "proj_width": 1024,
112
+ "num_steps": 10,
113
+ "use_cache": true,
114
+ "attention_implementation": "eager",
115
+ "freeze_vision_encoder": true,
116
+ "train_expert_only": false,
117
+ "train_state_proj": true,
118
+ "optimizer_lr": 2.5e-05,
119
+ "optimizer_betas": [
120
+ 0.9,
121
+ 0.95
122
+ ],
123
+ "optimizer_eps": 1e-08,
124
+ "optimizer_weight_decay": 1e-10,
125
+ "scheduler_warmup_steps": 1000,
126
+ "scheduler_decay_steps": 30000,
127
+ "scheduler_decay_lr": 2.5e-06
128
+ },
129
+ "output_dir": "outputs/train/2025-02-18/07-06-45_pi0",
130
+ "job_name": "pi0",
131
+ "resume": false,
132
+ "device": "cuda",
133
+ "use_amp": false,
134
+ "seed": 1000,
135
+ "num_workers": 4,
136
+ "batch_size": 8,
137
+ "steps": 100000,
138
+ "eval_freq": 20000,
139
+ "log_freq": 200,
140
+ "save_checkpoint": true,
141
+ "save_freq": 20000,
142
+ "use_policy_training_preset": true,
143
+ "optimizer": {
144
+ "type": "adamw",
145
+ "lr": 2.5e-05,
146
+ "weight_decay": 1e-10,
147
+ "grad_clip_norm": 10.0,
148
+ "betas": [
149
+ 0.9,
150
+ 0.95
151
+ ],
152
+ "eps": 1e-08
153
+ },
154
+ "scheduler": {
155
+ "type": "cosine_decay_with_warmup",
156
+ "num_warmup_steps": 1000,
157
+ "num_decay_steps": 30000,
158
+ "peak_lr": 2.5e-05,
159
+ "decay_lr": 2.5e-06
160
+ },
161
+ "eval": {
162
+ "n_episodes": 50,
163
+ "batch_size": 50,
164
+ "use_async_envs": false
165
+ },
166
+ "wandb": {
167
+ "enable": false,
168
+ "disable_artifact": false,
169
+ "project": "lerobot",
170
+ "entity": null,
171
+ "notes": null
172
+ }
173
+ }