Hugging Face's logo

hellorobotinc
/
diffusion_kitchen_cabinet_diagonal

Robotics
Safetensors
LeRobot
Model card Files Community
diffusion_kitchen_cabinet_diagonal
File size: 3,271 Bytes 
268b296
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
 
resume: false
device: cuda
use_amp: false
seed: 100000
dataset_repo_id: hellorobotinc/kitchen_cabinet_diagonal
video_backend: pyav
training:
  offline_steps: 200000
  online_steps: 0
  online_steps_between_rollouts: 1
  online_sampling_ratio: 0.5
  online_env_seed: ???
  eval_freq: -1
  log_freq: 250
  save_checkpoint: true
  save_freq: 10000
  num_workers: 16
  batch_size: 64
  image_transforms:
    enable: true
    max_num_transforms: 3
    random_order: false
    brightness:
      weight: 1
      min_max:
      - 0.8
      - 1.2
    contrast:
      weight: 1
      min_max:
      - 0.8
      - 1.2
    saturation:
      weight: 1
      min_max:
      - 0.5
      - 1.5
    hue:
      weight: 1
      min_max:
      - -0.05
      - 0.05
    sharpness:
      weight: 1
      min_max:
      - 0.8
      - 1.2
  grad_clip_norm: 10
  lr: 0.0001
  lr_scheduler: cosine
  lr_warmup_steps: 500
  adam_betas:
  - 0.95
  - 0.999
  adam_eps: 1.0e-08
  adam_weight_decay: 1.0e-06
  delta_timestamps:
    observation.images.gripper:
    - -0.16666666666666666
    - 0.0
    observation.images.head:
    - -0.16666666666666666
    - 0.0
    observation.state:
    - -0.16666666666666666
    - 0.0
    action:
    - -0.16666666666666666
    - 0.0
    - 0.16666666666666666
    - 0.3333333333333333
    - 0.5
    - 0.6666666666666666
    - 0.8333333333333334
    - 1.0
    - 1.1666666666666667
    - 1.3333333333333333
    - 1.5
    - 1.6666666666666667
    - 1.8333333333333333
    - 2.0
    - 2.1666666666666665
    - 2.3333333333333335
  drop_n_last_frames: 7
eval:
  n_episodes: 50
  batch_size: 50
  use_async_envs: false
wandb:
  enable: true
  disable_artifact: true
  project: diffusion-kitchen-diagonal
  notes: no depth
fps: 6
env:
  name: stretch_real
  task: stretch_baseX
  state_dim: 9
  action_dim: 9
  fps: ${fps}
  episode_length: 400
  gym:
    fps: ${fps}
repo_version: main
override_dataset_stats:
  observation.images.gripper:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.head:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
policy:
  name: diffusion
  n_obs_steps: 2
  horizon: 16
  n_action_steps: 8
  input_shapes:
    observation.images.gripper:
    - 3
    - 320
    - 320
    observation.images.head:
    - 3
    - 320
    - 320
    observation.state:
    - ${env.state_dim}
  output_shapes:
    action:
    - ${env.action_dim}
  input_normalization_modes:
    observation.images.gripper: mean_std
    observation.images.head: mean_std
    observation.state: min_max
  output_normalization_modes:
    action: min_max
  vision_backbone: resnet18
  crop_shape:
  - 320
  - 320
  crop_is_random: false
  pretrained_backbone_weights: null
  use_group_norm: true
  spatial_softmax_num_keypoints: 32
  down_dims:
  - 512
  - 1024
  - 2048
  kernel_size: 5
  n_groups: 8
  diffusion_step_embed_dim: 128
  use_film_scale_modulation: true
  noise_scheduler_type: DDPM
  num_train_timesteps: 100
  beta_schedule: squaredcos_cap_v2
  beta_start: 0.0001
  beta_end: 0.02
  prediction_type: epsilon
  clip_sample: true
  clip_sample_range: 1.0
  num_inference_steps: 100
  do_mask_loss_for_padding: false