Hugging Face's logo

notmahi
/
sample-tri-BimanualPutRedBellPepperInBin

Transformers
Safetensors
pytorch_model_hub_mixin
model_hub_mixin
Inference Endpoints
Model card Files Community
sample-tri-BimanualPutRedBellPepperInBin
File size: 4,219 Bytes 
70ba90f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
 
resume: false
device: cuda
use_amp: false
seed: 100000
dataset_repo_id: notmahi/tutorial-tri-BimanualPutRedBellPepperInBin
video_backend: pyav
training:
  offline_steps: 20000
  online_steps: 0
  online_steps_between_rollouts: 1
  online_sampling_ratio: 0.5
  online_env_seed: ???
  eval_freq: 0
  log_freq: 250
  save_checkpoint: true
  save_freq: 5000
  num_workers: 4
  batch_size: 32
  image_transforms:
    enable: false
    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.wrist_right_minus:
    - -0.03333333333333333
    - 0.0
    observation.images.wrist_left_plus:
    - -0.03333333333333333
    - 0.0
    observation.images.scene_right_0:
    - -0.03333333333333333
    - 0.0
    observation.images.scene_left_0:
    - -0.03333333333333333
    - 0.0
    observation.state:
    - -0.03333333333333333
    - 0.0
    action:
    - -0.03333333333333333
    - 0.0
    - 0.03333333333333333
    - 0.06666666666666667
    - 0.1
    - 0.13333333333333333
    - 0.16666666666666666
    - 0.2
    - 0.23333333333333334
    - 0.26666666666666666
    - 0.3
    - 0.3333333333333333
    - 0.36666666666666664
    - 0.4
    - 0.43333333333333335
    - 0.4666666666666667
  drop_n_last_frames: 7
eval:
  n_episodes: 50
  batch_size: 50
  use_async_envs: false
wandb:
  enable: false
  disable_artifact: false
  project: lerobot
  notes: ''
fps: 30
env:
  name: tri
  task: PutSpatulaInUtensilCrock-v0
  state_dim: 20
  action_dim: 14
  fps: ${fps}
  episode_length: 400
  gym:
    fps: ${fps}
override_dataset_stats:
  observation.images.wrist_right_minus:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.wrist_left_plus:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.scene_right_0:
    mean:
    - - - 0.485
    - - - 0.456
    - - - 0.406
    std:
    - - - 0.229
    - - - 0.224
    - - - 0.225
  observation.images.scene_left_0:
    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.wrist_right_minus:
    - 3
    - 480
    - 640
    observation.images.wrist_left_plus:
    - 3
    - 480
    - 640
    observation.images.scene_right_0:
    - 3
    - 480
    - 640
    observation.images.scene_left_0:
    - 3
    - 480
    - 640
    observation.state:
    - ${env.state_dim}
  output_shapes:
    action:
    - ${env.action_dim}
  input_normalization_modes:
    observation.images.wrist_right_minus: mean_std
    observation.images.wrist_left_plus: mean_std
    observation.images.scene_right_0: mean_std
    observation.images.scene_left_0: mean_std
    observation.state: min_max
  output_normalization_modes:
    action: min_max
  vision_backbone: vit_b_16
  pretrained_backbone_weights: IMAGENET1K_SWAG_LINEAR_V1
  resize_shape:
  - 256
  - 342
  crop_shape:
  - 224
  - 224
  crop_is_random: true
  use_spatial_softmax: false
  use_group_norm: true
  spatial_softmax_num_keypoints: 32
  down_dims:
  - 256
  - 512
  - 1024
  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: null
  use_ema: true
  ema_update_after_step: 0
  ema_min_alpha: 0.0
  ema_max_alpha: 0.9999
  ema_inv_gamma: 1.0
  ema_power: 0.75
  do_mask_loss_for_padding: false