Hugging Face's logo

TRI-ML
/
Foundry-VLA-1.7B-sim

Robotics
vla-foundry
foundry
vla_foundry
vla
diffusion-policy
flow-matching
Model card Files
xet
 Community
Foundry-VLA-1.7B-sim
File size: 4,185 Bytes
ef4a0a8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
action_dim: 20
action_fields:
- robot__action__poses__left::panda__xyz_relative
- robot__action__poses__right::panda__xyz_relative
- robot__action__poses__left::panda__rot_6d_relative
- robot__action__poses__right::panda__rot_6d_relative
- robot__action__grippers__left::panda_hand
- robot__action__grippers__right::panda_hand
allow_multiple_epochs: true
augmentation:
  enabled: true
  image:
    color_jitter:
      brightness: 0.2
      contrast: 0.4
      enabled: true
      hue:
      - -0.05
      - 0.05
      saturation: 0.2
    crop:
      enabled: true
      mode: random
      shape:
      - 224
      - 224
  point_cloud:
    color_jitter:
      brightness: 0.2
      contrast: 0.4
      enabled: false
      hue:
      - -0.05
      - 0.05
      saturation: 0.2
camera_names:
- scene_right_0
- scene_left_0
- wrist_left_plus
- wrist_right_minus
dataloader_in_order: false
dataset_cache:
  cache_dir: null
  cache_size_gb: null
  cache_verbose: null
  enabled: false
dataset_manifest:
- null
dataset_modality:
- robotics
dataset_statistics:
- null
dataset_weighting:
- 1.0
extrinsics_fields:
- extrinsics.scene_right_0
- extrinsics.scene_left_0
- extrinsics.wrist_left_minus
- extrinsics.wrist_left_plus
- extrinsics.wrist_right_minus
- extrinsics.wrist_right_plus
hf_fast_tokenizer_rayon_threads: 16
hf_fast_tokenizers_parallelism: true
image_indices:
- -1
- 0
image_names:
- scene_right_0_t-1
- scene_left_0_t-1
- wrist_left_plus_t-1
- wrist_right_minus_t-1
- scene_right_0_t0
- scene_left_0_t0
- wrist_left_plus_t0
- wrist_right_minus_t0
image_size: 224
img_num_tokens: 64
intrinsics_fields:
- intrinsics.scene_right_0
- intrinsics.scene_left_0
- intrinsics.wrist_left_minus
- intrinsics.wrist_left_plus
- intrinsics.wrist_right_minus
- intrinsics.wrist_right_plus
language_instruction_types:
- original
- randomized
- verbose
- alternative
lowdim_future_timesteps: 8
lowdim_past_timesteps: 1
mask_padded_images: true
max_text_seq_len: null
normalization:
  centered_norm: true
  enabled: true
  epsilon: 0.01
  field_configs:
    robot__action__grippers__left::panda_hand:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
    robot__action__grippers__right::panda_hand:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
    robot__action__poses__left::panda__rot_6d_relative:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
    robot__action__poses__left::panda__xyz_relative:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
    robot__action__poses__right::panda__rot_6d_relative:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
    robot__action__poses__right::panda__xyz_relative:
      enabled: true
      epsilon: 0.01
      method: percentile_1_99
      scope: per_timestep
  include_fields:
  - robot__action__poses__left::panda__xyz_relative
  - robot__action__poses__right::panda__xyz_relative
  - robot__action__poses__left::panda__rot_6d_relative
  - robot__action__poses__right::panda__rot_6d_relative
  - robot__action__grippers__left::panda_hand
  - robot__action__grippers__right::panda_hand
  lowdim_future_timesteps: 19
  lowdim_past_timesteps: 5
  method: percentile_1_99
  scope: per_timestep
num_workers: 8
pad_missing_images: true
point_cloud_num_points: 4096
pose_groups:
- name: left_panda_action
  position_key: robot__action__poses__left::panda__xyz
  rotation_key: robot__action__poses__left::panda__rot_6d
- name: right_panda_action
  position_key: robot__action__poses__right::panda__xyz
  rotation_key: robot__action__poses__right::panda__rot_6d
prefetch_factor: 8
processor: HuggingFaceTB/SmolVLM2-256M-Video-Instruct
processor_kwargs:
  max_image_size:
    longest_edge: 224
  size:
    longest_edge: 224
proprioception_dim: 0
proprioception_fields: []
seed: 42
seq_len: 2048
shuffle: true
shuffle_buffer_size: 2000
shuffle_initial: 500
tactile_fields: []
type: robotics
use_hf_fast_tokenizer: true
use_point_cloud: false
val_dataset_manifest: []
val_dataset_statistics: []
val_dataset_weighting: []