Hugging Face's logo

naavox
/
dit-grasp-3

Robotics
LeRobot
Safetensors
multi_task_dit
Model card Files
xet
 Community
naavox commited on
Commit
37fe883
·
verified ·
1 Parent(s): 3773170

Upload folder using huggingface_hub

Browse files
Files changed (7) hide show
  1. config.json +94 -0
  2. model.safetensors +3 -0
  3. policy_postprocessor.json +32 -0
  4. policy_postprocessor_step_0_unnormalizer_processor.safetensors +3 -0
  5. policy_preprocessor.json +67 -0
  6. policy_preprocessor_step_4_normalizer_processor.safetensors +3 -0
  7. train_config.json +230 -0
config.json ADDED
@@ -0,0 +1,94 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "multi_task_dit",
3
+ "n_obs_steps": 2,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 43
9
+ ]
10
+ },
11
+ "observation.images.gripper_camera": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 224,
16
+ 224
17
+ ]
18
+ }
19
+ },
20
+ "output_features": {
21
+ "action": {
22
+ "type": "ACTION",
23
+ "shape": [
24
+ 5
25
+ ]
26
+ }
27
+ },
28
+ "device": "cuda",
29
+ "use_amp": false,
30
+ "use_peft": false,
31
+ "push_to_hub": true,
32
+ "repo_id": "naavox/dit-grasp-3",
33
+ "private": null,
34
+ "tags": null,
35
+ "license": null,
36
+ "pretrained_path": "naavox/dit-grasp-1",
37
+ "horizon": 64,
38
+ "n_action_steps": 32,
39
+ "objective": "flow_matching",
40
+ "noise_scheduler_type": "DDPM",
41
+ "num_train_timesteps": 100,
42
+ "beta_schedule": "squaredcos_cap_v2",
43
+ "beta_start": 0.0001,
44
+ "beta_end": 0.02,
45
+ "prediction_type": "epsilon",
46
+ "clip_sample": true,
47
+ "clip_sample_range": 1.0,
48
+ "num_inference_steps": null,
49
+ "sigma_min": 0.0,
50
+ "num_integration_steps": 100,
51
+ "integration_method": "euler",
52
+ "timestep_sampling_strategy": "beta",
53
+ "timestep_sampling_s": 0.999,
54
+ "timestep_sampling_alpha": 1.5,
55
+ "timestep_sampling_beta": 1.0,
56
+ "hidden_dim": 512,
57
+ "num_layers": 6,
58
+ "num_heads": 8,
59
+ "dropout": 0.1,
60
+ "use_positional_encoding": false,
61
+ "timestep_embed_dim": 256,
62
+ "use_rope": true,
63
+ "rope_base": 10000.0,
64
+ "vision_encoder_name": "openai/clip-vit-base-patch16",
65
+ "use_separate_rgb_encoder_per_camera": false,
66
+ "vision_encoder_lr_multiplier": 0.1,
67
+ "image_resize_shape": null,
68
+ "image_crop_shape": [
69
+ 224,
70
+ 224
71
+ ],
72
+ "image_crop_is_random": true,
73
+ "text_encoder_name": "openai/clip-vit-base-patch16",
74
+ "tokenizer_max_length": 77,
75
+ "tokenizer_padding": "max_length",
76
+ "tokenizer_padding_side": "right",
77
+ "tokenizer_truncation": true,
78
+ "normalization_mapping": {
79
+ "VISUAL": "MEAN_STD",
80
+ "STATE": "MIN_MAX",
81
+ "ACTION": "MIN_MAX"
82
+ },
83
+ "optimizer_lr": 2e-05,
84
+ "optimizer_betas": [
85
+ 0.95,
86
+ 0.999
87
+ ],
88
+ "optimizer_eps": 1e-08,
89
+ "optimizer_weight_decay": 0.0,
90
+ "scheduler_name": "cosine",
91
+ "scheduler_warmup_steps": 0,
92
+ "do_mask_loss_for_padding": false,
93
+ "drop_n_last_frames": 31
94
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc2e54dd8fce6a27ae5fbd1d0515bca6e80215bf64761289d0e8d80297551c99
3
+ size 887699652
policy_postprocessor.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_postprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "unnormalizer_processor",
6
+ "config": {
7
+ "eps": 1e-08,
8
+ "features": {
9
+ "action": {
10
+ "type": "ACTION",
11
+ "shape": [
12
+ 5
13
+ ]
14
+ }
15
+ },
16
+ "norm_map": {
17
+ "VISUAL": "MEAN_STD",
18
+ "STATE": "MIN_MAX",
19
+ "ACTION": "MIN_MAX"
20
+ }
21
+ },
22
+ "state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
23
+ },
24
+ {
25
+ "registry_name": "device_processor",
26
+ "config": {
27
+ "device": "cpu",
28
+ "float_dtype": null
29
+ }
30
+ }
31
+ ]
32
+ }
policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90188879a871c5c48555ad51d975ef028ab546db4cd742f18d1f9190a5e374d9
3
+ size 8008
policy_preprocessor.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_preprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "rename_observations_processor",
6
+ "config": {
7
+ "rename_map": {}
8
+ }
9
+ },
10
+ {
11
+ "registry_name": "to_batch_processor",
12
+ "config": {}
13
+ },
14
+ {
15
+ "registry_name": "tokenizer_processor",
16
+ "config": {
17
+ "max_length": 77,
18
+ "task_key": "task",
19
+ "padding_side": "right",
20
+ "padding": "max_length",
21
+ "truncation": true,
22
+ "tokenizer_name": "openai/clip-vit-base-patch16"
23
+ }
24
+ },
25
+ {
26
+ "registry_name": "device_processor",
27
+ "config": {
28
+ "device": "cuda",
29
+ "float_dtype": null
30
+ }
31
+ },
32
+ {
33
+ "registry_name": "normalizer_processor",
34
+ "config": {
35
+ "eps": 1e-08,
36
+ "features": {
37
+ "observation.state": {
38
+ "type": "STATE",
39
+ "shape": [
40
+ 43
41
+ ]
42
+ },
43
+ "observation.images.gripper_camera": {
44
+ "type": "VISUAL",
45
+ "shape": [
46
+ 3,
47
+ 224,
48
+ 224
49
+ ]
50
+ },
51
+ "action": {
52
+ "type": "ACTION",
53
+ "shape": [
54
+ 5
55
+ ]
56
+ }
57
+ },
58
+ "norm_map": {
59
+ "VISUAL": "MEAN_STD",
60
+ "STATE": "MIN_MAX",
61
+ "ACTION": "MIN_MAX"
62
+ }
63
+ },
64
+ "state_file": "policy_preprocessor_step_4_normalizer_processor.safetensors"
65
+ }
66
+ ]
67
+ }
policy_preprocessor_step_4_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90188879a871c5c48555ad51d975ef028ab546db4cd742f18d1f9190a5e374d9
3
+ size 8008
train_config.json ADDED
@@ -0,0 +1,230 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "naavox/merged_224",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
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
+ "affine": {
62
+ "weight": 1.0,
63
+ "type": "RandomAffine",
64
+ "kwargs": {
65
+ "degrees": [
66
+ -5.0,
67
+ 5.0
68
+ ],
69
+ "translate": [
70
+ 0.05,
71
+ 0.05
72
+ ]
73
+ }
74
+ }
75
+ }
76
+ },
77
+ "revision": null,
78
+ "use_imagenet_stats": true,
79
+ "video_backend": "torchcodec",
80
+ "streaming": false
81
+ },
82
+ "env": null,
83
+ "policy": {
84
+ "type": "multi_task_dit",
85
+ "n_obs_steps": 2,
86
+ "input_features": {
87
+ "observation.state": {
88
+ "type": "STATE",
89
+ "shape": [
90
+ 43
91
+ ]
92
+ },
93
+ "observation.images.gripper_camera": {
94
+ "type": "VISUAL",
95
+ "shape": [
96
+ 3,
97
+ 224,
98
+ 224
99
+ ]
100
+ }
101
+ },
102
+ "output_features": {
103
+ "action": {
104
+ "type": "ACTION",
105
+ "shape": [
106
+ 5
107
+ ]
108
+ }
109
+ },
110
+ "device": "cuda",
111
+ "use_amp": false,
112
+ "use_peft": false,
113
+ "push_to_hub": true,
114
+ "repo_id": "naavox/dit-grasp-3",
115
+ "private": null,
116
+ "tags": null,
117
+ "license": null,
118
+ "pretrained_path": "naavox/dit-grasp-1",
119
+ "horizon": 64,
120
+ "n_action_steps": 32,
121
+ "objective": "flow_matching",
122
+ "noise_scheduler_type": "DDPM",
123
+ "num_train_timesteps": 100,
124
+ "beta_schedule": "squaredcos_cap_v2",
125
+ "beta_start": 0.0001,
126
+ "beta_end": 0.02,
127
+ "prediction_type": "epsilon",
128
+ "clip_sample": true,
129
+ "clip_sample_range": 1.0,
130
+ "num_inference_steps": null,
131
+ "sigma_min": 0.0,
132
+ "num_integration_steps": 100,
133
+ "integration_method": "euler",
134
+ "timestep_sampling_strategy": "beta",
135
+ "timestep_sampling_s": 0.999,
136
+ "timestep_sampling_alpha": 1.5,
137
+ "timestep_sampling_beta": 1.0,
138
+ "hidden_dim": 512,
139
+ "num_layers": 6,
140
+ "num_heads": 8,
141
+ "dropout": 0.1,
142
+ "use_positional_encoding": false,
143
+ "timestep_embed_dim": 256,
144
+ "use_rope": true,
145
+ "rope_base": 10000.0,
146
+ "vision_encoder_name": "openai/clip-vit-base-patch16",
147
+ "use_separate_rgb_encoder_per_camera": false,
148
+ "vision_encoder_lr_multiplier": 0.1,
149
+ "image_resize_shape": null,
150
+ "image_crop_shape": [
151
+ 224,
152
+ 224
153
+ ],
154
+ "image_crop_is_random": true,
155
+ "text_encoder_name": "openai/clip-vit-base-patch16",
156
+ "tokenizer_max_length": 77,
157
+ "tokenizer_padding": "max_length",
158
+ "tokenizer_padding_side": "right",
159
+ "tokenizer_truncation": true,
160
+ "normalization_mapping": {
161
+ "VISUAL": "MEAN_STD",
162
+ "STATE": "MIN_MAX",
163
+ "ACTION": "MIN_MAX"
164
+ },
165
+ "optimizer_lr": 2e-05,
166
+ "optimizer_betas": [
167
+ 0.95,
168
+ 0.999
169
+ ],
170
+ "optimizer_eps": 1e-08,
171
+ "optimizer_weight_decay": 0.0,
172
+ "scheduler_name": "cosine",
173
+ "scheduler_warmup_steps": 0,
174
+ "do_mask_loss_for_padding": false,
175
+ "drop_n_last_frames": 31
176
+ },
177
+ "output_dir": "/multitask_dit_data/tidy_modal_12",
178
+ "job_name": "multi_task_dit",
179
+ "resume": false,
180
+ "seed": 1000,
181
+ "cudnn_deterministic": false,
182
+ "num_workers": 13,
183
+ "batch_size": 400,
184
+ "steps": 35000,
185
+ "eval_freq": 20000,
186
+ "log_freq": 200,
187
+ "tolerance_s": 0.001,
188
+ "save_checkpoint": true,
189
+ "save_freq": 5000,
190
+ "use_policy_training_preset": true,
191
+ "optimizer": {
192
+ "type": "adam",
193
+ "lr": 2e-05,
194
+ "weight_decay": 0.0,
195
+ "grad_clip_norm": 10.0,
196
+ "betas": [
197
+ 0.95,
198
+ 0.999
199
+ ],
200
+ "eps": 1e-08
201
+ },
202
+ "scheduler": {
203
+ "type": "diffuser",
204
+ "num_warmup_steps": 0,
205
+ "name": "cosine"
206
+ },
207
+ "eval": {
208
+ "n_episodes": 50,
209
+ "batch_size": 50,
210
+ "use_async_envs": false
211
+ },
212
+ "wandb": {
213
+ "enable": false,
214
+ "disable_artifact": false,
215
+ "project": "lerobot",
216
+ "entity": null,
217
+ "notes": null,
218
+ "run_id": null,
219
+ "mode": null,
220
+ "add_tags": true
221
+ },
222
+ "peft": null,
223
+ "use_rabc": false,
224
+ "rabc_progress_path": null,
225
+ "rabc_kappa": 0.01,
226
+ "rabc_epsilon": 1e-06,
227
+ "rabc_head_mode": "sparse",
228
+ "rename_map": {},
229
+ "checkpoint_path": null
230
+ }