continuallearning commited on
Commit
6bd215a
·
verified ·
1 Parent(s): 36e4a3e

Upload folder using huggingface_hub

Browse files
config.json ADDED
@@ -0,0 +1,112 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "dit",
3
+ "n_obs_steps": 2,
4
+ "input_features": {
5
+ "observation.images.primary": {
6
+ "type": "VISUAL",
7
+ "shape": [
8
+ 3,
9
+ 256,
10
+ 256
11
+ ]
12
+ },
13
+ "observation.images.wrist": {
14
+ "type": "VISUAL",
15
+ "shape": [
16
+ 3,
17
+ 256,
18
+ 256
19
+ ]
20
+ },
21
+ "observation.state.cartesian": {
22
+ "type": "STATE",
23
+ "shape": [
24
+ 6
25
+ ]
26
+ },
27
+ "observation.state.gripper": {
28
+ "type": "STATE",
29
+ "shape": [
30
+ 1
31
+ ]
32
+ },
33
+ "observation.state.joints": {
34
+ "type": "STATE",
35
+ "shape": [
36
+ 7
37
+ ]
38
+ },
39
+ "observation.state.target": {
40
+ "type": "STATE",
41
+ "shape": [
42
+ 6
43
+ ]
44
+ },
45
+ "observation.state": {
46
+ "type": "STATE",
47
+ "shape": [
48
+ 20
49
+ ]
50
+ }
51
+ },
52
+ "output_features": {
53
+ "action": {
54
+ "type": "ACTION",
55
+ "shape": [
56
+ 7
57
+ ]
58
+ }
59
+ },
60
+ "device": "cuda",
61
+ "use_amp": false,
62
+ "use_peft": false,
63
+ "push_to_hub": true,
64
+ "repo_id": "continuallearning/dit_fft_pretraining_v1_seed1000",
65
+ "private": null,
66
+ "tags": null,
67
+ "license": null,
68
+ "pretrained_path": null,
69
+ "horizon": 16,
70
+ "n_action_steps": 8,
71
+ "normalization_mapping": {
72
+ "VISUAL": "MEAN_STD",
73
+ "STATE": "MIN_MAX",
74
+ "ACTION": "MIN_MAX"
75
+ },
76
+ "drop_n_last_frames": 7,
77
+ "tokenizer_max_length": 48,
78
+ "language_model_name": "openai/clip-vit-base-patch32",
79
+ "freeze_language_pretrained": true,
80
+ "vit_name": "facebook/dinov2-base",
81
+ "frequency_embedding_dim": 256,
82
+ "hidden_dim": 512,
83
+ "num_blocks": 6,
84
+ "num_heads": 16,
85
+ "dropout": 0.1,
86
+ "dim_feedforward": 4096,
87
+ "activation": "gelu",
88
+ "noise_type": "flow_matching",
89
+ "training_noise_sampling": "uniform",
90
+ "num_train_timesteps": 100,
91
+ "beta_schedule": "squaredcos_cap_v2",
92
+ "beta_start": 0.0001,
93
+ "beta_end": 0.02,
94
+ "prediction_type": "epsilon",
95
+ "set_alpha_to_one": true,
96
+ "steps_offset": 0,
97
+ "clip_sample": true,
98
+ "clip_sample_range": 1.0,
99
+ "use_encoder": false,
100
+ "n_encoder_layers": 6,
101
+ "num_inference_steps": 100,
102
+ "do_mask_loss_for_padding": false,
103
+ "optimizer_lr": 0.0001,
104
+ "optimizer_betas": [
105
+ 0.95,
106
+ 0.999
107
+ ],
108
+ "optimizer_eps": 1e-08,
109
+ "optimizer_weight_decay": 1e-06,
110
+ "scheduler_name": "cosine",
111
+ "scheduler_warmup_steps": 500
112
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17138a383e9cfeba7d163730aed269de11cd1b852eee1323ef6672c0658025e5
3
+ size 774622540
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
+ 7
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:d445019adcddba13beace30e04f099a49c9ea9a90f500b2717d0629bdf0a767a
3
+ size 6060
policy_preprocessor.json ADDED
@@ -0,0 +1,88 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "device_processor",
16
+ "config": {
17
+ "device": "cuda",
18
+ "float_dtype": null
19
+ }
20
+ },
21
+ {
22
+ "registry_name": "normalizer_processor",
23
+ "config": {
24
+ "eps": 1e-08,
25
+ "features": {
26
+ "observation.images.primary": {
27
+ "type": "VISUAL",
28
+ "shape": [
29
+ 3,
30
+ 256,
31
+ 256
32
+ ]
33
+ },
34
+ "observation.images.wrist": {
35
+ "type": "VISUAL",
36
+ "shape": [
37
+ 3,
38
+ 256,
39
+ 256
40
+ ]
41
+ },
42
+ "observation.state.cartesian": {
43
+ "type": "STATE",
44
+ "shape": [
45
+ 6
46
+ ]
47
+ },
48
+ "observation.state.gripper": {
49
+ "type": "STATE",
50
+ "shape": [
51
+ 1
52
+ ]
53
+ },
54
+ "observation.state.joints": {
55
+ "type": "STATE",
56
+ "shape": [
57
+ 7
58
+ ]
59
+ },
60
+ "observation.state.target": {
61
+ "type": "STATE",
62
+ "shape": [
63
+ 6
64
+ ]
65
+ },
66
+ "observation.state": {
67
+ "type": "STATE",
68
+ "shape": [
69
+ 20
70
+ ]
71
+ },
72
+ "action": {
73
+ "type": "ACTION",
74
+ "shape": [
75
+ 7
76
+ ]
77
+ }
78
+ },
79
+ "norm_map": {
80
+ "VISUAL": "MEAN_STD",
81
+ "STATE": "MIN_MAX",
82
+ "ACTION": "MIN_MAX"
83
+ }
84
+ },
85
+ "state_file": "policy_preprocessor_step_3_normalizer_processor.safetensors"
86
+ }
87
+ ]
88
+ }
policy_preprocessor_step_3_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d445019adcddba13beace30e04f099a49c9ea9a90f500b2717d0629bdf0a767a
3
+ size 6060
train_config.json ADDED
@@ -0,0 +1,246 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "continuallearning/pretraining_v1",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": false,
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": "dit",
85
+ "n_obs_steps": 2,
86
+ "input_features": {
87
+ "observation.images.primary": {
88
+ "type": "VISUAL",
89
+ "shape": [
90
+ 3,
91
+ 256,
92
+ 256
93
+ ]
94
+ },
95
+ "observation.images.wrist": {
96
+ "type": "VISUAL",
97
+ "shape": [
98
+ 3,
99
+ 256,
100
+ 256
101
+ ]
102
+ },
103
+ "observation.state.cartesian": {
104
+ "type": "STATE",
105
+ "shape": [
106
+ 6
107
+ ]
108
+ },
109
+ "observation.state.gripper": {
110
+ "type": "STATE",
111
+ "shape": [
112
+ 1
113
+ ]
114
+ },
115
+ "observation.state.joints": {
116
+ "type": "STATE",
117
+ "shape": [
118
+ 7
119
+ ]
120
+ },
121
+ "observation.state.target": {
122
+ "type": "STATE",
123
+ "shape": [
124
+ 6
125
+ ]
126
+ },
127
+ "observation.state": {
128
+ "type": "STATE",
129
+ "shape": [
130
+ 20
131
+ ]
132
+ }
133
+ },
134
+ "output_features": {
135
+ "action": {
136
+ "type": "ACTION",
137
+ "shape": [
138
+ 7
139
+ ]
140
+ }
141
+ },
142
+ "device": "cuda",
143
+ "use_amp": false,
144
+ "use_peft": false,
145
+ "push_to_hub": true,
146
+ "repo_id": "continuallearning/dit_fft_pretraining_v1_seed1000",
147
+ "private": null,
148
+ "tags": null,
149
+ "license": null,
150
+ "pretrained_path": null,
151
+ "horizon": 16,
152
+ "n_action_steps": 8,
153
+ "normalization_mapping": {
154
+ "VISUAL": "MEAN_STD",
155
+ "STATE": "MIN_MAX",
156
+ "ACTION": "MIN_MAX"
157
+ },
158
+ "drop_n_last_frames": 7,
159
+ "tokenizer_max_length": 48,
160
+ "language_model_name": "openai/clip-vit-base-patch32",
161
+ "freeze_language_pretrained": true,
162
+ "vit_name": "facebook/dinov2-base",
163
+ "frequency_embedding_dim": 256,
164
+ "hidden_dim": 512,
165
+ "num_blocks": 6,
166
+ "num_heads": 16,
167
+ "dropout": 0.1,
168
+ "dim_feedforward": 4096,
169
+ "activation": "gelu",
170
+ "noise_type": "flow_matching",
171
+ "training_noise_sampling": "uniform",
172
+ "num_train_timesteps": 100,
173
+ "beta_schedule": "squaredcos_cap_v2",
174
+ "beta_start": 0.0001,
175
+ "beta_end": 0.02,
176
+ "prediction_type": "epsilon",
177
+ "set_alpha_to_one": true,
178
+ "steps_offset": 0,
179
+ "clip_sample": true,
180
+ "clip_sample_range": 1.0,
181
+ "use_encoder": false,
182
+ "n_encoder_layers": 6,
183
+ "num_inference_steps": 100,
184
+ "do_mask_loss_for_padding": false,
185
+ "optimizer_lr": 0.0001,
186
+ "optimizer_betas": [
187
+ 0.95,
188
+ 0.999
189
+ ],
190
+ "optimizer_eps": 1e-08,
191
+ "optimizer_weight_decay": 1e-06,
192
+ "scheduler_name": "cosine",
193
+ "scheduler_warmup_steps": 500
194
+ },
195
+ "output_dir": "outputs/train/dit_fft_pretraining_v1_s1000",
196
+ "job_name": "dit_fft_pretraining_v1_s1000",
197
+ "resume": false,
198
+ "seed": 1000,
199
+ "num_workers": 16,
200
+ "batch_size": 128,
201
+ "steps": 200000,
202
+ "eval_freq": 0,
203
+ "log_freq": 100,
204
+ "tolerance_s": 0.0001,
205
+ "save_checkpoint": true,
206
+ "save_freq": 20000,
207
+ "use_policy_training_preset": true,
208
+ "optimizer": {
209
+ "type": "adam",
210
+ "lr": 0.0001,
211
+ "weight_decay": 1e-06,
212
+ "grad_clip_norm": 10.0,
213
+ "betas": [
214
+ 0.95,
215
+ 0.999
216
+ ],
217
+ "eps": 1e-08
218
+ },
219
+ "scheduler": {
220
+ "type": "diffuser",
221
+ "num_warmup_steps": 500,
222
+ "name": "cosine"
223
+ },
224
+ "eval": {
225
+ "n_episodes": 50,
226
+ "batch_size": 50,
227
+ "use_async_envs": false
228
+ },
229
+ "wandb": {
230
+ "enable": true,
231
+ "disable_artifact": true,
232
+ "project": "clare_rebuttal",
233
+ "entity": "470620104-technical-university-of-munich",
234
+ "notes": null,
235
+ "run_id": "coj580rq",
236
+ "mode": null
237
+ },
238
+ "peft": null,
239
+ "use_rabc": false,
240
+ "rabc_progress_path": null,
241
+ "rabc_kappa": 0.01,
242
+ "rabc_epsilon": 1e-06,
243
+ "rabc_head_mode": "sparse",
244
+ "rename_map": {},
245
+ "checkpoint_path": null
246
+ }