ralfroemer commited on
Commit
4c9ca58
·
verified ·
1 Parent(s): dada7dc

real_random_eps10_step10000 round 0 member 1

Browse files
real_random_eps10_step10000/round_000/member_01/config.json ADDED
@@ -0,0 +1,123 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "smolvla",
3
+ "n_obs_steps": 1,
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:0",
61
+ "use_amp": false,
62
+ "push_to_hub": true,
63
+ "repo_id": null,
64
+ "private": null,
65
+ "tags": null,
66
+ "license": null,
67
+ "pretrained_path": "${STORAGE_ROOT}/outputs/pretrain/hardware_pretrain_v2/seed_0/checkpoints/last/pretrained_model",
68
+ "chunk_size": 50,
69
+ "n_action_steps": 50,
70
+ "normalization_mapping": {
71
+ "VISUAL": "IDENTITY",
72
+ "STATE": "MEAN_STD",
73
+ "ACTION": "MEAN_STD"
74
+ },
75
+ "max_state_dim": 32,
76
+ "max_action_dim": 32,
77
+ "state_feature_indices": [
78
+ 0,
79
+ 1,
80
+ 2,
81
+ 3,
82
+ 4,
83
+ 5,
84
+ 6
85
+ ],
86
+ "resize_imgs_with_padding": [
87
+ 512,
88
+ 512
89
+ ],
90
+ "empty_cameras": 0,
91
+ "adapt_to_pi_aloha": false,
92
+ "use_delta_joint_actions_aloha": false,
93
+ "tokenizer_max_length": 48,
94
+ "num_steps": 10,
95
+ "use_cache": true,
96
+ "freeze_vision_encoder": false,
97
+ "freeze_text_model": false,
98
+ "train_state_proj": true,
99
+ "optimizer_lr": 5e-05,
100
+ "optimizer_betas": [
101
+ 0.9,
102
+ 0.95
103
+ ],
104
+ "optimizer_eps": 1e-08,
105
+ "optimizer_weight_decay": 1e-10,
106
+ "optimizer_grad_clip_norm": 10,
107
+ "scheduler_warmup_steps": 500,
108
+ "scheduler_decay_steps": 10000,
109
+ "scheduler_decay_lr": 5e-06,
110
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
111
+ "load_vlm_weights": true,
112
+ "add_image_special_tokens": false,
113
+ "attention_mode": "cross_attn",
114
+ "prefix_length": 0,
115
+ "pad_language_to": "max_length",
116
+ "num_expert_layers": 0,
117
+ "num_vlm_layers": 16,
118
+ "self_attn_every_n_layers": 2,
119
+ "expert_width_multiplier": 0.75,
120
+ "min_period": 0.004,
121
+ "max_period": 4.0,
122
+ "rtc_config": null
123
+ }
real_random_eps10_step10000/round_000/member_01/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dea2b59aada9e31bf5037c5e646635de0862013588233bda7489b14e411e0b5c
3
+ size 906712520
real_random_eps10_step10000/round_000/member_01/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": "IDENTITY",
18
+ "STATE": "MEAN_STD",
19
+ "ACTION": "MEAN_STD"
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
+ }
real_random_eps10_step10000/round_000/member_01/policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01b15e422b0661f1cdffeb2214f0dc0110ff63dd7b67b874e787784351ba4387
3
+ size 6060
real_random_eps10_step10000/round_000/member_01/policy_preprocessor.json ADDED
@@ -0,0 +1,103 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "smolvla_new_line_processor",
16
+ "config": {}
17
+ },
18
+ {
19
+ "registry_name": "tokenizer_processor",
20
+ "config": {
21
+ "max_length": 48,
22
+ "task_key": "task",
23
+ "padding_side": "right",
24
+ "padding": "max_length",
25
+ "truncation": true,
26
+ "tokenizer_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct"
27
+ }
28
+ },
29
+ {
30
+ "registry_name": "device_processor",
31
+ "config": {
32
+ "device": "cuda",
33
+ "float_dtype": null
34
+ }
35
+ },
36
+ {
37
+ "registry_name": "normalizer_processor",
38
+ "config": {
39
+ "eps": 1e-08,
40
+ "features": {
41
+ "observation.images.primary": {
42
+ "type": "VISUAL",
43
+ "shape": [
44
+ 3,
45
+ 256,
46
+ 256
47
+ ]
48
+ },
49
+ "observation.images.wrist": {
50
+ "type": "VISUAL",
51
+ "shape": [
52
+ 3,
53
+ 256,
54
+ 256
55
+ ]
56
+ },
57
+ "observation.state.cartesian": {
58
+ "type": "STATE",
59
+ "shape": [
60
+ 6
61
+ ]
62
+ },
63
+ "observation.state.gripper": {
64
+ "type": "STATE",
65
+ "shape": [
66
+ 1
67
+ ]
68
+ },
69
+ "observation.state.joints": {
70
+ "type": "STATE",
71
+ "shape": [
72
+ 7
73
+ ]
74
+ },
75
+ "observation.state.target": {
76
+ "type": "STATE",
77
+ "shape": [
78
+ 6
79
+ ]
80
+ },
81
+ "observation.state": {
82
+ "type": "STATE",
83
+ "shape": [
84
+ 20
85
+ ]
86
+ },
87
+ "action": {
88
+ "type": "ACTION",
89
+ "shape": [
90
+ 7
91
+ ]
92
+ }
93
+ },
94
+ "norm_map": {
95
+ "VISUAL": "IDENTITY",
96
+ "STATE": "MEAN_STD",
97
+ "ACTION": "MEAN_STD"
98
+ }
99
+ },
100
+ "state_file": "policy_preprocessor_step_5_normalizer_processor.safetensors"
101
+ }
102
+ ]
103
+ }
real_random_eps10_step10000/round_000/member_01/policy_preprocessor_step_5_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01b15e422b0661f1cdffeb2214f0dc0110ff63dd7b67b874e787784351ba4387
3
+ size 6060
real_random_eps10_step10000/round_000/member_01/train_config.json ADDED
@@ -0,0 +1,218 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "continuallearning/al_train_pretrainv2_real014",
4
+ "root": null,
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
8
+ "max_num_transforms": 1,
9
+ "random_order": false,
10
+ "tfs": {
11
+ "crop": {
12
+ "weight": 1.0,
13
+ "type": "RandomResizedCrop",
14
+ "kwargs": {
15
+ "size": [
16
+ 256,
17
+ 256
18
+ ],
19
+ "scale": [
20
+ 0.9,
21
+ 1.0
22
+ ],
23
+ "ratio": [
24
+ 0.9,
25
+ 1.1
26
+ ]
27
+ }
28
+ }
29
+ }
30
+ },
31
+ "revision": null,
32
+ "use_imagenet_stats": true,
33
+ "video_backend": "torchcodec",
34
+ "streaming": false,
35
+ "libero_tasks": null,
36
+ "metaworld_tasks": null,
37
+ "real_tasks": {
38
+ "0_put_bowl": "continuallearning/real_0_put_bowl_filtered_consolidated",
39
+ "1_stack_bowls": "continuallearning/real_1_stack_bowls_filtered_consolidated",
40
+ "4_put_lego_into_drawer": "continuallearning/real_4_put_lego_into_drawer_filtered_consolidated"
41
+ },
42
+ "train_repo_id": "continuallearning/al_train_pretrainv2_real014"
43
+ },
44
+ "env": null,
45
+ "policy": {
46
+ "type": "smolvla",
47
+ "n_obs_steps": 1,
48
+ "input_features": {
49
+ "observation.images.primary": {
50
+ "type": "VISUAL",
51
+ "shape": [
52
+ 3,
53
+ 256,
54
+ 256
55
+ ]
56
+ },
57
+ "observation.images.wrist": {
58
+ "type": "VISUAL",
59
+ "shape": [
60
+ 3,
61
+ 256,
62
+ 256
63
+ ]
64
+ },
65
+ "observation.state.cartesian": {
66
+ "type": "STATE",
67
+ "shape": [
68
+ 6
69
+ ]
70
+ },
71
+ "observation.state.gripper": {
72
+ "type": "STATE",
73
+ "shape": [
74
+ 1
75
+ ]
76
+ },
77
+ "observation.state.joints": {
78
+ "type": "STATE",
79
+ "shape": [
80
+ 7
81
+ ]
82
+ },
83
+ "observation.state.target": {
84
+ "type": "STATE",
85
+ "shape": [
86
+ 6
87
+ ]
88
+ },
89
+ "observation.state": {
90
+ "type": "STATE",
91
+ "shape": [
92
+ 20
93
+ ]
94
+ }
95
+ },
96
+ "output_features": {
97
+ "action": {
98
+ "type": "ACTION",
99
+ "shape": [
100
+ 7
101
+ ]
102
+ }
103
+ },
104
+ "device": "cuda:0",
105
+ "use_amp": false,
106
+ "push_to_hub": true,
107
+ "repo_id": null,
108
+ "private": null,
109
+ "tags": null,
110
+ "license": null,
111
+ "pretrained_path": "/cephhome/euralfr/projects/active_learning_vlas/outputs/pretrain/hardware_pretrain_v2/seed_1/checkpoints/last/pretrained_model",
112
+ "chunk_size": 50,
113
+ "n_action_steps": 50,
114
+ "normalization_mapping": {
115
+ "VISUAL": "IDENTITY",
116
+ "STATE": "MEAN_STD",
117
+ "ACTION": "MEAN_STD"
118
+ },
119
+ "max_state_dim": 32,
120
+ "max_action_dim": 32,
121
+ "state_feature_indices": [
122
+ 0,
123
+ 1,
124
+ 2,
125
+ 3,
126
+ 4,
127
+ 5,
128
+ 6
129
+ ],
130
+ "resize_imgs_with_padding": [
131
+ 512,
132
+ 512
133
+ ],
134
+ "empty_cameras": 0,
135
+ "adapt_to_pi_aloha": false,
136
+ "use_delta_joint_actions_aloha": false,
137
+ "tokenizer_max_length": 48,
138
+ "num_steps": 10,
139
+ "use_cache": true,
140
+ "freeze_vision_encoder": false,
141
+ "freeze_text_model": false,
142
+ "train_state_proj": true,
143
+ "optimizer_lr": 5e-05,
144
+ "optimizer_betas": [
145
+ 0.9,
146
+ 0.95
147
+ ],
148
+ "optimizer_eps": 1e-08,
149
+ "optimizer_weight_decay": 1e-10,
150
+ "optimizer_grad_clip_norm": 10,
151
+ "scheduler_warmup_steps": 500,
152
+ "scheduler_decay_steps": 10000,
153
+ "scheduler_decay_lr": 5e-06,
154
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
155
+ "load_vlm_weights": true,
156
+ "add_image_special_tokens": false,
157
+ "attention_mode": "cross_attn",
158
+ "prefix_length": 0,
159
+ "pad_language_to": "max_length",
160
+ "num_expert_layers": 0,
161
+ "num_vlm_layers": 16,
162
+ "self_attn_every_n_layers": 2,
163
+ "expert_width_multiplier": 0.75,
164
+ "min_period": 0.004,
165
+ "max_period": 4.0,
166
+ "rtc_config": null
167
+ },
168
+ "output_dir": "/cephhome/euralfr/projects/active_learning_vlas/outputs/iterative_fine_tuning/real_random_eps10_step10000/round_000/training/member_01",
169
+ "job_name": "real_random_eps10_step10000_round_000_member_01",
170
+ "resume": false,
171
+ "reinitialize_selected_layers": false,
172
+ "seed": 1001,
173
+ "num_workers": 8,
174
+ "batch_size": 32,
175
+ "steps": 10000,
176
+ "eval_freq": 0,
177
+ "log_freq": 20,
178
+ "save_checkpoint": true,
179
+ "save_freq": 20000,
180
+ "enable_val_loss": true,
181
+ "val_ratio": 0.1,
182
+ "val_freq": null,
183
+ "use_policy_training_preset": true,
184
+ "optimizer": {
185
+ "type": "adamw",
186
+ "lr": 5e-05,
187
+ "weight_decay": 1e-10,
188
+ "grad_clip_norm": 10,
189
+ "betas": [
190
+ 0.9,
191
+ 0.95
192
+ ],
193
+ "eps": 1e-08
194
+ },
195
+ "scheduler": {
196
+ "type": "cosine_decay_with_warmup",
197
+ "num_warmup_steps": 500,
198
+ "num_decay_steps": 10000,
199
+ "peak_lr": 5e-05,
200
+ "decay_lr": 5e-06
201
+ },
202
+ "eval": {
203
+ "n_episodes": 1,
204
+ "batch_size": 1,
205
+ "use_async_envs": false
206
+ },
207
+ "wandb": {
208
+ "enable": true,
209
+ "disable_artifact": true,
210
+ "project": "active-vla",
211
+ "entity": "lsy-tum",
212
+ "notes": null,
213
+ "run_id": "tbjk9s3t",
214
+ "mode": "online"
215
+ },
216
+ "checkpoint_path": null,
217
+ "rename_map": {}
218
+ }