Hugging Face's logo

robot-learning-group47
/
eval1_v10_15k_then_fixed93_2k

Safetensors
Model card Files
xet
 Community
nkim7 commited on
Commit
bde85be
·
verified ·
1 Parent(s): bcfbcd1

Upload eval1 v10 15k then fixed93 lightingonly 2k checkpoint

Browse files
Files changed (7) hide show
  1. config.json +117 -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 +105 -0
  6. policy_preprocessor_step_5_normalizer_processor.safetensors +3 -0
  7. train_config.json +213 -0
config.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "type": "smolvla",
3
+ "n_obs_steps": 1,
4
+ "input_features": {
5
+ "observation.state": {
6
+ "type": "STATE",
7
+ "shape": [
8
+ 6
9
+ ]
10
+ },
11
+ "observation.images.camera1": {
12
+ "type": "VISUAL",
13
+ "shape": [
14
+ 3,
15
+ 256,
16
+ 256
17
+ ]
18
+ },
19
+ "observation.images.camera2": {
20
+ "type": "VISUAL",
21
+ "shape": [
22
+ 3,
23
+ 256,
24
+ 256
25
+ ]
26
+ },
27
+ "observation.images.camera3": {
28
+ "type": "VISUAL",
29
+ "shape": [
30
+ 3,
31
+ 256,
32
+ 256
33
+ ]
34
+ },
35
+ "observation.images.empty_camera_0": {
36
+ "type": "VISUAL",
37
+ "shape": [
38
+ 3,
39
+ 480,
40
+ 640
41
+ ]
42
+ },
43
+ "observation.images.empty_camera_1": {
44
+ "type": "VISUAL",
45
+ "shape": [
46
+ 3,
47
+ 480,
48
+ 640
49
+ ]
50
+ }
51
+ },
52
+ "output_features": {
53
+ "action": {
54
+ "type": "ACTION",
55
+ "shape": [
56
+ 6
57
+ ]
58
+ }
59
+ },
60
+ "device": "cuda",
61
+ "use_amp": false,
62
+ "use_peft": false,
63
+ "push_to_hub": true,
64
+ "repo_id": "robot-learning-group47/eval1_v10_15k_then_fixed93_lightingonly",
65
+ "private": null,
66
+ "tags": null,
67
+ "license": null,
68
+ "pretrained_path": "/dev/shm/lerobot_policies/eval1_mix_nooutlier_fp32_noaug_015000/checkpoints/015000/pretrained_model",
69
+ "chunk_size": 50,
70
+ "n_action_steps": 50,
71
+ "normalization_mapping": {
72
+ "VISUAL": "IDENTITY",
73
+ "STATE": "MEAN_STD",
74
+ "ACTION": "MEAN_STD"
75
+ },
76
+ "max_state_dim": 32,
77
+ "max_action_dim": 32,
78
+ "resize_imgs_with_padding": [
79
+ 256,
80
+ 256
81
+ ],
82
+ "empty_cameras": 2,
83
+ "adapt_to_pi_aloha": false,
84
+ "use_delta_joint_actions_aloha": false,
85
+ "tokenizer_max_length": 48,
86
+ "num_steps": 10,
87
+ "use_cache": true,
88
+ "freeze_vision_encoder": true,
89
+ "train_expert_only": true,
90
+ "train_state_proj": true,
91
+ "optimizer_lr": 3e-05,
92
+ "optimizer_betas": [
93
+ 0.9,
94
+ 0.95
95
+ ],
96
+ "optimizer_eps": 1e-08,
97
+ "optimizer_weight_decay": 1e-10,
98
+ "optimizer_grad_clip_norm": 10.0,
99
+ "scheduler_warmup_steps": 500,
100
+ "scheduler_decay_steps": 10000,
101
+ "scheduler_decay_lr": 2.5e-06,
102
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
103
+ "load_vlm_weights": false,
104
+ "add_image_special_tokens": false,
105
+ "attention_mode": "cross_attn",
106
+ "prefix_length": -1,
107
+ "pad_language_to": "longest",
108
+ "num_expert_layers": -1,
109
+ "num_vlm_layers": 16,
110
+ "self_attn_every_n_layers": 2,
111
+ "expert_width_multiplier": 0.75,
112
+ "min_period": 0.004,
113
+ "max_period": 4.0,
114
+ "rtc_config": null,
115
+ "compile_model": false,
116
+ "compile_mode": "max-autotune"
117
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:003e7e2a55173357764dbcb32bf7d7f16b4d5fb194ec3b4268f8d46e3c07e30c
3
+ size 1197789224
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
+ 6
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
+ }
policy_postprocessor_step_0_unnormalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:287442861d5b2ab7b6af78b122213f68e9ed4b12d0cc56bc6d6a7dd92c322ca0
3
+ size 6608
policy_preprocessor.json ADDED
@@ -0,0 +1,105 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "policy_preprocessor",
3
+ "steps": [
4
+ {
5
+ "registry_name": "rename_observations_processor",
6
+ "config": {
7
+ "rename_map": {
8
+ "observation.images.front": "observation.images.camera1"
9
+ }
10
+ }
11
+ },
12
+ {
13
+ "registry_name": "to_batch_processor",
14
+ "config": {}
15
+ },
16
+ {
17
+ "registry_name": "smolvla_new_line_processor",
18
+ "config": {}
19
+ },
20
+ {
21
+ "registry_name": "tokenizer_processor",
22
+ "config": {
23
+ "max_length": 48,
24
+ "task_key": "task",
25
+ "padding_side": "right",
26
+ "padding": "max_length",
27
+ "truncation": true,
28
+ "tokenizer_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct"
29
+ }
30
+ },
31
+ {
32
+ "registry_name": "device_processor",
33
+ "config": {
34
+ "device": "cuda",
35
+ "float_dtype": null
36
+ }
37
+ },
38
+ {
39
+ "registry_name": "normalizer_processor",
40
+ "config": {
41
+ "eps": 1e-08,
42
+ "features": {
43
+ "observation.state": {
44
+ "type": "STATE",
45
+ "shape": [
46
+ 6
47
+ ]
48
+ },
49
+ "observation.images.camera1": {
50
+ "type": "VISUAL",
51
+ "shape": [
52
+ 3,
53
+ 256,
54
+ 256
55
+ ]
56
+ },
57
+ "observation.images.camera2": {
58
+ "type": "VISUAL",
59
+ "shape": [
60
+ 3,
61
+ 256,
62
+ 256
63
+ ]
64
+ },
65
+ "observation.images.camera3": {
66
+ "type": "VISUAL",
67
+ "shape": [
68
+ 3,
69
+ 256,
70
+ 256
71
+ ]
72
+ },
73
+ "observation.images.empty_camera_0": {
74
+ "type": "VISUAL",
75
+ "shape": [
76
+ 3,
77
+ 480,
78
+ 640
79
+ ]
80
+ },
81
+ "observation.images.empty_camera_1": {
82
+ "type": "VISUAL",
83
+ "shape": [
84
+ 3,
85
+ 480,
86
+ 640
87
+ ]
88
+ },
89
+ "action": {
90
+ "type": "ACTION",
91
+ "shape": [
92
+ 6
93
+ ]
94
+ }
95
+ },
96
+ "norm_map": {
97
+ "VISUAL": "IDENTITY",
98
+ "STATE": "MEAN_STD",
99
+ "ACTION": "MEAN_STD"
100
+ }
101
+ },
102
+ "state_file": "policy_preprocessor_step_5_normalizer_processor.safetensors"
103
+ }
104
+ ]
105
+ }
policy_preprocessor_step_5_normalizer_processor.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:287442861d5b2ab7b6af78b122213f68e9ed4b12d0cc56bc6d6a7dd92c322ca0
3
+ size 6608
train_config.json ADDED
@@ -0,0 +1,213 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": {
3
+ "repo_id": "robot-learning-group47/eval1_fixed_merged_remap",
4
+ "root": "/dev/shm/lerobot_datasets/eval1_fixed_merged_remap",
5
+ "episodes": null,
6
+ "image_transforms": {
7
+ "enable": true,
8
+ "max_num_transforms": 2,
9
+ "random_order": false,
10
+ "tfs": {
11
+ "brightness": {
12
+ "weight": 1.0,
13
+ "type": "ColorJitter",
14
+ "kwargs": {
15
+ "brightness": [
16
+ 0.85,
17
+ 1.15
18
+ ]
19
+ }
20
+ },
21
+ "contrast": {
22
+ "weight": 1.0,
23
+ "type": "ColorJitter",
24
+ "kwargs": {
25
+ "contrast": [
26
+ 0.85,
27
+ 1.15
28
+ ]
29
+ }
30
+ }
31
+ }
32
+ },
33
+ "revision": null,
34
+ "use_imagenet_stats": true,
35
+ "video_backend": "torchcodec",
36
+ "return_uint8": false,
37
+ "streaming": false
38
+ },
39
+ "env": null,
40
+ "policy": {
41
+ "type": "smolvla",
42
+ "n_obs_steps": 1,
43
+ "input_features": {
44
+ "observation.state": {
45
+ "type": "STATE",
46
+ "shape": [
47
+ 6
48
+ ]
49
+ },
50
+ "observation.images.camera1": {
51
+ "type": "VISUAL",
52
+ "shape": [
53
+ 3,
54
+ 256,
55
+ 256
56
+ ]
57
+ },
58
+ "observation.images.camera2": {
59
+ "type": "VISUAL",
60
+ "shape": [
61
+ 3,
62
+ 256,
63
+ 256
64
+ ]
65
+ },
66
+ "observation.images.camera3": {
67
+ "type": "VISUAL",
68
+ "shape": [
69
+ 3,
70
+ 256,
71
+ 256
72
+ ]
73
+ },
74
+ "observation.images.empty_camera_0": {
75
+ "type": "VISUAL",
76
+ "shape": [
77
+ 3,
78
+ 480,
79
+ 640
80
+ ]
81
+ },
82
+ "observation.images.empty_camera_1": {
83
+ "type": "VISUAL",
84
+ "shape": [
85
+ 3,
86
+ 480,
87
+ 640
88
+ ]
89
+ }
90
+ },
91
+ "output_features": {
92
+ "action": {
93
+ "type": "ACTION",
94
+ "shape": [
95
+ 6
96
+ ]
97
+ }
98
+ },
99
+ "device": "cuda",
100
+ "use_amp": false,
101
+ "use_peft": false,
102
+ "push_to_hub": true,
103
+ "repo_id": "robot-learning-group47/eval1_v10_15k_then_fixed93_lightingonly",
104
+ "private": null,
105
+ "tags": null,
106
+ "license": null,
107
+ "pretrained_path": "/dev/shm/lerobot_policies/eval1_mix_nooutlier_fp32_noaug_015000/checkpoints/015000/pretrained_model",
108
+ "chunk_size": 50,
109
+ "n_action_steps": 50,
110
+ "normalization_mapping": {
111
+ "VISUAL": "IDENTITY",
112
+ "STATE": "MEAN_STD",
113
+ "ACTION": "MEAN_STD"
114
+ },
115
+ "max_state_dim": 32,
116
+ "max_action_dim": 32,
117
+ "resize_imgs_with_padding": [
118
+ 256,
119
+ 256
120
+ ],
121
+ "empty_cameras": 2,
122
+ "adapt_to_pi_aloha": false,
123
+ "use_delta_joint_actions_aloha": false,
124
+ "tokenizer_max_length": 48,
125
+ "num_steps": 10,
126
+ "use_cache": true,
127
+ "freeze_vision_encoder": true,
128
+ "train_expert_only": true,
129
+ "train_state_proj": true,
130
+ "optimizer_lr": 3e-05,
131
+ "optimizer_betas": [
132
+ 0.9,
133
+ 0.95
134
+ ],
135
+ "optimizer_eps": 1e-08,
136
+ "optimizer_weight_decay": 1e-10,
137
+ "optimizer_grad_clip_norm": 10.0,
138
+ "scheduler_warmup_steps": 500,
139
+ "scheduler_decay_steps": 10000,
140
+ "scheduler_decay_lr": 2.5e-06,
141
+ "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
142
+ "load_vlm_weights": false,
143
+ "add_image_special_tokens": false,
144
+ "attention_mode": "cross_attn",
145
+ "prefix_length": -1,
146
+ "pad_language_to": "longest",
147
+ "num_expert_layers": -1,
148
+ "num_vlm_layers": 16,
149
+ "self_attn_every_n_layers": 2,
150
+ "expert_width_multiplier": 0.75,
151
+ "min_period": 0.004,
152
+ "max_period": 4.0,
153
+ "rtc_config": null,
154
+ "compile_model": false,
155
+ "compile_mode": "max-autotune"
156
+ },
157
+ "reward_model": null,
158
+ "output_dir": "outputs/train/eval1_v10_15k_then_fixed93_lightingonly_20260518_184910",
159
+ "job_name": "eval1_v10_15k_then_fixed93_lightingonly_20260518_184910",
160
+ "resume": false,
161
+ "seed": 1000,
162
+ "cudnn_deterministic": false,
163
+ "num_workers": 8,
164
+ "batch_size": 64,
165
+ "prefetch_factor": 2,
166
+ "persistent_workers": true,
167
+ "steps": 10000,
168
+ "eval_freq": 20000,
169
+ "log_freq": 20,
170
+ "tolerance_s": 0.0001,
171
+ "save_checkpoint": true,
172
+ "save_freq": 1000,
173
+ "use_policy_training_preset": true,
174
+ "optimizer": {
175
+ "type": "adamw",
176
+ "lr": 3e-05,
177
+ "weight_decay": 1e-10,
178
+ "grad_clip_norm": 10.0,
179
+ "betas": [
180
+ 0.9,
181
+ 0.95
182
+ ],
183
+ "eps": 1e-08
184
+ },
185
+ "scheduler": {
186
+ "type": "cosine_decay_with_warmup",
187
+ "num_warmup_steps": 500,
188
+ "num_decay_steps": 10000,
189
+ "peak_lr": 3e-05,
190
+ "decay_lr": 2.5e-06
191
+ },
192
+ "eval": {
193
+ "n_episodes": 50,
194
+ "batch_size": 8,
195
+ "use_async_envs": true
196
+ },
197
+ "wandb": {
198
+ "enable": true,
199
+ "disable_artifact": false,
200
+ "project": "lerobot",
201
+ "entity": null,
202
+ "notes": null,
203
+ "run_id": "kyzq0s2y",
204
+ "mode": null,
205
+ "add_tags": true
206
+ },
207
+ "peft": null,
208
+ "sample_weighting": null,
209
+ "rename_map": {
210
+ "observation.images.front": "observation.images.camera1"
211
+ },
212
+ "checkpoint_path": null
213
+ }